Introduction
1Recent mortality trends in industrialized countries raise several questions. The present situation differs in many respects from what demographers imagined 20 years ago. Who would have thought in the 1960s that infant mortality would drop to such a low level [1], or that differences between the expectation of life of the two sexes would continue to rise, and reach 8.2 years in France in 1980 compared with 5.8 years in 1950 (7.2 years and 4.8 years respectively in the Czech Republic) [2] ? Who would have dared to suggest that social inequality in the face of death, instead of gradually decreasing, would continue to grow ? Yet this is what the results of the most recent study conducted by INSEE on this subject indicate [3].
2But the most disturbing difference between past population forecasts and actual trends concerns the divergence observed during recent years between most Eastern European countries where, following the experience of the Soviet Union, mortality has been stationary or even risen, and Western Europe, where expectation of life has steadily increased.
3In an article published recently in Demografie, the author discussed the unfavourable mortality trend observed in the Czech Republic over the last 25 years and presented a component analysis by sex and age [4]. Whereas between World War II and 1960, the upward trend of the expectation of life had accelerated in the Czech Republic, as in all industrial countries, resulting in one of the lowest mortality levels in the world, the trend has been reversed during the last 25 years : the mean length of life has remained stationary for women and even decreased for men. This change has not affected all ages equally. Infant and child mortality have continued to decline. The first age groups to be affected by this reversal, during the early 1960s, were the elderly, followed rapidly by younger adults.
4In an attempt to put the surprising mortality trend in the Czech Republic in perspective and to investigate possible explanatory factors, we considered it useful on the one hand, to compare the trend with that in a Western European country, France, and on the other, to study the respective influence of the major causes of death. Therefore the first part of this paper will be devoted to a comparative study of total mortality, and in the second this comparison will be continued at the cause-of-death level.
I – Comparative study of overall mortality
5It may seem somewhat surprising to compare France, a country of 55 million inhabitants, and in which mortality differs from one region to another, to only part of Czechoslovakia (a country of 15 million inhabitants) : the Czech Republic (10 million). However, in Czechoslovakia, the strong regional variations in mortality are paralleled by historical, political and economic differences which give individual identities to the two components of the Federal Republic of Czechoslovakia : the Czech Republic and the Slovak Republic, for which statistical data have also traditionally been given separately. In comparing mortality with that of a Western European country, it seemed more interesting to focus on the Czech Republic, which became industrialized earlier than Slovakia [5].
A – Crude death rate
6In Figure 1, in which we present the crude death rates for France and the two Czechoslovak Republics since the turn of the century, the difference between the Czech and Slovak Republics is illustrated. Until World War II, the death rate was much higher in Slovakia, while since the 1950s the situation has been reversed. This diverging trend illustrates the time-lag of the demographic transition in the two republics. The crude death rate reflects not only different trends in mortality, but also differences in age structure, which are related to two different fertility patterns. For instance, the lower death rate in Slovakia since the 1950s is entirely due to a much younger age structure.
Annual variations in the crude death rate since 1900 : France and the two Republics of Czechoslovakia
Annual variations in the crude death rate since 1900 : France and the two Republics of Czechoslovakia
7From this point of view, the Czech Republic is much closer to France. There is, nevertheless, a considerable difference between the rates in the two countries. In the Czech Republic, where the transition began later than in France, mortality fell more rapidly during the first half of the century, but the age structure was younger, and this was reflected, after the 1920s, in a substantially lower crude death rate. After the troubled years of the second world war, this advantage persisted until the mid-1960s. However, since that time, although the age structure of the Czech population remained younger than that of France, the crude death rate was much higher. The spectacular turn-around of the mortality trend contrasts strongly with the pattern observed in France, which was extremely regular except during the war years. In recent years, despite an increasingly ageing population, France’s crude death rate has continued to decline.
B – Expectation of life
8Mortality trends are more accurately measured by expectation of life at birth (Figure 2 and Table 1). Until 1960, the mean length of life increased more rapidly in the Czech Republic than in France. In 1900, expectation of life was 43.4 years for males and 47.0 for females in France, compared with only 38.9 and 41.7 years in the Czech Republic. This difference gradually disappeared, but more quickly for males than for females. Men’s expectation of life in the Czech Republic caught up with that in France as early as the 1930s, while for females, it did so only in the mid-1950s ; by that time, the mean length of life for males was appreciably higher in the Czech Republic than in France. This situation (equality for females, Czech advantage for males) persisted until the early 1960s. Since then, however, the trend has been completely reversed, and in 1985, the mean length of life in France (71.3 years for males and 79.4 years for females) was once more considerably higher than in the Czech Republic (67.6 for males and 74.9 for females). The difference was almost as high as at the beginning of the century.
9The turning point clearly occurred in 1960. However, two somewhat dissimilar stages can be distinguished since then. During the 1960s, a certain slowing-down of the rise in life expectancy occurred in France, particularly for males, but during the same period in the Czech Republic, progress for females came to a complete halt and life expectancy of males declined sharply (by almost two years within a decade). During the 1970s and 1980s, progress was resumed in both countries. However, in France, the rapid rate of the 1950s was recovered, whereas the gain in the Czech Republic was much more modest, and in 1985 men’s life expectancy had not yet returned to the maximum achieved in 1960.
Annual variations in expectation of life at birth, by sex, since 1900 : France and the Czech Republic
Annual variations in expectation of life at birth, by sex, since 1900 : France and the Czech Republic
Expectation of life at birth since 1950 : France and the Czech Republic
Expectation of life at birth since 1950 : France and the Czech Republic
N.B. : France : data include "false stillbirths" for the whole period. Czech Republic : 1953-1964 data have been corrected for infant mortality (see note, Table 2).C – Age-specific mortality
10In the article in Demografie, it was shown that this unfavourable trend in the Czech expectation of life did not affect all age groups equally, but that it was particularly pronounced over age 40. To investigate differences between the two countries from this point of view, the data were broken down into three age groups, in each of which mortality depends on different aetiological processes : less than one year, 15-24 years and 40 years and over (Figure 3 and Table 2). The indices examined are the infant mortality rate (q0), the probability of dying within ten years of the 15th birthday (10q15) and expectation of life at the 40th birthday (e40).
1 – Infant mortality
11Since 1950, the infant mortality rate has steadily decreased in both countries and is at present so low that further changes can no longer have much influence on mean length of life. It is, however, of interest that its trend in the Czech Republic was much less regular than in France (Figure 3a). Whereas in France, the rate of decrease of the infant mortality rate has been almost constant since the war (from 51.9 p. 1,000 in 1950, it dropped to 8.3 p. 1,000 in 1985, a mean annual decrease of 5.2 %), in the Czech Republic two very different phases could be observed.
12Before examining this point, however, it is necessary to correct the data provided in the official publications of the Czech Republic to take into account changes introduced in 1953 and 1965 in the definition of live births.
13Between 1953 and 1964, infants born after less than 28 weeks gestation, who weighed less than 1,000 g or were less than 35 cm in length, were only considered as live-born if they lived for at least 24 hours. This restricted definition of live births was abandoned in 1965. A dual classification of births in 1964 according to the old and new definitions showed that this restriction had led to an under-estimation of infant mortality by 21 % [6]. We therefore increased the rates published for 1953-1964 accordingly.
14We should bear in mind that in France, the rates used here include "false stillbirths" [7] ; for the years before 1975, they are therefore corrected rates.
Comparative study of mortality trends in France and the Czech Republic for three age groups : less than one year, 15-24 years, and 40 years and over
Comparative study of mortality trends in France and the Czech Republic for three age groups : less than one year, 15-24 years, and 40 years and over
N.B. : France : infant mortality rates include "false stillbirths" for the whole period. Czech Republic : data between 1953 and 1964 have been corrected for a change in definition.Comparative study of mortality trends in France and the Czech Republic for three age groups : before age 1 (q0), between 15 and 25 (10q15) and after age 40 (e40)
Comparative study of mortality trends in France and the Czech Republic for three age groups : before age 1 (q0), between 15 and 25 (10q15) and after age 40 (e40)
15Immediately after World War II, the infant mortality rate in the Czech Republic was higher than in most other industrialized countries. A particularly active maternity and infant welfare policy resulted in a very rapid drop until the early 1960s. It is shown in Figure 3 that in 1950, the rate was still higher in the Czech Republic than in France (64 compared with 52 p. 1,000), but that this difference disappeared in 1953 and the trend was then reversed, so that in 1960 the Czech rate, even after correction, only came to 25 p. 1,000 compared with 27 in France.
16Since 1961, on the contrary, the reduction in infant mortality has been much slower than in France. The rates in the two countries were equal again in 1964, and since then there has been an increasing gap in favour of France. In 1985, the infant mortality rate still came to almost 13 p. 1,000 in the Czech Republic, whereas in France it was nearing 8 p. 1,000.
17The situation regarding mortality at ages 15-24 is very different. It is true that the
2 – Mortality at ages 15-24
18The situation regarding mortality at ages 15-24 is very different. It is true that the contribution of this age group to total mortality is almost negligible, even lower than that of infant mortality, but the highly specific pattern in this group merits attention. This time, it is the situation in France which was less favourable (Figure 3b). After a decline which remained fairly strong until 1960, the probability of dying within ten years of the 15th birthday (10q15) rose quite substantially for males and slightly less for females between 1960 and 1973. In 1974, a series of road safety measures brought this upward trend to a stop. Men’s mortality then remained stationary until 1980, while that of women declined. During the early 1980s, a new series of preventive measures, relating in particular to motor cycles, also led to a fall in men’s mortality.
19The Czech mortality rates in this age group, in contrast, decreased steadily, except in 1968/69. The probability of dying between the 15th and 25th birthdays, although initially higher than in France, reached a much lower level during the 1980s. The two curves crossed as early as the mid-1950s for females, but not until the end of the 1960s for males.
3 – Mortality after age 40
20Of the three indices shown in Figure 3, expectation of life at age 40 exerts by far the greatest influence on expectation of life at birth. During recent years, the trend has been substantially less favourable in the Czech Republic. Whereas between 1950 and 1960, mean length of life (Figure 2) was as high as in France for women and appreciably higher for men, life expectancy on the 40th birthday remained consistently lower than in France for women and was barely higher than in France for men (Figure 3c). Thus the more rapid decline of life expectancy at birth in the Czech Republic compared with that in France during the same period (Figure 2) was essentially due to the extremely rapid fall in infant mortality (Figure 3a).
21Between 1960 and 1970, on the other hand, life expectancies at birth and at the 40th birthday moved almost perfectly in parallel in the Czech Republic, whereas in France, particularly for men, expectation of life at the 40th birthday rose more slowly than life expectancy at birth. In other words, during this decade, men’s expectation of life in France only continued to improve because of a fall in mortality before the 40th birthday, and particularly in infant mortality (despite an increase in mortality at ages 15-24). In the Czech Republic, on the other hand, mortality before the age of 40 did not affect the reduction in life expectancy at birth, which was caused by the decrease in expectation of life at the 40th birthday.
22Since 1970, the resumption of the rapid increase in the French life expectancy at birth, for both sexes, is almost entirely attributable to a rise in expectation of life at the 40th birthday. In the Czech Republic, on the other hand, the slight improvement in life expectancy at birth owes nothing to mortality at ages 40 and over, as this has remained practically stationary for the last fifteen years, for both sexes : in fact, since 1969, men’s expectation of life at age 40 has remained slightly below its 1950 level !
23Finally, during the mid-1980s, expectation of life at the 40th birthday in the Czech Republic was 4.5 years lower than in France for females (36.5 against 41 years) and 3.5 years lower for males (30.5 against 34 years).
II – The importance of the different causes of death
24Although mortality over age 40 has been responsible for most of the deterioration recently observed in Czech mortality, it seemed preferable, for comparing cause-specific mortality in the two countries, to use synthetic standardized death rates for all age groups. On the one hand, in comparing mortality trends in these two countries, we have shown that mortality before the age of 40 cannot be considered completely neutral (this will become very clear for certain causes of death). But on the other hand, choosing standardized mortality rates (constructed here with a standard European population, presented in Annex 1) means using an index in which mortality in the higher age groups is preponderant.
A – Grouping causes of death
25The study of cause-specific mortality trends involves many practical problems, one of the most important of which is that caused by discontinuity in the statistical series introduced by each new Revision of the International Classification of Diseases (ICD). During the period considered in the present study (1950-1985), four successive revisions were made (the Sixth, Seventh, Eighth and Ninth Revisions). A systematic reconstruction of continuous time series based on the Detailed List of the Eighth ICD Revision has been carried out by INED [8] ; thus for France, re-grouping items into any useful sets was no problem. However, there has been no equivalent reconstruction for the Czech Republic. A priority task was therefore to obtain relatively consistent statistical series for the period covered by the last four Revisions of the ICD. This obviously reduced the possibilities of re-grouping items according to other criteria ; but it was also necessary to attempt to construct sets of causes which would be as homogeneous as possible from the point of view of pathology.
26In fact, we used a reclassification adopted in a similar type of study conducted jointly by INED and the Hungarian Central Statistical Office. In this reclassification, 35 basic categories were defined and were re-grouped, at a more general level, into the 9 broad groups of causes of death we have used for the present study. For two of these, we have also examined certain particularly interesting sub-groups. The composition of these groups of causes, for each ICD Revision, is presented in Annex 2.
27One of the most important criticisms to be levelled at the ICD, from an aetiological point of view, is that the section which deals with infectious diseases only includes a small part of all diseases caused by pathogenic organisms [9]. In an attempt to remedy this, we added to ICD Chapter I all acute diseases of the respiratory system, to form the first of the broad groups of causes used in the present study. Other infectious diseases remain in other chapters of the ICD, but the number of modifications introduced by each successive Revision makes it impossible, in the framework of this comparative study, to broaden the category of infectious diseases any further.
28Even at this level, we encountered difficulties with the statistical data for the Czech Republic, to which an additional correction had to be applied. Indeed, independently of the ICD Revisions, an important change in coding practice was introduced in 1960. Many of the deaths which, according to the ICD, should have been classified as item 502, chronic bronchitis, were until 1959 classified as item 491, pneumonia… Consistency required us to reclassify half the deaths originally under item 491 into item 502 for 1950-1959.
29The second broad group of causes consists of the remainder of the ICD chapter on diseases of the respiratory system, essentially, chronic diseases of the respiratory system.
30The third broad group, neoplasms, did not pose any definitional problems. Special attention will be paid to malignant neoplasms of the bronchus and lung, closely related to smoking and industrial pollution, and to malignant neoplasms of the breast, which play an important part in mortality from tumours among men and women respectively.
31The fourth and fifth groups divide the ICD chapter on diseases of the circulatory system into two sub-groups, the former consisting of cerebrovascular diseases and the latter of cardiovascular diseases. This division, which proved practicable with a few approximations, despite the various Revisions, makes it possible to distinguish between two quite different pathological categories. It would have been equally important to separate ischaemic heart disease from other cardiovascular diseases, but this was not possible with the Czech data.
32The sixth broad group covers diseases related to alcoholism, while the seventh consists of deaths from injuries and poisoning, with special attention to deaths from road accidents and suicide, which reflect social problems.
33The eighth group consists of other causes of death (diseases not included in the first six groups).
34Finally, the ninth group, ill-defined or unspecified causes of death, was only used in a preliminary stage of our analysis. To some extent, it reflects the statistical quality of the data used. From this point of view, the situation is better in the Czech Republic, where ill-defined deaths currently account for 0.5 % of all deaths, compared with 5 % in France. To retain these deaths in a group of their own might bias the interpretation of cause-specific mortality trends, as their proportion among all deaths has changed considerably over time (between 1950 and 1985, it dropped from 19 % to 5 % in France, and from 8 % to 0.5 % in the Czech Republic) [10]. The results presented below are those obtained after redistributing deaths from ill-defined causes, by sex and age group, proportionately among the other eight broad groups of causes (Table 3) [11].
35To analyze the trends observed in these broad groups of causes, we constructed eight semi-logarithmic graphs, drawn on the same scale for easier comparison.
B – Mortality trends for eight broad groups of causes of death
1 – Infectious diseases and acute diseases of the respiratory system
36Mortality from infectious diseases has decreased in both countries throughout the period. Consequently, though this group still accounted for a large proportion of all deaths in 1950, it is now of relatively minor importance (Figure 4). It is, however, noteworthy that whereas, apart from annual fluctuations, the trends of mortality from infectious diseases were very similar in both countries between 1950 and 1972, the Czech pattern has suddenly diverged for the last ten years. Since the beginning of the 1970s, the decline for both males and females has slowed down abruptly, to the extent that female mortality from these diseases is now higher than the corresponding male mortality in France. This is difficult to explain, as we have few elements for interpreting it. It is possible that once a certain level of mortality has been attained, the struggle against infectious diseases, which was relatively easy until then, may become more expensive and more difficult to organize. Another possibility is that a certain negligence in medical practice and hygiene has appeared in the Czech Republic, particularly serious at a time when new techniques that require meticulous asepsis (catheterization, immuno-depressive treatment, etc.) were developing. This could account for the slight rise in mortality from infectious diseases in the Czech Republic during the 1970s.
2 – Chronic diseases of the respiratory system
37Mortality from chronic diseases of the respiratory system has followed an entirely different course (Figure 5). In France, throughout the period, it increased very slightly for males and decreased appreciably for females. At a higher level, the trend is roughly the same for the Czech Republic until 1978. It is true that a sharp increase in male mortality from chronic respiratory diseases occurred in 1960, but this increase is probably more apparent than real, because the correction we made for chronic bronchitis (see II A) was seemingly inadequate for male mortality, whereas it was sufficient to establish a consistent series for women. This strong resistance of chronic diseases of the respiratory system to mortality progress, which is seen in both countries, is no doubt at least partly related to social factors. These diseases are sensitive to smoking behaviour and atmospheric pollution, and an aggravation of these factors greatly reduces the possibility of progress.
38Since 1978, however, a fairly sharp drop is observed in the Czech Republic for both sexes. This divergence from the French trend, six years after the movement in the opposite direction observed for infectious diseases (see IIB, 1), cannot be interpreted as explaining the latter. It is, however, possible that since the Ninth ICD Revision was introduced in the Czech Republic (1979), classification procedures have gradually been modified. A detailed study of the contents of the different ICD items would be needed to verify this.
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : infectious diseases and acute diseases of the respiratory system
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : infectious diseases and acute diseases of the respiratory system
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : chronic diseases of the respiratory system
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : chronic diseases of the respiratory system
3 – Neoplasms
39Apart from a slight anomaly in 1955 for Czech mortality from neoplasms (corresponding to a contrary deviation for "other diseases", apparent in Figure 11), mortality trends for this group are so regular that they appear almost linear, particularly for France (Figure 6). There is very little difference in male mortality between the two countries : the rise was slightly more rapid in the Czech Republic until the early 1970s, and then becomes almost imperceptible ; the starting and finishing levels are the same. Women’s mortality has differed more substantially between the two countries since the mid-1960s : in France, mortality from all neoplasms decreased regularly throughout the period, whereas in the Czech Republic it stagnated after 1960.
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : neoplasms
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : neoplasms
- all neoplasms
- neoplasm of lung (males) and breast (females)
40These differences are largely related to deaths from neoplasms of the bronchus and lung (males) and neoplasms of the breast (females).
41The level of men’s mortality from neoplasm of the lung (Figure 6b) was substantially higher in the Czech Republic than in France, and is counterbalanced by a lower level for all other neoplasms. A steep rise in mortality from this cause until around 1965 gave way to near-constancy after 1970. In France, on the contrary, mortality from neoplasm of the lung increased more or less steadily throughout the period. These diverging trends account for the slight difference between the two countries in male mortality from all neoplasms towards the middle of the period. The excess mortality from neoplasms of the bronchus and lung among Czech males is striking. The negative effects of smoking and atmospheric pollution are certainly greater than in France. However, the situation now seems to be stabilizing, whereas in France it is steadily deteriorating.
42For women’s mortality from neoplasm of the breast (Figure 6b), the levels of the two countries show less difference. However, once again, there is a fairly clear divergence in the trends, particularly after the mid-1960s. The rate which was lower in the Czech Republic at the beginning of the period, increased more rapidly than in France, where it stabilized after 1965. In France, medical progress – early diagnosis and improved treatment – put a stop to the rise in mortality from this cause, whereas it has continued to increase in the Czech Republic. Yet this disparity only accounts for part of the difference between the two countries in female mortality from all neoplasms since the 1960s.
4 – Cerebrovascular diseases
43One of the strongest contrasts between mortality patterns in France and the Czech Republic is observed for mortality from cerebrovascular diseases (Figure 7). For both sexes, the French and Czech curves cross during the 1960s. Cerebrovascular mortality in the Czech Republic, which was in 1950 much lower than in France, has reached a much higher level during the early 1980s.
44In France, the trend of mortality from this cause has declined continuously throughout the period, accelerating in the 1970s and 1980s. This can no doubt be related to gradual progress in therapy, intensified during the 1970s by improved supervision of hypertension and by the development of emergency and resuscitation units which enabled patients to be saved in cases of cerebral attacks to which they would previously have succumbed.
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : cerebrovascular diseases
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : cerebrovascular diseases
45In the Czech Republic, the trend followed the reverse direction, at least until the early 1970s, when the situation stabilized. The increase in cerebrovascular mortality during the 1960s is perhaps somewhat exaggerated. Indeed, a sudden rise is observed in 1968, when the Eighth Revision came into application, suggesting that a change in classification procedures may also be reflected in the statistical series. However, no symmetrical discontinuity appears in any of the other broad groups of causes used here, nor for the group of ill-defined causes. The changes could therefore only have affected several ICD items in different groups of causes. But even if we suppose that the sudden rise in 1968 is in part only apparent, the upward trend before and after this year remain. If a correction were made, it would reduce the increase observed between 1950 and 1973, but not eliminate it. Paradoxically, it may be argued that the rise in cerebrovascular mortality is related to improvements in the standard of living which occurred during the 1950s and 1960s. In a country where food was often difficult and expensive to obtain, easier buying could lead to new patterns of consumption and a less healthy diet.
46However, changing dietary habits alone cannot account for the rise in cerebrovascular mortality, at a time when, as we shall see below, cardiovascular mortality decreased. Further information is required, such as quantitative data on the consumption of different commodities, in particular salt, and on patterns of blood pressure in the two countries.
5 – Cardiovascular diseases
47Although less strongly contrasted than for the previous group, the cardiovascular mortality patterns were also very different in France and the Czech Republic (Figure 8). The levels differ so much that, despite a strong excess male mortality from this group of diseases, female mortality in the Czech Republic throughout the period was at least equal to, and sometimes higher than, the male mortality in France. The trends are also very different : in France, mortality dropped regularly throughout the period (more for females than males), whereas in the Czech Republic the rate remained more or less stationary for men, and stopped declining for women around the beginning of the 1970s. In the Czech Republic, the trend for both sexes has even started to rise since the end of the 1970s.
48Once again, dietary habits are no doubt partly responsible for blocking progress in the Czech Republic. Other factors involved may be greater stress and a higher smoking rate, which favour cardiovascular diseases. From this point of view, it would be interesting to separate ischaemic heart disease from other cardiovascular affections, but the statistical discontinuities introduced by the successive Revisions of the ICD rule this out.
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : cardiovascular diseases
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : cardiovascular diseases
6 – Alcoholism
49Mortality attributed to alcoholism has traditionally been much higher in France than in most other industrialized countries, including the Czech Republic. However, the pattern has changed in France : after a steep rise during the early 1950s, a drop was first seen in 1957 and 1958, followed by a slight increase during the early 1960s and a steady decline since the end of that decade. This pattern applies to both sexes, although at a much higher level for men (Figure 9).
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : alcoholism
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : alcoholism
50In the Czech Republic, the initial level was much lower and the trend more regular. Men’s mortality increased strongly over the period and that of women very slightly.
51A close correlation naturally exists between these mortality trends and alcohol consumption. In 1965 and 1982, the following consumption figures were recorded (in litres of pure alcohol per head) :
53Although these figures cover different types of consumption (mainly wine in France and mainly beer in the Czech Republic), the contrasting trends since 1965 are significant.
7 – Injuries and poisoning
54The pattern of deaths from injuries and poisoning, when considered globally, is the most similar for the two countries. The rate was very slightly higher in France than in the Czech Republic, with a considerable excess male mortality in both countries. The trend, for each sex and country, was a slight rise until 1970, followed by a slight decline to the end of the period (Figure 10a). When deaths in this group are broken down, however, much larger differences appear.
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : injuries and poisoning
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : injuries and poisoning
- all deaths attributed to injuries and poisoning
- road accidents
- suicide
55The increase between 1950 and 1970 in deaths from road accidents (Figure 10b) was much more rapid than for all deaths in this group, and the similarity between the countries was maintained. However, after 1970, the decline was much more abrupt in the Czech Republic (accelerating with the speed limits introduced in 1979) than in France (despite the measures taken in 1974). In France, the fall in women’s mortality even slowed down to mere constancy. Consequently, the trends are presently diverging in favour of the Czechs. Although very similar road safety measures have been adopted in both countries, they are no doubt more efficiently enforced in the Czech Republic, where traffic is less dense.
56The pattern for deaths from suicide is very different (Figure 10c). For both sexes and in both countries, mortality from this cause was practically constant until 1970, but the level in the Czech Republic was much higher. Since then, it has tended to rise in France, particularly since the late 1970s, and to decrease fairly sharply in the Czech Republic, so that in both countries exactly the same level was reached in 1982. The greater propensity for suicide traditionally attributed to Central European countries is therefore not supported by recent trends in the Czech Republic.
8 – Other causes of death
57The eighth group, which covers all other causes of death, contributes to broadening the gap between the two countries, since in the Czech Republic mortality from these causes has remained stationary since the mid-1950s, whereas in France it has decreased steadily since 1950. Lower in the former country at the beginning of the period, it has now reached a higher level than in France.
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : all other diseases
Standardized sex-specific mortality rates in France and the Czech Republic since 1950 : all other diseases
C – Overview of the changing structures of cause-of-death mortality
58The differences in cause-specific mortality trends revealed in the previous section are not equally important in causing the differences observed in total mortality. In Table 3, this point is illustrated by presenting the proportion of total mortality represented by deaths from each group of causes in 1950, 1985 and 1960, a turning-point for mortality in the Czech Republic.
59As we have shown, some groups of causes account for higher mortality in France than in the Czech Republic : alcoholism, injuries and poisoning (in particular road accidents). However, their relative weight in total mortality is low. In 1985, alcoholism in France only accounted for 5 % and 3 % respectively of the standardized total mortality rates for men and women, and injuries and poisoning for 11 % and 10 % respectively. The groups of causes for which mortality was higher in France therefore only accounted for between 13 and 16 % of all deaths, according to sex. It must also be noted that, although French excess mortality from alcoholism remains very high (but rapidly decreasing), it is fairly slight for deaths from injuries and poisoning. Thus the only group of causes for which mortality is much higher in France is alcoholism, which accounts for a mere 3-5 % of total mortality.
Weight of each of the eight broad groups of causes in the standardized total mortality rates for 1950, 1960 and 1985 : Czech Republic and France
Weight of each of the eight broad groups of causes in the standardized total mortality rates for 1950, 1960 and 1985 : Czech Republic and France
60On the other hand, the causes for which mortality was much higher in the Czech Republic account for a large proportion of the standardized total mortality rate. In 1985, cardiovascular diseases were responsible for over 35 % of all deaths in the Czech Republic for each sex. The addition of cerebrovascular diseases, for which the excess mortality of Czechs was just as strong, brings the total to 52 % for males and 56 % for females. Although less pronounced, the higher mortality from neoplasms in the Czech Republic adds to this, accounting for 21-23 % of total mortality according to sex. Finally, we must include the excess mortality from "other causes", which account for 10 % of total mortality.
61This situation is essentially a recent one, having begun gradually after 1960. Since this date, mortality from causes which have the greatest weight in total mortality, has either increased or remained stationary in the Czech Republic, while declining rapidly in France (except for male mortality from neoplasms, which has also risen in France). Mortality from cerebrovascular diseases has almost doubled in the Czech Republic, while being almost halved in France (Figure 7).
62Consequently, the cause-specific mortality patterns in the two countries have been considerably modified over time, and their differences are not the same now as they were 30 years ago. In both the Czech Republic and France, mortality from infectious diseases has fallen and only represents a fraction of its earlier weight in total mortality. For males, it dropped from 15 % in 1950 to 2.8 % in 1985 in France and from 14.9 % to 2.6 % in the Czech Republic. Conversely, deaths from neoplasms have become much more important in both countries, particularly for men (from 15.4 % to 22.8 % in the Czech Republic and from 15.8 % to 31.3 % in France).
63The weight of mortality from cerebrovascular diseases, on the other hand, has not changed in the same way : while it has substantially decreased in France (from 15.1 % to 10.4 % for men, from 17.6 % to 14.2 % for women), it has increased strongly in the Czech Republic (from 8.0 % to 15.4 % for men, from 10.0 % to 20.9 % for women), most of this gain occurring since 1960.
64Only mortality from cardiovascular diseases accounts for roughly the same proportion of total deaths throughout the period, in both countries and for both sexes. This proportion is much higher in the Czech Republic (around 36 %) than in France (around 24 %), and it should be borne in mind that these proportions apply to a total mortality which, since the 1960s, has been increasing in the Czech Republic and declining in France.
Conclusion
65The recent deterioration of mortality in the Czech Republic – essentially, higher risks of death for adults and the elderly – is therefore above all related to a rise in cerebrovascular mortality and, secondarily, to alcoholism. But these are not the only reasons for the gap which has been widening between this country and France over the past two decades. This is also, and perhaps mainly, due to the fact that cardiovascular mortality has declined substantially in France, while it has remained stationary in the Czech Republic (for women, a similar pattern is observed with neoplasms).
66A comparative study of this kind sheds light on certain aspects of current mortality trends in Eastern Europe. Cause-of-death analysis may also suggest a number of possible explanations, which however, at this stage, can be no more than hypotheses :
- changing dietary habits (influence of consumption of fat on cardiovascular diseases and salt on cerebrovascular diseases) ?
- health service organization (prevention of diseases of the circulatory system, organization of emergency care, etc.) ?
- social behaviour (alcoholism) ?
- development of stress factors related to social and economic conditions ?…
67To answer these questions, it would be necessary to correlate cause-specific mortality trends with certain socio-economic data, which are presently hard to obtain in an adequate form.
Age structure used for computing standardized rates(1)
Age structure used for computing standardized rates(1)
(1) Total of the populations of the following 26 European countries : Austria, Belgium, Bulgaria, Denmark, Spain, Finland, France, Greece, Hungary, Iceland, Ireland, Italy, Luxemburg, Malta, Norway, Netherlands, Poland, Portugal, GDR, FRG, Romania, United Kingdom, Sweden, Switzerland, Czechoslovakia, Yugoslavia.Items composing the nine groups of causes used : 6th, 7th, 8th and 9th Revisions(1) of the International Classification of Diseases (ICD)(2),(3)
Items composing the nine groups of causes used : 6th, 7th, 8th and 9th Revisions(1) of the International Classification of Diseases (ICD)(2),(3)
(1) The periods covered by these Revisions are the same in France and the Czech Republic : 1950-1957 for the 6th, 1958-1967 for the 7th, 1968-1978 for the 8th, and since 1979 for the 9th.(2) At this level, there is no change between the 6th and 7th Revisions.
(3) The Czech data do not provide deaths classed in the four-digit items of Chapter 16 which we have included here (in brackets). However, the three-digit items from which they originate are, in contrast to France, almost empty in the case of the Czech Republic : for this country, we therefore simply added them to the ill-defined causes.
Standardized mortality rates (p. 100 000) by group of causes since 1950 : Czech Republic
I Infectious diseases and acute diseases of the respiratory systemII Other diseases of the respiratory system
III Neoplasms
IIIa Neoplasm of lung
IV Cerebrovascular diseases
V Cardiovascular diseases
VI Alcoholism and cirrhosis of liver
VII Injuries and poisoning
VIIa Road accidents
VIIb Suicide
VIII Other diseases
IX Total mortality
(cont.)
I Infectious diseases and acute diseases of the respiratory systemII Other diseases of the respiratory system
III Neoplasms
IIIa Neoplasm of breast
IV Cerebrovascular diseases
V Cardiovascular diseases
VI Alcoholism and cirrhosis of liver
VII Injuries and poisoning
VIIa Road accidents
VIIb Suicide
VIII Other diseases
IX Total mortality
Standardized mortality rates (p. 100 000) by group of causes since 1950 : France
I Infectious diseases and acute diseases of the respiratory systemII Other diseases of the respiratory system
III Neoplasms
IIIa Neoplasm of lung
IV Cerebrovascular diseases
V Cardiovascular diseases
VI Alcoholism and cirrhosis of liver
VII Injuries and poisoning
VIIa Road accidents
VIIb Suicide
VIII Other diseases
IX Total mortality
(cont.)
I Infectious diseases and acute diseases of the respiratory systemII Other diseases of the respiratory system
III Neoplasms
IIIa Neoplasm of breast
IV Cerebrovascular diseases
V Cardiovascular diseases
VI Alcoholism and cirrhosis of liver
VII Injuries and poisoning
VIIa Road accidents
VIIb Suicide
VIII Other diseases
IX Total mortality
Notes
-
[*]
Translated by Linda Sergent.
-
[**]
Science Faculty, Charles University, Prague.
-
[***]
INED.
-
[****]
Alain Blum and Roland Pressat, « Une nouvelle table de mortalité pour l’URSS (1984-1985), Population, 1987, n° 6, 843-863.
-
[1]
The dominant theory at that time, put forward by many authors during the 1940s and 1950s, was that there was a minimum level beyond which improvements would not be possible. (Talacko, for instance, set it at 30 p. 1,000 for males and 20 p. 1,000 for females. The present rate is less than 10 p. 1,000). For further details, see :
- Jean Bourgeois-Pichat. – « Essai sur la mortalité biologique de l’homme », Population, 1952.
- Josef Talacko. – « Dynamická pozorování ve statistice úmrtnosti », Knihovna Statistickéko Obzoru, Prague, 1941 (sv. 43).
-
[2]
Jacques Vallin. – "Sex patterns of mortality : a comparative study of model life tables and actual situation with special reference to the cases of Algeria and France". In : Alain Lopez and Lado Ruzicka (eds.), Sex Differentials in Mortality : Trends, Determinants and Consequences, Canberra, 1983, pp. 443-476.
-
[3]
Guy Desplanques. – La mortalité des adultes : résultats de deux études longitudinales (période 1955-1980), Paris, INSEE, 1985, 212 p.
-
[4]
Jitka Rychtaříková. – "Vývoj úmrtnosti v ČSR podle pohlaví e vĕku v období 1950-1984", Demografie, 1987, n° 3.
-
[5]
During the nineteenth century and up to World War I, the Czech lands (Bohemia, Moravia, Silesia) were the most industrialized region of the old Austro-Hungarian Empire.
-
[6]
Milan Kučera. – "Mrtvorozenost, Kojenecká a novorozenecká úmrtnost v roce 1965 podle nové definice narozených", Demografie, 1966, n° 2, pp. 183-184.
-
[7]
Babies born alive but who died before the registration of their birth are considered as stillbirths by French law and registered as such. The National Institute of Statistics (INSEE) makes separate counts of these false stillbirths and of true stillbirths, but the former have been included in birth and death statistics only since 1975.
-
[8]
Jacques Vallin and France Meslé. – Les causes de décès en France de 1925 à 1978, Paris, INED, PUF, Travaux et Documents, n° 115, 1988, 607 p.
-
[9]
In France, barely 20 % of deaths from infectious diseases are actually classified in this chapter of the ICD (see Jacques Vallin and France Meslé. – Les causes de décès en France…, op. cit. note 8).
-
[10]
For 1953-1964, we have naturally added to this group the infant deaths corresponding to the correction made for changes in the definition of live births.
-
[11]
Ideally, a more sophisticated method would have been used, taking into account the affinities existing between ill-defined causes and other causes. However, for this period, when the part played by theses diseases in total mortality was relatively modest, proportionnal redistribution seemed an acceptable middle course.
