The Effect of Walking on Depressive and Anxiety Symptoms: Systematic Review and Meta-Analysis

Depression and anxiety are common mental disorders with substantial global burden. In 2023, global prevalence estimates were 5% for depression and 4% for anxiety, with lifetime experiences of depression and anxiety affecting 18.4% and 33.7% of people, respectively. These conditions account for large proportions of disability-adjusted life-years and include many individuals with subthreshold symptoms, particularly exacerbated during the COVID-19 pandemic. Available interventions (antidepressants, psychotherapy, physical exercise) are effective; among them, walking is safe, convenient, low-cost, and requires minimal equipment or training. It can be feasibly implemented for people with subthreshold symptoms or those who prefer nonpharmacologic options. While prior research has demonstrated health benefits of walking for cardiorespiratory health, blood pressure, lipids, chronic pain, weight, diabetes risk, and mortality, evidence on mental health outcomes—especially across different forms of walking (indoor vs outdoor, group vs individual, facilitated vs nonfacilitated, intensity, duration, frequency, pace, instructions)—has been limited. Previous meta-analyses included few RCTs, focused on specific populations or forms, and did not distinguish types of control groups. This systematic review and meta-analysis aims to comprehensively assess the effects of walking and its different forms on depressive and anxiety symptoms compared with inactive and active controls, providing detailed subgroup analyses to inform clinical practice and public health policies on walking promotion.

Prior systematic reviews indicated walking benefits for physical health (cardiorespiratory health, blood pressure, cholesterol, chronic pain, weight loss, diabetes risk, and mortality). For mental health, a meta-analysis of 8 RCTs (n=341) in depressed adults reported a significant, large effect of walking on depressive symptoms (SMD −0.86). Another meta-analysis (2015; n=110) found outdoor group walking benefited depressive symptoms (SMD −0.67). Other reviews targeted specific groups or forms (postpartum depression, nature walking). These earlier works were limited by small numbers of RCTs and participants, narrow populations or walking formats, and did not differentiate control types or detail walking characteristics (duration, intensity, frequency, pace, instructions). The current review addresses these gaps by including more recent RCTs, categorizing control conditions (inactive, active, other), and evaluating multiple walking formats and parameters via subgroup and meta-regression analyses. Comparison with prior work also notes consistency of walking’s benefits relative to inactive controls and comparability to other physical activity interventions, while highlighting the need to explore dose-response and mild-intensity walking effects.

Registration and reporting: Registered with PROSPERO (CRD42021247983) and conducted per PRISMA guidelines. Literature search: Systematic searches of English-language articles from inception to April 5, 2022 in MEDLINE (PubMed), CENTRAL, Embase, PsycINFO, AMED, CINAHL, and Web of Science, using terms related to walking/pedometers/step count, depression/anxiety, and RCTs. Reference lists of relevant reviews and backward/forward citation tracking were also performed. Screening and eligibility: Two authors independently screened titles/abstracts (duplicates removed via Covidence) and full texts; discrepancies resolved by a third author. Inclusion criteria: RCTs; any walking intervention; nonwalking controls; outcomes measured by depression/anxiety symptom scales at baseline and postintervention; English; full text available. Exclusions: mixed interventions (walking combined with other physical activities, supplements, or psychosocial interventions), single-bout or 1-day walking; control groups containing walking; outcomes measuring overall psychological well-being instead of specific depressive/anxiety symptoms; non-RCT designs; conference papers/letters; non-English; no full text. Data extraction: Two authors independently extracted study metadata (title, author, year, country), participant characteristics (age, gender, health condition), intervention details (duration; walking intensity, frequency, duration per session; instructions/training; location; format; motivation/pedometer use), and outcomes (depression/anxiety scales). Risk of bias: Cochrane Collaboration tool (excluding blinding of participants/personnel due to intervention nature). Items rated low/unclear/high; overall risk low if all items low, high if any item high, unclear if at least one item unclear and none high. Analysis: Standardized mean differences (SMDs) calculated from mean changes with SDs; missing SDs estimated from SEs/95% CIs or median baseline-post correlations per Cochrane Handbook. Controls categorized: inactive (no intervention, waitlist, usual care not treating depression/anxiety, brief physical activity instructions, placebo meetings), active (evidence-based interventions for depression/anxiety: moderate-intensity exercise, yoga, tai chi, meditation, CBT, stress management, art therapy, social interaction), other (light exercise, nutrition-related, regular education). Primary analysis compared walking vs inactive controls; secondary compared walking vs active or other controls. Meta-analysis used random-effects (DerSimonian-Laird), pooling SMDs with 95% CIs; heterogeneity assessed via I² and τ². Small-study effects assessed by funnel plots and Egger test. Subgroup analyses evaluated walking characteristics (frequency, duration per walk, intensity, pace, format, location), instructions/training, motivation (behavior change techniques), pedometer, dropout rate, sample size, mean age, baseline depression status, and risk of bias; significant subgroups (P<.05) entered into multivariate meta-regression. Software: Stata 16. Ethical considerations: Not applicable for a review; conducted per Declaration of Helsinki.

Study selection and characteristics: 75 RCTs included (n=8636 participants). Mean ages mostly 30.4–84.7 years; one study included adolescents (mean age 16.8) and was not meta-analyzed. Intervention durations ranged 10 days to 18 months; 59% lasted 3–6 months. Ten studies enrolled participants depressed at baseline (5 diagnosed; 5 screen-positive); none enrolled patients diagnosed with anxiety. Outcomes: 68 studies reported depressive symptoms; 39 reported anxiety; 32 reported both. Risk of bias: Low for random sequence generation in 52% of studies, allocation concealment 24%, blinding of outcome assessment 32%, incomplete outcome data 80%, selective reporting 99%, other bias 100%. Overall: 8% low risk, 69% unclear, 23% high. Primary meta-analyses (adults vs inactive controls): • Depression (n=44 RCTs): SMD −0.591 (95% CI −0.778 to −0.403); I²=84.8%; τ²=0.3008; P<.001. Significant reductions across most subgroups (frequency, duration per walk, location, individual format), except self-selected pace and group walking. Subgroup differences: greater effects in depressed vs nondepressed (SMD −1.863 vs −0.442; P=.002); lower dropout (0–10%) vs >10% (SMD −0.779 vs −0.399; P=.04); mean age >60 vs 30–60 years (SMD −0.885 vs −0.417; P=.04); interventions without motivation vs with motivation (SMD −0.763 vs −0.292; P=.006); no pedometer vs pedometer (SMD −0.766 vs −0.292; P=.01). Multivariate meta-regression: baseline depression associated with greater decline in depressive symptoms (SMD −1.158, 95% CI −1.851 to −0.466; P=.001; R²=37.3%). • Anxiety (n=26 RCTs): SMD −0.446 (95% CI −0.628 to −0.265); I²=81.1%; τ²=0.1530; P<.001. Significant reductions in most subgroups except self-selected pace and motivation. Subgroup differences: duration <3 months vs 3–6 months (SMD −0.837 vs −0.312; P=.03); following instructions vs not (SMD −0.691 vs −0.273; P=.03); without motivation vs with motivation (SMD −0.658 vs −0.153; P=.002). Meta-regression: no factor significantly associated (R²=17%; all P> .05). Small-study effects: Egger tests indicated asymmetry (depression t=−2.02, P=.005; anxiety t=−2.12, P=.01). Secondary analyses: • Walking vs active controls: No significant differences for depression (n=17 RCTs; SMD −0.126, 95% CI −0.343 to 0.092; I²=58%; τ²=0.1058; P=.26) or anxiety (n=14 RCTs; SMD −0.053, 95% CI −0.311 to 0.206; I²=67.7%; τ²=0.1421; P=.69). Egger tests showed no small-study effects (depression P=.28; anxiety P=.57). • Walking vs other controls: For depressive symptoms, walking was superior to light exercise (n=6; SMD −0.426, 95% CI −0.766 to −0.086; I²=55.9%; τ²=0.1148; P=.046), but not different from nutrition (n=3; SMD −0.091, 95% CI −0.541 to 0.358; I²=50.2%; τ²=0.0795; P=.69) or regular education (n=6; SMD −0.208, 95% CI −0.476 to 0.060; I²=59.6%; τ²=0.0622; P=.13). For anxiety, no significant difference vs light exercise (n=4; SMD −0.493, 95% CI −1.056 to 0.070; I²=65.1%; τ²=0.1611; P=.09). Overall, walking consistently reduced depressive and anxiety symptoms compared with inactive controls, with effects generally robust across formats and characteristics, and comparable to active evidence-based interventions.

The findings demonstrate that walking reduces depressive and anxiety symptoms in adults compared with inactive controls, addressing the research question of whether different forms of walking are effective mental health interventions. Benefits were observed across various walking characteristics (frequency, duration per walk, indoor/outdoor, individual/group) and participant profiles (depressed and nondepressed at baseline), with larger effects among those depressed at baseline—highlighted by meta-regression identifying baseline depression as a key moderator. The lack of differences versus active controls indicates walking is comparable to other evidence-based interventions (moderate-intensity exercises, yoga, CBT, etc.), supporting its use as an alternative or complementary option. Subgroup insights suggest practical optimization: shorter interventions (<3 months) and following instructions (eg, supervised sessions, structured components like warm-up/cooldown) were particularly beneficial for anxiety; guided or set pace and at least moderate intensity were associated with effectiveness, whereas self-selected pace showed weaker effects. Interventions without added motivation techniques or pedometers showed larger effects in this dataset, potentially reflecting confounding by supervision, intensity, or adherence differences; lower dropout rates coincided with larger effects, underscoring the importance of adherence. Indoor and outdoor walking were both effective, relevant for homebound individuals or constrained environments. Policy and practice implications include integrating moderate-intensity, guided-pace walking into clinical care and public health initiatives, supervising regular sessions when feasible to enhance adherence, and prioritizing depressed individuals where resources are limited.

Walking is an effective, accessible intervention to reduce depressive and anxiety symptoms compared with inactive controls, with effects comparable to active evidence-based interventions. Multiple formats (indoor or outdoor, individual or group) can be employed, with guided or set pace and at least moderate intensity recommended. Integrating walking into public health initiatives, clinical prevention and treatment plans, and workplace wellness programs can improve mental well-being. Future research should examine dose-response relationships (including mild-intensity walking), long-term maintenance effects beyond postintervention, adherence optimization strategies, trials in diagnosed depression/anxiety populations, and studies involving children and adolescents.

Primary analyses exhibited high between-study heterogeneity despite subgroup exploration, warranting cautious interpretation and generalization. Funnel plot asymmetry and Egger tests indicated small-study effects (potentially due to clinical heterogeneity, outcome reporting bias, publication bias, or chance). Most included studies had unclear (69%) or high (23%) risk of bias, often due to limited reporting on allocation concealment and outcome assessment blinding; however, subgroup analyses suggested no significant impact when excluding high risk studies. Effects were measured only postintervention, with no long-term follow-up maintenance assessed. There is a paucity of trials on mild-intensity walking, limiting dose-response insights. Most participants were not clinically diagnosed with depression or anxiety disorders, and few trials involved children or adolescents, constraining generalizability to these populations.