- Email: [email protected]
New anthropometric indices or old ones: Which is the better predictor of body fat?
G Model DSX 650 No. of Pages 7
Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Diabetes & Metabolic Syndrome: Clinical Research & Reviews journal homepage: www.elsevier.com/locate/dsx
Original article
New anthropometric indices or old ones: Which is the better predictor of body fat?$ Elham Ehrampousha,b , Peyman Arastehc,** , Reza Homayounfara,b,* , Makan Cheraghpourd , Meysam Alipourd, Mohammad Mehdi Naghizadeha , Maryam hadibarhaghtalaba , Sayed Hosein Davoodie , Alireza Askarif , Jalaledin Mirzay Razaze a
Noncommunicable Diseases Research Center, Fasa University of Medical Sciences, Fasa, Iran Health Policy Research Center, Shiraz University of Medical Sciences, Shiraz, Iran c Department of MPH, Shiraz University of Medical Sciences, Shiraz, Iran d Nutrition and Metabolic Diseases Research Center, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran e National Nutrition and Food Technology Research Institute, Faculty of Nutrition Sciences and Food Technology, Shahid Beheshti University of Medical Sciences, Tehran, Iran f Shiraz University of Medical Sciences, Shiraz, Iran b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 14 August 2016 Accepted 22 August 2016 Available online xxx
Background: The percent and distribution of body fat are important factors in the pathogenesis of numerous diseases. Our aim was to investigate common anthropometric indices in their relationship with body fat content. Methods: In a cross-sectional study 1360 healthy individuals (580 men and 780 women) in a cluster sampling, from Ahvaz, Iran, body fat content (using bioelectrical impedance) and anthropometric measurements [weight, waist circumference, a body shape index, abdominal volume index, body adiposity index, conicity, body mass index, hip circumference, waist to hip ratio and waist to height ratio] was obtained. The ROC curve analysis was used to compare each index with body fat percent. Results: Significant difference was found between men and women in all anthropometric parameters (p < 0.001). Women displayed higher percentages in the overweight and obese categories (33.6% vs. 32.9% and 26.4% vs. 22.1%, respectively). In both men and women, the strongest correlations were seen between body fat percent and BMI, AVI and WHtR (r > 7.9 and p < 0.001). BMI, WHtR and AVI in men and BAI, BMI and WHtR in women showed the most accuracy for estimating body fat percent, respectively. Conclusion: All anthropometric parameters could predict body fat percent with relatively good power, however BMI, WHtR and AVI are more powerful predictors. Based on our findings, we suggest using the AVI and WHtR instead of other indexes, as they are better able to assess the accumulation of fat in the abdominal area and are able to more accurately assess body fat percent, which are indicators of chronic disease. ã 2016 Diabetes India. Published by Elsevier Ltd. All rights reserved.
Keywords: Anthropometry Obesity Body mass index Body fat Weight-height ratio
$ Implication for health policy makers/practice/research/medical education: Older anthropometric indexes such as BMI have remained the most commonly used indexes for estimating disease risk. Recently it has become evident that old anthropometric indices are limited in many aspects for body fat and disease risk estimation. Replacing those indices with better and newer anthropometric indices that do not have the limitations of previous indices will aid in the correct estimation and prediction of disease development and overall body fat percent which itself is an indicator of chronic disease. * Corresponding author at: Noncommunicable Diseases Research Center, Fasa University of Medical Sciences, Fasa, Iran ** Corresponding author. E-mail addresses: [email protected], [email protected] (P. Arasteh), [email protected] (R. Homayounfar).
http://dx.doi.org/10.1016/j.dsx.2016.08.027 1871-4021/ã 2016 Diabetes India. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
2
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
1. Introduction Obesity defined as excess body fat, is a chronic and complex disease, which has long been recognized as one of the biggest problems in the field of public health. Increase in the prevalence of obesity leads to increased prevalence of obesity related complications such as cardiovascular disease [1], diabetes [2], some types of cancer [3], psoriasis [4], adverse pregnancy outcomes [5], earlier mortality [6] and many other health conditions. Abdominal obesity also known as central obesity is an excess of fat located in the abdominal area [7]. Anthropometry refers to the measurement of the human individual and has been used for identifying and understanding human physical variations. These measurements have simple, easy and effective characteristics that make them the first choice for nutritional evaluations. These indexes include: body mass index (BMI), waist circumference (WC), waist to hip ratio (WHR), waist to height ratio (WHtR), skinfold thickness, dual-energy X-ray absorption (DXA) and hydrostatic densitometry [8,9]. The point that should be considered regarding anthropometric indicators, in addition to their ability in estimating obesity or body fat, is their ability to detect the accumulation of fat in the abdominal area, which according to most studies is the most important factor affecting disease [8]. The negative impact of abdominal obesity on health is well recognized and although there have been tremendous advances in measuring body fat, there is still much debate regarding the most clinically efficient method of body fat assessment. In this study, we investigated common anthropometric indices regarding body fat content and further compared the available indices in order to better understand the more clinically valid and reliable measure for adiposity. 2. Materials and methods 2.1. Subjects and study protocol In a descriptive cross-sectional study from 2013 to 2014, the residents of Ahvaz, Iran were considered for inclusion in the study. For selecting the study sample, a cluster sampling technique was used. Twenty six districts in Ahvaz city were considered for the initial clusters. Then, the first digit of the postal codes of the citizens relating to the municipal classification was utilized to choose a random sample from the households in the city. Head – clusters were specified randomly using the list of the residing families and finally the map of each cluster along with its address were marked. The exact locations of the head – clusters in association with their addresses were specified on a map. Overall one thousand and four hundred persons were invited to participate in the study. Only participants who were in an apparently healthy state were enrolled in the study and any participant who had any systemic disease was excluded from the study population. For data gathering, 10 experienced nutritionists were trained for the interview process. Data gathering was done using face to face interviews and study related measurements were done at the participants' doors. All Participants were informed about the purpose and protocol of this study and each participant gave their consent to enter the study. 2.2. Measurements and calculations All anthropometric measurements were done in the morning, after an overnight fasting condition, at a similar time (9 a.m.), and according to the recommendations of the International Standards for Anthropometric Assessment (ISAK) [10]. Furthermore, all
measurements were performed by well – trained researchers to minimize coefficients of variation. Each measurement was made three times and the average value was calculated. Body weight was measured to the nearest 0.1 kg using an electronic scale (Seca 769 scale, Seca GMBH, Hamburg). Height was measured to the nearest 0.5 cm using a stadiometer (Seca 769 scale, Seca GMBH, Hamburg). BMI (Kg/m2) was calculated as weight (Kg) divided by squared height (m2). Abdominal waist and hip circumferences were measured using a flexible plastic tape. Waist circumference (WC) was measured at midpoint at the inferior border of the lowest ribs to the superior iliac crest. The measurement was done at the end of a normal expiration while the subject stood upright, with feet together and arms hanging freely at the sides. Hip circumference was measured over no restrictive underwear at the level of the maximum extension of the buttocks posteriorly in a horizontal plane, without compressing the skin. All anthropometric calculations were done according to the following equations: A body shape index ðABSIÞ ¼
WC BMI
2=3
height1=2
Abdominal volume index (AVI) = [2(WC)2 + 0.7(waist/hip)2]/1000
Body adiposity index ðBAIÞ ¼ Conicity ¼
Hip 18 Height1:5
WCðmÞ
qffiffiffiffiffiffiffiffiffiffiffiffiffi
0:109
WeightðkgÞ HeightðmÞ
Percentage of body fat was obtained by the Tetrapolar Bioelectrical Impedance Analysis (BIA) system (BF-350, Tanita Corp, Tokyo, Japan). Subjects stood on the metal contacts with bare feet and their body fat mass were determined. This measurement was repeated twice, and the average value was calculated and set. 2.3. Statistical analyses All statistical analysis including the anthropometric calculations were carried out using the Statistical Package for Social Sciences software (SPSS/IBM, Chicago, IL, USA), for windows version 22. All the data were tested for their normal distribution using the Kolmogorov – Smirnov test. Results are displayed as means and standard deviations (SD) and in percentages where appropriate. Student t-test for unpaired data was used to evaluate differences in anthropometric characteristics between two sexes. We categorized the participants according to sex and separately evaluated the existence of significant bivariate correlations among different anthropometric indices using the Pearson correlation coefficient. The diagnostic accuracy of each of the measurements for estimating abdominal fat was assessed using a receiver operating characteristic (ROC) curve analysis, reporting its sensitivity and specificity, positive predictive value (PPV), negative predictive values (NPV), positive likelihood ratio (PLR), negative likelihood ratio (NLR) and accuracy (Acc), for each of the two sexes separately. In this analysis body fat percent was considered as the continuous variable and other anthropometric variables were compared to body fat percent. A two-tailed p-value of less than 0.05 was considered statistically significant.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
3
Table 1 Anthropometric and baseline characteristics of participants. Variables
Men (n = 580)
Women (n = 780)
Overall (n = 1360)
P-value
Age – years Weight – Kg Height – m BMI – Kg/m2 BMI categories – no. (%) Underweight – BMI < 18.5 kg/m2 Normal weight – BMI 18.5- < 25 kg/m2 Overweight – BMI 25- < 30 kg/m2 Obese – BMI 30 kg/m2 BAI – Kg/m2 ABSI – m11/6Kg2/3 AVI Conicity Hip circumference – cm Waist circumference – cma Waist to Hip ratio Waist to height ratio Percent Fat
34.98 12.5 80.85 14.57 173.37 7.35 25.71 5.34
31.58 11.34 67.04 13.71 159.69 6.17 26.30 5.83
33 12 72.93 15.65 165.52 9.52 26.05 5.63
<0.001 <0.001 0.058
57 (9.8) 204 (35.2) 191 (32.9) 128 (22.1) 27.435 6.487 0.0797 0.0056 16.52 5.18 1.232 0.099 103.54 13.43 92.34 11.80 0.9063 0.0987 0.5483 0.0882) 24.53 6.64
63 (8.1) 249 (31.9) 262 (33.6) 206 (26.4) 33.096 5.443 0.0785 0.0065 15.30 4.48 1.233 0.106 104.02 11.09 85.82 13.19 0.8366 0.0825 0.5196 0.1013 35.81 5.83
120 (8.8) 453 (33.3) 453 (33.3) 334 (26.4) 30.682 6.539 0.0790 0.0062 15.82 4.83 1.233 0.103 103.81 12.14 88.60 13.02 0.8664 0.0961 0.5318 0.0969 31.00 8.33
<0.001 <0.001 <0.001 0.896 0.468 <0.001 <0.001 <0.001 <0.001
*All plus – minus values are means standard deviations; ABSI: A body shape index; AVI: Abdominal volume index; BAI: Body adiposity index; BMI: Body mass index.
3. Results Forty people either refused to participate in the study or had a condition which would preclude them from entering the study, so the final number of participants was 1360 (97.1%) with 580 men and 780 women. Baseline and anthropometric characteristics of participants are shown in Table 1. Significant difference was found between men and women in all anthropometric parameters (p < 0.001), except for BMI, conicity and hip circumference. As expected, men were taller and heavier. When taking BMI categories into account, women displayed higher percentages of individuals in the overweight and obese categories (33.6% vs. 32.9% and 26.4% vs. 22.1%, respectively). BAI values were significantly higher in female participants (p < 0.001), however WC, ABSI, AVI, WHR and WHtR were lower in women compared to men (p < 0.001) (Table 1). In both men and women the strongest correlations were seen between body fat percent and BMI, AVI and WHtR (r > 7.9 and p < 0.001). WC showed a strong correlation with AVI, WHtR, weight and BMI in both sexes (r > 0.8 and p < 0.001). Interestingly WHtR showed a strong correlation with most of the anthropometric indices including WC, AVI, BMI and conicity (r > 0.8 and p < 0.001) (Table 2). Table 3 displays the ROC curve analysis for anthropometric indices in their ability to estimate body fat (BIA) in men and
women. BMI, WHtR and AVI in men and BAI, BMI and WHtR in women showed the most accuracy for estimating body fat percent, respectively. Accuracy was highest for BMI (Acc: 0.86; sensitivity: 0.88; specificity: 0.81; PPV: 0.92; NPV: 0.72; PLR: 4.6 and NLR: 0.14) in men and WHtR (Acc: 0.72; sensitivity: 0.71; specificity: 0.78; PPV: 0.93; NPV: 039; PLR: 3.27 and NLR: 0367) for women (Fig. 1). 4. Discussion In here we evaluated and compared existing anthropometric indices for detecting body adiposity and based on our results we found that BMI and WHtR were the best choices for body fat estimation in males and females, respectively. To the best of the authors' knowledge this is the only population based study that compared available anthropometric indices to find valid and reliable measures for adiposity and especially central obesity in the Iranian population. BMI is the most well-known and widely used index to estimate body fat percentage. Many studies have examined the association between BMI and body fat percent among different groups of people, such as athletes [11], soldiers [12], young adults [13] and very obese subjects [14]. However, it is unknown whether the relationship between BMI and body fat percent is linear or curvilinear. There is uncertainty in the rate of increase in body fat percent from aging after adjusting for BMI [15], furthermore other
Table 2 Correlation matrix among anthropometric indices.
man
% Fat ABSI AVI BAI BMI Conicity Height HC WC Weight WHR WHtR
% Fat
ABSI
AVI
BAI
BMI
Conicity
Height
HC
WC
Weight
WHR
WHtR
1 0.697** 0.832** 0.640** 0.868** 0.631** 0.684** 0.405** 0.727** 0.450** 0.476** 0.802**
0.630** 1 0.575** 0.270** 0.162** 0.928** 0.045 0.338** 0.600** 0.174** 0.429** 0.562**
0.800** 0.421** 1 0.584** 0.877** 0.829** 0.014 0.711** 0.996** 0.862** 0.568** 0.946**
0.594** 0.015ns 0.717** 1 0.666** 0.485** 0.469** 0.860** 0.600** 0.412** 0.096* 0.723**
0.813** 0.080* 0.857** 0.835** 1 0.516** 0.205** 0.683** 0.871** 0.878** 0.456** 0.892**
0.434** 0.916** 0.744** 0.343** 0.323** 1 0.042 0.549** 0.849** 0.479** 0.547** 0.825**
0.615** 0.051ns 0.058ns 0.448** 0.158** 0.108** 1 0.035 0.003 0.284** 0.114** 0.297**
0.507** 0.007ns 0.778** 0.899** 0.860** 0.334** 0.013ns 1 0.723** 0.676** 0.133** 0.688**
0.736** 0.443** 0.996** 0.707** 0.847** 0.763** 0.053 0.770** 1 0.851** 0.563** 0.953**
0.535** 0.093** 0.837** 0.680** 0.946** 0.292** 0.164** 0.846** 0.830** 1 0.398** 0.724**
0.439** 0.698** 0.740** 0.180** 0.429** 0.841** 0.074* 0.172** 0.757** 0.417** 1 0.564**
0.79** 0.438** 0.974** 0.783** 0.852** 0.758** 0.265** 0.750** 0.976** 0.766** 0.742** 1
% Fat ABSI AVI BAI BMI Conicity Height HC WC Weight WHR WHtR
women
ABSI: A body shape index; AVI: Abdominal volume index; BAI: Body adiposity index; BMI: Body mass index; HC: hip circumference; WC: wasit circumference. p < 0.05. ** p < 0.001. *
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
4
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
Table 3 Diagnostic value of anthropometric indices in their ability to assess body fat percentage. Indices BMI BAI ABSI AVI Conicity WC WHR WHtR Weight HC
Men Women Men Women Men Women Men Women Men Women Men Women Men Women Men Women Men Women Men Women
Cut off points
Acc.
PPV
NPV
Sen.
Spe.
PLR
NLR
24.3795 24.3795 25.3455 31.4872 0.0768 0.07993 14.708 14.4505 1.1803 1.2195 85.75 82.25 0.8153 0.8152 0.4943 0.5016 68.5 64.25 96.5 101.25
0.862 0.689 0.819 0.722 0.750 0.517 0.836 0.659 0.836 0.624 0.836 0.692 0.845 0.672 0.836 0.726 0.853 0.643 0.845 0.612
0.925 0.963 0.931 0.976 0.831 0.902 0.901 0.936 0.865 0.908 0.901 0.930 0.851 0.896 0.901 0.932 0.894 0.948 0.883 0.957
0.722 0.374 0.628 0.405 0.545 0.258 0.686 0.342 0.741 0.306 0.686 0.364 0.818 0.33 0.686 0.395 0.742 0.336 0.733 0.320
0.881 0.639 0.809 0.672 0.821 0.450 0.869 0.620 0.916 0.594 0.869 0.669 0.952 0.672 0.869 0.713 0.904 0.590 0.904 0.543
0.813 0.898 0.844 0.932 0.563 0.796 0.750 0.823 0.625 0.748 0.75 0.789 0.562 0.673 0.75 0.782 0.719 0.864 0.687 0.898
4.697 6.267 5.178 9.880 1.877 2.206 3.475 3.505 2.444 2.359 3.475 3.172 2.177 2.058 3.475 3.275 3.216 4.340 2.894 5.324
0.147 0.401 0.226 0.352 0.318 0.691 0.175 0.462 0.134 0.543 0.175 0.420 0.085 0.487 0.175 0.367 0.133 0.474 0.139 0.509
Acu: Accuracy, PPV: positive predictive value, NPV: negative predictive value, Sen: sensitivity or true positive rate, Spe: specificity or true negative rate, PLR: Positive likelihood ratio, NLR: Negative likelihood ratio; BMI: body mass index; BAI: body adiposity index; ABSI: a body shape index; AVI: abdominal volume index; WC: waist circumference; WHR: waist to hip ratio; WHtR: waist to height ratio; HC: hip circumference. *All the indexes were compared with the amount of fat measured by the Tetrapolar Bioelectrical Impedance Analysis system as the gold standard.
inconsistencies like whether or not sex influences it are some issues that have limited the use of BMI in epidemiological studies. Even subjects with appropriate BMI may have high fat content, in addition there is some evidence that suggests a relationship between BMI greater than 25 with a lower risk of certain diseases such as osteoporosis and even associated with longevity. BMI cannot differentiate between muscle and fat tissue and cannot distinguish various components of fat which limits its use in estimating cardiovascular disease risk, since cardiovascular disease is more commonly associated with abdominal fat accumulation rather than subcutaneous fat [16,17]. In developing countries, WHO recommends that BMI should be used with caution because low BMI values also have a high risk for disease [18,19]. Hence it is difficult to determine the optimal range of BMI for health [20]. In our study BMI showed the strongest correlation with body fat also it had a strong relationship with weight, WC and WHtR, implying that this old index is still reliable in population studies. However, due to its reliance on weight, it can be misleading and inaccurate. So perhaps replacing BMI with new indexes that have the ability to distinguish the location of body fat, gives researchers the ability to better predict disease risk. WC in comparison to BMI is a better predictor of cardiovascular disease [21,22], type II diabetes (in women) [23,2] and metabolic syndrome [24]. WHO recommends that the relative risk of type II diabetes and cardiovascular disease is better diagnosed when using both BMI and WC [25]. This index has been used in the definition of metabolic syndrome, although use of it as a sole indicator of disease and body fat is problematic because it does not have a same manner for all body frames. A tall and athletic person is likely to have a large WC while this individual should be considered normal. The same WC would be considered as high for a short person, furthermore it also has a strong association with some genetic characteristics such as gender and race [26]. In our study WC showed a strong correlation with percentage of body fat (0.727 and 0.736 in men and women, respectively). WC was more associated with percent of body fat in women indicating a greater role of abdominal fat in total adiposity in women. WHR can underestimate abdominal obesity in overweight individuals who have a large hip circumference and overestimate it
in thin people with high WC. The main problem with this indicator is that waist and hip behave in the same way when weight is lost or gained. In weight gains, both measures increase together and so the impact of weight gain is underestimated by this index. Conceptually this index is not suitable to assess obesity, particularly the changes in weight [27–30]. The INTERHEART study [31] concluded that WHR was most associated with the risk of myocardial infarction among other indices. This relationship was seen in both sexes, all age groups, ethnic groups, smokers and nonsmokers in subjects with and without hyperlipidemia, diabetes and high blood pressure. On the other hand, they found that the association between BMI and myocardial infarction was weaker, especially among people who have hypertension or high ratio of ApoB/ApoA. In addition an increase in WHR compared to an increase in BMI, increased disease risk up to three folds. So when using BMI, we actually underestimate the impact of obesity on disease risk. Waist and hip circumference are independently associated with risk of myocardial infarction that makes both of these metrics suitable tools for epidemiological studies [20]. In our study, we did not document a strong correlation between body fat percentage and WHR in males and females, furthermore the correlation between WHR and other anthropometric indices was significant but weak. WHtR can be a perfect indicator of body fat, especially for adults in whom height does not undergo much change. So the fraction with a constant denominator could be a convenient and reliable index to assess and track changes. Taking height into consideration can partially reduce defects of waist measurement because it could eliminate the genetic variations in WC. Schneider et al. reported this index as a better predictor of cardiovascular risk than other anthropometric indices [29]. One study showed that WHtR had the highest correlation with risk of heart disease [16]. In our study, a strong correlation was seen between WHtR and body fat percent. The interesting thing about this indicator was its stronger correlation with body fat percent rather than weight, which makes it a better choice than BMI for population studies. Bergman et al. introduced BAI to replace BMI [32]. However, recently it has been proposed that BAI cannot overcome the limitations of BMI and its only advantage is that it does not require body weight in its calculations [24]. One study reported that BAI
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
1.0 0.9 0.8
Sensivity
0.7 0.6
1.0
0.8
0.8
0.6
0.6
0.4
BAI
0.4
0.2
Male Female
0.2
0.5
0.0
0.4
1.0
1.0
0.3
0.8
0.8 0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.0
ABSI Male Female 0.0
0.2
0.4
0.6
0.8
1.0
0.2
BMI
0.6
0.1
Male Female
0.4
Conicity
0.4
WC
0.2
Male Female
0.2
Male Female
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Specificity 1.0 0.9 0.8 0.7
Sensivity
1.0
5
0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.0
1.0
1.0
0.8
0.8
0.6
0.6
0.4
WHR
0.2
Male Female
0.0
0.5
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
0.4
WHtR
0.2
Male Female
0.0
0.4
1.0
1.0
0.3
0.8
0.8 0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.2
AVI
0.6
0.1
Male Female
0.4
Weight
0.4
Hip
0.2
Male Female
0.2
Male Female
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Specificity
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.0
0.2
0.4
0.6
0.8
1.0
Fig. 1. ROC curves of different anthropometric indices for the estimation of body fat percent.
may be less useful than BMI in assessing risk factors for metabolic syndrome. This study argued that WC and WHtR are better indicators of metabolic syndrome [33]. In addition, the BAI index is not able to distinguish the location of body fat. Since BAI, similar to BMI, does not take into consideration WC, we can conclude that BAI has the same limitations of BMI. It has been shown that BAI overestimates body fat in men and underestimates it in women [24]. In our study, BAI had a weak correlation with body fat percent compared to BMI, WC, AVI, ABSI and WHtR. However, its correlation with body fat percent was more than that of HC, weight, WHR and conicity. Considering the inability of BMI to estimate body shape and the strong correlation between WC and BMI, ABSI is able to adjust WC for height and weight. This index has little correlation with height, weight and BMI, while it shows a strong correlation with mortality rates [21]. So far, only two studies have examined this index. In one of them, Duncan and colleagues in Portugal reported this indicator to be better than BMI and WC and correlated it to systolic and diastolic blood pressure [34]. In another study, He and colleagues
compared this index with WC and BMI in its ability to predict diabetes in a Chinese population and found that WC and BMI were better indices for predicting the risk of diabetes [35]. In our study, this index had little ability to predict body fat percent (0.697 and 0.630 in men and women, respectively) and only had a strong correlation with conicity index. Conicity has been criticized since it was first introduced [36] as it does not take into consideration women’s obesity model. This largely relates to its mathematical model ranging from cylinder to double cone and has little sensitivity to women’s pear-shaped model of obesity and considers it a cylinder form. Our results suggest that this indicator is more associated with body fat in men compared to women (0.631 vs. 0.434), so using it to estimate body fat percentage for women is not very precise. On the other hand, the overall correlation of this indicator with body fat percent was low and we do not recommend using it for body adiposity assessments. AVI is one of the best indicators to assess the accumulation of fat in the abdominal region [37]. In our study after BMI, AVI showed the highest correlation with body fat percentage, furthermore it
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
6
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
showed the strongest correlation with WC (0.996 in both men and women) and WHtR (0.946 and 0.974 in men and women, respectively), which indicates its correlation to belly fat and its use instead of both these indexes. This index does not depend on weight, making it a good candidate to be used instead of BMI as well, especially in populations such as athletes. For clinical practice, it is necessary to have an index that can be representative of other induces to estimate disease risk. The results of our study indicate that all anthropometric parameters could predict body fat percent with relatively good power, however BMI, WHtR and AVI are more powerful predictors (highest accuracy in predicting body fat). Based on our findings, we suggest using AVI and WHtR instead of other indexes, as they are better able to assess the accumulation of fat in the abdominal area and are able to more accurately assess body fat percent, which are indicators of chronic disease [38]. Our study did have some limitations which should be considered. First is that the cross-sectional nature of our study did not allow us to completely evaluate the health conditions that may have been associated with increase in each of the anthropometric indexes on a long term basis. Secondly, disease development is a complex process and is not only associated with body fat percent and multiple factors affect the progression and initiation of disease. These two issues limited our ability to predict disease association, although the main goal of our study was to evaluate the association between anthropometric indexes and their association with body fat percent. We had a relatively large sample size which did not allow us to have a single examiner for all the measurements and inter-examiner differences is an existing dilemma. More comprehensive studies with long follow-ups considering disease development aside to body fat percent would give a more clear evidence to the superiority of each anthropometric index. Conflict of interest There is no conflict of interest to be declared regarding the manuscript. Disclosure No grant has supported this present study and the authors had no conflict of interest to be declared. Project funding was provided by the authors. Author contributions Conceptualization: Methodology: RH Software: MMN Validation: RH MMN Formal analysis: MMN Investigation: EE MC MH SHD JMR AA Resources: MA RH Data curation: MMN RH Writing (original draft preparation): RH PA Writing (review and editing): PA RH EE Visualization: MC Supervision: RH Project administration: RH Funding acquisition: Acknowledgement The authors would like to thank all the participants who patiently took part in the present study.
References [1] Zalesin KC, Franklin BA, Miller WM, Peterson ED, McCullough PA. Impact of obesity on cardiovascular disease. Med Clin North Am 2011;95(5):919–37. [2] Wang Y, Rimm EB, Stampfer MJ, Willett WC, Hu FB. Comparison of abdominal adiposity and overall obesity in predicting risk of type 2 diabetes among men. Am J Clin Nutr 2005;81(3):555–63. [3] Zhang C, Rexrode KM, van Dam RM, Li TY, Hu FB. Abdominal obesity and the risk of all-cause, cardiovascular, and cancer mortality sixteen years of followup in US women. Circulation 2008;117(13):1658–67. [4] Love TJ, Qureshi AA, Karlson EW, Gelfand JM, Choi HK. Prevalence of the metabolic syndrome in psoriasis: results from the National Health and Nutrition Examination Survey, 2003–2006. Arch Dermatol 2011;147(4):419. [5] Wendland EM, Duncan BB, Mengue SS, Nucci LB, Schmidt MI. Waist circumference in the prediction of obesity-related adverse pregnancy outcomes. Cad Saúde Pública 2007;23(2):391–8. [6] Seidell J. Waist circumference and waist/hip ratio in relation to all-cause mortality, cancer and sleep apnea. Eur J Clin Nutr 2009;64(1):35–41. [7] Cameron AJ, Dunstan DW, Owen N, Zimmet PZ, Barr E, Tonkin AM, et al. Health and mortality consequences of abdominal obesity: evidence from the AusDiab study. Med J Aust 2009;191(4):202–8. [8] Chen S, Guo X, Yu S, Zhou Y, Li Z, Sun Y. Anthropometric indices in adults: which is the best indicator to identify alanine aminotransferase levels? Int J Environ Res Public Health 2016;13(2):226. [9] Haghravan S, Keshavarz SA, Mazaheri R, Alizadeh Z, Mansournia MA. Effect of omega-3 PUFAs supplementation with lifestyle modification on anthropometric indices and vo2 max in overweight women. Arch Iran Med 2016;19(5)342–7 0161905/aim.008. [10] Ellis KJ, Bell SJ, Chertow GM, Chumlea WC, Knox TA, Kotler DP, et al. Bioelectrical impedance methods in clinical research: a follow-up to the NIH Technology Assessment Conference. Nutrition 1999;15(11):874–80. [11] Riewald S. Does the body mass index accurately reflect percent body fat in athletes? Strength Cond J 2008;30(1):80–1. [12] Mullie P, Vansant G, Hulens M, Clarys P, Degrave E. Evaluation of body fat estimated from body mass index and impedance in belgian male military candidates-Comparing two methods for estimating body composition. Mil Med 2008;173(3):266–70. [13] Brooks Y, Black DR, Coster DC, Blue CL, Abood DA, Gretebeck RJ. Body mass index and percentage body fat as health indicators for young adults. Am J Health Behav 2007;31(6):687–700. [14] Adams T, Heath E, LaMonte M, Gress R, Pendleton R, Strong M, et al. The relationship between body mass index and per cent body fat in the severely obese Diabetes. Obes Metab 2007;9(4):498–505. [15] Jackson AS, Stanforth PR, Gagnon J, Rankinen T, Leon AS, Rao DC, et al. The effect of sex, age and race on estimating percentage body fat from body mass index: the Heritage Family Study. Int J Obes Relat Metab Disord 2002;26:789– 96. [16] Melmer A, Lamina C, Tschoner A, Ress C, Kaser S, Laimer M, et al. Body Adiposity Index and other indexes of body composition in the SAPHIR study: association with cardiovascular risk factors. Obesity 2013;21(4):775–81. [17] Pischon T, Boeing H, Hoffmann K, Bergmann M, Schulze M, Overvad K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med 2008;359(20):2105–20. [18] Babai MA, Arasteh P, Hadibarhaghtalab M, Naghizadeh MM, Salehi A, Askari A, et al. Defining a BMI cut-off point for the Iranian population: the Shiraz Heart Study. PLoS One 2016;11(8):e0160639. [19] Zeng Q, He Y, Dong S, Zhao X, Chen Z, Song Z, et al. Optimal cut-off values of BMI, waist circumference and waist:height ratio for defining obesity in Chinese adults. Br J Nutr 2014;112(10):1735–44, doi:http://dx.doi.org/10.1017/ s0007114514002657. [20] Dimitriadis K, Tsioufis C, Mazaraki A, Liatakis I, Koutra E, Kordalis A, et al. Waist circumference compared with other obesity parameters as determinants of coronary artery disease in essential hypertension: a 6-year follow-up study. Hypertens Res: Off J Jpn Soc Hypertens 2016, doi:http://dx.doi.org/10.1038/ hr.2016.8. [21] Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS One 2012;7(7):e39504. [22] Freedman DS, Thornton JC, Pi-Sunyer FX, Heymsfield SB, Wang J, Pierson RN, et al. The body adiposity index (hip circumference height 1.5) is not a more accurate measure of adiposity than is BMI, waist circumference, or hip circumference. Obesity 2012. [23] Schulze MB, Heidemann C, Schienkiewitz A, Bergmann MM, Hoffmann K, Boeing H. Comparison of anthropometric characteristics in predicting the incidence of type 2 diabetes in the EPIC-potsdam study. Diabetes Care 2006;29 (8):1921–3. [24] López AA, Cespedes ML, Vicente T, Tomas M, Bennasar-Veny M, Tauler P, et al. Body adiposity index utilization in a Spanish Mediterranean population: comparison with the body mass index. PLoS One 2012;7(4):e35281. [25] Consultation W. Obesity: preventing and managing the global epidemic. World Health Organization Technical Report Series 2000894:. [26] Grundy SM, Brewer HB, Cleeman JI, Smith SC, Lenfant C. Definition of metabolic syndrome report of the National Heart, Lung, and Blood Institute/ American Heart Association Conference on scientific issues related to definition. Circulation 2004;109(3):433–8.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx [27] Dalton M, Cameron A, Zimmet P, Shaw J, Jolley D, Dunstan D, et al. Waist circumference, waist-hip ratio and body mass index and their correlation with cardiovascular disease risk factors in Australian adults. J Intern Med 2003;254 (6):555–63. [28] Esmaillzadeh A, Mirmiran P, Azizi F. Waist-to-hip ratio is a better screening measure for cardiovascular risk factors than other anthropometric indicators in Tehranian adult men. Int J Obes 2004;28(10):1325–32. [29] Schneider HJ, Glaesmer H, Klotsche J, Böhler S, Lehnert H, Zeiher AM, et al. Accuracy of anthropometric indicators of obesity to predict cardiovascular risk. J Clin Endocrinol Metab 2007;92(2):589–94. [30] Welborn T, Dhaliwal S. Preferred clinical measures of central obesity for predicting mortality. Eur J Clin Nutr 2007;61(12):1373–9. [31] Ehling A, Schäffler A, Herfarth H, Tarner IH, Anders S, Distler O, et al. The potential of adiponectin in driving arthritis. J Immunol 2006;176(7):4468–78. [32] Bergman RN, Stefanovski D, Buchanan TA, Sumner AE, Reynolds JC, Sebring NG, et al. A better index of body adiposity. Obesity 2011;19(5):1083–9.
7
[33] Snijder MB, Nicolaou M, Valkengoed IG, Brewster LM, Stronks K. Newly proposed body adiposity index (bai) by Bergman et al. is not strongly related to cardiovascular health risk. Obesity 2012;20(6):1138–9. [34] Duncan MJ, Mota J, Vale S, Santos MP, Ribeiro JC. Associations between body mass index, waist circumference and body shape index with resting blood pressure in Portuguese adolescents. Ann Hum Biol 2013;40(2):163–7. [35] He S, Chen X. Could the new body shape index predict the new onset of diabetes mellitus in the chinese population? PLoS One 2013;8(1):e50573. [36] Valdez R. A simple model-based index of abdominal adiposity. J Clin Epidemiol 1991;44(9):955–6. [37] Vuga M. Conceptual review of issues with practical abdominal obesity measures. . [38] Choo V. WHO reassesses appropriate body-mass index for Asian populations. Lancet (Lond, Engl) 2002;360(9328):235, doi:http://dx.doi.org/10.1016/s01406736(02)09512-0.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Diabetes & Metabolic Syndrome: Clinical Research & Reviews journal homepage: www.elsevier.com/locate/dsx
Original article
New anthropometric indices or old ones: Which is the better predictor of body fat?$ Elham Ehrampousha,b , Peyman Arastehc,** , Reza Homayounfara,b,* , Makan Cheraghpourd , Meysam Alipourd, Mohammad Mehdi Naghizadeha , Maryam hadibarhaghtalaba , Sayed Hosein Davoodie , Alireza Askarif , Jalaledin Mirzay Razaze a
Noncommunicable Diseases Research Center, Fasa University of Medical Sciences, Fasa, Iran Health Policy Research Center, Shiraz University of Medical Sciences, Shiraz, Iran c Department of MPH, Shiraz University of Medical Sciences, Shiraz, Iran d Nutrition and Metabolic Diseases Research Center, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran e National Nutrition and Food Technology Research Institute, Faculty of Nutrition Sciences and Food Technology, Shahid Beheshti University of Medical Sciences, Tehran, Iran f Shiraz University of Medical Sciences, Shiraz, Iran b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 14 August 2016 Accepted 22 August 2016 Available online xxx
Background: The percent and distribution of body fat are important factors in the pathogenesis of numerous diseases. Our aim was to investigate common anthropometric indices in their relationship with body fat content. Methods: In a cross-sectional study 1360 healthy individuals (580 men and 780 women) in a cluster sampling, from Ahvaz, Iran, body fat content (using bioelectrical impedance) and anthropometric measurements [weight, waist circumference, a body shape index, abdominal volume index, body adiposity index, conicity, body mass index, hip circumference, waist to hip ratio and waist to height ratio] was obtained. The ROC curve analysis was used to compare each index with body fat percent. Results: Significant difference was found between men and women in all anthropometric parameters (p < 0.001). Women displayed higher percentages in the overweight and obese categories (33.6% vs. 32.9% and 26.4% vs. 22.1%, respectively). In both men and women, the strongest correlations were seen between body fat percent and BMI, AVI and WHtR (r > 7.9 and p < 0.001). BMI, WHtR and AVI in men and BAI, BMI and WHtR in women showed the most accuracy for estimating body fat percent, respectively. Conclusion: All anthropometric parameters could predict body fat percent with relatively good power, however BMI, WHtR and AVI are more powerful predictors. Based on our findings, we suggest using the AVI and WHtR instead of other indexes, as they are better able to assess the accumulation of fat in the abdominal area and are able to more accurately assess body fat percent, which are indicators of chronic disease. ã 2016 Diabetes India. Published by Elsevier Ltd. All rights reserved.
Keywords: Anthropometry Obesity Body mass index Body fat Weight-height ratio
$ Implication for health policy makers/practice/research/medical education: Older anthropometric indexes such as BMI have remained the most commonly used indexes for estimating disease risk. Recently it has become evident that old anthropometric indices are limited in many aspects for body fat and disease risk estimation. Replacing those indices with better and newer anthropometric indices that do not have the limitations of previous indices will aid in the correct estimation and prediction of disease development and overall body fat percent which itself is an indicator of chronic disease. * Corresponding author at: Noncommunicable Diseases Research Center, Fasa University of Medical Sciences, Fasa, Iran ** Corresponding author. E-mail addresses: [email protected], [email protected] (P. Arasteh), [email protected] (R. Homayounfar).
http://dx.doi.org/10.1016/j.dsx.2016.08.027 1871-4021/ã 2016 Diabetes India. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
2
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
1. Introduction Obesity defined as excess body fat, is a chronic and complex disease, which has long been recognized as one of the biggest problems in the field of public health. Increase in the prevalence of obesity leads to increased prevalence of obesity related complications such as cardiovascular disease [1], diabetes [2], some types of cancer [3], psoriasis [4], adverse pregnancy outcomes [5], earlier mortality [6] and many other health conditions. Abdominal obesity also known as central obesity is an excess of fat located in the abdominal area [7]. Anthropometry refers to the measurement of the human individual and has been used for identifying and understanding human physical variations. These measurements have simple, easy and effective characteristics that make them the first choice for nutritional evaluations. These indexes include: body mass index (BMI), waist circumference (WC), waist to hip ratio (WHR), waist to height ratio (WHtR), skinfold thickness, dual-energy X-ray absorption (DXA) and hydrostatic densitometry [8,9]. The point that should be considered regarding anthropometric indicators, in addition to their ability in estimating obesity or body fat, is their ability to detect the accumulation of fat in the abdominal area, which according to most studies is the most important factor affecting disease [8]. The negative impact of abdominal obesity on health is well recognized and although there have been tremendous advances in measuring body fat, there is still much debate regarding the most clinically efficient method of body fat assessment. In this study, we investigated common anthropometric indices regarding body fat content and further compared the available indices in order to better understand the more clinically valid and reliable measure for adiposity. 2. Materials and methods 2.1. Subjects and study protocol In a descriptive cross-sectional study from 2013 to 2014, the residents of Ahvaz, Iran were considered for inclusion in the study. For selecting the study sample, a cluster sampling technique was used. Twenty six districts in Ahvaz city were considered for the initial clusters. Then, the first digit of the postal codes of the citizens relating to the municipal classification was utilized to choose a random sample from the households in the city. Head – clusters were specified randomly using the list of the residing families and finally the map of each cluster along with its address were marked. The exact locations of the head – clusters in association with their addresses were specified on a map. Overall one thousand and four hundred persons were invited to participate in the study. Only participants who were in an apparently healthy state were enrolled in the study and any participant who had any systemic disease was excluded from the study population. For data gathering, 10 experienced nutritionists were trained for the interview process. Data gathering was done using face to face interviews and study related measurements were done at the participants' doors. All Participants were informed about the purpose and protocol of this study and each participant gave their consent to enter the study. 2.2. Measurements and calculations All anthropometric measurements were done in the morning, after an overnight fasting condition, at a similar time (9 a.m.), and according to the recommendations of the International Standards for Anthropometric Assessment (ISAK) [10]. Furthermore, all
measurements were performed by well – trained researchers to minimize coefficients of variation. Each measurement was made three times and the average value was calculated. Body weight was measured to the nearest 0.1 kg using an electronic scale (Seca 769 scale, Seca GMBH, Hamburg). Height was measured to the nearest 0.5 cm using a stadiometer (Seca 769 scale, Seca GMBH, Hamburg). BMI (Kg/m2) was calculated as weight (Kg) divided by squared height (m2). Abdominal waist and hip circumferences were measured using a flexible plastic tape. Waist circumference (WC) was measured at midpoint at the inferior border of the lowest ribs to the superior iliac crest. The measurement was done at the end of a normal expiration while the subject stood upright, with feet together and arms hanging freely at the sides. Hip circumference was measured over no restrictive underwear at the level of the maximum extension of the buttocks posteriorly in a horizontal plane, without compressing the skin. All anthropometric calculations were done according to the following equations: A body shape index ðABSIÞ ¼
WC BMI
2=3
height1=2
Abdominal volume index (AVI) = [2(WC)2 + 0.7(waist/hip)2]/1000
Body adiposity index ðBAIÞ ¼ Conicity ¼
Hip 18 Height1:5
WCðmÞ
qffiffiffiffiffiffiffiffiffiffiffiffiffi
0:109
WeightðkgÞ HeightðmÞ
Percentage of body fat was obtained by the Tetrapolar Bioelectrical Impedance Analysis (BIA) system (BF-350, Tanita Corp, Tokyo, Japan). Subjects stood on the metal contacts with bare feet and their body fat mass were determined. This measurement was repeated twice, and the average value was calculated and set. 2.3. Statistical analyses All statistical analysis including the anthropometric calculations were carried out using the Statistical Package for Social Sciences software (SPSS/IBM, Chicago, IL, USA), for windows version 22. All the data were tested for their normal distribution using the Kolmogorov – Smirnov test. Results are displayed as means and standard deviations (SD) and in percentages where appropriate. Student t-test for unpaired data was used to evaluate differences in anthropometric characteristics between two sexes. We categorized the participants according to sex and separately evaluated the existence of significant bivariate correlations among different anthropometric indices using the Pearson correlation coefficient. The diagnostic accuracy of each of the measurements for estimating abdominal fat was assessed using a receiver operating characteristic (ROC) curve analysis, reporting its sensitivity and specificity, positive predictive value (PPV), negative predictive values (NPV), positive likelihood ratio (PLR), negative likelihood ratio (NLR) and accuracy (Acc), for each of the two sexes separately. In this analysis body fat percent was considered as the continuous variable and other anthropometric variables were compared to body fat percent. A two-tailed p-value of less than 0.05 was considered statistically significant.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
3
Table 1 Anthropometric and baseline characteristics of participants. Variables
Men (n = 580)
Women (n = 780)
Overall (n = 1360)
P-value
Age – years Weight – Kg Height – m BMI – Kg/m2 BMI categories – no. (%) Underweight – BMI < 18.5 kg/m2 Normal weight – BMI 18.5- < 25 kg/m2 Overweight – BMI 25- < 30 kg/m2 Obese – BMI 30 kg/m2 BAI – Kg/m2 ABSI – m11/6Kg2/3 AVI Conicity Hip circumference – cm Waist circumference – cma Waist to Hip ratio Waist to height ratio Percent Fat
34.98 12.5 80.85 14.57 173.37 7.35 25.71 5.34
31.58 11.34 67.04 13.71 159.69 6.17 26.30 5.83
33 12 72.93 15.65 165.52 9.52 26.05 5.63
<0.001 <0.001 0.058
57 (9.8) 204 (35.2) 191 (32.9) 128 (22.1) 27.435 6.487 0.0797 0.0056 16.52 5.18 1.232 0.099 103.54 13.43 92.34 11.80 0.9063 0.0987 0.5483 0.0882) 24.53 6.64
63 (8.1) 249 (31.9) 262 (33.6) 206 (26.4) 33.096 5.443 0.0785 0.0065 15.30 4.48 1.233 0.106 104.02 11.09 85.82 13.19 0.8366 0.0825 0.5196 0.1013 35.81 5.83
120 (8.8) 453 (33.3) 453 (33.3) 334 (26.4) 30.682 6.539 0.0790 0.0062 15.82 4.83 1.233 0.103 103.81 12.14 88.60 13.02 0.8664 0.0961 0.5318 0.0969 31.00 8.33
<0.001 <0.001 <0.001 0.896 0.468 <0.001 <0.001 <0.001 <0.001
*All plus – minus values are means standard deviations; ABSI: A body shape index; AVI: Abdominal volume index; BAI: Body adiposity index; BMI: Body mass index.
3. Results Forty people either refused to participate in the study or had a condition which would preclude them from entering the study, so the final number of participants was 1360 (97.1%) with 580 men and 780 women. Baseline and anthropometric characteristics of participants are shown in Table 1. Significant difference was found between men and women in all anthropometric parameters (p < 0.001), except for BMI, conicity and hip circumference. As expected, men were taller and heavier. When taking BMI categories into account, women displayed higher percentages of individuals in the overweight and obese categories (33.6% vs. 32.9% and 26.4% vs. 22.1%, respectively). BAI values were significantly higher in female participants (p < 0.001), however WC, ABSI, AVI, WHR and WHtR were lower in women compared to men (p < 0.001) (Table 1). In both men and women the strongest correlations were seen between body fat percent and BMI, AVI and WHtR (r > 7.9 and p < 0.001). WC showed a strong correlation with AVI, WHtR, weight and BMI in both sexes (r > 0.8 and p < 0.001). Interestingly WHtR showed a strong correlation with most of the anthropometric indices including WC, AVI, BMI and conicity (r > 0.8 and p < 0.001) (Table 2). Table 3 displays the ROC curve analysis for anthropometric indices in their ability to estimate body fat (BIA) in men and
women. BMI, WHtR and AVI in men and BAI, BMI and WHtR in women showed the most accuracy for estimating body fat percent, respectively. Accuracy was highest for BMI (Acc: 0.86; sensitivity: 0.88; specificity: 0.81; PPV: 0.92; NPV: 0.72; PLR: 4.6 and NLR: 0.14) in men and WHtR (Acc: 0.72; sensitivity: 0.71; specificity: 0.78; PPV: 0.93; NPV: 039; PLR: 3.27 and NLR: 0367) for women (Fig. 1). 4. Discussion In here we evaluated and compared existing anthropometric indices for detecting body adiposity and based on our results we found that BMI and WHtR were the best choices for body fat estimation in males and females, respectively. To the best of the authors' knowledge this is the only population based study that compared available anthropometric indices to find valid and reliable measures for adiposity and especially central obesity in the Iranian population. BMI is the most well-known and widely used index to estimate body fat percentage. Many studies have examined the association between BMI and body fat percent among different groups of people, such as athletes [11], soldiers [12], young adults [13] and very obese subjects [14]. However, it is unknown whether the relationship between BMI and body fat percent is linear or curvilinear. There is uncertainty in the rate of increase in body fat percent from aging after adjusting for BMI [15], furthermore other
Table 2 Correlation matrix among anthropometric indices.
man
% Fat ABSI AVI BAI BMI Conicity Height HC WC Weight WHR WHtR
% Fat
ABSI
AVI
BAI
BMI
Conicity
Height
HC
WC
Weight
WHR
WHtR
1 0.697** 0.832** 0.640** 0.868** 0.631** 0.684** 0.405** 0.727** 0.450** 0.476** 0.802**
0.630** 1 0.575** 0.270** 0.162** 0.928** 0.045 0.338** 0.600** 0.174** 0.429** 0.562**
0.800** 0.421** 1 0.584** 0.877** 0.829** 0.014 0.711** 0.996** 0.862** 0.568** 0.946**
0.594** 0.015ns 0.717** 1 0.666** 0.485** 0.469** 0.860** 0.600** 0.412** 0.096* 0.723**
0.813** 0.080* 0.857** 0.835** 1 0.516** 0.205** 0.683** 0.871** 0.878** 0.456** 0.892**
0.434** 0.916** 0.744** 0.343** 0.323** 1 0.042 0.549** 0.849** 0.479** 0.547** 0.825**
0.615** 0.051ns 0.058ns 0.448** 0.158** 0.108** 1 0.035 0.003 0.284** 0.114** 0.297**
0.507** 0.007ns 0.778** 0.899** 0.860** 0.334** 0.013ns 1 0.723** 0.676** 0.133** 0.688**
0.736** 0.443** 0.996** 0.707** 0.847** 0.763** 0.053 0.770** 1 0.851** 0.563** 0.953**
0.535** 0.093** 0.837** 0.680** 0.946** 0.292** 0.164** 0.846** 0.830** 1 0.398** 0.724**
0.439** 0.698** 0.740** 0.180** 0.429** 0.841** 0.074* 0.172** 0.757** 0.417** 1 0.564**
0.79** 0.438** 0.974** 0.783** 0.852** 0.758** 0.265** 0.750** 0.976** 0.766** 0.742** 1
% Fat ABSI AVI BAI BMI Conicity Height HC WC Weight WHR WHtR
women
ABSI: A body shape index; AVI: Abdominal volume index; BAI: Body adiposity index; BMI: Body mass index; HC: hip circumference; WC: wasit circumference. p < 0.05. ** p < 0.001. *
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
4
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
Table 3 Diagnostic value of anthropometric indices in their ability to assess body fat percentage. Indices BMI BAI ABSI AVI Conicity WC WHR WHtR Weight HC
Men Women Men Women Men Women Men Women Men Women Men Women Men Women Men Women Men Women Men Women
Cut off points
Acc.
PPV
NPV
Sen.
Spe.
PLR
NLR
24.3795 24.3795 25.3455 31.4872 0.0768 0.07993 14.708 14.4505 1.1803 1.2195 85.75 82.25 0.8153 0.8152 0.4943 0.5016 68.5 64.25 96.5 101.25
0.862 0.689 0.819 0.722 0.750 0.517 0.836 0.659 0.836 0.624 0.836 0.692 0.845 0.672 0.836 0.726 0.853 0.643 0.845 0.612
0.925 0.963 0.931 0.976 0.831 0.902 0.901 0.936 0.865 0.908 0.901 0.930 0.851 0.896 0.901 0.932 0.894 0.948 0.883 0.957
0.722 0.374 0.628 0.405 0.545 0.258 0.686 0.342 0.741 0.306 0.686 0.364 0.818 0.33 0.686 0.395 0.742 0.336 0.733 0.320
0.881 0.639 0.809 0.672 0.821 0.450 0.869 0.620 0.916 0.594 0.869 0.669 0.952 0.672 0.869 0.713 0.904 0.590 0.904 0.543
0.813 0.898 0.844 0.932 0.563 0.796 0.750 0.823 0.625 0.748 0.75 0.789 0.562 0.673 0.75 0.782 0.719 0.864 0.687 0.898
4.697 6.267 5.178 9.880 1.877 2.206 3.475 3.505 2.444 2.359 3.475 3.172 2.177 2.058 3.475 3.275 3.216 4.340 2.894 5.324
0.147 0.401 0.226 0.352 0.318 0.691 0.175 0.462 0.134 0.543 0.175 0.420 0.085 0.487 0.175 0.367 0.133 0.474 0.139 0.509
Acu: Accuracy, PPV: positive predictive value, NPV: negative predictive value, Sen: sensitivity or true positive rate, Spe: specificity or true negative rate, PLR: Positive likelihood ratio, NLR: Negative likelihood ratio; BMI: body mass index; BAI: body adiposity index; ABSI: a body shape index; AVI: abdominal volume index; WC: waist circumference; WHR: waist to hip ratio; WHtR: waist to height ratio; HC: hip circumference. *All the indexes were compared with the amount of fat measured by the Tetrapolar Bioelectrical Impedance Analysis system as the gold standard.
inconsistencies like whether or not sex influences it are some issues that have limited the use of BMI in epidemiological studies. Even subjects with appropriate BMI may have high fat content, in addition there is some evidence that suggests a relationship between BMI greater than 25 with a lower risk of certain diseases such as osteoporosis and even associated with longevity. BMI cannot differentiate between muscle and fat tissue and cannot distinguish various components of fat which limits its use in estimating cardiovascular disease risk, since cardiovascular disease is more commonly associated with abdominal fat accumulation rather than subcutaneous fat [16,17]. In developing countries, WHO recommends that BMI should be used with caution because low BMI values also have a high risk for disease [18,19]. Hence it is difficult to determine the optimal range of BMI for health [20]. In our study BMI showed the strongest correlation with body fat also it had a strong relationship with weight, WC and WHtR, implying that this old index is still reliable in population studies. However, due to its reliance on weight, it can be misleading and inaccurate. So perhaps replacing BMI with new indexes that have the ability to distinguish the location of body fat, gives researchers the ability to better predict disease risk. WC in comparison to BMI is a better predictor of cardiovascular disease [21,22], type II diabetes (in women) [23,2] and metabolic syndrome [24]. WHO recommends that the relative risk of type II diabetes and cardiovascular disease is better diagnosed when using both BMI and WC [25]. This index has been used in the definition of metabolic syndrome, although use of it as a sole indicator of disease and body fat is problematic because it does not have a same manner for all body frames. A tall and athletic person is likely to have a large WC while this individual should be considered normal. The same WC would be considered as high for a short person, furthermore it also has a strong association with some genetic characteristics such as gender and race [26]. In our study WC showed a strong correlation with percentage of body fat (0.727 and 0.736 in men and women, respectively). WC was more associated with percent of body fat in women indicating a greater role of abdominal fat in total adiposity in women. WHR can underestimate abdominal obesity in overweight individuals who have a large hip circumference and overestimate it
in thin people with high WC. The main problem with this indicator is that waist and hip behave in the same way when weight is lost or gained. In weight gains, both measures increase together and so the impact of weight gain is underestimated by this index. Conceptually this index is not suitable to assess obesity, particularly the changes in weight [27–30]. The INTERHEART study [31] concluded that WHR was most associated with the risk of myocardial infarction among other indices. This relationship was seen in both sexes, all age groups, ethnic groups, smokers and nonsmokers in subjects with and without hyperlipidemia, diabetes and high blood pressure. On the other hand, they found that the association between BMI and myocardial infarction was weaker, especially among people who have hypertension or high ratio of ApoB/ApoA. In addition an increase in WHR compared to an increase in BMI, increased disease risk up to three folds. So when using BMI, we actually underestimate the impact of obesity on disease risk. Waist and hip circumference are independently associated with risk of myocardial infarction that makes both of these metrics suitable tools for epidemiological studies [20]. In our study, we did not document a strong correlation between body fat percentage and WHR in males and females, furthermore the correlation between WHR and other anthropometric indices was significant but weak. WHtR can be a perfect indicator of body fat, especially for adults in whom height does not undergo much change. So the fraction with a constant denominator could be a convenient and reliable index to assess and track changes. Taking height into consideration can partially reduce defects of waist measurement because it could eliminate the genetic variations in WC. Schneider et al. reported this index as a better predictor of cardiovascular risk than other anthropometric indices [29]. One study showed that WHtR had the highest correlation with risk of heart disease [16]. In our study, a strong correlation was seen between WHtR and body fat percent. The interesting thing about this indicator was its stronger correlation with body fat percent rather than weight, which makes it a better choice than BMI for population studies. Bergman et al. introduced BAI to replace BMI [32]. However, recently it has been proposed that BAI cannot overcome the limitations of BMI and its only advantage is that it does not require body weight in its calculations [24]. One study reported that BAI
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
1.0 0.9 0.8
Sensivity
0.7 0.6
1.0
0.8
0.8
0.6
0.6
0.4
BAI
0.4
0.2
Male Female
0.2
0.5
0.0
0.4
1.0
1.0
0.3
0.8
0.8 0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.0
ABSI Male Female 0.0
0.2
0.4
0.6
0.8
1.0
0.2
BMI
0.6
0.1
Male Female
0.4
Conicity
0.4
WC
0.2
Male Female
0.2
Male Female
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Specificity 1.0 0.9 0.8 0.7
Sensivity
1.0
5
0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.0
1.0
1.0
0.8
0.8
0.6
0.6
0.4
WHR
0.2
Male Female
0.0
0.5
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
0.4
WHtR
0.2
Male Female
0.0
0.4
1.0
1.0
0.3
0.8
0.8 0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.2
AVI
0.6
0.1
Male Female
0.4
Weight
0.4
Hip
0.2
Male Female
0.2
Male Female
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Specificity
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.0
0.2
0.4
0.6
0.8
1.0
Fig. 1. ROC curves of different anthropometric indices for the estimation of body fat percent.
may be less useful than BMI in assessing risk factors for metabolic syndrome. This study argued that WC and WHtR are better indicators of metabolic syndrome [33]. In addition, the BAI index is not able to distinguish the location of body fat. Since BAI, similar to BMI, does not take into consideration WC, we can conclude that BAI has the same limitations of BMI. It has been shown that BAI overestimates body fat in men and underestimates it in women [24]. In our study, BAI had a weak correlation with body fat percent compared to BMI, WC, AVI, ABSI and WHtR. However, its correlation with body fat percent was more than that of HC, weight, WHR and conicity. Considering the inability of BMI to estimate body shape and the strong correlation between WC and BMI, ABSI is able to adjust WC for height and weight. This index has little correlation with height, weight and BMI, while it shows a strong correlation with mortality rates [21]. So far, only two studies have examined this index. In one of them, Duncan and colleagues in Portugal reported this indicator to be better than BMI and WC and correlated it to systolic and diastolic blood pressure [34]. In another study, He and colleagues
compared this index with WC and BMI in its ability to predict diabetes in a Chinese population and found that WC and BMI were better indices for predicting the risk of diabetes [35]. In our study, this index had little ability to predict body fat percent (0.697 and 0.630 in men and women, respectively) and only had a strong correlation with conicity index. Conicity has been criticized since it was first introduced [36] as it does not take into consideration women’s obesity model. This largely relates to its mathematical model ranging from cylinder to double cone and has little sensitivity to women’s pear-shaped model of obesity and considers it a cylinder form. Our results suggest that this indicator is more associated with body fat in men compared to women (0.631 vs. 0.434), so using it to estimate body fat percentage for women is not very precise. On the other hand, the overall correlation of this indicator with body fat percent was low and we do not recommend using it for body adiposity assessments. AVI is one of the best indicators to assess the accumulation of fat in the abdominal region [37]. In our study after BMI, AVI showed the highest correlation with body fat percentage, furthermore it
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
6
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx
showed the strongest correlation with WC (0.996 in both men and women) and WHtR (0.946 and 0.974 in men and women, respectively), which indicates its correlation to belly fat and its use instead of both these indexes. This index does not depend on weight, making it a good candidate to be used instead of BMI as well, especially in populations such as athletes. For clinical practice, it is necessary to have an index that can be representative of other induces to estimate disease risk. The results of our study indicate that all anthropometric parameters could predict body fat percent with relatively good power, however BMI, WHtR and AVI are more powerful predictors (highest accuracy in predicting body fat). Based on our findings, we suggest using AVI and WHtR instead of other indexes, as they are better able to assess the accumulation of fat in the abdominal area and are able to more accurately assess body fat percent, which are indicators of chronic disease [38]. Our study did have some limitations which should be considered. First is that the cross-sectional nature of our study did not allow us to completely evaluate the health conditions that may have been associated with increase in each of the anthropometric indexes on a long term basis. Secondly, disease development is a complex process and is not only associated with body fat percent and multiple factors affect the progression and initiation of disease. These two issues limited our ability to predict disease association, although the main goal of our study was to evaluate the association between anthropometric indexes and their association with body fat percent. We had a relatively large sample size which did not allow us to have a single examiner for all the measurements and inter-examiner differences is an existing dilemma. More comprehensive studies with long follow-ups considering disease development aside to body fat percent would give a more clear evidence to the superiority of each anthropometric index. Conflict of interest There is no conflict of interest to be declared regarding the manuscript. Disclosure No grant has supported this present study and the authors had no conflict of interest to be declared. Project funding was provided by the authors. Author contributions Conceptualization: Methodology: RH Software: MMN Validation: RH MMN Formal analysis: MMN Investigation: EE MC MH SHD JMR AA Resources: MA RH Data curation: MMN RH Writing (original draft preparation): RH PA Writing (review and editing): PA RH EE Visualization: MC Supervision: RH Project administration: RH Funding acquisition: Acknowledgement The authors would like to thank all the participants who patiently took part in the present study.
References [1] Zalesin KC, Franklin BA, Miller WM, Peterson ED, McCullough PA. Impact of obesity on cardiovascular disease. Med Clin North Am 2011;95(5):919–37. [2] Wang Y, Rimm EB, Stampfer MJ, Willett WC, Hu FB. Comparison of abdominal adiposity and overall obesity in predicting risk of type 2 diabetes among men. Am J Clin Nutr 2005;81(3):555–63. [3] Zhang C, Rexrode KM, van Dam RM, Li TY, Hu FB. Abdominal obesity and the risk of all-cause, cardiovascular, and cancer mortality sixteen years of followup in US women. Circulation 2008;117(13):1658–67. [4] Love TJ, Qureshi AA, Karlson EW, Gelfand JM, Choi HK. Prevalence of the metabolic syndrome in psoriasis: results from the National Health and Nutrition Examination Survey, 2003–2006. Arch Dermatol 2011;147(4):419. [5] Wendland EM, Duncan BB, Mengue SS, Nucci LB, Schmidt MI. Waist circumference in the prediction of obesity-related adverse pregnancy outcomes. Cad Saúde Pública 2007;23(2):391–8. [6] Seidell J. Waist circumference and waist/hip ratio in relation to all-cause mortality, cancer and sleep apnea. Eur J Clin Nutr 2009;64(1):35–41. [7] Cameron AJ, Dunstan DW, Owen N, Zimmet PZ, Barr E, Tonkin AM, et al. Health and mortality consequences of abdominal obesity: evidence from the AusDiab study. Med J Aust 2009;191(4):202–8. [8] Chen S, Guo X, Yu S, Zhou Y, Li Z, Sun Y. Anthropometric indices in adults: which is the best indicator to identify alanine aminotransferase levels? Int J Environ Res Public Health 2016;13(2):226. [9] Haghravan S, Keshavarz SA, Mazaheri R, Alizadeh Z, Mansournia MA. Effect of omega-3 PUFAs supplementation with lifestyle modification on anthropometric indices and vo2 max in overweight women. Arch Iran Med 2016;19(5)342–7 0161905/aim.008. [10] Ellis KJ, Bell SJ, Chertow GM, Chumlea WC, Knox TA, Kotler DP, et al. Bioelectrical impedance methods in clinical research: a follow-up to the NIH Technology Assessment Conference. Nutrition 1999;15(11):874–80. [11] Riewald S. Does the body mass index accurately reflect percent body fat in athletes? Strength Cond J 2008;30(1):80–1. [12] Mullie P, Vansant G, Hulens M, Clarys P, Degrave E. Evaluation of body fat estimated from body mass index and impedance in belgian male military candidates-Comparing two methods for estimating body composition. Mil Med 2008;173(3):266–70. [13] Brooks Y, Black DR, Coster DC, Blue CL, Abood DA, Gretebeck RJ. Body mass index and percentage body fat as health indicators for young adults. Am J Health Behav 2007;31(6):687–700. [14] Adams T, Heath E, LaMonte M, Gress R, Pendleton R, Strong M, et al. The relationship between body mass index and per cent body fat in the severely obese Diabetes. Obes Metab 2007;9(4):498–505. [15] Jackson AS, Stanforth PR, Gagnon J, Rankinen T, Leon AS, Rao DC, et al. The effect of sex, age and race on estimating percentage body fat from body mass index: the Heritage Family Study. Int J Obes Relat Metab Disord 2002;26:789– 96. [16] Melmer A, Lamina C, Tschoner A, Ress C, Kaser S, Laimer M, et al. Body Adiposity Index and other indexes of body composition in the SAPHIR study: association with cardiovascular risk factors. Obesity 2013;21(4):775–81. [17] Pischon T, Boeing H, Hoffmann K, Bergmann M, Schulze M, Overvad K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med 2008;359(20):2105–20. [18] Babai MA, Arasteh P, Hadibarhaghtalab M, Naghizadeh MM, Salehi A, Askari A, et al. Defining a BMI cut-off point for the Iranian population: the Shiraz Heart Study. PLoS One 2016;11(8):e0160639. [19] Zeng Q, He Y, Dong S, Zhao X, Chen Z, Song Z, et al. Optimal cut-off values of BMI, waist circumference and waist:height ratio for defining obesity in Chinese adults. Br J Nutr 2014;112(10):1735–44, doi:http://dx.doi.org/10.1017/ s0007114514002657. [20] Dimitriadis K, Tsioufis C, Mazaraki A, Liatakis I, Koutra E, Kordalis A, et al. Waist circumference compared with other obesity parameters as determinants of coronary artery disease in essential hypertension: a 6-year follow-up study. Hypertens Res: Off J Jpn Soc Hypertens 2016, doi:http://dx.doi.org/10.1038/ hr.2016.8. [21] Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS One 2012;7(7):e39504. [22] Freedman DS, Thornton JC, Pi-Sunyer FX, Heymsfield SB, Wang J, Pierson RN, et al. The body adiposity index (hip circumference height 1.5) is not a more accurate measure of adiposity than is BMI, waist circumference, or hip circumference. Obesity 2012. [23] Schulze MB, Heidemann C, Schienkiewitz A, Bergmann MM, Hoffmann K, Boeing H. Comparison of anthropometric characteristics in predicting the incidence of type 2 diabetes in the EPIC-potsdam study. Diabetes Care 2006;29 (8):1921–3. [24] López AA, Cespedes ML, Vicente T, Tomas M, Bennasar-Veny M, Tauler P, et al. Body adiposity index utilization in a Spanish Mediterranean population: comparison with the body mass index. PLoS One 2012;7(4):e35281. [25] Consultation W. Obesity: preventing and managing the global epidemic. World Health Organization Technical Report Series 2000894:. [26] Grundy SM, Brewer HB, Cleeman JI, Smith SC, Lenfant C. Definition of metabolic syndrome report of the National Heart, Lung, and Blood Institute/ American Heart Association Conference on scientific issues related to definition. Circulation 2004;109(3):433–8.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027
G Model DSX 650 No. of Pages 7
E. Ehrampoush et al. / Diabetes & Metabolic Syndrome: Clinical Research & Reviews xxx (2016) xxx–xxx [27] Dalton M, Cameron A, Zimmet P, Shaw J, Jolley D, Dunstan D, et al. Waist circumference, waist-hip ratio and body mass index and their correlation with cardiovascular disease risk factors in Australian adults. J Intern Med 2003;254 (6):555–63. [28] Esmaillzadeh A, Mirmiran P, Azizi F. Waist-to-hip ratio is a better screening measure for cardiovascular risk factors than other anthropometric indicators in Tehranian adult men. Int J Obes 2004;28(10):1325–32. [29] Schneider HJ, Glaesmer H, Klotsche J, Böhler S, Lehnert H, Zeiher AM, et al. Accuracy of anthropometric indicators of obesity to predict cardiovascular risk. J Clin Endocrinol Metab 2007;92(2):589–94. [30] Welborn T, Dhaliwal S. Preferred clinical measures of central obesity for predicting mortality. Eur J Clin Nutr 2007;61(12):1373–9. [31] Ehling A, Schäffler A, Herfarth H, Tarner IH, Anders S, Distler O, et al. The potential of adiponectin in driving arthritis. J Immunol 2006;176(7):4468–78. [32] Bergman RN, Stefanovski D, Buchanan TA, Sumner AE, Reynolds JC, Sebring NG, et al. A better index of body adiposity. Obesity 2011;19(5):1083–9.
7
[33] Snijder MB, Nicolaou M, Valkengoed IG, Brewster LM, Stronks K. Newly proposed body adiposity index (bai) by Bergman et al. is not strongly related to cardiovascular health risk. Obesity 2012;20(6):1138–9. [34] Duncan MJ, Mota J, Vale S, Santos MP, Ribeiro JC. Associations between body mass index, waist circumference and body shape index with resting blood pressure in Portuguese adolescents. Ann Hum Biol 2013;40(2):163–7. [35] He S, Chen X. Could the new body shape index predict the new onset of diabetes mellitus in the chinese population? PLoS One 2013;8(1):e50573. [36] Valdez R. A simple model-based index of abdominal adiposity. J Clin Epidemiol 1991;44(9):955–6. [37] Vuga M. Conceptual review of issues with practical abdominal obesity measures. . [38] Choo V. WHO reassesses appropriate body-mass index for Asian populations. Lancet (Lond, Engl) 2002;360(9328):235, doi:http://dx.doi.org/10.1016/s01406736(02)09512-0.
Please cite this article in press as: E. Ehrampoush, et al., New anthropometric indices or old ones: Which is the better predictor of body fat?, Diab Met Syndr: Clin Res Rev (2016), http://dx.doi.org/10.1016/j.dsx.2016.08.027