Medication Use Among a Cohort of Adolescents with An Autism Spectrum Disorder

Background:  While there is a current dearth in the literature evaluating psychotropic medication use among individuals with autism spectrum disorders (ASDs), medication use within this group has continued to rise (Aman, 2005). Psychotropic medications are being prescribed to allay symptoms with little empirical support (Oswald, 2007). The few studies that have examined psychotropic medication use of individuals with an ASD found significant relationships between medication use and demographic variables such as, intellectual disability, parent education and classroom placement (Aman, 2005).  However, there are two major gaps in this literature: 1) the relationships of medication use with demographic variables have only been examined cross-sectionally, thus there is limited knowledge of whether these relationships persist over time; and 2) little is known about other variables such as behavioral symptoms that may predict medication use within this population. 

Objectives:  1. To explore the relationship between demographic variables and psychotropic medication use among adolescents with an ASD at 3 time points. 2. To analyze behavioral predictors of psychotropic medication use among adolescents with an ASD at 3 time points 

Methods:  The sample [T1N1 = 87 (M = 12.5 years, SD=1.15), T2N2 = 62 (M =14.65  years, SD = 1.10), T3N2 = 63(M = 17.43 years, SD = 1.09)] is comprised of adolescents who were recruited through consecutive referrals for possible autism at age 2 to clinics in North Carolina and Chicago, and have been followed longitudinally as part of a larger study. Chi-square tests of independence were performed to examine the relationships between psychotropic medication use and demographic variables. Logistic regressions were used to analyze behavioral predictors of psychotropic medication use. 

Results:  At Time 1 and Time 2 a significant relationship was found between psychotropic medication use and classroom placement [T1: x²(2, N=87) = 10.89, p < .005, T2: x²(2, N = 62) = 6.14, p < .05]. At Time 1 and Time 3 a significant relationship was found between psychotropic medication use and diagnosis (Autism versus PDD-NOS) [T1: x²(1, N = 87) = 4.185, p<.05, T3: x²(1, N = 63) = 6.51]. At Time 1 and Time 3 a significant relationship was found between education of caregiver and psychotropic medication use [T1: x²(3, N = 87) = 11.81, p<.05, T3: x²(3, N = 63) = 7.92, p>.05]. Logistic regression models indicated varied relationships between behavioral predictors and medication use at each time point. Time 1 indicated that lethargy and stereotypy symptoms were significant predictors of psychotropic medication use (W=6.79, p>.01, W=5.49, p>.05). Time 2 indicated that irritability was a significant predictor of medication use (W=4.69, p>.05) Time 3 indicated no significant behavioral predictors of psychotropic medication use. 

Conclusions: The discussed analyses suggest that significant relationships exist between psychotropic medication use and demographic and behavioral variables. However, these relationships appear to shift over time rather than remaining constant. This research provides the foundation for future analysis of the patterns of medication use over time within a single cohort of individuals to assess whether stable relationships exist.