Short‐ and Long‐Term Impact of Exogenous Hormones on Thermoregulation in Women Authors: Cara Wall‐Scheffler, Martin Hora Published: 2026 ABSTRACT ABSTRACTObjectivesHow the body changes its methods of thermoregulation as part of the female ovarian cycle is a crucial aspect of human adaptive strategies to negotiate variable environments and tasks. The use of exogenous hormones to manage the ovarian cycle is a regular part of modern life but its long‐term impact on thermoregulation, particularly while moving, continues to be understudied.MethodsHere, we specifically track adults (N = 35) including men, women of known ovarian phases, and women on an estrogen‐progesterone combination pill, doing the same walking and resting protocol while their core body temperature was monitored continuously and their blood perfusion was monitored regularly. Blood perfusion was normalized to maximum blood flow, and lines were fit to the blood flow during exercise and during rest; the slope of these lines was used to analyze blood flow of the different endocrine groups.ResultsWomen on the pill showed significantly higher core temperatures (x¯ = 38.16°C) than women not on the pill (luteal x¯ = 37.85°C; follicular x¯ = 37.69°C) and men (x¯ = 37.80°C) (p = 0.02). Women in the luteal phase showed the most shallow slopes of blood perfusion (between −0.14 and 0.07), whereas all other endocrine groups showed slopes ranging from −0.31 to 0.15. Endocrine group significantly impacted the slope of blood perfusion during resting and walking (p = 0.059).ConclusionsGiven the different core temperatures, the similarity in the amount and pattern of blood perfusion between women on the pill and women in the follicular phase was not enough to substantially lower the core temperature of women on the pill. Another mechanism besides blood flow is necessary to explain the substantial difference in core body temperature maintained by women on the pill throughout the experimental period. These results suggest that core temperature does not automatically trigger blood perfusion as a mechanism of cooling for women on the pill. Additionally, women in the luteal phase must use other, non‐vascular, mechanisms to cool the body during exercise. Finally, the thermoregulatory responses of people on a combination pill do not match the thermoregulatory responses of people in the luteal phase; studies which conflate these two should use caution in their interpretation. Introduction The maintenance of body temperature involves multiple physiological systems; as such, fluctuations in body temperature offer key insights into the integration of long‐ and short‐term feedback loops, as well as the relationship between central and peripheral conditions (Silverthorn 2026). In particular, the relationship between longer‐term loops such as those related to the cycling of reproductive hormones, and shorter‐term loops such as those related to the immediate impact of exercise continue to be of interest in order to better understand the maintenance of body temperature homeostasis (Charkoudian 2003; Charkoudian and Stachenfeld 2016, 2014; Charkoudian et al. 1999; Chaseling et al. 2020; Gagnon and Kenny 2011, 2012a, 2012b; Silverthorn 2026) as well as metabolic rate (Curtis et al. 1996). The endocrine control of blood perfusion and hypothalamic set points is of particular interest for female physiology (Baker et al. 2020). During the estrogen‐dominated follicular phase, the thermoregulatory set‐point is lower than during the progesterone‐dominated luteal phase (Charkoudian and Stachenfeld 2014). Hypotheses vary as to how this lower set‐point is both initiated and maintained. The current data strongly suggest that estrogen promotes vasodilation, which subsequently allows for active blood perfusion to the periphery, thus cooling the body. This impact on perfusion via active vasodilation is generally agreed upon to explain the mechanistic reason for lower body temperatures during the follicular phase, particularly at rest (Charkoudian and Stachenfeld 2016). During the progesterone‐dominated mid‐luteal phase, the body's temperature is higher than during the follicular phase (Baker et al. 2020; Charkoudian and Stachenfeld 2014; Kolka and Stephenson 1997). This is generally believed to be due to a higher hypothalamic set‐point and is potentially maintained by an inhibitory effect of progesterone on vasodilation, particularly during exercise (Kolka and Stephenson 1997). Progesterone may also cause vasoconstriction to begin at a higher body temperature threshold (Charkoudian 2003), limiting blood perfusion, though this is still uncertain. Even more unclear is whether the constant hormone levels produced by an endocrine supplement, like an estrogen‐progesterone combination pill, impact the global and local conditions in the same way as the fluctuating levels of hormones work during the regular cycling between the follicular and luteal phases (Charkoudian and Stachenfeld 2014). In particular, though women on the combination pill generally show a higher core temperature (Baker et al. 2020), it is not clear if habitually high levels of progesterone, such as those when on the pill, show the same influence over systemic blood perfusion as acutely high levels of progesterone, such as those during the mid‐luteal phase. While studies on resting individuals generally show that estrogen promotes vasodilation and progesterone inhibits vasodilation, whether these effects vary during exercise remains less well understood (Charkoudian et al. 1999). Additionally, though many studies have investigated blood flow and blood perfusion during both rest and exercise (Charkoudian 2003; Charkoudian and Stachenfeld 2016, 2014; Charkoudian et al. 1999; Gagnon and Kenny 2011, 2012a, 2012b), most studies focus exclusively on blood flow through the forearm (Kolka and Stephenson 1997), with very few studies looking at trunk and shoulder blood flow as well. During exercise, systemic vasodilation occurs to send blood towards the surface of the skin in order to dissipate heat (Gagnon and Kenny 2012a). Though the threshold to start sending blood might be higher during luteal phase, the total amount of blood flow might also be higher when compared to follicular phase (Kolka and Stephenson 1997). Again, though it is assumed that these patterns are the same for an individual on the pill due to similarly higher core temperatures at rest, measures of the total pattern of blood flow during chronic versus acute levels of progesterone have shown conflicting results (Lei et al. 2019). Here we investigate the relationship between blood flow to the trunk and whole upper limb regions, and core body temperature during exercise in a sample of women and men. We have three primary hypotheses. Due to the hormonally‐mediated lowered responsiveness of blood vessels caused by high levels of progesterone (Charkoudian and Stachenfeld 2016), we hypothesize that during exercise, women experiencing the higher progesterone levels of the luteal phase will show a reduced pattern of blood flow changes in comparison with women in the follicular phase.Chronically high levels of progesterone, as found in women on the combination pill, may influence vasodilation and vasoconstriction patterns differently, leading to a potentially altered thermoregulatory situation for women while on the pill (Charkoudian and Stachenfeld 2014). We thus hypothesize that women on an estrogen‐progesterone combination pill will have significantly different blood perfusion patterns than women in either the luteal or follicular phases. These differences might include blood perfusion changes to the trunk, as well as to the limbs (Charkoudian et al. 1999).Because women in the follicular phase maintain a core body temperature more similar to males, and at least similar forearm blood perfusion patterns (Gagnon and Kenny 2012b), we hypothesize that women in the follicular phase will show similar blood flow patterns to men during both exercise and during the exercise recovery period. Due to the hormonally‐mediated lowered responsiveness of blood vessels caused by high levels of progesterone (Charkoudian and Stachenfeld 2016), we hypothesize that during exercise, women experiencing the higher progesterone levels of the luteal phase will show a reduced pattern of blood flow changes in comparison with women in the follicular phase. Due to the hormonally‐mediated lowered responsiveness of blood vessels caused by high levels of progesterone (Charkoudian and Stachenfeld 2016), we hypothesize that during exercise, women experiencing the higher progesterone levels of the luteal phase will show a reduced pattern of blood flow changes in comparison with women in the follicular phase. Chronically high levels of progesterone, as found in women on the combination pill, may influence vasodilation and vasoconstriction patterns differently, leading to a potentially altered thermoregulatory situation for women while on the pill (Charkoudian and Stachenfeld 2014). We thus hypothesize that women on an estrogen‐progesterone combination pill will have significantly different blood perfusion patterns than women in either the luteal or follicular phases. These differences might include blood perfusion changes to the trunk, as well as to the limbs (Charkoudian et al. 1999). Chronically high levels of progesterone, as found in women on the combination pill, may influence vasodilation and vasoconstriction patterns differently, leading to a potentially altered thermoregulatory situation for women while on the pill (Charkoudian and Stachenfeld 2014). We thus hypothesize that women on an estrogen‐progesterone combination pill will have significantly different blood perfusion patterns than women in either the luteal or follicular phases. These differences might include blood perfusion changes to the trunk, as well as to the limbs (Charkoudian et al. 1999). Because women in the follicular phase maintain a core body temperature more similar to males, and at least similar forearm blood perfusion patterns (Gagnon and Kenny 2012b), we hypothesize that women in the follicular phase will show similar blood flow patterns to men during both exercise and during the exercise recovery period. Because women in the follicular phase maintain a core body temperature more similar to males, and at least similar forearm blood perfusion patterns (Gagnon and Kenny 2012b), we hypothesize that women in the follicular phase will show similar blood flow patterns to men during both exercise and during the exercise recovery period. Materials and Methods Thirty‐five young adults participated in this research, 13 males and 22 females (mean age = 21.6 ± 1.6; Table 1). The females were initially divided into two groups, one group who was taking estrogen‐progesterone combination pills (N = 8), and one group who was freely cycling (N = 14). Of the group that was freely cycling, one group (N = 9) came to participate in the non‐menstruating follicular phase, and one group (N = 5) came to participate 7–10 days after a positive luteinizing hormone (LH) surge using a standard urine‐based LH test (Wondfo, Huangpu, Guangzhou, China) and were thus in mid‐luteal phase. There were no significant age differences between the people in the different groups (p = 0.365; Table 1). Participants exercised an average of 127 min/week; there were no significant differences between the amount of exercise accomplished and the different endocrine groups. Participants also had their body composition collected via bioimpedance (Biodynamics CIA 310e); there were no significant differences between the bioimpedance of the female participants across the three groups (Table 1). All participants signed written informed consent approved by Seattle Pacific University's Institutional Review Board. All data were collected in February in Seattle (Washington State, USA; a high latitude city in the northern hemisphere) and all participants had been residents in Seattle for at least 1 month prior to the study. TABLE 1Anthropometrics of the four groups.Group N Age (years)S.D.Mass (kg)S.D.Stature (m)S.D.Bioimpedance (ohms)S.D.Surface area (m2)S.D.SA:Mass ratio (m2/kg)S.D.Follicular921.00.056.78.891.60.07648.286.41.50.140.02660.002Luteal521.41.170.55.961.70.07611.660.51.70.110.02390.001Pill821.50.962.18.901.70.09701.977.91.50.150.02490.002Male1322.22.477.016.271.80.07514.073.01.80.150.02390.003 Note: Two‐tailed, independent t‐tests showed that males were significantly taller (p < 0.001), heavier (p = 0.007), and more dense (i.e., had less water) (p < 0.001) than females, but did not have significantly different SA:Mass ratios (p = 0.181). Between the female groups, there were no significant differences in stature (p = 0.443), bioimpedance (p = 0.139) or surface area (p = 0.078); however in regards to mass, luteal phase women were heavier than follicular phase women (p = 0.01) though no other differences in mass occurred (i.e., luteal phase women were not heavier than women on the pill (p = 0.09), nor were women on the pill heavier than women in follicular phase (p = 0.23)). Such similarities in body size and body composition may explain the similarities to behavior shown whilst accomplishing the protocol. Anthropometrics of the four groups. Note: Two‐tailed, independent t‐tests showed that males were significantly taller (p < 0.001), heavier (p = 0.007), and more dense (i.e., had less water) (p < 0.001) than females, but did not have significantly different SA:Mass ratios (p = 0.181). Between the female groups, there were no significant differences in stature (p = 0.443), bioimpedance (p = 0.139) or surface area (p = 0.078); however in regards to mass, luteal phase women were heavier than follicular phase women (p = 0.01) though no other differences in mass occurred (i.e., luteal phase women were not heavier than women on the pill (p = 0.09), nor were women on the pill heavier than women in follicular phase (p = 0.23)). Such similarities in body size and body composition may explain the similarities to behavior shown whilst accomplishing the protocol. Participants were asked to refrain from eating and drinking for 4 h prior to the protocol. The protocol always started between noon and 2 pm, so most people had eaten a small breakfast when they woke up, and no other food that day. Each participant was further asked to wear a tank top, shorts above their mid‐thigh, ankle socks, and their regular running shoes. People were asked not to wear any make‐up, and when applicable, hair was tied up above their neck. Mass, stature, and body composition via bioimpedance (Biodynamics CIA 310e) were taken (without shoes) at the same time as the data collection (Table 1). Additionally, whole body anthropometrics were taken, including lengths and circumferences of each segment, following Cross et al. (2008) method of measuring body surface area (Cross et al. 2008). The protocol consisted of participants swallowing a sensor that measured core body temperature made by MiniMitter (Philips Respironics, Bend, OR, USA). This tracked their body's temperature in real time, on a minute‐by‐minute basis. The participant waited at least an hour until the pill's temperature readings stabilized (no change in temperature across 10 readings), which meant the pill was in their duodenum and not their stomach prior to beginning the protocol. Once the pill had settled, participants moved into the testing area, which was a private stairwell, consisting of 41 stairs that moved between three floors. Each participant sat on a stool with their eyes closed, feet flat on the floor, and palms face up on their knees for 10 min. Then, they walked up and down the stairs for 5 min, paused for no more than 30 s while a side satchel (8.6 kg) was placed on them, walked for another 5 min, paused while the side satchel was removed, and then walked without the side satchel for a final 5 min (15 min of stair walking in total). After this they sat back on the stool and rested for 10 min (as above). Before and after every rest or pause, blood perfusion was measured using a white‐light probe (DSM II ColorMeter, Cortex Technology). This measures the amount of hemoglobin in blood flowing to the skin (Charkoudian 2003; Charkoudian et al. 1999; Chaseling et al. 2020) as a unit‐free index (Jaspers and Moortgat 2020) and has been used successfully in numerous studies on blood flow and perfusion. “It has been estimated that laser‐Doppler probes placed on [the] skin will sample at a maximum depth of 0.73 mm and volume of 2.7 mm3. As such, measurements of red blood cell flux primarily reflect flux through vessels of the upper dermal plexus” (Chaseling et al. 2020, p. R61). Blood flow was measured at five locations: the suprasternal notch, and the anterior aspects of biceps, forearm, wrist, and palm (Figure 1); participants maintained identical arm position for all measures (Figure 1) (Chaseling et al. 2020). Locations of each blood flow sampling area. Blood flow was collected before and after each rest (10 min periods) and during every 5‐min pause during stair walking. Areas were marked with a small circle around the testing site to ensure that measures were taken in the same place each time. Participants were stationary and their arm was in anatomical position for all measures. Figure drawn by Quinton Ouelette. Participants were asked to walk up and down 48 stairs, one step at a time, at the same pace for the entirety of the 15 min of walking. Speed was determined as the ratio of distance traveled (calculated by measuring the rise and the run of the steps) to the duration of the time it took to walk the distance. Participants followed a marked path up and down the steps to ensure the staircase was the same distance (22.9 m) for each walker. There were no significant differences between any group's mean speed (x¯ = 0.77 m/s) during walking (Table 2). Temperature (x¯ = 17.2°C) and humidity (x¯ = 50.4%) were monitored continuously through the experiment; independent t‐tests showed no significant differences between temperature and humidity during any group's session (Table 2). TABLE 2Experimental conditions for the different groups.Group N Speed (m/s)Standard deviationAmbient temperature (°C)Standard deviationAmbient humidity (%)Standard deviationFollicular90.810.1016.72.1550.86.89Luteal50.710.0516.92.4052.715.45Pill80.770.1117.81.5251.411.62Male130.770.1817.41.6246.610.12 Experimental conditions for the different groups. A repeated measures ANOVA was run in SPSS 31 for changes in core temperature across each 5 min of stair‐walking and the final 10 min of rest. Endocrine group was included in the model as a factor. For blood perfusion, all erythrocyte values were normalized to the maximum value for each body part for each person. Then, a linear regression was fit to the normalized perfusion values for each body region during exercise (the “Start” to “15 min” time points as noted on the x‐axis in Figure 2), and the slope calculated for each person. A line was also fit to the normalized perfusion values for each body region during resting (from the end of the 15 min of stair walking through the “End of Rest” as noted on the x‐axis of Figure 2), and the slope calculated for each person. The two slope values, for “stair‐walking” and “resting,” for each body region were included as the dependent variables in a GLM‐ANOVA for blood perfusion. Both endocrine group and body part were included as factors. For all models, the interaction between body mass and surface area was included as a covariate because of the role of individual body morphology on heat loss (Hora et al. 2022, 2020; Tilkens et al. 2007). Results The body's core temperature remained constant over the first five minutes of stair‐walking, before increasing steadily over the next ten minutes (Figure 2). The temperature then dramatically decreased during the ten‐minute rest. The absolute core temperatures significantly varied by endocrine group (p = 0.02, Table 3), with people on the pill having significantly higher core temperatures than all other groups (Figure 2). There was no significant impact of endocrine phase on the pattern of core temperature across the experiment (p = 0.466); in other words, there was not a significant interaction between stair‐walking time and endocrine phase (Figure 2). Absolute core temperatures across the five time points for each of the four endocrine groups, emphasizing the substantially higher core temperatures among women on the pill. Endocrine group was a significant main effect in the repeated measures ANOVA (p = 0.02, η 2 = 27.6%) though it did not significantly interact with how core temperature changed across the protocol (p = 0.466). Variability bars represent the 95% confidence interval. TABLE 3Mean overall core temperatures for each endocrine group.Group N Core temperature (°C)Standard deviationFollicular937.690.40Luteal537.850.36Pill838.160.38Male1337.800.35 Mean overall core temperatures for each endocrine group. Testing our three hypotheses focuses comparisons on the change in blood flow during the exercise portion of the protocol (the slope during exercise), versus the change in blood flow during the final rest period (the slope during resting). In the GLM‐ANOVA, the difference in slopes between resting and stair‐walking was significant (p = 0.003, η 2 = 85.1%). Per our hypotheses, endocrine group significantly interacts with the difference between resting and stair‐walking (p = 0.059, η 2 = 44.7%; Figure 3). There is also a significant interaction between the different regions of the body where blood perfusion is occurring and blood flow during stair‐walking and resting (p = 0.048, η 2 = 32.2%; Figure 4). This figure shows the different patterns of blood perfusion by endocrine group, as calculated by the slope of the normalized blood perfusion values (the y‐axis values). The red lines represent the steepness of the slope where blood is shunted during exercise (a high number means a steeper slope, and a greater increase in perfusion), and the blue lines represent where blood is shunted during rest (a number closer to zero represents a more shallow slope and less blood flow). Whereas women in luteal phase show the smallest slopes, and minimal variation in blood flow patterns for the neck between exercise and rest, other people show steeper slopes at the clavicle than at other parts of the body. The differences between endocrine groups and blood flow during rest and exercise are significant (p = 0.059). This figure shows the significant interaction (p = 0.048) between the difference in blood flow (the slopes of the normalized blood flow) during stair‐walking (positive slope values) and resting (negative slope values) to the different areas of the body by endocrine group. Discussion This research emphasizes the importance of long‐term (e.g., exogenous) and short‐term (e.g., endogenous) effects of fluctuating hormone load on blood perfusion during exercise and subsequent rest. Women on the pill show significantly higher core temperatures than other groups, particularly when exercising, but show a pattern of blood perfusion during exercise similar to women in follicular phase (and men) (Figure 3); however, women on the pill show substantially steeper, more negative slopes during post‐exercise rest than all other groups (Figure 4). Conversely, women in luteal phase, though having only slightly higher core temperatures than women in follicular phase (and men), show a reduced pattern of blood flow changes compared to the other groups (i.e., shallow slopes; Figure 3). This difference in slopes between women in luteal phase (shallow) and women on the pill (steep) is a key difference between short‐term and longer‐term progesterone response. Besides this, people generally show substantial diversion of blood to the chest and neck during exercise (the positive, steep slopes in red of Figure 3) with a rapid reversal of this pattern once exercise has stopped (the negative, steep slopes in blue of Figure 3). Previous work has suggested that women on the pill should show some physiological similarities with women in the luteal phase, given that both conditions have high levels of a combination of progesterone and estrogen (Baker et al. 2001); however, women on the pill may also show some chronic shifts not shown among naturally cycling women in the luteal phase, including substantially higher core temperatures as found here (Baker et al. 2001; Rogers and Baker 1996), as well as the steeper profusion slopes. If the changes occurring are triggered by changes in set‐points, it might either be because of the differences between endogenous and exogenous hormones (Baker et al. 2001; Kaciuba‐Uscilko and Grucza 2001) or because of the differences between short‐term changes in neural firing and long‐term changes in hypothalamic organization (Frascarolo et al. 1990). While the above might explain the different set‐points, and thus the ultimate trigger of the high absolute core temperatures among women on the pill, other work—specifically focused on changes in blood flow—might explain why women in luteal phase show different blood flow patterns, whereas women on the pill do not (Figure 3). In fact, whereas some suggest that pyrogen action in the brain might be the causal mechanism of the higher temperatures, others argue that the key control of core temperatures themselves is the actions of estrogen and progesterone on vasodilation and vasoconstriction (Charkoudian et al. 1999). While endocrine mediation of vasodilation has been established via numerous experimental protocols (Charkoudian 2003; Charkoudian et al. 1999), a key finding of this study is that chronic versus acute changes in progesterone—as caused by being on the pill or just being in luteal phase—influence different locations of vasodilation (e.g., the neck versus the hand, Figure 3) as well as the steepness of slopes. In particular, women in luteal phase show a lack of increased perfusion through the neck; increased perfusion across the neck is a hallmark response of the other groups. Women in luteal phase instead rely almost entirely on continuous blood flow through the upper limb that changes little from their maximum flow between resting and exercise conditions (Figure 3), potentially explaining their high core temperature at the end of the final rest (Figure 2). Habitual perfusion, without a sensitivity to exercise intensity, did not allow for the recovery of resting core temperatures within the time frame of this experiment. Given the different starting temperatures, the similarity in the amount and pattern of blood perfusion between women on the pill and women in follicular phase was clearly not enough to substantially lower the core temperature of women on the pill. Another mechanism besides blood flow is necessary to explain the substantial difference in core body temperature maintained by women on the pill throughout the experimental period. Blood perfusion clearly increases and decreases with work and load (which did not vary between the groups), not with absolute core body temperature. Other research suggesting that blood perfusion might in fact be driven by a cessation in the sweat (sudomotor) response caused by chronic pill usage (Lei et al. 2019) would explain the pattern of disconnect between core temperature and blood perfusion seen here. Future studies should include sweat rate measures to test this further. Limitations and Future Work The work presented here offers multiple ideas for furthering our understanding of the interactions between acute and chronic endocrine changes that influence thermoregulatory patterns, and as such exercise and post‐exercise responses in female bodies in particular. Given that this work focused on a repeated‐measures design of physical work, we chose to use a cross‐sectional sample of people in the different ovarian phases; however, future studies could track people over longer periods of time, could monitor people coming on and off specific doses of endocrine supplements via the pill, and people going in and out of the different ovarian phases (e.g., Curtis et al. 1996). Such long‐term studies could further elucidate some of the intricacies of long‐ and short‐term feedback loops crucial for both the establishment of set‐points, and the shifting in blood vessel responsiveness. Such longer‐term studies would need to manage changes in fitness, season, and protocol familiarity, but would clearly offer important information. Additionally, here we used a white‐light probe to assess blood perfusion at specific moments when the participant was not moving (Chaseling et al. 2020). Other studies which use a flowmetry doppler device (Charkoudian 2003; Charkoudian et al. 1999) can track changes in real time, though often need the participant to be perfectly still during the study, so would still focus on the resting periods between movements (until technology improves). Additionally, since these devices are measuring color, it is also possible that other aspects of hemoglobin concentration and capillary density (“vascularity”) matter for initial, baseline measures (Jaspers et al. 2017); here we controlled for this by normalizing to maximum hemoglobin concentration at each anatomical location. Studies which utilize multiple measurement systems, especially longitudinally, might be able to discern changes in longer‐term feedback mechanisms (e.g., making a higher density of capillaries) in response to endocrine supplementation. Author Contributions C.W.S. designed the study, collected data, performed analyses, wrote and edited the paper. M.H. contributed to conversations regarding the data analysis, as well as writing and editing the paper. Funding This writing of this paper was funded as part of a Fulbright Scholars program. Ethics Statement As noted in the methods, the study was approved by the university's federal‐wide‐assurance‐approved institutional review committee and all participants signed written informed consent prior to participating in the study. Conflicts of Interest The authors declare no conflicts of interest.