Gender Differentials of Metabolic Syndrome in Bangladesh taking Menopause into Consideration

Objective:To assess the gender differences of metabolic syndrome (MetS) in Bangladesh and factors associated with it. Materials and methods: Cross sectional study was performed on 227 premenopausal women, 48 menopausal women and 217 men. The gender differentials were studied with measurement of waist circumference, blood pressure and fasting blood for triglyceride (TG), high density lipoprotein (HDL), blood sugar (FBS).Results:Premenopausal women had 22.9% (95% CI: 17.9-28.8%), menopausal women had 43.8% (95% CI: 30.7-57.7%), and the men had 53.9% (95% CI: 47.3-60.4%) prevalence of MetS giving a 1:1.7:2.1 ratio of premenopausal, menopausal women and men prevalence respectively. More than 95% of study population had low HDL irrespective of gender and menopause status. Obesity was 5.3% in premenopausal women, 14.6% in menopausal women and 54.4% in men. HTN was 10.6% in premenopausal women, 35.4% in menopausal women, and 19.4% in men. High TG was 42.5% in premenopausal women, 41.7%, in menopausal women and 62.0% in men. High FBS was 22.0% in premenopausal women, 37.5% in menopausal women and 30.4% in men. Conclusion: Compared to premenopausal women, MetS in men was two times higher and nearly two times higher in menopausal women, pulling menopausal women prevalence almost equal to men. Menopausal women had high prevalence of HTN while obesity was more prevalent in men. Dyslipidemia should be a concern for the policy makers in search of a prevention program.


Introduction
Metabolic syndrome (MetS) or syndrome X or the deadly quintet had been given several definitions for establishing the diagnosis. The most updated criteria were agreed upon by different organizations worldwide has dyslipidaemia, hypertension, increased fasting glucose and abdominal obesity 1 . MetS had become a global concern because its prevalence has increased worldwide. In US, MetS had increased by 35% between 1988 and 2012, with more than a third of US adults having MetS 2 . In Bangladesh, a systematic review found that the prevalence was 30% and it was higher in females 3 . People with MetS were more likely to suffer from a heart attack and stroke 4 . There were different factors that affect the prevalence of MetS such as age, race socioeconomic status, physical activity, diet, geography and gender 5 . Different studies revealed disparity in the prevalence and components of MetS in respect to gender. For example, central obesity was more in women. Also, high triglycerides and impaired glucose tolerance were more predominant in men 6 . Some studies found gender differences in the treatment and response to treatment like less weight loss in females suffering from MetS 5 . There were some studies in Bangladesh on MetS but the knowledge about gender differences was scarce.Some studies focused on rural women 7 , some only on urban 8 areas and some only on hospital outpatients 9 or only rural pre-and post-menopausal women 10 . But no study so far investigated the prevalence on men, premenopausal and post-menopausal women hence leaving a knowledge gap in this field. In this paper we aimed to provide important gender specific information about MetS in Bangladesh to reject the null hypothesis that there is no difference in the prevalence of MetS among men, premenopausal and post-menopausal women.

Design and setting
The re-analysis was done using the dataset from cross-sectional study 11 to assess the difference of MetS in men, premenopausal women and postmenopausal women. With proper ethical approval and planning, we collected samples from 492 willing participants from one urban and one rural area, comprising around 46% premenopausal and 10% menopausal women, while the rest 44% were males. To avoid redundancy, we only put the analysis plan which was unique for this study.

Data analysis
The entered data was assessed for the baseline differences along three categories of respondents. We then plotted the components of MetS in similar fashion, first using the continuous variable followed by dichotomous categorization of the components. The analysis further followed the ratio of the prevalence in men, pre-and postmenopausal women. Keeping premenopausal status as reference as its prevalence was the lowest (in result section), we ran the logistic regression (LR) for MetS with odds ratio (OR), initially unadjusted and then adjusted for the factors which were significantly associated in preliminary analysis. In addition, we constructed the Box and Whisker's plot to assess the components accumulation of metabolic syndrome in these three groups of respondents by age. The numeric variables were expressed as mean and standard deviation (±SD) and the categorical variables as frequency and percentage. We used Confidence Interval Analysis (CIA) software to calculate confidence interval (CI) of the prevalence. A p-value of ≤0.05 was considered significant and above the cut-off value was taken not significant (ns).

Results
The baseline features of the respondents (Table 1) showed that age was not homogeneously distributed along the three groups of respondents as assumed because postmenopausal women were older than men and premenopausal women. There was no difference of monthly income, daily water intake or daily sleeping hours among these groups. Males were more educated compared to the other two groups (p<0.001). On the other hand, premenopausal and menopausal women were more involved with heavy occupation compared to significantly more menwith sedentary occupation (p<0.001). Postmenopausal women were more from rural areas, contrary to premenopausal women and men, who were significantly more from urban areas (p<0.001).  When we looked at the final category of MetS, it was observed that the prevalence of MetS was 22.9% in premenopausal women (95% CI: 17.9-28.8%), followed by 43.8% in postmenopausal women (95% CI: 30.7-57.7%), followed by 53.9% in men as highest prevalence (53.9%, 95% CI: 47.3-60.4%) among these three.We calculated the ratio of the prevalence to get a 1:1.7:2.1 ratio of premenopausal, postmenopausal women and men prevalence respectively. Both the prevalence of men (p<0.001) and postmenopausal women (p=0.003) were significantly higher that the premenopausal women but men and menopausal women were homogeneous in this regard (p=0.20, data not shown). When we checked for unadjusted LR (Table 4), the menopausal status, rural area and smoking were significantly associated with MetS and heavy working occupation had significantly reverse association. In adjusted LR, men (<0.001) and menopausal status (0.01) appeared as independent risk factors of being associated with MetS, adjusted for other variables which were not significant. We looked at the Box and Whisker's plot of the components of MetS by age classified by these three groups of respondents. It shows the difference of age in acquiring the components of MetS in these three groups. The prevalence of at least one component in premenopausal women (Figure 1) was found at the median age of 32-33 years, gradually adding up components with higher age, while they become a candidate of  Our study was the first to explore gender differences of MetS and its correlates in a rural and urban sample from Bangladesh. Our findings emphasized that gender had a fundamental role in the prevalence of MetS; (a) overall, males had the highest prevalence of MetS (54%) followed by postmenopausal (44%) and premenopausal (23%) females, (b) being a male or a postmenopausal female were independent risk factors of MetS, (c) the distribution of MetS components varied by gender; males were found to be more obese, had higher TG and lower HDL, while postmenopausal females had higher waist circumference, were more hypertensive and diabetic.
We other countries could be due to the inclusion of premenopausal women with menopausal women which might pull up the prevalence higher than the men. In our study, the higher prevalence of MetS among the males might be attributed to the different lifestyle patterns followed by males and females; we found significantly more men involved in sedentary occupation compared to females who were more engaged in occupation involving heavy work.
The current study highlighted menopause, in addition to male gender, to be an independent risk factor to develop MetS. This finding ties well with previous studies to established menopause as an independent predictor of MetS regardless of age or other comorbidities [23][24][25][26][27] . McNeill AM et. al reported that postmenopausal status increased the risk of MetS by 60%, despite adjusting for potential confounders (e.g. age, BMI, socioeconomic status, physical inactivity) 25 . Our results are also in accordance with findings from rural Bangladesh by Jasmin S. et. al. 28 who studied 1802 rural women and showed MetS to be more prevalent among postmenopausal (39.3%) than in premenopausal (16.8%) females. Additionally, MetS was significantly higher in postmenopausal than premenopausal females (OR: 1.78, CI: 1.26-2.51, p= 0.001) 28 . The proposed explanations go beyond statistics alone to the physiological and biological changes associated with menopause which delineates the transitional phase into sex hormone-depleted state characterized by estrogen deficiency 29 .
Menopause is accompanied with a redistribution of body fat resulting in central obesity and insulin resistance which were found to largely increase the prevalence of MetS in postmenopausal females 30 . Additionally, several biological changes take place during menopause and might aid indeveloping insulin resistance and MetS including oxidative stress, inflammatory changes, and postmenopausal alterations in adipocytokine production and lipid utilization 31,32 . These facts broadly elucidate the obvious acceleration in cardiovascular diseases (CVD) in females post menopause 33 .
Males and females had different patterns in the components of MetS. In our study, postmenopausal females had higher waist circumference, were more hypertensive and diabetic. A similar pattern of results was obtained in different studies over the globe. For example, a study utilized the database of the American National Health and Nutrition Examination Survey (NHANES) between 1999 and 2004, showed that women had higher mean SBP than men with around 80% of those women being postmenopausal 34 . Also, women had higher waist circumference and central obesity, elevated total cholesterol, lower HDL and more insulin resistance 31,34 . Moreover, multiple other studies showed that females had a significant higher prevalence of abdominal obesity compared to males while males had a higher prevalence of high TG levels 33,35 . Several factors might explain such trends including the fact that postmenopausal women lack estrogen which is known to cause vasodilation and lower the risk of HTN and CVD in premenopausal women 34,36 . It also explains our findings of how males acquired almost all of the components of MetS earlier compared to females with similar age category highlighting Estrogen protective roles in premenopausal females. Though, aging alone is a risk factor for developing HTN and diabetes, blood pressure changes related to aging are more profound among females compared to males 37