Of the 812 subjects with AD who were originally considered for this study, 21 (5%) had APOE
genotype Ε2Ε4. Because of the protective effects that have been previously reported for the APOE*2
], individuals with the Ε2Ε4 genotype were excluded from this analysis, although including them in the analyses produced similar results to those presented below. An additional subject with no neuropsychiatric data available was also excluded, leaving 790 subjects for the analysis. Table shows that 348 subjects (44%) had no APOE*4
allele, 368 (47%) had one Ε4 allele, and 74 (9%) had two Ε4 alleles. The sample was 33% male and had a mean education of 12 years, a mean illness duration of 4 years, a mean MMSE score of 17.6, and a mean CDR score of 1.29. The mean age at assessment decreased in relation to the number of APOE*4
alleles subjects possessed, from 78.1 years (SD 5.9) for those with no APOE*4
alleles to 72.8 years (SD 5.4) for those with two APOE*4
alleles, but otherwise there was no evidence of an association between the demographic variables and the number of APOE*4
Demographics and clinical characteristics of Alzheimer’s disease study sample
Table shows the frequency of reported neuropsychiatric symptoms in the four domains by number of APOE*4 alleles and the RR of a neuropsychiatric symptoms in a given domain by one or two APOE*4 alleles versus no APOE*4 alleles. Of the 790 subjects, 110 (14%) had missing values for at least one item in the BRSD. The occurrence of hallucinations was imputed for 6 (0.8%) subjects, the occurrence of delusions was imputed for 101 (13%) subjects, and the occurrence of aberrant motor behavior was imputed for 9 (1%) subjects; there were no missing agitation values, so no imputation was necessary for this domain. Across all 790 subjects, 99 (13%) experienced at least one occurrence of hallucinations, 540 (68%) experienced an occurrence of delusions, 327 (41%) experienced an occurrence of agitation, and 426 (54%) experienced an occurrence of aberrant motor behavior. In comparing subjects with different numbers of APOΕ*4 alleles, the occurrence of hallucinations tended to decrease as the number of APOΕ*4 alleles increased; we found that 53 (15%) subjects with no APOΕ*4 alleles experienced hallucinations, whereas only 4 (5%) subjects with two APOΕ*4 alleles experienced hallucinations. These results are illustrated by the decreasing RRs of hallucinations for one APOΕ*4 allele (0.71, 95% CI [0.50, 0.97]) and two APOΕ*4 alleles (0.32, 95% CI [0.00, 0.74]) versus no APOΕ*4 alleles after adjusting for age, gender, education, duration of illness, MMSE score, and CDR score at assessment. A post-hoc analysis modeling the number of APOE*4 alleles as continuous yielded a test for trend p-value of 0.0021 and an estimated relative decrease in risk of hallucinations of 34% for each increase in one APOE*4 allele, that is, a relative risk of 0.66, 95%CI (0.49, 0.86). There was no evidence of associations between the number of APOΕ*4 alleles and the occurrence of neuropsychiatric symptoms in the other three domains. Sensitivity analyses that imputed domains with missing items to the alternate values yielded similar results.
Frequency of AD subjects with at least one neuropsychiatric symptom in a given domain, and relative risk for the symptom in that domain by one and two APOΕ *4 alleles versus no APOE*4 allele
Of the 99 subjects who experienced hallucinations, 93 (94%) experienced at least one delusion as well, and of the 538 subjects who experienced delusions and had non-missing values for hallucinations, 445 (83%) experienced no hallucinations. The RR of delusions without hallucinations was 1.14 (95% CI [1.00, 1.29]) for subjects with one APOE*4 allele and 1.00 (95% CI [0.77, 1.29]) for subjects with two APOE*4 alleles versus subjects with no APOE*4 alleles, in contrast to the RR of hallucinations and delusions of 0.71, (95% CI [0.52, 0.97]) and 0.34 (95% CI [0.11, 1.05]) for one and two APOE*4 alleles respectively vs. no alleles. We conducted a secondary analysis comparing the 445 subjects with delusions and no hallucinations (D) to the 93 subjects with delusions and hallucinations (D+H). Our findings, summarized in Table , show that the D+H group had a significantly longer duration of illness, lower MMSE scores, and higher CDR scores compared to the D group. On the other hand, there were no significant differences in gender, age at assessment, and education between the two groups. Although there was no significant difference in APOΕ*4 allele frequency between the two groups, the D+H group tended to have lower APOΕ*4 frequencies than the D group. As Figure illustrates, the D+H group had significantly higher frequencies for each kind of delusion item than the D group. All comparisons were statistically significant (paranoid and imposter, P<.05, all others, P<.01).
Demographic and clinical characteristics of AD subjects with delusions only compared to subjects with delusions and hallucinations
Figure 1 Frequency of Delusion Type by Group. Legend: Frequency of specific delusion items by group (delusions and hallucinations vs. delusions only). This figure demonstrates that the delusions and hallucinations group had statistically significant higher frequencies (more ...)