Healthy aging does not compromise the augmentation of cardiac function during heat stress

Previous studies suggest that healthy aged humans are unable to maintain stroke volume during heat stress. We hypothesized that this is due to an inappropriate augmentation of cardiac function during heat stress. Contrary to our hypothesis, stroke volume was maintained during heat stress in healthy aged adults and was accompanied by a leftward shift of the Frank-Starling relation. These results suggest that cardiac function is appropriately augmented during heat stress in healthy aged humans.

Abstract

During heat stress, stroke volume is maintained in young adults despite reductions in cardiac filling pressures. This is achieved by a general augmentation of cardiac function, highlighted by a left and upward shift of the Frank-Starling relation. In contrast, healthy aged adults are unable to maintain stroke volume during heat stress. We hypothesized that this would be associated with a lack of shift in the Frank-Starling relation. Frank-Starling relations were examined in 11 aged [69 ± 4 (SD) yr, 4 men/7 women] and 12 young (26 ± 5 yr, 6 men/6 women) adults during normothermic and heat stress (1.5°C increase in core temperature) conditions. During heat stress, increases in cardiac output were attenuated in aged adults (+2.5 ± 0.3 (95% CI) vs. young: +4.5 ± 0.5 l/min, P < 0.01) because of an attenuated chronotropic response (+30 ± 4 vs. young: +42 ± 5 beats/min, P < 0.01). In contrast to our hypothesis, a leftward shift of the Frank-Starling relation maintained stroke volume during heat stress in aged adults (76 ± 8 vs. normothermic: 74 ± 8 ml, P = 0.38) despite reductions in cardiac filling pressure (6.6 ± 1.0 vs. normothermic: 8.9 ± 1.1 mmHg, P < 0.01). In a subset of participants, volume loading was used to return cardiac filling pressure during heat stress to normothermic values, which resulted in a greater stroke volume for a given cardiac filling pressure in both groups. These results demonstrate that the Frank-Starling relation shifts during heat stress in healthy young and aged adults, thereby preserving stroke volume despite reductions in cardiac filling pressures.

NEW & NOTEWORTHY

Previous studies suggest that healthy aged humans are unable to maintain stroke volume during heat stress. We hypothesized that this is due to an inappropriate augmentation of cardiac function during heat stress. Contrary to our hypothesis, stroke volume was maintained during heat stress in healthy aged adults and was accompanied by a leftward shift of the Frank-Starling relation. These results suggest that cardiac function is appropriately augmented during heat stress in healthy aged humans.

health agencies worldwide recognize the impending negative impact of rising global temperatures on human health and well-being (5, 15, 23a). Prolonged periods of elevated ambient temperatures, such as heat waves, are consistently associated with greater hospital admissions and mortality (11, 30, 34). A predicted rise in the number, duration, and intensity of heat waves (24, 29) will particularly affect the elderly, as they represent the population most vulnerable to heat-related morbidity and mortality (6). Although the underlying pathophysiology predisposing the elderly to a greater risk of heat-related morbidity and mortality remains unclear, impaired cardiovascular adjustments during heat stress are considered a primary contributing factor (19).

During heat stress, cutaneous vasodilation as well as redistribution of blood flow and volume from central, renal, and splanchnic circulations (7) increase skin blood flow to promote heat exchange with the environment. These adjustments place significant stress on the cardiovascular system, as large decreases in peripheral vascular resistance require adequate increases in cardiac output for blood pressure to be maintained. Notably, cardiac output can exceed 12 l/min in young adults exposed to the heat at rest (28). Such increases are mediated through an increase in heart rate, as stroke volume remains near normothermic values or increases slightly despite reductions in cardiac filling pressures (28). These responses are achieved through a general augmentation of cardiac function during heat stress, including enhanced systolic and diastolic functions (2, 3, 27, 32) and greater ejection fraction (7, 35). The net effect of these adjustments is characterized by a leftward and upward shift of the Frank-Starling relation during heat stress (4, 35).

Compared with young adults, increases in cardiac output during heat stress are blunted by ∼50% in healthy aged humans (25, 26). The precise mechanism(s) mediating this attenuated increase remains unclear. During heat stress, aged adults are capable of augmenting Doppler-derived indexes of systolic and diastolic functions (23) and have similar increases in muscle sympathetic outflow compared with young adults (12). Nonetheless, aged adults are unable to maintain stroke volume during heat exposure (25, 26). Importantly, this inability to maintain stroke volume fully accounts for the attenuated increase in cardiac output, as heart rate responses are generally—though not always (12, 14, 21, 22)—similar to the young (23, 25, 26). These results suggest that although Doppler-derived indexes of systolic function are augmented during heat stress in aged adults (23), cardiac contractility may not increase sufficiently to maintain stroke volume during heat-induced reductions in cardiac filling pressures. For example, reductions in cardiac β-adrenergic receptor sensitivity with aging (9, 10) may limit increases in inotropy during heat stress despite seemingly appropriate sympathetic outflow (12). Aged adults may therefore display compromised cardiac function that limits their ability to meet the cardiovascular demands imposed by heat stress.

In humans, integrated cardiac function can be assessed by examining the Frank-Starling relation (1, 20, 35). During heat stress, an inability of aged adults to maintain stroke volume—in parallel with a reduction in cardiac filling pressure—should be reflected in a lack of (or attenuated) shift of the Frank-Starling relation from normothermic conditions. The purpose of this study was to test the hypothesis that the Frank-Starling relation does not shift leftward during heat stress in healthy aged adults, thereby providing a mechanistic rationale for previously reported reductions in stroke volume during heat stress within this population.

MATERIALS AND METHODS

Participants.

Twelve healthy young and 11 aged adults (both men and women) participated in this study. Participant characteristics are presented in Table 1. All participants were nonsmokers, were free of known cardiovascular, respiratory, neurological, or metabolic diseases, and were not taking any related medications. Health status was determined by having the participants fill out a detailed medical history questionnaire and measuring their resting blood pressure and heart rate and performing a 12-lead electrocardiogram during a preliminary visit. Participants volunteered for one study visit that was performed at the same time of day for all participants. Participants were asked to refrain from strenuous physical activity for 24 h, as well as from caffeine and alcohol for 12 h prior to the study visit. The study and informed consent documents were approved by the Institutional Review Boards at the University of Texas Southwestern Medical Center and at Texas Health Presbyterian Hospital Dallas. Written informed consent was provided by all participants prior to their participation in the study.

Table 1.

Participant characteristics

	Young (n = 12)	Aged (n = 11)
Age, yr	26 ± 5 [20–36]	69 ± 4* [61–75]
Height, cm	165 ± 6 [158–177]	165 ± 8 [152–178]
Weight, kg	63.8 ± 9.7 [45.6–81.3]	68.5 ± 8.2 [54.3–78.6]
Body surface area, m2	1.70 ± 0.13 [1.45–1.99]	1.76 ± 0.14 [1.60–1.96]
Systolic blood pressure, mmHg	116 ± 11 [95–132]	130 ± 5*[116–138]
Diastolic blood pressure, mmHg	69 ± 7 [58–82]	77 ± 6*[64–84]
Heart rate, beats/min	65 ± 9 [48–82]	68 ± 8 [58–83]

Values are means ± SD [range].

^*P < 0.05 relative to young.

Measurements.

A 6F balloon-tipped fluid-filled catheter (Swan-Ganz; Edwards Life Sciences) was placed under fluoroscopic guidance through an antecubital vein into a branch of the pulmonary artery. The catheter allowed for the measurement of cardiac pressures, cardiac output by thermodilution, and pulmonary artery blood temperature. Cardiac pressures were referenced to atmospheric pressure with a pressure transducer (Transpac IV; ICU Medical) zeroed at 5 cm below the sternal angle. To determine pulmonary capillary wedge pressure (PCWP), the catheter balloon was inflated to obtain a successful wedge (determined visually) following which PCWP was determined at end-expiration. Cardiac output was measured by thermodilution (Tran 451N; GE Healthcare) upon injection of 10-ml ice-cooled saline. Pulmonary artery blood temperature was used as the measure of core temperature. Heart rate was obtained from an electrocardiogram that was interfaced with a cardiotachometer (CWE). Cardiac pressures, cardiac output, pulmonary artery blood temperature, and the electrocardiogram were displayed and continuously monitored on a patient monitor (Solar 8000i; GE Healthcare). Blood pressure was measured by automated auscultation of the brachial artery (Tango+; SunTech Medical). To manipulate core and skin temperatures, participants were dressed in a water-perfused tube-lined suit (Med-Eng) that covered the entire body except for the head, hands, feet, and forearms. Skin temperature was measured as a weighted average of six thermocouples attached to the skin surface (12).

Protocol.

After right heart catheterization, participants assumed the supine position on a patient bed with their lower body sealed at the waist within a custom-made lower body negative pressure (LBNP) chamber. For normothermic measurements, water maintained at 34°C circulated through the tube-lined suit. Baseline normothermic measurements were obtained after a minimum of 10 min of quiet rest. Thereafter, negative pressure was applied within the LBNP chamber to reduce cardiac filling pressures. Measurements were repeated during LBNP levels of 15 and 30 mmHg, following which negative pressure was stopped and baseline normothermic measurements were repeated after a 10-min recovery period. The temperature of the water circulating through the tube-lined suit was then increased to 50°C to increase skin and core temperatures. At an increase in core temperature of 1.5°C above normothermic values (∼60–90 min), baseline hyperthermic measurements were obtained, following which cardiac filling pressures were again reduced by applying 15 and 30 mmHg of LBNP. Measurements were obtained at each level of LBNP, following which negative pressure was removed and baseline hyperthermic measurements were repeated after 10 min of recovery. In a subset of young and aged adults (n = 4 per group), PCWP was returned to near pre-heat stress levels by rapid volume infusion (0.9% NaCl warm saline), at which point measurements were again obtained.

Data analysis.

Data were collected with acquisition hardware (MP150; Biopac) at a sampling frequency of 50 Hz. End-expiratory PCWP was determined as an average of 2–3 measurements per time point. Cardiac output was determined as the average of 2–4 measurements per time point. Stroke volume was calculated as cardiac output ÷ heart rate. Mean arterial pressure was calculated as diastolic blood pressure + 1/3 of pulse pressure. Peripheral vascular resistance was calculated as (mean arterial pressure − central venous pressure) ÷ cardiac output. Effective arterial elastance (E_a) was calculated as an index of cardiac afterload using the equation E_a = P_es ÷ SV, where P_es is end-systolic pressure and SV is stroke volume (18). End-systolic pressure was calculated as P_es = (2 × P_sys + P_dia) ÷ 3, where P_sys and P_dia are systolic and diastolic blood pressure, respectively (18). Baseline normothermic and hyperthermic measurements represent the average of the values measured before and after the LBNP periods. Frank-Starling relations were examined by plotting stroke volume as a function of PCWP.

Statistical analysis.

Group differences during LBNP were analyzed separately for the normothermic and hyperthermic conditions using a two-way mixed model analysis of variance (ANOVA) with the nonrepeated factor of age (young and aged) and the repeated factor of LBNP level (baseline, 15 mmHg, and 30 mmHg). Similarly, group differences during the heating period were analyzed using a two-way mixed model ANOVA with the nonrepeated factor of age (young and aged) and the repeated factor of core temperature (baseline, Δ0.6°C, Δ1.2°C, and Δ1.5°C). The level of significance was set at an alpha of P < 0.05, and a Holm-Sidak correction was applied for multiple comparisons. Single time point comparisons were analyzed using independent samples t-tests. Statistical analyses were performed using commercially available statistical software (Prism 6; GraphPad Software). All variables are reported as means ± 95% confidence intervals.

RESULTS

Normothermic baseline and reductions in cardiac filling pressures.

Normothermic baseline and LBNP data are presented in Table 2. At baseline, heart rate (P = 0.24), stroke volume (P = 0.07), and cardiac output (P = 0.19) were similar between groups. In contrast, mean arterial pressure and peripheral vascular resistance were greater in aged adults (both P < 0.01). Normothermic PCWP was greater in young adults (P = 0.03). During LBNP, heart rate and peripheral vascular resistance increased while cardiac output, stroke volume, PCWP, and mean arterial pressure decreased in both groups (all P < 0.01). Pulmonary capillary wedge pressure decreased to a lesser extent during LBNP in aged adults (age × LBNP interaction, P < 0.01).

Table 2.

Cardiovascular responses during normothermic LBNP

	Young			Aged
	Baseline	LBNP 15	LBNP 30	Baseline	LBNP 15	LBNP 30
Cardiac output, l/min	5.0 ± 0.4	4.4 ± 0.2	4.0 ± 0.2	4.6 ± 0.4	4.0 ± 0.4	3.6 ± 0.3
Heart rate, beats/min	58 ± 6	59 ± 6	68 ± 10	63 ± 6	64 ± 6	69 ± 6
Stroke volume, ml	89 ± 14	77 ± 10	63 ± 9	74 ± 8	63 ± 6	53 ± 5
MAP, mmHg	84 ± 3	81 ± 3	80 ± 4	96 ± 6*	94 ± 5*	93 ± 6*
Vascular resistance, dyn·s·cm−5	1,269 ± 127	1,421 ± 102	1,572 ± 108	1,628 ± 211*	1,859 ± 218*	2,051 ± 215*
PCWP, mmHg	10.6 ± 1.0	6.3 ± 1.3	3.4 ± 1.1	8.9 ± 1.1*	5.5 ± 1.0	3.8 ± 0.7

Values are means ± 95% confidence intervals. LBNP 15 and 30, lower body negative pressure levels of 15 and 30 mmHg; MAP, mean arterial pressure; PCWP, pulmonary capillary wedge pressure;

^*P < 0.05 relative to young at the indicated time point.

Whole-body heating.

Normothermic core temperature was slightly greater in young (36.7 ± 0.2°C) relative to aged (36.5 ± 0.1°C, P = 0.06) adults, and it remained greater during heat stress (at Δ1.5°C, young: 38.2 ± 0.2°C vs. aged: 38.0 ± 0.1°C, P = 0.04). However, core temperature increased by 1.5 ± 0.0°C in both groups (P = 0.17). Mean skin temperature was similar between young (34.8 ± 0.3 to 39.1 ± 0.2°C) and aged (34.7 ± 0.2 to 39.1 ± 0.3°C) adults throughout heating (P = 0.61). Increases in cardiac output during heating were attenuated in the aged group (P < 0.01, Fig. 1). Absolute values of heart rate were similar between groups during heating (P = 0.46, Fig. 1), although the increase from baseline was greater in young adults (+42 ± 5 vs. aged: +30 ± 4 beats/min, P < 0.01, Fig. 1). A significant age × core temperature interaction was observed for stroke volume (P = 0.02), as it slightly increased during heat stress in young adults but remained unchanged (relative to normothermia) in aged adults (Fig. 1). Nonetheless, the magnitude of the change in stroke volume from baseline to the end of heat stress did not differ between groups (P = 0.26, Fig. 1). Mean arterial pressure (Fig. 2), peripheral vascular resistance (Fig. 2), and arterial elastance (Fig. 3) decreased similarly in both groups during heating (age × core temperature interactions, P > 0.10), although absolute values remained greater in aged adults. A decrease in PCWP during heating was also observed in both groups (Fig. 3), with the decrease being greater in young adults (−4 ± 1 vs. aged: −2 ± 1 mmHg, P < 0.01).

Fig. 1.

Cardiac output (A), heart rate (C), and stroke volume (E) of young and aged adults exposed to whole-body passive heat stress, as well as individual changes from baseline at a 1.5°C increase in core temperature (B, D, and F). Values in A, C, and E and lines in B, D, and F are means ± 95% confidence intervals. †Significant main effect of core temperature. ‡Significant age × core temperature interaction. *P < 0.05 between groups.

Fig. 2.

Mean arterial pressure (A) and peripheral vascular resistance (B) of young and aged adults exposed to whole-body passive heat stress. Values are means ± 95% confidence intervals. †Main effect of core temperature. *P < 0.05 between groups.

Fig. 3.

Pulmonary capillary wedge pressure (PCWP, A) and arterial elastance (B) of young and aged adults exposed to whole-body passive heat stress. Values are means ± 95% confidence intervals. †Main effect of core temperature. ‡Significant age × core temperature interaction. *P < 0.05 between groups.

Hyperthermic reductions in cardiac filling pressures.

Hyperthermic LBNP data are presented in Table 3. Heart rate and peripheral vascular resistance increased while cardiac output, stroke volume, PCWP, and mean arterial pressure decreased in both groups during hyperthermic LBNP (P < 0.01). Age × LBNP interactions revealed that heart rate increased to a greater extent in young adults (P < 0.01), likely to compensate for a greater reduction in stroke volume (P < 0.01). The net result was a similar decrease in cardiac output between groups (P = 0.09). Mean arterial pressure and PCWP also decreased to a similar extent between groups (age × LBNP interactions, P > 0.58).

Table 3.

Cardiovascular responses during hyperthermic LBNP

	Young			Aged
	Baseline	LBNP 15	LBNP 30	Baseline	LBNP 15	LBNP 30
Cardiac output, l/min	9.4 ± 0.6	7.3 ± 0.6	6.1 ± 0.7	7.0 ± 0.5*	5.7 ± 0.5*	4.4 ± 0.4*
Heart rate, beats/min	100 ± 8	111 ± 9	123 ± 10	93 ± 7	96 ± 10	102 ± 12*
Stroke volume, ml	97 ± 12	68 ± 9	51 ± 8	76 ± 8*	60 ± 5	44 ± 6
MAP, mmHg	79 ± 3	72 ± 3	65 ± 5	89 ± 4*	81 ± 5*	74 ± 6*
Vascular resistance, dyn·s·cm−5	646 ± 52	774 ± 44	837 ± 54	985 ± 84*	1,121 ± 126*	1,344 ± 205*
PCWP, mmHg	6.7 ± 1.2	4.3 ± 1.2	2.8 ± 1.3	6.6 ± 1.0	4.5 ± 0.9	3.2 ± 0.8

Values are means ± 95% confidence intervals. LBNP 15 and 30, lower body negative pressure levels of 15 and 30 mmHg; MAP, mean arterial pressure; PCWP, pulmonary capillary wedge pressure;

^*P < 0.05 relative to young at the indicated time point.

Frank-Starling relations.

Frank-Starling relations are presented in Fig. 4. Aging itself (i.e., normothermia) shifted the Frank-Starling curve down and to the right because of lower absolute stroke volumes. Heat stress itself (i.e., prior to LBNP) shifted the Frank-Starling curve to the left in both groups resulting in a preservation of stroke volume despite reductions in PCWP. To examine if heat stress also caused an upward shift of the Frank-Starling relation compared with the normothermic curve, PCWP was returned to near pre-heat stress values by rapid saline infusion in a subset of four young (heat + saline: 11.0 ± 2.7 vs. preheating: 10.5 ± 2.0 mmHg) and four aged (heat + saline: 11.7 ± 4.2 vs. preheating: 9.1 ± 2.0 mmHg) adults. In each participant, a greater stroke volume was observed during heat stress upon restoration of PCWP. These data are illustrated in Fig. 5 and suggest that heat stress also shifted the Frank-Starling relation upward in both age groups.

Fig. 4.

Frank-Starling relations during normothermic and hyperthermic conditions in young (A) and aged (B) adults. The arrows indicate the baseline (i.e., pre-LBNP) values for each condition. Notice the leftward shift in the curve relating stroke volume to pulmonary capillary wedge pressure (PCWP) from normothermic to hyperthermic conditions in both groups such that stroke volume is preserved despite reductions in PCWP (prior to LBNP, indicated by the arrows).

Fig. 5.

Frank-Starling relations during normothermic and hyperthermic (HT) conditions in a subset of young (A, n = 4) and aged (B, n = 4) adults in whom cardiac filling pressures during heat stress were returned to near normothermic values by rapid saline infusion (HT + saline). Note that rapid saline infusion increased stroke volume to a level greater than that observed for a similar pulmonary capillary wedge pressure (PCWP) when normothermic (grey squares).

DISCUSSION

The current study examined the independent effect of human age on the Frank-Starling relation during heat stress. The main finding is that the Frank-Starling relation appropriately shifts during heat stress in healthy aged adults, resulting in a maintained stroke volume despite reductions in cardiac filling pressures. These findings suggest that integrated cardiac function is augmented during heat stress in healthy aged adults. However, increases in cardiac output remain limited because of an attenuated chronotropic response.

In young adults, stroke volume is maintained during heat stress despite reductions in cardiac filling pressures because of an augmentation of systolic and diastolic functions (3, 7, 27, 32). The net integration of these improvements is reflected by a leftward and upward shift of the Frank-Starling relation (4, 35). In contrast, the human cardiovascular adjustments to heat stress have long been considered altered by aging. Seminal studies by Minson and colleagues (25, 26) established that increases in cardiac output are attenuated by ∼50% in healthy aged adults during heat stress, because of an inability to maintain stroke volume as increases in heart rate were similar to those observed in young adults (25). Importantly, reductions in stroke volume in that study occurred only in aged adults despite similar reductions in central venous pressure between groups (25). These results suggested that healthy aged adults have compromised cardiac function during heat stress, resulting in an inability to maintain stroke volume and thus limiting the increase in cardiac output that can be achieved. We therefore hypothesized that the Frank-Starling relation does not shift leftward during heat stress in healthy aged adults, which would have provided a mechanistic rationale for their previously observed inability to maintain stroke volume (25). In contrast to our hypothesis, stroke volume was maintained near normothermic levels during heating (i.e., prior to LBNP) in aged adults, which is reflected by a leftward shift of the Frank-Starling relation (arrows in Fig. 4).

At present, a reason for the different findings between the current study and that of Minson et al. (25) is unclear. Minson et al. (25) heated participants until their limit of thermal tolerance, achieving a slightly greater increase in core temperature (∼1.8°C) than the one achieved in the current study (1.5°C). However, substantial reductions in stroke volume (≥10 ml) occurred at relatively mild (∼0.5°C) increases in core temperature in the study of Minson et al. (25). It is therefore unlikely that differences in the level of heating could explain the different findings. It should also be noted that Minson et al. (25) measured cardiac output using the acetylene-rebreathing technique, whereas we employed thermodilution. Whether such differences in methodology could explain the different findings remains to be determined. A final, important consideration is the fact that only men were included in the study of Minson et al. (25). Interestingly, it has previously been observed that stroke volume is maintained during passive heat stress in postmenopausal women (8), albeit no direct comparisons with a young or a male group were performed. When looking at individual changes in stroke volume from baseline (Fig. 1F), a reduction in stroke volume was observed in three young and five aged adults. Of the three young individuals, one was female whereas two of the five aged participants were female. Therefore the inclusion of female participants cannot explain differences between the current observations and those of Minson et al. (25). Regardless of these discrepancies, maintenance of stroke volume during heat stress in aged adults was also observed in a more recent study (14).

A shift of the Frank-Starling relation during heat stress likely occurs through changes in its parameters. Particularly relevant to heat stress are potential changes in cardiac preload, afterload, and contractility (36). In the current study, heat stress decreased cardiac preload as evidenced by the reductions in central venous and pulmonary capillary wedge pressures. A reduced cardiac preload, coupled with a preservation of stroke volume, is exemplified by the leftward shift of the Frank-Starling relation from normothermic to heat stress conditions (arrows in Fig. 4). Such a response could be mediated through reduced cardiac afterload and/or increased contractility. In the current study, heat stress reduced effective arterial elastance suggesting that cardiac afterload was reduced, thus potentially contributing to the maintenance of stroke volume. However, we believe the maintenance of stroke volume during heat stress is primarily due to an increase in cardiac contractility. A number of studies, some of which employed different methodologies, support the notion that heat stress improves cardiac contractility (2, 4, 7, 27). To examine changes in cardiac contractility with heat stress in the current study, cardiac filling pressures were returned to near normothermic levels in a subset of participants by rapid volume loading. This procedure increased stroke volume in every participant (young and aged), thus resulting in an elevated stroke volume relative to when participants were normothermic (grey squares in Fig. 5). Although obtained from a subgroup of participants, these data suggest that the Frank-Starling relation also shifts upward during heat stress in aged adults (relative to the normothermic curve), as previously demonstrated for young adults (4). If cardiac contractility did not increase with heat stress, we would have expected a similar stroke volume between normothermic and heat stress conditions at a given cardiac filling pressure. Therefore we suggest that cardiac contractility is augmented during heat stress in healthy aged adults, consistent with previously observed improvements in Doppler-derived indexes of systolic function during heating in this age group (23).

Despite evidence for an augmentation of cardiac function during heat stress, absolute values of cardiac output remained lower in aged adults. The lower absolute values are likely related to changes in cardiac function and structure that accompany sedentary aging. For example, sedentary aging (in the absence of heat stress) shifts the Frank-Starling relation downward and to the right because of diminished cardiac ventricular compliance (1). This observation was confirmed in the current study, and despite the leftward shift during heating, the Frank-Starling relation remained attenuated relative to that observed in young adults. Healthy aged adults therefore operate within an attenuated range of stroke volume for given cardiac filling pressures, which apparently limits the absolute values of cardiac output that can be achieved during heat stress.

Consistent with previous observations (14, 25, 26), the increase in cardiac output was attenuated by ∼50% in aged adults. This raises an interesting question: What is limiting the increase in cardiac output in aged adults during heat stress, despite an augmentation of cardiac function? In this study, the attenuated increase in cardiac output was primarily driven by an attenuated increase in heart rate, consistent with similar observations for this age group (12, 14, 21, 22). Increases in heart rate during heat stress occur mainly through changes in cardiac parasympathetic/sympathetic neural activity and to a lesser extent because of the direct effect of temperature on the heart (13, 17). Healthy aged adults have similar increases in muscle sympathetic outflow compared with young adults during heat stress (12), though it remains unknown if cardiac sympathetic activity is likewise similarly increased. Nonetheless, cardiac β-adrenergic receptor sensitivity decreases with human aging (9, 10), and thus it is possible that the attenuated increase in cardiac output during heat stress is due to blunted transduction of cardiac sympathetic outflow. An alternative or complementary possibility is less withdrawal of cardiac vagal tone during heating in aged adults, similar to what is observed during exercise (33). Furthermore, we cannot discount the possibility that aging affects the direct effect of temperature on the heart. It is also important to note that changes in the peripheral signaling pathways for cutaneous vasodilation impair the ability of the skin to vasodilate during heat stress in healthy aged adults (16). Therefore a final possibility is that the attenuated increase in cardiac output parallels lower perfusion of the skin circulation, although recent observations argue against this possibility (14). These possibilities represent interesting avenues for future research.

Perspectives.

Heat-related morbidity and mortality are disproportionately greater in the elderly (30, 31). The precise pathophysiology predisposing the elderly to a greater risk of heat-related morbidity and mortality remains unclear. Human aging per se has long been thought to compromise cardiovascular adjustments during heat exposure (25, 26). In addition, underlying cardiovascular disease is consistently identified as a risk factor for heat-related morbidity and mortality in the elderly (30, 31). For these reasons, altered cardiovascular responses to heat stress are considered a major contributing factor to the increased prevalence of heat-related morbidity and mortality in the elderly (19). The current results suggest that human age does not independently compromise the augmentation of cardiac function that occurs during heat stress. That said, an attenuated chronotropic response limits the increase in cardiac output that can be achieved in healthy aged humans. Future studies should examine the cardiovascular adjustments to heat exposure in clinical populations (e.g., heart failure) and in those with underlying risk factors for cardiovascular disease (e.g., hypertension and diabetes), as well as the effect of related medication use. In such populations, compromised cardiovascular function may limit the ability of the cardiovascular system to meet the demands of heat stress.

Considerations.

The magnitude of the leftward shift of the Frank-Starling relation could not be quantified statistically because of the number of data points obtained (i.e., baseline, LBNP 15 and 30 mmHg). In addition to these time points, stroke volume and cardiac filling pressure data points are needed at two additional levels of volume loading to examine the full range of the Frank-Starling relation (1, 20). However, this approach would require two separate study visits to avoid the confounding influence of volume loading while participants are normothermic on subsequent cardiovascular responses during heat stress. Nonetheless, a leftward shift of the Frank-Starling relation is defined, and clearly observed, by the maintenance of stroke volume despite reductions in cardiac filling pressure, which we statistically quantified in both age groups. Since this represented the main question of the study, we considered it unnecessary to expose healthy participants to the potential risks of a second right heart catheterization or of having the catheter remain in place until a second visit.

Conclusions.

The current study examined changes in the Frank-Starling relation during heat stress as a function of human age. Healthy aged adults display a leftward shift of the Frank-Starling relation during heating, as stroke volume is maintained despite reductions in cardiac filling pressures. These results demonstrate that human aging does not compromise the augmentation of cardiac function that occurs with heat stress. However, an attenuated chronotropic response limits the increase in cardiac output that healthy aged adults can achieve during heat exposure.

GRANTS

This study was supported by National Institute of General Medical Sciences Grant 068865, Department of Defense Grant W81XWH-12-1-0152, and Research Endowment Grant from the American College of Sports Medicine Foundation. D. Gagnon was supported by a Postdoctoral Fellowship from the Natural Sciences and Engineering Research Council of Canada.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

D.G., S.A.R., B.D.L., and C.G.C. conception and design of research; D.G., S.A.R., H.N., S.S., W.K.C. III, P.Y.P., D.S., and C.G.C. performed experiments; D.G. and H.N. analyzed data; D.G., S.A.R., H.N., S.S., W.K.C. III, P.Y.P., D.S., B.D.L., and C.G.C. interpreted results of experiments; D.G. prepared figures; D.G. drafted manuscript; D.G., S.A.R., H.N., S.S., W.K.C. III, P.Y.P., D.S., B.D.L., and C.G.C. edited and revised manuscript; D.G., S.A.R., H.N., S.S., W.K.C. III, P.Y.P., D.S., B.D.L., and C.G.C. approved final version of manuscript.