OBJECTIVE: To investigate the inter-relationship among diabetes, urinary tract infection (UTI), and urinary tract calculi (UTC). METHODS: This study used Taiwan's National Health Insurance claims data of ambulatory care visits and hospitalizations. A total of 12,257 newly diagnosed diabetes cases in 2000-2002 and 96,781 controls were followed to the end of 2007. The person-year approach with Poisson assumption was used to estimate the incidence density (ID) of UTC by diabetic status. Relative risk of UTC in relation to diabetes and UTI were estimated from Cox proportional hazard model with adjustment for sociodemographic variables and comorbidities. RESULTS: Over nearly 8 years of follow-up, 8.9% of diabetes and 7.2% of control subjects sought ambulatory care or were hospitalized for UTC, representing the ID of 14.4 and 11.4 per 1000 person-years, respectively. The multivariate analysis indicated that UTC risk was independently associated with diabetes (hazard ratio 1.18, 95% CI 1.10-1.27) and UTI (HR 1.68, 95% CI 1.60-1.76). The hazard ratio of UTC in relation to diabetes in men and women without UTI was 1.24 and 1.26, respectively. Diabetes may further increase the rate of UTC in women with UTI, with an hazard ratio increased from 1.79-2.12. Such additive effect by diabetes in men with UTI was only marginal (HR 1.68 vs 1.67). CONCLUSION: This cohort study provides epidemiologic support for the causal association between diabetes and UTC, which is independent of UTI. In addition, female patients with UTI accompanied by diabetes tended to be associated with a greater rate of UTC.
OBJECTIVE: To investigate the inter-relationship among diabetes, urinary tract infection (UTI), and urinary tract calculi (UTC). METHODS: This study used Taiwan's National Health Insurance claims data of ambulatory care visits and hospitalizations. A total of 12,257 newly diagnosed diabetes cases in 2000-2002 and 96,781 controls were followed to the end of 2007. The person-year approach with Poisson assumption was used to estimate the incidence density (ID) of UTC by diabetic status. Relative risk of UTC in relation to diabetes and UTI were estimated from Cox proportional hazard model with adjustment for sociodemographic variables and comorbidities. RESULTS: Over nearly 8 years of follow-up, 8.9% of diabetes and 7.2% of control subjects sought ambulatory care or were hospitalized for UTC, representing the ID of 14.4 and 11.4 per 1000 person-years, respectively. The multivariate analysis indicated that UTC risk was independently associated with diabetes (hazard ratio 1.18, 95% CI 1.10-1.27) and UTI (HR 1.68, 95% CI 1.60-1.76). The hazard ratio of UTC in relation to diabetes in men and women without UTI was 1.24 and 1.26, respectively. Diabetes may further increase the rate of UTC in women with UTI, with an hazard ratio increased from 1.79-2.12. Such additive effect by diabetes in men with UTI was only marginal (HR 1.68 vs 1.67). CONCLUSION: This cohort study provides epidemiologic support for the causal association between diabetes and UTC, which is independent of UTI. In addition, female patients with UTI accompanied by diabetes tended to be associated with a greater rate of UTC.