Background According to published data, between 1984 and 1994 mortalit
y rates in Russia initially underwent a rapid decline followed by an e
ven steeper increase. In 1994, male life expectancy at birth was 57.6
years, having fallen by 6.2 years since 1990, There has been concern t
hat such striking fluctuations in mortality are an artefact, although,
among other factors, alcohol consumption has been implicated. Methods
We analysed the age-specific and cause-specific patterns of mortality
decrease and increase by use of data from a newly reconstructed morta
lity series for Russia so that we could examine the plausibility of va
rious ex planations for the mortality trends. Findings All major cause
s of death, with the exception of neoplasms, showed declines in mortal
ity between 1984 and 1987 and increases between 1987 and 1994, In rela
tive terms, these tended to be largest for the age-group 40-50 years;
surprisingly, they were of the same magnitude among women and men, The
largest declines and subsequent increases in proportional terms were
observed for alcohol-related deaths and accidents and violence. Howeve
r, pronounced effects were also seen for deaths from infections, circu
latory disease, and respiratory disease. No substantial variations wer
e seen for neoplasms. Interpretation The stability of mortality from n
eoplasms in contrast to other causes over the period 1984-94 largely p
recludes the possibility that the changes in life expectancy are mainl
y an artefact, particularly one due to underestimation of the populati
on. Although factors such as nutrition and health services may be invo
lved, the evidence is that substantial changes in alcohol consumption
over the period could plausibly explain the main features of the morta
lity fluctuations observed, These results provide a major challenge to
public health in Russia and to our understanding of the determinants
of alcohol consumption and its role in explaining mortality patterns w
ithin and between many other countries.