Commentary: Plugging leaks and repelling boarders--where to next for the SS income inequality?
Commentary: Plugging leaks and repelling boarders--where to next for the SS income inequality?
复制标题
评论：堵漏、排斥寄宿生——党卫军收入不平等下一步何去何从？
DOI：
10.1093/ije/dyg318
复制
发表时间：
2003
影响因子：
7.7
通讯作者：
DaveySmith,George
中科院分区：
文献类型：
--
作者：
Lynch,John;Harper,Sam;DaveySmith,George
So, does Subramanian and Kawachi’s current analysis settle this dispute with Deaton and colleagues over race/ethnic composition? There are two main issues. First, their analysis demonstrates that ‘per cent black’does not remove the effects of income inequality on self-rated health. While this is convincing, it appears they did not adjust for mean income, as they have in previous analyses of US states. 8 The income inequality effects in the current paper were of the order of odds ratio (OR)= 1.35 (after individual-level adjustments) while in a similar analysis of the same data that was adjusted for state mean income, the OR was around 1.18. 8 Adjustment for some measure of average income has been standard in all studies on income inequality and health. In fact, in their analyses of self-rated health at the US metropolitan level they point out the importance of adjusting for mean income, as it confounds the association between income inequality and self-rated poor health. 9 However, in the current paper they may have been concerned about using more than one or two second-level predictors, given they were analysing US states and so had only 50 second-level units. Thus, all we can conclude is that health effects of income inequality remain after adjustment for ‘per cent black’, but that this income inequality effect was unadjusted for mean income. Second, the studies are difficult to compare because they differ in regard to the outcome, dataset, and modelling strategy—a fact Subramanian and Kawachi acknowledge in their paper. An important issue here relates to the imprecise way the word ‘health’is used—not just in the literature related to income inequality but in the social determinants field in general. Deaton studied mortality and Subramanian and Kawachi studied selfrated health. The usual approach for justifying the use of selfrated health as a valid outcome—followed here by Subramanian and Kawachi—is to cite studies10, 11 that show self-rated health is a strong predictor of mortality. Thus, the claim by Subramanian and Kawachi that their analysis settles the dispute relies on the assumption that mortality and self-rated poor health are reasonably interchangeable as outcomes. Mortality and morbidity are both important population health indicators, but there are several issues that may raise difficulties in considering them as equivalent, especially in aetiological studies. First, social exposures associated with self-rated health may not be associated with mortality. 12 For instance, while selfesteem was strongly associated with self-rated health, it did not predict mortality among a cohort of Finnish men. 13 In The Netherlands, Mackenbach and colleagues showed how a range of psychosocial factors which were related to self-assessed poor health were not associated with mortality. 14 Even with diseasespecific measures it has been shown that morbidity and mortality do not measure the same thing. For example, in one study, selfreported cardiovascular disease morbidity was related to daily stress, whereas cardiovascular disease mortality was not. 15 This suggests that a common tendency to report aspects of peoples’ lives as negative influences both the reporting of stress and the reporting of morbidity. Second, over the long term, mortality and morbidity transitions demonstrate countervailing trends, with declining mortality accompanied by increasing self-reported morbidity. 16, 17 This suggests that mortality and self-reported morbidity have somewhat different long-term determinants at the population level. The importance of determinants of trends has been demonstrated in regard to disentangling ischaemic and haemorrhagic stroke in relation to …
登录
查看更多内容
社区研究报告自评健康与死亡率之间的关联
DOI：
--
发表时间：
1999
期刊：
影响因子：
--
作者：
Y. Benyamini;E. Idler
通讯作者：
E. Idler
教育和职业社会阶层：哪个是死亡风险更重要的指标？
影响因子：
6.3
作者：
G. D. Smith;C. Hart;D. Hole;P. MacKinnon;C. Gillis;G. Watt;D. Blane;V. Hawthorne
通讯作者：
V. Hawthorne
DOI：
10.1136/bmj.312.7037.999
发表时间：
1996-04
期刊：
BMJ
影响因子：
--
作者：
G. Kaplan;E. Pamuk;J. Lynch;R. Cohen;J. Balfour
通讯作者：
G. Kaplan;E. Pamuk;J. Lynch;R. Cohen;J. Balfour
影响因子：
5
作者：
Idler, EL;Benyamini, Y
通讯作者：
Benyamini, Y
影响因子：
--
作者：
Leon, DA;Smith, GD
通讯作者：
Smith, GD