Abstract
We compared the parental age at birth of patients with Alzheimer disease (AD) with that of cognitively healthy control subjects. Within 206 carefully diagnosed AD patients, two groups were distinguished according to the likelihood of carrying a major gene for AD (MGAD). This likelihood was calculated by applying a Bayesian approach which incorporates data on aggregation of the disease, age at onset, and "censoring" ages within the family. All AD patients were ranked by MGAD probability. According to the sample's quartiles, two subgroups were defined representing the 52 individuals with the lowest and the 52 with the highest MGAD probability. Age at onset of dementia, education, and apolipoprotein E epsilon4 allele frequencies were not statistically different between the two groups. Fathers of patients with a low MGAD probability were significantly older (35.7+/-8.1 years) than fathers of both other groups (high MGAD probability 31.3+/-6.9 years, P=0.004; controls 32.6+/-6.8 years, P=0.04, n=50). The differences for mothers were less pronounced and not statistically significant. These findings suggest that increased paternal age is a risk factor for AD in the absence of a major gene, whereas increased maternal age and AD are associated only weakly and independently of genetic disposition.