Psychometric Properties of the Mood and Anxiety Symptom Questionnaire in Patients With Chronic Pain

Abstract

Previous studies have demonstrated that the Mood and Anxiety Symptom Questionnaire is able to discriminate between depressive and anxious symptoms. Factor analysis of the scale has revealed 3 factors: general distress, positive affect, and physiological hyper-arousal. However, the properties of the Mood and Anxiety Symptom Questionnaire have not been extensively examined in medical populations, in which the measurement of depressive and anxious symptoms is complicated by the presence of comorbid physical symptoms such as pain. Therefore, the purpose of the present study was to evaluate the psychometric properties of the Mood and Anxiety Symptom Questionnaire in patients with chronic pain. A total of 230 persons with chronic musculoskeletal pain were recruited from a multidisciplinary pain clinic (n=110) and a large urban community (n=120). Participants completed the Mood and Anxiety Symptom Questionnaire and were interviewed for the presence of depressive disorders using the Structured Clinical Interview for the DSM-IV. Factor analyses, scale reliability analyses, and discriminative function analyses were performed on the data. Utilizing principal factor analysis and orthogonal rotation, the factor structure of the Mood and Anxiety Symptom Questionnaire in both chronic pain samples was similar to previously reported factor analyses. Three factors emerged: positive affect, general distress, and physiological hyper-arousal. Coefficients of congruence calculated from the factor structures of each subsample of persons with chronic pain suggested that the factors were highly similar (0.83-0.93). Given that the factor structures were comparable in both populations, a final analysis was conducted on the entire sample. Again, this analysis revealed a 3-factor solution. Examination of the factor loadings and item-to-total correlations revealed that some of the original scale items failed to load on or correlate with a specific factor, or correlated highly with 2 or more factors (thus lacking discriminatory power). A revised scale, eliminating items with low factors loadings and/or high cross loadings, was examined. Reliabilities for the subscales were high, and the subscales demonstrated the ability to predict clinical diagnoses of depression in a hypothesized fashion. The results support the utility of the Mood and Anxiety Symptom Questionnaire with samples of patients with chronic pain and provide further support for the tripartite model of anxiety and depression.