In a large national sample of African American and white adults, individuals with lower levels of kidney function were more likely to have cognitive impairment compared to individuals with normal kidney function, independent of prevalent cardiovascular disease and cardiovascular risk factors. These results suggest that CKD, in addition to other modifiable cardiovascular risk factors, may be an important marker of cognitive impairment in US adults.
These results confirm and extend previous studies in elderly populations that had limited representation of African Americans and of individuals with advanced CKD. In the Health, Aging and Body Composition Study, we found that elderly individuals with CKD, defined similarly to the current study, had a 1.3 to 2.4-fold higher risk of cognitive decline over four years of follow-up, even after accounting for a number of confounding factors 4
. In a cross-sectional study of menopausal women participating in the Heart Estrogen/progestin Replacement Study (HERS), a lower estimated GFR was associated with poorer performance on tests of global cognition, executive function, language, and memory5
. When HERS participants were stratified by estimated GFR, only women with an estimated GFR <30 ml/min/1.73m2
had a significantly elevated prevalence of impairment. In the Cardiovascular Health Study, Seliger et al. found an association between serum creatinine concentration and the risk of incident dementia over a median six years of follow-up that was dependent on self-reported health status6
. Among older individuals with good or excellent health, an elevated serum creatinine concentration was associated with a 62% increased risk of dementia; however there was no association between kidney function and incident dementia in the subgroup of individuals with poor or fair self-reported health. Recently, Hailpern et al. reported that moderate CKD, defined as an estimated GFR of 30–59 ml/min/1.73m2
, was associated with poorer concentration and attention among 20–59 year old National Health And Nutrition Examination Survey (NHANES) participants15
Compared to these previous studies, the large sample size of REGARDS allowed us to examine the association between GFR and cognitive impairment across a wider spectrum of kidney function. Our results confirm that CKD is associated with an increased prevalence of cognitive impairment, and suggest that impairment may occur earlier in the course of kidney disease than previously recognized. Indeed, we found a significant increase in the prevalence of impairment for those with an estimated GFR <50 ml/min/1.73m2
, and a trend towards an increased prevalence of impairment among those typically classified as having normal or near-normal kidney function (estimated GFR 50–70 ml/min/1.73m2
), relative to participants with an estimated GFR 80–89 ml/min/1.73m2
. This finding should be interpreted with caution since these were exploratory analyses and the MDRD Study equation is known to be less accurate in this range16
. If confirmed in future studies, this would suggest that even small reductions in kidney function are associated with clinically significant consequences for cognitive functioning.
We also noted a significant, non-linear association between estimated GFR and the prevalence of cognitive impairment. Whether this observation reflects misclassification due to confounding from malnutrition or other factors is unclear. Nevertheless, it is worth noting since studies which do not account for this non-linear association may underestimate the true association between CKD and cognitive impairment. Future studies utilizing cystatin-c, a novel marker of kidney function which is currently being assayed in REGARDS17
, may clarify the association between kidney function and cognitive function among those with normal estimated GFR.
Individuals with CKD are frequently prescribed cumbersome medical regimens, and they must understand and weigh complex medical choices including the decision to undergo kidney transplantation, the decision to initiate dialysis, and the choice of dialysis modality. In addition to other factors, the high prevalence of cognitive impairment among persons with CKD may explain why practice guidelines for blood pressure management, preemptive vascular access placement, and other clinical targets remain difficult to achieve. While screening for cognitive impairment is generally not a routine part of CKD care, accurate identification and treatment of persons with cognitive impairment may facilitate improved adherence with dietary and pharmacologic therapies and aid in dialysis and long-term care planning.
The causes of cognitive impairment in persons with CKD cannot be determined from the current study, although these results suggest that traditional cardiovascular and stroke risk factors do not fully account for the association between reduced kidney function and cognitive impairment. In addition to traditional risk factors, CKD is also associated with a number of novel cardiovascular risk factors, including inflammation, oxidative stress, anemia, vascular calcification and hyperhomocysteinemia that may play an important role in the development and progression of cognitive impairment18–20
REGARDS participants were selected using population-based sampling strategies, therefore the results of this study should be broadly generalizable to the US adult population. Nevertheless, several limitations should be noted. These analyses were cross-sectional; therefore whether CKD is a marker for other factors that lead to cognitive impairment or a true causal risk factor cannot be concluded. Ongoing longitudinal studies of CKD and cognitive function in REGARDS and other cohorts will provide additional support for the hypothesis that CKD independently contributes to cognitive impairment. The Six-item Screener is a relatively insensitive measure of cognitive function and does not test different domains of cognitive function. The finding of such a strong association using a relatively insensitive measure of cognitive function only underscores the true strength of the association. The use of more sensitive measures of cognitive function, and in particular, measures of executive function associated with vascular causes of cognitive impairment, may have strengthened the associations reported here. Such measures were recently added to REGARDS follow-up assessments and will be available for future studies. Finally, while we adjusted for a large number of cardiovascular risk factors and for the presence of cardiovascular disease, residual confounding may still exist due to unmeasured comorbidity or misclassification.
Among African American and white US adults, lower levels of kidney function are associated with an increased prevalence of cognitive impairment, independent of traditional cardiovascular risk factors. The prevalence of impairment appears to increase early in the course of kidney disease; thus screening for impairment should be considered among all adults with CKD.
Given the high prevalence of CKD among US adults and the adverse consequences of cognitive impairment, future clinical trials to improve cognitive function should consider targeting this high risk population.