Journal Logo


Female Reproductive, Adrenal, and Metabolic Changes during an Antarctic Traverse


Author Information
Medicine & Science in Sports & Exercise: March 2019 - Volume 51 - Issue 3 - p 556-567
doi: 10.1249/MSS.0000000000001803


Women undertake increasingly physically demanding sports and employment but sex-related biological consequences of arduous exercise are poorly understood. Over the past 20 yr, emphasis on energy availability (EA, defined as energy intake minus exercise energy expenditure) has established low EA as a putative cause of the “female athlete triad,” where hypothalamic-pituitary-gonad (HPG) axis suppression in athletes leads to functional hypothalamic amenorrhea and/or impaired bone health (1). The term “female athlete triad” has been questioned because these phenomena can also affect men (2); however, women may have greater sensitivity to the effects of low EA than men, and there is a higher prevalence of disordered eating among women than men (3).

In the setting of military employment, it has been suggested that women may be at higher risk of psychological problems than men, such as posttraumatic stress disorder (1,4). There appears to be evidence suggesting a greater incidence of primary infertility in military women than age-matched civilians (5). Although these observations remain unexplored in terms of etiology, we recently proposed functional hypothalamic amenorrhea in military women might contribute to menstrual dysfunction, hypothesizing this could be mediated by a complex alteration in hormonal milieu, including reduced EA (1). Aspects of military training and employment other than exercise and reduced EA may also be likely to contribute to HPG axis suppression, for example, sleep deprivation and psychological stress (1,6,7).

Field studies of military training generally measure the effects of multiple concurrent stressors, making it difficult to delineate the effects of individual components like low EA, sleep deprivation or psychological stress (6). One highly researched model of the endocrine effects of a multi-stressor environment is US Army Ranger Training. Predominantly undertaken by men, Ranger training involves 61 d of strenuous exercise, sleep deprivation, total energy expenditure of around 4000 to 5000 kcal·d−1, routine energy deficit and widespread metabolic and hormonal deficiencies, for example, elevated fasting cortisol, reduced total testosterone and IGF-1 (7,8). Such changes have been demonstrated to be reversible upon re-feeding, cessation of stress and sleep derestriction (8). However, extremes of arduous exertion lasting this duration have not been widely researched in women.

We undertook an exploratory, observational study of the concurrent acute response and short-term recovery of female HPG and hypothalamic-pituitary-adrenal (HPA) axes (using basal and dynamic testing) in women undertaking an unprecedented, extremely arduous expedition to cross the Antarctic continental landmass, of similar duration to US Army Ranger training. The purpose of the crossing was to attempt to become the first all-female team to complete an unassisted Antarctic traverse using muscle power alone, and was not competitive, primarily research-focused or done to achieve a political or military training objective. The a priori hypothesis was that this expedition would induce an energy deficit, despite a comprehensive program of physical and nutritional preparation, with concurrent disturbances in HPG and HPA axes.



Six women participating in an unassisted Antarctic ski traverse expedition were invited to participate in the study 3 months beforehand. This was the first all-female team to attempt an unassisted Antarctic traverse. Individuals planned to haul sledges weighing 80 kg for 1700 km, expecting the crossing to take around 75 d. Selection and training for the expedition lasted 2 yr, the final team being selected from a pool of 250 women. Although none of the participants had been to Antarctica before, all had partaken in three preparatory expeditions in Norway, which aimed to simulate the crossing’s intensity and conditions (details can be found at Participation in the study was voluntary and independent of the expedition. All six women volunteered and provided written informed consent. Ethical approval was received form the Ministry of Defense Research Ethics Committee (827MoDREC/17). The study was conducted in accordance with the Declaration of Helsinki.

Experimental Design

The study design consisted of two preexpedition measurement sessions, 64 and 39 d before the expedition (visit pre-1 and pre-2, respectively) (Fig. 1A). Additional body composition measurements were undertaken separately from formal study visits, 16 d before and 5 d after the expedition. Follow-up visits were conducted 4 d after the expedition (immediately after arrival in Punta Arenas, Chile from Antarctica), and 15 d after the expedition, 36 h after return to the United Kingdom (visits post-1 and post-2, respectively). As part of a broader preparation schedule, participants were advised to gain 0.5 kg of body mass per week between visit pre-1 and the expedition (64 d or 9 wk; 4.5 kg). The expedition altitude profile and distance are indicated in Figure 1B. The maximum elevation above sea level was 2950 m.

Overview of experimental design and expedition. A, Timeline summary of major investigations. Saliva cortisol: five-point day curve was measured 40 to 34 d preexpedition and 18 to 24 d postexpedition (filled circle); morning and evening sampling undertaken 1, 5, and 10 d postexpedition (unfilled circle). Anthropometric examination: weight and skinfolds were undertaken 16 d preexpedition and 5 d postexpedition (filled triangle). Body fat was estimated by bioimpedance 1, 5, 10, and 20 d postexpedition (unfilled triangle). Questionnaires were undertaken 39 d preexpedition and 5 d postexpedition (diamond). Dynamic and basal blood tests: fasted blood sampling and dexamethasone-suppressed combined GnRH and ACTH-(1–24) test, 39 d preexpedition, and 4 to 5 and 15 to 16 d postexpedition (unfilled square). Body composition measured by DXA scan 64 and 39 d pre and 15 d postexpedition (filled square). B, Altitude profile of study. Dashed line: altitude. Solid line: elapsed ski distance. Target icons indicate altitude of study visits in Camberley, UK. GnRH, gonadotrophin-releasing hormone; ASL above sea level.

Dietary provision

Dietary provision for the expedition was estimated from changes in body mass during three training expeditions. During the expedition participants were provided with a complete diet providing average 20.9 ± 0.1 MJ·d−1 (4970 ± 25 kcal·d−1, or 70.8 ± 0.35 kcal·kg−1·d−1), comprising ~45% carbohydrate (7.7 ± 0.32 g·kg−1·d−1), ~45% fat (3.6 ± 0.07 g·kg−1·d−1) and ~10% protein (1.7 ± 0.35 g·kg−1·d−1). It is estimated (verbal communication) that participants consumed median 85% (range, 70%–99%) of the diet provided over the course of the expedition and did not share rations.


The schedule of measurements is illustrated in Figure 1. At visit pre-1, information including ethnicity, education, smoking habits, alcohol consumption, and a comprehensive medical reproductive and medication history taken including use and type of, and indication for, hormonal contraceptives was recorded. Reproductive and medication history and use of contraceptive questions were repeated after the expedition (visit post-1).

Psychological assessment

Questionnaires comprising six validated self-rating items on a Web-based application (SmartSurvey, Tewkesbury, UK) were completed at visits pre-2 and post-1 (Fig. 1). The psychosocial stress questionnaire was completed in an identical manner to Rosengren et al. (9), assessing the 6-month period before visit pre-2, and the 4-month period before visit post-1. Participants were asked to complete the Impact of Events Scale—Revised (IES-R) with reference to any major life event(s) identified (10). The Patient Health Questionnaire 9 (11) was chosen as a robust measure of depressive symptoms in military and civilian populations (12). We analyzed results on a continuous scale, to identify subtle differences in a low number of participants. The Beck Anxiety Inventory and Connor Davidson Resilience Scale 10 demonstrate similar consistency measuring anxiety and resilience, respectively, and were analyzed in the same manner (13,14). The BEDA-Q assesses risk of disordered eating concisely and consistently (15), and was scored according to the methods of Peric et al. (16). Total scores from each questionnaire were used for further analysis.

Weekly intraexpedition assessments

During the expedition, a weekly questionnaire was completed in the same manner as previous studies of female transantarctic expedition (Fig. 1) (17,18). This documented average perceived exertion, psychological stress, restfulness of sleep and confidence the team would complete the expedition (all on a Likert type-scale ranging from 1 [not at all] to 10 [the most possible]), and the average number of hours slept per night.

Body composition

Stature was measured at visit pre-1 (SECA Stadiometer 213, Birmingham, UK) and body mass was measured at every study visit (SECA Scales 874). Whole body and regional lean mass, fat mass and bone mineral content were measured using dual-energy x-ray absorptiometry (DXA) with participants wearing shorts and T-shirts at visits pre-1, pre-2 and post-2 (GE Lunar iDXA; GE Healthcare, Chalfont St Giles, UK) (Fig. 1).

Sixteen days before the expedition (separately from main study visits), and at visit post-1, skinfolds were measured at four sites (bicep, triceps, subscapular, supraspinatus) to the nearest mm by the same examiner using Harpenden calipers (BodyCare, UK) according to the method of International Society for the Advancement of Kinanthropometry (19). The average of three measurements taken from each site was used to calculate percentage body fat (19).

Body fat was measured by four-point bioimpedance (Omron BF511, Milton Keynes, UK) upon waking in the morning, 1, 5, 10, 15, and 18 to 24 d after the expedition.

Basal blood samples

After an overnight fast, a venous blood sample was collected at visits pre-2, post-1 and post-2 for measurements of metabolic, nutritional, reproductive and adrenal function.

Dynamic reproductive and adrenal cortex function

Dynamic reproductive and adrenal cortex function was measured at visits pre-2, post-1, and post-2. Participants first ingested 0.25 mg dexamethasone at 10:00 pm before a second overnight fast. This dose has been used to assess the sensitivity of the HPA axis to a near-physiological level of central negative feedback and to attempt to reduce the baseline variation in morning fasting cortisol before the prestimulation test cortisol (20,21). At 8:00 am, the following morning, a 21-gauge cannula was inserted into an antecubital or dorsal hand vein and a baseline blood sample was obtained before 10 μg Gonadorelin hydrochloride (Intrapharm, Maidenhead UK), followed by 1.0 μg adrenocortiocotrophin (ACTH)-(1-24) (tetracosactrin acetate as Synacthen®; Mallinckrodt, Dublin, Ireland), were injected followed by a 10-mL saline flush. ACTH-(1-24) was freshly diluted using 249-mL 0.9% NaCl (Baxter, UK), to which Synacthen® 250 μg in 1 mL had been added, shaken thoroughly and 1 mL of this mixture was injected using a 5-mL syringe to minimize contact with plastic. Venous blood was sampled through the cannula in ethylenediaminetetraacetic acid-containing tubes 20, 30, 40, and 60 min after drug administration. The doses of Gonadorelin, dexamethasone and ACTH-(1-24) were selected to mimic physiological levels of stimulation, as opposed to stimulation tests used clinically (and recommended in various clinical practice guidelines) which are intended to induce maximal axis stimulation and exclude endocrine insufficiency (e.g., 100 μg, 1 mg, and 250 μg, respectively) (20).

Hair and saliva cortisol

A 0.5-cm diameter hair sample was taken close to the scalp for measurement of cortisol at visit pre-2 (6 × 1 cm segments) and visit post-1 (4 × 1 cm segments). Hair grows at 1 cm per month, thus 1 cm represents 1 month of cortisol exposure (22).

Saliva was sampled by chewing on a synthetic swab for 1 min, which was placed in a plastic collection tube (Salivette®; Sarstedt, Nümbrecht, Germany). A detailed saliva day curve was measured at visits pre-2 and post-2 as follows: participants were woken at 07:00 and saliva sampled at 07:10, 08:20, 09:00, 09:30, 12:15, 13:30, 17:20, and 21:50. Evening and morning saliva sampling (last thing at night before going to sleep and immediately after waking the following morning) were also measured 1, 5, and 10 d after the expedition.

Laboratory Methods

Blood was collected in ethylenediaminetetraacetic acid, serum-separating gel and fluoride oxalate tubes (Monovette®; Sarstedt, Nümbrecht, Germany) and centrifuged at 5000 rpm for 5 min. Plasma and serum were stored at −80°C (after dry ice shipment to the UK of samples taken in Chile) until measurement.

Metabolic and nutritional markers

Thyroid-stimulating hormone (TSH), unbound thyroxine (fT4) and total T3 (tT3) were measured from gel-separated serum using Abbott® Architect analyzer (Abbott, Maidenhead, UK) according to manufacturer’s instructions. Insulin-like growth factor 1 (IGF-1), ferritin, insulin and C-peptide were determined from gel-separated serum using Roche® Cobas e411 analyzer (Roche Diagnostics, Welwyn Garden City, UK) according to manufacturer’s instructions. Creatinine, albumin, transferrin, calcium, zinc, iron, phosphate and magnesium were determined from gel-separated serum and glucose and lactate from plasma containing fluoride oxalate using commercial kits (Alpha Laboratories, Eastleigh, UK) adapted for use on a Cobas Fara centrifugal analyzer (Roche, UK). Leptin was measured by enzyme-linked immunosorbent assay (Quantikine, USA). Urea was determined from gel-separated serum using a commercial kit (Randox laboratories, UK) adapted for use on a Cobas Fara centrifugal analyzer. Homeostatic modeling assessment (HOMA) for beta cell function (HOMA-B), insulin sensitivity (HOMA-S), and insulin resistance (HOMA-IR) were calculated according to the methods of Levy et al. (23). Total 25-hydroxy vitamin D (25OHD) was measured using LC/ tandem MS using automated solid-phase extraction.

Additional data including resting energy expenditure and substrate utilization from direct calorimetry preexpedition and postexpedition are being published elsewhere.

Reproductive markers

Luteinizing hormone (LH), follicle-stimulating hormone (FSH), progesterone and estradiol were determined from plasma containing ethylenediaminetetraacetic acid using Abbot Architect® analyzer according to the manufacturer’s instructions. Inhibin B was measured by enzyme-linked immunosorbent assay (Beckman Coulter, High Wycombe, UK). Sex hormone-binding globulin (SHBG) and anti-Müllerian hormone (AMH) were determined from gel-separated serum using Roche® Cobas e411 analyzer according to manufacturer’s instructions. The rationale for these tests are summarized in the supplemental content (see Document, Supplemental Digital Content 1, Rationale for basal endocrine marker testing,

Adrenal markers

Cortisol, 17OH progesterone, testosterone, dihydroepiandrostenedione (DHEA) and androstenedione were measured using liquid chromatography mass spectrometry (LC/MS), by modifying internal standards from a protocol described previously (24). Hair was divided into 1 cm segments and powdered before cortisol extraction in each segment, representing 1 month averages, for a total of 10 months. Extraction and analysis by LC/MS was completed as described by Kirschbaum et al. (25). Saliva was stored at −80°C within 7 d of collection and was extracted and analyzed by LC/MS as described by Miller et al. (26).

Inter-assay %CV was <4% for Architect®, e411 and Fara assays and intra-assay %CV <10% for all enzyme-linked immunosorbent assay.

Statistical Analysis

Data are presented as individual data, or mean ± SD or median (interquartile range) for group comparison. Normality was assessed using Shapiro–Wilk test and nonnormally distributed data were log transformed before statistical analysis. Because of the small sample size, variables are presented as mean (95% confidence interval [CI]). Repeated-measures ANOVA was used to compare change in variables over time and pairwise comparisons were used where appropriate for statistically significant results. Paired t tests were used to compare the two preexpedition DXA scans, and single postexpedition variables with baseline. Preexpedition and postexpedition dichotomous questionnaire data were compared using χ2 test. One individual was excluded from analyses of basal reproductive hormones as she had commenced a combined contraceptive pill immediately before the expedition. Serum LH and FSH concentrations after injection of gonadotrophin releasing hormone and ACTH were described as absolute values, and as percentage change, by dividing concentrations after injection by the baseline concentration. This was done to allow comparison of within-subject change, because hormone-containing contraceptive use influenced baseline values. Area under the curve (AUC) was calculated using the trapezoidal rule. Within-subject changes in peak and AUC of cortisol and fold-rise in LH and FSH from baseline were compared from before to after the expedition.

Statistical analysis was performed using SPSS version 23.0 for Mac (IBM, USA). Significance was set at P < 0.05. For multiple variables assessed in the same domain, Bonferroni adjustment was made as follows: body composition, P < 0.01; basal reproductive markers, P < 0.005, adrenal markers, P < 0.05; metabolic markers, P < 0.002.


Description of Participants

Baseline characteristics of the cohort are shown in Table 1. The median (range) age was 32.8 yr (28.6–36.1 yr). Baseline questionnaires demonstrated high resilience, low depression and anxiety scores, and normal patterns of eating behavior. Fasting TSH, free T4, total T3, prolactin, LH:FSH ratio, androstenedione, total testosterone, DHEA, 17-OH progesterone, urea, sodium, potassium, chloride, and creatinine were within normal limits before the expedition (Table 2).

Characteristics of participants at visit pre-1.
Biochemical and hormonal parameters at baseline, 4 and 14 d after the expedition.

All participants used hormonal contraceptives during the expedition, intending to induce amenorrhea. One individual commenced levonogestrel 150 μg/ethinylestradiol 30 μg immediately before the expedition. One individual used Nexplanon® contraceptive implant while all others used a Mirena® intrauterine device. Five participants were amenorrheic during the expedition and one menstruated twice, stating this was less frequent than normal, within 4 to 10 d of due date.

Intraexpedition Rating Scales

Average scores for physical exertion scale were 5.5 ± 2.3/10 and stress level 3.7 ± 1.94/10, and level of confidence the team would complete the expedition 6.73 ± 1.81)/10; (6.35 ± 1.93 in weeks 1–3 and 7.11 ± 1.32 in weeks 5–8, P = 0.09). Average duration of sleep was 6.73 ± 1.75 h and rating of restfulness of sleep was 5.53 ± 2.05/10. Questionnaires after the expedition suggested moderately lower levels of psychosocial stress and financial stress, and fewer significant adverse events than before the expedition (P = 0.079).

Body Composition, Metabolic, and Nutritional Changes

Physical changes during the study are presented in Figure 2 and Supplemental Tables (see Table, Supplemental Digital Content 2, Anthropometric changes during the expedition,; see Table, Supplemental Digital Content 3, Regional lean, fat, and bone mass changes during the expedition, All participants gained body mass during the 2 months before the expedition, (average increase 2.6 ± 0.79 kg, or 3.69% ± 1.12% of body weight, P = 0.006), consisting of body fat (average increase 4.05% ± 0.96%, P < 0.0001), and lost body mass during the expedition (average loss 9.4 ± 2.31 kg, or 12.9% ± 3.17% of body weight, P < 0.0001). Body composition measured by DXA demonstrated a significant increase in total fat mass before (13.2 ± 2.11 vs 17.5 ± 2.52 kg, P < 0.001) and loss during the expedition (fat mass at visit post-2 was 12.1 ± 1.37 kg, P < 0.001), with these changes reflected in most regions (see Table, Supplemental Digital Content 3, Regional lean, fat and bone mass changes during the expedition, However, there was no difference in total lean mass or bone mineral content between visit pre-2 and visit post-2 (52.3 ± 2.10 vs 51.5 ± 3.04, P = 0.27), despite a 6.10% loss in lean mass from the legs. In the 15 d between the expedition and the follow-up DXA scan, fat mass estimated by bioimpedance tended to increase (see Table, Supplemental Digital Content 2, Anthropometric changes during the expedition, Regional DXA analysis showed statistically significant but modest decreases in android (area between the ribs and pelvis), gynoid (pelvis and upper thighs) and leg lean mass between visits pre-1 and pre-2, and loss of leg lean mass during the expedition (average 6.05 ± 1.11% decrease), but these did not impact the change in total lean mass (see Table, Supplemental Digital Content 3, Regional lean, fat and bone mass changes during the expedition, There was a small but statistically significant increase in total bone mineral content before the expedition (2.75 ± 0.13 kg vs 2.80 ± 0.13) kg, P = 0.005, but no change between visits pre-2 and post-2 (2.77 ± 0.12, P = 0.19 (see Table, Supplemental Digital Content 3, Regional lean, fat and bone mass changes during the expedition,

Anthropometric changes during the expedition. Data are mean ± SD. Shaded area: Duration of expedition. Circle with solid line: BMI. Square with dashed line: total body fat (%) by skinfold. Triangle pointing upward with dotted line: Total body fat (%) by bioelectrical impedance. Triangle pointing downward with dot-dash line: Total body fat (%) by DXA. Diamond with dot-dot-dash line: total lean mass (kg) by DXA.

Leptin decreased significantly after the expedition, thereafter increasing twofold from visits post-1 to post-2 (Table 2). Post hoc tests showed the change between visit pre-2 and post-1 was significant (P = 0.005), while there was no difference between pre-2 and post-2 (P = 0.39). Thyroid-stimulating hormone, free T4 and total T3 were normal preexpedition and remained unchanged after the expedition (Table 2). Fasted glucose, HOMA-B, HOMA-S and HOMA-IR, adjusted calcium, magnesium and phosphate did not change during or after the expedition (Table 2).

Questionnaire data demonstrated a marginal increase in BEDA-Q scores after the expedition, consistent with higher markers of disordered eating risk (see Table, Supplemental Digital Content 4, Pre and Post-Expedition psychological testing, Markers of nutritional status (albumin, magnesium, phosphate, iron, zinc), urea (Ln transformed) and electrolytes did not change during or after the expedition (Table 2).

Reproductive Function

Basal markers of reproductive function are displayed in Table 2. Estradiol tended to be lower at visit post-1, with a recovery noted by visit post-2. No differences between other sex steroids, LH or FSH were shown. Inhibin B and AMH did not differ between baseline and immediately after the expedition (P = 0.71 and P = 0.15, respectively, Table 2).

Dynamic LH and FSH responses before and after the expedition are shown in Figure 3. Fold rise in FSH and FSH AUC were log transformed before statistical analysis. LH and FSH fold rise and AUC during the test did not differ between visit pre-2 and visit post-1. At visit post-2, FSH had not changed from visit pre-1 (Fig. 3C; See Table, Supplemental Digital Content 5, Average values from dynamic endocrine function testing,, while there was a marked upward trend in LH, measured by AUC fold rise and peak fold rise (P = 0.055 and P = 0.071, respectively; Figure 3D, see Table, Supplemental Digital Content 5, Average values from dynamic endocrine function testing,

Dynamic gonadotrophin function before and after the expedition. Individuals represented by symbols. Actual concentrations of FSH (A) and LH (C) and fold difference (FSH, B; LH, D) from baseline concentrations of FSH (B) and LH (D) after 10 kg GnRH administration before, 5 d and 16 d after the expedition. FSH AUC and peak fold rise did not change across visits (P = 0.71 and P = 0.55, respectively). There was an upward trend in LH AUC and peak fold rise (P = 0.055 and P = 0.071, respectively). One individual (filled square) commenced levonogestrel 150 μg/ethinylestradiol 30 μg immediately before the expedition. One individual (unfilled circle) used Nexplanon® contraceptive implant while all others used a Mirena® intrauterine device.

Adrenal Cortex Function

Basal plasma cortisol did not change significantly during or after the expedition (Table 2).

Average hair cortisol before and during the expedition is shown in Figure 4B. Mean values are shown in Table, Supplemental Digital Content 5, Average values from dynamic endocrine function testing, Most participants demonstrated a significant increase in average cortisol levels during the expedition.

Dynamic, monthly average hair and diurnal saliva cortisol concentrations. A, adrenal response to ACTH-(1–24), 10 h after central suppression with 0.25 mg dexamethasone, before, and 5 d and 16 d after the expedition. Top row: cortisol concentrations. Bottom row: fold difference in cortisol from baseline. AUC and peak cortisol did not change between the three time points (P = 0.12 and P = 0.45, respectively). B, average monthly cortisol from 1-cm hair segments before and during the expedition (expedition represented by bracket). C, change in AUC and peak concentrations during the dynamic test before, and 5 and 16 d after the expedition. D, Saliva cortisol 36 to 40 d preexpedition and 18 to 24 d postexpedition (left panel) and diurnal cortisol 1, 4, and 10 d postexpedition (right panel). Individuals represented by symbols. Time: after ACTH-(1–24) administered. F, cortisol. **P < 0.001.

Individual dynamic plasma cortisol responses before and after the expedition are shown in Figure 4A. Both AUC and peak cortisol did not change between the three time points (P = 0.12 and P = 0.45, respectively, Figure 4B). Subjects demonstrated marked suppression of early morning cortisol after low-dose dexamethasone administration.

One participant demonstrated a more suppressed baseline in plasma cortisol than others (filled square symbol, Fig. 4). This individual also demonstrated markedly higher hair cortisol concentration through the expedition and two months beforehand.

Salivary cortisol in the days immediately after the exercise was blunted but by day 10 had recovered (Fig. 4C), reflected in a normal day curve which was unchanged from baseline (Fig. 4D).


With on-going debate as to whether women can endure extreme physical activity without detrimental effects on hormonal axes, given the finding of HPA and HPG axis suppression in extremely arduous exercise in men (e.g., in US Army ranger training (7,8)), we exploited the opportunity to examine the HPA and HPG axes among six women who completed a 1700-km ski expedition hauling 80-kg sledges up to 2950 m elevation. In doing so, the team broke several records including being the first all-female team to cross the Antarctic landmass unsupported. Our data demonstrate HPG and HPA axis resilience during extreme exertion despite significant fat loss. HPA axis basal function, sensitivity to central suppression and adrenal reactivity to ACTH did not change during or after the expedition, but demonstrated greater sensitivity to suppression from dexamethasone than anticipated from other studies using a similar protocol in older participants (20,21). Hair cortisol rose during the expedition as would be expected with sustained arduous exercise (27).

Coincidentally, the expedition duration (61 d) was identical to US Army Ranger training. Trainee Rangers are expected to cover around 322 km, carrying 30 to 41 kg. Although the expedition comprised a different form of exercise (skiing rather than walking or running), it was arguably noninferior in terms of effort or endeavor. One crucial difference is the 0 to 5 h of sleep per day expected during Ranger training (7), and deliberate psychological stress (28). which contrasts with the average 6.73 ± 1.75 h of sleep per night, albeit with poor perception of restfulness (in 24-h daylight), and modest weekly and whole-expedition stress ratings.

The primary drivers of adverse endocrine and metabolic changes in Ranger training appear to be nutritional deprivation (with loss of lean mass), psychological stress, sleep deprivation and exercise intensity. Nindl et al. (7,28) showed a 12.6% loss of body mass, 6% lean mass and 50% fat mass. The endocrine effects of negative energy balance are well-documented adaptations for survival and include suppression of the HPG axis and hypercortisolemia (1). In their meta-regression of field studies of arduous training, Murphy et al. (29) showed that the combination of training duration and low EA was inversely associated with physical performance (29), although it is difficult to delineate EA as a cause from the other factors described here.

A carefully calculated provision of approximately 21 MJ·d−1 (5000 kcal·d−1; ~45% carbohydrate, ~45% fat and ~10% protein), with significant fat gain before the expedition, plus a relatively low altitude and preservation of sleep, meant participants lost only fat mass, not lean mass. Sustained, submaximal exertion appears to have had the effect of preserving total lean mass, although leg lean mass reduced by 6.10%. This may relate to muscle fiber pennation rather than reduced mass per se; we were unable to confirm this by biopsy. Thus, weight loss was healthy, reinforcing the importance of appropriate nutrition preventing loss of lean mass and/ or hormonal disturbances, as has been shown in overtraining syndrome (30). As insufficient nutrition has been shown to cause multiple endocrine deficiencies in sports and exercise (2), we hypothesize that sufficient and appropriate nutrition had an important role in preventing changes to the HPA and HPG axes.

Calbet et al. (31) demonstrated that exercise maintains lean mass, during a 4-d extreme energy deficit in overweight men. Protein supplementation alone (1.5 g·kg−1·body mass−1·d−1) did not preserve lean mass, compared with carbohydrate. However, as demonstrated by Smith et al. in obese, sedentary women, a protein intake of 1.2 g·kg−1·d−1 mitigated loss of lean mass, compared with low protein intake (0.8 g protein·kg−1·d−1) during 10% weight loss over 27 wk (32). In men undertaking arduous military training, a mixed dietary supplement (5.1 MJ·d−1 (1220 kcal·d−1); ~45% carbohydrate, ~40% fat, ~15% protein) prevented 2 kg loss in lean mass, over 8 wk, compared with nonsupplemented controls (33). Despite a caloric deficit (indicated by weight loss), our participants maintained total lean mass, with an average protein intake of around 1.6 g·kg−1·d−1.

Low ambient temperatures induce brown adipose tissue (BAT) thermogenesis, mediated by catecholamine upregulation, acting as a sink for glucose and fatty acid uptake (34). Adaptive thermogenesis is upregulated by ß-3 adrenergic receptors, which are expressed in fat but not in muscle (35). Thus, the cold Antarctic environment could partially explain the high selectivity of substrate.

In mixed-sex Norwegian Ranger training involving 7-d food and sleep deprivation, women demonstrated greater fat utilization and glycogen preservation than men, implying greater capacity for endurance exercise (36). Estrogens appear to be responsible for this substrate dimorphism (37), while women subjectively claim better patrolling performance than men perhaps because of this metabolic advantage.

In addition to exercise and nutrition, the modest altitude of the expedition environment could have mitigated the loss in lean mass, compared with arduous expeditions at extreme altitudes, where hypobaric hypoxia contributes to loss of lean mass (38). Likewise, insufficient sleep, whether at altitude or as a programmed part of arduous training, could impede absorption of macronutrients and reduces gut readiness for daytime absorption (39), and it could be postulated that preservation of sleep contributed to the maintained total lean mass we observed.

No suppression of metabolic parameters such as thyroid hormones or elevated cortisol were seen during or after the expedition. Lean mass exerts a greater effect on resting metabolic rate and appetite than fat mass (40), and demonstrates a greater bidirectional relationship with androgens, and to a lesser extent estrogens, than fat mass (41). Thus, preservation of total lean mass might mitigate against some of the endocrine sequelae of negative energy balance. The decrease in leptin, followed by recovery postexpedition, was more pronounced than the changes we observed in body fat. Cold exposure itself may reduce leptin in women (42), but this appears to become effective only when cold exposure is sustained (43). The change in HPG axis function we observed did not correlate with leptin, as has been reported previously (44).

Dynamic attenuations in LH and sex steroids after an energy deficit may confer immediate survival benefits but may be associated with maladaptive suppression of hormonal axes and reproductive, bone or psychological sequelae if sustained (2). Luteinizing hormone was relatively suppressed before and during the expedition (reflecting hormonal contraception usage), but recovered by postexercise visit pre-2. There was no change in FSH before, during or after the expedition; this is consistent with studies of overtraining syndrome which generally demonstrate relatively normal FSH levels when LH is suppressed (reviewed in Cadegiani et al.) (45), and laboratory studies of reduced EA, which show normal levels relative to suppression of LH in response (46).

Cortisol reactivity and diurnal salivary cortisol were blunted relative to other studies, and may be an appropriate response to a high intensity of training (20,21). Alternatively, similar responses have been noted in dynamic testing of athletes during dysfunctional overtraining, also associated with elevated basal cortisol (reviewed in Cadegiani et al.) (45). Elevated hair cortisol concentrations are associated with exercise per se; whether the marked elevation during the expedition may represent an overtraining syndrome would be a pertinent question for future studies (27). The response of the HPA axis to central negative feedback is greater than has been described elsewhere (4,20). Yehuda et al. (4) reviewed the use of low-dose dexamethasone suppression in posttraumatic stress disorder (PTSD), showing PTSD was associated with increased central axis sensitivity. No suggestion of PTSD was noted from the psychological stress or IES-R assessments before or after the expedition, thus this may relate simply to age, fitness and lower volume of distribution of these participants compared with previous studies.

Similar exercise-associated patterns in the HPA and HPG axis were seen after restricted carbohydrate intake with aerobic and resistance activity (average 46 ± 9.1 MET and 4.7 ± 0.7 sessions per week, respectively), in normal body mass index (BMI) women over 20 wk (47). This regimen achieved an 11.9% weight loss with unchanged lean mass, and was associated with increased menstrual dysfunction, reduced testosterone, estradiol, free T3 and TSH and unchanged cortisol compared with weight-stable, exercising controls. While the degree of weight loss was similar to the present study, this intervention was achieved primarily through dietary restriction, since the exercise was less intense. The investigators also assessed recovery, demonstrating partial normalization of sex and thyroid hormones and leptin after 18 wk. As in the current study, mood profile was unaffected by the intervention, which might possibly account for the apparently stable cortisol responsiveness we observed.

Other correlates of overtraining syndrome include sleep deprivation and psychological stress (48). Psychological stress is a prominent feature of extreme physical endeavor. Therefore, while both stress and reduced EA may be shown to cause reproductive endocrine dysfunction independently, their impact in this context may be synergistic and it may be impossible to draw a distinction between them (1). The expedition required both significant mental and physical exertion, although perceived stress levels were modest through the expedition and anxiety, depression and psychosocial risk factor assessments did not change after the expedition.

It has been suggested the psychological stress of Ranger training results from nutrient and sleep deprivation, which serve to increase the arduousness of many military training formats (6). Sleep deprivation in isolation is associated with elevated evening cortisol, flattened cortisol day curve, reduced androgen secretion and higher sympathetic nervous system activity (39). Female sex hormones appear to be protective of the effect of sleep deprivation on cortisol blunting after psychosocial stress (49). The sustained moderate to high exercise intensity needed for a polar traverse represents a different form of exertion compared to US Ranger training, including its sustained, repetitive nature, austere environment, safety concerns and isolation. The stress and physical exertion scores reported during the expedition were consistent with previous arduous expeditions (17,18), while sleep diaries showed significantly longer sleep duration than would be expected in Ranger training (7,28), albeit of low perceived restfulness. Both increased sleep and the sustained, submaximal intensity of exercise could also account for the biological resilience we observed. Together with the nutritional strategy taken, and relatively reduced energy expenditure in women compared with men, these factors might have contributed to mitigating some of the negative psychological effects.

The major strength of our study is the unique nature of the expedition; this likely represents the first opportunity to study a cohort of female participants complete an endeavor of such a prolonged, arduous nature. Mitigating against low EA in women is important, since women appear to be at greater risk of low EA and its consequences than men (1,2). Previous studies of prolonged, arduous training have focused on male cohorts and recovery rates in women have not been studied. Furthermore, the effects of exercise or low EA on dynamic function of the HPA and HPG axes have not previously been studied in either sex.

Limitations to our study include the small number of participants. This is unavoidable on such extreme expeditions; we have attempted to mitigate this by a comprehensive characterization of the participants. The team is larger than any previous female-only transantarctic attempts, increasing the number of women who have crossed the continent from four to 10 (17,18). Other limitations include the natural limitations of a field study, such as 4-d delay in testing after the expedition. Every effort was made to overcome these using study visits shortly after the expedition arrived in Chile with imaging undertaken as soon as reasonably possible following the participants return to the UK. It was not logistically possible to repeat imaging immediately before and after the expedition, or use the same examiner to perform skinfolds in the UK and Chile, so we used the best feasible measures of body composition. The use of hormonal contraceptives, while representative of real-world hormonal milieu, do limit the interpretation of LH responses. For logistical and ethical reasons, dynamic tests of the HPA axis at a higher level (e.g., insulin tolerance test, corticotrophin releasing hormone test, desmopressin test) were not possible, however in future studies a maximal or two-bout exercise test could be considered. Calculation of cortisol awakening response would add merit to our study, but was not possible since participants were woken 10 min before the first saliva sample taken in the preexpedition and postexpedition day curves.

In conclusion, no short-term adverse effects were demonstrated from an unprecedented, successful transantarctic expedition in women. Cortisol reactivity and pituitary gonadotrophin reactivity were not impaired. We hypothesize these findings related to and preexpedition and intraexpedition nutrition, sleep provision, on the background of desirable selection characteristics, so that participants did not rate the expedition as subjectively stressful and lean mass was maintained.

The study was funded by an unrestricted grant from the UK Women in Ground Close Combat Review. The funders had no role in the study design. The authors are indebted to the Expedition ICE MAIDEN team for their exemplary participation in a very challenging study. We are grateful to the Wellcome Trust Clinical Research Facility (CRF), and particularly Jo Singleton, the CRF Mass Spec Core led by Natalie Homer, and to the Coventry NIHR CRF Human Metabolic Research Unit, University Hospitals Coventry and Warwickshire NHS Trust and the University of Warwick, Coventry, UK. We are grateful to the assay labs of Dr Forbes Howie and Dr Neil Johnson in the University of Edinburgh Centres of Reproductive Health and Cardiovascular Science, respectively. We thank Antarctic Logistic Expeditions Ltd. for supporting researchers in Antarctica and Chile. The authors would also like to thank the Surgeon General, Defense Medical Services and Expedition ICE MAIDEN Higher Management Committee for travel and subsistence costs.

None of the authors has conflicts of interest to declare, including professional relationships with companies or manufacturers who will benefit from the results of the present study. The results of the present study do not constitute endorsement by ACSM. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.


1. Gifford RM, Reynolds RM, Greeves J, Anderson RA, Woods DR. Reproductive dysfunction and associated pathology in women undergoing military training. J R Army Med Corps. 2017;163(5):301–10.
2. Mountjoy M, Sundgot-Borgen J, Burke L, et al. The IOC consensus statement: beyond the female athlete triad–relative energy deficiency in sport (RED-S). Br J Sports Med. 2014;48(7):491–7.
3. Gibbs JC, Williams NI, De Souza MJ. Prevalence of individual and combined components of the female athlete triad. Med Sci Sports Exerc. 2013;45(5):985–96.
4. Yehuda R. Current status of cortisol findings in post-traumatic stress disorder. Psychiatr Clin North Am. 2002;25(2):341–68, vii.
5. Interim report on the health risks to women in ground close combat roles 2016-20160706_ADR006101_Report_Women_in_Combat_WEB-FINAL.PDF. 2016. 65.
6. Hoyt RW, Friedl KE. Field studies of exercise and food deprivation. Curr Opin Clin Nutr Metab Care. 2006;9(6):685–90.
7. Nindl BC, Barnes BR, Alemany JA, Frykman PN, Shippee RL, Friedl KE. Physiological consequences of U.S. Army ranger training. Med Sci Sports Exerc. 2007;39(8):1380–7.
8. Henning PC, Scofield DE, Spiering BA, et al. Recovery of endocrine and inflammatory mediators following an extended energy deficit. J Clin Endocrinol Metab. 2014;99(3):956–64.
9. Rosengren A, Hawken S, Ounpuu S, et al. Association of psychosocial risk factors with risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364(9438):953–62.
10. Weiss DS. The Impact of Event Scale-Revised. In: Wilson JP, Keane TM, editors. Assessing psychological trauma and PTSD: a practitioner’s handbook. New York: Guildford Press; 1997. pp. 168–89.
11. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13.
12. Wells TS, Horton JL, LeardMann CA, Jacobson IG, Boyko EJ. A comparison of the PRIME-MD PHQ-9 and PHQ-8 in a large military prospective study, the millennium cohort study. J Affect Disord. 2013;148(1):77–83.
13. Campbell-Sills L, Stein MB. Psychometric analysis and refinement of the Connor–Davidson resilience scale (CD-RISC): validation of a 10-item measure of resilience. J Trauma Stress. 2007;20(6):1019–28.
14. Johnson DC, Polusny MA, Erbes CR, et al. Development and initial validation of the response to stressful experiences scale. Mil Med. 2011;176(2):161–9.
15. Martinsen M, Holme I, Pensgaard AM, Torstveit MK, Sundgot-Borgen J. The development of the brief eating disorder in athletes questionnaire. Med Sci Sports Exerc. 2014;46(8):1666–75.
16. Peric M, Zenic N, Sekulic D, Kondric M, Zaletel P. Disordered eating, amenorrhea, and substance use and misuse among professional ballet dancers: preliminary analysis. Med Pr. 2016;67(1):21–7.
17. Kahn PM, Leon GR. Group climate and individual functioning in an all-women Antarctic expedition team. Environ Behav. 1994;26(5):669–97.
18. Atlis MM, Leon GR, Sandal GM, Infante MG. Decision processes and interactions during a two-woman traverse of Antarctica. Environ Behav. 2004;36(3):402–23.
19. ISAK. International Standards for Anthropometric Assessment. Underdale, SA, Australia; 2001.
20. Reynolds RM, Walker BR, Syddall HE, et al. Altered control of cortisol secretion in adult men with low birth weight and cardiovascular risk factors. J Clin Endocrinol Metab. 2001;86(1):245–50.
21. Kajantie E, Eriksson J, Barker DJ, et al. Birthsize, gestational age and adrenal function in adult life: studies of dexamethasone suppression and ACTH1-24 stimulation. Eur J Endocrinol. 2003;149(6):569–75.
22. Stalder T, Kirschbaum C. Analysis of cortisol in hair–state of the art and future directions. Brain Behav Immun. 2012;26(7):1019–29.
23. Levy JC, Matthews DR, Hermans MP. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care. 1998;21(12):2191–2.
24. Stirrat LI, Walker JJ, Stryjakowska K, et al. Pulsatility of glucocorticoid hormones in pregnancy: changes with gestation and obesity. Clin Endocrinol (Oxf). 2018;88(4):592–600.
25. Kirschbaum C, Tietze A, Skoluda N, Dettenborn L. Hair as a retrospective calendar of cortisol production—increased cortisol incorporation into hair in the third trimester of pregnancy. Psychoneuroendocrinology. 2009;34(1):32–7.
26. Miller R, Plessow F, Rauh M, Gröschl M, Kirschbaum C. Comparison of salivary cortisol as measured by different immunoassays and tandem mass spectrometry. Psychoneuroendocrinology. 2013;38(1):50–7.
27. Gerber M, Brand S, Lindwall M, et al. Concerns regarding hair cortisol as a biomarker of chronic stress in exercise and sport science. J Sports Sci Med. 2012;11(4):571–81.
28. Friedl KE, Moore RJ, Hoyt RW, Marchitelli LJ, Martinez-Lopez LE, Askew EW. Endocrine markers of semistarvation in healthy lean men in a multistressor environment. J Appl Physiol (Bethesda, Md : 1985). 2000;88(5):1820–30.
29. Murphy NE, Carrigan CT, Philip Karl J, Pasiakos SM, Margolis LM. Threshold of energy deficit and lower-body performance declines in military personnel: a meta-regression. Sports Med. 2018;48:1–10.
30. Cadegiani FA, Kater CE. Body composition, metabolism, sleep, psychological and eating patterns of overtraining syndrome: results of the EROS study (EROS-PROFILE). J Sports Sci. 2018;36(16):1902–10.
31. Calbet JA, Ponce-González JG, Calle-Herrero J, et al. Exercise preserves lean mass and performance during severe energy deficit: the role of exercise volume and dietary protein content. Front Physiol. 2017;8:483.
32. Smith GI, Yoshino J, Kelly SC, et al. High-protein intake during weight loss therapy eliminates the weight-loss-induced improvement in insulin action in obese postmenopausal women. Cell Rep. 2016;17(3):849–61.
33. Fortes MB, Diment BC, Greeves JP, Casey A, Izard R, Walsh NP. Effects of a daily mixed nutritional supplement on physical performance, body composition, and circulating anabolic hormones during 8 weeks of arduous military training. Appl Physiol Nutr Metab. 2011;36(6):967–75.
34. Sidossis L, Kajimura S. Brown and beige fat in humans: thermogenic adipocytes that control energy and glucose homeostasis. J Clin Invest. 2015;125(2):478–86.
35. Krief S, Lönnqvist F, Raimbault S, et al. Tissue distribution of beta 3-adrenergic receptor mRNA in man. J Clin Investig. 1993;91(1):344–9.
36. Hoyt RW, Opstad PK, Haugen AH, DeLany JP, Cymerman A, Friedl KE. Negative energy balance in male and female rangers: effects of 7 d of sustained exercise and food deprivation. Am J Clin Nutr. 2006;83(5):1068–75.
37. Tarnopolsky MA. Sex differences in exercise metabolism and the role of 17-beta estradiol. Med Sci Sports Exerc. 2008;40(4):648–54.
38. Wandrag L, Siervo M, Riley HL, et al. Does hypoxia play a role in the development of sarcopenia in humans? Mechanistic insights from the Caudwell Xtreme Everest expedition. Redox Biol. 2017;13:60–8.
39. Potter GD, Skene DJ, Arendt J, Cade JE, Grant PJ, Hardie LJ. Circadian rhythm and sleep disruption: causes, metabolic consequences, and countermeasures. Endocr Rev. 2016;37(6):584–608.
40. Stubbs RJ, Hopkins M, Finlayson GS, Duarte C, Gibbons C, Blundell J. Potential effects of fat mass and fat-free mass on energy intake in different states of energy balance. Eur J Clin Nutr. 2018;72(5):698.
41. Carson JA, Manolagas SC. Effects of sex steroids on bones and muscles: similarities, parallels, and putative interactions in health and disease. Bone. 2015;80:67–78.
42. Ricci MR, Fried SK, Mittleman KD. Acute cold exposure decreases plasma leptin in women. Metab Clin Exp. 2000;49(4):421–3.
43. Iwen KA, Wenzel ET, Ott V, et al. Cold-induced alteration of adipokine profile in humans. Metab Clin Exp. 2011;60(3):430–7.
44. Corr M, De Souza MJ, Toombs R, Williams N. Circulating leptin concentrations do not distinguish menstrual status in exercising women. Hum Reprod. 2011;26(3):685–94.
45. Cadegiani FA, Kater CE. Hormonal aspects of overtraining syndrome: a systematic review. BMC Sports Sci Med Rehab. 2017;9:14.
46. Loucks AB, Thuma JR. Luteinizing hormone pulsatility is disrupted at a threshold of energy availability in regularly menstruating women. J Clin Endocrinol Metab. 2003;88(1):297–311.
47. Hulmi JJ, Isola V, Suonpää M, et al. The effects of intensive weight reduction on body composition and serum hormones in female fitness competitors. Front Physiol. 2017;7:689.
48. Meeusen R, Duclos M, Foster C, et al. Prevention, diagnosis, and treatment of the overtraining syndrome: joint consensus statement of the European College of Sport Science and the American College of Sports Medicine. Med Sci Sports Exerc. 2013;45(1):186–205.
49. Bassett SM, Lupis SB, Gianferante D, Rohleder N, Wolf JM. Sleep quality but not sleep quantity effects on cortisol responses to acute psychosocial stress. Stress (Amsterdam, Netherlands). 2015;18(6):638–44.


Supplemental Digital Content

Copyright © 2018 by the American College of Sports Medicine