Testosterone and obesity interact differently in men and women, review finds

Testosterone and obesity are locked in a two-way relationship, but the nature of that relationship flips almost completely depending on sex, according to a review published in Frontiers in Medicine on 28 October 2026. In men, obesity tends to drive testosterone down, and low testosterone in turn favors further fat accumulation. In women, obesity is more often associated with relative androgen excess, which can worsen fat distribution and metabolic dysfunction, particularly in the condition formerly known as polycystic ovary syndrome, now renamed polyendocrine metabolic ovarian syndrome, or PMOS, following a 2026 global consensus.

The review, written by Saki Komai, Takahiro Tsutsumi and Kyoichiro Tsuchiya of the University of Yamanashi in Japan, synthesizes basic and clinical research on how testosterone and adipose tissue regulate one another. The authors emphasize that the evidence is strongest in men, where interventional studies support body composition benefits of testosterone replacement therapy, while evidence in women is largely drawn from observational studies of PMOS and from experimental models. They also caution that body mass index does not fully capture visceral fat, ectopic fat or lean mass, and that these adiposity phenotypes should be considered separately.

The sexual dimorphism begins with the sources of the hormone. In men, roughly 95 percent of circulating testosterone is synthesized in testicular Leydig cells, with the remainder coming from the adrenal glands and peripheral conversion of adrenal androgen precursors. Production is governed by the hypothalamic-pituitary-gonadal axis: pulsatile gonadotropin-releasing hormone from the hypothalamus stimulates the pituitary to release luteinizing hormone and follicle-stimulating hormone, and luteinizing hormone in turn drives testosterone production in Leydig cells. Testosterone and its metabolite estradiol feed back on the hypothalamus and pituitary to keep the system in balance. Most circulating testosterone is bound to sex hormone-binding globulin and albumin, leaving only about 1 to 2 percent free.

In women of reproductive age, the picture is more distributed: approximately 25 percent of circulating testosterone comes from the ovaries, 25 percent from the adrenal glands, and about half from peripheral conversion of circulating androstenedione. Total testosterone concentrations are far lower in women than in men. A pooled analysis using liquid chromatography-mass spectrometry reported 95 percent reference ranges of 7.7 to 29.4 nanomoles per liter, roughly 222 to 848 nanograms per deciliter, in healthy young men aged 18 to 40, compared with 0 to 1.7 nanomoles per liter, roughly 0 to 49 nanograms per deciliter, in healthy menstruating women under 40. After natural menopause, the ovary remains hormonally active and contributes about half of circulating testosterone, with roughly 40 percent from peripheral conversion and 10 percent from direct adrenal production. Testosterone declines with age in women, with the steepest drop in earlier reproductive decades, but menopause itself does not appear to independently affect androgen levels.

In men, the review describes multiple mechanisms by which testosterone restrains fat accumulation. Activation of the androgen receptor in preadipocytes suppresses the expression of PPAR-gamma and C/EBP-alpha, transcription factors critical for adipocyte differentiation, thereby impeding the formation of mature fat cells. Testosterone also exerts anabolic effects on skeletal muscle, and a systematic review and meta-analysis of randomized trials showed that testosterone supplementation increased lean body mass in middle-aged and older men. Clinical observations reinforce the point: men undergoing androgen deprivation therapy for prostate cancer gain body fat, lose lean mass and gain weight, demonstrating that androgen signaling helps maintain body composition.

Obesity, in turn, is strongly associated with reduced testosterone in men, and the review traces several contributing pathways. A recent cross-sectional study of Japanese men found inverse associations of total testosterone with both body mass index and waist-to-hip ratio, and men with central obesity had significantly lower total and free testosterone than men without it. Increased adipose tissue mass raises expression of aromatase, the enzyme that converts testosterone into estradiol, potentially enhancing estradiol-mediated negative feedback on the hypothalamic-pituitary-gonadal axis, although the review notes that the contribution of this pathway remains incompletely established. Obesity-related reductions in sex hormone-binding globulin lower total testosterone substantially, but only partially explain the decline. Leptin resistance may impair the hypothalamic kisspeptin neurons that support pulsatile GnRH secretion, and chronic low-grade inflammation linked to visceral obesity may suppress the axis both centrally and directly at the testis, where pro-inflammatory cytokines such as tumor necrosis factor-alpha and interleukin-6 can inhibit Leydig cell steroidogenesis. Obstructive sleep apnea, common in men with obesity, is also associated with lower testosterone, particularly in severe cases. The authors stress that most human evidence for these pathways is observational and should not be read as proof of a single causal sequence.

The result is what the field calls male obesity-related secondary hypogonadism, or MOSH: a potentially reversible form of secondary hypogonadism in which low testosterone and obesity reinforce each other. Clinically significant weight loss can raise serum testosterone, with the magnitude of improvement generally proportional to the degree of weight reduction, making weight management the foundation of treatment. Testosterone replacement therapy may be considered in selected men with persistent symptoms and unequivocally low testosterone after appropriate evaluation, but the review is explicit that it should not replace interventions targeting the underlying obesity and associated metabolic disorders. The authors also flag an open question: obesity-related low testosterone includes phenotypes driven mainly by low sex hormone-binding globulin as well as those with reduced free testosterone and functional suppression of the axis, and criteria that distinguish these phenotypes or predict recovery after weight loss have not been established. Whether recovery depends primarily on total weight loss or on improvements in visceral fat, hepatic fat or insulin sensitivity remains unclear, and studies directly comparing lifestyle intervention, anti-obesity pharmacotherapy and metabolic bariatric surgery at equivalent weight reduction are limited.

In women, the direction reverses. Obesity may be associated with relative androgen excess, driven largely by insulin resistance and compensatory hyperinsulinemia. Insulin enhances ovarian androgen production by augmenting the steroidogenic response to luteinizing hormone, and it reduces hepatic production of sex hormone-binding globulin, increasing the free, biologically active fraction of circulating testosterone. Adipose tissue itself contributes through the enzyme AKR1C3, which converts androstenedione to testosterone and regulates local androgen availability. In women with PMOS, adipose AKR1C3-mediated androgen activation has been implicated in fat accumulation and metabolic dysfunction, although the evidence comes mainly from women with PMOS and experimental studies and should not be generalized to women with obesity who lack hyperandrogenic disorders.

PMOS serves as the clearest clinical illustration of androgen excess promoting metabolic dysfunction. Women with PMOS frequently show abdominal adiposity, obesity, insulin resistance and impaired glucose tolerance, and they have higher circulating androgens and a greater android-to-gynoid fat mass ratio than body mass index-matched controls. Although subcutaneous adipose AKR1C3 expression is not consistently increased, its expression is associated with fat accumulation in these women. Experimental work adds mechanistic weight: in a dihydrotestosterone-induced mouse model of PMOS, androgen exposure increased adiposity and adipocyte hypertrophy, and these changes were absent when androgen receptors were deleted specifically in adipose tissue, implicating adipose androgen receptor signaling in fat accumulation. The review characterizes PMOS as a vicious cycle in which obesity and hyperinsulinemia stimulate androgen production, and the resulting androgens further aggravate fat accumulation, adipose dysfunction and insulin resistance. The authors caution, however, that clinical studies in PMOS are predominantly observational, limiting causal inference.

Perhaps the most striking gap the review identifies lies in women without PMOS. Because research has focused on pathological hyperandrogenism, the metabolic role of testosterone within the physiological female range remains poorly understood. A cross-sectional study in the general female population showed an L-shaped inverse association between total testosterone and metabolic syndrome, though causality and an optimal concentration range are unknown. In healthy reproductive-aged women, circulating testosterone was inversely associated with body mass index, with higher concentrations in normal-weight than in overweight women, raising the possibility that relatively low testosterone within the physiological range accompanies an unfavorable adiposity phenotype. Whether low testosterone contributes to fat accumulation or merely reflects hormonal changes associated with increased adiposity has not been established in longitudinal human studies. The authors call for future work combining mass spectrometry-based measurement of testosterone, estradiol, sex hormone-binding globulin, free testosterone and 11-oxygenated androgens with longitudinal assessments of adiposity and insulin resistance, in both sexes. Their overall conclusion is that testosterone is neither uniformly beneficial nor harmful: its metabolic effects depend on sex, hormone concentration, bioavailability, conversion to other sex steroids, target tissue and the surrounding metabolic environment.

Subject of Research: The bidirectional relationship between testosterone and obesity and how it differs between men and women

Article Title: Sex differences in testosterone and obesity

Article References: Komai, S., Tsutsumi, T., & Tsuchiya, K. (2026). Sex differences in testosterone and obesity. Frontiers in Medicine, 13, Article 1992734. https://doi.org/10.3389/fmed.2026.1992734

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1992734

Keywords: testosterone, obesity, androgens, sex differences, PMOS, hypogonadism, adipose tissue, insulin resistance, SHBG, visceral fat, metabolic syndrome, weight loss