Heart rate variability dynamics during treatment for exertional heat strain when immediate response is not possible
Funding information:
This research was supported by the Natural Research Council of Canada (Discovery grant RGPIN-06313-2014 and Discovery grants Program Accelerator Supplements RGPAS-462252-2014 held by G.P.K.). G.P.K. is supported by a University of Ottawa Research Chair Award. The provision of financial support does not in any way infer or imply endorsement of the research findings by either agency. S.R.N. is supported by a Postdoctoral Fellowship from the Human and Environmental Physiology Research Unit.
Edited by: Michael Tipton
Linked articles: This article is highlighted in Viewpoint article by Casa et al. To read this paper, visit https://doi.org/10.1113/EP087759.
Abstract
New Findings
What is the central question of this study?
Does a delay in cold water immersion treatment affect the cardiac autonomic control of exertionally heat-strained individuals?
What is the main finding and its importance?
Cold water immersion is effective for treating exertionally heat-strained individuals even when treatment is commenced with a significant delay. However, that treatment delay leads to only partial/transient restoration of cardiac autonomic control. Therefore, we recommend that exertional heatstroke patients are continuously monitored for several hours even after core temperature has returned to normal values.
Immediate cold water immersion (CWI) is the gold-standard treatment for exertional heatstroke. In the field, however, treatment is often delayed, primarily owing to a delayed paramedic response and/or inaccurate diagnosis. We examined the effect of treatment (reduction of rectal temperature to 37.5°C) delays of 5 (short), 20 (moderate) and 40 (prolonged) min on cardiac autonomic control [as assessed via heart rate variability (HRV)] in eight exertionally heat-strained (40.0°C rectal temperature) individuals. Eleven HRV indices were computed that have been described commonly in the literature and characterize almost all known domains of the variability and complexity of the cardiopulmonary system. We found that the cardiac autonomic control (as assessed via HRV) of exertionally heat-strained individuals was significantly affected by the amount of time it took for the CWI treatment to be applied. Six out of 11 HRV indices studied, from all variability domains, displayed strong (P ≤ 0.005) time × delay interaction effects. Moreover, the number of significantly (P ≤ 0.005) abnormal (i.e. different from the short delay) HRV indices more than doubled (seven versus 15) from the moderate delay to the prolonged delay. Finally, our results demonstrated that a CWI treatment applied with delays of 20 and, primarily, 40 min did not lead to a full restoration of cardiac autonomic control of exertionally heat-strained individuals. In conclusion, this study supports CWI for treating exertionally heat-strained individuals even when applied with prolonged delay, but it highlights the importance of continued cardiac monitoring of patients who have suffered exertional heatstroke for several hours after restoration of core temperature to normal.
1 INTRODUCTION
Exertional heat stroke (EHS) is the second leading cause of non-traumatic death in competitive athletes (Maron, Doerer, Haas, Tierney, & Mueller, 2009), recently shown to be 10 times more prevalent than adverse cardiac events in a large cohort of 137,580 long-distance race participants (Yankelson et al., 2014). Exertional heat stroke is frequent also among labourers (Flouris et al., 2018a) and warfighters (Armed Forces Health Surveillance Center, 2015), leading to increased morbidity and mortality in these large populations. Understanding and treating this condition is becoming increasingly pertinent given: (i) the growing popularity of prolonged sport events, such as triathlons, marathons and half-marathons, across the world (estimated 19 million participants during 2013 in the USA alone; Running USA, 2014); (ii) the ever-increasing participation of older individuals (who have a limited capacity to dissipate heat during work and leisure activities in hot environments; Flouris et al., 2018b) in such events (Tsoutsoubi, Ioannou, Amorim, Tsianos, & Flouris, 2018); and (iii) the warming climate (Flouris & Kenny, 2017), which predisposes individuals to development of EHS.
Ample field and laboratory evidence has shown that immediate cold water (2–4°C) immersion (CWI) (Casa et al., 2007b; Demartini et al., 2015; Flouris et al., 2015; Proulx et al., 2003) with the aim of reducing the core temperature of the patient to near-normal resting levels is the best treatment for EHS (Casa et al., 2007b; Casa, Armstrong, Kenny, O'Connor, & Huggins, 2012; Flouris et al., 2014a; Flouris, Friesen, Carlson, Casa, & Kenny, 2015). Delay in the application of CWI treatment can lead to morbidity and mortality (Casa et al., 2005; Zeller, Novack, Barski, Jotkowitz, & Almog, 2011), because the length of time that core temperature remains above critical levels is linked to the severity and reversibility of multisystem organ failure in EHS (Casa et al., 2012; Hubbard et al., 1977; Zeller et al., 2011). Field data confirm a greater number of fatalities when CWI treatment is delayed (Casa et al., 2012; Zeller et al., 2011), and increased organ dysfunction, longer hospitalizations and/or a prolonged return-to play/work period in EHS victims who survive despite delays in treatments (Casa et al., 2012; Costrini, Pitt, Gustafson, & Uddin, 1979; O'Connor et al., 2010; Stearns et al., 2011; Zeller et al., 2011).
Given the importance of rapid cooling, all relevant guidelines highlight that the CWI treatment of EHS patients should commence as soon as possible. However, treatment is often delayed in real-life situations by ∼20 min owing to inaccurate diagnosis, lack of recognition of the condition and/or transport to treatment facilities (Casa et al., 2007a, 2012). Longer treatment delays of ∼40 min are also common owing to considerable waiting times for first responders (Carr et al., 2006, 2012; Marom et al., 2011). Victims of EHS who have been treated with delay may experience a longer period of cardiovascular compromise (Casa & Roberts, 2003; Epstein & Roberts, 2011). However, the impact of treatment delay on cardiac autonomic control remains unknown. Another issue that has not been studied relates to the impact on cardiac autonomic control of applying CWI treatment after a prolonged period of very high core temperature. A previous report using exertional heat strain as a model (which is not EHS per se, but can lead to it) showed that CWI treatment does not increase the risk for arrhythmias and can restore disturbances in cardiac autonomic control as effectively as natural recovery (Flouris et al., 2014a). However, the dynamics occurring in cardiac autonomic control when the application of CWI treatment is delayed have not been studied. This is an important issue to address, because EHS patients who are treated with delay often present with complete thermoregulatory system failure (Casa & Roberts, 2003; Epstein & Roberts, 2011). Application of a potent stimulus, such as CWI treatment, to a vulnerable individual who has suffered a prolonged period of neural, cardiac and thermoregulatory dysfunction may lead to cardiac autonomic dysregulation (Shattock & Tipton, 2012).
The aim of the present study was to examine the effect of treatment delays of 5, 20 and 40 min duration on heart rate variability [HRV; the most well-known non-invasive method for the assessment of cardiac autonomic control and for evaluating the alteration of the sympathetic and parasympathetic inputs (Carrillo et al., 2016; Dinas et al., 2011; Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996)] in exertionally heat-strained individuals (i.e. rectal temperature of 40.0°C). Given that the risks for morbidity and mortality associated with EHS rise as a function of treatment delay, we hypothesized that the length of time that body temperature remains above critical values would be a key criterion determining cardiac autonomic stability despite the application of CWI treatment.
2 METHODS
2.1 Ethical approval
The study conformed to the standard set by the Declaration of Helsinki and was approved by the University of Ottawa Health Sciences and Science Research Ethics Board (file no. H01-10-02). The study was not registered in a database. Written informed consent was obtained from all volunteers before their participation in the study. The present data were collected as part of a large study, and the main thermoregulatory data have been published elsewhere (Flouris et al., 2015).
2.2 Participants
The minimal required sample size was determined using G*Power v.3.1.9.2 (Faul et al., 2007) based on the previously reported (Leicht et al., 2009) change observed in low- (3.58 ± 0.65 versus 6.75 ± 0.97 msec) and high- (1.65 ± 0.72 versus 5.23 ± 1.17 msec) frequency power and in Pointcaré SD1 (3.01 ± 0.98 versus 17.12 ± 7.67 sec) and SD1/SD2 (0.07 ± 0.02 versus 0.27 ± 0.06) during CWI treatment applied to 11 exertionally heat-strained (core temperature, 40°C) healthy volunteers. Using these data, effect sizes (d) of 3.69, 3.54, 2.48 and 4.3 for the differences during CWI were expected. Based on the study protocol [three repeated measures × six protocol stages (see below)] and assuming a significance level of 0.005 and 0.995 power, six participants were required to detect between- or within-trial differences of a similar magnitude. In order to detect a reasonable departure from the null hypothesis with confidence, we recruited a total of eight healthy (non-smoking, free of any known cardiovascular, respiratory or metabolic diseases), physically active (exercised minimum of 30 min, three to five times per week) men. Their physical characteristics were as follows (mean ± SD): age, 30 ± 6 years; height, 180 ± 6 cm; mass, 79.6 ± 9.1 kg; maximal oxygen consumption, 59.8 ± 4.2 ml O2 kg−1 min−1; percentage body fat, 13.4 ± 3.0%; and body surface area, 1.99 ± 0.14 m2.
2.3 Experimental protocol
All participants completed one preliminary and three experimental sessions separated by a minimum of 48 h. During the preliminary session informed consent was obtained, followed by measurements of participants’ nude body mass, height, body density and maximal oxygen consumption. Nude body mass was measured using a calibrated digital high-capacity scale (IND560; Mettler Toledo Inc., Mississauga, ON, Canada), and body height was determined using a stadiometer (model 2391; Detecto, Webb City, MO, USA). Body density was measured by hydrostatic weighing and used to calculate lean body mass and percentage body fat with the Siri equation (Siri, 1956). Maximal oxygen consumption was measured by indirect calorimetry (MOXUS system; Applied Electrochemistry, Pittsburgh, PA, USA) during a graded exercise test performed on a treadmill (Woodway Desmo; Woodway USA, Inc., Waukesha, WI, USA) in thermoneutral conditions (22°C, 30% relative humidity). Participants were instructed to refrain from alcohol and the use of non-steroidal anti-inflammatory drugs for 48 h, severe or prolonged exercise for 24 h, caffeine for 12 h and food consumption for 2 h before each session. Water consumption was not restricted before or during each session. Experimental sessions were conducted in a random order at the same time of day for each participant to avoid circadian variations in core temperature.
Upon arrival at the laboratory, participants voided their bladder, inserted a temperature probe in their rectum and weighed themselves nude. Subsequently, they donned standardized athletic clothing (shorts and running shoes) and sat quietly while being instrumented. After instrumentation, participants underwent an experimental protocol consisting of six stages, the first of which was conducted in thermoneutral conditions, whereas the remaining five were conducted in heat stress conditions. During stage 1 (labelled ‘Baseline’ hereafter), participants remained resting for 20 min in thermoneutral conditions (22°C, 30% relative humidity). During stage 2 (labelled ‘Rest’ in the Results section), participants moved into a thermal chamber (Can-Trol Environmental Systems Ltd, Markham, ON, Canada) regulated at 40°C and 20% relative humidity, where they remained seated for an additional 20 min to record pre-exercise resting values. In stage 3 (labelled ‘Exercise’ hereafter), participants donned a nylon rain poncho covering the entire upper body and head to minimize evaporative heat loss and accelerate the heating process. They then ran continuously on a treadmill (Desmo HP; Woodway USA, Inc.) at ∼65% of their predetermined maximal oxygen consumption, until rectal temperature reached 40.0°C. Stage 4 (labelled ‘Delay’ hereafter) commenced after the cessation of exercise. During this stage, the nylon poncho was removed and replaced with a sleeveless nylon running jacket. The participants were then required to sit in an upright seated resting position inside the thermal chamber for a short (5 min), moderate (20 min) or prolonged (40 min) postexercise recovery period to simulate treatment delay. Thereafter, in stage 5 (labelled ‘Immersion’ hereafter), participants donned neoprene boots (DuPont, Wilmington, DE, USA) and entered a circulated (½ HP jet motor) iced (2°C) water bath (110 Gallon Stationary Whirlpool, model S-110-SL; Whitehall Manufacturing, City of Industry, CA, USA) located inside the thermal chamber. During the CWI, participants were immersed to the nipples, while in an upright seated position with the legs extended with both arms out of the bath, until rectal temperature was reduced to 37.5°C. Finally, in stage 6 (labelled ‘Recovery’ hereafter), participants exited the cold water bath; they were towel dried, and they sat upright inside the thermal chamber until rectal temperature was reduced to 36.5°C.
2.4 Core temperature
Oesophageal and rectal temperatures were measured throughout with general purpose thermocouple temperature probes (Mallinckrodt Medical Inc., St Louis, MO, USA) using previous methodology (Flouris et al., 2015) every 15 s using an HP Agilent data acquisition module (model 3497A) and simultaneously displayed and recorded in spreadsheet format with LabVIEW software (v.7.0; National Instruments Corporation, Austin, TX, USA). The average of oesophageal and rectal temperature was used as an index of core temperature.
2.5 Heart rate variability
- Low-frequency power, in the energetic HRV domain, is modulated primarily by the sympathetic nervous system but, depending on the context, can also reflect parasympathetic and/or baroreflex mechanisms (Dinas et al., 2011; Flouris et al., 2014b; Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996).
- High-frequency power, in the energetic HRV domain, is considered as a marker of vagal modulation (Dinas et al., 2011; Flouris et al., 2014b; Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996).
- Low- to high-frequency power ratio, in the energetic HRV domain, is generally considered as a measure of the global sympathovagal balance (Dinas et al., 2011; Flouris et al., 2014b; Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996), yet, given the uncertainty on low-frequency power interpretation, there is no consensus on the physiological meaning of this measure.
- Coefficient of variation, from the statistical HRV domain, which represents a sensitive measure of parasympathetic nervous system activity (Carrillo et al., 2016; Flouris et al., 2014b).
- Poincaré SD1 and SD2, from the geometric HRV domain, which are sensitive to parasympathetic nervous system activity (Carrillo et al., 2016; Flouris et al., 2014b).
- Largest Lyapunov exponent, from the invariant HRV domain, is an important index for detecting and characterizing chaos produced from a dynamic system (Flouris et al., 2014b; Liu et al., 2003).
- Shannon entropy, from the informational HRV domain, is sensitive to parasympathetic nervous system activity (Carrillo et al., 2016; Flouris et al., 2014b).
- Detrended fluctuation analysis α1 and α2 from the invariant HRV domain (Flouris et al., 2014b) are derived from a scaling analysis method to show the correlation properties of a signal and are associated with the risk for fatal cardiovascular events (Francesco et al., 2012; Ho et al., 1997; Huikuri et al., 2000).
- Hjorth's complexity, from the energetic HRV domain, which provides an estimate of spectral spread and, together with other types of complexity, is linked with disease manifestation and cardiovascular pathologies (Chen et al., 2015; Hjorth, 1970; Ho et al., 2011; Tang et al., 2015).
2.6 Statistical analysis
Owing to the multiple data points and varying duration (see previous subsection) of Exercise, Delay, Immersion and Recovery across the experimental sessions (given that the aim was to achieve a certain level of rectal temperature), it was not possible to test our hypothesis using a typical repeated-measures ANOVA. This analysis uses listwise deletion, and if one value is missing, the entire case is ignored. For our analysis, this means that we would ignore all data that did not include the entire group of participants (e.g. we would be able to analyse HRV only until the first participant reached a rectal temperature of 40.0°C during ‘Exercise’, ignoring HRV for all the other participants who required additional time to reach a rectal temperature of 40.0°C). Moreover, it was not possible to use a linear mixed model, because this analysis assumes that each participant has only one value at a given level or combination of levels of variables specified as repeated (in this case, only one HRV window per stage per session would be considered in the analysis).
In this light, we used a mixed-design ANOVA (also known as a split-plot ANOVA), with one fixed-effects factor termed ‘delay’ [representing the three different trials incorporating a ‘short’ (5 min), ‘moderate’ (20 min) or ‘prolonged’ (40 min) postexercise recovery period to simulate treatment delay] and one random-effects factor termed ‘stage’ (representing the six different stages of the protocol). To eliminate any bias caused by violations in the sphericity of the covariance matrix assumption, the results for main and interaction effects were adjusted to the degrees of freedom using the Greenhouse–Geisser index. To ensure that using a mixed-design ANOVA did not increase our probability of making a type I error, especially given the potential dependency in the fixed-effects factor of our analysis, we set a conservative statistical significance level at 0.005. This P value threshold also coincides with recent efforts to improve the reproducibility of scientific research and leads to reporting effects that are large enough for outcomes that are serious enough to make them worthy of further action, promoting better, more durable solutions (Benjamin et al., 2018). Moreover, to maximize the validity of our results, we used a Bonferroni correction in our post hoc tests and maintained the significance level at 0.005 also for those tests. All analyses were performed using the statistical software package SPSS v.20 for Windows (SPSS Inc., Chicago, IL, USA).
3 RESULTS
Table 1 presents the statistically significant (at the 0.005 level) main and interaction effects from the mixed-design ANOVA assessing the impact of treatment delay and protocol stage on core temperature, heart rate and indices of heart rate variability. As anticipated, the changes observed in core temperature (Fig. 1; which was, to a large extent, experimentally manipulated by the experimental protocol) and heart rate (Fig. 2) across the different protocol stages were affected by the treatment delay (interaction effects P ≤ 0.005). In turn, this was reflected in six out of 11 HRV indices studied from all variability domains that displayed strong stage × delay interaction effects (Table 1). Post hoc Student's paired t tests revealed that the number of significantly (P ≤ 0.005) abnormal (i.e. different from the short delay) HRV values more than doubled (seven versus 15) from the moderate delay to the prolonged delay (Figs. 2-4). Specifically, statistically significant (P ≤ 0.005) suppression of detrended fluctuation analysis α1 and Hjorth's complexity (4) were detected during the prolonged delay. During the subsequent Immersion stage, the coefficient of variation and Poincaré SD1 were significantly (P ≤ 0.005) increased, and a non-significant similar trend was observed in Poincaré SD2 (Fig. 3) during the prolonged delay. Finally, during the Recovery stage of the prolonged delay, the coefficient of variation, Poincaré SD1, Poincaré SD2 (Fig. 3), Shannon entropy and detrended fluctuation analysis α1 (Fig. 4) were significantly suppressed, whereas Hjorth's complexity was significantly increased (P ≤ 0.005).
| Main effect of delay | Interaction effect of delay × stage | |||
|---|---|---|---|---|
| Fd.f., P value | Partial η2/d | Fd.f., P value | Partial η2/d | |
| Core temperature | 7.9341.9, 0.001 | 0.181/0.941 | 4.3949.5, <0.001 | 0.379/1.562 |
| Heart rate | 6.2801.9, 0.002 | 0.024/0.314 | 5.0459.7, <0.001 | 0.089/0.625 |
| Low-frequency power | — | — | — | — |
| High-frequency power | 6.2752.0, 0.002 | 0.024/0.314 | — | — |
| Low-/high-frequency power ratio | 11.0701.8, <0.001 | 0.041/0.414 | — | — |
| Coefficient of variation | — | — | 5.5619.5, <0.001 | 0.098/0.659 |
| Poincaré SD1 | — | — | 5.6789.4, <0.001 | 0.099/0.663 |
| Poincaré SD2 | — | — | 5.8179.5, <0.001 | 0.102/0.674 |
| Largest Lyapunov exponent | — | — | — | — |
| Shannon entropy | — | — | 6.3019.7, <0.001 | 0.109/0.700 |
| DFA-α1 | 10.8692.0, <0.001 | 0.041/0.414 | 9.5139.9, <0.001 | 0.156/0.860 |
| DFA-α2 | — | — | — | — |
| Hjorth's complexity | 15.4181.9, <0.001 | 0.057/0.492 | 10.2379.4, <0.001 | 0.166/0.892 |
- Abbreviations: d, Cohen's effect size; DFA- α1/α2, detrended fluctuation analysis α1 and α2.
4 DISCUSSION
In real-life situations, delays in treatment of EHS victims are not uncommon. This can be the result of a failure to recognize the presence or seriousness of EHS and/or the absence of medical personnel to care for the victim or on-site equipment (Casa et al., 2012). Even if EHS is recognized promptly at the time of the incident, an individual can still suffer significant morbidity or death if cooling treatment is not applied (Armstrong et al., 2007; Casa et al., 2010; Hubbard et al., 1977; Zeller et al., 2011). The main findings of the present study were as follows: (i) the cardiac autonomic control (as assessed via HRV) of exertionally heat-strained individuals is significantly affected by the length of time it takes for the CWI treatment to be applied; and (ii) a CWI treatment applied with delays of 20 and, primarily, 40 min does not lead to a full restoration of the cardiac autonomic control of exertionally heat-strained individuals.
Our results suggest that individuals who have reached a high level of exertional heat strain (40.0°C rectal temperature, which is not EHS, but can lead to it) demonstrate a significant suppression of detrended fluctuation analysis α1 and Hjorth's complexity if not treated with CWI within 20 min. Disturbances in detrended fluctuation analysis α1 are strongly linked to the risk for fatal cardiovascular events (Ho et al., 1997; Huikuri et al., 2000) and myocardial ischaemia (Francesco et al., 2012) and may reflect the rapid decrease in arterial blood pressure experienced by EHS victims who have been treated with delay (Casa & Roberts, 2003; Epstein & Roberts, 2011). The suppression of Hjorth's complexity shows that heart rate is less regular and predictable and also points in the same direction, because it is linked to serious cardiovascular events, such as sudden cardiac death (Fujita et al., 2016), congestive heart failure (Ho et al., 2011) and stroke (Chen et al., 2015; Tang et al., 2015). Taken together, these results suggest that athletes, labourers, warfighters and other individuals who have reached a high level (40.0°C rectal temperature) of exertional heat strain might benefit from CWI to restore cardiac function.
In a previous study, we found that CWI treatment applied without delay to exertionally heat-strained individuals does not lead to cardiac autonomic imbalances, as assessed via HRV (Flouris et al., 2014a). In the present study, we found that the coefficient of variation and Poincaré SD1, both of which are influenced by vagal activity, were significantly increased when the CWI was applied with a prolonged delay of 40 min. The non-significant trends observed in high frequency power, low-/high-frequency power ratio and Poincaré SD2 also point in the same direction (Table 1). Overall, our HRV results show that CWI leads to parasympathetic nervous system reactivation, and this response was even more evident when the CWI was applied with a prolonged delay of 40 min. The parasympathetic branch of the autonomic nervous system is responsible for cardiovascular stability and restoration of homeostasis (Dinas et al., 2011). Accordingly, reduced vagal influence on cardiac autonomic control after exercise has been linked to ischaemic heart disease and can lead to malignant ventricular arrhythmias and sudden cardiac death (Billman, 2002). Therefore, our results suggest that CWI transiently improves the cardioprotective background even when applied with a delay of 40 min. Taken together with our previous finding that a 40 min delay in application of CWI to treat EHS does not affect the core cooling rate or the post-immersion core temperature after-drop (Flouris et al., 2015), our results are compatible with the conclusion that CWI is the most appropriate treatment for EHS.
Although our study is the first to show that CWI can be used to improve the cardioprotective background even when applied with a delay of 40 min, our findings during the Recovery stage of the protocol suggest that this effect is only transient. Hjorth's complexity was increased whereas Shannon entropy was decreased during Recovery, both suggesting that heart rate was more regular and predictable. However, a number of HRV indices linked to baroreflex sensitivity [Poincaré SD1, Poincaré SD2 and detrended fluctuation analysis α1 (Shaffer & Ginsberg, 2017)] were significantly suppressed during the Recovery stage when the CWI was applied with a prolonged delay of 40 min. Importantly, this was despite the fact that core temperature had returned to normal resting levels. Attenuated baroreflex sensitivity is closely linked with worsened disease state (Eckberg et al., 1971), cardiac mortality and sudden cardiac death in post-myocardial infarction and heart failure patients (La Rovere et al., 2001; La Rovere, Bigger, Marcus, Mortara, & Schwartz, 1998; Mortara et al., 1997), and with hypertension, diabetes, coronary artery disease, myocardial infarction and heart failure (La Rovere et al., 2008). It is important to note, however, that the impact of EHS and/or CWI treatment delay on baroreflex sensitivity should be confirmed through measurement of beat-by-beat fluctuations of arterial blood pressure and heart rate. This is the first study to reveal these findings, which might explain, at least in part, the cascade of events leading to death during EHS. That is, the present results suggest that the length of time that body temperature remains above critical values is a key criterion determining cardiac autonomic stability despite the application of CWI treatment. In turn, this has vast implications for morbidity and mortality, as shown in field studies reporting significantly lower survival of EHS victims who were treated with delay (Armstrong et al., 2007; Casa et al., 2010; Hubbard et al., 1977). In fact, studies show that the survival rate is negatively impacted when the duration of severe hyperthermia extends beyond 15–30 min (Armstrong et al., 2007; Casa et al., 2007b, 2010; Heled, Rav-Acha, Shani, Epstein, & Moran, 2004; Hubbard et al., 1977). Our HRV data show that a patient who has suffered a prolonged period of neural, cardiac and thermoregulatory dysfunction owing to treatment delay may still experience cardiac autonomic dysregulation. In this light, although our data support the use of CWI for treating EHS even when treatment is commenced with a significant delay, we highlight the importance of continuous cardiac monitoring of the patient for several hours, even after core temperature has returned to normal values.
In conclusion, the present study supports CWI as a treatment for exertionally heat-strained individuals even when applied with prolonged delay, but also provides evidence that the restoration of cardiac autonomic control is not complete after core temperature has returned to normal or near-normal values. Therefore, a strong recommendation of the present study is to continue cardiac monitoring of patients who have suffered EHS for several hours after core temperature has returned to a normal or near-normal level. Future studies should investigate the time required for full cardiac recovery after EHS or high exertional heat strain. Predicting a patient's risk of developing major and life-threatening complications is a key interest for physicians and researchers. In this light, the HRV indices used in the present study are highly researched biomarkers of prognostic value that might contribute to anticipation of a patient's clinical course and become a fundamental step towards better care.
ACKNOWLEDGEMENTS
The authors thank all the participants who volunteered for this study, and Mr Marc Carlson and Mr Mike Carter for their technical assistance.
COMPETING INTERESTS
All authors, except A.J.E.S. and C.L.H., have no competing interests to declare. A.J.E.S. is a patent holder, Director and shareholder of Therapeutic Monitoring Systems (TMS) Inc., focused on commercialization of variability-derived clinical decision support tools developed in Ottawa Hospital Research Institute (OHRI) Dynamical Analysis Laboratory. C.L.H. is a patent holder related to variability monitoring and physiological waveform analysis.
AUTHOR CONTRIBUTIONS
All measurements were performed in the Human Environmental Physiological Research Unit, University of Ottawa, ON, Canada. G.P.K. and A.D.F. conceptualized and designed the research. B.J.F. and G.P.K. performed experiments. A.D.F. and C.L.H. analysed data. A.D.F. prepared figures and drafted the manuscript. All authors interpreted experimental results and edited and revised the manuscript. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.
