Lifetime risk (LR), a cumulative incidence estimate, is a key metric of disease burden that reflects the impact of baseline prevalence, incidence, and mortality risk. Previous studies in developed countries estimated LR of diabetes (DM) to be ∼35%. Since 2008 both DM incidence and mortality in persons with DM have decreased. The impact of those changes on LR of DM and years of potential life lost (YPLL) to DM is not known. We used data from 558,225 participants aged ≥20 years from the National Health Interview Survey to estimate LRs and YPLLs up to age 80 years by period: 1997-2000, 2001-2005, 2006-2010, and 2011-2015. The age-specific incidence of diagnosed DM and mortality by DM were estimated using Poisson regression. Multistate difference equation accounting for competing risk was used to model LR and YPLL. Age-sex-race adjusted DM incidence (%) for the four periods was 0.6, 0.7, 0.8, and 0.7, respectively (overall P < .001). LR followed the change in incidence and decreased significantly during last period (Figure); YPLL steadily decreased from 12.0 (11.8-12.2) in 1997-2000 to 8.9 (8.7-9.1) in 2011-2015. In 2011-2015, LR (%) and YPLL (years) were 36.6 (95% CI 36.2-37.0) and 9.7 (9.5-9.9) for men, and 36.5 (36.0-36.9) and 8.2 (8.0-8.4) for women, 32.3 (32.0-32.8) and 8.4 (8.2-8.6) for whites, and 46.8 (46.1-47.4) and 10.6 (10.3-11.0) for blacks. In the United States, LR and YPLL have declined, but disparities still persist.


Y.J. Cheng: None. E.W. Gregg: None. R. Brinks: None. S.H. Saydah: None. A.L. Albright: None. G. Imperatore: None.

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. More information is available at