Persistent Increase of Prevalence of Metabolic Syndrome Among U.S. Adults: NHANES III to NHANES 1999–2006
OBJECTIVE To compare the prevalence in metabolic syndrome (MetSyn) between 1988–1994 and 1999–2006 among U.S. adults of different races or ethnicities.
RESEARCH DESIGN AND METHODS Analysis of data on 6,423 adult men and nonpregnant women aged ≥20 years from Third National Health and Nutrition Examination Survey (NHANES III) and 6,962 participants from the combined NHANES 1999–2006 were done. The revised National Cholesterol Education Program Adult Treatment Panel III definition was used to calculate MetSyn.
RESULTS Both the unadjusted prevalence (27.9 ± 1.1% to 34.1 ± 0.8%, P < 0.001) and age-adjusted prevalence (29.2 ± 1.0% to 34.2 ± 0.7%, P < 0.001) increased from NHANES III to NHANES 1999–2006, respectively. Although MetSyn prevalence was highest in Mexican Americans, significant increases in prevalence occurred among non-Hispanic whites and non-Hispanic blacks, especially among younger women.
CONCLUSIONS The persistent increase of MetSyn among U.S. adults is a serious public health concern because it raises the likelihood of increased prevalence of type 2 diabetes.
The metabolic syndrome (MetSyn) is a constellation of metabolic abnormalities and is associated with increased risk of developing diabetes (1), cardiovascular disease (2), and higher mortality from all causes (3). Among the few studies using nationally representative samples on MetSyn (4–9), Ford et al. (9) estimated an increasing trend of MetSyn prevalence by comparing the Third National Health and Nutrition Examination Survey (NHANES III) and NHANES 1999–2000 data. However, because of the smaller sample size of NHANES 1999–2000, the change in MetSyn prevalence for various subpopulations, which is necessary to track age and ethnicity specific trends, was not estimated. Therefore, the objective of this study is to compare the prevalence of MetSyn between NHANES III and NHANES 1999–2006 among U.S. adults of different races or ethnicities.
RESEARCH DESIGN AND METHODS
We identified the cases of MetSyn using the revised American Heart Association/National Cholesterol Education Program Adult Treatment Panel III definition (10), including medication uses for appropriate MetSyn criteria. Data for this study were obtained from public-use datasets of the NHANES III, NHANES 1988–1994 (data release 11#1A), and four continuous NHANES data releases: 1999–2000, 2001–2002, 2003–2004, and 2005–2006. Details of survey and laboratory procedure of NHANES are published elsewhere (11–13). Data from NHANES 1999–2006 were combined for this study to produce estimates of MetSyn for demographic subpopulations (e.g., sex-age-race/ethnicity) with greater statistical reliability. Because the data on fasting triglycerides and fasting glucose were required to identify MetSyn and those measurements were done on a subsample population, the sample weights for the subsample were used in this study.
The appropriate sample weights for combined NHANES 1999–2006 were constructed using National Center for Health Statistics guidelines (14). To maintain the consistency of blood pressure data between the two surveys, the procedure described by Ford et al. (9) was followed.
The continuous NHANES measured fasting glucose and serum triglycerides from blood samples drawn in the morning; therefore, only participants who attended a morning examination session for NHANES III were included in this analysis. Otherwise, the sample includes men and nonpregnant women aged ≥20 years who fasted for at least 8 h. The number of participants in the final analysis was 6,423 for NHANES III and 6,962 for NHANES 1999–2006. Statistical analyses to calculate prevalence were performed using the survey procedures in SAS software version 9.1 (SAS Institute, Cary, NC). The statistical significance of the change in MetSyn prevalence between the two surveys was examined by Student t test, in which the square root of the sum of the squared standard errors was utilized to calculate the pooled standard error of the difference in the mean.
The age-adjusted prevalence of four of the five metabolic abnormalities of MetSyn increased significantly between the surveys for women: abdominal obesity 46.0 ± 1.4% to 58.0 ± 1.1%, P < 0.001; hypertriglyceridemia 24.7 ± 1.2% to 27.6 ± 0.8%, P = 0.042; high blood pressure (HBP) 27.8 ± 0.9% to 36.6 ± 0.8%, P < 0.001; high fasting glucose 24.2 ± 1.2% to 29.2 ± 1.0%, P = 0.002. However, for men, age-adjusted prevalence significantly increased in abdominal obesity (30.4 ± 1.6% to 41.1 ± 1.1%, P < 0.001) and HBP (32.0 ± 0.8% to 40.0 ± 0.7%, P < 0.001) only. The age-adjusted prevalence of low HDL cholesterol significantly decreased in both sexes (men: 36.4 ± 1.7% to 27.6 ± 1.0%, P < 0.001; women: 39.6 ± 1.4% to 33.8 ± 1.1%, P = 0.001) between the surveys.
Both age-adjusted and age-specific prevalence of MetSyn for NHANES 1999–2006 were significantly higher than for NHANES III (Table 1). The unadjusted (P = 0.012) and age-adjusted (P = 0.046) prevalence increased significantly between the two surveys for men; however, there was no significant change in any of the three age-groups. For women, both unadjusted and age-adjusted (P < 0.001) prevalence increased significantly between the two surveys, with a significant increase noted in all three age-groups. Among non-Hispanic White (NHW) subjects, both men and women showed significant increases in unadjusted (men: P = 0.010; women: P = 0.001) and age-adjusted (men: P = 0.048; women: P = 0.007) prevalence of MetSyn. However, when classified by age-groups, only women aged 20–39 years showed significant increase (P = 0.010). Prevalence of MetSyn did not change significantly among non-Hispanic Black (NHB) men (P > 0.050) between the two surveys, but NHB women aged 20–39 years showed a significant increase in prevalence (P = 0.036). The age-adjusted prevalence of MetSyn in NHANES 1999–2006 was highest among Mexican Americans (men: 36.6 ± 1.9%; women: 42.6 ± 1.7%) with little change in this group from NHANES III. Using the unadjusted prevalence rates from combined sample population of NHANES 1999–2006, we estimated that about 32.4 million men and 35.3 million women in U.S. had MetSyn. Among U.S. adults with MetSyn, ∼50.6 million were NHW, ∼6.3 million were NHB, and ∼4.6 million were Mexican Americans.
The age-adjusted prevalence of U.S. adults reporting diabetes (other than pregnancy related) or having a fasting blood glucose ≥126 mg/dl significantly increased in both sexes (men: 8.1 ± 0.6% to 10.5 ± 0.6%, P = 0.005; women: 5.8 ± 0.6% to 8.5 ± 0.5%, P = 0.001) between the two surveys. The age-adjusted prevalence of MetSyn among U.S. men without diabetes did not change significantly (27.6 ± 1.4% to 30.6 ± 1.1%, P = 0.08); however, the prevalence significantly increased for women without diabetes (24.0 ± 1.2% to 29.4 ± 1.0%, P = 0.001), including women aged 20–39 years (10.0 ± 1.6% to 15.8 ± 1.2%, P = 0.003) and aged 40–59 years (25.8 ± 2.4% to 31.6 ± 1.7%, P = 0.049).
Ford et al. (9) estimated that ∼50 million U.S. adults in 1990 and ∼64 million in 2000 had MetSyn, representing a 28% increase in prevalence. From the combined NHANES 1999–2006 data, we estimated ∼68 million U.S. adults had MetSyn, or a further increase of 6%. The prevalence of MetSyn in U.S. adults in 1999–2006 was 34.1 ± 0.8% (after age adjustment 34.2 ± 0.7%), which is a significant increase from 1988–1994, and more so in women (28.4%) than in men (16.8%). Further, in both NHW and NHB the prevalence of MetSyn significantly increased in women, particularly younger women (aged 20–39 years). The increased prevalence of MetSyn was primarily due to increases in abdominal obesity and HBP.
An increase in MetSyn prevalence is expected to be followed by an increase in diabetes prevalence, though of a lesser magnitude. Between the two surveys, there was a 4.3% increase in age-adjusted prevalence of MetSyn among adults without diabetes and a 2.6% increase in diabetes. As we continue to see an increase in MetSyn, especially in certain ethnic groups and younger women, we will see a concomitant increase in diabetes and its comorbidities and associated medical costs.
No potential conflicts of interest relevant to this article were reported.
A.M. and G.L. contributed equally to study design, data analysis, and manuscript writing.
We thank Paul S. Fisk, MS, North Dakota State University, for his assistance with data analysis. We also thank Subrata K. Roy, PhD, Indian Statistical Institute, Kolkata, India, for his help during the review process.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Received May 7, 2010.
- Accepted September 24, 2010.
- © 2011 by the American Diabetes Association.
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.