Abstract

Background:

PNPLA3 rs738409 C>G is strongly associated with NAFLD and its severity. PNPLA3 expression is highly regulated by changes in energy balance; however, little is known about the interplay between PNPLA3, dietary factors, alcohol intake, and risk of liver-related death (LRD). We examined the effect of interactions between PNPLA3 rs738409 and dietary factors (macronutrients/coffee-tea intake) and alcohol consumption on the risk of LRD in a population-based US cohort.

Methods:

14,797 adults enrolled in NHANES III (1991–1994) were investigated. Participants with lack of PNPLA3 rs738409, dietary nutrients, and heavy alcohol intake (>2 or >3 drinks per day in women and men) were excluded. The remaining 4361 participants (study cohort) were linked to the National Death Index mortality data through December 2019. LRD was the study outcome. Associations between PNPLA3, dietary nutrients (% daily calories from fat, protein, carb, monounsaturated fatty acids [MUFAs], polyunsaturated fatty acids [PUFAs] and saturated fats, fiber, and cholesterol), non-heavy alcohol intake, and LRD were examined using weighted multivariable competing risk regression models, adjusted for age, sex, race/ethnicity, calorie intake, BMI, and physical activity. Multiplicative interaction terms between PNPLA3 and variables of interest were included in all models.

Results:

The PNPLA3 CC, CG, and GG genotypes frequencies were 55%, 36%, and 9%. There were 28 LRD (0.6%) over a mean 23.1 years follow-up. PNPLA3 G allele was associated with LRD (sHR: 2.8, 95% CI: 1.4-5.6). Non-heavy alcohol intake (sHR: 2.2, 95% CI: 1.07-4.5), % MUFAs (sHR: 0.94, 95% CI: 0.88-0.99), cholesterol intake (mg) (sHR: 1.001, 95% CI: 1.0004-1.002), and coffee-tea intake ≥3 cups/day (sHR: 0.38, 95% CI: 0.04-0.99) were also associated with risk of LRD.

Compared to wild-type PNPLA3 and non-drinkers, G allele carriage significantly increased the risk of LRD among drinkers (sHR: 3.5, 95% CI: 1.9-6.4) vs non-drinkers (sHR: 1.9, 95% CI: 0.7-5.1). Compared to wild-type PNPLA3, a higher risk of LRD was observed among G-allele carriers and higher consumption (top quartile) of saturated fat (sHR: 3.5, 95% CI: 1.2-9.7), and cholesterol (sHR: 3.0, 95% CI: 1.7-6.2). However, compared to wild-type PNPLA3, effects of the G-allele on the risk of LRD were significantly attenuated in those with higher consumption (top quartile) of coffee/tea (sHR: 0.06, 95% CI: 0.007-0.55) and MUFAs (sHR: 0.5, 95% CI: 0.1-0.9). The interactions between other macronutrients and PNPLA3 variation did not significantly impact the risk of LRD (Table 1).

Conclusion:

Non-heavy alcohol consumption and high intake of saturated fat and cholesterol adversely whereas higher caffeine/tea drinking favorably modify the risk of liver-related death associated with PNPLA3 rs738409. If supported by other studies, these observations will assist in the personalized management of NAFLD.