RESUMO
OBJECTIVE: The goal of the study was to examine the relations of general and diabetes-specific friend support and conflict to psychological and diabetes health among youth with type 1 diabetes. We examined gender as a moderator of these relations, and friend responsiveness and information-sharing as potential mediators. METHODS: Youth with type 1 diabetes (n = 167; M age 15.83 [SD = 0.78]; 50% female) were interviewed once in the Fall and once in the following Spring of the school year. Using multiple regression analysis, general friend support, general friend conflict, diabetes-specific support, and diabetes-specific conflict were investigated as simultaneous predictors of psychological and diabetes outcomes cross-sectionally and longitudinally over four months. RESULTS: Cross-sectionally friend conflict, including both general and diabetes-specific, was more predictive of outcomes than friend support. In cross-sectional and longitudinal analyses, gender was a significant moderator, such that several relations of general friend conflict to outcomes were significant for females but not nonfemales. Friend support revealed mixed relations to outcomes across cross-sectional and longitudinal analyses. Although we found links of friend relationship variables to mediators (perceived responsiveness; information sharing), we found little evidence of mediation. CONCLUSIONS: These findings show stronger evidence that conflictual friend relationships than supportive friend relationships are linked to health. Findings suggest that problematic friend relationships may have a stronger impact on the health of females than nonfemales. These results underscore the need to better understand the conditions under which friend support is helpful versus harmful and the reasons underlying these links.