Cognitive Domain Impairment and All-Cause Mortality in Older Patients Undergoing Hemodialysis

The highly prevalent cognitive impairment in hemodialysis patients is associated with all-cause mortality; however, the role of different cognitive domain impairments in this association is still not clarified. Our objective was to determine the association between cognitive domain impairment and all-cause mortality in elderly adult patients undergoing hemodialysis. We conducted a prospective cohort study including patients from 11 hemodialysis centers in Beijing. Baseline data were collected, and a series of neuropsychological batteries covering 5 domains of cognitive function were included for the assessment of cognitive function. According to the fifth version of the Diagnostic and Statistical Manual of Mental Disorders criteria (DSM-V), the patients were classified as normal, mild, and major cognitive impairment for global and domain cognitive function, then followed up for 1 year. Kaplan-Meier survival analysis was used to compare the difference in the cumulative survival rate in different cognitive domains. A multivariate Cox proportional hazards regression analysis was used to determine the association between global or domain cognitive impairment and all-cause mortality. A total of 613 patients were enrolled, the mean age was 63.82 +/- 7.14 years old, and 42.1% were women. After 49.53 +/- 8.42 weeks of follow-up, 69 deaths occurred. Kaplan-Meier plots demonstrated a significant association of cognitive impairment in memory, executive function, attention, and language domains with all-cause death. Multivariate Cox regression analysis showed that mild and major impairment of global cognition (HR = 2.89 (95% CI, 1.01-8.34), p = 0.049 and HR = 4.35 (95% CI, 1.55-12.16), p = 0.005, respectively), executive cognitive domain (HR = 2.51 (95% CI, 1.20-5.24), p = 0.014; HR = 3.91 (95% CI, 1.70-9.03), p = 0.001, respectively), and memory cognitive domain (HR = 2.13 (95% CI, 1.07-4.24), p = 0.031; HR = 3.67 (95% CI, 1.71-7.92), p = 0.001, respectively) were associated with all-cause mortality. Combined impairment of 3, 4, and 5 cognitive domains was associated with all-cause mortality [HR = 5.75 (95% CI, 1.88-17.57), p = 0.002; HR = 12.42 (95% CI, 3.69-41.80), p < 0.001; HR = 13.48 (95% CI, 3.38-53.73), p < 0.001, respectively]. We demonstrate an association between the executive and memory cognitive domain impairment and all-cause mortality in hemodialysis patients. Our data suggest that the impairments in these cognitive domains might help in the early identification of hemodialysis patients at risk of death.