Correlation of Body Mass Index and Waist Hip Ratio with lipid and hormone profile in women in menopausal transition

© 2014 Lejla Mešaliæ; licensee University of Sarajevo Faculty of Health Studies. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. UNIVERSITY OF SARAJEVO FACULTY OF HEALTH STUDIES ABSTRACT


INTRODUCTION
Menopausal transition is a period in life of a woman, with undefi ned beginning and duration, but with well-known changes in female organism.Menopausal transition is a period characterized by psychic, somatic and reproductive changes of capabilities at women (1).Hormonal profi le is changing during menopausal transition.Increased level of folliclestimulating hormone (FSH) stimulates follicles to grow, but those follicles mainly fail to reach the fi nal growth and maturity, which results with frequent anovulation.Progesterone production is 60 % lower than in reproduction period.In menopause, metabolic changes occur in diff erent tissues and organs as a result of changed hormonal profi le (2).Fat tissue is not only a passive fat depot refl ecting the energy balance and thermo-regulation, but is also a signifi cant endocrine organ (3).Th e main source of estrogen in menopause is estron generated from androstendion in peripheral tissues.Conversion processes of androgen to estrogens in menopause are not performed in fat tissue only, but also in central nervous system (4,5).One of the most important fat tissue hormones is leptin.During menopause, not only leptin, but also decreased level of growth hormone, E2 and androgens lead to changes of lipogenesis and lipolysis mechanisms, which lead to characteristic distribution of fat tissue in menopause (centripetal obesity).Lipid profi le during menopausal transition changes, but all mechanisms of those changes are not clarifi ed.One of the most important factors in that mechanism is the change of fat tissue distribution in postmenopause.Total cholesterol level increases through menopausal transition, with highest values in menopause.LDLD increases during menopausal transition, but it decreases after the menopause.Th e age is signifi cantly related to the changes in the triglyceride, total cholesterol and LDL level, while body mass index (BMI) is signifi cantly related to the changes of level of triglyceride, LDL, HDL (6).With menopause, HDL concentration decreases and HDL structure changes.HDL concentration is inversely proportional to abdominal obesity level (7).Menopause, age and increased distribution of abdominal fat tissue are three independent and important factors violating the lipoprotein profi le from the beginning of menopausal transition (8).Age, BMI and menopausal status are signifi cant indicators of the circulating lipoprotein level during menopausal transition (9).BMI is used to calculate the body surface and estimate the body weight.Doctors use BMI to estimate the risk of body weight and cardiac disease.New analysis of 40 studies, published in the Lancet magazine (10), which includes 250 000 patients, shows that the patients with BMI below 20 have higher mortality risk caused by cardiovascular diseases than the obese patients with BMI 30-35.Th e assessees who smoke have somewhat lower total cholesterol and signifi cantly lower HDL, although there is no diff erence in BMI distribution between the assessees who smoke and those who do not (11).E2 level is much lower at the obese, premenopausal women than at the premenopausal women of normal body weight.FSH level is also lower at the obese postmenopausal women than at the non-obese postmenopausal women (12).BMI impact to E2 and FSH varies in dependence on menopausal profi le (13).Waist-hip ratio, WHR is statistically important predictor of LDL concentration and cholesterol in relation to BMI at women in menopause, which confi rms the importance of fat tissue distribution as a risk factor for cardiovascular diseases at this group of women (14).Sultan and associates (15) state that waist-hip ratio may be used as screening for identifi cation of postmenopausal women with higher cardiovascular risk.Th e aim of this study was to determine the diff erences and connection of BMI and WHR with lipid and hormonal profi le of women in premenopause, perimenopause and postmenopause.

METHODS
Th is prospective, comparative study was conducted in period November 2009 -December 2010 and it included a total of 150 assessees aged 40-55, divided into three groups of 50 assessees each (premenopause, perimenopause and postmenopause).Th e assessees comprised women volunteers who agreed in writing to be included in the study.Each assessee was introduced with the character of the study implemented according to the generally accepted ethical standards for medical research.Qualifi cation criteria for inclusion into the study were: that the candidates do not take hormonal supplemental therapy, that they do not take medicines which could aff ect the lipid profi le, that they do not consume more than twenty cigarettes a day, that their BMI does not exceed 35kg/m².Processing of assessees comprised three phases: interview, taking blood sample, measuring BMI and WHR.For the purpose of analysis, the vein blood sample from cubital vein was taken, after which the blood was centrifuged, and obtained serum divided into two test tubes.Th e same day, concentration of lipid, lipoprotein, FSH, LH and E2 was determined.Concentration of the total cholesterol, triglyceride, LDL, HDL, VLDL, apo A, apo B and Lp (a) was also determined.Total cholesterol and triglyceride concentration was determined using enzymatic method at SIEMENS «Dimension RxL» machine.HDL and LDL were determined at the same machine utilizing direct "homogenous" method.VLDL was calculated with formula: VLDL = total cholesterol -HDL -LDL apo A, apo B and Lp (a) values were determined using immunoturbidimetry method at SIEMENS «Dimension RxL» machine, and reagents of the company "SENTINEL" were used.FSH, LH and E2 values were determined utilizing the hemiluminiscence method.Th e assessees' body weight and height were measured at the "SECA"scale with meter.Fattening status was assessed based on the Qeuetelet index (Devenport-Kaup modifi cation) or BMI where: BMI= body mass in kg/height in m².Based on the waist measuring at the narrowest place and hip measuring at the widest place, WHR was calculated according to the following formula: WH ratio = waist (cm)/hip (cm).

Statistical analysis
To compare the numerical variables among the assesse groups we used the variance analysis (ANO-VA) or Kruskal-Wallis non-parametric alternative.To analyse relation of BMI, WHR, menarche and reproductive age factors with lipid and hormonal profi le, we used linear regression model.Each factor was tested in bivariate model, adjusted to the group of assessees.Results are presented at regression coeffi cient with related 95% reliability interval (IP).Statistical signifi cance was confi rmed at p<0.05.Statistic programme PASW 18 (SPSS Inc., Chicago, Illinois, USA) was used for data processing.

RESULTS
Th e assessees included in the study were aged 40 to 55, with average age 48.1 ± 3.9.Statistically, the age between the groups diff ered signifi cantly: the assessees in perimenopause were 2 years older in average than those in premenopause (95% IP: 0.6-3.5 years), and the assessees in postmenopause were four years older in average than those in perimenopause (95% IP: 2.6-5.4 years) (Table 1).In addition, the assessees from the second group had somewhat lower WHR than the assessees from the third group, although the diff erence was not statistically signifi cant (average diff erence = 0.03, 95% IP: -0.003-0.06).Th ere is a signifi cant negative relation between BMI and HDL at assessees in premenopause (Table 2).If the regression coeffi cient had been standardised to interquartile value of BMI change (IQR=5), the diff erence of 0.18 mmol/L in medium HDL concentration between low and high BMI would have been suggested.Among the assessees in postmenopause, there is a signifi cant relation between BMI and cholesterol, as well as between BMI and VLDL.For interquartile value of BMI change, standard regression coeffi cient shows the increase of 0.67 mmol/L in medium cholesterol concentration and increase of 0.22 mmol/L in medium VLDL concentration.Th ere are no proofs of any relation between other lipid and BMI parameters.Multivariate regression analysis showed that the only important lipid fraction aff ected by BMI is HDL cholesterol.Considering the BMI value in relation to apo A, results indicate that there is a correlation between the groups of assessees in perimenopause and post-  Group I = assesses in premenopause; group II = assesses in perimenopause; group III = assesses in postmenopause.Group I = assesses in premenopause; group II = assesses in perimenopause; group III = assesses in postmenopause.
menopause according to the fi rst signifi cance order (p<0.001),while between the assessees in premenopause and postmenopause there is no signifi cant diff erence (p=0.516).Correlating BMI value in relation to apo B, there was statistically signifi cant diff erence between the assessees in perimenopause and premenopause (p=0.003), while between the assessees in postmenopause and premenopause there was no statistically signifi cant diff erence (p=0.367).
Obtained results regarding Lp (a) indicate the signifi cant correlation in relation to BMI, only among the asssessees in postmenopause and premenopause (p<0.001).Of three analysed hormones, signifi cant relation with BMI is noted only for E2 at women in premenopause (Table 3).For increase of unit in BMI, regression coeffi cient suggests 8.9-pecent decrease in medium E2 concentration.Th ere is a signifi cant negative relation between WHR and HDL at the assessees in premenopause and perimenopause (Table 4).At WHR change of 0.05, regression coeffi cient shows the diff erence of 0.11 mmol/L in medium HDL concentration at the assessees in premenopause, and the diff erence of 0.09 mmol/L in medium HDL concentration at the assessees in perimenopause.Th e only signifi cant relation between WHR and hormone parameters is proven for E2 at the assessees in premenopause (Table 5).Medium estradiol concentration decreased 1.29 times at WHR increase for 0.05.

DISCUSSION
Numerous studies addressed the issue of lipid changes during menopausal transition or related to menopausal changes of endogenous hormones.Although the numerous researches have been done on behaviour of lipids and lipoprotein during menopausal transition, their relation with sexual hormones and body mass is still being assessed.Th e average age in time of the last menstruation among the assessees included in this study is 48.One of the objectives of this study was to determine the relation of BMI and WHR with lipid and hormonal profi le of assessees in menopausal transition, and the results show that at the assessees in postmenopause, there is a positive correlation between BMI and total cholesterol, and BMI and VLDL cholesterol, and negative correlation between BMI and HDL in premenopause.
Th ere are no proofs of any relation between other lipid and BMI parameters.Other studies also indicate that women with higher BMI are exposed to the risk of higher lipid level, although skinner women can also have higher hormone-related LDL cholesterol during menopausal transition (16).Duration of postmenopause, as well as BMI in similar studies do not show signifi cant correlation with lipid, lipoprotein and Lp (a) concentration, while WHR shows signifi cant positive correlation with cholesterol, LDL and apo B ( 14) concentrations.According to Yamamoto and associates (11), there is no signifi cant correlation between BMI and serum Lp (a) value, and medium Lp (a) value shows the possible trend of increase at women over 40.Th is study showed signifi cant correlation between BMI and Lp (a) at the assessees in postmenopause and premenopause, which is confi rmed by contemporary knowledge on impact of higher Lp (a) to the increase of cardiovascular risk.We determined a signifi cant negative relation between WHR and HDL cholesterol at the assesses in premenopause and perimenopause, while in relation to other lipids, there were no correlations.According to Mešalić (14), among women in premenopause, WHR has a signifi cant negative correlation with HDL and Apo lipoprotein A concentration, which, considering the role of these two lipoproteins in occurrence of cardiovascular diseases, confi rms that even women with regular menstruations and higher WHR have the risk of cardiovascular diseases.Results of this study indicate negative correlation between WHR and estradiol at the assesses in premenopause and negative correlation between BMI and estradiol in premenopause.However, with increase of WHR and BMI, the estrogen level decreases in premenopause, based on which the women with higher cardiovascular risk may be identifi ed.Assessment results of the comprehensive study conducted by the National Health and Nutrition Examination Survey among the assessees aged 35-60 in period 1999-2002 (17) show that there are no signifi cant diff erences in total cholesterol, triglyceride, HDL, LDL cholesterol levels adjusted to the age, among menopausal periods at the group of women with normal BMI.Th e diff erence in HDL cholesterol values was noticed at the groups with normal and higher BMI.In the groups of assessees with normal BMI, LH and FSH hormones activ-ity was statistically diff erent than at the assessees in premenopause and perimenopause, and between the assessees in premenopause and postmenopause (17).SO, according to Azizi and Ainy (18) BMI, WHR and triglyceride value in blood does not show the signifi cant diff erences between the assessees in premenopause, perimenopause and postmenopause.Despite numerous studies, physiological role of Lp (a) has not been determined.Its atrogenous infl uence is attributed to enormous similarity of Lp(a) and LDL in structure, which are very cholesterol rich.In this study, Lp (a) was the highest at the assessees in premenopause and postmenopause.Somewhat diff erent results are presented by Kim and associates (19), who obtained lower Lp (a) values at perimenopausal assessees than at postmenopausal assessees.

CONCLUSION
BMI has a positive correlation with cholesterol concentration and VLDLD in postmenopause, positive correlation with apo A in perimenopause and postmenopause, and positive correlation with Lp (a) and apo B in premenopause and perimenopause, while with HDL and estradiol concentration it has a negative correlation in premenopause.WHR has negative correlation with HDL concentration in premenopause and perimenopause, and negative correlation with estradiol concentrations in premenopause.However, with increase of WHR and BMI, the estrogen level in premenopause decreases, based on which the women with higher cardio-vascular risks may be identifi ed.

TABLE 1 .
Main features of assessees11 Values are arithmetic median value ± SD as well as minimum and maximum.Group I= assessees in premenopause; group II= assessees in perimenopause; group III= assessees in postmenopause.

TABLE 2 .
Regression coeffi cient for lipid profi le in relation to BMI.

TABLE 3 .
Regression coeffi cient for hormonal profi le in relation to BMI.
1Values are arithmetic median value ± SD as well as minimum and maximum.Group I= assessees in premenopause; group II= assessees in perimenopause; group III= assessees in postmenopause.Group I = assesses in premenopause; group II = assesses in perimenopause; group III = assesses in postmenopause.

TABLE 4 .
Regression coeffi cient for lipid profi le in relation to WHR, calculated for WHR change of 0.05.

TABLE 5 .
Regression coeffi cient for hormonal profi le in relation to WHR.