Abstract
Background: While the prevalence of adult obesity and pediatric overweight (OW) is well known, incident obesity/OW is minimally described in the literature. We estimated 1-year incidence rates of adult obesity and pediatric OW among enrollees of a group-model managed care organization. We also examined demographic, clinical, and socioeconomic correlates of incident obesity/OW.
Methods: Enrollee-level data, including enrollment history and demographics, height and weight measurements and chronic disease history, were linked with US Census data based on residential address and geocoded to the census block group. Among enrollees with more than one BMI in a year, the average of BMIs was used for ascertaining adult obesity and the last BMI was used for ascertaining pediatric OW. Separate analyses were conducted for adult (18–74 years) and pediatric (2–17 years) enrollees. Associations of incident obesity/OW with enrollee and area-based residential characteristics were assessed using bivariate statistics.
Results: In 2006, 37.0% of adult enrollees (n=155,887) were obese. Among adult enrollees not obese in 2006 with BMI measured again in 2007 (n=67,200), 5.8% became obese by 2007. Incident obesity was associated with age (6.6% for 30–39 years, 3.4% for ≥70 years; P<0.01) and pre-existing diabetes (7.8%, P<0.01) and hypertension (6.7%, P<0.01). Incident obesity was higher among adults living in areas with higher percentage of black residents (P<0.01), adults with a high school education or less (P<0.01), and lower median household income (P=0.02). In 2006, 33.7% of child and adolescent enrollees (n=29,995) were OW. Among children and adolescents not OW in 2006 with BMI measured again in 2007 (n=10,743), 9.2% became OW by 2007. Incident OW was associated with age (11.4% for 2–9 years, 6.3% for 13–17 years; P<0.01), male gender (9.7%, P=0.04), and pre-existing diabetes (25.0%, P=0.01). Incident OW was higher among children and adolescents living in areas with higher percentage of black residents (P=0.01), adults with a high school education or less (P=0.01), and lower median household income (P<0.01).
Conclusions: Rates of incident obesity among adults and OW among children and adolescents will be compared to similarly calculated rates for metabolic syndrome conditions. Findings inform the need to target lifestyle intervention recruitment to specific subsets of enrollees at risk of becoming obese or OW.
- Received September 11, 2008.
