Resemblance of diet quality in families of youth with type 1 diabetes participating in a randomized controlled behavioral nutrition intervention trial in Boston, MA (2010-2013) : a secondary data analysis / Leah M. Lipsky, Denise L. Haynie, Aiyi Liu, Tonja R. Nansel

Background Parent-child diet quality resemblance is unknown in families of youth with type 1 diabetes, for whom nutrition is central to disease management. Objective Examine diet quality resemblance in families of youth with type 1 diabetes participating in a behavioral nutrition intervention trial and investigate whether treatment assignment or family meal frequency modifies resemblance. Design This is a secondary data analysis from an 18-month randomized controlled trial conducted August 2010 to May 2013. Participants/setting Parent-youth dyads (N=136, child age=12.3±2.5 years) were recruited from a northeast US diabetes center. Main outcome measures Parent and child Healthy Eating Index-2005 (HEI-2005, reflecting adherence to 2005 Dietary Guidelines for Americans) and whole plant food density (WPFD, reflecting intervention target foods) were calculated from 3-day food records collected every 6 months. Statistical analysis Linear random effects models adjusting for demographics and disease characteristics investigated parent-child diet quality resemblance. Separate models examined whether treatment assignment or family meal frequency modified resemblance. Three-way interaction terms examined whether resemblance changed over time by treatment assignment. Results Time-varying parent and child HEI-2005 and WPFD were positively associated (P<0.001), and there were no interactions with family meals. Parent-child HEI-2005 resemblance was similar across treatment groups; however, parent-child WPFD resemblance was stronger in the intervention (β±standard error [SE]=.30±.06) vs control families (β±SE=.12±.05). Parent-child HEI-2005 resemblance was similar over time by treatment assignment, whereas parent-child WPFD resemblance increased over time for families in the intervention group (three-way interaction term β±SE=.03±.01). Conclusions Parent and youth diet quality were positively correlated in families of youth with type 1 diabetes. Resemblance was stronger in the intervention group for target foods, but not for a general measure of diet quality. The lack of effect modification by family meal frequency suggests that family diet quality resemblance is not contingent on shared meals.

Nutrition.