Ethnic Group Differences in Waist Circumference Percentiles Among U.S. Children and Adolescents: Estimates from the 1999–2008 National Health and Nutrition Examination Surveys

Introduction

W aist circumference (WC) has been proposed as a criterion to define abdominal adiposity because it is highly related to insulin resistance and other cardiometabolic risk factors and because body mass index (BMI) may not perfectly reflect fatness. ¹ For example, in one study, only visceral fat [as measured by magnetic resonance imaging (MRI)], but not BMI or waist-to-hip ratio, was associated with fasting insulin levels and triglycerides in obese adolescent girls. ² Specifically, WC is a component of the metabolic syndrome, a constellation of elevated cardiometabolic risk factors, including blood pressure, lipids, and glucose in adults ¹ and in children and adolescents. ^{3 –10}

A recent study reported that adults with a normal BMI but a large WC had a higher risk of premature mortality, indicating that increased WC should be considered as a risk factor for mortality, in addition to BMI. ⁹ BMI is currently the standard method for defining overweight and obesity in children, adolescents, and adults. Yet, increased visceral or abdominal fat is positively associated with metabolic disease risk, ¹¹ independent of overall adiposity. ¹² Furthermore, in children, WC is a better indicator of visceral fat (body adipose tissue located within the abdominal cavity around the visceral organs) than is BMI. ^12,13

Although visceral fat can be assessed accurately by imaging techniques, such as computed tomography and magnetic resonance imaging, ¹³ these techniques are often not feasible for identifying people with abdominal obesity in large epidemiological studies, mass screenings, or clinical settings. WC is a simple, yet effective, way of measuring abdominal obesity in adults ¹⁴ and children ¹⁵ and may be a better predictor of cardiovascular disease risk than BMI in adults and children. ⁴ For WC to be useful in clinical applications, its distribution by age, sex, and ethnicity needs to be characterized. ¹⁶ Here we report waist circumference percentiles estimates for U.S. Mexican-American (MA), non-Hispanic black (NHB) and non-Hispanic white (NHW) 2 to 18 year olds by sex using the most current population-based data [1999–2008 National Health and Nutrition Examination Surveys (NHANES)] and statistical techniques applied to the NHANES III (1988–1994) data. ¹⁶

Methods

Results

Table 1 provides age-specific estimated WC values for the entire U.S. pediatric population. Overall, WC measurements increased as children get older, but this varied by age and sex. Boys and girls at or above the 90^th percentile have roughly the same WC percentile estimates until age 13, and then boys consistently have higher values than girls through age 18, whereas all other age-specific percentile values were about the same for both boys and girls. Overall, girls ages 12 and older that have a WC percentile at or above the 90^th percentile (90.0 cm) and girls ages 16 and older that have a WC at or above the 75^th percentile (90.6 cm) meet the adult WC cutoff criteria (88 cm) for the metabolic syndrome. All boys ages 16 and older that have a WC at or above the 90^th percentile (106.1 cm) meet WC criteria for adult metabolic syndrome (102 cm). The comparison of the quantile regressions between NHANES III ¹⁶ and NHANES 1999–2008 resulted in significant main effects for NHANES groups for all quantiles for both sexes (P<0.001 for all analyses). Odds ratios for comparisons of NHANES groups at each age indicated that the current NHANES were always higher than NHANES III. Significant differences are marked for each age in Table 1.

Table 2 provides ethnic group comparisons by gender for each WC percentile analyzed. Overall, MA boys and girls are more likely to be above all WC percentiles (10^th through 90^th) compared to their NHB and NHW counterparts. Specifically, MA boys were twice as likely versus NHB (OR=2.09, 95% CI 1.73–2.52) and almost twice as likely versus NHW (OR=1.86, 95% CI 1.51–2.29) to have a WC above the 90^th percentile.

Tables 3 –5 provide the ethnic group–specific WC percentile values by age and gender. NHB (89.1 cm) and MA (88.8 cm) girls ages 11 and older and NHW girls (88.7 cm) ages 12 and older who had a WC at or above the 90^th percentile meet adult WC criteria for the metabolic syndrome (88 cm). MA boys (105.1 cm) ages 14 and older and NHB (105.1 cm) and NHW boys (105.0 cm) ages 16 and older that had a WC at or above the 90^th percentile meet adult WC criteria for the metabolic syndrome (102 cm).

Discussion

In this analysis, we present the most currently available (1999–2008) population-based estimates for the 10^th, 25^th, 50^th, 75^th, and 90^th percentiles of WC for U.S. children and adolescents. In general, MA boys and NHB girls had higher WC measurements at increased percentiles and ages. Perhaps most striking is that, regardless of ethnic background, all adolescent girls ages 12 and older and above the 90^th percentile and ages 16 and older and above the 75^th percentile meet the adult WC cutoff criteria as a component of the metabolic syndrome. ³

Given the current childhood obesity epidemic, it is not surprising that our estimates for both boys and girls at all percentiles are larger compared to the published NHANES III estimates of 1988–1994. ¹⁶ This finding is consistent with others who report age-unadjusted (group) mean WC values greatly increased between the NHANES III and 2003–2004 surveys. ²²

Of great concern is the rate at which the upper percentiles (75^th and 90^th among both sexes) of WC increased from the NHANES III estimates to our current percentiles. Specifically, in the NHANES III analysis, Fernandez and colleagues ¹⁶ report that at the 75th percentile, 16- and 17-year-old MA and NHB girls had values exceeding 88 cm, the cutoff point for the metabolic syndrome in women. Similarly, at age 13, both MA and NHB girls in the 90th percentile of the distribution reached the 88-cm cutoff point. Our analysis shows that now, 15-year-old MA and NHB girls and 16-year-old NHW girls have 75^th percentile cutoff values above 88 cm, and the 90^th percentile estimate for all girls ages 11 and older is above this value. Similar patterns are seen in boys, although at older ages than are found in girls.

Likewise, the adult literature shows gender–ethnic group differences in mean WC have become wider since NHANES III, particularly when comparing NHB to NHW women. Specifically, one study based on NHANES analysis reported that NHB women would have central obesity and obesity prevalence of 90.8 and 70.7% by 2020. ²³ Other recent non-NHANES studies have reported no ethnic differences in African-American (AA) versus white men, but WC values are 5 cm higher in AA compared to white women. ²⁴ The authors state a need for ethnic-specific anthropometric thresholds to identify obesity-related health risks in women should be investigated further. These studies are important because if ethnic differences in BMI or WC as risk predictors are detected, then the public health guidelines targeting risk reductions must be made specific for different ethnic groups. Indeed, our findings reported here would support this conclusion in the U.S. pediatric population as well.

Whereas WC is currently one of the components of adult metabolic syndrome, incorporating WC as a routine measure in standard pediatric care is faced with several challenges to validity and reliability. An expert committee of the American Medical Association and the Centers for Disease Control and Prevention Task Force on Assessment, Prevention, and Treatment of Childhood Obesity was unable to recommend the routine clinical use of waist circumference for children because of incomplete information and the lack of specific guidance for its clinical application. ²⁵ In fact, the literature describes several different methods for measuring WC in children. ²⁶ There is no current standardization in WC measurement, and thus it has been suggested to use the iliac crest level because it only requires a single palpation. ²⁷ International organizations recommend the midpoint between the superior border of the iliac crest and the lowest rib. ²⁸ The National Institutes of Health and NHANES guidelines state that WC should be measured directly above the superior border of the iliac crest, ²⁹ and clinical settings often use the umbilicus and the minimal waist. ³⁰ Differences in location of measurement could certainly affect risk profile status. Although WC is easier to obtain than skinfold thickness measurements (but may be as equally unreliable), and can provide indirect information about visceral adiposity, standard reference values are currently unavailable to identify children at risk for cardiometabolic disease beyond the standard measure of BMI. This suggests the need the need for standardization and universal replication utilizing central training and/or nationally endorsed guidelines.

The literature reports a strong correlation between WC and cardiometabolic risk factors in both adults and children. Indeed this relationship is so convincing in the adult literature that WC is included in the adult definition instead of BMI. ¹ This has been implied for children as well. A recent joint council consensus statement from the American Heart Association ¹ states that “given the significant increase in waist circumference among U.S. children and adolescents over the past 2 decades, a marker of abdominal obesity should be considered as an important component of the pediatric metabolic syndrome definition.” This statement is supported by studies that consistently report that WC is a proxy measure of visceral fat that independently predicts insulin resistance, a powerful underlying mechanism of the metabolic syndrome. ³¹ Thus, body fat distribution, and central fat accumulation in particular, appears to be an important determinant of cardiometabolic risk, even in childhood, suggesting that perhaps WC should be included as a measurement in pediatric standard of care.

Certainly those children and adolescents who have a WC above the adult standard (88 cm for women, 102 cm for men) should be recognized by clinicians for a potential future cardiometabolic disease risk above normal. A recent CDC report predicts that the number of new diabetes cases each year will increase from 8 per 1,000 people in 2008 to 15 per 1,000 in 2050. ³² This sharp increase is attributed to the current obesity epidemic, increases in minority groups that are at high risk for type 2 diabetes, and people with diabetes living longer. Our national percentile estimates presented here suggest that this is a significant proportion of many pediatric practices, especially if they are serving a multiethnic community. Finally, the latest national pediatric WC percentile estimates provided here can serve as an important reference for future research that identifies specific cutoff values to predict adult onset cardiovascular disease, an area in much need of further research. Similarly, the national prevalence estimates of other metabolic syndrome components, such as blood pressure ³³ and lipids, have been updated. ³⁴

This analysis was conducted to generate age, gender, and ethnicity-specific percentile values for WC for the most current U.S. pediatric population available for analysis in NHANES. However, it should be noted that these are estimates based on a sample from 1999–2008 and offer a snapshot of the pediatric population for the United States for this time period only, a period of time during which rates of overweight and obesity had increased dramatically from prior intervals. Therefore, caution is advised in considering these estimates as ideal standards for clinicians and researchers to adhere to.

Conclusions

We have provided current WC percentile estimates for U.S. children between 2 and 18 years old by sex and ethnicity. All girls ages 12 and older and all boys ages 16 and older regardless of ethnic background that have a WC above the 90^th percentile meet the adult cutoff criteria for WC as a component of the adult metabolic syndrome. MA boys consistently have higher WC percentile-for-age and sex estimate values than their ethnic group counterparts. WC is a simple, inexpensive measure that can identify children at risk for cardiometabolic disease.