Background
Methods
Literature search strategy
Study selection
Inclusion and exclusion criteria
Data extraction
Risk of Bias
Analysis
Results
Characteristics of studies comparing exercise vs waiting list control groups
Author | Study Size | Diagnostic criteria | Intervention | Control Group | Were groups matched for time spent with the trainer? | Long Term Follow up | Outcome measure |
---|---|---|---|---|---|---|---|
Herring 2012 (a)[35] | n = 20 | Patients meeting DSM IV criteria for Generalised Anxiety Disorder. | Supervised aerobic exercise, twice per week for 8 weeks. | Untreated waiting list group.. | No | No | Penn State Worry Questionnaire (PSWQ) |
Herring 2012(b)
| n = 20 | Patients meeting DSM IV criteria for Generalised Anxiety Disorder | Resistance Exercise. | Untreated Waiting list control group | No | No | Penn State Worry Questionnaire (PSWQ) |
n = 54 | Patients with DSM IV social anxiety disorder as measured on the ADIS-IV-L scale | 3 sessions of aerobic exercise per week over two months. 2 individual and 1 group. Intensity and Heart rate not measured. Duration 10 weeks | Untreated SAD group | No | No | Liebowitz Social anxiety self-report scale (LSAS-SR), Social Interaction Anxiety Scale (SIAS-S) | |
n = 35 | Elevated Anxiety Sensitivity > 25 on the Anxiety Sensitivity Index | Six 20 min aerobic exercise sessions over two weeks on a treadmill. | Untreated waiting list control group. | No | Follow up at 3 weeks | Anxiety Sensitivity Index (ASI) Beck Anxiety Inventory (BAI) | |
n = 74 | Patients meeting DSM IV criteria for Generalised Anxiety Disorder, Panic Disorder or Social Phobia | 8 week programme of 30 min walking sessions, measured with pedometer, increasing to 5 sessions per week. Intensity not measured. Also had CBT and education | CBT and education | Yes, sessions with the exercise trainer were matched with CBT educational sessions in control group. | The Depression Anxiety Stress Scale (DASS 21) | ||
n = 51 | Students with clinically raised anxiety levels | Internet delivered programme, over a 10 week time period. Activity measured with an accelerometer. Counselling. | Counselling | No | No | State trait anxiety Inventory (STAI trait) | |
Broman-Fulks, 2008 [33] | n = 24 | Score higher than non-clinical mean on Anxiety sensitivity score | Aerobic exercise 6 twenty min. Sessions over a 2 week time period. Subjects fitted with polaris monitor. HR 60–90% maximum | waiting list control | No, controls attended to fill in the ASR-I only. | No | Anxiety Sensitivy Index (ASI) |
Wedekind 2010 [38] | n = 37 | DSM IV criteria for panic disorder with or without agoraphobia. | Exercise for 45 mins 3 times per week, initially walking increasing to running. Duration was 10 weeks. | Relaxation plus placebo | No | No | Clnical Global Impression Scale (CGI) |
n = 31 | Patients meeting the DSM III criteria for panic disorder and Agoraphobia | A four mile route to be walked, progressing to running 3 times per week for 10 weeks. Also one training session per week. | Placebo group | No | No | HADS, Panic and Agorsphobia scale (P&A) | |
n = 36 | Menopausal women with clinically raised anxiety according to the HRSA scale | Six month programme of 6o min sessions 3 times per weeks including aerobic and resistance exercise performed to 60–80% HR max. | Waiting list control group | No | No | Hospital Anxiety and Depression Scale (HADS) | |
n = 60 | Elevated Anxiety Sensitivity > 25 on the Anxiety Sensitivity Index | Six 20-min moderate intensity aerobic exercise sessions over two weeks (three sessions per week), | Untreated waiting list control group. | No | No | Anxiety Sensitivity Index (ASI) Beck Anxiety Inventory (BAI) |
Characteristics of studies of high intensity vs low intensity exercise
Author | Study Size | Diagnostic criteria | Intervention | Control group | Were groups matched for time spent with the trainer? | Long term follow up | Outcome measure |
---|---|---|---|---|---|---|---|
Broman-Fulks, 2004 [43] | n = 54 | Students with a score of 25 or more (0.75 SD over the mean) on the Anxiety Sensitivity Index | Six 20 min treadmill sessions at high intensity for 2 weeks. Polaris heart monitor to assess HR. 60–90% HR max. | Six 20 min walking at low intensity over two weeks. | Yes | Measurements repeated one week later | Anxiety Sensitivity Index (ASI), and State Trait Anxiety Inventory (STAI trait) |
n = 40 | Non-psychotic Inpatients with anxiety disorders | Jogging, 30 mins 3 or 4 times per week for 8 weeks at 70% HR max | Walking for 3 or 4 times per week over 8 weeks at a comfortable speed. | Yes | Yes, 6 months later | State Trait Anxiety Inventory (STAI trait) | |
n = 33 | Volunteers with anxiety levels in the “borderline” or “definite” range on HADS scale and/or raised score on the Profile of Moods States. | 10 weeks of one supervise and three unsupervised sessions. 20 mins of continuous walking at 60–65% HR max. | 10 weeks of one supervise and three unsupervised sessions. Strength, mobility and flexibility not above 50% HR max. | Yes | Yes at 3 months | State Trait Anxiety Inventory (STAI trait) | |
n = 47 | Participants had Panic Disorder according to DSMIV criteria | 30 mins on treadmill 3 times per week for 8 weeks to 70% VO2 max. | Low intensity stretching exercises for 30 mins 3 times per week for 8 weeks | Yes | Yes at 7 months | Hamilton Anxiety Scale (Ham-A), BAI | |
n = 79 | Patients meeting DSM III criteria for panic disorder with or without agoraphobia, Generalised anxiety disorder or social phobia. | Brisk walking or jogging to 70% VO2 max, Trained 3 times per week for 8 weeks. | Anaerobic training - strength, flexibility and relaxation - low intensity. | Yes | No | Comprehensive Psychopathological Rating Scale (CPRS), Phobic Avoidance Rating Scale (PARS) |
Risk of Bias in studies
Study | Random Sequence | Allocation Concealment | Blinding Participants | Blinding Outcome | Incomplete Outcome | Selective Reporting | Attention Bias |
---|---|---|---|---|---|---|---|
Herring et al., 2012 [35] | Blocked randomisation stratified according to medication use. | Clinicians performing the initial assessment blinded to allocation | Participants not blinded X | Not blinded. X | All allocated subjects completed the study. | All outcomes were reported. | Low intensity and high intensity groups matched for therapist time but not waiting list control group. |
Jazaieri et al., 2012 [34] | Randomised using Efrons randomisation procedure | No information (?) | Participants not blinded (X) | No information regarding assessor blinding (?) | No difference in attrition between groups. | Patients self-reported the quantity of exercise performed. (?) | patients were not matched for time spent with an instructor. (X) |
Smits et al. 2008 [36] | Computer Generated Random Sequence | No allocation concealment (X) | Participants were not blinded (X) | . Not blinded. (X) | No difference in attrition between groups. | All outcomes were reported. | Attention bias, patients were not matched for the spent with an instructor. (X) |
Merom et al. 2007 [23] | Computer Generated Randomisation | Performed by independent research centre | Participants not blinded (X) | Not Blinded (X) | No difference in attrition between groups. | All outcomes were reported. | Groups had matched time with therapist. |
Mailey et al., 2010 [37] | No information about method of randomisation (?) | No information (?) | Participants not blinded (X) | Not blinded (?) | Very little attrition in both groups. | All outcomes were reported. | Intervention and control groups not matched for time (X) |
Broman-Fulks, 2008 [33] | No information about the method of randomisation (?) | No information (?) | Participants were blinded as to the group they were in | Not Blinded (X) | All allocated subjects completed the study. | All outcomes were reported. | Attention bias, patients were not matched for the spent with an instructor. (X) |
Wedekind 2010 [38] | Blocked randomisation | Allocation performed by the hospital pharmacist | Participants not blinded (X) | Blind rater used to eliminate expectation differences (?) | No significant difference in drop outs between groups. | All outcomes were reported. | Subjects in the control group received time with a therapist doing relaxation training. |
Brooks et al. 1998 [39] | No information about method of randomisation (?) | No information (?) | Participants were blinded. | Investigators were blinded | Similar attrition rate | All outcomes were reported. | Time matched with therapist across groups |
Villaverde et al. 2012 [40] | Details given regarding the randomisation procedure | No information (?) | Participants not blinded (X) | Not Blinded (X) | Similar attrition rate between groups. | All outcomes were reported. | No time matching between groups (X) |
Medina et al. 2015 [41] | Not mentioned in the method section. (X) | No information (?) | Participants not blinded (X) | Not Blinded (X) | Similar attrition rate | All outcomes were reported. | No time matching between groups (X) |
Broman-Fulks, 2004 [43] | No information about method of randomisation (X) | No information (?) | Participants not blinded (X) | Not Blinded (X) | No attrition in the groups | All outcomes were reported. | Matched for time |
Sexton et al. 1989 [44] | No information about the method of randomisation (?) | No information (?) | Participants not blinded (X) | Not Blinded (X) | higher no. of dropout in the high intensity group (X) | All outcomes were reported. | Less risk of attention bias as both active interventions. |
Steptoe et al. 1989 [45] | No information about the method of randomisation (?) | No information (?) | Participants not blinded (X) | Not Blinded (X) | Similar attrition rate between groups. | Some outcomes were not reported (?) | Participants in both groups had similar time with therapists |
Gaudlitz et al. 2015 [46] | Randomised using block allocation | Assessment staff were blinded | Participants were blinded as to nature of the study | Study staff were blinded | Low attrition rate and all subjects accounted for | All outcomes were reported. | Time matched with therapist across groups |
Martinsen et al., 1989 [42] | Randomised using block allocation | No information (?) | Participants not blinded (X) | Assessors were not blinded | Low attrition rate | All outcomes were reported. | Groups were time matched |
Aerobic Exercise compared to Placebo for the treatment of Anxiety | ||||
---|---|---|---|---|
Patient or population: Patients with raised anxiety levels on a validated rating scale or diagnosed with Anxiety disorders. Intervention: Aerobic exercise for anxiety, Comparison: Non exercise Control Groups. | ||||
Outcomes | Risk with Aerobic exercise for anxiety | № of participants (studies) | Quality of the evidence (GRADE) | Comments |
Improvement in anxiety scores in patients who exercised compared to no exercise. Assessed with: Evidence based anxiety rating scales follow up: range 2 weeks to 10 weeks | 0.41 SMD lower (0.70 lower to 0.12 lower)* | 194 (6 RCTs) | ⨁⨁⨁◯MODERATE a, | The risk of bias noted is due to the lack of blinding of participants to the intervention in most of the studies. Also bias due to time spent with supervisor |
High intensity exercise compared to low intensity exercise for Anxiety Disorders | ||||
Patient or population: Anxiety Disorders Intervention: High intensity exercise Comparison: low intensity exercise | ||||
Outcomes | Risk with High intensity exercise | № of participants (studies) | Quality of the evidence (GRADE) | Comments |
Results of groups who undertook high intensity exercise compared with those who undertook low intensity exercise. Assessed with: Evidence based anxiety rating scales follow up: range 2 weeks to 10 weeks | 0.38 lower (0.68 lower to 0.08 lower)* | 174 (4 RCTs) | ⨁⨁⨁◯MODERATE a | The risk of bias noted is due to the lack of blinding of participants to the intervention in most of the studies. Time with supervisor was matched in these trials |
Long term High intensity exercise compared to Long term low intensity exercise for Anxiety Disorders | ||||
Patient or population: Anxiety Disorders, Intervention: Long term High intensity exercise,Comparison: Long term low intensity exercise | ||||
Outcomes | Risk with Long term High intensity exercise | № of participants (studies) | Quality of the evidence (GRADE) | Comments |
Improvement in anxiety levels in high intensity groups compared to low intensity groups over a longer time period. Assessed with: Evidence based anxiety scores follow up: range 3 months to 7 months | - 0.33 SMD lower (0.74 lower to 0.08 lower)* | 96 (3 RCTs) | ⨁⨁⨁◯MODERATE a | The risk of bias noted is due to the lack of blinding of participants to the intervention in most of the studies. Time with supervisor was matched in these trials |
Analysis
Exercise vs waiting list control group
Mean (SD) | Total | Mean (SD) | Total | SMD (95% CI) | Outcome | |
---|---|---|---|---|---|---|
Broman-Fulks, 2008 [33] | 27.92 (15.36) | 12 | 41 (25.68) | 12 | − 0.60 (− 1.42, 0.22) | Exercise lead to significant reductions in exercise sensitivity compared to the untreated group |
Jazaieri, 2012 [34] | 61.41 (28.64) | 25 | 65.42 (21.37) | 29 | −0.16 (− 0.69, 0.38) | Exercise lead to non-significant reductions in anxiety compared to the untreated group. |
Herring, 2012(a) [35] | 59.3 (7.38) | 10 | 65.5 (7.62) | 5 | −0.79 (− 1.71, 0.13) | Exercise lead to non-significant reductions in anxiety compared to the untreated group, |
Herring 2012(b) [35] | 61.10 (10.01) | 10 | 65.5 (7.62) | 5 | −0.44(−1.53,0.65) | Resistance exercise lead to non-significant reductions in anxiety compared to the untreated group, |
Smits, 2008 [36] | 10.19 (6.54) | 16 | 18.26 (10.24) | 19 | −0.92(−1.62,-0.22) | Exercise and exercise + CBT both lead to statistically significant reductions in both the ASI and BAI |
Merom, 2008 [23] | Not Reported | 38 | Not Reported | 36 | −0.16 (− 0.77, 0.45) | CBT and exercise lead to a greater, non-significant improvement in DASS-21 scores compared to CBT and education. |
Mailey, 2010 [37] | 44.05 (18.02) | 26 | 47.23 (9.29) | 25 | −0.22(− 0.77,0.33) | There was a small and non-significant improvement in anxiety and depression in the exercise group. |
Wedekind 2010 [38] | Not Reported | 20 | Not Reported | 17 | Not Reported | Exercise and relaxation both lead to reductions in anxiety, not statistically significant. F value = 3.7 |
11.5 | 16 | 22.8 | 15 | Not Reported | Exericise lead to significant improvement in symptoms but not as effective as Clomipramine F value = 13.4 | |
16.76 | 17 | 15.02 | 19 | Not Reported | There was a small non-significant improvement in the exercise group. | |
Not Reported | Not Reported | Not Reported | Not Reported | Not Reported | There was greater but non-significant improvement in anxiety sensitivity for those in Exercise compared to Waiting List control F value = 26.7 |
High intensity vs low intensity exercise
Mean (SD) | Total | Mean (SD) | Total | Std Mean Difference | Outcome | |
---|---|---|---|---|---|---|
Broman-Fulks,2004 [43] | 25.03 (9.71) | 29 | 28.56 (6.01) | 25 | −0.42(−0.96,0.12) | High intensity exercise led to more rapid reductions in anxiety sensitivity than low intensity exercise |
Sexton, 1989 [44] | 41.2 (11.3) | 17 | 46.2 (12.0) | 23 | −0.42(−1.05,0.22) | Both jogging and walking led to a reduction in anxiety. Jogging led to a greater reduction than walking but this was not statistically significant. |
Steptoe, 1989 [45] | 42.3 (11.5) | 17 | 46.5 (9.1) | 16 | −0.39 [−1.08, 0.30] | The moderate exercise led to greater reductions in anxiety than the low intensity attention placebo group. |
Gaudlitz, 2015 [46] | 11.9 (7.1) | 24 | 14.3 (9.4) | 23 | −0.29 [− 0.86, 0.29] | Higher Intensity Exercise and Low Intensity exercise both led to a reduction in anxiety scores. There was further improvement of anxiety over time with a medium-sized effect in the endurance training group, but not in the control group. |
36 | 43 | At the end of the study both groups had achieved significant reductions in scores compared with admission values The differences between groups were small and not statistically significant P > 0.1 |
Results from long term follow up scores
Length of follow up | High Intensity exercise Mean (SD) | Low Intensity exercise Mean (SD) | St Mean difference | |
---|---|---|---|---|
Sexton, 1989 [44] | 6 months | 43 (10.1) | 42 (16.8) | 0.07 (−0.55,0.70) |
Steptoe, 1989 [45] | 3 months | 39.9 (10.5) | 46.6 (10.8) | −0.60 (−1.56,0.36) |
Gaudlitz, 2015 [46] | 7 months | 8.5 (7.3) | 14.2 (9.8) | −0.66 (−1.31,0.00) |
−0.30 (− 0.72,0.12) |