Projected U.S. State-Level Prevalence of Adult Obesity and Severe Obesity
Level 3 - non-randomized controlled study
Statistical modeling and trend projection study using repeated cross-sectional surveys by design analogy
OpenAlex W2995420665 · doi:10.1056/nejmsa1909301
What was done
Researchers developed a multinomial regression modeling approach to project state-level and demographic subgroup-specific adult body-mass index (BMI) trends in the United States from 1990 through 2030. They analyzed self-reported BMI data from 6,264,226 adults in the Behavioral Risk Factor Surveillance System (BRFSS; 1993-1994 and 1999-2016) and corrected for quantile-specific self-reporting bias using physical measurements from 57,131 adults in the National Health and Nutrition Examination Survey (NHANES). Outcomes were categorized into underweight or normal weight (BMI <25), overweight (25 to <30), moderate obesity (30 to <35), and severe obesity (≥35). Model predictive validity was assessed using 1990-2010 data to forecast 2016 observed outcomes.
What was found
By 2030, 48.9% (95% CI, 47.7 to 50.1) of U.S. adults are projected to have obesity (BMI ≥30), exceeding 50% in 29 states and remaining above 35% in all states. Severe obesity (BMI ≥35) is projected to reach 24.2% (95% CI, 22.9 to 25.5) nationally and exceed 25% in 25 states. Nationally, severe obesity is projected to become the most prevalent BMI category among women (27.6%; 95% CI, 26.1 to 29.2), non-Hispanic black adults (31.7%; 95% CI, 29.9 to 33.4), and low-income adults (31.7%; 95% CI, 30.2 to 33.2).
Why it matters
This study provides state-level and demographic-specific forecasts showing that nearly half of U.S. adults will have obesity and a quarter will have severe obesity by 2030, highlighting severe disparities among vulnerable populations. These granular projections supply essential benchmarks for state and federal healthcare resource planning and targeted prevention policies.
Limits
Projections assume past secular trends will continue linearly without accounting for future structural, pharmacological, or policy interventions. The underlying data rely on self-reports from BRFSS, which, although statistically adjusted using NHANES measurements, remain subject to residual measurement and sampling errors. The abstract reports no longitudinal individual-level tracking.
Cited by
- contradicts The average American adult is obese, divorced, and has less than $1,000 in savings.