Bayesian spatio-temporal modeling of hyperlipidemia risk in Iran; A repeated cross-sectional analysis

Abstract

Abstract Background:There is little information about the use of hierarchical Bayesian approaches to accurately investigate the spatial distribution of relative risk of hyperlipidemia in Iran. In this research, we used hierarchical Bayesian models and examined the spatial distribution of relative risk of hyperlipidemia in separate provinces throughout Iran. Methods:in this study, all individuals with hyperlipidemia in all provinces of Iran in 2019. The main variables of the study included average age, gender, and number of cases of hyperlipidemia in each province. The population of each province was obtained from the Iranian Statistics Center, and was used to compute the disease prevalence, and the expected number of cases. Besag-York-Mollié (BYM) and Besag-York-Mollié 2 (BYM2) models to analyze data and Hamiltonian Monte Carlo method was also applied for parameter estimation. Results:The relative risk of hyperlipidemia was greater than 1 in 16% (95% CI: (0.304, 0.879)) of Iranian provinces (posterior probability > 0.8). Therefore, those aged >50 years old in Fars, Isfahan, Khuzestan, Razavi Khorasan and Tehran provinces were increasingly at risk of hyperlipidemia. The study found that women aged >70 years had the lowest average incidence of hyperlipidemia (RR=-0.86; 95% CI: (-1.13, -.0463)), while men aged 65-69 had the highest average incidence (RR=1.41; 95% CI: (-0.674, -0.129)). Conclusions: Our results show various relative risks of hyperlipidemia in different regions of the country, with some provinces at a higher risk. Moreover, the finding that women aged 70 years and above have the lowest average incidence of hyperlipidemia highlights the importance of early detection and management of the condition in younger age groups. Healthcare providers should focus on preventive measures, such as regular health screenings and lifestyle modifications, to reduce the risk of hyperlipidemia in high-risk populations.