Troubled adolescence and beyond; the characteristics of addiction in autism

Background:  Addiction, or Substance Use Disorder (SUD), among adults with Autism Spectrum Disorders (ASD) has never been studied before, contrary to SUD in ADHD. However, we know from clinical practice that comorbid SUD in ASD can present with a considerable impairment. Distinguishing between ASD and ADHD can be challenging, but becomes increasingly so in case of comorbid SUD. The approach to SUD in ASD is hampered by a lack of knowledge about this comorbid condition.

Objectives: The main objective of this study was to gain insight into the aspects of comorbid SUD in ASD on three levels: an observable phenotypical level, an underlying endophenotypical level, and a genetic level. A second objective was to determine how ASD differs from ADHD in adult patients on these three levels, with and without comorbid SUD.

Methods:  We studied 130 patients with ASD (n=75) or ADHD (n=55) with or without SUD. Firstly, on the phenotypical level, we looked at risk factors for developing SUD, prevalence, and outcome. Furthermore, we studied personality profiles with the Temperament and Character Inventory and autism spectrum profiles with the Autism Spectrum Quotient (AQ). Secondly, on the underlying neuropsychological (endophenotypical) level, we assembled data on intelligence profiles (WAIS III), attention tasks and executive functioning tasks. Thirdly, on the genetic level we studied five candidate genes in the different patient groups.

Results:  The results show that SUD is more common in ADHD than in ASD, but that the prevalence of SUD in the ASD group is comparable to the prevalence of SUD in other psychiatric patients. Early onset of smoking, adverse family history, and parental SUD, increase the risk on SUD significantly. On a phenotypical level, self-report questionnaires on personality and autistic symptoms suggest that the ASD-with-SUD subgroup presents with a better social orientation than the subgroup ASD-without-SUD. This effect is not seen in the ADHD patients. On the endophenotypical level, however, the ASD-with-SUD subgroup shows significantly more cognitive impairment than the ASD-without-SUD subgroup, whereas this SUD effect was not seen in the ADHD group. In our study there were also genetic differences between the ASD and ADHD groups in the polymorphisms of three of the five candidate genes we analyzed, independent of the SUD status. However, the numbers were small, and replication of the data is necessary

Conclusions:  The results show that comorbidity of SUD with ASD is less common than in ADHD, but that the consequences are at least as severe. Comorbid SUD is not always associated with a negative outcome, at least in the perception of the individual on a phenotypical level. Assessment of adults with ASD and SUD can be misleading when the neuropsychological functioning is not also taken into consideration. Finally, in the absence of SUD, ASD and ADHD in adults share many features on the (endo-) phenotypical level, but may be distinguishable at a genetic level. In the presence of SUD, the differences between ASD and ADHD are more pronounced.