Testosterone shields female mice from most metabolic and bone damage after ovary loss

When the ovaries stop producing hormones, whether through surgery, menopause, primary ovarian insufficiency, or chemotherapy, the female body begins to lose ground on two fronts at once: metabolism and bone. Estrogen depletion is famously tied to weight gain, insulin resistance, rising cholesterol, and osteoporosis. But what about testosterone? In women, elevated testosterone is usually painted as a villain, associated with the metabolic dysfunction seen in hyperandrogenic conditions such as polyendocrine metabolic ovarian syndrome, formerly known as PCOS. A new study from the University of Michigan, published in Frontiers in Endocrinology, complicates that picture dramatically. In female mice, chronic testosterone treatment prevented most of the metabolic and skeletal damage caused by surgical ovary removal, while leaving other deficiencies untouched, revealing that androgens and estrogens play strikingly distinct roles in female metabolic homeostasis.

The research team, led by Robin Kruger and senior author Ariella Shikanov, set out to answer a question that clinical studies have struggled to resolve: does testosterone worsen, mitigate, or simply modify the consequences of ovarian hormone loss? Human studies of exogenous testosterone administration show inconsistent effects on lipid profiles, insulin sensitivity, body weight, and blood pressure, and they are often constrained by retrospective designs, heterogeneous hormone regimens, and confounding variables such as participant age and baseline metabolic status. Crucially, none of these studies have examined potential differences in outcomes between patients with and without endogenous ovaries. A controlled animal model offered a way to isolate the direct physiological effects of testosterone and ovarian hormone loss on metabolism and the skeleton.

The experiment was elegantly simple in design. Eight-week-old adult female CD-1 mice underwent either sham surgery or ovariectomy, and at the same time were implanted subcutaneously with silastic capsules containing either 10 milligrams of testosterone powder or empty control capsules. This produced four experimental groups: intact mice with empty capsules, ovariectomized mice with empty capsules, intact mice with testosterone, and ovariectomized mice with testosterone, with seven to eight animals per group. After six weeks of treatment, the researchers measured fasting glucose, glucose tolerance, serum lipids, insulin, reproductive hormones, fat pad weights, adipocyte size, heart and liver phenotypes, and bone density and strength. Serum testosterone rose to the range typically observed in male mice in both testosterone-treated groups, confirming that the treatment produced a genuinely hyperandrogenic hormonal milieu.

The most dramatic result concerned glucose metabolism. Mice that lost their ovaries without testosterone replacement developed hyperglycemia and impaired glucose tolerance, consistent with previous studies of ovariectomy-induced metabolic dysfunction. Yet ovariectomized mice receiving testosterone retained entirely normal glucose function, indistinguishable from intact controls. Testosterone treatment was sufficient to prevent the ovariectomy-induced glucose metabolism defects, even though fasting insulin levels did not change proportionally. This suggests that testosterone supplementation alone can maintain glucose homeostasis in the absence of ovarian hormones, a finding with direct implications for how clinicians might think about metabolic risk after estradiol depletion.

Bone told an equally compelling story. Using nano-computed tomography of the femur and three-point bending tests of the radius, the team found that testosterone increased trabecular bone number, reduced trabecular spacing, and raised bone density measures regardless of whether the ovaries were present. Both testosterone-treated groups showed congruent trends toward greater maximum load and bending stiffness in mechanical testing. Interestingly, trabecular thickness was not increased despite the greater number of trabeculae, consistent with previous reports in mice and humans that testosterone induces formation of more but thinner trabeculae. Testosterone had minimal impact on cortical bone, supporting clinical evidence that cortical microstructure is not as strongly regulated by androgen signaling. Together, these data demonstrate that elevated testosterone is sufficient not only to prevent ovariectomy-associated bone loss but to raise trabecular bone density above baseline even without endogenous gonads.

But the story was not uniformly rosy. Testosterone could not prevent dyslipidemia in the absence of ovaries. Both ovariectomized groups, whether treated with testosterone or not, displayed elevated fasting total cholesterol and HDL cholesterol levels compared with intact mice. In mice, which carry most plasma cholesterol in HDL particles, elevated HDL more likely reflects overall hypercholesterolemia rather than a protective phenotype as it might in humans. Similarly, ovariectomy increased adipocyte size in both inguinal and perigonadal fat pads regardless of testosterone treatment, indicating adipocyte hypertrophy persisted despite the androgen excess. In contrast, testosterone-treated mice with intact ovaries showed decreased fat pad weight and smaller median adipocyte size, a pattern aligning with sexual dimorphism in body composition where males carry lower adiposity than females. The researchers concluded that lipid metabolism is more sensitive to estradiol levels than to testosterone levels, and that testosterone aromatized in non-ovarian tissues is not sufficient to maintain lipid homeostasis in females.

The heart added another layer of complexity. Testosterone treatment significantly increased heart weight in both ovary-intact and ovariectomized mice compared with untreated controls, supported by congruent trends toward increased heart longitudinal area and left ventricular cardiomyocyte size. Neither testosterone nor ovariectomy significantly impacted heart fibrosis, liver weight, collagen deposition, or hepatic lipid content over the six-week period. The researchers argue that these changes are more likely direct androgenic effects rather than consequences of reduced estradiol, because ovariectomized mice without testosterone did not exhibit the same phenotypes. This premise is consistent with prior studies showing that males generally have greater heart mass and higher triglyceride levels than females, and that testosterone can induce cardiomyocyte hypertrophy in vitro through androgen receptor signaling.

One of the most intriguing findings involved the uterus. Testosterone-treated mice with intact ovaries showed increased uterine weight and cross-sectional area, a uterotrophic effect suggesting significant aromatization of testosterone into estradiol and estrogenic signaling, despite the cessation of estrous cycling. Meanwhile, ovariectomized mice without testosterone showed the expected uterine regression. Remarkably, the uteri of testosterone-treated ovariectomized mice were indistinguishable from intact controls, avoiding both testosterone-induced hypertrophy and ovariectomy-induced atrophy. If the uterotrophic effect comes from aromatization into estradiol, this would indicate that peripheral tissues alone can generate bioactive levels of estradiol to sustain uterine morphology. Alternatively, the effect could reflect direct androgen receptor signaling in uterine tissue. Either mechanism suggests that testosterone treatment and ovariectomy do not produce identical hormonal environments, despite similar repression of estrous cycling, and such differences in estrogenic activity may explain how testosterone-treated intact mice differ from ovariectomized groups in metabolic outcomes such as adiposity and cholesterol.

The study faces limitations, particularly in determining the mechanisms governing the observed metabolic changes. A primary shortcoming is the inability to distinguish which effects of testosterone administration result from androgen signaling and which may be due to aromatization of testosterone into estradiol. Many peripheral tissues express aromatase, and this mechanism is undoubtedly at play even in ovariectomized animals. Previous studies have shown that aromatization is critical for the beneficial effects of androgen administration on rodent metabolism and skeletal health. Future studies may employ aromatase inhibition or androgen receptor knockout models to separate these two pathways, along with insulin tolerance testing, estradiol measurements, and quantification of metabolic gene expression. It is also worth noting that several metabolic endpoints, including fasting insulin, metabolic tissue fibrosis, and hepatic lipid accumulation, were unaffected by either testosterone or ovary status, likely reflecting the resistance of CD-1 mice to hepatic steatosis on a standard diet and highlighting the importance of genetic background and diet in determining susceptibility to reproductive hormone-associated metabolic disease.

Clinically, these distinctions matter enormously. For women with conditions characterized by high testosterone levels, including polyendocrine metabolic ovarian syndrome, congenital adrenal hyperplasia, environmental exposure, or androgen-producing tumors, several of which are common comorbidities with metabolic syndrome and obesity that worsen in peri-menopause, this study may inform which metabolic disease factors are likely to increase in risk after estradiol depletion and which may be compensated by high testosterone. The findings are also directly relevant to transmasculine patients receiving testosterone gender-affirming hormone therapy, especially those considering gender-affirming oophorectomy. The results suggest that testosterone therapy is sufficient to maintain normal glucose metabolism and bone density after oophorectomy, but that gonad retention may be protective with respect to lipid and adipose tissue regulation. Further research investigating the mechanisms of interaction between endogenous and exogenous hormones, effects in individuals with existing disease states such as obesity, diabetes, and cardiovascular disease, and potential mitigation strategies for metabolic risk factors will be crucial to support the health of hyperandrogenic patients.

Subject of Research: Endocrinology

Article Title: Testosterone treatment prevents most gonadectomy-induced metabolic and skeletal deficiencies in female mice

Article References: Kruger, R. E., Yohannes, H., Cho, E., Eisner, L., Pfau, D. R., Chan-Sui, R. K., Migotsky, N., Monteiro da Rocha, A., Killian, M. L., Moravek, M. B., Padmanabhan, V., & Shikanov, A. (2026). Testosterone treatment prevents most gonadectomy-induced metabolic and skeletal deficiencies in female mice. Frontiers in Endocrinology, 17, Article 1940100. https://doi.org/10.3389/fendo.2026.1940100

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1940100

Keywords: testosterone, ovariectomy, female mice, glucose metabolism, bone density, trabecular bone, dyslipidemia, adipocyte hypertrophy, estrogen, aromatization, hyperandrogenism, gender-affirming hormone therapy