Will I lose my heat acclimation in the taper?
Some: about 2.5% of the heart-rate and core-temperature gain per day without heat, and re-exposure brings it back fast. From 12 small studies, mostly in fit young men.
Daanen, Racinais and Périard asked how fast the main adaptations to heat acclimation are lost once heat exposure stops, and how fast they come back on re-exposure, because a heat block can disrupt the taper before a hot competition.
They reviewed 21 studies (449 participants, 21 of them women; six early studies among them) and regressed the percentage decay of each adaptation (0% = still fully acclimated, 100% = back to unacclimated) on the number of acclimation days, days without heat, daily exposure duration and heat intensity (WBGT) in 12 studies with usable data.
Heart-rate adaptation was lost at 2.3% per decay day and core-temperature adaptation at 2.6%; neither depended on the number of acclimation days (5 to 14), which the authors read as 5 or more days giving relatively stable adaptations. Longer daily exposures at lower heat intensity held the core-temperature adaptation better.
Sweat-rate decay was not related to decay days but was smaller after longer acclimation and hotter conditions.
Re-acclimation was much faster than decay: in one study of six women, each re-acclimation day undid about eight decay days for heart rate and about twelve for core temperature, and in other studies 1-4 days of re-exposure restored the acclimated state after 12-26 days away.
The authors suggest finishing the heat block before a 2-week taper, preferring heat exposure without exercise during the taper, and re-acclimating for 4-5 days before the event when possible.
Our reading: the direction (heart rate fades first, core temperature holds longer, re-acclimation is quick) is the authors' general picture, but single studies scatter widely (core-temperature decay 5% after 26 days in one, 43-69% after 2-3 weeks and 100% after 2 weeks in others); the per-day rates are a coarse fit to scattered, mostly small and uncontrolled data, and several numbers in the paper do not agree with each other (Key Results).
Protocol
Design: systematic review. Scopus searched from inception to 2 August 2017 (heat, acclim*, decay, temperature, human; English only) plus reference lists; one reviewer screened titles and abstracts; risk of bias by each author with the McMaster critical review form, then by consensus (scores 3 to 8 of 8, Table 2).
Inclusion: human participants; heat acclimation for at least 5 days, decay for at least 7 days in longitudinal studies, quantified exposure and duration; heart rate, core temperature or sweat rate measured.
Studies and participants: 21 full-text studies reviewed, n = 449 participants (428 men, 21 women; Table 1).Twelve studies entered the regressions (two of them report different variables of the same experiment); the others were population studies of people moving from the tropics, had too little data, used different people for the acclimated and unacclimated states, acclimated for only 2 days, or compared daily with every-other-day acclimation (Gill and Sleivert, kept for the qualitative review only).
Acclimation in the analysed studies: 5 to 14 days; WBGT 26 to 36 °C; 60 to 120 min a day. Decay intervals from 3 days to 5 weeks; designs cross-sectional (separate groups returning after different decay periods) and longitudinal (repeated heat-stress tests at least a week apart).
Decay metric: % decay = 100 × (acclimated value − value after decay) / (acclimated value − unacclimated value).
Analysis: stepwise multiple linear regression with % decay of end-exercise heart rate, end-exercise core temperature and sweat rate as dependent variables; number of acclimation days, number of decay days, daily exposure duration and WBGT as candidate predictors. No pooled effect sizes, confidence intervals or heterogeneity statistics.
Funding and interests: Daanen was supported by the European Union research programme Heat-Shield; no other funding; the authors declare no conflicts of interest.
Key Results
Regression equations after stepwise elimination (decay as % of the adaptation acquired):
| Adaptation | Equation (P for each term) | Data points | Fit | Loss per day without heat |
|---|---|---|---|---|
| End-exercise heart rate | 3.6 (P = 0.69) + 2.3 (P < 0.001) × decay days | 29 | r = 0.60, P < 0.001 | 2.3% |
| End-exercise core temperature | −126 (P = 0.11) + 2.6 (P = 0.007) × decay days − 1.2 (P = 0.007) × daily exposure (min) + 6.8 (P = 0.006) × WBGT | 27 | r = 0.69, P = 0.002 | 2.6% |
| Sweat rate | 964 (P < 0.001) − 27.7 (P = 0.002) × acclimation days − 18.2 (P = 0.002) × WBGT; decay days not retained | 13 (7 studies) | r = 0.82, P = 0.004 | not related to decay days |
Abstract summary: about 2.5% of the heart-rate and core-temperature adaptations are lost per decay day; 5 or more acclimation days suffice for both; longer daily exposure improves the core-temperature adaptation; longer acclimation and higher WBGT favour the sweat-rate adaptation.
Core-temperature example (authors): extending daily exposure from 60 to 90 min at WBGT 32 °C cut core-temperature decay from 45% to 9% (it matches the equation at about 10 decay days, our arithmetic; the paper does not say how many).
Scatter between studies (heart rate): 2-29% decay over 3-18 days in fit soldiers; about 30% after both 12 and 26 days; 40% after 1 week and 53% after 2 weeks; 54% after 3 weeks (with cold exposure); 63% after 5 days in women; 46% (summer) and 70% (winter) after 1 week and 99-100% after 3 weeks in another cohort. Core temperature: 5% after 26 days in one study, 43% after 2-3 weeks in two others.
Re-acclimation: in one study (6 women) heart rate = 148 + 0.76 × decay days − 5.86 × re-acclimation days (r = 0.86), and rectal temperature = 38.3 + 0.006 × decay days − 0.07 × re-acclimation days (r = 0.73), so one re-acclimation day undid about 8 decay days for heart rate and about 12 for core temperature (ratios 7.7 and 11.7, our arithmetic); by that fit a month without heat is made up by about 4 days of re-acclimation.Elsewhere: heart rate back to acclimated levels after 2 days of re-exposure following 12 days off and 4 days following 26 days; rectal temperature after 1 day following 26 days; a rectal-temperature plateau after 3.8 days following 2 weeks off and 5.3 days following 4 weeks.
Exercise capacity during decay (3 studies): tolerance time 115 min at the end of acclimation vs 116 min after 3 weeks (an insensitive test, most finished the 120 min); 46.5 vs 46.4 min after 1 week; a maximal ramp test lasting 12.4 min before acclimation, 14.2 min after, 13.7 min after 1 week and 12.7 min after 2 weeks of decay. The authors read this as performance benefits persisting 1-2 weeks.
Perception: RPE and thermal sensation rose only slightly during decay, even over a month (limited data).
The authors' 2-week prediction: for fully acclimated people (10 days of daily acclimation, 90 min, WBGT 30 °C) two weeks without heat would cost about 35% of the heart-rate adaptation, 6% of the core-temperature adaptation and 30% of the sweat-rate adaptation.
Numbers that do not add up:
The equations give 35.8% for heart rate and 6.4% for core temperature at 14 days (our arithmetic), matching the prediction; the sweat-rate equation as printed gives 964 − 27.7 × 10 − 18.2 × 30 = 141% and contains no decay-day term, so the "about 30%" for sweat rate does not follow from it.
The Limitations paragraph says decay in heart rate and core temperature is "strongly dependent on the number of HA days and to a lesser extent on the number of HAD days" and that the sweat-rate regression "is not significant"; the Results say the opposite on both counts (acclimation days dropped out of both equations; the sweat-rate equation has P = 0.004).
The abstract says eight studies investigated re-acclimation; the body says nine and names nine.
The sweat-rate regression lists seven studies, among them Flouris et al. [103], which Table 4 shows did not measure sweat rate and which reports the same experiment as Poirier et al. [42].
The Results say participants per analysed study ranged from 6 to 15; Table 1 lists 16 or 24 for four of the analysed studies (the per-group sizes were smaller).
"Only three of 20 studies included female participants" — Table 1 lists 21 studies, three with women.
The abstract says twelve studies "met the criteria and were systematically reviewed"; the body reviews 21 and uses 12 of them in the regressions.
Mechanisms
Authors' interpretation: the heart-rate adaptation decays faster than the core-temperature adaptation (their "general picture" across scattered studies; the regression slopes are similar, 2.3% vs 2.6% per decay day). Measured: sweat-rate change during acclimation itself was small and inconsistent (mean rise 21%, SD 13%; two studies found none), which makes its decay hard to quantify.
Authors' interpretation: re-acclimation is fast mainly because part of the adaptation is retained, and also because the rate of adaptation is higher on re-exposure; they cite a molecular heat-acclimation "memory". Decay may involve a gradual return of peripheral vasoconstriction (from heart-rate variability data showing parasympathetic dominance after acclimation).
Authors' interpretation: longer daily exposure gives a more durable core-temperature adaptation and hotter exposures a more durable sweat-rate adaptation, so they suggest keeping WBGT above about 30 °C and extending the daily exposure.
Measured, unexplained: in one study exercising in cool conditions after acclimation sped up decay compared with not exercising (heart rate 29-35% vs 17%; rectal temperature 24% vs 2% at 5 days); the authors could not explain it and note it runs against the older view that fitness and regular exercise help retain acclimation.
Measured, in one included study: after 12 days of maximal exercise in extreme heat with no cool recovery (15 men), core temperature had changed little during the block, yet end-exercise rectal temperature 3 days after it was 0.5 °C lower than before acclimation; the authors read this as adequate recovery being needed for the adaptation to appear.
Inference: that matches REF-3489, where the lower rate of core-temperature rise appeared 4-8 days after the block; part of the core-temperature gain can surface after a block rather than during it.
Practical Application
Will I lose my heat acclimation in the taper? Some: about 2.5% of the heart-rate and core-temperature gain per day without heat, and re-exposure brings it back fast. From 12 small studies, mostly in fit young men.
Put the heat block before the taper, not inside it (the authors' recommendation): after 10 days of acclimation, two weeks without heat would cost about a third of the heart-rate adaptation and little of the core-temperature adaptation (their regression estimate, not a measured taper). The base's "do not butt it against the start line" (SYNTH-04, from REF-3489) points the same way.
Top up inside the taper with heat, not with extra training (the authors' suggestion, not tested here): heat exposure without exercise during the taper, then 4-5 days of re-exposure just before a hot event if time allows.An abstract they cite reported that exercise-heat sessions every fifth day left heart rate and core temperature "better sustained" 25 days after acclimation; that is the nearest thing here to a maintenance dose. The base's 2-a-week maintenance remains practice (SYNTH-04).
Sweat rate is the least predictable adaptation: longer blocks and hotter sessions held it better; a short block may lose it soonest (inference from the sweat-rate regression, 13 data points).
Rehearse it first: the authors advise testing the whole sequence (block, taper, re-acclimation) before an important competition, because it may interfere with other pre-race needs.
Limitations
Small, old, heterogeneous studies. The analysed studies had 6 to about 24 participants, the review includes six early studies, and the protocols used walking, load carrying, cycling and rowing at WBGT 26-36 °C; the authors' own limitation is the lack of standardised acclimation and decay protocols. Most had no control group (Table 2).
A linear fit to scattered points. The per-day rates come from stepwise linear regressions on study-level data (heart rate r = 0.60, core temperature r = 0.69); individual studies range from almost no decay to complete loss in similar intervals, and decay percentages of −246% to +180% for resting core temperature show how unstable the metric is when the acclimation change is small.
Re-acclimation ratio from one study. The 8-12× figure is Stephens and Hoag's regression in 6 women, whose pre-acclimation heart rate was extrapolated (155 beats/min) because their data start on acclimation day 4.
Who. Mainly fit young men; three studies with women; none with older people; no competitive athletes; nothing on sex differences in decay (the authors' gap). Conditions during decay were rarely specified (two studies did).
Outcome. Heat-stress test physiology, not race performance; only three studies measured exercise capacity during decay.
Method. One database, one screener, English only; no pooled effect sizes or heterogeneity statistics.
Internal inconsistencies (listed under Key Results): the Limitations paragraph contradicts the Results on acclimation days and on the sweat-rate regression; 8 vs 9 re-acclimation studies; participants per study; 20 vs 21 studies; the sweat-rate 2-week prediction does not follow from its equation.
Funding: EU Heat-Shield support for one author; no conflicts declared.
How this entry was checked and changed
Why this level
read from the full text (Europe PMC, open access, CC BY 4.0). The rubric's A- line is an SR/MA "with serious heterogeneity or mostly small trials"; graded one step lower because the inputs are mostly uncontrolled longitudinal or cross-sectional heat-stress studies rather than controlled trials, no pooled effect size, confidence interval or heterogeneity statistic is reported, and the pooled rates come from stepwise regressions on 29 (heart rate), 27 (core temperature) and 13 (sweat rate) study-level data points from 12 small, mostly uncontrolled studies, only one database was searched with one screener, and the 8-12× re-acclimation figure rests on a single study of 6 women
Full text (checked 2026-10-26)
Adaptations in end-exercise HR decreased by 2.3% (P < 0.001) for every day of HAD; for end-exercise T c the daily decrease was 2.6% (P < 0.01) (full text)
Stepwise multiple linear regression analyses used 12 studies: 29 datapoints for HR, 27 for T c, and 13 for SR (full text)
HRA was 8-12 times faster than HAD at inducing adaptations in HR and T c, but no differences could be established for SR (full text)
A month of HAD could be compensated for by only 4 days of HRA (Stephens and Hoag, HR = 148 + 0.76 decay days − 5.86 re-acclimation days, r = 0.86) (full text)
Extending daily exposure from 60 to 90 min at a WBGT of 32 °C reduced HAD in T c from 45 to 9% (full text)
In the 2 weeks after HA a decay in HR (~ 35%), T c (~ 6%), and SR (~ 30%) can be expected for 10 days of daily HA, 90 min exposure, 30 °C WBGT; 4-5 days of HRA is advised before the event (full text)
Maximal ramp performance lasted 12.4 min before HA, 14.2 min after HA, 13.7 min after 1 week and 12.7 min after 2 weeks of HAD (full text)
Hein Daanen was supported by the European Union research program Heat-Shield; no conflicts of interest (full text)