Abstract Body: Introduction
Higher coffee consumption has been consistently associated with lower type 2 diabetes (T2D) risk but the underlying mechanisms are largely unknown. Our prior study showed higher coffee intake was associated with improved post-load glucose clarence and insulin sensitivity, and lower visceral adiposity.

Hypothesis
We hypothesized that coffee consumption was associated with decreased T2D risk, by lowering genetic susceptibility of T2D in pathophysiological pathways related to insulin sensitivity, insulin production, and central adiposity.

Methods
We analyzed data from 333,053 UK Biobank participants who were free of baseline chronic diseases. Coffee intake was assessed by a dietary questionnaire at baseline. We calculated seven partitioned polygenic scores (pPS) denoting 7 distinct physiological mechanisms underlying hyperinsulinemia, ranging from proinsulin, insulin secretion, to visceral adiposity. We examined the association of coffee intake, the pPS, and their interactions, with incident T2D, using Cox regression model.

Results
We identified 10,707 incident T2D cases during 13.3 years of follow-up. Higher coffee intake, as expected, was associated with lower risk of T2D; compared to participants who did not regularly drink coffee, those consuming 4-5 cups/day had a 21% lower risk of incident T2D (95% confidence interval, 0.74-0.85) after adjusting for known risk factors including BMI. Of all pPS, 5 significantly predicted T2D risk (P<0.05/7) but 2 pPS related to proinsulin and insulin secretion, did not (P>0.05). In multivariable adjusted analyses, we identified a significant multiplicative interaction between coffee intake and proinsulin pPS; this pPS, which was in a genetic cluster linked to lower visceral fat and higher post-load insulin response (Fig. A), was associated with lower T2D risk only among participants who were high coffee consumers (P interaction = 0.004; Fig. B). Furthermore, the positive associations between pPS for adiposity and pPS for visceral adiposity with T2D risk were attenuated among participants who consumed coffee daily vs. who did not (P interaction = 0.088 and 0.093, respectively).

Conclusions
Our data, coupled with our prior study, suggest that coffee consumption may be associated with lower T2D risk through its interactions with genetic pathways related to insulin secretion, post-load insulin sensitivity, and visceral adiposity. These results may inform tailored dietary recommendations to enhance T2D prevention.