Acta Psychiatr Scand 2015: 131: 350–359 All rights reserved DOI: 10.1111/acps.12371
© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd ACTA PSYCHIATRICA SCANDINAVICA
Exercise augmentation compared with usual care for post-traumatic stress disorder: a randomized controlled trial Rosenbaum S, Sherrington C, Tiedemann A. Exercise augmentation compared to usual care for post-traumatic stress disorder: a randomized controlled trial. Objective: To investigate the impact of a 12-week exercise programme in addition to usual care for post-traumatic stress disorder (PTSD). Method: An assessor-blinded randomized controlled trial was conducted among 81 participants with a DSM-IV-TR diagnosis of primary PTSD. Participants were recruited after admission to an in-patient programme at a private hospital. Participants were randomized to receive either usual care (n = 42), or exercise in addition to usual care (n = 39). The exercise intervention involved three, 30-min resistance-training sessions/week and a pedometerbased walking programme. Usual care involved psychotherapy, pharmaceutical interventions, and group therapy. Primary outcome was PTSD symptoms assessed via the PTSD checklist–civilian version (PCL-C). Secondary outcomes included symptoms of depression, anthropometry, physical activity, mobility, strength, and sleep quality. Results: Participants had a mean (SD) age of 47.8 years (12.1), 84% male. PTSD symptoms in the intervention group significantly reduced compared with the usual care group (mean difference = 5.4, 95% CI 10.5 to 0.3, P = 0.04, n = 58). There were significant between-group differences at follow-up for depressive symptoms, waist circumference, sleep quality, and sedentary time. Conclusion: This study provides the first evidence that an exercise intervention is associated with reduced PTSD and depressive symptoms, reduced waist circumference, and improved sleep quality.
S. Rosenbaum1,2,
C. Sherrington1, A. Tiedemann1 1 Musculoskeletal Division, The George Institute for Global Health and School of Public Health, University of Sydney, Sydney, NSW, and 2St John of God Health Care Richmond Hospital, Sydney, NSW, Australia
Key words: exercise; post-traumatic stress disorder; randomized controlled trial; physical activity Simon Rosenbaum, The George Institute for Global Health, Lv 13, 320 Kent St, Sydney, NSW 2000, Australia. E-mail: [email protected]
Accepted for publication October 27, 2014
Significant outcomes
• When augmented with usual care, structured exercise incorporating resistance training and walking improves symptoms for people with post-traumatic stress disorder (PTSD) beyond usual care alone.
• Exercise improves symptoms of depression and cardiometabolic risk in people with PTSD. Limitations
• A higher rate of loss to follow up in the physical assessments could be addressed by the inclusion of home visitation to collect follow-up data.
• While the higher inclusion of men in the current study is reflective of gender rates among police and servicemen who attended the in-patient facility, this does limit the generalizability of the findings to females.
350
Exercise and PTSD Introduction
Post-traumatic stress disorder (PTSD) can occur following exposure to a potentially traumatic event, causing clinically significant social and occupational impairment and characterized by symptoms of hyperarousal, re-experiencing, negative cognitions and mood, and avoidance (1). Lifetime prevalence of PTSD within the general population is estimated at between 5% and 10% (2), with occupation-specific rates among combat veterans and police officers significantly higher (e.g. up to 17%) (3, 4). Common comorbid psychiatric diagnoses include major depressive disorder and generalized anxiety disorder, which have both been the subject of comprehensive investigations as to the efficacy of exercise as an intervention (5, 6). People with PTSD are also more likely to suffer from metabolic disturbances and poor physical health (7). Despite the known benefits of exercise for improving cardiometabolic health (8), no studies to date have assessed the impact of exercise on physical health outcomes of people with PTSD. People with PTSD are also less likely to engage in regular physical activity compared with pre-PTSD levels (9). The antidepressive and anxiolytic effects of physical activity and exercise for people with serious mental illness are established (6, 10). Despite significant heterogeneity among interventions and study designs, exercise appears to be effective beyond the placebo effect for depressive and anxiety disorders (6). Given the positive impact of regular physical activity and exercise on physical health, clinical trials to investigate the impact of exercise in people with PTSD are justified. A 2010 Cochrane Collaboration review into the effect of sports and games on PTSD highlighted the lack of available evidence for exercise as a treatment or cotreatment option (11). The review identified five studies; however, the generalizability of the findings was limited due to significant methodological weaknesses, including a lack of randomization, small sample sizes, and the inclusion of interventions evaluating play-based therapy, considered a psychological intervention. One study that specifically examined an exercise intervention was a non-randomized trial of 12 female adolescents diagnosed with PTSD who completed a 5-week aerobic exercise programme (12). The authors found a significant reduction in symptoms as measured by the Childhood PTSD Symptom Scale (12), but the impact on measures of physical health was not examined. An interrelationship between weight loss and PTSD symptoms has been demonstrated in a prospective study of weight loss facility attendees (13),
suggesting the importance of accessing both psychological and physical health indicators. Sleep disturbances are common, debilitating symptoms of PTSD, with evidence suggesting that treatment strategies aimed at improving sleeprelated symptoms may also help to improve other symptoms of PTSD (14). Exercise has been found to be effective in improving sleep quality among the general population, and the antidepressive and anxiolytic effects of exercise have been proposed as potential mechanisms for this improvement (15). Despite this potential relationship, no previous studies have evaluated exercise as a strategy to improve sleep quality among people with PTSD. Aims of the study
The aim of this study was to investigate the impact of a 12-week combined resistance and walkingbased exercise programme as an augmentation to usual care for patients hospitalized for treatment for post-traumatic stress disorder. The effect of exercise on symptoms of post-traumatic stress (primary outcome), depression, anxiety and stress ratings, body composition, physical activity levels, sleep, mobility, and strength was assessed.
Material and methods Design
We applied an assessor-blinded randomized controlled trial (RCT) with ethical review provided by St John of God Health Care Ethics Committee (REF: 412) and The University of Sydney Human Research Ethics Committee (Protocol No: 14091). The research protocol (16) and the trial were registered with the Australian and New Zealand Clinical Trials Registry prior to commencement of participant recruitment (ACTRN 12610000579099). The study was designed with reference to the CONSORT extension for nonpharmacological treatment interventions (17) and took place between September 2010 and May 2012. To ensure allocation concealment, randomization to study groups was undertaken by an investigator (AT) not involved in recruitment using a block randomization sequence generated using random numbers in Excel and including randomly varying block sizes of 2, 4, and 6. Baseline assessment and questionnaires were completed prior to allocation of participants to a study group. An assessor blinded to group allocation collected outcome data at 12 weeks postrandomization. A priori sample size calculation determined that 80 participants (40 per 351
Rosenbaum et al.
Assessed for eligibility (n = 314)
Randomized (n = 81)
Allocated to usual care (n = 42) Received usual care (n = 42)
Excluded (n = 233) Did not meet inclusion criteria (n = 2) Out of area and unable to participate in follow up or declined to participate (n = 228) Pregnant (n = 3)
Allocated to intervention (n = 39) Received allocated intervention (n = 39)
Unable to contact (n = 14)
Unable to contact (n = 9)
Analyzed (n = 28)
Analyzed (n = 30)
Fig. 1. CONSORT 2010 flow diagram.
group) were required to allow 80% power to detect as significant at the 5% level a 5-point between-group difference on the primary outcome, the PTSD checklist–civilian version, with a standard deviation of 9.4 and a correlation between baseline and follow-up measures of 0.65 (18) allowing for 15% drop-out (Fig. 1). Participants
Eighty-one in-patients from the PTSD in-patient unit at St John of God Health Care’s Richmond Hospital, Australia, were recruited through direct contact with the lead author (SR), who introduced the study during regular ward meetings. Inclusion criteria were as follows: (i) men and women aged over 18 years; (ii) psychiatrist-confirmed DSM-IVTR diagnosis of primary PTSD; (iii) medical clearance to participate in an exercise programme; and (iv) cognitively able to provide consent to participate. Exclusion criteria were as follows: (i) medically unfit to participate in an exercise programme (for example recent acute cardiac event, unstable angina, acute embolus or infarction, or acute systematic infection); (ii) pregnant or planning pregnancy in the proceeding 12 months; and (iii) complex PTSD with trauma occurring in childhood only. 352
Intervention
Intervention details can be seen in the exercise reporting grid in Table S1 (19). The 12-week intervention consisted of a weekly supervised exercise session, two unsupervised home-based exercise sessions, and a walking programme facilitated by the provision of a pedometer and exercise diary. Supervised sessions were conducted in a gymnasium within the hospital and were supervised by an exercise physiologist (SR). Unsupervised sessions were home-based and utilized the same exercise programme as the supervised sessions. All required exercise equipment (exercise bands and pedometer) were provided to participants to ensure standardization between the supervised and unsupervised sessions. The exercise protocol was designed with reference to the American College of Sports Medicine (ACSM) guidelines for resistance training (20). It consisted of three sets of 10 repetitions of up to six exercises, with a rest period of between 30 and 60 s between sets. Exercise intensity was individualized as per ACSM recommendations and depending on results from the physical assessment, the International Physical Activity Questionnaire (IPAQ) and using the Borg Rating of Perceived Exertion Scale (RPE) (21, 22) in which participants were advised
Exercise and PTSD to aim for an exertion intensity of between 12 and 17 of 20 (i.e. ‘comfortably hard’ to ‘very hard’). An increase in load of up to 10% (based on RPE) was applied when the individual could perform the current workload for 1–2 repetitions over the desired number, based on ACSM guidelines (20, 23). Participants who were more physically active at baseline were expected to require less time on average to become familiar with the exercises and would therefore be able to achieve the recommended intensity and volume more quickly. Participants experiencing a higher severity of PTSD symptoms (based on participant report and liaison with clinical staff) were expected to be less engaged and less willing to complete sessions of longer duration, and the programme was adjusted accordingly by reducing the volume and/or intensity to maximize session adherence. Intervention group participants were provided with an Omron HJ109 pedometer (Omron, UE Square, Singapore) and exercise diary to quantify and record daily step count. The pedometer and exercise diary were used to help guide goal setting and act as a motivational tool. Participants were encouraged to aim for an ultimate daily target of 10 000 steps (24); however, this was individualized based on baseline physical activity levels to promote adherence and the achievement of realistic goals. Usual care
Both groups received usual care involving a combination of psychotherapy, pharmaceutical interventions, and group therapy facilitated by psychologists. These interventions included ‘arousal reduction’, for example, breathing retraining and relaxation skills, psychoeducation (involving information about PTSD and its treatment), cognitive therapy, and communication skills. Access to these treatments was equal across both groups for the duration of the study. Average length of stay for the in-patient programme was 3 weeks, with most patients continuing a less-intensive out-patient programme for several months. Outcome measures
Primary outcome. The primary outcome was PTSD symptoms as assessed by the PTSD checklist–civilian version (PCL-C) (25). The PCL-C is a 17-item self-report questionnaire that assesses the key symptoms of PTSD and is one of the most commonly used self-report measures of PTSD (26). Scores range from 17 to 85, with higher scores
indicating higher symptom severity and a score of 45 normally selected as the cutoff for a diagnosis of primary PTSD (25). Secondary outcomes. The Depression Anxiety and Stress Scale (DASS) is a 42-item self-report instrument that measures the related negative emotional states of depression, anxiety, and tension/stress. Higher scores on the DASS equate to greater symptom severity. For the depression domain, scores of 0–13 are considered normal or mild, 14– 20 moderate, 21–27 severe, and >28 extremely severe. For anxiety, 0–7 is considered normal, 8–9 mild, 10–14 moderate, 15–19 severe, and >20 extremely severe. For stress, 0–14 is normal, 15–18 mild, 19–25 moderate, 26–33 severe, and >34 extremely severe (27). The psychometric properties of the DASS have been comprehensively evaluated, and it has been found to be valid, consistent, and responsive to treatment (27). Given the poor documented physical health profile of people experiencing PTSD, measures of physical health were assessed including resting heart rate, blood pressure, body mass index (weight in kilograms/height in meters2), body fat percentage (obtained using a bio-impedance scale; Tanita, Kewdale, Western Australia, 2011), and waist circumference (defined as the point midway between the iliac crest and costal margin) by an assessor blinded to group allocation. Self-reported physical activity levels were assessed with the International Physical Activity Questionnaire–Short Form (IPAQ-SF) (28). The IPAQ-SF includes four questions that assess participation in vigorous and moderate intensity exercise, and time spent walking and sitting in the previous 7-day period. Sleep behaviour including total sleep time and barriers to sleep were assessed using the Pittsburgh Sleep Quality Index (29) and the Pittsburgh Sleep Quality Index Addendum for PTSD (30). Both instruments are self-report questionnaires that assess sleep quality over the past month. The PSQI has 10 items and total scores range from 0 to 21, with scores >5 associated with poor sleep quality (29). The PSQIA has 10 items and total scores range from 0 to 21, with a score of four previously shown to discriminate adults with PTSD (30). The 6-min walk test was used to assess cardiorespiratory fitness (31) and was administered by an assessor blinded to group allocation within the hospital gymnasium using an 8-m long walkway. Participants were instructed to walk the length of the walkway as many times as possible within the 6 min. Minimal encouragement was provided as per recommendations, and both RPE and total
353
Rosenbaum et al. Table 1. Baseline characteristics of participants in intervention and control groups. Means (standard deviations) unless stated otherwise
Characteristics
Intervention (n = 39)
Usual care (n = 42)
47.1 (11.3) 3 (8) 34 (88)
52.0 (12.7) 10 (24) 37 (89)
2 (5) 2 (5) 7 (18)
3 (7) 2 (5) 8 (19)
7 (18) 16 (41) 16 (41) 37 (95)
1 (2) 16 (38) 25 (60) 38 (91)
7 (18) 6 (15) 7 (18) 19 (49)
10 (23) 4 (8) 4 (8) 26 (61)
9 (23) 4 (10) 5 (14) 21 (53)
9 (21) 3 (7) 6 (15) 24 (57)
15 (39) 9 (23) 3 (8) 12 (30)
18 (42) 11 (26) 3 (8) 10 (24)
Age (years) Female gender, n (%) Trauma experienced through occupation, n (%) History, n (%) Diabetes Cardiovascular disease Hypertension Body mass index (kg/m2), n (%) Desired (18.5–24.9) Overweight (25–29.9) Obese (≥30) PCL-C ≥ 45,* n (%) DASS depression domain Normal mild (0–13) Moderate (14–20) Severe (21–27) Extremely severe (≥28) DASS anxiety domain Normal mild (0–9) Moderate (10–14) Severe (15–19) Extremely severe (≥20) DASS stress domain Normal mild (0–18) Moderate (19–25) Severe (26–33) Extremely severe (≥34)
PCL-C, post-traumatic stress disorder checklist–civilian; DASS-42, Depression Anxiety and Stress Scale. *Recommended diagnostic criteria.
number of laps were recorded at the end of each minute (31). Grip strength was assessed using an adjustable handgrip dynamometer (TTM Original Dynamometer, Tokyo, Japan) with a single assessment carried out using the dominant hand (32). Intervention adherence and adverse events
Attendance at the supervised exercise sessions was recorded by the exercise physiologist (SR). Participants were instructed to record their home-based exercise sessions on exercise diaries provided. Adverse events associated with exercise participation (e.g. muscle soreness, chest pain) that required medical intervention or interfered with activities of daily living for more than 2 days were recorded on the exercise diaries. Data analysis
Analyses used all available data as per an intention-to-treat approach. Data were analyzed using SPSS Version 21 (Armonk, NY, USA) and STATA Version 12.1 statistical software (StataCorp LP, College Station, TX , USA). The between-group 354
difference at follow-up [mean difference (MD) and 95% confidence interval] was calculated for all outcome measures using linear regression models with group as the independent variable, follow-up score on the outcome measures as the dependent variable, baseline score on the outcome measure as a covariate, and statistical significance set at P < 0.05. All data were normally distributed. Imputation methods were not applied given evidence of the subsequent risk of bias (33). Partial eta-squared (g2p ) effect sizes were calculated where a 0.01 to
© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd ACTA PSYCHIATRICA SCANDINAVICA
Exercise augmentation compared with usual care for post-traumatic stress disorder: a randomized controlled trial Rosenbaum S, Sherrington C, Tiedemann A. Exercise augmentation compared to usual care for post-traumatic stress disorder: a randomized controlled trial. Objective: To investigate the impact of a 12-week exercise programme in addition to usual care for post-traumatic stress disorder (PTSD). Method: An assessor-blinded randomized controlled trial was conducted among 81 participants with a DSM-IV-TR diagnosis of primary PTSD. Participants were recruited after admission to an in-patient programme at a private hospital. Participants were randomized to receive either usual care (n = 42), or exercise in addition to usual care (n = 39). The exercise intervention involved three, 30-min resistance-training sessions/week and a pedometerbased walking programme. Usual care involved psychotherapy, pharmaceutical interventions, and group therapy. Primary outcome was PTSD symptoms assessed via the PTSD checklist–civilian version (PCL-C). Secondary outcomes included symptoms of depression, anthropometry, physical activity, mobility, strength, and sleep quality. Results: Participants had a mean (SD) age of 47.8 years (12.1), 84% male. PTSD symptoms in the intervention group significantly reduced compared with the usual care group (mean difference = 5.4, 95% CI 10.5 to 0.3, P = 0.04, n = 58). There were significant between-group differences at follow-up for depressive symptoms, waist circumference, sleep quality, and sedentary time. Conclusion: This study provides the first evidence that an exercise intervention is associated with reduced PTSD and depressive symptoms, reduced waist circumference, and improved sleep quality.
S. Rosenbaum1,2,
C. Sherrington1, A. Tiedemann1 1 Musculoskeletal Division, The George Institute for Global Health and School of Public Health, University of Sydney, Sydney, NSW, and 2St John of God Health Care Richmond Hospital, Sydney, NSW, Australia
Key words: exercise; post-traumatic stress disorder; randomized controlled trial; physical activity Simon Rosenbaum, The George Institute for Global Health, Lv 13, 320 Kent St, Sydney, NSW 2000, Australia. E-mail: [email protected]
Accepted for publication October 27, 2014
Significant outcomes
• When augmented with usual care, structured exercise incorporating resistance training and walking improves symptoms for people with post-traumatic stress disorder (PTSD) beyond usual care alone.
• Exercise improves symptoms of depression and cardiometabolic risk in people with PTSD. Limitations
• A higher rate of loss to follow up in the physical assessments could be addressed by the inclusion of home visitation to collect follow-up data.
• While the higher inclusion of men in the current study is reflective of gender rates among police and servicemen who attended the in-patient facility, this does limit the generalizability of the findings to females.
350
Exercise and PTSD Introduction
Post-traumatic stress disorder (PTSD) can occur following exposure to a potentially traumatic event, causing clinically significant social and occupational impairment and characterized by symptoms of hyperarousal, re-experiencing, negative cognitions and mood, and avoidance (1). Lifetime prevalence of PTSD within the general population is estimated at between 5% and 10% (2), with occupation-specific rates among combat veterans and police officers significantly higher (e.g. up to 17%) (3, 4). Common comorbid psychiatric diagnoses include major depressive disorder and generalized anxiety disorder, which have both been the subject of comprehensive investigations as to the efficacy of exercise as an intervention (5, 6). People with PTSD are also more likely to suffer from metabolic disturbances and poor physical health (7). Despite the known benefits of exercise for improving cardiometabolic health (8), no studies to date have assessed the impact of exercise on physical health outcomes of people with PTSD. People with PTSD are also less likely to engage in regular physical activity compared with pre-PTSD levels (9). The antidepressive and anxiolytic effects of physical activity and exercise for people with serious mental illness are established (6, 10). Despite significant heterogeneity among interventions and study designs, exercise appears to be effective beyond the placebo effect for depressive and anxiety disorders (6). Given the positive impact of regular physical activity and exercise on physical health, clinical trials to investigate the impact of exercise in people with PTSD are justified. A 2010 Cochrane Collaboration review into the effect of sports and games on PTSD highlighted the lack of available evidence for exercise as a treatment or cotreatment option (11). The review identified five studies; however, the generalizability of the findings was limited due to significant methodological weaknesses, including a lack of randomization, small sample sizes, and the inclusion of interventions evaluating play-based therapy, considered a psychological intervention. One study that specifically examined an exercise intervention was a non-randomized trial of 12 female adolescents diagnosed with PTSD who completed a 5-week aerobic exercise programme (12). The authors found a significant reduction in symptoms as measured by the Childhood PTSD Symptom Scale (12), but the impact on measures of physical health was not examined. An interrelationship between weight loss and PTSD symptoms has been demonstrated in a prospective study of weight loss facility attendees (13),
suggesting the importance of accessing both psychological and physical health indicators. Sleep disturbances are common, debilitating symptoms of PTSD, with evidence suggesting that treatment strategies aimed at improving sleeprelated symptoms may also help to improve other symptoms of PTSD (14). Exercise has been found to be effective in improving sleep quality among the general population, and the antidepressive and anxiolytic effects of exercise have been proposed as potential mechanisms for this improvement (15). Despite this potential relationship, no previous studies have evaluated exercise as a strategy to improve sleep quality among people with PTSD. Aims of the study
The aim of this study was to investigate the impact of a 12-week combined resistance and walkingbased exercise programme as an augmentation to usual care for patients hospitalized for treatment for post-traumatic stress disorder. The effect of exercise on symptoms of post-traumatic stress (primary outcome), depression, anxiety and stress ratings, body composition, physical activity levels, sleep, mobility, and strength was assessed.
Material and methods Design
We applied an assessor-blinded randomized controlled trial (RCT) with ethical review provided by St John of God Health Care Ethics Committee (REF: 412) and The University of Sydney Human Research Ethics Committee (Protocol No: 14091). The research protocol (16) and the trial were registered with the Australian and New Zealand Clinical Trials Registry prior to commencement of participant recruitment (ACTRN 12610000579099). The study was designed with reference to the CONSORT extension for nonpharmacological treatment interventions (17) and took place between September 2010 and May 2012. To ensure allocation concealment, randomization to study groups was undertaken by an investigator (AT) not involved in recruitment using a block randomization sequence generated using random numbers in Excel and including randomly varying block sizes of 2, 4, and 6. Baseline assessment and questionnaires were completed prior to allocation of participants to a study group. An assessor blinded to group allocation collected outcome data at 12 weeks postrandomization. A priori sample size calculation determined that 80 participants (40 per 351
Rosenbaum et al.
Assessed for eligibility (n = 314)
Randomized (n = 81)
Allocated to usual care (n = 42) Received usual care (n = 42)
Excluded (n = 233) Did not meet inclusion criteria (n = 2) Out of area and unable to participate in follow up or declined to participate (n = 228) Pregnant (n = 3)
Allocated to intervention (n = 39) Received allocated intervention (n = 39)
Unable to contact (n = 14)
Unable to contact (n = 9)
Analyzed (n = 28)
Analyzed (n = 30)
Fig. 1. CONSORT 2010 flow diagram.
group) were required to allow 80% power to detect as significant at the 5% level a 5-point between-group difference on the primary outcome, the PTSD checklist–civilian version, with a standard deviation of 9.4 and a correlation between baseline and follow-up measures of 0.65 (18) allowing for 15% drop-out (Fig. 1). Participants
Eighty-one in-patients from the PTSD in-patient unit at St John of God Health Care’s Richmond Hospital, Australia, were recruited through direct contact with the lead author (SR), who introduced the study during regular ward meetings. Inclusion criteria were as follows: (i) men and women aged over 18 years; (ii) psychiatrist-confirmed DSM-IVTR diagnosis of primary PTSD; (iii) medical clearance to participate in an exercise programme; and (iv) cognitively able to provide consent to participate. Exclusion criteria were as follows: (i) medically unfit to participate in an exercise programme (for example recent acute cardiac event, unstable angina, acute embolus or infarction, or acute systematic infection); (ii) pregnant or planning pregnancy in the proceeding 12 months; and (iii) complex PTSD with trauma occurring in childhood only. 352
Intervention
Intervention details can be seen in the exercise reporting grid in Table S1 (19). The 12-week intervention consisted of a weekly supervised exercise session, two unsupervised home-based exercise sessions, and a walking programme facilitated by the provision of a pedometer and exercise diary. Supervised sessions were conducted in a gymnasium within the hospital and were supervised by an exercise physiologist (SR). Unsupervised sessions were home-based and utilized the same exercise programme as the supervised sessions. All required exercise equipment (exercise bands and pedometer) were provided to participants to ensure standardization between the supervised and unsupervised sessions. The exercise protocol was designed with reference to the American College of Sports Medicine (ACSM) guidelines for resistance training (20). It consisted of three sets of 10 repetitions of up to six exercises, with a rest period of between 30 and 60 s between sets. Exercise intensity was individualized as per ACSM recommendations and depending on results from the physical assessment, the International Physical Activity Questionnaire (IPAQ) and using the Borg Rating of Perceived Exertion Scale (RPE) (21, 22) in which participants were advised
Exercise and PTSD to aim for an exertion intensity of between 12 and 17 of 20 (i.e. ‘comfortably hard’ to ‘very hard’). An increase in load of up to 10% (based on RPE) was applied when the individual could perform the current workload for 1–2 repetitions over the desired number, based on ACSM guidelines (20, 23). Participants who were more physically active at baseline were expected to require less time on average to become familiar with the exercises and would therefore be able to achieve the recommended intensity and volume more quickly. Participants experiencing a higher severity of PTSD symptoms (based on participant report and liaison with clinical staff) were expected to be less engaged and less willing to complete sessions of longer duration, and the programme was adjusted accordingly by reducing the volume and/or intensity to maximize session adherence. Intervention group participants were provided with an Omron HJ109 pedometer (Omron, UE Square, Singapore) and exercise diary to quantify and record daily step count. The pedometer and exercise diary were used to help guide goal setting and act as a motivational tool. Participants were encouraged to aim for an ultimate daily target of 10 000 steps (24); however, this was individualized based on baseline physical activity levels to promote adherence and the achievement of realistic goals. Usual care
Both groups received usual care involving a combination of psychotherapy, pharmaceutical interventions, and group therapy facilitated by psychologists. These interventions included ‘arousal reduction’, for example, breathing retraining and relaxation skills, psychoeducation (involving information about PTSD and its treatment), cognitive therapy, and communication skills. Access to these treatments was equal across both groups for the duration of the study. Average length of stay for the in-patient programme was 3 weeks, with most patients continuing a less-intensive out-patient programme for several months. Outcome measures
Primary outcome. The primary outcome was PTSD symptoms as assessed by the PTSD checklist–civilian version (PCL-C) (25). The PCL-C is a 17-item self-report questionnaire that assesses the key symptoms of PTSD and is one of the most commonly used self-report measures of PTSD (26). Scores range from 17 to 85, with higher scores
indicating higher symptom severity and a score of 45 normally selected as the cutoff for a diagnosis of primary PTSD (25). Secondary outcomes. The Depression Anxiety and Stress Scale (DASS) is a 42-item self-report instrument that measures the related negative emotional states of depression, anxiety, and tension/stress. Higher scores on the DASS equate to greater symptom severity. For the depression domain, scores of 0–13 are considered normal or mild, 14– 20 moderate, 21–27 severe, and >28 extremely severe. For anxiety, 0–7 is considered normal, 8–9 mild, 10–14 moderate, 15–19 severe, and >20 extremely severe. For stress, 0–14 is normal, 15–18 mild, 19–25 moderate, 26–33 severe, and >34 extremely severe (27). The psychometric properties of the DASS have been comprehensively evaluated, and it has been found to be valid, consistent, and responsive to treatment (27). Given the poor documented physical health profile of people experiencing PTSD, measures of physical health were assessed including resting heart rate, blood pressure, body mass index (weight in kilograms/height in meters2), body fat percentage (obtained using a bio-impedance scale; Tanita, Kewdale, Western Australia, 2011), and waist circumference (defined as the point midway between the iliac crest and costal margin) by an assessor blinded to group allocation. Self-reported physical activity levels were assessed with the International Physical Activity Questionnaire–Short Form (IPAQ-SF) (28). The IPAQ-SF includes four questions that assess participation in vigorous and moderate intensity exercise, and time spent walking and sitting in the previous 7-day period. Sleep behaviour including total sleep time and barriers to sleep were assessed using the Pittsburgh Sleep Quality Index (29) and the Pittsburgh Sleep Quality Index Addendum for PTSD (30). Both instruments are self-report questionnaires that assess sleep quality over the past month. The PSQI has 10 items and total scores range from 0 to 21, with scores >5 associated with poor sleep quality (29). The PSQIA has 10 items and total scores range from 0 to 21, with a score of four previously shown to discriminate adults with PTSD (30). The 6-min walk test was used to assess cardiorespiratory fitness (31) and was administered by an assessor blinded to group allocation within the hospital gymnasium using an 8-m long walkway. Participants were instructed to walk the length of the walkway as many times as possible within the 6 min. Minimal encouragement was provided as per recommendations, and both RPE and total
353
Rosenbaum et al. Table 1. Baseline characteristics of participants in intervention and control groups. Means (standard deviations) unless stated otherwise
Characteristics
Intervention (n = 39)
Usual care (n = 42)
47.1 (11.3) 3 (8) 34 (88)
52.0 (12.7) 10 (24) 37 (89)
2 (5) 2 (5) 7 (18)
3 (7) 2 (5) 8 (19)
7 (18) 16 (41) 16 (41) 37 (95)
1 (2) 16 (38) 25 (60) 38 (91)
7 (18) 6 (15) 7 (18) 19 (49)
10 (23) 4 (8) 4 (8) 26 (61)
9 (23) 4 (10) 5 (14) 21 (53)
9 (21) 3 (7) 6 (15) 24 (57)
15 (39) 9 (23) 3 (8) 12 (30)
18 (42) 11 (26) 3 (8) 10 (24)
Age (years) Female gender, n (%) Trauma experienced through occupation, n (%) History, n (%) Diabetes Cardiovascular disease Hypertension Body mass index (kg/m2), n (%) Desired (18.5–24.9) Overweight (25–29.9) Obese (≥30) PCL-C ≥ 45,* n (%) DASS depression domain Normal mild (0–13) Moderate (14–20) Severe (21–27) Extremely severe (≥28) DASS anxiety domain Normal mild (0–9) Moderate (10–14) Severe (15–19) Extremely severe (≥20) DASS stress domain Normal mild (0–18) Moderate (19–25) Severe (26–33) Extremely severe (≥34)
PCL-C, post-traumatic stress disorder checklist–civilian; DASS-42, Depression Anxiety and Stress Scale. *Recommended diagnostic criteria.
number of laps were recorded at the end of each minute (31). Grip strength was assessed using an adjustable handgrip dynamometer (TTM Original Dynamometer, Tokyo, Japan) with a single assessment carried out using the dominant hand (32). Intervention adherence and adverse events
Attendance at the supervised exercise sessions was recorded by the exercise physiologist (SR). Participants were instructed to record their home-based exercise sessions on exercise diaries provided. Adverse events associated with exercise participation (e.g. muscle soreness, chest pain) that required medical intervention or interfered with activities of daily living for more than 2 days were recorded on the exercise diaries. Data analysis
Analyses used all available data as per an intention-to-treat approach. Data were analyzed using SPSS Version 21 (Armonk, NY, USA) and STATA Version 12.1 statistical software (StataCorp LP, College Station, TX , USA). The between-group 354
difference at follow-up [mean difference (MD) and 95% confidence interval] was calculated for all outcome measures using linear regression models with group as the independent variable, follow-up score on the outcome measures as the dependent variable, baseline score on the outcome measure as a covariate, and statistical significance set at P < 0.05. All data were normally distributed. Imputation methods were not applied given evidence of the subsequent risk of bias (33). Partial eta-squared (g2p ) effect sizes were calculated where a 0.01 to
