The Russian Epidemiological Crisis as Mirrored by Mortality Trends
by Vladimir M. Shkolnikov and France Meslé 
In the mid-1960s, life expectancy began to decrease in the USSR and in many Eastern European countries. This occurred several years after life expectancy improvements slowed or reversed in some Western countries, and was the first time since the beginning of the epidemiological transition when the evidence compelled specialists to recognize that mortality decline does not always accompany economic development. In the 1970s, improvements in life expectancy resumed in Western countries, but in Eastern European countries the epidemiological crisis continues until now. Jean Bourgeois-Pichat, and later other scholars, pointed out that an excess mortality of adult males was the principal factor in the general stagnation of life expectancy in the East.
Unfavorable mortality trends were particularly acute in Russia. Many specific features of Soviet mortality patterns have been analyzed and discussed elsewhere. However, some principal features remained hidden due to the traditional lack of Soviet mortality data. Recently, a large amount of new data was obtained as part of a joint Russian-French project on the reconstruction of continuous time series of Russian mortality by cause of death. Whereas previously data were only available for a few points in time (since 1986) and large classes of causes of death, complete and comparable annual mortality data are now available by age and by 185 diagnostic categories from 1965 to 1993.
With improved data availability, it has become possible to evaluate the principal components of the change in life expectancy for the Russian population over time. The influences of different causes of death on the general trend of Russian life expectancy have changed substantially. Since 1985, the effect of violent death has partly overshadowed the more fundamental negative influence of a long-term increase in cardiovascular mortality, which had dominated earlier (in the late 1960s and the 1970s). These new data provide the opportunity to not only better explore the large classes of causes of death, but also trends for more specific causes of death.
The enormous increase in mortality of 1992-1993 compels us to investigate Russia's historical mortality experience in order to detect something that could help us to understand what is really new and what is linked to the past. Therefore, the general goal of the present paper is to investigate the genesis of the unprecedented Russian epidemiological crisis of the 1990s using the new data on mortality trends and patterns. To accomplish that purpose, we chronologically analyze Russian mortality by age since 1938-39, and by age and cause of death since 1965. We start from the late 1930s so that we can correctly understand Russia's mortality pattern before World War II, and perhaps find early evidence of excess adult mortality. Then, we will analyze the components of the significant rise in life expectancy for the Russian population in the 1940s and 1950s.
The year 1965 introduced the modern epidemiological crisis in Russia. For the period of gradual mortality increase (1965-1980), we point out a divergence in mortality trends among different age groups, focusing on causes of male mortality in their active ages (15-64). Next, we analyze the dramatic mortality fluctuations that have occurred since 1985. They include at least three components. These are: 1) the long-term unfavorable mortality trend; 2) the mortality decrease and then increase caused by the anti-alcohol campaign of 1985-87; and 3) the sharp mortality increase of 1992-93. The paper concludes with some speculations on the main factors likely to be contributing to the recent epidemiological crisis in Russia.
General Trends in Life Expectancy at Birth in Russia
At the beginning of the 20th century, life expectancy for the Russian population was extremely low (30 years for males and about 32 years for females in 1896-1897). It was about 15 years below the life expectancy levels for France, England and Wales, or the United States. In the 1920s, after a period of great social shocks in 1914-1922, important improvements were achieved in Russia. During that period, life expectancy increased by 10 years in men and 13 years in women. However, in the 1930s very little progress was achieved in Russia due to the famine of 1933 and the political persecutions of Stalin's era. As a result, in the 1930s the gap between Russia and the United States became even wider--about 20 years for both males and females (Table 4.1).
Temporal Changes in Life Expectancy at Birth in Russia, United States, and Japan, by Gender
NOTE: Sources of Russian data:, 1896-1897: Goskomstat, 1987. 1926-1927: Novoselski and Paevski, 1930 (corrected at infant and older ages for underrecording of death rates. Corrections are described in Adamets, Shkolnikov, 1995). 1938-1939: Economic Archives of the Russian Federation, Section 1562, List No. 329, Document No. 151 (corrected, see Adamets, Shkolnikov, 1995). 1958-1959 and 1965: corrected Goskomstat's estimates (see Shkolnikov, Meslé, Vallin, 1995a or 1995b).
Just after World War II, life expectancy rose very rapidly in Russia, while in Western countries it was growing at approximately the same speed as before the war. Between 1938-1939 and 1958-1959 in Russia, life expectancy increased by over 20 years for both sexes--from 40 to 62 for men and from 47 to 70 for women. It was a period of successive efforts by the Soviet centralized health care system against infectious diseases using new antibiotics and mass vaccinations. By the mid-1960s, the gap in life expectancy between Russia and the United States dropped sharply to only 3 years for men and 1.5 years for women. In the 1960s, when excess mortality from infectious diseases had already diminished greatly and the "civilized ills" (alcoholism, smoking, traffic accidents, environmental pollution) were rising, progress in life expectancy both in Russia and in the United States nearly stopped. Whereas further progress recovered in the United States and in many other Western countries in the 1970s, in Russia the decrease in life expectancy generally continued over the 30 years since 1965, especially for males.
The comparison between Russia and Japan is also illustrative--until the 1960s the Russian and Japanese situations were rather similar. Both countries were very much behind Western countries before World War II, but they radically overcame their backwardness in the late 1940s and the 1950s. Since the 1960s, however, the situations in Russia and Japan developed in opposite directions. While Japan achieved spectacular improvements in the control of degenerative diseases and experienced rapidly increasing life expectancy, Russia exhibited a continuous rise in cardiovascular and accidental mortality and a lowering of life expectancy.
Sex differentials in life expectancy were growing over time in Russia, Japan, and the United States. Between the 1920s and the 1970s, the difference in life expectancy by gender increased from 5.5 to 10 years in Russia, from 2.5 to 7.4 years in the United States and from 1.8 to 5.4 years in Japan, with the difference always larger in Russia than in other countries. Since the mid-1980s, excess male mortality has increased significantly in Russia. As a result, the difference in life expectancy by sex increased to 13 years, versus 6.7 years in the United States or Japan.
Russian Mortality: 1938-39 and Postwar Improvements
Our knowledge of mortality in the Russian population before World War II is very limited. Data on deaths by age and sex are kept in the Economic Archive of the Russian Federation (list No. 329). There are also official life tables for 1938-1939, based on 1939 numbers relative to the 1939 census population. Adjustment of these life tables for mortality underreporting at older ages leads to a modest correction in life expectancy at birth by -0.1 years for males and by -0.4 years for females. However, the real mortality at the end of 1930s was probably substantially higher than these estimates due to incompleteness of death registration for infants and for adults in some regions (for instance, concealing the number of Stalin's victims and/or losses in military conflicts with Finland and Japan). We also incorporate a fragment of archive data on deaths by age and cause of death for the urban population of Russia in 1938 (cause-of-death recording did not exist in rural areas at that time), as well as mortality numbers by month for 1932-35. Certainly, these data are also approximate.
Russian mortality by age and cause of death compared with the United States. The ratios of Russian age-specific death rates (ASDR) in 1938-39 to those in the United States show that the major part of the 20-year difference in life expectancy between the two countries can be ascribed to extremely high levels of infant and child mortality in Russia (Figure 4.1). Generally, the ratio of Russian to U.S. death rates declines from younger to older ages. For males ASDRs in Russia are higher than those in the United States even at older ages, while for females at ages beyond 45-49 ASDRs are nearly identical in the two countries.
Figure 4.1--Ratio of Age-Specific Death Rates of the Russian Population (1938-1939) to Corresponding Death Rates of the U.S. Population (1939-1941)
In the late 1930s the infant mortality rate (IMR) was close to 200 per 1,000 births in Russia and to 50 per 1,000 in the United States. At ages 1-4 and 5-9 the relative difference in mortality is even higher, with Russian death rates about 15 times higher than those in the United States. Thus, the Russian mortality pattern at ages under 15 is similar to that of the "South" family in the Coale-Demeny model life tables (Figure 4.4). The "South" family is generally characterized by bad sanitary conditions, relatively early abandonment of breast-feeding, and widespread practice of supplementary feeding with unsterilized products. This exact situation has been described by Sergei Novoselski in his study on extremely high IMR in ethnic Russians at the beginning of 20th century.
Figure 4.2 shows for each age group the proportion of mortality due to four principal diagnostic categories of causes of death: 1) infectious diseases (tuberculosis, diarrheal diseases, all other food-born or air-born infections, influenza, rheumatism, pneumonia and other acute respiratory diseases, septicemia, inflammatory diseases, venereal diseases, etc.); 2) degenerative diseases (cardiovascular diseases, cancers, chronic respiratory, digestive, geneto-urinary, nervous diseases); 3) external causes (accidents and violence); and 4) other and unspecified causes. The top figure presents information for Russian males in 1938-1939 while the lower panel presents the same information for U.S. males in 1940.
Figure 4.2--Proportion of Four Principal Classes of Causes of Death for Males, by Age in the Russian Urban Population (1938) and in the United States (1940)
The proportion of deaths caused by infectious diseases was particularly high in Russia at ages 1 to 14 and accounts for the vast majority of deaths in that age range, especially before age 5. It is significantly higher than in the United States (Figure 4.2). High incidence of diarrheal diseases, measles, dysentery, scarlet fever, respiratory infections, and tuberculosis caused maximum losses in Russia. In both countries the proportion of deaths from external causes increased in young adult ages. However, in Russia this proportion is higher and its maximum occurs in an older age group (32 percent at age 25-29 in Russia versus 19 percent at age 20-24 in the United States).
The percentage of deaths due to other and unspecified conditions is higher in the United States than in Russia, while the percentage due to degenerative diseases in older ages is higher in Russia than in the United States. These differences probably reflect difference in cause-of-death recording between the two countries. Russian physicians may have been forced to record well-defined diagnoses even in cases where the cause of death was not perfectly evident. This could result in overrecording of degenerative diseases in older ages.
Seasonal variations in Russian mortality. For a better understanding of Russian mortality patterns it is useful to look at the seasonal variation of mortality in the 1930s. Three components can be extracted from the initial series of absolute numbers of deaths by month for Russia (1932-35, 1956-5) and the USSR (1936-37): these are a linear trend, regular seasonal waves, and irregular short term fluctuations (Figure 4.3). Here we will just focus on seasonality and not discuss the linear trend or irregular fluctuations (for instance, the enormous explosion of deaths, with its maximum in June, due to the famine of 1933). The regular seasonal variation in the 1930s included two peaks: a smaller one in winter-spring due to respiratory infections and a bigger one in summer-autumn due to food-borne infections. In the 1930s, the number of excess deaths in August was considerable--about 20-30 percent (Table 4.2). By the 1950s, the seasonal variation had already been greatly reduced as the summer peak completely disappeared, suggesting that food-borne infections had been substantially reduced as a cause of death. The maximum for the cold season remained rather stable from the 1930s to the 1950s.
NOTE: Excess deaths in Russia in 1933 (upper panel) correspond to period of famine in rural areas.
Figure 4.3--Variation in Number of Deaths by Month in the 1930s and the 1950s: Russia and the USSR/p>
Monthly Index of Mortality in Russia and the USSR in the 1930s and the 1950s
NOTE: Monthly index = (seasonal wave+trend)/trend
Extremely high child mortality in the late 1930s and its reduction after World War II. Comparisons with Coale-Demeny model life tables (MLT) reveal the peculiar age structure of prewar mortality in Russia (Figure 4.4). In 1938-39, infant and child mortality were relatively high in Russia, while adult mortality was moderate. However, there was a relative increase in male mortality between the ages of 40-44 and 60-64. The reason for that is not clear. We do not know whether our data include (completely or partially) deaths in Siberian camps and "special settlements," where mortality was certainly very high. In a previous paper we have shown that between 1926-27 and 1938-39 male mortality at able-bodied ages increased slightly in spite of decreases at younger ages. Perhaps these are some early signs of the increase in adult male mortality that has characterized Russia in recent decades. It is obvious, however, that in the 1930s this phenomenon was much less important than high infant and child mortality.
Figure 4.4--Ratios of Russian Age-Specific Death Rates in 1938-1939 to Corresponding Values from Coale-Demeny Model Life Tables
Generally, Russia's mortality pattern in 1938-39 contains the main elements of mortality patterns in traditional (preindustrial) societies: extremely high death rates for infants and children, prevalence of infectious and diarrheal diseases among causes of death, and summer seasonality. On the contrary, the United States before World War II had already experienced important gains in the reduction of premature mortality. The traditional Russian backwardness in the course of its epidemiological transition, as well as the demographic catastrophes of 1914-1920 (World War I and the Civil War) and 1933 (famine), determined this difference.
A radical reduction of the infant mortality rate (IMR) occurred in Russia in the late 1940s and the 1950s. Between 1940 and 1956, the IMR in Russia decreased by approximately 4 times, from 200 per 1,000 to 49 per 1,000. In the next decade, the IMR was cut in half to 25.5 per 1,000 in 1965 (Figure 4.5). As a result, the increase in life expectancy at birth was very large between 1938-39 and 1965--24.3 years for males and 27.1 for females (Table 4.3). Decreasing mortality at ages under 5 contributed approximately equal fractions of the improvement for males and females (16.5 and 16.8 years, respectively). The influence of mortality decline at adult ages on overall life expectancy improvements was much smaller and unequal by sex: 7.8 years for males and 10.3 years for females. Hence, during two postwar decades, Russian women gained significantly more than Russian men due to the more favorable mortality trends at adult ages.
Figure 4.5--Infant Mortality Rate in Russia: 1940-1993
Components of Life Expectancy Increase in 1938-1939 to 1956, by Age
NOTE: Each cell of this table shows the number of years by which life expectancy at birth increased between 1938-39 and 1965 due to mortality decrease in corresponding age group.
Mortality Trends Since 1965: The New Epidemiological Crisis
Age patterns of mortality. In the early 1960s, progress in life expectancy for the Russian population slowed considerably. Annual increments in increasing life expectancy became very small until finally improvements ceased. For males the year 1965 was the turning point, when reductions in mortality from infectious diseases no longer offset negative tendencies in chronic diseases and violence. For females this point occurred several years later. The decline in life expectancy was rather slow until the mid-seventies; but after that it accelerated, particularly for males.
Mortality trends since 1965 clearly differ by age. A general deterioration at adult ages, particularly for men, contrasts with the slow progress at ages under 15 (Figure 4.6). The probability of survival from birth to age 15 was increasing in spite of some short-term variations. Simultaneously, probabilities of survival from age 15 to 45, from 45 to 65, and from 65 to 80 have been decreasing. The decrements in survival probabilities over 1965-1992 are approximately equal for the three age intervals (0.05 for men and 0.02 for women). Below we shall see that the negative changes between ages 20 and 60 are to a great extent responsible for a general decrease in life expectancy at birth.
NOTE: Dotted lines show extrapolations of trends based on 1965-1980.
Figure 4.6--Probabilities of Survival: From Birth to Age 15, from 15-45, from 45-65, and from 65-80: Russia, 1965-1993
Gorbachev's anti-alcohol campaign, which began in 1985, caused a break from the long-term trends. In 1985-1987 survival probabilities increased by 0.04 for males (0.01 for females), but by 1992 they had returned to the 1984 level. The decrease in survival probabilities was rather slow in 1988-1991 but then accelerated in 1992. The dramatic mortality rise in 1993 resulted in a 0.04-0.05 decrease in survival probabilities for each of the three adult age groups and even for children (Figure 4.6). In 1985-1993, the survival probabilities clearly deviate from the linear trends of 1965-1984 (shown in dotted lines in Figure 4.6). Extrapolating on these very unfavorable trends leading up to 1993 results in values that are still better than the values actually observed in 1993.
The continuous growth in male mortality at active ages has led to a very special age pattern of male mortality in Russia. Comparison of Russian mortality patterns by age to the Coale-Demeny model life tables indicates clearly the spectacular excess mortality at adult ages since 1985. This excess mortality has been increasing over time and has its maximum at ages 35-45 (Figure 4.7). Whatever model is selected for comparison, the peculiarity of the Russian pattern is absolutely obvious. However, in contrast to the 1930s, infant mortality is lower than corresponding model values.
Figure 4.7--Ratios of Male Age-Specific Death Rates to Corresponding Death Rates from Coale-Demeny Model Life Tables (Families West and South)
The method of component analysis allows us to split the difference between life expectancy for the Russian population in 1965 and 1980 into its components, by age and cause of death (Table 4.4). Once again, increasing mortality at adult ages reveals itself as the main factor reducing life expectancy. The active ages (age 15 to 64) contribute 2.8 years to the total 3.1 year decrease for men. Excess mortality of men aged 15 to 44 contributes 1.3 years to the negative total. Most of these losses are due to cardiovascular (circulatory) mortality, injuries, and violence. Declining mortality from infectious, respiratory, neoplasmic, digestive, and other diseases for most age groups under 15 partly compensates for the prevailing negative changes at older ages.
Components of Life Expectancy Change in 1965-1980, by Age and Cause of Death
|Age||Infectious||Neo-plasms||Circulatory||Respiratory||Digestive||Other Diseases||Injury, Violence||Total|
|Age||Infectious||Neo-plasms||Circulatory||Respiratory||Digestive||Other Diseases||Injury, Violence||Total|
For women, the trend in life expectancy is much less unfavorable. Between 1965 and 1980, their life expectancy at birth decreased by 0.7 years. The components of that decline are different than those for males. The most important negative effect for women is attributable to circulatory diseases at older ages, while the negative effect of injury and violence is much smaller than for men. For both males and females, the trend in cancer mortality is surprisingly stable. It has even shown some positive influence on life expectancy.
Causes of Male Mortality at Active Ages
Given that the increasing excess mortality of adult males from cardiovascular diseases and external causes are the main factors in declining life expectancy for the Russian population, we will now focus our attention on males between the ages of 15 and 64. Age-adjusted (standardized) death rates for men aged 15-64 will be used to compare mortality levels. These standardized death rates, or SDRs, are weighted averages of age-specific death rates.
Table 4.5 presents a set of standardized death rates (SDRs) for twenty-one principal causes of death for Russian males at active ages. Below we will examine the situations in 1965 and 1980 (the beginning and end points of gradual mortality growth), 1987 (the year of minimum mortality due to anti-alcohol policies), and in 1993 (the most recent maximum). For comparison with a Western country, corresponding SDRs for the United States (1989) are also presented.
Standardized Death Rates, by Selected Causes of Death in Russia and in the United States: Males, Ages 15-64
|Death Rate Per 100,000||Ratio of Russian to 1989 U.S. Rates|
|Cause of Death||Russia 1965 (1)||Russia 1980 (2)||Russia 1987 (3)||Russia 1993 (4)||USA 1989 (5)||1965 (1)/(5)||1980 (2)/(5)||1987 (3)/(5)||1993 (4)/(5)|
|Infectious and parasitic diseases||47.63||23.31||14.58||25.00||5.07||9.4||4.6||2.9||4.9|
|- Cancer of stomach||56.49||36.17||30.91||26.83||2.90||19.5||12.5||10.7||9.3|
|- Cancer of intestine and rectum||5.65||8.85||9.21||10.53||8.20||0.7||1.1||1.1||1.3|
|- Cancer of bronchus and lung||37.91||46.55||53.25||54.99||33.92||1.1||1.4||1.6||1.6|
|- Cancer of prostatis||1.15||1.56||1.86||2.26||3.19||0.4||0.5||0.6||0.7|
|- Other neoplasms||45.26||49.48||56.05||64.75||48.44||0.9||1.0||1.2||1.3|
|- Hypertensive disease||8.46||5.14||3.46||6.07||4.60||1.8||1.1||0.8||1.3|
|- Ischaemic heart disease||74.37||151.78||134.28||210.70||65.66||1.1||2.3||2.0||3.2|
|- Other heart diseases||5.96||9.92||10.66||32.79||26.96||0.2||0.4||0.4||1.2|
|- Other circulatory||5.32||9.27||7.30||10.99||4.39||1.2||2.1||1.7||2.5|
|- Liver cirrhosis||5.88||13.18||8.38||14.23||11.74||0.5||1.1||0.7||1.2|
|- Other digestive diseases||13.21||16.86||12.75||20.90||5.99||2.2||2.8||2.1||3.5|
|- Road accidents||9.24||24.96||17.94||35.78||21.23||0.4||1.2||0.8||1.7|
|- Accidental poisonings||21.68||54.36||23.52||68.18||3.39||6.4||16.0||6.9||20.1|
|- Other Injury/Violence||72.29||98.87||53.46||135.83||3.49||20.7||28.3||15.3||38.9|
|Other causes of death||18.83||25.63||19.46||33.12||44.57||0.4||0.6||0.4||0.7|
|All causes combined||578.19||779.35||598.29||1003.49||349.37||1.7||2.2||1.7||2.9|
NOTE: Standardized death rate-age-adjusted death rate.
Russian male SDRs from all causes between the ages of 15 and 64 have been increasing since 1965. In 1987, mortality returned approximately to the level of 1965 (in fact, in 1987 it was a little higher), and finally, in 1993, it reached approximately twice the level in 1965. Most of this variation over time is related to changes in cardiovascular and accidental and violent mortality. In Russia, as in the United States, circulatory diseases account for approximately 33 percent of total active-age SDR, while the proportion due to injuries and violence is much higher in Russia (32 percent versus 16 percent in the United States). In contrast, the proportion of cancer mortality is greater in the United States (28 percent versus 19 percent).
Looking at specific causes of death, we find that, by 1993, mortality in active ages from infectious diseases, rheumatism, and hypertensive disease (reflecting Russia's initial delay in its epidemiological transition) and mortality from stomach cancer had dropped below the levels of the 1960s. On the other hand, Russia has lost its former advantages over the United States in mortality due to some cancers, liver cirrhosis, and traffic accidents. This is also true for "other heart disease" and homicide.
The leading role in the general increase in SDR for men aged 15 to 64 belongs to such causes of death as cerebrovascular disorders and ischaemic heart disease, lung cancer, accidental poisoning, suicide, and homicide. In 1965, SDRs for these causes were already high, but they became even higher in 1993. For instance, the ratio of Russian SDRs in 1993 to that in the United States in 1989 are: 3.2 for ischaemic heart disease, 7.6 for cerebrovascular disorders, 1.6 for lung cancer, 20.1 for accidental poisoning, 3.8 for suicide, and 3.7 for homicide.
The steady rise in the SDR for active ages was interrupted in the mid-1980s by a sharp decline caused by the anti-alcohol campaign (Figure 4.8). The rapid mortality decrease lasted only 3 years, from 1985 to 1987. Reduced mortality from external causes played the most important role in reducing total active-age mortality during the anti-alcohol campaign, although many other causes of death such as ischaemic heart disease, respiratory diseases, and liver cirrhosis contributed to the rapid changes in mortality since 1985. This will be discussed further below. The mortality level increased again after 1987, with the 1984 level nearly restored by 1992. In 1993, active-age mortality greatly exceeded that of the previous year, and grew beyond expectations based on long term trends (upper panel in Figure 4.8). (Similar results for survival rates were shown in Figure 4.6.)
NOTE: Standardized death rates per 100,000 at ages 15-65, 1965-1993
Figure 4.8--Trends in Russian Male Mortality in Active Ages: All Causes Combined and Principal Causes of Death
Some variations in mortality trends for men aged 15 to 64 within the broad classes of causes of death are noteworthy. For instance, different directions in trends for stomach cancer (decreasing) and lung cancer (increasing) resulted in relative stability of total mortality from neoplasms. Since 1988, deaths from certain causes rose sharply compared to others of the same class, e.g., "other heart diseases" among diseases of the circulatory system; homicide and accidental poisoning among external causes of death). The dramatic increase in "other heart diseases" was probably induced by improvements in diagnosis of several specific heart diseases within that classification.
Russian Mortality Trends in the International Context
Comparing Russian mortality by leading causes of death to that in Western countries helps us to detect the distinct epidemiological outcomes of two fundamentally different social systems and approaches to public health. Comparisons with Eastern European countries show many similarities due to similar social and medical care mechanisms, although some important peculiarities of Russian trends are also evident.
In this section, we compare the Russian trends in SDRs in 1970-1993 by circulatory diseases, external causes of death, neoplasms and respiratory diseases with those in the United States (1979-1989), with average SDRs weighted by population sizes for European Community countries (EC, 1970-1992), and with weighted average SDRs for Eastern and Central European countries (CEE, 1970-1992). Whatever cause of death we consider, mortality levels and trends are better for the EC and the United States than for CEE countries and Russia, particularly for men (Figure 4.9). Nevertheless, the general situation in CEE countries is substantially better than in Russia, in spite of the long-term worsening in mortality trends in many CEE countries. The 1993 rise in Russian mortality is clearly visible for all causes of death except cancers.
Figure 4.9--Time Trends in Standardized Death Rates for All Ages: Russia, European Community Countries, Countries of Central and Eastern Europe, and United States
The part played by circulatory diseases in the differences between groups of countries is overwhelming (Figure 4.9). Not only has this level always been higher in Russia, but it has been increasing, whereas trends in circulatory diseases have been improving in the EC and the United States.
The enormous level of male mortality from injuries and violence, and especially its rapid increase in the 1990s, sharply differentiates the Russian trend from the Eastern European and Western countries. For women the gap in violent mortality is much smaller, but again the Russian excess mortality is obvious, especially in recent years.
Cancer mortality in men has become more and more significant as a component of the general mortality gap between Russia and the West. Until the early 1980s, cancer mortality for men in Russia was higher than in the reference countries, but it was also rather stable. While in the 1970s cancer mortality was growing in EC, CEE, and in the United States, the difference between these countries and Russia became very small by 1979-1980. Thereafter, the Russian trend began to increase, while the EC and U.S. trends stabilized or even declined. Like Russia, the CEE average also exhibited a continuous increase. For females the difference in levels of cancer mortality across countries is much less significant; however in the 1980s and the 1990s, the Russian position is becoming relatively worse.
Mortality from respiratory diseases has been declining for both sexes and in all countries except the United States. Progress in Russia continued with some fluctuations until the 1990s, when it slowed down. The sudden 1993 increase returned mortality due to respiratory diseases to its 1984 level.
To examine the reasons for very low life expectancy in Russia in terms of causes of death, in Figure 4.10 we look at age- and cause-specific components of the difference in life expectancy between Russia and the United States, and between Russia and Hungary. In order to investigate the "normal" Russian pattern, we select the moment just preceding the unfavorable changes of 1992-1993. Later, we shall specially investigate the most recent mortality increase. The United States represents the Western countries, while Hungary represents the CEE countries. Hungary is chosen as a country of deep epidemiological crisis, which is probably rather similar to Russian's deterioration. During the 1970s and the 1980s, the values of life expectancy at birth in Hungary were not very different from that in Russia.
Figure 4.10--Contributions of Different Ages and Causes of Death to the Total Difference in Life Expectancy Between the United States and Russia in 1989, and Between Hungary and Russia in 1991
The United States versus Russia. The total difference in male life expectancy between the United States and Russia was 7.7 years in 1989. This amount consists of three main parts: 1) at ages above 40 there is a strong influence of high mortality from circulatory diseases in Russia and, to much smaller extent, mortality from neoplasms (Figure 4.10); 2) at younger adult ages (15-39) the role of excess mortality from injuries and violence is very clear for males; at middle ages (45-64) the negative contributions of external causes of death are also substantial; and 3) there is also a less important, negative effect from the relatively high Russian IMR, due to a high infant mortality rate from "other diseases" (e.g., perinatal conditions and congenital anomalies), and respiratory and infectious diseases. For all ages together, the contribution of circulatory diseases to the total difference in male life expectancy between the United States and Russia is 3.8 years, while the contribution by external causes is 2.4, and neoplasms add another 0.9 years.
For women, the structure of the gap in life expectancy (4.2 years in 1989) is much more uniform, because it depends mostly on excess mortality from circulatory diseases at older ages (Figure 4.10). Excess mortality from external causes is significant at active ages, but is much less important than for men.
The relatively low mortality of Russian women at older ages from neoplasms and respiratory and digestive diseases slightly compensates for the negative influence of very high cardiovascular mortality. There is a hypothesis that the coding practice at advanced ages in Russia differs from many other countries. Many physicians and statisticians in Russia believe that there is a tendency to over-record cardiovascular deaths and under-record deaths from neoplasms and some other specific conditions. However, in our previous studies we have not found any hard evidence for such an assumption.
Hungary versus Russia. Comparing the Russian male mortality pattern with that in Hungary, we discover a very peculiar structure of total difference in life expectancy (1.6 years in 1991). It depends mainly on external causes of death, while the contribution of circulatory diseases is rather low (Figure 4.10). Besides, mortality from certain chronic diseases, particularly diseases of the digestive tract, is substantially higher in Hungary than in Russia. This means that mortality from chronic diseases is very high in both countries, and is even worse in Hungary than in Russia for some causes of death. However, the enormously high Russian mortality from external causes in able-bodied ages reflects poorly on Russia in terms of the difference in male life expectancy.
For females the total difference in life expectancy between Russia and Hungary is rather small. In 1991, life expectancy for Russia was higher by 0.4 years than that for Hungary. The components related to external causes at young ages and circulatory diseases at older ages are unfavorable for Russia; however, the components related to cancers, digestive diseases, and other diseases are unfavorable for Hungary.
Differences in Life Expectancy by Sex
We have already emphasized several times the differences between male and female mortality. Russian male excess mortality is probably the highest in the world. As has been shown, during the rapid mortality decline in the early 1950s females gained considerably more than males due to better trends at adult ages. This tendency has been strengthening over time. From 1958-59 to 1993, the difference in life expectancy between males and females grew from 7.3 years (with life expectancy values corrected for mortality under-estimation in infant and older ages) to 13 years. To quantify the Russian mortality differentials by sex, we once again decompose the total difference in life expectancy according to leading classes of causes of death (Table 4.6).
Differences in Life Expectancy Between Females and Males, by Cause of Death
|Cause of Death||1965||1970||1975||1980||1984||1987||1992||1993|
|Injury, poisoning, and violence||2.95||3.68||3.83||4.09||3.87||2.50||4.26||4.94|
Throughout the 1959-93 period, more than 60 percent of the difference between males and females is due to only two classes of causes of death: injury, poisoning, and violence; and circulatory diseases. In each year, except the very unique year 1987, the injuries, poisoning, and violence explains around 35 percent of the total. The difference in mortality due to circulatory disease grew from 25 percent in 1965 to 30-32 percent in the 1980s and the 1990s.
In sum, the widening gap between the sexes is due to increasing male excess mortality in cardiovascular circulatory diseases and cancer, but also from external causes of death. The spectacular reduction of the difference between male and female life expectancies in 1987 (the difference fell from 11.3 in 1984 to 9.5 in 1987) is closely related to the diminution of the difference in mortality from injuries, poisoning, and violence. The recent increase in the gap from 1987 to 1993 is due to a new deterioration in injuries, poisoning, and violence as well as in cardiovascular diseases, all of which have been especially acute in men.
The Anti-Alcohol Campaign and Variations in Russian Mortality
Usually it is very difficult or even impossible to extract the impact of certain environmental factors or personal behavior on the mortality trend because, in actuality, all these factors are developing simultaneously. The anti-alcohol campaign of 1985-1987 provides the unique possibility to extract a pure effect of alcohol abuse on mortality. The action was so rapid and sharp that it is possible to assume that virtually nothing changed in other conditions of public health within that short period. In fact, these "other" conditions were changing rather slowly at that time.
In June 1985 the Gorbachev government organized a great experiment on the Soviet population. They tested the results of the introduction of a 2.5-fold reduction in state alcohol sales. They were able to directly implement this policy because in the mid-1980s state trade was the only legal source of alcohol in Russia. The main aspects of the anti-alcohol campaign were: reduction of state alcohol production; enforcement actions against distillation and distribution of homemade beverages ("samogon," or moonshine); elevation of state prices for alcohol in August 1985 and in August 1986; and further development of state institutions for compulsory treatment of alcoholism.
All of the anti-alcohol actions were directed to restriction of public access to alcohol, but they did not affect the motives and underlying factors of alcohol abuse. Thus, there was no hope to derive a long-term positive effect. Nevertheless, the anti-alcohol campaign brought about very strong-short-term fluctuations, which are clearly visible in Russian mortality trends. This paper points out only several principal findings, because we have already provided detailed analysis of the alcohol situation in Russia and mortality outcomes from the anti-alcohol campaign in another paper.
A sharp mortality drop is seen immediately after the introduction of anti-alcohol restrictions. The drop was much more substantial for males than for females. The number of avoided deaths, estimated by comparison with the long-term trends in ASDRs, exceeds 900,000 (about 600,000 among males and about 300,000 among females). In 1987 the Russian mortality rate reached its absolute minimum, and in 1988 the reverse (upward) trend began. In 1988-1991 the mortality rate increased rather moderately, whereas in 1992 (when actual alcohol consumption in Russia jumped from 12 to 14 liters of pure ethanol per capita) mortality increased sharply.
The overwhelming role of injuries and violence in active ages for the general mortality decrease in 1985-87 and subsequent increase in 1988-92 is obvious (Figure 4.11); the structures of mortality decrease and increase are similar. Death rates for males aged 20 to 54 exhibit very large decrements in 1985-87 and increments in 1987-92, compared to the minimum level of 1987. For instance, at ages 30-34 and 35-39 (ages of the biggest relative changes in death rates) the difference between ASDRs in 1980 and 1987 exceeds 0.7 of the ASDR value in 1987, and the increase from 1987 to 1992 is also about this magnitude. Injuries and violence are responsible for the predominant part of these changes. However, a part of the variation in male ASDRs, particularly at older ages, is due to cardiovascular diseases.
Figure 4.11--Proportions of Decrease and Increase in Russial Male ASDRs Associated With the Anti-Alcohol Campaign, 1985-1987
Between 1980 and 1987, life expectancy at birth increased by 3.5 years for males and by 1.3 years for females; then from 1987 to 1992 it decreased by 2.9 years for males and 0.6 years for females (Table 4.7). So, the improvement of 1985-87 is slightly bigger than the later deterioration, although the positive and negative contributions of external causes of death and digestive diseases (mostly liver cirrhosis) are exactly symmetrical.
Changes in Life Expectancy of Russian Population by Cause of Death: 1980-1987 and 1987-1992, by Gender
|Cause of Death||1980-1987||1987-1992||1980-1987||1987-1992|
|Injury and poisoning||2.01||-2.16||0.50||-0.56|
The leading influence of accidental and violent causes of death in 1985-1992 is obviously due to their exceptional level in Russia, which is far above any value observed elsewhere, but also due to the especially close relationship between these causes of death and alcohol consumption. So, external causes of death compose the most important part of alcohol-related mortality in Russia. This mortality pattern does not correspond to the main causes of alcohol-related deaths in other countries with high alcohol consumption. For example, in France the level of mortality from alcoholism is about 5 times the level in Russia, while mortality from accidental poisoning by alcohol is negligible in France and enormously high in Russia. Hence, alcohol abuse in Russia results mostly in immediate and acute consequences (e.g., accidents).
For an explanation of this phenomenon, we can refer to the influence of the Russian style of drinking. The immediate effect is often produced by drinking large doses of vodka in a short time with very little to eat. That can result in the loss of self control and in irresponsible or aggressive behavior.
However, we have to be cautious when comparing Russian death rates due to alcohol-related causes with those in Western countries because, in part, the difference depends on registration habits. It seems that many deaths caused by the acute effects of alcoholism are recorded in Russia as poisonings (by alcohol) or other accidents, without reference to the underlying alcoholism.
Another interesting matter concerns the significant reduction in mortality from cardiovascular and some other degenerative diseases during the anti-alcohol campaign. At first this seems surprising given the rather long pathogenesis of this sort of disease. However, a sudden reduction of alcohol consumption for people who are already predisposed to die from cardiovascular disease (for instance, cardiac patients) could certainly induce a rapid decrease in their mortality. Besides, before the campaign it is possible that some of the deaths from alcoholism were recorded as "acute heart disorders," according to the immediate cause of death.
Dramatic Mortality Increase in 1992-1993
In 1992-1993, life expectancy at birth for the Russian population took an unprecedented fall. Between 1991 and 1992, Russian males lost 1.5 years in life expectancy and females lost 0.5 years. As we have seen, a big share of this decrease is linked to the increase in alcohol consumption, following its reduction during the anti-alcohol campaign. However, in 1992 some new and alarming tendencies, not related to alcohol abuse, appeared in the Russian mortality pattern. In the following year they manifested themselves very clearly.
In 1993 the public health situation worsened so much that at first it seemed unbelievable. Between 1992 and 1993 life expectancy decreased by 3 years for males and by 1.8 years for females. By 1993, male life expectancy was 59 years, female life expectancy was 72 years. No country has exhibited such an abrupt change in peacetime.
In this section we shall analyze several important aspects of the increase in Russian mortality in 1992-93. First of all, we shall look at the absolute increase in deaths to try to evaluate the pure effect of rising mortality on the natural decrease in the Russian population. Then we shall discuss specific changes in Russian mortality by age (paying special attention to older age and infant mortality) and by cause of death. Changes in death rates for selected avoidable causes of death are to be considered as good indicators of deterioration in medical care practices. Finally, we apply a simple extrapolation of trends in ASDRs (assuming cohort effects on mortality to be negligible) in order to clarify what part of the 1993 life expectancy decrease may be associated with long-term mortality tendencies, which will be considered later.
Absolute mortality increase. The year 1992 was the first when population decline was registered in Russia. Certainly, the long-term fertility decline and population aging predisposed this event, although in recent years the decrease in the natural growth rate has accelerated due to a sharp fall in births and a sharp rise in deaths. In 1992 the annual number of births was lower than the annual number of deaths by 220,000; in 1993 this difference increased to 750,000.
Changes in the absolute number of deaths can be decomposed into two components: changes in age-specific death rates (the force of mortality) and changes in population size and age structure (aging). To extract a pure effect of mortality rates, we calculate the expected numbers of deaths for each year 1989-1993 period, applying age-specific death rates for the year under consideration to the population figures by age from the 1989 census ("expected deaths" in Table 4.8). The expected deaths correspond to the assumption of constancy in population size and age structure.
Recent Increase in Total Deaths for Both Sexes: Observed Values Versus Expected Values, Based on Census Population of 1989 (in thousands)
|Year||Total Deaths||Annual Increase|
NOTE: Expected death numbers are yielded by applying observed ASDRs to the census population figures of 1989.
Between 1989 and 1993, the observed annual number of deaths increased by 546,000 (from 1.583 million to 2.129 million). The increase in expected deaths, assuming the population size and age structure did not change between 1989 and 1993, is 432,000. The difference between the two figures (114,000) shows the influence of population aging. Hence, the majority of the increase in the annual deaths in 1989-93 is due to increases in age-specific mortality rates. The portion of the increase in the number of deaths due to increases in age-specific mortality rates increased over the 1989-93 period: over 90 percent of the increase in number of deaths between 1992 and 1993 is due to the increase in age-specific death rates. That is opposite to the overwhelming influence of population aging in 1990-1991. Between 1990 and 1991, aging accounted for over 90 percent of the increase in the number of deaths.
Age pattern of mortality increase. As noted above, the long-term tendency of overwhelming deterioration of life expectancy at active ages was interrupted by the anti-alcohol campaign. Deterioration was resumed in 1988 and was reinforced in 1993. The biggest relative increase in ASDRs occurred for men aged 25-44 and women aged 15-44, while in older and younger ages the increases are much smaller (Figure 4.12). For the ages of maximum relative mortality increase, the ratios of current male ASDRs to corresponding minimum levels of 1987 were about 1.3 in 1991, about 1.5-1.6 in 1992 and about 2.0 in 1993. For females aged 15 to 44, these ratios were about 1.2 in 1991, 1.35 in 1992, and 1.6 in 1993. Therefore, at these ages mortality rates were 20-100 percent higher in 1991-1993 than in 1989.
Figure 4.12--Age-Specific Death Rates in 1991, 1992, and 1993, Divided by Corresponding Death Rates in 1987, by Gender
Although the bell-shaped age patterns of excess mortality in 1991, 1992, and 1993 seem to be similar, there are indeed some peculiarities for the year 1993. First of all, the level of excess mortality in 1993 is enormously high. Besides, the relative mortality increase in middle and older ages (e.g., ages 45-74) is much more significant than in previous years. This is especially obvious for women, because the age of maximum excess mortality has moved from 15-19 in 1991-92 to 40-44 in 1993. In 1991-92 age-specific death rates at ages over 70 were slightly below the level of 1987, while in 1993 they were higher. The mortality increase in older ages between 1992 and 1993, visible in Figure 4.12, is seen even more clearly in terms of the age-specific death rates shown in Figure 4.13.
Figure 4.13--Male Death Rates for Different Ages: 1970-1993
Increase in infant mortality. We have already seen that over the period since 1965, except several years in the 1970s, the infant mortality rate (IMR) was declining in spite of unfavorable trends at adult ages. This long-term tendency was violated in the 1990s (Figure 4.13). In 1990 the IMR reached its minimum (17.4 per 1000 live births). In 1991 and 1992 it became slightly higher: 17.8 and 18.0, respectively. Infant mortality rates at this level are also characteristic for other former Soviet republics and Romania.
Some of the apparent deterioration in the IMR may have been induced by changes in the registration of live births, introduced at the beginning of 1993. That modification should, however, affect only mortality in the first month of life (neonatal mortality) and not data for later infancy. Data on infant mortality by age allow us to split the observed increase in IMR according to the subperiods of the first year of life. Infant deaths during the first month explain less than half of the total rise in IMR between 1992 and 1993, implying that over half of the increase is real and not due to the change in registration practice. Therefore, the value of IMR in 1993 should be at least 19 per 1,000 live births, according to the old registration procedure.
Certainly, a significant increase in IMR is a very alarming sign of a general deterioration in health conditions. Even "exogenous" causes of infant death (such as respiratory, infectious, and digestive diseases) increased substantially in 1993. However, as we have already seen, the increase in the infant mortality rate is substantially smaller than that for death rates in able-bodied ages.
Decrease in life expectancy. All the extremely unfavorable changes mentioned above have induced a strong negative influence on life expectancy at birth (Table 4.9). For males, the contribution of injuries and violence to the total decrease in life expectancy between 1992 and 1993 is still the highest (1.3 years). However, in contrast to 1988-1992, the contribution of cardiovascular diseases is also very high (1.1 years).
Components of Life Expectancy Decrease in 1992-1993, by Age and Cause of Death
|Age||Infectious||Neo-plasms||Circulatory||Respiratory||Digestive||Other Diseases||Injury, Violence||Total|
|Age||Infectious||Neo-plasms||Circulatory||Respiratory||Digestive||Other Diseases||Injury, Violence||Total|
Increasing mortality from circulatory diseases has its maximum negative impact at ages 45-64, while the highest negative impact of external causes occurs at ages 15-44. All other classes of causes of death, except cancers, contribute substantially to the general deterioration of life expectancy. Even infectious diseases, which had never before been a factor of worsening in postwar times, induced a loss of 0.1 years in male life expectancy at birth.
For women the contributions of circulatory diseases and external causes of death to the decrease in life expectancy between 1992 and 1993 differ from that for men due to a much more pronounced influence of cardiovascular diseases at ages over 45, rather than injuries and violence. Increasing mortality from circulatory diseases caused a loss of 0.94 years, while an increased number of deaths from external causes corresponds to a loss of 0.5 years.
Finally, it is important to point out that, in 1993, mortality from all causes of death combined brought significant losses in life expectancy, even at ages under 15 (0.1 years for males and 0.2 years for females).
Changes in specific causes of death. For a more comprehensive analysis of the reasons for the rapid mortality increase in 1992-93, we should look at the changing pattern of Russian mortality in terms of specific causes of death (Table 4.10). We shall try to find out some new tendencies in 1992-93 which were not characteristic in the previous period.
Standardized Death Rates for Selected Causes of Death: 1980, 1991, 1992, 1993 (per 100,000)
|Cancer of intestine and rectum||21.73||28.09||28.25||29.78||1.37||1.05||16.33||19.46||19.75||19.92||1.22||1.01|
|Cancer of bronchus and lung||84.69||103.74||104.83||105.67||1.25||1.01||9.48||10.57||10.80||10.77||1.14||1.00|
|Cancer of uterus||14.68||13.66||13.89||14.28||0.97||1.03|
|Cancer of prostatis||8.04||10.58||11.79||11.94||1.49||1.01|
|Other heart diseases||28.50||40.08||47.05||61.62||2.16||1.31||16.34||22.73||24.57||30.40||1.86||1.24|
|Other circulatory diseases||59.40||67.73||69.84||82.17||1.38||1.18||40.05||51.92||53.61||61.88||1.55||1.15|
|Other digestive diseases||30.44||26.44||32.16||35.38||1.16||1.10||11.83||11.84||13.42||14.00||1.18||1.04|
|Other injuries and violence||119.83||96.05||118.68||161.58||1.35||1.36||25.23||22.68||27.77||37.27||1.48||1.34|
|Other causes of death||65.80||69.31||72.31||82.77||1.26||1.14||40.07||49.59||50.57||56.62||1.41||1.12|
|All causes combined||1867.82||1683.29||1810.96||2145.15||1.15||1.18||958.61||884.76||919.49||1043.91||1.09||1.14|
The total increase in the standardized death rate (SDR) in 1992-93 was equal to 440/100,000 for males and 150/100,000 for females. More than 80 percent of these increases are due to circulatory diseases and external causes of death. The ratio of the increase in the SDR due to circulatory diseases to the change in SDR due to external causes is 1.4:1 for males and 3.5:1 for females.
Mortality from neoplasms is growing gradually over time without unexpected changes. Mortality from diseases of an infectious nature (including rheumatism and respiratory diseases) had been declining until the 1990s, but in 1992, and especially in 1993, it exhibited a substantial and unexpected rise. However, due to previous improvements, the death rate from infectious diseases in 1993 was less than that in 1980.
Mortality from diseases of the circulatory system was slowly increasing in the 1970s. In the 1980s, the situation was rather stable (except "other heart diseases" and "other circulatory diseases," which probably increased due to changing coding practices). A moderate increase resumed in 1992, though the SDRs were still below the levels of 1980, except "other" and cerebrovascular disorders. However, with the sharp increases between 1992 and 1993, cardiovascular mortality reached levels in 1993 substantially above the levels of 1980 (except for hypertensive disease in men and ischaemic heart disease in women).
Perhaps the most striking changes occurred in mortality from external causes of death. First of all, this class of causes of death has the highest relative rate of growth in the 1990s. As we have already seen (Figure 4.8), the mortality decrease during Gorbachev's anti-alcohol campaign corresponded to reduction in all types of violent deaths, but the subsequent reversal was induced by different dynamics. In the late 1980s and 1990s, growing alcohol consumption is no longer the only explanation of increased deaths from external causes. Homicides, suicides, accidental poisonings, and other causes of accidents and trauma exhibited a very large increase. In 1993, mortality from homicide, suicide, and accidental poisonings (by alcohol) were about the highest in the world. The Russian male SDR by suicide was higher than in Hungary (by 17 percent), and the Russian male SDR by homicide was 2.5-fold higher than in the United States. The Russian male SDR by accidental poisoning is at least 10-fold higher than that in any other country except the former Soviet republics. On the contrary, the SDR by traffic accidents rose sharply in 1987-1989, but it ceased to increase in the 1990s. The reasons for that are not clear. Perhaps it was due to shortages in oil supplies and/or activities of traffic police.
Increase in avoidable causes of death. Avoidable causes of death refers to causes of death amenable to medical intervention. A variety of mortality studies have confirmed that mortality due to certain diagnostic categories is particularly sensitive to medical care, either treatment or prevention.
Table 4.11 contains SDRs for selected avoidable causes of death. Mortality from many of these causes has been generally declining since 1965 (tuberculosis, intestinal infections, pneumonia, appendicitis, abdominal hernia, rheumatism, hypertensive disease, cancer of the uterus, and complications of pregnancy and childbirth). However, some avoidable causes of death were increasing or stagnating during the 1970s and the 1980s (bronchitis and lung emphysema, asthma, gastric ulcer, diabetes mellitus, and skin cancer).
Standardized Death Rates for "Avoidable" Causes of Death: 1965, 1980, 1985, 1990-1993 (per 100,000)
|Infectious and parasitic diseases||71.9||39.2||30.4||22.3||21.0||25.0||33.4||22.9||11.5||8.9||5.9||5.7||5.6||7.3|
|- Dysentery and other intestinal infections||2.9||2.8||2.7||1.4||1.5||1.4||2.0||1.9||2.1||2.3||1.2||1.3||1.0||1.4|
|- Bronchitis and lung emphysema||36.3||35.8||51.8||62.4||56.3||57.4||69.3||14.5||14.3||17.0||16.7||14.7||14.7||16.0|
|- Gastric ulcer||5.3||9.1||9.0||8.4||8.8||12.3||12.9||1.4||1.6||1.6||2.0||1.9||2.5||2.8|
|- Gastritis and duodenitis||0.7||0.7||0.4||0.4||0.4||0.5||0.6||0.5||0.2||0.2||0.1||0.1||0.2||0.2|
|- Abdominal hernia||3.9||1.8||1.3||1.0||1.1||1.3||1.4||1.8||1.7||1.4||1.3||1.4||1.9||1.7|
|- Gallstone disease and holecistitis||3.8||4.6||3.7||3.4||3.1||3.1||3.0||3.4||3.4||3.3||3.1||3.0||2.8||2.8|
|Diseases of genito-urinary system||21.6||20.6||21.8||20.6||20.2||21.0||20.7||6.6||7.5||8.6||9.0||8.9||9.2||9.2|
|Cancer of skin||1.6||2.2||2.3||2.8||2.7||2.9||2.9||1.1||1.5||1.6||1.9||1.9||2.1||1.9|
|Cancer of uterus||-||-||-||-||-||-||-||21.1||14.7||13.8||13.7||13.7||13.9||14.3|
|Complications of pregnancy and childbirth||-||-||-||-||-||-||-||3.8||2.0||1.8||1.2||1.2||1.0||1.0|
Obviously, a considerable deterioration occurred in 1992-1993. For many causes of death an increase in SDR is especially sharp in 1993. These are: tuberculosis, dysentery, pneumonia, bronchitis and lung emphysema, gastric ulcer, rheumatism, and hypertensive disease. Between 1991 and 1993, the SDR for tuberculosis increased by 51 percent among males and by 28 percent among females. For dysentery the rise was 33 percent and 8 percent; for pneumonia--99 percent and 39 percent; for bronchitis and lung emphysema--23 percent and 9 percent; for gastric ulcer--47 percent and 47 percent; for rheumatism--15 percent and 7 percent; for diabetes mellitus--29 percent and 23 percent; and for hypertensive disease--22 percent and 21 percent, respectively. As a result, mortality from many avoidable causes which were previously declining in 1993 returned to the levels characteristic of the early 1980s.
Observed levels of mortality in 1993 and expectations. Striking shifts in mortality of the Russian population indicate a transition from gradual deterioration in public health to a new and particularly acute stage of epidemiological crisis in Russia. The above consideration of mortality trends in Russia gives a clear impression that the level of mortality in 1993 lies substantially above the values expected from the long-term trends. To evaluate the magnitude of this difference we developed two simple scenarios of expected changes in age-specific death rates. The first corresponds to a linear continuation of trends observed in 1970-1984 (before the anti-alcohol campaign). The second scenario is based on the same data, but it describes each ASDR as A*(t)B, where t is a calendar year and A and B are parameters to be estimated. This function has an upper limit and therefore corresponds to mortality stabilization (in fact, a very slow increase) instead of an unlimited rise.
The first scenario leads to a predicted male life expectancy of 60.4 in 1993 and female life expectancy of 72.8, while the second scenario results in higher values--61.4 and 73.1, respectively. This means that, by extrapolating according to these scenarios, actual life expectancy in 1993 is lower than the predicted values by 1.4-2.4 years for males and by 0.8-1.1 years for females, and therefore a large part of the mortality increase in 1993 cannot be ascribed to long-term mortality trends. (A similar point was shown in Figure 4.6.) The deviations of observed life expectancies from their expected values are likely to be valid indicators of the real strength of the recent deterioration in the health of the Russian population.
Some Speculations on Factors Contributing to the Russian Epidemiological Crisis of the 1990s
The above analysis shows that the general lowering of Russian life expectancy at birth during the last three decades is mostly explained by a long-term increase in premature mortality from cardiovascular diseases and mortality from injuries and violence at ages 15-64. A rapid and especially sharp increase in cardiovascular mortality, mortality from injuries and violence, infectious and respiratory diseases, and some other causes of death occurred in 1992-93. Although a substantial part could be ascribed to restoration of the long-term mortality trends, another substantial part cannot be explained in such a way.
Michael Specter, in his recent article for the New York Times on Russian mortality increase, pays particular attention to environmental pollution. It is true that cancer mortality (especially the death rate from lung cancer) has been gradually increasing in Russia since the early 1980s. However, as we have seen, the impact of this process, as well as its influence on the overall life expectancy trend, is relatively small.
Certainly, new negative factors linked to the recent political and socio-economic transformations in Russian society may provide some explanation. Below we discuss what seem to be the more important of these factors.
First of all, the socio-economic transformations of the 1990s led to a general failure of Soviet state paternalism. Theoretically, the Soviet state had given an unprecedented set of social guarantees to the Russian people, including state promises of guaranteed housing, pensions, cheap public transport, cheap food, free medical care, etc. The Soviet economy and Soviet social policies were never really effective and only relatively poor standards of goods and services were provided. Nevertheless, the former situation was rather stable and people had always been sure of at least maintaining their quality of life standards in the foreseeable future. In the 1990s, the situation changed radically. Suddenly, the Russian people realized that the state would not be able to keep the former guarantees any longer and that they would have to take care of themselves. This created severe social stress. The situation has become particularly complicated for people aged 45-50. Many of them felt that it was too late or too difficult to change profession or occupation, and that it was impossible to earn enough money to maintain their former standards of living in their old profession.
A low value of individual life and individual health was characteristic of the former Soviet society. The priority of state aims and interests over personal needs and wishes taught people that their individual values were of minor importance. According to this ideology, there was no reason to pay much attention to one's future health. Many people believed that the state would help them in case of a serious health problem or any other disaster. Their resulting careless lifestyle has become especially dangerous under the new circumstances, when the general weakness of the Russian state has made its social and health efforts even more inadequate than in previous years.
Serious changes in living standards occurred in Russia in 1991-1992 (Figure 4.14). Real per capita income, which was growing in the 1970s and 1980s, fell sharply by 40 percent and returned to the level of the early 1970s. However, it is possible that the living conditions of certain population groups are currently much worse than in the early 1970s because the distribution of wages in the population is now much less uniform. In 1993, the difference between the highest and lowest income deciles was 11-fold and about a quarter of the total population was below the official poverty level. In the 1980s, it was a five-fold difference.
The registered decrease in wages for the Russian population in 1991-93 was very sharp. However, we have to be cautious about the quality of income statistics because, for many reasons, people tend to underreport their true incomes. It has been impossible to know how much money is passing from hand to hand, thus avoiding any type of registration.
In the 1990s, state food subsidies for meat and other products were removed and food markets were privatized. This led to a rapid rise in prices and a considerable decrease in daily protein intake by 23 percent in 1991-1992, according to Goskomstat. Total energy intake from food decreased, as well. The poorest part of the Russian population is likely to have significant nutritional deficiencies and particularly a shortage of vitamins.
In spite of shifts in nutrition, the Russian population generally continues to consume a high-protein and high-fat diet, and under-nutrition is clearly not a major health problem. Moreover, a survey of the nutrition of the Russian population, conducted by Barry Popkin, showed that the rate of obesity is high in Russia and that the percentage of adults who gained weight in 1992-93 is higher than the percentage of those who lost weight.
Slight decreases in the numbers of physicians and hospital beds (Figure 4.14) are not of principal importance because the general level of these indicators continues to be very high. For instance, the number of physicians per 10,000 inhabitants in Russia was about 45 in 1993, whereas it was 25 in Sweden, 28 in Denmark, 31 in Norway, 30 in France, 26 in Finland and 17 in Great Britain (in 1988-1992). However, it is possible that in many regions of Russia there is a lack of highly qualified specialized medical care. The institutions providing this type of medical services are traditionally located in Moscow, St. Petersburg, and several other big cities of Russia. Rapid growth in transportation prices and, in many cases, the necessity to pay for medical treatment leads to restricted access to advanced medical care for provincial residents.
The rapid decrease in the 1990s in the already low salaries for physicians and nurses constitutes a much more serious problem. The average salary in this sector was always about 70-75 percent of the general average for all branches of the economy combined. This situation contrasts sharply with the United States and other western countries where physicians' salaries are very high. It is obvious, however, that the 75 percent in the 1980s was much higher in real terms than the 75 percent in the 1990s, given the sharp reduction in the overall average real wage rate. The share of total budget outlays going to medicine, physical culture, and sport declined from 12.4 percent in 1970 to 9.4 percent in 1993. Numerous underpaid medical personnel are working in poorly equipped and poorly organized clinics that are receiving insufficient funds even for the continuation of normal functioning. Generally, the Russian medical care system is trying to survive in new conditions and is currently not able to undertake any significant modifications or improvements.
The growth in deaths from avoidable causes reflects a general weakening of Russian medical services. For instance, the return of diphtheria, a disease which had been virtually eliminated by the 1970s, is very illustrative. The number of deaths due to diphtheria was 10 in 1970 and 1981, 32 in 1989, 124 in 1992, and 482 in 1993. It seems that in the 1990s, even some basic medical and sanitary needs are not being treated in Russia.
The strong influence of excess alcohol consumption on Russian mortality has been already discussed in this paper. A great fluctuation in estimated real alcohol consumption (Figure 4.14) was caused by the anti-alcohol campaign of 1985-87. In 1992-93, a growth in real alcohol consumption probably was accelerated by social stress and the relatively slow increase in the prices of alcoholic beverages. Between December 1990 and December 1994 consumer prices increased by 2,020 times for all goods and services, by 2,154 times for food products, but only by 653 times for alcoholic beverages. This means that over this period, in relative terms, alcohol became over three times cheaper than these other products.
SOURCES: Real Per Capita Income and Daily Protein Intake provided by the Laboratory for Prognosis of Population Wages and Consumption, Institute for Economic Forecasting, Moscow; Physicians and Hospital Beds Per 10,000 Population from "Sources of Statistical Data," given in this paper; Real Alchohol Consumption, by Shkolnikov, Nemtsov, 1995, and evaluated by A. Nemtsov.
Figure 4.14--Real Per Capita Income, Daily Protein Intake, Number of Physicians and Hospital Beds per 10,000, and Real Alcohol Consumption, Each Compared to the Basic Level of 1970: Russia, 1970-1994
Growing criminality has brought a sharp rise in mortality from violence, now one of the highest in the world. In 1993, Russia's SDR for homicide was perhaps only exceeded by that for Colombia and the black male population of the United States. Between 1985 and 1993, the number of recorded crimes increased by 1.9 times, and the number of felony homicides increased by 2.4 times. Growing alcoholism, general weakness of the police and judicial system, political instability, and armed ethnic conflicts are the main factors behind the criminal explosion in the 1990s.
As such, a large part of the striking rise in mortality in the 1990s can be ascribed to the effects of a number of serious shifts in Russian society. All of them are closely interrelated. It is possible, however, to extract the two principal factors among them. These are: 1) lower living standards; and 2) social disorganization. We cannot discuss the links between these two here. However, we can say that the negative changes in living conditions themselves are not as bad as to lead to the massive deterioration in public health evident in the shifts in mortality by age and cause of death in 1992-93. In many countries, where living standards are much worse than in Russia--even in some developing countries--male life expectancy is significantly higher. We can suppose that some kind of complex interaction effect, coming from both principal factors, is responsible for the deteriorating health situation. In that case, the negative influence of the first component (real living conditions) is reinforced by the second group of factors (socio-psychological conditions) and vice versa.
1. Adamets, Sergei, and Vladimir Shkolnikov, "On Mortality Tables of the Population of the USSR, Russia, Ukraine and Byelorussia to the End of the 1930s," paper presented at the workshop, Population of the USSR in the 1920s in Light of Newly-Declassified Documentary Evidence, University of Toronto, January 27-29, 1994.
2. Anderson, Barbara, and Brian Silver, "Infant Mortality in the Soviet Union: Regional Differences and Measurement Issues," Population and Development Review, 12, 4, 1986, pp. 705-738.
3. Anderson, Barbara, and Brian Silver, "Trends in Mortality of the Soviet Population," Soviet Economy, Vol. 6, No. 3, 1990, pp. 191-251.
4. Andreev, Evgueny, "Metod komponent v analize prodoljitelnosti zhizni," [Method of the components in the life expectancy analysis]. Vestnik Statistiki, No. 3, March, 1992, pp. 42-47.
5. Andreev, Evgueny, "Prodoljitelnost' jizni i prichini smerti v SSSR." [Life expectancy and causes of death in the USSR], in A.G.Volkov, ed., Demograficheskiye protsessi v SSSR [Demographic Processes in the USSR], Moscow, Finansi I Statistika, 1990.
6. Blum, Alain, and Alain Monnier, "Recent Mortality Trends in the USSR: New Evidence," Population Studies, Vol. 43, No. 2, 1989, pp. 211-241.
7. Bourgeois-Pichat, Jean, "Mortality Trends in the Industrialized Countries," in Mortality and Health Policy, New York: United Nations, 1984.
8. Coale, Ansley J., and Paul Demeny, Regional Model Life Tables and Stable Populations, Princeton University Press, 1966.
9. Field, Mark, "The Health Crisis in the Former Soviet Union: A Report From the `Post-War' Zone," paper presented at the XIII International Conference on the Social Science and Medicine., Lake Balatone, Hungary, October 10-14, 1994.
10. Holland W.W., ed., European Community Atlas of Avoidable Death, Oxford University Press, Oxford, 1988.
11. Indirect Techniques for Demographic Estimation, Manual X, New York: United Nations, 1983.
12. Jozan, Peter, "Contrast in Mortality Trends," paper presented at the IUSSP General Conference, New Delhi, 1989.
13. Ksenofontova, Natalya Yu, "Nekotoryye tendentsii mladencheskoy smertnosti v posledneye desyatiletiye," [Some trends in infant mortality in the last decade] in: A.G.Volkov, ed., Demograficheskiye protsessi v SSSR [Demographic Processes in the USSR]. Moscow, Finansi I Statistika, 1990.
14. Meslé, France, "Mortality in Eastern and Western Europe: A Widening Gap," paper presented at the meeting of the British Society for Demographic Studies, London, April, 1993.
15. Meslé, France, Vladimir Shkolnikov, and Jacques Vallin, "Mortality by Cause in the USSR in 1970-1987: The Reconstruction of Time Series," European Journal of Population 8, 1992, pp. 281-308.
16. Meslé, France, Vladimir Shkolnikov, and Jacques Vallin, "Brusque Montée des Morts Violentes en Russie," Population, 3, 1994, pp. 780-790.
17. Meslé, France, Vladimir Shkolnikov, and Jacques Vallin, "DéveloppementÉconomique et Espérance de Vie: La Transition Sanitaire au Tournant des Années Soixante," paper presented at the General Congress of IUSSP, Montréal, August 24-September 1, 1993.
18. Novoselski (Sergei), Obzor glavneishih dannih po demografii i sanitarnoi statistiki Rossii. In, Kalendar vracha za 1916 god. Ch. II. [Review of the Principal Data in Demography and Sanitary Statistics of Russia. In Calendar of Physician for 1916, Part II, 1916, pp. 66-67.
19. Okolski, Marek, "East-West Mortality Differentials," in Alain Blum and J.L.Rallu, eds., European Population, Vol. 2, 1993, "Demographic Dynamics," John Libbey Ltd. (Materials of European Population Conference, Paris, October 21-25 1991.)
20. Poikolinen K., and J. Eskola, "The Effect of Health Services on Mortality: Decline in Death Rates from Amenable and Non-amenable Causes in Finland, 1969-81," Lancet, I, 1986, pp. 199-202.
21. Pollard, John, "Cause of Death and Expectation of Life: Some International Comparisons," in Vallin, Jacques, Stan D'Souza, and Alberto Palloni, eds., Measurement and Analysis of Mortality: New Approaches, pp. 291-313, Liege-Oxford: Oxford University Press, 1990.
22. Popkin, Barry, Lenore Kohlmeier, Namvar Zohoori, et al, "Nutritional Risk Factors in the Former Soviet Union," paper presented at the workshop, "Mortality and Adult Health Priorities in the New Independent States," Washington, D.C., November 17-18, 1994, forthcoming in Bobadilla, J.L., and C.A.Costello, eds., Mortality Patterns in the New Independent States and Adult Health Interventions, National Academy Press.
23. Pressat, Roland, "Contribution desÉcarts de Mortalité par ge àla Différence des Vies Moyennes," Population, Vol. 40, No. 4-5, 1985, pp. 766-770.
24. Preston, Samuel H., Nathan Keyfitz, and R. Schoen, Causes of Death: Life Tables for National Populations, New York-London: Seminar Press, 1972.
25. Ptukha, Mikhail, Otcherki po statistike naseleniya. (Essays on the Statistics of Population), Moscow, TsSU SSSR, 1960.
26. Rozenfeld, Boris, "The Crisis of Russian Health Care and Attempts of its Transformation," paper presented at the Workshop: Russia's Demographic Crisis in Comparative Perspective, June 5-6, 1995, RAND, Santa Monica, Calif. (chapter in this volume).
27. Rutstein, D.D., et al, "Measuring the Quality of Medical Care," New England Journal of Medicine, Vol. 11, 1976, pp. 582-588.
28. Shkolnikov, Vladimir M., France Meslé, and Jacques Vallin, "Recent Trends in Life Expectancy and Causes of Death in Russia (1970-1993)," in Bobadilla, J.L., and C.A.Costello, eds., Mortality Patterns in the New Independent States and Adult Health Interventions, National Academy Press (forthcoming).
29. Shkolnikov, Vladimir M., France Meslé, and Jacques Vallin, "La Crise Sanitaire en Russie (1970-1993), Population, No. 4-5, 1995, pp. 907-982.
30. Shkolnikov, Vladimir, and Alexander Nemtsov, "The Anti-alcohol Campaign and Variations in Russian Mortality," in Bobadilla, J.L., and C.A.Costello, eds., Mortality Patterns in the New Independent States and Adult Health Interventions, National Academy Press (forthcoming).
31. Shkolnikov, Vladimir, and Sergei Vassin, "Spatial Differences in Life Expectancy in European Russia in the 1980s," in Lutz, W., A. Volkov, and S. Scherbov, eds., Demographic Trends and Patterns in the Soviet Union Before 1991, New York-London: Routledge-IIASA, 1994, pp. 379-402.
32. Specter, Michael, "Climb in Russia's Death Rate Sets Off Population Implosion," New York Times, March 6, 1994.
33. Specter, Michael, "Plunging Life Expectancy Puzzles Russia," New York Times, August 2, 1995.
34. Vishnevsky, Anatoly, "Family, Fertility and Demographic Dynamics in Russia: Analysis and Forecast," paper presented at the workshop: Russia's Demographic Crisis in Comparative Perspective, June 5-6, 1995, RAND, Santa Monica, Calif. (chapter in this volume).
Sources of Statistical Data
1. Meslé, France, Vladimir M. Shkolnikov, Veronique Hertrich, et Jacques Vallin, "Tendances Récentes de la Mortalité par Cause en Russie 1965-1993," Paris: INED, forthcoming.
1. Narodnoye khozyastvo SSSR 1922-1982: Yubileynyy statisticheskiy yezhegodnik (The Economy of the USSR 1922-1982: Anniversary Statisticakl Yearbook). Moscow: Statistika, 1982.
2. Narodnoye khozyastvo SSSR za 70 let: Yubileynyy statisticheskiy yezhegodnik (The Economy of the USSR During the Last 70 Years: Anniversary Statisticakl Yearbook). Moscow: Statistika, 1987.
3. Narodnoye khozyastvo RSFSR v 1989 g. Statisticheskiy yezhegodnik (The Economy of the Russian Federation in 1989. Statistical Yearbook). Moscow, Goskomstat of the Russian Federation, 1990.
4. Rossiyskaya Federatsiya v 1992 Godu. Statisticheskiy yezhegodnik (The Russian Federation in 1992. Statistical Yearbook). Moscow, Goskomstat of the Russian Federation, 1993.
5. Standards of Living in the Russian Federation. Goskomstat of the Russian Federation, 1993.
Discussants: Elizabeth Frankenberg, RAND; and Ward Kingkade, U.S. Census Bureau
Unlike the fertility debate of the previous day, the discussion of Dr. Shkolnikov's paper showed little disagreement with the characterization of the mortality situation in Russia as an obvious and serious crisis. However, the discussants pointed out that further research is needed to understand regional and socio-economic differentials in mortality trends, especially with respect to the effects of such policies as the anti-alcohol campaign. The lessons of this campaign's initial success and eventual failure need to be closely studied in designing future policy measures, utilizing the detailed morbidity and mortality data that are now available. Specifically, more research is needed (and is possible) on such questions as regional and socio-economic differentials of the anti-alcohol campaign's effects. The recent mortality rise is undoubtedly linked to the degradation of the entire health care system, but the rapid criminalization and "alcoholization" of the country are the most important factors behind the dramatic increase in death rates, especially among men of working age. Thus the solution for the mortality problem should be sought not only in improving public health but also in strengthening public order.
 Dr. Vladimir M. Shkolnikov is Head of the Laboratory of Analysis and Forecasting of Mortality at the Center for Demography and Human Ecology, Institute of Economic Forecasting, Russian Academy of Sciences. Dr. France Meslé is Director of Research at L'institut National D'etudes Démographique (INED) in Paris, France.
 Meslé, France, and Jacques Vallin, "DéveloppementÉconomique et Espérance de Vie: La Transition Sanitaire au Tournant des Années Soixante," paper presented at the General Congress of IUSSP, Montréal, Canada, August 24-September 1, 1993.
 Meslé, France, "Mortality in Eastern and Western Europe: A Widening Gap," in Coleman, David, ed., European Population in the 1990s, Oxford University Press, 1995.
 Bourgeois-Pichat, Jean, "Mortality Trends in the Industrialized Countries," in Mortality and Health Policy, New York: United Nations, 1984.
Jozan, Peter, "Contrast in Mortality Trends," IUSSP-General Conference, New Delhi, 1989.
Okolski, Marek, "East-West Mortality Differentials," in Alain Blum and J.L.Rallu, eds., "Demographic Dynamics," European Population, Vol. 2, materials of European Population Conference, Paris, October 21-25, 1991, John Libbey Ltd, 1993.
 Anderson, Barbara, and Brian Silver, "Trends in Mortality of the Soviet Population," Soviet Economy, Vol. 6, No. 3, pp. 191-251, 1990. Andreev, Evgueny, Prodoljitelnost' jizni i prichini smerti v SSSR (Life Expectancy and Causes of Death in the USSR), in A.G.Volkov, ed., Demograficheskiye protsessi v SSSR (Demographic Processes in the USSR), Moscow, Finansi I Statistika, 1990. Blum, Alain, and Monnier, Alain, "Recent Mortality Trends in the USSR: New Evidence," Population Studies, Vol. 43, No. 2, 1989, pp. 211-241.
 Meslé, France, Vladimir Shkolnikov, and Jacques Vallin, "Mortality by Cause in the USSR in 1970-1987: The Reconstruction of Time Series," European Journal of Population, No. 8, 1992, pp. 281-308.
 For several years, very little data on causes of death were published in the former USSR. In the 1960s and the early 1970s only some aggregated data on cardiovascular and cancer mortality at the level of the whole USSR were published either in journal Vestnik Statistiki or in the statistical yearbooks Narodnoye Khozyaistvo SSSR and Naseleniye SSSR. During 1974-1986, the Soviet government forbid any publication on mortality and causes of death. Since 1988, a completely new statistical era arose (due to Gorbachev's Glasnost), and Goskomstat started to produce the statistical annuals Naseleniye SSSR (Naseleniye Rossii since 1992) containing only information about the large classes of causes of death (infectious diseases, neoplasms, circulatory diseases, respiratory diseases, and external causes of death).
 Meslé, France, Vladimir Shkolnikov, and Jacques Vallin, "Brusque Montée des Morts Violentes en Russie," Population, 3, 1994, pp. 780-790.
 Shkolnikov, Vladimir, France Meslé, and Jacques Vallin, "Recent Trends in Life Expectancy and Causes of Death in Russia (1970-1993)," in J.L. Bobadilla and C.A. Costello, eds., Mortality Patterns in the New Independent States and Adult Health Interventions, National Academy Press (forthcoming). Shkolnikov, Vladimir, France Meslé, and Jacques Vallin, "La Crise Sanitaire en Russie, 1970-1993," Population, No. 4-5, 1995, pp. 907-982.
 We have not yet studied the period before 1970 because the reconstructed cause-specific mortality rates for 1965-1969 became available only recently.
 Here we use the corrected (for underreporting of deaths in infancy and older ages) values of life expectancy at birth for Russia (Shkolnikov, Meslé, Vallin, 1994). The official Goskomstat estimates of life expectancy at birth for 1965-66 are 64.3 for males and 73.7 for females. Accordingly, the difference with the United States becomes even smaller--2.5 years for males and 0.3 years for females.
 Adamets, Sergei, and Vladimir Shkolnikov, "On Mortality Tables of the Population of the USSR, Russia, Ukraine, and Byelorussia to the End of the 1930s," paper presented at the workshop on "Population of the USSR in the 1920s in Light of Newly-Declassified Documentary Evidence," University of Toronto, January 27-29, 1994.
Correction for underestimation of death rates in infancy and older ages of Novoselski-Paevski's life tables for 1926-27 diminishes male life expectancy by 0.7 years, female life expectancy by 0.6 years.
 Coale and Demeny regional model life tables (MLT), Coale, Ansley J., and Paul Demeny, "Regional Model Life Tables and Stable Populations," Princeton Univ. Press, 1966, are used here. They were derived from a set of 192 life tables by sex recorded for actual populations. Therefore MLT provides a kind of aggregation of mortality experiences from different countries and different time periods. Four patterns of mortality were distinguished and were labelled "North," "South," "East," and "West" MLT family patterns. (Indirect Techniques for Demographic Estimation. Manual X, United Nations, New York, 1983).
 Novoselski, Sergei, 1916. "Obzor glavneishih dannih po demografii i sanitarnoi statistiki Rossii." In: Kalendar vracha za 1916 god. ch. II. [Review of the Principal Data in Demography and Sanitary Statistics of Russia. In: Physician's Calendar for 1916. Part II]. pp. 66-67.
Ptukha (Mikhail) 1960. Otcherki po statistike naseleniya. [Essays on the Statistics of Population] . Moscow, TsS USSSR.
 Source of mortality data by cause-of-death for the United States: Preston, Samuel H., Nathan Keyfitz, and R. Schoen, Causes of Death. Life Tables for National Populations. Seminar Press, New York-London, 1972.
 The method for decomposition of time series into components and the corresponding computer program were developed by Alexander Kovaldjy (Institute for Economic Forecasting, Moscow). These components are: (1) linear trend; (2) seasonal wave which is a regular periodic oscillation within a period of one year; and (3) irregular fluctuations, which are the differences between the observed values and the influence of components (1) and (2).
 See, Adamets, Shkolnikov, 1995.
 IMR values for 1940-1958 are collected and reconstructed by Sergei Zakharov (Center for Demography and Human Ecology, Moscow).
 Specialists tend to ascribe the increase in IMR in 1972-76 to changes in the registration of perinatal deaths. For more details, see, Anderson, Barbara, and Brian Silver, "Infant Mortality in the Soviet Union: Regional Differences and Measurement Issues," Population and Development Review, 12, 4, 1986, pp. 705-738; and Blum, Alain, and Alain Monnier, "Recent Mortality Trends in the USSR: New Evidence," Population Studies, Vol. 43, No. 2, 1989, pp. 211-241.
 We compare death rates of Russian males with that from MLT level 20. Life expectancy at birth for level 20 is equal to 63.6 in family West and 63.7 in family South. These values do not differ much from observed life expectancies in Russia in the 1960s-1980s.
 Andreev, Evgeny, 1982. Metod komponent v analize prodoljitelnosti zhizni. ["The component method in life expectancy analysis"]. Vestnik Statistiki No 3, March, 1985 pp. 42-47. Pressat, Roland. "Contribution des écarts de mortalité parâge àla différence des vies moyennes." Population, Vol. 40, No. 4-5, 1990, pp. 766-770. Pollard, John. "Cause of death and expectation of life: some international comparisons." In, Measurement and Analysis of Mortality: New Approaches, Vallin, Jacques, Stan D'Souza, Alberto Palloni, eds., pp. 291-313. Liege-Oxford, Oxford University Press.
 Here and below the "old" WHO European population standard (before its revision in 1992) is used. SDR values do not depend on the age structure of populations under study.
 See Shkolnikov, Meslé, Vallin, 1995a and 1995b.
 Sources of mortality data for EC and CEE: "Health for All," Eurostat PC, the personal computer system for data collecting and analysis. WHO, Regional Office for Europe, 1993.
 See Shkolnikov, Meslé, Vallin, 1995.
 Shkolnikov, Vladimir, and Alexander Nemtsov, "The Anti-Alcohol Campaign and Variations in Russian Mortality," in Bobadilla, J.L. and C.A.Costello, eds., Mortality Patterns in the New Independent States and Adult Health Interventions, National Academy Press (forthcoming).
 Shkolnikov, Vladimir, and Sergei Vassin, "Spatial Differences in Life Expectancy in European Russia in the 1980s," in Lutz, W, A.Volkov, and S. Scherbov, eds., Demographic Trends and Patterns in the Soviet Union Before 1991, Routledge-IIASA, New York-London: 1994, pp. 379-402.
 See Meslé, Shkolnikov, Vallin, 1994.
 See paper by A. Vishnevsky in this volume.
 The rules for registration of life births were changed on January 1 of 1993 according to a directive released by the Health Care Ministry of the Russian Federation and the Goskomstat on December 4, 1992. This reformation could effect some increase in early infant deaths due to more complete registration of live births. The old Soviet definition of life birth was more restrictive than the WHO definition. According to the old Soviet practice, children born before 28 weeks of gestation or whose weight is less than 1,000 grams or length is less than 35 centimeters were supposed not to be counted as either live births or infant deaths if they die before the end of their first week of life. For more details about the old Soviet system for registration of live births, see, Anderson, Silver, 1986. The new regulations state that live births in Russia are to be recorded at hospitals according to the WHO definition of a live birth. It is not clear, however, how much this change affected the practice of civil registration of births and infant deaths.
 Rutstein D.D., et al, "Measuring the Quality of Medical Care," New England Journal of Medicine, Vol. 11, 1976, pp.582-588. Poikolinen K., J. Eskola, "The Effect of Health Services on Mortality: Decline in Death Rates from Amenable and Non-amenable Causes in Finland," 1969-81. Lancet, I, Holland, W.W.,ed., European Community Atlas of Avoidable Death. Oxford University Press, Oxford, 1988, pp. 199-202.
 Specter, Michael, "Plunging Life Expectancy Puzzles Russia," New York Times, August 2, 1995.
 Field, Mark, "The Health Crisis in the Former Soviet Union: A Report from the `Post-War Zone,'" paper presented at the XIII International Conference on the Social Science and Medicine, Lake Balatone, Hungary, October 10-14, 1994. Michael Specter describes this situation as following: "The Soviets demanded that men in Russia sacrifice their lives for Communism. Nobody put the cost of life before the cost of building . . . society . . . we were taught to suffer, and we are taught that we will probably die in the next war. In that event, why worry about how you are going to survive to an old age?" Specter, Michael, "Climb in Russia's Death Rate Sets Off Population Implosion," New York Times, March 6, 1994.
 Popkin, Barry, Lenore Kohlmeier, Namvar Zohoori, et al, "Nutritional Risk Factors in the Former Soviet Union," paper presented at the workshop, "Mortality and Adult Health Priorities in the New Independent States," Washington, D.C., November 17-18, 1994, forthcoming in Bobadilla, J.L., and C.A.Costello, eds., Mortality Patterns in the New Independent States and Adult Health Interventions, National Academy Press.
 The system of obligatory medical insurance was introduced at the beginning of 1993. Until now it has not improved the financial situation in majority of medical care institutions. See paper by Boris Rozenfeld in this volume.
 See Shkolnikov, Nemtsov, 1995.