Development and Psychometric Properties of the 5-Item Diabetes Fatalism Scale Short Form (DFS-SF)

Objective: To create a Diabetes Fatalism Scale Short Form (DFS-SF) that captures the full construct with fewer items and a single factor. Methods: Data from 200 African American adults with type 2 diabetes collected as part of a randomized clinical trial were used. Five steps were followed: 1) correlations were run to identify items with the highest correlations with the overall DFS-12; 2) principal component analysis (unrotated) was used to identify the set of items that maximized variance on the single factor loading; 3) correlations were assessed between the DFS-12 and the new DFS-SF; 4) proportion of variance in the full scale explained by the short version was assessed; 5) discriminatory ability of the scale was evaluated by assessing the longitudinal relationship of DFS-SF with clinical outcomes, quality of life, self-care behaviors, and psychosocial variables. Results: Cronbach’s alpha for the 5-item DFS-SF scale was 0.72. The DFS-SF maintained the original dimensions of the DFS-12, including emotional distress, hopelessness, and powerlessness. The DFS-SF was significantly associated with worse glycemic control (β = 0.02, p = 0.047), lower quality of life (PCS β = -0.13, p = 0.048), (MCS β = -0.27, p < 0.001), higher BMI (β = 0.05, p = 0.03), worse general diet (β = -0.02, p = 0.02), less exercise (β = -0.02, p = 0.04), and lower medication adherence (β = -0.03, p = 0.002), self-efficacy (β = -0.34, p < 0.001), social support (β = -0.37, p = 0.001) and depressive symptoms (β = 0.11, p = 0.002). Conclusions: The DFS-SF is a single factor, 5-item scale that maps to the original dimensions of the DFS-12, accounts for 82% of the variance of the full scale, and is significantly correlated with clinical outcomes, quality of life, self-care behaviors, and psychosocial variables similar to the DFS-12. The DFS-SF provides a valid and reliable short form for use in clinical trials, and large-scale population-based studies where brevity is needed. Clinical Trial Registration: This study used data from a randomized clinical trial that was This study was registered with the National Institutes of Health Clinical Trials Registry on 13 March 2014 (ClinicalTrials.gov identifier# NCT02088658).