PREVALENCE OF AUTISTIC TRAITS IN A COGNITIVELY NORMAL AGING COHORT

Abstract

1.5% of young children have an autism spectrum disorder (ASD) but not included in such prevalence estimates are individuals with a subclinical “broad autism phenotype” (BAP), which has been found in approximately 4% among adults without a family history of ASD. Individuals with BAP are therefore likely present in most studies of cognitive aging, yet the impact of this is unknown. The Autism Spectrum Quotient (AQ) (Baron-Cohen et al), a validated screening questionnaire, was given to members of a cognitive aging cohort to determine the prevalence of individuals scoring in the BAP range, and to characterize their cognitive, behavioral, and personality profiles. The AQ was given during 428 consecutive member visits of the Arizona APOE cohort beginning January 2013, and was completed by 343 members. AQ scores were treated as a dichotomous variable with scores of 23 and higher defining high AQ. Cross-sectional characteristics were compared among groups by using the two-sample t-test / analysis of variance (ANOVA) F-test or Pearson chi-square test. 35 subjects scored in the BAP range of 23 or higher (AQ+) and included a higher proportion of men, reported history of depression and sleep disturbance but no difference in education, marital status, or estimated income (based on zip code) (Table 1). After adjusting for age, sex and education, the AQ+ group performed less well on most measures of mental speed, working memory and arithmetic (Table 2). Though behavioral measures remained within a normal range, mean scores on multiple measures including somatization, anxiety, and depression were higher and a measure of personal warmth was lower in the AQ+ group (Table 3). On the NEO-PI-R five factor personality inventory the AQ+ group exhibited higher neuroticism, and lower extraversion and agreeableness (Table 4). In our cognitively normal aging cohort, approximately 10% of subjects completing the AQ scored in the BAP range, and differ from AQ- subjects on multiple cognitive, psychological and personality measures. The impact of the broad autism phenotype on longitudinal cognitive trajectories and outcomes awaits further study.