Placebo and nocebo effects on sports and exercise performance: A systematic literature review update
Protocol
Update of the 2019 systematic review on placebo/nocebo in sport and exercise
Search of PubMed, Medline, Web of Science, EBSCO and Google Scholar for empirical studies published 2019 to end of May 2024
20 eligible studies with control or baseline-control conditions; interventions were nutritional, mechanical and other mixed ergogenic aids
Effect sizes (Cohen's d) calculated for placebo and nocebo separately
Key Results
Pooled placebo effect: d=0.67 (moderate to large), larger than in the earlier review
Placebo effects by intervention type: nutritional d=0.86, mechanical d=0.38, cream and gel d=0.05, open-label placebo d=0.16
Nocebo effects: d=1.20 — almost twice the pooled placebo effect, but based on only five measures from three studies
Nocebo appears stronger than placebo in sports — negative beliefs may hurt more than positive beliefs help, on a thin evidence base
Mechanism
Not from this paper (general placebo/nocebo literature): negative expectations raise anticipatory anxiety and alter pain and effort perception, which can produce measurable performance decrements. The review quantifies the effects; it does not test these pathways.
Practical Application
Across 20 sport/exercise studies (2019-2024), placebo interventions improved performance by a pooled d=0.67; nocebo effects were almost twice as large (d=1.20), though from only five measures in three studies — telling an athlete they are "not recovered" can plausibly impair performance beyond any physiological state
Inference, not tested here: when wearables show "poor recovery" → negative expectancy → nocebo → real performance drop
Coaches' language around data interpretation matters
Framing: "Your HRV is slightly below trend" vs "You're not recovered" — same data, different nocebo risk (the base's own practice, not tested in this review)
Not from this paper; the base's own practice: consider withholding negative wearable data before key sessions/races
Limitations
Included studies varied in design and quality
The nocebo estimate rests on five measures from three studies
Most nocebo studies used verbal/contextual manipulation, not wearable data specifically
Sports-specific nocebo research still limited
How this entry was checked and changed
Content correction 2026-09-27
Sport line "swimming" -> "general" (a cross-sport systematic review).
Named nocebo examples (McLemore 2020 d=1.8/2.5, Corsi 2019 d=0.67-1.51, Zech 2019 7-9%) -> not in the abstract, removed.
Added the abstract's caveat that the nocebo d=1.20 rests on five measures from three studies; added per-intervention placebo effect sizes from the abstract.
Mechanism (cortisol, motor cortex, cholecystokinin/dopamine) -> labelled as not from this paper.