Socioeconomic diVerentials in mortality among men within Great Britain: time trends and contributory causes

Study objective—To assess the size of mortality diVerentials in men by social class in Scotland as compared with England and Wales, and to analyse the time trends in these diVerentials. Subjects—Men from England and Wales and Scotland around each census from 1951 to 1981. Methods—Poisson regression analysis was used to calculate relative indices of inequality for disease specific and all cause mortality as a measure of mortality diVerentials between social classes. This measure is not dependent on the size of the social class groups, so it can be used to compare the magnitude of diVerentials over time periods during which the relative sizes of social class groups change. Main results—While overall death rates were higher in Scotland than in England and Wales around the 1951, 1961, and 1971 censuses the relative indices of inequality indicated smaller mortality diVerences between social classes in Scotland. Inequality, as indexed by the relative index of inequality, increased over time in both Scotland and England and Wales, but to a greater degree in Scotland, resulting in greater social class mortality diVerentials for Scotland in 1981 (the relative index of inequality increased from 1.40 to 2.43 for England and Wales, and from 1.22 to 2.57 for Scotland between 1951 and 1981). This greater increase in the magnitude of inequalities in all cause mortality in Scotland seemed to result from increasing social class diVerentials in cardiovascular disease, accidents and external causes, and “all other causes of death”. Examining the trends in overall death rates, it seems that the greater increase in social class diVerences in Scotland occurred because of the greater decrease in death rates among the privileged social groups, in combination with a smaller decrease (or a greater increase) in the death rates in the lower social class groups. Conclusions—This study has shown that trends in mortality and in inequalities in mortality diVer within Great Britain. Although death rates were higher in Scotland than in England and Wales, smaller mortality diVerentials by social class were found in Scotland over the period 1951 to 1971. By 1981, however, social class mortality diVerentials were greater in Scotland than in England and Wales. The greater increase in the social class diVerentials over time in Scotland, may have contributed to the worsening overall mortality profile in Scotland as compared with England and Wales that occurred between 1971 and 1981. (J Epidemiol Community Health 1998;52:214–218) Socioeconomic mortality diVerences in England and Wales have increased over the past 40 years. Recently, increased attention has been given to the international variation in the size of mortality diVerences associated with socioeconomic position. Kunst and Mackenbach show that countries diVer in the size of their socioeconomic mortality diVerences, with small diVerences being observed in Norway and Denmark and large diVerences in France. Scotland is known to have higher death rates than England and Wales. Carstairs and Morris argued that this higher mortality around the 1981 census is because of the much higher levels of deprivation in Scotland as compared with England and Wales. In another study it was shown that the mortality diVerentials between the west of Scotland and the south of England were largely explained by age, height, lung function, socioeconomic status, and smoking. The relative magnitude of socioeconomic mortality diVerentials within Scotland and England and Wales has not been examined previously, however. The purpose of this study is to assess the magnitude of the mortality diVerences by social class in Scotland as compared with England and Wales over the period 1951 to 1981, to determine whether the higher overall mortality rate in Scotland is accompanied by greater socioeconomic mortality diVerentials within Scotland.We investigate which causes of death have contributed to increasing social class mortality diVerences and whether these were diVerent in Scotland than in England and Wales. Methods All analyses were restricted to men aged 15 to 64 years of age because of diYculties in comparing social class measures in women, based on their own or their spouses occupation, over time. The population of 65 years and older was excluded, because the meaning of social class coded on death certificates for J Epidemiol Community Health 1998;52:214–218 214 Department of Social Medicine, Academic Medical Center, Amsterdam, the Netherlands P J Marang-van de Mheen L J Gunning-Schepers Department of Social Medicine, University of Bristol, United Kingdom G Davey Smith Department of Public Health, University of Glasgow, United Kingdom C L Hart Correspondence: Dr P J Marang-van de Mheen, Department of Social Medicine, Academic Medical Center, Meibergdreef 15, 1105 AZ Amsterdam, the