Familial Resemblance in Dietary Intakes of Children, Adolescents, and Parents: Does Dietary Quality Play a Role?

Abstract

Information on familial resemblance is important for the design of effective family-based interventions. We aimed to quantify familial correlations and estimate the proportion of variation attributable to genetic and shared environmental effects (i.e., familiality) for dietary intake variables and determine whether they vary by generation, sex, dietary quality, or by the age of the children. The study sample consisted of 1435 families (1007 mothers, 438 fathers, 1035 daughters, and 1080 sons) from the multi-center I.Family study. Dietary intake was assessed in parents and their 2-19 years old children using repeated 24-h dietary recalls, from which the usual energy and food intakes were estimated with the U.S. National Cancer Institute Method. Food items were categorized as healthy or unhealthy based on their sugar, fat, and fiber content. Interclass and intraclass correlations were calculated for relative pairs. Familiality was estimated using variance component methods. Parent-offspring (r = 0.11-0.33), sibling (r = 0.21-0.43), and spouse (r = 0.15-0.33) correlations were modest. Parent-offspring correlations were stronger for the intake of healthy (r = 0.33) than unhealthy (r = 0.10) foods. Familiality estimates were 61% (95% CI: 54-68%) for the intake of fruit and vegetables and the sum of healthy foods and only 30% (95% CI: 23-38%) for the sum of unhealthy foods. Familial factors explained a larger proportion of the variance in healthy food intake (71%; 95% CI: 62-81%) in younger children below the age of 11 than in older children equal or above the age of 11 (48%; 95% CI: 38-58%). Factors shared by family members such as genetics and/or the shared home environment play a stronger role in shaping children's intake of healthy foods than unhealthy foods. This suggests that family-based interventions are likely to have greater effects when targeting healthy food choices and families with younger children, and that other sorts of intervention are needed to address the intake of unhealthy foods by children.

Keywords: adolescence; diet quality; dietary intake; familial aggregation; familial resemblance; familiality; family study; healthy diet; shared environment; young children.

Figure 1

Flowchart showing the process leading to the final sample analyzed.

Figure 2

Parent–offspring, sibling correlations and spousal correlations for the sum of healthy and unhealthy food groups. Using Fisher’s r-to-z transformation, the confidence intervals are calculated using approximate standard errors [30]; the p-values are calculated according to [26]. The sample size was 1987 parent–offspring dyads, 750 sibling dyads, and 264 spouse pairs.

Figure 3

Parent–offspring and sibling correlations for the sum of healthy and unhealthy food groups separately for younger and older children. Using Fisher’s r-to-z transformation, the confidence intervals are calculated using approximate standard errors [30]; the p-values are calculated according to [25]. The sample size was 996 Parent–offspring dyads for the younger children, 991 Parent–offspring dyads for the older children, 192 younger sibling dyads, and 149 older sibling pairs.

Figure 4

Porportion of variation for the sum of healthy and unhealthy food groups explained by familial and non-familial factors for all subjects and separately for younger and older children. The sample size consisted of 1435 families (746 families in the younger age group, and 760 families in the older age group).