Autistic children may show irritability for various reasons, such as changes in routine or mood disorder [2, 10, 18]. Some of these children have severe irritability and do not respond well to FDA-approved medications like risperidone and aripiprazole. They are defined as drug refractory if they either (a) fail to respond to risperidone and aripiprazole treatment or (b) fail to respond to at least three psychotropic drugs at adequate doses and durations [14]. In our clinical sample, 8.2% of autistic children were found to have DRI. The average age of patients in the DRI group was 14 years and the risk of DRI increased with age. Additionally, autistic children who required very substantial support, had language impairment, co-occurring anxiety disorder, sleep difficulties, GIS and dental problems were at a significantly higher risk for DRI.
The prevalence of DRI in our study was lower than that reported by Adler et al. [14], who found a frequency of 39.5% among 135 patients with autism at their tertiary care clinic. This discrepancy may be due to the differences in the study populations. Adler et al. [14] used a more severe and selective sample of autistic children, whereas we used a more representative and diverse sample. To our knowledge, only a few studies have used the same definition of DRI as Adler et al. [14], but none of them have examined the clinical features and associated factors of this group [19, 20]. Future studies that investigate the characteristics of this group prospectively will provide important information for their diagnosis and management.
Age
Age may influence irritability in autistic children. Adolescence is a critical period for severe neuropsychiatric disorders such as schizophrenia, bipolar disorder, and catatonia [21]. Autistic adolescents also have more ADHD, anxiety disorders, destructive behaviors, bipolar disorder, depression, and psychotic disorders than younger children [22, 23]. Previous studies have shown that irritability worsens with age in autistic patients with severe symptoms or low intelligence [24], and that older age raises the risk of rehospitalization after discharge from an inpatient service [25]. In contrast, younger age has been associated with self-injurious behavior in two large community samples of autistic children [26]. Carroll et al. [27] found that autistic children with aggression were slightly younger than those without aggression (7.29 vs 8.61), but most of them responded well to risperidone treatment. Adler et al. [14] reported that DRI was more common in autistic individuals aged 12 years or older. In our study, the DRI group had a mean age of 14 years, which was significantly higher than the non-DRI group. Moreover, the likelihood of DRI increased with age. The sample characteristics may explain the differences in the findings. However, older age could be a major risk factor for DRI.
Sex, maternal education, family income
The sex ratio of the DRI and non-DRI groups were similar in our study. Studies in autism did not find significant sex differences in challenging behaviors [3, 28]. Maternal education and family income levels were significantly lower in the DRI group, but these differences were not statistically significant in the regression analysis. The association between socioeconomic factors and irritability in autism is unclear and may depend on other variables such as access to services or parental stress [18, 26].
Autism severity, intellectual and language impairment
As expected, most patients in the DRI group (94.3%) needed very substantial support, which increased their DRI risk. Language impairment was a significant risk factor for DRI, while ID was not. This may be because the majority of patients with ID also had language impairment. Previous studies have linked low IQ, language impairment, and severe autistic traits to poor outcomes; aggression, self-injurious behavior, and challenging behavior in autism [16, 27, 29–31]. Moreover, non-speaking children and those with low adaptive skills had more self-injurious behavior; impulsivity and social interaction deficits also increased chronic self-injurious behavior risk [32]. On the other hand, Williams et al. [33] found that verbal and non-verbal autistic inpatients had similar levels of challenging behavior, but the verbal group had more externalizing problems. However, this effect was reduced when adjusted for verbal abilities. A low adaptive score was still related to the problem behaviors severity. The findings may vary due to different sampling and assessment methods, but as autism morbidity increases, so do irritability and problem behaviors. Language impairment may affect communication and emotional regulation in autistic children, leading to increased irritability and frustration.
Co-occurring psychiatric disorders
Psychiatric disorders occur in approximately 70% of autistic people, 50% of whom have multiple psychiatric disorders [22, 34]. The most common co-occurring psychiatric disorders in autism are ADHD, anxiety and depressive disorders, oppositional defiant disorders, and CD [22]. Psychiatric disorders were found to play a role in the development of problem behaviors in two separate studies of autistic patients treated in inpatient services [5]. In our study, DRI group had more co-occurring psychiatric disorders and most of the DRI group had CD diagnosis. The coexistence of SIB and behavioral factors (aggression, hyperactivity, anxiety, and mood disturbances) has been proposed to be caused by a common etiology of impulse control failure [26]. As most children with DRI in our study were in adolescence and had ID and language impairments, the psychopathologies that increased during this period could have been classified as irritability.
The co-occurrence of anxiety disorders and sleep difficulties significantly increased the risk of DRI. Previous studies had shown that SIB is associated with hyperactivity, anxiety, and mood symptoms in autistic children [3, 27, 31, 35]. Autistic children with sleep difficulties have increased problem behaviors, and sleep difficulties are the most related factors among the other contributors [36]. Similarly, sleep difficulties are associated with self-injurious behavior [26] and self-injurious behavior decreases with improved sleep [37]. These findings emphasize the importance of conducting a comprehensive evaluation of co-occurring psychiatric disorders in autistic children with irritability. However, psychiatric diagnosis in autistic children is difficult, particularly when there are deficits in mental and verbal skills [34]. For instance, the high prevalence of CD in our sample could be attributed to clinicians' evaluation of problematic behaviors within this diagnosis.
Co-occurring physical conditions
Physical conditions are known to play a role in the development of behavioral problems in autistic children [38]. Physical conditions can coexist, and an increase in physical symptoms is associated with severe problem behaviors [39]. Among the co-occurring physical conditions in our sample, the DRI group had significantly more epilepsy, GIS and dental problems. Similarly, one-third of autistic inpatients in Guinchat et al.’s [5] study had a treatable physical condition. Autistic children, particularly those with language impairment, may exhibit problem behaviors and self-injurious behavior in response to pain and discomfort caused by co-occurring physical conditions [40].
Our study identified GIS and dental problems as risk factors for DRI. Previous studies have shown that autistic children have more GIS problems than neurotypical children [41], which may contribute to increased irritability and anxiety symptoms and reduced pharmacotherapy response in some autistic patients [42, 43]. Guinchat et al. [5] reported that GIS problems are prevalent and treatable co-existing medical conditions that can cause pain and irritability in hospitalized autistic children. Similarly, dental problems can cause pain and irritability, but autistic children have poorer oral health and less access to dental care than neurotypical children [44, 45]. Therefore, our findings emphasize the importance of thorough evaluation of co-occurring physical conditions in the treatment of the DRI group.
One strength of our study is the inclusion of multiple centers and the use of a large sample of autistic outpatients. To the best of our knowledge, only a few studies [2, 19, 46] have been conducted since Adler et al. [14] defined DRI. However, some significant limitations of our study should be considered when evaluating our findings. The study design, that is, retrospective chart review to identify DRI-related factors, is its most significant limitation. Another important limitation is the absence of standardized measures or methods for evaluating physical conditions and psychiatric disorders. Third, there may be selection bias because our sample was based on an outpatient group. As a result, our findings cannot be extended to all autistic children. Finally, due to the cross-sectional nature of our study, we were unable to assess the causal relationships between DRI and putative risk factors.
Despite these limitations, our findings indicate that DRI is common in autistic children and is associated with several risk factors such as age, autism severity, language impairment, anxiety disorder, sleep difficulties, GIS and dental problems. These findings suggest that DRI may be caused or worsened by a combination of developmental, psychological, and physical factors that affect communication and emotional regulation in autistic children. Therefore, it is important to conduct a comprehensive evaluation of co-occurring conditions and provide appropriate interventions for autistic children with DRI. Future longitudinal studies are needed to investigate the development and treatment of DRI in this population.