The Fibromyalgia Impact Questionnaire: A Useful Tool in Evaluating Patients with PostLyme Disease Syndrome

John Fallon, Danuta I. Bujak, Sara Guardino, and Arthur Weinstein

Objective. To determine the reliability and validity of a modied version of the Fibromyalgia Impact Questionnaire (FIQ) in evaluating patients with postLyme disease syndrome (PLDS). Methods. In this cross-sectional analysis 13 PLDS, 18 bromyalgia (FM), and 16 healthy controls (n 47) completed a modied FIQ containing items to evaluate physical impairment, symptom severity, and global well-being. Comparisons between groups were done using analysis of variance with a signicance level set at 0.05. Results. PLDS patients demonstrated statistically signicantly greater levels of impairment than controls in physical functioning, FIQ total score, global well-being, joint pain, fatigue, depression, ability to perform activities of daily living, and memory/concentration. FM patients demonstrated a statistically signicantly greater level of impairment than the control group in all categories, and the scores were signicantly higher than the PLDS group in the measurement of physical impairment, FIQ total score, muscle pain, and joint pain. Overall, the instrument possesses good reliability and validity, although adequacy of this instrument to measure impairment in the male PLDS population needs further elucidation. Conclusion. The results of this study suggest that the modied FIQ may be a useful tool in evaluating PLDS patients. The ndings suggest that there may be some differences in the etiopathology of the symptoms experienced by PLDS and FM patients. Key words. PostLyme disease syndrome; Fibromyalgia; Fibromyalgia Impact Questionnaire; Lyme disease.

INTRODUCTION
The persistence of arthralgia, myalgia, chronic fatigue, depression, and neurocognitive symptoms in a subset of Lyme disease patients, despite even repeated or prolonged antibiotic treatment, has been well documented (1 8). Several investigators have noted the similarity between these chronically symptomatic patients with postLyme disease syndrome (PLDS) and patients diagnosed with bromyalgia (FM) (2,3). As part of a larger study investigating the pathogenesis of PLDS, we examined the validity and reliability of a modied version of the Fibromyalgia Impact Questionnaire (FIQ; 9) as a tool for the clinical evaluation of patients with PLDS. In particular, we were interested in whether this form of the FIQ could be used to evaluate physical impairment, symptom severity,
0893-7524/99/$5.00

Supported by grants from NIAMS (1RO1AR-4313) and NIAID (NO1-AI-65308). John Fallon, MS, FNP, The Leinhard School of Nursing, Pace University, Pleasantville, New York; and Danuta I. Bujak, PhD, FNP (current address: University of Maryland, School of Nursing and School of Medicine, Baltimore), Instructor in Medicine, Sara Guardino, BA, Research Assistant, and Arthur Weinstein, MD (current address: The George Washington University Medical Center, Washington, DC), Professor of Medicine and Division Chief, Division of Rheumatic Diseases and Immunology, New York Medical College, Valhalla, New York. Address correspondence to Arthur Weinstein, MD, Division of Rheumatology, The George Washington University Medical Center, 2150 Pennsylvania Avenue, NW, Washington DC 20037. Submitted for publication December 18, 1997; accepted in revised form June 15, 1998. 1999 by the American College of Rheumatology.

42

Arthritis Care and Research

The FIQ in PostLyme Disease Syndrome 43

and global well-being in these patients and in determining the differences that might be demonstrated between PLDS and FM patients.

PATIENTS AND METHODS

Patients. Patient selection was determined by the standards used for the larger ongoing study of PLDS pathogenesis (10). The enrollment criteria for PLDS patients were: 1) meeting the established 1990 Centers for Disease Control criteria for Lyme disease at time of original illness; 2) having received at least one course of recommended oral or parenteral antibiotics for at least 23 weeks at recommended doses; 3) prior or current positivity to anti-Borrelial antibodies by Western blot; 4) 6-month or longer history of widespread musculoskeletal pain and fatigue that began simultaneously with or after the diagnosis of Lyme disease, without signicant prior history; 5) no evidence of severe clinical depression; and 6) no evidence of active Borrelial infection at evaluation and no other medical condition to account for the symptoms. For FM patients enrollment criteria were: 1) meeting the American College of Rheumatology criteria for bromyalgia (11); 2) no evidence of clinical depression or other systemic illness to account for symptoms; and 3) negative Lyme serology. For healthy controls, the following criteria were used: 1) age- ( 2 years), gender-, and racematched to the PLDS group; 2) in general, good health with no history of Lyme disease or FM; and 3) no symptoms that might indicate the presence of Lyme disease or FM. Instrument selection. The patients completed a modied version of the FIQ (9). The original FIQ consists of 19 items, the rst 10 of which are evaluated collectively to provide a single score of physical impairment. This physical impairment measurement focuses on the ability of patients to perform large muscle tasks during the previous week, with each item being rated on a 0 3 Likert-style scale ranging from always being able to perform a task 0, to never being able to perform a task 3. Item 11 has subjects identify the number of days in the past week that they felt good. Item 12 measures the number of days in the last week that subjects were forced to miss work because of their illness. The remaining 7 items on the original instrument consist of 10cm long horizontal anchored visual analog scales (VAS) on which the subjects rated the impact of

illness during the previous week on their ability to work, the severity of pain, fatigue, how rested they felt after sleeping, stiffness, anxiety, and depression. Subjects were instructed to skip items 12 and 13 (impact of illness on ability to do job) if they did not work outside the home. The principal modications to the FIQ used in this study consisted of the addition of two new visual analog scales one evaluating memory/concentration problems (item 22) and the other the ability to perform essential self-care activities of daily living (ADL) (item 21)and the division of the single scale evaluating pain into two distinct scales, one for muscle pain (item 14) and one for joint pain (item 15). The addition of the memory/concentration question was intended to assess the subjective intensity of the neurocognitive symptoms reported by many Lyme disease and PLDS patients, and to see how the reported level of impairment among PLDS subjects compared with that of FM patients. The two items assessing muscle and joint pain were designed not only to look for differences between PLDS and FM patients, but also to see to what degree patients could distinguish the nature of the pain they were experiencing. The only other structural change made for this study was to place each of the VAS questions on a different page of paper to guard against the answer for one question directly inuencing the answer for another due to direct comparison. The administration and scoring of the FIQ was performed according to the manner prescribed by Burckhardt et al (9). Instructions for completing the FIQ were explained to the subjects by one of the investigators, and the instrument was then selfadministered with an investigator available to help the subject with any questions. The PLDS and FM patients were seen at New York Medical College in 1995; the healthy controls were recruited by the researchers at various sites and were not evaluated at New York Medical College. This study was approved by the Institutional Review Board of New York Medical College as part of the larger study on the pathogenesis of PLDS. Data analysis. All data were coded and entered into a computer database. The data were analyzed using SPSS/PC-V6.0 software (12). Comparisons between patient groups were performed using analysis of variance with a signicance level set at P 0.05, with assessment of differences between individual items, physical impairment score, and FIQ total score (the sum of all items excluding 12 and 13) using the Scheffe post hoc test. Four distinct issues were considered in the ana-

44 Fallon et al

Vol. 12, No. 1, February 1999

Table 1. Demographic characteristics of participants (post-Lyme disease syndrome [PLDS], fibromyalgia [FM], and controls)
PLDS (n 13) Men/women Race Caucasian African American 5/8 12 (92.3%) 1 (7.7%) FM (n 18) 0/18 17 (94.4%) 1 (5.6%) Controls (n 16) 8/8 14 (87.5%) 2 (12.5%) 45.3 (2267) 16.2

impairment than the healthy controls in terms of physical impairment, global well-being, and FIQ total scores. In addition, the use of the alpha if item deleted provided additional insight into whether certain items on the FIQ might not have as much content or construct validity as would be desired.

RESULTS
In this cross-sectional analysis, 13 PLDS, 18 FM, and 16 healthy controls (age-, gender-, and racematched to the PLDS patients) (n 47) completed a modied version of the FIQ. Patients and controls. The demographic characteristics of participants (n 47) are summarized in Table 1. There were no statistically signicant differences among the 3 groups with regard to age, race, or education. As expected, the FM group differed from the other two groups in terms of gender, as all participants were women. In addition, the FM group differed from the PLDS group on illness duration, as mean length of time since disease onset was 7.95 versus 3.8 years for the latter. Results of the FIQ testing are summarized in Tables 2 and 3. PLDS patients demonstrated statistically signicantly higher scores (P 0.05) than controls in physical functioning, FIQ total score, global well-being, and VAS measures of joint pain, fatigue, depression, ability to perform activities of daily living, and impairment of memory and concentration, suggesting greater impairment. Differences between the two groups were also seen in severity of muscle pain, feeling rested after sleeping, stiffness, and anxiety, but these were not statistically signicant. FM patients demonstrated a statistically signicantly greater level of impairment than the control group in all categories, and the scores were signicantly higher than the PLDS group (P 0.05) in the measurement of physical impairment, FIQ total score,

Age, mean (range) 43.1 (2067) 46.6 (3563) Education, years, mean 15.8 15 Symptoms duration, years, mean (range) 3.8 (312.8)* 7.95 (130)*
* t-test P 0.01. NA not applicable.

NA

lysis and interpretation of the results: 1) the degree to which the FIQ possessed internal and external reliability for this patient population, 2) the extent to which the instrument possessed content and construct validity, and 3) the similarities and differences seen between FM and PLDS patients with this instrument. Reliability. Internal consistency of the FIQ for this cross-sectional study was assessed using Cronbachs alpha. In order to check for variance between items on the FIQ, a standardized item alpha was also generated as was an alpha if item deleted test. This was done in order to examine the reliability of the FIQ when each of the items was removed from the scale (13). Validity. In the original development and validation of the FIQ, the percentage of missing data for individual items on the instrument was used as the criteria for evaluating content validity (9). In addition to using the same criteria for our analysis, we assessed the degree of impairment reported by PLDS patients as compared with that experienced by the FM patients and the healthy controls. Construct validity of the FIQ for a FM population was determined to be satisfactory in the original study through factor analysis and correlations between items on the FIQ and items on the Arthritis Impact Measurement Scales instrument (9). Because of the similarities between the symptomatic manifestations of FM and PLDS, we chose to evaluate construct validity by examining the working hypothesis that the PLDS patients would demonstrate a statistically signicantly (P 0.05) higher level of

Table 2. Fibromyalgia Impact Questionnaire (FIQ) scores*

PLDS (n 13) 50.12 2.23 5.28 FM (n 18) 98.03 4.88 7.23 Controls (n 16) 14.02 0.07 0.72

FIQ FIQ total score Physical impairment (total) Global well-being

* Analysis of variance 0.05. PLDS post-Lyme disease syndrome; FM bromyalgia. Signicantly different from controls. Signicantly different from PLDS and controls.

Arthritis Care and Research

The FIQ in PostLyme Disease Syndrome 45

Table 3. Fibromyalgia Impact Questionnaire visual analog scale (VAS) scores by group*
PLDS (n 13) 3.33 4.15 6.62 5.34 4.58 3.42 4.92 4.62 5.62 FM (n 18) 6.62 7.01 7.74 7.52 6.70 4.99 4.31 6.18 5.74 Controls (n 16) 1.03 0.85 2.98 2.76 1.91 1.31 0.90 0.73 0.76

ing, global well-being, and symptom severity than the healthy controls.

VAS item no. (010 cm) 14. 15. 16. 17. 18. 19. 20. 21. 22. Muscle pain Joint pain Fatigue Rest Stiffness Anxiety Depression ADL impairment Memory/concentration problem

DISCUSSION
This is the rst study that has used the FIQ to evaluate patients with PLDS. The FIQ was selected as a tool for this study for several reasons. First, the instrument contains items to assess the broad range of symptomatic distress frequently seen in patients with PLDS: physical impairment, diminished global well-being, arthralgia, myalgia, fatigue, difculty performing ADLs, stiffness, anxiety, depression, and memory/concentration problems. Second, the FIQ was designed specically for use in the clinical assessment of FM patients and has established test retest reliability as well as content and construct validity for a specic patient population, and thus would seem to be an ideal choice for an instrument to examine the similarities and differences between spontaneous FM and the bromyalgia-like symptoms seen in PLDS patients. Third, it was attractive as a concise tool that could be self-administered and completed by subjects within approximately 5 minutes. External reliability of the FIQstability of the instrument determined through testretest correlationswas established by Burckhardt et al in their original development and validation study on FM patients (9). Initial studies have indicated that PLDS patients tend to have chronic symptoms that persist over months and years similar to those observed in FM (1,3,10). Therefore, the testretest reliability of the FIQ was deemed sufciently well established to justify extrapolation to the PLDS patient population for the purposes of this study. The internal consistency of the instrument using Cronbachs alpha indicated that items 12 and 13 were not providing reliable information. In analyzing these ndings we believe these items were problematic for two reasons. The rst had to do with interpretation of the questions by subjects. Although subjects were instructed not to respond to items 12 or 13 if they did not work outside the home, some of those who did not work provided a number of days missed from work in the last week. The second problem had to do with the time limitation of one week in item 12, as a number of patients commented that their illness had caused some change in their ability to work but that this change was not necessarily reected in the number of days missed from work in the past week. Burckhardt agrees that item 12 should be deleted from the instrument because it seems to

* Analysis of variance P 0.05. PLDS post-Lyme disease syndrome; FM bromyalgia; ADL activity of daily living. Signicantly different from controls. Signicantly different from PLDS and controls.

muscle, and joint pain. The severity of symptoms did not correlate with their duration (P 0.05). Reliability. The overall alpha coefcient with all items included was 0.72 (n 35). Twelve cases were dropped from analysis due to lack of data for item 12 (subjects did not work outside home). An additional analysis showed that excluding item 12 only raised the alpha to 0.74 when all 47 cases were included, but excluding only item 13 (impact of illness on ability to do job) raised the alpha to 0.93 (12 cases dropped). Excluding both items 12 and 13 improved internal reliability by raising the alpha to 0.94 (all 47 cases analyzed). While analysis of variance showed no statistically signicant difference between PLDS, FM, and healthy controls in regard to days of work missed (item 12), both the PLDS and FM groups demonstrated a statistically signicantly higher level of impairment in their ability to do their jobs (item 13). Content validity. Of the 10 items comprising the physical impairment subscale (items 110), only item 9, ability to do yardwork, and item 4, ability to wash dishes, had more than 2 missing cases. A total of 9 subjects (19.1%) did not do yard work, with 1 PLDS subject, 5 FM subjects, and 3 controls leaving this item blank. Four subjects, 3 FM and 1 control, did not wash dishes. Construct validity. Construct validity of this modied FIQ for the PLDS population was supported by conrmation of our hypothesis that these patients would demonstrate a statistically signicantly greater degree of impairment in physical function-

46 Fallon et al

Vol. 12, No. 1, February 1999

lack the accuracy and sensitivity to provide a good correlation with symptom severity (Burckhardt CS: personal communication). The internal consistency of the instrument also improved signicantly with the deletion of item 13. It is likely that the misinterpretation of items 12 and 13 by some subjects contributed to this nding. Because our study was initiated with items 12 and 13 we did not change the method of administration as neither item is included in the nal summation of scores that produces a total FIQ score. Overall, the FIQ demonstrated acceptable content validity in this study. However, several male patients commented that the tasks comprising the physical functioning item were not truly representative of the gross motor tasks that they considered important in their daily or weekly routine. This was important because of the gender differences in the patient populations affected by FM and PLDS. Whereas more than 87% of FM patients tend to be women (12), and the development and validation of the FIQ was done with all-female samples (9), the PLDS population does not appear to have as pronounced a gender skew (1). In this study, while all the FM patients were women, 5 of the 13 PLDS patients (38.5%) were men. Analysis of variance among all 3 groups on the physical functioning item indicated that for several of the tasks listed, the PLDS population experienced relatively little impairment. On item 2 (ability to do laundry with a washer and dryer), item 4 (ability to wash dishes/cooking utensils by hand), and item 6 (ability to make beds), there was no statistically signicant difference between the PLDS and control groups. By contrast, the FM group reported a statistically signicantly greater amount of impairment than the controls for every task, and were signicantly more impaired than PLDS patients in their ability to wash dishes (item 4), vacuum rugs (item 5), make beds (item 6), walk several blocks (item 7), visit friends and relatives (item 8), and do yard work (item 9). However, the PLDS subjects demonstrated a statistically signicantly greater amount of impairment than the control group in physical functioning, and the content validity of this item of the FIQ is supported by the small amount of missing data. Because a number of the items may be gender-biased, it remains uncertain whether the physical functioning score of the FIQ in its current form is sufciently sensitive for evaluating men with PLDS. However, only one man in the PLDS group omitted a possibly gender-biased item. Our ndings indicate that the modied version of the FIQ used in this study may be a useful tool in

measuring the impairment in physical functioning, global well-being, and symptom severity associated with PLDS. The FM subjects demonstrated greater levels of impairment than the PLDS patients, with statistically signicant differences seen in physical impairment, higher FIQ total scores, and the severity of muscle and joint pain. Only on item 20, which measures depression, did the PLDS patients score higher than the FM patients, but the observed difference was statistically not signicant. These results support the concept that PLDS is similar to FM but is milder in its clinical manifestations. However, the items added to our modied version of the FIQ did illustrate some differences between these patients. For example, while the FM patients experienced a statistically signicantly greater amount of both muscle and joint pain than the PLDS patients, the PLDS subjects experienced more joint than muscle pain. The reason for this difference is unclear. On the other hand, the overall similarity of scores might reect a similar basis for their symptoms, and the difference may be related to the duration of these symptoms. The performance on two items added to the FIQ (the interference with ADLs and the severity of memory/concentration problems) showed no statistically signicant difference between the FM and PLDS groups, although the FM subjects demonstrated more impairment than PLDS subjects. The relatively greater amount of difculty experienced by FM patients in performing ADLs was in keeping with the overall higher levels of impairment reported by these subjects. While subjective symptom reports and objective measures of neurocognitive function may be abnormal in Lyme disease (14) and in patients with PLDS (2,15), neurocognitive impairment is not a dening feature of FM, but abnormalities have been described (16). In analyzing the results of this study, two factors must be kept in mind. The rst is the question of construct validity of the FIQ in evaluating the physical impairment experienced by PLDS patients. To determine this, the known-groups technique was used (13). Contrasting the scores of the PLDS, FM, and healthy control groups that differed on the measured attributes allowed us to show that the intergroup differences can be assessed with this instrument. It would also be interesting for future longitudinal studies to look at the issue of testretest reliability for this instrument with the PLDS population and to help determine the stability of the instrument to measure PLDS symptoms over time. This is being measured at present in a prospective

Arthritis Care and Research

The FIQ in PostLyme Disease Syndrome 47

study of patients with early Lyme disease at our institution (17). Second, it is important to bear in mind that the FIQ was not created with the intention of developing a diagnostic set of norms for patients with bromyalgia, according to Burckhardt (Burckhardt CS: personal communication). Rather, Burckhardt stated that the FIQ was designed as an aid to clinical evaluation for the purposes of assessing therapeutic interventions for bromyalgia (9). Our results suggest that this instrument in its current form can provide the same function for PLDS patients. It is anticipated that further clinical studies already in progress will help to clarify these issues and provide greater insight into the differences seen between FM and PLDS patients.

7.

8.

9.

10.

11.

We thank Dr. Barbara Orlowski for assistance with statistical analysis and Dr. Lillie M. Shortridge-Baggett for assistance with the preparation of the manuscript.

12. 13.

REFERENCES
1. Asch ES, Bujak DI, Weiss M, Peterson MGE, Weinstein A. Lyme disease: an infectious and postinfectious syndrome. J Rheumatol 1994;21:454 61. 2. Bujak DI, Weinstein A, Dornbush RL. Clinical and neurocognitive features of the post Lyme syndrome. J Rheumatol 1996;23:13927. 3. Dinerman H, Steere AC. Lyme disease associated with bromyalgia. Ann Intern Med 1992;117:2815. 4. Rahn DW, Malawista SE. Lyme disease. In: Koopman WJ, editor. Arthritis and allied conditions. 13th ed. Baltimore: Williams and Wilkins; 1997. p. 232139. 5. Logigian EL, Kaplan RF, Steere AC. Chronic neurologic manifestations of Lyme disease. N Engl J Med 1990;22:1438 44. 6. Pster HW, Preac-Mursic V, Wilske B, Schielke E, Sorgel F, Einhaupl KM. Randomized comparison of

14.

15.

16.

17.

ceftriaxone and cefotaxime in Lyme Neuroborreliosis. J Infect Dis 1991;163:311 8. Steere AC, Dwyer ED, Winchester R. Association of chronic Lyme arthritis with HLA-DR4 and HLA-DR2 alleles. N Engl J Med 1990;323:219 23. Wahlberg P, Granlund H, Nyman D, Panelius J, Seppala I. Treatment of late Lyme Borreliosis. J Infect Dis 1994;29:255 61. Burckhardt CS, Clark SR, Bennett RM. The Fibromyalgia Impact Questionnaire: development and validation. J Rheumatol 1991;8:728 33. Weinstein A, Dornbush RL, Guardino SD, Bujak DI, Norton D, Masdeu J, et al. Post Lyme syndrome and bromyalgia share clinical, psychological, and neurocognitive features [abstract]. Arthritis Rheum 1997;40 Suppl 9:S45. Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Goldenberg DL, et al. The American College of Rheumatology 1990 criteria for the classication of bromyalgia: report of the multicenter criteria committee. Arthritis Rheum 1990;33:160 72. SPSS Inc. SPSS for Windows: base systems users guide, release 6.0. Chicago: SPSS Inc.; 1993. Polit D, Hungler B. Nursing research: principles and methods. 5th ed. New York: JB Lippincott; 1995. Halperin JJ, Pass HL, Anand AK, Volkman DJ, Dattwyler RJ. Nervous system abnormalities in Lyme disease. Ann N Y Acad Sci 1988;539:24 34. Kaplan RF, Meadows ME, Vincent LC, Logigian EL, Steere AC. Memory impairment and depression in patients with Lyme encephalopathy: comparison with bromyalgia and nonpsychotically depressed patients. Neurology 1992;42:12637. Mountz JM, Bradley LA, Modell JG, Alexander RW, Triana-Alexander M, Aaron LA, et al. Fibromyalgia in women: abnormalities of regional cerebral blood ow in the thalamus and the caudate nucleus are associated with low pain threshold levels. Arthritis Rheum 1995;38:926 38. Bujak DI, Weinstein A, Weinberg S. Cognitive appraisal as a predictor of outcome in early Lyme disease [abstract]. Arthritis Rheum 1997;40 Suppl 9:S222.