The British Journal of Psychiatry

1–8. doi: 10.1192/bjp.bp.115.171397

Effectiveness of group body psychotherapy

for negative symptoms of schizophrenia:
multicentre randomised controlled trial
S. Priebe, M. Savill, T. Wykes, R. P. Bentall, U. Reininghaus, C, Lauber, S. Bremner,
S. Eldridge and F. Röhricht

Background Results
Negative symptoms of schizophrenia have a severe impact In total, 275 participants were randomised. The adjusted
on functional outcomes and treatment options are limited. difference in negative symptoms was 0.03 (95% CI –1.11 to
Arts therapies are currently recommended but more 1.17), indicating no benefit from body psychotherapy. Small
evidence is required. improvements in expressive deficits and movement disorder
symptoms were detected in favour of body psychotherapy.
No other outcomes were significantly different.
Aims
To assess body psychotherapy as a treatment for negative Conclusions
symptoms compared with an active control (trial registration: Body psychotherapy does not have a clinically relevant
ISRCTN84216587). beneficial effect in the treatment of patients with negative
symptoms of schizophrenia.
Method Declaration of interest
Schizophrenia out-patients were randomised into a None.
20-session body psychotherapy or Pilates group. The primary
outcome was negative symptoms at end of treatment. Copyright and usage
Secondary outcomes included psychopathology, functional, B The Royal College of Psychiatrists 2016. This is an open
social and treatment satisfaction outcomes at treatment end access article distributed under the terms of the Creative
and 6-months later. Commons Attribution (CC BY) licence.

Schizophrenia is a severe mental health disorder that affects activity, and all body psychotherapy groups were conducted by the
approximately 0.7% of the population.1 Symptoms include same therapist. Three earlier trials on body-oriented psycho-
positive symptoms such as hallucinations, disordered thinking therapy not included in the NICE review suggested improvements
and delusions, and negative symptoms that include expressive in various outcomes including negative symptoms,10–12 however,
deficits such as blunted affect and impoverished speech, and all had significant methodological shortcomings.
experiential deficits such as asociality, anhedonia, and avolition.2,3 There are a number of advantages to evaluating this particular
Negative symptoms have been found to have a profound impact form of arts therapy as a treatment for schizophrenia. First, it is
on long-term outcomes,4,5 but current treatment options are recognised that patients with schizophrenia can experience a range
limited. In a review by the National Institute for Health and Care of body disturbances such as desomatisation, abnormal bodily
Excellence (NICE) in the UK,6 arts therapies – an umbrella term sensations and motor impairments.13,14 Consequently, providing a
for all non-verbal creative therapies such as art therapy, music form of therapy that focuses on the body may help to address such
therapy and body psychotherapy – were identified as the only type disturbances. Second, to our knowledge this is the only form of arts
of therapy with justified claims to reduce negative symptoms. therapy where a treatment manual specific to the treatment of
Consequently, it was recommended that clinicians should consider negative symptoms has been produced that details a theoretical
referring people with schizophrenia for arts therapies.6,7 However, model, mode of action and a standardised therapy structure. Beyond
the review was based on only six small-scale trials, meaning more its possible clinical effectiveness, body psychotherapy is relatively
evidence is needed. Since the publication of NICE guidelines one inexpensive, can be combined flexibly with other treatment methods,
large trial of conventional art therapy has been completed and may appeal to patients who are difficult to engage in other
(MATISSE) that found no significant treatment effect on negative treatments given its novel approach. In order to examine the
symptoms.8 Following MATISSE, the aim of the present study was effectiveness of body psychotherapy as a treatment for negative
to evaluate the effectiveness of a different type of arts therapy, symptoms we conducted a full-scale, randomised controlled trial
namely body psychotherapy, as a treatment for negative symptoms (RCT) comparing a manualised form of the intervention with a
of schizophrenia. Body psychotherapy is a form of therapy that well-defined, physically active control condition, namely Pilates.
involves an explicit theory of body–mind functioning designed Pilates is a structured physical fitness programme involving
to improve emotional, cognitive, physical and social integration. stretching and controlled movement. The specific components
In an earlier trial where this therapy was evaluated,9 a significant of body psychotherapy under investigation were the focus on body
reduction in negative symptoms was detected in the body experience at a cognitive and emotional level, the facilitation of
psychotherapy group in comparison with a supportive counselling emotional group interactions, and the link between movement
control group. The effect size was large, and was maintained months and emotion. The components common to both interventions
later. However, this study was relatively small (45 participants), did include the non-specific effects on non-emotional group
not control for the non-specific effects of supported group physical interactions, group facilitator attention and physical activity.

1
Priebe et al

end of the study to ensure scores remained sufficiently concordant

Method
throughout, using videotapes of the assessments.
Design and participants
This study was an assessor-masked, two-arm, RCT, approved by Procedure
the Camden and Islington National Research Ethics Committee
Potential participants were approached by their clinicians for
(Ref:H0722/44), and the trial is registered (ISRCTN84216587). A
consent to be contacted by a researcher. If they agreed, the
detailed study design description is available in the published
researcher arranged a meeting during which an explanation of
protocol.15 Participants were recruited from mental health
the study was provided, informed consent obtained and an
community services in the UK. The inclusion criteria were:
eligibility assessment completed. Once approximately 16 eligible
diagnosis of schizophrenia (ICD-10 codes F20.0–F20.9),16 aged
participants were recruited a full baseline assessment, which
18–65 years, a Positive and Negative Syndrome Scale (PANSS)
included a second PANSS assessment, was undertaken within
negative subscale score 518,17 no change of antipsychotic
1 month prior to the group start date. The assessments were
medication for 6 weeks, a willingness and ability to consent
typically conducted in the participants’ home, or the local
and participate, and a sufficient command of English to
community mental health team site, and took 40–120 min to
complete the research interviews and actively participate in group
complete. Once all baseline assessments were completed a list of
interactions in English. Participants were randomised into a
identification codes was sent to the PCTU via the trial manager
manualised, 20-session body psychotherapy group, or a 20-session
for randomisation, approximately 1 week before the groups were
beginner’s-level Pilates class, in addition to standard care.
going to start. Participants were then notified of their group
allocation by the relevant group facilitators. After group
completion participants were assessed at end of treatment, within
Randomisation and masking
1 month of the groups’ completion, and again 6 months later.
Randomisation was conducted by the Pragmatic Clinical Trials Participants were paid £25 expenses for each assessment attended.
Unit (PCTU) independently through a computer-generated
sequence. Participants were randomly allocated, with equal
probability, to the intervention or control group, stratified by Experimental and control conditions
study centre, in batches using randomly permuted blocks of four The treatment under investigation was body psychotherapy, as
and six, starting each batch at the start of a new block to preserve outlined in the manual.9,28 Body psychotherapy has a long
balance. The chief investigator, all assessors and the trial tradition in psychiatry, going back to the beginning of the 20th
statistician were masked to the treatment allocation until all century, and has been influenced by psychodynamic psycho-
end-of-treatment data were collected and the statistical analysis therapies, dance movement psychotherapy, and techniques
plan was signed off. To maintain masking, baseline assessments designed to address body image disturbances.29
took place prior to randomisation. The main goals of body psychotherapy as a treatment for
negative symptoms in chronic schizophrenia are: to reconstruct
a coherent ego structure through grounding and bodily awareness;
Outcomes to strengthen self-referential processes as a prerequisite for safe
Outcomes were assessed at baseline, end of treatment and social interaction and reality testing; to widen and deepen the
6 months after treatment completion. The primary outcome was range of emotional responses to environmental stimuli; to
the PANSS negative subscale17 at the end of treatment. Secondary improve boundary demarcation, enabling differentiation between
outcomes were general psychopathology and positive symptoms self and other; and to help patients explore a range of expressive
measured with the PANSS;17 the Clinical Assessment Interview and communicative behaviours with the aim of reducing
for Negative Symptoms (CAINS)3 expression and experience emotional withdrawal and improving prosocial capabilities
subscales; subjective quality of life using the Manchester Short Each session comprised five discrete sections. The first section
Assessment of Quality of Life (MANSA);18 objective social aimed to facilitate communication between patients, and draw
situation using the SIX;19 depression using the Calgary scale;20 patients’ focus towards the body. The second section focused on
the number of social contacts using the Social Network Scale physical experiences and movements, exploring the personal and
(SNS);21 and a measure of patient activity using four items from general space. The third section addressed specific body image
the Time Use Survey (TUS).22 All were completed at each disturbances such as boundary loss and desomatisation. The
assessment point. Treatment satisfaction was measured at end of fourth section centred on creativity and tasks requiring patients
treatment using the Client Satisfaction Questionnaire (CSQ).23 to use their bodies and movement as a source of expression and
Extrapyramidal symptoms (EPS) were evaluated at each stage pleasure. In the final section, patients reflect on events, thoughts
using the Simpson Angus Scale (SAS),24 however, given logistical or feelings that may have been brought up by the group.
constraints three items were not assessed (leg pendulousness, head Both body psychotherapy and Pilates groups were delivered in
dropping and glabella tap). Given evidence that suggests that the 20 sessions of 90 min each, over a 10-week period, held twice a
PANSS negative subscale includes some items that relate to week on non-consecutive days. This duration of treatment was
cognitive, rather than negative symptoms, the alternative PANSS deemed appropriate given the therapy had been manualised for
Marder negative symptom subscale was also evaluated.25 For an 20 sessions, was long enough to result in significant medium
economic evaluation of the intervention (which will be reported and large treatment effects in two recent trials of body psycho-
elsewhere), data were obtained using the Client Service Receipt therapy,9,30 and in a review on music therapy 16 sessions were
Inventory (CSRI)26 and the EQ-5D.27 Data from the CSRI were sufficient to result in medium-effect size improvements in
used to calculate the defined daily dose (DDD) of prescribed negative symptoms.31 Groups contained between seven and ten
antipsychotic medication. All researchers were trained in participants. To limit the impact of any one therapist or instructor
conducting the full PANSS assessment prior to assessing patients. each one ran a maximum of two groups.
The interrater agreement between all researchers conducting Each body psychotherapy group was facilitated by an
PANSS interviews was assessed at the beginning, middle and Association of Dance Movement Psychotherapy (ADMP)

2
 Group body psychotherapy for negative symptoms of schizophrenia

accredited therapist trained to deliver the manualised inter- groups was low; however, the rate of screening required to identify
vention, supported by a co-facilitator. Each therapist received eligible participants was higher than anticipated (see Fig. 1). This
three supervision sessions held by a senior therapist per group. was because of a number of factors. First, a relatively large number
Adherence to the manual was assessed using an adherence scale of screened patients were found to be ineligible, because of either
that we produced for this study (see online supplement DS1) by an incorrect diagnosis or insufficient negative symptoms. Second,
therapists evaluating four randomly selected sessions of each once all group therapy/control places were provisionally filled,
group (one from each quartile). The scale considered both the no more participants were approached unless a participant
content of the sessions, assessing whether therapists adhered to subsequently dropped out. Consequently, a number of
the format and utilised the techniques and objects appropriately, participants were initially screened as potentially eligible, but were
and their competence, assessing their ability to foster a cohesive not approached as a result of the lack of available spaces on the
therapeutic environment and their ability to translate the activities trial in their area. Of those randomised, 266 (96.7%) were assessed
undertaken as strategies to address specific negative symptoms. at end of treatment, and 255 (92.7%) went on to complete the
The active control condition was beginner’s Pilates classes, 6-month follow-up.
which was described to participants as a physical health and fitness The baseline characteristics of the sample are presented in
intervention. All classes were facilitated by a Register of Exercise Table 1. Participants presented with moderate levels of negative
Professionals (REPS) level-three qualified Pilates instructor, symptoms (PANSS negative score 23.1, s.d. = 4.4). The mean level
assisted by a co-facilitator. Prior to starting, instructors received of interrater reliability for the PANSS was high (PANSS total
a brief training session from an experienced clinician. A brief intraclass coefficient 0.85). Assessor masking was maintained prior
Pilates guide was developed based on the Pilates Union Matwork to the primary outcome assessment in 94.3% of cases.
Manual.32 The guide provided a summary of how to run the Participants attended significantly more body psychotherapy
groups, and a loosely structured exercise plan. Props (other than sessions (body psychotherapy median 11, interquartile range
mats and head blocks), music and activities designed to encourage (IQR) = 5–17; Pilates median 8, IQR = 1–15; P = 0.01). In total,
group interactions was not permitted. Group interactions were 106 participants (75.7%) attended at least five body psychotherapy
expected to occur as they commonly do when people conduct sessions, the level defined as adhering to treatment in the CACE
activities in a group. However, instructors were advised not to analysis. Therapist adherence to the manual was relatively high
initiate or promote any interactions. with a mean score of 17.6 (out of 20; s.d. = 0.21).

Analysis plan Primary outcome

A 20% reduction in the PANSS score has been used as an indicator Outcomes are shown in Table 2. There was a small reduction in
of clinically significant improvement in the past, which would be a mean PANSS negative symptoms between baseline and end of
difference of approximately three points given the eligibility treatment in both groups (within-group mean reduction in the
criteria. To detect this difference with a standard deviation of 5, body psychotherapy group 1.5 (s.d. = 3.5); Pilates group 1.6
with 90% power for 5% significance, 58 patients were required (s.d. = 3.8)). After controlling for baseline scores, study centre
in each arm. To allow for clustering by group, an intracluster and therapy group, no significant difference between the
correlation coefficient (ICC) for treatment group of 0.1, and seven experimental and control condition was detected (adjusted mean
patients per group with analysable data at the end of treatment difference = 0.03, 95% CI 71.11 to 1.17, P = 0.959. Model-based
gives an inflation factor of 1.6, meaning 93 participants in each ICC = 0.099).
arm were required. At 6 months we anticipated a 31% drop-out,
so recruiting 256 participants would leave 88 per arm at 6 months,
and 91% power to detect a difference of three points. A total of Secondary outcomes
128 patients per arm, i.e. 16 groups of approximately 8 patients A significant mean difference reduction in the SAS (70.65, 95%
in each arm, gave 94% power for the end-of-treatment analysis, CI 71.13 to 70.16, P = 0.009, ICC50.001), which measures
assuming 87.5% of patients have analysable data. Estimates for extrapyramidal symptoms, and the CAINS expression subscale
the standard deviation ICC for treatment and study drop-out were (70.62, 95% CI 71.23 to 0.00, P = 0.049, ICC = 0.022), which
based upon the findings from the exploratory trial.9 measures asociality, anhedonia and avolition was detected in the
The primary outcome was the PANSS negative subscale at body psychotherapy arm in comparison with the Pilates group
the end of treatment, using an available case analysis following at the end of treatment. No other significant differences were
intention-to-treat principles. Mixed-effects models fitted by found at this stage. In an analysis of the multiply imputed data-
restricted maximum likelihood with fixed effects for the inter- sets, no substantial differences in the results were evident,
vention, baseline PANSS negative scores and centre, and random although the reduction in the CAINS expressive subscale was no
effects for therapy groups were used. To evaluate the impact of longer below the P = 0.05 threshold for significance (70.60,
missing data, multiple imputation of the data-set was performed 95% CI 71.22 to 0.02, P = 0.056, ICC = 0.026).
and the analysis replicated. A simple complier-average causal At the 6-month follow up, no significant mean difference in
effect (CACE) analysis was completed,33 defining adherence as the PANSS negative score was detected between conditions
attending at least five body psychotherapy sessions. Planned (70.18, 95% CI 71.68 to 1.31, P = 0.812, ICC = 0.137). There
subgroup analyses examined whether there were differences in was a significant mean difference in the SAS at 6-month follow
response between those with higher negative symptoms at base- up (70.50, 95% CI 70.94 to 70.07, P = 0.028, ICC = 0.007),
line, and a longer duration of illness. Analyses were completed using but no other significant differences were detected. In the CACE
Stata version 12. analysis, no significant mean difference in the PANSS negative
score between body psychotherapy and Pilates was detected
Results (70.13, 95% CI 71.41 to 1.64). In the secondary outcomes, only
a significant mean difference in the SAS was detected (70.82, 95%
In total 275 participants were randomised, recruited from CI 71.51 to 70.12). As an exploratory outcome, an additional
December 2011 until June 2013. The study attrition rate for both CACE analysis on the primary outcome was conducted where

3
Priebe et al

1371 patients identified as meeting

inclusion criteria by clinicians

1015 patients excluded

7 534 not contacted/unsuitable
481 patients declined
6
356 patients recruited

81 patients excluded
33 no longer eligible
7 33 no longer wish to take part/
not contactable
6 15 no longer able to take part
Enrolment 275 assessed and
randomised
(34 groups)

6 6
Baseline 140 body psychotherapy 135 Pilates
assessment (17 groups) (17 groups)

3 patients lost 6 patients lost

2 withdrew
8 7
6 6 4 withdrew
Assessment 1 patient deceased 2 not contactable
following 137 end of treatment 129 end of treatment
10-week 137 assessed 127 assessed
treatment 2 not contactable

5 patients lost
2 withdrew
6 patients lost 7 1 not contactable
2 withdrew 8 1 too unwell
4 not contactable 1 moved
Assessment 6 6
6 months
post-treatment 131 6-month follow-up 124 6-month follow-up
131 assessed 124 assessed

Fig. 1 CONSORT diagram.

those adherent to treatment were defined as those that attended at Strengths and limitations
least 10 body psychotherapy sessions, and again no significant dif- The study retention rates were excellent, with 92.7% of
ference was detected (0.15, 95% CI 71.89 to 2.19). In pre- participants remaining in the study until its end. The large sample
planned subgroup analyses, no significant differences in response sizes and minimal drop-out meant the study was highly powered
were detected between patients with higher negative symptoms at to detect a clinically important difference in the primary outcome
baseline, or a longer duration of illness. No serious adverse events (>94%). This, together with the non-significant result suggest that
related to either intervention were reported. these findings go further than just failing to reject the null
hypothesis, and instead can be interpreted as evidence of the
intervention having no clinically important benefit.
Discussion The intraclass coefficient scores on the PANSS between
assessors was high (PANSS intraclass coefficient 0.85), with no
No significant differences between body psychotherapy and evidence of rater drift. Participants randomised to body
Pilates were detected in the PANSS negative symptom subscale. psychotherapy attended a median of 11 sessions, which is
A statistically significant improvement in the body psychotherapy relatively high given participants typically experienced high social
arm was detected in the CAINS expression subscale and withdrawal and motivation deficits. Approximately 40% of
movement disorder symptoms. However, the small effect sizes participants in the body psychotherapy condition attended at least
mean these improvements are unlikely to reflect relevant clinical 75% of the sessions offered, which compares very favourably with
benefits. There was no significant difference on other outcomes. the MATISSE trial evaluating art therapy with a similar patient
Given that the confidence interval excludes a clinically meaningful group.8 The Pilates groups were also well attended, enabling a
difference in negative symptoms on the PANSS, and the high comparison that appropriately controls for the non-specific
statistical power, these results support the conclusion that body effect of regular group activity. This relatively high attendance is
psychotherapy is not an effective treatment for patients with likely to be attributable to the logistical support provided by the
negative symptoms of schizophrenia as compared with Pilates as co-facilitators, which included the provision of taxis to those
an active control. who required it. The body psychotherapy intervention was

4
 Group body psychotherapy for negative symptoms of schizophrenia

Table 1 Descriptive statistics of participant characteristics at baseline, for experimental and control condition
Body psychotherapy group Pilates group Total
Variable (n = 140) (n = 135) (n = 275)

Centre, n (%)
East Londona 41 (29) 40 (30) 81 (29)
North East Londona 8 (6) 8 (6) 16 (6)
South London 36 (26) 32 (24) 68 (25)
Manchester 23 (16) 23 (17) 46 (17)
Liverpool 32 (23) 32 (24) 64 (23)
Age, years: mean (s.d.) 41.1 (10.1) 43.3 (11.1) 42.2 (10.7)
Gender, n (%)
Men 103 (74) 100 (74) 203 (74)
Woman 37 (26) 35 (26) 72 (26)
Ethnicity, n (%)b
White 71 (52) 67 (53) 138 (52)
Black 39 (29) 38 (30) 77 (29)
Asian 13 (9) 16 (13) 29 (11)
Other 14 (10) 6 (5) 20 (8)
Employment, n (%)c
Unemployed 131 (94) 132 (98) 263 (96)
Other 8 (6) 3 (2) 11 (4)
Living situation, n (%)c
Alone 83 (60) 73 (54) 156 (57)
With others 56 (40) 62 (46) 118 (43)
Number of children, median (IQR) 0 (0–1) 0 (0–1) 0 (0–1)
Duration of illness, median (IQR) 12.6 (8.8) 12.7 (9.5) 12.6 (9.1)
Number of hospital admissions, median (IQR) 3.9 (3.8) 4.0 (4.2) 4.0 (4.0)
Medication: defined daily dose, mean (s.d.) 1.48 (1.11) 1.71 (1.28) 1.59 (1.20)

IQR, interquartile range.

a. These two centres were treated as one for the purposes of the stratified randomisation.
b. As a result of missing data total n for the body psychotherapy group is 137, for the Pilates group 127 and for all participants 264.
c. As a result of missing data total n for the body psychotherapy group is 139 and for all participants 274.

manualised and therapists were largely adherent to treatment therapy,31 the relationship between the number of sessions and
guidelines, allowing the intervention to be evaluated as it had been improvements in negative symptoms was curvilinear, with small
designed. effect sizes found in as few as three sessions, and medium effects
One limitation is that in the Pilates groups emotional group in 16 sessions. In the prespecified CACE analysis no significant
interactions, although discouraged, may also have occurred. In differences between the groups were detected when those adherent
addition, although the focus on body experience at a cognitive to treatment were defined as attending at least five sessions. In an
and emotional level may not be explicitly addressed in Pilates, exploratory analysis of our data the difference was also highly
an emphasis on centring, concentration and breathing may have non-significant when this threshold was increased to a minimum
implicitly fostered such links. A link between movement-based attendance of ten sessions. If the lack of effect is attributable to
exercises such as Pilates and mindfulness, which may help address insufficient dose, it would be reasonable to expect at least a trend
negative symptoms, has been proposed.34,35 However, the small towards symptom improvements as participants received more
within-group changes detected suggest that neither group was sessions, however, this was not detected.
effective, as opposed to both being equally effective. A pre–post
reduction of 1.5 points in the PANSS negative subscale was half
the level prespecified as an indicator of clinically meaningful Comparisons with existing literature
change, and was comparable with the 1.3 point reduction found These findings are in contrast to the exploratory trial where
in the supportive counselling group evaluated in the exploratory significant improvements in negative symptoms were found in
trial.9 The reduction found is consistent with changes in the body psychotherapy group compared with supportive
treatment-as-usual study arms in a recent meta-analysis that counselling.12 This study was the only one identified of sufficient
examined the within-group changes of negative symptoms over quality to be included in a recent Cochrane review of dance
time,36 suggesting the improvements observed were spontaneous, therapy for schizophrenia.37 In the context of arts therapies as a
and did not reflect any therapeutic effect. Given the symptom whole, our findings contradict the current NICE arts therapy
change in the Pilates group was similar to control conditions from review,6 instead mirroring those reported in the MATISSE art
other clinical trials that aimed to treat negative symptoms, it therapy trial.8 Collectively, these two trials could be considered
suggests that adopting Pilates as a comparator was appropriate, to have one of two implications for the NICE recommendations,
with the findings generalisable to other active control conditions dependent upon how the concept of arts therapies itself continues
presuming they do not provide any additional clinical benefit over to be defined. If we continue to evaluate arts therapies as a
treatment as usual either. singular treatment ‘type’ as is the case in the present NICE
Another possible limitation is the relatively short duration of review,6,7 then incorporating the findings from the current study
the treatment under investigation. Although it remains unclear and MATISSE would result in the current evidence base suggesting
whether more prolonged exposure to therapy may result in that arts therapies are not an effective treatment for negative
changes to negative symptoms, this should be considered unlikely. symptoms of schizophrenia. If arts therapies are instead
In the meta-analysis on the dose–response effect of music recognised as heterogeneous, each with a different model of

5
6
Priebe et al

Table 2 Descriptive statistics and complete case analysis of outcome measures over the three time points by condition
Body psychotherapy group Pilates group Between-group differencesb
At end of treatment 6 months post-treatment
End of End of
10-week 6 months post- 10-week 6 months post- Adjusted mean differ- Adjusted mean differ-
Outcomea Baseline treatment treatment Baseline treatment treatment ence/IRR (95% CI)c ICCd ence/IRR (95% CI)c ICCd

PANSS, mean (s.d.)

Negative (n = 254) 23.3 (4.3) 21.8 (5.4) 21.7 (5.7) 23.1 (4.4) 21.5 (4.7) 21.7 (5.1) 0.03 (71.11 to 1.17) 0.099 70.18 (71.68 to 1.31) 0.137
Positive (n = 253) 14 (5.1) 13.1 (4.7) 13.4 (4.7) 14.1 (4.7) 13.3 (4.2) 13.6 (4.9) 0.06 (70.71 to 0.84) 50.001 70.12 (71.03 to 0.79) 50.001
General (n = 249) 32.9 (8.3) 30.2 (8) 30.1 (8.1) 32.5 (8.1) 29.9 (7.3) 30.4 (7.5) 0.32 (71.31 to 1.94) 0.096 70.70 (73.07 to 1.67) 0.205
Marder negative (n = 253) 22.2 (4.7) 20.7 (5.7) 20.2 (5.7) 21.9 (5) 20.3 (5.1) 20.1 (5.6) 0.23 (70.86 to 1.32) 0.678 0.04 (71.38 to 1.45) 0.075
CAINS, mean (s.d.)
Experience (n = 246) 22.1 (5.6) 20.5 (5.8) 20.8 (6.7) 21.5 (5.5) 19.8 (5.8) 20.6 (6.2) 0.05 (71.13 to 1.22) 0.037 70.04 (71.48 to 1.40) 0.041
Expression (n = 253) 8 (3.5) 7.3 (3.7) 7.1 (4) 7.5 (3.9) 7.5 (4.1) 7.1 (4.3) 70.62 (71.23 to 0.00) 0.022 70.27 (71.05 to 0.50) 0.023
Calgary, mean (s.d.) (n = 253) 4.8 (4.2) 3.9 (4.3) 4.1 (4.1) 4.6 (4.6) 3.9 (4.3) 4.2 (4.2) 70.01 (70.72 to 0.71) 50.001 70.20 (71.18 to 0.79) 0.086
MANSA, mean (s.d.) (n = 254) 4.4 (0.9) 4.5 (0.9) 4.6 (1) 4.4 (0.9) 4.6 (0.9) 4.5 (0.9) 70.11 (70.27 to 0.58) 50.001 0.10 (70.12 to 0.32) 0.050
SNS, median (IQR) (n = 232)
Relatives seen 2 (1.0–3.0) 3 (1.0–4.0) 2 (1.0–4.0) 2 (1.0–4.0) 2 (1.0–4.0) 2 (1.0–4.0) 1.13 (0.89 to 1.32) – 0.96 (0.80 to 1.15) –
Friends seen 1 (0.0–2.0) 1 (0.0–2.0) 0.5 (0.0–2.0) 1 (0.0–2.0) 1 (0.0–2.0) 1 (0.0–2.0) 0.94 (0.80 to 1.42) – 0.91 (0.63 to 1.30) –
Total number seen 3 (2.0–5.0) 4 (3.0–6.0) 4 (2.0–6.0) 4 (2.0–5.0) 4 (3.0–6.0) 4 (2.0–6.0) 0.83 (0.69 to 1.14) – 0.97 (0.85 to 1.12) –
TUS, median (IQR) (n = 254)
Number of activities 3 (1.0–6.0) 3 (1.0–7.0) 3 (1.0–7.0) 3 (1.0–6.0) 2 (1.0–7.0) 2 (1.0–7.0) 1.03 (0.89 to 1.42) – 1.04 (0.81 to 1.33) –
Time spent (hours) 1.5 (0.0–3.5) 1.5 (0.3–4.0) 1.5 (0.0–3.0) 1.8 (0.3–4.0) 2 (0.3–4.5) 1.5 (0.2–3.8) 1.03 (0.80 to 1.32) – 0.96 (0.73 to 1.25) –
SAS, mean (s.d.) (n = 229) 1.7 (2.1) 1.2 (1.7) 1.2 (1.5) 2.3 (2.7) 2.1 (2.9) 1.9 (2.4) 70.65 (71.13 to 70.16) 50.001 70.50 (70.94 to 70.07) 0.007
SIX, mean (s.d.) (n = 254) 2.4 (1.1) 2.5 (1.1) 2.5 (1) 2.3 (1.1) 2.5 (1.1) 2.5 (1.2) 70.02 (70.17 to 0.20) 50.001 70.10 (70.27 to 0.08) 50.001
CSQ, mean (s.d.) (n = 237) – 25.3 (4.6) – – 25.9 (4) – 70.68 (71.80 to 0.44) – – –

PANSS, Positive and Negative Syndrome Scale; CAINS, Clinical Assessment Interview for Negative Symptoms; Calgary, Calgary Depression Scale; MANSA, Manchester Short Assessment of Quality of Life; SNS, Social Network Scale; TUS, Time Use Survey;
SAS, Simpson Angus Scale; CSQ, Client Satisfaction Questionnaire.
a. The number of participants completing each outcome at 6 months (and for the CSQ at end of treatment) is included in brackets.
b. Models adjusted for baseline measure of outcome, study centre and a random effect for therapy group (except CSQ).
c. All values are adjusted mean difference except those for SNS and TUS where incident rate ratios (IRRs) are reported.
d. Values are model-based intracluster correlation coefficients (ICCs). ICCs not calculated for count outcomes.
 Group body psychotherapy for negative symptoms of schizophrenia

action, then it suggests that the current evidence base upon which
References
NICE concludes that arts therapies may be helpful for negative
symptoms may be inappropriate. If we presume the latter, then 1 Saha S, Chant D, Welham J, McGrath J. A systematic review of the
full-scale trials in other arts therapies such as music therapy prevalence of schizophrenia. PLoS Med 2005; 2: e141.
may be merited in this particular patient group. Although 2 Kirkpatrick B, Fenton WS, Carpenter WT Jr, Marder SR. The NIMH-MATRICS
small-scale investigations have suggested that music therapy consensus statement on negative symptoms. Schizophr Bull 2006; 32: 214–9.
may be effective,38 given that promising results in small-scale 3 Horan WP, Kring AM, Gur RE, Reise SP, Blanchard JJ. Development and
investigations have not been replicated either here or in MATISSE, psychometric validation of the Clinical Assessment Interview for Negative
Symptoms (CAINS). Schizophr Res 2011; 132: 140–5.
it suggests that caution should be advised in interpreting such
4 Hunter R, Barry S. Negative symptoms and psychosocial functioning in
findings. schizophrenia: neglected but important targets for treatment. Euro
In the secondary outcomes a small, significant improvement Psychiatry 2012; 27: 432–6.
in the body psychotherapy group was detected in expressive 5 Ho BC, Nopoulos P, Flaum M, Arndt S, Andreasen NCJ. Two-year outcome in
symptoms at end of treatment measured by the CAINS, and in first-episode schizophrenia: predictive value of symptoms for quality of life.
movement disorder symptoms both at end of treatment and Am J Psychiat 1998; 155: 1196–201.

6 months later. For both findings it is important to consider that 6 National Collaborating Centre for Mental Health. Schizophrenia: The NICE
Guideline on Core Interventions in the Treatment and Management of
multiple testing with the risk of an inflated type I error was Schizophrenia in Adults in Primary and Secondary Care (CG82). NICE, 2009.
conducted. However, the fact that a difference was detected in this
7 National Collaborating Centre for Mental Health. Psychosis and
scale, in contrast with the PANSS, may be important given one of Schizophrenia in Adults: The NICE Guideline on Treatment and Management
the main aims of the Collaboration to Advance Negative Symptom (CG178). NICE, 2014.
Assessment for Schizophrenia (CANSAS)39 was to develop new 8 Crawford MJ, Killaspy H, Barnes TR, Barrett B, Byford S, Clayton K, et al.
scales that are sufficiently sensitive to detect negative symptom Group art therapy as an adjunctive treatment for people with schizophrenia:
a multicentre pragmatic randomised trial. BMJ 2012; 344: e846.
change in clinical trials.2 Second, this finding may provide further
9 Rohricht F, Priebe SG. Effect of body-oriented psychological therapy on
evidence for the importance of measuring expressive and
negative symptoms in schizophrenia: a randomized controlled trial. Psychol
experiential features of negative symptoms separately given they Med 2006; 36: 669–78.
represent separate constructs.40 The change in movement disorder 10 Darby J. Alteration of some body image indexes in schizophrenia. J Consult
symptoms should be interpreted with much caution since an Clin Psych 1970; 35: 116–21.
incomplete scale was used. Although it is intuitive to consider that 11 Nitsun M, Stapleton JH, Bender M. Movement and drama therapy with long-
a treatment that focuses specifically on the body may help alleviate stay schizophrenics. Br J Med Psychol 1974; 47: 101–19.
movement-related symptoms, this finding should be re-examined 12 Goertzel V, MAY PR, Salkin J, Schoop T. Body-ego technique: an approach to
in a trial focused on such outcomes before drawing firm the schizophrenic patient. J Nerv Ment Dis 1965; 141: 53–60.

conclusions. 13 Priebe S, Röhricht F. Specific body image pathology in acute schizophrenia.

Psychiatry Res 2001; 101: 289–301.
In conclusion, overall, this study does not support group body
14 Röhricht F, Priebe S. Disturbances of body experience in schizophrenic
psychotherapy as a treatment for negative symptoms of patients [in German]. Fortschr Neurol Psychiatr 1997; 65: 323–36.
schizophrenia. Reviewing the effectiveness of different arts therapy
15 Priebe S, Savill M, Reininghaus U, Wykes T, Bentall R, Lauber C, et al.
modalities separately may be informative to determine whether Effectiveness and cost-effectiveness of body psychotherapy in the treatment
existing guidelines should be more cautious in recommending of negative symptoms of schizophrenia – a multi-centre randomised
art and body psychotherapy specifically, or whether this extends controlled trial. BMC Psychiatry 2013; 13: 26.

to arts therapies as a whole. 16 World Health Organization. International Statistical Classification of Diseases
and Related Health Problems (Tenth Revision) (ICD-10). WHO, 1992.
17 Kay SR, Flszbein A, Opfer LA. The positive and negative syndrome scale
S. Priebe, FRCPsych, M. Savill, PhD, Unit for Social and Community Psychiatry, WHO
(PANSS) for schizophrenia. Schizophrenia Bull 1987; 13: 261–76.
Collaborative Centre for Mental Health Services Development, Queen Mary University
of London, London, UK; T. Wykes, PhD, Institute of Psychiatry, Kings College London, 18 Priebe S, Huxley P, Knight S, Evans S. Application and results of the
London, UK; R. P. Bentall, PhD, Department of Psychiatry, University of Liverpool, Manchester Short Assessment of Quality of Life (MANSA). Int J Soc Psychiatry
Liverpool, UK; U. Reininghaus, MSc, DiplPsych, Institute of Psychiatry, Kings College 1999; 45: 7–12.
London, London, UK and Department of Psychiatry and Psychology, School for Mental
Health and Neuroscience, Maastricht University, The Netherlands; C. Lauber, MD, 19 Priebe S, Watzke S, Hansson L, Burns T. Objective social outcomes index
Services Psychiatriques, Jura Bernois – Bienne-Seeland, Bellelay, Switzerland; (SIX): a method to summarise objective indicators of social outcomes in
S. Bremner, PhD, Brighton and Sussex Medical School, University of Sussex, mental health care. Acta Psychiatr Scand 2008; 118: 57–63.
Brighton, UK; S. Eldridge, MSc, Centre for Primary Care and Public Health, Queen
Mary University of London, London, UK; F. Röhricht, MD, MRCPsych, Unit for Social 20 Addington D, Addington J, Maticka-Tyndale E. Assessing depression in
and Community Psychiatry, WHO Collaborative Centre for Mental Health Services schizophrenia: the Calgary Depression Scale. Br J Psychiatry 1993; 163
Development, Queen Mary University of London, London, UK (suppl 22): 39–44.

Correspondence: S. Priebe, Unit for Social and Community Psychiatry, WHO 21 Dunn M, O’Driscoll C, Dayson D, Wills W, Leff J. The TAPS Project. 4: an
Collaborative Centre for Mental Health Services Development, Queen Mary observational study of the social life of long-stay patients. Br J Psychiatry
University of London, London E13 8SP, UK. Email: [email protected] 1990; 157: 842–8.
22 Office for National Statistics. The United Kingdom 2000 Time Use Survey –
First received 17 Jun 2015, final revision 4 Nov 2015, accepted 5 Nov 2015
Technical Report. ONS, 2003.
23 Nguyen TD, Attkisson CC, Stegner BL. Assessment of patient satisfaction:
development and refinement of a service evaluation questionnaire. Eval
Program Plann 1983; 6: 299–313.
Funding 24 Simpson G, Angus J. A rating scale for extrapyramidal side effects. Acta
Psychiat Scand 1970; 45: 11–9.
This work was supported by the National Institute for Health Research – Health Technology
Assessment (NIHR-HTA) programme (grant number: 08/116/68). The views and opinions 25 Marder SR, Davis JM, Chouinard G. The effects of risperidone on the five
expressed therein are those of the authors and do not necessarily reflect those of the dimensions of schizophrenia derived by factor analysis: combined results of
HTA programme, NIHR, NHS or the Department of Health. the North American trials. J Clin Psychiat 1997; 58: 538–46.
26 Chisholm D, Knapp MRJ, Knudsen HC, Amaddeo F, Gaite L, Van Wijngaarden
B. Client Socio-Demographic and Service Receipt Inventory – European
Acknowledgements Version: development of an instrument for international research. EPSILON
Study 5. Br J Psychiatry 2000; 177 (suppl 39): s28–33.
The authors would like to thank Stavros Orfanos, Ciara Banks, Erica Eassom, Tabitha Dow, 27 Szende A, Oppe M, Devlin NJ. EQ-5D Value Sets: Inventory, Comparative
Rebecca Stockley, Josie Davies and Nina Papadopoulos for their involvement in the project. Review and User Guide. Springer, 2007.

7
Priebe et al

28 Rohricht F. Body-Oriented Psychotherapy in Mental Illness: A Manual for 35 Khoury B, Lecomte T, Gaudiano A, Paquin K. Mindfulness interventions for
Research and Practice. Hogrefe, 2000. psychosis: a meta-analysis. Schizophr Res 2013; 150: 176–84.
29 Röhricht F. Body oriented psychotherapy. The state of the art in empirical 36 Savill M, Banks C, Khanom H, Priebe S. Do negative symptoms of
research and evidence-based practice: a clinical perspective. Body Mov schizophrenia change over time? A meta-analysis of longitudinal data.
Dance Psychother 2009; 4: 135–56. Psychol Med 2015; 45: 1613–27.
30 Röhricht F, Papadopoulos N, Priebe S. An exploratory randomized controlled
37 Ren J, Xia J. Dance therapy for schizophrenia. Cochrane Database Syst Rev
trial of body psychotherapy for patients with chronic depression. J Affect
2013; 10: CD006868.
Disord 2013; 151: 85–91.
31 Gold C, Solli HP, Krüger V, Lie SA. Dose–response relationship in music 38 Mössler K, Chen X, Heldal TO, Gold C. Music therapy for people with
therapy for people with serious mental disorders: systematic review and schizophrenia and schizophrenia-like disorders. Cochrane Database Syst
meta-analysis. Clinical Psychol Rev 2009; 29: 193–207. Rev 2011; 12: CD004025.

32 Newham E. Pilates Union UK: Comprehensive Matwork Manual. Pilates Union 39 Blanchard JJ, Kring AM, Horan WP, Gur R. Toward the next generation of
UK, 2010. negative symptom assessments: the collaboration to advance negative
33 Angrist JD, Imbens GW, Rubin DB. Identification of causal effects using symptom assessment in schizophrenia. Schizophrenia Bull 2011; 37: 291–9.
instrumental variables. J Am Stat Assoc 1996; 91: 444–55. 40 Blanchard JJ, Cohen AS. The structure of negative symptoms within
34 Adams M, Caldwell K, Atkins L, Quin R. Pilates and mindfulness: a qualitative schizophrenia: implications for assessment. Schizophr Bull 2006; 32: 238–45.
study. J Dance Educ 2012; 12: 123–30.

8
Data supplement to Priebe et al. Effectiveness of group body psychotherapy for negative
symptoms of schizophrenia: multicentre randomised controlled trial. Br J Psychiatry doi:
10.1192/bjp.bp.115.171397

Supplement DS1
Body Psychotherapy Adherence Scale (BPT-AS)

The BPT-AS is composed of 10 items and has been developed in order to test adherence to protocol and
treatment distinctiveness. Each of the 10 items is rated on a 0 to 2 point scale (0 = no evidence, 1= limited
evidence, 2=definite evidence).
Most BPT-AS items assess therapist behaviours specific to BPT in chronic schizophrenia (BPT-CS), e.g.
movement interventions, bodily self-awareness exercises; the first three items relate to aspects of the therapy
practice that BPT shares with other group therapies (e.g. group cohesion, therapeutic environment, provision
of therapy rationale, conveying core themes).

(Please circle each score as appropriate)

1. THERAPEUTIC ENVIRONMENT
Is there evidence that the therapist has created an appropriate therapeutic environment that enables
a positive therapeutic relationship/alliance?
0 =No Evidence 1 = limited evidence 2 = clear evidence

2. GROUP COHESION:
Is there evidence that therapist’s actions facilitate the cohesiveness and shared identity of the
treatment group?
0 =No Evidence 1 = limited evidence 2 = clear evidence

3. PROVISION OF BPT RATIONALE:

Does the therapist provide patients with an explanation for why the performance of specific BPT
tasks/interventions will help them to address specific symptoms of their illness?
0 =No Evidence 1 = limited evidence 2 = clear evidence

4. PROGRESSING THERAPY THROUGH FIVE SECTIONS:

Is the therapist following the 5-section structure format of the session?
0 =No Evidence 1 = limited evidence 2 = clear evidence

5. EXTENDED BODY-SELF-AWARENESS EXERCISES/INTERVENTIONS:

To what extent does the therapist use self-awareness exercises (e.g. body check-in/ body
exploration)?
0 =No Evidence 1 = limited evidence 2 = clear evidence
6. MOVEMENT-BASED SELF-EXPRESSION EXERCISES/INTERVENTIONS:
To what extent does the therapist use movement exercises to foster self-expression and/or express
and communicate emotions?
0 =No Evidence 1 = limited evidence 2 = clear evidence.

7. MOVEMENT BASED SOCIAL INTERACTION EXERCISES:

To what extent does the therapist use movement exercises to encourage social interaction?
0 =No Evidence 1 = limited evidence 2 = clear evidence

8. USE OF TOOLS / OBJECTS IN THERAPY:

To what extent does the therapist use a range of different tools/objects in therapy?
0 =No Evidence 1 = limited evidence 2 = clear evidence

9. CONSISTENCY OF PRACTICE REVIEW:

Does the therapist appropriately review the session at the end of the group?
0 =No Evidence 1 = limited evidence 2 = clear evidence

10. BODILY COPING STRATEGIES IN RELATION TO SPECIFIC NEGATIVE SYMPTOMS:

Does the therapist introduce and/or relate to different body based coping strategies for responding to
negative symptoms?
0 =No Evidence 1 = limited evidence 2 = clear evidence

Any observations, comments or reflections:

Effectiveness of group body psychotherapy for negative
symptoms of schizophrenia: multicentre randomised controlled
trial
S. Priebe, M. Savill, T. Wykes, R. P. Bentall, U. Reininghaus, C Lauber, S. Bremner, S. Eldridge and F.
Röhricht
BJP published online May 5, 2016 Access the most recent version at DOI: 10.1192/bjp.bp.115.171397

Supplementary Supplementary material can be found at:

Material http://bjp.rcpsych.org/content/suppl/2016/04/25/bjp.bp.115.171397.DC1.html

References This article cites 0 articles, 0 of which you can access for free at:
http://bjp.rcpsych.org/content/early/2016/04/21/bjp.bp.115.171397#BIBL
Reprints/ To obtain reprints or permission to reproduce material from this paper, please
permissions write to [email protected]

P<P Published online 2016-05-05T00:05:13-07:00 in advance of the print journal.

You can respond /letters/submit/bjprcpsych;bjp.bp.115.171397v1

to this article at
Downloaded http://bjp.rcpsych.org/ on May 16, 2016
from Published by The Royal College of Psychiatrists

Advance online articles have been peer reviewed and accepted for publication but have not yet
appeared in the paper journal (edited, typeset versions may be posted when available prior to
final publication). Advance online articles are citable and establish publication priority; they are
indexed by PubMed from initial publication. Citations to Advance online articles must include the
digital object identifier (DOIs) and date of initial publication.

To subscribe to The British Journal of Psychiatry go to:

http://bjp.rcpsych.org/site/subscriptions/