Comparison of individual-level and contextual-level socioeconomic status indicators in schoolchildren in Ireland: a repeated cross-sectional survey

Abstract

BackgroundEpidemiological studies mostly focus on individual-level measures of socioeconomic status (SES). Contextual measures can provide a wider picture of a person's environment. We aimed to investigate whether family SES measures are associated with school SES in schoolchildren. MethodsWe included a nationally representative cross-sectional sample of 2929 children aged 6–7 years (51·3% girls) who participated in one of the three measurement rounds (2010, 2012, or 2015) of the Irish Childhood Obesity Surveillance Initiative. Family SES indicators on parental education (high or low), parental occupation (employed or unemployed), housing type (detached or semi-detached house, terraced house, or apartment), and housing ownership (rented or owned) were reported by parents. Schools were split into disadvantaged and non-disadvantaged according to the Department of Education and Skills definition. χ2 tests were conducted to test associations. Ethical approval was obtained by the UCD Human Research Ethics Committee. Parents or children's guardians signed an informed consent form. FindingsThe proportion of mothers with low education level was significantly higher in disadvantaged schools than in non-disadvantaged schools (2010, 62·5% [n=35] vs 36·7% [n=348], p<0·001; 2012, 61·3% [n=49] vs 37·5% [n=346], p<0·001; 2015, 64·9% [n=50] vs 33·9% [n=270], p<0·001), irrespective of the measurement round. Similar results were observed with fathers' education and school SES. Disadvantaged schools had significantly higher percentages of unemployed parents than did non-disadvantaged schools in 2012 (mothers 54·3% [n=44] vs 41·7% [n=384], p=0·028; fathers 27·4% [n=20] vs 13·0% [n=115], p=0·001) and 2015 (46·7% [n=35] vs 30·2% [n=239], p=0·003; 18·7% [n=12] vs 7·4% [n=55], p=0·002), but not in 2010. The percentage of families owning a house was significantly lower in disadvantaged schools than in non-disadvantaged schools (2010 64·8% [n=35] vs 83·8% [n=816], p<0·001; 2012 53·1% [n=43] vs 80·9% [n=758], p<0·001; 2015 55·3% [n=42] vs 83·4% [n=665], p<0·001). InterpretationSchool SES was significantly related to individual-level SES measures over the three study rounds. These findings show that contextual and compositional SES indicators are equally important and should be considered. Ideally, studies should simultaneously examine the effects of both individual-level and contextual-level SES variables. FundingNone.