Restricted Repetitive Behaviors and Interests and the Female Autism Phenotype: An Autism Speaks Autism Treatment Network (AS-ATN) Study.

Background: Autism Spectrum Disorder (ASD) continues to be diagnosed at a greater rate in males than in females with current prevalence rates estimated at 4-5:1 (males; CDC, 2016). Although clinical characteristics of ASD manifest differently in males and females (e.g. Bölte et al, 2011; Carter et al, 2007), few studies have investigated sex differences in ASD’s two core domains: social–communication and restricted, repetitive behaviors, interests and activities (RRBs). Additionally, compared to the social-communication domain, considerably less is known about the cause, development and impact of RRBs in children with ASD, including possible sex differences.

Objectives: The aim of this study is to examine potential differences in clinically identified (ADOS) RRB symptoms using a large sample of age and IQ-matched females and males with ASD.

Methods: Data were extracted from the Autism Speaks Autism Treatment Network (ATN) registry. The study sample included 513 females and 513 males with an ASD diagnosis (Autistic Disorder, Asperger’s Disorder) propensity score-matched 1:1 on age in 24mo intervals and IQ (≥70 vs. <70). Age (<6 yrs, 6-<12 yrs), race, ethnicity, IQ, caregiver education, evidence of lost skills, and autism diagnosis were tabulated overall and by sex. Fisher’s exact tests evaluated if these characteristics differ by sex. ADOS RRB domain calibrated severity score and its individual items were described overall and by sex in four age-IQ groups: 1. IQ≥70 and < 6 yrs, 2. IQ≥70 and 6-<12 yrs, 3. IQ<70 and <6 yrs, 4. IQ<70 and 6-<12 yrs. T-tests assessed whether RRB differs by sex within each of these groups. Means and standard deviations were reported. RRB individual item scores (1 or 2 vs. 0) were tabulated by sex, and Fisher’s exact tests determined if these items differ by sex.

Results: Demographic characteristics, evidence of lost skills, or ASD diagnosis did not differ significantly by sex (see Table 1). RRB total score differed by sex (p=0.022) in the IQ<70-age 6-<12 group, with females reporting lower RRB than males, but no significant sex differences were found for other groups. In the IQ≥70-age<6 group females had a higher percentage of abnormal hand/finger/other complex mannerisms (p=0.038) and abnormal intonation of vocalizations or verbalizations (p=0.049). In the IQ≥70-age 6-<12 group females also had a higher percentage of abnormal hand/finger/other complex mannerisms (p=0.037; see Table 2). No item scores differed significantly by sex in either of the IQ<70 groups.

Conclusions:  Findings from the largest dataset to date of matched young females and males with ASD indicate that there are sex differences in overall RRBs and in particular sub-domains for selected groups. These differences add to the growing literature regarding the need for sex-specific algorithms in autism diagnostic instruments which may then help improve our ability to identify misdiagnosed and under-diagnosed females with ASD.