Kidney Markers and Cognitive Impairment: SPRINT Study Baseline Data

Chronic kidney disease (CKD) among older adults affects as many as 26 million individuals in the United States. As the worldwide population ages, the prevalence of CKD increases, possibly reflecting the accompanying increase in cardiovascular risk factors such as hypertension, diabetes, and obesity. CKD stage 3 or 4 affects nearly 11% of individuals 60 to 69 years of age and more than one in three people ≥70 years of age.

CKD is associated with an increase in cardiovascular risk. The risk includes cerebrovascular circulation, and the frequency of cerebrovascular disease manifested by acute and subclinical stroke is also increased in individuals with CKD. Findings from previous studies have suggested that cerebrovascular disease may link CKD and cognitive impairment. It is unclear whether albuminuria and reduced glomerular filtration rate (GFR) are each associated with different cognitive domain profiles.

Daniel E. Weiner, MD, MS, and colleagues conducted a cross-sectional cohort study utilizing baseline data from SPRINT (Systolic Blood Pressure Intervention Trial). The study was designed to examine the relationships among kidney markers, cognitive function, and cerebrovascular disease. The study participants were at high risk for cardiovascular disease, without diabetes or known stroke. The current study evaluated participants in the SPRINT cognition substudy, SPRINT-Memory and Cognition in Decreased Hypertension (SPRINT-MIND).

The researchers tested the hypothesis that albuminuria and reduced GFR are independently correlated with poorer cognitive function and a larger burden of abnormal white matter volume. They reported results of the study in the American Journal of Kidney Diseases [2017;70(3):357-367]. The primary outcomes of interest were cognitive function, a priori defined as five cognitive domains based on 11 cognitive tests using Z scores, and abnormal white matter volume quantified by brain magnetic resonance imaging (MRI).

There were 2921 participants in SPRINT who received the extended cognitive testing. Of those, 214 had essential data missing, resulting in a final cohort of 2707 participants with complete essential data. Mean estimated GFR (eGFR) at baseline was 70.8 mL/min/1.73 m2, median baseline urine albumin-creatinine ratio (ACR) was 9.7 mg/g, mean age was 68.4 years, 36.7% were women, 30.5% were black, and 29.1% had a history of cardiovascular disease.

Levels of albuminuria were normal in 51.2% of participants, 28.7% had high-normal albuminuria (urine ACR, 10-29 mg/g), and 2.9% had severe albuminuria (urine ACR, 30-299 mg/g). Eighteen point two percent (n=493) had eGFR of 45 to <60 mL/min/1.73m2 and 11.3% (n=306) had eGFR <45 mL/min/1.73 m2. Participants with higher urine ACR were older, had prevalent cardiovascular disease, higher systolic blood pressure, and lower eGFR; lower eGFR was associated with older age, white race, prevalent cardiovascular disease, lower diastolic blood pressure, and higher urine ACR.

Raw scores on cognitive tests were stratified by urine ACR and eGFR. In unadjusted analyses, there was an association between higher urine ACR and significantly worse performance in all cognitive domains (global cognitive function, executive function, memory, attention/concentration, and language). Following partial and full adjustment, the association between urine ACR and cognitive function remained robust, with the exception of language and a borderline statistically significant association with memory. Each doubling of urine ACR had the same association with cognitive performance as being 7 months older for global cognitive function, 10 months for executive function, 6 months for memory, and 14 months for attention/concentration.

Lower eGFR was associated with worse performance in global cognitive function, executive function, attention/concentration, and memory in unadjusted analyses. Following multivariable adjustment, the associations were attenuated, but remained statistically significant for global cognition and memory.

A subset of SPRINT-MIND participants underwent MRI (n=637); they were younger and had less cardiovascular disease and higher eGFRs than those who did not undergo MRI. In multivariable models, there was an association between higher urine ACR with significantly larger abnormal white matter volume. Following adjustment for age, intracranial volume, sex, race, education, and scanner, eGFR was not associated with abnormal white matter volume.

The researchers cited some limitations to the study, including the cross-sectional design that limits the ability to draw conclusions regarding causality or reverse causation. Other than education, there were no data on other social factors that may have confounded the association between kidney disease markers and cognitive function. The associations are statistically significant, but it is unclear whether they are clinically significant. In older individuals particularly, estimates of GFR based on serum creatinine may be confounded by muscle mass. Finally, many cognitive tests do not precisely map to specific cognitive domains, making categorization of cognitive tests into specific categories somewhat arbitrary.

In conclusion, the researchers said, “In a large population of nondiabetic community-dwelling adults with risk factors for cardiovascular disease, CKD markers are associated with worse cognitive performance. Most notable, the presence of albuminuria, even at very low levels, is associated with worse global cognitive function and worse performance on tests evaluating executive function and attention/concentration domains, while concurrently identifying a higher burden of abnormal brain white matter disease. These findings suggest that vascular disease may mediate these inter-relationships and identify patients with kidney disease as a high-risk population for cognitive impairment.”

Takeaway Points

  1. In a subset of SPRINT participants, researchers conducted a cross-sectional cohort study to examine associations among kidney markers, cognitive function, and cerebrovascular disease.
  2. Following multivariable adjustment, those with higher urine albumin-creatinine ratio had worse performance on tests of global cognitive function, executive function, memory, and attention/concentration.
  3. There was an independent association between lower estimated glomerular filtration rate and worse global cognitive function and memory.