Secondary Logo

Journal Logo

Anthropometric Predictors of Hemoglobin A1c among Adults

NHANES 2003 to 2004 and 2013 to 2014

Nicolo, Michele L.1; Compher, Charlene W.2; Shewokis, Patricia A.3,4; Boullata, Joseph I.3; Sukumar, Deeptha3; Smith, Sinclair A.5; Volpe, Stella L.2

Translational Journal of the American College of Sports Medicine: November 1, 2019 - Volume 4 - Issue 21 - p 242–247
doi: 10.1249/TJX.0000000000000107
Original Investigation

ABSTRACT Nearly 10% of adult Americans have type 2 diabetes mellitus (DM), yet 25% are undiagnosed. Our purpose was to compare body mass index (BMI), waist circumference (WC), and waist-to-height ratio (WHtR) as predictors for type 2 DM in adults 40 to 59 yr of age. We hypothesized that BMI, WC, and WHtR would predict hemoglobin A1c (HbA1c) ≥6.5%, diagnostic of diabetes. Data from the National Health and Nutrition Examination Survey for 2003 to 2004 (N = 1069) and 2013 to 2014 (N = 906) were used in logistic regression models. There were differences in ethnic/racial distribution in the 2003 to 2004 and 2013 to 2014 sample. BMI, WC, and WHtR were higher in 2003 to 2004 than 2013 to 2014 (BMI, 29.5 vs 25.7 kg·m−2; WC, 99.8 vs 87.1 cm; WHtR, 0.59 vs 0.55, respectively, each P < 0.0001). In 2003 to 2004, WC (odds ratio = 2.65, 95% confidence interval = 1.57–4.48) and WHtR (odds ratio = 2.91, 95% confidence interval = 1.69–5.04) predicted HbA1c ≥ 6.5%, but BMI did not. In 2013 to 2014, BMI, WC, or WHtR did not predict HbA1c ≥ 6.5%. WC or WHtR may suggest risk of type 2 DM in some populations.

1Institute of Health Promotion and Disease Prevention, Keck School of Medicine, University of Southern California, Los Angeles, CA

2School of Nursing, University of Pennsylvania, Philadelphia, PA

3Department of Nutrition Sciences, Drexel University, Philadelphia, PA

4School of Biomedical Engineering, Science and Health Systems, Drexel University, Philadelphia, PA

5Department of Health Sciences, Drexel University, Philadelphia, PA

Address for correspondence: Stella L. Volpe, Ph.D., R.D., A.C.S.M.-C.E.P., F.A.C.S.M., Department of Nutrition Sciences, Drexel University, Three Parkway Building, 1601 Cherry Street, Mail Stop 31030 Philadelphia, PA 19102 (E-mail:

Key Points

  • Links between obesity and type 2 DM have been established
  • This study suggests WC and WHtR may be a stronger approach to identifying the risk for type 2 DM in adults living in the United States than using BMI.
Back to Top | Article Outline


The prevalence of type 2 diabetes mellitus (type 2 DM) in the United States is approximately 29 million, and about 25% of the population remains undiagnosed (1,2). The risk of developing type 2 DM increases with age, becoming more substantial after 45 yr of age (3). The prevalence of type 2 DM in adults increased from 6.6% in 2003 to 7% in 2004 (4), then to 8.7% by 2013 and 8.4% by 2014, and continued to increase in 2015 to 8.7% (4,5).

Obesity is considered a risk factor for insulin resistance and development of type 2 DM (6–8). Although the mechanism is not well understood, it is suggested that an increase in fat tissue interferes with glucose metabolism, accentuating insulin resistance (2,6,9–15). In 2003 to 2004, the prevalence of obesity among adults, 20 yr of age and older, was about 32% (16), increasing to 36.5% by 2014 (17). Prevalence of obesity is greater among adults between 40 and 59 yr of age (16,18). In 2014, the prevalence of obesity among adults in this age-group was 40.2% (19) and continued to increase to 42.8% in 2016 (18).

Body mass index (BMI) is an accepted surrogate measure of total body fat. Published standards of care have recognized a BMI greater than 25 kg·m−2 as a risk factor for developing type 2 DM (20,21). Body fat distribution cannot be determined through use of BMI (22). In addition, BMI categories classifying overweight or obesity are not consistent among all ethnic populations (23,24). Central adiposity, measured using waist circumference (WC), is associated with an accumulation of visceral fat, increasing the likelihood of developing insulin resistance (25–29). Like WC, waist-to-height ratio (WHtR) may be used to identify individuals with increased central adiposity relative to their height. A WHtR greater than 0.5 is associated with increased risk of metabolic diseases including cardiovascular disease and type 2 DM, which is consistent among women and men, all ages and ethnicities (30–32). Although not considered standard practice, WHtR may be a more practical measurement to determine type 2 DM risk compared with BMI.

Since 2008, hemoglobin A1c (HbA1c) measure ≥6.5% has been considered an acceptable diagnostic method for type 2 DM by the American Diabetes Association (33). As a strategy to identify simple, feasible screening tools for public health practice, we evaluated whether BMI, WC, or WHtR predicted HbA1c ≥ 6.5% in an older adult U.S. sample, using data from the National Health and Nutrition Examination Survey (NHANES) for the years 2003 to 2004 and 2013 to 2014. We hypothesized that BMI, WC, and WHtR would predict HbA1c ≥ 6.5%.

Back to Top | Article Outline


This study was conducted under regulatory approval by Drexel University’s Institutional Review Board. Existing data were used from the databases for NHANES 2003 to 2004 and 2013 to 2014. The NHANES is a cross-sectional survey used to evaluate nutrition, health, and health behaviors of civilian, noninstitutionalized persons living in the United States (34). By contrast to the 2003 to 2004 survey, the 2013 to 2014 survey included Asian Americans, and individuals with incomes less than 130% above the poverty line were oversampled (34). However, because Asian Americans could not be identified as a discrete group in the 2003 to 2004 sample, our analyses were limited to non-Hispanic White, non-Hispanic Black, and Hispanic (including Mexican American) participants. Nonpregnant adults, between 40 and 59 yr of age, were included in the present study. The sample was further restricted to include only participants who had the following variables: HbA1c% and measured body weight, standing height, and WC to permit computation of BMI and WHtR. BMI was calculated by dividing weight measured in kilograms by height measured in meters squared (35). WHtR was calculated by dividing the measured WC (cm) by the standing height (cm). A total of 1069 and 906 adults, 40 to 59 yr of age, from NHANES 2003 to 2004 and NHANES 2013 to 2014, respectively, were included in the analyses.

Back to Top | Article Outline

Statistical methods

Continuous variables were reported as mean ± SD, and categorical variables as frequency and percentage. Continuous variables for BMI, WC, and WHtR were standardized using z-scores. Established criteria for BMI categories and values of WC and WHtR were used to describe the two samples (26,36,37).

Our primary aim was to determine whether BMI, WC, or WHtR predicted HbA1c ≥ 6.5%. We conducted a univariate binary logistic with z-scores for BMI, WC, and WtHtR to determine the odds ratio (OR) and 95% confidence interval (CI) for HbA1c ≥ 6.5%, while adjusting for sex, race, and ethnicity. A multiple logistic regression model was built to determine the comparative strength of predicting HbA1c ≥ 6.5%, including BMI, WC, and WHtR. Models were adjusted to account for multicollinearity between WC and WHtR. For all analyses, the Statistical Package for the Social Sciences software (version 24; SPSS Inc., Chicago, IL) was used, and P < 0.05 was considered statistically significant.

Back to Top | Article Outline


Characteristics of the two samples are in Table 1. The mean age and percentage of women and men included in our study were similar in both samples. The percentage of non-Hispanic White adults was greater (55.5% vs 47.3%, respectively), but the percentage of Hispanics was lower in 2003 to 2004 than that in 2013 to 2014 (22.1% vs 28.5%), P < 0.0001.



The body composition measures of the two samples are listed in Table 1. BMI, WC, and WHtR were significantly higher in 2003 to 2004 compared with 2013 to 2014 (P < 0.0001 for each measure). Reference categories for BMI, WC, and WHtR are displayed in Table 2. In 2003 to 2004, there were fewer underweight (0.7% vs 20.6%) and healthy weight (23.6% vs 29.3%) individuals but more overweight (36.5% vs 23.6%) and obese (39.2% vs 26.5%) participants compared with 2013 to 2014 (P < 0.0001). More adults had a WC greater than 80 cm (91.7% vs 63.5%, respectively; P < 0.0001) and/or a WHtR greater than 0.5 (87.7% vs 67.2%, respectively; P < 0.0001) in 2003 to 2004 compared with 2013 to 2014.



The logistic regression models are listed in Table 3. In the univariate models, BMI was positively associated with HbA1c ≥ 6.5% (OR = 1.51, 95% CI = 1.26–1.82, P < 0.0001), WC (OR = 1.70, 95% CI = 1.39–2.08, P < 0.0001), and WHtR (OR = 1.76, 95% CI = 1.45–2.14, P < 0.0001) in 2003 to 2004. In the model adjusted for BMI, sex, race, and ethnicity, WC (OR = 2.65, 95% CI = 1.57–4.48, P < 0.0001) and WHtR (OR = 2.91, 95% CI = 1.69–5.04, P < 0.0001) were positively associated with HbA1 ≥ 6.5%. However, in 2013 to 2014, BMI, WC, or WHtR was not associated with HbA1 ≥ 6.5%.



Back to Top | Article Outline


In this large sample of noninstitutionalized adults from NHANES 2003 to 2004, WC and WHtR predicted type 2 DM, but BMI did not. There were more non-Hispanic White and fewer Hispanic adults included in NHANES 2003 to 2004 than that in the 2013 to 2014 sample. In the 2013 to 2014 sample, BMI, WC, or WHtR did not predict type 2 DM, but all three anthropometric variables were significantly lower compared with 2003 to 2004.

It is important to note that several public health initiatives targeting DM awareness and education were introduced between 2002 and 2014 (Fig. 1) (38). These campaigns may have increased the number of individuals diagnosed and treated for type 2 DM by 2013 to 2014, although the NHANES data do not enable the identification of individuals treated with diabetes medications. Less than 2% of both samples reported insulin use for diabetes management. In addition, greater than 90% of the data for oral diabetes medication use was missing.

Figure 1

Figure 1

The Affordable Care Act (39) was implemented in 2010. Casagrande et al. (40) investigated changes in insurance coverage among adults diagnosed with DM in the United States. On the basis of data collected through the National Health Interview Surveys between 2009 and 2016, they reported that insurance coverage among adults with DM increased by 5.4% (P < 0.001), and adults covered by Medicaid increased by 4.8% (P < 0.05) (40). In addition, Casgrande et al. (40) reported that medical costs were reported to have decreased among adults reporting a household income less than $35,000 per year (P = 0.004). The education and screening practices associated with greater access to health care may have helped frame the positive changes in health outcomes associated with type 2 DM. If a larger portion of the population were treated for type 2 DM in 2013 to 2014, the HbA1c might have been reduced.

Oversampling of certain populations participating in NHANES 2013 to 2014 data collection, including adults meeting criteria for below the federal poverty level, may have played a role in the findings (34). The major increase in the number of individuals with a BMI < 18.5 kg·m−2 in 2013 to 2014 compared with 2003 to 2004 is an unusual trend given the overall increase in obesity in the same time frame (16,41). However, individuals with limited energy intake in the context of food insecurity and/or work that requires greater energy expenditure may be less likely to be overweight or obese.

Although race and ethnicity were included as control variables in our analyses, the greater number of Hispanic adults participating in NHANES 2013 to 2014 compared with 2003 to 2004, concurrent with a smaller number of non-Hispanic White people, may explain differences in anthropometric measures between the two NHANES cycles. The oversampling of these groups was intended to fill a gap in our public health knowledge about these populations. Genetic predisposition for type 2 DM among individuals, including Mexican Americans, may have contributed to the observed differences between the two samples (42).

On the basis of NHANES 2003 to 2004, increased abdominal fat, indicated by a greater WC and/or greater WHtR, may capture more adults at risk for type 2 DM than BMI. Adults with increased central or abdominal fat may be overlooked because their BMI or direct measure of WC does not meet the criteria for obesity (2). Others have reported findings to support this observation. Ashwell and Gibson (22) suggest that using the WHtR ratio may identify more adults with an “early risk” for cardiometabolic disease, including type 2 DM.

Other investigators considered use of BMI, WC, and WHtR in varied sex, race, and ethnic samples. Huerta et al. (43) reported that in 37,733 Spanish adults, WHtR had a greater predictive risk for type 2 DM than BMI among women, whereas BMI was a better predictive risk for type 2 DM among men. In the U.S. Atherosclerosis Risk in Communities Study, WC was a stronger predictor for type 2 DM in White and Black women than BMI or WHtR (29). However, in White and Black men, WHtR was a stronger for type 2 DM than BMI or WC. In our study, we included sex, race, and ethnicity as control variables and, thus, did not investigate differences in risk for type 2 DM using anthropometric measures between women and men or among racial or ethnic groups.

BMI has been implemented as a screening tool for the risk of type 2 DM (44); however, our data suggest that relying on BMI alone may result in the exclusion of some individuals at risk for type 2 DM because BMI does not consider abdominal adiposity. Body fat distribution measures, such as WC and WHtR, could be used to identify individuals at risk for type 2 DM who meet the criteria for healthy BMI but have greater abdominal fat distribution.

This study had several strengths. Our study included data from a large sample from NHANES, which is a nationally representative survey of health and health behaviors of individuals living in the United States. Data included in this study were based on objective measurements performed by trained staff. In addition, our study is the first to investigate anthropometric predictors of HbA1c% in two samples at 10-yr intervals in adults at greater risk for type 2 DM because of their older age. We used rigorous statistical approaches, including accounting for multicollinearity between body composition variables.

Our study had several limitations as well. Although we analyzed a large sample using information from a cross-sectional survey, including objective measures, these concurrent measures cannot imply causality between central adiposity and type 2 DM. Although including adults 40 to 59 yr of age was a strength of our study, restricting our sample to adults 40 to 59 yr of age limits the generalizability of the results to the larger population. We included data from two points in time and did not investigate differences among the population or trends in HbA1c over the intervening years. Our sample was limited to participants who identified as non-Hispanic White, Black, Hispanic, and Mexican American and would not apply to others. The use of oral diabetes medications was not included in our analyses because of missing data; thus, individuals who were treated for type 2 DM with medications and thus have lower HbA1c values may have been included. The implementation of diabetes education initiatives and the Affordable Care Act may have resulted in greater likelihood of diagnosis and treatment of the disease between 2003 and 2014.

Back to Top | Article Outline


In conclusion, although there are several anthropometric methods available to screen for risk of type 2 DM, measures such as BMI may exclude a large number of individuals at risk. WC and WHtR may be alternative methods of identifying the risk for metabolic diseases, including type 2 DM. These measures may capture individuals who otherwise might be overlooked due to having a healthy BMI. Longitudinal studies, where serial anthropometric and HbA1c% measures are collected, may provide a clearer understanding of the relationship between body fat distribution and the risk for type 2 DM. Further exploration of optimal predictors of type 2 DM in Hispanic adults and those of low socioeconomic status is also important.

The authors have no funding to disclose and no disclosure to report. The authors confirm that the views of this paper do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Centers for Disease Control and Prevention. National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States. Atlanta (GA): U.S. Department of Health and Human Services; 2014.
2. Carey DG, Jenkins AB, Campbell LV, et al. Abdominal fat and insulin resistance in normal and overweight women: direct measurements reveal a strong relationship in subjects at both low and high risk of NIDDM. Diabetes. 1996;45(5):633–8.
3. Cowie CC, Rust KF, Byrd-Holt DD, et al. Prevalence of diabetes and high risk for diabetes using A1C criteria in the U.S. population in 1988–2006. Diabetes Care. 2010;33(3):562–8.
4. Centers for Disease Control and Prevention. Diagnosed Diabetes, Age-Adjusted Percentage, Adults with Diabetes. 2015 February 2 [cited 2018 August 27, 2018]. Available from:
5. Unwin N, Gan D, Whiting D. The IDF Diabetes Atlas: providing evidence, raising awareness and promoting action. Diabetes Res Clin Pract. 2010;87(1):2–3.
6. Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature. 2006;444(7121):840–6.
7. Van Itallie TB. Health implications of overweight and obesity in the United States. Ann Intern Med. 1985;103(6 Part 2):983–8.
8. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA. 2003;289(1):76–9.
9. Frayn KN. Visceral fat and insulin resistance—causative or correlative? Br J Nutr. 2000;83(Suppl 1):S71–7.
10. Dandona P, Aljada A, Bandyopadhyay A. Inflammation: the link between insulin resistance, obesity and diabetes. Trends Immunol. 2004;25(1):4–7.
11. Leong KS, Wilding JP. Obesity and diabetes. Best Pract Res Clin Endocrinol Metab. 1999;13(2):221–37.
12. Philipsen A, Jorgensen ME, Vistisen D, et al. Associations between ultrasound measures of abdominal fat distribution and indices of glucose metabolism in a population at high risk of type 2 diabetes: the ADDITION-PRO study. PLoS One. 2015;10(4):e0123062.
13. Weyer C, Funahashi T, Tanaka S, et al. Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metabol. 2001;86(5):1930–5.
14. Bergman RN. Non-esterified fatty acids and the liver: why is insulin secreted into the portal vein? Diabetologia. 2000;43(7):946–52.
15. Peiris AN, Struve MF, Struve MF, et al. Glucose metabolism in obesity: influence of body fat distribution. J Clin Endocrinol Metab. 1988;67(4):760–7.
16. Ogden CL, Carroll MD, Curtin LR, et al. Prevalence of overweight and obesity in the United States, 1999–2004. JAMA. 2006;295(13):1549–55.
17. Ogden CL, Fakhouri TH, Carroll MD, et al. Prevalence of obesity among adults, by household income and education—United States, 2011–2014. MMWR Morb Mortal Wkly Rep. 2017;66(50):1369–73.
18. Hales CM, Carroll MD, Fryar CD, et al. Prevalence of obesity among adults and youth: United States, 2015–2016. NCHS Data Brief. 2017;(288):1–8.
19. Ogden CL, Carroll MD, Fakhouri TH, et al. Prevalence of obesity among youths by household income and education level of head of household—United States 2011–2014. MMWR Morb Mortal Wkly Rep. 2018;67(6):186–9.
20. Klein S, Sheard NF, Pi-Sunyer X, et al. Weight management through lifestyle modification for the prevention and management of type 2 diabetes: rationale and strategies. A statement of the American Diabetes Association, the North American Association for the Study of Obesity, and the American Society for Clinical Nutrition. Am J Clin Nutr. 2004;80(2):257–63.
21. American Diabetes Association. Screening for type 2 diabetes. Diabetes Care. 2004;27(Suppl 1):s11–4.
22. Ashwell M, Gibson S. Waist-to-height ratio as an indicator of ‘early health risk’: simpler and more predictive than using a ‘matrix’ based on BMI and waist circumference. BMJ Open. 2016;6(3):e010159.
23. Rush E, Goedecke JH, Jennings C, et al. BMI, fat and muscle differences in urban women of five ethnicities from two countries. Int J Obes (Lond). 2007;31(8):1232.
24. Sumner AE, Ricks M, Sen S, et al. How current guidelines for obesity underestimate risk in certain ethnicities and overestimate risk in others. Current Cardiovascular Risk Reports. 2007;1(2):97–101.
25. Janssen I, Katzmarzyk PT, Ross R. Waist circumference and not body mass index explains obesity-related health risk. Am J Clin Nutr. 2004;79(3):379–84.
26. Janssen I, Katzmarzyk PT, Ross R. Body mass index, waist circumference, and health risk: evidence in support of current National Institutes of Health guidelines. Arch Intern Med. 2002;162(18):2074–9.
27. Ganz ML, Wintfeld N, Li Q, et al. The association of body mass index with the risk of type 2 diabetes: a case-control study nested in an electronic health records system in the United States. Diabetol Metab Syndr. 2014;6(1):50.
28. Hardy DS, Stallings DT, Garvin JT, et al. Best anthropometric discriminators of incident type 2 diabetes among white and black adults: a longitudinal ARIC study. PLoS One. 2017;12(1):e0168282.
29. Hardy DS, Stallings DT, Garvin JT, et al. Anthropometric discriminators of type 2 diabetes among White and Black American adults. J Diabetes. 2017;9(3):296–307.
30. Ashwell M, Clarke G. Obesity risk: importance of the waist-to-height ratio. Nurs Stand. 2009;23(41):49.
31. Ashwell M. Waist to height ratio and the Ashwell® shape chart could predict the health risks of obesity in adults and children in all ethnic groups. Nutr Food Sci. 2005;35(5):359–64.
32. Rodea-Montero ER, Evia-Viscarra ML, Apolinar-Jiménez E. Waist-to-height ratio is a better anthropometric index than waist circumference and BMI in predicting metabolic syndrome among obese Mexican adolescents. Int J Endocrinol. 2014;2014:195407.
33. Saudek CD, Brick JC. The clinical use of hemoglobin A1c. J Diabetes Sci Technol. 2009;3(4):629–34.
34. Johnson CL, Dohrmann SM, Burt VL, et al. National Health and Nutrition Examination Survey: sample design, 2011–2014. Vit Health Stat 2. 2014;(162):1–33.
35. Centers for Disese Control and Prevention. National Health and Nutrition Examination Survey (NHANES): Anthropometry Procedures Manual. Atlanta (GA): Centers for Disease Control and Prevention; 2004.
36. Romero-Corral A, Somers VK, Sierra-Johnson J, et al. Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes (Lond). 2008;32(6):959.
37. World Health Organization. Waist Circumference and Waist–Hip Ratio: Report of a WHO Expert Consultation, Geneva, 8–11 December 2008. World Health Organization; 2011.
38. Saaddine JB, Cadwell B, Gregg EW, et al. Improvements in diabetes processes of care and intermediate outcomes: United States, 1988–2002. Ann Intern Med. 2006;144(7):465–74.
39. Rak S, Coffin J. Affordable Care Act. J Med Pract Manage. 2013;28(5):317–9.
40. Casagrande SS, McEwen LN, Herman WH. Changes in health insurance coverage under the Affordable Care Act: a national sample of U.S. adults with diabetes, 2009 and 2016. Diabetes Care. 2018;41(5):956–62.
41. Hales CM, Carroll MD, Fryar CD, et al. Prevalence of obesity among adults and youth: United States, 2015–2016. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics; 2017.
42. Horikawa Y, Oda N, Cox NJ, et al. Genetic variation in the gene encoding calpain-10 is associated with type 2 diabetes mellitus. Nat Genet. 2000;26(2):163.
43. Huerta JM, Tormo MJ, Chirlaque MD, et al. Risk of type 2 diabetes according to traditional and emerging anthropometric indices in Spain, a Mediterranean country with high prevalence of obesity: results from a large-scale prospective cohort study. BMC Endocr Disord. 2013;13:7.
44. Meigs JB, Wilson PW, Fox CS, et al. Body mass index, metabolic syndrome, and risk of type 2 diabetes or cardiovascular disease. J Clin Endocrinol Metabol. 2006;91(8):2906–12.
© 2019 American College of Sports Medicine