Depressive and bipolar disorders in persons with intellectual disability and low-functioning autism spectrum disorder. The development and first validation of a new diagnostic tool (SPADD-M)

Background: Current literature reports the frequent co-occurrence of psychiatric disorders (PDs) in adults with Intellectual and Developmental Disorders (IDD), such as Intellectual Disability (ID) and Low-Functioning Autism Spectrum Disorder (LF-ASD). Unfortunately, the adaptations of diagnostic criteria for PDs as well as the standardization of the psychiatric diagnostic process in this population still represent an unmet need, with relevant implications for both the management of mental health issues and the research in this field. In fact, the assessment of PDs in people with IDD requires appropriate modifications in respect to the general population, to adjust for cognitive dysfunctions, language and communication limitations, sensory impairments, skill deficits, impairment of adaptive behavior, and physical disabilities. Moreover, research and clinical experience highlight the interpretation of problem behaviors (PBs) to interfere with the psychiatric diagnostic process. Among PDs, depressive and bipolar disorders (DDs and BDs, respectively), seem to be very common and associated to a variety of behavioral, medical and treatment issues in these populations. Indeed, the prevalence rates of DDs and BDs in large ID population-based studies are estimated to be about 5% and 2.5%, respectively. The prevalence rates are at least doubled in the overall ASD population and even higher rates can be found in clinical studies. These data indicate the co-occurrence of mood disorders (MDs) in intellectually disabled persons to be a relevant clinical issue, but some difficulties may arise as far as a certain number of patients, mostly those with lower cognitive and adaptive functioning, have atypical presentations. To implement the validity of clinical diagnosis and in order to operationalize the adaptations and descriptions of MDs in ID, the design and experimentation of specifically addressed assessing tools is required. Aim: The main objective of the present study was to develop and to validate the version for DDs and BDs of the Systematic Psychopathological Assessment for persons with Intellectual and Developmental Disabilities (SPAIDD-M). Secondary aims of the present study were the investigation of the prevalence and clinical features of MDs in people with ID and LF-ASD, including demographic, anamnestic, familial, and clinical variables as well as the course and clinical specificities of MDs in this population. Method: 233 adults with ID and with an eventual co-occurrence of ASD, aged 16-65 years, were recruited among those attending the residential and clinic-rehabilitative facilities of the San Sebastiano Foundation in Florence and of the wide network of the Research and Clinical Center (CREA). The sample underwent a complex anamnestic evaluation implying the collection of demographics, psychosocial, familial, medical and psychiatric information via a semi-structured interview developed ad hoc. The SPAIDD-M 1.2 was completed for all the participants. For the evaluation of concurrent validity, 197 participants (85.3%) were administered with the Diagnostic Assessment for the Severely Handicapped (DASH-II), and 141 (68.4%) were assessed with DSM-5 criteria for Major Depressive Episode, Mania and Hypomania, adapted according to the definition of DM-ID-2. Mixed features of major depressive episode (MDE) and (hypo)manic episode were evaluated in 110 probands. The overall sample was stratified based on ID severity as described by the following three groups: 1. Borderline intellectual functioning (BIF) and mild ID; 2. Moderate ID; 3. Severe and profound ID. The test-retest reliability was evaluated in reference to three study participants (one for each ID level) through VII the administration of the tool to the same informants at baseline and after two-to-three months. The inter-rater reliability was evaluated through a special session which included eight different professionals rating for the same study participant. χ² for categorical variables, ANOVA and F-variance for continuous variables tests were used for comparisons between groups. The psychometric properties evaluated were: internal consistency, inter-rater reliability, face validity, criterion validity, test-retest reliability, concurrent validity. To explain the variance, a factorial analysis was performed. For other clinical variables, stepwise backward procedure logistic regression models were elaborated. Results: SPAIDD-M 1.2 showed good psychometric properties. The face validity, which underwent two revisions, resulted to be comprehensible for most evaluators. Some difficulties remained about the completion of the items relative to mixed features. The internal consistency was very good, with a Cronbach’s  of .937, as well as an acceptable inter-rater reliability expressed by a Cohen’s K coefficient ranging from .870 to .575. Criterion validity was also good for both the major depressive and the (hypo)manic episodes. The concurrent validity was found to be high with moderate correlations between SPAIDD-M and DASH-II scores, and strong correlations between SPAIDD-M and DSM-5 diagnoses. A factor analysis identified four main factors explaining about the 38% of the score variance. The sample was very homogeneous regarding all the demographic, socio-economic, familial, medical and psychiatric variables explored. The comparisons of the psychiatric diagnoses indicated the rates of MDs to be equally distributed across the three ID groups, whereas this homogeneity was not evident before the assessment provided in the study protocol. Indeed, before the study, the severe/profound group had been correctly diagnosed only in the half of the cases as well as other participants in the other groups had to be diagnostically reviewed. In our sample, MDs were associated to a higher number of mental health issues in the personal and family history as well as a higher use of psychotropics. LF-ASD resulted to be at higher risk for MDs, mostly BD type I, and the affective illness was more frequently associated to catatonia, mixed features and rapid cycling. The analyses of differences between DDs and BDs as presented in ID did not identify any statistically significant feature, and both clinical conditions resulted to be equally associated to high rates of psychotic, catatonic and mixed symptoms in a more frequent way than it is reported in the general population. The diagnosis of BD was associated to a complex pattern of comorbidities including ADHD, impulse control disorders, binge eating and the presence of a familial psychiatric burden. By the contrary, DDs were associated to anxiety disorders and family history of ASD. Limitations of the study: Sample size, selection, referral and recall biases. Concerns regarding the use of behavioral equivalence. Difficulties in the differential diagnosis with other psychiatric and iatrogenic conditions beyond DDs and BDs. Conclusions: The study provides preliminary information regarding the psychometric properties and the effectiveness of specifically addressed assessment in the ID and LF-ASD population. The prevalence data are not generalizable to the overall ID population. Some clinical information has been drawn regarding the peculiar presentations of MDs in persons with ID pointing towards a pathoplastic effect of the basal neurodevelopmental condition. Further research is needed to operationalize the assessment of DDs and BDs in ID and LF-ASD.