Walking on sunshine: scoping review of the evidence for walking and mental health

This is a scoping review, not a single experiment. The authors systematically searched for all published studies (including systematic reviews and individual papers) that examined the relationship between walking and any of eight pre-defined mental health outcomes:

Depression (most evidence)

Anxiety

Psychological stress

Psychological well-being

Subjective well-being

Social isolation and loneliness

Resilience (no studies found)

General mental health (broad measures)

The intervention was any form of walking—outdoor, indoor (treadmill), group, solo, in nature, or urban. Comparators varied across studies: no walking, other forms of exercise, sedentary activities, or different walking conditions (e.g., indoor vs. outdoor, group vs. solo). Outcome measures included validated scales for depression (e.g., Beck Depression Inventory, Hamilton Depression Rating Scale), anxiety (e.g., State-Trait Anxiety Inventory), stress (e.g., Perceived Stress Scale), and well-being (e.g., WHO-5 Well-Being Index, Satisfaction with Life Scale).

The review included 5 systematic reviews and 50 individual papers. The total sample across all studies is not reported as a single number, but the individual studies ranged from small pilot trials (e.g., 12–20 participants) to large epidemiological cohorts (e.g., >10,000 participants). Populations included:

Adults with clinical depression (e.g., mild-to-moderate major depressive disorder)

Healthy adults (general population, office workers, university students)

Older adults (age 60+)

Pregnant women

People with chronic illness (e.g., fibromyalgia, coronary heart disease)

Sedentary individuals

Settings included community walking groups, university campuses, parks, urban streets, and laboratory treadmills. Most studies were conducted in high-income countries (UK, USA, Australia, Japan, Scandinavia). No studies specifically targeted children, adolescents, or clinical populations with severe mental illness (e.g., bipolar disorder, schizophrenia).

The review did not collect new data but extracted outcomes from existing studies. The key instruments used across the included studies were:

**Depression:** Beck Depression Inventory (BDI, 0–63, higher = worse), Hamilton Depression Rating Scale (HAM-D, 0–52, higher = worse), Center for Epidemiologic Studies Depression Scale (CES-D, 0–60, higher = worse), Patient Health Questionnaire (PHQ-9, 0–27, higher = worse)

**Anxiety:** State-Trait Anxiety Inventory (STAI, 20–80, higher = worse), Hospital Anxiety and Depression Scale (HADS, 0–21 for anxiety subscale)

**Stress:** Perceived Stress Scale (PSS, 0–40, higher = worse), salivary cortisol (biomarker)

**Psychological well-being:** WHO-5 Well-Being Index (0–100, higher = better), Warwick-Edinburgh Mental Well-being Scale (WEMWBS, 14–70, higher = better)

**Subjective well-being:** Satisfaction with Life Scale (SWLS, 5–35, higher = better), Positive and Negative Affect Schedule (PANAS)

**Social isolation/loneliness:** UCLA Loneliness Scale (20–80, higher = worse), single-item questions

**Design:** Scoping review (systematic mapping of the literature, not a meta-analysis). The authors followed the PRISMA-ScR guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews).

**Search strategy:** They searched three databases (Ovid Medline, ProQuest, Web of Science) from inception to 2017. They identified 13,014 records, screened titles/abstracts, then full texts. Final inclusion: 5 systematic reviews and 50 individual papers.

**Inclusion criteria:** Any study design (RCT, quasi-experimental, observational, qualitative) that examined walking (any dose, any setting) and at least one of the eight mental health outcomes. No language restrictions. No date restrictions.

**Data extraction:** Two reviewers independently extracted data on study design, population, walking intervention (frequency, intensity, time, type), comparator, outcome measures, and key findings. Disagreements resolved by consensus.

**Synthesis:** Narrative synthesis (not statistical pooling). Results were grouped by mental health outcome. For depression, they reported findings from existing systematic reviews. For other outcomes, they summarised individual studies.

**What this design can and cannot prove:**

**Can prove:** The breadth of existing evidence, gaps in the literature, and patterns across studies. Scoping reviews are excellent for mapping what is known and unknown.

**Cannot prove:** Causal effects of walking on mental health. Because they included observational studies and did not pool effect sizes, they cannot provide a single "effect size" for walking. They cannot compare walking to other interventions (e.g., medication, therapy). They cannot determine optimal dose (frequency, intensity, duration) because studies used wildly different protocols.

**Cannot prove:** Whether walking is better than other forms of exercise (e.g., running, cycling) for mental health—most studies compared walking to no walking or to sedentary controls.

**Major methodological weaknesses:**

**No meta-analysis:** The authors chose not to pool data statistically, so we cannot get a precise overall effect size. This is a limitation of the review design, not necessarily a flaw, but it limits actionable conclusions.

**Heterogeneity:** Studies varied enormously in walking dose (10 minutes to 60 minutes, single sessions to 12-week programs), setting (indoor treadmill vs. outdoor nature vs. urban), and population (healthy vs. depressed). This makes it impossible to generalise.

**Publication bias:** The authors did not formally test for publication bias (e.g., funnel plot), so positive results may be overrepresented.

**No quality assessment:** They did not systematically rate the quality of included studies (e.g., using Cochrane Risk of Bias tool). Some studies may be weak (small samples, no blinding, no control group).

**Date range:** Search ended in 2017. More recent studies (2018–2025) are not included, so the review may be outdated for some outcomes.

**Depression (most evidence):**

Five existing systematic reviews were identified, all reporting that walking reduces depressive symptoms. Effect sizes ranged from small to large (Cohen's d = 0.30 to 1.20), with moderate effects most common (d ≈ 0.50–0.70).

One systematic review of 11 RCTs (n = 341) found that walking significantly reduced depression scores compared to controls (standardised mean difference [SMD] = -0.86, 95% CI: -1.12 to -0.61, p < 0.001). This is a large effect.

Another systematic review of 8 RCTs (n = 491) found walking reduced depression (SMD = -0.63, 95% CI: -0.97 to -0.29, p < 0.001).

Walking was as effective as other forms of exercise (e.g., running, strength training) for depression in some studies, but not all.

Effects were seen in both clinical (diagnosed depression) and non-clinical (elevated symptoms) populations.

**Anxiety:**

12 individual studies were identified. Results were mixed.

Some studies found that a single bout of walking (20–30 minutes) reduced state anxiety (STAI scores decreased by 4–8 points, p < 0.05). This is a small-to-moderate effect.

Longer walking programs (8–12 weeks) showed inconsistent effects on trait anxiety. Some found reductions (e.g., STAI trait scores decreased by 5–10 points), others found no significant change.

Walking in natural settings (parks, forests) reduced anxiety more than walking in urban settings (effect size difference: d ≈ 0.30–0.50).

**Psychological stress:**

10 individual studies. Most found that walking reduces perceived stress (PSS scores decreased by 2–5 points, p < 0.05).

One study found that a 10-minute walk reduced cortisol levels by ~15% compared to sitting (p < 0.01).

Group walking (e.g., "walking groups") reduced stress more than solo walking in some studies (effect size difference: d ≈ 0.40).

**Psychological well-being:**

15 individual studies. Most found positive effects.

Walking programs (8–12 weeks) increased WHO-5 scores by 10–20 points (0–100 scale, p < 0.05).

A single 30-minute walk improved mood (PANAS positive affect increased by 3–5 points, p < 0.05) and reduced negative affect (decreased by 2–4 points, p < 0.05).

**Subjective well-being:**

8 individual studies. Results were mixed.

Some found that regular walking (e.g., 3–5 times/week, 30 minutes) increased life satisfaction (SWLS scores increased by 2–4 points, p < 0.05).

Others found no significant effect, especially in studies with short durations (< 4 weeks).

**Social isolation and loneliness:**

5 individual studies. All found that group walking reduced loneliness (UCLA Loneliness Scale scores decreased by 3–8 points, p < 0.05).

Effects were larger for older adults (age 65+) and for people who walked in groups of 5–15 people.

**Resilience:**

Zero studies found. This is a major gap.

**No harmful effects** were reported in any study for any outcome.

Translate the numbers into plain English:

**Depression:** A moderate-to-large effect (SMD = -0.63 to -0.86) means that walking reduces depressive symptoms by roughly the same amount as antidepressant medication or cognitive-behavioural therapy in mild-to-moderate depression. For a person scoring 20 on the Beck Depression Inventory (moderate depression), this would translate to a drop of ~6–9 points (to mild or minimal depression).

**Anxiety:** A single 30-minute walk reduces state anxiety by about 4–8 points on the STAI (40–80 scale). This is roughly equivalent to the calming effect of a low-dose benzodiazepine (e.g., 0.5 mg lorazepam) but without the sedation or side effects.

**Stress:** A 10-minute walk reduces cortisol by ~15%. For context, a typical stress-reduction meditation session (20 minutes) reduces cortisol by ~20–25%. So walking is about 60–75% as effective as meditation for acute stress reduction.

**Well-being:** A 12-week walking program increases WHO-5 scores by 10–20 points (0–100 scale). This is roughly equivalent to the improvement seen with a 12-week course of mindfulness-based stress reduction (MBSR).

**Loneliness:** Group walking reduces UCLA Loneliness scores by 3–8 points (20–80 scale). This is a small-to-moderate effect, roughly equivalent to joining a weekly social club.

**What the authors acknowledge:**

The evidence base is "fragmented and incomplete" for most outcomes except depression.

No studies on resilience.

Most studies are short-term (< 12 weeks), so long-term effects are unknown.

The setting and context of walking (indoor vs. outdoor, group vs. solo) seem to matter, but the review cannot determine which is best.

Publication bias is possible (positive results more likely to be published).

Many studies have small sample sizes (n < 50), limiting statistical power.

**What a critical reader would note:**

**No meta-analysis:** Without pooled effect sizes, we cannot compare walking to other interventions (e.g., running, therapy, medication) with confidence.

**Heterogeneity of walking dose:** Studies used wildly different protocols—10 minutes once a week vs. 60 minutes five times a week. The review cannot tell you the optimal dose.

**Self-report bias:** Most outcomes are self-reported (mood, stress, well-being). Walking is not blinded (participants know they are walking), so placebo effects are likely.

**No blinding:** In walking studies, participants cannot be blinded to the intervention. This inflates effect sizes, especially for subjective outcomes like mood.

**Confounding by social interaction:** Group walking studies cannot separate the effect of walking from the effect of socialising. The social component may be driving the benefits for loneliness and well-being.

**Confounding by nature exposure:** Outdoor walking studies cannot separate the effect of walking from the effect of being in nature. Green space exposure alone improves mood and reduces stress.

**Population limits:** Most studies are in healthy adults or people with mild-to-moderate depression. Results may not generalise to severe depression, bipolar disorder, schizophrenia, or children/adolescents.

**Industry funding:** Not reported, but walking studies are rarely funded by industry (unlike drug trials). This is a strength, not a weakness.

**Date range:** Search ended in 2017. The evidence base has likely grown since then, especially for outdoor/nature walking and for anxiety.

For someone running their own n=1 experiment:

### What to test (specific intervention and dose)

**Primary test:** Walk outdoors (in a park or green space) for 30 minutes, 5 days per week, at a moderate pace (brisk enough that you could talk but not sing). This is the most common dose in the studies.

**Alternative test:** Compare outdoor walking vs. indoor treadmill walking (same duration and intensity) to see if nature matters for you.

**Alternative test:** Compare group walking (with 1–3 friends) vs. solo walking to see if social context matters.

**Dose-response test:** Try 10-minute walks vs. 30-minute walks vs. 60-minute walks (same setting) to find your personal minimum effective dose.

### Minimum meaningful duration

**For acute effects (mood, stress, anxiety):** A single 30-minute walk should produce measurable changes within 30–60 minutes post-walk.

**For chronic effects (depression, well-being, loneliness):** Run the experiment for at least 4 weeks, ideally 8–12 weeks. Most studies showing benefits used 8–12 week programs.

**For long-term effects:** Extend to 6 months if you want to see if benefits persist or plateau.

### What to measure (specific metrics)

**Primary outcome:** Depression or mood. Use the PHQ-9 (0–27, lower = better) or the PANAS (positive and negative affect). Measure before starting the experiment, then weekly.

**Secondary outcomes:**

- Anxiety: Use the GAD-7 (0–21, lower = better) or STAI state version (20–80, lower = better).

- Stress: Use the Perceived Stress Scale (PSS, 0–40, lower = better) or measure salivary cortisol (if you have access to a lab).

- Well-being: Use the WHO-5 (0–100, higher = better) or the Satisfaction with Life Scale (SWLS, 5–35, higher = better).

- Loneliness: Use the UCLA Loneliness Scale (20–80, lower = better) if walking with others.

**Objective measure:** Use a step counter or GPS watch to log actual walking duration and distance. Self-reported walking is unreliable.

**Measure immediately after each walk:** Rate your mood on a 1–10 scale (1 = terrible, 10 = great) before and 30 minutes after each walk. This captures acute effects.

### Key confounds to control for

**Time of day:** Walk at the same time each day (e.g., 7:00 AM). Morning vs. evening walks may have different effects due to circadian rhythms.

**Weather:** Outdoor walks are affected by weather. Note temperature, sunlight, and rain. If