Introduction

The purpose of this study was to perform neuropsychological assessments to detect vascular risk factors for cognitive decline in The Atherosclerosis Risk in Communities cohort (ARIC). ARIC is a large biracial, multi-site, longitudinal investigation of initially middle-age individuals.

Methods

Three cognitive assessments were administered to approximately 11, 000 individuals (approximately 8,700 white and 2,300 black) at two ARIC visits (ARIC Visit 2 and Visit 4) separated by six years. Visit 2 was considered the baseline. The subjects ranged in age between 47 to 70 years at the Visit 2 assessments. Women outnumbered men, with approximately 6,100 women and 4,800 men. Educational level varied widely with 6% having less than a ninth-grade education, 45% 9 to 12 years of education, 9% post-secondary vocational education, and 39% with at least some college education.

Cognitive Testing

Trained interviewers administered the tests in a quiet setting and standardized order. Change scores were calculated by subtracting the subjects' Visit 4 score from subjects' baseline score for each test. Mean change scores were tabulated by age level. The test battery included the:

• Delayed Word Recall (DWR), a test of verbal learning and recent memory that requires the subject to recall 10 common nouns following a 5-minute interval in which subjects generate sentences using the words.
• Digital Symbol Subtest (DSS) of the Wechsler Adult Intelligence Scale-Revised, a timed paper and pencil test that requires timed translation of numbers using a key provided.
• Word Fluency (WF), a test in which participants generate as many words a possible (excluding proper names and places) beginning with a particular letter of the alphabet in 60 seconds. Three separate 1-minute periods were used for each of the letters F, A and S.

Risk Factor Assessment

Baseline risk factors were assessed using the following criteria:

• Hypertension was considered to be present if systolic BP was >= 140 mm Hg, if diastolic BP was >= 90 mm Hg or if antihypertensive medications were being used.
• Diabetes was defined as a fasting blood glucose >/= 126 mg/dL (>/= 7 mmol/L), a nonfasting glucose >/= 200 mg/dL (>/= 11 mmol/L), a self-report of diabetes, or treatment for diabetes.
• Hyperlipidemia was defined as a low-density lipoprotein (LDL) cholesterol >/= 140 mg/dL (>/= 3.6 mmol/L), or the use of cholesterol-lowering agents.
• Participants were questioned about their use prescription and non-prescription medications in order to assess use of non-steroidal anti-inflammatory drugs, including aspirin (NSAIDS), and CNS-active medications.
• · Carotid wall intima-media thickness (IMT) was measured bilaterally using a high-resolution B-mode ultrasound. For analysis, IMT measurements were divided into tertiles (<0.65mm, 0.65-0.76mm and >0.76mm).

Results

The average time of follow-up was 6.0 (standard deviation, or SD =0.3) years with a range of 3.6 to 8.8 years. All mean change scores declined for the entire cohort over the duration of follow-up, but declines were relatively small - equivalent to 0.1 SD for DWR, 0.2 SD for DSS and 0.05 SD for WF - with a greater decline for older than younger participants.

Baseline Risk Factors and Test Score Change
Diabetes mellitus at baseline was associated with a greater decline in both DSS and WF. Diabetics' scores declined 0.9 more symbols for the DSS and 0.7 more symbols for the WF than non-diabetic subjects' scores (p < 0.05). The presence of hypertension at baseline was only associated with a decline in the DSS score. Hypertensive subjects had a 0.4 greater symbol decline in the test score than non-hypertensive subjects (p < 0.05).

There were 89 incident strokes in the group. The mean scores for this group declined four- to nine-fold more in the follow-up interval compared to subjects with no incident stroke (p <0.05). On average, participants who had incident stroke had a decline of 0.9 more words on the DWR, 6.7 more symbols on the DSS, and 4.3 more words on the WF compared to those who did not have incident stroke. A history of stroke at baseline was not associated with cognitive decline over the follow-up period.

Smoking status, carotid wall IMT, smoking, NSAID use at baseline were not associated with a decline in scores for any of the cognitive tests.

Baseline Risk Factors and Test Scores Stratified by Age
To determine the relationship between age and cognitive decline, the cohort was stratified into two groups: those 47 to 57 years of age and those >= 58 years of age.

• Diabetes was associated with cognitive decline for both the younger and the older age groups, with a decline on both the DSS and the WF (p<0.05).
• Hypertension was associated with a decline on the DSS alone for older participants only (p< 0.05).

A comparison of 4,000 subjects who did not return for Visit 4 with the cohort of 11,000 used for this report showed significant differences demographically. The baseline risk factors and cognitive status were also significantly worse. The authors note that the attrition of more impaired individuals may have lessened the ability to show the association between cognitive decline and risk factors.

Conclusions

In a middle age and young-elderly cohort, the decline in cognitive function is greater in subjects with diabetes mellitus, hypertension and incident stroke. The authors note that the changes are small and probably not clinically significant to the participants, but from a pathologic perspective, they "...offer evidence of the initiation of cognitive impairment associated with diabetes and hypertension." These finding suggest that effective treatment of diabetes and hypertension before age 60 might decrease cognitive impairment in later years.