Podium: What has poverty to do with heart attacks?

From a speech by the University College professor to the British Association Annual Festival of Science

A MAJOR unsolved public health problem is the persisting social inequalities in morbidity and mortality in all countries. This is simply the result of worse health for the deprived and better health for the non-deprived. The Whitehall study of British civil servants showed that mortality rates follow a social gradient. The slope of the gradient is not fixed. In England and Wales, between the Seventies and the Nineties, there was a rise in the relative difference between social classes, in deaths from cardiovascular disease and suicide, as well as from all other causes.

The search for understanding these changing social differentials is likely to lead to enhanced understanding of aetiology as well as opportunities for action to improve health in society.

Insights from behavioural genetics suggest that, relative to environmental factors, genetic predisposition may be more important in determining individual differences in behaviour.

Two types of evidence make clear that there are environmental causes of coronary heart disease. The first is the changing of these rates over time. The rise in coronary heart disease mortality this century, and now the well documented fall by as much as 50 per cent, cannot be the result of genetic drift. It is changed nurture not nature. Similarly, the changing social class distribution of coronary heart disease is the result of changed environment.

There is also clear evidence of genetic predisposition to coronary heart disease. But how do we reconcile evidence on the social determinants with that of genetic predisposition?

The standard way is to portion out variance into genes and environment. This approach has three limitations.

First, the relative contribution of genes and environment depend on circumstances. Among monozygotic twins where "genes" are perfectly controlled, 100 per cent of the difference between twins will be environmental, after allowing for imprecision of measurement. Among unrelated adoptive siblings, brought up alongside each other, much of the difference between them has to be genetic.

Second, there may be important interactions between genes and environment We know from experiments in humans that although plasma cholesterol in a group can be changed by manipulating diet, there are large individual differences in response, which may be genetic in origin.

Third, proportion of variance may give a misleading guide to the social and practical importance of a cause of variation in disease. In the Whitehall study, smoking accounted for about 4 per cent of variance in lung cancer mortality. Yet 95 per cent of those lung cancer deaths were in smokers. Is this a contradiction? No. It means that most smokers do not die of lung cancer and, among smokers, there must be factors, to which genes may contribute, that determine which smokers do develop lung cancer.

Given that we do not judge the social gradient in disease to be caused by differentials in genetic susceptibility, the important questions relate to understanding causes of the social gradient in health and how far its slope can be made less steep. Our approach to understanding how social position is translated into ill health derives from three theoretical approaches to social stratification: material conditions, power relations, and social standing.

The first suggests that people in absolute deprivation will have worse health because of their material deprivation. However, none of the Whitehall group is poor. It may be that differences in income and status are important for the other two reasons.

The second, power relations, is a psycho-social concept. It suggests that people of higher status have more control over the conditions in which they live and work. Evidence from the Whitehall II study showed that people who have less control over their jobs have a higher incidence of coronary heart disease than others.

The third factor, general social standing, relates to lifestyle. Men in low grades are more likely to smoke, less likely to pursue vigorous leisure activity and eat less fruit and vegetables. This is part of the explanation for the social gradient in disease that act in addition to psycho-social factors.

Given the demonstrated importance of psycho-social factors in predicting disease it can be hypothesised that there is a virtuous circle: the factors that improve health will also improve general well being and prosperity of society.