University of Michigan Ann Arbor, MI, United States
Vincent L. Chen, MD, MS, Xiaomeng Du, MS, Yanhua Chen, PhD, Annapurna Kuppa, MS, Brian Halligan, PhD, Elizabeth Speliotes, MD, PhD, MPH University of Michigan, Ann Arbor, MI
Introduction: Unhealthy dietary patterns are linked to increased risk of non-alcoholic fatty liver disease (NAFLD), as are variants in genes including PNPLA3 and TM6SF2. However, whether dietary patterns exacerbate or ameliorate genetic risk for NAFLD is not well-established. We aimed to identify interactions between diet and genetic risk factors for NAFLD.
Methods: We utilized UK Biobank. Hepatic steatosis was quantified by liver fat percentage (LFP) by proton density fat fraction and dietary patterns using 24-hour recall surveys. Dietary patterns of red meat, fruit, vegetable, whole grain, and dairy were defined as above vs. below US 2015-2020 guidelines; we also calculated the Modified Mediterranean Diet Score (MMED) in which a higher score indicates a more Mediterranean-style diet. Genetic predictors were PNPLA3-rs738409-G and TM6SF2-rs58542926-T, two alleles well-established to associate with NAFLD risk. We excluded participants with excess alcohol intake or missing LFP, dietary, or genetic data. We first conducted linear regression with dependent variable of LFP and independent variables of dietary patterns (high vs. low intake of dietary components, or MMED) or genetic predictor (PNPLA3, TM6SF2). We then created interaction models which each included a dietary variable, genetic variable, and interaction term of dietary*genetic variables.
Results: We included 28125 participants (52% female, median age 66 years). In non-interaction models, high red meat intake associated with increased LFP while higher MMED and high vegetable, fruit, or whole grain intake associated with lower LFP; LFP effect sizes ranged from 0.26-0.57%. PNPLA3 and TM6SF2 risk alleles also associated with higher LFP: effect size on LFP was 0.64% per PNPLA3 allele and 1.13% per TM6SF2 allele. PNPLA3 interacted with MMED and vegetable intake to amplify their beneficial effect (p = 0.05). Among patients with no PNPLA3 risk alleles, 19% of individuals with MMED in the 25th percentile and 12% of those with MMED in the 75th percentile had NAFLD (Figure). Among those with two PNPLA3 risk alleles, these values were 37% and 22%, respectively, indicating a far greater effect of diet among those at higher genetic risk. TM6SF2 also interacted with MMED and vegetable intake to accentuate their beneficial effects on LFP (p< 0.05).
Discussion: Beneficial dietary patterns were enhanced by PNPLA3 and TM6SF2 risk alleles. Dietary interventions may be more effective in patients at elevated genetic risk.
Figure: Percentage of patients with non-alcoholic fatty liver (defined at liver fat percentage >5%), based on Modified Mediterranean Diet Score and PNPLA3-rs738409 genotype
Vincent Chen indicated no relevant financial relationships.
Xiaomeng Du indicated no relevant financial relationships.
Yanhua Chen indicated no relevant financial relationships.
Annapurna Kuppa indicated no relevant financial relationships.
Brian Halligan indicated no relevant financial relationships.
Elizabeth Speliotes indicated no relevant financial relationships.
Vincent L. Chen, MD, MS, Xiaomeng Du, MS, Yanhua Chen, PhD, Annapurna Kuppa, MS, Brian Halligan, PhD, Elizabeth Speliotes, MD, PhD, MPH. P1815 - PNPLA3 and TM6SF2 Risk Alleles Amplify Effects of Diet on Hepatic Steatosis, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.