Abstract
The hepatic steatosis index (HSI), a noninvasive tool for assessing hepatic steatosis, has been linked to hypertension. This study aimed to investigate their nonlinear relationship and modifying factors. Using 2017-2020 NHANES data, we analyzed 7723 adults aged ≥ 20 years. Pearson chi-square test and Student t test were used to compare baseline characteristics across HSI quartiles (Table 1). Multivariable logistic regression assessed HSI-hypertension associations, while threshold regression identified inflection points. Subgroup analyses evaluated effect modifications. Baseline Characteristics: Higher HSI quartile (Q4) had a significantly different age distribution (P < .001), with the highest quartile (Q4) being younger on average (48.34 ± 16.28 years) than Q2 and Q3, female predominance (66.25%), and lower socioeconomic status (income-to-poverty ratio 2.38 ± 1.47, P < .001). Metabolic comorbidities showed dose-response trends, with diabetes prevalence rising from 5.33% (Q1) to 23.08% (Q4) (P < .001). Primary analysis: Each HSI unit was associated with a 4% higher risk of hypertension (odds ratio [OR] = 1.04, 95% confidence interval [CI]: 1.03-1.05). Q4 had a 2.65-fold higher risk than Q1 (95% CI: 2.23-3.15). Nonlinearity: A significant inflection point emerged at HSI = 50.33, with steeper risk slope below threshold (OR = 1.06, P < .001) versus null association above (OR = 0.99, P = .181). Subgroup Heterogeneity: Stronger associations occurred in males (β = 1.06 vs females 1.02), older adults (≥60 years), and coronary heart disease patients (threshold lowered to 38.67) (P interaction < 0.05). HSI exhibits a threshold-based nonlinear association with hypertension, demonstrating stronger associations in males, older adults, and cardiovascular disease subgroups, supporting its utility as a marker in metabolic hypertension risk stratification.