The factor structure and predictive validity of the Kessler Psychological Distress Scale (K10) in children and adolescents

Abstract

ObjectiveTo systematically investigate the factor structure and predictive validity of the Kessler‐10 (K10) in an Australian child and adolescent population.MethodConfirmatory factor analyses were undertaken on data from 2,967 youth aged 11–17-years from the second Australian Child and Adolescent Survey of Mental Health and Wellbeing. Receiver operating characteristic analyses and penalised logistic regressions were used to explore the ability of K10 factor scores to predict Diagnostic Interview Schedule for Children Version IV (DISC‐IV) identified mental disorders.ResultsA bifactor model with two‐group factors (Anxiety and Depression) provided the best fit for youth with past‐month mental disorders. A four‐factor solution (Nervousness, Agitation, Fatigue, and Negative Affect) fit best for youth without past‐year mental disorders; however bifactor analyses indicated that these factors accounted for little common variance after controlling for the effects of general distress. When controlling for age and gender, Agitation factor scores predicted DISC‐IV behavioural disorder diagnoses and neither Anxiety scores nor Nervousness scores were predictive of DISC‐IV diagnosed social phobia, separation anxiety, obsessive–compulsive disorder, casting doubt over the suitability of K10 items to capture anxiety symptoms in youth.ConclusionThe K10 is “unidimensional enough” in the Australian child and adolescent community population. Associations between K10 and mental disorders are mainly represented by items comprising Negative Affect and Agitation factors, which raises the possibility that briefer versions might be equally useful, either for capturing the general distress factor or a version that is more predictive of mental disorders in youth.