BackgroundThe prevalence rates of obesity and smoking are increasing globally and behave like epidemics. There is evidence that the social distribution of risk factors changes as countries develop, with risk spreading down the social gradient over time. However, how or why this occurs is not clear, and there is a scarcity of consensus between academic disciplines over interventions to address these epidemics, particularly for obesity. Multicountry studies can help to elucidate how the social gradient of risk factors changes over time, identify target groups, and generate hypotheses regarding the causes of these epidemics, including how individual-level factors interact with large-scale economic change. This study examines income inequalities in these two key non-communicable disease risk factors (smoking and obesity), comparing England with six middle-income countries to identify common and specific patterns. MethodsWe used nationally representative data from the Health Survey for England (HSE) 2009 (n=3740), and the WHO Study on Global Ageing and Adult Health 2010 (SAGE). SAGE is unique in providing standardised data for six low-income and middle-income countries (LMICs): China (n=14 888), Ghana (n=5565), India (n=12 198), Mexico (n=2737), Russia (n=4670), and South Africa (n=4225). In both studies, body-mass index (BMI) was calculated from anthropometric measurements as weight/height2 and smoking was based on self-reported smoking status (present smoker: yes or no). Income was derived from household earnings (HSE) or wealth (SAGE), which is a reasonable alternative to income in LMICs. Individuals were classified into income quintiles on the basis of their country-specific distribution of income. Relative and absolute inequalities in total household income for smoking (present or non-smoker) and obesity (BMI ≥30 kg/m2) were calculated, respectively, as the age-adjusted relative index of inequality and the slope index of inequality (RII and SII) by sex and country for adults aged at least 18 years. FindingsIn terms of obesity, England was the only country where poor women were at a disadvantage, with a large magnitude of relative inequality in obesity (RII 1·90, 95% CI 1·41–2·59; p<0·0001), and there was no gradient of inequality in obesity for men. By contrast, in the lower income countries, both men and women who were poor were at lower risk of obesity, showing higher food insecurity. In terms of smoking, poor men and women were at a disadvantage in all countries, with England displaying the highest magnitude of relative inequality (RII 3·4, 95% CI 2·7–4·1; p<0·0001 in women and RII 3·10, 95% CI 2·27–4·21; p<0·0001 in men). The results for all the countries are supplied in the appendix. The study was limited by the cross-sectional nature of the data, and different country non-response rates might have introduced bias. We did not explore the role of possible confounders, such as marital status and number of children, or the role of national culture in this analysis; however, the aim of the study was to identify overall patterns that have not been documented previously. InterpretationAlthough the specific pattern of inequality in obesity among women in England suggests that strategies might need to be tailored to a high-income setting, the similarity in patterns of inequality in smoking across countries suggests that the social determinants are universal. Why smoking levels are higher among the poor across a sample of countries with very different levels of economic development needs further investigation. An alarming implication is that the smoking gradient seems to reverse before the obesity gradient, suggesting that low-income groups might forego essential needs (food) before non-essential needs (cigarettes). This raises questions as to the potentially harmful nature of the global consumer environment. FundingThe first author completed the work under funding from the Wellcome Trust (Research Training Fellowship grant number WT088536MA). SAGE surveys are funded by WHO. HSE is funded by the Health and Social Care Information Centre, UK.
Read full abstract