Liu · Mayo Clinic proceedings 2022 · Repeated cross-sectional study · n=19,941

Trends in Metabolic Phenotypes According to Body Mass Index Among US Adults, 1999-2018.

Cited 22 times in the scientific literature.

Level 4 - case-series / case-control

Repeated cross-sectional survey analysis of nationally representative data

PubMed 35691704 · doi:10.1016/j.mayocp.2022.02.013 · record verified 2026-08-29

What was done

The authors examined temporal trends and sociodemographic differences in metabolic phenotypes among 19,941 noninstitutionalized US adults aged 20 and older (mean age 46.9 years; 50.6% female) using NHANES data from 1999-2000 through 2017-2018. Phenotypes were classified by combining BMI categories (underweight, normal weight, overweight, obese) with metabolic health status. Based on the 2009 joint metabolic syndrome statement, metabolic health was defined in the primary analysis as having 2 or fewer components (and 0 components in secondary analysis) among: elevated waist circumference, fasting glucose ≥100 mg/dL, blood pressure ≥130/85 mm Hg, triglycerides ≥150 mg/dL, and low HDL cholesterol.

What was found

Between 1999 and 2018: - Metabolically healthy normal weight declined from 33.2% to 25.1% (difference: -8.09%; 95% CI, -12.5% to -3.70%; P < .001). - Metabolically healthy obesity (MH-OB) increased from 9.92% to 14.1% (difference: 4.17%; 95% CI, 1.13% to 7.21%; P < .001). - Metabolically unhealthy obesity (MU-OB) increased from 19.0% to 26.4% (difference: 7.41%; 95% CI, 2.67% to 12.2%; P < .001). - Metabolically unhealthy normal weight decreased from 3.77% to 2.10% (difference: -1.67%; 95% CI, -3.22% to -0.12%; P = .006). - Prevalence remained stable for metabolically healthy underweight (overall 1.62%), metabolically healthy overweight (overall 22.2%), metabolically unhealthy underweight (overall 0.06%), and metabolically unhealthy overweight (overall 11.2%). - MH-OB prevalence rose among non-Hispanic Whites (7.58% to 12.0%, P < .001) and Hispanics (12.2% to 18.4%, P = .01), but remained stable in non-Hispanic Blacks (22.6%, P = .62).

Why it matters

This study documents a broad deterioration in the cardiometabolic profile of the US population over two decades, driven by a shrinking proportion of metabolically healthy normal-weight individuals and a rising burden of obesity across all demographic groups.

Limits

The repeated cross-sectional design cannot evaluate individual transitions between metabolic phenotypes over time or link phenotypes to longitudinal clinical endpoints. Using threshold-based cutoffs for metabolic health (≤2 components) may overlook continuous shifts in cardiovascular risk markers. Institutionalized adults were not included.

Cited by