Incident acute kidney injury and the risk of dementia: the stockholm creatinine measurements (scream) project

Abstract Background and Aims Preclinical studies suggest that acute kidney injury (AKI) leads to biochemical and pathological changes in the brain. However, whether developing AKI associates with the risk of dementia is not well explored. We here aimed to evaluate the association between developing AKI and the subsequent risk of dementia in general population. Method We included all adults with age ≥65 years, free from dementia diagnosis and with known kidney function in Stockholm during 2006-2019. The exposure was AKI (time varying): prevalent history of AKI was ascertained by clinical diagnoses, and new/incident AKI events during observation by both clinical diagnoses and transient creatinine elevations according to KDIGO criteria. The outcome was dementia diagnosis, first via confirmed cases in the Swedish registry of cognitive/dementia disorders (SveDem) and enriched with cases identified by two issued diagnoses of dementia in outpatient care or initiation of specific anti-dementia medications. We considered death, migration, and end of follow-up as censoring events and also explored risk associations with specific dementia subtypes. The association between developing AKI and study outcomes was evaluated through time-varying multivariate Cox regression, accounting for recurrent AKI events through weighted cumulative exposure metrics. Results We included 354,612 individuals with median 76 years (55% women). During median follow-up of 8.1 (IQR: 3.7, 12.1) years, there were 48,624 (14%) persons suffering at least one episode of AKI, and 47,765 (13%) diagnoses of dementia were recorded. The rate of dementia cases was 37.13 per 1000 person-year (95% CI 36.06-38.22) after developing AKI, which was approximately 2 times higher than the rate observed during the periods without AKI (18.47 [18.3-18.64]). After multivariable adjustment, developing AKI was associated with a 48% higher risk of having dementia [adjusted HR 1.48 (95%CI 1.43-1.53)]. This pattern was consistent for the dementia subtypes: the HRs for Alzheimer's dementia, Parkinson's disease dementia, and vascular dementia were 1.14 (1.07-1.21), 1.90 (1.37-2.65) and 1.24 (1.15-1.33) higher, respectively, after developing AKI. Among persons with recurrent AKIs, each additional AKI event was associated a 4% higher risk of dementia (adjusted HR: 1.04, 95% CI: 1.03-1.05). We observed similar results across subgroups of sex, baseline kidney function, history of hypertension, diabetes, and cardiovascular disease. Conclusion In this region-representative cohort, participants who experienced AKI were at increased risk of receiving a diagnosis of dementia.