The purpose of this survey was to establish a database of habitual food and drink consumption in a representative sample of Irish adults, aged 18–64 years. The present paper describes the sampling protocol, response rate and characteristics of the survey population in terms of sex and age groups, seasonality, geographical location, marital status, social class, socio-economic status and education level.

A cross-sectional food consumption survey was carried out. In the Republic of Ireland, a nationally representative sample of adults was randomly selected with a validated two-stage clustered design, using the electoral register as the sampling frame. This method produced a self-weighting or ‘epsem’ sample of individuals, where each adult who was registered to vote had an equal opportunity of being selected. Similarly, in Northern Ireland, a two-stage random sampling procedure was used. The sampling frame was the electoral register, and the sample was stratified by urban/rural and by an index of material deprivation, to ensure representation of each sector of the community. The recruitment procedure was the same in the North and South. An introductory letter with an information leaflet was posted to each selected individual and these were followed up by a visit from a fieldworker, who invited participation in the survey.

Northern Ireland and Republic of Ireland between 1997 and 1999.

The response rate, which is the percentage of the total number of people who completed a 7-day food diary (n = 1379) out of the total eligible sample (n = 2177), was 63%. Non-respondents and dropouts constituted 34% and 3%, respectively, of the total eligible sample. Compared with the most recent census figures available, the sample was generally found to be representative in terms of sex and age group profiles, geographical location, marital status, seasonality, social class, socio-economic group and education level. Data on sex and age group and geographical location were collected from non-respondents for comparison with the survey sample. There were no apparent differences between them.

The North/South Ireland Food Consumption Survey has established a Samplino relational database of habitual food and drink consumption, in addition to data on Response rate habitual physical activity, anthropometric measurements, socio-demographic factors, Demographics lifestyle, health status indicators and attitudes, in a nationally representative sample Socio-economic factors of the population of the island of Ireland.

Smoking initiation and persistence are clearly associated with factors commonly thought to be environmental in origin, including socio-economic status. However, twin models that incorporate gene–environment correlation and gene×environment interaction have not been applied to elucidate the genetic or environmental role that socio-economic status plays in smoking initiation and nicotine dependence.

Twin structural equation modelling was used to examine gene–environment correlation and gene×environment interaction of one index of socio-economic status, educational attainment, with smoking initiation and nicotine dependence among 5119 monozygotic and 4295 dizygotic male–male Vietnam-era twins from the Vietnam Era Twin Registry, a national registry of twin pairs who served in the military during the Vietnam era.

Educational attainment correlated significantly with smoking initiation (r=−0.27, p<0.001). Additive genetic (p=0.011), shared environment (p=0.002) and unique environment (p=0.027) components contributed to the correlation between educational attainment and smoking initiation. Educational attainment also significantly moderated the variance in smoking initiation (p<0.001), suggestive of gene×environment interaction, although the interaction with the additive genetic, shared environmental and unique environmental components could not be resolved due to multicollinearity. In contrast, educational attainment neither correlated with nor moderated nicotine dependence, once smokers had initiated.

Our study suggests that educational attainment is associated with smoking initiation, in part due to gene–environment correlation and gene×environment interaction. However, once smoking initiation is taken into account, there are no effects – be they gene–environment correlation or gene×environmental interaction – of educational attainment on nicotine dependence.