OBJECTIVES. Abdominal obesity may be a better predictor than overall obesity for the risk of cardiovascular disease and type 2 diabetes. Waist circumference and waist-height ratio are 2 simple, yet effective, surrogate measures of abdominal obesity. We sought to examine the recent trends in mean waist circumference and waist-height ratio and prevalence of abdominal obesity among children and adolescents aged 2 to 19 years in the United States.
METHODS. Representative samples of the civilian, noninstitutionalized US population from the National Health and Nutrition Examination Survey conducted during 4 time periods, 1988–1994 (ie, National Health and Nutrition Examination Survey III), 1999–2000, 2001–2002, and 2003–2004, were examined to estimate the mean waist circumference and waist-height ratio of boys and girls in 4 different age groups. Data from the 3 most recent National Health and Nutrition Examination Surveys were combined to establish a National Health and Nutrition Examination Survey 1999–2004 category.
RESULTS. Categorized by age group, the unadjusted mean waist circumference for boys increased between National Health and Nutrition Examination Survey III and National Health and Nutrition Examination Survey 1999–2004 from 50.7 cm (aged 2–5 years), 61.9 cm (aged 6–11 years), 76.8 cm (aged 12–17 years), and 81.3 cm (aged 18–19 years) to 51.9, 64.5, 79.8, and 86.6 cm, respectively. During the same time periods and within the same age groups, the unadjusted mean waist circumference for girls increased from 51.0, 61.7, 75.0, and 77.7 cm to 51.8, 64.7, 78.9, and 83.9 cm, respectively. The relative change in waist-height ratio was similar to waist circumference at each age group for both boys and girls. Using the 90th percentile values of waist circumference for gender and age, the prevalence of abdominal obesity increased by 65.4% (from 10.5% to 17.4%) and 69.4% (from 10.5% to 17.8%) for boys and girls, respectively.
CONCLUSIONS. Mean waist circumference and waist-height ratio and the prevalence of abdominal obesity among US children and adolescents greatly increased between 1988–1994 and 1999–2004.
Abdominal obesity, a state of excessive accumulation of both central subcutaneous and visceral fat, has emerged as an important predictor for metabolic complications and adverse health effects; it has been linked to the metabolic syndrome, type 2 diabetes, and cardiovascular disease in both adult men and women1–3 and increased cardiovascular and metabolic risks in children and adolescents.4–6 Furthermore, abdominal obesity is a critical component of the National Cholesterol Education Program Adult Treatment Panel III7 and International Diabetes Federation8 criteria for the definition and diagnosis of metabolic syndrome in adults and has been used to define the metabolic syndrome in adolescents.9
Although visceral fat (body adipose tissue located within the abdominal cavity around the visceral organs) can be accurately assessed by imaging techniques, such as computed tomography and MRI,10 using these techniques to identify people with abdominal obesity in large epidemiological studies, mass screenings, or clinical settings may not be feasible. Waist circumference (WC) and waist-height ratio (WHtR) are simple, yet effective, ways of measuring abdominal obesity in adults11 and children12 and may be better predictors of cardiovascular disease risk than body mass index (BMI) in adults13 and children.14 In particular, WC is a better indicator of visceral fat than BMI in children.15
The prevalence of overall obesity as measured by BMI has increased dramatically among children and adolescents in the United States during the past 2 decades16,17; however, little is known about the secular trends in abdominal obesity in the United States. Data from the National Health and Nutrition Examination Survey (NHANES) III and from the years 1999 to 2000 show an increase in mean WC among children in the United States.18 Other countries, such as the United Kingdom,19–21 Australia,22 and Spain,23 have reported a significant increase in WC in children and adolescents. Thus, this report examines the most recent national data and trends for WC and WHtR, 2 important surrogate measures for abdominal obesity, among children and adolescents in the United States.
Study Population and Design
This study examined data from the NHANES conducted at the following 4 time periods: 1988 to 1994 (ie, NHANES III),24 1999 to 2000,25 2001 to 2002,26 and 2003 to 2004.27 For all 4 of the NHANES studies, the participants were selected using a multistage, stratified sampling design to be representative of the civilian, noninstitutionalized US population. After being interviewed in their homes, participants were invited to be examined in a mobile examination center.
Body measurements of survey participants were taken using the Centers for Disease Control and Prevention standardized methods and equipment throughout the NHANES surveys to ensure comparability of anthropometric measures over time.24–27 Participants were measured in private at the mobile examination center by 2 health technicians (an examiner and a recorder). The examiner positioned the participant, took each measurement, and read the measurement aloud to the recorder, who repeated the number and entered it into the automated system or hard copy form. The automated system was designed to function as a quality control measure by minimizing possible measuring and recording errors. To control or minimize the common error in anthropometrics because of body positioning or in reading and recording the measurements, the examiner and the recorder were trained using the Centers for Disease Control and Prevention standard procedures for obtaining measurements. The recorder assisted the examiner with positioning of the participants and the reading process to minimize the errors.
WC was measured using a steel measuring tape to the nearest 0.1 cm at the high point of the iliac crest at minimal respiration when the participant was in a standing position.24–27 The examiner stood behind the participant, palpated the hip area for the right iliac crest, marked a horizontal line at the high point of the iliac crest, and crossed the line to indicate the midaxillary line of the body (Fig 1). The examiner then stood on the participant's right side and placed the measuring tape around the trunk in a horizontal flat surface at the level marked on the right side of the trunk. The recorder observed the participant to ensure that the tape was parallel to the floor and that the tape was snug but did not compress the skin.
Weight was measured on a Toledo self-zeroing weight scale (Toledo Scale, Inc, Columbus, OH), and height was measured with a stadiometer to the nearest millimeter. WHtR was calculated as the ratio of waist (cm) and height (cm). Other variables included in our analyses were gender, age, and race/ethnicity (ie, non-Hispanic white, non-Hispanic black, Mexican American, and other).
This study estimated the means of WC and WHtR and prevalence of abdominal obesity of boys and girls aged 2 to 19 years who had participated in NHANES during the following 4 time periods: 1988–1994, 1999–2000, 2001–2002, and 2003–2004. Data from the 3 most recent NHANES were also combined to establish an NHANES 1999–2004 category. The 90th percentile values of WC for gender and age generated in NHANES III were used as cutoff values to identify subjects with abdominal obesity in each of the survey periods.9 A WHtR cutoff of 0.5 was used to define abdominal obesity for both 6- to 19-year-old boys and girls.28 This cutoff value was not applied to very young children (aged 2–5 years) in our study, because it may overestimate the number considered at risk.28 Linear trends in the mean of WC and WHtR and in the prevalence of abdominal obesity across the 4 time periods were tested using linear regression and logistic regression models by gender and 4 age categories, respectively. The total differences in estimates between the NHANES III (1988–1994) and NHANES 1999–2004 surveys were tested by using the pooled t test (t statistic = difference in the means or prevalence divided by pooled SE of the 2 estimates).29 All of the analyses were performed by using SUDAAN 9.0 software (Research Triangle Institute, Research Triangle Park, NC) to account for the complex sampling design.
Between NHANES III (1988–1994) and NHANES 1999–2004, the overall unadjusted mean WC increased by 3.7 cm (1.4 in) among both boys and girls. The largest increase in mean WC occurred among 18- to 19-year-old boys (5.3 cm or 2.1 in) and 18- to 19-year-old girls (6.2 cm or 2.4 in). Increases in mean WC in other age groups were 1.2 cm (2–5 years), 2.6 cm (6–11 years), and 3.1 cm (12–17 years) for boys and 0.8 cm (2–5 years), 3.1 cm (6–11 years), and 3.9 cm (12–17 years) for girls, respectively.
A steady increase in the means of WC occurred across the 4 time periods overall for both boys and girls (P < 0.0001 for linear trends) and among all of the gender- and age-specific groups (all Ps < 0.05; Table 1). Increasing age was associated with a larger increase in both absolute differences and relative changes for both sexes. The largest relative changes in WC occurred among the 18- to 19-year-old boys (6.6%; P < .0001) and girls (8.0%; P < .0001). Similar to WC, the WHtR also increased throughout the 4 time periods (all Ps < 0.01 for trends) except for 2- to 5-year-old girls. The relative changes in WHtR increased over age, with the largest relative change occurring among 18- to 19-year-old boys (6.4%; P < .001) and girls (8.1%; P < .0001).
The cutoff values of WC for ≥90th percentile by gender and age in NHANES III (1988–1994) were calculated (Appendix) and applied in the 3 most recent NHANES surveys and the combined NHANES 1999–2004 to estimate the prevalence of abdominal obesity (Table 2). Overall relative increase in the prevalence of abdominal obesity was 65.4% (P < .0001) for boys and 69.4% (P < .0001) for girls. A linear increasing trend in the prevalence of abdominal obesity occurred overall for boys and girls (P < 0.0001 for trends). The relative increase in the prevalence of abdominal obesity was higher among older children and adolescents. Between NHANES III (1988–1994) and NHANES 1999–2004, the largest relative increase in the prevalence of abdominal obesity occurred among 2- to 5-year-old boys (84.0%; P < .0001) and 18- to 19-year-old girls (126.2%; P = .0001).
Using the WHtR value of 0.5 as a cutoff point, the prevalence of abdominal obesity was higher than that using the ≥90th percentile in all of the age groups for both sexes (Table 2). Among boys, the prevalence of abdominal obesity was ∼20% in each age group in NHANES III (1988–1994). However, the prevalence among 18- to 19-year-old boys became larger than other age groups in the 3 recent surveys. Among girls, the prevalence of abdominal obesity was higher among older adolescents than younger children. This pattern persisted in the 3 recent surveys.
The relative increase in WC in each age group across different gender and racial/ethnic subpopulations is shown in Fig 2. Among non-Hispanic white boys, the largest relative increase in WC occurred in the 18- to 19-year-olds, whereas among non-Hispanic black and Mexican American boys, the largest relative increase in WC occurred in adolescents aged 12 to 17 years. However, among non-Hispanic white, non-Hispanic black, and Mexican American girls, the largest relative increase occurred in the 18- to 19-year-olds. The WHtR was higher among 2- to 5-year-olds and 18- to 19-year-olds across all of the gender- and race/ethnicity-specific subgroups (Fig 3). However, the relative increase of WHtR was similar and consistent with that of WC across gender, race/ethnic, and age subgroups.
In contrast to the means of WC or WHtR, the large relative changes in the prevalence of abdominal obesity seemed to occur at slightly different age groups (Fig 4). Among non-Hispanic white and Mexican American boys, the large relative increase in the prevalence occurred among 2- to 5-year-olds and 12- to 17-year-olds, whereas among non-Hispanic black boys, the large relative increase occurred among 12- to 17-year-olds and 18- to 19-year-olds. Among non-Hispanic white and non-Hispanic black girls, the large relative increase occurred among 6- to 11-year-olds and 18- to 19-year-olds, whereas among Mexican American girls, the large relative increase occurred among 12- to 17-year-olds and 18- to 19-year-olds. The patterns of the relative changes in the prevalence of abdominal obesity as defined by the 90th percentile values of WC were similar across gender, race/ethnicity, and age specific subgroups.
In NHANES III, the correlation coefficient between WC and WHtR was 0.47 (P < .0001), 0.82 (P < .0001), 0.93 (P < .0001), and 0.93 (P < .0001) among children and adolscents aged 2 to 5, 6 to 11, 12 to 17, and 18 to 19 years, respectively. The agreement between the 2 approaches of defining abdominal obesity (cutoff values of WC for 90th percentile versus WHtR cutoff value of 0.5) was poor for 2- to 5-year-olds (κ = 0.23), but good for 6- to 11-year-olds (κ= 0.66), 12 to 17-year-olds (κ= 0.61), and 18 to 19-year-olds (κ= 0.56). The correlation coefficients and the κ statistics were similar across gender and over time.
The data from the 2 most recent NHANES surveys, 2001–2002 and 2003–2004, provide updated evidence of a continuing increase in the mean WC and WHtR and the prevalence of abdominal obesity among US children and adolescents. As the epidemic of overall obesity measured by BMI increased, the WC and WHtR, 2 surrogate measures of abdominal obesity, correspondingly increased. Our results are of particular concern, because visceral adiposity, measured by WC and WHtR, increases the risk for obesity-associated morbidity and mortality in children and adults.1–5 Increases in energy intake,30 portion sizes,31 sweetened beverage intake,32 and tobacco use33 and declines in sufficient vigorous physical activity,34 despite declines in sedentary behaviors (eg, hours of television viewing),34 likely contributed to the secular trends in mean WC and WHtR. In addition, growing evidence indicates a potentially causal relationship between stress and abdominal obesity.35 However, it is unknown whether chronic stress activated by psychosocial and socioeconomic handicaps, depression, anxiety, smoking, and alcohol may also contribute to the increasing trends of WC and WHtR among children and adolescents.
The increase in the mean WC and WHtR and the prevalence of abdominal obesity occurred consistently across different gender, racial/ethnicity, and age subgroups. A particularly interesting finding was that the relative increase was larger among older children and adolescents. The largest increases in mean WC and WHtR occurring among youth aged 18 to 19 years are in agreement with data from the Behavioral Risk Factor Surveillance System, which showed that between 1991 and 1998, the greatest increase in the prevalence of obesity was among people aged 18 to 29 years.36 These trends suggest that young adults should receive high priority for intervention efforts to reduce obesity, particularly abdominal obesity.
Data on secular trends in WC and WHtR are scarce, particularly in children and adolescents. In British children aged 2 to 5 years, there was an average increase of 1.05 cm (0.11 cm/year) for boys and 1.81 cm (0.20 cm/year) for girls between 1987 and 1995–1998.19 In British youth aged 11 to 16 years, the mean WC increased by 6.9 cm (0.35 cm/year) for boys from 1977 to 1997 and 6.2 cm (0.62 cm/year) for girls from 1987 to 1997.20 In Australian children aged 7 to 8 years to 12 to 13 years, the WC z score increased by 0.74.22 Our results indicate a similar or relatively smaller change among US boys (2–5 years: 0.12 cm/year; 12–17 years: 0.31 cm/year) and girls (2–5 years: 0.08 cm/year; 12–17 years: 0.39 cm/year) than British children. Substantially increasing trends in WC were also reported in Spanish and British adolescents.21,23
Using the cutoff values of WC for the 90th percentile for gender and age in NHANES III, the prevalence of abdominal obesity during the 4 time periods was comparable to the prevalence of overall overweight as defined by BMI in children and adolescents in the United States.16,17 Thus far, there is no consensus on the cutoff values of WC for identifying children with abdominal obesity. WC percentile reference data based on large population surveys have been generated in children and adolescents in Australia,37 Canada,38 Italy,39 the Netherlands,40 Spain,41 the United Kingdom,42 and the United States.43,44 Several cutoff values for WC in children have been suggested, such as age- and gender-specific 75th percentile to define moderate waist value,41,43 and 90th percentile6,9,43 or 95th percentile38,41 to define high waist value or abdominal obesity. In our study, we chose the cutoff values of WC for 90th percentile to define abdominal obesity, because children with a WC ≥90th percentile were more likely to have clustering of multiple cardiovascular risk factors than those with a WC <90th percentile.6,9 Because of lack of a standardized technique for measuring WC in children, caution needs to be taken when comparing WC percentile reference data between studies. A possible limitation of these data is the lack of a gold standard research method, such as cross-sectional computed tomography or MRI scans to compare the WC measures in these subjects; no such measure was collected for the previous or current NHANES studies.
To the best of our knowledge, our study is the first report on the recent trends in WHtR among US children and adolescents. Researchers proposed recently that WHtR might be a better predictor of risk for cardiovascular disease than BMI or WC for the following reasons: (1) WHtR is more highly correlated with visceral fat mass45 and clustering of cardiovascular risk factors in children14 and adults46; (2) WHtR may be a more accurate tracking indicator of fat distribution and accumulation by age, because it accounts for the growth in both WC and height over age, particularly in children and adolescents; and (3) the value of WHtR is free of measurement units and is in a close agreement between males and females at each age group. As evidenced in our study, the values of WHtR seem to be closer between boys and girls and between age groups than the WC values. The consistent patterns in increasing mean WHtR and WC over time suggest that WHtR is as sensitive as WC in monitoring changes. Thus, WHtR may be a potentially useful surrogate measure for abdominal obesity across different age, gender, or racial/ethnic subpopulations.
A common cutoff value (eg, ≥0.5) for adults46 might be used to identify people with abdominal obesity. However, using the WHtR cutoff value of 0.5 in children and adolescents28 may generate a higher prevalence of abdominal obesity than that using ≥90th percentile for WC by gender and age as evidenced in our study. In particular, among very young children (aged 2–5 years), the correlation between WC and WHtR was low, the agreement of classification for abdominal obesity using the 2 approaches was poor, and the prevalence of abdominal obesity using this criterion was approaching 50%. Taken together, using the WHtR cutoff value of 0.5 to define abdominal obesity in very young children (aged 2–5 years) may not be valid. Therefore, future research is warranted to determine the appropriate cutoff or threshold values using standardized WC or WHtR reference data for identifying children at risk of abdominal obesity in relation to clustering of metabolic and cardiovascular risk factors or obese-related diseases, such as type 2 diabetes.
These findings provide updated and useful information that further strengthens our understanding of the obesity epidemic among children and adolescents in the United States and sheds light on the importance of using WC or WHtR in monitoring the trends of abdominal obesity in public health and clinical practice. Previous studies have shown that increased mortality risk related to excess body fat is mainly because of abdominal adiposity. This finding raises concerns about how the increases in the mean WC and WHtR that we demonstrated may affect future adverse health outcomes among children during childhood years and later as adults in the United States. Because relatively few health care professionals routinely measure WC, efforts are needed to promote the measurement of this important anthropometric parameter as a “vital sign” in pediatric primary care practice.
- Accepted June 5, 2006.
- Address correspondence to Chaoyang Li, MD, PhD, Centers for Disease Control and Prevention, 4770 Buford Highway, MS K66, Atlanta, GA 30341. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
Dr Li had full access to all of the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.
- ↵Bjorntorp P. Abdominal fat distribution and the metabolic syndrome. J Cardiovasc Pharmacol.1992;20(suppl 8) :S26– S28
- Carey VJ, Walters EE, Colditz GA, et al. Body fat distribution and risk of non-insulin-dependent diabetes mellitus in women. The Nurses' Health Study. Am J Epidemiol.1997;145 :614– 619
- ↵Despres JP, Moorjani S, Lupien PJ, Tremblay A, Nadeau A, Bouchard C. Regional distribution of body fat, plasma lipoproteins, and cardiovascular disease. Arteriosclerosis.1990;10 :497– 511
- ↵Esmaillzadeh A, Mirmiran P, Azizi F. Clustering of metabolic abnormalities in adolescents with the hypertriglyceridemic waist phenotype. Am J Clin Nutr.2006;83 :36– 46
- ↵International Diabetes Federation. The IDF consensus worldwide definition of the metabolic syndrome. Available at: www.idf.org/webdata/docs/MetSyndrome_FINAL.pdf. Accessed April 3, 2006
- ↵Taylor RW, Jones IE, Williams SM, Goulding A. Evaluation of waist circumference, waist-to-hip ratio, and the conicity index as screening tools for high trunk fat mass, as measured by dual-energy X-ray absorptiometry, in children aged 3–19 y. Am J Clin Nutr.2000;72 :490– 495
- ↵Zhu S, Wang Z, Heshka S, Heo M, Faith MS, Heymsfield SB. Waist circumference and obesity-associated risk factors among whites in the third National Health and Nutrition Examination Survey: clinical action thresholds. Am J Clin Nutr.2002;76 :743– 749
- ↵Ford ES, Mokdad AH, Ajani UA. Trends in risk factors for cardiovascular disease among children and adolescents in the United States. Pediatrics.2004;114 :1534– 1544
- ↵McCarthy HD, Ellis SM, Cole TJ. Central overweight and obesity in British youth aged 11–16 years: cross sectional surveys of waist circumference. BMJ.2003;326 :624
- ↵Rudolf MC, Greenwood DC, Cole TJ, et al. Rising obesity and expanding waistlines in schoolchildren: a cohort study. Arch Dis Child.2004;89 :235– 237
- ↵Moreno LA, Sarria A, Fleta J, Marcos A, Bueno M. Secular trends in waist circumference in Spanish adolescents, 1995 to 2002. Arch Dis Child.2005;90 :818– 819
- ↵Centers for Disease Control and Prevention. The Third National Health and Nutrition Examination Survey (NHANES III 1988–94) Reference Manuals and Reports [CD-ROM]. Bethesda, MD: National Center for Health Statistics; 2005
- ↵Centers for Disease Control and Prevention. NHANES 1999–2000 public data release file documentation. Available at: www.cdc.gov/nchs/about/major/nhanes/nhanes99-00.htm. Accessed December 15, 2005
- ↵Centers for Disease Control and Prevention. NHANES 2001–2002 public data release file documentation. Available at: www.cdc.gov/nchs/about/major/nhanes/nhanes01-02.htm. Accessed December 15, 2005
- ↵Centers for Disease Control and Prevention. NHANES 2003–2004 public data release file documentation. Available at: www.cdc.gov/nchs/about/major/nhanes/nhanes2003-2004/nhanes03_04.htm. Accessed December 15, 2005
- ↵Glantz SA. Primer of Biostatistics. 6th ed. New York, NY: McGraw-Hill Medical; 2005
- ↵Weitzman M, Cook S, Auinger P, et al. Tobacco smoke exposure is associated with the metabolic syndrome in adolescents. Circulation.2005;112 :862– 869
- ↵Centers for Disease Control and Prevention. Youth risk behavior surveillance: United States, 2003. MMWR CDC Surveill Summ.2004;53 :1– 96
- ↵Ashwell M, Cole TJ, Dixon AK. Ratio of waist circumference to height is strong predictor of intra-abdominal fat. BMJ.1996;313 :559– 560
- Copyright © 2006 by the American Academy of Pediatrics