Significant Gender X Autism Status Interactions in Middle Childhood On Variables Related to Difficult Temperament

Background: Temperament is an understudied construct in the autism literature and in the handful studies to date few researchers have investigated gender differences. Furthermore, in the childhood temperament literature most researchers have relied on instruments designed to measure either Thomas and Chess or Rothbart’s conceptualizations of temperament.

Objectives: This study seeks to determine whether temperament variables measured with the new Inventory of Children’s Individual Differences- Short Form (ICID; Deal, Halverson, Martin, Victor, & Baker, 2005) differentiate children with ASD from typical children, as well as determine whether there are any significant ASD status X gender interactions on temperament variables in middle childhood.

 Methods: Here we report a 2 (ASD versus typical)* X 2 (Gender) age controlled MANCOVA comparing  113 ASD (93 male; 20 female) and 372 control children (163 male; 209 female) between the ages of 8 and 12 (Middle Childhood; MC).  Data from ASD children was collected on-line via the Interactive Autism Network and data from typical children was taken from the norming sample data set.   

Results: Pillai’s Trace statistics indicated significant ASD X gender interactions [F(1, 15) = 2.032, p < .05], significant gender differences [F(1, 15) = 1.321, p < .05], and significant differences between ASD and typical children: [F (15,1) = 39.574, p < .001].  Follow up Wesch’s F tests indicated significant gender X diagnostic status interactions for antagonism [(F(1, 426)= 5.06, p < .05], consideration [(F(1, 426)= 4.02, p < .05], intelligence [(F(1, 426)= 4.04, p < .05], negative emotion (anger) [(F(1, 426)= 12.81, p < .05], and strong willed [(F(1, 426)= 4.27, p < .05].  Diagnostic group differences were found with ASD children scoring lower on achievement [F(1, 426) = 118.21, p < .001], compliance [F(1, 426) = 151.18, p < .001], consideration [F(1, 426) = 227.39, p < .001], intelligence [F(1, 426) = 43.12, p < .001], openness [F(1, 142.54) = 203.57, p < .001], positive emotion [F(1, 426) = 66.30, p < .001], and sociability [F(1, 426) = 355.95, p < .001] and higher on antagonism [F(1, 426) = 44.08, p < .001], distractible [F(1, 426) = 15.59, p < .001], fear/insecure [F(1, 426) = 44.50, p < .001], negative emotion (anger) F(1, 426) = 92.20, p < .001]shy [F(1, 426) = 259.24, p < .001], and strong willed [F(1, 426) = 61.88, p < .001]. No group differences were found for activity level [F(1, 426) = 1.71, p = .192] or organized [F(1, 426 ) = 3.03, p =.082].

Conclusions: These results indicate that in MC females with ASD are perceived by their parents as having more difficult temperaments and as being less intellectually capable than boys with ASD. Furthermore, there were widespread temperament differences between children with ASD and typical children with largest differences found on sociability and shyness. Future research may seek to determine whether the sources of the gender X diagnostic status interactions are due to true behavioral differences using more objective methods.

*This project is funded by a grant from the Organization for Autism Research