Racial-ethnic disparities in the association between risk factors and diabetes: The Northern Manhattan Study

To identify risk factors (RF) for diabetes within a multiethnic cohort and to examine whether race-ethnicity modified their effects.

Participants in the Northern Manhattan Study without diabetes at baseline were studied from 1993 to 2014 (n=2430). Weibull regression models with interval censoring data were fit to calculate hazard ratios and 95% confidence intervals for incident diabetes. We tested for interactions between RF and race-ethnicity.

During a mean follow-up period of 11years, there were 449 diagnoses of diabetes. Being non-Hispanic black (HR 1.69 95% CI 1.11-2.59) or Hispanic (HR 2.25 95% CI 1.48-3.40) versus non-Hispanic white, and body mass index (BMI; HR 1.34 per SD 95% CI 1.21-1.49) were associated with greater risk of diabetes; high-density lipoprotein cholesterol (HR 0.75 95% CI 0.66-0.86) was protective. There were interactions by race-ethnicity. In stratified models, the effects of BMI, current smoking, and C-reactive protein (CRP) on risk of diabetes differed by race-ethnicity (p for interaction <0.05). The effects were greater among non-Hispanic whites than non-Hispanic blacks and Hispanics.

Although Hispanics and non-Hispanic blacks had a greater risk of diabetes than whites, there were variations by race-ethnicity in the association of BMI, smoking, and CRP with risk of diabetes. Unique approaches should be considered to reduce diabetes as traditional RF may not be as influential in minority populations.