The health effects of relative status

This thesis explores the effect of relative status within a social group on health. Previous research in the area has found a status gradient for many health outcomes using a wide variety of status measures. As simple explanations failed to account for this gradient, more complex psychosocial explanations were proposed. Generally, these theories state that individuals derive some health benefit from comparison with those of lower status, with the converse also being true. However, empirical evidence for this theory has been derived from research with abstract social categories such as census data or occupational grades within an organization, rather than with a reference group that an individual is likely to know. Because of this, the possibility that a group against which comparison is made might also have some influence on the behaviors and attitudes of the individual has not been explored. In this thesis, egocentric social groups to a geodesic distance of two are constructed for individuals in the Framingham Heart Study panel dataset. The status properties of these social groups are found to have significant effects on the individual's health. Tests are done using education, occupational prestige, BMI, and a latent status variable as measures of status. Health outcomes of self-rated health, blood pressure, and a latent health variable are used. In most cases, the influence of the social group is found to be stronger than any comparison effect. However, an important counterexample is found for younger females, in which comparison on body type appears to be stronger than any influence effect. Further tests on the mechanisms of said effect are found to be suggestive but not conclusive. The finding that influence tends to be more important than comparison runs contrary to existing theory, which would predict that comparison would be the most important. Research methods that fail to consider this influence effect are incomplete and likely to provide misleading evidence. The influence effect has been shown at the local level, but the conclusions have implications for the broader topic of explaining the observed society-wide status gradients in health outcomes. Individuals tend to be surrounded by a social group similar to themselves, providing an environment that reinforces behaviors and attitudes consistent with their position in the status hierarchy. Previous researchers have not tended to consider this social effect, making it an unmeasured effect that might explain some of the status gradient that is left when measurable behaviors are controlled for. Thus, a strong local influence effect could be easily interpreted as a comparison effect at the macro level.