Summary of Search, “Best Evidence” Appraised, and Key Findings

1. •We searched for studies that used CWI (water temperature <20°C), the gold standard treatment for exertional heatstroke,¹ to treat hyperthermic males and females (eg, T_rec ≥39.5°C).
2. •Nine studies met the search criteria; however, 6 studies were excluded because data were not separated by gender. Three studies^2–4 met our predetermined inclusion and exclusion criteria for this critically appraised topic (CAT) study.
3. •Two studies^2,3 observed clinically meaningful differences, with females having higher T_rec cooling rates. The third study⁴ observed no differences in T_rec cooling rates between males and females (Table 1). The average effect size for all studies analyzed in this CAT^2–4 was 2.4 (ranged from 0 to 3.9).
4. •The average T_rec cooling rates across all examined studies^2–4 were 0.18 (0.05) and 0.24 (0.03)°C/min for males and females, respectively. Although hyperthermic females cooled ∼33% faster than males, both genders cooled faster than recommendations for exertional heat stroke (EHS) victims (ie, 0.15°C/min).⁵

Inclusion

1. •Following exercise, subjects were hyperthermic with a T_rec ≥ 39.5°C (103.1°F).
2. •Studies used CWI (≤20°C [68°F]) to treat hyperthermia.
3. •Body core temperature was measured with T_rec since this is the only valid, clinically useful body temperature site for an exercising human.⁶
4. •Limited to studies published in English in the last 15 years (1993–2018).
5. •Limited to studies classified as Oxford Centre for Evidence-Based Medicine level 3 or higher.

Exclusion

1. •Studies that did not test both male and female subjects or differentiate cooling rates by gender.
2. •Studies that measured body core temperature by oral, axillary, esophageal, tympanic, or temporal thermometry.
3. •Studies that cooled hyperthermic subjects by other means (eg, cooling vests, passive cooling, ice towels).
4. •Studies that used percooling (ie, cooling during exercise) or precooling (ie, cooling before exercise).

Results of Search

Two retrospective cohort studies^3,4 on EHS victims and 1 quasi-experimental laboratory study² met the predetermined inclusion and exclusion criteria. Each study was reviewed and independently categorized using the 2011 Oxford Centre for Evidence-Based Medicine (Table 2).

Best Evidence

The studies in Table 2 were identified as the best evidence and selected for inclusion in this CAT. These studies were included because they were considered level 3 evidence or higher and included data to allow comparison of gender differences in CWI T_rec cooling rates.

Implications for Practice, Education, and Future Research

Two main clinical observations can be derived from this CAT. First, hyperthermic male and female athletes do not have similar T_rec cooling rates. Hyperthermic females included in this CAT cooled ∼33% faster than males. Males required CWI for 4.3 (3.0) minutes longer than females. Second, T_rec cooling rates for men and women were either acceptable (0.08–0.15°C/min) or ideal (>0.15°C/min).⁵ All 3 studies^2–4 observed ideal cooling rates for females, whereas only 2 studies^3,4 observed ideal cooling rates in males. In 1 study,² males cooled at only acceptable T_rec cooling rates. These results suggest that no additional precautions need to be taken based on gender; however, clinicians should be aware that males may cool slightly slower than females when being treated with CWI.

The primary reason for differences in T_rec cooling rates between genders is likely related to body surface area (BSA), lean body mass (LBM), and body surface area-to-lean body mass (BSA/LBM) ratios. Women, typically, have higher BSA than males.⁷ Thus, females may have a greater capacity to dissipate heat by nonevaporative mechanisms during CWI. Conversely, males are typically heavier and have more LBM than females,^2,7 which would tend to favor higher metabolic activity and slower cooling rates during CWI. Collectively, this means BSA/LBM ratios tend to be higher in women than men. Several investigators^2,8 have noted relationships between the BSA/LBM ratio and T_rec cooling rates of hyperthermic subjects. Friesen et al⁸ observed male subjects with a high BSA/LBM cooled 1.7 times faster than males with a low BSA/LBM. Similarly, Lemire et al² hypothesized that females cooled faster than males because of differences in BSA/LBM ratios, since the body surface area-to-mass (BSA/M) ratio and BSA were matched between genders. In fact, BSA/LBM was highly correlated with the CWI T_rec cooling rate (r = .7) when gender data was pooled.² Future research should confirm Friesen et al’s⁸ observations in female subjects to provide greater evidence for the contribution of the BSA/LBM ratio to T_rec cooling rates.

Two other minor variables may explain why females cooled faster than males. First, females, in general, have higher adiposity than males.^2,7,9 However, simply having more adipose tissue is an unlikely explanation for differences in cooling rate between genders. A difference of 10% adiposity did not affect T_rec cooling rates in hyperthermic men cooled in 8°C water (13% body fat = 0.23 [0.09]°C/min; 23% body fat = 0.20 [0.09]°C/min).¹⁰ However, individuals with more fat mass tend to shiver less during cold exposure, which may result in less metabolic activity in the lean tissue and may result in faster cooling rates.² The influence of adiposity on T_rec cooling rates in hyperthermic men and women with >25% body fat has not yet been tested. Second, increases in circulating plasma hormones, especially during the luteal phase of menses, may impact the sensitivity of the shivering response, skin vasoconstrictor response, and/or metabolic activity in women.¹¹ Experiments that attempt to control resting body core temperature typically test females during their follicular phase to decrease variability and these hormonal responses. Thus, T_rec cooling rates of women tested during the follicular phase are likely to be less impacted by menses than those of women tested in the luteal phase.

We acknowledge 2 main limitations of our CAT. First, only one of our studies² reported subjects’ mass, body fat, BSA/M ratio, and BSA/LBM ratio. Therefore, it was not possible to compare the effects of anthropometric factors as they relate to gender between all 3 studies in our CAT. Because only 1 study² controlled for anthropomorphics between genders, and these factors greatly influence T_rec cooling,^8,7 we placed greater weight on the results of this study² for the final conclusion in our CAT. Second, while Armstrong et al³ had 4 females diagnosed with EHS in their study, only 1 was included in the CWI treatment group. Consequently, we were only able to compare the T_rec cooling rates of the lone female subject to the 13 male EHS victims who were also treated with CWI.

In conclusion, hyperthermic males and females (some with EHS) have different CWI T_rec cooling rates. Hyperthermic females cooled faster than males, which is likely attributed to their higher BSA/LBM ratio. However, future research may wish to answer the following questions: (1) Do females with different BSA/LBM ratios cool at different rates like their male counterparts?⁸ (2) When BSA/LBM ratios are matched between men and women, are T_rec cooling rates equivocal? (3) What is the effect of >25% body fat on T_rec cooling rates in men and women? Regardless of differences in T_rec cooling rates between genders, the cooling rates were either acceptable or ideal⁵ in all studies examined. The fact that 100% of the men and women who experienced EHS (n = 295) in both retrospective cohort studies^3,4 survived without any noted complications attests to the life-saving capabilities of immediate CWI. Therefore, we conclude CWI is highly effective, regardless of gender, and must be the clinical gold standard for treating EHS victims. This CAT should be reviewed and revised upon the completion of additional studies as they relate to T_rec cooling rates between genders.