OBJECTIVES: Considerable variation has been reported in the prevalence and correlates of challenging behaviour (CB) in adults with intellectual disabilities (ID). To provide a robust estimate of prevalence, we identified the entire administrative population of adults with ID in a defined geographical area and used a behaviour assessment tool with good psychometric properties.

METHODS: Data from 265 adults who were known to services were collected using a demographic survey tool and the Behavior Problems Inventory - Short Form. The prevalence of self-injurious, aggressive/destructive, stereotyped, and overall CB was evaluated. We explored the potential of developing cumulative risk indices (CRI) to inform longitudinal research and clinical practice.

RESULTS: The prevalence of overall CB was 18.1% (95% CI: 13.94-23.19%). The prevalence of self-injurious behaviour was 7.5% (95% CI: 4.94-11.37%), aggressive-destructive behaviour 8.3% (95% CI: 5.54-12.25%), and stereotyped behaviour 10.9% (95% CI: 7.73-15.27%). Communication problems and severity of ID were consistently associated with higher risk of CBs. CRIs were significantly associated with CBs, and the five methods of CRI development produced similar results.

CONCLUSIONS: Findings suggest a multi-element response to CB is likely to be required that includes interventions for communication and daytime activity. Exploratory analyses of CRIs suggested these show promise as simple ways to capture cumulative risk in this population. Subject to longitudinal replication, such a tool may be especially useful in clinical practice to identify adults who are priority for interventions and predict future demand on services.

PRACTITIONER POINTS: The prevalence of challenging behaviour (CB) was 18.1% in this total population study. Stereotypy was the most frequent type of CB. Communication difficulties and severe-profound intellectual disabilities were most systematically related to the presence of CB. Establishing the effect of multiple risk factors is likely to identify people who are priority for interventions. Addressing multiple, rather than singular risks, is likely to be more efficacious. We tested five different methods of putting together a multiple risk index. All methods provided a reasonable association with CB. The most user-friendly method was the additive cumulative risk index (CRI). Limitations This is a cross-sectional design which enabled factors currently associated with CB to be identified for the whole cohort, but these variables may not be those conferring risk for the development or maintenance of CB over time. Future longitudinal research is required to replicate these CRI analyses before concluding about the CRI method with the highest predictive validity.