Potential Measurement Properties of a Questionnaire for Eating-Related Distress Among Advanced Cancer Patients With Cachexia: Preliminary Findings of Reliability and Validity Analysis.

Abstract

Background: There are no validated tools for measuring eating-related distress among patients with advanced cancer. Objective: The aim of the study was to investigate the potential measurement properties of a questionnaire for eating-related distress. Methods: This is a secondary analysis. We performed an exploratory factor analysis for factorial validity and calculated Cronbach's α for internal consistency. Patients were classified into the 2 groups categorized using the international cachexia criteria. The total scores of each factor and all items were calculated, and comparisons were performed for known-group validity. We performed statistical correlation analysis for concurrent validity, convergent validity, and discriminant validity using Pearson's product moment correlation coefficient. Results: A total of 140 patients responded. Three factors were identified. The values of Cronbach's α were 0.90, 0.89, and 0.86, respectively. Patients were classified into 2 groups: Non-cachexia/Pre-cachexia (n = 57) and Cachexia/Refractory cachexia (n = 83). Significant differences were observed in the total scores of each factor and all items: (factor 1) 7.5 vs. 11.0, p < 0.001; (factor 2) 8.0 vs. 13.0, p < 0.001; (factor 3) 5.0 vs. 10.0, p < 0.001; (all items) 20.0 vs. 35.0, p < 0.001, respectively. The total scores of each factor and all items significantly correlated with the Edmonton Symptom Assessment System-revised and the nutrition impact symptoms: 0.62 (p < 0.001) and 0.63 (p < 0.001), respectively. Scaling success rates were 100% in factor 1, 2, and 3. Conclusion: The questionnaire appears to be useful. Eating-related distress in patients with advanced cancer had 3 factors.