Validity of the methods evaluating the association between psychosocial stress and diabetes incidence and/ or control

Abstract Aims: Assessment of psychosocial stress remains a major obstacle in understanding its association with diabetes. We examined the validity of methods used to measure psychosocial stress in relation to diabetes incidence and control. Methods: 125 studies of psychosocial stress and diabetes published between 1983 and 2017 were evaluated. Evaluation included the stress type definition, study’s design, methodology of stress assessment and diabetes outcomes (incidence and/or control).Results: An association between stress and diabetes outcomes was found in 89% of studies. Prospective designs were commonly employed (38%), followed by cross-section designs (18%). More studies (64%) focused on psychosocial stress than sleep disorder related stress (22%). More studies explored diabetes incidence as an outcome (74%) than diabetes control (22%). The most common measure for stress assessment was a self-administered questionnaire (64%), followed by physiological indicators (7%). An analysis of the validity of the measures used, indicates that 67% of the studies used valid measures, with self-administered questionnaires exhibited a similar validity as other tools. Interestingly, the validity of tools used to evaluate sleep disorders severity was generally higher than those used to evaluate psychosocial stress.Conclusion: The majority of studies evaluating the association between anteceding stress and diabetes outcome applied valid measures.