We make use of multiple waves of National Income Dynamics Study data, from 2008 to 2015, to
investigate the socio‐economic factors that correlate with the prevalence and onset of diabetes. Our
analysis follows a cohort of 3470 older adults aged forty and above, who are interviewed four times
over a six year period. We use linear probability models and estimate the likelihood of diabetes as a
function of age, race, gender, education, income, exercise, and obesity. Our primary findings are that
age and obesity correlate strongly with diabetes, while income does not have a statistically significant
effect, conditional on the other covariates. Our regression estimates indicate that, of individuals who
reported not being diabetic in Wave 1, those who were obese and morbidly obese were 12.9 and 16.7
percentage points more likely to have experienced the onset of diabetesrespectively, relative to those
with a BMI in the healthy range. In addition, frequent exercise does appear to have a slight protective
effect against the onset of diabetes, and there is some evidence that better educated people have a
lower risk of onset of the disease.