G-3749

Association between latent class Risk factors of Hypertension and Incidence of Hypertension: A Prospective Cohort Study

 Mobina Vatankhah ¹ © ℗, Azam Namdar ², Mehdi Sharafi ³   

 
Abstract

Introduction: A significant worldwide health burden, hypertension is caused by a complicated interaction between behavioral and metabolic risk factors. Conventional studies frequently evaluate these elements separately, failing to account for the combined effects of their clustering. The purpose of this study was to look into the ways that different latent profiles of combined risk variables affect the prevalence of adult hypertension. Methods and Materials: This prospective cohort study included 8,074 adults (mean age 46.98 ± 8.98 years; 50% female) without hypertension at baseline. Latent class analysis (LCA) was applied to identify unobserved subgroups based on metabolic and behavioral variables, including BMI, WHR, lipid profile, physical activity, smoking, alcohol use, drug use, and shift work. Cox proportional hazards models were used to estimate the risk of incident hypertension across latent classes, adjusting for potential confounders (age, sex, marital status, education, socioeconomic status, and physical activity). Results: During the follow-up period, 319 new cases of hypertension were identified, yielding an overall incidence of 4%. Incidence was significantly associated with older age (P0.001), female gender (P0.001), lower physical activity (P=0.014), obesity (P0.001), and elevated waist-to-hip ratio (WHR) (P0.001). Adverse lipid parameters, including low HDL (P=0.007) and elevated TyG index (P=0.009), were also linked to hypertension, along with behavioral factors such as smoking (P=0.001), drug use (P0.001), and alcohol consumption (P=0.012). LCA identified five latent classes with distinct risk profiles. Class 1 (23.6%) was characterized by smoking and drug use and served as the reference. Class 3 (49.0%) included individuals with central obesity and high WHR, while Class 5 (16.2%) had the most severe metabolic profile, including elevated TyG index, obesity, and dyslipidemia. Adjusted Cox regression showed significantly higher risk of hypertension in Class 3 (HR=1.93; 95% CI: 1.28–2.91; P=0.001) and Class 5 (HR=2.55; 95% CI: 1.65–3.93; P0.001), compared to Class 1. Classes 2 and 4 exhibited elevated risk, but associations were not statistically significant. Conclusion and Discussion: This study highlights the added value of latent class modeling in identifying high-risk subgroups for hypertension. Individuals with clustered metabolic abnormalities—particularly obesity, high WHR, and dyslipidemia—showed the highest risk. These findings support the need for tailored prevention strategies targeting specific risk combinations rather than isolated factors and emphasize early screening for metabolic syndrome traits in at-risk populations.

Keywords: Hypertension, Latent Class Analysis, Obesity, Metabolic Syndrome, Risk Factors