(1) To determine the extent to which an individual's childhood social circumstances and region of residence influence their dietary pattern at age 43 years and (2) to establish the extent to which an individual adopts the dietary pattern of their social and regional circumstances at age 43 years.

Longitudinal study of a social class stratified, random sample of all legitimate, singleton births in the week of 3–9 March 1946.

England, Scotland and Wales.

The 3187 survey members who provided sociodemographic information at age 4 years in 1950 and sociodemographic and dietary data (48-hour dietary recall) at 43 years in 1989.

People who remained in the non-manual social class consumed significantly higher amounts of food items correlated with the factor health aware (items include high-fibre breakfast cereals, wholemeal breads, apples and bananas) than those who remained in the manual social class. Those who made the transition from manual social class in childhood to non-manual social class at age 43 years partly adopted the distinctive dietary patterns of the non-manual social classes. Consumption of items in the factors refined (items include whole-fat milk, white bread, sugar and butter) and sandwich (items include tomatoes, lettuce, onions, bacon and ham) did not differ by social class or regional mobility.

This work suggests that although adult dietary patterns are developed as a result of childhood influences, these patterns can be modified as a result of social and regional transitions. Such changes in dietary patterns may influence susceptibility to disease.