P Aveyard1, S Manaseki, J Chambers. 1. Department of Public Health, Birmingham Health Authority. p.n.aveyard@bham.ac.uk
Abstract
BACKGROUND: Social capital describes the notion that the social processes in an area can lead to benefits in health. As Super Profiles describe the social character of an area and they are easy for health authorities to use, they could provide a simple method for local assessment of how social organization affects health. METHODS: We calculated the expected mean birthweight for the enumeration districts of Birmingham based upon marital status, registration details of the child, year of birth, the mother's country of birth, fetal sex and deprivation as judged by the Townsend score using data from 138,696 live-born singleton births for the years 1986-1996 inclusive. We classified enumeration districts into Target Markets, derived from Super Profiles. For each Target Market, we calculated the observed mean birthweight and the difference and 95 per cent confidence interval between the observed and expected birthweights. We used information in Super Profiles to speculate about the social processes that led to some Target Markets having mean birthweights that were significantly different from those expected. RESULTS: Fifteen of the 40 Target Markets had significant differences between predicted and observed mean birthweight, but these differences were less than 50 g. There were no common characteristics of Target Markets that were consistently advantageous for birthweight and none that were disadvantageous. CONCLUSION: The information in the Super Profiles does not illuminate the way that social processes affect health, and the variation in mean birthweight between areas explained by social processes as measured by Super Profiles is small.
BACKGROUND: Social capital describes the notion that the social processes in an area can lead to benefits in health. As Super Profiles describe the social character of an area and they are easy for health authorities to use, they could provide a simple method for local assessment of how social organization affects health. METHODS: We calculated the expected mean birthweight for the enumeration districts of Birmingham based upon marital status, registration details of the child, year of birth, the mother's country of birth, fetal sex and deprivation as judged by the Townsend score using data from 138,696 live-born singleton births for the years 1986-1996 inclusive. We classified enumeration districts into Target Markets, derived from Super Profiles. For each Target Market, we calculated the observed mean birthweight and the difference and 95 per cent confidence interval between the observed and expected birthweights. We used information in Super Profiles to speculate about the social processes that led to some Target Markets having mean birthweights that were significantly different from those expected. RESULTS: Fifteen of the 40 Target Markets had significant differences between predicted and observed mean birthweight, but these differences were less than 50 g. There were no common characteristics of Target Markets that were consistently advantageous for birthweight and none that were disadvantageous. CONCLUSION: The information in the Super Profiles does not illuminate the way that social processes affect health, and the variation in mean birthweight between areas explained by social processes as measured by Super Profiles is small.