Higher FHR was associated with greater NAFLD prevalence in adults with type 2 diabetes mellitus (T2DM). The observational analysis, published in Diabetes, Metabolic Syndrome and Obesity, evaluated the relationship between FHR and NAFLD risk using regression and nonlinear modeling approaches.

The study included 718 adults with T2DM recruited from Shenzhen People’s Hospital in China. Participants were stratified into quartiles based on FHR. Multiple linear regression assessed associations between FHR and NAFLD, and a generalized additive model tested for nonlinearity. Subgroup analyses examined consistency across clinical characteristics. Discriminative performance was evaluated using area under the curve analysis.

After adjustment for relevant variables, FHR was positively associated with NAFLD (OR 1.30; 95% CI 1.15-1.48). A nonlinear association was observed, with an inflection point at an FHR value of 1.23. Effect estimates differed across this threshold, with odds ratios of 3.07 (95% CI 1.51-6.24) below and 1.20 (95% CI 1.05-1.37) above the inflection point. Stronger associations were observed in subgroups with systolic blood pressure <140 mmHg, alanine transaminase >40 U/L, fasting blood glucose ≤7 mmol/L, urinary albumin-to-creatinine ratio ≤30 mg/g, triglycerides ≤1.7 mmol/L, and a history of alcohol consumption.

FHR demonstrated greater discriminative ability for NAFLD (AUC 0.697) compared with fasting C-peptide (AUC 0.649) or high-density lipoprotein cholesterol alone (AUC 0.635).

These findings show that FHR is nonlinearly associated with NAFLD in adults with T2DM, with higher risk observed when values exceed the identified threshold.