CoChat
Sign Up
Cited by 66Physical Activity and HealthTreatment of Major DepressionMeta-analysis and systematic reviews

Challenges establishing the efficacy of exercise as an antidepressant treatment: a systematic review and meta-analysis of control group responses in exercise randomised controlled trials

Stubbs, Brendon, Vancampfort, Davy, Rosenbaum, Simon, Ward, Philip B., Richards, Justin, Ussher, Michael, Schuch, Felipe B.·2016·Sports Medicine
DOI 10.1007/s40279-015-0441-5
Open this paper in CoChat — free account

Abstract

Background Whilst previous meta-analyses have demonstrated that control group responses (CGRs) can negatively influence antidepressant efficacy, no such meta-analysis exists in exercise randomised controlled trials (RCTs).Objective The aim of this study was to conduct a systematic review and meta-analysis investigating CGRs and predictors in control groups of exercise RCTs among adults with depression.Methods Three authors acquired RCTs from a previous Cochrane review (2013) and conducted updated searches of major databases from January 2013 to August 2015. We included exercise RCTs that (1) involved adults with major depressive disorder (MDD) or depressive symptoms; (2) measured depressive symptoms pre- and post-intervention using a validated measure [e.g. Hamilton Depression Scale (HAM-D)]; and (3) included a non-active control group. A random effects meta-analysis calculating the standardised mean difference (SMD) together with 95 % confidence intervals (CIs) was employed to determine CGR.Results Across 41 studies, 1122 adults with depression were included [mean (SD) age 50 (18) years, 63 % female]. A large CGR of improved depressive symptoms was evident across all studies (SMD -0.920, 95 % CI -1.11 to -0.729). CGRs were elevated across all subgroup analyses, including high quality studies (n = 11, SMD -1.430, 95 % CI -1.771 to -1.090) and MDD participants (n = 18, SMD -1.248, 95 % CI = -1.585 to -0.911). The CGR equated to an improvement of -7.5 points on the HAM-D (95 % CI -10.30 to -4.89). In MDD participants, increasing age moderated a smaller CGR, while the percentage of drop-outs, baseline depressive symptoms and a longer control group duration moderated a larger CGR (i.e. improvement) (p < 0.05).Conclusion In order to demonstrate effectiveness, exercise has to overcome a powerful CGR of approximately double that reported for antidepressant RCTS.

AI snapshot

Written from the paper’s full text.

TL;DR

This systematic review and meta-analysis found that control groups in exercise randomized controlled trials for depression experience large improvements in depressive symptoms, with a pooled SMD of -0.920 across 41 studies and 1,122 adults. The authors conclude that exercise must overcome a control group response approximately double that reported in antidepressant trials to demonstrate effectiveness.

Key findings

  • Across 41 studies and 1,122 adults with depression, control groups showed a large improvement in depressive symptoms: SMD -0.920 (95% CI -1.11 to -0.729); trim-and-fill adjusted SMD -1.09 (95% CI -1.31 to -0.873).
  • Control group responses remained elevated in subgroups: high-quality studies (N=11) SMD -1.430 (95% CI -1.771 to -1.090); MDD participants (N=23) SMD -1.161 (95% CI -1.410 to -0.912); wait-list/TAU controls (N=26) SMD -0.642 (95% CI -0.853 to -0.431).
  • The control group improvement equated to -7.5 points on the HAM-D (95% CI -10.30 to -4.89), or -4.50 points in wait-list/TAU studies (95% CI -8.18 to -0.864), described as approximately double the improvement reported for antidepressant RCT control groups (4 points) and above the NICE 3-point clinically meaningful threshold.
  • In MDD participants, increasing age moderated a smaller control group response, while percentage of drop-outs, baseline depressive symptoms, and longer control group duration moderated a larger response (p<0.05).

Methods

  • Systematic review and meta-analysis following MOOSE/PRISMA and a PROSPERO protocol (CRD42015025333); RCTs were drawn from a previous Cochrane review and updated searches of major databases from 01/2013 to 08/2015.
  • Inclusion criteria: adults with MDD or depressive symptoms; exercise as the active arm; validated pre-post depression measure; non-active control group (e.g., TAU, wait-list, light-intensity activity); peer-reviewed journal article or published PhD dissertation.
  • Random-effects meta-analysis calculated SMDs with 95% CIs, with sensitivity analyses, subgroup analyses, meta-regression, publication bias assessment, trim-and-fill adjustment, and fail-safe N calculations.

Limitations

  • Stated by the authors: heterogeneity was present in most analyses; important participant data were often unavailable (e.g., illness duration, psychotropic medication use) and some RCTs were ineligible due to missing data; long-term control group responses could not be systematically considered; and the time period encompassed changes in diagnostic criteria for MDD.
  • Stated/evident: participants cannot be blinded to exercise, so expectations may introduce error to observed exercise effects; control group types varied widely (wait-list, TAU, placebo pill, psychotherapy, ECT, etc.), which may influence the control group response.
  • Evident from the text: 30 of 41 included studies were judged low quality/high risk of bias, and some subgroup analyses relied on very few studies (e.g., ECT, meditation, light therapy each N=1), limiting precision.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1007/s40279-015-0441-5

Cites

90 of 90 references verified0 citation errors · 0 not checkable

Statistics

11 tests recomputed · 0 with the reported and recomputed p on opposite sides of .05

Claims the paper makes

0 of 22 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Abstract

    Across 41 studies, control groups in exercise RCTs for depression experienced a large improvement in depressive symptoms (SMD -0.920, 95% CI -1.11 to -0.729).

    “A large CGR of improved depressive symptoms was evident across all studies (SMD -0.920, 95% CI=-1.11 to -0.729).”
    • Also stated in: § Results · boosted
      “Data was pooled from 41 studies and established that the control group experienced a large significant improvement (i.e., reduction) in depressive symptoms (SMD -0.920, 95% CI -1.11 to -0.729, p<0.001, Q=634, p<0.01) (figure 2).”
    • Also stated in: § Discussion · boosted
      “The present review found consistent evidence that control groups in exercise RCTs experience pronounced improvements in depressive symptoms.”
    Quote in the paperSubstantial claim
  • Findingown data
    § Abstract

    In MDD participants, increasing age moderated a smaller control group response.

    “In MDD participants, increasing age moderated a smaller CGR”
    • Also stated in: § Results · plain
      “Mean age (β=0.0498, 95% CI 0.0108 to 0.0888, z=2.51, p=0.01, R² = 0.34), the percentage of drop outs (β=-0.0197, 95% CI-0.0367 to -0.0027, z=-2.27, p=0.02, R² = 0.21), baseline depressive symptoms (β=-0.0535, 95% CI -0.1067 to-0.0002, z=-1.97, p=0.04, R² = 0.18) and longer control group duration moderated the control group response in MDD studies (β=0.0592, 95% CI 0.0003 to 0.1181, z=1.97, p=0.04,”
    Quote in the paperNot a substantial claim
  • Findingown data
    § Abstract

    High quality studies (low risk of bias) had a control group response of SMD -1.430 (95% CI -1.771 to -1.090).

    “CGR were elevated across all subgroup analyses, including high quality studies (N=11, SMD=-1.430, 95% CI -1.771 to -1.090)”
    • Also stated in: § Results · plain
      “found a larger control group response in high quality studies (SMD -1.430, 95% CI -1.771 to -1.090, p<0.001; Q=167, p<0.01)”
    Quote in the paperSubstantial claim
  • Findingown data
    § Abstract

    In participants with MDD, the pooled control group response was SMD -1.161 (95% CI -1.410 to -0.912).

    “MDD participants (N=23 SMD=-1.161, 95%CI=-1.410 to -0.912)”
    • Also stated in: § Results · plain
      “the pooled control groups response across 23 studies in participants with MDD was -1.161 (95% CI-1.41 to -0.912, Q=354, p<0.01, Begg=-0.26, p=0.08, Egger=-2.19, p=0.13)”
    Quote in the paperSubstantial claim
  • Findingown data
    § Abstract

    The control group response equated to an improvement of -7.5 points on the HAM-D (95% CI -10.30 to -4.89).

    “The CGR equated to an improvement of -7.5 points on the HAM-D (95% CI -10.30 to -4.89).”
    • Also stated in: § Results · plain
      “Data from 13 studies established that control group participants improved by -7.59 points (95% CI -10.30 to-4.889, p<0.001) on the HAM-D scale.”
    Quote in the paperSubstantial claim
  • May combine two claims. 11
  • Classified as probably not a substantial claim. 13

  • May combine two claims.

    Findingown data
    § Results

    High risk of bias on allocation concealment, ITT analyses and assessor blinding were not significant moderators across all studies or in MDD alone (p>0.05).

    “Categorical meta-regression analysis found that studies judged high risk of bias on allocation concealment, ITT analyses and assessor blinding were not significant moderators across all studies or in those in MDD alone (p>0.05”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results

    In a multivariate meta-regression of MDD studies, none of the moderators were significant, but percentage drop-out showed marginal significance (β=-0.0179, p=0.06), and the model explained over half of the heterogeneity (R2=0.59).

    “When all were entered into a multivariate meta-regression model, none were significant moderators but percentage drop-out demonstrated a marginal significance (β=-0.0179, 95% CI -0.0366 to 0.0008, z=-1.88, p=0.06) and the model explained over half of the heterogeneity (R² = 0.59).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results

    The Begg test indicated publication bias but the Egger test did not.

    “The Begg (Tau =-0.32, p=0.003) but not the Egger tests indicated publication bias (intercept =1.5, p=0.2).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning
    § Table 3

    Control group response is larger in inpatient trials than in outpatient/community trials.

    “Inpatient | 4 | -1.783 | -2.40 | -1.161 | <0.0001 | 85 | 91 | <0.01”
    • Also stated in: § Table 3 · plain
      “Outpatient/ community | 36 | -0.884 | -1.048 | -0.639 | <0.0001 | 91 | 462 | <0.01”
    • Also stated in: § Discussion · hedged
      “A hypothesis for the larger control group response in inpatient trials might be that inpatients benefit from the multidisciplinary treatment”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

    Findingown data
    § Table 3

    Control group response is greater in people with a formal MDD diagnosis than in those with depressive symptoms only.

    “MDD | 24 | -1.161 | -1.410 | -0.912 | <0.0001 | 93 | 354 | <0.01”
    • Also stated in: § Table 3 · plain
      “Depressive symptoms | 14 | -0.481 | -0.738 | -0.224 | =0.001 | 88 | 120 | <0.01”
    • Also stated in: § Table 2 · plain
      “MDD diagnosis | 24 | -0.6319 | -1.2088 | -0.0549 | 0.03 | 0.13”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Discussion

    Some heterogeneity was encountered in most analyses, but meta-regression analyses explained large portions of it.

    “First, we encountered some heterogeneity in most of the analyses. However, we were able to explain large portions of this with our meta-regression analyses.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § Introduction

    Exercise interventions, when added to TAU, have to demonstrate greater symptom reductions than potent control interventions, and failure to do so may be interpreted as a lack of efficacy.

    “Therefore, while RCTs of antidepressant medication typically contend with placebo responses of inert pills, exercise interventions, when added to TAU, have to demonstrate greater symptom reductions than potent control interventions, and failure to do so may be interpreted as a lack of efficacy.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: TAU is frequently a proven and effective intervention for depressive symptoms.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    A recent Cochrane review found that exercise improved symptoms of depression (SMD -0.62, 95% CI -0.81 to -0.42).

    “A recent Cochrane review [21] found that exercise improved symptoms of depression (standardised mean difference (SMD) -0.62, 95% CI -0.81 to -0.42).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [21] Cooney GM, Dwan K, Greig CA, et al. Exercise for depression. Cochrane. 2013;9(CD00436):6.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given
    § Methods

    This systematic review adhered to the MOOSE guidelines and PRISMA statement, following a predetermined published protocol (PROSPERO registration CRD42015025333).

    “This systematic review adhered to the MOOSE guidelines [23] and PRISMA statement [24], following a predetermined published protocol (PROSPERO registration CRD42015025333).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [23] Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283(15):2008–12. source

      Not checked in the free run.

    • cites [24] Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement. PLoS Clin Trials. 2009;6(7):1–6.

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data
    § Results

    Overall, eleven studies were judged to be of good methodological quality and at low risk of bias, and the remaining 30 were of low quality (high risk of bias).

    “Overall, eleven studies were judged to be of good methodological quality and at low risk of bias [12, 14, 37, 38, 43, 53, 54, 67, 69, 73, 74] and the remaining 30 were of low quality (high risk of bias).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    A meta-analysis quantified the placebo response for psychological outcomes in exercise studies in the general population, finding a mean effect size of 0.20 (95% CI 0.02 to 0.41).

    “Recently, a meta-analysis [20] quantified the placebo response for psychological outcomes in exercise studies in the general population, finding a mean effect size of 0.20 (95% CI 0.02 to 0.41).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [20] Lindheimer J, O’Connor P, Dishman R. Quantifying the placebo effect in psychological 5 outcomes of exercise training: a meta-analysis of randomized trials. Sports Med. 2015;45(5):693–711. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Introduction

    A large placebo response has been reported in non-pharmacological interventions for depression, with an effect size of 0.82 (95% CI 0.63 to 1) for control group responses in rTMS trials.

    “A large placebo response has also been reported in non-pharmacological interventions for depression, with a recent meta-analysis demonstrating an effect size of 0.82 (95% CI 0.63 to 1) for control group responses in trials of repetitive transcranial magnetic stimulation [19].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [19] Brunoni AR, Lopes M, Kaptchuk TJ, et al. Placebo response of non-pharmacological and pharmacological trials in major depression: a systematic review and meta-analysis. PLoS One. 2009;4(3):1–10. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    The Cochrane review results were criticised for questionable restrictions to high quality trials resulting in a shrinkage of the effect size.

    “However the results were criticised for questionable restrictions to high quality trials resulting in a shrinkage of the effect size [22].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [22] Ekkekakis P. Honey, I shrunk the pooled SMD! Guide to critical appraisal of systematic reviews and meta-analyses using the Cochrane review on exercise for depression as example. Ment Health Phys Act. 2015;8:21–36. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given
    § Discussion

    The time period examined encompassed significant changes in the diagnostic criteria used to identify patients with MDD and this may have influenced the results.

    “Finally, the time period that we examined encompassed significant changes in the diagnostic criteria used to identify patients with MDD and this may have influenced the results.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data
    § Discussion

    Some important participant data was not available across most studies, so it was not possible to investigate the influence of these data (e.g., illness duration, psychotropic medication use).

    “Second, some important participant data was not available across most studies and it was therefore not possible to investigate the influence of these data (e.g. illness duration, psychotropic medication use).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation
    § Discussion

    Participants with lower expectations are more likely to drop-out from the study protocol than those who do expect a benefit.

    “Therefore, it is plausible that participants with lower expectations are more likely to drop-out from the study protocol than those who do expect a benefit.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [81] Benedetti F. Placebo effects: understanding the mechanisms in health and disease. New York: Oxford University Press; 2009.

      Not checked in the free run.

    • Rests on an assumption: The relationship between initial expectations and adherence observed in a general population walking program applies to participants in exercise RCTs for depression.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § Discussion

    Lack of data with a number of RCTs meant they were not eligible for inclusion in the meta-analysis.

    “In addition, lack of data with a number of RCTs meant they were not eligible for inclusion in our meta-analysis.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Discussion

    A recent correlational study [81] in the general population found that initial expectations predicted adherence to a 2-week walking program.

    “A recent correlational study [81] in the general population found that initial expectations predicted adherence to a 2-week walking program.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [81] Benedetti F. Placebo effects: understanding the mechanisms in health and disease. New York: Oxford University Press; 2009.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § Discussion

    Due to limitations in the dataset, long term control group responses in the RCTs could not be considered in a systematic manner.

    “Third, due to limitations in the dataset we could not consider the long term control group responses in the RCTs in a systematic manner.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § Discussion

    Of all control group methods used, a waiting list appears to have the least improvement, although this is still in the medium effect size range.

    “Of all control group 'methods' used, a waiting list (N=11) appears to have the least improvement, although this is still in the medium effect size range.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Discussion

    NICE states that an improvement of 3 points on the HAM-D is clinically meaningful.

    “with the National Institute for Health and Care Excellence (NICE) stating an improvement of 3 points is clinically meaningful [77].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [77] Nelson ME, Rejeski WJ, Blair SN, et al. Physical activity and public health in older adults: Recommendation from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc. 2007;39(8):1435–45. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Discussion

    Considering the substantial control group response that exercise RCTs must overcome and the absolute benefits, critiques concerning the efficacy of exercise for depression may be premature, particularly in light of other health benefits of an active lifestyle.

    “Thus, considering the substantial control group response that exercise RCTs must overcome to show a benefit in depressive symptoms and the benefits in absolute terms, critiques concerning the efficacy of exercise for depression (e.g., [13]) may be premature, particularly in light of the plethora of other health benefits of an active lifestyle [10, 78, 79].”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [13] Chalder M, Wiles NJ, Campbell J, et al. Facilitated physical activity as a treatment for depressed adults: randomised controlled trial. BMJ. 2012;344:e2758. source

      Not checked in the free run.

    • cites [10] Naci H, Ioannidis JPA. Comparative effectiveness of exercise and drug interventions on mortality outcomes: metaepidemiological study. BMJ. 2013;347:f5577. source

      Not checked in the free run.

    • cites [78] Blair SN. Physical inactivity: the biggest public health problem of the 21st century. Br J Sports Med. 2009;43(1):1–2.

      Not checked in the free run.

    • cites [79] Bridge JA, Birmaher B, Iyengar S, et al. Placebo response in randomized controlled trials of antidepressants for pediatric major depressive disorder. Am J Psychiatry. 2009;166(1):42–9. source

      Not checked in the free run.

22 of 22

  • The paper relies on this without support.

    TAU is frequently a proven and effective intervention for depressive symptoms.

    “Thus, TAU is frequently a proven and effective 'intervention' for depressive symptoms.”

  • The paper relies on this without support.

    Control group participants in exercise trials for depression do not receive a conventional placebo.

    “In exercise trials among people with depression, control group participants do not receive a conventional placebo.”

  • The paper relies on this without support.

    The control group response in exercise studies is comparable to the placebo response in antidepressant trials.

    “The control group response in exercise studies is approximately double that reported in antidepressant meta-analyses.”

  • The paper relies on this without support.

    The relationship between initial expectations and adherence observed in a general population walking program applies to participants in exercise RCTs for depression.

    “A recent correlational study [81] in the general population found that initial expectations predicted adherence to a 2-week walking program.”
    “Therefore, it is plausible that participants with lower expectations are more likely to drop-out from the study protocol than those who do expect a benefit.”

  • The paper relies on this without support.

    Participants cannot be blinded to exercise training.

    “Due to the inability to blind participants to exercise training, expectations about the intervention are likely to introduce error to the observed effect of exercise [83]”
5 of 5

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The smaller effect of exercise when control groups have more participants taking antidepressants is expected because the difference in improvement is smaller when exercise is compared with effective treatments or with untreated controls.

    “This is expected as the difference on the magnitude of the improvement on depressive symptoms is smaller when exercise is compared with effective treatments, such as the use of pharmacological antidepressants, or when compared with controls without any treatment.87”

    From Exercise as medicine for depressive symptoms? A systematic review and meta-analysis with meta-regression (2023) · cites this paper as [87]

Statistics that don’t add up

Each reported p-value recomputed from its test statistic and degrees of freedom.

  • Inconsistent p-value: z=1.97, p = .04

    Minor

    The paper reports p = .04 here; recomputing from the test statistic and its degrees of freedom gives p = .049 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p = .04 · recomputed p = .049

  • Inconsistent p-value: z=-1.97, p = .04

    Minor

    The paper reports p = .04 here; recomputing from the test statistic and its degrees of freedom gives p = .049 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p = .04 · recomputed p = .049

2 of 2

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Doesn’t add upz = -1.97reported p = .04 · recomputed p = .049

    “z=-1.97, p = .04”

  • Doesn’t add upz = 1.97reported p = .04 · recomputed p = .049

    “z=1.97, p = .04”

  • Consistentz = -4.61reported p = .001 · recomputed p < .001

    “Z=-4.61, p = .001”

    The paper prints p at a floor (such as p < .001)

  • Consistentz = -0.4reported p = .688 · recomputed p = .689

    “z=-0.4, p = .688”

  • Consistentz = -2.15reported p = .03 · recomputed p = .032

    “z=-2.15, p = .03”

  • Consistentz = -2.32reported p = .02 · recomputed p = .020

    “z=-2.32, p = .02”

  • Consistentz = -1.46reported p = .14 · recomputed p = .144

    “z=-1.46, p = .14”

  • Consistentz = 1.58reported p = .11 · recomputed p = .114

    “z=1.58, p = .11”

  • Consistentz = 2.51reported p = .01 · recomputed p = .012

    “z=2.51, p = .01”

  • Consistentz = -2.27reported p = .02 · recomputed p = .023

    “z=-2.27, p = .02”

  • Consistentz = -1.88reported p = .06 · recomputed p = .060

    “z=-1.88, p = .06”

11 of 11

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Longer duration is said to increase improvement despite a positive regression coefficient

    Major

    A positive coefficient means a less negative control-group response as duration increases, consistent with the paper's interpretation of positive age coefficients as smaller improvement. The abstract instead says longer duration moderated a larger improvement.

  • Discussion reverses the stated baseline-severity association

    Major

    The abstract says higher baseline depressive symptoms moderated a larger improvement in MDD, but the discussion says higher symptoms moderated a smaller response.

  • MDD baseline-symptom regression results differ between text and table

    Major

    For the MDD analysis, the text reports β=−0.0535 and p=0.04, whereas Table 2 reports β=−0.0005 and p=0.008.

  • Multivariate MDD coefficient differs between text and table

    Major

    The text and Table 2 give different MDD-diagnosis coefficients and confidence intervals for the main multivariate meta-regression.

  • MDD duration regression results differ between text and table

    Major

    The MDD duration analysis has β=0.0592 and p=0.04 in the text, but β=0.0513 and p=0.006 in Table 2.

  • Comorbidity subgroups account for only 25 of 26 restricted studies

    Minor

    The wait-list/usual-care analysis includes 26 RCTs, but its reported subgroups without and with comorbidities contain 20 and five studies, totaling 25. The unrestricted comorbidity subgroups account for all 41 studies.

    Reported 26 · recomputed 25

  • Depression-classification study counts disagree

    Minor

    The results describe 23 MDD studies and 15 depressive-symptom studies, while Table 3 assigns 24 and 14 studies to those same categories.

  • Publication-type counts disagree

    Minor

    The characteristics section reports 35 peer-reviewed studies, whereas Table 3 reports 34 peer-reviewed studies and seven theses among the 41 studies.

  • Age-regression confidence interval changes sign between text and table

    Minor

    The text gives a negative lower CI bound for the main-analysis age coefficient, while Table 2 gives a positive lower bound.

  • Positive inpatient SMD falls outside its negative confidence interval

    Minor

    The wait-list/usual-care inpatient SMD is printed as positive 1.99, although both CI bounds are negative. The CI midpoint is −2.02, indicating at least a sign error in the reported estimate.

    Reported 1.99 · recomputed -2.02

  • NICE statement cites a different reference

    Minor

    The statement about NICE cites reference 77, but reference 77 is an article on physical activity in older adults; the NICE publication is listed as reference 76.

11 of 11

Cites

1 retracted

Flagged first, then by citations

1 of the 91 references couldn’t be matched to a known paper.

Research works printed but not linked: 5

  1. Bonnet LH. Effects of aerobic exercise in combination with cognitive therapy on self reported depression [dissertation]. Hempstead: Hofstra University; 2005.
  2. Epstein D. Aerobic activity versus group cognitive therapy: an evaluative study of contrasting interventions for the alleviation of clinical depression. Reno: University of Nevada; 1986.
  3. Orth DK. Clinical Treatments for Depression. Morgantown: West Virginia University; 1979.
  4. National Institute for Health and Care Excellence. Depression: management of depression in primary and secondary care. National Institute for Health and Care Excellence: London; 2004.
  5. Blair SN. Physical inactivity: the biggest public health problem of the 21st century. Br J Sports Med. 2009;43(1):1–2.

Cited by

Sorted by citations

Related

Authors
Stubbs, B., Vancampfort, D., Rosenbaum, S., Ward, P. B., Richards, J., Ussher, M., & Schuch, F. B.
Affiliations
  • National Health Service (GB)
  • KU Leuven (BE)
  • UNSW Sydney (AU)
  • The University of Sydney (AU)
  • St George's, University of London (GB)
  • Universidade Federal do Rio Grande do Sul (BR)
Corresponding author
Stubbs, Brendon
ORCID
Venue
Sports Medicine
Publisher
Springer Science and Business Media LLC
Published
2015-12-26
Volume
46
Issue
5
Pages
699-713
ISSN
  • 0112-1642 (ISSN-L)
  • 1179-2035
DOI
10.1007/s40279-015-0441-5
OpenAlex
W2238093882
Semantic Scholar
e4e3fa97d2d010f8ef676b5ce6ba836ebec203a1
Cited by
66
References
91
Citations by year
2026: 4 · 2025: 6 · 2024: 6 · 2023: 5 · 2022: 4 · 2021: 3 · 2020: 8 · 2019: 3 · 2018: 5 · 2017: 13 · 2016: 9
FWCI
1.10
Type
journal article
Language
en
Open access
closed
Licence
www.springer.com/tdm (tdm)
Locations
  • Sports Medicine (publishedVersion)
  • PubMed (publishedVersion)
  • St George's Online Research Archive (St George's University of London) (acceptedVersion)
  • Stirling Online Research Repository (University of Stirling) (submittedVersion, public-domain)
  • Research Portal (King's College London) (submittedVersion)
  • Lirias (KU Leuven) (submittedVersion)
Topics
  • Health Sciences › Medicine › Physiology › Physical Activity and Health
  • Health Sciences › Medicine › Pharmacology › Treatment of Major Depression
  • Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
Keywords
physical activity, antidepressant efficacy, major depressive disorder, depressive symptoms, placebo effect, Hamilton Depression Rating Scale, age, dropout
MeSH terms
Depressive Disorder, Exercise Therapy, Female, Humans, Male, Middle Aged, Randomized Controlled Trials as Topic, Control Groups

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review