RESUMEN
The current study aimed to examine the effects of a conversation map (CM) psychosocial intervention on diet and exercise health beliefs and health behaviors among people with diabetes (PWD). Based on the Health Belief Model, this large-scale randomized controlled trial (N = 615) investigated whether an added 1-h theory-driven CM intervention (N = 308), compared with usual shared-care service only (N = 307), could significantly better improve PWD's diet and exercise health beliefs and health behaviors at 3-month posttest. Multivariate linear autoregression analysis demonstrated that, controlling for baseline, the CM group had significantly better diet (ß = .270) and exercise (ß = .280) health behaviors at 3-month posttest than the control group. The intervention effects on health behavior change were primarily mediated through desired changes in targeted health beliefs, as informed by the theory. For example, as regards diet, the CM group had significantly greater increases in perceived susceptibility (ß = .121), perceived benefits (ß = .174), and cues to action (ß = .268), as well as greater decreases in perceived barriers (ß = -.156), between pretest and 3-month posttest. In conclusion, future diabetes care may integrate brief theory-driven CM interventions, as in this study, into current shared-care practice to help PWD improve diabetes self-management health behaviors more effectively. Implications for practice, policy, theory, and research are discussed.