Association between indicators of childhood social status and smoking in an elderly population – the CARLA Study

B Schumann; A Kluttig; D Tiller; K Werdan; J Haerting; H Greiser

doi:10.1055/s-0030-1266207

Das Gesundheitswesen, Table of Contents

Association between indicators of childhood social status and smoking in an elderly population – the CARLA Study

B Schumann ¹, A Kluttig ², D Tiller ², K Werdan ², J Haerting ², H Greiser ³

¹Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, Berlin
²Martin-Luther-Universität Halle-Wittenberg, Halle/Saale
³Deutsches Krebsforschungszentrum, Heidelberg

Abstract

Background: Several studies show an association of childhood socio-economic status (SES) with adult smoking; however, mostly paternal occupation only or aggregated indices are used. We investigated the role of different childhood SES indicators for adult smoking behaviour. Methods: Cross-sectional data of the CARLA Study (Cardiovascular disease, Living and Ageing in Halle) including 1779 men and women aged 45–83 were analysed. The four indicators of childhood SES were school education and occupational position of each parent. We used logistic regression models to calculate sex-stratified age-adjusted odds ratios (OR) with 95% confidence intervals (CI) of current smoking by the individual childhood SES indicators, the highest SES category being the reference. Results: Twenty-three percent of men and 14.4% of women were current smokers. In men, school education of both parents was statistically significantly associated with smoking. Men with unknown paternal education had the highest risk (OR 2.3, CI 1.1–4.6). With increasing paternal education, ORs decreased. For men whose mother had a low or medium education, ORs were lower compared to those with highly educated mothers. After adjustment for the three other childhood SES indicators and adult SES, social differences partly increased. Small differences were observed by father's occupation, while men with a blue collar mother had a nearly 2fold risk of being a smoker than men with white collar mothers. In women, none of the childhood SES indicators was statistically significantly associated with adult smoking, although there was a strong social gradient for maternal education. Women with unknown maternal education had a 2.4fold risk of smoking; the lowest risk was found in women whose parent was self-employed or farmer. Conclusions: For men, parental education was associated with adult smoking; however, different patterns emerged for father's and mother's education. No statistically significant relation of childhood socio-economic indicators with smoking was observed in women.