The association between kidney function, cognitive function, and structural brain abnormalities in community-dwelling individuals aged 50+ is mediated by age and biomarkers of cardiovascular disease

Abstract

Aim Cognitive impairment has been associated with kidney function and chronic kidney disease. Whether this association is due to accelerated cardiovascular disease or an independent specific kidney function effect related to toxins is unclear. We investigated the impact of an array of clinical factors, inflammatory biomarkers, and cardiovascular biomarkers on the association between kidney function, cognitive function, and structural brain abnormalities. Methods and results We used data from the first and third waves of the TILDA Study, a population-representative prospective cohort of Irish adults aged 50 years and over, based on stratified random sampling (n = 3,774). The MRI sub-study included participants who consented to MRI brain imaging in addition to the health assessment. Multivariable linear and mixed-effect longitudinal regression models were fitted separately for each kidney marker/eGFR-equation after adjusting for baseline age and demographics, clinical vascular risk factors, and biomarkers. Unadjusted analyses showed an association between low eGFR, cognitive dysfunction, and cognitive decline (p < 0.001 for all kidney markers). Kidney function markers were also associated with white matter disease (OR = 3.32 (95% CI: 1.11, 9.98)), total grey matter volume (β=-0.17, 95% CI -0.27 to -0.07), and regional grey matter volumes within areas particularly susceptible to hypoxia (p < 0.001 for all). All the associations decreased after adjusting for age and were also diminished after adjusting for CVD biomarkers. Age and CVD-biomarker score were significant mediators of the adjusted associations between eGFR and cognitive status. These results remained consistent for cross-sectional and longitudinal outcomes and specific cognitive domains. Conclusions Decreased kidney function was associated with cerebrovascular disease. The association appeared to be mediated predominantly by age and the combination of CVD markers (namely NT-proBNP and GDF15), supporting the idea that shared biological pathways underline both diseases. Further mechanistic studies of the specific molecular mechanisms that lead to both kidney and cognitive decline are warranted. Translational perspective In our prospective cohort study, age, and concentrations of biomarkers previously related to CVD were found to be mediators of the association between mild to moderately decreased kidney function and neurocognitive abnormalities. In particular, a score composed of combined levels of GDF15 and NT-proBNP, was associated with cystatin C; and was mediating the adjusted association between kidney function and cognitive decline. Inflammation biomarker score showed borderline mediating effect. These results suggest potential causal vascular pathways that underline both development of early kidney function decline and cognitive decline. Additionally, CVD biomarkers could be evaluated in discriminating individuals at particular risk of dementia in the general population, who may benefit from intensive CVD risk factor modification.