Materials and methods. Our case-control study included 82 women aged 46–60 years; Group 1 consisted of 35 women with MS, while Group 2 comprised 47 women without MS. We collected anamnesis on the basis of outpatient medical records using a specially developed form, performed anthropometry and laboratory tests, and assessed the severity of climacteric syndrome using the modified menopausal index (MMI).
Results. In patients with MS vs. the control group, the homeostasis model assessment of insulin resistance (HOMA-IR) index, visceral adiposity index (VAI) and MMI were significantly higher (p<0.001, p<0.001 and p=0.003, respectively), while the levels of follicle-stimulating hormone (FSH) and prolactin were lower (p=0.008 and p=0.011, correspondingly). Using the multiple logistic regression method, we developed a screening model for assessing the probability of developing MS in women. The model included five parameters: the presence of somatic comorbidity, VAI, MMI, FSH content and prolactin level. The ROC curve was used to evaluate the model. The area under the ROC curve (AUC) of the model was 0.909 with a 95% confidence interval (CI) 0.835–0.984 (sensitivity: 87.1%; specificity: 78.6%).
Conclusion. Our predictive model for the development of metabolic syndrome in women in the early postmenopausal period allowed identifying patients with increased cardiometabolic risk, thereby contributing to timely initiation of personalized preventive measures and treatment.
Introduction
After the onset of menopause, the incidence of metabolic syndrome (MS) increases, reaching, according to epidemiological studies, 31–55% of cases [1–3]. Menopausal MS (MMS) in women is characterized by fast weight gain resulting in the development of visceral adiposity, insulin resistance, dyslipidemia, and arterial hypertension (HTN). Estrogen deficiency triggers endothelial dysfunction and the development of mild hyperandrogenism. As a result, there is a transition from the gynoid to the android type of adipose tissue distribution. It actively produces inflammatory markers with pronounced atherogenic properties [4]. It is known that some adipokines secreted by adipocytes, including adiponectin and leptin, are associated with the development of MS [5]. At the same time, it was suggested that vasomotor symptoms during the menopausal transition and early postmenopause, especially in patients with moderate and severe MS, are associated with a bigger risk of cardiovascular diseases (CVD) [6, 7].
Unfortunately, there is still low awareness of the CVD risk in women after the end of reproductive function. Consequently, it is essential to continue the search for informative predictors of the MMS development in order to improve personalized prevention and ensure early risk management of diseases associated with MS.
Objective: To develop an effective predictive model of the MS development in women in their early postmenopausal period.
Materials and methods
Our case-control study included 82 women who signed informed consent to participate in it. Data were collected at the Department of Disease Prevention, City Clinical Hospital No. 1, Chelyabinsk, Russia, from 2022 to 2024. The study was approved by the Ethics Committee of the South Ural State Medical University of the Russian Federation Ministry of Healthcare (protocol No. 9 of September 22, 2022).
Group 1 (cases) consisted of 35 women with MS, while Group 2 (controls) included 47 women without MS.
Inclusion criteria for the study were as follows: postmenopausal women undergoing +1a, +1b, +1c and +2 stages of reproductive aging sensu the reproductive aging classification in women, STRAW+10 (Stages of Reproductive Aging Workshop) [8]; and age of ≤60 years.
Exclusion criteria were as follows: use of weight loss medications, undergoing menopausal hormone therapy (MHT), taking systemic glucocorticoids during the last 6 months; types 1 and 2 diabetes mellitus; autoimmune diseases; and somatic and gynecological diseases in the acute or decompensated stage (during investigation of biochemical markers).
All women were evaluated for age at the onset of menopause, obstetric and gynecological history, the presence of chronic somatic diseases based on outpatient medical records, and anthropometric measurements (weight, height) for calculating the body mass index (BMI, kg/m²). Waist circumference (WC, cm) was measured along the midaxillary line at the midpoint between the iliac crest and the lower edge of the last rib, hip circumference (HC, cm) was assessed along the largest circumference around the buttocks. To identify the severity of abdominal obesity, we calculated the WC/HC ratio and visceral adiposity index (VAI) based on following formula:
VAI = [WC/36.58 + (1.89×BMI)] × (TG/0.81) × (1.52/HDL-C),
where TG stands for triglycerides, and HDL-C is high-density lipoprotein cholesterol [9].
To identify the presence and severity of climacteric syndrome, we employed the modified menopausal index (MMI) by Kupperman as adapted by E.V. Uvarova (1983). Laboratory tests included measuring biochemical parameters of blood plasma (glucose, uric acid, alanine aminotransferase, aspartate aminotransferase, total cholesterol, HDL-C, low-density lipoproteins [LDL], and TG) and hormonal parameters (follicle-stimulating hormone [FSH], prolactin, total testosterone, sex hormone-binding globulin [SHBG], thyroid-stimulating hormone [TSH], and insulin). Free androgen index (FAI) was calculated as:
FAI = total testosterone (nmol/L)/SHBG (nmol/L) × 100.
Homeostasis model assessment of insulin resistance (HOMA-IR) index was calculated using the following formula:
HOMA-IR = fasting glucose (mmol/L) × fasting insulin (mIU/L)/22.5.
For measuring biochemical parameters, we used the ChemWell automated biochemical analyzer (Awareness Technology, USA). The content of FSH, prolactin, total testosterone, SHBG, TSH, and insulin was measured in the serum of venous blood by the Adaltis Personal Lab automated ELISA analyzer (Italy) using the enzyme-linked immunosorbent assay method in accordance with the recommendations of the manufacturer of the reagent kits (Alcor-Bio and Vector-Best, Russia).
The presence of MS was established according to 2013 clinical recommendations by the Russian Federation Ministry of Healthcare, Recommendations for the Management of Patients with Metabolic Syndrome.
Statistical analyses were performed using the StatTech v. 4.8.0 software (StatTech LLC, Russia). Quantitative indicators were assessed for compliance with the normal distribution using the Shapiro-Wilk test. Quantitative indicators with normal distribution are presented as arithmetic means (M) and standard deviations (SD). In the absence of normal distribution, quantitative data were described using the median (Me) and the lower and upper quartiles [Q₁; Q₃]. Categorical data are presented as counts and percentages. Comparison of two groups by a normally distributed quantitative parameter with equal variances was performed using Student’s t-test; in other cases, the nonparametric Mann–Whitney U test was employed. Comparison of percentages in the analysis of four-field contingency tables was performed using Pearson’s χ² criterion (for values of the expected phenomenon exceeding 10). Construction of a predictive model for the probability of a certain outcome was executed using the multiple logistic regression method. The measure of certainty indicating the part of the variance that can be explained by logistic regression was the Nagelkerke’s R² coefficient. The ROC curve analysis method was used to assess the discrimination power of quantitative features in predicting a certain outcome. The cutoff point value of a quantitative feature was determined by the highest value of the Youden Index. Differences were considered statistically significant at p<0.05.
Results
The median age of women with MS was 57 [54; 58] years, while in the control group it was 55 [51; 57] years (p=0.057). The duration of postmenopause (6 [3; 9.5] years in Group 1 and 5 [2; 7] years in Group 2) did not differ between the groups (p=0.106). Anthropometric parameters and severity of climacteric syndrome are presented in Table 1. Female patients with MS had higher overall MMI scores (p=0.003), as well as higher scores of the neurovegetative (p=0.003), and metabolic and endocrine (p<0.001) MMI components.
Table 1. Anthropometric parameters and indices of the modified menopausal index in the study groups
Parameter | Group | р | |
1: women with MS (n=35) | 2: women without MS (n=47) | ||
Body weight, kg, Me [Q₁; Q₃] | 80.00 [76.45; 89.00] | 64.50 [58.00; 73.50] | <0.001 |
Body mass index, Me [Q₁; Q₃] | 31.23 [29.15; 34.09] | 23.92 [22.26; 26.91] | |
Waist circumference (WC), cm, M (SD) | 99.14 (9.52) | 82.02 (9.05) | |
WC/Hip circumference, M (SD) | 0.94 (0.08) | 0.83 (0.08) | |
Visceral adiposity index, Me [Q₁; Q₃] | 1.91 [1.04; 3.27] | 1.08 [0.77; 1.41] | |
Modified menopausal index, score, Me [Q₁; Q₃] | 23 [16.5; 27.5] | 15 [8.5; 22.5] | 0.003 |
Neurovegetative symptoms, score, Me [Q₁; Q₃] | 11 [9.5; 16] | 7 [4; 13] | |
Metabolic and endocrine symptoms, score, Me [Q₁; Q₃] | 5 [3; 6] | 3 [1; 4] | <0.001 |
Psychoemotional symptoms, score, Me [Q₁; Q₃] | 5 [3; 7.5] | 5 [3; 6] | 0.368 |
HTN is one of the MS components; therefore, its frequency in Group 1 was predictably higher. However, more frequent presence of somatic comorbidity in Group 1 is noteworthy: the most typical combination was HTN and cholelithiasis (Table 2).
Table 2. Incidence of major chronic somatic diseases in the study groups
Disease | Group, count (%) | р | |
1: women with MS (n=35) | 2: women without MS (n=47) | ||
Hypertension (HTN) | 26 (74.3) | 6 (12.8) | <0.001 |
Osteoarthritis | 5 (14.3) | 6 (12.8) | 1.000 |
Arrhythmia | 1 (2.9) | 2 (4.3) | 1.000 |
Coronary artery disease, angina | 2 (5.7) | 0 | 0.179 |
Cholelithiasis | 6 (17.1) | 3 (6.4) | 0.161 |
Thyroid disease | 5 (14.3) | 6 (12.8) | 1.000 |
Respiratory diseases | 4 (11.4) | 4 (8.5) | 0.718 |
Varicose veins | 5 (14.3) | 6 (12.8) | 1.000 |
Anemia | 2 (5.7) | 5 (10.6) | 0.693 |
Presence of ≥2 somatic diseases | 24 (68.6) | 13 (27.7) | <0.001 |
Combination of HTN and cholelithiasis | 5 (14.3) | 0 | 0.012 |
Combination of HTN and thyroid disease | 4 (11.4) | 2 (4.3) | 0.394 |
Analysis of cardiometabolic parameters revealed higher mean concentrations of glucose, insulin, uric acid, triglycerides and lower concentrations of HDL in the group of women with MS. At the same time, these parameters did not go beyond the reference values in both groups, while total cholesterol and LDL were above normal in both groups. As for hormonal status, we observed lower concentrations of FSH (p=0.008) and prolactin (p=0.011) in women with MS vs. the control group. Levels of TSH, total testosterone, SHBG, and FAI did not differ statistically significantly between the groups (Table 3).
Table 3. Intergroup comparisons of cardiometabolic parameters and hormonal profile
Parameter | Group | р | |
1: women with MS (n=35) | 2: women without MS (n=47) | ||
Glucose, mmol/L, Me [Q₁; Q₃] | 5.4 [5.1; 6.1] | 5.1 [4.7; 5.6] | 0.013 |
Insulin, mIU/L, Me [Q₁; Q₃] | 11.2 [8.0; 16.4] | 6.6 [4.9; 8.3] | <0.001 |
Homeostasis model assessment of insulin resistance index, Me [Q₁; Q₃] | 2.8 [1.9; 4.4] | 1.5 [1.1; 2.0] | |
Alanine aminotransferase, U/L, Me [Q₁; Q₃] | 25.6 [18.2; 31.7] | 20.0 [16.8; 27.0] | 0.066 |
Aspartate aminotransferase, U/L, M (SD) | 23.4 (6.9) | 23.8 (5.9) | 0.854 |
Uric acid, μmol/L, M (SD) | 295.6 (61.6) | 230.0 (69.0) | <0.001 |
Total cholesterol, mmol/L, Me [Q₁; Q₃] | 6.7 [5.7; 7.9] | 6.2 [5.3; 6.9] | 0.090 |
Low-density lipoproteins, mmol/L, Me [Q₁; Q₃] | 4.5 [3.4; 5.2] | 3.9 [3.0; 4.5] | 0.074 |
High-density lipoproteins (HDL), mmol/L, M (SD) | 1.5 (0.4) | 1.7 (0.4) | 0.01 |
Non-HDL cholesterol, mmol/L, Me [Q₁; Q₃] | 5.1 [4.0; 5.9] | 4.3 [3.6; 5.1] | 0.019 |
Triglycerides, mmol/L, Me [Q₁; Q₃] | 1.5 [0.9; 2.0] | 1.0 [0.8; 1.2] | <0.001 |
Follicle-stimulating hormone, IU/mL, M (SD) | 67.9 (31.5) | 87.6 (33.0) | 0.008 |
Prolactin, mIU/L, | 189.0 [126.0; 279.5] | 273.0 [185.8; 421.6] | 0.011 |
Total testosterone, nmol/L, Me [Q₁; Q₃] | 0.8 [0.4; 1.0] | 0.5 [0.4; 0.9] | 0.391 |
Sex hormone-binding globulin, nmol/L, Me [Q₁; Q₃] | 31.8 [21.0; 50.2] | 37.4 [16.9; 82.7] | 0.536 |
Free androgen index, %, Me [Q₁; Q₃] | 2.2 [1.2; 3.7] | 1.8 [0.8; 4.1] | 0.314 |
Thyroid-stimulating hormone, μIU/mL, Me [Q₁; Q₃] | 1.3 [0.9; 1.8] | 1.5 [0.8; 1.9] | 0.966 |
We developed a model to quantify the relationship of clinical and laboratory parameters with the likelihood of MMS presence in women in their early postmenopausal period by means of using stepwise multiple logistic regression (Table 4).
Table 4. Logistic regression results
Variable | Odds ratio | 95% confidence interval | p |
Presence of ≥2 somatic diseases: | 9.108 | 1.828–45.422 | 0.007 |
Visceral adiposity index | 5.345 | 1.470–19.453 | 0.011 |
Modified menopausal index, score | 1.087 | 1.007–1.174 | 0.033 |
Follicle-stimulating hormone, mIU/mL | 0.968 | 0.944–0.993 | 0.012 |
Prolactin, mIU/L | 0.993 | 0.987–0.998 | 0.012 |
The developed regression model is statistically significant (p<0.001). Nagelkerke’s coefficient of determination is 64.4%. Hence, the observed relationship is described by the following equations:
P =1 / (1 + e-z)
z = –0.968 + 2.209 X1 + 1.676 X2 + 0.084 X3 – 0.033 X4 – 0.007 X5,
where P is the estimate of the probability of MS presence; z is the value of the logistic function, X1 is the presence of ≥2 somatic diseases, X2 is the VAI, X3 is the MMI (score), X4 is FSH concentration (mIU/mL) and X5 is prolactin content (mIU/L).
For additional assessment of the model quality, we carried out ROC analysis. The quality of the ROC analysis was assessed based on the area under the ROC curve (AUC). The latter was 0.909; 95% confidence interval 0.835–0.984 (p<0.001) (Figure).
The threshold value of probability estimates P at the cutoff point, which corresponded to the highest value of the Youden Index, was 0.352. The presence of MS was predicted at P exceeding this value or equal to it. The sensitivity and specificity of the resulting predictive model were 87.1% and 78.6%, respectively.
Discussion
According to large-scale studies conducted on various populations, the VAI is an accurate, inexpensive and widely available marker for MS screening [10, 11]. In our study, the VAI demonstrated high predictive accuracy for measuring the risk of MS in women in their early postmenopausal period. Since the diagnosis of MS includes anthropometric and biochemical parameters, the VAI is a more reliable tool than the WC to HC ratio, since it includes biochemical parameters (such as HDL and TG) rather than solely body weight and fat distribution (BMI and WC).
The Kupperman MMI modified by E.V. Uvarova separately analyzes three categories of symptoms: neurovegetative, metabolic and endocrine and psychoemotional. It provides a more detailed characterization of the clinical picture of climacteric syndrome and the severity of its individual components, which has a number of advantages in terms of the comprehensive assessment of MMS risk factors. In our study, MMI showed higher overall scores, as well as its individual component scores (neurovegetative, metabolic and endocrine) in women with MMS vs. women without this syndrome.
Several studies established a higher prevalence of MS in postmenopausal vs. perimenopausal women, which suggests a protective role of sex hormones in the pathogenesis of CVD [2, 4]. In two systematic reviews assessing the relationship between endogenous estrogens and CVD risk, conflicting results were obtained: some studies confirmed the expected cardioprotective effects of estrogens, while other studies failed to do so, and some of them showed that endogenous estrogens may be associated with a higher risk of CVD [12, 13]. These contradictory results can be explained by the data of the Study of Women’s Health Across the Nation (SWAN), which revealed four different trajectories of changes in estradiol (E2) levels and three trajectories of changes in FSH levels during the menopausal transition: they had a close correlation with BMI and ethnicity. Regardless of their ethnicity, obese women exhibited a flat trajectory (slow decrease) of E2 and a slow increase of FSH. Higher E2 levels after menopause reflect enlarged adipose tissue mass with accelerated aromatization in obese women (the conversion of testosterone to E2), while increased BMI has an inhibitory effect on gonadotropin secretion [14].
In our study, FSH content was significantly lower in women with MS vs. control women. Several studies assessed the relationship between FSH levels and MS. E. Jung et al. (2020) established that serum FSH levels inversely correlated with cardiometabolic risk factors in Korean postmenopausal women, and the researchers suggested that low FSH levels may be a predictor of CVD in them [15]. S.W. Lee et al. showed that FSH levels were inversely associated with HOMA-IR, body fat mass measured by bioelectrical impedance, and the incidence of MS; but were directly associated with serum adiponectin levels [16]. According to a 2023 study by Y. Chen et al., high serum FSH and luteinizing hormone concentrations were associated with a lower risk of MS in postmenopausal women [17]. Conflicting results were obtained in a 2020 study by C. Zhang et al., in which high FSH levels were directly associated with the risk of MS in perimenopausal women [18].
In our study, women with MS had lower prolactin levels compared with women without metabolic disorders. A 2022 analysis of studies by Y. Macotela et al. revealed that moderately high serum prolactin levels were associated with a lower prevalence of metabolic diseases [19]. Another analysis demonstrated that both low and high prolactin levels were associated with metabolic disorders and the development of MS [20]. Taking into account the conflicting results showing both favorable and detrimental effects of prolactin on metabolic homeostasis, the role of the latter in the development of metabolic disorders in postmenopausal women requires further clarification.
Conclusion
The developed predictive model of MMS development with the assessment of predictors of somatic comorbidity, menopausal symptoms, and anthropometric and laboratory parameters allows identifying patients with MMS in their early postmenopausal period, and therefore, with an existing elevated cardiometabolic risk and a more severe course of climacteric syndrome. This must be taken into account for the timely initiation of personalized treatment. The limitations of our study are caused by its small sample size. Therefore, studies with greater power and a long observation period are required to further confirm the effectiveness of the model.
Conflict of interest. The authors declare no conflicts of interest.
Author contributions. All authors contributed equally to the preparation of the article.
Kytikova OY, Antonyuk MV, Kantur TA, et al. Prevalence and biomarkers in metabolic syndrome. Obesity and Metabolism 2021; 18(3): 302-12. (In Russ.) https://www.doi.org/10.14341/omet12704
Hallajzadeh J, Khoramdad M, Izadi N, et al. Metabolic syndrome and its components in premenopausal and postmenopausal women: A comprehensive systematic review and meta-analysis on observational studies. Menopause 2018; 25(10): 1155. https://www.doi.org/10.1097/GME.0000000000001136
Ebtekar F, Dalvand S, Gheshlagh RG. The prevalence of metabolic syndrome in postmenopausal women: A systematic review and meta-analysis in Iran. Diabetes Metab Syndr. 2018; 12(6): 955-60. https://www.doi.org/10.1016/j.dsx.2018.06.002
Pu D, Tan R, Yu Q, et al. Metabolic syndrome in menopause and associated factors: A meta-analysis. Climacteric 2017; 20(6): 58391. https://www.doi.org/10.1080/13697137.2017.1386649
Everson-Rose SA, Barinas-Mitchell EJM, El Khoudary SR, et al. Adipokines and subclinical cardiovascular disease in post-menopausal women: Study of women’s health across the nation. J Am Heart Assoc. 2021; 10(7): e019173. https://www.doi.org/10.1161/JAHA.120.019173
Muka T, Oliver-Williams C, Colpani V, et al. Association of vasomotor and other menopausal symptoms with risk of cardiovascular disease: A systematic review and meta-analysis. PLoS One 2016; 11(6): e0157417. https://www.doi.org/10.1371/journal.pone.0157417
Armeni A, Anagnostis P, Armeni E, et al. Vasomotor symptoms and risk of cardiovascular disease in peri- and postmenopausal women: A systematic review and meta-analysis. Maturitas. 2023; 171: 13-20. https://www.doi.org/10.1016/j.maturitas.2023.02.004
Harlow SD, Gass M, Hall JE, et al. Executive summary of the Stages of Reproductive Aging Workshop + 10: Addressing the unfinished agenda of staging reproductive aging. Menopause 2012; 19(4): 387. https://www.doi.org/10.1097/gme.0b013e31824d8f40
Amato MC, Giordano C, Galia M, et al. Visceral adiposity index. Diabetes Care 2010; 33(4): 920-2. https://www.doi.org/10.2337/dc09-1825
Bijari M, Jangjoo S, Emami N, et al. The accuracy of visceral adiposity index for the screening of metabolic syndrome: A systematic review and meta-analysis. Int J Endocrinol. 2021; 2021: 6684627. https://www.doi.org/10.1155/2021/6684627
Lazzer S, D’Alleva M, Isola M, et al. Cardiometabolic Index (CMI) and Visceral Adiposity Index (VAI) highlight a higher risk of metabolic syndrome in women with severe obesity. J Clin Med. 2023; 12(9): 3055. https://www.doi.org/10.3390/jcm12093055
Crandall CJ, Barrett-Connor E. Endogenous sex steroid levels and cardiovascular disease in relation to the menopause: A systematic review. Endocrinol Metab Clin North Am. 2013; 42(2): 227-53. https://www.doi.org/10.1016/j.ecl.2013.02.003
Gulamhusein N, Miranda KT, Ahmed SB, et al. Measurements of postmenopausal serum estradiol levels and cardiovascular events: A systematic review. CJC Open 2024; 6(2 Part B): 347-54. DOI: https://www.doi.org/10.1016/j.cjco.2023.11.010
Tepper PG, Randolph JF, McConnell DS, et al. Trajectory clustering of estradiol and follicle-stimulating hormone during the menopausal transition among women in the Study of Women’s Health Across the Nation (SWAN). J Clin Endocrinol Metab. 2012; 97(8): 2872-80. https://www.doi.org/10.1210/jc.2012-1422
Jung ES, Choi EK, Park BH, et al. Serum follicle-stimulating hormone levels are associated with cardiometabolic risk factors in post-menopausal Korean women. J Clin Med. 2020; 9(4): 1161. https://www.doi.org/10.3390/jcm9041161
Lee SW, Hwang IS, Jung G, et al. Relationship between metabolic syndrome and follicle-stimulating hormone in postmenopausal women. Medicine 2022; 101(18): e29216. https://www.doi.org/10.1097/MD.0000000000029216
Chen Y, Wang C, Sun B, et al. Associations of follicle-stimulating hormone and luteinizing hormone with metabolic syndrome during the menopausal transition from the National Health and Nutrition Examination Survey. Front Endocrinol. 2023; 14: 1034934. https://www.doi.org/10.3389/fendo.2023.1034934
Zhang C, Zhao M, Li Z, et al. Follicle-stimulating hormone positively associates with metabolic factors in perimenopausal women. Int J Endocrinol. 2020; 2020: 7024321. https://www.doi.org/10.1155/2020/7024321
Macotela Y, Ruiz-Herrera X, Vázquez-Carrillo DI, et al. The beneficial metabolic actions of prolactin. Front Endocrinol. 2022; 13: 1001703. https://www.doi.org/10.3389/fendo.2022.1001703
Zaidalkilani AT, Al-Kuraishy HM, Al-Gareeb AI, et al. The beneficial and detrimental effects of prolactin hormone on metabolic syndrome: A double-edge sword. J Cell Mol Med. 2024; 28(23): e70067. https://www.doi.org/10.1111/jcmm.70067