Abstract
Background
Albuminuria and estimated glomerular filtration rate (eGFR) are each associated with increased risk for cognitive impairment, but their joint association is unknown.
Study Design
Prospective cohort study.
Setting and Participants
A US national sample of 19,399 adults without cognitive impairment at baseline participating in the REGARDS )REasons for Geographic And Racial Disparities in Stroke) study.
Predictors
Albuminuria was assessed by the urine albumin-creatinine ratio (UACR) and GFR was estimated using the CKD-EPI (CKD Epidemiology Collaboration) equation.
Outcomes
Incident cognitive impairment was defined as a score of 4 or less on the Six-item Screener at the last follow-up visit.
Results
Over a mean follow-up of 3.8 ± 1.5 years, UACR 30 – 299 and ≥300 mg/g were independently associated with 31% and 57% higher risk for cognitive impairment, respectively, relative to individuals with UACR <10 mg/g. This finding was strongest among those with high eGFR and attenuated at lower levels (P=0.04 for trend). Relative an eGFR ≥60 ml/min/1.73m2, eGFR <60 ml/min/1.73m2 was not independently associated with cognitive impairment. However, after stratifying by UACR, eGFR <60 ml/min/1.73m2 was associated with 30% higher risk for cognitive impairment among participants with UACR <10 mg/g but not higher UACR levels (P=0.04 for trend).
Limitations
single measure of albuminuria and eGFR, screening test of cognition
Conclusions
When eGFR was preserved, albuminuria independently associated with incident cognitive impairment. When albuminuria was <10 mg/g, low eGFR independently associated with cognitive impairment. Albuminuria and low eGFR are complementary but not additive risk factors for incident cognitive impairment.
Keywords: albuminuria, chronic kidney disease, cognitive impairment
Chronic kidney disease (CKD), usually marked by the presence of persistent albuminuria or an estimated glomerular filtration rate (eGFR) <60 ml/min/1.73m2, is increasingly recognized as an elevated risk state for cognitive impairment and dementia 1. As CKD is common, affecting an estimated 13% of adults in the United States 2, and potentially modifiable, defining the relation between markers of CKD and risk for cognitive impairment is critical in this large at-risk population.
Both albuminuria and low eGFR are associated with elevated risk for cognitive impairment 3–7. However, previous studies have produced limited and conflicting information about whether albuminuria and low eGFR are associated with cognitive impairment independent of the other, owing to investigations focused on one marker but not both, cross-sectional study designs, use of dichotomous measures for albuminuria and eGFR, and study populations with a small number of participants who have both risk factors. A better understanding of these relationships could aid in risk stratification of patients with CKD and may also provide unique insight into the causal pathways that link abnormalities of kidney structure and function with cognitive impairment. Importantly, recent investigations have highlighted independent associations between albuminuria, eGFR and macrovascular outcomes 8–11 that have prompted an international working group to consider revising the CKD classification system 12. Determining whether similar relationships exist between albuminuria, eGFR and outcomes such as cognitive impairment that are strongly linked with microvascular disease 13, may help to inform these efforts.
We evaluated the association between albuminuria, eGFR, and the incidence of cognitive impairment in a large nationwide sample of United States (US) adults participating in the REGARDS (REasons for Geographic And Racial Disparities in Stroke) Study. We hypothesized that higher levels of albuminuria and lower levels of eGFR would be associated with higher risk for incident cognitive impairment independent of the other.
Methods
Study population
The REGARDS study is a national sample of 30,239 community-dwelling adults aged 45 years and older in the US population. Recruitment of the REGARDS cohort has been previously described 14. Briefly, REGARDS participants were recruited from a commercially available nationwide list of over 250 million individuals and 120 million households (Genesys, Inc) so that approximately half of the cohort was targeted to be black, half male, and half from the stroke belt/buckle, an area of the southeastern US with stroke mortality rates 50% higher than the rest of the US. Exclusion criteria included race other than black or white, active treatment for cancer, medical conditions preventing long-term participation, severe cognitive impairment judged by the telephone interviewer, residence in or inclusion on a waiting list for a nursing home, or inability to communicate in English. For these analyses, we included all subjects enrolled between December 18, 2003 (the date cognitive function testing was incorporated into the REGARDS baseline interview) and the end of study enrollment, November 1, 2007 (25,084 subjects). Among these subjects, 697 individuals were missing cognitive data at baseline, 1621 were missing serum creatinine or urinary albumin-creatinine ratio (UACR) values, and 43 individuals had self-reported kidney failure or an estimated GFR <15 ml/min/1.73m2. We further excluded 1943 individuals with cognitive impairment at baseline (defined below) and 1381 individuals without an assessment of cognitive function during follow-up, leaving 19,399 individuals in the analytic cohort.
Measurement of albuminuria and kidney function
During the baseline telephone interview demographic characteristics, health conditions and use of antihypertensive or diabetes medications, as well as verbal informed consent were obtained. During the subsequent in-home baseline examination anthropometric measurements, an electrocardiogram, phlebotomy, and written informed consent were obtained. Blood samples were shipped to a central laboratory for determination of serum creatinine and glucose. Serum creatinine measurements were calibrated to a method traceable to creatinine determined by isotope dilution mass spectrometry (IDMS) as previously described 15, and these values were then employed to estimate GFR using the CKD-EPI (CKD Epidemiology Collaboration) equation 16. Estimated GFR values were categorized as ≥90 ml/min/1.73m2, 60 to <90 ml/min/1.73m2, 45 to <60 ml/min/1.73m2, and <45 ml/min/1.73m2. UACR was measured on a random spot urine sample using the BN ProSpec Nephelometer (Dade Behring), and categorized as <10 mg/g, 10 to <30 mg/g, 30 to <300 mg/g and ≥300 mg/g.
Assessment of cognitive function
Starting on December 18, 2003 a six item cognitive screening examination was incorporated into the REGARDS baseline telephone interview and administered to all participants enrolled on or after that date, and then during annual follow-up telephone interviews. Designed for either in-person or telephone administration, the Six-item Screener (Item S1, available as online supplementary material) 17 is a validated test of global cognitive function that includes recall and orientation items derived from the widely used Mini-Mental State Exam.18 Scores on the Six-item Screener range from 0–6. We defined incident cognitive impairment as a score of 4 or less during the last available REGARDS telephone interview. In a previous study, a score of 4 or less had a sensitivity of 74.2% to 84.0% and a specificity of 80.2% to 85.3% for a diagnosis of cognitive impairment based on extensive clinical evaluation in community and clinical samples.17
Covariates
Race was defined by participant self-report. Education was categorized as less than high school education, high school education, some college, and professional. Geographic region of residence was categorized as stroke belt, stroke buckle, or other region, as previously defined.14 Hypertension was defined as a systolic blood pressure ≥140 mm Hg, a diastolic blood pressure ≥90 mm Hg, or the self-report of use of anti-hypertensive medications. Diabetes was defined as fasting glucose ≥126 mg/dL, non-fasting glucose ≥200 mg/dL, or self-reported current treatment for diabetes. Cerebrovascular disease was defined as self-report of stroke at baseline. Prevalent coronary heart disease was defined as ECG evidence of a prior myocardial infarction or self-report of a myocardial infarction, coronary artery bypass surgery, coronary angioplasty or coronary stenting. Alcohol and tobacco use were categorized as current versus past or never use.
Statistical Analysis
We used ANOVA and logistic regression as appropriate to assess differences in baseline characteristics across UACR categories. We used logistic regression models to evaluate the association, expressed as an odds ratio (OR) and 95% confidence interval (95% CI), between UACR and eGFR categories with incident cognitive impairment. We evaluated three different models: (1) an unadjusted model, (2) a model adjusted for all characteristics listed in the first table, and (3) a model adjusted for UACR and eGFR in addition to all characteristics listed in the first table. We constructed similar models using eGFR and log transformed UACR as continuous variables rather than categorical variables. Individuals with a urinary albumin excretion less than the detection limit for the assay used (2 mg/L) were assigned values of 1 mg/L. We also tested for interactions between UACR and eGFR, separately, with age (<65 versus ≥65 years, the median age at baseline in the REGARDS study), sex, and race. We determined the association between UACR and cognitive impairment within each eGFR stratum after dichotomizing UACR as <30 mg/g and ≥30 mg/g. Similar analyses were conducted to determine the association between eGFR <60 versus ≥60 ml/min/1.73m2 and cognitive impairment in each UACR stratum. The trend in the association between UACR and incident cognitive impairment across eGFR categories and between eGFR and incident cognitive impairment across UACR categories was assessed by including multiplicative interaction terms. Subsequently, we determined the joint association between albuminuria and eGFR with incident cognitive impairment by comparing the odds for incident cognitive impairment in eight categories: UACR <30 and ≥30 mg/g and eGFR ≥90, 60 to <90, 45 to <60, and <45 ml/min/1.73m2. Participants with UACR <30 mg/g and eGFR ≥90 ml/min/1.73m2 were used as the reference group.
In sensitivity analyses we used previously defined sex-specific UACR cut-points (UACR ≥ 25 mg/g in women or ≥ 17 mg/g in men) rather than the uniform cut-point of ≥30 mg/g 19. We also substituted a more stringent definition of incident cognitive impairment, a Six-item Screener score of four or less at each of the participant’s last two assessments (n=17,320 with two or more follow-up assessments), to determine if this materially changed the results. To assess whether exclusion of individuals missing follow-up data influenced the results, we conducted sensitivity analyses by classifying all individuals missing follow-up data as cases of cognitive impairment, and separately, as cognitively intact at follow-up. Lastly, as the exact date a participant developed cognitive impairment was not known (i.e., it was assessed annually), we also used interval censored regression models to calculate hazard ratios for developing cognitive impairment 20. All analyses were performed with SAS v9.1 (Cary, NC).
Results
Albuminuria ≥10 mg/g was present in 7037 (36.2%) study participants (Table 1). Individuals with albuminuria were older, more likely to be black and less likely to have a college education. They had a higher prevalence of most cardiovascular risk factors and a lower prevalence of current alcohol use. The mean eGFR was 86.3 ± 19.2 ml/min/1.73m2. The distribution of albuminuria by eGFR categories is shown in Figure 1. Participants excluded from the analytic cohort due to missing follow-up cognitive assessments were older and more likely to have albuminuria and low eGFR (Table S1).
Table 1.
Baseline characteristics, by albuminuria status.
| Baseline characteristics | Urine albumin-creatinine ratio | P-value for trend | |||
|---|---|---|---|---|---|
| <10 mg/g | 10 to <30 mg/g | 30 to <300 mg/g | ≥300 mg/g | ||
| No. (row percentage) | 12,362 (63.7%) | 4,437 (22.9%) | 2,181 (11.2%) | 419 (2.2%) | |
|
| |||||
| Age (years) | 62.9 +/− 9.1 | 66.0 +/− 9.5 | 66.5 +/− 9.9 | 65.8 +/− 9.4 | <0.001 |
|
| |||||
| Male sex | 40.9 | 35.2 | 44.0 | 49.6 | 0.2 |
|
| |||||
| Black race | 36.0 | 36.4 | 47.0 | 64.0 | <0.001 |
|
| |||||
| Region | |||||
| Non-belt/Buckle | 43.4 | 42.6 | 45.1 | 45.1 | Ref |
| Stroke Belt | 35.0 | 35.4 | 34.8 | 33.7 | 0.5 |
| Stroke Buckle | 21.7 | 22.0 | 20.2 | 21.2 | 0.3 |
|
| |||||
| Education | |||||
| < High school | 8.5 | 11.8 | 15.1 | 14.1 | Ref |
| High school | 24.5 | 26.7 | 27.1 | 30.8 | <0.001 |
| Some College | 27.3 | 27.9 | 27.1 | 30.6 | <0.001 |
| College/Professional | 39.7 | 33.6 | 30.7 | 24.6 | <0.001 |
|
| |||||
| Coronary Heart Disease | 13.3 | 18.9 | 23.0 | 34.2 | <0.001 |
|
| |||||
| Stroke | 3.5 | 5.9 | 8.1 | 12.4 | <0.001 |
|
| |||||
| Diabetes | 13.5 | 23.7 | 36.8 | 60.7 | <0.001 |
|
| |||||
| Hypertension | 51.0 | 64.4 | 74.4 | 87.8 | <0.001 |
|
| |||||
| Current smoking | 13.0 | 14.2 | 16.5 | 21.6 | <0.001 |
|
| |||||
| Current alcohol use | 55.7 | 51.4 | 46.6 | 41.8 | <0.001 |
|
| |||||
| eGFR (ml/min/1.73m2) | 87.8 +/−17.3 | 85.9 +/− 19.3 | 81.6 +/− 23.4 | 66.8 +/− 28.2 | <0.001 |
|
| |||||
| eGFR category | |||||
| ≥90 ml/min/1.73m2 | 49.2 | 46.4 | 39.4 | 24.3 | Ref |
| 60–89 ml/min/1.73m2 | 44.2 | 43.3 | 41.1 | 32.7 | <0.001 |
| 45–59 ml/min/1.73m2 | 5.2 | 7.5 | 12.0 | 19.1 | <0.001 |
| <45 ml/min/1.73m2 | 1.4 | 2.8 | 7.5 | 23.9 | <0.001 |
|
| |||||
| Six-item screener score | <0.001 | ||||
| 5 points | 20.0 | 22.6 | 25.5 | 29.6 | |
| 6 points | 80.0 | 77.4 | 74.5 | 70.4 | |
Note: Data cover included study participants only. Continuous data presented as mean +/− standard deviation; categorical data as percentage.
Abbreviations: eGFR, estimated glomerular filtration rate; Ref, reference
Figure 1.
Prevalence of albuminuria according to eGFR categories among included participants at baseline. There were N=9102 (47%), 8419 (43%), 1317 (7%) and 561 (3%) participants with eGFR ≥90, 60–89, 45–59, and <45 ml/min/1.73m2, respectively. Note: albuminuria is expressed as urine albumin-creatinine ratio (ACR, in mg/g). eGFR - estimated glomerular filtration rate.
The average duration of follow-up was 3.8 ± 1.5 years over which time 1549 subjects (8.0%) developed cognitive impairment. Cognitive impairment developed in 6.6 %, 9.0 %, 12.2 % and 15.0 % of participants with UACR <10 mg/g, 10 to <30 mg/g, 30 to <300 mg/g, and ≥300 mg/g, respectively. After adjustment for age, sex, race, education, region, diabetes, hypertension, baseline cardiovascular disease, baseline stroke, smoking and alcohol use, UACR of 30 to 299 and ≥300 mg/g, compared to UACR <10 mg/g, was associated with ORs for incident cognitive impairment of 1.32 (95% CI, 1.12 – 1.55) and 1.60 (95% CI, 1.18 – 2.16) (Table 2). When log-transformed UACR was analyzed as a continuous variable, each two-fold increase in UACR was associated with 7% higher odds for developing cognitive impairment (OR, 1.07; 95% CI, 1.03–1.10) after multivariable adjustment. These associations were similar after additional adjustment for eGFR and baseline Six-item Screener score. There was no evidence for statistical interaction between UACR and age, sex, or race on the odds of incident cognitive impairment (P-values >0.1).
Table 2.
Association of UACR and eGFR with incident cognitive impairment
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| UACR | |||
| <10 mg/g | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) |
| 10-<30 mg/g | 1.40 (1.23 – 1.58) | 1.14 (0.99 – 1.30) | 1.14 (0.99 – 1.30) |
| 30 to <300 mg/g | 1.97 (1.70 – 2.28) | 1.32 (1.12 – 1.55) | 1.31 (1.12 – 1.55) |
| ≥ 300 mg/g | 2.49 (1.89 – 3.29) | 1.60 (1.18 – 2.16) | 1.57 (1.15 – 2.14) |
| Per UACR doubling | 1.16 (1.13 – 1.19) | 1.07 (1.03 – 1.10) | 1.07 (1.03 – 1.10) |
| eGFR | |||
| ≥ 90 ml/min/1.73m2 | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) |
| 60 to <90 ml/min/1.73m2 | 1.55 (1.38 – 1.74) | 1.01 (0.89 – 1.15) | 1.01 (0.88 – 1.15) |
| 45 to <60 ml/min/1.73m2 | 2.38 (1.98 – 2.85) | 1.13 (0.92 – 1.39) | 1.09 (0.88 – 1.34) |
| <45 ml/min/1.73m2 | 2.74 (2.13 – 3.51) | 1.15 (0.87 – 1.52) | 1.05 (0.79 – 1.39) |
| eGFR (/20 ml/min/1.73m2 decrease) | 1.37 (1.30 – 1.44) | 1.04 (0.97 – 1.11) | 1.02 (0.95 – 1.09) |
Note: Values shown are odds ratios (95% confidence interval). mean follow-up duration is 3.8 years.
Model 1: Crude
Model 2. Adjusted for age, race, sex, education, region, diabetes, hypertension, cardiovascular disease, stroke, smoking and alcohol use.
Model 3. Model 2 + UACR (<10, 10 to <30, 30 to <300, ≥300 mg/g) or eGFR (<45, 45 to <60, 60 to <89, ≥90 ml/min/1.73m2)
Abbreviations: eGFR, estimated glomerular filtration rate; UACR, urine albumin-creatinine ratio
Cognitive impairment developed in 6.0 %, 8.9 %, 13.1 % and 14.8 % of participants with eGFR ≥90, 60 to <90, 45 to <60 and <45 ml/min/1.73m2, respectively. After adjustment for age, sex, race, education, region, diabetes, hypertension, baseline cardiovascular disease, baseline stroke, smoking and alcohol use, there was no association between eGFR categories or eGFR as a continuous variable and incident cognitive impairment (Table 2). For example, compared to the referent group with eGFR ≥90 ml/min/1.73m2, eGFR <45 was associated with 5% higher odds for incident cognitive impairment (OR, 1.05; 95% CI, 0.79 – 1.39). The ORs for incident cognitive impairment increased modestly in the eGFR range of 60 to <90 ml/min/1.73m2 (ORs of 0.96 [95% CI, 0.83–1.11] for eGFR 75 to <90 and 1.10 [95% CI, 0.94–1.30] for eGFR 60 to <75 ml/min/1.73m2), but there was no evidence for a higher risk for impairment at very high eGFR values (i.e. >110 ml/min/1.73m2). Adjustment for albuminuria and additional adjustment for Six-item Screener score did not change these findings. There was no evidence for statistical interaction between eGFR and age, sex, or race on the odds of incident cognitive impairment (P-values >0.1).
In analyses stratified by eGFR, ACR ≥30 mg/g (versus <30 mg/g) was associated with higher odds ratios for incident cognitive impairment among individuals with an eGFR of 60 to <90 ml/min/1.73m2 (OR, 1.37; 95% CI, 1.11 – 1.69) and ≥90 ml/min/1.73m2 (OR, 1.44; 95% CI, 1.12 – 1.85) (Table 3). However the association was diminished with lower eGFR level (P=0.04 for trend across eGFR categories ≥90, 60 to <90, 45 to <60, <45 ml/min/1.73m2). In analyses stratified by ACR, eGFR <60 ml/min/1.73m2 (versus >60 ml/min/1.73m2) was associated with higher odds for incident cognitive impairment among individuals with ACR <10 mg/g (OR, 1.30; 95% CI, 1.02 – 1.66), but not at higher ACR levels (P=0.04 for trend across ACR categories <10, 10 to <30, 30 to <300 and 300 mg/g) (Table 4). When we categorized participants according to UACR and eGFR, individuals with UACR ≥30 mg/g and eGFR ≥60 ml/min/1.73m2 and individuals with UACR <30 mg/g and eGFR <45 ml/min/1.73m2 had the highest adjusted odds for incident cognitive impairment compared to the reference group with UACR <30 mg/g and eGFR ≥90 ml/min/1.73m2 (Figure 2).
Table 3.
Association of albuminuria with incident cognitive impairment, by eGFR stratum
| Unadjusted | Multivariable-adjusted* | |
|---|---|---|
| EGFR category | ||
| ≥ 90 ml/min/1.73m2 | 1.83 (1.45 – 2.32) | 1.44 (1.12 – 1.85) |
| 60 to <90 ml/min/1.73m2 | 1.88 (1.55 – 2.28) | 1.37 (1.11 – 1.69) |
| 45 to <60 ml/min/1.73m2 | 1.28 (0.90 – 1.82) | 1.10 (0.74 – 1.62) |
| <45 ml/min/1.73m2 | 0.90 (0.56 – 1.44) | 0.80 (0.48 – 1.36) |
| P-value for trend | 0.006 | 0.04 |
Except where indicated, values shown are odds ratio (95% confidence interval). Albuminuria defined as UACR ≥30 mg/g. The referent category is UACR <30 mg/g. P-value for trend represents the interaction between eGFR as a continuous variable and UACR ≥ 30 mg/g.
Adjusted for age, race, sex, education, region, diabetes, hypertension, cardiovascular disease, stroke, smoking and alcohol use.
Abbreviations: eGFR, estimated glomerular filtration rate; UACR, urine albumin-creatinine ratio
Table 4.
Association of eGFR <60 ml/min/1.73m2 with incident cognitive impairment, by UACR stratum
| Unadjusted | Multivariable-adjusted* | |
|---|---|---|
| UACR category | ||
| <10 mg/g | 2.13 (1.71 – 2.66) | 1.30 (1.02 – 1.66) |
| 10 to <30 mg/g | 1.80 (1.36 – 2.40) | 1.02 (0.74 – 1.40) |
| 30 to <300 mg/g | 1.30 (0.96 – 1.76) | 0.87 (0.62 – 1.22) |
| ≥300 mg/g | 1.00 (0.58 – 1.71) | 0.76 (0.41 – 1.40) |
| P-value for trend | <0.001 | 0.04 |
Note: Except where indicated, values shown are odds ratio (95% confidence interval). The referent category is eGFR ≥60 ml/min/1.73m2. P-value for trend represents the interaction between ACR as a log-transformed continuous variable and eGFR <60 ml/min/1.73m2.
Adjusted for age, race, sex, education, region, diabetes, hypertension, cardiovascular disease, stroke, smoking and alcohol use.
Abbreviations: eGFR, estimated glomerular filtration rate; UACR, urine albumin-creatinine ratio
Figure 2.
Adjusted odds ratios for cognitive impairment over 3.8 years, by estimated glomerular filtration rate (eGFR, in ml/min/1.73m2) and urine albumin-creatinine ratio (ACR, in mg/g). The referent category is eGFR ≥90 ml/min/1.73m2 and ACR <30 mg/g. Odds ratios are adjusted for age, sex, race, education, region, diabetes, hypertension, cardiovascular disease, stroke, smoking and alcohol use. Bars represent 95% confidence intervals.
N=297, 976, 7385, and 8141, respectively for categories ACR <30 mg/g and eGFR <45, 45–59, 60–89, and ≥90 ml/min/1.73m2. N=264, 342, 1033, and 961, respectively, for categories ACR ≥30 mg/g and eGFR <45, 45–59, 60–89, and ≥90 ml/min/1.73m2.
We conducted several sensitivity analyses. When we utilized sex-specific UACR cut-points, the multivariable adjusted OR for incident cognitive impairment associated with albuminuria was 1.29 (95% CI, 1.02–1.63), 1.46 (95% CI, 1.20–1.76), 1.07 (95% CI, 0.74–1.57), and 0.78 (95% CI, 0.46–1.32) for adults with eGFR ≥90 ml/min/1.73m2, 60 to <90 ml/min/1.73m2, 45 to <60 ml/min/1.73m2 and <45 ml/min/1.73m2, respectively. When we excluded individuals with stroke at baseline, the results were also similar (Table S2). When we utilized a more stringent definition of cognitive impairment requiring a Six-item Screener score ≤4 at each of the last two assessments, the magnitude of the association between albuminuria and incident cognitive impairment was strengthened, but was unchanged for eGFR. For example, among individuals with UACR 30 to <300 mg/g and ≥300 mg/g (versus < 10 mg/g), the odds ratio for cognitive impairment was 1.33 (95% CI, 0.97 – 1.82) and 2.24 (95% CI, 1.31–3.82), respectively, after multivariable adjustment. When we utilized a time to event model, the results were unchanged for UACR and strengthened for eGFR (Table S3). Because subjects excluded from the analytic cohort due to missing follow-up cognitive assessments were more likely to have UACR ≥300 mg/g and more likely to have eGFR <45 ml/min/1.73m2, we also repeated the analyses after including individuals with missing follow-up cognitive assessments and classifying them as impaired, and separately, as cognitively intact. The association between albuminuria and incident cognitive impairment was similar (data not shown), while the association between eGFR <45 ml/min/1.73m2 and incident cognitive impairment was strengthened when all individuals with missing follow-up assessments were classified as impaired (OR, 1.32; 95% CI, 1.05–1.65).
Discussion
In a large cohort of black and white adults, when eGFR was preserved, albuminuria independently associated with higher risk for developing cognitive impairment. When albuminuria was <10 mg/g, low eGFR independently associated with higher risk for developing cognitive impairment.
Our findings confirm the importance of albuminuria and eGFR as risk factors for cognitive impairment, and expand on previous studies to demonstrate their joint association and gradients in risk over a wide range of albuminuria and eGFR levels. In the Rancho Bernardo Study involving 1345 community-dwelling adults with a mean age of 75, ACR ≥30 mg/g, but not eGFR, was independently associated with cognitive decline among men, with no association between ACR, eGFR and cognitive decline among women 4. In a cross-sectional study of 335 homebound elders with a mean age of 73, Weiner et al. noted albuminuria, assessed as a continuous variable, but not eGFR, was associated with lower cognitive function and a higher prevalence of brain small vessel ischemic disease 5. In cross-sectional analyses from the CRIC (Chronic Renal Insufficiency Cohort) Study which included 3591 adults with a mean age of 58 and eGFR <70 ml/min/1.73m2, lower eGFR, but not albuminuria, was associated with lower cognitive function 21. Differences in the study populations or in the definition of cognitive impairment may explain the conflicting findings of previous studies. For example, the Six-item Screener used in this study evaluates memory and orientation but not executive function, a cognitive domain linked with cerebrovascular disease.
In the CKD classification paradigm, albuminuria is considered an early marker of kidney disease and low eGFR a late marker. In the current study, however, the risk for incident cognitive impairment was higher among adults with albuminuria and eGFR ≥60 ml/min/1.73m2 (currently classified as CKD stage 1–2) as compared to individuals without albuminuria and eGFR 45–59 ml/min/1.73m2 (currently classified as CKD stage 3), and comparable to or higher than the risk of incident cognitive impairment among those with eGFR <45 ml/min/1.73m2 (CKD stages 3–4). These findings are important for several reasons. First, most older adults with CKD die before progression to ESRD 22. Thus, it is important for revised CKD classification methods to accurately reflect the risk for clinical outcomes beyond ESRD. Second, they highlight the heterogeneity of risk within the same CKD stage and in particular, the disproportionately high risk relative to assigned CKD stage among those with albuminuria and normal or near-normal eGFR. Third, it suggests that the pathways which link albuminuria and low eGFR with cognitive impairment are only partially overlapping. Albuminuria is a marker of microvascular dysfunction 23, whereas low eGFR may also be a marker of other novel risk factors, such as disorders of mineral metabolism, anemia, oxidative stress, altered drug disposition and others that contribute to cognitive impairment through vascular and non-vascular mechanisms. If the presence of albuminuria and low eGFR identify different pathways of risk, then this may imply a role for different therapeutic strategies to delay cognitive decline.
There may be several possible explanations for the observation that albuminuria is more strongly associated with incident cognitive impairment at higher compared to lower eGFR levels. Creatinine-based GFR estimating equations, including the CKD-EPI equation, overestimate GFR in individuals with low muscle mass, and may misclassify kidney function in malnourished individuals who are at high risk for cognitive impairment 16. Alternatively, the phenotype of albuminuria with high eGFR may identify individuals with glomerular hyperfiltration. This pattern has been described in patients with and without diabetes 24, 25, and may be a marker of systemic microvascular hemodynamic abnormalities that contribute to accelerated loss of kidney and cognitive function. While it is possible that competing mortality risk may explain the attenuation in risk for cognitive impairment at low eGFR, we do not think this is the case because our results did not change substantively in sensitivity analyses that included individuals with missing follow-up data due to death or other causes.
Interestingly, the attenuation in risk associated with albuminuria at lower eGFR is not unique for the outcome of cognitive impairment. For example, in a large community-based study, Hemmelgarn et al. noted that the magnitude of the association between heavy albuminuria and risk for mortality, myocardial infarction or ESRD was reduced at low eGFR as compared to normal eGFR 9. In a meta-analysis of community-based cohorts, Matsushita et al. reported a similar interaction between albuminuria and eGFR in several cohorts, although this interaction was not statistically significant after pooling data 10. However, in contrast to our results, these studies found that albuminuria remained independently associated with a higher risk for mortality even among those with moderate or severe reductions in eGFR (<45 ml/min/1.73m2). A similar, independent association of albuminuria and eGFR with mortality has been noted in REGARDS 26. Thus it seems more likely that our findings reflect differences in microvascular versus macrovascular outcomes rather than differences in the REGARDS cohort versus other community-based cohort studies. Owing to a relatively small sample of adults with eGFR <45 ml/min/1.73m2 after stratifying for albuminuria we cannot exclude the possibility that albuminuria may also be independently associated with cognitive impairment in this group.
Strengths of the REGARDS study include the large national sample of black and white adults with standardized measures of albuminuria and serum creatinine at baseline and prospective assessment of cognitive function using a validated measure. This study also has several limitations. First, albuminuria is known to have day-to-day variability 27. However, misclassification of some individuals would be expected to bias the results towards the null. Second, the Six-item Screener is a relatively insensitive measure of cognitive function and it does not assess executive function, a cognitive domain linked with vascular diseases. It may also misclassify some individuals due to its narrow scoring range. We attempted to reduce this misclassification by using different definitions of cognitive impairment, with consistent results. Attrition as a result of death or loss to follow-up may have influenced the study results. We performed several sensitivity analyses after including individuals missing follow-up data, with similar findings, suggesting these results are robust to several assumptions we made in the analyses. Finally, although we adjusted for several potential confounders in these analyses, there may be residual confounding from unmeasured factors.
In summary, we found that albuminuria and low eGFR were complementary but not additive risk factors for the development of cognitive impairment. Our findings provide additional support for incorporating information about albuminuria into CKD classification methods, and suggest that the pathways mediating the association between albuminuria and low eGFR with cognitive impairment may be partially distinct.
Supplementary Material
Table S1: Characteristics of those included in analytic cohort versus those excluded.
Table S2: Association of UACR and eGFR with incident cognitive impairment, excluding individuals with stroke at baseline.
Table S3: Association of UACR) and eGFR with time to first cognitive impairment.
Item S1: The Six-Item Screener.
Acknowledgments
These findings were reported as an oral presentation on November 20, 2010 at the American Society of Nephrology meeting in Denver, CO.
The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at http://www.regardsstudy.org.
Support: This research project is supported by a cooperative agreement U01 NS041588 from the National Institute of Neurological Disorders and Stroke, National Institutes of Health, Department of Health and Human Service. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Neurological Disorders and Stroke or the National Institutes of Health. Representatives of the funding agency have been involved in the review of the manuscript but not directly involved in the collection, management, analysis or interpretation of the data. Dr. Kurella Tamura is supported by K23AG028952 from the National Institute of Aging. Additional funding was provided by an investigator-initiated grant-in-aid from Amgen Inc to Dr Warnock. Dr. Bradbury is an employee of Amgen Inc. The manuscript was sent to Amgen for internal review before submission for publication, but the company did not have any role in the design and conduct of the study or the collection, management, data analysis, or interpretation of the data.
Footnotes
Financial Disclosure: The authors declare that they have no other relevant financial interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Kurella Tamura M, Yaffe K. Dementia and cognitive impairment in ESRD: diagnostic and therapeutic strategies. Kidney Int. 2011;79:14–22. doi: 10.1038/ki.2010.336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Coresh J, Selvin E, Stevens LA, et al. Prevalence of chronic kidney disease in the United States. Jama. 2007;298:2038–47. doi: 10.1001/jama.298.17.2038. [DOI] [PubMed] [Google Scholar]
- 3.Barzilay JI, Gao P, O’Donnell M, et al. Albuminuria and decline in cognitive function: The ONTARGET/TRANSCEND studies. Arch Intern Med. 2011;171:142–50. doi: 10.1001/archinternmed.2010.502. [DOI] [PubMed] [Google Scholar]
- 4.Jassal SK, Kritz-Silverstein D, Barrett-Connor E. A prospective study of albuminuria and cognitive function in older adults: the Rancho Bernardo study. Am J Epidemiol. 2010;171:277–86. doi: 10.1093/aje/kwp426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Weiner DE, Bartolomei K, Scott T, et al. Albuminuria, Cognitive Functioning, and White Matter Hyperintensities in Homebound Elders. Am J Kidney Dis. 2008 doi: 10.1053/j.ajkd.2008.08.022. [DOI] [PMC free article] [PubMed]
- 6.Kurella M, Chertow GM, Fried LF, et al. Chronic kidney disease and cognitive impairment in the elderly: the health, aging, and body composition study. J Am Soc Nephrol. 2005;16:2127–33. doi: 10.1681/ASN.2005010005. [DOI] [PubMed] [Google Scholar]
- 7.Seliger SL, Siscovick DS, Stehman-Breen CO, et al. Moderate renal impairment and risk of dementia among older adults: the Cardiovascular Health Cognition Study. J Am Soc Nephrol. 2004;15:1904–11. doi: 10.1097/01.asn.0000131529.60019.fa. [DOI] [PubMed] [Google Scholar]
- 8.Hallan S, Astor B, Romundstad S, Aasarod K, Kvenild K, Coresh J. Association of kidney function and albuminuria with cardiovascular mortality in older vs younger individuals: The HUNT II Study. Arch Intern Med. 2007;167:2490–6. doi: 10.1001/archinte.167.22.2490. [DOI] [PubMed] [Google Scholar]
- 9.Hemmelgarn BR, Manns BJ, Lloyd A, et al. Relation between kidney function, proteinuria, and adverse outcomes. Jama. 2010;303:423–9. doi: 10.1001/jama.2010.39. [DOI] [PubMed] [Google Scholar]
- 10.Matsushita K, van der Velde M, Astor BC, et al. Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet. 2010;375:2073–81. doi: 10.1016/S0140-6736(10)60674-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.O’Hare AM, Hailpern SM, Pavkov ME, et al. Prognostic implications of the urinary albumin to creatinine ratio in veterans of different ages with diabetes. Arch Intern Med. 2010;170:930–6. doi: 10.1001/archinternmed.2010.129. [DOI] [PubMed] [Google Scholar]
- 12.Eckardt KU, Berns JS, Rocco MV, Kasiske BL. Definition and classification of CKD: the debate should be about patient prognosis--a position statement from KDOQI and KDIGO. Am J Kidney Dis. 2009;53:915–20. doi: 10.1053/j.ajkd.2009.04.001. [DOI] [PubMed] [Google Scholar]
- 13.Lopez OL, Jagust WJ, Dulberg C, et al. Risk factors for mild cognitive impairment in the Cardiovascular Health Study Cognition Study: part 2. Arch Neurol. 2003;60:1394–9. doi: 10.1001/archneur.60.10.1394. [DOI] [PubMed] [Google Scholar]
- 14.Howard VJ, Cushman M, Pulley L, et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25:135–43. doi: 10.1159/000086678. [DOI] [PubMed] [Google Scholar]
- 15.Kurella Tamura M, Wadley V, Yaffe K, et al. Kidney function and cognitive impairment in US adults: the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study. Am J Kidney Dis. 2008;52:227–34. doi: 10.1053/j.ajkd.2008.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604–12. doi: 10.7326/0003-4819-150-9-200905050-00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Callahan CM, Unverzagt FW, Hui SL, Perkins AJ, Hendrie HC. Six-item screener to identify cognitive impairment among potential subjects for clinical research. Med Care. 2002;40:771–81. doi: 10.1097/00005650-200209000-00007. [DOI] [PubMed] [Google Scholar]
- 18.Folstein MF, Folstein SE, McHugh PR. “Mini-mental state” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–98. doi: 10.1016/0022-3956(75)90026-6. [DOI] [PubMed] [Google Scholar]
- 19.Mattix HJ, Hsu CY, Shaykevich S, Curhan G. Use of the albumin/creatinine ratio to detect microalbuminuria: implications of sex and race. J Am Soc Nephrol. 2002;13:1034–9. doi: 10.1681/ASN.V1341034. [DOI] [PubMed] [Google Scholar]
- 20.Finkelstein DM. A proportional hazards model for interval-censored failure time data. Biometrics. 1986;42:845–54. [PubMed] [Google Scholar]
- 21.Kurella Tamura M, Xie D, Yaffe K, et al. Vascular risk factors and cognitive impairment in chronic kidney disease: the Chronic Renal Insufficiency Cohort (CRIC) Study. Clin J Am Soc Nephrol. doi: 10.2215/CJN.02660310. in press. [DOI] [PMC free article] [PubMed]
- 22.O’Hare AM, Choi AI, Bertenthal D, et al. Age affects outcomes in chronic kidney disease. J Am Soc Nephrol. 2007;18:2758–65. doi: 10.1681/ASN.2007040422. [DOI] [PubMed] [Google Scholar]
- 23.Stehouwer CD, Henry RM, Dekker JM, Nijpels G, Heine RJ, Bouter LM. Microalbuminuria is associated with impaired brachial artery, flow-mediated vasodilation in elderly individuals without and with diabetes: further evidence for a link between microalbuminuria and endothelial dysfunction--the Hoorn Study. Kidney Int Suppl. 2004:S42–4. doi: 10.1111/j.1523-1755.2004.09211.x. [DOI] [PubMed]
- 24.Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiation and progression of diabetic and other glomerulopathies. Am J Med. 1982;72:375–80. doi: 10.1016/0002-9343(82)90490-9. [DOI] [PubMed] [Google Scholar]
- 25.Pinto-Sietsma SJ, Janssen WM, Hillege HL, Navis G, De Zeeuw D, De Jong PE. Urinary albumin excretion is associated with renal functional abnormalities in a nondiabetic population. J Am Soc Nephrol. 2000;11:1882–8. doi: 10.1681/ASN.V11101882. [DOI] [PubMed] [Google Scholar]
- 26.Warnock DG, Muntner P, McCullough PA, et al. Kidney function, albuminuria, and all-cause mortality in the REGARDS (Reasons for Geographic and Racial Differences in Stroke) study. Am J Kidney Dis. 56:861–71. doi: 10.1053/j.ajkd.2010.05.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Miller WG, Bruns DE, Hortin GL, et al. Current issues in measurement and reporting of urinary albumin excretion. Clin Chem. 2009;55:24–38. doi: 10.1373/clinchem.2008.106567. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Characteristics of those included in analytic cohort versus those excluded.
Table S2: Association of UACR and eGFR with incident cognitive impairment, excluding individuals with stroke at baseline.
Table S3: Association of UACR) and eGFR with time to first cognitive impairment.
Item S1: The Six-Item Screener.