(OR21-02-26) Genetic Modification of LDL-C Response to Change in Saturated Fat Intake: A Secondary Analysis of the DIETFITS Trial

Objectives: Low-density lipoprotein cholesterol (LDL-C) responses to dietary interventions vary considerably across individuals, but the determinants of this variability remain incompletely understood. We conducted a secondary analysis of the DIETFITS trial to examine the independent and interactive effects of genetic risk and change (Δ) in saturated fat (SFA) intake on 6-month ΔLDL-C.

Methods: DIETFITS was a 12-month randomized clinical trial of 609 adults aged 18-50 years with overweight or obesity (BMI 28-40 kg/m²), assigned to a healthy low-carbohydrate (HLC) or healthy low-fat (HLF) diet. Analyses included genotyped participants with complete baseline and 6-month LDL-C and diet data (HLC n=223; HLF n=208). Genetic effects on LDL-C were summarized using a polygenic score for LDL-C (PGS-LDL). Multivariable linear regression models were fit separately by diet group, including both PGS-LDL and ΔSFA (% energy), to estimate independent associations with ΔLDL-C. Both PGS-LDL and ΔSFA were standardized, with coefficients representing mg/dL difference in ΔLDL-C per 1-SD. Models adjusted for age, sex, genetic PCs, baseline LDL-C, and 6-month Δweight.

Results: Substantial heterogeneity in LDL-C response was observed in both diet groups. At 6 months, percent ΔLDL-C ranged from -36% to +126% in HLC (mean ± SD 4.7 ± 19) and from -53% to +145% in HLF (-2.3 ± 22). In both diet groups, higher genetic risk and greater ΔSFA were independently associated with greater ΔLDL-C. Per 1-SD increment, PGS-LDL and ΔSFA were associated with a 6.0 (SE 1.4) and 5.1 (1.3) mg/dL higher ΔLDL-C, respectively, in HLC, and 4.2 (1.3) and 4.8 (1.3) mg/dL higher ΔLDL-C in HLF (all p< 0.01). A significant PGS-LDL x ΔSFA interaction was observed in HLC (p-interaction=0.0001) but not in HLF (p=0.33). In HLC, the association between ΔSFA and ΔLDL-C strengthened across PGS-LDL tertiles: -1.6 mg/dL (95% CI: -5.6, 2.4) in the lowest, 5.0 (0.5, 9.4) in the middle, and 10.5 (6.3, 15.0) in the highest risk tertile. This pattern was not observed in HLF.

Conclusions: Both genetic background and ΔSFA intake independently contributed to LDL-C response following a HLC or HLF diet intervention. In the context of a HLC diet, genetic propensity towards higher LDL-C amplified the LDL-C response to ΔSFA. These findings support the role of genetic background in LDL-C responsiveness to dietary interventions.

Funding Sources: AHA; NIDDK