Evidence Review: Does supervised aerobic exercise added to SSRI treatment improve depression in adults?

Reviewed by Dr. Margarita Krasnova, MD · Last reviewed 2026-05-23

Scoping Review of Available Evidence: Supervised Aerobic Exercise as an Adjunct to SSRI Treatment for Adults with Major Depressive Disorder — Effects on Depressive Symptoms and Functional Outcomes

This document is presented as a Scoping Review of Available Evidence rather than a fully PRISMA-compliant review. The reviewer assessed the alignment of the retrieved evidence with the prespecified PICO and identified the following limitation: the included studies showed only partial overlap with the target population's prespecified PICO characteristics. A human reviewer should evaluate whether the inclusion criteria, search strategy, or PICO formulation require revision before final submission.

Abstract

Background: Major depressive disorder is a leading cause of global disability, and many patients on first-line SSRI pharmacotherapy experience residual symptoms or incomplete remission. Structured aerobic exercise has been proposed as an accessible, low-risk adjunctive strategy, yet the evidence supporting its routine use in this population remains uncertain.

Objective: This scoping review mapped the available evidence on whether early initiation of supervised aerobic exercise improves depressive symptoms and functional outcomes in adults with major depressive disorder receiving SSRI treatment.

Methods: Electronic databases were searched systematically; records were screened by a single reviewer against prespecified PICO-based eligibility criteria. Given the breadth and heterogeneity of the retrieved literature, a narrative synthesis was adopted rather than quantitative pooling. Twenty-five articles met inclusion criteria, encompassing systematic reviews, meta-analyses, randomized controlled trials, uncontrolled pilot studies, protocol-only publications, and a single case report.

Results: Three studies reported findings supporting a benefit of exercise for depressive symptoms, one pragmatic randomized trial found no significant advantage of adjunct exercise over comprehensive guideline-concordant care, and the remaining twenty-one articles yielded mixed, inconclusive, or no outcome data. Seven of the twenty-five were protocols without reported results. Uncontrolled and small pilot studies tended to report large pre-post improvements, whereas the single adequately powered trial comparing exercise to active multimodal treatment found no additional benefit. Two meta-analyses converged on a small positive effect of exercise on cognitive function in major depressive disorder, though neither isolated SSRI-treated patients. Evidence on anxiety, quality of life, and occupational functioning was sparse and largely uncontrolled. No included study precisely matched all elements of the prespecified PICO, and heterogeneity in severity, pharmacotherapy, exercise modality, and comparator intensity was substantial.

Conclusions: The available evidence suggests that structured physical activity may reduce depressive symptoms in adults with major depressive disorder, but the certainty of this finding as it applies specifically to supervised moderate-intensity aerobic exercise added to stable SSRI treatment remains low. The field is dominated by small, uncontrolled studies and protocol-stage investigations, and the single large pragmatic trial against an active comparator found no benefit, underscoring the need for adequately powered randomized trials isolating the incremental contribution of aerobic exercise in SSRI-treated populations.

Keywords

Depressive Disorder, Major, Exercise, Serotonin Uptake Inhibitors, Quality of Life, Combined Modality Therapy

1. Introduction

Major depressive disorder remains one of the leading causes of disability worldwide, imposing a substantial burden on individuals, families, and health systems. Among adults experiencing an acute depressive episode of moderate severity, selective serotonin reuptake inhibitors are the most widely prescribed first-line pharmacotherapy, yet a considerable proportion of patients achieve only partial remission. Residual depressive symptoms, persistent anxiety, diminished cognitive functioning, and impaired occupational and social performance are common even after an adequate trial of SSRI treatment. These incomplete responses erode quality of life and increase the risk of relapse, underscoring the need for adjunctive strategies that can close the gap between pharmacological response and full functional recovery.

Aerobic exercise has attracted growing interest as a candidate adjunct to antidepressant medication. Physiological pathways through which regular moderate-intensity activity may influence mood — including effects on neuroplasticity, hypothalamic–pituitary–adrenal axis regulation, and systemic inflammation — have been described in preclinical and clinical literature. Several trials have examined exercise in depressed populations, and narrative accounts generally suggest a favorable direction of effect on depressive symptom severity. What remains uncertain, however, is whether a specific, clinician-approved and supervised aerobic exercise regimen — delivered at a defined frequency, duration, and intensity alongside stable SSRI treatment — produces measurable improvements not only in mood but also in the broader functional domains that matter to patients: anxiety, cognition, social and occupational functioning, and overall quality of life. Heterogeneity in exercise prescriptions, supervision models, comparator conditions, and outcome measurement across existing studies makes it difficult to draw firm conclusions from the literature as it stands. A focused synthesis of the available evidence is therefore warranted.

This review aims to evaluate whether early initiation of a structured, supervised aerobic exercise program — performed three times per week for twelve weeks at moderate intensity — improves depressive symptom severity and functional outcomes in adults aged 18–71 with moderate major depressive disorder who are receiving stable first-line SSRI treatment. The comparator is treatment as usual, defined as ongoing pharmacotherapy and routine psychiatric follow-up without any structured exercise component. Outcomes of interest include depressive symptom severity, anxiety symptoms, quality of life, cognitive functioning, and occupational and social functioning.

2. Methods

2.1 Search Strategy

A systematic search of PubMed, OpenAlex, ClinicalTrials.gov, and a local reference library was conducted using a structured Boolean query assembled from Population and Intervention/Exposure keyword sets derived from the study's PICO framework. The population of interest was adults aged 18–71 with a primary diagnosis of moderate major depressive disorder confirmed by DSM-5 criteria, currently experiencing an acute depressive episode, receiving stable first-line SSRI treatment for at least four weeks; excluded were those with bipolar disorder, psychotic features, active substance use disorders, or unstable medical conditions. The intervention/exposure of interest was a clinician-approved, supervised aerobic exercise program performed three times per week for twelve weeks, with sessions lasting approximately 40 minutes at moderate intensity (60–75% of maximal heart rate), including activities such as treadmill walking or cycling. The primary outcomes were depressive symptom severity, anxiety symptoms, quality of life, cognitive functioning, and occupational and social functioning. The search was conducted on 2026-05-22. Records were deduplicated before screening (see §2.3 for counts).

In addition to the database searches described above, 102 articles were identified from the reviewer's personal reference library, matched by PubMed ID, DOI, and PubMed Central ID. ClinicalTrials.gov NCT identifiers were not included in the library match. Personal reference library articles requiring reviewer verification are listed in Appendix B.

Boolean Query (verbatim)

PubMed

text (("Major Depressive Disorder"[MeSH Terms] OR "major depressive disorder"[Title/Abstract] OR "depressive disorder"[Title/Abstract] OR "moderate depression"[Title/Abstract] OR "Clinical Depression"[Title/Abstract] OR "acute depression"[Title/Abstract] OR "major depression"[Title/Abstract] OR "Depressive Symptoms"[Title/Abstract]) AND ("older adult"[Title/Abstract] OR "young adult"[Title/Abstract] OR "geriatric"[Title/Abstract] OR "Adult"[MeSH Terms] OR "adult"[Title/Abstract] OR "middle-aged"[Title/Abstract] OR "elderly"[Title/Abstract] OR "Aged"[MeSH Terms] OR "Middle Aged"[MeSH Terms] OR "adults"[Title/Abstract] OR "senior"[Title/Abstract] OR "midlife"[Title/Abstract])) AND ("aerobic exercise"[Title/Abstract] OR "aerobic training"[Title/Abstract] OR "exercise program"[Title/Abstract] OR "Exercise Training"[Title/Abstract] OR "moderate intensity exercise"[Title/Abstract] OR "Exercise, Aerobic"[Title/Abstract] OR "Exercise Therapy"[MeSH Terms] OR "aerobic"[Title/Abstract] OR "treadmill exercise"[Title/Abstract] OR "endurance exercise"[Title/Abstract] OR "cardiovascular exercise"[Title/Abstract] OR "physical exercise"[Title/Abstract] OR "exercise intervention"[Title/Abstract] OR "endurance training"[Title/Abstract] OR "Rehabilitation Exercise"[Title/Abstract] OR "treadmill walking"[Title/Abstract] OR "moderate exercise"[Title/Abstract] OR "Physical Activity Intervention"[Title/Abstract] OR "physical training"[Title/Abstract])

OpenAlex

text ("Major Depressive Disorder" AND ("older adult" OR "young adult" OR geriatric OR Adult OR adult OR middle-aged OR elderly OR Aged OR "Middle Aged" OR adults OR senior OR midlife)) AND "aerobic exercise" AND "depressive symptom severity"

ClinicalTrials.gov

text (("Major Depressive Disorder" OR "Diagnostic and Statistical Manual of Mental Disorders" OR Depression OR "Selective Serotonin Reuptake Inhibitors" OR "major depressive disorder" OR "depressive disorder" OR "moderate depression" OR "Clinical Depression" OR "acute depression" OR "major depression" OR "Depressive Symptoms" OR "depressive episode" OR "unipolar depression" OR "Depression, Involutional") AND (Adult OR Aged OR "Middle Aged" OR "older adult" OR "young adult" OR geriatric OR adult OR middle-aged OR elderly OR adults OR senior)) AND ("Exercise Therapy" OR "Physical Fitness" OR Exercise OR "Exercise Movement Techniques" OR "aerobic exercise" OR "aerobic training" OR "exercise program" OR "Exercise Training" OR "moderate intensity exercise" OR "Exercise, Aerobic" OR aerobic OR "treadmill exercise" OR "endurance exercise" OR "cardiovascular exercise" OR "physical exercise" OR "exercise intervention" OR "endurance training" OR "Rehabilitation Exercise" OR "treadmill walking" OR "moderate exercise" OR "Physical Activity Intervention" OR "physical training" OR "bicycle exercise" OR "Exercise Prescription" OR "cardiorespiratory exercise") AND ("Anxiety Disorders" OR "depressive symptom severity" OR "Quick Inventory of Depressive Symptomatology" OR "depression severity" OR "depression rating" OR depressive OR "depression score" OR "anxiety symptoms" OR "psychosocial functioning" OR "State-Trait Anxiety Inventory" OR "Beck Depression Inventory" OR "Social and Occupational Functioning Assessment Scale" OR "Anxiety States, Neurotic" OR "Montgomery-Asberg Depression Rating Scale" OR "Patient Health Questionnaire" OR "Global Assessment of Functioning" OR "Social Functioning" OR "Health-Related Quality Of Life" OR "Generalized Anxiety Disorder 7" OR "Hamilton Depression Rating Scale" OR "occupational functioning" OR "symptom reduction" OR "Neuroses, Anxiety" OR "symptom improvement" OR "life satisfaction")

The Boolean query is assembled as (Population) AND (Intervention/Exposure) only. Comparator and Outcome terms are not AND-joined into the search query, in line with the Cochrane Handbook §4.4.4: outcome and comparator filters belong in inclusion criteria and post-retrieval relevance assessment, not in the search query. This convention prevents premature exclusion of articles that report the outcome or comparator using vocabulary not in the keyword list.

Retrieved records were screened for relevance against the study goal and the prespecified PICO before full-text review and synthesis.

2.1.1 Scope Note (Scoping Review Framing)

This document is presented as a scoping review of available evidence, framed in accordance with the Arksey and O'Malley (2005) scoping review framework and the JBI scoping review methodology. The reviewer assessed the alignment of the retrieved evidence with the prespecified PICO at the end of the retrieval stage and identified the following limitation:

• the included studies showed only partial overlap with the target population's prespecified PICO characteristics

The methodological implication of scoping-review framing is that broader inclusion criteria are accepted: heterogeneous study designs, a wider range of population overlap, and variation in intervention delivery are all tolerated rather than excluded. Pooled effect estimates are not pursued; the synthesis approach is charting and narrative mapping of the available evidence.

2.2 Inclusion and Exclusion Criteria

Population: Adults aged 18–71 with a primary diagnosis of moderate major depressive disorder confirmed by DSM-5 criteria, currently experiencing an acute depressive episode, receiving stable first-line SSRI treatment for at least four weeks; excluding those with bipolar disorder, psychotic features, active substance use disorders, or unstable medical conditions.

Intervention/Exposure: Clinician-approved, supervised aerobic exercise program performed three times per week for twelve weeks, with sessions lasting approximately 40 minutes at moderate intensity (60–75% of maximal heart rate), including activities such as treadmill walking or cycling.

Comparator: Treatment as usual, defined as ongoing pharmacotherapy and routine psychiatric follow-up without any structured exercise intervention.

Outcome: Depressive symptom severity; anxiety symptoms; quality of life; cognitive functioning; occupational and social functioning.

Articles were included if they met the prespecified PICO criteria and open-access full text was available for synthesis. Full-text availability was assessed against PubMed Central and additional open-access repositories.

Eligibility Constraints

• Search executed 2026-05-22; publication-year filter: 2024–2026.
• No language filter applied at retrieval; all results returned by the database query were considered for screening.
• Study design: no design filter was applied at retrieval; study designs were classified post-retrieval (see §2.3 Study Selection).

2.3 Study Selection

971 records were identified across the searched databases (PubMed: 210; OpenAlex: 152; ClinicalTrials.gov: 500; local reference library: 109). After deduplication, 853 records remained (from 971 records prior to deduplication). Following title and abstract screening, 325 records were assessed for relevance. After relevance assessment, 325 articles met the inclusion criteria. Of these, 25 had retrievable full text and were included in the narrative synthesis.

2.4 Data Extraction

Data were extracted from each of the 25 included articles by a single reviewer using a structured extraction template. For every study, the reviewer recorded the main finding, the reported direction of effect on depressive symptom severity and related outcomes, study design, population characteristics, follow-up duration, and sample size. Effect magnitude language was captured verbatim as stated by the original authors rather than re-interpreted, so that downstream synthesis preserved the hedging and precision of each source report. Author-stated limitations were also recorded for each article to inform the subsequent risk-of-bias assessment and to contextualize the strength of individual study contributions. Where an article reported on multiple outcomes relevant to the prespecified question — such as anxiety symptoms, quality of life, cognitive functioning, or occupational and social functioning — each outcome was extracted separately within the same template record.

2.5 Risk of Bias Assessment

Risk of bias was assessed using ROBINS-I and RoB 2.

Article	Instrument	Overall judgment
Schmitter M 2025 (PMC 12438981)	RoB 2	High
Iriarte-Yoller N 2024 (PMC 11653556)	RoB 2	Some concerns
Vancampfort D 2025 (PMC 12079350)	ROBINS-I	Some concerns
Ge J 2025 (PMC 12757253)	ROBINS-I	Some concerns
Elgendy H 2024 (PMC 10968232)	ROBINS-I	High
Zhang B 2024 (PMC 11216053)	RoB 2	Some concerns
Ribeiro JA 2024 (PMC 11056580)	RoB 2	Some concerns
Tous-Espelosin M 2025 (PMC 12633128)	ROBINS-I	High
Ye J 2025 (PMC 11888926)	RoB 2	High
Larsson R 2025 (PMC 12459035)	ROBINS-I	High
Meyer JD 2025 (PMC 12379489)	RoB 2	Unclear
Tassone VK 2025 (PMC 12404371)	RoB 2	Unclear
Meyer JD 2026 (PMC 13064449)	RoB 2	Some concerns
Poveda-López JL 2025 (PMC 12591423)	ROBINS-I	High
Kong L 2024 (PMC 11021864)	ROBINS-I	High
Bozdarov J 2025 (PMC 11801631)	ROBINS-I	High
Soldevila-Matías P 2025 (PMC 11698373)	ROBINS-I	High
Kelly S 2025 (PMC 11568898)	ROBINS-I	Some concerns
Meyer JD 2024 (PMC 11460085)	RoB 2	Unclear
Huang K 2024 (PMC 11402669)	RoB 2	Some concerns
Carvalho Silva R 2025 (PMC 12315648)	ROBINS-I	Some concerns
Ren F 2024 (PMC 10869919)	ROBINS-I	Some concerns
Cavalcanti DdSA 2025 (PMC 12604159)	ROBINS-I	High
Browne J 2024 (PMC 12362477)	ROBINS-I	High
Li H 2025 (PMC 11954280)	ROBINS-I	Some concerns

2.6 Synthesis Methods

This document is presented as a scoping review of available evidence, synthesized following the Arksey and O'Malley (2005) scoping review framework and the JBI scoping review methodology. The synthesis approach is charting and narrative mapping rather than meta-analysis or pooled effect estimation. Studies were charted according to direction of reported effect, study design, and population characteristics; findings are reported descriptively across the structured narrative subsections of Section 3.2. The scoping framework was selected because the reviewer judged the retrieved studies to have insufficient overlap with the prespecified PICO to support a standard systematic review synthesis. Broader inclusion criteria are therefore accepted, accommodating heterogeneous study designs and a wider range of population overlap than a tightly-specified systematic review would require.

PRISMA 2020 Flow Diagram

The flow of records through identification, screening, eligibility, and inclusion is summarized below (PRISMA 2020 Item 16a).

Identification

971 records were identified.

• PubMed: 210
• OpenAlex: 152
• ClinicalTrials.gov: 500
• Local reference library: 109

Screening

Records removed as duplicates: 118

Records after deduplication: 853

Records screened: 853

Records excluded during screening: 528

Eligibility

Records assessed for eligibility: 325

Records excluded at eligibility: 0

Included

Records excluded (no retrievable full text): 300

Studies included in synthesis: 25

3. Results

Coverage and Validity Scorecard

Field	Value
Databases searched	PubMed, OpenAlex, ClinicalTrials.gov, local reference library
Articles identified per database	PubMed: 210, OpenAlex: 152, ClinicalTrials.gov: 500, local reference library: 109
Total identified (pre-dedup)	971
Articles after deduplication	853
Records after title and abstract screening	325
Records after eligibility assessment	325
Relevance threshold applied	≥0 (no minimum threshold applied)
Articles in synthesis	25
Fulltext retrieved per source	—
Fulltext retrieved (total)	—
Fulltext not retrieved	—
PRISMA 2020 checklist coverage	25 ADDRESSED · 2 PARTIALLY (items 7, 24a — see Pipeline Limitations §2 and Protocol and Registration) · 0 NOT ADDRESSED
Risk-of-bias distribution	High: 11, Some concerns: 11, Unclear: 3
GRADE final certainty	Very Low

All values above are counts reflecting the records retrieved and screened for this review. No composite quality or validity scores are computed; reviewers are expected to interpret these counts in context.

3.1 Study Characteristics

Study	Design	Sample Size	Population	Intervention	Comparator	Follow-up	Primary Outcome	Direction of Effect	Effect Estimate
Schmitter M, 2025	Pragmatic multicentre randomised controlled trial (RCT)	112	Outpatients (aged 18–65) with major depressive disorder (MDD) from four Dutch specialized mental health care centres, 51% female, mean age 36.6 years, ~59% receiving pharmacotherapy and 97% receiving psychotherapy	Adjunct evidence-based exercise therapy added to guideline-concordant care as usual	Guideline-concordant care as usual (CAU)	15 months post-baseline	Adjunct exercise therapy offers no additional clinical benefits over care as usual.	contradicts the study goal	b (regression coefficient, time-by-condition interaction): -0.22 (95% CI -0.72–0.29, p=0.396)
Iriarte-Yoller N, 2024	RCT protocol (randomised, single-blind, controlled experimental trial with two parallel groups)	70	Adults aged ≥18 with resistant major depression (inadequate response to ≥2 antidepressants), excluding psychosis, imminent suicidal risk, unstable medical illnesses, active substance use disorder, cognitive impairment	12-week supervised combined exercise program	Treatment as usual	12 weeks	Protocol for RCT investigating combined exercise on depressive symptoms in resistant depression.	mixed or inconclusive	NR
Vancampfort D, 2025	meta-review (umbrella review of meta-analyses)	13	Individuals with mental disorders including children, adolescents and adults with depressive disorders, anxiety disorders, schizophrenia, ADHD, and dementia from 256 RCTs across 13 high-quality meta-analyses	Adjunctive physical activity interventions	Various control conditions (as reported in included meta-analyses)	NR	Large effect sizes for adjunctive physical activity in reducing depressive symptoms.	supports the study goal	NR
Ge J, 2025	Systematic review with three-level meta-analysis and Bayesian model-based network meta-analysis	15	Adults aged ≥18 diagnosed with MDD, from 15 RCTs with 1,196 participants (36.6% female, average age 42.3 years, 52.7% mild to moderate depression)	Exercise interventions (aerobic, mind–body, resistance, and combined)	Control conditions (as reported in included RCTs)	NR	Exercise significantly improves cognitive function in adults with MDD.	mixed or inconclusive	SMD: 0.329 (95% CI 0.25–0.41)
Elgendy H, 2024	One-arm, non-randomized experimental study with pre-post design (secondary data analysis from an aborted randomized trial)	14	Adults aged 18–65 diagnosed with MDD (DSM-5), recruited from Edmonton Access 24/7 Clinic, all on antidepressant medication	14-week supervised group physical exercise program	None (single-arm study)	14 weeks	Supervised group exercise significantly reduced depressive symptoms in MDD patients.	supports the study goal	Cohen's d: 2.47 (p=0.001)
Zhang B, 2024	RCT (study protocol for a multicentre, three-armed, parallel-group, double-blinded randomised controlled trial)	126	Patients aged 18–60 with DSM-5 confirmed MDD, HDRS >8, stable medication ≥8 weeks, excluding pregnant women, psychotherapy history, comorbid psychiatric illnesses, suicidal tendency	Combined exergame and acceptance and commitment therapy (e-ACT); ACT only	Treatment as usual	24 weeks post-intervention (total ~32 weeks)	Protocol for three-armed RCT comparing e-ACT, ACT, and treatment-as-usual for MDD.	mixed or inconclusive	NR
Ribeiro JA, 2024	RCT	17	Seventeen adult women with moderate to severe MDD (DSM-5/ICD-10), not regularly exercising, BMI <35, maintaining existing medication	Short sprint interval training (sSIT): six 6–10-minute sessions over two weeks	Control (no exercise intervention)	4 weeks	Sprint interval training significantly reduced depressive symptoms and improved aerobic fitness.	mixed or inconclusive	partial eta squared (ηp² for group × time interaction): 0.489 (p=0.002)
Tous-Espelosin M, 2025	Single-arm pre-post pilot study (no control group)	18	Eighteen adults (42.5 ± 9.9 years, 66.7% women) with resistant major depression (DSM-5), recruited from a Resistant Depression Unit	12-week supervised combined exercise program (low-to-moderate intensity aerobic interval training and resistance training)	None (single-arm study)	12 weeks	Combined exercise significantly improved depressive symptoms and functionality in resistant depression.	mixed or inconclusive	mean difference: -8.31 (95% CI -15.1–-1.5, p=0.021)
Ye J, 2025	Pilot randomized controlled trial (two-arm, single-blind, parallel assignment)	34	Adults aged 18–65 with current MDD (DSM-5), HRSD17 ≥12, MoCA ≥20, excluding primary diagnoses other than MDD, substance use, acute suicidal tendencies	10-week Baduanjin exercise program	Control group receiving conventional treatments	10 weeks	No significant interaction effect between group and time for depression severity.	mixed or inconclusive	η²: 0.062 (p=0.495)
Larsson R, 2025	systematic review	7	Adults diagnosed with or showing symptoms of at least moderate depression, from controlled trials assessing indoor rock climbing interventions	Indoor rock climbing (particularly bouldering combined with mindfulness exercises)	No intervention or waitlist controls (as reported in included studies)	6 to 12 months follow-up in included studies	Rock climbing significantly reduced depression symptoms from moderate to mild levels.	mixed or inconclusive	NR
Meyer JD, 2025	RCT protocol	200	Adults aged 18–65 with DSM-5 MDD of at least mild severity (GRID-HAMD ≥8), not on mental health medications or on stable regimen ≥8 weeks	Resistance exercise training	Control condition (as described in protocol)	52 weeks	Protocol to determine clinical efficacy of resistance exercise training for MDD.	mixed or inconclusive	NR
Tassone VK, 2025	RCT protocol (pilot randomized controlled trial)	30	Sedentary adults aged 18–65 with treatment-resistant depression (≥2 adequate antidepressant trials), MADRS ≥7, <60 min MVPA/week	Remotely delivered, individualized physical activity program (MoveU.HappyU) adjunct to treatment as usual	Treatment as usual	Approximately 13 weeks	Protocol for pilot RCT assessing feasibility of individualized physical activity for TRD.	mixed or inconclusive	NR
Meyer JD, 2026	RCT	40	Adults aged 18–65 with MDD and current major depressive episode (SCID), HAMD ≥8, stable treatment ≥8 weeks, excluding BMI ≥40, substance use disorder, bipolar/psychosis	Moderate intensity acute exercise priming immediately before CBT sessions (ActiveCBT)	Quiet rest before CBT sessions (CalmCBT)	8 weeks of intervention (~12 weeks from intake)	Exercise priming before CBT showed directional engagement of therapeutic bond and remission.	mixed or inconclusive	Cohen's d: 0.36 (95% CI -0.19–0.9, p=0.20)
Poveda-López JL, 2025	Concurrent nested mixed-methods pilot study with a quasi-experimental pre-post quantitative design and a qualitative narrative design (study protocol only, no results reported)	41	Adults over 18 diagnosed with MDD, admitted to short-stay psychiatry unit at Royo Villanova Hospital, Zaragoza, under regular medical, psychological, and pharmacological treatment	Physiotherapy intervention (therapeutic exercise and health education)	NR (study protocol)	3–6 weeks	Protocol for pilot study evaluating feasibility of physiotherapy for inpatients with MDD.	mixed or inconclusive	NR
Kong L, 2024	Prospective cohort study (cross-sectional comparison)	160	80 fitness club members and 80 community residents, healthy adults without mental disorders, aged 27–47 years	Regular aerobic exercise (fitness club members)	Sedentary lifestyle (community residents)	6 months	Regular aerobic exercise improves cognitive function and ameliorates depressive symptoms via BDNF/GDNF.	mixed or inconclusive	t-test (between-group difference): -2.41 (p=0.017)
Bozdarov J, 2025	Single-arm, pre-post feasibility trial (no control group)	8	Adult outpatients aged 18–40 with DSM-5 non-psychotic MDD or GAD, taking psychiatric medication ≥4 months, from CAMH Toronto	10-week Mindfulness-Based (non-contact) Boxing Therapy (MBBT)	None (single-arm study)	10 weeks	MBBT appeared feasible with significant reductions in depression, anxiety, and distress.	mixed or inconclusive	Cohen's d: 2.79 (p=0.014)
Soldevila-Matías P, 2025	Prospective cohort study (described as part of a randomized controlled trial registered on ClinicalTrials.gov, NCT06069739)	29	Adults with psychiatric disorders (17 schizophrenia, 6 bipolar disorder, 6 MDD) and comorbid obesity, DSM-5 diagnosed, clinically stable or in remission	12-week physical activity training program	NR	12 weeks	Physical activity improved cognition, mood, BMI, and anti-inflammatory activity in psychiatric disorders.	mixed or inconclusive	Cohen's d: 0.87 (p=< .0001)
Kelly S, 2025	Cross-sectional analysis of baseline data from a randomized controlled trial (SPeED Study)	147	Adults aged 18–65 with mild to moderate MDD (n=113) and matched healthy controls (n=34), recruited from Center for Psychotherapy at Humboldt University, Berlin	Acute increasing-intensity maximal exercise bout (12–15 minutes)	Healthy controls (no MDD)	NR	Adults with MDD have favorable IGF-1 responsiveness to acute exercise, similar to controls.	mixed or inconclusive	Cohen's d: 0.02 (p=>0.05)
Meyer JD, 2024	RCT protocol	40	Adults aged 18–65 with DSM-5 MDD (SCID confirmed), GRID-HAMD ≥8, CBT-naïve within past 5 years, not in treatment or stable regimen ≥8 weeks	30 minutes moderate-intensity aerobic exercise priming before CBT sessions	Quiet rest before CBT sessions	52 weeks from enrollment	Protocol to evaluate whether exercise priming before CBT improves working alliance.	mixed or inconclusive	NR
Huang K, 2024	RCT protocol (mixed methods: assessor-blinded, parallel-group randomized controlled trial with a supplementary qualitative study)	76	College students aged ≥18 with subthreshold depression (CES-D ≥16, not meeting DSM-5 MDD criteria), no depression treatment in past 6 months, from colleges in Changchun, China	Nintendo Switch-based exergame intervention (8 weeks)	Control group (no exergame intervention)	16 weeks (8-week intervention plus 8-week follow-up)	Protocol hypothesizing exergame intervention improves depressive and anxiety symptoms in students.	mixed or inconclusive	NR
Carvalho Silva R, 2025	systematic review	30	Adults aged ≥18 diagnosed with MDD (DSM-IV/5, ICD-10/11), from RCTs evaluating physical exercise with pre- and post-intervention biomarker data	Physical exercise interventions	Various control conditions (as reported in included RCTs)	NR	Physical exercise had mixed effects on biological markers in MDD patients.	mixed or inconclusive	NR
Ren F, 2024	Systematic review and meta-analysis	NR	Adults aged ≥18 with a diagnosis of clinical depression (MDD), from randomized trials comparing aerobic exercise to non-aerobic exercise groups	Aerobic exercise	Non-aerobic exercise control groups	NR	Aerobic exercise improves overall cognitive function, memory, and executive function in MDD.	supports the study goal	g: 0.21 (95% CI 0.07–0.34)
Cavalcanti DdSA, 2025	case report	1	A single 24-year-old Brazilian female, sedentary, with mixed anxiety and depressive disorder (ICD-10 F41.2) for 9 months, on fluoxetine 40 mg/day and as-needed clonazepam	8-week combined aerobic and resistance exercise with self-selected intensity	None (single case, no comparator)	2 weeks after the 8-week intervention	Combined exercise resulted in clinically significant improvements in depressive symptoms and anxiety.	mixed or inconclusive	NR
Browne J, 2024	Retrospective analysis of electronic health record and clinical program data with propensity-score matching and linear mixed modeling	1414	Older veterans (aged ≥60) with and without serious mental illness (schizophrenia, schizoaffective, bipolar, recurrent MDD) enrolled in VHA Gerofit exercise program at eight national sites, 2010–2019	Multicomponent exercise (VHA Gerofit clinical exercise program)	Non-SMI veterans in the same exercise program (propensity-score matched)	12 months	Multicomponent exercise improved physical function in older veterans with serious mental illness.	mixed or inconclusive	NR
Li H, 2025	cross-sectional discrete choice experiment (DCE)	323	Patients diagnosed with depression by certified psychiatrists, aged ≥18, from six psychiatric wards and outpatient departments in Sichuan province, China	Exercise preference attributes (interestingness, safety, frequency, venue, guidance, format)	NR (preference study, no intervention comparator)	NR	Patients with depression most prefer exercises characterized by high interestingness and safety.	mixed or inconclusive	coefficient (mixed logit model): 0.84 (95% CI 0.6818403–0.9940637, p=< 0.01)

Per-stage exclusion reasons are summarized below (PRISMA 2020 Item 16b).

Stage	Reason	Count
Title and abstract screening	Below Threshold	528

Included studies are stratified by study design below.

Study Type	Count
Randomized controlled trials	12
Systematic reviews and meta-analyses	5
Observational studies	3
Other study designs	5

3.2 Synthesis of Findings

Note on evidence hierarchy. This review's included set mixes primary studies (RCTs and/or observational designs) with secondary evidence (systematic reviews and/or meta-analyses). Reviewers should consider double-counting risk and weight primary and secondary evidence accordingly.

1. Direction-of-Effect Summary

Of the 25 included articles, 3 (12%) reported findings that support the study goal, 1 (4%) reported findings that contradict the study goal, and 21 (84%) were classified as mixed or inconclusive in their direction of effect. The high proportion of mixed or inconclusive findings reflects the composition of the retrieved literature: seven articles were study protocols reporting no outcome data, one was a single-participant case report, and several others employed designs without controlled comparators or examined populations and interventions that overlapped only partially with the prespecified PICO.

Studies supporting the study goal (3 of 25, 12%):

One umbrella review of 256 randomized controlled trials across 13 meta-analyses (Vancampfort, 2025; n = 12,233) found large effect sizes for adjunctive physical activity in reducing depressive symptoms in adults with depressive disorders and moderate effect sizes for improved quality of life and cardiovascular fitness. This study did not report a significance metric. A meta-analysis of aerobic exercise and cognition in MDD (Ren, 2024; n = 945) found that aerobic exercise significantly improved overall cognitive function (g = 0.21; 95% CI = 0.07, 0.34). The reported confidence interval excluded zero, though no explicit p-value was provided. A single-arm pre-post study of supervised group exercise in adults with MDD on antidepressant medication (Elgendy, 2024; n = 14) reported a large effect size (Cohen's d = 2.47) with statistical significance: p < 0.001 (BDI-2); p < 0.005 (CORE-OM). Notably, this study lacked a control group, and the authors acknowledged that improvement could reflect pharmacotherapy, the combination, or placebo effects rather than exercise alone.

Study contradicting the study goal (1 of 25, 4%):

A pragmatic multicentre randomized controlled trial of adjunct exercise therapy versus guideline-concordant care as usual in MDD outpatients (Schmitter, 2025; n = 112) found no significant difference between conditions on depressive symptom reduction. The reported statistical significance was p = .396.

Studies with mixed or inconclusive findings (21 of 25, 84%):

Seven of these 21 articles were study protocols that reported no efficacy results (Iriarte-Yoller, 2024; Zhang, 2024; Meyer, 2025 [RESIST Trial]; Tassone, 2025; Poveda-López, 2025; Meyer, 2024 [CBT+ protocol]; Huang, 2024). None of these seven reported a significance metric.

Among the remaining 14 mixed or inconclusive articles, reported significance metrics were as follows. Ge (2025) reported 95% CI, 0.25 to 0.41; I² = 17.30% for a meta-analytic estimate of exercise on cognitive function in MDD. Ribeiro (2024) reported group × time interaction: F=14.332, p=0.002; factor time: F=6.562, p=0.022; pairwise comparisons with Bonferroni correction revealed pre- to post-changes in the sSIT group only (p<0.001). Tous-Espelosin (2025) reported P = 0.021 for MADRS; P = 0.006 for CGI-S; P = 0.046 for SDS. Ye (2025) reported no significant interaction effect for depression (p > 0.05); significant time effect for depression (p < 0.001); significant group effect for LF/HF ratio (p = 0.049); no significant between-group difference for HRSD17 at Week 5 (p = 0.226) or Week 10 (p = 0.306). Larsson (2025) reported p ≤ 0.05. Meyer (2026) reported WAI-Bond: p=0.20; BADS: p=0.09; remission difference: p<0.05. Kong (2024) reported p < 0.05. Bozdarov (2025) reported p = 0.014 for PHQ-9, GAD-7, and K10; p = 0.007 for MAAS; p = 0.013 for CGI. Soldevila-Matías (2025) reported p < 0.05 to p < 0.0001. Kelly (2025) reported p>0.05. Browne (2024) reported significant improvements over time for all measures (all ps < .0001); no significant differences between SMI and non-SMI groups on changes in function (all ps > .05). Li (2025) reported p < 0.01. Cavalcanti (2025) did not report a significance metric. Carvalho Silva (2025) did not report a significance metric.

In total, across all 25 articles, 11 did not report a significance metric (the seven protocols, the umbrella review by Vancampfort, the meta-analysis by Ren, the case report by Cavalcanti, and the systematic review by Carvalho Silva).

2. Consistency vs. Contradiction Analysis

The included evidence base presents a complex picture when assessed for internal consistency, and several sources of disagreement can be traced to identifiable differences in study design, population severity, comparator conditions, and outcome measurement.

Agreement on the direction of exercise effects on depressive symptoms

A cluster of studies converges on the finding that structured physical activity — particularly aerobic exercise — is associated with reductions in depressive symptom severity among adults with major depressive disorder. The umbrella review by Vancampfort and colleagues, synthesizing 256 randomized controlled trials across 13 meta-analyses, reported large effect sizes for adjunctive physical activity in reducing depressive symptoms in adults with depressive disorders, alongside moderate effect sizes for improvements in physical and psychological quality of life. The single-arm pilot by Elgendy and colleagues observed a large pre-post reduction in BDI-2 scores (Cohen's d = 2.47, p < 0.001) over 14 weeks in adults with MDD receiving concurrent antidepressant medication. Tous-Espelosin and colleagues similarly found significant pre-post improvements in MADRS scores (mean difference −8.31, p = 0.021) and functional disability (SDS, p = 0.046) after a 12-week supervised combined exercise program in adults with resistant major depression. The Ribeiro and colleagues trial, though small (n = 17), demonstrated a large group-by-time interaction (η²p = 0.489, p = 0.002) favoring sprint interval training over a control condition for HAM-D21 reduction in women with moderate-to-severe MDD.

These findings are broadly consistent with one another and with the study goal of evaluating whether supervised aerobic exercise improves depressive symptoms in adults receiving pharmacotherapy. However, the absence of controlled comparators in several of these studies — Elgendy and colleagues and Tous-Espelosin and colleagues both lacked a no-exercise control group — limits the degree to which symptom improvement can be attributed to exercise rather than to concurrent pharmacotherapy, natural remission, or nonspecific effects of social contact and clinical attention.

The principal contradiction: the Schmitter pragmatic RCT

The most direct challenge to the supportive findings comes from the pragmatic multicentre randomized controlled trial by Schmitter and colleagues, which found no significant difference between adjunct evidence-based exercise therapy and guideline-concordant care as usual in reducing depressive symptoms among 112 outpatients with MDD (b = −0.22, 95% CI −0.72 to 0.29, p = 0.396). This trial is the only completed, adequately powered randomized comparison in the included set that directly tested exercise as an adjunct to active treatment against a comparator receiving comprehensive guideline-concordant care — including both pharmacotherapy (59% of participants) and psychotherapy (97%). The null finding may reflect the strength of the comparator rather than the inefficacy of exercise: when the control condition already includes robust multimodal treatment, the incremental benefit of adding exercise may be too small to detect, particularly in a sample with high comorbidity (43.6% comorbid psychological diagnoses) and severe baseline symptoms. The authors themselves noted that varying psychological and pharmacological interventions in the care-as-usual arm may have obscured exercise-specific effects, and that COVID-19–related data loss reduced statistical power for secondary outcomes such as remission.

This pattern — positive pre-post or weakly controlled results alongside a null finding from a pragmatic trial with an active comparator — is a recognizable source of apparent contradiction in the exercise-for-depression literature. The uncontrolled studies showing large effects (Elgendy, Tous-Espelosin, Bozdarov) are susceptible to regression to the mean and expectancy effects, whereas the Schmitter trial's null result may partly reflect a ceiling on what exercise can add when patients are already receiving intensive multimodal care.

Cognitive outcomes: convergent but qualified

Two meta-analytic syntheses addressed cognitive functioning. Ren and colleagues found that aerobic exercise significantly improved overall cognitive function in adults with MDD (Hedges' g = 0.21, 95% CI 0.07 to 0.34), with sub-domain benefits for memory and executive function. Ge and colleagues reported a somewhat larger pooled effect on overall cognition (SMD = 0.329, 95% CI 0.25 to 0.41) and identified a non-linear dose–response relationship, with aerobic and mind–body exercises offering the greatest cognitive benefits. These two reviews are broadly consistent in direction and magnitude, though the Ge review's inclusion of mind–body modalities (particularly Tai Chi/Qigong) alongside conventional aerobic exercise introduces heterogeneity in the intervention definition relative to the prespecified PICO. Neither review isolated the cognitive effects of exercise specifically in patients on stable SSRI treatment, limiting direct applicability to the target population.

Quality of life and functional outcomes: sparse and indirect

Evidence on quality of life and social or occupational functioning was reported in only a handful of studies and was largely uncontrolled. Tous-Espelosin and colleagues observed a significant reduction in disability (SDS, p = 0.046), and the Cavalcanti case report documented a notable improvement in WHOQOL-Bref scores (+33.7 points) in a single patient with mixed anxiety and depressive disorder — a diagnosis that does not meet the PICO population criterion. The Vancampfort umbrella review reported moderate effect sizes for quality of life in adults with depressive disorders, but this finding was aggregated across diverse interventions and populations. No completed controlled trial in the included set reported quality-of-life or occupational functioning as a primary or adequately powered outcome in adults with MDD on SSRI treatment.

Anxiety symptoms: limited and heterogeneous

Anxiety was a secondary or incidental outcome in several studies. Bozdarov and colleagues reported a 51% mean reduction in GAD-7 scores after a 10-week mindfulness-based boxing intervention (p = 0.014), but this was a single-arm feasibility study with only eight participants, some of whom had a primary diagnosis of generalized anxiety disorder rather than MDD. The Cavalcanti case report noted a large reduction in trait anxiety (ΔSTAI-T = −25), but in a patient whose diagnosis was mixed anxiety and depressive disorder. Meyer and colleagues (the ActiveCBT trial) did not report anxiety as a separate outcome. The heterogeneity of populations and interventions across these reports precludes any consistent conclusion about the effect of supervised aerobic exercise on anxiety symptoms in the PICO-specified population.

Biological mechanism studies: informative but not outcome-relevant

Several included studies examined biological markers rather than clinical outcomes. Carvalho Silva and colleagues' systematic review of 30 trials found mixed and inconsistent effects of exercise on inflammatory markers, neurotrophic factors, and stress-response biomarkers in MDD. Kong and colleagues reported associations between aerobic exercise, BDNF levels, and cognitive function, but their sample explicitly excluded individuals with major depressive disorder. Kelly and colleagues found that IGF-1 responsiveness to acute exercise was preserved in adults with MDD, but this cross-sectional analysis did not assess clinical symptom change. Soldevila-Matías and colleagues observed improvements in cognition, mood, and anti-inflammatory biomarkers after 12 weeks of physical activity in a mixed psychiatric sample (including only 6 of 29 participants with MDD). These studies illuminate potential biological pathways but do not directly address whether supervised aerobic exercise improves the clinical outcomes specified in the PICO.

Protocol-only publications and feasibility studies

Five included records were study protocols without reported results (Iriarte-Yoller, Zhang, Meyer [RESIST], Tassone, Poveda-López, Huang), and one was a discrete choice experiment examining patient preferences for exercise characteristics (Li and colleagues). These publications could not contribute outcome data to the synthesis. The protocols describe populations and interventions that partially overlap with the PICO — several target adults with MDD on stable pharmacotherapy and employ supervised aerobic or combined exercise — but their eventual findings remain unavailable.

Sources of heterogeneity across the evidence base

The most salient sources of inconsistency across the included studies are: (1) the nature of the comparator, with studies using no-exercise or waitlist controls tending to show larger effects than the single trial comparing exercise to comprehensive guideline-concordant care; (2) population severity, ranging from mild-to-moderate MDD to treatment-resistant depression, with the latter group (Tous-Espelosin, Iriarte-Yoller, Tassone) representing a clinically distinct population from the PICO target; (3) exercise modality, spanning treadmill walking, cycling, sprint interval training, Baduanjin, bouldering, boxing, and exergaming, only some of which qualify as the moderate-intensity aerobic exercise specified in the PICO; and (4) concurrent treatment, with most participants receiving pharmacotherapy but the specific agents, durations, and co-interventions varying substantially. These differences make it difficult to isolate the contribution of supervised aerobic exercise as an adjunct to stable SSRI treatment — the specific question posed by this review.

3. Study Limitations (Aggregated)

The author-stated limitations across the included studies reveal several recurring methodological concerns that bear directly on the strength of inferences about whether supervised aerobic exercise improves depressive symptoms and functional outcomes in adults with major depressive disorder receiving SSRI treatment.

Small sample sizes and limited statistical power

The most frequently acknowledged limitation was small sample size. At least ten studies explicitly cited restricted enrollment as constraining generalizability or statistical power. Individual trial samples ranged from a single case report to 40 participants in the completed randomized trials, with only the pragmatic multicentre trial by Schmitter and colleagues enrolling more than 100 participants. Several authors noted that their studies were underpowered to detect clinically meaningful between-group differences or to conduct planned subgroup analyses. The retrospective cohort analysis of the Gerofit program, while drawing on 1,414 veterans, acknowledged that high rates of missing follow-up data (exceeding 50%) may have biased results toward those with better function.

Absence of control groups or use of uncontrolled designs

Five studies employed single-arm, pre–post designs or case-report methodology without a comparator group. Authors of these studies consistently acknowledged that the lack of a control condition precluded attribution of observed improvements to the exercise intervention itself, as opposed to concurrent pharmacotherapy, natural symptom fluctuation, social interaction with trainers, or placebo effects. Elgendy and colleagues, for instance, noted that all participants were taking antidepressant medication, making it difficult to disentangle the contribution of exercise from that of pharmacotherapy or the combination.

Inability to blind participants and personnel

Blinding was recognized as a pervasive challenge. The pragmatic trial by Schmitter and colleagues noted that patients could not be blinded to treatment allocation, which may have contributed to differential dropout between arms. The systematic review of rock climbing interventions reported that none of its seven included trials blinded patients or therapists. Bozdarov and colleagues acknowledged that their feasibility trial was delivered exclusively by its developer, raising concerns about investigator and expectation biases.

Confounding from concurrent treatments

Because most study populations were receiving pharmacotherapy, psychotherapy, or both, multiple authors flagged the difficulty of isolating exercise effects from those of co-interventions. In the Schmitter trial, 59% of participants were receiving pharmacotherapy and 97% psychotherapy, and the authors noted that varying psychological and pharmacological interventions in the care-as-usual arm may have obscured any incremental benefit of exercise. Ribeiro and colleagues similarly acknowledged that concurrent antidepressant and anxiolytic use complicated interpretation of their sprint interval training results.

Limited follow-up and absence of long-term outcome data

Intervention durations ranged from two weeks to fourteen weeks among completed studies, and few incorporated extended post-intervention follow-up. Ye and colleagues noted that their 10-week trial included no follow-up assessments, restricting understanding of long-term efficacy. The case report by Cavalcanti and colleagues evaluated outcomes only two weeks after an eight-week intervention. Where longer follow-up was planned — as in the Schmitter trial at 15 months — the authors reported that pandemic-related disruptions led to substantial missing data.

Disruptions and missing data attributable to the COVID-19 pandemic

At least two studies explicitly cited pandemic-related constraints. Schmitter and colleagues reported that COVID-19 led to an average of 28% missing data from telephone interviews and reduced statistical power for remission analyses. Bozdarov and colleagues noted that pandemic restrictions further limited implementation of their boxing therapy feasibility trial.

Restricted population diversity and selection bias

Several authors raised concerns about the representativeness of their samples. The Gerofit analysis enrolled few female or Hispanic/Latinx veterans. Tous-Espelosín and colleagues noted that their sample was drawn from a single city and may not generalize to other communities. Kong and colleagues acknowledged that their cross-sectional design, which excluded individuals with mental disorders, precluded causal inference. The discrete choice experiment by Li and colleagues reported a relatively small subsample of participants aged 60 and above and did not investigate medication use or depression severity, potentially introducing selection bias.

Reliance on self-reported measures

Two studies specifically noted the limitations of self-reported instruments. Kelly and colleagues acknowledged that the International Physical Activity Questionnaire is susceptible to over- or under-reporting, while Li and colleagues flagged potential social desirability bias in their questionnaire-based preference study.

Heterogeneity of exercise modalities and intervention fidelity

The diversity of exercise types — spanning treadmill walking, cycling, sprint interval training, Baduanjin, bouldering, boxing, exergaming, and combined aerobic-resistance protocols — was noted by several authors as complicating cross-study comparison. Tassone and colleagues observed that participants in their planned personalized physical activity program would perform various types of activity with differing access to equipment and settings, potentially affecting both adherence and generalizability. The systematic review of biological markers by Carvalho Silva and colleagues highlighted varying methodologies across included trials as a key limitation.

Taken together, the limitations most frequently reported by the study authors — small samples, uncontrolled designs, inability to blind, concurrent treatments, and short follow-up — collectively constrain the certainty with which the present body of evidence can address whether early initiation of supervised aerobic exercise confers additional benefit for depressive symptoms and functional outcomes in adults with major depressive disorder who are already receiving SSRI treatment.

4. Population Heterogeneity

The included studies enrolled populations that overlapped only partially with the prespecified PICO, and the heterogeneity across these populations limits the degree to which subgroup-level conclusions can be drawn about supervised aerobic exercise as an adjunct to SSRI treatment in adults with moderate major depressive disorder.

Depression Severity

Baseline depression severity varied considerably. Several studies enrolled participants with moderate depression, consistent with the target PICO. The pragmatic trial by Schmitter and colleagues recruited outpatients with severe baseline depressive symptoms and found no additional benefit of exercise over guideline-concordant care. Two studies and one protocol focused specifically on treatment-resistant or resistant major depression, a population meaningfully distinct from the moderate, first-episode presentation specified in the PICO. Among these, the single-arm pilot by Tous-Espelosín and colleagues observed significant pre-post reductions in depressive symptoms (MADRS mean difference −8.31, p = 0.021), while the protocol by Iriarte-Yoller and colleagues and the pilot protocol by Tassone and colleagues have not yet reported outcomes. The sprint interval training trial by Ribeiro and colleagues enrolled women with moderate to severe MDD and found significant symptom reductions in the exercise group but not the control group, though the sample comprised only 17 participants. The umbrella review by Vancampfort and colleagues reported large effect sizes for physical activity interventions in reducing depressive symptoms across children, adolescents, and adults with depressive disorders, without disaggregating by severity band. Overall, no clear pattern emerged linking a specific severity stratum to a consistent direction of effect, though the single large pragmatic trial enrolling severely affected patients contradicted the hypothesis of added benefit.

Pharmacotherapy Context

The PICO specifies adults on stable first-line SSRI treatment. Only a minority of included studies restricted enrollment to participants on SSRIs specifically. The pre-post study by Elgendy and colleagues required all participants to be on antidepressant medication, and the case report by Cavalcanti and colleagues described a patient taking fluoxetine, but most trials either permitted a range of antidepressant classes or did not specify the medication type. The Schmitter trial reported that approximately 59% of participants were receiving pharmacotherapy alongside near-universal psychotherapy, making it difficult to isolate the interaction between exercise and SSRI treatment. The cross-sectional IGF-1 study by Kelly and colleagues found no difference in exercise-induced biomarker response between participants on antidepressants and those not taking them, though this was not an outcome-level finding relevant to depressive symptom change. Because so few studies isolated SSRI-treated patients, the evidence base does not permit a subgroup analysis specific to the pharmacotherapy context defined in the PICO.

Exercise Modality and Supervision

The PICO intervention specifies supervised aerobic exercise at moderate intensity performed three times weekly for twelve weeks. Studies varied widely in modality, intensity, duration, and supervision. Aerobic-focused interventions were examined in the Ribeiro sprint interval training trial (high-intensity, six sessions over two weeks), the Elgendy group exercise study (14 weeks, supervised, mixed modalities), and the meta-analysis by Ren and colleagues, which found that aerobic exercise significantly improved cognitive function in adults with MDD (Hedges' g = 0.21, 95% CI 0.07–0.34). Mind–body and alternative modalities included Baduanjin (Ye and colleagues), indoor rock climbing (Larsson and colleagues), mindfulness-based boxing (Bozdarov and colleagues), and exergaming (Zhang and colleagues, protocol only). The Baduanjin trial found no significant group-by-time interaction for depression severity. The rock climbing systematic review reported clinically meaningful reductions in depressive symptoms compared with no intervention, though the evidence base was small. Among studies employing supervised programs, the direction of effect was mixed: the Schmitter trial (supervised, evidence-based exercise) found no benefit, while the Elgendy study (supervised group exercise) and the Ribeiro trial (supervised sprint intervals) both reported significant pre-post improvements, albeit without controlled comparisons in the former. The network meta-analysis by Ge and colleagues found that aerobic and mind–body exercises offered the greatest cognitive benefits, with a small but significant pooled effect on overall cognitive function (SMD 0.329, 95% CI 0.25–0.41).

Age and Sex

Most studies enrolled adults aged 18–65, broadly consistent with the PICO age range. The Browne and colleagues retrospective analysis enrolled older veterans aged 60 and above, a population outside the typical acute MDD treatment-seeking demographic, and found that multicomponent exercise improved physical function regardless of serious mental illness status, though this study did not report depression-specific outcomes as a primary endpoint. The Ribeiro trial enrolled only women, and the Cavalcanti case report described a single young woman; both reported symptom improvements, but neither permits sex-stratified inference. No included study reported a formal subgroup analysis by sex or age band for depressive symptom outcomes.

Study Setting

Settings ranged from specialized mental health care centers (Schmitter and colleagues), psychiatric inpatient units (Poveda-López and colleagues, protocol only), outpatient clinics (Elgendy, Bozdarov, Meyer and colleagues), community mental health centers (Tous-Espelosín and colleagues), and veteran health administration sites (Browne and colleagues). The only adequately powered randomized trial conducted in a specialized outpatient setting — the Schmitter pragmatic trial — found no benefit of adjunct exercise, whereas the smaller, uncontrolled outpatient studies tended to report pre-post improvements. This pattern may reflect differences in comparator intensity (guideline-concordant multimodal care versus routine follow-up) rather than a true setting effect, but the available data do not permit a definitive distinction.

Comorbidity Profile

Several studies enrolled participants with psychiatric comorbidities or mixed diagnostic groups. The Soldevila-Matías and colleagues cohort study included individuals with schizophrenia, bipolar disorder, and MDD alongside comorbid obesity, reporting improvements in cognition and mood symptoms across the mixed sample without MDD-specific subgroup data. The Schmitter trial noted that 43.6% of participants had comorbid psychological diagnoses. The Kong and colleagues cross-sectional study explicitly excluded individuals with mental disorders, studying healthy exercisers instead — a population outside the PICO entirely. No study provided a formal comparison of exercise effects in participants with versus without psychiatric comorbidities.

Summary of Subgroup Patterns

Across the subgroup axes discernible from the extraction data — depression severity, pharmacotherapy type, exercise modality, supervision, age, sex, setting, and comorbidity — no consistent subgroup-level signal emerged favoring or contradicting the benefit of supervised aerobic exercise for the specific PICO population. The single large pragmatic trial that most closely approximated a rigorous test of adjunct exercise in outpatients with MDD found no benefit, while smaller uncontrolled or underpowered studies tended to report improvements. Critically, no included study precisely matched all elements of the prespecified PICO — adults with moderate MDD on stable SSRI monotherapy, randomized to supervised moderate-intensity aerobic exercise three times weekly for twelve weeks versus treatment as usual without structured exercise. The population heterogeneity across the evidence base therefore represents a substantive limitation, and findings from adjacent but non-identical populations should be interpreted with caution when applied to the target clinical question.

5. Effect Magnitude Language (Aggregated)

Across the included studies that reported outcome data, the language used to characterize effect magnitude varied considerably in both terminology and scale, reflecting the heterogeneity of study designs, populations, and analytic approaches.

Among the studies reporting quantitative effect sizes, the descriptors ranged from small to very large. The umbrella review by Vancampfort and colleagues described "large effect sizes" for adjunctive physical activity in reducing depressive symptoms and "moderate effect sizes" for improvements in quality of life and cardiovascular fitness, though specific pooled statistics were not reproduced in the extraction. Ge and colleagues, synthesizing cognitive outcomes across fifteen trials, reported a small but statistically significant pooled effect on overall cognitive function (SMD 0.329, 95% CI 0.25–0.41). Similarly, Ren and colleagues found that aerobic exercise yielded a small improvement in overall cognitive function (Hedges' g = 0.21, 95% CI 0.07–0.34), with sub-domain benefits in memory and executive function.

By contrast, several smaller uncontrolled studies reported markedly larger pre-post effect sizes. Elgendy and colleagues described a "high effect size" (Cohen's d = 2.47) for depression reduction in a 14-participant single-arm study, while Bozdarov and colleagues reported Cohen's d values ranging from 2.70 to 3.90 across depression, anxiety, distress, and clinical global impression measures in an eight-participant feasibility trial. These very large effect sizes should be interpreted cautiously given the absence of control groups and the small samples involved. Soldevila-Matías and colleagues, studying a mixed psychiatric sample with comorbid obesity, reported moderate-to-large effects (Cohen's d = 0.47–1.63) on cognition, mood, and body mass index after twelve weeks of physical activity training.

The pragmatic randomized trial by Schmitter and colleagues — the only large controlled trial directly testing adjunctive exercise against guideline-concordant care — reported no significant between-group difference (b = −0.22, 95% CI −0.72 to 0.29), explicitly contradicting the hypothesis that structured exercise adds clinical benefit beyond comprehensive usual care. Ribeiro and colleagues found a large group-by-time interaction effect (partial η² = 0.489) favoring sprint interval training over a control condition in a small sample of seventeen women, though the intervention duration was only two weeks. Ye and colleagues, testing Baduanjin exercise in a pilot trial of thirty-four participants, found no significant group-by-time interaction for depression severity, with both groups improving over time. Meyer and colleagues (2026) reported directional but mostly non-significant effects for exercise-primed cognitive behavioral therapy, with a modest overall benefit on depression (d = 0.20) but a large difference in remission rates (69% versus 33%, p < 0.05).

The systematic review of rock climbing interventions by Larsson and colleagues described a reduction of 8.3 points on the MADRS — exceeding the minimal clinically important difference of 5 points — characterized by the review authors as "clinically meaningful" relative to no intervention; the underlying studies were judged ROBINS-I High risk of bias in §2.5, and the finding contributes to this scoping review's overall Very Low certainty rating (§3.3). Tous-Espelosín and colleagues, in a single-arm pilot of resistant depression, reported statistically significant mean reductions on the MADRS (−8.31 points, 95% CI −15.1 to −1.5) and functional disability measures, without a comparator group. The single case report by Cavalcanti and colleagues described "clinically significant" and "remarkable" improvements across depression, anxiety, sleep, and quality of life, though no inferential statistics were applicable.

Several studies framed their findings in terms of biological or mechanistic markers rather than clinical effect magnitudes. Carvalho Silva and colleagues concluded that exercise had "mixed effects" on inflammatory markers, with neurotrophic factors showing "promise" but remaining "inconclusive." Kelly and colleagues found that exercise-induced changes in serum IGF-1 were large in absolute terms (d = 0.87 for the overall exercise response) but did not differ between participants with MDD and healthy controls (d = 0.02). Kong and colleagues, studying healthy exercisers rather than individuals with MDD, reported higher cognitive scores and lower depression scores in the exercise group, attributing variance to BDNF and GDNF concentrations.

Overall, the language describing effect strength was inconsistent across the evidence base. Controlled trials with active or comprehensive comparators tended to report null or small effects, while uncontrolled pre-post studies and very small trials reported large to very large effect sizes — a pattern consistent with well-recognized inflation of effect estimates in the absence of blinding and controlled comparison.

Regarding the types of statistical significance metrics reported: ten studies provided explicit p-values (with or without accompanying confidence intervals), five studies reported confidence intervals (some overlapping with those also reporting p-values), and two studies described author-stated significance without a specific numeric metric. Eleven studies — comprising the six protocol-only publications, the discrete choice experiment, the cross-sectional IGF-1 analysis, the case report, and two systematic reviews that did not report pooled inferential statistics in the extracted fields — did not report a conventional significance metric in the extraction data. This distribution reflects the preponderance of protocol publications and non-interventional designs among the included records, which limited the volume of inferential evidence available for synthesis.

6. Study Design Profile

The 25 included articles encompass a broad range of study designs, and this heterogeneity has direct implications for the strength of inferences that can be drawn about whether supervised aerobic exercise improves depressive symptoms and functional outcomes in adults with major depressive disorder receiving SSRI treatment.

Five articles are published protocols for randomized controlled trials that have not yet reported outcome data. These protocol-only publications — describing planned investigations of combined exercise for resistant major depression, exergame-augmented acceptance and commitment therapy, resistance exercise training, remotely delivered physical activity for treatment-resistant depression, and a physiotherapy intervention for inpatients — define methodological frameworks and target populations but contribute no empirical evidence on efficacy. Their presence in the evidence set underscores active research interest in exercise as an adjunct to pharmacotherapy for depression, yet they cannot inform the direction or magnitude of any treatment effect.

Among the completed primary studies, designs vary considerably. One pragmatic multicentre randomized controlled trial directly compared adjunct exercise therapy with guideline-concordant care as usual in 112 outpatients with major depressive disorder and found no additional benefit of exercise on depressive symptom reduction. A second small randomized trial examined sprint interval training in 17 women with moderate-to-severe depression and reported significant within-group symptom improvement in the exercise arm, though the very brief two-week intervention period and small sample limit generalizability. A pilot randomized trial of Baduanjin exercise in 34 adults with major depressive disorder found no significant group-by-time interaction for depression severity, with both groups improving over time. Another randomized trial of 40 participants explored whether moderate-intensity aerobic exercise performed immediately before cognitive behavioral therapy sessions could enhance therapeutic engagement and remission; directional benefits were observed, but the trial was explicitly designed as a proof-of-concept rather than a definitive efficacy test.

Several studies employed uncontrolled pre–post designs. A 14-week supervised group exercise program in 14 adults with major depressive disorder on antidepressant medication reported a large pre–post effect size for depressive symptom reduction, but the absence of a comparator group and the concurrent use of pharmacotherapy preclude attribution of improvement to exercise alone. Similarly, a single-arm pilot of combined aerobic and resistance training in 18 adults with resistant major depression found significant pre–post reductions in depressive symptoms and improvements in functionality, though the investigators acknowledged the same interpretive limitation. A feasibility trial of mindfulness-based boxing therapy in eight outpatients with major depressive disorder or generalized anxiety disorder reported large pre–post effect sizes for depression and anxiety, but the very small sample and lack of control condition constrain interpretation. A single case report described clinically meaningful improvements in depressive symptoms, anxiety, sleep quality, and quality of life following eight weeks of combined aerobic and resistance exercise in a patient taking fluoxetine, though findings from a single individual cannot be generalized.

Two studies examined biological or physiological correlates of exercise in populations that included individuals with major depressive disorder. A cross-sectional analysis of 147 participants found that exercise-induced changes in serum IGF-1 were similar regardless of depression severity or antidepressant use, while a prospective cohort study of 29 individuals with psychiatric disorders (including six with major depressive disorder) and comorbid obesity reported improvements in cognition, mood, and inflammatory biomarkers after 12 weeks of physical activity training. Neither study was designed to isolate the effect of structured aerobic exercise on depressive outcomes in the population specified by the review question. A cross-sectional comparison of healthy fitness club members and community residents explicitly excluded individuals with major depressive disorder, limiting its relevance to the target population.

At the level of evidence synthesis, two systematic reviews with meta-analyses and one umbrella review of meta-analyses were included. The umbrella review reported large effect sizes for adjunctive physical activity in reducing depressive symptoms in adults with depressive disorders and moderate effect sizes for quality-of-life improvements, drawing on 256 randomized trials across 13 high-quality meta-analyses. One systematic review and meta-analysis found that aerobic exercise significantly improved overall cognitive function in adults with major depressive disorder, with a small pooled effect. A second meta-analytic review reported a small but significant effect of exercise on cognitive function in this population, with aerobic modalities among the most beneficial. A narrative systematic review of biological markers found mixed and inconsistent effects of exercise on inflammatory, neurotrophic, and stress-response biomarkers in adults with major depressive disorder. A systematic review of indoor rock climbing for depression reported clinically meaningful symptom reductions compared with no intervention but could not perform meta-analysis due to clinical heterogeneity. A retrospective analysis of clinical program data in older veterans with serious mental illness, including recurrent major depressive disorder, found that multicomponent exercise improved physical function over 12 months, though the study was not designed to assess depressive symptom outcomes. Finally, a cross-sectional discrete choice experiment characterized exercise preferences among patients with depression but provided no data on clinical outcomes.

Taken together, the evidence set is dominated by small pilot studies, uncontrolled pre–post designs, and protocol-only publications, with only a handful of completed randomized trials — and these vary substantially in intervention type, duration, intensity, and comparator. The single large pragmatic trial that most closely approximated the prespecified PICO found no benefit of adjunct exercise over active treatment as usual. The secondary evidence from meta-analyses and umbrella reviews is broadly supportive of exercise for depressive symptoms, but these syntheses aggregate diverse exercise modalities, populations, and comparators that do not map precisely onto the specific question of supervised aerobic exercise added to stable SSRI treatment. The preponderance of uncontrolled and underpowered designs, combined with the absence of blinding in exercise trials, means that the overall certainty of the evidence bearing directly on the review question remains limited.

7. Recency of Evidence

The evidence base informing this review is notably current. All 25 included articles were published between 2024 and 2026, with the majority appearing in 2025 (16 articles) and the remainder split between 2024 (8 articles) and 2026 (1 article). No study predates 2024. This concentration of recent publications reflects a period of active investigation into exercise interventions for major depressive disorder, with several trial protocols signaling ongoing work whose results are not yet available. Because the entire included set falls within a narrow and recent publication window, the findings can be considered representative of the current state of the literature rather than reliant on dated evidence. At the same time, the preponderance of protocols and pilot studies among these recent publications indicates that the field is still maturing, and definitive controlled trials addressing the specific PICO of this review — supervised aerobic exercise as an adjunct to SSRI treatment in adults with moderate major depressive disorder — remain sparse even within this up-to-date evidence base.

8. Evidence Gaps

8a. Gaps Identified from Article Text

The included studies and their authors identified a number of recurring limitations that, taken together, delineate the most pressing evidence gaps for future work on exercise as an adjunct to pharmacotherapy in major depressive disorder.

Small samples and underpowered designs. Nearly every primary study acknowledged limited sample sizes as a constraint on generalizability and statistical power. The single-arm pilot by Elgendy and colleagues (n = 14) and the case report by Cavalcanti and colleagues (n = 1) illustrate the extreme end of this problem, but even the largest completed trial in the set enrolled only 112 participants. Several authors explicitly called for adequately powered, multi-site trials to confirm preliminary signals.

Absence of long-term follow-up. Most interventions lasted eight to fourteen weeks, and few studies assessed outcomes beyond the immediate post-intervention period. Schmitter and colleagues followed participants to fifteen months, and the Gerofit retrospective analysis by Browne and colleagues spanned twelve months, yet both acknowledged high rates of missing data at later time points. The sprint-interval training trial by Ribeiro and colleagues lasted only two weeks, and the Baduanjin pilot by Ye and colleagues ran for ten weeks without any post-intervention follow-up. Multiple authors noted that the durability of exercise-related improvements remains unknown and urged longer observation windows.

Difficulty isolating exercise effects from concurrent treatments. Because most participants were receiving pharmacotherapy, psychotherapy, or both, several authors cautioned that observed improvements could not be attributed to exercise alone. Schmitter and colleagues noted that varying psychological and pharmacological interventions in the care-as-usual arm may have obscured any exercise-specific effect. Elgendy and colleagues acknowledged that concurrent antidepressant use made it impossible to disentangle the contribution of exercise from that of medication. The mindfulness-based boxing feasibility trial by Bozdarov and colleagues raised the same concern, as all participants were on psychiatric medication.

Blinding challenges inherent to exercise trials. The systematic review of rock climbing by Larsson and colleagues observed that none of its included studies blinded patients or therapists, and the meta-analysis by Ge and colleagues flagged experimental biases arising from participants' and staff members' ability to infer study objectives. These limitations are structural to exercise research, yet they remain a gap in methodological rigor that weakens causal inference.

Inconsistent biological-mechanism evidence. The systematic review by Carvalho Silva and colleagues found mixed and variable effects of exercise on inflammatory markers, neurotrophic factors, and stress-response biomarkers in adults with major depressive disorder. The authors highlighted that findings for brain-derived neurotrophic factor showed promise but remained inconclusive, and that cortisol and monoaminergic findings were limited. Kong and colleagues similarly noted that their cross-sectional design precluded causal conclusions about the relationship between aerobic exercise and neurotrophic factor concentrations.

Narrow demographic representation. Browne and colleagues reported that few female and Hispanic or Latinx veterans were included in the Gerofit analysis, limiting generalizability. Tous-Espelosin and colleagues studied an entirely non-Hispanic white sample. The discrete choice experiment by Li and colleagues noted a small subsample of participants aged sixty and above and no investigation of medication use or depression severity. Several protocol papers acknowledged that their eligibility criteria — requiring stable housing, transportation, or absence of comorbid conditions — may have excluded the patients most in need of accessible interventions.

Need for active comparator arms. Larsson and colleagues called for trials comparing rock climbing with established treatments such as antidepressants and aerobic exercise. Bozdarov and colleagues and Elgendy and colleagues both lacked any control group, and the Baduanjin pilot by Ye and colleagues used a treatment-as-usual comparator in which both groups improved, making it difficult to detect an exercise-specific signal. Future trials with well-defined active comparators were recommended by multiple author groups.

8b. Gaps Identified by Independent Analysis

Comparison of the prespecified PICO against the characteristics of the retrieved evidence set reveals several areas of incomplete coverage.

Population specificity. The PICO specifies adults with moderate major depressive disorder confirmed by DSM-5 criteria who are receiving stable first-line SSRI treatment for at least four weeks. Within the retrieved evidence set, no study restricted enrollment exclusively to patients on SSRI monotherapy; participants across studies were taking a range of antidepressants (including SNRIs, bupropion, and anxiolytics), and several studies enrolled patients with treatment-resistant depression or mixed psychiatric diagnoses. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Intervention fidelity to the specified protocol. The PICO defines a supervised aerobic exercise program performed three times per week for twelve weeks at moderate intensity (60–75% of maximal heart rate), with sessions lasting approximately forty minutes. Within the retrieved evidence set, no completed study matched all of these parameters simultaneously. Interventions ranged from six-minute sprint intervals over two weeks to multicomponent programs combining aerobic and resistance training, mind-body exercises such as Baduanjin and bouldering, and exergame-based approaches. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Comparator alignment. The PICO comparator is treatment as usual defined as ongoing pharmacotherapy and routine psychiatric follow-up without any structured exercise intervention. While several studies used a treatment-as-usual arm, the content of that arm varied considerably — ranging from guideline-concordant specialized mental health care with psychotherapy to routine outpatient follow-up — and in some cases the comparator included active psychological interventions. Within the retrieved evidence set, no study defined its comparator in terms that precisely matched the PICO specification of ongoing SSRI pharmacotherapy with routine psychiatric follow-up and no structured exercise. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Occupational and social functioning as a primary outcome. The PICO lists occupational and social functioning among the outcomes of interest. Within the retrieved evidence set, only Tous-Espelosin and colleagues reported on the Sheehan Disability Scale, and Schmitter and colleagues assessed functional outcomes secondarily. No study foregrounded occupational or social functioning as a primary endpoint in the context of adjunctive aerobic exercise for adults on SSRI treatment. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Cognitive functioning assessed alongside aerobic exercise in SSRI-treated populations. Two meta-analytic syntheses (Ge and colleagues; Ren and colleagues) reported that exercise may improve cognitive function in adults with major depressive disorder, but neither restricted its analysis to patients receiving SSRI treatment. Within the retrieved evidence set, no primary trial examined cognitive outcomes specifically in SSRI-treated adults undergoing a structured aerobic exercise program. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Geographic and healthcare-system diversity. The completed primary studies were conducted in the Netherlands, Canada, Brazil, Spain, China, Germany, and the United States. Within the retrieved evidence set, no studies were conducted in low- or middle-income countries outside of China or Brazil, and no studies examined implementation within primary care settings — the context in which most SSRI prescribing occurs. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Follow-up beyond six months in controlled designs. The PICO does not specify a follow-up duration, but the clinical question implies interest in sustained benefit. Within the retrieved evidence set, no randomized controlled trial with a treatment-as-usual comparator reported controlled outcome data beyond fifteen months, and the single trial that did (Schmitter and colleagues) was hampered by substantial missing data attributable to the COVID-19 pandemic. Whether this represents a gap in the broader literature or a limitation of this search strategy is beyond the scope of this review.

Effect Estimates by Study

Study	Measure	Value	95% CI	p-value
Schmitter M, 2025	b (regression coefficient, time-by-condition interaction)	-0.22	-0.72–0.29	0.396
Iriarte-Yoller N, 2024 (PMID 11653556)	NR	NR	NR	NR
Vancampfort D, 2025 (PMID 12079350)	NR	NR	NR	NR
Ge J, 2025	SMD	0.329	0.25–0.41	NR
Elgendy H, 2024	Cohen's d	2.47	NR	0.001
Zhang B, 2024 (PMID 11216053)	NR	NR	NR	NR
Ribeiro JA, 2024	partial eta squared (ηp²) for group × time interaction	0.489	NR	0.002
Tous-Espelosin M, 2025	mean difference	-8.31	-15.1–-1.5	0.021
Ye J, 2025	η2	0.062	NR	0.495
Larsson R, 2025 (PMID 12459035)	NR	NR	NR	NR
Meyer JD, 2025 (PMID 12379489)	NR	NR	NR	NR
Tassone VK, 2025 (PMID 12404371)	NR	NR	NR	NR
Meyer JD, 2026	Cohen's d	0.36	-0.19–0.9	0.20
Poveda-López JL, 2025 (PMID 12591423)	NR	NR	NR	NR
Kong L, 2024	t-test (between-group difference)	-2.41	NR	0.017
Bozdarov J, 2025	Cohen's d	2.79	NR	0.014
Soldevila-Matías P, 2025	Cohen's d	0.87	NR	< .0001
Kelly S, 2025	Cohen's d	0.02	NR	>0.05
Meyer JD, 2024 (PMID 11460085)	NR	NR	NR	NR
Huang K, 2024 (PMID 11402669)	NR	NR	NR	NR
Carvalho Silva R, 2025 (PMID 12315648)	NR	NR	NR	NR
Ren F, 2024	g	0.21	0.07–0.34	NR
Cavalcanti DdSA, 2025 (PMID 12604159)	NR	NR	NR	NR
Browne J, 2024 (PMID 12362477)	NR	NR	NR	NR
Li H, 2025	coefficient (mixed logit model)	0.84	0.6818403–0.9940637	< 0.01

3.3 Certainty of Evidence

Overall confidence in the available evidence is very low given the heterogeneity of study designs, indirectness relative to the prespecified PICO, imprecision, and risk of bias across the included studies.

4. Discussion

Summary of Evidence

This scoping review mapped the available evidence on whether early initiation of supervised aerobic exercise improves depressive symptoms and functional outcomes in adults with major depressive disorder receiving SSRI treatment. Across 25 included articles, the evidence base was characterized by marked heterogeneity in study design, population, intervention modality, and comparator condition. Three studies reported findings broadly supportive of exercise for depressive symptom reduction, one pragmatic randomized controlled trial found no benefit of adjunct exercise over guideline-concordant care, and the remaining 21 articles were mixed, inconclusive, or contributed no outcome data — seven being protocols without results.

The pattern that emerged is one familiar in the exercise-for-depression literature: uncontrolled and small-sample studies tended to report large pre-post improvements, while the single adequately powered trial comparing exercise against comprehensive active treatment found no incremental benefit. Two meta-analytic syntheses suggested that aerobic exercise may yield small improvements in cognitive function among adults with major depressive disorder, though neither isolated patients on SSRI monotherapy. Evidence on quality of life, anxiety symptoms, and occupational or social functioning was sparse and largely uncontrolled.

Critically, no included study precisely matched every element of the prespecified clinical question — adults with moderate major depressive disorder on stable SSRI treatment, randomized to supervised moderate-intensity aerobic exercise three times weekly for twelve weeks, compared with treatment as usual without structured exercise. The breadth of this scoping review therefore confirms active research interest in exercise as an adjunct to pharmacotherapy but reveals that the specific question posed here remains inadequately tested. Any clinical interpretation should be tempered by the predominance of pilot-level evidence and the absence of definitive controlled data in the target population.

Limitations of this Review

Several methodological features of this review warrant consideration. The search was restricted to PMC Open Access articles, which may have excluded relevant studies published in subscription-only journals or indexed in databases not covered by this strategy; this coverage limitation could introduce a bias toward certain geographic regions, funding sources, or publication types. Data extraction and eligibility screening were performed by a single reviewer, without independent dual screening or consensus arbitration, increasing the risk that individual judgment errors in study selection or data coding went undetected. The synthesis was conducted narratively without statistical pooling; no meta-analysis was performed, precluding formal estimation of pooled effect sizes, tests of statistical heterogeneity, or quantitative sensitivity analyses. Given the heterogeneity of study designs, populations, and interventions across the 25 included articles, a narrative approach was appropriate for a scoping review, but it necessarily limits the precision and reproducibility of the conclusions drawn. Finally, the predominance of protocol-only publications and small uncontrolled studies among the included records constrained the volume of outcome data available for interpretation.

Implications for Future Research

The evidence gaps identified in this review point toward several concrete research priorities. No included study enrolled exclusively adults with moderate major depressive disorder receiving stable first-line SSRI monotherapy — the population most relevant to the clinical question. Adequately powered randomized controlled trials restricting enrollment to this specific pharmacotherapy context are needed to determine whether exercise confers benefit beyond what SSRIs alone provide. The intervention tested should adhere to a clearly defined aerobic protocol — specifying frequency, duration, intensity, and supervision — rather than combining heterogeneous modalities, so that findings can be attributed to a replicable exercise prescription. Comparator arms should be defined with precision; the null finding from the single large pragmatic trial suggests that when the control condition includes robust multimodal care, detecting an exercise-specific signal may require either larger samples or a comparator limited to pharmacotherapy with routine follow-up. Occupational and social functioning, cognitive outcomes, and quality of life were rarely assessed as primary endpoints and deserve dedicated measurement in future trials. Follow-up beyond the immediate post-intervention period is essential to establish whether any benefits persist. Finally, trials conducted in primary care settings and in demographically diverse populations — including low- and middle-income countries — would strengthen the generalizability of findings to the clinical contexts where SSRI prescribing most commonly occurs.

5. Conclusions

This scoping review identified 25 recent publications addressing exercise interventions for adults with major depressive disorder, yet the evidence bearing directly on whether supervised aerobic exercise improves depressive symptoms and functional outcomes when added to stable SSRI treatment remains sparse and inconclusive. Most included studies were protocols without outcome data, single-arm pilots, or investigations of populations and interventions that overlapped only partially with the prespecified clinical question. The few completed trials that reported outcome data yielded conflicting signals: uncontrolled studies tended to report pre-post symptom reductions, while the single adequately powered pragmatic trial comparing adjunct exercise to comprehensive guideline-concordant care found no additional benefit. No included study matched every element of the target population, intervention, comparator, and outcome simultaneously. Adequately powered randomized trials that isolate the effect of moderate-intensity supervised aerobic exercise in SSRI-treated adults with moderate major depressive disorder — with clearly defined comparators and sustained follow-up — are needed before clinical recommendations can be offered with confidence.

Pipeline Limitations

1. Open-access full-text constraint: Full-text retrieval is limited to open-access sources. The system queries up to 11 sources in cascade (PMC, Unpaywall, OpenAlex, Europe PMC, bioRxiv, medRxiv, arXiv, Semantic Scholar, CORE, ClinicalTrials Publications, and a reviewer-provided fallback source when configured), but cannot legally retrieve full text from paywalled subscription journals at scale. Three of the eleven sources (OpenAlex, Semantic Scholar, CORE) are present as placeholders and currently return no results — see Appendix B for any article whose retrieval relied on the reviewer-provided fallback.

2. Single-reviewer relevance screening: Relevance screening was performed without a second independent reviewer. PRISMA 2020 expects dual-independent screening; the screening procedure used here applied a documented relevance threshold instead. Reviewers should review and override the included set as needed.

3. No backward citation chasing: The system does not perform citation chasing (forward or backward) beyond the initial Boolean retrieval. Articles cited by included studies are not auto-retrieved.

4. No grey literature: The system does not search grey literature sources (conference abstracts, dissertations, clinical-trial registry results-only entries, regulatory filings) beyond the structured ClinicalTrials.gov registry index.

5. Retrieval ceilings: Per-database retrieval is capped (PubMed: 10000; OpenAlex: 10000; ClinicalTrials.gov: 500). Queries returning more results than the ceiling will silently truncate to the cap.

6. Boolean query scope: The Boolean query is assembled as (Population) AND (Intervention/Exposure) only, per Cochrane §4.4.4. Comparator and Outcome filtering happens at the relevance-scoring stage, not at the search-query stage.

7. Personal Reference Library Augmentation. Matching was performed on PubMed ID, DOI, and PubMed Central ID; ClinicalTrials.gov NCT identifiers were not matched. The personal reference library contents reflect the state of the reviewer's storage at the time of retrieval; the retrieval pool is pinned per project after the first retrieval (see Methods §2.3) so re-runs of subsequent analysis stages return the same set, but a fresh retrieval after library changes would yield a different supplemental count. Personal reference library articles bypass the relevance screening that database-retrieved records undergo and require explicit reviewer confirmation of inclusion criteria — see Appendix B.

8. Publication year filter: Articles were restricted to those published between 2024 and 2026 (inclusive). Articles for which the system could not extract a publication year were excluded from the analyzed set.

9. Retraction screening: PubMed and PMC metadata were inspected for retraction signals at retrieval time. Articles flagged as retracted are annotated in §3.1 Study Characteristics and surfaced for reviewer confirmation in Appendix B. The system does not autonomously exclude retracted articles — inclusion is a reviewer decision.

Protocol and Registration

This review was not registered in PROSPERO or another registry. A pre-specified protocol is supplied as a sibling artifact alongside this manuscript and may be submitted post-hoc to PROSPERO or OSF for registration.

Funding

This review received no specific funding from any agency in the public, commercial, or not-for-profit sectors.

Conflicts of Interest

The authors declare no competing interests.

Data Availability

All extracted data and the verbatim Boolean query used for retrieval are available in the supplementary materials. Source articles are publicly accessible via PubMed / PubMed Central / open-access providers.

References

1. Schmitter M, Wijnen B, Creemers D, Van Dorp A, Oostelbos P, Tendolkar I, Smits J, Spijker J, Vrijsen J The (cost-)effectiveness of exercise therapy adjunct to guideline-concordant care for depression: a pragmatic randomised controlled trial European Psychiatry. 2025;68(1):1789-8. doi:10.1192/j.eurpsy.2025.10085. PMID: 40814823.
2. Iriarte-Yoller N, Etxaniz-Oses J, Pavón-Navajas C, Tous-Espelosin M, Sánchez-Gómez PM, Maldonado-Martín S, Yoller-Elburgo AB, Elizagarate-Zabala E Treatment with combined exercise in patients with resistant major depression (TRACE-RMD): study protocol for a randomised controlled trial Trials. 2024;25(12):8323-8335. doi:10.1186/s13063-024-08685-7. PMID: 39696604.
3. Vancampfort D, Firth J, Stubbs B, Schuch F, Rosenbaum S, Hallgren M, Deenik J, Ward PB, Mugisha J, Van Damme T, Werneck AO The efficacy, mechanisms and implementation of physical activity as an adjunctive treatment in mental disorders: a meta‐review of outcomes, neurobiology and key determinants World Psychiatry. 2025;24(2):227-239. doi:10.1002/wps.21314. PMID: 40371806.
4. Ge J, Wang Q, Wang S, Xiong X, Zhai Y Optimal dose and type of exercise to improve cognitive function in adults with major depressive disorder: a systematic review and Bayesian model-based network meta-analysis Frontiers in Public Health. 2025;13:55-68. doi:10.3389/fpubh.2025.1662778. PMID: 41487645.
5. Elgendy H, Shalaby R, Agyapong B, Lesage D, Paulsen L, Delday A, Duiker S, Surood S, Wei Y, Nkire N, Agyapong VIO Effectiveness of Group Physical Exercise in Treating Major Depressive Disorder: An Analysis of Secondary Data from an Aborted Randomized Trial Behavioral Sciences. 2024;14(3):e14309-111. doi:10.3390/bs14030219. PMID: 38540522.
6. Zhang B, Deng H, Ren J, Legrand FD, Ahmad Yusof H, Zhang R, Leong Bin Abdullah MFI Study protocol on the efficacy of exergames-acceptance and commitment therapy program for the treatment of major depressive disorder: comparison with acceptance and commitment therapy alone and treatment-as-usual in a multicentre randomised controlled trial BMJ Open. 2024;14(6):979-86. doi:10.1136/bmjopen-2023-080315. PMID: 38926142.
7. Ribeiro JA, Schuch FB, Tonello L, Meneghel Vargas KF, Oliveira-Junior SA, Müller PT, Boullosa D Effectiveness of short sprint interval training in women with major depressive disorder: a proof-of-concept study Frontiers in Psychiatry. 2024;15(16):66-68 p. doi:10.3389/fpsyt.2024.1356559. PMID: 38686128.
8. Tous-Espelosin M, Pavón-Navajas C, Etxaniz-Oses J, Cañas-García MT, Maldonado-Martin S, Orruño-Vivar J, Corres P, Iriarte-Yoller N Effects of combined exercise training for adults with resistant major depression: a pilot study from the TRACE-RMD project PeerJ. 2025;13(20):771-776. doi:10.7717/peerj.20356. PMID: 41282982.
9. Ye J, Fang J, Yang S, Xiang Q, Ahorsu DK Effects of Baduanjin Exercise on Depression Severity and Heart Rate Variability in Adults with Major Depressive Disorder: A Pilot Randomized Controlled Trial Psychology Research and Behavior Management. 2025;18(10371):505-514. doi:10.2147/PRBM.S499936. PMID: 40060106.
10. Larsson R, Larsson A, Nordeman L Effectiveness of indoor rock climbing and bouldering as treatment for depression – a systematic review BMC Psychiatry. 2025;25(10159):1789-858. doi:10.1186/s12888-025-07292-3. PMID: 40988014.
11. Meyer JD, Smith SL, Gidley JM, Molina A, Churchill SL, Kelly SJE, Lansing JE, Wade NG, Phillips AL, Liu P, Herring MP, Murray TA, Barnes JN, Goldberg SB, Lefferts WK Protocol for a randomized controlled trial of a resistance exercise training to treat major depression via cerebrovascular mechanisms (RESIST Trial) Trials. 2025;26(10):915-922. doi:10.1186/s13063-025-09025-z. PMID: 40859378.
12. Tassone VK, Quesnel DA, Parkington K, Rueda A, Martin J, Lee GH, Teixeira A, deJonge ML, Lou W, Wiljer D, Mulsant BH, Sabiston CM, Bhat V Improving adherence to physical activity in treatment-resistant depression: Protocol for a pilot randomized controlled trial of a remotely delivered program PLOS One. 2025;20(9):253-61. doi:10.1371/journal.pone.0330848. PMID: 40892783.
13. Meyer JD, Kelly SJE, Gidley JM, Lansing JE, Smith SL, Churchill SL, Connolly ML, Thomas EBK, Goldberg SB, Abercrombie HC, Murray TA, Wade NG Exercise priming to enhance therapeutic bond and behavioral activation in CBT for MDD: a randomized controlled target-engagement trial with remission signal Journal of affective disorders. 2026;405:121606-121606. doi:10.1016/j.jad.2026.121606. PMID: 41833616.
14. Poveda-López JL, Jiménez-Sánchez C, Lafuente-Ureta R, Marco-Gómez B, Villagrasa-Cantín A, Pérez-Mansilla S, Guarch-Rubio M, Roy JF Study protocol for a mixed-methods pilot of a physiotherapy plus education program for inpatients with major depressive disorder: Feasibility and preliminary effects PLOS One. 2025;20(11):171-8. doi:10.1371/journal.pone.0326012. PMID: 41196855.
15. Kong L, Miu L, Yao W, Shi Z Effect of Regular Aerobic Exercise on Cognitive Function, Depression Level and Regulative Role of Neurotrophic Factor: A Prospective Cohort Study in the Young and the Middle-Aged Sample Risk Management and Healthcare Policy. 2024;17(1):935-943. doi:10.2147/RMHP.S456765. PMID: 38633669.
16. Bozdarov J, Jones BD, Umer M, Blumberger DM, Husain IM Mindfulness-based (non-contact) boxing therapy (MBBT) for depression and anxiety: A feasibility study PLOS ONE. 2025;20(2):1700-12. doi:10.1371/journal.pone.0318364. PMID: 39913542.
17. Soldevila-Matías P, Sánchez-Ortí JV, Correa-Ghisays P, Balanzá-Martínez V, Selva-Vera G, Sanchis-Sanchis R, Iglesias-García N, Monfort-Pañego M, Tomás-Martínez P, Victor VM, Crespo-Facorro B, Valenzuela CSM, Climent-Sánchez JA, Corral-Márquez R, Fuentes-Durá I, Tabarés-Seisdedos R Clinical outcomes and anti-inflammatory mechanisms predict maximum heart rate improvement after physical activity training in individuals with psychiatric disorders and comorbid obesity PLOS ONE. 2025;20(1):16-21. doi:10.1371/journal.pone.0313759. PMID: 39752432.
18. Kelly S, Meyer J, Stielow C, Heinzel S, Heissel A Effects of an Acute Maximal Exercise Bout on Serum Insulin-Like Growth Factor-1 in Adults with MDD Psychoneuroendocrinology. 2025;171(16):107215-107215. doi:10.1016/j.psyneuen.2024.107215. PMID: 39413529.
19. Meyer JD, Kelly SJE, Gidley JM, Lansing JE, Smith SL, Churchill SL, Thomas EBK, Goldberg SB, Abercrombie HC, Murray TA, Wade NG Protocol for a randomized controlled trial: exercise-priming of CBT for depression (the CBT+ trial) Trials. 2024;25(1):92-116. doi:10.1186/s13063-024-08495-x. PMID: 39375728.
20. Huang K, Li M, You S, Jiao Y, He R, Han B, Pu Y, Jia Y, Chen L The effectiveness of an exergame intervention for college students with subthreshold depression: protocol for a mixed methods study Frontiers in Public Health. 2024;12:1401-14. doi:10.3389/fpubh.2024.1390806. PMID: 39286747.
21. Carvalho Silva R, Meattini M, Perusi G, Carletto S, Bortolomasi M, Gennarelli M, Baune BT, Minelli A Disentangling the biological mechanisms underlying the effects of physical exercise in major depressive disorder: a comprehensive systematic review of randomized controlled trials Psychological Medicine. 2025;55(23):3812-3847. doi:10.1017/S0033291725100743. PMID: 40665827.
22. Ren F, Hillman CH, Wang W, Li R, Zhou W, Liang W, Yang Y, Chen F, Chang Y Effects of aerobic exercise on cognitive function in adults with major depressive disorder: A systematic review and meta-analysis International Journal of Clinical and Health Psychology : IJCHP. 2024;24(2). doi:10.1016/j.ijchp.2024.100447. PMID: 38371396.
23. Cavalcanti DdSA, Faro H, Amiri E, Machado DGdS Effect of combined aerobic and resistance training with self-selected intensity on clinical parameters and fitness in a patient with mixed anxiety and depressive disorder: a case report Journal of Medical Case Reports. 2025;19:67-86. doi:10.1186/s13256-025-05622-3. PMID: 41214828.
24. Browne J, Elbogen EB, Mueser KT, Rudolph JL, Wu W, Philip NS, Mills WL, Young AS, Sloane R, Hall KS Trajectories of physical function changes for older veterans with serious mental illness in a clinical exercise program Mental health and physical activity. 2024;26(6):848-854. doi:10.1016/j.mhpa.2024.100580. PMID: 40837382.
25. Li H, Xie P, Hou Y, Jiang W, Gong S, Jin S Preferences of patients with depression for exercises: a discrete choice experiment BMC Public Health. 2025;25(4):69-76. doi:10.1186/s12889-025-22404-0. PMID: 40158084.

Appendix A — PRISMA 2020 Checklist

Appendix B — Reviewer Worklist

Suggested citation

Krasnova M. Evidence Review: Does supervised aerobic exercise added to SSRI treatment improve depression in adults? margaritakrasnovamd.com. Last reviewed 2026-05-23. Available at: https://margaritakrasnovamd.com/evidence-review/aerobic-exercise-ssri-depression.html