The aim of this study was to determine the impact on risk perceptions of disclosing genetic test results used to estimate the risk of Alzheimer's disease (AD). Adult children (n = 149) of people with AD were randomized to one of two groups--Intervention group: lifetime risk estimates of AD based on age, gender, family history, and Apolipoprotein E (APOE) genotype;
Control group: lifetime risk estimates of AD based on the same risk factors excluding APOE genotype. Perceptions of personal risk (PPR) for AD were assessed six weeks after risk assessments. PPR were correlated with actual lifetime risk estimates (r = 0.501; p < 0.0001). After controlling for lifetime risks communicated to participants, age, and number of affected relatives, PPR scores among those with an epsilon4-positive test result (the test result associated with increased AD susceptibility) (adjusted mean: 3.4 (SD: 0.7)) were not different from the PPR scores in the CONTROL GROUP (adjusted mean: 3.4 (SD: 0.7) (F1,91= 1.98; p = 0.162). Again, controlling for lifetime risk estimates, age, and number of affected relatives, the PPR score of those receiving an epsilon4-negative test result was significantly lower (adjusted mean: 3.1 (SD: 0.8)) than those in the CONTROL GROUP (adjusted mean: 3.4 (SD: 0.7) (F1,95 = 6.23; p = 0.014). Perceptions of risk of developing AD are influenced by genetic test disclosure in those receiving epsilon4-negative, but not those receiving epsilon4-positive test results. Despite the reduced perceptions of risk in the former group, there was no evidence of false reassurance (i.e., perceiving risks as equal to or lower than population risks of AD), although this possibility should be assessed in other testing contexts.