Natural bodybuilding is a sport that blends physical skill with artistic presentation (1). Bodybuilding competitions are judged on an athlete’s muscular size, symmetry, definition, and posing skills (2). To obtain the desired muscle definition, athletes undergo a period of calorie restriction and increased activity energy expenditure to lose fat mass, coupled with nutrition and resistance training practices to preserve skeletal muscle (3). After an extended period of energy restriction, athletes often increase their calorie, carbohydrate, and sodium intake several days before competition to increase muscle glycogen stores and intracellular fluid to accentuate muscle volume, known as “peaking” (4). Collectively, these strategies encompass a bodybuilder’s contest preparation (“contest prep”).

Natural bodybuilding is unique in that athletes willingly subject themselves to semi-starvation to achieve their contest condition, cognizant of the disruptions to physiological and psychological function, which are largely “adaptable” states of low energy availability (EA; transient, reversible signs of body system suppression with minimal long-term adverse health effects) (5, 6), but can manifest into “problematic” low EA (signs of body system suppression with long-term adverse health effects) in some bodybuilders (7) and other physique competitors (8).

Since the first prospective studies associated low EA with reduced triiodothyronine (T3) and luteinizing hormone (LH) dysregulation in healthy, sedentary women three decades ago (9, 10), mirroring observations seen in amenorrhoeic female athletes (11, 12), a growing body of research has surfaced on indicators of low EA in athletes. Consolidating low EA research with athlete testing, validation, and usability, the IOC REDs Clinical Assessment Tool Version 2 (IOC REDs CAT2) was developed to detect signs of problematic low EA in athletes (13).

Two primary indicators of low EA in Step 2 of the IOC REDs CAT2 for males are clinically low and sub-clinically low (within the lowest 25% quartile of the reference range) total or free T3 and testosterone (14). During chronic states of energy restriction in natural bodybuilders, T3 and testosterone levels reach clinically low values (5, 6); however, both can recover to within normal physiological ranges and baseline values in 1–3 months following an increase in energy intake and body mass (BM) post contest prep (5, 6). Experimental research in a male combat athlete and soldiers following a chronic state of energy restriction has shown that energy repletion that far exceeds total daily energy expenditure rates restores T3 and testosterone levels to within normal physiological ranges and baseline values within several days to a week, albeit with pronounced increases in BM (15, 16).

Although the diagnostic measures of T3 and testosterone serve as primary indicators of problematic low EA in male athletes, the measurements themselves adapt quickly to changes in energy intake and EA. While substantial energy repletion intakes and rapid increases in BM are practiced by combat and physique athletes following the termination of a phase of energy restriction (16–18) they do not reflect the acute energy repletion (“refeed”) practices of physique athletes during continued energy restriction (19, 20) nor the modest fluctuations in energy balance and EA of other athletes between training (and matches) and rest days (21–24).

A lack of research exists on the transient nature of T3 and testosterone levels after short-term, modest energy repletion following a period of energy restriction. Understanding how T3 and testosterone levels respond to acute changes in energy intake following a phase of energy restriction and low EA would clarify their responsiveness and assist physicians in accurately timing and interpreting these diagnostic measurements of problematic low EA in certain athletes. In addition, such insights would help discern whether the dietary practice of refeeding (2 days) in bodybuilding is efficacious for attenuating endocrine adaptations during ongoing energy restriction (25).

Reductions in T3 have been shown to coincide with stepwise suppressions in REE and menstrual disturbance (26), the latter being a primary marker of low EA in women (27). The assessment of low EA using REE is commonly depicted as a REE_ratio of measured REE (mREE) via indirect calorimetry to a predictive estimate of REE (pREE) with a cut-off ratio of <0.90 (mREE/pREE) (27). However, REE remains a “potential” marker in the IOC REDs CAT2 assessment model due to a lack of consistency in the literature (13). This discrepancy may be partly due to predictive equations for determining the REE_ratio being non-specific to the population studied, exercise being performed close to measurement, and the lack of weight stabilisation before an intervention (27).

Therefore, this case study serves as a lens to assess the effects of two primary indicators of low EA in a male bodybuilder during chronic energy restriction and 2 days of modest energy repletion in the athlete’s peaking phase. It also provides further information on REE and REE_ratio during chronic energy restriction using an athlete-specific predictive equation and adopting recommended pre-assessment procedures. Lastly, we offer insight into field-based measurements accessible to the research-active practitioner on aspects of physiology, psychology, and strength performance during a chronic phase of energy restriction. Collectively, our findings offer further empirical evidence to contribute to the shortage of research on low EA in male athletes and assist in contextualizing low EA markers using tools such as the IOC REDs CAT2.

Data are presented as arithmetic means and standard deviations (±). Data was analyzed using Microsoft Excel (version 2021, Microsoft Corporation, WA, USA). Normality was determined using the Shapiro–Wilk test. Pearson’s correlation (normal distribution) and Spearman’s rank correlation (non-normal distribution) were used to test associations for oral temperature, resting heart rate and BM. Statistical significance was set at an alpha of p < 0.05. As internal replicate measurements of all endocrine markers were not performed, the inter-assay CoV% provided by Roche Diagnostics for the cobas e 602 analyzer was used as a proxy measure. The typical error (TE) was estimated to determine the variability of each endocrine measure and calculated as the inter-assay CoV% multiplied by the mean measurements by the assay manufacturer divided by 100. The TE multiplied by 2 was used to discern a value indicative of a meaningful change.

This case study assessed the nature of two primary markers of low EA (free T3 and TT) following 2 days of modest energy repletion. Moreover, we provide further insight into endocrine, physiological, psychological, and strength performance outcomes following a prolonged period of BM loss. The athlete’s free T3 and TT fell into clinically low (2.7 pmol/L⁻¹) and sub-clinically low (9.1 nmol/L⁻¹) ranges by week 18, respectively. Previous research on male bodybuilders, a combat athlete and military personnel during chronic phases of energy restriction have demonstrated similar reductions (5, 6, 15, 16). Considering the secretion of LH regulates the production of testosterone in males via the anterior pituitary gland secondary to the pulsatile release of gonadotropin-releasing hormone (GnRH) within the hypothalamus, it was surprising that LH, albeit trending downwards, remained largely unchanged between baseline and week 18 compared to TT. Similarly, we anticipated free T4 production to decrease per free T3 and REE and SHBG production to increase per a reduction in TT; however, neither demonstrated meaningful changes. Discrepancies between LH, SHBG and testosterone have been previously documented in male athletes (37, 38) and could have been due to the small number of measurements taken. Moreover, discrepancies in T4 with decreases in T3 have also been reported in the literature (9, 39). However, LH levels from baseline to the athlete’s third measurement (2 days of energy repletion) fell by −23.7%, in conjunction with a −50.0% reduction in TT, suggesting that a reduction in GnRH mediated the decrease in TT.

After 2 days of modest energy repletion, free T3 increased by 18.5%, whereas TT levels continued to decrease by −20.9%. In a recent study of a male combat athlete, a two-day ad-libitum increase in calories (64–89 kcal.kg FFM⁻¹ BM.d⁻¹) led to clinically low testosterone levels increasing more than two-fold, returning close to baseline values in a week (16). Similar increases in free testosterone have been demonstrated in explorers 3 days after completing an 850-km cross-country skiing expedition to the North Pole. While the explorer’s energy repletion intake was not measured in the 3 days following the expedition, a 5 kg weight regain at 5 days indicates a pronounced calorie increase above energy balance (40). The discrepancy between our results and those mentioned is likely due to both studies’ energy repletion intake exceeding total daily energy expenditure requirements and thus reaching optimal EA values to support the restoration of endocrine function. In contrast, if we consider the athlete’s pronounced reduction in REE, our intake likely achieved close to energy balance but below adequate EA to support the restoration of the hypothalamic–pituitary-gonadal axis. Thus, we can speculate that the continued reduction in TT following 2 days of modest energy repletion could suggest that the hypothalamic–pituitary-gonadal axis requires a greater energy and/or carbohydrate intake to respond during periods of short-term energy repletion following energy restriction compared to the hypothalamic–pituitary-thyroid axis; however, due to the study design and limitations described below, this warrants further study.

A reduction in REE appears to be a consistent finding in case studies during contest prep, with reductions in REE ranging from −179 kcal to roughly ~1,000 kcal (5, 6, 20, 30) and a decrease in RER (30). While the use of a predictive equation specific to physique athletes provided a close REE_ratio at baseline (0.99) and detected energy deficiency at week 18 (0.78) with the cut-off ratio of <0.90 (mREE/pREE) (27), due to the substantial decrease in the athlete’s REE, common predictive equations used in athletes such as the Harris-Benedict (41) (0.81) and Ten-Haaf (42) (0.75) equations also produced an mREE/pREE ratio of <0.90. From week 1–17, the athlete averaged a weekly BM loss of −0.30 kg, which is lower than the average (−0.42 kg) of other male physique athletes during contest prep (43) but aligns with current evidence-based recommendations (32). Notably, the athlete’s lowest weight was recorded at week 15 (76.3 kg); however, during weeks 16–18, the athlete complained of gastrointestinal (GI) symptoms (bloating and constipation), which likely contributed to the athlete’s lack of BM loss in the final 2 weeks. Upon highlighting these GI complaints, a review of the athlete’s dietary intake was performed to see if any potential dietary culprits of GI distress could be identified (e.g., uncharacteristically high amounts of FODMAP-rich foods) (44), but no abnormal changes from the athlete’s previous 15 weeks of dietary intake had occurred. GI disturbance has been implicated in the REDs physiological model and reported in male athletes with secondary exercise dependence and disordered eating (14, 45); however, further studies are warranted in male athletes. In the final weeks leading up to the competition, the athlete moved house while maintaining his case study, video blogging, and work commitments. In the athlete’s video blog during this time, he mentions stress may have been the cause of his GI disturbance, which is empirically supported (46).

The athlete’s oral temperature decreased throughout the intervention period in accordance with BM losses (r = 0.674, p = 0.002; Figure 1b), as per previous research during a period of energy restriction (36). A similar relationship was demonstrated with the athlete’s RHR (r = 0.560, p = 0.016), which was reduced by 9 beats.min⁻¹ by week 16, akin to previous case reports during contest prep (6, 30). As shown in Figure 1d, the athlete’s mood state remained largely undisturbed during contest prep. However, vigor and anger increased slightly above average in the final 2 weeks. Mood disturbances have been previously reported during contest prep (7), with pronounced increases reported in one physique athlete in the final stage before competition (6). While a heightened preoccupation with food could have been partly causal of the athlete’s mood disturbance, per previous research (7), it may have also been a result of changes in his personal life noted earlier, and the emotional toll on his relationships—both of which were mentioned in the athlete’s video blogs. The athlete demonstrated decreases in absolute 1 RM in the squat, bench, and deadlift from baseline by −10.0%, −7.4%, and −8.3%, while their relative 1 RM decreased by −5.4%, −2.6%, and −3.6%, respectively. Notably, these changes are within the CoV (0.5–12.1%) of a recent systematic review on the retest reliability of the 1RM test (47); thus, some of the reduction in strength could simply be a product of day-to-day variability.

While our study measured indicators of EA, we did not directly measure EA due to discrepancies in skinfold body fat prediction equations (29) and dietary energy intake and exercise energy expenditure both being highly susceptible to error (48, 49). Moreover, it is likely that these measurements alone are not a true reflection of an athlete’s EA, considering non-exercise activity thermogenesis is not factored into the equation and can surpass exercise activity expenditure values in certain athletes (50). Several limitations are worthy of note with this research, which the reader must consider when evaluating the finding’s validity and translatability to practice. The athlete was recruited for the intervention just 2 weeks before its commencement, which prevented test re-test reliability measurements for oral temperature, morning RHR, BRUMS, and 1RM testing. Furthermore, the athlete’s baseline BRUMS assessment was performed on just 1 day, which could have been influenced by transient external factors and the reason for the discrepancy between scores at baseline and during the intervention. Internal quality control metrics of the endocrine measurements were not performed; thus, the inter-assay CoV was derived from the assay manufacturer as a proxy measure. The Metallyzer 3B (Cortex) system used to assess RER at rest in our study has been shown to overestimate carbohydrate oxidation by 53% and underestimate the energy derived from fat by 25%, albeit in an exercise context (51). While our results indicate a shift towards greater fat oxidation similar to previous research on a physique athlete during contest prep (30), caution is advised when interpreting the RER values due to the potential for measurement inaccuracies. Furthermore, the athlete’s free T3 levels were assessed using the immunoassay method, which, compared to the gold-standard method of liquid chromatography–tandem mass spectrometry, frequently overestimates free T3 levels at the low reference interval (52). Lastly, the athlete’s TT levels were measured at three single time points, when it is recommended to take two consecutive measures at each time point due to day-to-day variation (53).

Our case study offers insight into the nature of two primary indicators of low EA in males (free T3 and TT) as per the IOC REDs CAT2 during chronic energy restriction and a short-term period of modest energy repletion. Moreover, this study offers further insight into the effects of “potential indicators” of low EA (e.g., REE) during 18 weeks of contest prep (13). The athlete adopted evidence-based weight-loss recommendations for athletes yet incurred declines in metabolic and reproductive hormones, resulting in a risk status of “Moderate to High” (Orange) according to the IOC REDs CAT2 Severity/Risk Assessment Tool (13, 14). The extent to which these alterations resulted from the athlete’s low EA load or degree of leanness was not determined and requires further insight; however, the results are pertinent to practitioners working with athletes adopting similar weight-loss approaches. The implications of the transient nature of free T3 and the unresponsiveness of TT following a 2 days modest energy repletion warrant further study and could have a meaningful impact on assessing low EA in males. We successfully documented endocrine, physiological, psychological, and performance indicators while the athlete adopted evidence-based guidelines for contest prep. Importantly, our case study was undertaken using assessment methods accessible to many research-active practitioners, which we hope will allow more practitioners to report their inductions from practice to inform future research, bridging the gap between science and practice (54).

The athlete felt he achieved his best contest prep condition and won his physique category. Reflecting on the intervention the day before the competition, the athlete questioned whether 18 weeks of energy restriction, necessitating constraints on social events, travel, and causing a strain on his close relationships, coupled with an ongoing heightened fixation with food, was worth the reported changes in body composition. Several days after the show, the athlete mentioned that his motivation, as a result of no longer working towards the competition, had noticeably reduced. A week post-contest prep, the athlete increased his weight by ~1.5 kg but was not preoccupied with the rise in BM, unlike previous post-contest prep experiences. He credited this change in perspective to viewing his increased BM as an indicator of returning to a condition commensurate with achieving his best lifting performances.