The menopausal transition includes substantial hormonal and physiological changes that can markedly impair quality of life, specifically by vasomotor symptoms like night sweats and hot flashes. These symptoms affect around 75% of females and may begin several years before the final menstrual period (FMP), persisting up to a decade. Variability is noted in individuals based on ethnicity, race, and socioeconomic factors. Estrogen withdrawal and hypothalamic dysfunction are recognized contributors to symptom development. Emerging evidence suggests that metabolic health, including obesity and insulin resistance, may also influence the onset and severity of menopausal symptoms. However, the role of insulin as an early marker of metabolic dysfunction in predicting menopausal symptom patterns remains unclear.
A study published in the Journal of Clinical Endocrinology & Metabolism aimed to investigate whether fasting insulin levels during perimenopause predict the timing, duration, and incidence of physical menopausal symptoms and to compare the predictive strength of insulin with body mass index (BMI). The study also examined how insulin and BMI relate to longitudinal changes in the reproductive hormones, which include estradiol (E2), follicle-stimulating hormone (FSH), and testosterone (T), across the menopause transition.
The analysis used longitudinal data from the Study of Women’s Health Across the Nation (SWAN), which followed women from premenopause/perimenopause for up to 10 years. Secondary analyses were conducted using anonymized publicly available data. Metabolic measures were assessed at age 47 to reduce confounding from age-related increases in insulin. Participants were involved if they attended the age-47 visit and had complete data on fasting insulin, fasting glucose, and BMI. Women were excluded if they used glucose-lowering medications, underwent hysterectomy or oophorectomy, or had missing key variables. The final analytic sample included 704 women. Fasting insulin values were log-transformed to improve normality.
Menopausal symptoms like hot flashes, night sweats, cold sweats, and vaginal dryness were self-reported annually and analyzed for age onset, duration across visits, and incidence. Hormonal trajectories relative to the FMP were estimated by mixed effects models. Statistical analyses included linear regression models for continuous symptom outcomes, Cox proportional hazards models for symptom incidence, and multivariable models adjusted for race/ethnicity, income, and smoking status. Statistical significance was defined as P <0.05.
Participants had a mean fasting insulin level of 10.1 µIU/mL (standard deviation [SD]: 6.7), a BMI of 27.0 (SD 6.6), a mean fasting glucose level of 89.5 mg/dL, and a mean homeostatic model assessment of insulin resistance (HOMA-IR) of 2.31. Most women were asymptomatic at age 47, and the mean age at FMP was 51 years.
In adjusted linear regression models, higher fasting insulin levels were significantly linked to earlier onset of hot flashes (β = −1.14 years, P < 0.001) and night sweats (β = −0.69 years, P = 0.035), as well as longer duration of hot flashes (β = 0.62, P = 0.032) and cold sweats (β = 0.38, P = 0.018). These findings indicate that elevated insulin levels at age 47 predict earlier onset and prolonged duration of vasomotor symptoms. BMI showed similar but weaker associations: after adjustment, only the earlier onset of hot flashes (β = −0.093, P < 0.001), earlier onset of night sweats (β = −0.047, P = 0.045), and longer duration of cold sweats (β = 0.025, P = 0.031) remained statistically significant.
Cox proportional hazards analyses demonstrated that insulin was a stronger predictor of vasomotor symptoms incidence than BMI. Each standard deviation increase in log-transformed insulin (about 5.6 µIU/mL) was associated with 14% higher hazard of hot flashes (hazard ratio [HR]: 1.14, 95% confidence interval [CI]: 1.05 to 1.24, P = 0.002) and a 20% higher hazard of cold sweats (HR: 1.20, 95% CI: 1.06 to 1.35, P = 0.004). When insulin and BMI were involved simultaneously, insulin remained independently linked to hot flash risk (HR: 1.15, P = 0.007), whereas BMI was no longer significant. In models incorporating insulin, BMI, and fasting glucose, only insulin independently predicted hot flash incidence (HR: 1.13, P = 0.027).
Hormonal trajectory analyses revealed that higher BMI at age 47 was associated with a smaller decline in oestradiol levels and a slower rise in FSH during menopause, consistent with adiposity-related estrogen feedback. In contrast, elevated insulin levels were associated with greater increases in testosterone, suggesting a hyperandrogenic profile.
Overall, this study identifies fasting insulin as a significant predictor of earlier onset and greater severity of vasomotor symptoms during perimenopause, surpassing the predictive value of BMI. These findings position insulin as a key metabolic marker to manage symptomatic and hormonal changes in midlife women and highlight potential opportunities for targeted metabolic interventions.
Reference: Faria Athar, Sarah Gregory, Emma J Houston, Nicole M Templeman. Insulin levels early in perimenopause inform vasomotor symptom incidence across the menopausal transition. J Clin Endocrinol Metab. 2026;dgaf699. doi:10.1210/clinem/dgaf699
