Slowing Conversion From MCI to Alzheimer's Disease
Figure 1 shows the participants included in the analyses. There were 2,520 participants with MCI at baseline who had two or more follow-up visits, 1,964 (77.9%) of whom had hypertension and 545 (21.6%) who did not. Of participants with hypertension, 273 (13.9%) were African American, and 1,691 (86.1%) were Caucasian. Four hundred eighty-eight participants were taking RAS-acting medications at all visits and 296 were RAS-acting medication nonusers at all visits. Of the 296 RAS-acting medication nonusers, 113 were taking a calcium channel blocker, 174 a beta-blocker, and 108 a diuretic. Some of the RAS-acting medication nonusers were taking a combination of non-RAS-acting medications. Of RAS-acting medication users, 312 were taking BBB-crossing medications and 124 were taking non-BBB-crossing formulations.
(Enlarge Image)
Figure 1.
Consort diagram of participants included in the present study. Analyses include only participants treated for hypertension at each study visit. Participants with hypertension but not taking a medication and those without hypertension were not included. Participants who alternated between centrally acting and non-centrally acting medications across visits were excluded from the blood–brain barrier group analysis but were included in the renin-angiotensin system– (RAS) vs non-RAS-acting medication analyses. AA = African American; BP = blood pressure.
Table 1 lists baseline demographic and clinical variables according to RAS-acting medication user group. Both groups were well educated (15 years) and elderly (mean age 75), and most had been diagnosed with amnestic MCI (79% of RAS-acting medication users, 84% of nonusers). Approximately one-quarter of participants were taking a medication indicated for AD. RAS-acting medication users had higher systolic BP (P < .001) and were more likely to have self-reported diabetes mellitus (P < .001) than nonusers. There was one difference in cognitive test performance at baseline, with RAS-acting medication nonusers performing better on Digit Span Forward (P = .01; data not shown). There were no between-group differences in incident stroke at baseline or during follow-up.
There were racial differences in baseline demographic, vascular, and cognitive scores. Caucasians were more highly educated (P = .001). African Americans had a greater prevalence of diabetes mellitus (P < .001), and Caucasians self-reported more depression (P = .02). Caucasians scored better than African Americans on eight of 10 cognitive tests. In contrast, the CDR-SOB score was significantly higher (indicating greater disease severity) for Caucasians than African Americans. Of participants taking an antihypertensive medication, African Americans were more likely to be prescribed RAS-acting medications (P = .02).
Whether there were differences in conversion rate from MCI to AD was analyzed for RAS-acting and non-RAS-acting medication users and for centrally acting medication users and non-centrally acting medication users over an average of 3 years. Two hundred eighty of the 784 participants taking an antihypertensive medication converted to AD, which is consistent with reported conversion rates of 10% to 30% annually.
One hundred sixty-one RAS-acting medication users (33.0%), 119 nonusers (40.2%), 98 centrally acting medication users (30.7%), and 48 non-centrally acting medication users (40.0%) converted to AD. Adjusted results show that the conversion rate from MCI to AD was significantly lower for RAS-acting medication users than for nonusers (P = .04) and that, of RAS-acting medication users, those using a centrally acting medication converted less often than those using a non-centrally acting medication (P = .06). There was no significant interaction between Caucasians and African Americans for either model, although the power to detect such an interaction was low because of the small numbers of African Americans who converted (n = 35). No violation of the proportional hazards assumption was found for either model.
Table 2 compares the cognitive performance and functional ability of RAS-acting medication users with those of nonusers. Results revealed a beneficial effect of RAS-acting medication use over time on the CDR-SOB (P = .005) and Digit Span Forward (P = .02).
Next, whether participants taking BBB-crossing RAS-acting medications would have greater cognitive protection than those taking non-centrally acting medications was explored (Table 3, Figure 2). Three hundred twelve participants were taking BBB-crossing RAS-acting medications at baseline, and 124 were taking non-BBB-crossing medications. Results regarding rate of decline showed beneficial cognitive and functional effects on the Boston Naming Test (P < .01), MMSE (P < .01) and CDR-SOB (P < .01), with participants taking BBB-crossing RAS-acting medications having less cognitive decline than those taking non-BBB-crossing RAS-acting medications.
(Enlarge Image)
Figure 2.
Change in cognitive and functional score according to renin-angiotensin system (RAS)-acting medication use and central activation. CDR SOB = Clinical Dementia Rating Score of Boxes. MMSE = Mini-Mental State Examination
Because of lack of statistical power, it was not possible to test for racial differences in conversion rates to AD, but to determine whether there was an effect of the interaction between RAS-acting medication use and race on cognitive and functional ability, a three-way interaction (test by race by time) was examined for all cognitive tests and the CDR-SOB. Although results generally did not reveal a consistent effect of race on the CDR-SOB or Digit Span Forward, the three-way interaction had an effect on the MMSE (P = .05), category fluency test (P = .04), and Digit Span Backward (P = .03), with African Americans taking RAS-acting medications having more-robust cognitive benefits than Caucasians taking RAS-acting medications, although these associations were not considered significant with the conservative P-value of .027 and should be verified in future studies. Finally, a second three-way analysis (test by race by time) was performed to test for racial effects regarding centrally acting medications. Unlike the comparison between RAS-acting medication users and nonusers, comparisons of centrally acting medication users with non-centrally acting medication users did not reveal an effect of race on functional or cognitive ability.
Results
Participants
Figure 1 shows the participants included in the analyses. There were 2,520 participants with MCI at baseline who had two or more follow-up visits, 1,964 (77.9%) of whom had hypertension and 545 (21.6%) who did not. Of participants with hypertension, 273 (13.9%) were African American, and 1,691 (86.1%) were Caucasian. Four hundred eighty-eight participants were taking RAS-acting medications at all visits and 296 were RAS-acting medication nonusers at all visits. Of the 296 RAS-acting medication nonusers, 113 were taking a calcium channel blocker, 174 a beta-blocker, and 108 a diuretic. Some of the RAS-acting medication nonusers were taking a combination of non-RAS-acting medications. Of RAS-acting medication users, 312 were taking BBB-crossing medications and 124 were taking non-BBB-crossing formulations.
(Enlarge Image)
Figure 1.
Consort diagram of participants included in the present study. Analyses include only participants treated for hypertension at each study visit. Participants with hypertension but not taking a medication and those without hypertension were not included. Participants who alternated between centrally acting and non-centrally acting medications across visits were excluded from the blood–brain barrier group analysis but were included in the renin-angiotensin system– (RAS) vs non-RAS-acting medication analyses. AA = African American; BP = blood pressure.
Table 1 lists baseline demographic and clinical variables according to RAS-acting medication user group. Both groups were well educated (15 years) and elderly (mean age 75), and most had been diagnosed with amnestic MCI (79% of RAS-acting medication users, 84% of nonusers). Approximately one-quarter of participants were taking a medication indicated for AD. RAS-acting medication users had higher systolic BP (P < .001) and were more likely to have self-reported diabetes mellitus (P < .001) than nonusers. There was one difference in cognitive test performance at baseline, with RAS-acting medication nonusers performing better on Digit Span Forward (P = .01; data not shown). There were no between-group differences in incident stroke at baseline or during follow-up.
There were racial differences in baseline demographic, vascular, and cognitive scores. Caucasians were more highly educated (P = .001). African Americans had a greater prevalence of diabetes mellitus (P < .001), and Caucasians self-reported more depression (P = .02). Caucasians scored better than African Americans on eight of 10 cognitive tests. In contrast, the CDR-SOB score was significantly higher (indicating greater disease severity) for Caucasians than African Americans. Of participants taking an antihypertensive medication, African Americans were more likely to be prescribed RAS-acting medications (P = .02).
Conversion Rates to AD as a Function of RAS-acting Medication use
Whether there were differences in conversion rate from MCI to AD was analyzed for RAS-acting and non-RAS-acting medication users and for centrally acting medication users and non-centrally acting medication users over an average of 3 years. Two hundred eighty of the 784 participants taking an antihypertensive medication converted to AD, which is consistent with reported conversion rates of 10% to 30% annually.
One hundred sixty-one RAS-acting medication users (33.0%), 119 nonusers (40.2%), 98 centrally acting medication users (30.7%), and 48 non-centrally acting medication users (40.0%) converted to AD. Adjusted results show that the conversion rate from MCI to AD was significantly lower for RAS-acting medication users than for nonusers (P = .04) and that, of RAS-acting medication users, those using a centrally acting medication converted less often than those using a non-centrally acting medication (P = .06). There was no significant interaction between Caucasians and African Americans for either model, although the power to detect such an interaction was low because of the small numbers of African Americans who converted (n = 35). No violation of the proportional hazards assumption was found for either model.
Effect of RAS-acting Medications on Cognitive and Functional Ability
Table 2 compares the cognitive performance and functional ability of RAS-acting medication users with those of nonusers. Results revealed a beneficial effect of RAS-acting medication use over time on the CDR-SOB (P = .005) and Digit Span Forward (P = .02).
Effect of BBB-crossing RAS-Acting Medications on Cognitive and Functional Ability
Next, whether participants taking BBB-crossing RAS-acting medications would have greater cognitive protection than those taking non-centrally acting medications was explored (Table 3, Figure 2). Three hundred twelve participants were taking BBB-crossing RAS-acting medications at baseline, and 124 were taking non-BBB-crossing medications. Results regarding rate of decline showed beneficial cognitive and functional effects on the Boston Naming Test (P < .01), MMSE (P < .01) and CDR-SOB (P < .01), with participants taking BBB-crossing RAS-acting medications having less cognitive decline than those taking non-BBB-crossing RAS-acting medications.
(Enlarge Image)
Figure 2.
Change in cognitive and functional score according to renin-angiotensin system (RAS)-acting medication use and central activation. CDR SOB = Clinical Dementia Rating Score of Boxes. MMSE = Mini-Mental State Examination
Race and Longitudinal Decline According to RAS use
Because of lack of statistical power, it was not possible to test for racial differences in conversion rates to AD, but to determine whether there was an effect of the interaction between RAS-acting medication use and race on cognitive and functional ability, a three-way interaction (test by race by time) was examined for all cognitive tests and the CDR-SOB. Although results generally did not reveal a consistent effect of race on the CDR-SOB or Digit Span Forward, the three-way interaction had an effect on the MMSE (P = .05), category fluency test (P = .04), and Digit Span Backward (P = .03), with African Americans taking RAS-acting medications having more-robust cognitive benefits than Caucasians taking RAS-acting medications, although these associations were not considered significant with the conservative P-value of .027 and should be verified in future studies. Finally, a second three-way analysis (test by race by time) was performed to test for racial effects regarding centrally acting medications. Unlike the comparison between RAS-acting medication users and nonusers, comparisons of centrally acting medication users with non-centrally acting medication users did not reveal an effect of race on functional or cognitive ability.
SHARE
