Socioeconomic Status and Health: The Challenge of the Gradient

Citation: Adler, et al. (1994) Socioeconomic Status and Health: The Challenge of the Gradient.
Internet Archive Scholar (search for fulltext): Socioeconomic Status and Health: The Challenge of the Gradient
Tagged: uw-madison (RSS), wisconsin (RSS), sociology (RSS), demography (RSS), prelim (RSS), qual (RSS), WisconsinDemographyPrelimAugust2009 (RSS), socioeconomic class (RSS)

Summary

Hypothesis

The basic hypothesis is that incremental improvements in SES lead to better health outcomes. Previous studies have tended to focus on SES in dichotomous terms–those below the poverty line were compared with those above the poverty line (see Antonovsky 1967; Haan, Kaplan and Syme 1989). Although poverty is clearly associated with poor health outcomes, Adler et al. argue that the relationship between SES and health is more complex. Generally speaking, SES has been associated with better health outcomes at every level. Therefore, it is important to view the relationship between SES and health as a gradient rather than a simple, dichotomous function.

Data

This is a review article. No new data were analyzed.

Evidence for the Gradient

The original Whitehall study (Marmot et al., 1984) found that grade of employment among British civil servants was associated with mortality. "Compared to mortality risk of the top administrators . . . relative risk or mortality was 1.6 for the professional-executive grades, 2.2 for the clerical grades, and 2.7 for the lowest grades" (p. 16). Other studies have found similar relationships between SES variables (e.g., occupation, education and income) and health outcomes (e.g., Susser, Watson and Hopper 1985; Adelstein 1980; Kitagawa and Hauser 1973; Pappas, Queen, Hadder and Fisher 1993).

Possible Causal Mechanisms

Assuming that the relationship is "real" (i.e., not spurious or misspecified), there are a number of possible mechanisms by which SES affects health, including:

1. Health behaviors: (e.g., SES—>smoking—>lung cancer)
2. Physical activity: (e.g., SES—>obesity—>CHD)
3. Psychological Characteristics, including: (a). SES—>depression—>CHD (e.g., Murphy et al. 1991; Booth-Kewley and Feiedman 1987) (b). SES—>hostility—>CHD (e.g., Barefoot et al. 1991; Dembroski et al. 1989) # Psychological Stress: "Higher placement in the SES hierarchy can reduce stress . . . in two ways. First, higher SES diminishes the likelihood that individuals will encounter negative events. [Second], as individuals descend the SES hierarchy, they may have fewer social and psychological resources . . ."
4. Social Ordering: (e.g., SES—>low control—>stress—>CHD)

Limitations of Previous Research

1. Dichotomous treatment of SES
2. Measurement–most studies measure SES with only one variable (e.g., income or education)
3. Myopic focus on a particular level of SES. For instance, most studies have examined individual data or ecologic data, but not both simultaneously (Dr. Robert's research on SES, Communities and Health is a possible exception).
4. Simplistic analyses–basic regression analyses cannot untangle the complex web of relationships between SES and health. (Adler et al. suggest GOM models or tree-structured regressions).

Conclusions

The relationship between SES and health appears to be real and very important. In order to improve our understanding, more refined methods of analysis must be adopted. The authors close by saying that "Social class is among the strongest known predictors of illness and health and yet is, paradoxically, a variable about which very little is known" (p.22).