§ Claim under review · Health
"A systematic review and meta-analysis of 18 randomized controlled trials involving 827 women found that Pilates was associated with reduced depressive symptoms (Ju et al., Medicine (Baltimore), 2023;102(41):e35419)"
Verdict
Mostly accurate
Confidence
HighSummary
The study cited in this post is real and the numbers check out. A 2023 systematic review by Ju and colleagues, published in Medicine (Baltimore), did pool 18 randomized controlled trials with 827 female participants and found that Pilates was linked to a significant reduction in depression symptom scores. The post's wording is careful, it says "associated with" rather than "causes," and it explicitly states Pilates does not cure depression and is not a replacement for mental health care, which matches what the study authors themselves concluded. Two pieces of context are missing. First, most of the women in those trials were not recruited for clinical depression, they were women with conditions like multiple sclerosis, breast cancer, chronic low back pain, type 2 diabetes and schizophrenia, and depression scores were tracked as an outcome. Second, the authors rated the evidence as moderate, only 4 of the 18 trials were top quality, and the benefit was weaker in some groups. The paper also contains an internal inconsistency between the effect size reported in its abstract and the one in its results section, which is a minor quality flag. Overall this is a fairly reported citation with some useful nuance left out.
The readings
key figures from the evidencerandomized controlled trials in the meta-analysis
female participants pooled in the meta-analysis
abstract effect size for depression symptom reduction
Why this verdict
Evidence
The cited paper exists exactly as referenced. The citation details in the post (authors, title, journal, year, volume, issue, article number) all match the published record.
The paper's own abstract states: "18 RCTs with 827 female patients were included. The methodological quality of the RCTs was considered an A level in 4 studies, B level in 13, studies, and C level in 1 study investigation." On the primary outcome, the meta-analysis reported moderate evidence that Pilates significantly improved the severity of depression symptoms (SMD = –0.73; 95% CI –0.86 to –0.59; P < .01) and anxiety symptoms (SMD = –0.62; 95% CI –0.79 to –0.46) .
The authors' stated conclusion is deliberately modest: Pilates exercise could reduce levels of depression and anxiety in female patients with depression and anxiety symptoms, and can be used as a potential ancillary program to improve those symptoms . The word "ancillary" is the authors' own, and it aligns with the post's framing of Pilates as an addition rather than a replacement.
METHODOLOGY AND CONTEXT
Design: Systematic review with meta-analysis of RCTs, prospectively registered (PROSPERO CRD42023426522).
Search: Five electronic databases (Scopus, MEDLINE/PubMed, Embase, Web of Science, and the Cochrane Library) were searched up to January 2023 for RCTs on Pilates in female patients with depressive disorders or elevated depression levels; primary outcome was depression severity, secondary was anxiety; analysis in Stata 15.1, risk of bias assessed with the PEDro scale . Note an internal inconsistency: the full text states the databases were searched from inception through 2023 with the final retrieval day being July 1, 2023 , which does not match the abstract's "up to January 2023."
Screening: After abstract and full-text screening, 18 studies were removed for not including Pilates and depression (n = 2) or for no or unclear outcome measures (n = 16), leaving 18 studies with 827 female patients .
Outcome instruments: Depression was measured with the Beck Depression Inventory in 10 studies, the Hospital Anxiety and Depression Scale in 5, the Calgary Depression Scale for Schizophrenia in 1, plus the Quick Inventory of Depressive Symptomatology, the Edinburgh Postnatal Depression Scale and the General Health Questionnaire . These are symptom-rating scales, not diagnostic confirmation of remission.
Population composition (important context not in the post): The included samples were largely women with chronic medical conditions in whom depression scores were an outcome, not cohorts recruited primarily for major depressive disorder. The paper's own subgroup structure reflects this: subgroups were categorized as young adults, middle-aged and old adults, patients with multiple sclerosis, patients with metabolic diseases, and patients with chronic low back pain . Independent peer-reviewed description of this meta-analysis characterizes it as covering the effect of Pilates on depression in women with various medical conditions such as MS, schizophrenia, chronic low back pain, type 2 diabetes, and breast cancer .
Effect was not uniform across subgroups. The authors themselves report that while findings supported benefit in middle-aged and older women, it seemed to be less useful for patients with metabolic disorders .
Statistical inconsistency within the paper: the abstract reports SMD –0.73 (95% CI –0.86 to –0.59), while the results section reports a different figure for the same 18-study pooled depression analysis: the meta-analysis of 18 studies showed Pilates significantly reduced depression symptoms (SMD = −0.71; 95% CI −0.90 to −0.52; P = .02; heterogeneity: I2 = 44%) . Both point estimates are similar and both are statistically significant, but the confidence intervals and P values differ between the abstract and the body of the same paper. That is a quality flag.
Journal context: Medicine (Baltimore) is a legitimate PubMed-indexed Wolters Kluwer open-access journal, but it is a very high-volume general journal rather than a specialist psychiatry, psychology or sports medicine journal, and its editorial selectivity has been the subject of bibliometric scrutiny. This does not make the paper wrong. It means it carries less weight than a Cochrane review or a specialist journal meta-analysis.
Background knowledge, not a retrieved source: exercise trials cannot blind participants to their own intervention, and depression outcomes are self-reported. This combination is well known to inflate effect sizes in exercise-and-mood meta-analyses, particularly where control groups receive no active comparator. This is a general methodological caveat, offered as context, not as a specific retrieved finding about this paper.
Findings
✓ What's accurate 7
- The cited study exists, is real, is peer reviewed, and is indexed in PubMed. Citation details match exactly.
- "18 randomized controlled trials" is correct.
- "827 women" is correct.
- "Associated with reduced depressive symptoms" is correct and, if anything, more cautious than the paper's own wording of "significantly improved."
- The paper's inclusion criteria did specify females with depressive symptoms or depressive disorders, so the caption's "women with depressive disorders or elevated depressive symptoms" tracks the paper's stated criteria.
- The post's caveats are accurate and consistent with the authors' own conclusion that Pilates is a "potential ancillary program."
- The effect size reported is moderate to large by conventional standards, so "meaningful reduction" is a fair characterization of what the paper claims.
≈ What's misleading 5
- **Omitted qualifier (minor):** the post does not convey that the pooled participants were predominantly women with chronic physical or neurological conditions such as multiple sclerosis, breast cancer, chronic low back pain, type 2 diabetes and schizophrenia, rather than a general population of women with clinical depression. A reader is likely to picture "827 depressed women," which is not quite the sample.
- **Omitted qualifier (minor):** the post does not mention that the benefit was not consistent across all subgroups, with the authors noting weaker results for women with metabolic disorders.
- **Omitted qualifier (minor):** the strength of the underlying evidence is characterized by the authors themselves as "moderate," with most included trials rated B-level quality on the PEDro scale and only 4 of 18 rated A-level. The post presents the finding without any evidence-quality signal.
- **Marketing framing (mild):** the on-image headline "Science shows Pilates can be a powerful ally for your mental health" is stronger than the paper supports on its own. The caption walks this back substantially.
- The post does not mention that the paper contains an internal discrepancy between its abstract and results section on the primary effect estimate, or between its two stated search cutoff dates. Most reposting accounts would not catch this, but it bears on how much weight the finding should carry.
? What's uncertain 4
- Whether the discrepancy between the abstract SMD (–0.73, CI –0.86 to –0.59) and the results-section SMD (–0.71, CI –0.90 to –0.52) reflects a typographical error, a fixed versus random effects difference, or a data-handling error. The paper does not explain it in the text retrieved.
- The full limitations section as written by the authors could not be retrieved in full within the available search budget, so their own stated caveats are only partially documented here.
- Whether any of the included trials used active exercise comparators versus waitlist or usual-care controls, which materially affects how much of the effect is attributable to Pilates specifically rather than to attention, group contact, or general activity.
- Durability of effect. No long-term follow-up data was located; the subgroup analysis covered intervention durations of 8, 12 and 16 weeks only. ---
Sources
2 of 3 linked to records**Ju M, Zhang Z, Tao X, Lin Y, Gao L, Yu W. "The impact of Pilates exercise for depression symptoms in female patients: A systematic review and meta-analysis." Medicine (Baltimore). 2023 Oct 13;102(41):e35419**
**Publisher full text (Lippincott/Ovid), Medicine journal**
**Secondary citations of the paper in later peer-reviewed literature** (BMC Psychology 2025 reformer Pilates RCT; IBS/Pilates trial discussion; Healthcare 2025 observational Pilates study)