Objective: To develop a validated, caregiver-based measurement scale to assess sexual changes across several domains in a sample of 86 patients with penetrating traumatic brain injury (TBI) and 65 patients with neurodegeneration due to frontotemporal dementia and corticobasal syndrome.
Methods: A new measure, the Sexual Symptoms in Neurological Illness and Injury Questionnaire (SNIQ), was constructed. Dimensionality, monotonicity, item discrimination power, and scalability were evaluated using nonparametric Mokken item response theory (IRT) methodology.
Results: Three primary domains were established. The domains presented with sufficient reliability (rho .70 to .80), while meeting the Mokken IRT criteria of medium scalability. The domains were labeled 'Prosocial sexual behaviour' (H = .42), 'Sexual interest' (H = .50), and 'Inappropriate sexual behaviour' (H = .41). A fourth dimension emerged, 'Detachment' (H = .47), but with very few items.
Conclusions: Construct validity was established for groups of items pertaining to three unique aspects of sexuality. These findings support further use of the SNIQ in assessing and researching sexual behaviours in patients with dementia and brain injury.
Keywords: Inhibition; brain injury; construct validity; dementia; sexuality.