Physiological Responses to Heat Acclimation: A Systematic Review and Meta-Analysis of Randomized Controlled Trials.

Author:Rahimi, Gholam R. Mohammed


Hot and humid ambient environments affect the human physiological response to physical activity (No and Kwak, 2016; Tucker et al., 2004), and exercise-induced increases in core temperature ([T.sub.C]) and increased ambient temperature result in premature fatigue and loss of athletic performance (Al-Nawaiseh et al., 2013). Muscle and skin blood flow (Sawka et al., 2011), aerobic capacity (Tucker et al., 2004), early onset of anaerobic threshold (Tyka et al., 2009), stimulation and accumulation of stress hormones (Jones et al., 2010), increased anaerobic glycolysis, increased use of intramuscular glycogen and lactate accumulation (Tan et al., 2018), are all factors associated with increased body temperature during exercise. However, competitive athletes compete in a variety of environmental conditions, which emphasizes the importance of reducing the potential adverse effects of a hot environment on athletic performance (Al-Nawaiseh et al., 2013). Heat acclimatization (HA) has been undertaken by endurance athletes to enhance tolerance and exercise performance in hot conditions (Guy et al., 2016; Periard et al., 2015a). Generally, HA has been undertaken at sub-maximal intensities during exposure to elevated temperature and/or humidity (Taylor, 2014; Tyler et al., 2016). Some of these studies have reported beneficial responses, such as decreases in heart rate (HR), [T.sub.C] and skin temperatures ([T.sub.S]) (Brade et al., 2013).

Alterations in critical physiological parameters such as increased plasma volume (PV), reduced exercise HR ([HR.sub.E]) (Kelly et al., 2016; Sawka et al., 2011), lower resting and exercise [T.sub.C] (Sawka et al., 2011; Tyler et al., 2016), and improved maximal cardiac output (Sawka et al., 1985; Shvartz et al., 1977), can all lead to increased performance (Guy et al., 2016) via short and medium duration HA programs. These adaptations could be beneficial for consequent performance in the heat, as well as in the cold, where possible fluid loss may be considerable (Corbett et al., 2014). The apparent dose-response to HA proposes that 15 days or more is required to optimize performance (Guy et al., 2015). Nevertheless, PV, HR and [T.sub.C] adjustments can take place as quickly as four days, and sustained thermal adaptation needs to be maintained by regular exertional exposure to hot climates (Weller et al., 2007).

Numerous elite sporting competitions are programmed in geographical locations that involve exposure to hot and humid environments, such as the 2020 Tokyo Olympics and 2022 Qatar World Cup. Hence, it is essential that athletes should be readied for such competitions, especially those who live and exercise in cold environments or are unaccustomed to heat stress (Milne and Shaw, 2008).

A previous meta-analysis on HA efficacy was conducted (Tyler et al., 2016), but only included investigations up until February 2016 and also included non-randomized, controlled trials. Additional randomized controlled trials (RCTs) have since been published, and this argues in favor of carrying out an updated meta-analysis. This update includes data pooling for several outcomes that were previously not carried out because of insufficient outcome data. The aim of this meta-analysis was to update previous pooled analyses, using only level 1 (RCT) evidence, on outcome measures relating to thermoregulatory adaptations attributed to HA with athletic performance in RCTs.


Search strategy

This systematic review and meta-analysis has been rep orte d using the PRISMA guid eline s (Liberati et al., 2009). Accordingly, using a PubMed search strategy (1966 to Nov 1, 2018), we identified relevant articles by the following keywords: "acclimation", "acclimatization", "heat", "acclimation and performance", "temperature", "exercise training in heat". Then, after the initial screening, the references of all studies based on the inclusion and exclusion criteria were examined to find additional studies.

Study selection

Two reviewers separately looked at the titles and read the abstracts and filtered relevant articles to be included. The four-phase (identifying, screening, qualification and inclusion) method identified were used in the PRISMA report to diminish the number of primary search results.

Inclusion/exclusion criteria

Study design: full text articles of controlled trials and RCTs of heat exercise training of natural and artificial HA, excluding review articles, conference abstracts and study protocols.

Comparison intervention: study protocols that used heat training (HOT) with thermo-neutral training (NEUTRAL), in a pre-post design, excluding the studies on acute interventions (e.g., single-session interventions) and also water immersion interventions.

Population: Men and women (age [greater than or equal to] 18 years) who identified as triathletes, runners, endurance athletes, cyclists or team sport athletes from both elite and sub elite competitive levels.

Publications: English language manuscripts published in specialised English journals.

Outcome measures

The outcome measurements of this meta-analysis were;

Time trial (TT) performance (in seconds): in included studies TT performance were measured with the test of maximal leg cycle exercise test (time to reach exhaustion), 5 km TT performance, 20-km cycling, and running 3 km TT on a motorized treadmill.

Maximum oxygen uptake (V[O.sub.2max] in ml/kg/min), Exercise ([HR.sub.E]), time trials ([HR.sub.TT]) and maximal heart rate ([HR.sub.M]).

Core ([T.sub.C]) and mean skin temperature ([T.sub.S]): in included studies [T.sub.C] was measured in the rectal (Guy et al., 2016; Lorenzo et al., 2010; Sunderland et al., 2008; Willmott et al., 2016), gastrointestinal (Chalmers et al., 2016; Petersen et al., 2010; Schmit et al., 2018), and oesophageal (Nielsen et al., 1993) sites. In addition, end exercise values (Lorenzo et al., 2010; Nielsen et al., 1993) and also delta temperature from pre-post values (Chalmers et al., 2016; Guy et al., 2016; Petersen et al., 2010; Schmit et al., 2018; Willmott et al., 2016) were used for T.sub.C and T.sub.S in included studies.

Thermal comfort ([T.sub.Comf]): in included studies [T.sub.Comf] was determined according to the 5-point scale, 8-point scale or a 10-point scale.

Plasma volume (PV in percent): change in PV (%) was estimated using the method of Dill and Costill 1974 or PV was calculated from body mass by the equation of Sawka et al 1992.

Blood lactate (mmol.L).

Rate of perceived exertion (RPE): in included studies RPE was measured using the Borg and Kaijser 2006 scales.

Statistical analysis

For all included studies, we summarized the effect size for any outcome by measuring the mean difference between the heat and neutral condition from before and after the intervention. If multiple articles were published from the same dataset then we checked the data in order to avoid using the same results for the same outcome measure on more than one occasion. Results were analyzed by weighted mean difference (MD), if the measurement method or reporting was identical. For outcomes using different measurement or reporting techniques a standardized mean difference (SMD) was used. All analyses were conducted using Review Manager 5.3 (The Nordic Cochrane Centre, Copenhagen, Denmark). Extracted outcome data employed were change in the mean [+ or -] SD. In studies that reported SE data, these were converted to SD. A random-effects inverse variance was employed. When a standardized mean difference (SMD) was used the guideline for commentary was used (Cohen, 1988), with 0.2 described as small, 0.5 medium and 0.8 as large. Where an article contained a control group and more than one HA group, we separately labelled each HA groups and adjusted the sample size of the control group according to the number of HA groups. We presented meta-analysis using forest plots and applied a 5% level of significance to describe the significance of results.

Heterogeneity: To evaluate the heterogeneity among the studies, the I2 statistic was employed, with values > 50% showing substantial heterogeneity (Higgins et al., 2003). The risk of publication bias was assessed using the Egger plot (Egger et al., 1997). Any analysis of heterogeneity depends on the number of trials included in a metaanalysis, which is generally small, and this limits the statistical power of the test. We therefore based evidence of asymmetry on P

Study quality: Study quality and reporting was assessed using the validated TESTEX scale (Table 1) (Smart et al., 2015). This is a validated 15-point scale which evaluates quality of the study (5 points maximum) and reporting (10 points maximum). A study with a TESETX quality score of less than 10 was deemed of low quality.


Study characteristics

Figure 1 displays the selection process employed to include manuscripts in our meta-analysis. Of 617 possibly associated articles reclaimed from the search, 113 were animal studies and a further 489 were excluded by title or abstract, leaving 19 full text articles. A further 4 were excluded as duplicate studies, two more were excluded as they used immersion water protocol and 2 used an acute protocol, leaving 11 studies for the meta-analysis.

The characteristics of the included articles are shown in Table 2. The eleven included...

To continue reading