Letter: Statin use in patients with type 2 diabetes has lower risk of hip fractures

We read with great interest the article 'Statin use in patients with type 2 diabetes has lower risk of hip fractures: A Taiwan national population-based study' by Tien-Ching Lee et al.1 They described that statin use was a significant and independent protective factor of hip fractures with an adjusted HR of 0.69. In addition, they noted a significant dose-response relationship between statin use and hip fracture risk. However, we believe there are several points that require further discussion. First, patients newly diagnosed with diabetes could have varying degrees of severity, and studies have shown a positive correlation between the severity of diabetes and the risk of hip fracture.2, 3 Ignoring this variable in the study may result in selection bias. It would be more appropriate to mitigate this bias by matching the severity levels of the two groups with the Diabetes Complications Severity Index, as suggested by Hsu et al.2 Second, in this multivariate analysis, the authors adjusted for 22 variables but did not conduct collinearity analysis.4, 5 Collinearity refers to the high correlation between two or more variables which could result in adverse effects on the main influencing factors and consequently affect the target variable. For instance, hypertension and antihypertensive medication in this study may provide redundant information. Therefore, identifying the possibility of collinearity for each variable is important. The variance inflation factor can be calculated for each variable to check for collinearity. Common solutions include removing variables with high Variance inflation factor values or finding ways to combine several variables with collinearity into one variable. Furthermore, building upon the previous point, the authors adjusted for 22 variables, but potential interactions among these variables were not taken into consideration. This omission may lead to an impact on the adjusted values, potentially overestimating the protective effect of statins on hip fractures. For instance, the combined use of glucocorticoids and proton pump inhibitors may have an additive effect on the risk of hip fractures.6 The authors could perform subgroup analysis to examine the interactions of known risk factors.7 Although a causal and definitive effect of statins on preventing hip fractures in patients with type 2 diabetes will require future studies to reach a final conclusion, we thank Tien-Ching Lee et al. for their excellent article and look forward to their response. All authors have read and approved this manuscript. The authors declare no conflicts of interest. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1002/dmrr.3756. Research data are not shared.