We used data from seven 2-year cycles of the National Health and Nutrition Examination Survey (NHANES) starting with 1999-2000 and concluding with 2011-2012.⁴ NHANES is a national health survey of the civilian non institutionalized US population in which a sample is selected by using a complex, multistage, probability sampling design. The examination response rates across the survey cycles ranged from 69.5% to 79.6%. The surveys received institutional review board approval and participants provided written informed consent.

In the mobile examination center, the waist circumference of participants was measured just above the iliac crest to thenearest1 mm using a steel measuring tape. Abdominal obesity was defined as a waist circumference greater than 102 cm in men and greater than 88 cm in women.²

Because the unadjusted and age-adjusted results were similar, we report only the latter. Tests for linear trend were conducted using orthogonal polynomial coefficients. Analyses, which take into account the complex sampling design of the surveys, were conducted using SAS version 9.3 (SAS Institute Inc) and SUDAAN version 11.0.0 (Research Triangle Institute) software. Two-sided P < .05 was considered statistically significant.

Data from 32 816 men and nonpregnant women aged 20 years or older were analyzed. The overall age-adjusted mean waist circumference increased progressively and significantly from 95.5 cm (95% CI, 94.2-96.8 cm) in 1999-2000 to 98.5 cm (95% CI, 97.5-99.4 cm) in 2011-2012 (Table 1). Significant increases occurred in men, women, non-Hispanic whites, non-Hispanic blacks, and Mexican Americans.

Particularly large increases between the first and last surveys were observed in non-Hispanic white women aged 40 to 49 years (6.6 cm), non-Hispanic black men aged 30 to 39 years (8.1 cm), Mexican American men aged 20 to 29 years (8.7 cm), Mexican American women aged 70 years or older (11.2cm), and non-Hispanic black women aged 30 to 39 years (11.6 cm).When limited to the period from 2003 to 2012, during which levels of general obesity plateaued, significant increases in mean waist circumference were still noted among all adults (P = .02 for linear trend), women (P ≤ .01 for linear trend), non-Hispanic blacks (P = .02 for linear trend), and Mexican Americans (P = .01 for linear trend).

The overall age-adjusted prevalence of abdominal obesity increased significantly from 46.4% (95% CI, 42.1%-50.8%) in 1999-2000 to 54.2% (95% CI, 51.3%-57.0%) in 2011-2012 (Table 2). Significant increases were present in men, women, non-Hispanic whites, non-Hispanic blacks, and Mexican Americans.

Previous analyses of data from NHANES show that the prevalence of obesity calculated from body mass index (BMI) did not change significantly from 2003-2004 to 2011-2012.⁵ Positive developments in energy expenditure and intake have given hope that the decades-long increase in the prevalence of obesity in the United States may have crested.

In contrast, our analyses using data from the same surveys indicate that the prevalence of abdominal obesity is still increasing. The reasons for increases in waist circumference in excess of what would be expected from changes in BMI remain speculative, but several factors, including sleep deprivation, endocrine disruptors, and certain medications, have been proposed as potential explanations.⁶

Limitations to this analysis include the absence of data for Asians prior to 2011-2012, specific Asian populations, and specific Hispanic populations. Because of the nature of the sampling design, institutionalized adults were not included in the surveys.

At a time when the prevalence of obesity may have reached a plateau, the waistlines of US adults continue to expand. Our results support the routine measurement of waist circumference in clinical care consistent with current recommendations as a key step in initiating the prevention, control, and management of obesity among patients.²