Risk of dementia in patients with end-stage renal disease under maintenance dialysis - A nationwide population-based study with consideration of competing risk of mortality

Background: Dementia is prevalent in the end-stage renal disease (ESRD) population. However, it is still not clarified whether ESRD is one of the etiology of dementia or its attributable effect on the cumulative risk of dementia. Meanwhile, the effect of competing risk of mortality should be taken into consideration when performing epidemiologic analyses among populations with high risk of mortality. Methods: By using the National Health Insurance Research Database (1998-2010), we identified 927,142 non-ESRD individuals and 99,158 ESRD patients to investigate the effect of ESRD on the risk of dementia. Age- and sex-specific incidence rates (IRs) and cumulative incidence rates (CIRs) were first compared between these two cohorts. Competing risk analyses including cause-specific and subdistribution proportional hazards models were then constructed with adjustments for potential confounders. Results: The overall IR and CIR of dementia were much higher in the ESRD group than in the non-ESRD group (10.73 vs. 1.40 per 1000 person-years and 0.061 vs. 0.017, respectively, both P < 0.0001). Results from the multivariable cause-specific hazard models suggested that ESRD was one of the etiological factors for dementia (cause-specific hazard ratio [csHR]: 2.06 [95% CI: 1.95-2.17]). However, the subdistribution HR (sdHR) of ESRD was 0.51 (95% Cl: 0.49-0.54), which indicated the lower cumulative incidence risk of dementia in ESRD patients. The inverse relationship between csHR and sdHR could be explained by the high mortality rate in the ESRD population. These findings were also essentially consistent across various subgroup analyses according to selected confounders, as well as in the analyses that limited dementia diagnoses made by neurologists or psychologists. Conclusions: Although ESRD appears directly associated with the risk of dementia, the high competing mortality means that primary prevention of comorbidity associated with dementia may be more effective in reducing overall dementia in the general population, which may also potentially reduce the incidence of ESRD and prevent death from multimorbidity when affected by ESRD.