Background: Evidence from other studies has show large, systematic differences between the health of social groups. It is not clear whether this relationship applies equally to all areas of health need. We assess whether a variety of areas of ill health show positive correlations with increasing socioeconomic disadvantage, and whether there are indicators of socio-economic disadvantage that are better than others at predicting the prevalence of specific morbidities at a population level.
Methods: The prevalence of a range of common morbidities was determined by a postal questionnaire sent to 16,750 subjects (response rate 79 per cent), and compared with socio-economic information obtained from the 1991 Census.
Results: There was substantial variation in the degree to which the various morbidities were related to the socioeconomic variables. When compared with socio-economic variables, long-term limiting illness, respiratory conditions and depression had high correlations of +0.8 or more. Cardiovascular conditions were less related (r = +0.60 to +0.79). None of the disorders of the gastrointestinal system showed a high correlation with socio-economic variables. There was also substantial variation in the degree of correlation of the socio-economic measures with each area of morbidity. The measures that showed the highest correlations were in respect of household characteristics such as car ownership and single parent households. Variables describing household amenities such as lacking a bath or central heating were least related to the morbidity measures.
Conclusions: Some areas of morbidity show strong associations with socio-economic disadvantage, but others show only modest or no relationship. The optimum choice of socio-economic variable as a proxy for health need depends on the area of illness being considered.