A nationwide US analysis reveals that prediabetes raises death risk only in younger adults, pointing to the urgent need for targeted screening and prevention programs in the prime working years.
Research Letter: Demographics, Lifestyle, Comorbidities, Prediabetes, and Mortality. Image Credit: Neirfy / Shutterstock
In a recent letter published in the JAMA Network Open, researchers at the University at Buffalo evaluated whether demographic, lifestyle, and comorbidity factors modify the association between prediabetes and all-cause mortality in United States (US) adults.
Background
Prediabetes is a quiet warning light that often flashes without symptoms. Prediabetes raises cardiovascular disease risk and ties to higher all-cause mortality in the US, but headlines seldom explain for whom the danger is greatest.
Factors such as age, race and ethnicity, smoking, alcohol use, and chronic conditions complicate the story, and programs must know who needs help first. Clinicians, employers, and families need guidance to target prevention dollars where they matter most.
Further research is required in order to pinpoint which groups bear the highest mortality risk and to clarify how demographic, lifestyle, and comorbidity factors shape that risk.
About the study
The analysis followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidance and used National Center for Health Statistics (NCHS) files linked to the National Death Index (NDI) from the National Health and Nutrition Examination Survey (NHANES).
Adults aged 20 years or older who completed the interview and examination, had a valid mortality linkage, and were in the 2005-2018 cycles were eligible. Prediabetes was confirmed by self-report or hemoglobin A1c (HbA1c) 5.7%-6.4%. Covariates included demographics, lifestyle behaviors, and comorbid conditions.
Race and ethnicity were self-reported and categorized as non-Hispanic White, non-Hispanic Black, or other (which included Mexican American, other Hispanic, Asian, multiracial, or other race not classified as non-Hispanic White or non-Hispanic Black).
Associations with all-cause mortality were estimated using multivariable Cox proportional hazards models in stages: demographics only; demographics plus lifestyle; and demographics, lifestyle, and comorbidities. In the demographics-only model, the HR dropped to 0.88 (95% CI, 0.80–0.98) before increasing slightly with further adjustments.
Effect modification was assessed in strata by age (20-54, 55-74, ≥75 years) and by race and ethnicity. NHANES design features and weights yielded representative estimates.
Analyses were conducted in R (R Foundation for Statistical Computing) 4.4.1, with P values and significance at P < .05. Because this retrospective analysis used deidentified data, Institutional Review Board (IRB) approval was not required under Title 45 Code of Federal Regulations (CFR) §46.102(e).
Study results
Among 38,093 US adults in NHANES, 9,971 (26.2%), representing more than 51 million people, had prediabetes. Most of those with prediabetes were female and aged 20-54 years. Unadjusted models showed higher all-cause mortality for adults with prediabetes compared with those without (hazard ratio [HR], 1.58; 95% confidence interval [CI], 1.43-1.74).
After sequential adjustment for demographics, lifestyle, and comorbidities, the association was no longer significant (fully adjusted HR, 1.05; 95% CI, 0.92-1.19).
Significant interaction terms emerged for age group and for race and ethnicity. In age-stratified models, mortality risk was statistically significant only among adults aged 20-54 years (HR, 1.68; 95% CI, 1.25-2.20). No significant associations were detected within race-and-ethnicity strata (non-Hispanic Black: HR 1.02; non-Hispanic White: HR 1.06; other: HR 0.81).
Weighted mortality was 10.4% among adults with prediabetes versus 7.4% among those without. Adults with prediabetes carried heavier comorbidity burdens: hypertension in 43.4% versus 28.3%; heart disease in 10.0% versus 5.9%; and a history of cancer in 13.2% versus 9.1%. Mean body mass index (BMI) was approximately 29 in both groups.
Smoking patterns were similar across groups, though former smoking was somewhat more common in prediabetes (27.3% vs 23.6%). The prediabetes group also contained a larger share of people in midlife (40.6% aged 55-74 years) and fewer younger adults (46.8% aged 20-54 years) than the normoglycemic group.
These distributions show how demographic and clinical profiles can obscure the true effect of prediabetes on mortality until they are accounted for statistically.
Taken together, the results indicate that background factors, who people are, how they live, and the conditions they already manage, explain much of the apparent link between prediabetes and death at the population level. Yet the elevated HR among younger adults stands out, suggesting that lifestyle behaviors, limited access to health care, life stage challenges, stronger genetic predispositions, or early-life physiologic stress or delayed diagnosis may be especially consequential before age 55.
For individuals, this means that being told “prediabetes” in one’s 20s, 30s, or early 40s should prompt action on diet, physical activity, sleep, and screening rather than watchful waiting.
For communities and employers, the findings support convenient and affordable prevention programs designed for busy working-age adults, for example, flexible, virtual, and peer-led offerings scheduled outside standard clinic hours.
Conclusions
To summarize, in a nationally representative sample, prediabetes alone did not predict higher all-cause mortality after accounting for demographics, lifestyle, and comorbidities, but younger adults (20-54 years) remained at increased risk. For clinicians and health systems, this points to early identification and age-tailored prevention as practical levers: prioritize screening, timely referrals, and programs that fit working-age schedules.
Employers and communities can reinforce these efforts by offering lifestyle support.
The authors note that limitations of the study include its cross-sectional design, reliance on some self-reported data, lack of longitudinal tracking, and inability to infer causality from observational data. For individuals, the message is actionable: a prediabetes label in early adulthood warrants changes in diet, physical activity, sleep, and stress management to reduce risk and protect long-term health.
Journal reference:
- Ekwunife, O., Wang, X., Fraser, R., Campbell, J. A., Walker, R. J., Jacobs, D., & Egede, L. E. (2025). Demographics, Lifestyle, Comorbidities, Prediabetes, and Mortality. JAMA Netw Open. 8(8). DOI:10.1001/jamanetworkopen.2025.26219, https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2837340