The concern of most recent books is almost exclusively methodological 10— 14 —the health of populations has become a footnote to a detailed exposition of how to calculate a multivariably adjusted effect estimate from a study with appropriate sampling, and then how to apply a billiard-ball view of causation to your study results. Research in this tradition existed before the mid-century—for example in RA Bolt published a study relating public health expenditure to infant mortality rates across cities in the US.
The potential impact of medical therapies was illustrated with data on the changing social class distribution of diabetes deaths at age 20—34 years, when most cases would be type 1 or insulin-dependent diabetes Table 1.
Over the subsequent decade, mortality rates fell in all social classes, but to a much greater extent for social classes I and II than for social classes IV and V, leading to a cross-over in social class patterning of diabetes mortality. The suggestion here is that the more privileged social groups benefit at an earlier stage from the introduction of insulin, and that while insulin had a dramatic effect on diabetes mortality, some benefit much more than others. Differences in the quality of medical care were examined through case-fatality rates Table 2 , which were considerably lower in teaching hospitals than non-teaching hospitals for several important health problems where treatment manifestly could affect outcome.
While differences in characteristics of the patients and their diseases could account for much of this, it suggested that differences in medical care resources and procedures also produced differences in health outcomes. Gross variation in medical practice was utilized as a way of indicating that, in at least some places, optimal care was not being delivered. Thus Morris suggested that the substantial variation in tonsillectomy rates Table 3 indicated over-treatment in some places, which could be contributing to wasted health service expenditure.
Morris considered that it was important to quantify the need for health care in the population. Population studies—of people's needs for health care and their demands for it—were said to be required. Health services were not seen as being outside of the remit of epidemiology, and indications, process, outcome and costs could, in Morris's view, be quantified.
Woven throughout Uses are a myriad of examples of insightful thinking about epidemiology which have been incorporated in the later development of the discipline. Here there is room for just a few examples. The great potential of population-based approaches to disease prevention—as opposed to interventions targeting the relatively small number of high-risk individuals—has been given considerable emphasis in recent public health policy. Recognition of the fact that clinical trials have tended to be too small to detect plausible effects of medical treatments which, while not great, may have substantial population impact, has had considerable influence in recent years.
There has been considerable recent interest in the concept that groups possess properties over and above the sum of the properties of individuals, and that these may influence disease risk. But this figure postulates a function of the group as a whole, in this instance psychological morale. While the seven functions of epidemiology were treated comprehensively in Uses , the key issue in epidemiology was seen to be the uncovering of causes of disease.
In the first edition, the chapter on aetiology covered about a third of the book, which had increased to about a half by the third edition. Furthermore, much of the material in the chapters regarding the other uses of epidemiology refers to how they can contribute to understanding the causes of disease.
One concern of the Uses of Epidemiology that has tended to atrophy in more recent epidemiological textbooks is with the history and geography of disease. The book starts out with a lively summation of disease trends in Britain. It was particularly concerned with the increasing male-female disparity in death rates, Figure 2 with little indication of any improvement in male death rates from the s through to the s, a period during which female death rates declined consistently. The important contribution of ischaemic heart disease and lung cancer to this increasing disparity was made clear.
These two conditions—together with peptic ulcer—were causes of an increasing proportion of deaths from the mid-century onwards, and therefore received much attention in the book and influenced its thinking. Changes in disease rates were also taken to indicate the environmental dependency of disease burden. While many of the problems facing epidemiology in the mid-century have been solved, some remain resolutely intractable.
One of the striking findings reported in Uses related to the changing prevalence of coronary atheroma during the period when deaths from ischaemic heart disease increased dramatically. If anything, there was a decrease in the level of atheroma over this period.
This led to the hypothesis that factors relating to blood clotting were of importance.
Epidemiology of cardiovascular disease
Table 4 presents data from autopsies of young men dying in the Korean war early s and Vietnam war late s. These data have been frequently cited as demonstrating the high prevalence of atherosclerosis in early adulthood, and the importance of early intervention. It has less often been noted that there was a substantial decline in prevalence between the early s and late s.
The data suggest that the recent decline in adult ischaemic heart disease mortality could have been influenced by changes in onset of the early stages of the disease in childhood. This notion was discussed in several places in Uses. Table 5 shows how social circumstances in childhood—indexed by father's social class—specifically influence the risk of cardiovascular disease mortality in later life. These data come from students attending Jerry Morris's alma mater , Glasgow University. Therefore continuity between childhood and adulthood social circumstances is unlikely to account for the association.
Furthermore, other socially patterned causes of death do not show the same association as cardiovascular disease, suggesting that lifestyle and socioeconomic factors in adulthood—which influence other causes of death in addition to cardiovascular disease—do not generate this association. Another study from Scotland has demonstrated that different causes of death show different relative strengths of association with deprivation in childhood and adulthood.
For some of these associations we have a reasonable basis for judging why the findings are as they are. For example, in this study smoking was more strongly associated with adulthood social circumstances than childhood circumstances, 34 and as smoking is the major determinant of lung cancer risk, the disease would be expected to be strongly socially patterned by adulthood social class.
Conversely, stomach cancer is related to Helicobacter pylori infection, an infection generally acquired in childhood and related to overcrowded housing, large family size, absence of running water or an indoor toilet, and the inability to maintain adequate hygiene practices. Thus childhood social circumstances would be expected to influence the risk of stomach cancer in adulthood, as was found. The associations seen at an individual level in prospective epidemiological studies can be considered with respect to the historical and geographical trends in disease, as advocated by Jerry Morris.
It is noteworthy that stomach cancer and stroke—both diseases related to deprivation in childhood—have shown markedly declining rates over the century in Britain, in tandem with improving material circumstances, falling family size and reduced overcrowding. The risk of mortality from these diseases declines as cohorts who experienced improved conditions in their childhood become older adults.
It is not surprising that the declines in stomach cancer and stroke mortality in several countries demonstrate cohort effects. Furthermore, when an indicator of socio-environmental conditions—infant mortality rate—from 70 years previously is examined in relation to mortality rates across countries, strong correlations are seen for those causes of death known to be influenced by childhood deprivation: stroke, stomach cancer and tuberculosis Table 6.
It is noticeable how this reference from to these non-specific adrenal cortical processes has developed little over subsequent decades and is still invoked to account, for example, for social class differences in disease. Indeed, the demonstration that one process increases risk across a wide range of health outcomes among disadvantaged socio-demographic groups would relegate disease-specific investigations to being of secondary importance, with the key task being the identification and characterization of the underlying increased susceptibility.
Data from several sources suggest that such a focus would miss the true complexity of socioeconomic differentials in health. When particular causes of ill-health and death are examined there is a considerable degree of heterogeneity in their association with socioeconomic position. Figure 3 presents data relating to cancer from the Whitehall study of London civil servants, among whom there was a marked gradient in the association between employment grade and all-cause mortality. However, for the 13 specific cancer sites examined grade-related risk varied by site.
The low-grade civil servants had a greater mortality risk for seven of the cancer sites, the higher grades had a greater risk for six. Similar findings with respect to the heterogeneity of site-specific cancer risk with socioeconomic position have come from other studies. In Table 7 , data for a wider range of causes of death are presented from a mortality follow-up of a third of a million men in the US. For some causes of death—including AIDS, homicide, respiratory disease, diabetes and rheumatic heart disease—there are large differentials, with relative risks greater than 1.
The bottom decile income group had mortality rates 2—6 times higher than the rates for the highest decile income group for these causes. For a large number of causes of death—many of them relatively minor contributors to all-cause mortality—there were weak or reversed gradients between income and risk. For example, dying in flying accidents was markedly more likely for higher income men—presumably because those who earned more could afford to fly more. The marked heterogeneity in the strength and even direction of the associations between socioeconomic position and cause-specific mortality draws attention to the need for explanatory models which account both for the overall and specific health effects of socioeconomic position, by considering how lifetime socioeconomic position structures the distribution of risk factors for a range of outcomes over time, and how this can vary by geographical location and birth cohort.
A striking phenomenon, mentioned above, is the tendency for the most important causes of death to demonstrate the most marked socioeconomic gradients. Indeed, as particular causes of death have become more important health problems over the course of this century, the tendency for them to be concentrated among the most deprived tends to become greater. Table 8 presents data on male lung cancer from to In when lung cancer caused one per cent of deaths it showed no social class gradient; by there was a marked gradient—with the mortality rate in social class V men 4.
A similar picture is seen with respect to social class differences in coronary heart disease during the period of rapid increase in this condition as a cause of death. It reflects the ability of favourable social circumstances to allow some people to avoid identified noxious exposures. The influence of these exposures occurs against the background of less avoidable exposures for example poor growth, health and development in childhood to determine the overall pattern of disease.
It should be remembered in this regard that even lung cancer—a disease for which a particularly important adult risk factor can be identified —may show socio-demographic differentials over and above those created by smoking. Human bodies in different social locations become crystallized reflections of the social experiences within which they have developed. The socially patterned nutritional, health and environmental experiences of the parents and of the individuals concerned influence birthweight, height, weight and lung function, for example, which are in turn important indicators of future health prospects.
These biological aspects of bodies and the histories of bodies should be viewed as frozen social relations, rather than as asocial explanations of health inequalities which, once accepted, exclude the social from consideration. Comprehending the ways in which the social becomes biological—and the biological in turn becomes part of the social world—must be a central aspect of an agenda aimed at improved understanding of how health inequalities arise and how they can potentially be reduced.
The first edition of Uses 4 was much concerned with an increase in peptic ulcer in Britain, 4 and the marked international differences in the prevalence of peptic ulcer were also noted. An analysis of data from 19 countries showed similar cohort patterns in all countries, with some variation between countries in when the rises and falls started. However, much of the research carried out in the s, s and s on peptic ulcer related to psychological factors: it was the classic stress-related disease.
In the third edition of Uses , Morris recognized that there was no better theory for peptic ulcer than stress hypotheses at the time, but was clearly very dissatisfied with these. The prevalence of infection is declining in a cohort-specific fashion in countries with declining peptic ulcer incidence. This advance was made by a pathologist and a clinician, with no input from the extensive body of epidemiological research on this important public health topic. As epidemiology enters the 21st century its traditional uses remain of considerable importance in the post-genome world.
The importance of the style of thinking advocated by Jerry Morris is increased by the tendency of epidemiology to concentrate more and more at the individual rather than population level. Putting together individuals and their historical and social contexts still has much to offer to the imaginative traveller. Impact of a new therapy. Death rates per million from diabetes at 20—35 years of age. Males, England and Wales.
Working of health services
Relation of adult mortality age 65—74 years in — with infant mortality at time of birth and at time of death for 27 countries. Lung cancer mortality — social class differences and contribution to total mortality among men of working age.
Relative rates with overall mortality rate among men of working age at each time point as baseline. Frequency of mining accidents in relation to size of pit—number of miners. Source: Morris. Mortality in middle-age, — The contribution of coronary heart diseases and lung cancer. England and Wales. Cancer in the Whitehall Study: relative rate for low employment grade versus high employment grade. Source: Davey smith et al. Cover of Health by JN Morris. First published by the Association for Education in Citizenship, ; reproduced with permission of the author.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Article Navigation. Close mobile search navigation Article Navigation.
- Causation in epidemiology: association and causation.
- Milton to Ouida: A Collection of Essays.
- Determining causation in epidemiology.!
- Navigation menu.
- Lucy in the Mind of Lennon (Inner Lives).
- What is Epidemiology?!
- Hawaiian Crossover; Conversations with a god?!
Volume Article Contents. Working of health services. Prescient ideas. In search of causes. Oxford Academic. Google Scholar. Cite Citation. Permissions Icon Permissions. Box 1. Table 1. View Large. Table 2. Case-mortality in teaching and non-teaching hospitals a. England and Wales, Table 3.
Table 4. Table 5. All analyses adjusted for quintile of year of birth. Source: Davey Smith et al. Table 6. Table 7. Table 8. Figure 1. View large Download slide. Figure 2. Figure 3. A paper prepared for a meeting to honour Professor Jerry Morris on his 90th birthday. Morris JN. London: English Universities Press, Murphy S. Soc Hist Med. Uses of epidemiology. Br Med J ; Aug — Uses of Epidemiology. Edinburgh: Livingstone, Greenwood M. London: Williams and Norgate Ltd, Taylor I, Knowelden J.
Principles of Epidemiology. Susser M.
- Insolvency or Recovery?.
- Follow us on.
- A Love to Last Forever (The Brides of Gallatin County Book #2);
- FBA Hot List - The Best Amazon FBA Tips for 2013;
- Hannah says Hello!;
- Epidemiology and Public Health | Boundless Microbiology.
Epidemiol Rev. Edinburgh: Churchill Livingstone, Ahlbom A, Norell S. Introduction to Modern Epidemiology. Chestnut Hill: Epidemiology Resources Inc. Rothman KL. Modern Epidemiology. Boston: Little, Brown and Co. Norell SE. Workbook of Epidemiology. New York: Oxford University Press, Basic Epidemiological Methods and Biostatistics. Boston: Jones and Bartlett Publishers, Szklo M, Nieto FJ.
Epidemiology: Beyond the Basics. Gaithersburg, Maryland: Aspen Publishers, Stallones RA. To advance epidemiology. Annu Rev Public Health. Consistency of findings. Have the same findings must be observed among different populations, in different study designs and different times? Specificity of the association. There must be a one to one relationship between cause and outcome. Biological gradient.
Chapter 5. Disease Outbreaks | Medical Epidemiology, 4e | AccessMedicine | McGraw-Hill Medical
Change in disease rates should follow from corresponding changes in exposure dose-response. According to Rothman , while Hill did not propose these criteria as a checklist for evaluating whether a reported association might be interpreted as causal, they have been widely applied in this way. Rothman contends that the Bradford - Hill criteria fail to deliver on the hope of clearly distinguishing causal from non-causal relations.
For example, the first criterion 'strength of association' does not take into account that not every component cause will have a strong association with the disease that it produces and that strength of association depends on the prevalence of other factors. In terms of the third criterion, 'specificity', which suggests that a relationship is more likely to be causal if the exposure is related to a single outcome, Rothman argues that this criterion is misleading as a cause may have many effects, for example smoking.
The fifth criterion, biological gradient, suggests that a causal association is increased if a biological gradient or dose-response curve can be demonstrated. However, such relationships may result from confounding or other biases. According to Rothman, the only criterion that is truly a causal criterion is 'temporality', that is, that the cause preceded the effect. Note that it may be difficult, however, to ascertain the time sequence for cause and effect. The process of causal inference is complex, and arriving at a tentative inference of a causal or non-causal nature of an association is a subjective process.
For a comprehensive discussion on causality refer to Rothman. Association or Causation: evaluating links between 'environment and disease'. Bull World Health Organ, Oct. Hill, AB, The environment and disease; association or causation? Proc R Soc Med ; Skip to main content. Create new account Request new password. You are here Epidemiology for Practitioners. Introduction Learning objectives: You will learn basic concepts of causation and association. Read the resource text below. Resource text A principal aim of epidemiology is to assess the cause of disease.
That is, the observed association may in fact be due to the effects of one or more of the following: Chance random error Bias systematic error Confounding Therefore, an observed statistical association between a risk factor and a disease does not necessarily lead us to infer a causal relationship.
Related Epidemiology Causes of Diseases
Copyright 2019 - All Right Reserved