Examining clinical correlates, treatment outcomes and mediators in young people with comorbid obsessive–compulsive disorder and autism spectrum disorder

Obsessive–compulsive disorder (OCD) is a common psychiatric condition affecting approximately 0.25–3% of children and adolescents [1, 2]. It is marked by substantial distress and functional impairment, including poor educational, social, and/or family functioning [3, 4]. The prevalence of OCD is considerably higher among young people with autism spectrum disorders (ASD) than among the general population. A meta-analysis of data from 31 studies estimated the prevalence of OCD among youth with ASD to be over 17% [5]. Despite the high rates of comorbidity, little is known about the clinical presentation and treatment response of young people with OCD and ASD (hereafter referred to as ‘OCD + ASD’), compared to those without ASD (hereafter referred to as ‘OCD’).

Only a small number of studies have examined the clinical characteristics of youth with OCD + ASD, and findings have been inconsistent. No differences in OCD symptom severity [6‐8] or insight [8] have been reported in studies comparing youth with OCD + ASD versus OCD. However, higher levels of functional impairment [8, 9] and family accommodation (FA) [8] in youth with OCD + ASD have been observed in some studies. High levels of FA have been consistently reported across the OCD literature [10], yet very few studies have examined FA in youth with ASD [8, 11]. FA has been found to predict treatment outcomes in young people with OCD [12‐14] and those with OCD + ASD [15]. Additionally, youth with OCD + ASD have been found to have higher levels of comorbidity compared to those with OCD [8], specifically higher rates of diagnosed attention deficit hyperactivity disorder or inattention/hyperactivity symptoms [6‐8], externalising disorders [8], peer problems [6] and social and separation anxiety [7]. Those with OCD + ASD have also been found to have lower prosocial behaviors compared to those with OCD [6]. Finally, some studies have found that individuals with OCD + ASD less frequently report sexual obsessions and checking, washing, repeating, games and superstitious compulsions [7, 8], whereas one study found no significant differences in OCD obsessions and compulsions [6]. Of note, most of the studies exploring clinical characteristics had relatively small sample sizes of youth with OCD + ASD (n = 12–70). Whilst Martin and colleagues [9] included a large cohort of young people (n = 7922), which is a strength of this study, our understanding of the wider clinical presentation of youths with OCD + ASD is nevertheless still limited as they used only a single item measure of global functioning. Given the limited available evidence, more research is needed to define and establish the OCD clinical phenotype of this common subgroup of patients. Clinical characterization of this group is important in order to address their needs, including tailoring therapeutic options.

Cognitive behavioral therapy (CBT) is the evidence-based psychological treatment for OCD [16‐18]. Whilst single case studies/series and small trials (n = 1–58) have found adults and young people with OCD + ASD to benefit from CBT for OCD [19], there is indication in studies with small samples (n = 22–25) that the outcomes (OCD symptom reduction) may be poorer compared to typically developing youths with OCD [8, 20]. Yet, the reasons for the difference in outcomes, above and beyond having a diagnosis of ASD, remain unclear. Establishing the factors that account for the poorer OCD treatment outcomes in young people with OCD + ASD can provide a useful focus for clinical adaptations of current evidence-based practices to enhance treatment efficacy and ultimately patient outcomes.

Accordingly, the current study employed the largest sample to date of well-characterised youth with OCD and OCD + ASD to examine and compare these groups on clinical characteristics and treatment outcomes. The study also sought to investigate for the first-time putative mediators accounting for differences in CBT treatment outcomes between the two groups. Based on previous research, we expected higher levels of functional impairment [8, 9], family accommodation [8], and concurrent psychopathology [6‐8] in the OCD + ASD group compared to the OCD group. We also predicted differences in OCD symptom dimensions between the groups, possibly the OCD + ASD group reporting less sexual obsessions and checking, washing, repeating, games and superstitious compulsions [7, 8]. Second, given the existing evidence thus far, it was hypothesized that youth with OCD + ASD would have poorer treatment outcomes than those with OCD at end of treatment [8, 20]. Finally, given the lack of available studies, no specific hypotheses were advanced with respect to variables that may mediate the differential response to treatment in these groups.

To our knowledge, this is the largest study to date comparing the clinical phenotype and treatment outcomes for young people with OCD with and without comorbid ASD using a broad range of outcome measures and the first one to examine factors that may explain differences in treatment outcomes, above and beyond having a diagnosis of ASD.

Regarding the clinical presentation, both groups showed similar levels of overall OCD severity, including severity of obsessions and compulsions. Together with previous studies [6‐8], such findings suggest that OCD affects individuals with ASD to a similar degree to youth without ASD, with both groups endorsing equally time-consuming, distressing and interfering obsessions and compulsions. However, a number of differences in clinical characteristics were observed in this study. Specifically, those with OCD + ASD presented with higher levels of family accommodation, poorer insight into their OCD, worse global functioning, higher levels of comorbid psychopathology, in particular hyperactivity, emotional and peer problems, lower levels of prosocial behavior and were more likely to be prescribed medication than those with OCD without ASD.

Results regarding family accommodation are consistent with the high rates of parental accommodation of OCD symptoms as well as parental accommodation of restrictive and repetitive behaviors (RRBs) reported previously in samples of youth with ASD and those with comorbid OCD and ASD [8, 11]. The finding that people with ASD + OCD had higher levels of family accommodation might be due to parents accommodating both RRBs and OCD symptoms. Alternatively, given parents may be accustomed to accommodating their child's ASD symptoms, it makes it easier for them to accommodate OCD. Equally, parents’ knowledge and skills to distinguish RRBs from OCD compulsions may contribute to higher parental accommodation. Family accommodation may be an important avenue to address in future research and in clinical settings; for example, a better understanding of the factors that drive or maintain family accommodation in parents of children with OCD + ASD may help address the needs and difficulties of youth with this comorbidity. Albeit preliminary, the findings highlight avenues for potential additional parent-based intervention and guidance for how to address RRBs versus OCD rituals.

It is noteworthy that youth with OCD + ASD presented with poorer insight into their OCD symptoms compared to youth without ASD. This novel finding contrasts with that of Griffiths and colleagues [8] who reported no significant difference in insight despite using the same measure. Due to the smaller sample (n = 25 in each group) their study may have been underpowered to detect small effect sizes as were observed for insight in the current study. Whilst in need of replication, this finding might be associated with cognitive rigidity and difficulties with metacognition often reported in ASD, leading to difficulties in acknowledging the excessive and unreasonable nature of their obsessions and compulsions. Further research is needed to explore insight among those presenting with OCD and ASD. Consistent with previous studies [8, 9], the OCD + ASD group had greater functional impairment compared to the OCD group. However, what remains to be examined is how much of the impairment is due to OCD versus ASD as the measure used in this study only captured overall functional impairment. It is likely that there is a cumulative burden of having co-occurring OCD and ASD, which may be further compounded by the elevated levels of other psychopathological difficulties (e.g. attentional problems) that were found in the current study. It would be important for future research to investigate functional impairment among youth with co-morbid OCD and ASD in greater depth, for example, using assessment tools that allow to delineate OCD- versus ASD-related impairment. Whilst Griffiths and colleagues [8] measured OCD-related functional impairment, they did not examine ASD-related impairment. Knowing how much each of these diagnoses contribute to greater functional impairment may give ideas of avenues of support for those with a dual diagnosis. Additionally, it would be interesting to explore the impact of specific autistic symptoms on OCD treatment outcomes and functional impairment, including the severity of impairments in social communication, and repetitive and restricted behaviors.

Like previous studies, we found higher rates of hyperactivity symptoms [6‐8], peer problems [6], emotional problems [7] and lower prosocial behaviors [6] in young people with OCD + ASD compared to those with OCD. Contrary to previous research [8], we failed to find differences in conduct problems. The latter may be characteristic of the sample seen in the clinic and would need to be replicated to ascertain whether this is a clinical characteristic of those with OCD + ASD. Overall, these results are unsurprising as we know young people with ASD are likely to have high levels of psychopathology [35]. The finding that youth of OCD + ASD had higher peer problems and lower prosocial behaviour is characteristic of the ASD profile [36]. These findings have important clinical implications in that they highlight that youth with OCD + ASD will present with additional difficulties; however, the finding that this does not impact on CBT outcomes (discussed below) is important as it highlights that this may not need to be specifically addressed to treat OCD symptoms.

Finally, the reason for increased use of medication in those with a dual diagnosis is unclear since both groups showed a comparable severity of OCD symptoms. It is possible that those with OCD + ASD are more likely to be prescribed medication due to the greater functional impairment or due to a poor or partial response to previous courses of CBT. Alternatively, it is possible that clinicians may be more inclined to offer medication to those presenting with this dual diagnosis [9]. Further research is warranted to clarify the treatment history of those with comorbid OCD + ASD which may help understand reasons for increased medication use and its potential impact on CBT outcomes.

Consistent with a previous investigation [8], we found young people with OCD + ASD reported fewer games/superstitious compulsions compared to the OCD group. This may be due to young people with ASD struggling with abstract thinking such as that involved in superstitious related compulsions. The current study also found the OCD + ASD group reported fewer aggressive and religious obsessions, which has not been reported before. This is at odds with studies that have found no differences in symptom dimensions between the groups [6] or differences found in other dimensions, namely sexual obsessions, checking, washing and repeating compulsions [7]. It is difficult to ascertain what may account for differences across studies as they all used the same measure (CY-BOCS). Given the inconsistencies in the literature, our findings are in need of replication in future research.

Both groups showed a reduction in OCD symptoms after treatment. However, as hypothesized, and in line with previous research, the OCD + ASD group showed less improvement at the end of treatment compared to the OCD group [20]. This was echoed with lower rates of response and remission rates in the OCD + ASD group compared to the OCD group. The response rate in the OCD + ASD group found in this study (51%) was broadly comparable to those reported in the adult and pediatric literature, whereby rates range from 45% [15] to 66% [37]. The remission rates in the current study (31%) were also found to be in line with previous studies, ranging from 9% [20] to 52% [37]. Our findings provide additional evidence in a naturalistic setting in support of CBT for OCD as an effective treatment for youth people with OCD + ASD. In addition, this study supports previous research showing a reduced treatment response among youth with ASD + OCD and highlights the continued need for rigorous research to optimize CBT treatment for those with this common dual diagnosis [8, 9, 20]. Several approaches have been suggested to enhance outcomes in this challenging group, including specific ASD-adaptations, increasing the quantity and intensity of CBT sessions and delivering home-based sessions [11, 38‐40].

Multiple mediators were examined in an attempt to explore which factors, over and above having a diagnosis of ASD, may account for the group difference in treatment outcomes. Global functional impairment and being on medication were found to partly account for the poorer treatment outcomes in youth with OCD + ASD compared to those with OCD only. Indeed, we found that functional impairment mediated 19% of the effect of a dual diagnosis on CBT outcomes; an even higher proportion of the difference in outcomes was mediated by being on medication (31%). This is a novel finding that encourages more clinical consideration and research. It is possible that those with co-occurring OCD + ASD may have a cumulative burden of having both OCD and ASD, causing greater functional impairment and resulting in poorer CBT outcomes. This raises the question of whether promoting functioning (e.g. optimizing educational support, facilitating engagement in leisure/social activities) could provide a helpful platform for CBT to enhance outcomes. Being on medication also mediated the differential treatment outcomes. It is possible medication may be a proxy for a measure of a complexity (e.g. poorer insight, treatment resistance) that was not captured in this study and that interferes with treatment outcome. Taken together, our results provide preliminary support in favor of further examination of functional impairment and pharmacological approaches for young people with OCD + ASD.

Surprisingly, family accommodation did not account for differential treatment outcomes between those with OCD versus OCD + ASD, despite the significant group differences in family accommodation and previous studies showing family accommodation to predict CBT outcomes [12‐15]. The current findings may reflect that family accommodation was successfully addressed in treatment; a pre- and post-treatment measure of family accommodation would have been helpful to ascertain this and could be considered in future research. Insight was significantly poorer in the OCD + ASD group and has in previous studies been found to predict treatment outcomes [41]. However, the fact that insight did not account for differential treatment outcomes may be due to insight being measured on a single-item scale of 0–4, so there potentially was not sufficient variance to fully test this association. Future research should utilize more detailed measures of insight, such as the Brown Belief Assessment Scale (BABS) [42]. Alternatively, insight could have improved over the course of treatment as has been found in previous studies [41]. Limited by sample size, our analyses were unpowered to test treatment group by sex interactions in response to treatment. Future studies should properly examine the role of sex in this effect. It is interesting to find that concurrent psychopathology did not predict CBT outcomes. This  highlights that having ASD may  account for compromised outcomes as opposed to comorbidities, indicating that treatment needs to be modified to support this group with the aspects of treatment they may struggle with, e.g. extended psychoeducation on emotions for those with alexithymia.

The present findings have important implications. Clarifying the association between ASD, functional impairment and medication prescription could be an important direction for future research. The finding that functional impairment and being on medication emerged as mediators for the difference in CBT outcomes for youth with OCD + ASD also suggests that these may be important avenues for improving outcomes for this patient group. A better understanding of such constructs will help refine our OCD treatment protocols for this patient group, for instance, introducing elements to better address impairments associated to ASD (e.g. sensory sensitivity, emotion psychoeducation) or dropping other elements that do not contribute to change. Indeed, the treatment focus is typically on reducing OCD symptom severity; however, the results may highlight the potential value of improving functioning to enhance treatment outcome.

There are several limitations to the current study. The study was conducted in a National and Specialist clinic, so there may be a bias in referrals where the OCD is the most prominent issue and, therefore, we may not have seen a full range of young people on the autism spectrum; most notably, we did not have young people with global learning disabilities. We may have also had a higher proportion of young people with more severe and treatment resistant OCD referred to us compared to other studies, which may have impacted our results. Only a subgroup of the sample had treatment outcome data available. Although those with and without treatment data did not differ on many clinical characteristics (e.g. age of onset,  CY-BOCS scores etc),  those with data had lower levels of family accommodation, conduct problems and functional impairment, representing possibly a less severe subgroup of patients. The lack of information on medication for 8% of young people and the type of medication (SSRIs or other) for 35% also limits our conclusions regarding the role of medication in mediating outcomes. Another limitation is the lack of a control group (e.g. OCD with another comorbidity) which did not allow us to explore if the results were specific to having an ASD comorbidity versus having any comorbidity. This would be important to include in future research. The lack of structured diagnostic assessments of comorbidities in this study is a further limitation. Additionally, the measure of family accommodation measure was self-report and, therefore, it may be difficult to establish if it captured family accommodation of OCD symptoms only or whether parents included ASD symptoms too. The measure of functional impairment was a clinician-rated global measure, which did not allow us to distinguish disorder-specific impairment nor to get parent or child reports of impairment. Finally, OCD assessments were carried out by expert clinicians specialised in OCD and experienced in assessing and disentangling ASD ego-syntonic RRBs from ego-dystonic OCD compulsions; however, we cannot completely rule out that CY-BOCS may have been influenced by RRBs. A measure such as the ASD modified version of the CY-BOCS [43] could be used in future studies to help address this potential limitation.

Despite these limitations, the present study remains the largest study to date comparing the clinical characteristics and treatment outcomes for youth with OCD + ASD compared to those without ASD. To the best of our knowledge, this is also the first study to examine putative mediators accounting for the difference in treatment outcomes between youth with OCD + ASD versus OCD only. The results highlight that the presentation of OCD in those with comorbid OCD and ASD does not significantly differ from those without ASD in terms of symptom severity. However, those with comorbid OCD + ASD present with higher levels of family accommodation, functional impairment and poorer insight and are more likely to be on medication. Those with OCD + ASD also present poorer treatment outcomes compared to youth with OCD, with functional impairment and medication status emerging as significant mediators of outcome differences and as potential avenues in future research on the development and adaptation of treatment for those with comorbid OCD and ASD.