Unexpected Mortality Trends

Trends in mortality in the last few decades in the industrialised countries (and elsewhere) show some curious paradoxes.

1. In the 1950s there appeared to be a level in infant mortality below which it would be difficult to descend, and the most advanced countries were thought to be quite close to attaining it. Yet deaths in the first year of life, expressed as a percentage of live births, continued to decline. In England and Wales this percentage fell by about one-quarter between 1956 and 1967; and by three-fifths by 1984, mainly due to reduction in perinatal mortality; there was a rather smaller diminution in the USA over that period, but in the Soviet Union also the corresponding ratio fell by nearly one-half, according to official figures.

Infant mortality rates per thousand in 1967, for the sexes combined, are shown in the table below for a selection of countries, arranged in increasing order, and the two right-hand columns give the proportions to those rates of the corresponding data for 1976 and for 1984 (or in one or two instances an adjacent year). The decline has been very substantial throughout, and it has not faltered, except perhaps for Czechoslovakia in the late 1960s and early 1970s.

An exception has been the experience in the Soviet Union, in which according to official figures the infant mortality rate rose a little between 1967 and 1976, and since then has done no more than regain its 1967 level. An analysis by Anderson and Silver¹ suggests that the rise may he attributable to an improvement in the completeness of death registration, and hence spurious, hut they do not question the general lack of improvement over the past 20 years; they point out however that considerable areas of the Soviet Union are hardly as well developed industrially as are the countries shown in the table, and that the data collected are still incomplete.

Forty years ago Bourgeois-Pichat² distinguished, as components of infant mortality, exogenous mortality (related to risks of infection, feeding and hygiene) which was rapidly declining, and endogenous mortality which had barely changed and which he thought would he much more difficult to alter. There was at that time little doubt that the exogenous part would continue to decline, but the endogenous part would not and this would greatly hinder further progress. The 1970s proved the contrary. In all western countries, even the most advanced, not only did infant mortality continue to decline at the same proportionate rate as in earlier decades, but endogenous mortality, for some 15 years, declined more quickly than exogenous, making up for a net slowing of the progress in the latter.

2. At adult ages, especially in men, progress seemed to stand still in the l960s in many countries which were far from the lower biological limit to which populations were thought to be able to approach with success in the struggle against exogenous causes. At certain ages at that time the risks of death tended to increase in numerous countries, and the expectation of life diminished, surprisingly contradicting the general secular trend towards a fall in mortality. This phenomenon was particularly clear in Eastern Europe, and in 1974 the authorities in the USSR ceased to publish the figures required to calculate expectations of life. Anderson and Silver³ reckon that while death rates fell for Russian men in their twenties they rose at higher ages, and some calculations by Blum and Monnier⁶ suggest that the increase, taking 1958/59 as the standard, was as much as 30 per cent by 1984/85 over the age range 40-70. This in contrast to the women’s experience, in which any rises were much more modest. Some indirect estimates by Pichat⁷ for the early 1980s appeared to be very severe, but they have been confirmed by renewed official publications following perestroika: the expectation of life at birth fell from 70.1 in 1964/65 to 67.4 in 1978/79, a loss of more than 2½ years – a trend of unprecedented severity which remains largely unexplained.⁸ A similar trend occurred in other countries of the Eastern Bloc, though slightly later, as the following figures illustrate for five countries combined with equal weight (Bulgaria, Czechoslovakia, East Germany, Hungary and Poland):

The Western countries also experienced a certain slowing of progress – indeed sometimes a slight reversal, mostly in the 1960s – but the expectation of life again rose quite rapidly after 1970. The following are some figures for Sweden; they show a simple index of mortality over the age range 47-72 for various years:

The data for individual years are naturally more open to variation than those for quinquennial groupings, but on the whole the figures reasonably illustrate the general trend. There was a tendency for the rate of improvement to be more rapid at the younger ages than at the older over the range in question.

For the United States of America, over the middle part of life (ages 25-74) the following approximate indexes of percentage improvement per decennium (etc) have been experienced since the beginning of the Second World War:

Once again it is necessary to treat the figures for individual years with more caution than those for a multi-year group, hut basically the same features are shown as for Sweden; also, in general, the rate of advance is better at the younger ages than at the older within the range illustrated.

In England and Wales, a fairly steady pace of progress has been possible, as the following rough guides show, but again the advance has recently been more rapid and the younger have fared better than the older in the development:

Sex and Age	1951/555	1961/65	1977	1984
Men 20-44	100	87	76	63
Men 45-64	100	96	85	74
Men 65-84	100	98	94	83
Women 20-44	100	76	62	56
Women 45-64	100	90	85	75
Women 65-84	100	91	86	71

3. A feature strongly evident in the foregoing figures is the much more favourable, or less unfavourable, set of changes for women than those for men over the whole period. So, while in the West the differences in behaviour and way of life between the sexes have greatly diminished, the excess male mortality has greatly increased at many ages. When at the beginning of the 1960s Coale and Demeny⁹ constructed their tables of types of mortality, they envisaged a rapid reduction in the disparity in expectation of life between sexes in the most advanced countries. In fact, the contrary has occurred. In France, for example, the mean length of life from birth increased for women from 73.6 in 1960 to 78.4 in 1980, but that for men grew only from 67.0 to 70.2. The difference between the sexes thus rose by more than 1½ years, from 6.6 years to 8.2. It is true that this difference seems to have since stabilised (in 1987 a mean duration for men of 72.0 and for women 80.3) but it is astonishing that it was able to increase by so much during these decades. Data from recent experience in some other countries are as follows:

Country	Year	Expectation of life at birth
Country	Year	Males	Females
USA	1983	71.0	78.3
Australia	1984	72.6	79.1
England and Wales	1982-84	71.6	77.6
Sweden	1984	73.8	79.9
USSR	1971-72	64.0	74.0
Czechoslovakia	1984	67.1	74.3

All these developed nations show a considerable disparity between the sexes, but that is not so evident for some developing countries, as the following official statistics seem to show:

Mexico	1979	62.1	66.0
Peru	1980-85	56.8	60.5
China	1980-85	66.7	68.9
Philippines	1980-85	60.2	63.7

That the difference is largely attributable to middle and later life is illustrated by the expectations of life remaining after age 40:

USA	1983	34.1	40.2
Mexico	1975-80	32.4	35.4

4. It is paradoxical that although economic development and industrialisation are responsible for big improvements in length of life, as may he seen from the following figures for areas in increasing stages of advance:

yet within a well developed country it is in the less-industrialised parts that mortality is lowest:

The example of Italy is interesting,¹⁰ for it is no longer in the North, economically the most developed part, with the highest standards of living, that expectation of life is greatest. The centre and the south, traditionally much poorer areas, now have quite a pronounced advantage in adult mortality. There are, however, one or two complications in this picture, as the following data show:⁵

The advantage in the south applies only in adulthood; in childhood the reverse is true. A study of the copious data set out in the United Nations Demographic Year Books shows that this sort of picture is found in many developed countries; the change-over from rural inferiority to rural superiority in vitality often does not occur until middle life is reached, at 40 or even 50 for men though perhaps at a different age for women. In the less developed countries, however, the urban areas tend to have the advantage at all ages: the following round figures are based on recent data supplied to the United Nations by Egypt, Israel, Pakistan and Puerto Rico:

Age group	Approximate percentage of rural to urban mortality
	Males	Females
5-14	250	180
30-44	130	120
55-69	130	120

There is some disparity between men and women in this statement, and in general for many countries the urban-rural differences are appreciably smaller for females than for males, and sometimes even in the opposite direction. One must suppose that the different occupations and way of life the sexes follow have a big influence here.

5. After the Second World War, the rapid disappearance of social inequality was expected by many, yet this continues in spite of forty years of social security and welfare states. The results of the English study of 1950/51 on social differences in mortality led some to think of a time when economic status would no longer be an important influence on mortality in the more advanced countries. Subsequent investigations show, however, that this is far from being achieved. Because of changes of method, and alterations of classification, it is not easy to measure with any precision how much change has taken place; it was written in l976⁴ that "a general study of a large number of social class mortality investigations made at different times has suggested that in developed countries, differentials are tending to decline in strength". Others, however, believe the contrary. Certainly mortality has fallen in all categories, but the economically most favoured groups may have profited by this diminution more than the others. Desplanques¹¹ found that, in men, mortality between ages 35 and 60 declined by 20 per cent for all non-manual occupations between 1955-59 and 1975-80 but fell by only 15 per cent for all manual occupations. Perhaps recent swings to the right in politics in many places may have had something to do with the difference between the findings of ten years ago and those of today?

6. To sum up, trends in mortality can no longer he explained by the pre dominance of a single factor such as the conquest of infectious diseases, or by a simple recourse to general medical progress and an improved standard of life. The reasons for change today are much more complex, and predictions cannot properly he based on a straight extrapolation from past tendencies for all causes of death in combination. A more delicate analysis of the numerous factors at work is required, however difficult this must be.

Biology and Society