Population  2002/1

Population

2002/1 (Vol. 57)

pages200

doi 10.3917/popu.201.0171

publisher I.N.E.D

E-mail Alerts

Please provide your e-mail address to receive a Table Of Content alert every time a new issue is published.


you're reading

Mortality in Europe: the Divergence Between East and West

byFrance Meslé[*]By the same author

France Meslé, Institut national d’études démographiques, 133, bd Davout, 75980 Paris, Cedex 20, tel: 33 0(1) 56 06 21 43, fax: 33 0(1) 56 06 21 99

e-mail: mesle.at.ined.fr

andJacques Vallin[*]By the same author
Page 157-197
1

From 1965 to 1995, average European male life expectancy — with the European part of the former USSR included among European countries — barely increased. In fact, if one looks at western, northern and southern Europe, life expectancy gained six years, whereas it gained only one year in the countries of central Europe and actually lost six years in the former Soviet Union… France Meslé and Jacques Vallin describe in detail the age-specific, sex-specific and cause-specific mortality trends over these 30 years, during which the former Soviet Union and eastern Europe went through many political upheavals. In the non-formerly Communist countries, life expectancy gains were mostly due to the decline of infant mortality (an area where further gains will hereafter be limited) and of the mortality of adults aged 30 to 59 and especially over 60 (half of the gains). In the former USSR, losses affected all ages above 15, among both males and females, and the main culprits are cardiovascular diseases and violent deaths.

2

Between 1950 and 1995, the European life expectancy pattern underwent deep transformations (Vallin and Meslé, 2001). In 1950, the northwestern part of Europe had much higher life expectancy levels than the Mediterranean and eastern regions, where the standard of living was generally lower; by 1965, this opposition was already much less marked, after a spectacular rise in the life expectancies of southern and eastern Europe. Indeed, in less than 20 years, the latter had nearly succeeded in catching up with the countries of northern and western Europe (see Figure 1). Since the mid-1960s, however, the situation has once again changed entirely. The countries of eastern Europe, then governed by Communist regimes, were struck by a health crisis that considerably hampered their progress, and in some cases life expectancy even declined, especially among males; at the same time, after a period of stagnation in the 1960s, the west began to progress once again thanks to new advances in the treatment of cardiovascular diseases. By 1995, this divergent movement had transformed the European life expectancy map, and the new line of separation now corresponded to that of the former Iron Curtain.

Map of life of all expectancies in Europe, in 1950 and 1965 and 1995, by sex (in years)
Source: Vallin and Meslé, 2001.
3

Figure 2 illustrates the shift from a period of general convergence, which ended towards the mid-60s, to one of divergence between east and west: the changes in life expectancy are shown for four countries representative of four different health trends in Europe: Sweden, Italy, Poland and Russia. After World War II and until the 1960s, the life expectancies of all four countries converged strongly. In particular, Russia’s life expectancy shot up between the end of the 1940s and the end of the 1950s, catching up with the others. Within a decade, the difference between the life expectancy of Russia and Sweden decreased from over twenty years to under ten for males, and even to 3 years for females. For Italy and Poland, although the trend began earlier, the gap with Sweden was also considerably reduced. Thus, in the early 1960s, it seemed that in compliance with Abdel Omran’s pattern of epidemiologic transition (1971), European life expectancies were all converging towards a maximum that Sweden had by then almost reached.

Annual trend in life expectancy at birth in Sweden, Italy, Poland and Russia, since World War II (in years)
Sources: Andreev et al. (1998) and Meslé et al. (1996) for Russia; national statistical yearbooks for the other countries.
4

In reality, the subsequent divergence was caused by two major events. First, advances in health care were not limited to infectious diseases, as Abdel Omran had seen it. Progress in the treatment of cardiovascular disease and some forms of cancer on the one hand, and on the other the progress made in the prevention of certain “man-made” diseases, such as alcoholism, smoking or accidents, have made the increase in life expectancy possible. In Sweden, where life expectancy had apparently reached a ceiling, a sudden increase was observed in male life expectancy, which had been stagnating, while female life expectancy continued to rise. The western European countries continued to converge, as Mediterranean countries continued to advance (shown here by the case of Italy), reducing the historical gap between north and south. The second major event, however, is that the Communist countries, which had been progressing very fast in the area of infectious diseases, were totally unable to follow the new pathway adopted by the west. Thus, in 1995, after thirty years of near stagnation (among females) or even decline (among males), life expectancy in Russia is now ten years below that of Sweden for females and almost 20 years for males.

5

The divergence between east and west is a major aspect of European demography at the end of the 20th century. Our aim is to show that this phenomenon involves a profound divergence in age-specific mortality patterns; one of the reasons for this divergence is that the Communist countries did not participate in the cardiovascular revolution that enabled the West to make new advances in the progress of life expectancy.

I - Age-specific mortality patterns: increasing differences

6

The comparison within each country between life expectancy at birth in 1995 and in 1965 clearly shows that these divergent trends have split Europe in two (see Figure 3).

Life expectancy in 1995 compared to life expectancy in 1965 (in years)
Source: Vallin and Meslé, 2001.
7

On the one hand, for both males and females, the countries of northern [1]  Denmark, Finland, Iceland, Norway, Sweden.[1], western [2]  Austria, Belgium, France, Germany, Ireland, Luxembourg,...[2] and southern Europe (including the countries of the Balkans) [3]  Albania, Bosnia, Croatia, Greece, Italy, Macedonia,...[3] are located far above the diagonal line marking an absence of change between 1965 and 1995. In these countries, longevity made great strides during these thirty years. On the contrary, life expectancy having declined in all the European countries of the former USSR [4]  Belarus, Estonia, Latvia, Lithuania, Moldova, Russia,...[4] during the same period, at least as concerns males, these countries are situated far below the diagonal. The situation is not quite as bad among females, but except for Estonia and Lithuania that are located right on the diagonal, the others have all slightly regressed. The countries of central Europe [5]  Bulgaria, Czech Republic, Hungary, Poland, Romania,...[5] are about half-way between these two clearly divergent groups, near the diagonal for males and slightly above it for females.

1 - Four groups of countries selected through hierarchical analysis

8

The contrasting trends in European mortality cannot be limited to a simple divergence of life expectancy levels: indeed, age-specific changes in mortality are very different. In order to show these differences, we carried out a hierarchical analysis of age-specific death probabilities for a group of 28 European countries with a population of at least one million for which data for the years 1965 and 1995 were accessible; we distinguished between East Germany and West Germany for practical reasons linked to data availability, and also because we expected significant differences between these two regions of present-day Germany, their historical development having been very different until the 1990s [6]  The 28 countries selected are the following: Austria,...[6]. For each of these countries, we were able to collect the probability of dying figures by five year age groups [7]  More precisely at: 0 year, 1-4 years, 5-9 years, 10-14,...[7], during periods of several years centered around 1965 and 1995. The hierarchical analysis [8]  For more information on this type of analysis, see...[8] was performed [9]  Special thanks to Bénédicte Garnier and Arnaud Bringé...[9] on the logarithms of the probabilities, after a principal components analysis which made it possible to reduce the noise by limiting the final analysis to the axes that account for most of the variance, in other words the first 10 axes [10]  The analysis was performed with the SPAD software.[10]. Figure 4 illustrates the results obtained for each sex in 1965 and 1995.

Dendrograms resulting from the hierarchical analysis of age-specific probabilities of dying in 28 European countries, in 1965 and 1995, by sex
Source: Life tables published by national statistics institutes or computed by the authors.
9

In 1965, the situations among both males and females were varied enough to establish a distinction between several groups of countries in the dendrograms (i.e. tree diagrams), but no clear oppositions appear. One group does seem to stand a little apart from the rest; it is made up, however, of countries that are not all geographically close (Denmark, Greece, Netherlands, Norway and Sweden for males; Netherlands, Norway, Sweden and Switzerland, for females); moreover, the distance between the first and the second node is not much longer than that separating the two following nodes, and this pattern is repeated up to the last branches of the tree. In fact, starting from the furthest knot, one can increase the number of partitions at each level without reaching a stage where a final partition becomes obvious. A more or less large number of groups of countries can be isolated in this manner, but without bringing out any sharp contrast between them.

10

On the contrary, in 1995 the dendrograms show a striking opposition between males of eastern and western Europe (except for Portugal, whose situation is similar to that of eastern Europe). Among females, the opposition is almost as striking, but the two groups of countries are different: the countries of the former USSR, along with two countries of eastern Europe (Romania and Bulgaria) are in opposition with all the others. In both cases, the second node is very distant from the first. One must descend quite far in the distance levels in order to obtain a partition in three groups, and reach very close to the terminal elements in order to obtain a partition in four or five groups.

11

In 1995, among men, the partition in three groups clearly separates, within the eastern European group, the former USSR (Estonia, Latvia, Lithuania, Russia, Ukraine) from the other former communist countries (Bulgaria, the Czech Republic, GDR, Hungary, Poland, Romania, Yugoslavia), to which one must add Portugal. The case of Portugal is in a way an exception confirming the rule. Indeed, as will be seen below, the main factor of divergence in the mortality patterns corresponds to the modification of the risk of dying at adult ages. In this respect, however, the situation of Portugal is rather atypical for western Europe, since mortality among young adults is abnormally high, as compared to that of other age groups. In order to divide the third group of males into two sub-sets, one must go very far down in the dendrogram, where one can find two rather large groups, well defined geographically: a northern group, running from Western Germany to the British Isles and the Scandinavian countries, and a group of Mediterranean and Alpine countries, to which one must add Belgium.

12

Among females, the divisions suggested by the 1995 dendrogram are slightly different. With three groups, the distances are quite long but the geographical continuity is not as marked. Central Europe does stand in opposition to most western countries but it remains associated with Portugal, the United Kingdom, Ireland and Denmark.

13

In fact, as we know, the differences in mortality have traditionally been greater among males than among females, and this is why we decided to select the groups of countries to be used for the analysis on the basis of the hierarchy among males. We could thus be satisfied with the radical opposition of two groups, east and west. However, as we have seen, both groups contain geographically disjointed sub-groups, even though this means going quite far down in the dendrogram. As a result, we selected the following groups:

  • countries of “Mediterranean and Alpine Europe”: Austria, Belgium, France, Greece, Italy, Spain and Switzerland (the inclusion of Belgium does not destroy the group’s geographical continuity);

  • countries of “northern Europe”: Denmark, Finland, West Germany, Ireland, Netherlands, Norway, the United Kingdom, and Sweden;

  • countries of “central Europe”: Bulgaria, the Czech Republic, East Germany, Hungary, Poland, Romania and Yugoslavia (removing Portugal in order to preserve the geographical continuity of the group);

  • countries of the “former USSR”: Estonia, Latvia, Lithuania, Russia and the Ukraine.

2 - The differentiation of age-specific mortality patterns

14

For each of the above-mentioned groups, we calculated the average age-specific death probabilities by weighting the national probabilities by the size of the population, in 1965 and 1995. Figure 5 shows to what extent mortality patterns changed between those two dates. Among females, in 1965, the four mortality profiles were almost identical. Among males, in 1965, the countries of the former USSR already displayed a specific pattern of excess mortality among adults (between 20 and 60 years), whereas the difference between the central European group and the two western groups was very small.

Age-specific death probabilities by 5-year group for the four groups of countries resulting from the male dendrogram, in 1965 and 1995, by sex
Source: Life tables computed by the authors on the basis of national data.
15

In 1995, the situation differs radically. Among both males and females, the gap has considerably widened between the countries of the former USSR and those of northern and Mediterranean Europe, and central Europe has also moved apart. This change is much more marked among males than among females and can be observed at almost all ages, but especially among adults; not only the mortality levels, but also the mortality patterns are very different.

16

Figures 6 and 7 show in detail the difference between the patterns specific to each group of countries. Figure 6, in which the mortality pattern observed in 1995 is compared to that observed in 1965, shows how age-specific mortality patterns have diverged. Whereas the risk of dying was lower in 1995 among males at all ages than in 1965 in northern Europe and Mediterranean Europe, adult death risks increased in the other two groups. Among females, the mortality pattern of central Europe is less divergent from that of northern and Mediterranean Europe, but nevertheless, at adult ages (35-65 years), the mortality decline is limited.

Ratio of 1965 probability of dying to that of 1995 for the four groups of countries resulting from the male dendrogram, by sex
Source: Mortality tables calculated by the authors on the basis of national data.
17

Beyond this marked opposition between eastern and western Europe, especially among males, the overall picture must be qualified. First, as concerns the west, mortality barely declined in Mediterranean Europe among young men aged 25-35; the decline of mortality at older ages, however, was steeper in southern than in northern Europe; at more advanced ages, Mediterranean Europe’s advantage appears even more clearly among females, for whom it starts at 35. In addition, in eastern Europe, the overall disadvantage of adults is especially marked among males and in the former USSR. In the countries of the former USSR, male mortality between the ages of 40 and 50 was twice as high in 1995 as 30 years before, and at all ages between 15 and 65, the risk of dying had increased by more than 50%. Among females, the trend is not as strongly marked. In central Europe, mortality really increased only among males aged 30 to 65.

18

In Figure 7, we obtain another picture of the same phenomenon by relating the risks of dying of each group of countries to the average risks for all of Europe, both in 1995 and in 1965. Whereas in 1965, none of the four groups is very far from the average, we can nevertheless distinguish a specific pattern for the USSR, different from that of the other three groups. The opposition is even more striking in 1995 for both males and females, despite the deterioration observed in central Europe. This is for a large part due to the relative weight of the former USSR, especially Russia, in the total population of Europe, as well as to the fact that in the former USSR, the overall situation has deteriorated more than in central Europe, even though the latter did not follow the trends observed in northern and southern Europe since 1965.

Ratio of probabilities of dying for the four groups of countries to the average probabilities of dying for all of Europe, in 1965 and 1995, by sex
Source: Life tables computed by the authors on the basis of national data.

3 - Contribution of the various age groups to the variations in life expectancy

19

In order to understand the role of the different age-specific mortality patterns in the divergence of life-expectancy trends, we calculated the impact of the mortality changes at each age on life-expectancy gains or losses between 1965 and 1995. In order to do this, we applied John Pollard’s method (1982). To make the results more accessible, we re-grouped them by age as follows: the newborn, less than one year old; children aged 1 to 14; young people aged 15 to 29; adults aged 30 to 59; adults aged 60 to 74; and the elderly, aged 75 and over [11]  The impact of a mortality decline at a given age on...[11].

20

Life expectancy at birth, for all of Europe, obtained on the basis of weighted averages of the probabilities of dying, rose from 67.0 years in 1965 to 67.7 in 1995 among males, and from 73.5 to 76.9 years among females (see Table 1). In thirty years, males only gained less than one year on average, and females gained a little over three years. These rather limited gains are clearly due to the losses incurred in the east, which to a large extent offset the advances made in the west. Among females, mortality decline in all age groups contributed to lengthening female life expectancy. However, the impact remains very low at ages 15-29 (barely a tenth of a year) and at ages 30-59 (0.2 years). The greatest gain by far is obtained thanks to the reduction of infant mortality (1.2 years), followed by the gains at ages 60-74 and 75 and over (0.8 years for each age group).

Table 1 –  Contribution of changes in age-specific mortality to the gains and losses in life expectancy between 1965 and 1995 for the four groups of European countries (in years)
21

Among males, the decline of infant mortality has had a similar if not greater impact than among females (1.3 years), but this gain is cancelled out by the negative impact of the mortality increase at 30-59 (–1.25 years). Male longevity decreased slightly as well, because of the increase in male mortality at 15-29 years (–0.1 years) and has not gained much through decreases at other ages. This average pattern of change, however, — which we are presenting here because, to our knowledge, it has not yet been described — represents of course nothing more than the outcome of contrary trends that are easier to analyze to the extent that they pertain to relatively homogenous groups.

22

On the male side, during these 30 years, life expectancy at birth gained 6.7 years in Mediterranean Europe and 5.6 years in northern Europe, but only 1.3 years in central Europe; conversely, the former USSR lost 6.3 years (see Table 1). The decrease of infant and juvenile mortality has led to gains in life expectancy in all of Europe. Indeed, thanks only to the decline of infant mortality, central Europe gained over two years, and Mediterranean Europe about as much; northern Europe, however, gained only half as much, and the former USSR barely over half a year. The main difference between the two large groups of countries is that in the west, the mortality decline at other ages added quite a significant amount to the gains obtained at young ages, whereas in the east, these added gains are practically negligible, when they are not replaced by downright losses (Figure 8).

Contribution of age-specific mortality changes to the gains and losses in life expectancy between 1965 and 1995, for the four groups of European countries (in years)
Source: Life tables computed by the authors on the basis of national data.
23

In Mediterranean Europe, the progress of male longevity is due more to the decline of mortality after age 60 (2.9 years) than to that under 15 (2.5 years). This is also true in northern Europe, though the gains are not quite as large (2.4 years over 60 and 1.6 under 15). In central Europe, because of the increase in male mortality between the ages of 30 and 75, life expectancy lost almost all it had gained from the decrease in mortality at all other ages (–1.5 years against +2.8). In the countries of the former USSR, all the age groups over 15 are affected by a mortality increase resulting in a decline of life expectancy. These losses are very heavy at 30 59 (–4.5 years). Overall, these countries would have lost 7 years of life expectancy if the decline of infant and child mortality had not added one year.

24

If, in northern Europe, females have gained almost exactly the same number of years of life expectancy as males (5.6 years), everywhere else the gender gap has grown wider. In southern Europe, female life expectancy gained 7.7 years, that is, one year more than male life expectancy. In central Europe, females have gained 4.1 years, nearly three times more than males, and in the former USSR, they lost only 2 years, or over three times less than males (see Table 1).

25

In all countries, the decrease in infant and child mortality contributed to the increase in female life expectancy, just like it did for male life expectancy. These gains, which are slightly lower than those observed among males, follow the same geographical variations. Three main points can be noted with respect to female mortality trends as compared to male trends: in the first place, in both groups of western countries, life expectancy gains were particularly high at older ages: of the 5.6 years gained in northern Europe, 3.2 (57%) were gained thanks to the decrease in mortality after 60, and in southern Europe, the mortality decrease after 60 accounted for 4.2 years (54%) out of 7.7. Next, in central Europe, the impact is positive at all ages, even if it is rather slight for young adults. Finally, the losses incurred by females in the countries of the former USSR are much smaller. Females lost only 1.2 years of life expectancy because of the increase in mortality at 30-59 years, against 4.5 years for males.

II - Eastern Europe: the cardiovascular revolution has yet to begin

26

Unfortunately, at the level of causes of death, it is impossible to continue the analysis of the differences on the basis of the four groups determined by the hierarchical analysis, because of the great number of discrepancies in the available data. Although WHO regularly updates its database on cause-specific mortality for most European countries [12]  Mortality data, http:// www. who. int/ whosis.[12], the 28 countries referred to here for the hierarchical analysis are far from being evenly represented. In particular, they are not all listed for the period 1965-1995; in addition, during this period, the International Classification of Diseases (ICD) underwent three revisions [13]  Eighth, Ninth and Tenth Revisions.[13], and each of these took effect at different times in different countries, resulting in a break in the continuity of statistics. For these reasons, we decided to select within each of the geographical groups mentioned above, one country among the most populated for which sufficient data were available: the United Kingdom for northern Europe, France for Mediterranean Europe, Poland for central Europe, and Russia for the former USSR. For each of these countries, we were able, through various reconstructions [14]  For France: Vallin and Meslé, 1988,1998; for the United...[14], to establish continuous series of age-specific mortality data covering the entire period.

1 - Main trends in causes of death

27

Figure 9 shows the annual trends of standardized mortality rates [15]  Established on the basis of the rates by five-year age...[15] for five large groups of causes of death since 1950: infectious and respiratory diseases [16]  Chapters I (items 001 to 139) and VIII (items 460 to...[16], neoplasms [17]  Chapter II (items 140 to 239) of ICD-9.[17], cardiovascular diseases [18]  Chapter VII (items 390 to 459) of ICD-9.[18], other diseases [19]  Chapters III to VI (items 240 to 389) and chapters...[19] and violent deaths [20]  Chapter XVII (items 800 to 999) of ICD-9.[20]. The deaths attributed to an ill-defined cause [21]  Chapter XVI (items 780 to 799) of ICD-9.[21] were redistributed proportionately [22]  Over this period, the proportion of deaths due to an...[22].

Standardized mortality rates since 1950 for 5 large groups of causes of death in the United Kingdom, France, Poland, Russia, both sexes*
*

Figure 9 and the following figures are constructed as semi-logarithmic scales in order to make it possible to compare the paces of change at very different mortality levels.

Source: Vallin and Meslé, 2001.

a - The decline of infectious and respiratory diseases

28

We grouped the causes of death included in chapters I (infectious and parasitic diseases) and VIII (diseases of the respiratory system) of ICD-9 in order to provide a global picture of the trends in mortality due to infectious diseases, since mortality due to the diseases listed in chapter VIII is to a large extent conditioned by mortality due to infectious diseases of the respiratory system (influenza, pneumonia, acute bronchitis in particular). Mortality due to the latter group of diseases has steadily declined over the past fifty years, though it is still governed by the fluctuations of influenza epidemics (see graph 1 of Figure 9). However, two important re marks must be made at this point.

29

On the one hand, in the United Kingdom, whose mortality for this group of causes was the lowest of the four countries in the 1950s, the decline was very slow, and today its mortality due to these causes is higher than in the other three countries. At the end of the 1990s, the British standardized rate was twice as high as that of Poland or France. This trend, which is specific to Britain, is nevertheless partly due to an artifact. British statistics classify as pneumonia a host of chronic illnesses, cardiovascular diseases in particular, for which pneumonia represents only the terminal phase. In 1984, the Office of Population Censuses and Surveys (OPCS) made an attempt to improve the definition of the primary cause of death (OPCS, 1985), and this effort resulted in a sudden drop in the incidence of pneumonia (Meslé, 1995); after 1992, however, British statistics went back to their previous practice (OPCS 1995), which means that as of 1993, the mortality curve due to infectious and respiratory diseases followed the previous trend. In this case, we adjusted the crude data in order to produce a curve based on a constant definition, even if the British definition cannot be compared to that of other countries [23]  We are aware that the adjustment should be done the...[23].

30

On the other hand, although in Russia, as in the other three countries, the general trend is one of decline, the years 1980-1990 were marked by a strong fluctuation, in this case not due to an artifact. Indeed, Gorbachev’s anti-alcohol campaign had a significant impact on mortality from infectious and respiratory diseases; thanks to this campaign, in 1985-86, Russian males gained three years of life expectancy and Russian females over one year. Conversely, in 1991-1994, mortality due to these causes was affected by the economic and social crisis linked to the brutal transition to a market economy, and the losses in life expectancy were 5.7 and 3.0 years respectively. Now that the crisis is over, the standardized mortality rate due to infectious and respiratory diseases has gone back to the overall trend that preceded this wide fluctuation.

b - The stability of neoplasms

31

Compared to other causes of death, mortality due to neoplasms seems stable (see graph 2 of Figure 9). In addition, levels are extremely similar from one country to another. This stability in time and homogeneity in space actually masks variations in opposite directions among the main forms of cancer, depending on their etiology. Thus, in all countries, mortality due to cancer of the stomach or the uterus has been constantly declining since the 1950s. Conversely, lung cancer, which is directly linked to smoking, increased strongly everywhere in the 1950s and 1960s. These opposing trends are also subject to strong geographical variations, as shown in Figure 10, which illustrates lung cancer mortality trends since 1965.

Standardized rates of mortality from lung cancer since 1965 in the United Kingdom, France, Poland, and Russia, by sex
Sources: WHO database for Poland and the United Kingdom; Vallin-Meslé database for France; Meslé et al. (1996) for Russia.
32

As we know, this type of mortality varies considerably by sex, and for this reason it is important to treat males and females separately. The difference between countries is also considerable, especially for females. With respect to males, a reversal has been observed in recent years. The United Kingdom was first to change towards the end of the 1960s; today, the British standardized mortality rate from lung cancer is lower than it was in the early 1960s and it is the lowest of the four countries studied, whereas it was the highest in 1965. In France, there is no clear reversal, but the increase was curtailed towards the end of the 1970s. In Russia and Poland, the rate continued to increase rapidly until the end of the 1980s; in recent years, the increase has been slowing down in Poland, and in Russia there may be an incipient reversal (if the change observed in the 1990s reflects reality). These differences among countries and trends are for the most part due to differences in smoking habits and anti-smoking policies.

33

The United Kingdom was the first country to react to the growing tobacco addiction, following a famous survey carried out among physicians on the smoking habits and incidence of cancer among their patients; the first consequence of this survey was the fact that doctors themselves became aware of the danger and reduced their own consumption (Royal College, 1962). The ensuing prevention campaign led to a drop in smoking and a decisive decrease in mortality due to lung cancer.

34

In France, anti-smoking campaigns began only later. Nevertheless, thanks to the policies developed by the French Ministers of Health, Simone Veil in the 1970s and Claude Evin in the 1980s, who implemented several series of measures (awareness campaigns, prohibition of cigarette advertising, higher taxes on tobacco, prohibition of smoking in public places, etc.), the increase in male mortality due to lung cancer was curbed.

35

In Poland and Russia, however, no such measures were taken until the early 1980s when the first such campaigns were launched in Poland (Przewozniak and Zatonski, 1993), a fact that probably explains why lung cancer mortality continued to increase until the end of the decade.

36

On the other hand, the fact that the trend was recently reversed in Russia is more difficult to understand, since no extensive anti-smoking measures have been implemented, and since the limited available data (McKee et al., 1998) does not show any decline in consumption. The economic crisis and the breakdown of industrial production that followed the collapse of the communist regime and the transition towards a market economy may be part of the reason, since a great many polluting activities were massively reduced. However, it is hard to tell whether this apparent decrease in mortality from lung cancer in Russia corresponds to actual fact.

37

Female mortality from lung cancer is much lower than that of men in all countries, but its increase began much more recently in all countries as well. In countries where anti-smoking measures were implemented and had a significant effect on smoking habits among men, these positive results came later among women. In the United Kingdom, for example, the increase in female mortality from lung cancer was curbed only towards the end of the 1980s, and the decrease has only barely begun in recent years. In France, the trend is clearly on the increase [24]  Since the early 1990s, we can observe a slight decrease...[24], although at a level that is three times lower than the one observed in the UK. There are two reasons why women are less receptive to anti-smoking campaigns. First, female consumption evolves at much lower levels than male consumption and concerns a more specific section of the population, less sensitive to the arguments put forth in the awareness campaigns; secondly, smoking can be perceived in a sense as a symbol of emancipation, and it could be that in this case, the symbol is more important than the health concerns.

38

In Poland, female mortality from lung cancer began to increase earlier than in France, and there is no sign of decrease. This makes the apparent decrease observed in Russia among both males and females all the more suspect. The decline of mortality due to lung cancer in Russia may be the result of a statistical artifact and may not correspond to reality.

c - Divergence in mortality from cardiovascular diseases

39

The divergence between mortality rates in the west and in the east is mainly due to cardiovascular diseases. Indeed, in all countries and for the entire period, this group of diseases represents the main cause of death, and it is also here that the divergence in mortality trends appears most obviously. Whereas in the beginning of the 1960s, the standardized rate of mortality from cardiovascular diseases was about the same in Russia, Poland and the United Kingdom, it later increased in the first two countries and decreased in the third; as a result, the Polish rate is twice as high as the British rate while the Russian rate is three times higher (see graph 3 of Figure 9). Over the entire period, France remained at a much lower level and, better still, since the decrease of mortality due to cardiovascular diseases was more rapid there than in the United Kingdom, the gap between the two countries widened considerably over 50 years (from 30% in 1950 to 60% in 1996). This divergence between France and the United Kingdom is typical of the advantage, mentioned earlier, of Mediterranean countries over northern countries in terms of survival at very old ages. In France, the standardized mortality rate due to cardiovascular disease is three times lower than in Poland and four times lower than in Russia.

40

It is interesting to note, however, that cardiovascular mortality has undergone a recent reversal in Poland. It ceased to grow in the mid-1980s, and since 1990 has been declining at a pace similar to that of the United Kingdom. Poland is still far from having caught up with western countries, but this reversal may be a sign that, like the Czech Republic or Hungary, where the same phenomenon has been observed (Vallin and Meslé, 2001), Poland is about to come out of the long crisis affecting eastern Europe since the mid-1960s.

41

Given the crucial impact of this group of causes on European mortality trends, it must be examined in greater detail. Several problems of observation are encountered, however. Indeed, even if we limit our observation to a very small number of sub-groups of medical causes of death, a serious problem lies in the difference between interpretations of medical nomenclature. Figure 11 shows the trends since 1965 for four groups of cardiovascular disease: ischaemic heart disease, other heart diseases, cerebrovascular diseases, and other diseases of the circulatory system.

Standardized rates of mortality from four types of cardiovascular diseases since 1965 in the United Kingdom, France, Poland and Russia, both sexes
Sources: WHO database for Poland and the United Kingdom; Vallin-Meslé database for France; Meslé et al. (1996) for Russia.
42

Two discrepancies are immediately visible in this figure. The first one pertains to France and the United Kingdom. In the United Kingdom, mortality due to ischaemic heart disease is much higher than in France — higher than expected given the differences observed above, in the context of cardiovascular diseases taken globally — while mortality from other heart diseases is much lower than in France. What appears clearly is that deaths from cardiovascular diseases are not identified in the same way in the United Kingdom and in France: the former country tends to classify heart diseases more often in the category of ischaemic heart disease and the latter more often among “other heart diseases”. It could be that in France, the apparent convergence until the mid-1980s of mortality due to ischaemic heart disease and to other heart diseases is due, at least in part, to the gradual improvement of the diagnosis of ischaemic heart disease. In any case, it seems that the comparison between both countries should be based on both forms of disease taken together.

43

Similarly, the distinction established by Poland between cerebrovascular diseases and other diseases of the circulatory system differs from that made in other countries. Though the first group accounts for a large share of mortality in all the other countries, whereas the second plays a secondary role, in Poland it seems to come in last as a cause of death, whereas the other diseases of the circulatory system largely surpass all other cardiovascular diseases. Once again, in order to compare the four countries selected, it appears necessary to group cerebrovascular diseases together with the other diseases of the circulatory system. This was done in Figure 12.

Standardized rates of mortality from heart diseases and cerebrovascular diseases and other diseases of the circulatory system since 1965 in the United Kingdom, France, Poland, and Russia, for both sexes
Sources: WHO database for Poland and the United Kingdom; Vallin-Meslé database for France; Meslé et al. (1996) for Russia.
44

Two observations should be made at this point. The diseases that account for most of the large gap between eastern and western Europe are cerebrovascular diseases and other diseases of the circulatory system (see graph 2 of Figure 12). Whereas in the early 1950s, the four countries were at relatively close levels, the difference between Poland or Russia and the United Kingdom is now 4 to 1, and between Poland or Russia and France even 5 to 1. Conversely, the recent decline of these diseases triggered the reversal of the trend for all cardiovascular diseases in Poland in the early 1990s (see Figure 9).

45

Of course, heart diseases also play a role in this divergence, but to a lesser extent (see graph 1 of Figure 12). In Poland, mortality due to heart disease did not increase during the entire period (1965-1995) and it has even slightly decreased since the early 1980s. On the other hand, these diseases play a significant role not only in the difference between France and the United Kingdom, but also in the divergence between Poland and Russia, so much so that today, mortality rates from this cause are similar in Poland and in the United Kingdom.

46

In Russia, the fluctuations caused by the 1985 anti-alcohol campaign and the 1992-1994 economic and social crisis seem to have had a greater impact on heart diseases than on cerebrovascular diseases. This may seem surprising, since it is generally considered that cerebrovascular diseases are more vulnerable to alcohol consumption. Recent studies have shown, however, that the effect of alcohol on cardiovascular diseases depends on the type of consumption (Britton et al., 1998). A regular consumption of beverages with low alcohol content, such as wine in France, favors the development of cerebrovascular diseases but can also protect against heart disease. Conversely, the consumption of large quantities of beverages with high alcohol content such as vodka (binge drinking) in Russia can cause acute heart attacks. This sheds some light on why sudden variations in the consumption of alcohol, such as the 1985 drop in Russia, had an immediate effect on mortality from heart disease.

47

Since 1950, France and the United Kingdom have in common the fact that they experienced two phases of an evolution. At first, until the end of the 1960s, the two main components of cardiovascular mortality stagnated among males and declined only slightly among females (Vallin and Meslé, 2001). It appears that in those days, from a medical and social point of view, not much was achieved regarding this group of diseases that, owing to the reduction of infectious diseases, became the main cause of death. The trends observed then in France and the United Kingdom were very typical of what was happening in all industrialized countries. Beginning in the early 1970s, however, the different subgroups of cardiovascular diseases began to decline one after the other at a growing pace. The trend began with cerebrovascular diseases in the early 1970s, thanks in particular to the generalization of blood pressure check-ups (along with campaigns against alcoholism in France). Then heart diseases (both ischaemic and non-ischaemic) also began to decrease thanks to the development of new technologies and therapies at the turn of the 1980s: coronary surgery, betablockers, thrombolytic drugs, the organization of emergency services (e.g. the SAMU in France), monitoring and treatment of hypercholesterolemia, and last but not least the drop in tobacco consumption. Nevertheless, the gap between the two countries widened during this period. Some researchers have described what they call “the French paradox” (Criqui and Ringel, 1994); in fact, what we have is a more global difference between the Northern and Southern countries of western Europe, probably linked to lifestyle and in particular to diet (Keys, 1986; Kushi et al., 1995).

48

Eastern Europe, however, did not share in these advances, that made an obvious difference in the west starting in the 1970s and gained momentum during the 1980s; on the contrary, mortality due to cardiovascular diseases increased in the east, where the communist regimes relied almost too exclusively on the centralized administration of modern health care, whereas the struggle against cardiovascular diseases, unlike that against infectious diseases, requires important changes in individual behaviour and the active participation of citizens in the management of their own health care. In addition, the economies of these countries were involved in a ruinous arms race and space competition with western countries, and as a result, the means necessary to create an efficient health system in the area of chronic diseases were unavailable (long-term care of serious diseases requiring costly therapies, widespread use of sophisticated technical equipment, creation of a dense network of emergency medical services, etc.)

d - Other diseases

49

The fourth graph of Figure 9 illustrates mortality trends due to all other diseases — a residual group, quite heterogeneous by definition, and for this reason of little interest. Nevertheless, in three of the four countries, France, Poland and the United Kingdom, an overall decline since the 1950s can be noted. We also observe that mortality from all these causes is much higher in France than in the United Kingdom. This is mainly due to a higher mortality rate for diseases of the digestive system, which in turn are linked to chronic alcoholism, a common problem in France. The relatively low rate for this group of causes in Russia until the early 1990s is probably due to a widespread underestimation of all these causes, since Soviet statistics tend to attribute these deaths to cardiovascular diseases. On the other hand, the 1993-1994 increase in mortality linked to these causes reflects the impact of the economic and social crisis on a large number of diseases included in this category.

e - Violent deaths: a contrasting picture

50

After cardiovascular diseases, violent deaths are the second factor of divergence between east and west. In the west, after a historical phase of expansion linked for the most part to the perverse effects of industrial development, and of road traffic in particular, the mortality from violent causes began to decline towards the end of the 1960s, a little earlier in the United Kingdom than in France (see graph 5 of Figure 9). Western countries were thus able to curb the rise of what Abdel Omran (1971) called “man-made diseases”, which he considered part of the final phase of his “epidemiologic transition”. As we now know, this “final” phase turned out not to be the last, and the curbing of violent deaths, together with that of cardiovascular diseases, led to the epidemiological revolution of the 1970s, which in most industrialized countries made it possible for life expectancy to resume its increase.

51

In the East, on the contrary, the mortality from violent causes continued to rise rapidly until the early 1980s. Since then, Russia and Poland have gone their separate ways. In Poland, the frequency of violent deaths increased, though less rapidly, with many fluctuations until the early 1990s, at which point the rate seems to have begun to decrease. As happened with cardiovascular mortality, Poland may now be entering the new phase of health progress, 25 years after the west. In Russia [25]  Death due to alcoholism is for a large part included...[25], the very wide fluctuation of the 1980-1990 decade makes it difficult to discern long-term trends, that appear to remain negative; indeed, this group of causes was more than any other affected by the 1985 anti-alcohol campaign and the economic and social crisis of 1993-94. But in the long run, even after the decline of recent years, the standardized mortality rate from violent causes is still twice as high in 1998 as it was in 1965.

52

Among different forms of violent death, homicide plays an exceptional role in eastern European countries and especially in Russia. Figure 13 illustrates the homicide trends for males, who are clearly more vulnerable to this type of death than females. Over the entire period, the standardized rate is much higher in Russia than in France and in the United Kingdom (about 10 times higher in the mid-1960s and 30 to 40 times higher in the 1990s).

Standardized rates of mortality from homicide since 1965 in the United Kingdom, France, Poland and Russia, males
Sources: WHO database for Poland and the United Kingdom; Vallin-Meslé database for France; Meslé et al. (1996) for Russia.
53

Compared to other forms of violent deaths, deaths by homicide were less affected by the anti-alcohol campaign and on the contrary exacerbated by the economic and social crisis of 1993-1994. In Poland, the situation, which in the mid-1960s was similar to that of the western countries, continued to deteriorate until the early 1990s, but today seems to have stabilized at a much lower level than in Russia.

2 - Contribution of the evolution of disease to the increase or decline of life expectancy

54

In order to evaluate the contribution of different groups of causes of death to the divergence of life expectancies since 1965, when the “great divide” occurred between east and west, it is necessary to separate this period into at least two sub-periods. From 1965 to 1984, the divergence between east and west remained constant and the comparison between the four selected countries is quite simple; but since 1984, as a result of the wide fluctuations observed in Russia and the recent reversal in Poland, each country has been experiencing different trends and a different pace of change.

a - From 1965 to 1984

55

In order to make our point clearer, we will focus on male life expectancy, the situation being similar though less marked for females. From 1965 to 1984, life expectancy gained about three and a half years in France and in the United Kingdom, whereas in Poland it stagnated and in Russia it lost about three years. In the United Kingdom, most of the progress came from the decline in mortality from diseases of the circulatory system, thanks to which life expectancy gained 1.4 years, or 42% of the total gain (see Table 2). In France, the decline of cardiovascular mortality also led to a gain of 1.4 years, but unlike the United Kingdom, France lost 0.4 years because of the increase in cancer mortality, and gained much more than the United Kingdom (1.4 years instead of 0.7) from a decrease in the group of “other diseases”. In fact, during this period, the decline in the latter group of causes concerned mainly young children; the difference between France and the United Kingdom is thus essentially due to the contribution of the infant mortality decline (see Figure 14). The difference in cancer mortality trends affects almost exclusively the mortality of adults aged 45 to 59.

Table 2 –  Contribution of changes in large groups of causes of death to the gains and losses in male life expectancy between 1965 and 1984 in the United Kingdom, France, Poland and Russia (in years)
56

Diseases of the circulatory system had the greatest impact on life expectancy trends in Russia as well, but this time the impact was negative, since of the 2.8 years lost by Russia between 1965 and 1984, 1.8 can be ascribed to the increase of cardiovascular mortality increase; these losses are for the most part concentrated in the 40-70 year age group (see Figure 14). Violent deaths (including, in Russia, a large share of mortality due to alcoholism) also led to a considerable loss in life expectancy (1.4 years) affecting mainly younger adults (aged 20 to 50). The impact of other groups of causes of death is comparatively low, and in the end, these heavy losses are only slightly offset by the 0.4 year gain in the area of infectious diseases, which, as shown in Figure 14, is mainly due to the decline in infant mortality.

Contribution of changes in large groups of causes of death to the gains and losses in male life expectancy between 1965 and 1984 in the United Kingdom, France, Poland and Russia (in years)
Source: Calculations made by the authors from data cited in Figure 10.
57

In Poland, despite the fact that life expectancy remained just about constant during the entire period, the situation does not differ greatly from the Russian situation (see Table 2), with one difference: Poland was able to gain 2.2 years thanks to the decrease in mortality from infectious and respiratory diseases, itself the consequence mainly of the decline in infant mortality (see Figure 14). These gains were almost entirely cancelled out by the losses — almost as heavy as in Russia — from diseases of the circulatory system (–1.5 years), whereas the lower impact of violent deaths goes along with a similarly negative impact of neoplasms (– 0.6 years for each).

b - Since 1984

58

After 1984, while Western European countries continued to advance, a sharp difference began to emerge in the east between Polish and Russian trends. Russian life expectancy underwent very wide fluctuations, while Polish life expectancy stagnated until 1988 and even declined significantly between 1988 and 1991, before beginning to increase in 1992. For this reason, we established several sub-periods for each country in order to account for the impact of different causes of death on the increase or decrease of life expectancy.

Poland: the situation improves

59

Between 1981 and 1991 the Poles lost 0.7 years of life expectancy, but from 1991 to 1996 they gained over 2 years (see Table 3).

Table 3 –  Contribution of changes in large groups of causes of death to the gains and losses in male life expectancy between 1984 and 1991 and between 1991 and 1996 in Poland (in years)
60

The 1984-1991 trend prolonged the earlier trend (see Figure 15) in the sense that Poland still continued to lose years in life expectancy at adult ages (which were still to a large extent offset by gains acquired through the decline in infant mortality). Cardiovascular diseases, violent deaths and cancers continued to take a heavy toll on the expectation of life at birth (respectively –0.6, –0.3 and –0.2 years).

Contribution of changes in large groups of causes of death to the gains and losses in male life expectancy between 1984 and 1991 and between 1991 and 1996 in Poland (in years)
Source: Calculations made by the authors from data cited in Figure 10.
61

The 1991-1996 period marked a complete reversal, with an overall decline of cardiovascular diseases. Life expectancy increased at all adult ages (+1.1 years overall). It seems that a large share of this progress was due to improved diet. Indeed, since milk, butter, and red meat had become too expensive, they were in part replaced by fruit, vegetables and vegetable fats (Zatonski, 2000). Thus, between 1989 and 1994, the consumption of butter per person and per year dropped from 8.8 kg to 3.8 kg, while the consumption of vegetable fats increased from 8 to 13 kg (Sekula et al., 1996). In addition, the decrease in violent deaths led to an increase in life expectancy at all ages where this cause had previously had a negative impact (+0.5 years). As for mortality from cancer, it remained stable during this period and its impact on the variation of life expectancy became almost negligible.

The great Russian fluctuation

62

In order to follow the variations of life expectancy in Russia since 1984, it is important to make a distinction between the consequences of the anti-alcohol campaign implemented by Gorbachev and those of the economic and social crisis of 1992-1994. Thanks to the anti-alcohol policies, life expectancy first rose suddenly (1984-1987), but this rise was followed by a decline as the campaign lost steam (1987-1992). During the following two years (1992-1994), the transition to a market economy accelerated the decrease in life expectancy, but as the population adapted to its new conditions, life expectancy once again began to rise (1994-1997).

63

From 1984 to 1987, most of the life expectancy gains were the result of Gorbachev’s anti-alcohol policy, since they mainly reflect a decline in violent deaths between the ages of 25 and 60 (see Figure 16). This decline, which as we know was entirely due to the reduction of alcoholism, is related to a smaller decrease in the mortality from cardiovascular diseases between the ages of 45 and 70 (also because of restrictions on alcohol consumption). Overall, in three years, the drop in the incidence of violent deaths led to an increase in life expectancy of 1.8 years, and the drop of cardiovascular diseases added 0.7 years (see Table 4).

Table 4 –  Contribution of changes in large groups of causes of death to the fluctuations of male life expectancy in Russia since 1984 (in years)
64

During the following years (1987-1992), the loosening of the anti-alcohol measures triggered the opposite phenomenon: violent deaths became more frequent, as well as cardiovascular diseases. This period is the symmetrical opposite of the previous one (see graphs 1 and 2 of Figure 16), since the increase in violent deaths led to a loss of 2.2 years of life expectancy, and that of cardiovascular mortality to a loss of 0.6 years.

65

After 1992, life expectancy at first declined even faster, but finally it began to rise, once again very symmetrically, but with a different cause-specific pattern (see graphs 3 and 4 of Figure 16). From 1992 to 1994, all causes at almost all ages are responsible for the drop in life expectancy. Violent deaths continue to play an important role, always at adult ages (20 to 60 years), but cardiovascular diseases are just as important at all ages from 30 to 80. Almost all the other groups of causes (infectious and respiratory diseases, diseases of the digestive system, other diseases) play a significant role, with the sole exception of neoplasms. We observe a general resurgence of all diseases tied to the deterioration of living conditions, and their impact on the decline in life expectancy is directly proportional to their traditional epidemiological weight. Overall, out of a total loss of 4.5 years of life expectancy, violent deaths and diseases of the circulatory system account for 1.7 years each, diseases of the respiratory system for 0.4 years, diseases of the digestive system and infectious diseases for 0.2 years each and other diseases for 0.3 years (see Table 4).

Contribution of changes in mortality by age and large groups of causes of death to the fluctuations of male life expectancy in Russia since 1984 (in years)
Source: Calculations made by the authors from data cited in Figure 10.
66

Conversely, starting in 1994, when the population began to adapt to the crisis situation, life expectancy almost returned to its previous level, thanks to a decrease in the same causes that had previously been increasing.

67

Unlike what seems to be happening in Poland, for Russia it is not clear whether this trend represents a lasting reversal of the negative trends that preceded the great fluctuations of the 1980s-1990s. Indeed, the 1987 life expectancy level is still below that of 1984. We also know that the pace of the increase slowed down considerably in 1998 and that there was a new drop in 1999.

Increased life expectancy for the oldest old in western Europe

68

The situation is much simpler in France and the United Kingdom, since life expectancy continued to increase steadily during the entire period. A distinction between the sexes is useful for the purpose of exploring a new area of progress: life expectancy gains at very old ages.

69

From 1984 to 1997, males gained three years of life expectancy in the United Kingdom and almost 3.5 in France, while females gained 2.1 and 2.9 years respectively (see Table 5). Thus, during this period and especially in the United Kingdom, the gap between males and females diminished; a significant factor in this progress is the reduction of mortality due to cardiovascular disease among the oldest old. Overall, depending on country and sex, the decline in diseases of the circulatory system led to a life expectancy gain of 1.4 to 2 years. In comparison to Figure 14 (1965-1984), Figure 17 (1984-1997) shows the impact of reducing the incidence of such diseases on the extension of male life expectancy. The comparison between males and females also clearly shows to what extent the life expectancy of females, which is higher than that of males, has gained mainly from the decline in cardiovascular diseases among the oldest old.

Table 5 –  Contribution of changes in large groups of causes of death to gains and losses in life expectancy between 1984 and 1997 in France and in the United Kingdom, by sex (in years)
Contribution of changes in mortality by age and large groups of causes of death to gains and losses in life expectancy between 1984 and 1997 in France and in the United Kingdom, by sex (in years)
Source: Calculations made by the authors from data cited in Figure 10.

Conclusion

70

As we have seen, over a period of thirty years, the European mortality landscape has undergone deep transformations, with a growing gap between east and west and a closing of the gap between north and south.

71

In the east, the situation is disastrous for the countries of the former USSR, such as Russia, that have lost several years of life expectancy. Since the mid-1960s, these countries have not been able to curb the increase in man-made diseases, nor have they been able to acquire the means to combat mortality from cardiovascular diseases. The result is not only a decrease in life expectancy, but also a deep distortion of the age-specific mortality curve, with a severe handicap at adult ages. Though the fluctuation caused by the 1985 anti-alcohol campaign generated some improvement, it was short-lived; since then, there has been no sign of any genuine turn for the better, and it is to be feared that the countries of the former USSR have gone back to their long-term negative trends. In some countries of central Europe, where the life expectancy decline was not as considerable, it seems on the contrary that there have been some encouraging signs of improvement since the beginning of the 1990s, namely a decline in man-made diseases and mortality due to cardiovascular diseases. However, what we have observed in Poland cannot be generalized to all the countries of central Europe: whereas the Czech Republic, Slovakia and Hungary are clearly following an upward trend, Bulgaria and Romania are still lagging behind.

72

Eastern Europe’s failure to keep up with western countries during the past thirty years is all the more striking since the latter countries made a great deal of progress during those years, despite some wavering in the 1960s. Southern Europe has completely caught up with northern Europe and mortality patterns have become increasingly homogenous, to such an extent that the hierarchical analysis carried out for all of Europe masked the remaining differences. However, not all differences have been erased. The countries of southern Europe have a distinct advantage over their northern counterparts with respect to survival at very old ages, thanks to the massive decline in mortality from cardiovascular diseases, in particular among females. On the other hand, they are still burdened with excess mortality around age 20, essentially as a result of automobile accidents, especially among males.


REFERENCES

  • Andreev Evgueny, 1982, “Metod komponent v analize prodoljitel’nostijihni” [Component method in the analysis of life expectancy], Vestnik Statistiki, n° 3, March, pp. 42-47.
  • Andreev Evgueny M., Darskij Leonid E., Khar’kova Tatiana L., 1998, Demozrafqueskaw qsmorqw Rossqq: 1927-1959 [Population History of Russia: 1927-1959], Moscow, Informatika, 187 p.
  • Baudier François, Velter Annie, 1998, “Tabac”, in: Jacques Arènes, Marie-Pierre Janvrin, François Baudier (dir.), Baromètre santé 97/98. Jeunes, Vanves, Comité français d’éducation pour la santé, pp. 158-177.
  • Britton Annie, McKee Martin, Leon David, 1998, Cardiovascular disease and heavy drinking: a systematic review, London, LSHTM, 92 p.
  • Criqui Michael H., Ringel Brenda, 1994, “Does diet or alcohol explain the French paradox?”, The Lancet, vol. 344, n° 8939-8940, pp. 1719-1723.
  • Hertrich Véronique, Meslé France, 1998, “Mortalité et politiques socio-économiques: le cas de 4 pays d’Europe centrale et orientale”, in: Morbidité, mortalité: problèmes de mesure, facteurs d’évolution, essai de prospective, Colloque international de Sinaia (2-6 septembre 1996), Paris, AIDELF, pp. 395-412.
  • Keys Ancel., Menotti Alessandro, Karvonen Martti J. et al., 1986, “The diet and the 15-year death rate in the seven country study”, American Journal of Epidemiology, vol. 124, pp. 903-915.
  • Kushi L. H., Lenart E. B., Willett W. C., 1995, “Health implications of Mediterranean diets in light of contemporary knowledge. 1. Plants foods and diary products. 2. Meats, wine, fats and oils”, American Journal of Clinical Nutrition, vol. 61, pp. 1407S-1427S.
  • Lebart Ludovic, Morineau Alain, Piron Marie, 199, Statistique exploratoire multidimensionnelle, Paris, Dunod, 440 p.
  • McKee Martin, Bobak M., Rose R., Shkolnikov Vladimir, Chenet Laurent, Leon David, 1998, “Patterns of smoking in Russia”, Tobacco Control, n° 7, pp. 22-26.
  • Meslé France, 1995, “L’enregistrement des causes de décès en Europe”, in: Josianne Duchêne, Guillaume Wunsch (ed.), Collecte et comparabilité des données démographiques et sociales en Europe, Louvain-la-Neuve, Academia/L’Harmattan, pp. 401-431.
  • Meslé France, Hertrich Véronique, 1997, “Mortality trends in Europe: the widening gap between East and West”, in 23rd International Population Conference, Beijing 1997, Liège, IUSSP, pp. 479-508.
  • Meslé France, Shkolnikov Vladimir, Hertrich Véronique, Vallin Jacques, 1996, Tendances récentes de la mortalité par cause en Russie, 1965-1994, Paris, Ined, 140 p. + 2 disquettes. (Données statistiques n° 2).
  • Meslé France, Vallin Jacques, 1993, “Causes de décès: de la 8e à la 9e révision, deux cas différents, la France et l’Angleterre”, in Alain Blum, Jean-Louis Rallu (eds.), European Population. II. Demographic Dynamics, Paris, John Libbey/Ined, pp. 421-445.
  • Omran Abdel R., 1971, “The epidemiologic transition: a theory of the epidemiology of population change”, Milbank Memorial Fund Quarterly, vol. 49, n° 4, pp. 509-538.
  • OPCS, 1985, Mortality statistics, cause. England and Wales, 1984, London, Office of Population Censuses and Surveys (Series DH2, n° 16).
  • OPCS, 1995, Mortality statistics, cause. England and Wales, 1993, London, Office of Population Censuses and Surveys, 169 p. (Series DH2, n° 20).
  • Pollard John H., 1982, “The expectation of life and its relationship to mortality”, Journal of Institute of Actuaries, vol. 109, pp. 225-240.
  • Przewozniak Krysztof, Zatonski Witold, 1993, “Tobacco smoking in Poland in the years 1923-87”, Polish Population Review, n° 3, pp. 103-124.
  • Royal college of physicians, 1962, Smoking and health: summary and report of the Royal College of Physicians of London on smoking in relation to cancer of the lung and other diseases, New York, Pitman Publishing Co.
  • Sekula W., Niedzialek Z., Figurska K., Morawska M., Boruc T., 1996, Food consumption in Poland converted into energy and nutrients, 1950-1995, Warsaw, Instytut Zywnosci i Zywienia.
  • Shkolnikov Vladimir, Meslé France, Vallin Jacques, 1995, “La crise sanitaire en Russie”, Population, 50 (4-5), pp. 907-982.
  • Vallin Jacques, Meslé France, 1988, Les causes de décès en France de 1925 à 1978, Paris, Ined/Puf (Travaux et Documents, Cahier 115), 608 p.
  • Vallin Jacques, Meslé France, 1998, “Comment suivre l’évolution de la mortalité par cause malgré les discontinuités de la statistique. Le cas de la France de 1925 à 1993”, in Gérard Pavillon (ed.), Enjeux des classifications internationales en santé, Paris, Éditions Inserm (Coll. Questions en santé publique), pp. 113-156.
  • Vallin Jacques, Meslé France, 2001, “Trends in mortality in Europe since 1950: age-, sex- and cause-specific mortality”, in Trends in Mortality and Differential Mortality, Strasbourg, Council of Europe Publishing (Population Studies n° 36), pp. 31-186.
  • WHO, 1992, World Health Statistical Yearbook, Geneva, WHO.
  • Zatonski Witold A., 2000, Development of the health situation in Poland (against the background of other countries in Central and Eastern Europe). Is democracy healthier? The « health miracle on the Vistula », Warsaw, M. Sklodowska-Curie Cancer Center and Institute of Oncology, Health Promotion Foundation, 91 p.
  • Databases

Notes

[*] Institut national d’études démographiques (INED), Paris.Translated by Zoe Andreyev.

[1] Denmark, Finland, Iceland, Norway, Sweden.

[2] Austria, Belgium, France, Germany, Ireland, Luxembourg, Netherlands, Switzerland, United Kingdom.

[3] Albania, Bosnia, Croatia, Greece, Italy, Macedonia, Malta, Portugal, Slovenia, Spain, Yugoslavia.

[4] Belarus, Estonia, Latvia, Lithuania, Moldova, Russia, Ukraine.

[5] Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovakia.

[6] The 28 countries selected are the following: Austria, Belgium, Bulgaria, Czech Republic, Denmark, Estonia, Finland, France, East Germany (GDR), West Germany (FRG), Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Netherlands, Norway, Poland, Portugal, Romania, Russia, Spain, Sweden, Switzerland, Ukraine, United Kingdom and Yugoslavia.

[7] More precisely at: 0 year, 1-4 years, 5-9 years, 10-14, etc. up to 85-89 years.

[8] For more information on this type of analysis, see Lebart et al. (1995)

[9] Special thanks to Bénédicte Garnier and Arnaud Bringé for their technical advice and assistance.

[10] The analysis was performed with the SPAD software.

[11] The impact of a mortality decline at a given age on the progression of life expectancy depends not only on the mortality rate itself and the extent of the decline but also on the age considered. The same decline, applied to a same rate, will of course have a greater impact if it concerns infants under 1 year of age than older persons.

[12] Mortality data, http:// www. who. int/ whosis.

[13] Eighth, Ninth and Tenth Revisions.

[14] For France: Vallin and Meslé, 1988,1998; for the United Kingdom: Meslé and Vallin, 1993; for Poland: Hertrich and Meslé, 1998, Meslé and Hertrich, 1997; for Russia: Shkolnikov et al., 1995, Meslé et al., 1996.

[15] Established on the basis of the rates by five-year age groups and the WHO European population standard (1992).

[16] Chapters I (items 001 to 139) and VIII (items 460 to 519) of the 9th Revision of the ICD.

[17] Chapter II (items 140 to 239) of ICD-9.

[18] Chapter VII (items 390 to 459) of ICD-9.

[19] Chapters III to VI (items 240 to 389) and chapters IX to XV (items 520 to 779) of ICD-9.

[20] Chapter XVII (items 800 to 999) of ICD-9.

[21] Chapter XVI (items 780 to 799) of ICD-9.

[22] Over this period, the proportion of deaths due to an ill-defined cause is low enough to allow a comparison in time and space to be made thanks to a simple proportional distribution.

[23] We are aware that the adjustment should be done the other way around, using the data from the years 1985 to 1992, but this is a rather difficult task, which is beyond the scope of this article; in addition, such a reconstruction would be an exception to the general principle governing the reconstructions used here, which are based on homogenous series, according to the most recent definitions.

[24] Since the early 1990s, we can observe a slight decrease in female tobacco consumption (Baudier and Velter 1998), but this decline has not yet had any incidence on mortality trends.

[25] Death due to alcoholism is for a large part included among violent deaths in Russia. Indeed, very often the cause of death is acute alcoholism, which in this case is considered a form of poisoning. The case may be similar in Poland for acute alcoholism.

Abstract

English

After a period of general convergence, the 1960s were marked by the divergence between the life expectancies of eastern European countries, where all progress came to a halt, and those of the rest of Europe where health care made large strides. A hierarchical analysis of age-specific mortality patterns shows that this divergence goes together with the development of very different patterns of age at death; in the countries of eastern Europe, and especially in the USSR, excess mortality at adult ages is spectacularly high.
Cause-specific analysis reveals the decisive role played by two kinds of diseases. On the one hand, “man-made diseases” (alcoholism, smoking, car accidents, etc.) have continued to increase in the east, whereas they were curbed in the west starting in the 1960s. On the other hand, eastern Europe was unable to join the cardiovascular revolution that had enabled the west to increase its life expectancy levels. The considerable divergence between eastern and western Europe should not hide the differences that still remain among western countries. Indeed, mortality patterns are changing in the west, and the traditional opposition between north and south is undergoing radical transformations.

Outline

  1. Age-specific mortality patterns: increasing differences
    1. Four groups of countries selected through hierarchical analysis
    2. The differentiation of age-specific mortality patterns
    3. Contribution of the various age groups to the variations in life expectancy
  2. Eastern Europe: the cardiovascular revolution has yet to begin
    1. Main trends in causes of death
      1. The decline of infectious and respiratory diseases
      2. The stability of neoplasms
      3. Divergence in mortality from cardiovascular diseases
      4. Other diseases
      5. Violent deaths: a contrasting picture
    2. Contribution of the evolution of disease to the increase or decline of life expectancy
      1. From 1965 to 1984
      2. Since 1984
  3. Conclusion

To cite this article

France Meslé and Jacques Vallin "Mortalité en Europe : la divergence Est-Ouest", Population 1/2002 (Vol. 57).
URL  www.cairn.info/revue-population-2002-1-page-157.htm
DOI 10.3917/popu.201.0171.

Back to Top

Feedback

Why is this article not available in English?

Cairn International Edition is a service dedicated to helping a non–French–speaking readership to browse, read, and discover work published in French journals. You will find English full–text translations, in addition to French version already available on Cairn regular edition. Full text translations only exist for a selection of articles.

If you are interested in having this article translated into English, please enter your email address and you will receive an email alert when this article has been translated.

Message sent

Your email address has been saved.
We will notify you when this article becomes available in English.