Commentary: Socioeconomic status: more than a confounder?

W Ahrens

University of Bremen, Bremen Institute for Prevention, Research and Social Medicine, D 28359 Bremen, Germany. E-mail: ahrens{at}bips.uni-bremen.de

Larynx cancer is among the three cancers displaying the strongest gradient of mortality for various measures from low to high social categories, especially in men, with a 2- to 19-fold relative risk of dying from the disease in the lowest as compared with the highest category in different countries.1 The incidence shows a similar pattern with 1.5- to 4-fold relative risks in most countries for men but weaker gradients in women where relative risks below 1.5 were observed.

Menvielle et al.2 report on a hospital-based case-control study that was originally focused to study occupational exposures. In this re-analysis the authors show that much of the social difference between laryngeal cancer cases and controls can be explained by alcohol consumption, smoking, and occupational agent exposures.

Various indicators have been used to describe social inequality or socioeconomic status (SES), often based on occupation. Menvielle et al.2 show that methods developed in the social sciences (for refined examples see refs 3,4) can give useful input to epidemiological studies. However, the classification they have adopted is rather crude and the occupational exposure index was derived from exactly the same variable as the SES indicator ‘manual/non-manual’. Therefore both dimensions are inherently correlated and this could lead to some overadjustment. However, this approach may also leave some residual confounding due to the fact that the exposure variable does not integrate all known and suspected carcinogens of the larynx (acid mists and coal tar products are ignored) and due to the methodological limitations of job-exposure matrices, i.e. the problem of exposure misclassification caused by non-consideration of the within-job variability.

Other classifications of SES are based on school education, professional degree, employment position, type of housing tenure, or income, or a combination of them. None of them can be regarded as sufficient in itself. Depending on the chosen indicator and the way it is created, different but correlated dimensions of social inequality such as occupational or social prestige, wealth, material living conditions, knowledge, and social life circumstances are measured. Many factors directly or indirectly affecting health are involved, like diet, smoking, alcohol consumption, occupational and environmental exposures, physical activity, physical and psychological stress, access to healthcare, and levels of resources. Again, many of these factors are highly correlated and thus it is not surprising that different determinants of social differences tend to give parallel results in terms of cancer incidence and mortality in general5 and for the risk of laryngeal cancer in particular.1 This is also shown in the present paper where both the classification according to education as well as the one based on occupation show similar effects.2

The third dimension which is considered is that of downward and upward social mobility. This is intriguing although it remains questionable whether the mobility per se rather than its subjective perception is a meaningful measure. Jahn et al.6 show that desired changes are favourable while involuntary changes are associated with an increased risk for lung cancer.

Summarizing, the inclusion of occupational exposures in addition to smoking and alcohol in order to clarify the role of SES seems to be an innovative approach and provides some further evidence that the observed differences are not completely explained by these known main causes of the disease. However, the present study does not resolve whether social factors play an independent role in the causation of laryngeal cancer. In addition to the problems discussed above, lack of control for dietary factors may also explain the remaining social differences in disease risk. Residual confounding due to misclassification of alcohol exposure can also not be ruled out here, although the association of alcohol consumption with social class is not consistent.7

In general, SES indicators have been shown to be useful in describing social inequality in health. However, they refer to a vertical classification that only reflects part of the social patterning of society and may disregard relevant influencing factors. There is increasing recognition of the relevance of horizontal structuring that happens in societies today with a trend towards individualization and diversification of lifestyles. New dimensions like job satisfaction, job security, working conditions, and social support and their relationship to health-related outcomes warrant consideration.

Is SES merely a confounder? To understand the pathways that lead to social inequalities in health, the links in the chain need to be identified, e.g. the mechanisms by which social position acts on smoking behaviour. Here, smoking may rather be considered as an intermediate factor in the causal pathway from SES to disease. Thus, identification of social differences in disease risks that are unexplained by known causes may serve as a starting point in the search for the causes of cancer. This is not possible if such factors are only treated as confounders.

The evidence that a ‘residual’ effect of social factors on disease occurrence does exist that is not directly explained by material exposures or health-related behaviour is still controversial. What remains unclear are the pathways leading from social determinants to health. From a neo-material perspective social differences influence health simply through material resources, as has been put forward by Lynch et al.:8

Interpretation of links between income inequality and health must begin with the structural causes of inequalities, and not just focus on perceptions of that inequality. Reducing health inequalities and improving public health in the 21st century requires strategic investment in neo-material conditions via more equitable distribution of public and private resources.

This view is opposed by Marmot and Wilkinson9 who argue that:

Social dominance, inequality, autonomy, and the quality of social relations have an impact on psychosocial wellbeing and are among the most powerful explanations for the pattern of population health in rich countries.

The application of analytical epidemiological approaches addressing these pathways may contribute to a better understanding of the mechanisms involved. This may include the investigation of psychosocial factors, work stress, and social support that are mediated through neuro-endocrine mechanisms. However, in this regard the evidence linking such factors to social differences in health is stronger for cardiovascular disease than it is for cancer.10

The prevention of disease occurrence means that we have to search for causal factors that are modifiable so that effective disease prevention is possible without exact knowledge of the causal mechanisms. Investigation of social factors by interdisciplinary approaches may identify specific targets for intervention. Greatest advances may be expected from social and economic changes affecting lifestyle and exposure factors in a favourable direction. This is particularly true in diseases of multifactorial origin like cancer.


    References
 Top
 References
 
1 Faggiano F, Partanen T, Kogevinas M, Boffetta P. Socioeconomic differences in cancer incidence and mortality. In: Kogevinas M, Pearce N, Susser M, Boffetta P (eds). Social Inequalities and Cancer. Lyon: International Agency for Research on Cancer, IARC Sci.Publ. 1997, pp. 65–176.

2 Menvielle G, Luce D, Goldberg P, Leclerc A. Smoking, alcohol drinking, occupational exposures and social inequalities in hypopharyngeal and laryngeal cancer. Int J Epidemiol 2004;33:799–806.[Abstract/Free Full Text]

3 Ganzeboom HBG, Treiman DJ. Internationally comparable measures of occupational status for the 1988 International Standard Classification of Occupations. Soc Sci Res 1996;25:201–39.[CrossRef][ISI]

4 Wolf C. The ISCO-88 International Standard Classification of Occupations in cross-national survey research. Bulletin de Methodologie Sociologique 1997;54:23–40.

5 Pukkala E. Cancer Risk by Social Class and Occupation. Basel: Karger, 1995.

6 Jahn I, Becker U, Jöckel K-H, Pohlabeln H. Occupational life course and lung cancer risk in men. Findings from a socio-epidemiological analysis of job-changing histories in a case-control study. Soc Sci Med 1995;40:961–75.[CrossRef][ISI][Medline]

7 Moller H, Tonnesen H. Alcohol drinking, social class and cancer. In: Kogevinas M, Pearce N, Susser M, Boffetta P (eds). Social Inequalities and Cancer. Lyon: International Agency for Research on Cancer, IARC Sci.Publ. 1997, pp. 251–63.

8 Lynch JW, Smith GD, Kaplan GA, House JS. Income inequality and mortality: importance to health of individual income, psychosocial environment, or material conditions. BMJ 2000;320:1200–04.[Free Full Text]

9 Marmot M, Wilkinson RG. Psychosocial and material pathways in the relation between income and health: a response to Lynch et al. BMJ 2001;322:1233–36.[Free Full Text]

10 Chandola T, Marmot M. Social Epidemiology. In: Ahrens W, Pigeot I (eds). Handbook of Epidemiology. Heidelberg: Springer, 2004 (in print).





This Article
Extract
Full Text (PDF)
All Versions of this Article:
33/4/806    most recent
dyh203v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by Ahrens, W
PubMed
PubMed Citation
Articles by Ahrens, W