Family resemblance in fat intake in the Netherlands GIJ Feunekes, A Sta¯eu, C de Graaf and WA van Staveren Department of Food Technology and Nutritional Sciences, Division of Human Nutrition and Epidemiology, Wageningen Agricultural University, PO Box 8129, 6700 EV Wageningen, The Netherlands Objectives: To assess family resemblance in fat intake in a representative sample of Dutch families. Design and subjects: Households (n  1077) with children between 1 and 30 y old were selected from the data set of the Dutch National Food Consumption Survey 1992. Two-day diet records were available for all household members. Results: Pearson correlation coef®cients for fat and fatty acid intakes (En%) ranged from r  0.51 to r  0.61 between parents, and from r  0.52 to r  0.72 between siblings. The mean associations in fat and fatty acid intake (En%) between mothers or fathers and children ranged from r  0.37 to r  0.50, and they were surprisingly similar for children from 1±3 y of age up to children above 21 y of age. Associations were consistently high for foods eaten at home, and weak for foods eaten outside of the home. Similar within-family associations were found in a set of 1052 households of the Dutch National Food Consumption Survey of 1987. Reported adherence to a therapeutic diet by one of the parents did not erase within-family intake correlations, suggesting that family resemblance is a dynamic phenomenon. Conclusion: Dutch parents and children living together resemble each other in short term intake of fats and fatty acids. Sponsorship: This study was supported by the Ministry of Agriculture, Nature Management and Fisheries. Descriptors: fat intake; dietary intake; social environment; family resemblance Introduction Atherogenesis is a process that starts early in life, and coronary heart disease risk factors of young people may persist into adulthood (Lauer et al, 1988; Jacobs et al, 1992; Klag et al, 1993). Prevention of coronary heart disease should therefore start early in life (Lapinleimu et al, 1995; Shea et al, 1993). Aggregation of fat intake within nuclear families who live together has been reported for husband and wife pairs (Lee & Kolonel, 1982; Nelson et al, 1980; Eastwood et al, 1982; Oliveria et al, 1992; Pe Ârusse et al, 1988; Patterson et al, 1988) and for parents and children (Oliveria et al, 1992; Pe Ârusse et al, 1988; Patterson et al, 1988). Associations in dietary fat intake within nuclear families are expected to be due to a combination of factors. In the ®rst place, the cohabitional effect: family members living in the same household have access to about the same foods (Garn et al, 1979). Further, family members may in¯uence each other by means of their attitudes and norms regarding fat intake, and they may act as a model for each others eating behaviour. In¯uences within families are multidirec- tional. For example, children may imitate eating behaviour of parents and siblings, but they may also in¯uence which foods are purchased or prepared. Shared food preferences may cause some resemblance in intake. In a meta-analysis of seven studies weak resemblances in food preferences (r  0.19) have been found between parents and children (Borah-Giddens & Falciglia, 1993). Also, genetic factors may in¯uence the energy and nutrient intake (de Castro, 1993a; de Castro, 1993b; Wade et al, 1981). Whether family resemblance in dietary intake changes with increasing age of children has not been subject to study yet. In the current study the association of intakes of fat, fatty acids, cholesterol and energy was assessed within a large representative sample of Dutch nuclear families with children from 1±30 y of age. The data of the 1992 Dutch National Food Consumption Survey were used. The analyses were repeated on comparable data of the Dutch National Food Consumption Survey of 1987. We hypothesised that there was at least a moderate association in intakes of fats and cholesterol within these families. The association between intakes of parents and children was expected to decline with increasing age of the children, because the children acquire more in¯uence on their own diet and in¯uences of peers may increase (Lau et al, 1990; Patterson et al, 1988; Sallis & Nader, 1988). Methods Sample Data from the Dutch National Food Consumption Surveys (DNFCS) of 1987 and 1992 (Hulshof & van Staveren, 1991; Lo Èwik et al, 1994; Anonymous, 1993), were used to assess the resemblance in dietary intake within families. The households in these surveys were derived from an existing panel, which consisted of a strati®ed probability sample of non-institutionalised households in the Nether- lands (Hulshof & van Staveren, 1991; Anonymous, 1993). There were 2203 households in the survey of 1987, 2475 households in the survey of 1992, and 453 of the house- holds participated in both surveys. Households with at least one parent and one child were selected. Children above the age of 30 y, were removed from the dataset; their parents were also removed when there were no younger children in the household. No information about a biological relation- ship between members of the households was available. Correspondence: Dr GIJ Feunekes, Unilever Research Laboratory, Consumer Science Unit, PO Box 114, 3130 AC Vlaardingen, The Netherlands. European Journal of Clinical Nutrition (1997) 51, 793±799 ß 1997 Stockton Press. All rights reserved 0954±3007/97 $12.00