Objective: To test the validity and reliability of the Thai version of the Diabetes-39 (D-39) instrument.

Design: Correlational study.

Materials and methods: Translation of the English questionnaire into Thai employed the translation-back translation technique. Preliminary testing included assessment by 3 experts, think-aloud testing in 5 patients and pretesting in 30 patients. A large-sample study was conducted in 397 diabetic patients at 6 community hospitals in southern Thailand. The majority of subjects (93.2%) completed the scale through an interview.

Results: Exploratory factor analytic results revealed a six-factor structure underlying the D39. The newly identified factor was other health problems and diabetic complications, comprising of 3 items. Thirty-one items from the 36 remaining items clustered into 5 factors identified by the previous studies. The size of nearly all factor loadings were about 0.40 or greater than 0.40. All six dimensions of the D-39 showed reliability indices greater than 0.70. The patterns of the relationship between the D-39 and the SF-36 or self-perception of disease severity were consistent with those hypothesized. All dimensions of the D-39 could reliably discriminate among subjects with/without comorbidities or those with/without insulin injection or those with/without complications. However, the effect size was rather small, less than 0.50.

Conclusions: The D-39 Thai version shows an acceptable level of reliability and validity. The study suggests 6 factors underlying the D-39. However, further studies in a different sample are needed to confirm theresult on factor structure.