They used new statistical methods derived from the field of econometrics to control for several factors that could provide alternate explanations for the relationship between sugar intake and diabetes, including obesity, overweight, sedentary behavior, other calorie sources, and a long list of socioeconomic measures.