Key takeaway
Across 22 randomised trials (375 participants, mostly men), one cold-water immersion after exercise did not change maximal isometric strength (g 0.08, 95% CI −0.11 to 0.26) and was followed by weaker jumping straight afterwards (g −0.68), gone by 24–48 hours. Lower soreness and creatine kinase looked favourable but weakened in sensitivity checks, and certainty was low to very low.
Study design
Trial meta-analysis
Reading basis
Full text
Reading focus
Design & applicability
The research question
Cold-water immersion is a standard recovery tool in elite sport, where congested schedules leave little time between sessions and matches. Earlier reviews reported benefits for soreness but pooled many time points and disagreed on strength, and one trial suggested immersion can hurt performance in the short term, which hints that its effect changes over the hours after exercise. No review had quantified what a single immersion does at 0, 24, 48 and 72 hours. This meta-analysis set out to measure how one acute bout of cold-water immersion affects strength, explosive jumping, a muscle-damage blood marker and soreness at each of those time points, in order to find the best window for using it.
Who and how
This is a systematic review and meta-analysis of randomised controlled trials, registered on PROSPERO before data extraction and following Cochrane and PRISMA guidance. The authors searched PubMed, Web of Science, Cochrane Library and Embase up to 1 September 2025, screening 833 records down to 22 trials published between 2008 and 2025. These included 375 participants (362 males and 13 females, aged 15.6 to 48.9 years): 13 studies of professional athletes and 9 of healthy non-athletes, with 9 parallel-group trials and 13 crossover trials that all used washout periods. Eligible trials gave a single cold-water immersion within 1 hour after exercise and compared it with seated passive recovery at 20–25 °C; trials adding other recovery methods or supplements were excluded. Exercise ranged from sprint intervals and match simulations to a half marathon and eccentric strength work, and immersion ranged from 5.0 to 20.0 °C for 6 to 20 minutes.
Four outcomes were pooled at 0, 24, 48 and 72 hours after exercise: maximal voluntary isometric contraction (MVIC, peak force), countermovement jump (CMJ, jump performance), creatine kinase (CK, a blood marker of muscle damage) and soreness on a visual analogue scale (VAS). The effect measure was Hedges' g from pre-to-post change scores (negligible below 0.2, small 0.2–0.5, medium 0.5–0.8, large above 0.8), with fixed- or random-effects models chosen by heterogeneity, tests for differences between time points, and leave-one-out checks. Because studies contributed several effect sizes, the authors added a cluster-robust analysis by study, assessed publication bias with funnel plots, Egger's test and trim-and-fill, and graded certainty with GRADE. Study quality was rated with the PEDro scale (all 22 trials scored 6–8 out of 10) and the Cochrane RoB 2 tool (all 22 rated high risk, driven by the impossibility of blinding participants).
What it found
- 1A single immersion did not change maximal isometric strength at any time. Six studies (15 effect sizes) gave g = 0.08 (95% CI −0.11 to 0.26, p = 0.42), with no heterogeneity (I² = 0.00%) and moderate certainty; none of the four time points differed (all p ≥ 0.36), and the test for differences between time points gave p = 0.88.
- 2Jumping was worse immediately after exercise, then recovered. Across 11 studies (23 effect sizes) the overall effect was null (g = 0.01, 95% CI −0.25 to 0.28, p = 0.92), but it differed by time (p = 0.004): g = −0.68 at 0 hours (95% CI −1.16 to −0.20, p = 0.01), 0.38 at 24 hours (95% CI −0.02 to 0.77, p = 0.06) and 0.06 at 48 hours (95% CI −0.22 to 0.33, p = 0.68). Certainty was low and heterogeneity substantial (I² = 64.49%).
- 3Soreness was lower after immersion in the main analysis, but the finding is fragile. Nine studies (24 effect sizes) gave g = −0.58 (95% CI −0.99 to −0.16, p = 0.006) with very low certainty and I² = 87.32%; only 0 hours was significant on its own (g = −0.62, 95% CI −1.20 to −0.04, p = 0.04). With clustering by study the effect was g = −0.363 (95% CI −0.842 to 0.115, p = 0.118), and a fixed-effect model gave g = −0.22 (95% CI −0.37 to −0.07).
- 4The creatine kinase benefit did not survive bias adjustment. Seven studies (19 effect sizes) gave g = −0.43 (95% CI −0.67 to −0.19, p < 0.001), strongest at 24 hours (g = −0.64, 95% CI −1.08 to −0.20, p < 0.01), but Egger's test flagged small-study effects (p < 0.001), trim-and-fill shrank it to g = −0.160 (95% CI −0.443 to 0.124), and clustering by study gave g = −0.348 (95% CI −0.791 to 0.095, p = 0.103). Certainty was low.
- 5The null results held up, and by 48–72 hours the functional measures were back near baseline. With clustering by study, MVIC stayed at g = 0.076 (95% CI −0.121 to 0.272, p = 0.367) and CMJ at g = −0.002 (95% CI −0.223 to 0.220, p = 0.987), and excluding studies that enrolled women changed no direction. The authors read this as outcome-specific effects rather than a uniform benefit.
Limits & uncertainty
- 1The authors note a profound sex imbalance (362 males against 13 females), so the results cannot be extended to women, and that participants cannot be blinded, which likely inflates a subjective outcome such as soreness. The Cochrane tool rated all 22 trials as high risk of bias for this reason.
- 2The authors note that several trials were very small (the smallest had 7 participants), that Egger's test and trim-and-fill showed publication bias and small-study effects for CK and VAS, and that the cluster-robust analysis rests on a limited number of studies, so it is supportive rather than definitive.
- 3The authors note considerable heterogeneity in water temperature, depth, duration and exercise type, and too few trials for meta-regression or dose subgroups, so the best protocol remains loosely defined. They also looked only at single immersions followed for up to 72 hours, so repeated use and long-term adaptation are not addressed.
- 4Our observation: the time-dependent jump result is a subgroup contrast with low certainty and wide intervals, and the text gives totals rather than how many trials contribute to each time point. The review also tested four outcomes at four time points with no described adjustment for multiple comparisons.
- 5Our observation: the proposed mechanisms, such as slower nerve conduction, were inferred from earlier experiments rather than measured in these trials, as the authors acknowledge, so the reasons for the immediate jump deficit remain untested here.
What to take from it
The most direct use is deciding whether to take a cold dip straight after exercise. The authors advise against routine immediate use when another sprint, jump or other power task is coming within a few hours, since jumping was lower immediately afterwards (though gone by 24–48 hours). If it is the last session of the day and you simply want to feel less sore, a cold dip may help, but that finding has very low certainty and weakened under stricter analyses.
Do not read lower soreness or lower CK as stronger or faster-recovering muscles: maximal strength recovery did not change. The review looked at single sessions, and the authors point to other meta-analyses suggesting that routine use after resistance training may blunt muscle growth and strength gains, which this review did not test. The evidence is almost entirely in men, so apply it to women with extra caution.
Reading your own data
Test power the way the trials did: a countermovement jump just after the dip, then 24 and 48 hours later, on a day with immersion and on a day without it after similar sessions. Record soreness on a 0–10 scale separately, because in these trials soreness and jump performance did not move together, and log water temperature and minutes (the trials used 5.0 to 20.0 °C and 6 to 20 minutes). Do not lean on watch recovery scores or HRV here, since none of the trials measured them.
How to read this evidence
Pooling trials can improve precision, but compare populations, controls, bias and uncertainty. Network rankings include indirect evidence and do not identify a universal best treatment.
Sources & attribution
Temporal dynamics of muscle strength recovery following acute cold-water immersion: a systematic review and meta-analysis
PeerJ · 2026-07-16
DOI: 10.7717/peerj.21537 ↗PubMed · 42473449 ↗MoveVitals editorial interpretation · educational content · no individual clinical review recorded
Sources checked 2026-11-01
Practical guides
Read in context
Cold-water immersion after hard exercise: less soreness, better power, no strength gain
In a meta-analysis of 52 randomised trials, cold-water immersion after high-intensity exercise was followed 24 hours later by lower soreness (SMD −0.89), higher perceived recovery (0.66) and better muscular power (0.22), with no effect on strength. Participants could not be blinded, most were men, and many pooled results rest on few studies with high heterogeneity.
Sports Medicine
Short-term HRV was higher during slow breathing in 962 military men, injured or not
In 962 military men (median age 37.6 years), every short-term heart rate variability (HRV) measure was higher during slow breathing at six breaths a minute than during spontaneous breathing, injured or not; in fast resting breathers RMSSD rose from 15.57 to 21.89 ms. It was one session in a fixed order, so it shows an immediate effect, not lasting benefit.
European Journal of Applied Physiology
Acute:chronic workload ratio: no evidence it can reduce injury risk, authors argue
The authors argue there is no evidence for using the acute:chronic workload ratio (ACWR) in training-load management or to reduce injury risk: no study has properly estimated a causal effect, and the ratio is statistically inaccurate and ambiguous. This is a methodological critique, not a new data analysis, so it does not show that training load is unrelated to injury.
International Journal of Sports Physiology and Performance
Heart rate variability: 24-hour, 5-minute and ultra-short readings are not interchangeable
An expert overview of heart rate variability (HRV) concludes that 24-hour, 5-minute and ultra-short readings measure different things, so norms for one cannot be applied to another; pooled 5-minute norms from 21,438 healthy adults averaged 50 ms for SDNN and 42 ms for RMSSD. It is a narrative review, not a trial, and it tests no consumer wearable.
Frontiers in Public Health

