The present study assessed the age-related differences in RSR distribution in young (18–30 years) and older (60–80 years) individuals during rest and exercise in a hot environment (32 ± 0.1 °C/50.5 ± 0.8% RH). Whole body sweat maps were created for both age groups to illustrate the distribution pattern over the body. The main findings were that, although working at the same fixed rate of heat production (200 W m−2), older individuals had (1) significantly lower GSL than the young group and (2) significantly lower RSR at almost all body regions during rest and at the hands, legs, ankles, and feet during exercise. The lower RSR observed in the leg region of older individuals coincided with a higher Tsk, indicative of lower evaporative cooling. Furthermore, despite a significantly higher increase in Tgi than the young group, older individuals felt slightly cooler throughout both the rest and exercise period (particularly during the pad application), suggesting a reduced sensitivity to temperature increase. These findings add to the body of literature that suggests older individuals are at increased risk of heat-induced illness and injury, compared to their younger counterparts, due to a lower GSL, a disproportional reduction in RSR at the lower extremities, a reduced thermal sensitivity, and impaired defence against a rise in Tgi.
Gross sweat loss
Despite the use of a fixed heat production protocol, substantial inter- and intra-individual variation was observed in GSL in the present study, in agreement with the previous work (Smith and Havenith
2011,
2012). GSL increased significantly from rest to exercise as expected, caused by an increase in metabolic heat production and a proportional increase in sweat loss.
The previous studies have observed a higher GSL in participants with higher
VO
2max values and thus higher fitness levels, when exercising at a fixed load (W) (Havenith et al.
1995) and a %
VO
2max (Havenith et al.
1998; Smith and Havenith
2011,
2012). However in the present study, all participants were exercising at the same fixed heat production (200 W m
−2) to avoid bias when comparing across age groups (Cramer and Jay
2014), and as expected, no relationship was observed between GSL and predicted
VO
2max. This finding supports the conclusions of previous studies which state that GSL does not differ between fit and unfit individuals during exercise at a fixed heat production (Jay et al.
2011; Cramer et al.
2012; Gagnon and Kenny
2012), particularly at rates of 250 W m
−2 or lower (Gagnon and Kenny
2012). The findings are, however, equally consistent with Smith and Havenith’s observations working at equal %
VO
2max (higher GSL with higher fitness), as they showed an excellent correlation between GSL and metabolic rate. Thus, in a scenario where people work at a %
VO
2max (e.g., while running in a race), given the higher metabolic rates in the fitter (faster) people, they are expected to produce more sweat, as indeed observed.
GSL was significantly lower in the older group compared to the young group during both rest and exercise in the present study. This finding is in agreement with the previous literature during passive heating (Inoue et al.
1995,
1998; Inoue and Shibasaki
1996) and exercise (Anderson and Kenney
1987; Kenney and Anderson
1988), supporting the notion that with increasing age comes a decrease in sweat gland function (Inoue and Shibasaki
1996), which will be discussed in detail below. In the present study, the observed age-related difference in GSL, in the absence of a relationship between GSL and predicted
VO
2max, could propose a true age-related change in sweating mechanisms and not fitness level-based differences. However, further research is required to assess GSL during an actual
VO
2max test to confirm this. For the rest period, the observed lower GSL in the older group may be at least partially related to their lower heat production. While resting metabolic rate was not measured in this study, a lower metabolic rate at rest was observed in the older group in the previous work (Coull
2019) with these age groups (older: 43.0 ± 4.6 vs young: 57.1 ± 6.7 W m
−2; supine).
Regional sweat rate
This is the first study to provide RSR data over the whole body surface in both young and older individuals (see Fig.
5). From these data, it is clear that there is RSR variation both within and between age groups and body segments, which is in accordance with the previous literature (Kuno
1956; Inoue et al.
1991; Cotter et al.
1995; Inoue and Shibasaki
1996; Havenith et al.
2008; Machado-Moreira et al.
2008a; Smith and Havenith
2011,
2012). Despite this inter-individual variation in RSR, a number of similar patterns were consistently observed.
As expected, RSR increased in parallel with GSL from rest to exercise at all body regions and the distribution patterns remained similar within both age groups.
In the young group, the highest RSR were observed at the posterior torso, followed by the anterior torso, legs/feet then arms/hands. In particular, the upper region of the posterior torso had the highest RSR and the feet had the lowest. This pattern mirrors the findings from previous sweat mapping research in young males (Smith and Havenith
2011), albeit with lower absolute values, due to differences in the exercise intensities employed. Similar distribution patterns have also been observed with studies assessing variation within body segments, including the hand/arms (Smith et al.
2007) and torso (Havenith et al.
2008; Machado-Moreira et al.
2008a) and other studies assessing multiple regions (Cotter et al.
1995; Smith et al.
2013a).
The older group showed a similar RSR pattern for both rest and exercise, with the exception of the extremities. The legs were shown to have a lower RSR than the arms, particularly during exercise. Compared to the younger group, older individuals had significantly lower RSR at several regions of the torso at rest and at all leg regions at both rest and during exercise. A significant body of literature conducted by Inoue and colleagues has consistently observed lower RSR at the leg regions (single sweat capsule sample at thigh) in older individuals (Inoue et al.
1991,
1995,
1998,
1999a,
b; Inoue and Shibasaki
1996). However, within some of the aforementioned studies, the decrements were also noted at several other body regions including the back (Inoue
1996; Inoue et al.
1998), chest, and forearm (Inoue et al.
1998). After conducting a substantial amount of research in this specific area, Inoue and Shibasaki (
1996) concluded that the age-related decline in heat loss effector function is likely to occur successively in cutaneous vasodilation, followed by sweat output per gland and density of active sweat glands. These decrements are suggested to proceed from the lower extremities, posterior upper body, anterior upper body, and lastly to the head (Inoue and Shibasaki
1996).
Other previous research has aimed to elucidate the exact physiological mechanisms responsible for the age-related differences in sweat rate (Tankersley et al.
1991; Inoue et al.
1999b; Smith et al.
2013a). Some studies have previously postulated that ageing per se has no influence on the sweat response, and instead, the decrement in sweat rate is related to the expected reduction in physical fitness and/or habitual physical activity level (Drinkwater et al.
1982; Smolander et al.
1990; Havenith et al.
1995). However, further studies identified that such declines in sweating still exist when young and older individuals are matched for aerobic fitness and activity levels (Tankersley et al.
1991; Armstrong and Kenney
1993; Inoue et al.
1999a; Smith et al.
2013a). Several factors may contribute to the lower sweat rates observed in older individuals including decreased heat-activated sweat gland function (Anderson and Kenney
1987; Kenney and Anderson
1988; Inoue et al.
1991; Inoue and Shibasaki
1996; Smith et al.
2013a), lower sensitivity to acetylcholine (Kenney and Fowler
1988; Inoue et al.
1999b), and a decreased thermal sensitivity (Natsume et al.
1992).
There is increasing evidence to suggest that a lower sweat gland output, caused by progressive atrophy of the sweat gland itself, is the primary contributing factor of the age-related changes in the sweating response (Sato and Timm
1988; Inoue et al.
1999b,
2004; Kenney and Munce
2003; Shibasaki et al.
2013; Smith et al.
2013a). As discussed above, it may be that this occurs prior to a decrease in heat-activated sweat gland density, as the previous studies have observed similar sweat gland densities across age groups (Anderson and Kenney
1987; Inoue et al.
1991; Inoue and Shibasaki
1996). Although this theory seems to be well documented, there are still questions surrounding the regional pattern of this age-related decline (Smith et al.
2013a). A peripheral-to-central decline has been suggested to be the most logical hypothesis (Kenney and Munce
2003). While we observed higher RSR at the lower extremities than the upper, indeed, it seems that RSR at the lower legs is reduced by 52% more than the upper legs with ageing (difference between age group upper and lower values during exercise period, Figs.
5,
6). As the pattern for the arms is different, the present study, perhaps, provides more support for the differentiating (between extremities) theory of Inoue and colleagues.
To assess RSR over the whole body, it was decided that the use of technical absorbent was the most appropriate method. Although this method is not new to the literature, it is rarely utilised in studies assessing RSR, which is surprising considering that it provides an inexpensive, easy-to-use alternative to the ventilated capsule method (Morris et al.
2013). Ventilated capsules only cover small areas of skin (2–13 cm
2), which may not be representative of the whole body segment (Havenith et al.
2008). In this context, it should be noted that many of the aforementioned RSR studies were limited to ~ 2 to 6 sweat capsules over the entire body. Consequently, their conclusions drawn upon age-related changes in RSR distribution, specifically over the extremities, are taken from a single small sample from the arm and the leg (typically the thigh). Utilising technical absorbents, the present and earlier data (Smith and Havenith
2011,
2012), show that large RSR variance exists over the extremities, and therefore, we propose that RSR inferences based on a limited number of capsules should be interpreted with caution.
Regional skin temperature
Regional fluctuation is evident in
Tsk as a result of a number of factors, including alterations in skin blood flow and evaporative cooling of the skin from sweating. Despite this association, there was no significant relationship between the rise in
Tsk and RSR in the present study at rest or during exercise, in agreement with the previous research (Cotter et al.
1995; Smith and Havenith
2011,
2012). Local
Tsk significantly increased during the rest period at all body regions in the older group and all except the upper back in the young group. It is noteworthy that both age groups
Tsk decreased at the majority of regions during exercise, which may be explained by the increase in evaporative cooling and a higher air velocity (1.5 m s
−1) compared to rest. Despite whole body towel drying to mitigate the build-up of excess sweat, continued evaporative sweat loss during the infrared images may also have contributed to the lower
Tsk observed after exercise.
Age-related differences in regional
Tsk were observed during both rest and exercise in the current study (ESM3). Older individuals had a significantly higher
Tsk at the leg regions when compared to the young, which coincides with the lower RSR observed in this region. Together, this is indicative of a lower evaporative cooling in the older compared to younger group. For visual purposes, Fig.
7 illustrates these age-related differences in
Tsk (at the end of the rest period) in the form of a body map. When combining Fig.
7 with the sweat maps data, it is evident that the legs are the most affected body area in relation to the decline in thermoregulation with age.
As the sweat measurement with absorbent pads involves covering large areas of skin, the impact of
Tsk changes in the sampling period must be considered due to the risk of artificially increasing RSR. At rest, the
Tsk of some regions increased significantly by ~ 0.4 °C while during the exercise sample period,
Tsk increased by up to ~ 0.6 °C. This is a relatively small rise in
Tsk when compared to the average increase observed in the previous studies (Smith and Havenith
2011,
2012), using the same pad application technique (> 1 °C). An increase in
Tsk is arguably unavoidable when utilising this technique even despite the short application periods of the absorbent material (5 min). However, the authors of the aforementioned studies concluded that regional sweat variation could not be explained by observed regional variations in
Tsk, which also holds true for the present findings. In support of this, a recent study underlines the limited role of
Tsk in sweat control (Ravanelli et al.
2020).
Gastrointestinal temperature, thermal sensation, and comfort
During exposure to the heat, the older individuals had a significantly greater rise in
Tgi during both rest and exercise, when compared to the young group. This was evident despite all participants working at the same fixed rate of heat production. The increased heat strain observed in the older group is a result of a decreased ability to dissipate heat through vasomotor adjustments and sweating and has been observed in numerous previous studies, albeit not implementing fixed heat production protocols (Anderson and Kenney
1987; Sagawa et al.
1988; Inoue et al.
1991; Dufour and Candas
2007; Smith et al.
2013a). The rise in
Tgi observed does not seem threatening under controlled laboratory conditions; however, many older individuals typically spend longer durations exposed to heat stress than in this study, especially during the summer months, and thus are at increased risk of heat-induced illnesses and injury (Waldock et al.
2018).
Despite having a significantly higher increase in
Tgi, the older group felt slightly cooler throughout the trial and more comfortable at the end of the rest period, rating lower values than their younger counterparts. This was especially evident for the period when applying the absorbent pads and stretch clothing to the skin, as the younger group reported a significant rise in thermal sensation and became more uncomfortable, whereas the older group did not. The inability to report a change in thermal sensation and comfort when adding a layer of clothing highlights the vulnerability of older individuals in warm conditions and supports the previous evidence of a reduced whole body thermal sensitivity (Natsume et al.
1992; Taylor et al.
1995; Tochihara et al.
2011; Takeda et al.
2016) and thermal comfort (Natsume et al.
1992; Taylor et al.
1995; Waldock et al.
2018) in the aged. The combination of impaired autonomic and behavioural responses further increases the susceptibility of the older population and is a cause for concern in the current climate (Kenney et al.
2014; Waldock et al.
2018).
Limitations
The present study assessed age-related differences in RSR using a whole body mapping approach with technical absorbents in male individuals. Unlike sweat capsules, this technique does not allow continuous monitoring of RSR development; hence, no analysis of the dynamics of sweat generation is possible. This disadvantage was accepted given the goal of covering a large part of the body for measurement, which is complicated with capsules (Taylor and Machado-Moreira
2013). Also the requirement to test the body in two parts may increase variability and thus reduce statistical power. To minimise this impact, all RSRs were standardised against the mean GSL variation of each trial. As work rates to achieve the same fixed heat production (200 W m
−2) were defined in preliminary tests, and metabolic rate was not measured during the exercise period of the main trials, this may have introduced some variation in the actual workloads. However, as the intensity of the exercise was light-to-moderate, substantial changes in mechanical efficiency (from the submaximal calculations to the main trial) were not expected throughout the exercise. Finally, as this study was conducted solely on male participants within a fixed environmental condition (32 °C/50% RH), generalisation of results to females and to other climate types may require careful consideration. Future research should aim to investigate age-related differences in RSR in females and within a wider range of ambient conditions.
Application
The findings of the present study have several applications from both a health-based and practical view point. The observed age-related declines in subjective and objective responses in the heat put older individuals at an increased risk of illness, injury, and, in extreme cases, mortality. Therefore, it may be necessary to revisit safety guidelines for working, exercising, and rest in hot conditions in those over the age of 65 years. Alternatively, aiming to alleviate these declines could promote better health in older individuals. The sweat mapping data presented within this study can be useful for clothing design, whereby different areas could be targeted to increase sweat evaporation, enhance cooling, and improve comfort. Moreover, the design of healthcare products and appliances may be tailored to individual needs based on the RSR of older individuals. For example, hospital beds and chairs that patients spend a large amount of time lying or sitting on could be designed to reduce irritation or the development of pressure sores caused partly from sweat accumulation. Finally, the current data relate directly into the design of thermal/sweating manikins and modelling in thermal physiology, providing more realistic sweat distribution patterns for young and older individuals.