IJE Advance Access originally published online on April 30, 2007
International Journal of Epidemiology 2007 36(2):384-386; doi:10.1093/ije/dym054
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary: ‘Epidemiologic transition’ interrupted or sweep to the second stage of ‘health transition’?
Institut National d’Etudes Démographiques, 133 Boulevard Davout, Paris cedex 20, France. E-mail: vallin{at}ined.fr
Accepted 22 February 2007
Do it twice and you will get different results. That's a very common law when measuring demographic indicators in developing countries where current statistics are incomplete. And it is very likely that the third time will give yet another result! It is not enough to make good demographers despair. It even presents a good opportunity to outstanding specialists like the authors of ‘Epidemiologic transition interrupted: a reassessment of mortality trends in Thailand’1 to try to approach the true measure of mortality in such a country. To approach, not to reach absolutely, of course. Indeed, several questions still await answers. But the most questionable aspect of the results presented here is not to what extent reassessment fits the reality but to what extent trends reassessed justify the main title: ‘Epidemiologic transition interrupted’.
The work done by the authors of the article significantly improves our knowledge about recent trends in Thai mortality. They make a very important clarification of the complex set of contradicting data produced by direct and indirect methods applied to vital statistics, census data and the results of different types of surveys. They give us a clearer idea of what must be rejected, what can be used and what can be improved by adequate correction. Nevertheless, as usual, good answers bring new questions.
Vital statistics are clearly incomplete as they are in many developing countries. However, Thailand has a long tradition of civil registration of births and deaths based on an administrative system rather well spread out over the whole country. Even incomplete data could be of great use if we could estimate its level of completeness and the pace of its improvement with time. Simple assumptions could then be enough to reconstruct annual series of relatively good quality, which would be very helpful to follow the growing impact of epidemics like HIV-AIDS and to monitor the results of policy measures adopted to try to stop them.
Indeed, the article discusses at length the under-registration of deaths in the vital statistics, and, at least for child mortality, on the basis of the most reliable surveys, the authors came to the final conclusion that U5MR was 58 per 1000 in the 1980s, 30 in the 1990s and 23 in the 2000s. That means that vital statistics missed 64% child deaths in the 1980s, 57% in the 1990s and 52% in the 2000s. Nevertheless, no attempt is made to reconstruct annual time series. Such a reconstruction should have shown a rather straightforward decline in child mortality more clearly than the curve of U5MR given by uncorrected vital statistics. Not only was child mortality not affected by HIV-AIDS epidemics fundamentally, but its continuous decline was more pronounced than indicated by vital statistics, the coverage of which improved. It can only be said that in very recent years the decline slows down due to the delayed impact of AIDS on infant and child mortality. One thing is striking however: why U5MR reached a peak in 1998 so sharply while adult mortality fell down to a minimum, due to what the authors call the ‘system problems of 1997–98’. The fact is all the more strange because infant mortality rate also goes down at the same moment. Since U5MR includes IMR, the 1998 peak must be even sharper for the probability of death between ages 1 and 5. Why? Was there any special change in the registration system that could act differently for 1–5 age group than for the others? Or was there any sudden rise in some infantile disease? It would be very interesting to know more about this.
More generally, it would be interesting to know more about the "system problems of 1997–98’. The authors only say it is ‘more a data collection than registration’ problem. But, what happened? Is it just a question of temporary losses for the two years, or a more or less permanent degradation that affects all the successive years? Though not explicitly stated, it seems that the authors’ opt for temporary losses when, for example, they state: ‘the 1995–96 SPC survey estimates of vital registration coverage would not have captured the system problems in 1997–98 and would underestimate mortality for the 1990s’. Consequently, it is quite realistic to consider that the decline in adult mortality observed since 1999 is, at least, as true as the rise observed in the mid 1980s until 1996. It could even be argued that the recent decline is underestimated while the former increase was overestimated, if we have to consider progressive improvement of the registration coverage. Indeed, this would not be enough to question the reality of the impact of HIV/AIDS epidemics. What we really don't know is when the impact peaked precisely: in 1996? Or somewhere between 1996 and 1999? And at what level? Whatever the answer, it seems obvious that adult mortality peaked somewhere in 1996–99 and that after the sharp increase of the late 1980s–early 1990s, a significant backward surge has been going on since the end of the 1990s. This is quite apparent from uncorrected vital statistics in term of trends. It would have been wonderful if the authors had also tried to produce annual estimations based on corrected vital statistics.
They preferred to discuss different life expectancy estimations for two global periods: 1980–90 and 1990–2000. The results are very interesting, but they obviously lead to the conclusion that life expectancy is declining and that ‘Epidemiologic transition (is) interrupted’. Indeed, the period 1990–2000 includes the greatest part of the impact of HIV epidemics and only the very beginning of the decrease. To conclude differently, and more correctly, I think, it would have been better to compare 1990–96 with 1999–2004, or even better to try to correct vital statistics and to give annual estimates of life expectancy from 1980 to 2004. But was it possible?
Whatever the answer, the idea of an ‘Epidemiologic transition interrupted’ must be discussed. In his seminal article of 1971, Abdel Omran conceived the ‘epidemiologic transition’ as the transition from a first ‘age of pestilence and famine’ to a third age of ‘degenerative and man made diseases’ through a second age of ‘receding pandemics’.2 Of course, Abdel Omran was wrong in the idea that ‘epidemiologic transition’ will stop at the third age, but not more wrong than most demographers of the time who could hardly predict the forthcoming cardiovascular revolution, and such error has no direct relation with our discussion. The main mistake of the Omran's epidemiologic transition theory related to it having led us to believe that any population, when modernizing, will go straightforwardly to that third age, without any interruption. Reality is more complex. The main shortcoming of the theory is to underestimate both the dynamics of infectious diseases, as if no new diseases could appear, and the link between the capacity of each population to improve its health status with its success in improving its economic, social, cultural and political capacities. I would be much more comfortable with a larger concept of ‘health transition’, promoted by Julio Frenk and colleagues3 in the first issue of the ‘Health Transition Review’, that would not only open the door to successive transitions (like the cardiovascular revolution) after the classical Omran epidemiologic transition, but also clearly link each step to societal changes in the economical, social, cultural and political fields.4,5 Thus, the health transition includes several ‘epidemiologic transitions’ the first of which (at least during modern times) is the Omran one. Furthermore, the ‘health transition theory’ itself is better understood if two main factors that influence each step greatly are taken into consideration. The first one is that diseases, especially infectious diseases, had in the past and still have their own dynamics, independently from human action. Mirko Grmek6 thus tried to explain the long history of human health through the concept of ‘pathocenosis’, which takes into consideration the diversity and the changes in pathological profiles of historical populations according to the link between diseases and their environment.7 In that perspective, to perform any step of the health transition, a society must not only be successful in fighting against existing diseases, but also in preventing any new diseases to spread.8
The second point is that not all populations are able to benefit from any major technical, economical or social progress which, theoretically, opens the door to a major health improvement, and to a new step in the health transition. For a few decades, after World War II, antibiotics gave the false impression that efficient medical technologies could be transferred to any population very easily, independently of their socio-economic and political situation, and bring all populations towards the maximum of life expectancy. At Omran's theory time, such a limit was thought to be of 75 years.9 It was without taking into account major new improvements letting some countries enter a new step of the health transition while others had not completed the first one. Thus, African populations were still far under such a limit when, in the 1970s, it was rapidly crossed by western European countries, thanks to new weapons developed against cardiovascular diseases, while Eastern European countries that were very close to catching up the West in the mid 1960s, ceased to progress at the point where no more significant life expectancy increase could result from additional reduction of infectious diseases.10 Some biologists and/or demographers 11,12 then established a new theoretical limit for life expectancy at 85 years, subject to criticism from others13–15 who refused to consider any fixed limit. Indeed today, female life expectancy crossed that new limit in countries like Japan, and will soon in France, Sweden, etc ... thanks to unprecedented improvements in survival of the elderly. But, as for other major improvements, only a few pioneers succeeded while female life expectancy at oldest ages has tended to stagnate since the 1980s in countries like the United States or The Netherlands.13 In reality, each time a major change occurs in the way to fight or to prevent diseases, some pioneer populations are ready to take benefit immediately, according to their socio-economic or political situation while others need time to fulfil the required conditions to benefit. And when, finally, they do benefit, some of the first one or possibly others get new ways to make new progresses and a new gap opens between those who are able to take benefit and those who are not ready.14
Have HIV/AIDS epidemics really interrupted the ‘epidemiologic’ transition in Thailand? The question clearly refers to the Omran's original concept of epidemiologic transition. However, to clarify the question, it is better to consider Omran's ‘epidemiologic transition’ as a first step of the ‘health transition’. Indeed, when hit by HIV, Thailand was roughly at the end of that first step, very little improvement of life expectancy could be expected from additional reduction of infectious diseases. And it was not clear if it was already entering the cardiovascular revolution, but much more likely it was entering the Omran's age of ‘man-made diseases’, with a rapid increase in violent deaths (especially traffic accidents) as reported by the authors of the article. The latter fact is a good reason to avoid attributing the whole adult mortality increase of the years 1986–96 to AIDS alone. There is no doubt that AIDS played a major role, but it is very plausible that its impact was less than suggested by all-cause mortality data.
As the authors say: ‘Thailand is a good example of a country where health policy is responsive, flexible and effective. This is very well illustrated by the HIV/AIDS epidemic where preventive measures have limited the spread of the epidemic. These policy responses have led to a generalized decline in young adult mortality from HIV since the late 1990s, and this is expected to continue.’ In other words, Thailand was able to check very rapidly the new plague. It seems to me quite inappropriate to state that the epidemiologic transition is interrupted. The country faced the plague almost as soon and as efficiently as the North American or West European countries did. The case of Thailand is very different from that of African countries that lost 10, 15, 20 years of life expectancy because of AIDS. In their case, the first step of health transition was clearly interrupted because they were simply not ready to face what Abdel Omran did not imagine: the emergence of a new deadly disease.
For Thailand, the question is not to know if AIDS interrupted a step already achieved. It is more to know if that country is close to taking the second step of the health transition. Is that country on its way to follow the Western industrialized countries experience? Or is its capacity for further life expectancy improvement curtailed by insufficient ability to check man-made diseases and to fight cardiovascular diseases? The question is of great importance for all developing countries that have already reached life expectancies of about 70 years. Will they progress further? Will they be stopped in the near future? Answering the question would require a good estimation of cause-of-death mortality trends. I strongly support the call of the authors for solid data to allow such questions to be answered.
 |
References |
|---|
 Top References |
|---|
1 Hill K, Vapattanawong P, Prasartkul P, Porapakkham Y, Limm SS, Lopez AD. Epidemiologic transition interrupted: a reassessment of mrotality trends in Thailand, 1980–2000. Int J Epidemiol.
2 Omran AR. The epidemiologic transition: a theory of the epidemiology of population change. Milbank Memorial Fund Quarterly (1971) 49:509–38.[CrossRef][Web of Science][Medline]
3 Frenk J, Bobadilla JL, Stern C, Frejka T, Lozano R. Elements for a theory of the health transition. Health transition review (1991) 1:21–38.[Medline]
4 Mesle F, Vallin J. Développement économique et espérance de vie: la transition sanitaire au tournant des années soixante. In: Congrès international de la population, Vol. 2 (1993) Montréal: Liège, UIESP. 365–82.
5 Mesle F, Vallin J. The health transition: trends and prospects. In: Demography, analysis and synthesis. A treatise in demography—Caselli G, Vallin J, Wunsch G, eds. (2006) New York: Elsevier, 247–60. 2.
6 Grmek MD. Préliminaires d’une étude historique des maladies. Annales ESC (1969) 24:1473–83.
7 Biraben JN. Les pathocénoses en Europe. In: Histoire des maladies (1999) Paris: Association économie et santé, pag. mult. 1–36. In: Grmek M (coor.). (Séminaire sur l'histoire des maladies, Courchevel, 18–20 mars 1996).
8 Vallin J. Diseases, deaths, and life expectancy. In: Genus (2005) 61:279–96. (Proceedings of the International Conference "Trends and problems of the world population in the XXI century. 50 years since Rome 1954, Rome, 26–28 May 2005).
9 Vallin J, Berlinguer G. From endogenous to the maximum human life span. In: Demography, analysis and synthesis. A treatise in demography—Caselli G, Vallin J, Wunsch G, eds. (2005) New York: Elsevier, 94–116. 2.
10 Vallin J, Mesle F, Council of Europe. Trends in mortality in Europe since 1950: age-, sex- and cause-spécific mortality. In: Trends in mortality and differential mortality (2001) Strasbourg: Council of Europe Publishing. 31–186. 334 p. (Population Studies n° 36).
11 Fries JF. The compression of morbidity. Milbank Quarterly (1983) 61:397–419.[Web of Science]
12 Olshansky SJ, Carnes BA, Cassel C. In search of Mathuselah: estimating the upper limits to human longevity. Science (1990) 250:634–40.
13 Vaupel JW, Lundstrom H. Longer life expectancy? Evidence from Sweden of reductions in mortality rates at advanced ages. In: Studies in the economics of aging—Wise DA, ed. (1994) Chicago: University of Chicago Press, 79–104.
14 Kannisto V, Lauritsen J, Thatcher AR, Vaupel JW. Reductions in mortality at advanced ages: several decades of evidence from 27 countries. Population and Development Review (1994) 20:793–810.[CrossRef][Web of Science]
15 Oeppen J, Vaupel JW. Broken limits to life expectancy. Science (2002) 1029–31.
16 Mesle F, Vallin J. Diverging trends in female old-age mortality: the United States and the Netherlands versus France and Japan. Population and Development Review (2006) 32:123–45.[CrossRef][Web of Science]
17 Vallin J, Mesle F. Convergences and divergences in mortality: A new approach to health transition. Demographic Research (2004) 12–43. (Special Collection 2. Article 2, Determinants of Diverging Trends in Mortality, available at http//www.demographic-research.org/special/2/2/).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Mooney Infectious Diseases and Epidemiologic Transition in Victorian Britain? Definitely Soc Hist Med, December 1, 2007; 20(3): 595 - 606. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||