Diet, gym, supplements, or maybe it is all in your mind? A systematic review and meta-analysis of studies on placebo and nocebo effects in weight loss in adults

The study addresses whether placebo and nocebo effects influence weight loss outcomes in adults. Rising obesity and central adiposity prevalence (e.g., NHANES 2011–2018 showing androidal obesity increasing from 54.5% to 59.1%) make effective interventions a public health priority. While weight loss primarily depends on energy intake versus expenditure, psychological factors—including expectations and treatment context—may shape outcomes. Placebo effects are improvements due to psychosocial context (difference between placebo and non-treatment/natural history groups), while nocebo effects are worse outcomes due to negative expectations or context. The review aimed to fill a gap, as no prior systematic review comprehensively assessed placebo/nocebo effects in weight loss. Research questions: (1) Is a placebo effect present in adult weight loss? (2) What is its magnitude? (3) Is a nocebo effect present? (4) What is its magnitude? (5) Is there a difference between placebo and nocebo effect magnitudes in adult weight loss?

The authors situate the work within extensive evidence that placebo and nocebo effects affect many clinical outcomes, though their role in weight loss is less described. Placebo effects are defined as benefits due to psychosocial context, distinct from placebo responses, which also include natural history and regression to the mean. Nocebo effects represent adverse outcomes compared to non-treatment/natural history groups. Prior publications suggest placebo effects may influence outcomes beyond symptom relief, potentially including weight reduction and appetite regulation. Expectation-based mechanisms—especially verbally induced expectations—can produce physiological changes. Conversely, expectancy tied to medicalized contexts (e.g., supplements) may foster illusory invulnerability and risk compensation, potentially yielding nocebo-like outcomes for weight loss. A recent review (Chin et al., 2022) focused only on pharmacological obesity trials and placebo responses; the present study broadens scope to pharmacological and nonpharmacological placebo interventions and emphasizes true placebo/nocebo effects (versus responses), including body composition and fat distribution metrics.

Design: Systematic review and meta-analysis, preregistered on PROSPERO (registration noted; protocol preregistered 14 Jan 2022) and conducted per PRISMA guidance, with a supplementary search in August 2023. Search strategy: PICOS framework with MeSH and natural-language terms; databases searched included PubMed, Cochrane, PsycINFO, PsycARTICLES, TripDatabase, and Embase. No publication timeframe limits; English-language only. Inclusion criteria: adult humans in weight loss programs or studies examining placebo/nocebo effects in weight loss; at least two groups (placebo and relevant control—no-treatment/natural history); reporting body weight or derived indicators before and after intervention. Possible inclusion of placebo-controlled pharmacological trials required a natural history/no-intervention group. Exclusion criteria: one-arm studies; pediatric populations; animal studies; nonprimary research (reviews, meta-analyses, secondary publications, letters, conference abstracts). Study selection: Two independent reviewers screened titles/abstracts and full texts; disagreements resolved by discussion or third reviewer arbitration. Risk of bias: Two reviewers used the Cochrane Risk of Bias Tool (random sequence generation, allocation concealment, blinding of participants/personnel and outcome assessors, incomplete outcome data, selective reporting, and other biases). Data extraction: Two independent reviewers used a standardized, pilot-tested form to extract sample size; age; sex; baseline and postintervention body weight/derived indicators; measurement methods and calculations; intervention description (including nature of placebo); duration; methods and outcomes for subjective satisfaction (if any). Synthesis and analysis: Narrative synthesis included all eligible studies. Meta-analysis included studies with comparable outcomes, reporting, and methodology. Continuous outcomes were analyzed using standardized mean difference (SMD) with 95% confidence intervals under a random-effects model. Statistical significance threshold p≤0.05. Software: Review Manager v5.4 and Microsoft Excel 2019. Sensitivity analyses: Excluded smallest-sample study to assess robustness.

- Studies included: Five randomized controlled studies met criteria; four were meta-analyzed (Crum & Langer 2007; Merikle & Skanes 1992; Panayotov 2019; Stanforth et al. 2011). Tippens et al. (2014) was excluded from the meta-analysis due to incompatible reporting (change scores only). Interventions ranged 4–12 weeks and included sham supplementation, suggested hypocaloric diet and increased exercise, and placebo subliminal audiotapes. Primary outcomes: body mass, BMI, waist circumference (WC), waist-to-hip ratio (WHR), and body fat percentage (BF%). - Risk of bias: Several domains showed unclear risk, including sequence generation, allocation concealment, blinding, and attrition; small samples were common. - Body weight: Meta-analysis of final values showed no significant difference between placebo and control (SMD [95% CI]: −0.07 [−0.38; 0.24]). Individual studies showed mixed results; in some, greater weight reduction occurred in placebo vs control, but not consistently. - BMI: Pooled analysis (three studies) showed no statistically significant difference, with a trend favoring placebo (SMD [95% CI]: −0.24 [−0.58; 0.11]). - Body fat percentage: Pooled analysis (three studies) approached significance for greater reduction with placebo (SMD [95% CI]: −0.27 [−0.61; 0.07]). In several studies, experimental/placebo groups had significant baseline-to-final reductions not seen in controls. - WC and WHR: WC (two studies) showed greater reduction in control vs placebo; WHR (one study) showed greater reduction in placebo vs control with significant within-group change for placebo. - Sensitivity analyses: Excluding the smallest study yielded similar results: body weight SMD −0.08 [−0.40; 0.25]; BMI SMD −0.21 [−0.57; 0.15]; BF% SMD −0.26 [−0.62; 0.10]. Restricting to two methodologically similar studies (Crum & Langer; Stanforth) for body mass yielded SMD −0.11 [−0.47; 0.25]. - Nocebo finding: One study (Tippens et al.) suggested expectancy around oral weight-loss agents may be disadvantageous, indicating a possible nocebo effect (e.g., illusory invulnerability among participants believing strongly in supplement efficacy was associated with less weight loss). Overall, preliminary trends indicate possible placebo effects (especially for BMI and BF%) and possible nocebo effects in medically framed interventions.

The review’s findings suggest that placebo mechanisms may modestly enhance weight-loss-related outcomes, particularly BMI reduction and fat distribution/composition changes, when interventions involve non-medical contexts (e.g., beliefs about caloric value or exercise). These effects align with expectation-driven and context effects known to influence physiological outcomes. Conversely, medically framed placebo interventions (e.g., pill/supplement contexts) may induce a nocebo-like effect through illusory invulnerability or risk compensation, potentially undermining adherence and resulting in less favorable weight outcomes, as indicated by Tippens et al. The study differentiates true placebo/nocebo effects (versus responses) and highlights that benefits seen in placebo arms of pharmacological trials may partly reflect factors beyond inactive treatment effects. While pooled effects were not statistically significant, consistent trends favoring placebo for BMI and BF% and WHR improvement in one study indicate that psychosocial context can influence objective anthropometric outcomes. These insights underscore the importance of communication and expectancy management in weight-loss interventions and suggest that integrating psychological and behavioral strategies could augment lifestyle-based programs.

Preliminary evidence indicates that placebo interventions may enhance beneficial outcomes in adult weight-loss programs, particularly in non-pharmacological contexts, while medically framed interventions (e.g., tablets/supplements) may evoke nocebo effects that hinder weight loss. Given the limited number of studies and methodological constraints, definitive conclusions cannot be drawn. Clinicians and allied professionals should consider leveraging positive communication and expectancy to support lifestyle interventions. Future research should include well-powered randomized trials with rigorous methodology, comprehensive anthropometry (including detailed body composition analyses), and clear reporting to more precisely estimate the magnitude of placebo and nocebo effects in weight loss.

- Few eligible primary studies; small sample sizes, limiting power and generalizability. - Multiple domains with unclear risk of bias: random sequence generation (including site-level rather than participant-level randomization in some studies), allocation concealment, blinding of participants/personnel and outcome assessors, and attrition reporting. - Heterogeneous interventions and outcomes; only a subset suitable for meta-analysis. - Meta-analysis used final values rather than change scores due to unavailable standard deviations for changes, which may affect precision (though consistent with Cochrane guidance when change SDs are unavailable). - Limited data on subjective satisfaction and adherence; potential contextual and expectancy effects not uniformly measured. - English-language restriction may introduce language bias.