Cross-informant agreement between parent-reported and adolescent self-reported problems in 25 societies.

We used population sample data from 25 societies to answer the following questions: (a) How consistently across societies do adolescents report more problems than their parents report about them? (b) Do levels of parent-adolescent agreement vary among societies for different kinds of problems? (c) How well do parents and adolescents in different societies agree on problem item ratings? (d) How much do parent-adolescent dyads within each society vary in agreement on item ratings? (e) How well do parent-adolescent dyads within each society agree on the adolescent's deviance status? We used five methods to test cross-informant agreement for ratings obtained from 27,861 adolescents ages 11 to 18 and their parents. Youth Self-Report (YSR) mean scores were significantly higher than Child Behavior Checklist (CBCL) mean scores for all problem scales in almost all societies, but the magnitude of the YSR-CBCL discrepancy varied across societies. Cross-informant correlations for problem scale scores varied more across societies than across types of problems. Across societies, parents and adolescents tended to rate the same items as low, medium, or high, but within-dyad parent-adolescent item agreement varied widely in every society. In all societies, both parental noncorroboration of self-reported deviance and adolescent noncorroboration of parent-reported deviance were common. Results indicated many multicultural consistencies but also some important differences in parent-adolescent cross-informant agreement. Our findings provide valuable normative baselines against which to compare multicultural findings for clinical samples.

Relation of emotional and behavioral problems with body mass index in preschool children: the Generation R study.

OBJECTIVE: Although problem behavior in children and adolescents has frequently been associated with overweight, it is unclear whether this relationship is already present in early childhood. We hypothesized that problem behavior is positively related to body mass index (BMI) in children of preschool age and that eating behavior explains part of this relation. METHODS: The study was embedded in the Generation R Study, a population-based cohort with data available on BMI and problem behavior for 3137 children aged 3 to 4 years. Problem behavior was measured with the child behavior checklist (CBCI), and eating behavior was assessed using the child eating behaviour questionnaire (CEBQ). Linear regression analyses were conducted to assess the association between the CBCI (expressed as z-scores). CEBQ, and BMI standard deviation scores (BMI-SDS), Bootstrapping was used to formally test mediation. RESULTS: Children with higher levels of emotional problems had a lower BMI-SDS after adjustment for relevant covariates (e.g., ß [95% confidence interval {Cl}] for mother report of emotional problems = -0.04 [-0.07, -0.001], father report = -0.04 [-0.08, -0.001]). Behavioral problems were not associated with BMI. Emotional and behavioral problems were not associated with underweight or overweight if studied categorically. The effect estimate for the relation of emotional problems with BMI-SDS attenuated to nonsignificance after adjustment for specific eating behaviors, i.e., they were accounted for by satiety responsiveness, fussiness, and emotional undereating. CONCLUSION: In this population-based study, emotional problems in preschoolers were negatively related to BMI, and this relation was fully explained by food avoidant eating behaviors.