fibromialgia dolor cronico estudio
fibromialgia dolor cronico estudio
fibromialgia dolor cronico estudio
Departamento de Psicología
Universitat de les Illes Balears
Palma, 2008
A todos los que me ayudaron a crecer
Agradecimientos
En cuarto lugar, a mis compañeros, desde las raíces más profundas, Pedro
Ruiz, Mourad Akaârir, Jose Olaya, Marta Salvat, Mónica Söderberg, Jose Periáñez,
Dani Adrover, Érika Nyhus, Nicole Blay, Marta Santandreu, Alfonso ‘Chiqui’
Rodríguez-Cotes, Ruth Hernández, Tina Alorda, Rubén Miranda, César Walteros,
Mirna Frascarelli, Inma Riquelme, Marga Adrover, Cati Ibáñez y otros alumnos
colaboradores (Lola Juste, Paloma Fajardo, Rosa Oliva, Pilar, Jéssica, Salvador
López, Christian Peake, Thomas Kaufmayer, Mathias); hasta los nuevos brotes,
Ignacio Cifre, Miguel Ángel Muñoz, Inma Tur, Mercedes Martínez, Ana Mª González y
Noemí Sánchez. Por otro lado, agradecer también la inestimable aportación, personal
y académica, de los compañeros que he tenido oportunidad de conocer en cursos,
talleres y/o congresos a lo largo de estos años, en especial a Alfonso Barrós-
Loscertales.
Por último, y más importante, a mis familiares y amigos (entre los que se
cuentan muchos de mis compañeros): el humus, la luz y el agua de mi vida. En
especial a mis padres, a mi hermano y a Antonia. A mis abuelos, tíos y primos. A
M.A.R., por lo que fue y ahora es. A Lina Florit, por su buen arte y su calidez. A Kuk,
por compartir su amor por la música, la naturaleza y los animales, por su paciencia,
por acercarme a sus amigos (mis amigos), por cuidarme y hacerme reír… por llenar
mi vida.
Índice
Este trabajo aborda una cuestión con repercusiones sociales y económicas muy
importantes dentro del ámbito de la salud, ya que según datos estimados
recientemente, el dolor crónico causa en Europa una pérdida de 500 millones de días
de trabajo, 34 billones de euros y el sufrimiento de un tercio de los europeos
(Breivik, Collet, Ventafridda, Cohen & Gallacher, 2006). Por otro lado, la fibromialgia
constituye un síndrome doloroso que, por su desconocida etiología, representa un
reto tanto a nivel socio-sanitario como a nivel científico.
1. Introducción
El dolor cumple una función biológica adaptativa, ya que tiene una finalidad
beneficiosa para la integridad del organismo, al funcionar como una señal de alarma
que nos avisa de que se ha producido una herida o de que estamos enfermos.
Además, este tipo de dolor, el dolor agudo, tiene carácter temporal, ya que lo
habitual es que vaya remitiendo con el tratamiento adecuado y desaparezca
finalmente cuando acaba el proceso de curación. Pero, existen múltiples casos en los
que la percepción de dolor se prolonga mucho más allá del momento final del
proceso de enfermedad o la curación de la herida. En otras ocasiones, puede no
haberse producido ninguna herida ni conocerse daño orgánico responsable del dolor.
En estos casos, el dolor no cumple ninguna función útil, muy por el contrario, cuando
C. Sitges 12
específica del dolor central que lo distingue del dolor neuropático, provocado por
lesiones de los nervios periféricos, es el tiempo de inicio de los síntomas dolorosos,
que suele ocurrir semanas o incluso meses después de la lesión neurológica causante
(Cailliet, 1995).
Por otro lado, según la duración del dolor, éste se puede clasificar en: agudo,
si la duración es inferior a 6 meses (APA, 2003). Según la IASP (Vallejo y Comeche,
1994), se refiere a un dolor de reciente comienzo y duración probablemente
limitada, que generalmente tiene una relación temporal y causal con una lesión o
enfermedad. O crónico, si la duración es superior o igual a 6 meses (APA, 2003).
Según la IASP (Vallejo y Comeche, 1994), se refiere a un dolor que persiste a lo largo
de periodos más allá del tiempo de cicatrización de la lesión, frecuentemente sin una
causa claramente identificable.
De todos modos, ya hemos visto que la distinción entre dolor agudo y dolor
crónico no responde sólo a criterios temporales de duración, ya que hoy en día, el
dolor agudo se identifica con el síntoma propiamente dicho, mientras que el dolor
crónico se considera una enfermedad que precisa un abordaje mucho más complejo.
Según Vallejo y Comeche (1999), el dolor agudo tiene una finalidad beneficiosa para
Estudio de la actividad cerebral en pacientes con fibromialgia 15
la integridad del organismo, ya que actúa como una señal de alarma, mientras que el
dolor crónico no. Además, el dolor agudo tiene carácter temporal, ya que lo habitual
es que vaya remitiendo con el tratamiento adecuado y finalmente desaparezca; el
dolor crónico es de carácter persistente. En el caso del dolor agudo se produce un
aumento de la activación autonómica (p.e., taquicardia, sudor) y se asocia con la
ansiedad, mientras que el dolor crónico se relacionaría más con la depresión. Por
otro lado, actualmente, se considera que en el dolor influyen aspectos tanto físicos
como psicológicos y que, por tanto, es necesario adoptar un enfoque multifactorial
para explicar su etiología.
En lo que se refiere al impacto sobre la calidad de vida, más del 40% de los
sujetos con dolor crónico manifiestan que su dolor impacta en sus actividades diarias,
desde hacer ejercicio a dormir. Además, el 79% de los pacientes encuestados
señalaron que su dolor incrementaba durante el día debido a su actividad. Por otro
lado, en cuanto al impacto en el ámbito laboral, uno de cada cinco sujetos con dolor
crónico ha perdido el trabajo a consecuencia de su dolor. Aquellos que están
empleados se ven forzados a tomar una media de más de 15 días libres en el trabajo
cada año a causa del dolor (basado en una estimación de un periodo de 6 meses). Y,
en cuanto al impacto emocional, uno de cada cinco sujetos con dolor crónico ha sido
diagnosticado de depresión como consecuencia de su dolor; teniendo España la tasa
más alta de depresión (29%) a consecuencia del dolor, seguida de Noruega (28%).
Además, más de la mitad de sujetos con dolor crónico informan de sentimientos de
impotencia o de inhabilidad para pensar o funcionar normalmente (Breivik et al.,
2006).
En otro estudio realizado a nivel nacional (Catalá et al., 2002), se apunta que
el dolor, y en particular el dolor crónico, tiene una alta prevalencia en la población
española general y un impacto significativo en las relaciones laborales (provocando
una limitación en la actividad ocupacional del 11%; un 35.8% en población >65 años y
Estudio de la actividad cerebral en pacientes con fibromialgia 17
El 29.6% de la población española entre los 18-95 años informó haber tenido
dolor el día antes (el 37.6% de mujeres y el 20.9% de hombres), y el 43.2% la semana
antes de la entrevista. Los lugares más comunes de dolor eran las extremidades
inferiores (22.7%) y la espalda (niveles cervical y lumbar) (21.5%), seguidas de la
cabeza (20.5%), y las causas más comunes de dolor crónico eran artritis/reumatismo
(18.3%) y migraña/dolor de cabeza (10.5%). La frecuencia de dolor incrementaba con
la edad, alcanzando el 42.6% en población mayor de 65 años. Entre las personas que
se quejaron de dolor durante el último día o la última semana, la duración de los
síntomas era superior a 3 meses (dolor crónico) en el 54% de los casos, representando
el 23.4% de la población española general (Catalá et al., 2002).
Por otro lado, según la Enquesta de Salut de les Illes Balears (Conselleria de
Sanitat i Consum, 2001), en la que se evaluaban también 5 dimensiones de la calidad
de vida relacionadas con la salud, el problema más prevalente es la presencia de
dolor o malestar, que afecta al 31% de la población mayor de 14 años, seguido de la
ansiedad o depresión (19.5%), los problemas de movilidad (15%), las dificultades para
realizar las actividades cotidianas (13%) y un 6% explican problemas de cuidado
personal. En las personas de más de 64 años estos problemas son siempre más
frecuentes en las mujeres, que manifiestan dolor/malestar (63%) y problemas para
caminar (49%), mientras que en los hombres estos porcentajes son del 46% y del 36%,
respectivamente. Por islas, la valoración es peor en Mallorca excepto en los
C. Sitges 18
(a) (b)
Las vías de proyección descendente (ver Gráfica 1b), que forman parte de los
sistemas inhibidores endógenos, implican a las estructuras mediales del tronco
cerebral, desde el diencéfalo medial hasta los núcleos del rafe en el bulbo. Son
particularmente activas las áreas de la sustancia gris periacueductal. Estas
estructuras proyectan, de forma descendente, hacia el núcleo magno del rafe,
situado en el bulbo, y éste finalmente emite sus axones por los cordones laterales
C. Sitges 20
hasta el asta posterior de la médula. Además, existen otros núcleos del tronco que
también participan en este sistema (Aliaga y Santacana, 1994).
© Martin (1999)
En resumen, el dolor casi siempre empieza con un estímulo nocivo que puede
ser mecánico, térmico, eléctrico o químico. Las fibras nerviosas aferentes implicadas
en la transmisión nociceptiva se dividen en los grupos A, B (ambas mielinizadas) y C
(no mielinizadas). El grupo A se clasifica además en fibras alfa, beta, gamma y delta.
El grupo A-beta, A-delta y las fibras C transmiten el tacto y la presión, el dolor agudo
y el dolor crónico, respectivamente (ver Tabla 1). Las fibras nerviosas primarias
aferentes transmiten estas señales de dolor por el asta dorsal de la médula espinal.
Las señales nociceptivas del asta dorsal de la médula espinal se transmiten al tálamo
y al prosencéfalo donde las características emocionales y desagradables del dolor se
C. Sitges 22
activo que filtra, selecciona y modula los inputs. Los factores psicológicos, que
fueron previamente desestimados como ‘reacciones de dolor’ fueron vistos como
parte integrante del procesamiento del dolor y, de este modo, se abrieron nuevas
vías de control del dolor (Melzack, 1999).
Por otro lado, los primeros estudios realizados sobre el fenómeno del miembro
fantasma (Melzack, 1989; Melzack et al., 1997 citado en Melzack, 1999) llevaron a
cuatro conclusiones, que apuntaban a nuevo concepto de sistema nervioso: (1) ya
que el miembro fantasma (u otra parte del cuerpo) se experimenta tan real, es
razonable concluir que el cuerpo que siente está sustentado por los mismos procesos
neuronales en el cerebro, que están normalmente activados y modulados por inputs
corporales, pero que pueden actuar en ausencia de ellos; (2) todas las cualidades que
se experimentan normalmente en el cuerpo, incluyendo el dolor, se pueden sentir
también en ausencia de los inputs corporales, ya que los orígenes de los patrones que
subyacen a las cualidades de la experiencia están en redes neuronales en el cerebro;
(3) el cuerpo es percibido como una unidad y es identificado como el ‘yo’ (‘self’),
distinto de otras personas y del mundo que nos rodea, por lo que esta experiencia de
unidad con diversos sentimientos está producida por procesos neuronales centrales y
no puede estar originada por el sistema nervioso periférico o la espina dorsal; (4) los
procesos cerebrales que subyacen al body-self están construidos por especificación
genética, aunque pueden ser modificados por la experiencia.
Por último, algunos estudios recientes, han mostrado que en el cerebro de los
pacientes con dolor crónico existen cambios estructurales, más concretamente,
atrofias (Apkarian et al., 2004) y neuroquímicos, más concretamente, reducciones de
N-acetil aspartato y glucosa (Grachev, Fredrickson & Apkarian, 2000) en la corteza
prefrontal dorsolateral. Parece ser que es el lugar de mayor neurodegeneración y de
potencial muerte celular en los pacientes con dolor crónico (Apkarian et al., 2004),
lo que podría tener efectos negativos en el sistema descendente inhibitorio y
contribuir a su estado crónico de dolor. Por otro lado, Schmidt-Wilcke et al. (2006)
además de mostrar que los pacientes con dolor crónico de espalda presentan un
decremento de la materia gris en la corteza prefrontal dorsolateral derecha, éstos
también muestran una reducción en el tronco cerebral y en la corteza somato-
sensorial derecha, así como un incremento en la materia gris en el putamen y en el
tálamo posterior izquierdo.
© Kiriazis (2004)
las astas posteriores. Este fenómeno lleva, a su vez, al incremento del dolor
(hiperalgesia), al incremento de la respuesta de dolor ante estímulos no dolorosos
(alodinia) y al dolor espontáneo. Con frecuencia, los pacientes con FM presentan
estos fenómenos clínicos de la sensibilización central, que indican un incremento de
la percepción de dolor (Staud & Domingo, 2001). Varios estudios muestran que el
input hacia las vías nociceptivas centrales es anormal en los pacientes con FM (Staud
& Domingo, 2001; Staud, 2002; Staud et al., 2003).
Según Staud & Domingo (2001), el daño tisular no sólo causará activación a
corto plazo de las neuronas de la médula espinal, sino que frecuentemente también
tendría lugar una sensibilización central a largo plazo por vía excitatoria de
aminoácidos y neuropéptidos como el glutamato, el aspartato y la sustancia P.
Aunque muchos pacientes con FM no muestran evidencia de daño tisular, los niveles
de sustancia P medidos en su fluido cerebroespinal se han encontrado de dos a tres
veces más elevados que en sujetos controles normales (Russell et al., 1994 citado en
Staud & Domingo, 2001). Otro de los mecanismos del dolor central depende de los
receptores del N-metil-D-aspartato (NMDA) en el asta dorsal de la médula espinal,
que pueden desencadenar la liberación de óxido nítrico, ciclooxigenasa y quinasas.
Los receptores de NMDA, que están presentes en el SNC, también se pueden
encontrar en la médula rostral ventromedial. La médula rostral ventromedial
C. Sitges 32
presentes en muchos de los pacientes con FM, cuando recibían estímulos dolorosos
generados por láser tanto en los puntos sensibles (tender points) como en puntos
control, que podría deberse a la presencia de una sensibilización periférica de las
fibras C, junto con una sensibilización central generalizada en estos pacientes
(Granot et al., 2001). Otro estudio reciente, realizado con estimulación eléctrica
intramuscular e intracutánea mostró, de nuevo, umbrales de dolor más bajos y
mayores amplitudes en el componente N80, indicativas de un procesamiento
sensorial aumentado en los pacientes con FM (Diers et al., 2008). Otros estudios
llevados a cabo con estímulos auditivos han mostrado que los pacientes con FM
presentaban menores amplitudes en los potenciales evocados auditivos (PEAs) N1, P2
(Carrillo de la Peña, Vellet, Pérez & Gómez-Perretta, 2006) y P3 (Ozgocmen et al.,
2003), que podrían relacionarse con un déficit serotoninérgico presente en estos
pacientes. Por último, Stevens, Batra, Köter, Bartels & Schwarz (2000), mostraron
que los pacientes con FM presentaban una potencia relativa (μV2) baja en la
frecuencia alfa rápida (9.5-12.5 Hz) y beta lenta (12.5-19.5 Hz) en condiciones de
reposo (con ojos abiertos), aunque la potencia de alfa rápida se incrementaba
ligeramente durante y después de la inmersión del brazo derecho (dominante) en una
cubeta de agua helada (cold pressor test). Como se muestra en los anteriores
estudios, el umbral sensorial y de dolor es más bajo en los pacientes con FM,
sintiéndose el agua fría mucho más dolorosa e intolerable antes que el grupo control.
Para Hart, Martelli & Zasler (2000), el proceso de dolor significativo o más
severo requiere control atencional central consciente, por tanto, los sujetos con un
nivel de dolor bajo podrían distraer la atención lejos del dolor y dirigirla hacia una
tarea, alcanzando un determinado grado de analgesia. Eccleston (1994) y Grigsby,
Rosenberg & Busenbark (1995) conceptualizan el dolor como un estímulo perceptivo
que demanda atención, y la atención como un recurso finito y unitario. El dolor
compite por los recursos limitados de la atención, de tal modo que afecta al
funcionamiento de las tareas que implican el procesamiento y la integración de otra
información. De esta forma, es probable que el dolor interrumpa el funcionamiento
de una tarea exigente debido a un mayor desgaste de los recursos atencionales.
Asimismo, Eccleston & Crombez (1999) teorizaron que la interrupción de la atención
causada por el dolor está mediada por las características relacionadas con el dolor
(p. e., el valor de la amenaza del dolor) y por las demandas ambientales (p. e.,
despertar emocional). El dolor, por lo tanto, se selecciona para la acción dentro de
ambientes afectivos y motivacionales complejos para impulsar el escape. Otros
investigadores han postulado que los sesgos atencionales en estados crónicos de dolor
C. Sitges 36
Asimismo, Tracey & Mantyh (2007) señalan que para las personas que sufren
tanto dolor agudo como crónico, el estado emocional tiene un gran impacto en la
percepción de dolor resultante y en la habilidad para afrontarlo. Muchos estudios han
intentado entender cómo la anticipación y la ansiedad causan un aumento en la
experiencia de dolor (Hsieh, Stone-Elander & Ingvar, 1999; Ploghaus et al., 1999,
2000, 2001; Porro et al., 2002, 2003; Song et al., 2006, citados en Tracey & Mantyh,
2007). Algunas de las regiones críticas implicadas en amplificar o exacerbar la
experiencia de dolor incluyen el complejo entorrinal, la amígdala, la ínsula anterior y
las cortezas prefrontales (Tracey & Mantyh, 2007). Por otro lado, según Tracey &
Mantyh (2007), los trastornos depresivos a menudo acompañan al dolor persistente.
Aunque la relación exacta entre depresión y dolor es desconocida, con un debate
sobre si una condición lleva a la otra o si existe una diátesis subyacente, algunos
estudios, como el de Neugebauer, Li, Bird & Han (2004), han intentado aislar
regiones cerebrales, como la amígdala, que pueden estar mediando en su
interacción.
Según Williams & Gracely (2007), otro factor cognitivo común que se conoce
que modula las quejas de dolor es el catastrofismo, un estilo atribucional o un
comportamiento en el que el dolor es caracterizado como tremendo, horrible e
insoportable. Este constructo incorpora magnificación de los síntomas relacionados
con el dolor, rumiación sobre el dolor, sentimientos de impotencia y pesimismo sobre
las consecuencias relacionadas con el dolor (Edwards, Bingham, Bathon &
Haythornthwaite, 2006), y se define como un conjunto de procesos negativos
emocionales y cognitivos (Sullivan et al., 2001). El catastrofismo parece que juega un
papel importante en el desarrollo de la cronicidad del dolor. Se pensaba que era un
síntoma de depresión pero actualmente es reconocido como un factor independiente
que está sólo parcialmente asociado con la depresión. Se ha sugerido que aumenta la
percepción de dolor, incrementando la atención hacia los estímulos dolorosos y
aumentando las respuestas emocionales de dolor. Por otro lado, el catastrofismo
modula la actividad evocada por el dolor en numerosas estructuras relacionadas con
la anticipación del dolor (corteza fronto-medial contralateral, cerebelo ipsilateral),
la atención hacia el dolor (giro cingulado anterior contralateral, corteza prefrontal
dorsolateral bilateral), las respuestas emocionales (claustro ipsilateral,
Estudio de la actividad cerebral en pacientes con fibromialgia 37
Aunque, por otro lado, según Montoya & Larbig (2006), también hay datos
empíricos del apoyo social que sugieren efectos opuestos perjudiciales en los
pacientes con dolor crónico. El comportamiento solícito se define como una conducta
que indica la preocupación por la condición física y psicológica del paciente, el
fomento del descano y hacerse cargo de las tareas o desalentar otras formas de
actividad en situaciones estresantes de trabajo o tareas diarias de casa. Muchos
estudios han encontrado que elevadas respuestas solícitas de personas significativas
para el paciente, así como la satisfacción con las relaciones sociales, están asociadas
positivamente a grandes niveles de dolor y discapacidad entre los pacientes con dolor
crónico, ya que la atención de la pareja o el refuerzo es contingente a las quejas de
dolor (Block, Kremer & Gaylor, 1980; Fillingim, Doleys, Edwards & Lowery, 2003;
Flor, Breitenstein, Birbaumer & Fürst, 1995; Giardino, Jensen, Turner, Ehde &
Cardenas, 2003; Gil, Keefe, Crisson & van Dalfson, 1987; Keefe et al., 1999; Papas,
Estudio de la actividad cerebral en pacientes con fibromialgia 39
Robinson & Riley, 2001; Paulsen & Altmaier, 1995; Romano et al., 1995; Schmaling,
Smith & Buchwald, 2000; Thieme, Gromnica-Ihle & Flor, 2003; Turk, Flor & Rudy,
1987, citados en Montoya & Larbig, 2006).
Según Williams & Gracely (2007), aunque la cognición parece modular la experiencia
de dolor, también es probable que el dolor interfiera con la capacidad de pensar y
procesar la información. Una conocida queja de los pacientes con FM es la de un
estado de alteración cognitiva que ha sido denominada como ‘fibro niebla’ (‘fibro
fog’). Los déficits cognitivos observados en la FM se parecen a los encontrados en el
envejecimiento. Por ejemplo, los pacientes con FM tienden a completar las medidas
de memoria de trabajo con una competencia que es similar a la de los controles
sanos que son 20 años más mayores (Park, Glass, Minear & Crofford, 2001; Bangert et
al., 2003). Bangert et al. (2003) muestran que para alcanzar un nivel de ejecución
similar al de los sujetos sanos, los pacientes con FM necesitan emplear muchos más
recursos cerebrales. En este sentido, los pacientes con FM mostraban una activación
neuronal más extendida en regiones frontales y parietales, incluyendo activación
bilateral en el giro frontal medio y una activación del lado derecho en el giro fronto-
medial, lóbulo parietal superior y giro precentral. En una revisión de los estudios de
neuroimagen realizados sobre el dolor agudo, llevada a cabo por Apkarian, Bushnell,
Treede & Zubieta (2005), se muestra que las áreas cerebrales más comúnmente
activadas para el dolor en los sujetos sanos son la corteza somatosensorial primaria
(SI), la corteza somatosensorial secundaria (SII), la corteza insulada (IC), la corteza
cingulada anterior (ACC), la corteza prefrontal (PFC) y el tálamo. Lo más interesante
es que la simple anticipación del dolor activa regiones similares (PFC, ínsula anterior,
ACC). Estas regiones están implicadas en la formación de las representaciones
cognitivas y afectivas del dolor, involucrando memorias de acontecimientos pasados,
comprensión del presente y de las futuras implicaciones de los acontecimientos
señalados por el dolor (Koyama, McHaffie, Laurient & Coghill, 2005).
fluencia verbal (Park et al., 2001). Aunque, a pesar de que los pacientes con FM
muestran niveles más elevados de ansiedad, no parecen existir diferencias en su
rendimiento cognitivo, en comparación con otros grupos de pacientes con dolor
crónico, de origen musculoesquelético (Dick, Eccleston & Crombez, 2002). Parece ser
que estos déficits cognitivos presentes en diversas patologías que implican dolor
crónico, están relacionados con procesos tales como la intensidad del dolor percibido
o la ansiedad-rasgo (Grace et al., 1999), aunque no existe unanimidad sobre la
influencia de esta última como factor modulador sobre el rendimiento cognitivo (Park
et al., 2001). Asimismo, Sletvold, Stiles & Landro (1995) encontraron déficits en
pruebas que requerían atención, procesamiento rápido de información y velocidad
psicomotora en pacientes con FM. Aunque, posteriormente, utilizando la misma
muestra de sujetos que en el estudio anterior, Landro, Stiles & Sletvold (1997)
encontraron que sólo aquellos pacientes con FM que tenían una historia de vida de
depresión mayor mostraban deterioro en memoria en relación con los controles
normales. El grupo total de pacientes con dolor rindieron de forma similar que los
pacientes con depresión mayor, y mostraron deterioro en pruebas de recuerdo y
fluencia verbal. Esos resultados se podrían explicar por el hecho de que los pacientes
con dolor tienden a sobreestimar sus problemas cognitivos en una escala de auto-
valoración (Grace et al., 1999), y por la importancia de la influencia de variables
emocionales (ansiedad y depresión) en el rendimiento cognitivo.
Todas las técnicas utilizadas en los estudios se han aplicado en datos de medidas
centrales, y están basadas en el análisis de los potenciales evocados (PEs)1, el
análisis (lineal) espectral y el análisis no lineal (dimensión fractal y entropía), que
pasarán a ser definidas a continuación.
1
Potenciales evocados, del inglés event-related potentials.
Estudio de la actividad cerebral en pacientes con fibromialgia 41
Los PEs resultantes son relativamente complejos ya que cuentan con diversos
componentes de diferente amplitud, polaridad (positiva (P) o negativa (N)) y latencia
(tiempo que transcurre entre la aplicación del estimulo y la obtención de algún tipo
de respuesta) (Nicolau et al., 1995). Los estímulos sensoriales en todas las
modalidades están asociados con una serie de desviaciones en los potenciales
evocados que están relacionadas con la transmisión de la información desde el
sistema sensorial periférico a la corteza y/o con la llegada de la información a la
corteza. Para algunas modalidades, las latencias más tempranas de las desviaciones
son extremadamente cortas (unos pocos milisegundos) y, en este caso, estas
desviaciones reflejan la transmisión de información sensorial en las vías sensoriales.
Para todas las modalidades, muchos de estos componentes sensoriales son
modificables, por ejemplo, por manipulaciones atencionales (Rugg & Coles, 1996).
2001), postulan que la información aferente activaría el hipocampo; CA3, las células
piramidales y las interneuronas se activarían con el primer tono, pero durante el
intervalo entre estímulos (inter-stimulus interval, ISI) habría una inversa relación
hiperbólica entre las células piramidales y la actividad de las interneuronas, en
consonancia con la inhibición de las células piramidales por las interneuronas.
Aunque parece que el filtro sensorial no se limita a una sola área cerebral o a un solo
componente de latencia media, y que la estimulación repetitiva y el ISI afectan a los
componentes cerebrales. Otras estructuras cerebrales que podrían estar implicadas
son la región temporo-parietal (áreas de Brodmann 22 y 2) y la corteza prefrontal
(áreas de Brodmann 6 y 24) (Grunwald et al., 2003; Knight, Staines, Swick & Chao,
1999).
2
Valores normativos para mujeres de población española.
Estudio de la actividad cerebral en pacientes con fibromialgia 45
los promedios fueron filtrados digitalmente (filtro de paso bajo, 30 Hz) y se les
corrigió la línea base antes de realizar las medidas estadísticas sobre la amplitud de
los componentes: P50 (20-80 ms después de la aparición de los estímulos en los
electrodos C3/C4), N80 (60-110 ms en C3/C4) y P200 (135-260 ms en Cz), para los
PESs, y P200 (120-250 ms en Pz), en el caso de los PEVs.
resultados obtenidos en otros estudios recientes, llevados a cabo por Babiloni et al.
(2006, 2008) con estimulación dolorosa eléctrica o mediante láser en sujetos sanos,
mostraron importantes implicaciones en el estudio del dolor crónico, ya que la
desincronización del ritmo alfa (8-12 Hz) o lo que es lo mismo, la reducción de su
potencia, supondría un mecanismo subyacente de los procesos anticipatorios que
preceden a la integración de la información dolorosa y motora a nivel cortical, que
en el caso de los pacientes con dolor crónico podría estar sobreactivado, exagerando,
de este modo, la activación anticipatoria de la corteza sensorio-motora ante
cualquier e insignificante estímulo doloroso.
vez a los procesos internos del sistema (Francis, 1995). Las representaciones
neuronales o los mapas corticales son volubles, alterables por las experiencias de la
vida. Son dinámicos y aparentemente auto-organizados en múltiples escalas de
tiempo y espacio. Los procesos competitivos y cooperativos entre la conectividad
intrínseca y las señales de entrada (inputs) externas, junto con las fluctuaciones
locales en las interacciones celulares, esculpen conjuntamente estos mapas a lo largo
del tiempo. Incluso el cerebro adulto tiene la capacidad de una rápida reorganización
funcional (Kelso, 1999).
ln (N-1)
FD= (1)
ln (N-1) + ln (d/L)
En el método de Sevcik (2, 3), primero los datos se normalizan para estar
dentro de una unidad cuadrada por el reescalamiento de la abscisa (eje de tiempo) y
de la ordenada (señal de EEG) del espacio de datos:
s(i)-smin
i’=i/N, s’(i’)= (2)
smax-smin
donde s(i) y s’(i’) son las señales de EEG originales y normalizadas, respectivamente,
en el punto de datos i; smax y smin son los valores máximo y mínimo de la señal; i=1, 2,
3, ... , N es el número serial de los puntos de datos e i’ es el normalizado. La FD se
calcula como:
ln (L)
FD=1+ (3)
ln (2 (N-1))
Por otro lado, Heath (2000) también apunta que la D2 necesita ser
complementada por otros índices cuantitativos que indiquen cómo los puntos del
atractor divergen y se contraen en el tiempo. Como es el caso del máximo exponente
de Lyapunov, que indica cuán rápido arbitrariamente puntos cercanos en la
representación del espacio de fase de un proceso dinámico divergen en el tiempo.
Como resultado, el exponente máximo de Lyapunov, junto con el espectro completo
de los exponentes de Lyapunov, uno por cada dirección ortogonal en el espacio de
fase, ayudan a clasificar un proceso dinámico como un punto fijo, cíclico o caótico.
Según Stam (2005), las dimensiones son medidas estáticas de los atractores que no
ofrecen información de la evolución de las trayectorias a lo largo del tiempo. Los
exponentes de Lyapunov, que indican la divergencia (exponentes positivos) o
convergencia (exponentes negativos) exponencial de las trayectorias cercanas al
atractor, y las medidas de entropía, que dan información de la tasa de pérdida de
información de la dinámica del atractor, por otro lado, pueden ser consideradas
medidas ‘dinámicas’ de la complejidad de éste. La entropía es igual a la suma de
todos los exponentes positivos de Lyapunov, y una entropía positiva indica un sistema
caótico dinámico. Fue introducida una amplia variedad de algoritmos para el cálculo
de las medidas de entropía, como la propuesta por Grassberger & Procaccia (1983),
que mostraron que ésta podía ser determinada mediante la integral de correlación, o
el pronóstico no lineal (Pezard et al., 1997; Wales, 1991), la entropía aproximada
Estudio de la actividad cerebral en pacientes con fibromialgia 53
(Pincus, 1991), la entropía de la máxima probabilidad (Schouten, Takens & Van den
Bleek, 1994), las tasas de entropía de grano grueso (Palus, 1996) y la entropía de
multiresolución (Torres, Anino, Gamero & Gemignani, 2001). En sistemas de gran
extensión es también posible determinar la entropía espacial, definida como la
pérdida de información por unidad de longitud (Van der Stappen, Schouten & Van den
Bleek, 1994).
jτ
yj(τ)=1/τ Σ xi (1)
i=(j-1)τ+1
Figura 2. Serie temporal simulada u[1], ..., u[n] que ilustra el procedimiento para
calcular la entropía muestral (SampEn) para el caso en el que el patrón de longitud,
m, es 2, el criterio de similitud, r, es 20 (r es un valor real positivo que normalmente
se encuentra entre el 10% y el 20% de la desviación muestral de las series
temporales). Las líneas punteadas horizontales alrededor de los puntos de datos u[1],
u[2] y u[3] representan u[1] ± r, u[2] ± r y u[3] ± r, respectivamente. Dos valores de
datos están emparejados el uno con el otro, es decir, son indistinguibles, si la
diferencia absoluta entre ellos es ≤r. Todos los puntos verdes representan los puntos
de los datos que están emparejados con el punto del dato u[1]. Igualmente, todos los
puntos rojos y azules están emparejados con los puntos de los datos u[2] y u[3],
respectivamente. Si consideramos la plantilla de secuencia del componente-2 verde-
rojo (u[1], u[2]) y del componente-3 verde-rojo-azul (u[1], u[2], u[3]), para el
Estudio de la actividad cerebral en pacientes con fibromialgia 55
segmento mostrado hay dos secuencias verde-rojo, (u[13], u[14]) y (u[43], u[44]),
que se emparejen con la plantilla de secuencia (u[1], u[2]), pero sólo una secuencia
verde-rojo-azul que se empareje con la plantilla (u[1], u[2], u[3]). Estos cálculos se
repetirían para la siguiente plantilla de secuencia del componente-2 y -3, que serían,
(u[2], u[3]) y (u[2], u[3], u[4]), respectivamente. El número de secuencias que se
emparejarían con las plantillas de los componentes -2 y -3 serían contadas y añadidas
de nuevo a los valores previos. Este procedimiento se repetiría para todas las otras
posibles plantillas de secuencias, (u[3], u[4], u[5]), ..., (u[N-2], u[N-1], u[N]), para
determinar el ratio entre el número total de plantilla de emparejamientos del
componente-2 y el número total del -3. La SampEn es el algoritmo natural de este
ratio y refleja la probabilidad de que las secuencias que están emparejadas la una
con la otra para los dos primeros puntos de datos también coincidan con los del
siguiente punto (Costa et al., 2005).
Parece ser que el EEG normal refleja una débil, pero significativa, estructura
no lineal. También se ha sugerido que el EEG normal podría reflejar una dinámica
crítica, cercana a la bifurcación entre dos tipos de atractores: el atractor punto y el
atractor cíclico. Probablemente, las épocas lineales de tipo alfa I podían ser
explicadas por un atractor punto en el modelo y las épocas no lineales de tipo alfa II
podían serlo por un ruidoso atractor cíclico (Stam, 2005). En la depresión mayor, se
han descrito anormalidades en el EEG durante el sueño y un incremento de la
predictibilidad en el EEG durante el día, aunque el número de estudios es aún un
poco limitado (Nandrino et al., 1994; Pezard et al., 1996; Röschke, Mann & Fell,
1994). Por otro lado, en numerosos estudios realizados por Aftanas y colaboradores
se explora casi todo el espectro de medidas no lineales de EEG con el fin de
caracterizar cambios en la función cerebral relacionados con la emoción y el estado
C. Sitges 56
de ánimo (Aftanas et al., 1994, 1997a,b, 1998, 2002). En primer lugar, estos autores
muestran un incremento en la dimensión durante una tarea de imaginación
comparada con otra perceptiva; la condición emocional iba asociada con un
incremento más posterior en la dimensión (Aftanas, Koshkarov, Mordvintsev &
Pokrovskaja, 1994). Usando la previsión no lineal, mostraron que las emociones
negativas estaban asociadas con una alta predictibilidad del EEG especialmente en
regiones posteriores (Aftanas et al., 1997a). También mostraron que el índice de
entropía de Kolmogorov y el exponente más grande de Lyapunov estaban
incrementados después de ver películas positivas o negativas, en comparación con
ver películas neutras (Aftanas, Lotova, Koshkarov, Popov & Makhnev, 1997b). Usando
la dimensión mutua (Dm), como una medida de acoplamiento no lineal, mostraron
que las emociones negativas estaban asociadas con un decremento del acoplamiento
de las áreas frontales izquierdas, mientras que las emociones positivas estaban
asociadas con un incremento del acoplamiento en áreas más posteriores centrales
(Aftanas et al., 1998). También mostraron que un estado de meditación parece estar
asociado con un decremento en la dimensión de complejidad (Aftanas &
Golocheikine, 2002). Algunos estudios de otros autores usaban varios tipos de
estimulación para investigar cambios en la complejidad cerebral. Como Kondakor et
al. (1997), que mostró que un simple procesamiento visual (ojos cerrados
comparados con ojos abiertos) estaba asociado con un incremento en la dimensión de
complejidad global. Por otro lado, Lutzenberger, Flor & Birbaumer (1997), mostraron
como la memoria para el dolor personal estaba caracterizada por un incremento en
la dimensión de complejidad.
Objetivo 4. Estudiar la dinámica cerebral de los pacientes con dolor crónico desde la
perspectiva del análisis lineal, mediante el cálculo del espectro de frecuencias, y no
lineal, mediante el cálculo del índice de entropía multiescalar y de la dimensión
fractal.
3. Publicaciones
reflejado en los componentes más tardíos (N80, entre los 60-110 ms; P200, entre los
135-260 ms).
la información no-dolorosa relacionada con el dolor (Flor et al., 1992, 1997; Flor &
Birbaumer, 1994; Lutzenberger et al., 1997). Numerosas investigaciones (Pearce et
al., 1990; Pincus, Pearce, McClelland, Farley & Vogel, 1994; Pincus & Newman, 2001;
Roelofs, Peters, Zeegers & Vlaeyen, 2002), han mostrado, más concretamente, que
los pacientes con dolor crónico atienden selectivamente a las palabras relacionadas
con el dolor, hacen más asociaciones relacionadas con el dolor ante palabras
ambiguas y tienen más probabilidades de recuperar palabras relacionadas con el
dolor de su memoria. Partiendo de la hipótesis del condicionamiento clásico de las
respuestas de dolor, que propone que éstas podrían dar lugar a la formación de
memorias relacionadas con el dolor amplias y fácilmente excitables (Linton, Melin &
Götestam, 1985 citado en Flor, Knost & Birbaumer, 1997) y en el modelo hebbiano de
ensamblaje celular (Hebbian cell assembly model) (Hebb, 1949), se podría asumir
que un estímulo que está asociado con una experiencia de dolor podría excitar los
ensamblajes celulares relacionados con el dolor y crear una experiencia de dolor
incluso en la ausencia de estimulación dolorosa externa (Flor et al., 1997).
ACR ⫽ American College of Rheumatology; ANOVA ⫽ analysis of iological mechanisms in chronic pain, but also highlight their
variance; BDI ⫽ Beck Depression Inventory; EEG ⫽ electroenceph- importance for FM pain.
alogram; FM ⫽ fibromyalgia; fMRI⫽functional magnetic resonance Previous studies have also examined the influence of ex-
imaging; IAPS ⫽ International Affective Picture System; MPI ⫽ perimental mood induction on pain perception (10 –12) and
West Haven–Yale Multidimensional Pain Inventory; MPQ ⫽
brain functioning (13) in healthy subjects. Basically, it has
McGill Pain Questionnaire; MSK ⫽ musculoskeletal; SE ⫽ standard
error of the mean; SEP ⫽ somatosensory-evoked potentials; STAI ⫽
been found that subjective pain ratings can be enhanced by
State-Trait Anxiety Inventory. viewing unpleasant slides (10), negative emotional films (11),
or reading negative emotional statements (12). Furthermore, a
INTRODUCTION recent study using functional magnetic resonance imaging
(fMRI) demonstrated that the induction of a negative mood
T he diagnosis of fibromyalgia (FM) as defined by the
American College of Rheumatology (ACR) (1) includes
criteria such as chronic widespread musculoskeletal pain, hy-
state by presentation of fearful faces can also alter the brain
processing of nonpainful visceral information within the an-
peralgesia and/or allodynia, morning stiffness, insomnia, fa- terior cingulate gyri and the insula (13). Thus, it seems that the
tigue, cognitive problems, and often depression and headache. emotional context in which the stimulation occurs could mod-
Although no pathophysiological process has yet been found to ulate the central processing of sensory information. However,
account for these symptoms, recent studies have demonstrated little is known about the extent to which a negative emotional
altered brain processing of nociceptive information in patients context can influence the brain processing of somatic infor-
with FM (2–5), which might result from hyperexcitability of mation in chronic pain and, particularly, in FM.
the central nervous system (6). In addition, previous studies In the present study, we investigated whether sensory and
have emphasized the role of emotions and psychological stres- cognitive processing of tactile stimuli could differ as a func-
sors in the origin and maintenance of FM, and suggested that tion of the context induced by external emotional stimulation
these patients might be particularly vulnerable to the effects of in FM and patients with musculoskeletal pain as a result of
negative mood (7–9). In this sense, there has been some identifiable somatic lesions (MSK), which served as an emo-
evidence showing that negative affect is directly associated tionally less disturbed comparison group. For this purpose,
with increased pain perception in patients with FM (8,9). emotional context was manipulated using affective slides and
Thus, these studies do not only suggest that enhanced pain asking patients to empathize with the mood state elicited by
sensitivity and negative affect constitute relevant psychophys- the pictures when they were simultaneously performing a
somatosensory detection task at a nonpainful body location.
By using such an experimental task, it was also possible to
From the Department of Psychology and Research Institute of Health examine the influence of negative affect on cognitive process-
Sciences (IUNICS), University of the Balearic Islands, Palma, Spain (P.M., ing in patients with chronic pain. In addition, brain activity
C.S., N.B.); the Medical Unit for Disability Assessment, Social Security
Agency, Palma, Spain (M.G.-H., R.I., D.C.); and the Pain Clinic, General was recorded from the electroencephalogram (EEG) using
Hospital, Palma, Spain (M.T.). somatosensory-evoked potentials (SEPs). In accord with pre-
Address correspondence and reprint requests to Pedro Montoya, PhD, vious studies, SEP recordings have provided an easy, reliable,
Edificio Beatriu de Pinós, Department of Psychology, Cra. de Valldemossa
km 7.5, 07122 Palma, Spain. E-mail: [email protected] and robust tool for examining the central nervous hyperexcit-
Received for publication December 17, 2004; revision received June 11, ability associated with chronic pain (3–5), as well as some
2005. cognitive variables involved in somatosensory information
Research was supported by the Spanish Ministerio de Ciencia y Tecnologı́a,
and European Funds (Fondos FEDER) (grant BSO2001-0693). processing such as attention (14,15). In this sense, it has been
DOI: 10.1097/01.psy.0000188401.55394.18 suggested that early SEP components within the first 100 ms
after stimulus onset (i.e., P50, N80) originate in primary Patients received two stimulation blocks; each consisted of 560 stimuli (480
sensory cortex and are related to sensory and attentional frequent and 80 deviant stimuli) of 100-ms duration with a constant pressure
of two bars and a variable interstimuli interval of 550 ms (⫾50 ms). None of
processing of somatosensory information, whereas SEP com-
the patients reported discomfort as a result of the tactile stimulation.
ponents at later latencies are linked to complex cognitive During each stimulation block (duration 8 minutes), patients were viewing
processes such as, for instance, shifting attention to deviant a sequence of pictures selected from the International Affective Picture
stimuli in an oddball task (i.e., P200). Based on previous System (IAPS) (20). The IAPS constitutes a standardized and exhaustively
evidence about central hyperexcitability and negative affect in investigated set of affective pictures containing more than 600 items. The
content of the pictures ranges from explicit sexual material, to human injury
FM, it was expected that the experimental induction of a
and surgical slides, to pleasant pictures of children and wildlife. In the present
negative mood state during the processing of nonpainful in- experiment, patients were viewing 40 pictures with negative valence (unpleas-
formation would result in more increased pain perception and ant content) and 40 pictures with positive valence (pleasant content). The
significant changes of brain activity within FM than patients order of the two stimulation blocks was counterbalanced between the subjects
with MSK pain. Particularly, we were interested in examining within each group; for half of the participants (12 patients with FM, 8 patients
with MSK), the unpleasant picture sequence was presented followed by the
early and late SEP components to find out whether there are
pleasant picture sequence, whereas the other half (15 patients with FM, 8
any group differences in the effects of mood induction on the patients with MSK) viewed the affective pictures in the reverse order. Each
early attentional and late cognitive processing of tactile infor- picture was presented for 6 seconds and was followed by a 6-second blank
mation. screen. One deviant and six frequent tactile stimuli were delivered during each
slide presentation, as well as during the blank screen. Patients were instructed
PATIENTS AND METHODS to ignore tactile stimulation and to pay attention to the slides trying to imagine
Participants experiencing themselves in the situations described by the pictures. Subjects
were seated in front of a computer screen in a sound-attenuated room, and
Patients with a diagnosis of chronic musculoskeletal pain for at least 6
instructed to keep eye movements and blinks to a minimum during the
months who had normal hearing, corrected visual acuity, and no history of
experiment.
head trauma or drug abuse were recruited from the pain management unit of
SEPs were recorded from 32 electrodes placed in accordance with the
a tertiary care hospital and the Medical Unit for Disability Assessment of the
international 10 to 20 system and with reference electrodes at ear mastoids.
Spanish Social Security Agency in Palma (Spain). Forty-three righthanded
Nevertheless, for the purposes of the present experiment, only nine electrodes
female patients aged between 40 and 65 years, from an initial sample of 84,
located over the midline (Fz, Cz, Pz), the left (F3, C3, P3), and the right
agreed to participate in this study. At the time of the examination, 24 patients
hemisphere (F4, C4, P4) were analyzed. An electrooculogram channel was
(19 patients with FM and 5 patients with MSK) were involved in a litigation
obtained by placing one electrode above and one below the left eye. Ground
process for financial compensation resulting from disability. Patients were
was placed anteriorly to the location of the Fpz electrode. Electrode imped-
classified according to the primary diagnosis recorded in their medical chart
ance was measured to be less than 10 k⍀. The signals were amplified by a
in two groups: FM (n ⫽ 27) and MSK pain disorders resulting from identi-
BrainAmp amplifier at a sampling rate of 1000 Hz with high and low pass
fiable somatic lesions (n ⫽ 16). The diagnosis of fibromyalgia was further
filter setting at 0.10 Hz and 70 Hz, respectively. A 50-Hz notch filter was also
confirmed by an experienced and external rheumatologist. The MSK group
applied. EEG was segmented in epochs of 700-ms duration (⫺100 ms to 600
was composed by patients with rheumatoid arthritis (n ⫽ 6) and patients
ms relative to stimulus onset). Averaging was performed for frequent and
diagnosed with radiculopathy with a herniated disk (n ⫽ 10). All patients,
deviant stimuli separately, and only SEPs elicited by tactile stimulation during
except in two cases, were taking regular pain medication, including antide-
the presentation of slides were used. All average waves were digitally filtered
pressants and myorelaxants; 48.8% of the patients were taking anxiolytic
(30 Hz low pass) and baseline-corrected before statistical measures of com-
drugs.
ponent amplitude were computed.
At the time of recruitment, patients were verbally informed about the
Previous work has shown that nonpainful tactile stimuli evoke a typical
details of the study and noted that the participation in the study was not linked
SEP with readily identifiable components at several latencies between 20 and
to their possible litigation process. A specifically designed patient information
400 ms (14,15,21). According with these studies and based on visual inspec-
leaflet was also given, and after agreeing to participate, each patient provided
tion of the mean waveforms, we calculated the maximal amplitudes (baseline-
written consent. The study was in accordance with the Declaration of Helsinki
to-peak) of the following peaks from individual waveforms and for each
(1991) and was approved by the Ethics Committee of the University of
condition: P50 (defined as the maximum positive amplitude in the time
Balearic Islands (Spain).
window 20 – 80 ms after stimulus onset at electrodes C3/C4), N80 (maximum
negative amplitude in the time window 60 –110 ms after stimulus onset at
Clinical Pain Assessment
C3/C4), and P200 (maximum positive amplitude in the time window 135–260
All patients underwent an extensive medical and psychological assess- ms after stimulus onset at Cz). Additionally, patients were asked to rate their
ment, including examination of tender points and assessment of clinical pain current clinical pain using a 10-cm visual analog scale at three points during
characteristics through self-report questionnaires. The Spanish versions of the the experiment; at the beginning (after stimulation device was tested), after
Beck Depression Inventory (BDI) (16), the State-Trait Anxiety Inventory viewing the unpleasant and after viewing the pleasant slides.
(STAI) (17), the West Haven–Yale Multidimensional Pain Inventory (MPI)
(18), and the McGill Pain Questionnaire (MPQ) (19) were completed. Data Analysis
The study followed a randomized factorial mixed design with the within-
Somatosensory Stimulation Task and Recording subjects factors “type of stimulus” (frequent versus deviant), “emotional
of Brain Activity context” (viewing unpleasant versus pleasant slides), and “electrode loca-
Somatosensory-evoked potentials elicited on tactile stimulation were re- tions” (9 electrodes), and the between-subjects factor “group” (patients with
corded following a so-called oddball paradigm. During this experimental task, FM versus patients with MSK pain). The effects of these factors on SEP
two types of stimuli are presented in a random series such that one of them amplitudes were separately examined for each component (P50, N80, and
occurs infrequently (oddball or deviant stimuli). In the present experiment, P200) with multivariate repeated-measures analyses of variance (ANOVA).
480 stimuli were applied to the right hand (frequent stimuli) and 80 stimuli In addition, to further test the topographical effects on SEP amplitudes,
were applied to the left hand (deviant stimuli) in a random order. All stimuli subsequent multivariate repeated-measures ANOVAs were carried out sepa-
were delivered using a pneumatic stimulator consisting of a small membrane rately for midline (Fz, Cz, Pz), and lateral electrode locations over the left (F3,
attached to the second digit by a plastic clip and fixated with adhesive strips. C3, P3) and the right hemisphere (F4, C4, P4). Sociodemographic and
TABLE 1. Clinical and Sociodemographic Characteristics of the Two questionnaire data were analyzed with two-sample t tests to examine differ-
Sample of Patients ences between the two groups (patients with FM versus patients with MSK
pain) on these variables. For pain ratings, an ANOVA for repeated measures
MSK (n ⫽ 16) FM (n ⫽ 27) was performed with the factors “emotional context” (baseline, viewing un-
pleasant versus pleasant slides), “presentation order” (pleasant first versus
Age (yrs) unpleasant first), and “group” (patients with FM versus patients with MSK
Mean (SD) 49.19 (8.27) 51.33 (6.77) pain). Only statistically significant results were reported.
Range 37–64 40–63
Marital status, n (%) RESULTS
Single 3 (17.6%) 1 (3.7%) Comparison of Sociodemographic and
Married 10 (58.8%) 18 (66.7%)
Pain Characteristics
Separated 2 (11.8%) 4 (14.8%)
Widow 1 (5.9%) 4 (14.8%) Descriptive statistics of the self-report measures are shown
Education, n (%) in Table 1. Pairwise comparisons between patients with FM
⬍8 yr 2 (11.8%) 4 (14.8%) and patients with MSK pain revealed that both groups were
8–12 yr 14 (87.5%) 21 (77.8%) comparable in all the clinical and sociodemographic variables,
⬎12 yr 0 (0.0%) 2 (7.4%) with the exception of pain duration in years (t[41] ⫽ 3.11, p ⬍
Medication, n (%) .01). Nevertheless, both groups did not differ on pain medi-
Antidepressants 11 (78.6%) 19 (70.4%) cation or clinical pain ratings.
Analgesics/muscular 10 (66.7%) 20 (74.1%)
relaxants/NSAIDs
Anxiolytics 8 (50.0%) 13 (48.1%) Subjective Pain Ratings
Duration of pain (yrs) The ANOVA for the subjective ratings of current clinical
Mean (SD)* 3.88 (3.41) 9.63 (9.38) pain yielded a significant main effect of “emotional context”
Range 0.83–13.00 1.00–40.00
(F [2, 74] ⫽ 4.98, p ⬍ .05, Greenhouse-Geisser corrected),
Pain intensity (10-cm VAS)
Mean (SD) 5.14 (1.18) 4.66 (1.69)
reflecting an overall increase in pain ratings after either pos-
Range 3.60–8.00 1.60–7.50 itive or negative affective slides as compared with baseline
State-Trait Anxiety ratings. Nevertheless, the primary result of interest was a
Inventory (raw score) significant “emotional context” ⫻ “group” interaction (F [2,
Mean (SD) 30.53 (15.34) 25.76 (9.61) 74] ⫽ 3.43, p ⬍ .05, Greenhouse-Geisser corrected), showing
Range 5–57 9–48 that patients with FM had higher ratings after viewing the
Beck Depression Inventory unpleasant slides than after viewing pleasant slides or baseline
(raw score)
Mean (SD) 18.13 (8.88) 23.88 (9.08) (F [2, 48] ⫽ 6.68, p ⬍ .01, Greenhouse-Geisser corrected)
Range 7–40 8–47 (Figure 1). By contrast, no significant differences on current
McGill Pain Questionnaire pain ratings were found in patients with MSK pain. In fact,
(no. of adjectives) only 4 of 16 patients with MSK pain reported increased
Sensory 8.75 (2.14) 9.81 (1.36) subjective pain ratings after viewing unpleasant in comparison
Affective 1.87 (0.72) 2.18 (0.73) to pleasant slides. No significant effects were found resulting
Evaluative 1.00 (0.00) 0.96 (0.19) from the factor “presentation order” or its interaction with
Miscellaneous 2.46 (0.74) 2.41 (0.84)
other factors, suggesting that the observed emotional modu-
West Haven–Yale
Multidimensional Pain lation of pain ratings in FM was not influenced by the order of
Inventory (0–6) slides (unpleasant first versus pleasant first).
Social support 4.00 (2.10) 3.75 (1.52)
Affective distress 3.75 (1.11) 4.04 (1.11)
Interference 1 3.01 (1.76) 3.52 (1.60)
Interference 2 5.34 (0.64) 5.20 (0.76)
Pain intensity 4.52 (0.97) 4.56 (0.96)
Life control 3.03 (1.73) 3.53 (1.31)
Distracting responses 3.59 (1.50) 3.71 (1.53)
Solicitous responses 2.68 (1.87) 2.29 (1.50)
Punishing responses 1.04 (1.51) 1.15 (1.62)
Household chores 2.69 (1.80) 3.44 (1.25)
Activities away from 2.17 (1.15) 1.64 (1.04)
home
Outdoor work 0.83 (1.52) 0.92 (1.72)
Social activities 2.06 (1.24) 1.45 (1.20)
* p ⬍ .05.
MSK ⫽ musculoskeletal pain; FM ⫽ fibromyalgia; SD ⫽ standard deviation; Figure 1. Subjective ratings of ongoing pain for patients with musculoskel-
etal (MSK) and fibromyalgia (FM) at the beginning of the experiment and
NSAIDs ⫽ nonsteroidal antiinflammatory drugs; VAS ⫽ visual analog scale.
after viewing the pleasant and the unpleasant slides.
Figure 2. Scalp distribution of somatosensory-evoked potentials (SEPs) elicited by nonpainful tactile stimulation at the second digit of the left (deviant stimuli)
or the right hand (frequent stimuli) when patients were viewing either unpleasant or pleasant slides. Waveforms correspond to the SEPs at electrodes C4 and
C3 for deviant and frequent stimuli, respectively. Maps represent the scalp distribution at each peak (P50, N80, P200). Time is in milliseconds. MSK indicates
musculoskeletal pain; FM ⫽ fibromyalgia.
0.17; pleasant slides: mean ⫽ 0.44 V, SE ⫽ 0.12) (F [6, case, these results appear to be in line with extensive literature
10] ⫽ 0.75, NS). indicating that emotion and pain are strongly associated (22).
The multivariate ANOVA on N80 amplitudes (60 –110 ms Our findings are also in agreement with previous work sug-
after stimulus onset) yielded significant effects of “emotional gesting that pain-related negative affect might contribute to
context” ⫻ “type of stimulus” ⫻ “group” (F [1, 41] ⫽ 4.63, clinical pain, particularly in FM (7,9), and that pain cata-
p ⬍ .05), and “emotional context” ⫻ “type of stimulus” (F [1, strophic thoughts might play an important role in increased
41] ⫽ 4.66, p ⬍ .05). Subsequent ANOVAs carried out pain perception of these patients (8).
separately for each group showed that deviant stimuli elicited Together with the lack of significant effects resulting from
more enhanced N80 amplitudes during pleasant slides the presentation order of the slides and the comparable initial
(mean ⫽ ⫺0.78 V, SE ⫽ 0.39) as compared with unpleasant pain ratings of both groups, these results provide further
slides (mean ⫽ 0.15 V, SE ⫽ 0.26) in patients with FM support for the notion of an abnormal processing of body
(F [1, 26] ⫽ 11.97, p ⬍ .01); but no differences were found in information associated with negative mood states in patients
patients with MSK pain (unpleasant slides: mean ⫽ ⫺0.63 with FM. Recently, Philipps et al. (13) using nonpainful
V, SE ⫽ 0.59; pleasant slides: mean ⫽ ⫺0.47 V, SE ⫽ esophageal stimulation and the presentation of fearful versus
0.54) (F [1, 15] ⫽ 0.15, NS). By contrast, no differences neutral faces demonstrated a similar effect on a measure of
resulting from “emotional context” were found on N80 am- perceived discomfort in healthy volunteers. The authors found
plitudes elicited by frequent stimuli, neither in the patients that the intensity of the negative emotional context could
with FM (F [1, 26] ⫽ 0.37, NS) nor the patients with MSK modulate subjective responses and, more interestingly, modify
pain (F [1, 15] ⫽ 0.64, NS). the neural activity to esophageal stimulation within cerebral
The multivariate ANOVA on P200 amplitudes (135–260 regions involved in the processing of visceral sensations.
ms after stimulus onset) yielded no significant effects involv- The present study also demonstrated differential effects of
ing the factors “emotional context” or “group.” experimental mood induction on brain correlates of somato-
sensory information processing in patients with FM and MSK
DISCUSSION
pain. Specifically, it was observed that patients with FM
In the present study, we investigated the affective modu-
displayed a significant enhancement of somatosensory P50, as
lation of somatosensory brain processing among patients with
well as a reduction of N80 amplitudes, when viewing unpleas-
FM and patients with chronic MSK pain resulting from iden-
ant slides in comparison to viewing pleasant slides, whereas
tifiable somatic lesions by using a laboratory-based mood
no differential effect was observed in patients with MSK pain
induction task and the recording of SEPs elicited by nonpain-
at these latencies. In addition, nonpainful tactile stimulation
ful tactile stimulation. Patients were instructed to empathize
elicited enhanced P50 amplitudes in patients with FM in
with the mood state induced by pleasant and unpleasant slides,
comparison to patients with MSK. Earlier P50 and N80 brain
whereas frequent and infrequent target stimuli were simulta-
neously delivered in a randomized order to the hands. Based responses analyzed in this experiment had a latency of 20 to
on previous studies, it was assumed that patients with FM 110 ms after stimulus onset, representing the primary evoked
would be more vulnerable to the influence of negative affect cortical response to somatic stimulation (23). Scalp topogra-
than other patients with chronic MSK pain with somatic phy also showed that the spatial distribution of both SEP
lesions (7). Therefore, it was hypothesized that the experimen- components yielded its maximum over the contralateral hemi-
tal mood induction would elicit differential effects on brain sphere to stimulated body side within the primary somatosen-
correlates of somatosensory information processing and sub- sory cortex (S1). These findings suggest the existence of a
jective pain ratings of patients with FM and patients with highly selective perceptual mechanism in FM, involved in
MSK pain. the sensory processing of nonpainful tactile information in the
Although we found an overall increase in pain ratings after central nervous system, which might be influenced by the
either positive or negative affective slides, behavioral data affective characteristics of the stimulation context. This seems
provided evidence that current pain was differentially modu- to be in agreement with some evidence showing that early
lated by the emotional context in the two patient groups. brain activity can be modulated by psychological factors.
Considering that the selected affective slides were clearly Thus, for instance, it has been found that some laboratory
different on valence and quite similar on arousal ratings, the stressors such as a cold-pressor test or a mental arithmetic task
fact that current pain significantly increased in FM after may attenuate the auditory P50 response (24,25). Recently, it
viewing the unpleasant slides may suggest an abnormal vul- has been also reported the existence of a visual-to-auditory
nerability of these patients to the negative emotional context crossmodal sensory gating phenomenon at very early process-
in which pain occurs. By contrast, we observed no influence ing stages during speech processing (26), as well as a negative
of the negative emotional context on subjective pain ratings of correlation between social anhedonia and somatosensory P50
patients with MSK pain. Thus, the main effect of emotional amplitude in a S1–S2 paradigm among patients with schizo-
context on pain ratings remains puzzling and might suggest phrenia (27). Therefore, although it remains unclear what
that the arousal level elicited by the affective slides could also mechanisms may account for the P50 enhancement observed
play a relevant role by modulating pain experience. In any in FM during the unpleasant pictures, our data suggest that
affective visual stimuli could differentially modulate the early mood states can modulate central nervous excitability thresh-
processing of somatosensory information in these patients. olds without conscious cognitive processing in chronic pain
The present findings are also consistent with previous states. These findings point toward the importance of consid-
evidence, supporting the hypothesis of an abnormal activation ering a biopsychosocial model, integrating affective, cogni-
of pain-related brain regions in FM (2,28 –32). For instance, tive, and social factors, to understand the brain mechanisms
some fMRI studies have shown that patients with FM, in involved in the origin and maintenance of chronic pain.
comparison to healthy control subjects, reported more subjec-
tive pain and displayed more enhanced activation of pain- The authors thank three anonymous reviewers and Dr. J. Richard
related brain regions (including somatosensory cortices, Jennings for their helpful comments on a previous version of the
insular cortex, anterior and posterior cingulate) at similar manuscript.
levels of pressure (2,28,29). Plasticity in both the somatosen-
sory and the motor system has been also observed in neuro- REFERENCES
pathic and musculoskeletal pain such as phantom limb pain 1. Wolfe FW, Smythe HA, Yunas MB, Bennett RM, Bombardier C, Gold-
(33–35), low back pain (36), or complex regional pain syn- enberg DL, Tugwell P, Campbell SM, Ables M, Clark P, Fam AG, Farber
drome (37). In these studies, a significant positive relationship SJ, Fiechtner JJ, Franklin CM, Gatter RA, Hamty D, Lessard J, Licht-
bronn AS, Masi AT, McCain GA, Reynolds WJ, Romano TJ, Russell IJ,
was found between pain intensity and the amount of cortical Sheon RP. The American College of Rheumatology 1990 criteria for the
reorganization of the somatosensory cortex, reflecting a pro- classification of fibromyalgia. Arthritis Rheum 1990;33:160 –72.
cess of central hyperexcitability that might be responsible for 2. Gracely RH, Petzke F, Wolf JM, Clauw DJ. Functional magnetic reso-
nance imaging evidence of augmented pain processing in fibromyalgia.
the maintenance of chronic pain in these patients. Similarly, in Arthritis Rheum 2002;46:1333– 43.
the present study, we found an abnormal brain processing of 3. Lorenz J, Grasedyck K, Bromm B. Middle and long latency somatosen-
bodily information under an aversive context followed by an sory evoked potentials after painful laser stimulation in patients with
fibromyalgia syndrome. Electroencephalogr Clin Neurophysiol 1996;
enhanced perception of clinical pain in FM. Recently, we have 100:165– 8.
also provided a further demonstration that social support may 4. Gibson SJ, Littlejohn GO, Gorman MM, Helme RD, Granges G. Altered
reduce subjective pain ratings and early magnetic brain re- heat pain thresholds and cerebral event related potentials following pain-
ful CO2 laser stimulation in subjects with fibromyalgia syndrome. Pain
sponses to nonpainful stimulation over the somatosensory 1994;58:185–93.
cortex in FM (38). Thus, it is tempting to assume that the P50 5. Granot M, Buskila D, Granovsky Y, Sprecher E, Neumann L, Yarnitsky
might be reflecting some kind of plastic changes in brain D. Simultaneous recording of late and ultra-late pain evoked potentials in
fibromyalgia. Clin Neurophysiol 2001;112:1881–7.
functioning related to pain perception and the processing of 6. Desmeules JA, Cedraschi C, Rapiti E, Baumgartner E, Finckh A, Cohen
nonpainful information in FM. P, Dayer P, Vischer TL. Neurophysiologic evidence for a central sensi-
Basically, our results point to a disturbed cognitive processing tization in patients with fibromyalgia. Arthritis Rheum 2003;48:1420 –9.
7. Davis MC, Zautra AJ, Reich JW. Vulnerability to stress among women in
of somatosensory nonpainful information in FM and provide chronic pain from fibromyalgia and osteoarthritis. Ann Behav Med
further support for the notion that FM is characterized by an 2001;23:215–26.
abnormal brain processing of nociceptive, as well as nonnoci- 8. Geisser ME, Casey KL, Brucksch CB, Ribbens CM, Appleton BB,
Crofford LJ. Perception of noxious and innocuous heat stimulation
ceptive information (39). Nevertheless, it remains unclear among healthy women and women with fibromyalgia: association with
whether these abnormal mechanisms in the processing of so- mood, somatic focus, and catastrophizing. Pain 2003;102:243–50.
matosensory information within FM could also help to explain 9. Staud R, Robinson ME, Vierck CJ Jr, Cannon RC, Mauderli AP, Price
DD. Ratings of experimental pain and pain-related negative affect predict
the enhanced pain sensitivity and the clinical pain reported by clinical pain in patients with fibromyalgia syndrome. Pain 2003;105:
these patients. In this sense, some authors have proposed that 215–22.
attentional factors could also mediate the influence of emotions 10. Meagher MW, Arnau RC, Rhudy JL. Pain and emotion: effects of
affective picture modulation. Psychosom Med 2001;63:79 –90.
on pain in humans (40,41). Although our data do not provide 11. Weisenberg M, Raz T, Hener T. The influence of film-induced mood on
direct support for an attentional bias on FM pain, it seems that pain perception. Pain 1998;76:365–75.
emotional factors might alter brain processing of nonpainful body 12. Zelman DC, Howland EW, Nichols SN, Cleeland CS. The effects of
induced mood on laboratory pain. Pain 1991;46:105–11.
information in FM. Other possible shortcoming of the present 13. Phillips ML, Gregory LJ, Cullen S, Coen S, Ng V, Andrew C, Giampietro
study was the lack of a healthy control group to compare the V, Bullmore E, Zelaya F, Amaro E, Thompson DG, Hobson AR, Wil-
effects of the mood induction on brain activity. Nevertheless, liams SC, Brammer M, Aziz Q. The effect of negative emotional context
on neural and behavioural responses to oesophageal stimulation. Brain
healthy control subjects are difficult to compare with clinical 2003;126:669 – 84.
groups because they cannot be matched according to illness- 14. Eimer M, Forster B. Modulations of early somatosensory ERP compo-
related factors, particularly pain history, medication and treat- nents by transient and sustained spatial attention. Exp Brain Res 2003;
151:24 –31.
ment, social and personal restrictions, and many other factors. To 15. Kida T, Nishihira Y, Wasaka T, Nakata H, Sakamoto M. Passive en-
clarify all these issues, further research should investigate hancement of the somatosensory P100 and N140 in an active attention
whether brain correlates of pain processing can be also modu- task using deviant alone condition. Clin Neurophysiol 2004;115:871–9.
16. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for
lated by emotional factors in FM, as well as in other chronic pain measuring depression. Arch Gen Psychiatry 1961;56:561–71.
diseases and healthy subjects. 17. Spielberger CD, Gorsuch RL, Lushene RE. The State-Trait Anxiety
Summarizing, our data reveal a significant influence of Inventory (STAI): Test Manual. Palo Alto, CA: Consulting Psychologists
Press; 1970.
emotional context on pain processing among patients with FM 18. Kerns RD, Turk DC, Rudy TE. The West Haven-Yale Multidimensional
and patients with MSK pain. It also indicates that affective Pain Inventory (WHYMPI). Pain 1985;23:345–56.
19. Melzack R. The McGill Pain Questionnaire: major properties and scoring Vogt B, FitzGerald L, Mandelkern MA. Brain responses to visceral and
methods. Pain 1975;1:277–99. somatic stimuli in patients with irritable bowel syndrome with and
20. Lang PJ, Bradley MM, Cuthbert BN. International Affective Picture without fibromyalgia. Am J Gastroenterol 2003;98:1354 – 61.
System (IAPS): Technical Manual and Affective Ratings. Gainesville: 32. Wik G, Fischer H, Bragee B, Kristianson M, Fredrikson M. Retrosplenial
The Center for Research in Psychophysiology, University of Florida; cortical activation in the fibromyalgia syndrome. Neuroreport 2003;14:
1997. 619 –21.
21. Zopf R, Giabbiconi CM, Gruber T, Muller MM. Attentional modulation 33. Montoya P, Ritter K, Huse E, Larbig W, Braun C, Topfner S, Lutzen-
of the human somatosensory evoked potential in a trial-by-trial spatial berger W, Grodd W, Flor H, Birbaumer N. The cortical somatotopic map
cueing and sustained spatial attention task measured with high density and phantom phenomena in subjects with congenital limb atrophy and
128 channels EEG. Brain Res Cogn Brain Res 2004;20:491–509. traumatic amputees with phantom limb pain. Eur J Neurosci 1998;10:
22. Keefe FJ, Lumley M, Anderson T, Lynch T, Studts JL, Carson KL. Pain 1095–102.
and emotion: new research directions. J Clin Psychol 2001;57:587– 607. 34. Birbaumer N, Lutzenberger W, Montoya P, Larbig W, Unertl K, Topfner
23. Hari R, Reinikainen K, Kaukoranta E, Hamalainen M, Ilmoniemi R, S, Grodd W, Taub E, Flor H. Effects of regional anesthesia on phantom
Penttinen A, Salminen J, Teszner D. Somatosensory evoked cerebral limb pain are mirrored in changes in cortical reorganization. J Neurosci
magnetic fields from SI and SII in man. Electroencephalogr Clin Neu- 1997;17:5503– 8.
rophysiol 1984;57:254 – 63. 35. Flor H, Elbert T, Knecht S, Wienbruch C, Pantev C, Birbaumer N, Larbig
24. Johnson MR, Adler LE. Transient impairment in P50 auditory sensory W, Taub E. Phantom-limb pain as a perceptual correlate of cortical
gating induced by a cold-pressor test. Biol Psychiatry 1993;33:380 –7. reorganization following arm amputation. Nature 1995;375:482– 4.
25. Yee CM, White PM. Experimental modification of P50 suppression. 36. Flor H, Braun C, Elbert T, Birbaumer N. Extensive reorganization of
Psychophysiology 2001;38:531–9. primary somatosensory cortex in chronic back pain patients. Neurosci
26. Lebib R, Papo D, de Bode S, Baudonniere PM. Evidence of a visual-to- Lett 1997;224:5– 8.
auditory cross-modal sensory gating phenomenon as reflected by the 37. Pleger B, Tegenthoff M, Schwenkreis P, Janssen F, Ragert P, Dinse HR,
human P50 event-related brain potential modulation. Neurosci Lett 2003; Volker B, Zenz M, Maier C. Mean sustained pain levels are linked to
341:185– 8. hemispherical side-to-side differences of primary somatosensory cortex
27. Arnfred SM, Chen AC. Exploration of somatosensory P50 gating in in the complex regional pain syndrome I. Exp Brain Res 2004;155:115–9.
schizophrenia spectrum patients: reduced P50 amplitude correlates to 38. Montoya P, Larbig W, Braun C, Preissl H, Birbaumer N. Influence of
social anhedonia. Psychiatry Res 2004;125:147– 60. social support and emotional context on pain processing and magnetic
28. Cook DB, Lange G, Ciccone DS, Liu WC, Steffener J, Natelson BH. brain responses in fibromyalgia. Arthritis Rheum 2004;50:4035– 44.
Functional imaging of pain in patients with primary fibromyalgia. 39. Montoya P, Pauli P, Batra A, Wiedemann G. Altered processing of
J Rheumatol 2004;31:364 –78. pain-related information in patients with fibromyalgia. Eur J Pain 2005;
29. Giesecke T, Gracely RH, Grant MA, Nachemson A, Petzke F, Williams 9:293–303.
DA, Clauw DJ. Evidence of augmented central pain processing in idio- 40. Dehghani M, Sharpe L, Nicholas MK. Selective attention to pain-related
pathic chronic low back pain. Arthritis Rheum 2004;50:613–23. information in chronic musculoskeletal pain patients. Pain 2003;105:
30. Gracely RH, Geisser ME, Giesecke T, Grant MA, Petzke F, Williams 37– 46.
DA, Clauw DJ. Pain catastrophizing and neural responses to pain among 41. Keogh E, Ellery D, Hunt C, Hannent I. Selective attentional bias for
persons with fibromyalgia. Brain 2004;127:835– 43. pain-related stimuli amongst pain fearful individuals. Pain 2001;91:
31. Chang L, Berman S, Mayer EA, Suyenobu B, Derbyshire S, Naliboff B, 91–100.
a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s
Research Report
A R T I C LE I N FO AB S T R A C T
Article history: The present investigation was aimed to evaluate the influence of emotional valence on
Accepted 7 November 2005 brain correlates of non-painful somatosensory processing. For this purpose, we examined
Available online 20 December 2005 changes on the somatosensory-evoked potentials (SEP) elicited by frequent and deviant
tactile stimuli (probability 14%) when subjects were viewing affective pictures. Twenty
Theme: healthy volunteers aged between 19 and 47 years old participated in the study. The P50, N80,
Neural basis of behavior and P200 components of the SEP, as well as the P200 component of the visual-evoked
Topic: potentials (VEP) elicited by the affective pictures were analyzed. Overall, a significant P50
Motivation and emotion amplitude reduction was observed when subjects were viewing unpleasant pictures, in
comparison to pleasant pictures. Furthermore, larger SEP amplitudes were obtained in
Keywords: response to the deviant than to the frequent stimuli. In addition, unpleasant pictures
Somatosensory-evoked potentials elicited larger P200 amplitudes of the VEP than pleasant. Data suggest that affective stimuli
Emotions may modulate the early processing of somatosensory information in the brain, probably
Oddball task reflecting the existence of an adaptive perceptual/attentional mechanism to motivationally
Affective pictures relevant stimuli.
© 2005 Elsevier B.V. All rights reserved.
1. Introduction occur (Bradley and Lang, 2000). Using affective slides depicting
scenes of varying emotional valence, several laboratories have
It is well-known that emotions may influence our sensory demonstrated that the startle response is enhanced during
experience. Thus, for instance, it has been observed that unpleasant compared to neutral and pleasant pictures (Brad-
negative mood states, such as depression or anxiety, can ley and Lang, 2000; Ehrlichman et al., 1995). Recently, we have
enhance pain severity (Keefe et al., 2001; McWilliams et al., even shown that social support given by significant others
2003), or that emotional traumatic events may lead to may reduce subjective pain perception and modulate brain
recurrent and vivid recollections of sensory experiences responses to innocuous tactile stimuli in fibromyalgia patients
(Hendler et al., 2003; Newport and Nemeroff, 2000). Recent (Montoya et al., 2004).
research has also provided increasing experimental evidence Electrophysiological studies have also revealed that affec-
suggesting an interaction among emotion and sensory tive stimuli can modulate the late components of the event-
processing. One of the most consistent finding has been that related potentials (ERPs). Specifically, enhanced P200 ampli-
sensory reflexes, such as the acoustic startle eyeblink reflex, tudes and a sustained later positivity have been observed in
can be modulated by the emotional context in which they response to emotionally salient (i.e., pleasant or unpleasant)
0006-8993/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.brainres.2005.11.019
206 BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –21 2
compared to neutral stimuli (Carretie et al., 2001, 2003; elicited during the tactile stimulation appeared to affect
Cuthbert et al., 2000; Montoya et al., 1996). The study of differentially the P50 amplitude elicited by the deviant and
brain mechanisms governing affective modulation of sensory the frequent stimuli. Indeed, a significant ‘emotional
processing in humans has also attracted some attention. context’ × ‘type of stimulus’ interaction effect was found
Recent brain research has shown that the visual cortex might (F[1,32] = 7.46, P b 0.05), indicating that the P50 enhance-
be differentially activated as a function of valence using ment associated with pleasant pictures was elicited by
affective stimulus presentation (Lang et al., 1998; Reiman et deviant stimuli (multivariate contrast: F[1,32] = 7.89,
al., 1997; Taylor et al., 2000). Thus, significant increases in P b 0.01), but not by frequent stimuli (multivariate contrast:
regional blood flow of the visual cortex have been found F[1,32] = 0.45, NS).
when participants viewed unpleasant films in comparison to The somatosensory P50 amplitude showed a centropar-
neutral films (Reiman et al., 1997) or when they imagined ietal-dominant scalp distribution contralateral to the stimu-
traumatic compared to neutral non-affective experiences lation side for the deviant stimuli during pleasant pictures,
(Rauch et al., 1996). Lang et al. (1998) and Sabatinelli et al. with maximum amplitudes at electrodes CP4 (2.94 μV), and
(2004) have shown similar changes in the occipital gyri, the C4 (2.49 μV). Statistical analyses revealed a significant main
right fusiform gyrus, and the right inferior and superior effect of the factor ‘electrode site’ (F[18,576] = 30.66, P b 0.001)
parietal lobules, when subjects were processing affective and a significant interaction between the factors ‘type of
pictures. Moreover, a recent study found that a negative stimulus’ and ‘electrode site’ (F[18,576] = 23.61, P b 0.001) on
emotional context compared with a neutral one can influ- P50 peak amplitudes. To further analyze topographical
ence the activity upon ventral and dorsal anterior cingulate effects, univariate ANOVAS were computed to test the effects
gyrus, as well as the insular responses to non-painful of ‘emotional context’ and ‘type of stimulus’ on P50
esophageal stimulation (Phillips et al., 2003). These data amplitudes at each electrode location. Significant effects of
further suggest that emotional inputs might undergo more the factor ‘emotional context’ were found at F4, C4, CP3, and
elaborated processing than non-affective stimuli at a very P3, indicating that tactile stimuli elicited enhanced P50
early stage in cortical afferent analysis. amplitudes during viewing pleasant pictures. Significant
In the present experiment, we investigated the effects of effects of the factor ‘type of stimulus’ were also found at
emotion on brain processing of somatosensory information. electrodes Cz, Pz, CPz, C3, CP3, C4, P4, CP4, T4, T6, and O2,
For this purpose, somatosensory-evoked potentials (SEPs) revealing that deviant stimuli elicited greater P50 amplitudes
elicited by innocuous tactile stimulation were recorded than frequent stimuli. In addition, significant interaction
when healthy subjects were viewing a sample of emotionally effects between the factors ‘emotional context’ and ‘type of
arousing pictures from the International Affective Picture stimulus’ were found at Pz, CPz, P3, CP3, O1, F4, C4, P4, T6, O2,
System (IAPS) (Lang et al., 1999). Subjects were asked to showing again that differences due to emotional context
empathize with the mood state elicited by the pictures. were present on P50 amplitudes elicited by deviant, but not
According with the motivational priming hypothesis (Lang et by frequent stimuli (see Fig. 2).
al., 1998), it was predicted that the brain processing of tactile A similar analysis of variance was carried out on P50
information would be modulated by the emotional context in amplitudes elicited by tactile deviant and frequent stimuli
which it occurs. Particularly, we expected that the emotional during the inter-stimulus interval (blank screen). Significant
context induced by unpleasant pictures would reduce ampli- main effects of the factors ‘type of stimulus’ (F[1,32] = 9.28,
tudes of early SEP components elicited by tactile stimuli in P b 0.01), ‘electrode site’ (F[18,576] = 18.59, P b 0.001), as well as a
comparison with the emotional context induced by pleasant interaction effect between the factors ‘type of stimulus’ and
pictures. ‘electrode site’ (F[18,576] = 25.93, P b 0.001) were found. The P50
amplitudes were larger to the deviant compared with the
frequent stimuli at electrodes Cz, C4, Pz, P4, CPz, CP4, T4, T6,
2. Results and O2. Again, the P50 amplitude showed a centroparietal-
dominant scalp distribution contralateral to the stimulation
2.1. SEP elicited by deviant and frequent tactile stimuli side for the deviant stimuli, with maximum amplitudes at
electrodes CP4 (2.75 μV, during pleasant pictures, and 2.59 μV,
Within the first 250-ms interval after stimulus onset, during unpleasant pictures). However, in this case, no
somatosensory-evoked potential (SEP) waveforms elicited significant effect of the emotional context was found on
by both types of stimuli were characterized by a positive somatosensory P50 amplitudes during the inter-stimulus
peak (P50), followed by a negative (N80) and a second interval, thus, confirming that the observed affective modu-
positive peak (P200) (see Fig. 1). The scalp topography of lation of this SEP component was specifically linked to the
the P50 and N80 components clearly indicated that brain view of affective pictures.
activity elicited by tactile stimuli was more prominent over The somatosensory N80 peak amplitude (60–110 ms after
centroparietal regions of the hemisphere contralateral to stimulus onset) showed a centrotemporal-dominant scalp
stimulation than over the ipsilateral hemisphere. distribution contralateral to the stimulation side, with max-
The somatosensory P50 peak amplitudes (20–80 ms after imum amplitudes at electrode C4 (−2.16 μV and −2.13 μV, for
stimulus onset) were overall larger when viewing pleasant deviant stimuli during pleasant and unpleasant pictures,
as compared to unpleasant pictures (F[1,32] = 5.10, P b 0.05), respectively). The statistical analyses of the N80 amplitudes
and larger for deviant compared to frequent tactile stimuli revealed a significant effect of the factor ‘electrode site’ (F
(F[1,32] = 11.40, P b 0.01). Moreover, the emotional context [18,576] = 9.50, P b 0.001) and a significant interaction between
BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –2 12 207
Fig. 1 – Somatosensory-evoked potentials (SEP) at electrode C3 for frequent (A) and at C4 for deviant stimuli (B), when
participants were viewing pleasant and unpleasant pictures. Scalp maps represent the electrical distribution of the P50, N80,
and P200 components for tactile stimuli during the mood-inducing conditions.
the factors ‘type of stimulus’ and ‘electrode site’ (F ‘electrode site’ (F[18,576] = 55.93, P b 0.001), as well as a
[18,576] = 13.79, P b 0.001). To further analyze the topographical significant interaction between the factors ‘type of stimulus’
effects, univariate ANOVAS were computed to test the effects and ‘electrode site’ (F[18,576] = 25.16, P b 0.001). Univariate
of ‘type of stimulus’ at each electrode location. The N80 ANOVAs confirmed that P200 amplitudes were larger for the
amplitudes were significantly larger for deviant than for deviant compared with the frequent stimuli at all electrodes.
frequent stimuli at electrodes C3, C4, P3, P4, CP3, CP4, T4, No significant effects involving the factor ‘emotional context’
and T6. No significant effects involving the factor ‘emotional were found.
context’ were found.
The somatosensory P200 peak amplitude (135–260 ms 2.2. VEP elicited by pleasant and unpleasant pictures
after stimulus onset) showed a centroparietal-dominant
scalp distribution with maximum amplitudes at electrode The scalp topography of the visual P200 peak amplitude (135–
Cz (7.54 μV and 7.29 μV, for deviant stimuli during unpleasant 260 ms after stimulus onset) indicated that brain activity
and pleasant pictures, respectively). The statistical analyses elicited by unpleasant stimuli was larger than the activity
of the P200 amplitudes revealed significant effects of the elicited by pleasant stimuli (F[1,32] = 4.36, P b 0.05) (Fig. 3). In
factors ‘type of stimulus’ (F[1,32] = 48.00, P b 0.001) and addition, univariate ANOVAs revealed that these differences
208 BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –21 2
Fig. 2 – Mean amplitudes and standard errors of the somatosensory P50 component elicited by frequent (A) and deviant (B)
tactile stimuli under the two mood-inducing conditions.
on P200 amplitudes of the VEP were located at electrodes P3, embedded within an unpleasant context in comparison with
CP3, T5, T6, and O2 (see Fig. 4). P50 elicited by deviant stimuli during a pleasant context. The
topographical analyses also revealed that the P50 and N80
components elicited to the deviant stimuli showed a centro-
3. Discussion parietal-dominant scalp distribution, with maxima ampli-
tudes in the hemisphere contralateral to the stimulation side.
Somatosensory-evoked potentials (SEP) to tactile stimulation In addition, P200 amplitudes of the visual-evoked potentials
were examined in an oddball paradigm in which frequent and (VEP) elicited by unpleasant pictures were larger than those
rarely interspersed deviant stimuli were applied to the index elicited by pleasant pictures.
finger of the two hands, and simultaneously, attention was The first somatosensory brain response peaked around 50
directed away from stimuli instructing subjects to view ms after stimulus onset and was stronger over the contralat-
affective pictures. Larger P50, N80, and P200 amplitudes of eral hemisphere to stimulation side. According with previous
the SEPs were overall observed in response to deviant in electrophysiological recordings, it is likely that this brain
comparison to frequent tactile stimuli. Nevertheless, reduced response was generated in the primary somatosensory cortex
P50 amplitudes were found when deviant stimuli were (Hari et al., 1990). Other studies have also shown an
BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –2 12 209
Fig. 3 – Visual-evoked potentials (VEP) at electrode O2 and scalp distribution elicited by pleasant and unpleasant pictures.
Scalp maps represent the electrical distribution of the P200 component elicited by the affective pictures.
enhancement of the somatosensory P50 amplitude when and Belger, 1999). In this sense, this gating would facilitate the
subjects were instructed to attend to the stimulation site control of information processing by a filtering mechanism,
(Zopf et al., 2004). Recently, we have reported an attenuation of allowing individuals to attend to relevant stimuli and to ignore
P50 amplitudes among fibromyalgia patients associated with irrelevant information.
the presence of patients' significant others (Montoya et al., Although the precise relationship between the magnitude
2004). Thus, it may be that the observed P50 reduction during of the P50 attenuation and the strength of inhibitory brain
the unpleasant emotional context in the present study might circuits has yet to be clarified, there is some evidence pointing
be reflecting an attentional gating mechanism, which allows towards the modulation of this gating mechanism by psycho-
the brain to ‘gate out’ incoming irrelevant somatosensory logical factors. Thus, for instance, it has been found that some
input and to ‘gate in’ motivationally relevant information. laboratory stressors, such as a cold-pressor test or a mental
This interpretation seems to be in accordance with previous arithmetic task, may attenuate the auditory P50 response
findings indicating that the P50 reduction elicited by repetitive (Johnson and Adler, 1993; Yee and White, 2001). Recently, it
stimuli in a S1–S2 paradigm constitutes a reliable index of a has been also reported the existence of a visual-to-auditory
sensory gating phenomenon (Arnfred and Chen, 2004; Boutros cross-modal sensory gating phenomenon at very early
Fig. 4 – Mean amplitudes and standard errors of the visual P200 component elicited by pleasant and unpleasant pictures.
210 BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –21 2
processing stages during speech processing (Lebib et al., 2003), in response to deviant stimuli as compared with frequent
as well as a negative correlation between social anhedonia stimuli. Based on the onset latency and scalp distribution of
and somatosensory P50 amplitude in a S1–S2 paradigm among this component, similar to the P300 component, changes
schizophrenic patients (Arnfred and Chen, 2004). Therefore, elicited by tactile deviant stimuli could be interpreted as
although it remains unclear what mechanisms may account reflecting a more complex cognitive functioning associated
for the P50 reduction observed during the unpleasant pictures with memory or stimulus evaluation (Polich and Herbst, 2000).
in the present study, our data suggest that affective visual In summary, the present study demonstrated that early
stimuli could differentially modulate the early processing of somatosensory information processing can be modulated by
somatosensory information in healthy subjects. emotional influences. In particular, it has been observed that
The interpretation of an affective modulation of somato- an unpleasant context elicited a significant attenuation of a
sensory processing information is also indirectly supported by sensory-specific somatosensory ERP component within the
our finding that unpleasant pictures elicited larger P200 first 50 ms after stimulus onset. This modulation of somato-
amplitudes of the VEPs than pleasant ones. In the present sensory processing could be interpreted as reflection of an
study, we used two sets of emotional pictures with similar initial sensory gating mechanism that allows the filtering of
ratings of arousal and from a validated pool of affective motivationally relevant information. While this hypothesis
pictures (Lang et al., 1999). Therefore, the observed differences needs to be confirmed by further research, it seems that the
on the visual P200 amplitudes between unpleasant and dual-task as used in the present study may constitute a
pleasant pictures should be only attributable to the different powerful tool to explore the interaction between emotion and
stimulus valence. These findings are in agreement with cognitive processes in clinical populations, such as chronic
previous data, showing that negative emotional pictures elicit pain patients and persons with increased vigilance to bodily
higher P200 amplitudes and shorter P200 latencies than sensations or anxiety disorders.
positive pictures (Carretie et al., 2001, 2003). Consistent with
our previous interpretation, the authors suggested that the
visual P200 component might be considered as an attention-
4. Experimental procedures
related index of stimulus ability to catch subjects' attention. In
this sense, the enhancement of the visual P200 amplitude Thirty-three healthy female students with normal or corrected-to-
could be reflecting an attentional bias towards negative normal vision whose age ranged from 19 to 47 years (mean
stimuli during emotional processing. Thus, it seems conceiv- age = 26.5 years, SD = 7.97) participated in the experiment. All
able that the somatosensory P50 attenuation elicited by the participants gave written informed consent according with the
Declaration of Helsinki. The experimental procedure was ap-
unpleasant pictures observed in this study could be due to the
proved by the local institutional ethics committee.
strong orienting and attentional effect raised by the unpleas-
ant pictures. Nevertheless, additional empirical research is 4.1. Somatosensory stimuli
necessary to more precisely determine the effects of attention
and salience on somatosensory gating. A possible shortcom- Non-noxious tactile stimuli were delivered at the 2nd digit of the
ing of the present experiment is that deviant stimuli were right (frequent stimulus) and the left hand (deviant stimulus) by
means of a pneumatic stimulator, consisting of a small membrane
always applied to the left hand (right hemisphere), and that
attached to the body surface by a plastic clip and fixated with
the comparative effect on sensory gating elicited by stimuli at
adhesive strips. Stimuli were applied using an oddball paradigm
the right hand could not be explored. Further research should with probabilities of 0.86 and 0.14 for frequent and deviant stimuli,
also address the question of whether the sensory gating respectively. Each stimulation block consisted of 560 stimuli of 100
effects are extensive to other body locations and to other ms duration with an approximate pressure of 2 bars and a variable
sensory modalities. inter-stimuli interval of 550 ± 50 ms.
Although the somatosensory P50, N80, and P200 ampli-
4.2. Affective stimuli
tudes overall increased for the deviant tactile stimuli com-
pared with the frequent stimuli, it is likely that different brain During the somatosensory stimulation task, subjects were simul-
processes accounted for these differences in the processing of taneously viewing a sequence of pictures selected from the
somatosensory information. Thus, considering that the devi- International Affective Picture System (IAPS) (Lang et al., 1999).
ant and the frequent stimuli were applied to different body The IAPS is a set of photographs which contents ranges from
locations, it is conceivable that the changes on P50 amplitudes explicit sexual material, to human injury and surgical slides, to
were only reflecting the differential activation of the S1 pleasant pictures of children and wildlife. Two counterbalanced
blocks of each forty pictures were presented. In one block, the
cortices. With regard to the changes on the somatosensory
pictures were always positive (pleasant content), whereas in the
N80 amplitudes, our data appear to be in agreement with other block, the pictures were always negative (unpleasant
previous studies linking the enhancement of the contralateral content). Each individual picture was viewed a total of two times
N80 component with sustained attention to a body location by the participants. Pictures were displayed for 6 s and followed by
(Eimer and Forster, 2003; Hotting et al., 2003). In this sense, a 6-s blank screen.
some authors suggested that the early contralateral negativity
4.3. Experimental tasks
during sustained attention could be interpreted as reflection
of a spatially selective mechanism, which could be related
Subjects received the tactile stimulation when they were viewing
with the sensory gating of activity in S1 cortex (Eimer and the affective pictures. The presentation order of the picture
Forster, 2003). Finally, the present study also revealed a sequences was counterbalanced between the subjects: for half of
significant enhancement of somatosensory P200 amplitudes the participants, the aversive pictures were presented followed by
BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –2 12 211
the pleasant pictures; while the other half viewed the affective Bradley, M.M., Lang, P.J., 2000. Affective reactions to acoustic
pictures in the reverse order. The delivery of tactile stimuli was stimuli. Psychophysiology 37, 204–215.
programmed to allow only one deviant stimulus during each 6-s Carretie, L., Martin-Loeches, M., Hinojosa, J.A., Mercado, F., 2001.
picture presentation and one deviant during the 6-s blank screen. Emotion and attention interaction studied through
Patients were instructed to ignore tactile stimulation and to pay event-related potentials. J. Cogn. Neurosci. 13, 1109–1128.
attention to the slides trying to imagine experiencing them- Carretie, L., Hinojosa, J.A., Mercado, F., 2003. Cerebral patterns of
selves in the situations described by the pictures. Subjects were attentional habituation to emotional visual stimuli.
seated in a reclining chair in a sound attenuated room and Psychophysiology 40, 381–388.
instructed to keep eye movements and blinks to minimum Cuthbert, B.N., Schupp, H.T., Bradley, M.M., Birbaumer, N., Lang,
during the experiment. P.J., 2000. Brain potentials in affective picture processing:
covariation with autonomic arousal and affective report. Biol.
4.4. EEG recording and analysis Psychol. 52, 95–111.
Ehrlichman, H., Brown, S., Zhu, J., Warrenburg, S., 1995. Startle
Electroencephalography (0.1–40 Hz) was recorded from 19 scalp reflex modulation during exposure to pleasant and unpleasant
electrodes (Fz, F3, F4, FCz, Cz, C3, C4, CPz, CP3, CP4, Pz, P3, P4, T3, odors. Psychophysiology 32, 150–154.
T4, T5, T6, O1, O2) with linked-earlobe reference. Horizontal and Eimer, M., Forster, B., 2003. Modulations of early somatosensory
vertical electrooculograms (EOG) were recorded bipolarly from the ERP components by transient and sustained spatial attention.
outer canthi of both eyes. Electrode impedance was kept below 10 Exp. Brain Res. 151, 24–31.
kΩ. The digitization rate was 1000 Hz. Hari, R., Hamalainen, H., Hamalainen, M., Kekoni, J., Sams, M.,
Somatosensory-evoked potentials (SEP) to frequent (tactile Tiihonen, J., 1990. Separate finger representations at the
stimuli delivered to the right hand) and deviant stimuli (left hand) human second somatosensory cortex. Neuroscience 37,
were separately averaged relative to a 100-ms prestimulus 245–249.
baseline for both task conditions (viewing unpleasant vs. pleasant Hendler, T., Rotshtein, P., Yeshurun, Y., Weizmann, T., Kahn, I.,
pictures). Peak amplitudes were computed for the following SEP Ben-Bashat, D., Malach, R., Bleich, A., 2003. Sensing the
components: P50 (20–80 ms after stimulus onset at electrodes C3/ invisible: differential sensitivity of visual cortex and amygdala
C4), N80 (60–110 ms at C3/C4), and P200 (135–260 ms at Cz). To to traumatic context. NeuroImage 19, 587–600.
investigate longer latency effects of the emotional context on Hotting, K., Rosler, F., Roder, B., 2003. Crossmodal and intermodal
somatosensory processing, mean amplitudes were also computed attention modulate event-related brain potentials to tactile
between 210 and 460 ms after stimulus onset. In addition, visual- and auditory stimuli. Exp. Brain Res. 148, 26–37.
evoked potentials (VEP) elicited by pleasant and unpleasant Johnson, M.R., Adler, L.E., 1993. Transient impairment in P50
pictures were also analyzed. The averaging window spanned auditory sensory gating induced by a cold-pressor test. Biol.
1500 ms from 100 ms prior to picture onset. After baseline Psychiatry 33, 380–387.
correction, the P200 peak amplitudes (120–250 ms at Pz) were Keefe, F.J., Lumley, M., Anderson, T., Lynch, T., Studts, J.L., Carson,
determined for all electrodes. K.L., 2001. Pain and emotion: new research directions. J. Clin.
SEP amplitudes were statistically analyzed using multivariate Psychol. 57, 587–607.
analyses of variance for repeated measures with the factors Lang, P.J., Bradley, M.M., Fitzsimmons, J.R., Cuthbert, B.N., Scott,
‘emotional context’ (unpleasant vs. pleasant slides), ‘type of J.D., Moulder, B., Nangia, V., 1998. Emotional arousal and
stimuli’ (frequent vs. deviant), and ‘electrode site’ (19 electrodes). activation of the visual cortex: an fMRI analysis.
Multivariate analyses of variance for the components of the VEPs Psychophysiology 35, 199–210.
included the factors ‘emotional context’ (unpleasant vs. pleasant Lang, P.J., Bradley, M.M., Cuthbert, B.N., 1999. International
slides) and ‘electrode site’ (19 electrodes). Greenhouse–Geisser affective picture system (IAPS): technical manual and affective
adjustments to the degrees of freedom were used. ratings. NIMH-Center for the Study of Emotion and Attention,
University of Florida, Gainesville.
Lebib, R., Papo, D., de Bode, S., Baudonniere, P.M., 2003. Evidence of
a visual-to-auditory cross-modal sensory gating phenomenon
Acknowledgments as reflected by the human P50 event-related brain potential
modulation. Neurosci. Lett. 341, 185–188.
Research was supported by the Spanish Ministerio de Ciencia McWilliams, L.A., Cox, B.J., Enns, M.W., 2003. Mood and
y Tecnología and European Funds (Fondos FEDER) (grants anxiety disorders associated with chronic pain: an
BSO2001-0693 and SEJ2004-01332). examination in a nationally representative sample. Pain
106, 127–133.
Montoya, P., Larbig, W., Pulvermuller, F., Flor, H., Birbaumer, N.,
1996. Cortical correlates of semantic classical conditioning.
Appendix A. Supplementary data Psychophysiology 33, 644–649.
Montoya, P., Larbig, W., Braun, C., Preissl, H., Birbaumer, N., 2004.
Supplementary data associated with this article can be found Influence of social support and emotional context on pain
in the online version at doi:10.1016/j.brainres.2005.11.019. processing and magnetic brain responses in fibromyalgia.
Arthritis Rheum. 50, 4035–4044.
Newport, D.J., Nemeroff, C.B., 2000. Neurobiology of PTSD. Curr.
REFERENCES Opin. Neurobiol. 10, 211–218.
Phillips, M.L., Gregory, L.J., Cullen, S., Coen, S., Ng, V., Andrew, C.,
Giampietro, V., Bullmore, E., Zelaya, F., Amaro, E., Thompson,
Arnfred, S.M., Chen, A.C., 2004. Exploration of somatosensory P50 D.G., Hobson, A.R., Williams, S.C., Brammer, M., Aziz, Q.,
gating in schizophrenia spectrum patients: reduced P50 Cohen, S., 2003. The effect of negative emotional context on
amplitude correlates to social anhedonia. Psychiatry Res. 125, neural and behavioural responses to oesophageal stimulation.
147–160. Brain 126, 669–684.
Boutros, N.N., Belger, A., 1999. Midlatency evoked potentials Polich, J., Herbst, K.L., 2000. P300 as a clinical assay: rationale,
attenuation and augmentation reflect different aspects of evaluation, and findings. Int. J. Psychophysiol. 38, 3–19.
sensory gating. Biol. Psychiatry 45, 917–922. Rauch, S.L., Bessel, A., Fisler, R.E., Alpert, N.M., Orr, S.A., Savage,
212 BR A I N R ES E A RC H 1 0 6 8 ( 2 00 6 ) 2 0 5 –21 2
C.R., Fischman, A.J., Jenike, M.A., Pitman, R.K., 1996. A Taylor, S.F., Liberzon, I., Koeppe, R.A., 2000. The effect of graded
symptom provocation study of posttraumatic stress disorder aversive stimuli on limbic and visual activation.
using positron emission tomography and script-driven Neuropsychologia 38, 1415–1425.
imagery. Arch. Gen. Psychiatry 53, 380–387. Yee, C.M., White, P.M., 2001. Experimental modification of P50
Reiman, E.M., Lane, R.D., Ahern, G.L., Schwartz, G.E., 1997. suppression. Psychophysiology 38, 531–539.
Neuroanatomical correlates of internally and externally Zopf, R., Giabbiconi, C.M., Gruber, T., Muller, M.M., 2004.
generated human emotion. Am. J. Psychiatry 154, 918–925. Attentional modulation of the human somatosensory evoked
Sabatinelli, D., Flaisch, T., Bradley, M.M., Fitzsimmons, J.R., Lang, potential in a trial-by-trial spatial cueing and sustained spatial
P.J., 2004. Affective picture perception: gender differences in attention task measured with high density 128 channels EEG.
visual cortex? NeuroReport 15, 1109–1112. Brain Res. Cogn. Brain Res. 20, 491–509.
ARTHRITIS & RHEUMATISM
Vol. 54, No. 6, June 2006, pp 1995–2003
DOI 10.1002/art.21910
© 2006, American College of Rheumatology
Objective. To examine brain activity elicited by Conclusion. Our findings suggest that in FM
repetitive nonpainful stimulation in patients with fibro- patients, there is abnormal information processing,
myalgia (FM) and to determine possible psychophysio- which may be characterized by a lack of inhibitory
logic abnormalities in their ability to inhibit irrelevant control to repetitive nonpainful somatosensory informa-
sensory information. tion during stimulus coding and cognitive evaluation.
Methods. Fifteen female patients with a diagnosis
of FM (ages 30–64 years) and 15 healthy women (ages The syndrome of fibromyalgia (FM) constitutes a
39–61 years) participated in 2 sessions, during which chronic musculoskeletal pain disorder characterized by
electrical activity elicited in the brain by presentation of widespread lowered pain threshold, fatigue, muscle stiff-
either tactile or auditory paired stimuli was recorded ness, and emotional distress (1). Recent research has
using an electroencephalogram. Each trial consisted of examined the possibility that pain and tenderness in FM
2 identical stimuli (S1 and S2) delivered with a random- might be linked to altered neurobiologic mechanisms
ized interstimulus interval of 550 msec (ⴞ50 msec), that function during the processing of somatosensory
which was separated by a fixed intertrain interval of 12 information (2–8). In this regard, studies examining
seconds. Event-related potentials (ERPs) elicited by 40 sensitivity to experimentally induced pain have consis-
trials were averaged separately for each sensory modal- tently demonstrated that FM patients show more en-
ity. hanced pain sensitivity within minutes after sustained
Results. ERP amplitudes elicited by the somato- nociceptive stimulation (increased wind-up responses)
sensory and auditory S2 stimuli were significantly re- than do healthy controls (2,3). We have also found that
duced compared with those elicited by S1 stimuli in the FM patients have lowered pressure–pain thresholds to
healthy controls. Nevertheless, significant amplitude repeated measurements within a session and over days
reductions from S1 stimuli to S2 stimuli were observed (4). Moreover, it has been suggested that FM patients
in FM patients for the auditory, but not the somatosen-
might have deficiencies in central inhibitory mecha-
sory, modality.
nisms, such as diffuse noxious inhibitory control (5,6) or
the endogenous pain inhibitory system (7). Indeed, some
Supported by the Spanish Ministerio de Ciencia y Tecnologı́a studies have found that, unlike in healthy subjects, tonic
(BSO2001-0693 and SEJ2004-01332) and by funds from the European conditioning nociceptive stimulation does not reduce the
Union (Fondos FEDER).
1
Pedro Montoya, PhD, Carolina Sitges, MS, Alfonso perception of noxious or innocuous stimuli in FM pa-
Rodrı́guez-Cotes, MS: Research Institute on Health Sciences (IU- tients (6). Taken together, these data seem to support
NICS), University of the Balearic Islands, Palma, Spain; 2Manuel the hypothesis that FM should be considered a pain
Garcı́a-Herrera, MD, PhD, Raúl Izquierdo, MD, Dolores Collado,
MD: Medical Unit for Disability Assessment (UMEVI), Social Secu- disorder characterized by enhanced sensitization of the
rity Agency, Palma, Spain; 3Magdalena Truyols, MS: General Hospi- central nervous system, which leads to long-term neuro-
tal, Palma, Spain. plastic changes and to abnormal processing of nocicep-
Address correspondence and reprint requests to Pedro
Montoya, PhD, Department of Psychology, Research Institute on tive information (3,8).
Health Sciences (IUNICS), University of the Balearic Islands, Car- Nevertheless, little is yet known about the brain
retera de Valldemossa, km 7.5, 07122 Palma, Spain. E-mail: pedro. mechanisms involved in the processing of nonpainful
[email protected].
Submitted for publication September 20, 2005; accepted in somatosensory information in FM. Previous brain imag-
revised form March 9, 2006. ing studies have reported an abnormal reduction of
1995
1996 MONTOYA ET AL
regional cerebral blood flow in thalamic and caudate PATIENTS AND METHODS
nuclei of patients with FM during rest (9,10). Recently, Study subjects. Fifteen female patients with a main
2 studies using functional magnetic resonance imaging diagnosis of fibromyalgia (FM) (mean ⫾ SD age 49.67 ⫾ 8.24
revealed that FM patients exhibit enhanced responses to years) and 15 healthy women (mean ⫾ SD age 48.0 ⫾ 5.87
painful and nonpainful stimulation in multiple areas of years) participated in the study. Subjects were excluded from
the brain, such as the somatosensory cortices, insula, the study if they were pregnant, had neurologic disease, or
were taking opioids. FM patients were included in the study if
putamen, anterior cingulate cortex, and cerebellum, as
they fulfilled the classification criteria of the American College
compared with healthy control subjects (11,12). In addi- of Rheumatology (1), with a minimum of 11 tender points (of
tion, brain responses to painful simulation in FM pa- a total of 18 specific tender points), and had pain as their
tients were characterized by reduced thalamic activity dominant symptom. FM patients underwent an extensive med-
relative to that in the healthy controls, which was ical assessment by an experienced rheumatologist to confirm
that they met the inclusion criteria. Four patients reported
interpreted as an abnormal inhibitory mechanism in-
comorbid pain diseases: 2 had osteoporosis, 1 had back pain,
duced by persistent excitatory input associated with and 1 had trigeminal neuralgia. All study participants com-
ongoing pain. pleted the Beck Depression Inventory (19) and the State-Trait
In the present study, we further addressed the Anxiety Inventory (20). FM patients also completed the West
question of abnormal brain processing of nonpainful Haven–Yale Multidimensional Pain Inventory (21). Patients
were allowed to continue taking their long-term medications
somatosensory information in FM patients by examining including those specifically for FM.
event-related potentials (ERPs) from the electroenceph- At the time of recruitment, subjects were verbally
alogram (EEG). Early and late ERP components elic- informed about the details of the study, noting that their
ited by the repetition of identical stimuli have been participation in the study was not linked to their possible
litigation process. A patient information leaflet designed spe-
frequently used to examine the brain’s ability to inhibit cifically for the study was given to each subject, and after they
sensory information (13–15). In this regard, several agreed to participate, each subject provided written consent.
psychophysiologic studies have consistently shown that The study was conducted in accordance with the Declaration
ERP amplitudes elicited by the second stimulus in a of Helsinki (1991 version) and was approved by the Ethics
paired stimulus task are significantly reduced in healthy Committee of the University of the Balearic Islands.
Somatosensory and auditory stimulation protocol.
individuals, reflecting some kind of physiologic habitua- The entire experiment consisted of 2 recording runs with a
tion to repetitive irrelevant stimuli (13,14). It has been 2-minute break between runs. The order of the runs was
argued that the first stimulus of the pair could activate counterbalanced between the subjects within each group.
some inhibitory brain pathways that suppress the re- Forty trials of either somatosensory or auditory stimulation
were presented in each run. A trial consisted of 2 identical
sponse to the second stimulus presented a short time stimuli (S1 and S2) delivered with a randomized interstimulus
later (e.g., 500 msec). Moreover, it has been observed interval of 550 msec (⫾50 msec), and separated by a fixed
that patients with some psychiatric diseases (e.g., schizo- intertrain interval of 12 seconds.
phrenia, bipolar depression, posttraumatic stress disor- Nonpainful somatosensory stimuli were applied using
der, cocaine abuse) and some pain conditions (migraine, a commercially available pneumatic stimulator (Biomagnetic
Technologies, San Diego, CA), which consisted of a small
premenstrual syndrome) show a reduced habituation of membrane attached to the body surface by a plastic clip and
early ERP responses as compared with healthy controls fixated with adhesive strips, as described previously (22). These
(16–18). Thus, the observed ERP attenuation in healthy stimuli were applied with a constant pressure of 2 bars and a
individuals may reflect some kind of brain mechanism by duration of 100 msec. Auditory stimuli were 2 tones (1,000 Hz,
92 dB peak-equivalent sound pressure level [SPL], 100 msec in
which incoming irrelevant information is “gated out,” duration), which were delivered binaurally through earphones.
and this mechanism might be altered under some psy- Both somatosensory and auditory stimuli were masked by
chopathologic conditions (16). white noise (87 dB SPL). At the end of the experiment, the
We examined the auditory and somatosensory subjects were asked whether they had experienced any pain or
discomfort from the stimulation techniques. None of the
gating mechanism in FM patients and healthy controls.
subjects reported either pain or discomfort.
We hypothesized that the two groups would differ in Recording of brain activity by EEG. The EEG was
their habituation of brain responses to repetitive sensory recorded from 32 electrodes placed in accordance with the
stimulation. Based on previous evidence about pain and International 10-20 System, and with reference electrodes at
central hyperexcitability in FM, it was expected that the the mastoid processes. Only data from 9 electrodes located
over the midline (Fz, Cz, and Pz), the left hemisphere (F3, C3,
FM patients might show a reduced brain response to the and P3), and the right hemisphere (F4, C4, and P4) were
second stimulus in the somatosensory, but not in the statistically analyzed for the purposes of the present study. An
auditory, paired stimulus paradigm. electrooculogram channel was obtained by placing 1 electrode
REDUCED BRAIN HABITUATION IN FIBROMYALGIA 1997
above and 1 electrode below the left eye. Ground was placed ponent), and mean amplitude during the period 160–360 msec
anteriorly to the location of the Fpz (midprefrontal) electrode. (late positive complex [LPC]). Similarly, amplitudes of the
Electrode impedance was measured to be ⬍5 k⍀. The signals auditory ERPs were obtained in the following time windows
were amplified by a BrainAmp amplifier (Brain Products, after stimulus onset: negative peak amplitude during the
Munich, Germany) at a sampling rate of 1,000 Hz, with period 125–175 msec (N100 component), and positive peak
high-pass and low-pass filter settings at 0.10 Hz and 70 Hz, amplitude during the period 175–275 msec (P200 component).
respectively. A 50-Hz notch filter was also applied. The EEG Statistical analysis. ERP data were statistically ana-
was segmented in epochs of 600-msec duration (–100 msec to lyzed according to a randomized factorial mixed design, using
500 msec relative to the stimulus onset). Averaging was
the between-subjects factor “group” (FM patients versus
performed separately for S1 and S2 stimuli. All averaged waves
healthy controls) and the within-subjects factors “stimulus” (S1
were digitally filtered (30-Hz low-pass filter) and baseline-
corrected before measures of component amplitude were versus S2) and “electrode location” (9 electrodes). The effects
computed. of these factors on somatosensory and auditory ERP ampli-
Data reduction. Amplitudes of the somatosensory tudes were separately examined for each time window using
ERPs elicited by the first and the second tactile stimuli were multivariate repeated-measures analysis of variance
computed in 3 time windows after initiation of the stimulus for (ANOVA). In addition, the relationship between sensory
each subject and each condition: positive peak amplitude gating and psychological variables, such as depression and
during the period 60–110 msec (P50 component), negative anxiety, was examined by computing Spearman’s correlations
peak amplitude during the period 110–160 msec (N100 com- across all subjects between the psychometric scores and the
1998 MONTOYA ET AL
Figure 1. Somatosensory brain activity in response to repetitive stimulation in patients with fibromyalgia
and healthy control subjects. The waveforms represent the somatosensory event-related potentials (ERPs)
at electrode Cz elicited by the first stimulus (black line) and the second stimulus (blue line) in the paired
stimulus task. Brain maps represent the scalp distribution of ERP amplitudes elicited by the first stimulus
(black border) and the second stimulus (blue border) at time latencies of 80 msec (P50) and 135 msec
(N100), and at the time range 160–360 msec, after initiation of the stimulus. Vertical broken line shows
the temporal occurrence of the stimulus (0 msec); blue shaded area shows the time window between 160
and 360 msec. AM ⫽ amplitude.
ratio of the amplitudes of S2 to S1 for each somatosensory and (N100), and sustained positivity starting at 160 msec
auditory ERP measure. (LPC).
RESULTS The amplitudes elicited by S1 and S2 stimuli for
the different ERP components averaged across all elec-
Clinical characteristics. Table 1 shows the socio-
demographic and clinical characteristics of the FM trodes are shown in Table 2. As predicted, the ERP
patients and healthy control subjects. The mean dura- amplitudes of these components were clearly reduced
tion of FM symptoms was 13.5 years (range 0.6–30 following presentation of S2, indicating a robust gating
years). FM patients reported higher levels of depression effect after stimulus repetition. Overall, these reductions
(t[28] ⫽ 6.64, P ⬍ 0.001) and anxiety (t[28] ⫽ 5.01, P ⬍ in amplitude in response to S2 as compared with S1 were
0.001) compared with healthy controls. confirmed by large and statistically significant main
Somatosensory gating effects. Figure 1 illustrates effects of “stimulus” for P50 (F[1,28] ⫽ 11.68, P ⬍ 0.01),
the mean somatosensory ERPs elicited by the first and N100 (F[1,28] ⫽ 33.84, P ⬍ 0.001), and LPC (F[1,28] ⫽
the second stimuli at electrode Cz in FM patients and 45.00, P ⬍ 0.001) amplitudes.
healthy controls. These ERP waveforms were character- Nevertheless, the presentation of tactile S2
ized by a positive deflection at ⬃80 msec after stimulus clearly resulted in differential amplitude reductions of
onset (P50), followed by a negative peak at 130 msec some somatosensory ERP components for FM patients
REDUCED BRAIN HABITUATION IN FIBROMYALGIA 1999
Table 2. Somatosensory and auditory ERP amplitudes elicited by the first and second stimuli in FM patients and
healthy controls*
FM patients (n ⫽ 15) Healthy controls (n ⫽ 15)
S1 S2 P S1 S2 P
Somatosensory ERP
P50 3.38 ⫾ 0.87 2.71 ⫾ 0.65 – 4.49 ⫾ 0.85 1.42 ⫾ 0.47 ⬍0.001
N100 ⫺5.91 ⫾ 1.23 0.10 ⫾ 0.74 ⬍0.001 ⫺5.05 ⫾ 1.41 ⫺0.65 ⫾ 0.66 ⬍0.01
LPC 5.04 ⫾ 0.59 3.73 ⫾ 0.54 – 8.46 ⫾ 0.97 2.39 ⫾ 0.75 ⬍0.001
Auditory ERP
N100 ⫺10.03 ⫾ 0.96 ⫺5.61 ⫾ 0.44 ⬍0.001 ⫺12.10 ⫾ 1.3 ⫺3.72 ⫾ 0.5 ⬍0.001
P200 6.64 ⫾ 1.18 3.70 ⫾ 0.63 ⬍0.05 9.84 ⫾ 1.23 4.41 ⫾ 0.62 ⬍0.001
* Amplitudes of the somatosensory and auditory event-related potentials (ERPs) elicited by the first and the second
stimuli (S1 and S2) were computed in 3 time windows after initiation of stimulus for each fibromyalgia (FM) patient,
each healthy control subject, and each condition (see Patients and Methods for details). Values are the mean ⫾ SEM
V for the positive peak amplitude (P50 and P200), the negative peak amplitude (N100), and the late positive complex
(LPC). P values indicate significant differences between ERP amplitudes elicited by the first and second stimuli.
and healthy controls. Thus, significant “stimulus” times to S1 (F[1,28] ⫽ 4.71, P ⬍ 0.05) than did healthy
“group” interactions were found for P50 (F[1,28] ⫽ 4.80, controls.
P ⬍ 0.05) and LPC (F[1,28] ⫽ 18.77, P ⬍ 0.001) Auditory gating effects. Figure 3 illustrates the
amplitudes, confirming that FM patients and healthy mean auditory ERPs elicited by the first and the second
controls differed in their brain processing of repetitive stimuli at electrode Cz in FM patients and healthy
information in the S1/S2 paradigm. Post hoc analyses controls. In this case, the ERP waveforms were charac-
further showed significant reductions in P50 (F[1,28] ⫽ terized by a negative peak at 100 msec (N100) after
15.73, P ⬍ 0.001) and LPC (F[1,28] ⫽ 60.94, P ⬍ 0.001) stimulus onset, and a positive deflection at ⬃200 msec
amplitudes from S1 to S2 in healthy controls, but not in (P200) after stimulus onset. ANOVA for the peak
FM patients (Figure 2). Post hoc analyses also revealed amplitudes yielded a significant main effect of “stimu-
that FM patients had lower LPC amplitudes in response lus” for N100 (F[1,28] ⫽ 53.72, P ⬍ 0.001) and P200
Figure 2. Reduced somatosensory gating in patients with fibromyalgia (FM) as compared with healthy
control subjects. Mean and SEM amplitudes in the P50 component and in the time range 160–360 msec
averaged across the electroencephalogram electrodes Fz, F3, F4, Cz, C3, C4, Pz, P3, and P4 are shown.
2000 MONTOYA ET AL
Figure 3. Auditory brain activity in response to repetitive stimulation in patients with fibromyalgia and
healthy control subjects. The waveforms represent the auditory event-related potentials (ERPs) at
electrode Cz elicited by the first stimulus (black line) and the second stimulus (blue line) in the paired
stimulus task. Brain maps represent the scalp distribution of ERP amplitudes elicited by the first stimulus
(black border) and the second stimulus (blue border) at time latencies of 155 msec (N100) and 235 msec
(P200) after initiation of the stimulus. Vertical broken line shows the temporal occurrence of the stimulus
(0 msec).
(F[1,28] ⫽ 27.57, P ⬍ 0.001) amplitudes, indicating that P ⬍ 0.05) and anxiety (rs ⫽ 0.51, P ⬍ 0.01) scores,
overall brain responses to auditory S2 were reduced in indicating that high levels of emotional distress were
comparison with S1. Although no main effects of associated with an increased S2 to S1 ratio (i.e., reduced
“group” were found for the N100 and P200 amplitudes, habituation). No significant correlations were obtained
a significant “stimulus” times “group” interaction effect for the other somatosensory or auditory ERP ampli-
was obtained for the auditory N100 amplitudes (F[1,28] tudes.
⫽ 5.15, P ⬍ 0.05), indicating that the stimulus repetition
resulted in differential effects on brain activity in FM DISCUSSION
patients and healthy controls. Moreover, post hoc ana- In the present study, we investigated the effects
lyses showed that auditory N100 amplitudes were signif- of repetitive non-nociceptive stimulation on habituation
icantly reduced from S1 to S2 in both groups (F[1,28] ⫽ of brain activity among patients with fibromyalgia and
12.81, P ⬍ 0.01 in FM patients and F[1,28] ⫽ 46.06, P ⬍ healthy controls by using an S1/S2 paired stimulus
0.001 in healthy controls) and that responses to S2 were paradigm and the recording of event-related potentials.
significantly lower in FM patients than in healthy con- Consistent with previous studies (2), we found that
trols (F[1,28] ⫽ 7.99, P ⬍ 0.01) (see also Table 2). amplitudes of early and late ERP components were
Correlations between sensory gating and emo- significantly attenuated when the same stimulus was
tional variables. Significant correlations were observed repeated within a short time interval (500 msec) in
for the somatosensory LPC amplitudes between the healthy controls. Moreover, it was observed that the
ratio of S2 to S1 amplitudes and depression (rs ⫽ 0.43, amplitude attenuation from S1 to S2 in the somatosen-
REDUCED BRAIN HABITUATION IN FIBROMYALGIA 2001
sory modality was significantly reduced in FM patients responses to the second tactile stimuli in FM patients
compared with healthy controls. In contrast, brain re- compared with healthy controls. Thus, the present study
sponses to auditory S2 were significantly attenuated shows that the gating of somatosensory responses is
relative to S1 in both groups. These findings suggest that abnormally reduced in FM patients, suggesting an im-
responses to somatosensory S2 appear to be inhibited, paired short-term habituation among these patients as
or “gated,” by the effects of S1 in healthy controls, but compared with healthy controls. These data further
not in FM patients. extend our previous findings (4,22,27) of an abnormal
The attenuation effect of the event-related brain brain processing of nonpainful somatosensory informa-
responses following stimulus repetition in healthy sub- tion, rather than a generalized information processing
jects is a well-known psychophysiologic phenomenon dysfunction, in patients with FM.
called “sensory gating” (13,14). Essentially, this habitu- Moreover, we hypothesized that FM patients
ation of the cortical response to S2 has been interpreted might show gating deficits at several stages of the
as reflecting an attentive capability of the brain to somatosensory information processing. Thus, somato-
recognize and to filter redundant and irrelevant incom- sensory P50 responses in this study showed a latency of
ing information. In this regard, it has been suggested 80 msec and represent the primary evoked cortical
that the first stimulus of a paired S1/S2 paradigm could response to somatic stimulation (28), whereas the so-
activate some inhibitory brain pathways that suppress matosensory N100 amplitudes are assumed to be gener-
the response to the second stimulus presented a short ated in the secondary somatosensory cortex and modu-
time later (e.g., 500 msec) (13). Indeed, animal research lated by attentional processes (29). In contrast, sustained
has demonstrated that the first stimulus in such a paired late positive amplitudes during the time range 160–360
stimulus paradigm activates excitatory pyramidal cells, msec have been associated with more complex cognitive
as well as inhibitory hippocampal interneurons, that functioning, such as memory or stimulus evaluation (30).
would suppress the activity in the pyramidal neurons Therefore, the present data might indicate that FM
elicited by subsequent presentations of a second identi- patients present habituation deficits that might be re-
cal stimulus (23). Neuropharmacologic research has lated to the coding, as well as to the cognitive evaluation,
revealed that cholinergic, dopaminergic, and noradren- of nonpainful somatosensory information.
ergic neurotransmitter systems are involved in the mod- In this regard, our findings add to a growing
ulation of brain responses to repetitive stimulation (16). literature in which FM patients have been shown to have
Clinical research has also found that patients some deficits of nociceptive information processing rel-
diagnosed as having schizophrenia showed little or no ative to healthy controls, such as enhanced sensitivity to
attenuation effect to the second stimulus, providing a repetitive pain pressure (4,31), abnormal maintenance
physiologic mechanism for the observed inability of of pain sensations after repetitive thermal stimulation
these patients to filter, or “gate,” thoughts and irrelevant (3), or deficits in endogenous pain inhibitory system
information from the environment (24). Two other (5–7). Moreover, it has been suggested that hyperalgesia
studies have shown that patients with migraine had a and allodynia in FM, as well as in other chronic pain
similar attenuation deficit of brain responses in an states, are the behavioral consequences of central sensi-
auditory paired stimulus task as compared with the tization. Thus, it would be possible that the observed
responses of healthy controls (17,18). Another recent disruption of the inhibitory brain mechanism involved in
study has shown that sensory gating was disrupted when the early processing of non-nociceptive repetitive stim-
healthy subjects were simultaneously involved in a com- ulation might be a further consequence of those neuro-
peting task that elicits emotional distress (25). Based on plastic changes due to central sensitization associated
available evidence, it has been argued that the sensory with chronic pain.
gating phenomenon might play a crucial role in the brain The hypothesis of a somatosensory gating deficit
processing of incoming information as indicators of in FM is also consistent with previous findings concern-
attentional and arousal deficits (26). ing brain activation elicited by pain stimuli in these
We found that FM patients and healthy controls patients. Thus, recent neuroimaging research has re-
differed in their habituation of brain responses to repet- vealed that FM patients compared with healthy controls
itive and irrelevant nonpainful somatosensory informa- show enhanced activity in areas of the brain that are
tion, but not in their responses to repetitive and irrele- involved in pain processing (somatosensory, prefrontal,
vant auditory information. This was reflected by a cingulate, and insular cortices), but decreased thalamic
markedly lower inhibition of the P50 and LPC cortical blood flow in response to painful (11), and nonpainful
2002 MONTOYA ET AL
(12) stimuli. Those investigators postulated that a lack of FM patients and healthy controls. Basically, a reduced
thalamic activity would result from persistent excitatory habituation to somatosensory stimuli was observed dur-
afferent pain signaling and would sharpen subjective ing the early and the late stages of information process-
pain sensation in patients with chronic pain, such as FM ing in FM patients as compared with healthy controls.
patients. Our findings further suggest that this abnormal We suggest that these findings might indicate a lack of
brain functioning in FM might also lead to an altered inhibitory control to repetitive nonpainful information
processing of nonpainful body information, which could during stimulus coding and cognitive evaluation in FM.
result in excessive pain sensitivity, or allodynia, in these In addition, our data provide further support for the
patients. hypothesis that FM might be characterized by specific
The present data also indicate some deficits in deficiencies in brain correlates of nonpainful somatosen-
the cognitive processing of repetitive nonpainful so- sory information processing.
matosensory information in FM. It has been suggested
that patients with subjective health complaints, such as
FM, might be characterized by a cognitive bias, which REFERENCES
would enhance the processing of information that is 1. Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C,
closely related to their complaints (32). Thus, it seems Goldenberg DL, et al. The American College of Rheumatology
1990 criteria for the classification of fibromyalgia: report of the
possible that the reduced habituation of the late ERP Multicenter Criteria Committee. Arthritis Rheum 1990;33:160–72.
components in FM could be related to an impaired 2. Arendt-Nielsen L, Graven-Nielsen T. Central sensitization in
ability to screen out irrelevant information arising from fibromyalgia and other musculoskeletal disorders. Curr Pain
Headache Rep 2003;7:355–61.
the body. Our finding that FM displayed a sensory gating 3. Price DD, Staud R. Neurobiology of fibromyalgia syndrome.
deficit in the late ERP amplitudes only for the somato- J Rheumatol Suppl 2005;75:22–8.
sensory, but not for the auditory, modality seems to lend 4. Montoya P, Pauli P, Batra A, Wiedemann G. Altered processing of
pain-related information in patients with fibromyalgia. Eur J Pain
further support to this view. Nonetheless, the fact that 2005;9:293–303.
FM patients had significantly more reduced LPC ampli- 5. Lautenbacher S, Rollman GB. Possible deficiencies of pain mod-
tudes in response to S1 than did healthy controls may ulation in fibromyalgia. Clin J Pain 1997;13:189–96.
6. Kosek E, Hansson P. Modulatory influence on somatosensory
also have contributed to the observed group differences perception from vibration and heterotopic noxious conditioning
in the attenuated response to S2. It could also be argued stimulation (HNCS) in fibromyalgia patients and healthy subjects.
that high levels of distress associated with pain in FM Pain 1997;70:41–51.
7. Julien N, Goffaux P, Arsenault P, Marchand S. Widespread pain in
could lead to a reduced gating effect in these patients. fibromyalgia is related to a deficit of endogenous pain inhibition.
The correlation data from the present study are partly Pain 2005;114:295–302.
consistent with this hypothesis. We observed that high 8. Desmeules JA, Cedraschi C, Rapiti E, Baumgartner E, Finckh A,
Cohen P, et al. Neurophysiologic evidence for a central sensitiza-
levels of depression or anxiety were associated with tion in patients with fibromyalgia. Arthritis Rheum 2003;48:
reduced habituation of somatosensory LPC amplitudes, 1420–9.
but not with habituation of other somatosensory or 9. Kwiatek R, Barnden L, Tedman R, Jarrett R, Chew J, Rowe C, et
al. Regional cerebral blood flow in fibromyalgia: single-
auditory brain measures. Thus, it also seems that emo- photon–emission computed tomography evidence of reduction in
tional factors might modulate the cognitive processing of the pontine tegmentum and thalami. Arthritis Rheum 2000;43:
repetitive nonpainful somatosensory stimulation in FM 2823–33.
10. Mountz JM, Bradley LA, Modell JG, Alexander RW, Triana-
patients (27). Such interpretations are speculative, how- Alexander M, Aaron LA, et al. Fibromyalgia in women: abnor-
ever, and await further empirical confirmation. malities of regional cerebral blood flow in the thalamus and the
One limitation of the present study is that ⬎50% caudate nucleus are associated with low pain threshold levels.
Arthritis Rheum 1995;38:926–38.
of our FM patients were taking antidepressants at the 11. Gracely RH, Petzke F, Wolf JM, Clauw DJ. Functional magnetic
time of the experiment. Given that these drugs affect resonance imaging evidence of augmented pain processing in
several neurotransmitter systems, the possibility that fibromyalgia. Arthritis Rheum 2002;46:1333–43.
12. Cook DB, Lange G, Ciccone DS, Liu WC, Steffener J, Natelson
effects of repetitive stimulation on brain activity may BH. Functional imaging of pain in patients with primary fibromy-
also be mediated by the medication cannot be com- algia. J Rheumatol 2004;31:364–78.
pletely ruled out. Additional research is needed to 13. Boutros NN, Belger A. Midlatency evoked potentials attenuation
and augmentation reflect different aspects of sensory gating. Biol
determine whether our findings may be generalized to Psychiatry 1999;45:917–22.
FM patients who are not taking antidepressant medica- 14. Johnson MR, Adler LE. Transient impairment in P50 auditory
tions. sensory gating induced by a cold-pressor test. Biol Psychiatry
1993;33:380–7.
In summary, we found significant differences in 15. Lebib R, Papo D, de Bode S, Baudonniere PM. Evidence of a
the cortical correlates of somatosensory gating between visual-to-auditory cross-modal sensory gating phenomenon as
REDUCED BRAIN HABITUATION IN FIBROMYALGIA 2003
reflected by the human P50 event-related brain potential modula- LE, et al. The genetics of sensory gating deficits in schizophrenia.
tion. Neurosci Lett 2003;341:185–8. Curr Psychiatry Rep 2003;5:155–61.
16. Leonard S, Adams C, Breese CR, Adler LE, Bickford P, Byerley 25. Yee CM, White PM. Experimental modification of P50 suppres-
W, et al. Nicotinic receptor function in schizophrenia. Schizophr sion. Psychophysiology 2001;38:531–9.
Bull 1996;22:431–45. 26. Cromwell HC, Anstrom K, Azarov A, Woodward DJ. Auditory
17. Siniatchkin M, Kropp P, Gerber WD. What kind of habituation is inhibitory gating in the amygdala: single-unit analysis in the
impaired in migraine patients? Cephalalgia 2003;23:511–8. behaving rat. Brain Res 2005;1043:12–23.
18. Ambrosini A, De Pasqua V, Afra J, Sandor PS, Schoenen J. 27. Montoya P, Sitges C, Garcia-Herrera M, Izquierdo R, Truyols M,
Reduced gating of middle-latency auditory evoked potentials Blay N, et al. Abnormal affective modulation of somatosensory
(P50) in migraine patients: another indication of abnormal sensory
brain processing among patients with fibromyalgia. Psychosom
processing? Neurosci Lett 2001;306:132-4.
Med 2005;67:957–63.
19. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An
inventory for measuring depression. Arch Gen Psychiatry 1961;4: 28. Hari R, Reinikainen K, Kaukoranta E, Hamalainen M, Ilmoniemi
561–71. R, Penttinen A, et al. Somatosensory evoked cerebral magnetic
20. Spielberger CD, Gorsuch RL, Lushene RE. The State-Trait fields from SI and SII in man. Electroencephalogr Clin Neuro-
Anxiety Inventory (STAI): test manual. Palo Alto (CA): Consult- physiol 1984;57:254–63.
ing Psychologists Press; 1970. 29. Eimer M, Forster B. Modulations of early somatosensory ERP
21. Kerns RD, Turk DC, Rudy TE. The West Haven-Yale Multidi- components by transient and sustained spatial attention. Exp
mensional Pain Inventory (WHYMPI). Pain 1985;23:345–56. Brain Res 2003;151:24–31.
22. Montoya P, Larbig W, Braun C, Preissl H, Birbaumer N. Influence 30. Polich J, Herbst KL. P300 as a clinical assay: rationale, evaluation,
of social support and emotional context on pain processing and and findings. Int J Psychophysiol 2000;38:3–19.
magnetic brain responses in fibromyalgia. Arthritis Rheum 2004; 31. Lautenbacher S, Rollman GB, McCain GA. Multi-method assess-
50:4035–44. ment of experimental and clinical pain in patients with fibromyal-
23. Miller CL, Freedman R. The activity of hippocampal interneurons gia. Pain 1994;59:45–53.
and pyramidal cells during the response of the hippocampus to 32. Eriksen HR, Ursin H. Subjective health complaints, sensitization,
repeated auditory stimuli. Neuroscience 1995;69:371–81. and sustained cognitive activation (stress). J Psychosom Res
24. Freedman R, Olincy A, Ross RG, Waldo MC, Stevens KE, Adler 2004;56:445–8.
Journal of Affective Disorders 104 (2007) 73 – 82
www.elsevier.com/locate/jad
Research report
Abnormal brain processing of affective and sensory pain
descriptors in chronic pain patients ☆
Carolina Sitges a,b , Manuel García-Herrera c , Miquel Pericás d , Dolores Collado c ,
Magdalena Truyols e , Pedro Montoya a,b,⁎
a
Department of Psychology, University of Balearic Islands, Spain
b
Research Institute on Health Sciences (IUNICS), Palma, Spain
c
Medical Unit for Disability Assessment, Social Security Agency, Palma, Spain
d
Medical College of the Balearic Islands, Palma, Spain
e
Pain Clinic, General Hospital, Palma, Spain
Received 26 September 2006; received in revised form 21 February 2007; accepted 23 February 2007
Available online 16 April 2007
Abstract
Objective: Previous research has suggested that chronic pain patients might be particularly vulnerable to the effects of negative mood
during information processing. However, there is little evidence for abnormal brain processing of affective and sensory pain-related
information in chronic pain. Behavioral and brain responses, to pain descriptors and pleasant words, were examined in chronic pain
patients and healthy controls during a self-endorsement task.
Methods: Eighteen patients with fibromyalgia (FM), 18 patients with chronic musculoskeletal pain due to identifiable physical injury
(MSK), and 16 healthy controls were asked to decide whether word targets described their current or past experience of pain. The number
of self-endorsed words, elapsed time to endorse the words, and event-related potentials (ERPs) elicited by words, were recorded.
Results: Data revealed that chronic pain patients used more affective and sensory pain descriptors, and were slower in responding
to self-endorsed pain descriptors than to pleasant words. In addition, it was found that affective pain descriptors elicited
significantly more enhanced positive ERP amplitudes than pleasant words in MSK pain patients; whereas sensory pain descriptors
elicited greater positive ERP amplitudes than affective pain words in healthy controls.
Conclusions: These data support the notion of abnormal information processing in chronic pain patients, which might be
characterized by a lack of dissociation between sensory and affective components of pain-related information, and by an
exaggerated rumination over word meaning during the encoding of self-referent information about pain.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Chronic pain; Fibromyalgia; Musculoskeletal pain; Emotion; Pain descriptors; Evoked potentials
☆
Contributors: Authors Pedro Montoya and Carolina Sitges designed the study and wrote the protocol. Authors Miquel Pericás, Dolores Collado
and Magdalena Truyols managed the literature searches and analyses. Authors Manuel García-Herrera and Pedro Montoya undertook the statistical
analysis, and author Carolina Sitges wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.
⁎ Corresponding author. Beatriu de Pinós Building, Department of Psychology, Cra. de Valldemossa km 7.5, 07122 Palma, Spain. Tel.: +34
971172646; fax: +34 971172309.
E-mail address: [email protected] (P. Montoya).
0165-0327/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jad.2007.02.024
74 C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82
affective pain-related words compared to patients with Questionnaire (FIQ) (Burckhardt et al., 1991), the Pain
pain attributable to a known biomedical cause. Anxiety Symptoms Scale (PASS) (McCracken et al.,
1992), and the Pain Vigilance and Awareness Question-
2. Methods naire (PVAQ) (McCracken, 1997).
Thirty-six patients suffering from chronic musculo- Stimuli consisted of 20 affective (e.g., fearful, nau-
skeletal pain (26–63 years old, 32 females), and 16 seating), and 20 sensory pain descriptors (e.g., sharp,
healthy volunteers (39–66 years old, 15 females) par- hot-burning), and 30 pleasant, non-pain-related words
ticipated in the study. All patients had pain for at least (e.g., beautiful, friendly) each presented during 300 ms
6 months as the dominant symptom complaint, and were on a 17″ computer screen (1 m viewing distance
recruited from the Pain Management Unit of a tertiary care approximately). Words were selected from the McGill
hospital and the Medical Unit for Disability Assessment Pain Questionnaire (Melzack, 1975; Lázaro et al.,
of the Spanish Social Security Agency in Palma (Spain). 1994), and from those used in previous research (Larbig
Healthy controls were recruited among the administrative et al., 1996; Flor et al., 1997). Pain patients were
staff of the University of the Balearic Islands. instructed to assess whether the presented word was
Pain patients were classified into two groups related or not to their current experience of pain by
according to the primary diagnosis recorded in their pressing a button (YES versus NO decision). Healthy
medical chart: patients with fibromyalgia (FM, n = 18), controls were asked to remember a past experience of
and patients with chronic pain due to identifiable persistent pain (e.g., headache, toothache). The order of
physical injuries (rheumatoid arthritis, radiculopathy, word type was randomised. Each word was presented
herniated disk) (MSK, n = 18). All patients underwent twice, resulting in 140 trials. The inter-trial interval
an extensive medical assessment by an experienced between button response and stimulus onset was
rheumatologist (MP) to confirm the fulfillment of ACR randomised between 1800 and 2000 ms.
classification criteria for FM (Wolfe et al., 1990) with a
minimum of 11 out of 18 tender-points. Participants 2.4. EEG recording
were excluded from the study if they were pregnant, had
neurological disease, or were taking opioids. Brain electrical activity was recorded with an
Patients were allowed to continue taking long-term electrode cap from 32 sites placed according to the
medications. At the time of the study, 11 patients (4 FM international 10–20 system. Electrodes were also placed
and 7 MSK) were involved in a litigation process for on mastoids as reference, and above the left eye and
financial compensation due to disability. Patients were below the right eye to record eye movement. Ground
verbally informed about the details of the study and was placed anteriorly to the location of the FCz
explicitly notified that the participation in the study was electrode. All impedances were kept below 10 kΩ.
not linked to their possible litigation process. A specif- The signals were registered by a BrainAmp MR
ically designed information leaflet was also given to all amplifier at a sampling rate of 1000 Hz, with high and
subjects, and after their agreement to participate, a low pass filter settings at 0.10 Hz and 70 Hz,
written consent was obtained. The study was in accor- respectively. A 50 Hz notch filter was also applied.
dance with the Declaration of Helsinki (1991) and was
approved by the Ethics Committee of the University of 2.5. Procedure
the Balearic Islands.
Anxiety and depression questionnaires were filled
2.2. Self-report measures out after written consent was obtained. The other
psychological questionnaires were completed in a
All participants completed the Spielberger State previous session up to 1 month before the EEG session.
Anxiety Inventory (STAI-S) (Spielberger et al., 1970) After EEG electrodes were attached and the subject was
and the Beck Depression Inventory (BDI) (Beck et al., instructed about the task, four practice trials with
1961). In addition, pain patients completed the McGill irrelevant words were presented. Then, 140 trials (40
Pain Questionnaire (MPQ) (Melzack, 1975), the affective, 40 sensory and 60 pleasant words) were
West Haven–Yale Multidimensional Pain Inventory presented in a random sequence on the computer screen.
(WHYMPI) (Kerns et al., 1985), the Fibromyalgia Impact Reaction times (RTs) and visual event-related potentials
76 C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82
Table 1
Demographic and psychological data for chronic pain patients and healthy controls
FM (n = 18) MSK (n = 18) Healthy (n = 16) Significance level
Age (years)
Mean (SD) 49.44 (6.54) 46.39 (9.24) 49.20 (8.60) .563
Range 37–61 26–63 39–66
Gender, n (%) .064
Male 0 (0%) 4 (22.2%) 1 (6.3%)
Female 18 (100%) 14 (77.8%) 15 (93.8%)
Education, n (%) .184
b8 years 8 (44.4%) 2 (11.1%) 4 (25.0%)
8–12 years 8 (44.4%) 10 (55.6%) 7 (43.8%)
N12 years 2 (11.1%) 6 (33.3%) 5 (31.3%)
Litigation process, n (%) 4 (22.2%) 7 (38.9%) – .291
Medication, n (%)
Antidepressants 9 (60.0%) 11 (73.3%) – .456
Analgesics/muscular relaxants/NSAIDs 11 (73.3%) 12 (70.6%) – .869
Anxiolytics 10 (66.7%) 9 (52.9%) – .447
Pain intensity (0–10)
Mean (SD) 6.02 (2.09) 6.25 (1.81) – .746
Range 2.0–9.3 3.5–10.0 –
Pain duration (years)
Mean (SD) 11.73 (12.54) 6.37 (7.29) – .144
Range 2.0–40.0 1.3–30.0 –
STAI (state)
Mean (SD) 31.39 (10.80) 37.00 (12.94) 15.07 (9.96) .001
Range 15–50 13–57 2–45
BDI
Mean (SD) 25.88 (10.10) 22.75 (11.64) 6.07 (4.99) .001
Range 8–44 7–45 0–20
MPQ
Sensory 9.93 (1.03) 8.24 (2.44) – .018
Affective 2.53 (0.64) 2.00 (0.61) – .022
Miscellaneous 2.33 (0.98) 2.06 (1.06) – .467
WHYMPI (0–6) (mean, SD)
Social support 3.75 (1.52) 4.63 (1.56) – .116
Affective distress 3.95 (1.10) 4.00 (1.02) – .894
Interference social activities 3.89 (1.70) 4.14 (1.01) – .621
Interference daily activities 5.22 (0.81) 5.00 (0.77) – .445
Pain intensity 4.86 (0.68) 4.71 (1.00) – .635
Life control 3.25 (1.68) 3.31 (1.44) – .916
Distracting responses 3.74 (2.03) 4.05 (1.60) – .637
Solicitous responses 2.01 (1.60) 2.56 (1.83) – .378
Punishing responses 0.62 (1.37) 1.62 (1.61) – .074
Household chores 3.04 (1.61) 3.02 (1.71) – .980
Activities away from home 1.42 (1.17) 2.23 (1.22) – .066
Outdoor work 0.77 (1.57) 1.47 (1.34) – .182
Social activities 1.69 (1.85) 1.96 (1.09) – .610
FIQ (0–10)
Physical impairment 1.82 (0.55) 1.78 (0.47) – .801
Feel good 6.20 (1.15) 6.25 (1.48) – .917
Work missed 3.64 (2.27) 3.00 (2.66) – .498
Do job 7.93 (2.17) 7.60 (1.71) – .638
Pain 7.81 (2.04) 7.39 (1.41) – .508
Fatigue 7.36 (2.83) 8.01 (1.92) – .447
Rested 8.15 (2.70) 7.32 (2.50) – .384
Stiffness 8.28 (2.15) 5.96 (3.67) – .041
Anxiety 6.79 (2.93) 5.28 (3.58) – .203
Depression 6.26 (2.83) 6.05 (3.16) – .843
C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82 77
Table 1 (continued )
FM (n = 18) MSK (n = 18) Healthy (n = 16) Significance level
PASS (0–5)
Cognitive 2.76 (1.35) 2.86 (1.34) – .843
Physiologic 2.24 (1.15) 2.07 (1.10) – .689
Escape/avoidance 2.46 (0.79) 2.69 (1.13) – .524
Phobic appraisal 2.19 (1.11) 2.49 (1.05) – .446
PVAQ 51.34 (15.56) 51.76 (15.96) – .941
(ERPs) elicited by each word stimuli were recorded. and healthy controls. One-way analyses of variance
After the experiment, patients were debriefed about the (ANOVAs) revealed significant differences among the
purpose of the study. three groups on depression (F[2,47] = 19.82, P b.001)
and anxiety scores (F[2,48] = 15.08, P b.001). Tukey
2.6. Data reduction and analysis post-hoc comparisons also indicated that both FM and
MSK patients were more depressed (P b.001) and
The percentage of self-endorsed words and the anxious (P b.001) than healthy controls.
average reaction time (RT) for each Word (sensory pain- Further differences between patients with fibromy-
related words, affective pain-related words, and pleasant algia (FM) pain and patients with musculoskeletal pain
words) were obtained for each subject. Data was analyzed due to identifiable physical injuries (MSK) on self-
using a 3 (Word) × 2 (Response) × 3 (Group) mixed-model report measures of pain were assessed in additional
repeated-measures analysis of variance (ANOVA). statistical analyses. Student's t-test revealed that FM
EEG was segmented in epochs of 1000 ms duration patients used significantly more sensory (t[30] = 6.23,
(from 100 ms before the onset of stimulus to 900 ms P b.05) and affective adjectives from the McGill Pain
after). Averaging was performed separately for each Questionnaire (MPQ) (t[30] = 2.41, P b.05), and that
word type. All average waves were digitally filtered they had higher scores in the Stiffness FIQ scale (t[30] =
(30 Hz high cutoff) and baseline corrected before statis- 2.14, P b.05) than the MSK patients.
tical measures of component amplitude were computed.
Ocular movements were corrected using a standardized 3.2. Number of self-endorsed words
regression method (Gratton et al., 1983). Trials with
amplitudes outside the range of ± 50 μV were auto- Fig. 1 displays the average percentage of words self-
matically excluded. The ERP amplitudes for the P200 endorsed (YES response) or not (NO response) as
component (defined as the maximum baseline-to-peak descriptive for the experience of pain. More than 70% of
amplitude, in the time-window 200–350 ms, after the the sensory and affective pain words were considered
onset of stimulus at electrode O2), and the mean ampli- descriptive for the experience of pain in FM and MSK
tude in the time-window 500–800 ms after stimulus patients, whereas only 36–38% of those words were con-
onset were calculated for each electrode position. sidered as pain descriptors in healthy controls. Significant
Twelve electrodes located at the right (F4, C4, CP4,
TP8, P4, O2) and the left hemisphere (F3, C3, CP3, TP7,
P3, O1) were used for statistical analyses. For each
component, ERP amplitudes were examined in a Brain
region (frontal, central, centro-parietal, temporo-parietal,
parietal and occipital) × Hemisphere (left versus right) ×
Word (sensory, unpleasant, pleasant words) × Group
(FM, MSK, healthy controls) mixed-model repeated-
measures ANOVA.
3. Results
Table 1 displays clinical and sociodemographic Fig. 1. Percentage of self-endorsed words for healthy controls and
characteristics of chronic pain patients (FM and MSK) chronic pain patients.
78 C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82
words (P b.05) only for healthy controls (Fig. 4). response to pleasant words at parietal electrode locations
Moreover, topographical analyses revealed significant than FM and MSK patients (all Ps b .05).
differences between sensory and affective pain descrip-
tors at frontal, central, and centro-parietal electrode lo- 4. Discussion
cations (all Ps b .05) in this group. Furthermore, healthy
controls had greater P200 amplitudes at parietal The aim of the present study was to investigate brain
electrodes than MSK pain patients (P b.05) in response correlates of information processing in chronic pain
to sensory pain-related words. patients. For this purpose, we asked pain patients and
For the LPC amplitudes in the time 500–800 ms, healthy controls to decide whether some words were
significant effects of Word × Group (F[4,94] = 3.04, appropriate or not for describing their experience of pain
P b.05) (Fig. 4) and Brain region (F[5,235] = 4.80, (self-endorsement task). We expected significant group
P b.01) were found. Post-hoc comparisons revealed that differences in the processing of sensory and affective
LPC amplitudes elicited by affective pain descriptors pain descriptors versus pleasant words, indicating a
were significantly greater than amplitudes elicited by more elaborated description of the experience of pain in
pleasant words (P b.05) for MSK pain patients. Moreover, patients than in healthy controls. The observation was
LPC amplitudes elicited by sensory pain descriptors were that chronic pain patients chose more sensory and
overall more increased than those elicited by affective affective pain words, and reacted slower during
pain descriptors (P b.05) for healthy controls. In addition, responses to pain descriptors than healthy controls. In
topographical analyses further showed that healthy addition, with MSK pain patients, the electrophysiolog-
controls had greater LPC amplitudes, in response to ical data showed that affective pain descriptors elicited
sensory pain descriptors, at central and centro-parietal greater LPC amplitudes, than pleasant words. Moreover,
electrode locations than FM patients (all Ps b .05). Finally, chronic pain patients displayed more reduced ERP
healthy controls also had greater LPC amplitudes in amplitudes than healthy controls, in response to sensory
pain descriptors and pleasant words.
Our behavioral findings are in line with previous
studies showing; a) that chronic pain patients self-
endorsed more negative illness words and fewer pleas-
ant words than did healthy controls (Read and Pincus,
2004; Denton et al., 2005); and b) that they responded to
target stimuli more slowly than healthy controls when
affective and sensory pain descriptors are used as cues
(Pearce and Morley, 1989; Snider et al., 2000; Kuhajda
et al., 2002). In the current study, chronic pain patients
exhibited slower response times to sensory and affective
descriptors than to pleasant words, particularly during
NO responses. In contrast, no differential effect due to
word category was observed in healthy controls. These
findings suggest the presence of an information pro-
cessing bias in chronic pain patients, characterized by an
exaggerated rumination over word meaning, during the
encoding of self-referent information about pain.
Nevertheless, our data do not provide support for the
notion of a selective attentional bias towards affective
pain descriptors as postulated by cognitive theories
(Pincus and Morley, 2001). The sensory and affective
pain descriptors used here were similar to those used in
Fig. 4. Mean amplitudes of the P200 peak component and late positive cognitive bias experiments, and participants were
complex (LPC) (time-window between 500 and 800 ms after stimulus required to respond to target stimuli. However, the dif-
onset) elicited by affective, and sensory pain descriptors, as well as ferences between our self-endorsement task and other
positive words for chronic pain patients, and healthy controls. ERP
amplitudes were averaged across EEG electrodes located at several
experimental tasks used in cognitive bias experiments
brain regions (frontal, central, centro-parietal, temporo-parietal, might be explained by methodological differences.
parietal and occipital). These concern the goal-directedness and the level of
80 C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82
processing necessary to react to the targets. The purpose hyper-vigilance) might worsen pain in chronic pain
of the present self-endorsement task is explicit and patients (Vlaeyen and Linton, 2000). The present study
participants are instructed to explicitly encode the is, therefore, in agreement with previous research
content of the words, whereas the purpose of the dot- supporting a dissociation of brain mechanisms involved
probe task is implicit and participants are instructed to in the processing of sensory and affective dimensions of
process irrelevant information (for instance, to press a pain (Rainville et al., 1997; Hofbauer et al., 2001).
key as fast as possible when a letter appears) ignoring Nevertheless, we also observed that sensory and
the content of the pain words. These two paradigms led affective pain words did not differentially influence
to different information processing strategies being ERP amplitudes in FM or MSK pain patients. Moreover,
followed in chronic pain patients. sensory pain words elicited greater P200 and LPC
Event-related potential (ERP) data showed that with amplitudes in healthy controls than in chronic pain
MSK pain patients, affective pain descriptors elicited patients. Thus, it may be argued that chronic pain patients
more enhanced amplitudes of the late positive complex could be characterized as having reduced engagement of
(LPC) than did pleasant words. In contrast, in healthy attentional and motivational resources to the processing of
controls, sensory pain descriptors elicited more positive pain descriptors. Our findings also seem to be in
P200 and LPC amplitudes than did affective pain agreement with previous data examining brain activity
descriptors. An enhancement of positive ERP amplitudes elicited by unpleasant pain-related and neutral words in
elicited by emotional stimuli has been consistently FM (Montoya et al., 2005a), chronic back pain (Flor et al.,
observed in brain research about affective processing 1997), and amputees with chronic phantom limb pain
(Montoya et al., 1996; Pauli et al., 1997; Cuthbert et al., (Larbig et al., 1996). In those studies, pain-related and
2000; Keil et al., 2002). Emotional and neutral word neutral words were presented either at individual
stimuli appear to activate different brain networks at an perception thresholds or embedded in a lexical decision
early processing stage (100–140 ms after stimulus onset) task, and they elicited similar ERP amplitudes in the time
(Ortigue et al., 2004). Moreover, a long-lasting positive between 400 and 600 ms after onset of stimulus in pain
shift beginning around 300 ms after stimulus has been patients. The fact that both implicit and explicit
observed following highly arousing emotional stimuli experimental tasks yielded similar findings suggests that
(e.g., pictures with erotic or violent content) as compared this pattern of brain activity might be characteristic of
to pleasant ones. Thus, it has been suggested that early information processing among chronic pain patients.
enhanced ERP amplitudes (100–250 ms after stimulus Thus, all these findings provide support for the hypothesis
onset) in response to words might be linked to the of a lack of affective brain modulation in chronic pain
allocation of attentional resources during language patients, which could be interpreted as a difficulty to
processing (Pulvermuller, 2001). However, an enhance- disengage from pain-related stimuli (Asmundson et al.,
ment of LPC amplitudes might reflect the engagement of 2005). In a similar way, previous brain research has
the motivational system for a more complete processing of shown that patients with affective disorders, such as panic
salient stimuli (motivational priming hypothesis) (Lang disorder (Windmann et al., 2002) or post-traumatic stress
et al., 1998). It follows that the observed ERP amplitude disorder (Felmingham et al., 2003), failed to differentiate
enhancement in MSK pain patients and in healthy controls between emotional stimuli. Taken together, these findings
might be reflecting a differential engagement of atten- support the notion of a dysfunctional inhibitory modula-
tional and motivational brain systems for the processing of tion of affective information processing (Windmann et al.,
pain descriptors. Sensory pain descriptors appeared to be 2002), which could lead to behavioral difficulties for
motivationally relevant inputs for healthy controls, disengaging attention from pain-related and threat stimuli
whereas affective pain descriptors were apparently highly in those patients.
arousing stimuli for some chronic pain patients. This There are some shortcomings of the present study that
interpretation seems to be in agreement with previous must be borne in mind in interpreting these results. The
research showing that healthy controls increase attention samples were relatively small and although they were
towards sensory pain descriptors when expectancies of large enough to detect large effect sizes, smaller effect
threat are low (Boston and Sharpe, 2005). Focusing sizes may have been obscured, leading to Type II errors.
attention on sensory components of pain might reduce the Nonetheless, the sample sizes here are comparable to
experience of acute pain, while focusing on affective others reported in the literature (range between 12 and 40
components might increase pain perception (Ahles et al., patients). No measure was made of illness or pain beliefs
1983; Dar and Leventhal, 1993). In contrast, it has been in chronic pain patients. Therefore, although the hypoth-
suggested that attention to sensory pain sensations (e.g., esis that several schemas would modulate information
C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82 81
processing biases in chronic pain patients is plausible, it Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., Erbaugh, J., 1961.
remains to be explored. An inventory for measuring depression. Arch. Gen. Psychiat. 56,
561–571.
In summary, the present study revealed differences in Boston, A., Sharpe, L., 2005. The role of threat-expectancy in acute
brain processing of pain-related information between pain: effects on attentional bias, coping strategy effectiveness and
chronic pain patients and healthy controls. These response to pain. Pain 119, 168–175.
differences were the number of descriptors chosen for Burckhardt, C.S., Clark, S.R., Bennett, R.M., 1991. The fibromyalgia
impact questionnaire: development and validation. J. Rheumatol.
describing the pain experience, the response time to
18, 728–733.
endorse the descriptors, and the pattern of brain activity Cuthbert, B.N., Schupp, H.T., Bradley, M.M., Birbaumer, N., Lang, P.J.,
elicited by pain descriptors. In our opinion, these 2000. Brain potentials in affective picture processing: covariation
findings suggest that pain patients might have some with autonomic arousal and affective report. Biol. Psychol. 52,
difficulties in disengaging attention from pain-related 95–111.
stimuli during the encoding of this information, probably Dar, R., Leventhal, H., 1993. Schematic processes in pain perception.
Cogn. Ther. Res. 17, 341–357.
due to some dysfunctional inhibitory modulation of Dehghani, M., Sharpe, L., Nicholas, M.K., 2003. Selective attention to
affective information processing (Montoya et al., 2006). pain-related information in chronic musculoskeletal pain patients.
Interestingly, we also observed that FM and MSK Pain 105, 37–46.
patients did not differ either in the processing of self- Denton, F.J., Sharpe, L., Schrieber, L., 2005. Cognitive bias in
endorsed pain descriptors or in the level of pain intensity, systemic lupus erythematosus. Eur. J. Pain 9, 5–14.
Felmingham, K.L., Bryant, R.A., Gordon, E., 2003. Processing angry
depression, anxiety, and other psychological variables. and neutral faces in post-traumatic stress disorder: an event-related
However, MSK patients exhibited an enhancement of potentials study. Neuroreport 14, 777–780.
brain activity in response to affective compared to Flor, H., Knost, B., Birbaumer, N., 1997. Processing of pain- and
sensory pain descriptors, whereas FM patients failed to body-related verbal material in chronic pain patients: central and
show any differential effect of word category on ERP peripheral correlates. Pain 73, 413–421.
Gratton, G., Coles, M.G., Donchin, E., 1983. A new method for off-
amplitudes. Thus, it seems that factors other than the line removal of ocular artifact. Electroencephalogr. Clin. Neuro-
presence of chronic pain symptoms might be influencing physiol. 55, 468–484.
the cognitive processing of affective and sensory stimuli Hofbauer, R.K., Rainville, P., Duncan, G.H., Bushnell, M.C., 2001.
in these patients. Future research should be addressed to Cortical representation of the sensory dimension of pain.
examine the generalization of these results in a more J. Neurophysiol. 86, 402–411.
Keil, A., Bradley, M.M., Hauk, O., Rockstroh, B., Elbert, T., Lang, P.J.,
diverse sample of chronic pain patients, and to other 2002. Large-scale neural correlates of affective picture processing.
types of sensory and affective stimuli. Psychophysiology 39, 641–649.
Kerns, R.D., Turk, D.C., Rudy, T.E., 1985. The West Haven–Yale
Acknowledgements Multidimensional Pain Inventory (WHYMPI). Pain 23, 345–356.
Kuhajda, M.C., Thorn, B.E., Klinger, M.R., Rubin, N.J., 2002. The
effect of headache pain on attention (encoding) and memory
This study was supported by grants from the Spanish (recognition). Pain 97, 213–221.
Ministerio de Educación y Ciencia (Plan Nacional de Lang, P.J., Bradley, M.M., Fitzsimmons, J.R., Cuthbert, B.N., Scott, J.D.,
I+D+I) and European Funds (FEDER) (ref: BSO2001- Moulder, B., Nangia, V., 1998. Emotional arousal and activation of
0693 and SEJ2004-01332). the visual cortex: an fMRI analysis. Psychophysiology 35, 199–210.
Larbig, W., Montoya, P., Flor, H., Bilow, H., Weller, S., Birbaumer, N.,
1996. Evidence for a change in neural processing in phantom limb
References patients. Pain 67, 275–283.
Lázaro, C., Bosch, F., Torrubia, R., 1994. The development of a Spanish
Ahles, T.A., Blanchard, E.B., Leventhal, H., 1983. Cognitive control Questionnaire for assessing pain: preliminary data concerning
of pain: attention to the sensory aspects of the cold-pressor reliability and validity. Eur. J. Psychol. Assess. 10, 141–151.
stimulus. Cogn. Ther. Res. 7, 159–178. McCracken, L.M., 1997. ‘Attention’ to pain in persons with chronic
Apkarian, A.V., Sosa, Y., Krauss, B.R., Thomas, P.S., Fredrickson, B.E., pain: a behavioural approach. Behav. Ther. 28, 271–284.
Levy, R.E., Harden, R.N., Chialvo, D.R., 2004. Chronic pain patients McCracken, L.M., Zayfert, C., Gross, R.T., 1992. The Pain Anxiety
are impaired on an emotional decision-making task. Pain 108, Symptoms Scale: development and validation of a scale to measure
129–136. fear of pain. Pain 50, 67–73.
Asmundson, G.J., Kuperos, J.L., Norton, G.R., 1997. Do patients Melzack, R., 1975. The McGill Pain Questionnaire: major properties
with chronic pain selectively attend to pain-related informa- and scoring methods. Pain 1, 277–299.
tion?: preliminary evidence for the mediating role of fear. Pain Mitchell, P., Andrews, S., Ward, P., 1993. An event-related potential
72, 27–32. study of semantic congruity and repetition in a sentence-reading
Asmundson, G.J., Wright, K.D., Hadjistavropoulos, H.D., 2005. task: effects of context change. Psychophysiology 30, 496–509.
Hypervigilance and attentional fixedness in chronic musculoskel- Montoya, P., Larbig, W., Pulvermuller, F., Flor, H., Birbaumer, N.,
etal pain: consistency of findings across modified Stroop and dot- 1996. Cortical correlates of semantic classical conditioning.
probe tasks. J. Pain 6, 497–506. Psychophysiology 33, 644–649.
82 C. Sitges et al. / Journal of Affective Disorders 104 (2007) 73–82
Montoya, P., Pauli, P., Batra, A., Wiedemann, G., 2005a. Altered Read, J., Pincus, T., 2004. Cognitive bias in back pain patients
processing of pain-related information in patients with fibromy- attending osteopathy: testing the enmeshment model in reference
algia. Eur. J. Pain 9, 293–303. to future thinking. Eur. J. Pain 8, 525–531.
Montoya, P., Sitges, C., Garcia-Herrera, M., Izquierdo, R., Truyols, Roelofs, J., Peters, M.L., Zeegers, M.P.A., Vlaeyen, J.W.S., 2002. The
M., Blay, N., Collado, D., 2005b. Abnormal affective modulation modified Stroop paradigm as a measure of selective attention
of somatosensory brain processing among patients with fibromy- towards pain-related stimuli among chronic pain patients: a meta-
algia. Psychosom. Med. 67, 957–963. analysis. Eur. J. Pain 6, 273–281.
Montoya, P., Sitges, C., García-Herrera, M., Rodríguez-Cotes, A., Roelofs, J., Peters, M.L., Fassaert, T., Vlaeyen, J.W.S., 2005. The role
Izquierdo, R., Truyols, M., Collado, D., 2006. Reduced brain of fear of movement and injury in selective attentional processing
habituation to somatosensory stimulation in patients with fibro- in patients with chronic low back pain: a dot-probe evaluation.
myalgia. Arthritis Rheum. 54, 1995–2003. J. Pain 6, 294–300.
Ortigue, S., Michel, C.M., Murray, M.M., Mohr, C., Carbonnel, S., Snider, B.S., Asmundson, G.J., Wiese, K.C., 2000. Automatic and
Landis, T., 2004. Electrical neuroimaging reveals early generator strategic processing of threat cues in patients with chronic pain: a
modulation to emotional words. Neuroimage 21, 1242–1251. modified Stroop evaluation. Clin. J. Pain 16, 144–154.
Pauli, P., Dengler, W., Wiedemann, G., Montoya, P., Flor, H., Spielberger, C.D., Gorsuch, R.L., Lushene, R.E., 1970. The State–
Birbaumer, N., Buchkremer, G., 1997. Behavioral and neurophys- Trait Anxiety Inventory (STAI): Test Manual. Consulting
iological evidence for altered processing of anxiety-related words Psychologists Press, Palo Alto, CA.
in panic disorder. J. Abnorm. Psychol. 106, 213–220. Staud, R., Robinson, M.E., Vierck Jr., C.J., Cannon, R.C., Mauderli, A.P.,
Pearce, J., Morley, S., 1989. An experimental investigation of the Price, D.D., 2003. Ratings of experimental pain and pain-related
construct validity of the McGill Pain Questionnaire. Pain 39, negative affect predict clinical pain in patients with fibromyalgia
115–121. syndrome. Pain 105, 215–222.
Pearce, S.A., Isherwood, S., Hrouda, D., Richardson, P.H., Erskine, A., Vlaeyen, J.W.S., Linton, S.J., 2000. Fear-avoidance and its con-
Skinner, J., 1990. Memory and pain: tests of mood congruity and sequences in chronic musculoskeletal pain: a state of the art. Pain
state dependent learning in experimentally induced and clinical 85, 317–332.
pain. Pain 43, 187–193. Windmann, S., Sakhavat, Z., Kutas, M., 2002. Electrophysiological
Pincus, T., Morley, S., 2001. Cognitive-processing bias in chronic evidence reveals affective evaluation deficits early in stimulus
pain: a review and integration. Psychol. Bull. 127, 599–617. processing in patients with panic disorder. J. Abnorm. Psychology
Pincus, T., Newman, S., 2001. Recall bias, pain, depression and cost in 111, 357–369.
back pain patients. Br. J. Clin. Psychol. 40, 143–156. Wolfe, F.W., Smythe, H.A., Yunas, M.B., Bennett, R.M., Bombardier, C.,
Pincus, T., Pearce, S., McClelland, A., Farley, S., Vogel, S., 1994. Goldenberg, D.L., Tugwell, P., Campbell, S.M., Ables, M., Clark, P.,
Interpretation bias in response to ambiguous cues in pain patients. Fam, A.G., Farber, S.J., Fiechtner, J.J., Franklin, C.M., Gatter, R.A.,
J. Psychosom. Res. 38, 347–353. Hamty, D., Lessard, J., Lichtbronn, A.S., Masi, A.T., McCain, G.A.,
Pulvermuller, F., 2001. Brain reflections of words and their meaning. Reynolds, W.J., Romano, T.J., Russell, I.J., Sheon, R.P., 1990. The
Trends Cogn. Sci. 5, 517–524. American College of Rheumatology 1990 criteria for the classifica-
Rainville, P., Duncan, G.H., Price, D.D., Carrier, B., Bushnell, M.C., tion of fibromyalgia. Arthritis Rheum. 33, 160–172.
1997. Pain affect encoded in human anterior cingulate but not
somatosensory cortex. Science 277, 968–971.
Editorial Manager(tm) for European Journal of Pain
Manuscript Draft
Manuscript Number:
Title: Linear and nonlinear analyses of EEG dynamics during non-painful somatosensory
processing in chronic pain patients
Keywords: EEG activity; multiscale entropy; fractal dimension; FFT; chronic pain
Order of Authors: Carolina Sitges, MSc; Xavier Bornas, PhD; Jordi Llabrés, PhD; Miquel Noguera,
PhD; Pedro Montoya, PhD
Abstract: The aim of the study was to explore electroencephalographic (EEG) activity of chronic
pain patients, evaluating the influence of emotional valence on brain correlates of somatosensory
processing. For this purpose, EEG activity elicited by non-painful pneumatic stimuli were recorded
when subjects were viewing affective pictures. Nineteen patients with musculoskeletal pain and 21
healthy subjects participated in the experiment. Multiscale entropy (MSE), fractal dimension (FD),
and fast Fourier transform (FFT) were applied to characterize variability and complexity of EEG
activity. The analysis revealed significant enhancement of entropy on P4 compared to P3 in chronic
pain patients, suggesting a differential flexibility of the right compared to the left parietal
hemisphere. Analysis of fractal dimension revealed significantly higher values at P4 than at P3 only
under the unpleasant pictures viewing condition for chronic pain patients. By contrast, healthy
subjects did not present any significant differences due to hemisphere or affective conditions on
nonlinear complexity measures. Finally, analyses of EEG band power revealed that chronic pain
patients failed to show an affective modulation in theta and beta activity over sensorimotor cortices
and temporal regions. These findings suggest that sustained pain might lead to an abnormal
activation of brain networks related to emotional processing of somatosensory information among
chronic pain patients.
Abstract
Abstract
dimension (FD), and fast Fourier transform (FFT) were applied to characterize
P4 than at P3 only under the unpleasant pictures viewing condition for chronic
pain patients. By contrast, healthy subjects did not present any significant
measures. Finally, analyses of EEG band power revealed that chronic pain
patients failed to show an affective modulation in theta and beta activity over
patients.
Cover Letter
Department of Psychology
Sincerely,
Carolina Sitges1, Xavier Bornas1, Jordi Llabrés1, Miquel Noguera2 & Pedro
Montoya1
1
University Institute of Health Sciences Research (IUNICS), University of
pain
de Valldemossa km 7.5, 07122 Palma, Spain. Phone: +34 971 172646. Fax:
dimension (FD), and fast Fourier transform (FFT) were applied to characterize
P4 than at P3 only under the unpleasant pictures viewing condition for chronic
pain patients. By contrast, healthy subjects did not present any significant
measures. Finally, analyses of EEG band power revealed that chronic pain
patients failed to show an affective modulation in theta and beta activity over
patients.
2
1. Introduction
Chronic pain refers to pain that persists for more than six months and as such
constitutes one of largest economic burden for the health system in the
unclear, it has been suggested that long-term plastic changes along different
available evidence further reveals that the emotional context in which the body
information in chronic pain patients (Geisser et al., 2003; Staud et al., 2003;
EEG time series may effectively reflect dynamic changes of brain activity
dynamical systems has been applied to EEG and MEG data in order to capture
the macroscopic spatial and temporal dynamics of brain activity (Elbert et al.,
1994; Stam, 2005). In this sense, previous research has shown that emotional
complexity has been reported in chronic pain patients during the recall of
al., 1997); whereas reduced complexity has been associated with states of
3
dimensional complexity in EEG might reflect the activation of widespread
the subjective experience of pain and relative EEG power in healthy controls.
Thus, enhanced pain perception has been related to increased power of beta
rhythm, and to reduced power of delta, theta and alpha rhythms (Chang et al.,
2002a, 2002b, 2003, 2004). Lowered power in the alpha rhythm over primary
subjects’ evaluation of pain intensity (Babiloni et al., 2006). Early reports have
also revealed that chronic pain patients showed less alpha asymmetry than
spatial tasks, cold pressor stimulation) (De Benedittis and De Gonda, 1985;
Stevens et al., 2000), and that neurofeedback training to increase alpha waves
1975).
pain patients and healthy controls during the affective modulation of non-painful
dimension), as well as in spectral band power of the EEG. For this purpose, an
stimuli.
4
2. Materials and methods
2.1. Participants
months who had normal hearing, corrected visual acuity, and no history of head
trauma or drug abuse were recruited from the Pain Management Unit of a
tertiary care hospital and the Medical Unit for Disability Assessment of the
details of the study during the recruitment and noted that their participation was
leaflet was also given, and after agreeing to participate, each subject provided
written consent. The study was in accordance with the Declaration of Helsinki
(1991) and was approved by the Ethics Committee of the University of the
Multidimensional Pain Inventory (MPI) (Kerns et al., 1985), and the McGill Pain
5
2.3. Somatosensory stimulation task and recording of brain activity
Somatosensory stimulation was applied to the index finger of both hands using
2nd digit by a plastic clip and fixated with adhesive strips. During the
random series such that one of them was frequently delivered to the right hand
(86% of the trials), and the other was delivered to the left hand (14% of the
trials) (oddball paradigm). One stimulation block consisted of 560 stimuli (480
received two stimulation blocks: one when they were viewing forty pictures with
negative valence (unpleasant content), and other when they were viewing forty
pictures with positive valence (pleasant content). Each picture was presented
for 6 seconds and was followed by a 6-second blank screen. One deviant and
six frequent tactile stimuli were delivered during each slide presentation, as well
as during the blank screen. All pictures were selected from the International
Affective Picture System (IAPS) (Lang et al., 1997), and were the same as used
elsewhere (Montoya et al., 2005b). The order of the two stimulation blocks was
a sound attenuated room, and instructed to keep eye movements and blinks to
minimum during the experiment. This task has provided to be useful for the
6
study of affective modulation during somatosensory brain processing (a detailed
2005b; a similar task using painful and non-painful stimuli was also described
the present experiment only 15 electrodes located over the midline (Fz, Cz, Pz),
the left (F3, F7, C3, T3, P3, O1) and the right hemisphere (F4, F8, C4, T4, P4,
electrode above and one below the left eye. Ground was placed anteriorly to
10 kΩ. The signals were amplified with a BrainAmp amplifier at a sampling rate
of 1000 Hz, with high and low pass filter setting at 0.10 Hz and 70 Hz,
Data were resampled at 500 Hz, and corrected for blinking and eye movement
artifacts (Gratton et al., 1983). For nonlinear analysis, multiscale entropy (MSE)
was calculated for 2-minutes EEG segments using the software developed by
Costa et al. (2002). This type of entropy measure has been shown to be useful
7
approach of the MSE algorithm, more complex time series, with features at
various scales, will exhibit entropy values that strongly depend on the time
scale, while a less complex signal will show a constant entropy value. For
calculation purposes, the parameter r of the MSE algorithm was set to 20% of
the standard deviation and the parameter m was set to 2, values that are within
the usual range suggested to measure this kind of entropy (Pincus, 1995). The
series were coarse-grained up to scale 20, so that the shortest time series had
3000 points.
al. (2005), implements the Katz and the Sevcik algorithms in Matlab code. In the
same sense that MSE looks at entropy at different temporal scales, the FD
For linear analysis, fast Fourier transform (FFT) was computed on EEG
segments of 512 data points, obtaining the power spectra on the following
frequency bands: delta (2-4 Hz), theta1 (4-6 Hz), theta2 (6-8 Hz), alpha1 (8-10
Hz), alpha2 (10-12 Hz), beta1 (12-18 Hz) and beta2 (18-22 Hz).
Basically, the study followed a factorial mixed design with the within-subjects
and the between-subjects factor ‘group’ (chronic pain patients vs. healthy
8
were carried out separately for midline (Fz, Cz, Pz), and lateral electrode
locations over the left (F3, F7, C3, T3, P3, O1) and the right hemisphere (F4,
F8, C4, T4, P4, O2). Demographic and questionnaire data (STAI and BDI) were
9
3. Results
indicated that chronic pain patients were more depressed and anxious than
healthy controls.
scale factors (sf) 1, 5, 10, 15, 20 at parietal electrode locations (P3 and P4)
(Figure 1). Significant interaction effects between the factors hemisphere and
Bonferroni corrected). Sample Entropy was higher in the right than in the left
parietal electrode for any scale factor. The same analysis was carried out for
other electrodes over centro-parietal brain regions (C3, C4, Cz, and Pz), but no
10
3.3. Fractal Dimension
(unpleasant vs. pleasant) and hemisphere (left vs. right), and the between-
subject factor group (healthy volunteers vs. chronic pain patients) were
computed on all frequency bands for each brain region (frontal, fronto-lateral,
to test the effects on all frequency bands over midline electrodes (see Figure 3
and 4, and Table 3). A first observation is that the total power at any frequency
Delta (2-4 Hz). Significant main effects of emotional context were found
unpleasant than pleasant images. Higher delta activity was observed in control
11
was also found in temporal electrodes (F(1,38)=4.97, p<.05), showing greater
delta activity over the right than over the left hemisphere (p<.01) in healthy
Theta1 (4-6 Hz) and theta2 (6-8 Hz). A significant interaction between
group and hemisphere in theta1 was found for temporal electrodes (T3/T4)
activity over the right than over the left (p<.01), whereas no differences
between group, hemisphere and emotional context was found for central
the left than over the right hemisphere during unpleasant images (p<.05) in
healthy controls, but not in chronic pain patients. Higher theta2 activity was also
observed over the left than over the right hemisphere for central (C3/C4)
Alpha1 (8-10 Hz) and alpha2 (10-12 Hz). No significant effects were
F(1,30)=4.54, p<.05), showing that healthy controls had higher activity than
Beta1 (12-18 Hz) and beta2 (18-22 Hz). A significant main effect of
chronic pain patients. Greater beta2 activity was elicited by unpleasant than by
way interaction between group, hemisphere and emotional context was found
12
for temporal electrodes (T3/T4) (F(1,37)=5.59, p<.05), showing that pleasant
images elicited greater beta2 activity over the right hemisphere than over the
left hemisphere in healthy controls (p<.05), but not in chronic pain patients.
13
4. Discussion
The present study revealed that chronic pain patients and healthy controls differ
stimulation was applied to the hand in affective contexts. Thus, we found that
chronic pain patients displayed more entropy and fractal dimension over the
right than over the left parietal hemisphere; whereas healthy controls did not
patients at several EEG frequency bands (delta, theta1, and beta2). These
findings are in agreement with previous data indicating that chronic pain
whereas fractal dimension of the EEG has been related to the activity and
processes, such as handgrip force (Liu et al., 2005) or memory for pain
information about the flexibility that the system needs to respond to cognitive
demands. In this context, the higher entropy values observed in chronic pain
between the left and the right posterior parietal brain in these patients when
14
stimulation was applied to the right hand, whereas infrequent stimulation was
applied to the left hand. Thus, it could be argued that occurrence of stimulation
at the left hand was less predictable than stimulation at the right hand, leading
to a higher alertness state. In our opinion, the higher entropy found in the right-
sided parietal EEG activity could reveal this system's mode of brain operation.
On the contrary, repetitive stimulation (like the one delivered to the right hand)
can be easily predicted by the system, which therefore does not need to be in a
further to speculate that the origins of this pattern of brain dynamics could be in
the personal pain history of chronic pain patients. By repeatedly coping with
painful stimuli in their lives, these patients would have developed a flexible brain
pain patients when they were viewing unpleasant pictures, but not pleasant
during a particular task (Lutzenberger et al., 1995). This measure has been
handgrip force (Liu et al, 2005), respectively. In this sense, our results might
indicate a more enhanced activation of cell assemblies over the right than over
the left parietal cortex in chronic pain patients when non-painful stimulation was
15
pain patients might be characterized by an abnormal processing of non-painful
with the hypothesis that negative contextual information could enhance pain
of alpha and beta EEG frequency bands over parietal brain regions as
Hz) and slow beta power (12.5-19.5 Hz) have been found in patients with
chronic pain during resting states (Chen et al., 1998), together with reduced
Golocheikine, 2002). In the present study, several aspects of the EEG pattern
Thus, pain induction using the cold pressor test (Chang et al., 2002b), an
Hz) and alpha2 (11-14 Hz) over parietal brain regions with respect to baseline,
but also significant increases of delta (0.5-4 Hz), theta (4-8 Hz) and beta2 (14-
25 Hz) over frontal and temporal areas. Although the significance of these
still controversial (Stevens et al. 2000; Apkarian et al., 2005), it has been
argued that reductions in alpha activity over posterior brain regions could be
16
the attention system during sustained painful stimulation (Chang et al., 2002a).
chronic pain might indicate that persistent pain, as it occurs in chronic pain
group. In healthy controls, unpleasant images elicited greater delta and beta2
activity than pleasant images over frontal brain regions, as well as greater
theta1 and theta2 activity over the left than over the right sensorimotor cortex.
delta and beta activity was also observed in chronic pain patients over frontal
brain regions, but no modulation appeared in theta and beta activity over
during tonic pain have been related to the processing of motivational and
affective pain components (Chen et al., 1989; Pauli et al., 1999; Chang et al.,
2002b). Recently, several studies have also linked an increased frontal activity
al., 2006; Schweinhardt et al., 2008). Thus, it might be interpreted that chronic
17
pain patients display an abnormal pattern of EEG activation over brain regions
First, it should be noted that many patients with chronic pain in the present
drugs prescribed for chronic pain on sensory and cognitive brain processing. A
range 2-18 Hz) (Sarnthein et al., 2006). To clarify this issue, future studies
patients. Second, chronic pain patients scored significantly higher than controls
18
could be possible that the findings reported here were mainly due to the effects
of mood rather than to the effects of pain. Further research should investigate
whether chronic pain patients and patients with affective disorders without pain
measures of EEG time series may contribute to increasing the insight into brain
patients displayed increased entropy and fractal dimensionality values over the
right parietal region, suggesting that non-painful stimulation of the left hand
chronic clinical pain conditions have distinct, but overlapping brain activation
19
Acknowledgments
and valuable comments. This research was supported by the Spanish Ministerio
20
References
2006; 27:241-253.
Aftanas LI, Golocheikine SA. Non-linear dynamic complexity of the human EEG
Aftanas LI, Lotova NV, Koshkarov VI, Makhnev VP, Mordvinstev YN, Popov,
Aftanas LI, Lotova NV, Koshkarov VI, Pokrovskaja VL, Popov SA, Makhnev VP.
Aftanas LI, Lotova NV, Koshkarov VI, Popov SA, Makhnev VP. Nonlinear
Apkarian AV, Bushnell MC, Treede RD, Zubieta JK. Human brain mechanisms
of pain perception and regulation in health and disease. Eur J Pain 2005;
9:463-484.
21
high-resolution electroencephalography study. Eur J Neurosci 2003;
18:1692-1700.
Baliki MN, Chialvo DR, Geha PY, Levy PY, Harden RN, Parrish TB, Apkarian
AV. Chronic pain and the emotional brain: specific brain activity
aversive auditory arousal versus muscle pain: specific EEG patterns are
associated with human pain processing. Exp Brain Res 2002a; 147:387-
393.
correlation of EEG activities during cold pressor test in man. Brain Res
22
Chang PF, Arendt-Nielsen L, Graven-Nielsen T, Svensson P, Chen AC.
Elbert T, Ray WJ, Kowalik ZJ, Skinner JE, Graf KE, Birbaumer N. Chaos and
1994; 74:1-37.
Geisser ME, Casey KL, Brucksch CB, Ribbens CM, Appleton BB, Crofford LJ.
Gratton G, Coles MG, Donchin E. A new method for off-line removal of ocular
193-199.
23
Kemp AH, Gray MA, Silberstein RB, Armstrong SM, Nathan PJ. Augmentation
Kerns RD, Turk DC, Rudy TE. The West Haven-Yale Multidimensional Pain
Lang PJ, Bradley MM, Cuthbert BN. International affective picture system
Liu JZ, Yang Q, Yao B, Brown RW, Yue GH. Linear correlation between fractal
dimension of EEG signal and handgrip force. Biol Cybern 2005; 93:131-
140.
EEG during memory for personal pain in chronic pain patients. Neurosci
24
Melzack R, Perry C. Self-regulation of pain: the use of alpha-feedback and
hypnotic training for the control of chronic pain. Exp Neurol 1975; 46:452-
469.
67:957-963.
39:991–999.
5:110-117.
25
Schweinhardt P, Kalk N, Wartolowska K, Chessell I, Wordsworth P, Tracey I.
Spielberger CD, Gorsuch RL, Lushene RE. The State-Trait Anxiety Inventory
1970.
Staud R, Robinson ME, Vierck CJ Jr, Cannon RC, Mauderli AP, Price DD.
clinical pain in patients with fibromyalgia syndrome. Pain 2003; 105: 215-
222.
Veldhuijzen DS, Kenemans JL, van Wijck AJ, Olivier B, Kalkman CJ, Volkerts
Zhuo M. Cortical excitation and chronic pain. Trends Neurosci 2008; 31:199-
207.
26
Figure legends
Figure 1. Mean SampEn values for scale factors (sf) 1, 5, 10, 15, 20 on P3
(solid line) and P4 (dashed line) for healthy controls and chronic pain patients.
Figure 2. Mean scores and standard errors (bars) of fractal dimension (FD) at
electrodes P3 and P4 for healthy controls and chronic pain patients during non-
Figure 3. Mean values and standard errors (bars) of FFT analysis on delta (2-4
Hz), alpha 2 (10-12 Hz), and beta 1 (12-18 Hz) at central and parietal
Figure 4. (a) Average of EEG power spectra (2-22 Hz) for healthy controls and
chronic pain patients, and (b) topographical distributionof EEG power spectra
on the scalp.
27
Table 1
* p≤.001
Table 2
vs. chronic pain patients) x condition (unpleasant vs. pleasant) x localization (P3
vs. P4) MANOVAs computed on SampEn values obtained with the MSE
3-way MANOVAs
F p F p
Table 3. Interaction effects revealed by the higher order group (healthy controls
(left vs. right) x localization (P3, P4, C3, C4, O1, O2) MANOVAs and interaction
effects revealed by three-way group (healthy controls vs. chronic pain patients)
F p F p F p F p
4. Resultados
En Montoya, Sitges et al. (2005), los principales resultados fueron los siguientes:
únicamente el grupo de pacientes con fibromialgia (FM) mostró una mayor amplitud
en el primer componente positivo, P50 (20-80 ms), de los PEs somatosensoriales
(PESs) en los electrodos laterales (F3, C3, P3, F4, C4, P4), mientras visualizaban la
serie de imágenes desagradables (unpleasant), con respecto a la serie de imágenes
agradables (pleasant). Este mismo grupo de pacientes también mostró una menor
amplitud del primer componente negativo, N80 (60-110 ms), durante la presentación
del estímulo infrecuente (deviant) mientras visualizaba la serie de imágenes
desagradables, con respecto a la otra serie de imágenes agradables. No se encontró
ninguna diferencia en estos dos componentes (P50, N80) en el grupo de pacientes con
dolor crónico de origen orgánico (MSK), así como tampoco se hallaron diferencias
estadísticamente significativas, debidas al contexto emocional (imágenes
desagradables, agradables) o al tipo de estímulo (frecuente, infrecuente), en el
segundo componente positivo, P200 (135-260 ms), para ninguno de los dos grupos de
pacientes. Los mapas de densidad de corriente mostraron que la distribución de
dichos componentes (P50, N80) era parietal, por tanto, involucraba probablemente a
la corteza somatosensorial primaria (SI), y estaban situados en el hemisferio
contralateral a la zona corporal (dedo índice derecho o izquierdo) estimulada. En
cuanto al nivel de dolor informado, de nuevo, únicamente el grupo de pacientes con
FM mostró un incremento después de visualizar las imágenes con contenido
desagradable.
Los resultados del siguiente estudio, Montoya & Sitges (2006), muestran que,
en el primer componente positivo, P50 (20-80 ms), de los PESs, la serie de imágenes
agradables (pleasant), comparada con la serie de imágenes desagradables
(unpleasant), y el estímulo infrecuente (deviant), comparado con el estímulo
frecuente producen una mayor amplitud. Los mapas de densidad de corriente
muestran que dicho componente presenta una distribución centroparietal,
contralateral a la zona corporal (dedo índice derecho o izquierdo) estimulada, con
las amplitudes máximas en CP4 y C4 (es decir, en la zona centroparietal derecha,
que es contralateral al dedo índice izquierdo, donde se presentaba el estímulo
infrecuente). Por otro lado, también se analizó dicho componente durante el
intervalo entre estímulos, en el que los sujetos no recibían estimulación visual pero
seguían recibiendo estimulación táctil, y se mostró que su amplitud era mayor ante
Estudio de la actividad cerebral en pacientes con fibromialgia 135
En Sitges et al. (2007) se muestra, en primer lugar, que más del 70% de las
palabras afectivas y sensoriales fueron consideradas descriptivas de la experiencia
del dolor para los pacientes con dolor crónico de origen orgánico (MSK) y para los
pacientes con fibromialgia (FM), mientras que sólo un 37% de esas palabras fueron
C. Sitges 136
consideradas como descriptores de su dolor para los pacientes sanos (sin dolor
crónico). En cuanto a los tiempos de reacción, los pacientes con MSK tardan más
tiempo en responder ‘sí’ ante las palabras agradables, con respecto a los
descriptores afectivos; y menos tiempo para responder ‘no’ ante las palabras
agradables, con respecto a los adjetivos sensoriales. Los pacientes con FM tardan
menos en responder ‘no’ ante las palabras agradables, en comparación con los
adjetivos sensoriales y los afectivos. Los sujetos sanos no muestran diferencias en los
tiempos de reacción en ninguno de los casos. Y, en cuanto a los PEVs, en el
componente P200 (~250 ms) se mostraron amplitudes significativamente mayores
ante la presentación de los descriptores sensoriales, en comparación con los
afectivos, sólo en el grupo de sujetos sanos. Los análisis topográficos mostraron
diferencias significativas entre los descriptores sensoriales y afectivos en los
electrodos frontales, centrales y centro-parietales sólo en ese grupo de sujetos. Los
sujetos sanos también mostraron, en dicho componente, mayores amplitudes en los
electrodos parietales en respuesta a los descriptores sensoriales, en comparación con
el grupo de pacientes con MSK. En cuanto al componente LPC (500-800 ms), se
mostró que las amplitudes producidas por los descriptores afectivos eran
significativamente mayores que las producidas por las palabras agradables en los
pacientes con MSK, y que las amplitudes debidas a los descriptores sensoriales eran
mayores que las generadas por los descriptores afectivos en los sujetos sanos. Los
análisis topográficos mostraron que los sujetos sanos tenían mayores amplitudes en
dicho componente, en respuesta a los descriptores sensoriales, en los electrodos
centrales y centro-parietales, en comparación con los pacientes con FM. Por otro
lado, los sujetos sanos también mostraron mayores amplitudes en dicho componente,
en respuesta a las palabras agradables, en los electrodos parietales, en comparación
con los dos grupos de pacientes con dolor crónico. Por último, cabe comentar que
tanto los pacientes con dolor crónico de origen orgánico (MSK) como los pacientes
con FM presentaban mayores niveles de depresión y ansiedad que el grupo de sujetos
sanos. Además, los pacientes con FM obtenían mayores puntuaciones en la dimensión
sensorial y afectiva del McGill Pain Questionnaire (MPQ; Melzack, 1975), y en la
escala de agarrotamiento del Fibromyalgia Impact Questionnaire (FIQ; Burckhardt,
Clark & Bennett, 1991).
En Sitges et al. (2008), tanto las técnicas de análisis lineales como las no
lineales muestran diferencias entre los dos grupos de sujetos en casi todas las
medidas. En primer lugar, en cuanto al índice de entropía multiescalar, únicamente
Estudio de la actividad cerebral en pacientes con fibromialgia 137
5. Discusión general
Los resultados del primer estudio (Montoya, Sitges et al., 2005), aportan evidencia de
un procesamiento sensorial y cognitivo diferente en el grupo de pacientes con FM,
con respecto al grupo de pacientes con MSK. Los pacientes con FM muestran un
mecanismo perceptivo selectivo, que parece estar influenciado por factores
emocionales (en este caso, por el contexto afectivo generado mediante la
presentación de imágenes). El mismo paradigma llevado a cabo con sujetos sanos
(Montoya & Sitges, 2006), muestra la influencia diferencial del contexto emocional
sobre el procesamiento somatosensorial temprano (P50) de estímulos infrecuentes no
nociceptivos y sobre el procesamiento visual más tardío (P200). Más concretamente,
se pudo observar que un contexto emocional desagradable atenúa de forma
significativa el componente P50, que podría interpretarse como el reflejo de un
mecanismo de filtro sensorial inicial afectado por un sesgo atencional hacia estímulos
aversivos (Smith et al., 2006). Por otro lado, los estímulos negativos (imágenes
desagradables) provocan un PE, originado en la corteza de asociación visual, de
mayor amplitud y menor latencia que el originado tras la presentación de estímulos
positivos, debido a la movilización de recursos atencionales dirigidos al
procesamiento de imágenes afectivas (ver Carretié et al., 2001, 2003). A pesar de no
haberse tomado en cuenta a la hora de realizar los análisis estadísticos, en la Fig 1b
de este segundo estudio se muestra un tercer componente, que diferencia a los
pacientes con dolor crónico de los sujetos sanos. Se trata de un componente positivo
que aparece ante la presentación del estímulo infrecuente alrededor de los 300 ms
(comparar Figure 2 (deviant) de Montoya, Sitges et al. (2005) y Fig 1b de Montoya &
Sitges (2006)). Probablemente, se trate de la denominada P3b (ver Polich, 2007); un
componente que se relaciona con la respuesta cerebral al estímulo infrecuente, y
con estructuras temporales y parietales.
Una de las limitaciones de este estudio (Montoya, Sitges et al., 2006), que
aparece también en muchos de los estudios realizados con pacientes, incluidos
también los que se pasarán a discutir a continuación, es que los sujetos sanos no
representan un buen grupo de comparación, ya que no comparten las mismas
historias clínicas, farmacológicas ni sociales. En concreto, en este trabajo el grupo de
pacientes con FM presentaba mayores niveles de depresión y ansiedad, en
comparación con el grupo de sujetos sanos, que además se comprobó que
correlacionaban positivamente con el incremento de la ratio de S2 con respecto a S1,
reduciendo de este modo el efecto de habituación. Por otro lado, numerosos estudios
apuntan los efectos de diversos fármacos antidepresivos (tricíclicos, bloqueadores de
los receptores 5-HT2, inhibidores de la recaptación de la serotonina) y ansiolíticos
sobre diversos procesos cognitivos, como la memoria de trabajo, la atención, la
detección de errores, y sus correspondientes correlatos cerebrales (PEs auditivos,
C. Sitges 140
visuales) (Andereri et al., 2008; De Bruijn, Hulstijn, Verkes, Ruigt & Sabbe, 2004;
Jongsma et al., 1998; Riba, Rodríguez-Fornells, Münte & Barbanoj, 2005; Saletu,
Anderer, Di Padova, Assandri, & Saletu-Zyhlarz, 2002; Van Laar et al., 2002;
Veldhuijzen et al., 2006). Una posible solución a esta limitación sería utilizar otro
grupo de pacientes, para hacer de sujetos control, con características lo más
similares posibles a los pacientes con dolor crónico, como es el caso de los pacientes
con depresión mayor.
El estudio de los PEs, tal como señala Marco-Pallarés (2005), además de las
ventajas comentadas en el apartado 2.1.1, también presenta diversas limitaciones,
que son objeto de controversia y hacen que las conclusiones a las que se llega con
esta técnica sean a menudo cuestionadas. Una limitación importante del estudio de
la actividad eléctrica (o magnética) es el hecho de que la resolución espacial que se
obtiene es del orden de los centímetros y, por tanto, muy inferior a la estimada con
otras técnicas de neuroimagen, como la resonancia magnética funcional (fMRI) o la
tomografía por emisión de positrones (PET), que es del orden de los milímetros. Por
otro lado, a diferencia de estas técnicas donde se obtiene la localización y la
intensidad de las fuentes del parámetro medido, la información obtenida mediante
EEG/MEG no permite mostrar una información directa de la localización de las
fuentes cerebrales que producen los potenciales, ya que los electrodos están
situados en el cuero cabelludo y, por tanto, relativamente lejos de las fuentes que
generan los potenciales y además están separadas por capas de material con
diferentes propiedades eléctricas. La segunda limitación de los PEs está relacionada
con el estudio de su evolución temporal. En general, la mayoría de los PEs se
generan por la activación de muchas áreas cerebrales diferentes, por lo que la
actividad eléctrica observada en el cuero cabelludo es el efecto de la información
solapada de la actividad eléctrica procedente de muchas áreas cerebrales, cada una
de las cuales con una dinámica propia y que dan como resultado un PE que a menudo
tiene una estructura aparentemente unitaria. Y, la tercera gran limitación, que
aparece en el estudio de los PEs, viene determinada por el tratamiento estadístico
de los datos originales que se usan para obtenerlos, ya que se obtienen promediando
un número elevado de ensayos para poder eliminar la información considerada ruido
y obtener la respuesta evocada cerebral. Pero el EEG no es sólo ruido, ya que al
realizar un espectro de frecuencias se puede observar que la forma principal de éste
es de 1/f pero sobre el que aparecen picos a determinadas frecuencias, denominadas
bandas.
resolución espacial (BESA, LORETA) como de la evolución temporal de los PEs (ICA,
PCA), que se comentarán en el siguiente apartado.
6. Conclusiones finales
• Por último, parece ser que el cerebro de los pacientes con dolor crónico (MSK)
mostraría mayor nivel de complejidad, así como una mayor dependencia a las
condiciones iniciales, como muestran los mayores valores obtenidos en
dimensión fractal y entropía muestral en la zona parietal derecha, y
‘desincronizado’, como muestran los menores valores de potencia en casi todo
el espectro de frecuencias en las zonas centrales, temporales y parietales, que
el de los sujetos sanos, a la hora de participar en una tarea de procesamiento
de estímulos táctiles influenciada por contextos afectivos positivos y negativos.
Estudio de la actividad cerebral en pacientes con fibromialgia 147
7. Referencias bibliográficas
Abeles, A. M., Pillinger, M. H., Solitar, B. M. & Abeles, M. (2007). Narrative review:
the pathophysiology of fibromyalgia. Annals of Internal Medicine 146, 726-
734.
Ademoglu, A., Micheli-Tzanakou, E. & Istefanopulos, Y. (1997). Analysis of pattern
reversal visual evoked potentials by spline wavelets. IEEE Transactions of
Biomedical Engineering 44, 881-890.
Adler, L. E., Olincy, A., Waldo, M., Harris, J. Q., Griffith, J., Stevens, K. et al.
(1998). Schizophrenia, sensory gating and nicotinic receptors. Schizophrenia
Bulletin 24, 189-202.
Aftanas, L. I. & Golocheikine, S. A. (2002). Non-linear dynamic complexity of the
human EEG during meditation. Neuroscience Letters 330, 143-146.
Aftanas, L. I., Koshkarov, V. I., Mordvinstey, Y. N. & Pokrovskaja, V. L. (1994).
Dimensional analysis of human EEG during experimental affective experience.
International Journal of Psychophysiology 18, 67-70.
Aftanas, L. I., Lotova, N. V., Koshkarov, V. I., Makhnev, V. P., Mordvinstey, Y. N. &
Popov, S. A. (1998). Non-linear dynamical complexity of the human EEG
during evoked emotions. International Journal of Psychophysiology 28, 63-67.
Aftanas, L. I., Lotova, N. V., Koshkarov, V. I., Pokrovskaja, V. L., Popov, S. A. &
Makhnev, V. P. (1997a). Non-linear analysis of emotion EEG: calculation of
Kolmogorov entropy and the principal Lyapunov exponent. Neuroscience
Letters 226, 13-16.
Aftanas, L. I., Lotova, N. V., Koshkarov, V. I., Popov, S. A. & Makhnev, V. P. (1997b).
Nonlinear forecasting measurements of the human EEG during evoked
emotions. Brain Topography 10, 155-162.
Aliaga, L. y Santacana, E. (1994). Dolor. Madrid: IDEPSA.
Ambrosini, A., De Pasqua, V., Áfra, J., Sandor, P. S. & Schoenen, J. (2001). Reduced
gating of middle-latency auditory evoked potentials (P50) in migraine
patients: another indication of abnormal sensory processing? Neuroscience
Letters 306, 132-134.
American Psychiatric Association (1980). DSM-III: Manual diagnóstico y estadístico de
los trastornos mentales. Barcelona: Masson.
American Psychiatric Association (1995). DSM-IV: Manual diagnóstico y estadístico de
los trastornos mentales. Barcelona: Masson.
C. Sitges 148
Başar, E., Demiralp, T., Schürmann, M., Başar-Eroglu, C. & Ademoglu, A. (1999).
Oscillatory brain dynamics, wavelet analysis, and cognition. Brain & Language
66, 146–183.
Başar-Eroglu, C., Başar, E., Demiralp, T. & Schürmann, M. (1992). P300-response:
possible psychophysiological correlates in delta and theta frequency channels.
A review. International Journal of Psychophysiology 13, 161-179.
Boutros, N. N. & Belger, A. (1999). Midlatency evoked potentials attenuation and
augmentation reflect different aspects of sensory gating. Biological Psychiatry
45, 917-922.
Breivik, H., Collett, B., Ventafridda, V., Cohen, R. & Gallacher, D. (2006). Survey of
chronic pain in Europe: prevalence, impact on daily life, and treatment.
European Journal of Pain 10, 287-333.
Burckhardt, C. S., Clark, S. R. & Bennett, R. M. (1991). The Fibromyalgia Impact
Questionnaire: Development and validation. Journal of Rheumatology 18, 728-
734.
Cailliet, R. (1995). Síndromes dolorosos. Dolor y mecanismos de manejo. México: El
Manual moderno.
Carretié, L. e Iglesias, J. (1995). Psicofisiología. Fundamentos metodológicos.
Madrid: Pirámide.
Carretié, L., Hinojosa, J. A. & Mercado, F. (2003). Cerebral patterns of attentional
habituation to emotional visual stimuli. Psychophysiology 40, 381–388.
Carretié, L., Martín-Loeches, M., Hinojosa, J. A. & Mercado, F. (2001). Emotion and
attention interaction studied through event-related potentials. Journal of
Cognitive Neuroscience 13, 1109–1128.
Carrillo de la Peña, M. T., Vallet, M., Pérez, M. I. & Gómez-Perretta, C. (2006).
Intensity dependence of auditory-evoked cortical potentials in fibromyalgia
patients: A test of the generalized hypervigilance hypothesis. The Journal of
Pain 7, 480-487.
Català, E., Reig, E., Artés, M., Aliaga, L., López, J. S. & Segú, J. L. (2002).
Prevalence of pain in the Spanish population: telephone survey in 5000 homes.
European Journal of Pain 6, 133-140.
Chang, P-F., Arendt-Nielsen, L., Graven-Nielsen, T. & Chen, A. C. N. (2003).
Psychophysical and EEG responses to repeated experimental muscle pain in
humans: Pain intensity encodes EEG activity. Brain Research Bulletin 59, 533–
543.
C. Sitges 150
Diers, M., Koeppe, C., Yolmaz, P., Thieme, K., Markela-Lerenc, J., Schiltenwolf, M.
et al. (2008). Pain ratings and somatosensory evoked responses to repetitive
intramuscular and intracutaneous stimulation in fibromyalgia syndrome.
Journal of Clinical Neurophysiology 25, 153-160.
Eccleston, C. (1994). Chronic pain and attention: A cognitive approach. British
Journal of Clinical Psychology 33, 535–547.
Eccleston, C. & Crombez, G. (1999). Pain demands attention: A cognitive-affective
model of the interruptive function of pain. Psychological Bulletin 125, 356–
366.
Edwards, R., Bingham III, C. O., Bathon, J. & Haythornthwaite, J. A. (2006).
Catastrophizing and pain in arthritis, fibromyalgia, and other rheumatic
diseases. Arthritis Care & Research 55, 325–332.
Evers, A. W., Kraaimaat, F. W., Greenen, R., Jacobs, J. W. & Bijlsma, J. W. (2002).
Longterm predictors of anxiety and depressed mood in early rheumatoid
arthritis: a 3 and 5 year follow up. Journal of Rheumatology 29, 2327-2336.
Flor, H. & Birbaumer, N. (1994). Acquisition of chronic pain: psychophysiologic
mechanisms. American Pain Society Journal 32, 119–127.
Flor, H., Birbaumer, N., Schugens, M. M. & Lutzenberger, W. (1992). Symptom-
specific psychophysiological responses in chronic pain patients.
Psychophysiology 29, 452–460.
Flor, H., Knost, B. & Birbaumer, N. (1997). Processing of pain- and body-related
verbal material in chronic pain patients: central and peripheral correlates.
Pain 73, 413-421.
Francis, S. E. (1995). Chaotic phenomena in Psychophysiological self-regulation.
Robertson, R. & Combs, A. (eds.). Chaos theory in Psychology and life
sciences (pp. 253-265). Mahwah, New Jersey: Lawrence Erlbaum Associates,
Publishers.
Freedman, R., Waldo, M., Bickford-Winner, P. & Nagamoto, H. T. (1991). Elementary
neuronal dysfunction in schizophrenia. Schizophrenia Research 4, 233–243.
Giovengo, S. L., Russell, I. J. & Larson, A. A. (1999). Increased concentrations of
nerve growth factor in cerebrospinal fluid of patients with fibromyalgia. The
Journal of Rheumatology 26, 1564-1569.
Grace, G. M., Nielson, W. R., Hopkins, M. & Berg, M. A. (1999). Concentration and
memory deficits in patients with fibromyalgia syndrome. Journal of Clinical
& Experimental Neuropsychology 21, 477-487.
C. Sitges 152
Gracely, R. H., Geisser, M. E., Giesecke, T., Grant, M. A. B., Petzke, F., Williams, D.
A. et al. (2004). Pain catastrophizing and neural responses to pain among
persons with fibromyalgia. Brain 127, 835-843.
Grachev, I. D., Fredrickson, B. E. & Apkarian, A. V. (2000). Abnormal brain chemistry
in chronic back pain: an in vivo proton magnetic resonance spectroscopy
study. Pain 89, 7-18.
Granot, M., Buskila, D., Granovsky, Y., Sprecher, E., Neumann, L. & Yarnitsky, D.
(2001). Simultaneous recording of late and ultra-late pain evoked potentials in
fibromyalgia. Clinical Neurophysiology 112, 1881-1887.
Grassberger, P. & Procaccia, I. (1983). Measuring the strangeness of strange
attractors. Physica D 9,189--208.
Gratton, G., Coles, M.G.H., & Donchin, E. (1983). A new method for off-line removal
of ocular artifact. Electroencephalography & Clinical Neurophysiology 55,
468-484.
Gray, C. M. & Singer, W. (1989). Stimulus-specific neuronal oscillations in orientation
columns of cat visual cortex. Proceedings of the National Academy of Sciences
of the USA 86, 1698-1702.
Grigsby, J., Rosenberg, N. L. & Busenbark, D. (1995). Chronic pain is associated with
deficits in information processing. Perceptual & Motor Skills 81, 403–410.
Grunwald, T., Boutros, N. N., Pezer, N., von Oertzen, J., Fernández, G., Schaller, C.
et al. (2003). Neuronal substrates of sensory gating within the human brain.
Biological Psychiatry 53, 511-519.
Hari, R., Reinikainen, K., Kaukoranta, E., Hämäläinen, M., Ilmoniemi, R., Penttinen,
A. et al. (1984). Somatosensory evoked cerebral magnetic fields from SI and
SII in man. Electroencephalography & Clinical Neurophysiology 57, 254–263.
Hart, R. P., Martelli, M. F., & Zasler, N. D. (2000). Chronic pain and
neuropsychological functioning. Neuropsychology Review 10, 131-149.
Heath, R. A. (2000). Nonlinear dynamics. Techniques and applications in Psychology.
Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers.
Hebb, D. O. (1949). The organization of behavior: A neuropsychological theory. New
York: Wiley-Interscience.
Heinrichs, M., Baumgartner, T., Kirschbaum, C. & Ehlert, U. (2003). Social support
and oxytocin interact to suppress cortisol and subjective responses to
psychosocial stress. Biological Psychiatry 54, 1389-1398.
Jasper, H. A. (1958). The ten–twenty system of the International Federation.
Electroencepholography & Clinical Neurophysiology 10, 371–375.
Estudio de la actividad cerebral en pacientes con fibromialgia 153
Landro, N. I., Stiles, T. C. & Sletvold, H. (1997). Memory functioning in patients with
primary fibromyalgia and major depression and healthy controls. Journal of
Psychosomatic Research 42, 297–306.
Lang, P. J., Bradley, M. M. & Cuthbert, B. N. (1997). International Affective Picture
System (IAPS): Technical manual and affective ratings. Gainesville, FL:
University of Florida.
Legrain, V., Guérit, J-M., Bruyer, R. & Plaghki, L. (2002). Attentional modulation of
the nociceptive processing into the human brain: selective spatial attention,
probability of stimulus occurrence, and target detection effects on laser
evoked potentials. Pain 99, 21-39.
Liu, J. Z., Yang, Q., Yao, B., Brown, R. W. & Yue, G. H. (2005). Linear correlation
between fractal dimension of EEG signal and handgrip force. Biological
Cybernetics 93, 131-140.
Lorenz, J. (1998). Hyperalgesia or hypervigilance? An evoked potential approach to
the study of fibromyalgia syndrome. Zeitschrift für Rheumatologie 57, S2:19-
22.
Lorenz, J., Grasedyck, K. & Bromm, B. (1996). Middle and long latency
somatosensory evoked potentials after painful laser stimulation in patients
with fibromyalgia syndrome. Electroencephalography & Clinical
Neurophysiology 100, 165–168.
Lutzenberger, W., Flor, H. & Birbaumer, N. (1997). Enhanced dimensional complexity
of the EEG during memory for personal pain in chronic pain patients.
Neuroscience Letters 266, 167-170.
Makeig, S., Debener, S., Onton, J. & Delorme, A. (2004). Mining event-related brain
dynamics. TRENDS in Cognitive Sciences 8, 204-210.
Marco-Pallarés, J. (2005). Augment de la resolució espacial i estudi de la dinàmica
temporal de les fonts de corrent dels potencials evocats cerebrals. Tesis
doctoral para optar al título de Doctor en Neurociencias, Departamento de
Psiquiatría y Psicobiología Clínica, Universitat de Barcelona, España.
Martin, J. H. (1999). Neuroanatomía. Madrid: Prentice Hall.
Martínez-Lavín, M., Hermosillo, A. G., Rosas, M. & Soto, M. E. (1998). Circadian
studies of automatic nervous balance in patients with fibromyalgia: A heart
rate variability analysis. Arthritis & Rheumatism 41, 1966-1971.
Martínez-Lavín, M., Vidal, M., Barbosa, R. E., Pineda, C., Casanova, J. M. & Nava, A.
(2002). Norepinephrine-evoked pain in fibromyalgia. A randomised pilot study
(ISCRTN 70707830). BMC Musculoskeletal Disorders 3, 1-6.
Estudio de la actividad cerebral en pacientes con fibromialgia 155
Melzack, R. & Wall, P. D. (1965). Pain mechanisms: a new theory. Science 150, 971-
979.
Melzack, R. (1975). The McGill Pain Questionnaire: major properties and scoring
methods. Pain 1, 277-299.
Melzack, R. (1999). From the gate to the neuromatrix. Pain 82, S121-S126.
Merskey, H. & Bogduk, N. (eds.) (1994). Classification of Chronic Pain: Descriptions
of Chronic Pain Syndromes and Definitions of Pain Terms. Seattle: IASP Press.
Miller, C. L. & Freddman, (1995). The activity of hippocampal interneurons and
pyramidal cells during the response of the hippocampus to repeated auditory
stimuli. Neuroscience 69, 371-381.
Montoya, P. & Larbig, W. (2006). Abnormal somatosensory brain processing in
fibromialgia: Neurophysiological and behavioral evidence. Pederson, J. A.
(ed.). New research on fibromyalgia (pp.129-156). New York: Nova Science
Publishers, Inc.
Montoya, P. & Sitges, C. (2006). Affective modulation of somatosensory-evoked
potentials elicited by tactile stimulation. Brain Research 1068, 205-212.
Montoya, P., Sitges, C., García-Herrera, M., Izquierdo, R., Truyols, M., Blay, N. et al.
(2005). Abnormal affective modulation of somatosensory brain processing
among patients with fibromyalgia. Psychosomatic Medicine 67, 957-963.
Montoya, P., Sitges, C., García-Herrera, M., Izquierdo, R., Truyols, M. y Collado, D.
(2006). Consideraciones acerca de las alteraciones de la actividad cerebral en
pacientes con fibromialgia. Reumatología Clínica 2, 251-260.
Montoya, P., Sitges, C., García-Herrera, M., Rodríguez-Cotes, A., Izquierdo, R.,
Truyols, M. & Collado, D. (2006). Reduced brain habituation to sensory
stimulation in patients with fibromyalgia. Arthritis & Rheumatism 54, 1995-
2003.
Nandrino, J. L., Pezard, L., Martinerie, J., el Massioui, F., Renault, B., Jouvent, R. et
al. (1994). Decrease of complexity in EEG as a symptom of depresión.
NeuroReport 12, 528-530.
Navarro, J. (2007). Dinámicas No Lineales: Algunas Técnicas de Análisis y Software
Libre. Extraído el 3 de noviembre de 2008 de
http://complejidad.bligoo.com/content/view/221417/Software_Libre_y_Dina
micas_Complejas.html
Neugebauer, V., Li, W., Bird, G. C. & Han, J. S. (2004). The amygdale and persistent
pain. Neuroscientist 10, 221–234.
C. Sitges 156
Pezard, L., Nandrino, J. L., Renault, B., el Massioui, F., Allilaire, F., Muller, J. et al.
(1996). Depression as a dynamical disease. Biological Psychiatry 39, 991–999.
Pincus, S. M. (1991). Approximate entropy as a measure of system complexity.
Proceedings of the National Academy of Sciences of USA 88, 2297-2301.
Pincus, T. & Newman, S. (2001). Recall bias, pain, depression and cost in back pain
patients. British Journal of Clinical Psychology 40, 143–156.
Pincus, T., Fraser, L. & Pearce, S. (1998). Do chronic pain patients ‘Stroop’ on pain
stimuli? British Journal of Clinical Psychology 37, 49-58.
Pincus, T., Pearce, S., McClelland, A., Farley, S. & Vogel, S. (1994). Interpretation
bias of ambiguous stimuli in chronic pain patients. Journal of Psychosomatic
Research 38, 347-353.
Polich, J. & Herbst, K. L. (2000). P300 as a clinical assay:rationale, evaluation, and
findings. International Journal of Psychophysiology 38, 3-19.
Polich, J. (2007). Updating P300: An integrative theory of P3a and P3b. Clinical
Neurophysiology 118, 2128-2148.
Purves, D., Augustine, G. J., Fitzpatrick, D., Katz, L. C., LaMantia, A. S. &
McNamara, J. O. (2001). Invitación a la neurociencia. Madrid: Editorial Médica
Panamericana.
Quian-Quiroga, R. (1998). Quantitative analysis of EEG signals: time-frequency
methods and chaos theory. PhD thesis, Institute of Physiology and Institute of
Signal Processing, Medical University of Lübeck, Germany.
Reid, G. J., Lang, B. A. & McGrath, P. J. (1997). Primary juvenile fibromyalgia.
Psychological adjustment, family functioning, coping, and functional
disability. Arthritis & Rheumatism 40, 752-760.
Riba, J., Rodríguez-Fornells, A., Münte, T. & Barbanoj, M. J. (2005). A
neurophysiological study of the detrimental effects of alprazolam on human
action monitoring. Cognitive Brain Resarch 25, 554-565.
Richman J. S. & Moorman J. R. (2000). Physiological time-series analysis using
approximate entropy and sample entropy. American Journal of Physiology:
Heart & Circulatory Physiology 278, H2039-H2049.
Rivera, J., Alegre, C., Ballina, F. J., Carbonell, J., Carmona, L., Castel, B. et al.
(2006). Documento de consenso de la Sociedad Española de Reumatología
sobre la fibromialgia. Reumatología clínica 2, S55-66.
Roelofs, J., Peters, M. L., Maurice P. A. Zeegers, M. P. A. & Vlaeyen, J. W. S. (2002).
The modified Stroop paradigm as a measure of selective attention towards
C. Sitges 158
Schürmann, M., Başar-Eroglu, C., Kolev, V. & Başar, E. (1995). A new metric for
analyzing single-trial event-related potentials (ERPs): application to human
visual P300 delta response. Neuroscience Letters 197, 167-170.
Schürmann, M., Başar-Eroglu, C., Kolev, V. & Başar, E. (2001). Delta responses and
cognitive processing: single-trial evaluations of human visual P300. Journal of
Psychophysiology 39, 229-239.
Schwarz, M. J., Späth, M., Müller-Bardoff, H., Pongratz, D. E., Bondy, B. &
Ackenheil, M. (1999). Relationship of substance P, 5-hydroxyndole acetic acid
and tryptophan in serum of fibromyalgia patients. Neuroscience Letters 259,
196-198.
Shelhamer, M. (2007). Nonlinear dynamics in Physiology. A state-space approach.
London: World Scientific.
Sitges, C., Bornas, X., Llabrés, J., Noguera, M. & Montoya, P. (2008). Linear and
nonlinear analyses of EEG dynamics during non-painful somatosensory
processing in chronic pain patients. European Journal of Pain (in review).
Sitges, C., García-Herrera, M., Pericás, M., Collado, D., Truyols, M. & Montoya, P.
Abnormal brain processing of affective and sensory pain descriptors in chronic
pain patients. Journal of Affective Disorders 104, 73-82.
Sletvold, H., Stiles, T. C. & Landro, N. I. (1995). Information processing in primary
fibromyalgia, major depression and healthy controls. Journal of
Rheumatology 22, 137–142.
Smith, N. K., Larsen, J. T., Chartrand, T. L., Cacioppo, J. T., Katafiasz, H. A. &
Moran, K. E. (2006). Being bad isn't always good: Affective context moderates
the attention bias toward negative information. Journal of Personality &
Social Psychology 90, 210-220.
Sociedad Española del Dolor (2003). No sufras en silencio. Extraído el 26 de octubre,
2004 de http://www.sedolor.es/00_php/IE/sedolor.php?id=00000&cat=0&sec=
Stam, C.J. (2005). Nonlinear dynamical analysis of EEG and MEG: Review of an
emerging field. Clinical Neuropsychology 116, 2266-2301.
Staud, R. & Domingo, M. A. (2001). Evidence for abnormal pain processing in
fibromyalgia syndrome. Pain Medicine 2, 208-215.
Staud, R. & Spaeth, M. (2008). Psychophysical and neurochemical abnormalities of
pain processing in fibromyalgia. CNS Spectrums 13, S12-17.
Staud, R. (2002). Evidence of involvement of central neural mechanisms in
generating fibromyalgia pain. Current Rheumatology Reports 4, 299-305.
C. Sitges 160
Staud, R., Robinson, M. E., Vierck Jr., C. J., Cannon, R. C., Mauderli, A. P. & Price,
D. D. (2003). Ratings of experimental pain and pain-related negative affect
predict clinical pain in patients with fibromyalgia syndrome. Pain 105, 215-
222.
Staud, R., Vierck, C. J., Cannon, R. L., Mauderli, A. P. & Price, D. D. (2001).
Abnormal sensitization and temporal summation of second pain (wind-up) in
patients with fibromyalgia syndrome. Pain 91, 165-175.
Staud, R., Vierck, C. J., Robinson, M. E. & Price, D. D. (2006). Overall fibromyalgia
pain is predicted by ratings of local pain and pain related negative affect:
possible role of peripheral tissues. Rheumatology 45, 1409-1415.
Stevens, A., Batra, A., Kötter, I., Bartels, M. & Schwarz, J. (2000). Both pain and EEG
response to cold pressor stimulation occurs faster in fibromyalgia patients
than in control subjects. Psychiatry Research 97, 237-247.
Stucky, C. L., Gold, M. S. & Zhang, X. (2001). Mechanisms of pain. Proceedings of the
National Academy of Sciences 98, 11845–11846.
Sullivan, M. J., Thorn, B., Haythornthwaite, J. A., Keefe, F., Martin, M., Bradley, L.
A. et al. (2001). Theoretical perspectives on the relation between
catastrophizing and pain. The Clinical Journal of Pain 17, 52–64.
Sutton, S., Braren, M., Zubin, J. & John, E. R. (1965). Evoked-potential correlates of
stimulus uncertainty. Science 150, 1187-1188.
Thelen, E. & Smith, L. B. (2000). A dynamic systems approach to the development of
cognition and action. Cambridge, Massachusetts: MIT Press/Bradford book.
Torres, M. E., Anino, M. M., Gamero, L. G. & Gemignani, M.A. (2001). Automatic
detection of slight changes in nonlinear dynamical systems using
multiresolution entropy tools. International Journal of Bifurcation & Chaos
11, 967–981.
Tortora, G. J. & Grabowski, S. R. (1998). Principios de anatomía y fisiología. Madrid:
Harcourt Brace.
Tracey, I. & Mantyh, P. W. (2007). The cerebral signature for pain perception and its
modulation. Neuron 55, 377-391.
Tracey, I. (2005). Nociceptive processing in the human brain. Current Opinion in
Neurobiology 15, 478-487.
Urban, M. O. & Gebhart, G. F. (1999). Supraspinal contributions to hyperalgesia.
PNAS 96, 7687-7692.
Vallejo, M. A. (2000). Emociones y dolor. Revista Sociedad Española del Dolor 7, 3-5.
Estudio de la actividad cerebral en pacientes con fibromialgia 161
Yordanova, J., Devrim, M., Kolev, V., Ademoglu, A. & Demiralp, T. (2000). Multiple
time-frequency components account for the complex functional reactivity of
P300. NeuroReport 11, 1097-103.