Journal of Obesity & Weight Loss Therapy
Open Access

Among parameters defining the health-related quality of life weight to height, the ratio is the oldest. Although the first instances of its applications were reported 183 years ago, its renaissance came in the early ’70s of the last century and is owed to Keys and coworkers who introduced the term Body Mass Index (BMI) [1]. Since then, BMI became the primary descriptor of obesity in thousands of scientific reports. Notwithstanding its popularity, a variety of studies have found shortcomings in its clinical applicability and opened an advent for the search of new clinically viable indices of obesity. Among those employed in practice, the waist circumference (WC) [2], the waist-tohip ratio (WHR) [3], and the waist-to-height ratio (WHtR) [2] provide an attractive alternative to BMI.

Over a decade ago, Aswell and Hsieh [4] delineated WHtR superior to Body Mass Index (BMI) for early warning of cardiovascular health risks. This observation was further confirmed by his study [2], indicating WHtR not only better than BMI but also WC and WHR for prediction of risk of cardiovascular diseases (CVD) in women. Notwithstanding the apparent clinical attractivity of WHtR, there is a dearth of reports on correlations between WHtR and serum lipid levels in elderly postmenopausal women. Thus there are only a few reports on relations between WHtR and serum lipid levels [5], total cholesterol (TC), triglycerides (TG), and high-density cholesterol (HDL-C) and all of those follow traditionally established WHtR threshold of 0.5 [5].

Although the exact BMI cutoffs indirectly defining cardiovascular disease risks are adopted by the World Health Organization, the analysis of scientific literature unfolds its fallibility in elderly women’s health assessment. It is due to the multitude of age-dependent factors among which the menopausal transition plays a pivotal role [3,6]. This observation is strengthened by reports indicating that adverse lipid levels changes are not associated with an increase in BMI [3]. However, they are coupled to a rise in WC [6], which in turn is negatively associated to the cardiovascular health (CVH) [2,3] and levels of water and fat-soluble vitamins [6].

To address the issues mentioned above, we undertook the study on relations among BMI, WC, WHR, WHtR, blood pressure, and serum lipid levels [TC, TG, HDL-C, and low-density lipoprotein (LDL-C)] in a healthy group of postmenopausal women.

Participants

This study was performed on postmenopausal women for whom the menopausal status was defined by the time passed from the last period; the postmenopausal women are these for whom the last period occurred >12 months before this study [3]. A total of 1,400 women living in Gdansk, Poland metropolitan area, were contacted and asked to participate in this study. Out of 1,480 eligible subjects, 450 refused to sign the consent form, 1,030 provided written consent, and only 645 took part in the experiment. The study sample encompassed the age range between 56 and 83 years, with the mean age of 6.97 years, the average height of 160.57 cm, and the average body mass of 69.16 kg. A general practitioner performed the health interview on the day of measurement. The health interview elucidated women who did not and do not suffer from congestive heart disease or kidney disease, and for whom the body mass is relatively stable, i.e., the bodyweight has not varied more than 2 kg over six months before this study. All the studied women were not in risk of hypertension. The provincial medical chamber approved the study protocol.

Anthropometric measurements

Anthropometric parameters including body mass (BM), height (Ht), waist circumference (WC), hip circumference (HC), Body Mass Index (BMI), waist-to-hip ratio (WHR), and waist-to-height ratio (WHtR) were collected during this study. Bodyweight was assessed to the nearest 0.2 kg body height to the nearest 0.5 cm. All circumferences were measured accordingly to the WHO protocol using a tailor’s tape measure. Waist circumference was measured at the midpoint between the lower margin of the least palpable rib and the top of the iliac crest. Hip circumference was measured around the widest portion of the buttocks, with the tape parallel to the floor. All circumferences were measured with an accuracy of 0.5 cm. BMI, WHR, and WHtR were calculated using the following formulas: BMI = BM (kg)/Ht² (m²), WHR = WC (cm)/HC (cm), and WHtR = WC (cm)/Ht (cm).

Serum lipids

An analysis of serum lipid levels was performed accordingly to the ATP III protocol. Lipid analysis was performed in a 12-hour fasting subject who was resting in a sitting position for at least five minutes before phlebotomy. Blood for serum analysis was collected in tubes without anticoagulant, for plasma analysis in tubes with EDTA. Total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), highdensity lipoprotein cholesterol (HDL-C), and triglycerides (TG), were determined using automatic biochemistry analyzer (Roche Cobas 8000 modular analyzer Series C701, Mannheim, Germany).

The mean of samples and statistical inferences were estimated using bootstrap analysis based on 10,000 replications. Relative differences between means were analyzed using “natural” relative difference employing natural logarithm, denoted as log percent (L%) [3].

The clustering of the data was a two-level process; the first level was hierarchical clustering was employed at the second level two-step clustering technique was used. The hierarchical cluster was obtained using Ward’s method for calculation of a distance between cluster pairs. The initial number of the cluster was assessed using a dendrogram. This procedure was followed by the TwoStep cluster. The quality of clusters was assessed using a silhouette plot.

Statistical differences were assessed using confidence intervals (CI) of means between specific groups of data estimated using Bootstrap analysis [2]. Bootstraps was obtained using repeated random samples of the same size as the original sample drawn with replacements from the original data. The number of bootstrap re-samples for the construction of CIs was equal to 10,000. The CIs were calculated by the bias-corrected and accelerated (Bca) method [2]. The statistically significant difference between samples was defined by the lack of overlap of 95% CIs of specific samples. Correlations between specific parameters were analyzed using a linear regression where predictors were always defined by an ordinate on a specific graph.

Table 1 comprises descriptive statistics of the study sample. Table 2 delineates the stratification of the study sample accordingly to currently defined standards for BMI, WC, WHR, and WHtR [2]. Table 3 compares the stratification of the study sample as a function of a level of TC, TG, HDL-C, and LDL-C. Figure 1A-G depicts the correlations between BMI, WC, HC, WHR, and WHtR. Table 4 encloses the stratification of BMI as a function of WHtR-cluster, and Table 5 envelopes the results of WHtR-cluster dependent differences in serum lipid levels.

Table 1: Anthropometric parameters of a study sample.

Table 2: Sample stratification accordingly to current golden standards for Body Mass Index (BMI), Waist Circumference (WC), Waist-to-Hip Ratio (WHR), and Waist-to-Height Ratio (WHtR).

Table 3: Sample stratification accordingly to the National Cholesterol Education Program.

Table 4: BMI stratification as a function of WHtR cluster.

Table 5: Differences in serum lipid levels stratified by WHtR clusters.

Figure 1 (A-Z): The correlations between BMI, WC, HC, WHR, and WHtR.

The age range of the studied sample is between 56 and 83 years, height range between 147 and 178 cm, body mass between 43.3 and 102.5 kg, waist circumference between 59 and 114 cm, hip circumference between 73 and 128 cm, BMI between 18.97 and 39.88, the waist-tohip ratio between 0.68 and 1.02, and the waist-to-height ratio between 0.39 and 0.70. There are 43 women within the normal (healthy) BMI range, 65 subjects within the normal WC range, 102 subjects within the normal WHR range, and 31 subjects within the normal WHtR range. Twenty-eight subjects are within the desirable TC range, 78 subjects within the normal TG range, 22 subjects within the normal HDL-C range, and 19 subjects within the optimal LDL-C range.

There are the following relations between BMI, WC, HC, WHR, and WHtR: a) WC → f(BMI), R2adj=72%, F(1,100)=248.848, p=.0, b) HC → f(BMI), R2adj=40%, F(1, 100)=85.263, p=.0,c) WHR → f(BMI), R2adj=8.1%, F(1,100)=8.603, p=.04,d) WHtR → f(BMI), R2adj=80%, F(1,100)=370.969, p=.0,e) HC → f(WC), R2adj=46%, F(1,100)=85.263, p=.0,f) WHtR → f(HC), R2adj=44%, F(1,100)=78.899, p=.0, and g) WC → f(WHtR), R2adj=90%, F(1, 100)=895.342, p=.0

WHtR clustering resulted in the three clusters: A) 0.39 ≤ cluster 1 < 0.50, B) 0.50 ≤ cluster 2 < 0.58, and C) 0.58 ≤ cluster 3 ≤ 0.7. The BMI ranges encompassed by specific WHtR cluster are as follows: cluster 1 encompass BMI from 21.88 kg/m² to 23.10 kg/m2, cluster 2 from 26.25 kg/m² to 27.56 kg/m², and cluster 3 from 31.37 kg/m² to 33.87 kg/ m2. The statistical analysis unfolds statistically significant differences between these BMI brackets (Table 4).

A similar analysis performed for serum lipid levels unfolds the lack of statistically significant differences for TC, TG, and LDL-C as a function of the WHtR cluster. However, there are statistically significant differences in HDL-C levels. Thus, WHtR cluster 1 encompasses HDL-C levels between 2.01 mmol/L and 2.23 mmol/L, cluster 2 between 1.75 mmol/L and 1.97 mmol/L, and cluster 3 between 1.54 mmol/L and 1.80 mmol/L.

In this study, we analyzed the relations between the waist-to-height ratio (WHtR), serum lipid levels, and BMI in the sample comprising postmenopausal women 56 to 83 years of age. We undertook this topic to extend our understanding of the relevancy of WHtR for the screening of cardio-metabolic risks in elderly (postmenopausal) women. This specific sample was also chosen because the plethora of current studies [1-4] unfolds the menopausal transition as a complex medical phenomenon encompassing intercalating age-, psychologicaland physiological- (hormonal) dependent changes. These parameters are also the direct cause of detrimental health changes resulting in depression and an increase in body mass and obesity.

A cursory review of parameters currently employed for prediction of cardiovascular risk unfolds height as the intrinsic parameter of the most popular two: WHtR and BMI. Although the recent meta-analysis of the clinical suitability for prediction of cardio-metabolic risks of waist-to-height ratio [2] showed WHtR better than WC or BMI, some of the recent studies unfolded descriptors not containing height, as an intrinsic parameter, i.e., WC and WHR, as potentially more informative for prediction of cardiovascular risk [2,3].

In this study, we observe and confirmed [1] that an increase in WC is positively and strongly correlated with an increase in BMI, indicating that both WC and BMI may potentially exchangeably be used for obesity-related health risks. Additionally, the battery of the presented observations rises - in our opinion - the need for further validation of WHtR as the universal mean for prediction of the cardiovascular risk.

Currently, all the reports traditionally suggest [2,3,7], 0.5 as the golden cut-off standard for health related WHtR changes. However, this study indicates the use of two-thresholds approach (WHtR<0.5: optimal, 0.50 ≤ WHtR<0.58 : borderline, and 0.58 ≤ WHtR : high) as more appropriate for prediction of cardiovascular risk. The thesis of two WHtR thresholds presented in this study is further strengthened via a strong correlation with WHO-BMI obesity scale in the following fashion: WHtR cluster 1 encompasses “normal” BMI range, cluster 2 “pre-obese” range, and cluster 3 “obese I” BMI range.

The results presented in this study lead to the following conclusions: (1) BMI and WC may exchange ably be used for prediction of the cardiovascular health, (2) WHtR cut-off value should be adjusted to the specific gender and age group as shown in this study for postmenopausal women.