Menopause is defined as the permanent cessation of menstrual periods resulting from the loss of ovarian follicles, along with serum estrogen deficiency. The average age of menopause is between 46 and 52 years, but it differs by race [1]. Several previous studies have shown that the mean menopausal age varies from 46.9 to 50.4 years in Korean women [2,3]. Both menopause and its accompanying sex hormone changes have been reported to be related to several metabolic changes, including the prevalence of obesity, dyslipidemia, insulin resistance, type 2 diabetes, and increased risk of cardiovascular disease [4,5]; however, the exact mechanisms underlying these associations are not yet fully understood. Thus, it is important to monitor and manage high-risk groups for chronic diseases in middle-aged women.

Previous studies have presented conflicting results on whether menopause is directly related to obesity or body fat distribution in middle-aged women. A recent systemic review proposed that the change in fat mass quantity between pre-menopausal and postmenopausal women was predominantly attributed to increasing age, rather than the menopausal transition itself [6]. Conversely, several studies have suggested that a decrease in endogenous estrogen levels during menopause may increase the total amount of overall body fat, particularly central fat, also referred to as trunk or torso fat, in postmenopausal women [7–10]. To clarify the factors associated with body fat distribution, further research using more accurate measurements such as dual-energy X-ray absorptiometry (DEXA) or computed tomography (CT), rather than anthropometric measurements such as body mass index (BMI) or waist circumference, is needed [11].

Central or visceral obesity, prevalent among menopausal women, has been recognized as a strong predictor of insulin resistance, hyperinsulinemia, type 2 diabetes, and cardiovascular risk [12–14]. However, most previous studies have been based on anthropometric measurements, and how specific changes in body fat distribution in each body part after menopausal transition are related to cardiometabolic disorders, including insulin resistance, remains unclear.

Therefore, this study aimed to investigate body fat distribution according to menopausal status and its relationship with insulin resistance among middle-aged Korean women using nationwide population-based data.

A total of 3,468 Korean women aged 40–60 years were included in this study: 1,979 premenopausal women and 1,489 postmenopausal women. A comparison of the baseline characteristics according to menopausal status is shown in Table 1. The mean ages of the premenopausal and postmenopausal women were 45.21±3.81 years and 54.76±3.55 years, respectively. Postmenopausal women have higher parity, BMI, waist circumference, and prevalence of chronic diseases, more frequent use of oral contraceptive and menopausal hormone therapy, lower education levels, less intensive physical activity, higher carbohydrate and lower fat uptake, higher blood pressure, higher fasting glucose levels, and generally worse lipid profiles than premenopausal women. Total fat mass was significantly higher in postmenopausal women than that in premenopausal women (20.08±4.92 kg and 19.37±4.89 kg, respectively, P<0.001). The trunk, arm, and head fat percentages were significantly higher in postmenopausal women than in premenopausal women (Fig. 1). However, no significant difference was observed in the leg fat percentage according to menopausal status among middle-aged women.

The results of multivariable regression analysis for HOMA-IR are presented in Table 2. Trunk fat percentage (estimate, 0.02; P=0.003) showed a significant positive correlation with HOMA-IR, whereas the leg fat percentage (estimate, −0.03; P<0.001) showed a negative correlation after adjusting for several confounding factors. In addition, BMI, waist circumference, diabetes mellitus, and triglyceride levels were positively correlated with HOMA-IR, whereas age, intensive physical activity, and high-density lipoprotein cholesterol level were negatively correlated with HOMA-IR in this multivariable analysis. In addition, menopausal status, current menopausal hormone therapy, carbohydrate or fat uptake, household income, and education levels were not associated with HOMA-IR (Table 2). In the subgroup analysis according to menopausal status, the leg fat percentage showed a significant negative correlation with HOMA-IR in both pre- and postmenopausal women (estimate, −0.03; P<0.001 and estimate, −0.03; P=0.001, respectively), whereas the trunk fat percentage showed a positive correlation only in premenopausal women (estimate, 0.02; P=0.010; Table 3).

Using a nationally representative, population-based survey, this study showed that body fat distribution differed according to menopausal status among middle-aged women and was also significantly associated with insulin resistance after adjusting for several confounding factors. Trunk fat percentage and leg fat percentage were inversely associated with insulin resistance, as measured using HOMA-IR. In particular, leg fat negatively correlated with HOMA-IR regardless of menopausal status, whereas trunk fat was only associated with HOMA-IR in premenopausal women. Predicting and managing the metabolic health of middle-aged women should consider analyzing body fat distributions that show different patterns depending on menopause, rather than only using anthropometric measurements such as BMI in health checkups for those populations.

Various methodologies have been used to assess total body fat quantity or distribution, including BMI, waist circumference, DEXA, CT, and magnetic resonance imaging (MRI), in studies on women’s cardiometabolic health. Previous cross-sectional studies using DEXA showed that postmenopausal women had a greater proportion of central fat than premenopausal women [8,16]. Another study using single-slice CT also showed an increase in visceral abdominal fat with menopause [17]. In this study, we measured body fat distribution using DEXA, which is a useful technique that can accurately determine discrete fat mass cross-sectionally. It is more accurate than BMI and waist circumference to assess body composition, is less expensive and less invasive than MRI, and exposes patients to smaller doses of radiation than CT. Furthermore, DEXA is frequently performed for osteoporosis screening among middle-aged women in South Korea; therefore, body fat distribution can be easily and accurately measured along with bone mineral density in those populations. Although anthropometric measurements can be used to predict insulin resistance among middle-aged women, body fat distribution can also be measured freely using DEXA in those populations and provide valuable information regarding their metabolic health, including insulin resistance. However, further studies are needed to validate how our findings can be used to manage the health of middle-aged women.

Central adiposity has been associated with cardiovascular risk and results in adverse metabolic consequences such as dysglycemia, dyslipidemia, hypertension, and cardiovascular disease [14]. In line with these findings, this study also showed that trunk fat percentage had an independent adverse effect on insulin resistance, whereas leg fat percentage had a protective effect. The protective properties of leg fat against cardiometabolic disease risk have been suggested in several studies. Studies on adipose tissue physiology in vitro and in vivo suggest distinct properties of the leg fat depot with regard to lipolysis and fatty acid uptake [18]. It appears to be more passive than the abdominal depot and exerts its protective properties by long-term fatty acid storage, which affects insulin resistance [18]. Similarly, a larger hip circumference was associated with a lower prevalence of diabetes mellitus and dyslipidemia in a previous large population-based study [19]. Isolated adipocytes from upper-body obese women respond less to insulin activity, inhibiting lipolysis in the adipose tissue when compared with those of lower-body obesity [20]. These findings indicate that changes in body fat distribution are associated with insulin resistance and could be a marker of cardiometabolic disease risk.

To the best of our knowledge, this is the first large population-based study using a nationwide database to reveal a significant association between insulin resistance and body fat distribution according to menopausal status in middle-aged Korean women. However, this study has several limitations. First, this was a cross-sectional study that did not allow the investigation of causality. Determining the causal relationship between insulin resistance and menopause is difficult. Second, although many confounding factors were assessed and statistically controlled for, the possibility of residual confounding cannot be ruled out. Further studies are required to confirm the relationship between insulin resistance and menopause. Finally, menopausal status was not classified in detail and was determined based only on self-reported questionnaires. Further longitudinal studies with detailed definitions of menopausal status and hormonal assessments are needed.

Menopausal women have not only higher body weight, but also a different fat distribution, compared with premenopausal women. Body fat distribution was independently correlated with insulin resistance in middle-aged women. Although BMI has been used to evaluate the metabolic health of middle-aged women, our findings indicate that body fat distribution should also be assessed along with general obesity. Further studies are warranted to investigate the role of DEXA as a screening tool to predict high-risk groups for cardiometabolic diseases among middle-aged women.