Figures
Abstract
Background
The Body Mass Index (BMI) and Waist Circumference (WC) are well-used anthropometric predictors for cardiovascular diseases (CVD), but their validity is regularly questioned. Recently, A Body Shape Index (ABSI) and Body Roundness Index (BRI) were introduced as alternative anthropometric indices that may better reflect health status.
Objective
This study assessed the capacity of ABSI and BRI in identifying cardiovascular diseases and cardiovascular disease risk factors and determined whether they are superior to BMI and WC.
Design and Methods
4627 Participants (54±12 years) of the Nijmegen Exercise Study completed an online questionnaire concerning CVD health status (defined as history of CVD or CVD risk factors) and anthropometric characteristics. Quintiles of ABSI, BRI, BMI, and WC were used regarding CVD prevalence. Odds ratios (OR), adjusted for age, sex, and smoking, were calculated per anthropometric index.
Results
1332 participants (27.7%) reported presence of CVD or CVD risk factors. The prevalence of CVD increased across quintiles for BMI, ABSI, BRI, and WC. Comparing the lowest with the highest quintile, adjusted OR (95% CI) for CVD were significantly different for BRI 3.2 (1.4–7.2), BMI 2.4 (1.9–3.1), and WC 3.0 (1.6–5.6). The adjusted OR (95% CI) for CVD risk factors was for BRI 2.5 (2.0–3.3), BMI 3.3 (1.6–6.8), and WC 2.0 (1.6–2.5). No association was observed for ABSI in both groups.
Citation: Maessen MFH, Eijsvogels TMH, Verheggen RJHM, Hopman MTE, Verbeek ALM, Vegt Fd (2014) Entering a New Era of Body Indices: The Feasibility of a Body Shape Index and Body Roundness Index to Identify Cardiovascular Health Status. PLoS ONE 9(9): e107212. https://doi.org/10.1371/journal.pone.0107212
Editor: Yan Gong, University of Florida, United States of America
Received: March 10, 2014; Accepted: August 14, 2014; Published: September 17, 2014
Copyright: © 2014 Maessen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All data are included within the paper.
Funding: Dr. Thijs M. H. Eijsvogels is financially supported by the Netherlands Organization for Scientific Research (Rubicon Grant 825.12.016). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Overweight and obesity are the fifth leading cause of global death and are an increasing worldwide health problem. In 2008, more than 1.4 billion adults suffered from overweight and within this group approximately 36% adults is obese [1]. Overweight and obesity are associated with an increased risk for cardiovascular diseases (CVD), type 2 diabetes mellitus and premature death [2]–[9]. Therefore, an early detection of overweight or obesity is considered necessary to prevent CVD [10].
The Body Mass Index (BMI) and Waist Circumference (WC) are currently recommended by several guidelines to classify overweight and obesity [10], [11]. Indeed, an increase in BMI or WC has been shown to be a risk factor for CVD [10]. However, previous studies also demonstrated that the discriminative capacity of BMI is not optimal, as this calculation cannot distinguish between adipose tissue and lean body mass [12]–[16]. WC alternatively, has been shown to be a good predictor for abdominal adipose tissue [17], [18], but currently it is unclear to what extent the range of WC depends on body size [19], [20]. This has led to the idea that by combining traditional measures (e.g. height, weight, BMI, or WC) a better body index could be designed, which takes body shape into account [21]–[24].
Recently, two new body indices have been introduced [22], [23]. In 2012, Krakauer et al. developed ‘A Body Shape Index’ (ABSI), which is based on waist circumference (m), BMI (kg⋅m−2), and height (m) [22]. According to the authors, a high ABSI relates to a greater fraction of abdominal adipose tissue and appears to be a significant risk factor for premature death [22]. Other studies have suggested that ABSI is able to predict the onset of diabetes mellitus [25] and that it could be used to evaluate physical health status of adolescents [26]. In 2013, Thomas et al. developed the Body Roundness Index (BRI), which is a new geometrical index that combines height (m) and waist circumference (m) to predict the percentage of body fat and to evaluate health status [23]. However, it is unknown whether ABSI and BRI can determine the presence or risk of cardiovascular diseases.
Therefore, the aim of this study was to assess the capacity of the novel indices A Body Shape Index (ABSI) and Body Roundness Index (BRI) to identify cardiovascular diseases and cardiovascular disease risk factors in the Dutch population and to determine whether these anthropometric indices are superior to the Body Mass Index and Waist Circumference. We postulate that ABSI and BRI can better identify CVD than BMI or WC.
Methods
Ethics Statement
The study adhered to the Declaration of Helsinki. The Local Committee on Research Involving Human Subjects (CMO) of the region Arnhem and Nijmegen approved the study, and all subjects gave their written informed consent.
Study design and study population
The Nijmegen Exercise Study (www.nijmegenexercisestudy.com) is a large-population based study conducted by the Department for Health Evidence and the Department of Physiology of the Radboud University Medical Center in June 2011. The overall aim of the Nijmegen Exercise Study is to examine the effects of physical activity on general health and various disease outcomes in the general population.
In 2011, participants of the International Nijmegen Four Days Marches, the largest multi-day walking event in the world, were eligible to participate in the Nijmegen Exercise Study. By means of a passive recruitment strategy (Four Days Marches newsletter and internet advertisement), Dutch speaking adults were invited to complete an online survey. Participants were asked about their date of birth, anthropometric measures (weight, height, and WC), and whether they were diagnosed with CVD. In addition, participants were questioned about their lifestyle factors (physical activity and smoking habits).
Anthropometric measures
Self-reported body height, weight, and waist circumference.
Participants reported their height (centimetre), waist circumference (centimetre), and body weight (kilogram) in the online survey. Each anthropometric measurement was accompanied with detailed instructions. These data were used to calculate the BMI, ABSI, and BRI according to standardized formulas. BMI was based on weight (kg) and height (m) and calculated using formula [1] [27].(1)
A body shape index.
The ABSI was based on height (m), BMI (kg⋅m−2), and waist circumference (m) and calculated using formula [2] [22].(2)
The Body Roundness Index.
The BRI was based on height (m) and waist circumference (m). First, the eccentricity (ε) of the body was determined using formula [3]. Eccentricity (non-dimensional value) quantifies the degree of circularity of an ellipse, and it ranges between zero (perfect circle) to one (a vertical line) [23].(3)
Subsequently, the BRI was calculated using formula [4]. Described by Thomas et al., values closer to 1 are related to leaner individuals, whereas larger values are associated with rounder individuals [23].(4)
Presence of cardiovascular diseases
Participants were asked whether their physician diagnosed (yes/no) presence of CVD (1. myocardial infarction and 2. stroke) or CVD risk factors (1. hypertension and 2. hypercholesterolemia). Participants were categorized in the CVD or CVD risk factor group if one of these questions were replied positively, while they were allocated to the control group if all answers were negative. Participants who reported CVD as well as CVD risk factors were allocated to the CVD group.
Lifestyle factors
All participants reported whether they had followed a training program prior to the Nijmegen Four Days Marches and/or whether they performed other sport activities (e.g. playing soccer, tennis, etc.). Based on the duration (hours/week), frequency, and intensity (low, moderate, high) of these exercise activities, the average time and intensity level of physical activity per day was calculated. The recommendations of the American Heart Association were used to determine if a participant fulfilled the physical activity criteria [28].
Participants reported whether they were current smoker, ex-smoker or non-smokers and were categorized as (ex)-smoker or non-smoker.
Calculations and statistical analysis
Group characteristics were analysed using descriptive techniques. Continuous variables (age, height, weight, ABSI, BMI, BRI, and WC) were reported in tables as mean and differences between the CVD, CVD risk factor, and control group were analysed using one-way analysis of variance (ANOVA) with a Bonferroni post-hoc test. Categorical data were reported in proportions and differences tested by Pearson’s chi-squared test (sex, physical activity, myocardial infarction, stroke, hypertension, hypercholesterolemia, and smoking). Statistical significance was assumed at P<0.05.
Pairwise correlation coefficients between the continuous variables height, weight, ABSI, BMI, BRI, and WC were assessed by calculating Pearson correlation coefficients. To examine the discriminative power of the anthropometric indices for CVD or CVD risk factors the Area under the Receiver Operating Characteristic (AROC) curves were calculated. The AROC represents a measure of accuracy of an anthropometric index to discriminate between subjects with or without CVD or CVD risk factors [29].
Quintiles of BMI and BRI were created and the prevalence of CVD and CVD risk factors was calculated in each quintile. Since ABSI is strongly correlated with age and sex [22], ABSI was stratified for age (per year), after which ABSI quintiles were determined within each age group for males and females separately. To calculate the prevalence of CVD and CVD risk factors per ABSI quintile, subjects within the same ABSI quintile were merged. For WC the quintiles were stratified by sex and the prevalence of CVD and CVD risk factors in each quintile was calculated. Logistic regression was used to estimate the odds ratio for suffering from CVD or CVD risk factors per quintile, both unadjusted and adjusted for sex, age, and lifestyle factors (smoking and physical activity). For all four anthropometric indices, the lowest quintile was set as reference. All statistical analyses were performed using SPSS 20.0 (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp) software.
Results
5742 participants completed the online questionnaire. After the exclusion of participants who were pregnant, had underweight (calculated BMI of <18.5), or with missing data on anthropometric measures, the study population consisted of 4627 (80.6%) participants (Figure 1). This included 2425 men (57.5±11.4 years of age) and 2202 women (50.2±12.6 years of age). 1332 (27.7%) participants reported CVD or CVD risk factors, of which 32.9% (438) were female. Hypertension and hypercholesterolemia were reported in respectively 19.5% and 14.6% of the study population. Myocardial infarction and stroke were reported in respectively 2.6% and 1.4% of the study population (Table 1).
5742 participants completed the online questionnaire. 1115 participants were excluded because of missing data (n = 1065), pregnancy (n = 7), or BMI <18.5 (n = 43). The final group sample consisted of 4627 participants.
CVD, CVD risk factor and control group characteristics
The participants with CVD and CVD risk factors were older (62.9±8.3 vs. 59.6±8.8 vs. 51.6±13.0 year; P<0.001) compared the control group. The CVD and CVD risk factor group reported a higher body weight (81.6 (±11.5) vs. 78.1 (±13.1) vs. 74.5±12.6 kg; P<0.001) and smoked more often (71.5% vs. 61.6% vs. 49.5%; P<0.001). There were more participants physically active within the CVD risk factor group compared the control group (67.4 vs. 59.2; P<0.001), whereas the CVD and control group were comparable (66.5 vs. 59.2; P = 0.06) (Table 1). Furthermore, participants with CVD and CVD risk factors demonstrated higher ABSI, BMI, BRI, and WC values compared to the control group (Table 1).
Correlation and AROC scores
ABSI and BRI were positively and significantly correlated to height, weight, BMI and WC (Table 2; Figure 2). Within the CVD group, the AROC scores were 0.63 for ABSI, 0.68 for BRI, 0.64 for BMI, and 0.69 for WC (Figure 3A). Within the CVD risk factor group, the AROC scores were 0.57 for ABSI, 0.63 for BRI, 0.60 for BMI, and 0.61 for WC (Figure 3B).
A. Correlation between ABSI and BMI. Increase in ABSI and increase in BMI show a poor yet significant correlation (r = 0.087, P<0.001). B. Correlation between ABSI and WC. Increase in ABSI and increase in WC show a significant correlation (r = 0.678, P<0.001). C. Correlation between BRI and BMI. Increase in BRI and increase in BMI show a significant correlation (r = 0.780, P<0.001). D. Correlation between BRI and WC. Increase in BRI and increase in WC show a significant correlation (r = 0.898, P<0.001). E. Correlation between BRI and ABSI. Increase in BRI and increase in ABSI show a significant correlation (r = 0.642, P<0.001).
A. Area under the Receiver Operating Characteristic curve of ABSI, BRI, BMI, and WC to identify subjects with cardiovascular diseases. B. Area under the Receiver Operating Characteristic curve of ABSI, BRI, BMI, and WC to identify subjects with cardiovascular disease risk factors.
Prevalence of CVD and CVD risk factors
The prevalence of CVD increased per quintile for all four anthropometric indices (1st quintile vs. 5th quintile): ABSI 3.9% vs. 4.3%, BRI 0.8% vs. 6.5%, BMI 1.0% vs. 6.1%, WC 1.0% vs. 6.1% (P<0.05) (Table 3A). The prevalence of CVD risk factors increased per quintile for all four anthropometric indices (1st quintile vs. 5th quintile): ABSI 23.2% vs. 24.1%, BRI 11.6% vs. 36.9%, BMI 15.0% vs. 34.2%, WC 15.0% vs. 34.2% (P<0.05) (Table 3B).
Odds of cardiovascular diseases
The unadjusted odds of CVD prevalence increased with increasing quintiles for ABSI (ORunadjusted = 1.1 [0.7−1.8]), BRI (ORunadjusted = 9.1 [4.1−20.1]), BMI (ORunadjusted = 6.7 [3.3−13.5]), and WC (ORunadjusted = 3.4 [1.9−6.3]) (Table 4). Physical activity was non-significant in the logistic regression analysis and was therefore excluded from further analysis. After adjustment for age, sex, and smoking, the odds ratio of CVD was still significant for BRI (ORadjusted 3.2 [1.4−7.2]), BMI (ORadjusted = 3.3 [1.6−6.8]), and WC (ORadjusted = 3.0 [1.6−5.6]). However, the odds ratios did not vary across the ABSI quintiles (ORadjusted = 1.1 [0.7−1.8]) (Table 4).
Odds of cardiovascular disease risk factors
The unadjusted odds of CVD risk factor prevalence increased with increasing quintiles for ABSI (ORunadjusted = 1.1 [0.8−1.3]), BRI (ORunadjusted = 4.5 [3.5−5.7]), BMI (ORunadjusted = 2.9 [2.3−3.7]), and WC (ORunadjusted = 2.4 [2.0−3.0]) (Table 5). After adjustment for age, sex, and smoking, the odds ratio of CVD was still significant for BRI (ORadjusted 2.5 [2.0−3.3]), BMI (ORadjusted = 2.4 [1.9−3.1]), and WC (ORadjusted = 2.0 [1.6−2.5). However, the odds ratios did not vary across the ABSI quintiles (ORadjusted = 1.1 [0.9−1.3]) (Table 5).
Discussion
We demonstrated that the prevalence of CVD and CVD risk factors significantly increased across the quintiles for ABSI, BRI, BMI, and WC. However, after adjusting for age, sex, and smoking only BRI, BMI, and WC could identify CVD or CVD risk factors and not ABSI. In contrast to our hypothesis, neither ABSI nor BRI are superior measures compared to BMI and WC to determine the presence of CVD or CVD risk factors. Our findings indicate that the novel index ABSI is not suitable for identifying CVD or CVD risk factors, while the BRI showed similar capabilities as BMI and WC.
A Body Shape Index
The ABSI was created to produce a quantitative measure to estimate the health of body shape independently of body size (height, weight, and BMI) [22]. While Krakauer et al. demonstrated that ABSI predicts premature mortality better than BMI or WC [22], our results suggest that ABSI cannot be used to distinguish between individuals with and without CVD or CVD risk factors. Compared with BRI, BMI, and WC, the ABSI demonstrated the lowest predictive power and had no significant association with CVD or CVD risk factors prevalence after adjustment for age, sex, and smoking. Thus, ABSI is not a suitable measurement to identify CVD and CVD risk factors.
A possible explanation for the contrasting findings between our data and Krakauer et al., is the endpoint variable, namely the prevalence of CVD and CVD risk factors versus premature death. Alternatively, subject characteristics may explain the dissimilarities between our study and the one of Krakauer et al. [30]. Our study population has approximately the same ABSI values as Krakauer i.e. 0.081±0.0058 m11/6⋅kg−2/3 vs. 0.081±0.0053 m11/6⋅kg−2/3, but surprisingly our study population had a lower BMI (3rd quintile: 23.7–25.2 vs. 25.6–28.4) and lower WC (3rd quintile females: 82–87 vs. 88–97; males: 94–98 vs. 94–101). The explanation why ABSI is similar is that our study population comprising Dutch subjects were taller compared to the American study population of Krakauer et al. The estimated average body height is 1.70 meters in the study of Krakauer et al., whereas our study population had an average body height of 1.75 meters. This suggests that body height might confound the capacity of ABSI to identify CVD in our study population. Future studies should further investigate the limits of ABSI and especially study the impact of body height on the calculation of ABSI.
Body Roundness Index
We are the first to study the capacity of BRI to identify CVD and CVD risk factors. The BRI was developed to predict both body fat and the percentage visceral adipose tissue by using WC in relation to height, which allows estimation of the shape of the human body figure as an ellipse or oval [23]. We demonstrated that the BRI is capable to identify CVD and CVD risk factors (Table 3). This is in agreement with previous studies [21], [24] which also related WC to body height (i.e. waist-to-height ratio [WtH-ratio]). The spearman rank test revealed a perfect nonlinear relation between BRI and WtH-ratio (r = 1; P = 0.00). This indicates that both body indices are related by an one-to-one nonlinear transformation and strictly means that it does not matter what index (BRI or WtH-ratio) is used to determine CVD. However, as demonstrated by Thomas et al., the advantage of the BRI over the WtH-ratio is that it also can be used to estimate the amount of body fat percentage and gives therefore a better impression of physical health status. Although the adjusted OR of the BRI was higher compared to BMI and WC, this did not reach statistical significance. Nevertheless, we recommend that future studies should study the sensitivity of BRI to determine cardiovascular health risks in a clinical setting. With the development of computerized software that is accessible via internet websites, Tablets or Smartphone applications, novel body measures with complex algorithms can easily be used by physicians [22], [23]. Therefore, the BRI has the potential to improve the detection, evaluation, and progression of CVD and CVD risk factors.
Methodological considerations
Our study compared the ABSI and BRI mutually and against the BMI and WC in a large heterogeneous study population. In our study, participants self-reported their weight and WC using our written instructions, whilst in the study of Krakauer et al. trained personnel measured WC. Although, it is likely that participants made some measurement errors, prior studies demonstrated that self-report of body weight and waist circumference is a validated and feasible method [31], [32]. We therefore suppose that self-reporting has had a minor influence on our findings. Additionally, the outcome variable (CVD) was self-reported; participants were only allowed to report CVD or CVD risk factors if they were diagnosed by a physician. Possibly some participants had a history of CVD and did not report this. This could have led to an underestimation of our results. Therefore, the capacity of BRI, BMI, and WC to identify CVD or CVD risk factors could somewhat be higher. Finally, one might argue that the cross-sectional nature of our study is suboptimal to study the predictive capacity of anthropometric characteristics. However, in our study design we demonstrated that the BRI as well as BMI and WC were able to identify participants with CVD. It is a first step in validating BRI in relation to CVD and therefore we believe that the used method is appropriate. In future studies, the longitudinal relation between BRI and CVD incidence should be studied.
Conclusion
In the current study, we demonstrated that CVD and CVD risk factors prevalence increased per quintile across a heterogeneous population for ABSI, BRI, BMI, and WC. Nonetheless, only BRI, BMI, WC, and not ABSI, could significantly determine the presence of CVD and CVD risk factor after correction for sex, age and smoking. The capacity of BRI to mathematically model the human body shape gives an adequate impression of the cardiovascular health status, but was not superior to BMI or WC.
Acknowledgments
We recognize the excellent help of the organization of the International Nijmegen Four Days Marches, Monique Eijgenberger, Erik Brummelkamp, Maurits Hoogerwerf, and Jan Verbeek. In addition, we thank Dick Thijssen and Jos van Dijck for critically reviewing the online questionnaire.
Author Contributions
Conceived and designed the experiments: TE MH AV. Performed the experiments: MM RV TE FV. Analyzed the data: MM AV FV. Wrote the paper: MM FV TE RV MH AV. Data interpretation: MM FV TE RV MH AV.
References
- 1.
WHO (2013) Obesity and overweight - Fact sheet No. 311. Geneva: World Health Organisation.
- 2. Hubert HB, Feinleib M, McNamara PM, Castelli WP (1983) Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation 67: 968–977.
- 3. Must A, Spadano J, Coakley EH, Field AE, Colditz G, et al. (1999) The disease burden associated with overweight and obesity. JAMA 282: 1523–1529.
- 4. Wilson PW, D’Agostino RB, Sullivan L, Parise H, Kannel WB (2002) Overweight and obesity as determinants of cardiovascular risk: the Framingham experience. Arch Intern Med 162: 1867–1872.
- 5. Van Gaal LF, Mertens IL, De Block CE (2006) Mechanisms linking obesity with cardiovascular disease. Nature 444: 875–880.
- 6. Kahn SE, Hull RL, Utzschneider KM (2006) Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 444: 840–846.
- 7. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, et al. (2001) Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 344: 1343–1350.
- 8. Flegal KM, Kit BK, Orpana H, Graubard BI (2013) Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA 309: 71–82.
- 9. Adams KF, Schatzkin A, Harris TB, Kipnis V, Mouw T, et al. (2006) Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. N Engl J Med 355: 763–778.
- 10.
WHO (2000) Obesity: preventing and managing the global epidemic. Report of a World Health Organization consultation. World Health Organ Tech Rep Ser 894: i–xii, 1–253.
- 11.
Van Binsbergen JJ, Langens FNM, Dapper ALM, Van Halteren MM, Glijsteen R, et al. (2010) NHG-Standaard Obesitas M95. NHG-Standaarden. Available: https://www.nhg.org/standaarden/volledig/nhg-standaard-obesitas. Accessed 2014 Aug 28.
- 12. Romero-Corral A, Somers VK, Sierra-Johnson J, Thomas RJ, Collazo-Clavell ML, et al. (2008) Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes (Lond) 32: 959–966.
- 13. Wunderlich S, Brusca J, Bai Y (2012) Overweight and Obesity among Older Adults: How should Obesity be Determined. J Obes Wt Loss Ther 2: 2.
- 14. Kok P, Seidell JC, Meinders AE (2004) [The value and limitations of the body mass index (BMI) in the assessment of the health risks of overweight and obesity]. Ned Tijdschr Geneeskd 148: 2379–2382.
- 15. Nevill AM, Stewart AD, Olds T, Holder R (2006) Relationship between adiposity and body size reveals limitations of BMI. Am J Phys Anthropol 129: 151–156.
- 16. Prentice AM, Jebb SA (2001) Beyond body mass index. Obes Rev 2: 141–147.
- 17. Pouliot MC, Despres JP, Lemieux S, Moorjani S, Bouchard C, et al. (1994) Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol 73: 460–468.
- 18. Janssen I, Heymsfield SB, Allison DB, Kotler DP, Ross R (2002) Body mass index and waist circumference independently contribute to the prediction of nonabdominal, abdominal subcutaneous, and visceral fat. Am J Clin Nutr 75: 683–688.
- 19.
WHO (2008) WHO Expert Consultation on Waist Circumference and Waist-Hip Ratio, Geneva, 8–11 December 2008. World Health Organization Library Cataloguing-in-Publication Data.
- 20. Hsieh SD, Yoshinaga H (1999) Do people with similar waist circumference share similar health risks irrespective of height? Tohoku J Exp Med 188: 55–60.
- 21. Browning LM, Hsieh SD, Ashwell M (2010) A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0.5 could be a suitable global boundary value. Nutr Res Rev 23: 247–269.
- 22. Krakauer NY, Krakauer JC (2012) A new body shape index predicts mortality hazard independently of body mass index. PLoS One 7: e39504.
- 23. Thomas DM, Bredlau C, Bosy-Westphal A, Mueller M, Shen W, et al. (2013) Relationships between body roundness with body fat and visceral adipose tissue emerging from a new geometrical model. Obesity (Silver Spring) 21(11): 2264–2271.
- 24. Ashwell M, Gunn P, Gibson S (2012) Waist-to-height ratio is a better screening tool than waist circumference and BMI for adult cardiometabolic risk factors: systematic review and meta-analysis. Obes Rev 13: 275–286.
- 25. He S, Chen X (2013) Could the new body shape index predict the new onset of diabetes mellitus in the chinese population? PLoS One 8: e50573.
- 26. Duncan MJ, Mota J, Vale S, Santos MP, Ribeiro JC (2013) Associations between body mass index, waist circumference and body shape index with resting blood pressure in Portuguese adolescents. Ann Hum Biol 40(2): 163–167.
- 27. Keys A, Fidanza F, Karvonen MJ, Kimura N, Taylor HL (1972) Indices of relative weight and obesity. J Chronic Dis 25: 329–343.
- 28. Haskell WL, Lee IM, Pate RR, Powell KE, Blair SN, et al. (2007) Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation 116: 1081–1093.
- 29. Zweig MH, Campbell G (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 39: 561–577.
- 30. Deurenberg P, Yap M, van Staveren WA (1998) Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab Disord 22: 1164–1171.
- 31. Dekkers JC, van Wier MF, Hendriksen IJ, Twisk JW, van Mechelen W (2008) Accuracy of self-reported body weight, height and waist circumference in a Dutch overweight working population. BMC Med Res Methodol 8: 69.
- 32. Rimm EB, Stampfer MJ, Colditz GA, Chute CG, Litin LB, et al. (1990) Validity of self-reported waist and hip circumferences in men and women. Epidemiology 1: 466–473.