Response to: Critical commentary on the association between thiazolidinedione use and dementia risk in patients with type 2 diabetes

We thank Wasif et al for their comments¹ on our article that explored the association between thiazolidinedione (TZD) use and dementia risk in patients with type 2 diabetes mellitus (T2DM).² In their letter, the authors addressed the potential confounding effects of age of diabetes onset and mental disorders, which were suggested to be associated with dementia risk in recent studies,^{3, 4} and thus they suggested subgroup analyses according to participants' age at T2DM onset and status of mental disorders.¹ In our analyses we controlled age and years of T2DM at baseline²; thus the issue of collinearity would arise in the multivariate and inverse probability weighted model if we controlled age of T2DM onset, which was the difference between baseline age and diabetes duration. Moreover, comorbidities were adjusted using the Charlson comorbidity index, which did not contain mental disorders.²

In response to the authors' concerns, we conducted subgroup analyses according to their suggestions¹ (Table 1). Age of diabetes onset was defined as the age when the patient was first diagnosed with T2DM. In addition, history of mental disorders was identified using the International Classification of Diseases, Tenth Revision (all codes of F10–F99 except F70–F79) according to a previous study.⁴ Consistent with previous studies, we observed higher incidence of dementia in patients of older age and with history of mental disorders.^{3, 4} But no significant effect modification was observed. Compared with use of alpha-glucosidase inhibitors, TZD use was significantly associated with dementia incidence in different age groups, with a hazard ratio (HR) of 0.44 (95% confidence interval [CI], 0.21–0.91), 0.62 (95% CI, 0.39–0.98), and 0.47 (95% CI, 0.31–0.71) for T2DM patients aged <55 years, 55–65 years, and >65 years, respectively. Similarly, TZD use was consistently associated with dementia risk in terms of history of mental disorders, with an HR of 0.53 (95% CI, 0.30–0.94) and 0.50 (95% CI, 0.36–0.69) for T2DM patients with and without history of any mental disorders, respectively.

We agree with the authors that we cannot fully rule out biases caused by unmeasured confounding and inaccurate information.⁵ Therefore, we conducted several sensitivity analyses using different dementia identification rules combing diagnosis and prescriptions of antidementia drugs. Results under various outcome definitions were consistent (the Table 1), suggesting that our results were robust against potential outcome misclassification.² Furthermore, we did not adjust full medication history and some other potential confounders, such as dietary patterns, hence we calculate the E-value as a sensitivity analysis. The E-value for the observed HR of 0.51 in our primary analysis was 3.33, which was the minimum strength that an unmeasured confounder needed to be associated with both the exposure and outcome to explain away the observed association.²

However, our analyses should be interpreted in the context of its observational nature. Future prospective studies with more refined measurement of dementia cases and potential confounders are needed to give further insights into the potential role of TZDs in reducing dementia incidence.

Houyu Zhao conceived of and designed the work. Hongbo Lin, Peng Shen, and Yexiang Sun acquired the data. Houyu Zhao analyzed the data. Houyu Zhao drafted the manuscript. Lin Zhuo and Siyan Zhan critically revised the manuscript for important intellectual content. Siyan Zhan, Hongbo Lin, and Peng Shen supervised the study. Houyu Zhao and Siyan Zhan obtained the funding. All authors were responsible for the interpretation of the data, and revised, and gave final approval of the manuscript. Siyan Zhan is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

This work was supported by a grant from the China Postdoctoral Science Foundation (Grant No. 2022M710251) and a grant from the National Natural Science Foundation of China (Grant No. 81973146). The funder had no role in the study design, data collection, data analysis, data interpretation, writing of the report, or decision to submit the article for publication. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit the article for publication.

The authors have no conflicts of interest to declare.