https://orcid.org/0000-0001-8261-3457
Abstract
Although gonadal axis dysregulation from energy deficit is well recognized in women, the effects of energy deficit on the male gonadal axis have received much less attention.
To identify relevant articles, we conducted PubMed searches from inception to May 2021.
Case series and mechanistic studies demonstrate that energy deficit (both acutely over days or chronically over months) either from inadequate energy intake and/or excessive energy expenditure can lower serum testosterone concentration as a result of hypothalamic-pituitary-testicular (HPT) axis dysregulation in men. The extent to which this has clinical consequences that can be disentangled from the effects of nutritional insufficiency, concomitant endocrine dysregulation (eg, adrenal and thyroid axis), and coexisting comorbidities (eg, depression and substance abuse) is uncertain. HPT axis dysfunction is primarily the result of loss of GnRH pulsatility resulting from a failure of leptin to induce kisspeptin signaling. The roles of neuroendocrine consequences of depression, hypothalamic-pituitary-adrenal axis activation, proinflammatory cytokines, Ghrelin, and genetic susceptibility remain unclear. In contrast to hypogonadism from organic pathology of the HPT axis, energy deficit-associated HPT dysregulation is functional, and generally reversible by restoring energy balance.
The clinical management of such men should aim to restore adequate nutrition and achieve and maintain a healthy body weight. Psychosocial comorbidities must be identified and addressed. There is no evidence that testosterone treatment is beneficial. Many knowledge gaps regarding epidemiology, pathophysiology, and treatment remain and we highlight several areas that require future research.
Reproductive function is tightly coupled to energy availability and so it is not surprising that states of energy deficit are associated with dysregulation of the hypothalamic-pituitary-testicular axis (HPT) reflected in postpubertal men by a reduced serum testosterone concentration. Energy deficit may be absolute when a reduction in energy intake leads to progressive weight loss, or relative when increased energy expenditure is not matched by sufficient energy intake to maintain body fat stores, even if weight is not lost (for example, in the case-intensive engagement in resistance exercise to increase muscle mass). The effects of energy deficit on the reproductive function, although well recognized in women, have received relatively scant attention in men. In this minireview, we discuss clinical and pathogenetic aspects of HPT axis dysregulation associated with absolute or relative energy deficit. Typically, the HPT axis dysregulation is functional (ie, not from organic HPT axis pathology, such as in the case of, for example, Klinefelter syndrome or a pituitary adenoma), and therefore potentially reversible with restoring the energy deficit and achieving and maintaining a healthy body weight. To identify relevant articles, we conducted PubMed searches from inception to May 2021, using the keywords testosterone, androgens, hypogonadism, nutrition, anorexia, anorexia nervosa, exercise, energy deficit, reversible energy deficit in sport, caloric restriction, and starvation. We focussed on human studies, citing key preclinical evidence where equivalent data from men are lacking.
Effects of an Absolute Energy Deficit on the HPT Axis
HPT axis dysregulation from inadequate energy intake has been described in animal models and men as early as the 1930s. Studies in starved male rodents demonstrated low pituitary gonadotrophic activity, and Mulinos and Pomerantz introduced the term “pseudohyphysectomy” to describe the effects of starvation on male reproductive function (1). An absolute energy deficit may result from lack of access to food, or illness such as gastrointestinal disease leading to poor intake or malabsorption, cancer, and cardiac disease. A reduction in energy intake may be behavioral to address a distorted perception of body appearance (eg, anorexia nervosa). Effects of caloric restriction on the HPT axis can be acute, occurring within days, before changes in body composition have occurred, or chronic, occurring within weeks to months.
Effects of an Acute Energy Deficit
Studies in healthy young (n = 8, age 20-32 years) men between 84% to 119% of height-adjusted average body weight have reported that fasting (water only) reduces serum testosterone and LH within 48 hours (2). Total testosterone was reduced from 13.8 nmol/L (398 ng/dL) to 10.9 nmol/L, and LH from 2.9 to 1.1 IU/L, with an accompanying decrease in LH pulse frequency (2). In healthy young (n = 6-8, age 19-35 years) men, within 25% of normal body weight undergoing 3 to 5 days of fasting (water only), both serum total testosterone and free testosterone (assayed by solid phase radioimmunoassay) declined by 50%, and LH by 30%, with a 3-fold decrease in the calculated LH secretion rate because of reductions of both the number of LH secretory bursts and the mass of LH secreted per burst. These biochemical changes could be completely prevented by pulsatile GnRH infusion (3, 4), suggesting that acute fasting inhibits hypothalamic GnRH secretion. These studies demonstrated that caloric restriction can suppress the central drive to the HPT axis before a substantial amount of body weight loss has occurred. Interestingly, Bergendahl et al reported that healthy older men are less susceptible to the HPT axis-suppressive effects of short-term fasting (water only for 3.5 days) compared with young men (5). Although after fasting in young men (n = 8, mean age 28 years, body mass index [BMI] 26 kg/m2), total and free testosterone (measured by radioimmunoassay) fell by 46% and 40%, and 24-hour endogenous LH production rate by 2.1-fold, there were no significant fasting associated changes in serum testosterone and LH production rate in older men (n = 8, mean age 67 years, BMI 26 kg/m2) (5). That younger men are more susceptible to fasting-induced HPT axis suppression than older men is in contrast with most other evidence suggesting that older age increases the risk of HPT axis dysfunction in response to physiologic stimuli and pathogenic stressors (6, 7). This may be, in addition to the observation that young men are more susceptible to the HPT axis-suppressive effects of inflammation than older men (see Proinflammatory Cytokines) (8), 1 of the contributing reasons why energy deficit-associated functional hypogonadism has been predominantly reported in younger men (9, 10).
Effects of a Chronic Energy Deficit
Although most studies have reported short-term (days) effects of energy deficits on the male HPT axis, relatively less is known about long-term effects of caloric restriction in men. An early case series of 28 Indian men (mean age 41 years) with severe chronic protein-calorie malnutrition (mean body weight 33 kg), but no evidence of other clinical disease, reported clinical features of hypogonadism (diminished body hair and beard growth and small soft testes) accompanied by low total testosterone (reduced by ~50% compared with healthy controls). Interestingly, in some men serum LH was elevated and inappropriately in others (11). However, in more recent case series (n = 3-31) men with severe chronic energy deficit have both low serum testosterone and inappropriately low serum LH (9, 12-16), suggesting that, similar to acute caloric restriction, chronic caloric restriction leads to central suppression of the HPT axis. Of note, in contrast to acute caloric restriction, the responsiveness of the HPT axis to exogenous GnRH in men with chronic anorexia is impaired (12, 13). However, GnRH responsiveness normalizes with weight gain, suggesting that, as with acute caloric restriction, HPT axis changes with chronic caloric restriction are functional, and reversible with restauration of the energy deficit (see Management).
More recent studies have examined the effects of more modest degrees of energy restriction on the HPT axis. One cross-sectional study comparing 24 middle-aged men (mean age 52 years) practicing long-term (average 7.5 years) caloric restriction to age-matched controls eating an ad libitum diet reported that, compared with men eating ad libitum, calorie-restricted men had a significantly lower BMI (19.1 vs 26.7 kg/m2), lower body fat (9.7% vs 23.2%), lower serum total testosterone (12.0 vs 17.6 nmol/L), and higher SHBG (23.8 vs 14.2 nmol/L). Free testosterone concentrations were not reported in this study (17). In a randomized controlled trial (RCT) of healthy nonobese adults subjected to 2 years of 25% caloric restriction achieving an 11.5% weight loss, a secondary analysis among men (n = 66) showed no difference in serum total testosterone and LH concentrations between the calorie restricted and ad libitum groups at either at 12 months or at study end. However, men in the calorie-restricted group had significantly higher concentrations of SHBG at both time points, leading to a significantly lower calculated free testosterone at 12 months (between-group difference 100 pmol/L [2.89 ng/dL]) but not at study end (18). Of note, men in the caloric-restricted group reported improvements in sexual drive and arousal, suggesting that the modest changes in serum free testosterone were biologically insignificant (18).
The effects of intermittent fasting, a popular dietary approach used for weight loss and overall health, on the HPT axis have been addressed in very few studies, and only in combination with exercise. In an 8-week study, 34 trained athletes undergoing a standardized exercise program were randomized to either time-restricted feeding (100% of energy needs consumed within 8 hours) or kilocalorie- and macronutrient-matched normal diet (consumed during waking hours). Serum testosterone, free triiodothyronine and IGF-1 decreased in the time-restricted feeding, but not in the normal diet group (19). However, in a similarly designed 4-week study, with the exception that men were subjected to a 25% caloric deficit, testosterone significantly decreased in both the time-restricted feeding and normal diet groups (20).
In summary, the findings suggest that both acute and chronic caloric restriction suppress the HPT axis at the central level, and that the severity of HPT axis suppression correlates with the severity of the caloric deficit.
Men affected by chronic energy deficiency typically have a low BMI. Unlike in women, where loss of menstrual function is obvious and prompts action, dysregulation of the HPT axis under these circumstances in men may go unnoticed. Systematic studies assessing the prevalence of HPT axis dysregulation associated with energy deficit are lacking, and the condition is almost certainly underrecognized and underreported. A 2019 review described clinical features in 23 men with HPT axis dysregulation from energy deficit, largely because of reduced energy intake, who were identified mostly from single case reports (10). Common clinical features included low libido, fatigue, bradycardia, and, in some cases, reduced exercise performance and stress fractures. Median age (range) was 20 years (16-33), BMI 15.9 kg/m2 (12.5-20.5), and total testosterone 3 nmol/L (0.6-21.3 mmol/L) (10). In a case-control study of 31 men with anorexia nervosa aged 21 ± 1 years and a BMI of 15.8 ± 0.4 kg/m2, total testosterone was 9.2 nmol/L, compared with 23.5 nmol/L in healthy (BMI 22.9 ± 0.5 kg/m2) age-matched controls (15).
Overall, the available evidence suggests that energy deficit from reduced energy intake is sufficient to cause severe HPT axis suppression; in such men, BMI is frankly low (<20 kg/m2). Symptoms such as fatigue, flat mood, and decrease in motivation or exercise performance may dominate, and the decrease in or loss in sexual desire may not always be mentioned. Such men not uncommonly display multiple endocrine abnormalities, including dysregulation of the adrenal, thyroid and GH/IGF-1 axes (10), and there is a high prevalence of comorbid depression (21). The extent to which symptoms reflect these other disorders is not known. In contemporary studies, testicular volume is commonly normal (9) (unless exogenous androgens are abused), although historical studies in men with severe energy deficit have reported testicular atrophy (11). Total testosterone concentrations are typically frankly low <7 nmol/L (9, 10, 22) and can be near castrate, with inappropriately low to low normal LH (10, 22). Inhibin concentrations have been reported to be low (9) or normal (15); however, whether fertility is compromised is not well established, and data regarding sperm analyses are extremely limited. In a contemporary series of 10 men with energy deficit-associated HPT axis suppression, of the 3 men who could produce a semen sample for analysis, all samples showed a normal sperm count (9). Clearly, the effects of energy deficit on spermatogenesis and fertility require further study.
Effects of a Relative Energy Deficit on the HPT Axis: Role of Excessive Exercise
The HPT axis suppressive effects of reduced energy intake are exacerbated by increased energy expenditure (ie, excessive exercise), emphasizing the concept that the net energy deficit is the key driver of HPT axis suppression (23). Although in some experimental settings, acute exercise is associated with increases in serum testosterone (for review, see (24)), HPT-axis suppressive effects of excessive exercise were recognized in men in the 1970s (25). This led to the concept that, as in women, inadequate energy intake and excessive energy expenditure (ie, excessive exercise) have additive effects on gonadal axis suppression in men. In 2014, the International Olympic Committee replaced the term “female athletic triad” with the gender-neutral term “relative energy deficit in sport” (26).
One of the early studies assessing the effects of acute strenuous exercise was conducted in 14 men (aged 27-58 years) participating in a noncompetitive marathon and reported that, compared with baseline, serum testosterone decreased by 60% when remeasured after completion of the run (25). In a study of 14 well-trained male cyclists who completed a 1230-km ultra-endurance cycling event, serum testosterone fell by -67% ± 18%; in this small study there was no correlation between the testosterone decline and the estimated energy balance, which ranged from a 11 000-kcal deficit to a 3500-kcal surplus (27). Similar reductions of serum testosterone after excessive endurance exercise have been reported, for example, following participation in ironman triathlon (28) and ultra-marathon events (29). However, reduction of serum testosterone following strenuous endurance exercise are relatively short-lived, with recovery to baseline concentrations within days after the exercise event (28).
Chronic extensive exercise has also been linked to HPT-axis suppression in men. In an early cross-sectional study, men running at least 64 km per week (n = 31) had significantly lower total (by ~20%) and free calculated testosterone (by ~30%) compared with sedentary controls (n = 19) of similar age and weight (30). In a controlled study comparing highly trained men (n = 6) running 125 to 200 km per week, although serum testosterone concentrations were similar to those of healthy age-matched controls (n = 13) running less than 5 km per week, runners had, compared with controls, significantly reduced spontaneous LH pulse frequency, amplitude, and reduced LH responses to exogenous GnRH (31). Whether the HPT-axis suppressive effects of excessive exercise led to symptomatic androgen deficiency is not well established. In a case control study, long-distance runners (n = 9) had not only lower serum testosterone (9.2 nmol/L vs 16.2 nmol/L in sedentary controls), but also a higher burden of androgen deficiency-like symptoms, assessed by the nonspecific aging male symptom score (32).
Additional stressors, such as sustained workload, sleep deprivation, and thermal strain may contribute to HPT axis suppression. For example, in healthy men participating in the intensive 8-week US Army Ranger military training course, in conjunction with an 8% loss of body weight, serum testosterone decreased by ~70%, and this was accompanied by other endocrine markers of energy deficit, such as an increase in SHBG, cortisol, and in the proinflammatory cytokines IL-6 and IL-8, and decreases in free triiodothyronine and IGF-1, respectively (33). Of note, changes were rapidly reversible, with body weight and hormonal parameters returning to baseline within 2 to 6 weeks of course completion (33). In another cohort of men participating in the US Army Ranger course, testosterone fell from a baseline of 16.3 nmol/L to 2.2 nmol/L, but refeeding, even with continuation of the other stressors, led to prompt testosterone recovery, suggesting that the energy deficit, not the exercise in itself, is the key factor in HPT axis suppression even in a multistressor environment (23).
Obsessive attention to exercise, combined with some degree of restrictive eating, may not result in a low BMI, although there may be a history of weight loss (9). If significant resistance exercise is being undertaken, there may be a toned and muscular appearance; however, percent body fat is reduced (9). A recent retrospective study at a large academic center identified 10 cases of male functional hypogonadism (defined biochemically) resulting from excessive exercise and/or weight loss over a period of 20 years (9). Affected men were young, 18 to 33 years old, with decreased training performance and low libido the most common presenting symptoms. Mean BMI was 20.3 ± 2.1 kg/m2, serum total testosterone 4.3 nmol/L, and body fat ranged from 4.1% to 13% (9).
Some men suffer from muscle dysmorphia or bigorexia, a condition recently recognized by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, and may engage in excessive resistance exercise such as weightlifting. Such men may abuse exogenous androgens, which can contribute to suppression of the endogenous HPT axis (34).
In summary, although the intensity and the amount of exercise required to suppress the HPT axis is not well defined and likely depends on concomitant energy intake, the overall energy deficit, and on individual susceptibility, the available evidence suggests that strenuous endurance exercise, both acutely and chronically, can suppress the HPT axis, most likely because of increased energy expenditure leading to a net energy deficit. Overall, the literature is limited by focus on convenience samples and reported cases likely represent the severe end of the spectrum. Population-based studies are required to determine the true prevalence, and the effects of more modest energy deficits on the male HPT axis.
Mechanisms by Which Energy Deficit Affects the HPT Axis
Potential mechanisms by which energy deficit may affect the HPT axis are summarized in Fig. 1.
Mechanisms of energy deficit-induced dysregulation of hypothalamic-pituitary-testicular function. Depicted are potential mechanisms by which energy deficit may lead to dysregulation of the hypothalamic-pituitary-testicular (HPT) axis. The best-established mechanism is the is the failure of leptin to induce kisspeptin signalling, essential for GnRH pulsatility. Other potential mechanisms include hypothalamic-pituitary-adrenal axis activation, including increased opioid signalling derived from proopiomelanocortin stimulated by corticotrophin-releasing hormone (CRH), and enhanced proinflammatory cytokine and Ghrelin signalling. Circulating SHBG concentrations, produced by the liver increase during energy deficit, but how this affects testosterone availability to androgen-responsive target tissues requires further study. Solid lines show pathways for which there is compelling evidence in humans. The dashed lines show pathways where evidence from studies in humans is less clear.
Leptin
Energy sufficiency is signalled to hypothalamic neurons by leptin, a hormone secreted by adipose tissue in proportion to the amount of fat stored in adipocytes. Serum leptin concentrations fall as fat mass decreases; however, acute restriction of energy intake results in a rapid fall in serum leptin concentration before any decrease in fat mass (35). A carefully conducted experimental study in healthy young men reported that, during short-term (72 hours) fasting, leptin replacement prevented the starvation-induced changes in the male HPT axis, with full restoration of LH pulsatility characteristics and serum testosterone (35). Although leptin receptors are expressed at multiple levels of the male HPT axis, preclinical studies suggest that leptin action at the hypothalamus is sufficient to maintain reproductive function (36). However, mice with a selective deletion of the leptin receptor in GnRH neurons exhibit a normal reproductive phenotype suggesting that the leptin effect is indirect (37).
Kisspeptin
Recent data have reported that the effect of leptin is mediated via a group of kisspeptin expressing neurons in the hypothalamus. Although its upstream regulatory factors are incompletely understood, kisspeptin has emerged as a key regulator of GnRH release (38). Of note, kisspeptin does not act in isolation. Kisspeptin activity is tightly controlled by peripheral factors that signal energy availability, such as leptin, and arcuate nucleus (ARC) kisspeptin neurons express leptin receptors (38). In rodents, leptin increases kisspeptin mRNA and protein in the ARC (39), and ARC kisspeptin neurons mediate pulsatile LH secretion in male mice (40). In the male rhesus monkey, fasting decreases hypothalamic kisspeptin and kisspeptin receptor expression (41). Kisspeptin stimulates the male HPT axis in healthy men (42, 43) and in men with type 2 diabetes and modest reductions in serum testosterone (44). Overall, these data suggest that in men, low leptin reduces kisspeptin-induced activation of the HPT axis.
Stress
Preclinical evidence suggests that depending on the type of stressor (energy deprivation, and/or coexisting inflammatory, depression or anxiety) multiple only partially understood, central nervous signalling mechanisms can play a role in suppressing the HPT axis, acting largely at the level of kisspeptin-expressing hypothalamic ARC neurons; for review see (45). The best-characterized mechanisms with respect to energy deficit-associated suppression of the HPT axis include components of the hypothalamic-pituitary-adrenal (HPA) axis (corticotropin-releasing hormone, pro-opiomelanocortin cleavage products, and cortisol) and proinflammatory cytokines (3, 4, 23, 33).
Hypothalamic-Pituitary-Adrenal Axis
Energy deficit, especially if occurring in a multistressor environment (23, 33) not only suppresses the HPT axis but concomitantly activates the HPA axis. The current evidence is insufficient to determine whether HPT axis suppression and HPA axis activation are caused by energy deficit as the in-common regulator or whether there is cross-talk between the 2 axes. One study of young men reported that the HPT axis-suppressive effects of 48 hour fasting occurred in the absence of measurable changes in circulating cortisol, letting the authors conclude that activation of the HPA axis is unlikely to be the cause of the fasting-induced HPT axis suppression (2). However, more prolonged fasting for 5 days almost doubles cortisol production rates (46) and chronic energy deprivation (eg, anorexia nervosa is associated with an activation of the HPA axis from corticotrophin releasing hormone hypersecretion) (12, 47). One potential mechanism by which HPA axis activation could suppress the HPT axis is via an increase of endogenous opioids such as β-endorphin derived from proopiomelanocortin, the pituitary precursor of ACTH. Opioids (either endogenous or exogenous) have been shown to have a tonic inhibitory effect on the HPT axis at the level of the hypothalamus, via GnRH downregulation (48), and in healthy men, administration of the opioid receptor antagonist naloxone not only stimulates LH and testosterone secretion (49), but also antagonizes the gonadal steroid-mediated negative feedback on GnRH secretion (50). Although naloxone has been reported to lead to gonadal axis activation in women with weight loss-related hypothalamic amenorrhea in some studies (51), whether similar mechanisms operate during energy deficit has not been studied in men.
Proinflammatory Cytokines
In an experimental randomized study, infusion of a relatively low dose of the pro-inflammatory cytokine IL-2 reduced LH’s feedforward drive on testosterone secretion, leading to a reduction in both pulsatile and basal testosterone secretion in young men (n = 17, median age 24 years); of note, the IL-2-induced HPT-axis suppressive effects were more marked in younger than older men (n = 18, mean age 63 years) (8). Given that energy deficit is associated with low-grade inflammation, especially in a multistressor environment (23, 33), this may be, in addition to the age-dependent HPT axis-suppressive effect on fasting (5), explain at least in part why energy deficit-induced HPT axis suppression may manifest predominantly in younger men. Of note, a proof-of-concept RCT in obese men has reported that anti-inflammatory treatment with an IL-1 antagonist increases serum testosterone, albeit modestly (by 1.2 nmol/L over 4 weeks) (52). Thus, although the data suggest that an energy deficit-associated proinflammatory state may be mechanistically linked to HPT axis suppression, whether anti-inflammatory treatment can stimulate the HPT axis in energy-deprived men has not been tested, and the role of HPA-axis activation in HPT-axis suppression needs further study.
Ghrelin
Ghrelin, synthesized predominately in the stomach, is a putative signal of energy deficit, stimulating hunger and hence a functional antagonist of leptin. Circulating ghrelin is elevated during energy restriction, malnutrition, and in anorexia nervosa, and has been reported to suppress the HPT axis in preclinical studies (53). In men, the evidence is limited to short, proof-of-concept studies. In a placebo-controlled randomized crossover study of 10 healthy young men (mean age 26 years), ghrelin administration over 3 hours significantly reduced LH pulse amplitude as well as frequency, and reduced serum testosterone (54). Likewise, in a controlled study, 7 healthy men (mean age 26 years), ghrelin infusion abolished LH pulsatility (55). In this study, ghrelin did not decrease the LH response to GnRH, letting the authors speculate that the HPT axis-suppressive effects of ghrelin occur upstream of GnRH neurons. In summary, although preclinical and preliminary human data are consistent with the hypothesis that energy deficit-associated increases in ghrelin may be mechanistically linked to HPT axis suppression, studies testing whether manipulation of endogenous ghrelin concentrations affect the HPT axis are lacking.
Genetic Susceptibility
As in women, the susceptibility of the male gonadal axis to low energy availability is variable, and the reasons for this are not understood. In women with functional hypothalamic amenorrhea, an increased frequency of rare variants in GnRH-associated genes has been reported (56). The largest contemporary cohort of men with HPT-axis dysregulation from energy deficit underwent screening for variants in 14 genes known to underlie isolated GnRH activity. Rare variants were identified in 2 of the 10 men (9). Whether such genetic variants modulate the sensitivity of the HPT axis to energy deficit remains to be clarified in larger studies.
Management
Since the 1970s, clinical studies have reported that clinical and biochemical features of HPT-axis dysregulation from energy deficit, even if chronic, can recover with refeeding and restoration of body weight (9-11, 13). Participants in a strenuous 8-week intensive military training course had complete recovery of HPT-axis function within 2 to 6 weeks (23, 33). Moreover, correcting the energy deficit with refeeding was sufficient to restore normal serum testosterone, even though other stressors were ongoing (23). In men with anorexia nervosa during weight regain, increases in circulating leptin closely associate with HPT-axis recovery (14). Thus, restoring the energy deficit is the key to achieving HPT-axis recovery. In men in whom HPT-axis dysfunction persists despite restoration of the energy deficit, other causes (eg, organic HPT axis pathology, covert exogenous androgen, other substance abuse and depression) should be excluded.
Given evidence that excessive endurance exercise can suppress the male HPT axis even in men with normal weight (9, 57), downscaling of excessive training regimens is usually necessary in addition to nutritional intervention. In men with body dysmorphic disorders, associated depression or exogenous androgen dependence, psychiatric input may be required.
RCTs in obese men have reported that testosterone treatment may prevent the energy deficit-associated loss of lean mass (58) and the energy deficit-associated increase in bone resorption (59). Moreover, in a recent proof-of-concept RCT of testosterone treatment in young men (mean age 25 years, n = 50) subjected to 28 days of exercise- and diet-induced energy deficit equal to 55% of total energy deficit, testosterone treatment increased lean mass compared with placebo, by 2.5 kg (95% CI, 1.6-3.3). However, testosterone treatment had no effect on measures of muscle function (60) or, in a secondary analysis, on self-reported measures of appetite (61). Overall, there is no evidence that testosterone treatment in men with HPT-axis dysregulation from energy deficit improves clinically important outcomes, and whether testosterone treatment promotes weight regain remains unknown. Indeed, given that malnutrition-associated HPT-axis dysregulation is functional and reversible, testosterone treatment will inhibit endogenous HPT-axis recovery and is not recommended. Likewise, there is no evidence to suggest that selective estrogen receptor modulators, or aromatase inhibitors (which carry a risk of further reducing bone density), are beneficial for the treatment of HPT-axis dysregulation because of energy deficit.
Although, with the exception of case reports (62), similar studies in men with energy deficit are not available, in an RCT in 20 women with hypothalamic amenorrhoea resulting from energy deficit, recombinant methionyl human leptin (metreleptin) treatment given over 36 weeks, compared with placebo, significantly increased circulating serum estradiol, and induced menstruation in 7/10 metreleptin-treated women compared with 2/9 women receiving placebo (63). Of note, however, metreleptin-treated women lost significantly more body fat compared with placebo-treated women, and 4 of the 10 metreleptin-treated women required decreased doses because of weight loss (63). This is consistent with studies in lean, leptin-sensitive women reporting that leptin treatment over 6 months reduces energy intake, associated with lipid catabolism and loss of body weight, mainly from loss of adipose tissue (64). Consistent with this, in a 46-day study of 24 overweight men (mean BMI 28.8 kg/m2) randomized to long-acting pegylated human recombinant leptin or matching placebo on the background of a very low-energy diet, whereas the decline in mean LH concentrations was attenuated in the leptin group compared with the placebo group, leptin treatment led to led to significant additional weight loss (2.8 kg) compared with placebo (65). Hence, potential benefits of leptin treatment on HPT dysfunction from energy deficit (35) may be counteracted by reductions in appetite and loss of body weight. In a small proof-of-concept study in 15 women with anorexia nervosa, a single infusion of Ghrelin did not increase appetite measure by visual analog scales, letting the authors conclude that Ghrelin is unlikely to be effective as a single appetite stimulatory treatment for patients with anorexia nervosa (66). To our knowledge, no study has assessed the effects of Ghrelin treatment in men with HPT-axis dysregulation from energy deficit.
Of note, men with energy deficit commonly have reduced IGF-1 concentrations, and in healthy young men with experimental androgen deficiency, graded doses of testosterone add-back increased muscle mass and strength, and changes in muscle mass and performance were positively correlated with changes in circulating IGF-1 (67). Moreover, in a study of healthy young men with experimental androgen deficiency, treatment with recombinant GH ans/or with recombinant IGF-1 maintained lean mass and protein synthesis rate (68). Rodent studies, however, have reported that IGF-1 receptor signaling in muscle may not be obligatory to mediate the anabolic effects of testosterone on skeletal muscle (69), and there is no evidence that GH treatment has any benefits in men with HPT axis because of energy deficit. Of note, the energy deficit-associated GH resistance is, similar to the other endocrine alterations seen during energy deficit, considered to be adaptive, maintaining euglycemia through increased lipolysis. GH treatment can lead to decreased fat mass, an undesirable effect in states of energy deficit, and cannot be advocated. Indeed, similar to what is observed for the HPT axis, the male GH/IGF-1 axis promptly recovers with restauration of the energy deficit (23, 33).
Summary and Future Directions
Evidence from case series and experimental studies show that low energy availability (either from inadequate caloric intake and/or excessive exercise) can impair HPT-axis function in men. However, the true population frequency of energy-associated HPT-axis dysregulation is not known and requires observational studies in the general population. It is likely that this condition is underrecognized, and better epidemiologic data will help to raise public awareness of this issue. This is particularly important given significant societal pressure to attain what is perceived to be the “ideal” male body image. Of note, although this review focusses on postpubertal men, a recent retrospective cohort study of 683 boys presenting with delayed puberty to a tertiary referral center has estimated that 3.7% of cases were caused by a relative energy deficit (disordered eating and/or excessive exercise) (70).
Currently, most of the evidence stems from men subjected to pronounced energy deficit, and the effects of lesser energy deficits on the HPT axis remain to be clarified. Moreover, given the nonspecificity of androgen-deficiency-like symptoms and signs, the degree of which clinical features (ie, fatigue, low libido, low mood, loss of muscle mass) are due to the HPT axis suppression in itself and/or modified by coexisting comorbidities or concomitant hormonal alterations (eg, adrenal, thyroid, GH/IGF-1 axis dysregulation) occurring as a consequence of energy deficit requires further study. Of note, in states of energy deficit, serum SHBG is commonly increased, and this can affect the interpretation of total testosterone concentrations. However, whether free testosterone (ideally measured by equilibrium dialysis) is a better reflection of androgen status in men with energy deficits than total testosterone is an open question to be addressed in future studies. Of note, the effects of energy deficit on sperm concentration and fertility rates remain unknown, and constitute an important area of further research
The best studied mechanisms implicated in HPT-axis dysregulation in states of energy deficit involves altered leptin/kisspeptin signalling and, quite possibly a pro-inflammatory state, leading to functional and potentially reversible suppression of the HPT axis predominantly at the hypothalamic level. The role of other factors such as HPA-axis activation and Ghrelin requires further study (Fig. 1). In addition, energy deficit is associated with several additional endocrine changes including hepatic GH resistance leading to lowered IGF-1, and sick euthyroid syndrome. Whether these endocrine alterations affect the HPT axis and/or the clinical phenotype associated with energy deficit requires further study. Overall, the current evidence is insufficient to determine whether changes in the HPT axis and the concomitant other hormonal alterations are simply co-occurring consequent to energy deficit as the common denominator or whether there is crosstalk between the HPT axis and the other hormones’ axes. The biological significance and potential clinical consequences of these relationships between HPT axis dysregulation and changes in other hormones requires further study. Disentangling the complex picture of these interrelated phenomena will be difficult and will require carefully designed preclinical experiments and mechanistic studies in men.
Moreover, factors determining inter-individual susceptibility to energy deficit-associated HPT-axis dysregulation remain unknown. Although younger age and genetic changes in genes implicated in HPT-axis development and function could play a role, such preliminary findings require confirmation. Whether genetic alterations in other pathways, such as those involved in energy regulation or stress-related signalling pathways could modulate susceptibility, has not been studied but would be of interest.
The management of such men should focus on achieving a healthy body weight and a healthy level of physical activity in the context of holistic, individualized patient-centered care addressing associated comorbidities. For some men, a multidisciplinary approach involving endocrinologists, mental health professionals, dietitians, and exercise physiologists may be necessary. However, prospective data demonstrating the efficacy of such an approach are lacking. Moreover, there are no data on clinical management strategies for men that fail such measures, and no data on long-term health outcomes, which is of particular relevance given that most men presenting with energy deficit-associated HPT-axis dysregulation are relatively young. A desire to preserve skeletal muscle and bone mass while waiting for recovery and weight restoration, may prompt prescription of testosterone. We recommend against this on the basis that (1) there is currently no clear evidence of benefit and (2) testosterone may have unforeseen longer term consequences by inhibiting recovery of the HPT axis and may be difficult to withdraw. Further research is required. Whether agents that act centrally to activate the HPT axis, such as selective estrogen receptor modulators are beneficial and safe, remains to be determined. Likewise, experimental therapies, such as leptin treatment, require further research.
Abbreviations
- ARC
arcuate nucleus
- BMI
body mass index
- HPA
hypothalamic-pituitary-adrenal
- HPT
hypothalamic-pituitary-testicular
- RCT
randomized controlled study.
Acknowledgments
Funding:This minireview was not supported by any funding.
Additional Information
Disclosures: M.G. has received research funding from Bayer, Otzuka, and speaker’s honoraria from Besins Health Care and Novartis. G.A.W. has received research funding for testosterone pharmacology studies (Lawley, Bayer, Lilly), speakers (Besins, Bayer), and consultancy (Elsevier) fees.
Data Availability
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
References
Author notes
Both authors contributed equally to this work.
