Dear Editor, We have read with great interest the article written by Hsieh et al., 1 which reports the association of topical corticosteroids (TCSs) use with an increased risk of osteoporosis and major osteoporotic fracture from a population-based case-control study in Taiwan. Subsequently, we would like to highlight certain key concerns we have with regard to the article.Firstly, the study used a case-control design. Considering osteoporosis is a silent asymptomatic disease, Dual-energy X-ray absorptiometry (DXA) is crucial for the diagnosis of osteoporosis among asymptomatic patients. Past study has reported that DXA screening was underused in women at increased fracture risk, as 7-year cumulative incidence of DXA screening was 58.8% among women aged 60-64 years with ≥1 risk factor in a regional health system in the US. 2 However, the prevalence of DXA screening is even lower in Taiwan. In Taiwan, screening for osteoporosis was not supported in policy, and most of the enrollees of Taiwan NHI must pay the cost of DXA evaluation by themselves unless osteoporosis diagnosis or osteoporotic fracture is evident. 3 According to a previous population-based study in Taiwan conducted by Hsu et al. 4 in 2022, only 2.9% of patients receive prior DXA testing. Therefore, a prospective study design is suggested in order to minimize misclassification bias.Secondly, there are some factors increasing the chance of being prescribed medication, including TCSs. In a past study in Greece, Pappa et al. discovered that older age, university education, visits to physician, comorbidity, or poor physical and mental health were associated with a higher likelihood of using drugs, while visits to physicians, smoking, and obesity increased the likelihood of major polypharmacy. 5 Also, the use of over-the-counter (OTC) topical agents is very popular in Taiwan, 6 and OTC drugs used were not recorded in National Health Insurance Research Database, which could also cause misclassification bias. We therefore suggest the authors adjust factors such as health service utilization or health literacy to better evaluate the association between TCSs used and research outcome.Lastly, although the author adjusted some comorbidities and medications which could lead to outcomes in analysis,