HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Electronic Letters to:

RESEARCH AND PRACTICE: Arline T. Geronimus, Margaret Hicken, Danya Keene, and John Bound "Weathering" and Age Patterns of Allostatic Load Scores Among Blacks and Whites in the United States Am J Public Health 2006; 96: 826-833 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

Electronic letters published:

Difficulties in the Interpretation of Patterns of Racial Differences in Allostatic Load

James P Scanlan   (26 February 2006)

What's Missing from the Weathering Hypothesis?

Nigel M Thomas   (24 January 2006)

Difficulties in the Interpretation of Patterns of Racial Differences in Allostatic Load 26 February 2006
James P Scanlan, Attorney James P. Scanlan, Attorney at Law Send letter to journal: Re: Difficulties in the Interpretation of Patterns of Racial Differences in Allostatic Load jps{at}jpscanlan.com James P Scanlan	Geronimus et al. find support for the weathering hypothesis in a perceived increase in the racial disparity in allostatic load over the life course.1 That racial disparities in the consequences of stress would increase over time seems entirely plausible. Possibly the effects sizes of the difference between average allostatic loads of blacks and whites might show an increase with age. But the use of a dichotomized measure in the manner employed in this study has methodological problems. In general, relative differences in experiencing an outcome tend to increase (while relative differences in avoiding the outcome tend to decline) as the outcome grows more prevalent, solely for reasons related to the shape of the risk distributions of each group.2,3 Such tendency makes it difficult to discern whether disparities are increasing with age in any meaningful sense with respect to an outcome, like having an allostatic load of 4 or greater, that grows more prevalent with age. That is, a true tendency for disparities in high allostatic load levels to increase with age – such as might be reflected in increasing effects sizes – may not be sufficient to outweigh the statistical tendency for the disparity in rates of exceeding a certain allostatic load level to decline simply because allostatic loads increase with age. The Geronimus study seems nevertheless to find that the relative black-white differences are increasing with age. But this seeming increase is a consequence of the use of odds ratios in circumstances where an outcome is sufficiently prevalent that relative odds do not approximate relative risks. The underlying rates of having allostatic levels of 4 or greater shown in parentheses in Geronimus’s Table 1 make clear that, while for each gender the black-white ratio of the odds of having an allostatic load of 4 or greater is highest for the oldest age group, the black-white relative risk of having such an allostatic load is lowest for that group. Similar interpretative issues are involved with the adjusted figures. The same parenthetical figures also make clear that the relative risk of having allostatic levels below 4 is highest in the oldest age group. The fact that one reaches opposite conclusions about whether the size of the relative difference increases with age depending on whether one examines the rates of experiencing the outcome or the rates of avoiding the outcome highlights the difficulties in interpreting health disparity patterns using dichotomous measures. It warrants note that some commentators prefer odds ratios precisely because, unlike relative risks, odds ratios show the same proportionate difference whether one examines the presence or absence of an outcome.4,5 However, like the relative risks of experiencing or avoiding an outcome, odds ratio tend to change systematically as an outcome grows more or less prevalent regardless of whether there is a meaningful change in the relative situation of two groups.6,7 Thus, odds ratios do not provide an effective means of identifying differences between the relative situations of blacks and whites in different age groups that are not simply the consequence of greater prevalence of an outcome among the older age groups. Hence, where, as in the case of allostatic load, comparisons of the size of disparities can be measured by continuous measures rather than dichotomous measures, use of the continuous measure seems the preferred approach. 1. Geronimus A, Hicken M, Keene D, and Bound J. Weathering and Age Patterns of Allostatic Load Scores Among Blacks and Whites in the United States. Am J Public Health. 2006;96:XXX-XXX. 2. Scanlan JP. Race and Mortality. Society. 2000;37:19-35. 3. Scanlan JP. Divining difference. Chance. 1994;7(4):38-9, 48. 4. Gastwirth J. Statistical reasoning in law and public policy (Vol. 1). Boston: Academic Press, 1988. 5. Cornfield J. A method of estimating comparative rates from clinical data. Applications to cancer of the lung, breast and cervix. J Nat Cancer Inst. 1951; 11: 1269 75. 6. Scanlan JP. Can we actually measure health disparities? Chance. 2006:19(_): ____ (in press). 7. Scanlan JP. The perils of provocative statistics. The Public Interest. 1991;102:3 14.
What's Missing from the Weathering Hypothesis? 24 January 2006
Nigel M Thomas, Student Teachers College, Columbia University Send letter to journal: Re: What's Missing from the Weathering Hypothesis? nt170{at}columbia.edu Nigel M Thomas	Geronimus1 et al’s (2005) finding is a curious one. The authors considered “whether US Blacks experience early health deterioration, as measured across biological indicators of repeated exposure and adaptation to stressors” and found evidence of the weathering hypothesis among Blacks. Blacks had higher cumulative risk metrics than did Whites. These racial differences were not explained by poverty. Poor Black women had the highest allostatic load scores followed by non-poor Black women. The finding suggests that within the racial and gender stratification of the United States, black women are relatively more likely than other groups to engage in “John Henryism” --persistent high-effort coping with the pernicious effects of discrimination.2 What is missing from this discussion, though, is a synthesis of life expectancy tables with the weathering hypothesis and allostatic load scores. The “weathering hypothesis” suggests that health may decline in early adulthood as a physical consequence of cascading socio-economic stressors. 1 Allostatic load captures the wear and tear the body experiences as it strives to achieve stability in disruptive environments.3 Life expectancy- -the statistical projection of the length of an individual’s life span based upon probabilities and assumptions of living conditions, and other affecting factors—has been described as the best indicator of population health. 4 This is the paradox: Black women live longer than White men. 5 The disparity in life expectable at birth is even starker when Black women are compared with Black men. Black men experience disproportionately higher death rates in all the leading causes of death. 5 And Geronimus has documented elsewhere that whereas one third of Black girls at 15 will not reach their 65th birthday, the figure is two thirds for Black boys the same age. 6This is a paradox that demands better explication if we are to truly advance our understanding of how stress-related chronic diseases contribute to excess mortality in marginalized populations in the United States and elsewhere. References 1 Geronimus A, Hicken M, Keene D, and Bound J. Weathering and Age Patterns of Allostatic Load Scores Among Blacks and Whites in the United States AJPH published December 27, 2005, 10.2105/AJPH.2004.060749. 2 James S, Schultz A, and van Olphen J. Social capital, poverty and community health: An exploration of linkages. In Saegert S, Thompson JP, and Warren M, Eds, Social Capital and Poor Communities. New York: Russell Sage Foundation Press. 2001. 3 McEwen B. Stress, Adaptation, and Disease: Allostasis and Allostatic Load. Annals of the New York Academy of Sciences 840:33-44 (1998). 4Pan American Health Organization. Health in the Americas. 2002, Volume I: 456 pp; Volume II: 600 pp., ISBN 92 75 11587 7. 5 Health disparities experienced by black or African Americans--United States. (2005). MMWR Morb Mortal Wkly Rep, 54(1), 1-3. 6 Geronimus A and Thompson JP. To Denigrate, Ignore, or Disrupt: Racial Inequality in Health and the Impact of a Policy Induced Breakdown of African-American Communities. Du Bois Review. 2004, 1:2, 247-279.