Investigation of risk factors for osteoporosis with a focus on hypertension and estimation of the causal effect of hypertension on osteoporosis using causal forest.

Abstract

The current study aimed to comprehensively investigate the factors that most significantly increase the likelihood of developing osteoporosis, which is of great importance for aging populations. To this end, we focus on hypertension (HT) and examine its interaction and causal effect on osteoporosis. Using an administrative claims database, a nested case-control study and time-to-event analysis were conducted focusing on Japanese individuals aged ≥65 years. The results of the nested case-control study showed that rheumatoid arthritis (RA) had the highest odds ratio (OR = 1.961, 95% CI = 1.85-2.078), followed by HT (OR = 1.722, 95% CI = 1.659-1.787). In the time-to-event analysis, RA had the highest hazard ratio (HR = 2.133, 95% CI = 1.972-2.308), followed by chronic kidney disease (CKD) (HR = 1.473, 95% CI = 1.354-1.602), chronic obstructive pulmonary disease (HR = 1.46, 95% CI = 1.323-1.611), and HT (HR = 1.269, 95% CI = 1.21-1.331). Additionally, significant interactions were observed when HT co-existed with CKD, disorders of lipoprotein metabolism and other lipidemias (DLM), and RA. Moreover, the summary causal tree results of the conditional average treatment effect (CATE) using a causal inference approach revealed that the subgroup with DLM = 0, diabetes mellitus (DM) = 0, and RA = 0 exhibited the highest estimated CATE of 0.372, suggesting a strong independent causal effect of HT on osteoporosis in this group.