Social factors and increase in mortality in Russia in the 1990s: prospective cohort study

BMJ 2003; 326 doi: (Published 05 June 2003) Cite this as: BMJ 2003;326:1240
  1. S L Plavinski (splavinskij{at}, dean1,
  2. S I Plavinskaya, leading researcher2,
  3. A N Klimov, principal researcher2
  1. 1 College of Public Health, Medical Academy for Postgraduate Studies, Saint Petersburg, Russia,
  2. 2 Department of Biochemistry, Institute for Experimental Medicine, Saint Petersburg, Russia
  1. Correspondence to: S L Plavinski
  • Accepted 7 April 2003


Objective To determine the association between social factors and the increase in mortality in Russia in the 1990s.

Design Prospective population cohort study.

Setting Saint Petersburg, Russia.

Participants Two cohorts of men aged 40–59 years randomly selected from district voting list: 3907 screened in 1975–7 and 1467 in 1986-8.

Main outcome measures Education, various health related measures, alcohol intake. Mortality in subsequent 10 years.

Results There was no recorded increase in mortality in men with university degrees. The relative risk in the second cohort compared with the first was 0.92 (95% confidence interval 0.67 to 1.24). For participants with only high school education it was significantly higher in the second cohort (1.32, 1.02 to 1.71). The most pronounced differences were found among participants with the lowest level of education, in which the relative risk was 1.75 (1.44 to 2.12). The same pattern held for coronary vascular disease and cancer mortality.

Conclusion In Russia men in the lower socioeconomic groups were most affected by the sharp increases in mortality in the 1990s.


The rapid increase in mortality in Russia in the 1990s was a development previously not reported in any economically developed country. For men life expectancy decreased from about 64 years in 1990 to 59 years in 1993.1 There is still controversy as to who suffered most from concurrent economic breakdown. Several large prospective studies on mortality from ischaemic heart disease in Russia offer a unique opportunity to identify subpopulations that were more affected.


The study sample comprised two cohorts of men living in the Petrogradsky district of Leningrad (now Saint Petersburg). The first cohort was 5000 randomly selected men from the 1974 voting list who were born from 1916 to 1935. The response rate was 78% (3907 men were screened). The second cohort was selected from the 1985 voting list (men born from 1927 to 1946). In total 1000 men aged 40–49 and 1000 aged 50–59 were randomly selected for screening. The response rate was 71% for 40–49 year olds and 76% for 50–59 year olds (total 1467 men screened). The screening procedure has been described in detail elsewhere24 and was the same for the two groups.

The follow up study began in January 1979. If the state registration organ (ZAGS) indicated that participants were no longer registered at the designated addresses we tried to contact them if they had moved away or contacted their relatives or neighbours if they had died. Overall loss to follow up was 3%. In the second cohort we were unable to get data for 15 men who had died.

The first cohort was followed up for a mean of 18.1 years, and there were 1890 deaths. The second cohort was followed up for a mean of 11.2 years, and there were 323 deaths. The first day of follow up was the day the last participant from this cohort was screened: 6 July 1977 for the first cohort and 21 August 1988 for the second cohort (see webextra figure). We have presented data for the first 10 years of follow up.

We used Kaplan-Meier survival curves and calculated relative risks and confidence intervals from person time data. We used the exact Poisson method to calculate confidence intervals.5 Statistical analysis was performed with SAS system, version 6.12 for Windows (SAS Institutes, Cary, NC).


The two groups screened closely resembled each other in terms of age, place of residence, structure, behavioural risk factors, and biological indicators (table 1). The number of people with education less than high school decreased because the second cohort included only those born after the Bolshevik revolution, when high school education became mandatory.

Table 1

General characteristic of two screened groups of men. Figures are means (SD) unless stated otherwise

View this table:

Table 2 and the figure show our main results. Among participants with university education all cause mortality in both cohorts was almost the same: 12.8 per 1000 person years of observation for the first cohort and 11.7 for the second (relative risk 0.92, 95% confidence interval 0.67 to 1.24). Mortality among men with only high school education was the same in both cohorts for the first six years of follow up (figure), but then diverged and at the end of follow up was significantly higher in the second cohort (1.32, 1.02 to 1.71). The largest differences were in men with the least education (less than high school). All cause mortality was higher almost from the beginning of the follow up, steadily rising with time and reaching 48.6 per 1000 person years for the second cohort at the end of follow up. Ten years before it was 43% lower (27.8 per 1000 person years; 1.75, 1.44 to 2.12). The increase in mortality has been noted for cardiovascular disease and cancer, though it was not significant for coronary heart disease. Increases in mortality from cardiovascular disease and cancer were highest among men with the least education (1.99, 1.49 to 2.63; and 1.78, 1.20 to 2.58, respectively). Surprisingly, there was no significant increase in the rate of accidental/violent deaths.

Table 2

Mortality in two cohorts according to cause of death during 10 years of follow up

View this table:

All cause mortality among two cohorts according to level of education (group 1 screened in 1975-7, group 2 screened in 1986-8)

What is already known on this topic

In the mid-1990s the former Soviet Union experienced unprecedented increases in mortality especially among men of working age and in the urban population

Increase in mortality associated with lower educational attainment has been noted in Russia

What this study adds

The increase in mortality in Russia did not affect all social groups equally

The greatest increases were among those with the least education

Alcohol consumption may account for some but not all of this increase

We divided participants into two groups: those who drank more than 150 g of alcohol during the week before screening and those who drank less. In both groups mortality increased in the 1990s (see webextra table A). This increase was more pronounced than the increase in mortality associated with increased alcohol consumption within each cohort. Among men with the least education, mortality in the 1990s increased 60-80% compared with 8-22% increase associated with increased alcohol consumption (the relative risk for alcohol related mortality was 1.27, 1.02 to 1.57, in the first cohort and 1.08, 0.73 to 1.57, in the second cohort).


In 1980 the Black report first presented evidence to show the links between the socioeconomic environment and health and wellbeing.6 In the United Kingdom between 1976–81 and 1986–92 the mortality gap between socioeconomic classes I and II and classes IV and V increased.7 The higher mortality among people with little education has also been well documented in Russian epidemiological studies.3 8

Our data show that the increase in mortality in the mid-1990s disproportionately affected men with low education, with an increase of almost 75% from the level of the mid-1980s. The reason why social changes have struck mostly the least educated could be that for those sectors of society, breakdown of the socialist state could engender a sense of catastrophe.9 The mediator between stress and mortality could be alcohol. Russian men commonly use alcohol because it “helps them to forget everyday cares and difficulties.”10 Also alcohol related problems were more common among the less educated respondents. However, our analysis shows that the increase in mortality in Russia was socially determined, and though alcohol may play a part in this process it is not the sole factor.


  • Embedded ImageA figure showing details of recruitment and follow up and a table showing alcohol intake can be found on

    We thank all those who participated in data collection and follow up of participants: G Ilyina, A Katrushenko, V Khoptiar, I Klenina, V Konstantinov, B Lipovetsky, E Magracheva, T Maslova, G Mirer, N Muchina, N Nikulcheva, N Parfenova, D Shestov, Y Slepenkov, V Tryufanov, L Vassilieva, N Zhukovskaya.

  • Contributors SLP proposed the initial hypotheses, analysed the data. and wrote the paper. SIP participated in data collection and analysis, provided expert knowledge of the mortality classification and determinants of mortality, and provided ideas for study execution and analysis. ANK conceived the study, supervised and directed data collection, and is guarantor. All authors commented on the paper. ANK was the principal investigator in the larger project of which this study is part.

  • Funding Cohort study was funded partially by NIH grants NO-1HR12243-L/HR/NHLBI, NO-1HV08112/HV/NHLBI and by the Soviet (then Russian) Academy of Medical Sciences. The guarantor accepts full responsibility for the conduct of the study, had access to the data, and controlled the decision to publish.

  • Competing interests None declared.

  • Ethical approval The USSR Lipid Research Clinics cohort study has been carried out under a government-to-government agreement between the United States and the Union of Soviet Socialist Republics on a joint programme in cardiovascular diseases, signed in 1972.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
View Abstract