ABSTRACT
LAY ABSTRACT: Autistic people are more likely than non-autistic people to experience mental health difficulties. The Strengths and Difficulties Questionnaire is often used to screen for these difficulties and to otherwise make important decisions about mental health treatment and research in populations of autistic people. However, this study suggests that parent-reported Strengths and Difficulties Questionnaire scores may not be useful for comparing autistic and non-autistic adolescents at 11, 14 and 17 years old, as well as screening for mental health conditions in autistic adolescents. In addition, several items may be more likely to be endorsed by parents of autistic 17-year-olds than by parents of non-autistic 17-year-olds (and vice versa), which might suggest caution is needed when comparing groups on specific items.
ABSTRACT
BACKGROUND: The nine-item Patient Health Questionnaire (PHQ-9) and the seven-item Generalised Anxiety Disorder scale (GAD-7) scales are routinely used in research and clinical practice. Whilst measurement invariance of these measures across gender has been demonstrated individually in general population studies and clinical samples, less is known about invariance of the distinct but correlated latent factors ('depression' and 'anxiety'). The current study assessed measurement invariance of these constructs across males and females seeking treatment for common mental health disorders. METHODS: Data were provided from eight psychological treatment services in London, England. Data from initial assessments with the services where individual items on the PHQ-9 and GAD-7 were available were included in analyses. Measurement invariance was explored across self-identified genders, with 'male' and 'female' categories available in the dataset. Sensitivity analyses were conducted using propensity score matching on sociodemographic and clinical variables. RESULTS: Data were available for 165,872 patients (110,833 females, 55,039 males). There was evidence of measurement invariance between males and females in both the full sample and a propensity score matched sample (n = 46,249 in each group). CONCLUSIONS: Measurement invariance of the correlated depression and anxiety factors of PHQ-9 and GAD-7 were indicated in this sample of individuals seeking psychological treatment for CMHDs. These results support the use of these measures in routine clinical practice for both males and females. This is of particular importance for assessing the prevalence of clinically significant levels of symptoms as well as comparing treatment outcomes across genders.
