A preliminary study of apolipoprotein E genotype and psychiatric manifestations of Alzheimer's disease

The epsilon 4 isoform of the apolipoprotein (apo) E gene is associated with AD. ^[1] Evaluation of the clinical phenotypes of AD with isoforms of the APOE gene may offer insight into possible roles of the apoE4 protein in AD. Lopez-Alberola et al. ^[2] found no differences in the psychiatric, demographic, neurologic, or neuroimaging features of 31 AD patients with the APOE4/ - genotype and 31 AD patients with other APOE genotypes. However, the study was small, control subjects were not included, and the frequency of other alleles was unspecified.

We hypothesized that patients with AD and the epsilon 4 allele would be phenotypically different from patients with AD without the epsilon 4 allele and from control subjects. We focused on the presence or absence of psychiatric manifestations because the frequency of depression and psychosis is high in patients with AD ^[3-5] and disturbances in neuronal circuitry and neurotransmitter imbalances are common to both AD and psychiatric illnesses and therefore proteins such as apoE may influence psychiatric manifestations in AD.

Methods.

Data were available from a previously described, community based, longitudinal study of the elderly, representative of individuals over 65 residing in the area. ^[6] All subjects had a neuropsychological battery, a 17-item Hamilton Depression Rating Scale (HDRS), and a screen for psychosis adapted from the Screening Clinical Interview with the DSM-III-R (SCID). A diagnosis of probable AD was made using the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease Related Disorders Association criteria. The severity of cognitive impairment was assessed by the clinical dementia rating scale (CDR), the total recall score on the Bushke selective reminding memory test, and the Blessed cognitive score. ^[6,7]

Patients were excluded from the study if they had genotypes other than APOE3/3 or APOE3/4, because these genotypes were less frequent or if they had possible AD (implying a potential additional cause for their dementia). We included patients with AD and a CDR between 0.5 (very mild AD) ^[8] and 2 (moderate AD). Subjects were divided into four categories based on the APOE3/3 or APOE3/4 genotype and the presence or absence of AD.

For depression, we evaluated the total HDRS score. For psychosis, we used questions dealing with delusions and hallucinations from the adapted SCID pooled for an overall psychosis score with 1 point for each affirmative answer (maximum score of 3). Scores were evaluated both as continuous and categorical variables. For the categorical variables, we used a cutoff score of 10 or more for the HDRS (60% sensitivity and 88% specificity for the diagnosis of major depression in patients with AD) ^[9] and psychosis was rated as present or absent.

Demographic variables were analyzed using the Student's t-test or chi squared analysis and Fisher's exact tests. With control subjects of the APOE3/3 genotype as reference, the scores as continuous outcome variables were analyzed using analysis of variance, with and without adjusting for age, education and severity of dementia (using CDR, total recall, or Blessed score). The presence or absence of the dichotomized categorical outcome variables of HDRS and psychosis scores were analyzed using chi squared analysis, Fisher's exact tests, and logistic regression analysis both before and after adjusting for age, education, and severity of dementia. The analysis was repeated among patients using APOE3/3 patients as reference.

Results.

Of 181 eligible subjects, 46 had AD (APOE3/3 [26]; APOE3/4 [20]) and 135 were control subjects (APOE3/3 [98]; APOE3/4 [37]) Table 1. We evaluated ethnicity because the association between APOE4 and AD may differ among races. ^[1] There was no difference in the outcome variables of psychosis and HDRS scores among the different ethnic groups. Patients with AD and the APOE3/4 genotype show higher HDRS and psychosis scores when compared with APOE3/3 control subjects when scores were analyzed both as continuous Table 2 and as categorical Table 3 variables. The difference remained significant after adjusting for age, education, and severity of dementia and was not seen when APOE3/4 control subjects or APOE3/3 patients were compared with APOE3/3 control subjects. When compared with patients with the APOE3/3 genotype, patients with the APOE3/4 genotype show higher HDRS and psychosis scores when the scores were analyzed as continuous variables, and these results approached significance when the scores were analyzed as categorical variables (see Table 2 and Table 3).

Discussion.

Our results suggest that patients with AD and the APOE3/4 genotype may have more psychiatric impairment than their peers with the APOE3/3 genotype or control subjects of either genotype. The at least threefold increase in psychiatric symptomatology does not appear to be due to differences in age, education, or severity of dementia as it persisted after adjusting for them.

APOE4 increases risk of AD either by altering the properties of beta-amyloid or tangle formation, ^[1] and this may predispose patients with the epsilon 4 allele to increased frequency of psychiatric symptoms. Patients with AD and psychosis have an overall trend toward increased numbers of senile plaques (SPs) and tangles throughout the brain with more SPs in the presubiculum and increased neurofibrillary tangles in the middle frontal cortex as compared with AD patients without psychosis. ^[3] Patients with AD and depression show accelerated degeneration of the locus ceruleus and substantia nigra. ^[4] Autopsies on patients with AD and the epsilon 4 allele ^[10] reveal overall increased numbers of plaques and tangles when compared with patients with AD and the APOE3/3 genotype, with significant increases in SPs in the middle frontal, superior temporal, and inferior parietal areas and the CA1 area of the hippocampus that is near the presubiculum, although the substantia nigra and locus ceruleus were not examined. ^[10] The similarities in the neuropathologic profiles of patients with AD and the epsilon 4 allele and patients with AD and psychopathology may help explain the increased frequency of psychiatric symptoms among patients with AD and the epsilon 4 allele.

Neurochemically, patients with AD and psychosis have more noradrenaline in the substantia nigra and less serotonin in the subicular region of the brain. ^[3] Patients with AD and depression have reduced noradrenaline in the neocortex and trends toward reduced levels in all brain areas for both serotonin and noradrenaline. ^[5] The neurochemical pattern of patients with AD and the APOE isoforms has not been studied.

Patients with the epsilon 4 allele have an earlier onset of AD. ^[1] We did not examine disease duration. Thus, the increased frequency of psychiatric symptoms may be explained by a longer duration of AD among our APOE3/4 patients. However, this explanation is unlikely as symptoms such as depression and psychosis are poorly correlated with AD duration. ^[3-5] Also, by adjusting for age and severity of dementia, we effectively adjusted for differences in ages of onset of AD among genotypes. Hence, our results cannot be explained simply by differences in ages of onset or of duration of AD between groups. Alternatively, if we postulate the psychiatric symptoms to be influenced by SP deposition, SP deposition in AD is not correlated with illness duration. ^[1]

Further evaluation of the clinical, neuropathologic, and neurochemical differences in AD with different isoforms of APOE may provide valuable information as to the role of the epsilon 4 allele in the pathogenesis of AD.