Reliability and validity of the Turkish version of the problem areas in diabetes (PAID) survey: Results from diabetes MILES: Turkey

Background and Aims: To 1) assess the reliability and validity of the Turkish version of the Problem Areas in Diabetes (PAID), PAID-5 and PAID-1 survey among Turkish adults with type 1 or type 2 diabetes, 2) estimate the level of elevated diabetes-specific distress (DD), and 3) determine the demographic and clinical correlates of DDsymptom severity. Methods and Results: From 2017-2019, 252 adults with type 1(n = 80) (T1DM) or type 2 diabetes (n = 172) (T2DM) self-reported demographic factors, DD(PAID) and related psychological and clinical questionnaires. We examined PAID internal consistency, structural and convergent validity. Associations of measures with DD were explored with hierarchical linear regression analysis. PCA yielded a 3-factor solution for PAID-20 and a 2-factor solution for PAID-5. Cronbach's alpha for PAID/PAID-5 subscales ranged from 0.63-0.90. All PAID versions correlated most strongly to BIPQ and HFS. The prevalence of elevated distress (PAID-20 >= 33) was 40% in T1DM and 15% in T2DM. DD severity was significantly positively associated (p < 0.01) with more negative diabetes perceptions (BIPQ) and type 1 diabetes. Conclusions: The Turkish translation of the PAID and its short forms appeared to have satisfactory psychometric properties. Elevated diabetes distress was more common in T1DM.