Psychosis is one of the most typical and persistent neuropsychiatric manifestations
of AD where nearly half of AD patients experiences psychotic symptoms such as hallucinations
and/or delusions (Ropacki & Jeste., 2005). Psychotic AD patients tend to have poorer
overall health (Bassiony et al., 2000) and experience greater cognitive and functional
impairments (Chui et al., 1994; Fischer et al., 2012). These qualities contribute
to a reduced quality of life for patients and their caregivers. Researchers from our
group found that psychosis risk in neuropathologically-diagnosed AD patients is significantly
higher in females who are homozygous positive for APOE ε4 with the presence of Lewy body pathology (Kim et al., 2017). These results point
to a sexual dimorphism in AD psychosis where the ε4 allele may play a larger role
in influencing psychosis development in females. We hypothesize that AD patients with
this high-risk profile (female and homozygous positive for APOE ε4) are more susceptible to psychosis, potentially leading to more severe symptoms
as well as earlier disease onset respectively. As such, this exploratory study aims
to validate the sex-specific effect of APOE ε4 status (i.e. high-risk profile) on psychosis in a clinical cohort.
We first obtained all required Research Ethics Board approval from St. Michael’s Hospital
and the National Alzheimer’s Coordinating Centre (NACC). A series of parallel analyses
were conducted via SPSS (Version 220.127.116.11). For each analysis, we first conducted
a descriptive analysis of our sample including age, sex, cognitive status, psychosis
status, and presence of the relevant alleles. Patients with CNS neoplasm, non-AD-induced
dementia, and significant psychiatric co-morbidity were excluded as per the inclusion/exclusion
criteria for NACC.
Selected participants from the NACC database had a presumptive etiological diagnosis
of AD from a clinician. They also had available APOE ε4 status and Neuropsychiatric Inventory Questionnaire (NPI-Q) data. Subjects were
categorized by high-risk profile. Subjects with either hallucinations or delusions
as defined by the NPI-Q were considered psychotic (Figure 1).
Procedure & Analyses
We investigated the effect of having a high-risk profile (female and homozygous positive
for APOE ε4) on age and neuropsychiatric test scores at initial visit, as well as presence
and severity of psychotic symptoms (delusions or hallucinations).
To investigate the effect of the high-risk profile on age, we conducted a one-way
between-subjects analysis of variance in the high-risk vs. non-high-risk conditions.
Subjects with the high-risk profile (M = 70.43, SD = 7.75) were significantly younger
than those without the profile (M = 74.69, SD = 9.76) at the time of their initial
visit [F(1, 16202) = 139.05, p < .001] (Figure 2).
To investigate the effect of the high-risk profile on the presence of delusions and
hallucinations, we conducted a binary logistical regression for each psychotic symptom.
We found that the high-risk profile significantly predicted the presence of delusions,
b = 0.302, df = 1, p = .004. There was no significance for hallucinations.
Lastly, to study the effect of the high-risk profile on the severity of delusions
and hallucinations, we conducted an ordinal logistical regression for each psychotic
symptom. Subjects with the high-risk profile had a predicted increase of 0.388 in
the log-odds of being in a more severe category for delusions, p = .041. Similarly, high-risk individuals had a predicted increase of 0.638 in the
log-odds of being in a more severe category for hallucinations, p = .046.
From our results, the high-risk profile of being female and homozygous positive for
APOE ε4 seems to be associated with an earlier age of initial visit, an increased presence
of delusions and more severe psychotic symptoms. In terms of psychosis in AD, our
findings are in line with our prior work suggesting APOE ε4 may play a modulatory role in the development of psychosis in female patients
with AD, leading to earlier onset of symptoms, increased prevalence of delusions and
more severe psychotic symptoms (Kim et al., 2017).
Taken together, we believe that this high-risk profile may be a significant step towards
understanding the mechanisms that underly AD and psychosis in AD. Future directions
include elucidating the sex-specific cellular mechanisms on APOE ε4 status.
We wish to acknowledge the generosity of the St. Michael’s Hospital Foundation and
the Heather and Eric Donnelly endowment.