Subjects
A detailed description of the design, methods, subject selection criteria, and baseline characteristics for the EROS study is available in a depository (
https://osf.io/bhpq9/). The study was approved by the ethical committee of the Federal University of São Paulo (approval number: 1093965). For the analysis in the present article, we included two groups: ATL and NPAC. Selected results from both the EROS-HPA axis [
20] and the EROS-STRESS [
21] arms of the EROS study were included in this analysis.
We recruited subjects through social media (Facebook and Instagram); the aim of the recruitment process was to include ATL and NPAC who were eligible for the study. A preliminary analysis of the candidates was done by e-mail correspondence, and included questions regarding age, sex, and approximate body weight and height. Based on the candidate’s responses, their approximate body mass index (BMI) was calculated. If no exclusion criteria were identified, an individual interview was then scheduled.
At the interview, body weight and height were verified using high precision weight and height scales. Questions regarding other conditions, use of medications or hormones, and characteristics of the sport (in the case of athletes) were also asked, and age was confirmed by verification of an identity card. All subjects had to fulfill the following inclusion criteria at this point: male; aged between 18 and 50 years; BMI between 20.0 and 32.9 kg/m2; absence of previous psychiatric disorders and use of centrally-acting drugs; and absence of any hormonal therapy in the previous 6 months.
The following additional inclusion criteria regarding training aspects were required from all ATL regarding training level: exercise at least four times a week, for a total of > 300 min a week, with moderate-to-vigorous training intensity (self-perception comparing own training with that of others, based on the Talk Test), and continuous training for the current sport(s) for at least 6 months, without interruption for > 30 days. The quantification of training load was recorded by the coach of each athlete, but not in a systematic way, as the recruitment occurred transversally in order to collect real-life training data (i.e., it was not controlled). We required a minimum amount of physical activity for potential exercise-induced adaptions.
NPAC were required to: (
1) fulfill the initial inclusion criteria, (
2) be sedentary (without any physical activity, including light exercises) for at least 3 years, and (
3) no history of exercise that would fulfill the criteria for ATL.
Candidates who fulfilled criteria were selected. After signing a written consent, the remaining subjects underwent biochemical examination to exclude confounding disorders and prevent inclusion of subjects with altered basal and stimulated hormone levels due to inflammation, infection, kidney disease, lipid metabolism abnormalities, vitamin deficiencies, or obvious hormonal dysfunctions. In ATL, exams were performed from 36 to 48 h after the last training session. The biochemical inclusion criteria used is listed in Table
1.
Table 1
Biochemical inclusion criteria for the EROS study
Ultrasensitive C-reactive protein (CRP) | <3 mg/dL | Latex-intensified immunoturbidimetry |
Erythrocyte sedimentation rate (ESR) | <25 mm/h | Automated spontaneous sedimentation method |
Creatinine (and TFG) | <1.5 mg/dL (> 60 mL/min) | Jaffe enzymatic assay |
Hematocrit | 36–54% | Automated assay |
Neutrophils | 1000–9000 /mm3 | Automated assay |
Creatine kinase (CK) | < 5000 U/L | Calorimetric activity assay; International Federation of Clinical Chemistry |
Alanine aminotransferase (ALT) | <50 U/L | Calorimetric activity assays |
Aspartate aminotransferase (AST) | <50 U/L | Calorimetric activity assays |
Ferritin | 20–1000 ng/dL | Chemiluminescence assay |
Vitamin B12 | >180 pg/mL | Chemiluminescence assay |
Fasting glucose | <100 mg/dL | Enzymatic assay of hexokinase |
Total testosterone | >200 ng/dL | Chemiluminescence assay |
TSH | <5 μIU/mL | Chemiluminescence assay |
Design
After the selection process, the subjects underwent basal biochemical tests, hormonal responses to stimulation tests, and evaluation of sleep, psychological, social and eating patterns, and analysis of the body metabolism and composition, as part of the different arms of the EROS study [
20‐
23], with a maximum interval from the beginning to the end of the tests of 10 days. All tests and the collection of blood for analysis were medically supervised. Blood or plasma collection tubes were checked before and after each collection to ensure that an appropriate tube type was used for the biomarker and that each subject was properly identified. Following collection, the tubes were immediately centrifuged or analyzed to prevent loss of quality of the collected material. For the present study, we analyzed the hormonal responses to stimulation tests.
We evaluated the peripheral component of the hypothalamic-pituitary-adrenal (HPA) axis with a CST, which directly evaluates adrenal responses to synthetic adrenocorticotropic hormone (ACTH) [
24], and whose impaired responses are indicative of primary adrenocortical dysfunction. Then we evaluated the central component of the HPA axis with an ITT, which evaluates the integrity of the hypothalamus and pituitary [
24], and whose normal response requires absence of dysfunctions in all levels of the HPA axis (hypothalamus, pituitary, and adrenals). Whenever adrenal responses are normal to the CST, any abnormality observed in an ITT must be located in the hypothalamus or in the pituitary [
24]. We also evaluated growth hormone (GH) and prolactin responses to the ITT, as well as glucose changes and clinical behavior during a hypoglycemic episode. These tests were part of the “EROS-HPA axis” [
20] and “EROS-Stress” [
21] arms of the EROS study.
Methodology
For the CST, subjects were required to fast for 8 h and to have had the last training session at least 72 h (3 days) prior to the test, and to arrive at the laboratory at 7:30 h in the morning. They sat in blood-drawing chairs and rested for 30 min. Ten mL of blood (divided into two EDTA tubes) was collected before and 30 and 60 min after an intravenous administration of 250 μg of cosyntropin (as recommended by the guidelines of Endocrinology societies [
24]) for analysis of cortisol.
Subjects underwent the ITT 48 h after the CST, following an 8-h fasting and a minimum of 120 h without exercising. Subjects arrived at the laboratory at 7:30 h in the morning, were seated in a blood-drawing chair, and rested for 30 min. Then, 0.1 IU/kg of regular insulin was administered intravenously after blood collection (10 mL in ethylenediaminetetraacetic acid (EDTA) and plasma tubes) at time zero (baseline). Capillary glucose was checked every 5 min from time 10 min after insulin administration, or whenever subjects reported symptoms. Blood for time one was collected when: 1) capillary glucose was < 30 mg/dL without symptoms; 2) subjects classified symptoms of hypoglycemia as moderate to severe (
5‐
10) regarding either adrenergic (shakiness, cold sweating, heart palpitations, or pallor) or neuroglycopenic (sleepiness, mood changes, or unrest) symptoms, or both; or 3) if capillary glucose was < 45 mg/dL in presence of any symptom. If after 40 min none of these three criteria was achieved, an extra 0.05 IU/kg of regular insulin was administered intravenously; and again after an additional 40 min, if none of the criteria was achieved. Finally, if hypoglycemia did not occur, the subject would be withdrawn from the study due to likely insulin resistance (which makes it unfeasible to perform a proper ITT test). However, none of the patients required a third dose of insulin.
After time one blood collection, 10 mL of 50% glucose solution was infused intravenously and high-glycemic index and pure carbohydrate food (lemon or strawberry dairy-free sorbet, Diletto, Brazil) was offered ad libitum. Ten mL of blood was collected again, 30 min after the hypoglycemic episode (time two). In all blood samples we determined cortisol, ACTH, GH, prolactin and glucose, as well as the absolute ACTH/cortisol ratio at all times during ITT [
24‐
26]. During the ITT we evaluated the time-to-hypoglycemia (min) since insulin administration and self-reported intensity of adrenergic and neuroglycopenic symptoms on a scale of 0 to 10 (0 = asymptomatic, 10 = severe symptoms). We did not perform the 60 min for all athletes, as the first round (a “pilot” evaluation, with three sedentary and three healthy athletes) did not disclose differences for any of the hormones between 30 and 60 min after hypoglycemia. Also, protocols for ITT admit variations regarding time for the blood collection, and whether the time for blood collect depended the hypoglycemic episode or not.
Due to the risk of severe hypoglycemia during ITT, subcutaneous glucagon pens were always available (GlucaGen HypoKit, 1 μg, NovoNordisk), as well as 20 mL-syringes containing 50% glucose solution and an automated external defibrillator (AED).
Basal and hypoglycemia-induced serum cortisol, plasma ACTH, serum GH and serum prolactin levels were determined by commercially available electrochemiluminescence assays, that were previously validated, standardized and tested. The detection limits for ACTH and GH were 5.0 pg/mL and 0.05 μg/L, respectively, but there was no minimal analytical limit for the other markers. The intra- and inter-assay coefficients of variability of all the biochemical markers measured in all arms of the EROS study were below 3.0 and 3.5%, respectively.
We evaluated basal and hypoglycemia-induced and absolute changes in the levels of cortisol, ACTH and prolactin during the ITT (from time zero to time two); GH changes were not determined because of its wide pulse amplitude. Mean time-to-hypoglycemia and intensity of adrenergic and neuroglycopenic symptoms were also compared between groups.
In addition, a 7-day dietary record with specific calorie and macronutrient account, and self-reported sleeping patterns, social and psychological characteristics, basal muscular, inflammatory, immunologic and hormonal parameters, and body composition and metabolism were evaluated in all selected participants.
Post-hoc joint analysis
We joined the responses to CST and ITT, obtained from the EROS-HPA axis and EROS-STRESS arms of the EROS, including analysis.
We did not compare the magnitude of the responses between GH, prolactin, cortisol, and ACTH, as hormones have distinct behaviors with respect to the physiological amplitude of responses to stimulations. Instead, we compared differences between ATL and NPAC responsiveness within each hormone, through the ratios between hormonal responses in athletes and non-athletes.
We analyzed the context of the findings, from which we proposed the new hypothesis of hormonal conditioning process that athletes undergo.