**Background:** Low fruit and vegetable intake in children increases risk for chronic noncommunicable diseases, yet children worldwide consume far less than recommended. Family meals are a key setting for shaping children's dietary habits, and observational meta-analyses have identified longer mealtime duration as the component most strongly associated with better nutritional health. This randomized clinical trial tested whether experimentally extending family mealtime duration increases children's fruit and vegetable intake.
**Methods:** From November 2016 to May 2017, 50 parent-child dyads (children aged 6–11 years; mean age 8 years; 50% girls; parents: 72% mothers, mean age 43 years) were recruited from a volunteer database at the Max Planck Institute for Human Development. Each dyad completed two laboratory evening meals in randomized order (AB/BA block design): a regular condition (same duration as the family's reported regular mealtime; mean 20.83 minutes) and a longer condition (50% more time; mean ~10 minutes extra). Meals were a typical German evening meal with sliced bread, cold cuts, bite-sized fruits and vegetables, dessert (chocolate pudding or fruit yogurt and cookies), water, and one sugar-sweetened beverage, all reflecting the child's preferences. Participants were told the exact time the investigator would return but not that the study concerned mealtime duration. Video recordings from two angles were coded by blinded raters for food consumption (number of pieces of fruits/vegetables, bread/cold cuts, dessert, beverages), interpersonal communication, and eating rate (bites/min). Interrater reliability was high (ICC 0.934–0.997). Primary outcomes were fruit and vegetable intake; secondary outcomes included consumption of other foods, eating rate, satiety (visual analog scale for parents, picture rating scale for children), and mealtime atmosphere. Statistical analyses used paired t-tests, Wilcoxon signed-rank tests, and multilevel modeling.
**Key Results:** Children ate significantly more fruits (mean difference [MD] 3.32 pieces; 95% CI 0.96 to ∞; Cohen's d=0.33; p=0.01) and vegetables (MD 4.05 pieces; 95% CI 2.19 to ∞; Cohen's d=0.52; p<0.001) in the longer condition, totaling approximately 7 additional pieces (~1 portion or 100 g). Consumption of bread and cold cuts did not differ significantly (MD 30.4 kcal; p=0.22). Children drank significantly more water (MD 54.2 mL; 95% CI 24.73–83.67; p<0.001) and sugar-sweetened beverages (MD 36.5 mL; 95% CI 5.53–67.47; p=0.02) in the longer condition. Linear mixed models fit better than logarithmic models for both fruits and vegetables, indicating continued consumption over time. For vegetables, a significant cross-level interaction showed children had already eaten more by the end of the regular mealtime in the longer condition (b=0.01; p=0.001); for fruits, the interaction was negative (b=−0.01; p<0.001), suggesting additional fruit was consumed during the extra time. Eating rate was significantly lower in the longer condition (MD −0.72 bites/min; 95% CI −0.56 to ∞; Cohen's d=1.08; p<0.001), and children reported significantly greater satiety (p<0.001). Dessert consumption did not differ significantly (MD −12.3 kcal; p=0.07). No significant differences were found in interpersonal communication or self-rated atmosphere.
**Clinical Implications:** Extending family meals by approximately 10 minutes led children to consume about one additional portion of fruits and vegetables, an increase associated with a 6–7% reduction in cardiometabolic disease risk per daily portion. The effect was specific to fruits and vegetables, not other foods, and was accompanied by slower eating and greater satiety—both linked to lower obesity risk. The intervention is simple, inexpensive, and low-threshold. Limitations include the laboratory setting (limited generalizability), limited ethnic and socioeconomic diversity (82% of parents had academic-track secondary education), and the unknown durability of effects over time. Further research in diverse, naturalistic settings over longer periods is needed.