The Influence of Gender and Speech on Autism Symptomatology as Measured by the ADI-R

Background: Autistic Disorders (AD) is characterized by a higher prevalence in males with a male/female ratio of 4:1. The ADI-R is the gold standard instrument used for research and clinical diagnosis of AD (Lord, 1995).  Individuals are diagnosed with AD if they meet criteria in all three domains of the ADI-R: social, communication, and repetitive Stereotyped Behaviors (RSB). Symptom differences between males and females diagnosed with AD by the ADI-R are not always obvious.  This study examines differences between the ADI-R domains in males and females with AD.

Objectives: To evaluate symptom differences in persons diagnosed with AD using the ADI-R domains in males and females who are verbal, verbal with difficulty (VWD), and non verbal.

Methods: The study included 266 subjects (207 males, 59 females) with an ADI-R diagnosis of AD.  Level of speech was based on ADI-R criteria for verbal vs. nonverbal.  Multivariate ANOVA used to evaluate the effect of gender and level of speech for the Social (B1, B2, B3, B4), Communication (C1, CV2, CV3, C4) and RSB (D1, D2, D3, D4) subdomains and domain total scores. For the Communication, subdomains CV2 and CV3 were omitted for nonverbal or VWD individuals.

Results: The omnibus multivariate test on the ADI-R Social domain showed there were no overall gender effects. However, there was an effect for level of speech, (F=3.048, p=0.002). Between subjects analysis showed a main effect for level of speech and the Social domain and 2 of the 4 subdomains: B3 (F_(1,265)=4.491, p=0.012), B4 (F_(1,265)=10.416, p=0.000) and Social total score (F_(1,265)=7.080, p=0.001). Post-hoc comparisons using Scheffe indicate there was a significant difference between verbal and nonverbal groups on B3 (lack of shared enjoyment), B4 (lack of socio-emotional reciprocity), and overall social score. The verbal group also differed significantly from the VWD group on B4 (lack of socio-emotional reciprocity). In the Communication domain, there was significant gender difference for VWD group on C4 (F_(1,37)=5.655, p=.023) with females exhibiting less deficits in social play.  The verbal group showed significant differences on the CV2 (conversation skills) subdomain, F_(1,174)=3.898, p=0.050), with females exhibiting less deficits. There were no gender differences in the Nonverbal group. The omnibus multivariate test on the ADI-R RSB domain indicated no overall gender effects. There was an effect for level of speech, (F=8.659, p=0.000). The between subjects analysis showed a main effect for level of speech and the RSB domain and the 3 of the 4 subdomains: D1 (F_(1,265)=5.840, p=0.003), D2 (F_(1,265)=20.129, p=0.000), D4 (F_(1,265)=5.488, p=0.005) and RSB total score (F_(1,265)=6.410, p=0.002). Additionally, there were significant differences between the verbal and nonverbal groups on D1 (preoccupations), D2 (routines), D4 (preoccupations with objects) and the overall RSB score.

Conclusions: The diagnosis of an ASD by ADI-R represents a complex pattern of behaviors with males and females achieving diagnosis frequently by different sets of symptoms, influenced strongly by speech.  This profile of symptoms more prevalent in females than males needs to be replicated in a larger sample and has implications for shaping early intervention strategies in males and females.