The period from 1965 to 2020 in Russia was marked by unprecedentedly high alcohol-related and overall mortality, making it unique in the country’s demographic history. Despite the significance of this issue, alcohol-related causes of death and overall mortality trends have often been studied in isolation, without simultaneously considering their long-term dynamics. This study aims to comprehensively analyze alcohol-related and overall mortality trends over this period and identify key stages marked by changes in mortality patterns.

Using Rosstat data, standardized mortality rates for men and women due to alcohol poisoning and all causes were calculated for 1965–2020. Correlation and regression analyses were performed to examine the relationship between overall and alcohol-related mortality. The dynamics of these indicators were compared over the entire period and across individual stages.

The results revealed three complete cycles of synchronous fluctuations in alcohol and overall mortality in Russia: 1965–1988, 1988–1998, and 1998–2019, with peaks in 1980, 1994, and 2003, respectively. The historical maximum mortality rate occurred in 1994. Between 1965 and 1994, overall mortality increased by 47%, while alcohol poisoning mortality increased by 361%. By the early 2010s, both indicators had returned close to their 1965 levels.

Analysis indicated that each shift in mortality trends was associated with the emergence of new, often multiple, factors affecting alcohol consumption and mortality, specific to periods of growth or decline. A strong synchronicity between overall and alcohol-related mortality was observed (R = 0.84; 95% CI: 0.73–0.91), despite alcohol poisoning accounting for only 1.4% of all deaths. Two exceptions were noted: 1965–1980 among women, and 2005–2019 among individuals with severe alcohol dependence, reflecting specific drinking behaviors.

Four post-World War II anti-alcohol campaigns (1958, 1972, 1985, and 2000) were also analyzed. Only the 2000 campaign achieved a significant and sustained reduction in alcohol consumption and mortality. The synchronicity between alcohol-related and overall mortality in Russia indicates a substantial contribution of alcohol to excess mortality and population decline. Despite notable declines in alcohol consumption and overall mortality between 2005 and 2019, both indicators remain high. Russia continues to rank among the highest globally in alcohol-related mortality, showing disproportionately high mortality relative to alcohol consumption levels.

causes of death, standardized mortality rate, mortality dynamics, alcohol poisoning, alcohol consumption, alcohol behavior, anti-alcohol policy

There is a widespread perception that Russia’s severe alcohol problems have a centuries-old history, but this is only partially accurate and requires careful consideration of the phenomenon’s scope and historical context. On the eve of 1861, serfs, together with “state” peasants bound to land belonging to the treasury, constituted the overwhelming majority of the population. Due to their enslaved status, this population experienced minimal problems related to alcohol. Drunkenness and strong spirits were largely confined to the wealthy, the elite, and the court nobility. At the other extreme, the majority of habitual drinkers were townspeople, artisans, and merchants, although urban populations were relatively small, ranging from 4% to 12% at different times.

At the national level, Russia first encountered serious alcohol-related problems in the mid-19th century, following the emancipation of the peasants in 1861, the introduction of new technologies leading to cheaper alcohol production, and a change in alcohol policy with the replacement of tax farms by excise taxes in 1863. Alcohol-related issues intensified at the turn of the 19th and 20th centuries with the rise of capitalism and the mass migration of peasants to cities, turning them into industrial workers. However, even then, Russia did not rank among the top ten European countries in terms of alcohol consumption (Dmitriev 2001). Consumption in Tsarist Russia did not exceed 6 liters per person per year (l/h/y), and generally fluctuated between 2–5 liters in the European part of the country (Dmitriev 2001). Similar figures are reported by modern authors (Lisitsyn & Kopyt 1983), who estimate consumption at 2.8–3.4 l/h/y between 1890 and 1910, although these data remain incomplete. It is important to note that the severity of alcohol-related problems is determined not only by consumption levels but also by socioeconomic conditions, which may increase vulnerability to alcohol (Boyd et al. 2022).

Information on alcohol-related mortality in Tsarist Russia was recorded and published, but in a limited and incomplete form. A generalization of these data for the years 1870–1894 was carried out in recent studies by comparing them with mortality data from 2008–2009 (Andreev, Bogoyavlenskii, Stickley 2011; Andreev, Bogoyavlenskii, Stickley 2013; Andreev, Churilova 2024). According to these authors, the levels of alcohol-related mortality in the late 19th century and in the early 21st century were similar. However, this conclusion required numerous assumptions, some of which are difficult to accept unconditionally.

The most important limitation is the discrepancy in diagnostic classification of deaths. At that time, no accounting unit corresponded to the current categories of the International Classification of Diseases, 10th revision (ICD-10), such as X45 (“Accidental alcohol poisoning and exposure”) or Y15 (“Alcohol poisoning and exposure, undetermined intent”). The authors had access only to absolute numbers of “sudden deaths from diseases and painful attacks, including from drunkenness” for 1870–1894. They interpreted these deaths as “sudden deaths without signs of violence in an obvious state of intoxication” and included them as alcohol-related mortality. However, the key point is that the classification emphasized the suddenness of death, with diseases being the primary cause and deaths from drunkenness recorded only as an inclusion. This methodological limitation makes it difficult to directly compare these historical data with modern statistics over a 130-year interval.

The quality of death diagnosis in 19th-century Russia was likely affected by the underdevelopment of the medical system. In 1880, there were, on average, 58.000 people per medical district (Frenkel 1913), and by 1894, there were only 1.3 doctors per 100,000 population (Zatravkin 2017), compared with 51 doctors per 100.000 in 2022 (Russian Statistical Yearbook 2023). With such limited medical personnel in Tsarist Russia, it is reasonable to assume that physicians could not diagnose all alcohol-related deaths, and these responsibilities were often delegated to police officers in cities and to village officials in rural areas, where the majority of the population lived.

During the revolutionary years, alcohol consumption rose sharply. However, the Bolshevik government quickly suppressed and contained drunkenness through administrative and repressive measures, maintaining consumption at approximately 2 l/h/y (Deichman 1956). During the Great Patriotic War, vodka production decreased by half and beer production by threefold (National Economy of the USSR 1990), effectively reducing alcohol-related problems. The role of the so-called “People’s Commissar’s 100 grams” in subsequent alcohol history has been exaggerated (Shunyakov 2016). Pre-war and first post-war sales levels were similar – 1.9 and 1.85 l/h/y, respectively (Deichman 1956).

In the early post-war years, alcohol consumption from state sources remained limited due to widespread poverty, which, combined with other factors, led to a significant increase in moonshining. This was exacerbated by the crop failure of 1946 and the resulting famine (Ellman 2005), prompting government intervention (Appendix 1). Nevertheless, alcohol consumption continued to rise, reaching 4.2 l/h/y in 1961–1964 (Holmes 2017). In response, the first post-war anti-alcohol campaign was launched in 1958 on the initiative of N.S. Khrushchev, during which 52.000 people were convicted for producing or selling moonshine (Zhirnov 2012). However, legislative and administrative measures (see Appendix 1) failed to curb the further growth of moonshining.

To replenish the state budget, the production and sale of alcoholic beverages were increased, the range of products expanded, and inexpensive wines were imported from Algeria, Romania, Hungary, Czechoslovakia, and China. Additional contributing factors included wartime devastation, widespread poverty, human and health losses, and increasing urbanization. The prevailing post-war mindset, shaped by expectations of “another life,” also influenced drinking behaviors (Senyavskaya 1999). On the other hand, this period of social optimism was partially offset by economic growth, as reflected in GDP, which rose from $1.053.7 in 1960 to $2.345.7 in 1980 (Kalabekov 2024).

By the mid-1960s, these factors contributed to an unprecedented increase in alcohol sales and consumption in Russia, marking the beginning of a new fifty-year period in the country’s alcohol history. Sales of vodka and other distilled spirits rose from 82.1 million dal in 1940 to 230.4 million dal in 1970 (Kalabekov 2024). State sales of alcoholic beverages in the RSFSR increased from 4.6 liters per capita per year in 1960 to 8.2 liters in 1970 (Kalabekov 2024). Independent estimates of real alcohol consumption were even higher: 9.8 (Treml 1982) and 11.3 l/h/y (Nemtsov & Shelygin 2014) in 1960, and 12.0 and 13.3 l/h/y in 1970, respectively. Consumption from illegal sources was also substantial, estimated at 6.7 and 5.0 l/h/y for the 1960s and 1970s, respectively (Nemtsov & Shelygin 2014). Key documents (Appendix 1) illustrate the tense alcohol situation during this period.

In 1972, the second anti-alcohol campaign was launched, prohibiting street drinking and restricting the sale of strong alcohol from 11:00 AM to 7:00 PM. Despite these measures, by 1980, alcohol consumption reached record levels: official sales amounted to 10.5 l/h/y, while independent estimates of actual consumption reached 14.0 (Treml 1982) and 14.9 l/h/y (Nemtsov & Shelygin 2014).

The first notable change in the trend of alcohol consumption occurred in 1979–1980, driven by the increasing social and health consequences of high consumption. This led to the emergence of anti-alcohol factors, including state and local policy measures (Appendix 1). Notably, the Soviet government established a drug addiction service in 1976 and significantly expanded and strengthened the resuscitation service. Additionally, between 1965 and 1980, the progressive increase in alcohol-related mortality contributed to a reduction in the number of heavy drinkers, who were primarily responsible for the severity of alcohol problems.

Other contributing factors included the 1979 energy crisis and the threefold increase in world oil prices (Statistical Review 2009), which improved living standards and reduced poverty. Alcohol consumption is inherently a limiting phenomenon; its growth cannot continue indefinitely. By 1979–1980, these factors collectively halted the growth of consumption, albeit at a very high level – 14.0 l/h/y (Treml 1982) or 14.9 l/h/y (Nemtsov 2009). At this point, the USSR became one of the world leaders in alcohol consumption. Drunkenness caused significant material losses, which the Soviet leadership viewed as a contributing factor to the degradation of the Soviet economy. On the initiative of Yu.V. Andropov in 1982, a new cycle of anti-alcohol measures was launched (Norström 2011), culminating in the 1985 campaign.

As a result of the 1985 campaign, unlike previous initiatives, alcohol consumption was significantly reduced (Nemtsov & Shelygin 2014; Treml 1997; Norström 2011). This was the first time that the country’s leadership accepted a reduction in budget revenues to achieve public health goals by limiting alcohol production and sales, reducing the number of retail outlets, and raising prices. In addition, drinking alcoholic beverages in public places and alcohol advertising were banned. Government alcohol sales fell by 63%, reaching 3.9 l/h/y. The estimated reduction in real alcohol consumption was smaller, at 28% (to 10.2 l/h/y (Nemtsov & Shelygin 2014)), as illegal production reached near-industrial scale (Nemtsov 2009), and unrecorded consumption increased from 3.7 to 6.3 l/h/y. Other independent estimates of actual consumption were consistent with these figures (Treml 1997; Norström 2011).

By 1988, however, economic difficulties – including budget deficits and wage arrears – forced authorities to expand alcohol sales. The growing industrial production of illegal alcohol largely offset the decline in legal sales (Nemtsov 2009). Combined with increased government sales, this contributed to a rebound in alcohol consumption and, in effect, marked the winding down of the campaign from 1988 onward. Several factors contributed to the campaign’s failure, including insufficient preparation of the all-Union effort and extremely tight timelines – less than one month was allocated for planning (see Appendix 1). Moreover, the campaign failed to gain popular support, and, as Edwards (2006) argues, successful national alcohol policies require strong backing at both national and local levels. Seven years after the campaign began, by 1992, real alcohol consumption had returned to 1984 levels, estimated at 13.81, 14.0, and 14.72 l/h/y according to three independent sources (Nemtsov & Shelygin 2014; Treml 1997; Norström 2011).

The market reforms initiated by the new Russian leadership in 1992 led to alcohol prices, like those of other goods, being liberalized. However, the characteristics and harmful properties of alcohol were not taken into account (Babor 2010), and no measures were implemented in advance to limit its availability. This resulted in a sharp increase in alcohol consumption (Nemtsov & Shelygin 2014; World Health Organization 2018; Manthey et al. 2019), culminating in 1994, when Russia set a world record for alcohol-related mortality (Vishnevsky 2017; Nemtsov et al. 2019). The forced reduction of alcohol consumption during the 1985 campaign had saved over a million lives, but market liberalization ultimately caused a far greater number of deaths (Nemtsov 2009). This effect was driven by the accumulation of mortality among heavy drinkers who had survived the campaign but remained at high risk during the subsequent years (Andreev 2002).

After 1994, alcohol consumption began to decline. This was largely due to the sharp impoverishment of the population following market reforms (Nemtsov 2009), the restoration of the state alcohol monopoly (see Appendix 1), and other state interventions involving the Ministry of Internal Affairs, the FSB, and the Customs Committee. Another important factor in the mid-1990s was the reduction in the number of heavy drinkers – the primary consumers of alcohol and main contributors to alcohol-related mortality – due to high mortality.

The decline in consumption was short-lived. By 1999, a new growth phase began following the 1998 financial default, which was accompanied by ruble devaluation, reduced real incomes and savings, rising prices, unemployment, and a sharp decline in living standards (Nemtsov 2009). However, this post-crisis recession was brief, soon replaced by large-scale economic growth (Kalabekov 2024), facilitated by rising world oil prices. The liberalization of the ruble exchange rate and other factors reduced wage arrears as early as 1999, and the standard of living began to recover. These changes contributed to a reversal of the previous declining trend and the resumption of growth in alcohol consumption until 2000.

Between 1992 and 2000, the criminalization of the alcohol market, initiated by the 1985 anti-alcohol campaign, intensified. With the onset of market reforms, alcohol-related crime evolved into organized criminal activity, often involving human casualties (Puzyrev 2021). According to some estimates, in 1997–1998 the share of illegally produced alcohol exceeded 50% of total consumption (Nemtsov & Shelygin 2014). In 1999, losses to the state budget from the illegal circulation of liquor and vodka products amounted to 17.9 billion rubles in excise taxes alone, while total estimated losses reached 25–30 billion rubles (Kuvalin 2009). The legal alcohol market’s share declined from 80% in 1992 to 15% in 1997, and the contribution of alcohol excise taxes to GDP, already marginal, fell to 0.3% (Kosmarskaya 1998). These factors contributed to President B.N. Yeltsin issuing a decree to strengthen the regulation of alcohol production and circulation (see Appendix 1). However, these measures did not significantly reduce violations in the alcohol market nor eliminate the budget deficit.

In 2000, the change in the presidency of the Russian Federation was accompanied by intensified measures in alcohol policy. One of the first steps was the government order (Appendix 1) establishing a specialized administrative body, Rosspirtprom, which was granted control over state-owned alcohol and liquor enterprises, as well as broad fiscal powers to regulate the alcohol market. The main objectives of the new policy were pragmatic: replenishing the state budget by reducing the share of illegal alcohol, monopolizing the market, and encouraging consumers to shift to legally produced alcohol.

At the initial stage, Rosspirtprom focused on displacing small and medium-sized market participants, which were considered the primary source of illegal alcohol. Simultaneously, a policy of private–state monopolization of production was implemented, intended to strengthen control over the alcohol market by concentrating assets in the hands of a limited group of owners. An additional measure was the adoption of the Code of the Russian Federation on Administrative Offences in 2001 (see Appendix 1), a significant portion of which addressed violations related to the circulation of alcoholic beverages.

These measures succeeded in halting the growth of alcohol consumption but were deemed insufficient, as budget revenues increased only modestly. This led to the adoption of three legislative acts in 2005: limiting the number of beer retail outlets, introducing new excise stamps, and – critically – increasing the minimum authorized capital for market participants (Appendix 1). These measures gave the policy the characteristics of an anti-alcohol campaign.

On January 1, 2006, the first version of the Unified State Automated Information System (EGAIS), which relied on the new excise stamps, became operational. However, the system was implemented without adequate preparation, and the production of new excise stamps was significantly delayed. These issues caused logistical disruptions and disorganization in the alcohol market, leading to a sharp decline in alcohol consumption in 2006–2007. By the end of 2007, excise stamps were issued for all types of alcoholic beverages and container sizes, helping to stabilize the market.

The results of the anti-alcohol campaign launched in 2000 were multifaceted, but the most significant was the unprecedented decline in alcohol consumption, both in scale and duration, beginning in 2006. This success was recognized by the World Health Organization (WHO) (Alcohol policy impact case study 2019) and The Lancet (Russia’s alcohol policy: a continuing success story 2019), where it was cited as an example of effective government regulation of alcohol consumption (Neufeld & Rehm 2013). A major achievement of the campaign was the reduction in the criminalization of the alcohol market and the dismantling of the extensive network of organized crime in this sector, resulting in a substantial decrease in the share of illegal alcohol and, consequently, an overall decline in alcohol consumption (Alcohol policy impact case study 2019).

Unlike the 1985 campaign, which was shorter (four years versus twenty-five years) and concentrated the majority of anti-alcohol measures in a single year (Sumenkova & Katomina 2020), the 2000 campaign was characterized by its prolonged and phased implementation. Measures were introduced gradually, building upon each other and creating a cumulative effect. This gradual approach makes it difficult to assess the effectiveness of individual measures – for example, attributing the one-third reduction in alcohol consumption directly to the introduction of the 2009 Concept (Appendix 1) remains debated in several studies (Neufeld et al. 2020; Salagai et al. 2021).

A key component of the campaign’s success was the active participation of regional administrations, partly due to the accumulated social and economic problems associated with alcohol. An illustrative example is the reduction of alcohol sales hours at night, initially implemented voluntarily in response to a government proposal and later extended through legislation (Kolosnitsyna 2024). Unlike the 1985 campaign, government measures during the 2000 campaign did not provoke resistance from the population; instead, they were accompanied by substantial public support (Kolosnitsyna & Dubynina 2019), which can be attributed to their gradual and distributed implementation over time.

An important factor contributing to the decline in alcohol consumption during this period was the improvement in population well-being, driven by a significant increase in global hydrocarbon prices. For example, oil prices quadrupled between January 2000 and January 2008 (Chen et al. 2021). Between 1999 and 2013, real disposable incomes nearly tripled. In addition, the 1998 financial default led to the normalization of monetary and market relations, which further contributed to improving the economic situation (Kuvalin 2009). Notably, this period is distinguished from global trends by a decline in alcohol consumption occurring simultaneously with rising incomes, likely due to the steady intensification of anti-alcohol policies at both federal and regional levels (Appendix 1).

Due to the limited effectiveness of Rosspirtprom, it was replaced in 2008 by the Federal Service for Alcohol Market Regulation (RAR), whose powers were significantly expanded. One of RAR’s key initiatives was the introduction of the second version of the Unified State Automated Information System (EGAIS) in 2016, which continues to operate as of September 2025. Beginning January 1, 2010, the Alcohol Market Regulator introduced a minimum retail price for vodka to combat counterfeit products, and in 2011, night-time sales restrictions on alcoholic beverages were established at the legislative level (Appendix 1). Nearly half of Russia’s constituent entities further tightened these restrictions – for instance, in the Republic of Tyva, alcohol sales were prohibited from 3:00 PM until 11:00 AM the following day. The combined effect of these measures contributed to a sustained decline in alcohol consumption until the onset of the COVID-19 pandemic in 2020.

In summary, until the mid-1860s, alcohol regulation in Russia was primarily focused on finding optimal approaches to taxation. Beginning in the mid-19th century, systematic statistics on alcohol consumption began to be collected, allowing for an analysis of how the tsarist government, the church, and the population (Dmitriev 2001), and later the Soviet leadership, responded to periodic increases in alcohol consumption, which reached 4–5 l/h/y.

Since the mid-1960s, Russia entered a completely new stage in its alcohol history, characterized by unprecedented cyclical fluctuations in consumption, peaking at 18 l/h/y in 1994. The severity of the problem was such that a level of 9 l/h/y was considered relatively acceptable (Salagai et al. 2021). Each increase or decrease in consumption was driven by a complex interplay of factors, including rising or falling incomes, government policies that either counteracted or promoted consumption growth, changes in the structure of beverages consumed, psychological stress within the population, and other social and economic influences.

* * *

Alcohol-related mortality can be conditionally divided into two main groups. The first group includes cases in which alcohol is the sole or primary cause of death. This category encompasses, for example, accidental alcohol poisoning (X45) and alcoholic liver disease (K70), as well as somatic, neurological, and other diseases that exhibit specific alcoholic pathomorphological signs (Table 1).

The second group comprises cases in which alcohol abuse acts as a contributing factor, aggravating the course of somatic, neurological, or psychiatric diseases. Such pathologies may develop prior to the onset of alcohol abuse or as a consequence of prolonged consumption. According to estimates by P.O. Kuznetsova, in 2018 the number of alcohol-dependent deaths in Russia reached 195,500 (Kuznetsova 2020), representing 10.7% of all deaths that year (1.828.910 people). Of these, 25% were entirely attributable to alcohol, amounting to 48.900 deaths in 2018. This aligns closely with Rosstat data, which recorded 48.800 deaths from the same causes (Demographic Yearbook 2019).

For the period 2011–2021, detailed data on the composition of alcohol-related deaths indicate that three diagnoses predominated: alcoholic cardiomyopathy, alcohol poisoning (X45, Y15, X65), and alcoholic liver disease (including cirrhosis, hepatitis, and fibrosis), collectively accounting for more than 80% of all mortality directly caused by alcohol (Zamyatnina 2022). In the 15–49 age group, alcohol-related mortality was even higher, representing 34% of deaths among men and 20% among women (GBD… 2018).

In the United States, the total number of alcohol-related deaths is approximately 140.000 per year (average for 2015–2019) (Esser et al. 2022), despite the fact that the U.S. population is more than twice that of Russia (as of 2018). However, such comparisons should be made with caution, as Russian estimates of alcohol-related mortality (Kuznetsova 2020) are based on relative risk calculations from three large Siberian cities (Zaridze et al. 2009) and may not fully reflect mortality nationwide.

Mortality directly attributable to alcohol, particularly accidental alcohol poisoning, has been the subject of repeated research. The most detailed analysis was conducted in Mortality from External Causes in Russia Since the Mid-20th Century, edited by A.G. Vishnevsky (Vishnevsky 2017). This work identified the historical peak of alcohol mortality in 1994, with 55.500 excess deaths, and compared Russian rates with other countries. The results demonstrated that Russia leads globally in the number of accidental alcohol poisonings: rates among men were at least four times higher than in Finland, with similarly large differences observed among women. Moreover, over the fifty-year period from 1965 to 2014, life expectancy in Russia increased by only two years, and alcohol-related mortality contributed almost nothing to this gain.

According to WHO estimates, Russia ranks second in the world for the share of alcohol-attributable mortality – 17.4% among men and 4.3% among women – second only to Belarus (18.1% and 4.3%, respectively) (Global Status Report… 2018). This indicates that alcohol-related losses in Russia are disproportionately high relative to alcohol consumption, which had significantly decreased by 2018, by which time Russia was no longer a global leader in consumption (Global Status Report… 2018). This discrepancy has been described in comparative studies of wealthy and poor populations and is referred to as the “alcohol harm paradox” (Boyd et al. 2022): higher-income individuals often consume more alcohol but experience fewer alcohol-related harms, whereas lower-income groups suffer more, even at lower consumption levels. This pattern is closely linked to differences in living standards, nutrition quality, and the effectiveness of healthcare systems.

One of the key challenges in studying alcohol-related mortality is the underreporting of such deaths. Some cases are recorded under other causes of death, such as “Injuries of undetermined intent” or “Diseases of the circulatory system” (Andreev 2016; Semenova et al. 2018). Moreover, the decline in alcohol-related mortality in Russia has been accompanied by an increase in poisonings of undetermined intent (Y15) (Semenova et al. 2018). While the Y15 category is numerically documented, it is difficult to estimate how many alcohol-related deaths are recorded under “Diseases of the circulatory system” (excluding I42.6, alcoholic cardiomyopathy) or other somatic categories. As a result, the true extent of misclassification remains unclear, complicating accurate assessments of alcohol-related mortality. According to E.M. Andreev (2016), the accuracy of recording alcohol-related deaths may also be affected by subjective factors, such as the preferences of management seeking to reduce reported rates of socially significant mortality.

The aim of this study is to conduct a comprehensive analysis of the dynamics of alcohol-related and overall mortality rates in Russia from 1965 to 2020, as well as to identify patterns in their interrelationship and their association with external social phenomena influencing the level and structure of mortality.

The analysis of mortality in Russia was conducted using annual cause-of-death data from the Russian Fertility and Mortality Database (RosBRiS) of the Center for Demographic Research at the Russian School of Economics, covering the period 1965–2020. The year 1965 was selected as the starting point because RosBRiS provides data on deaths due to alcohol poisoning beginning in that year.

In parallel with alcohol-related mortality, all-cause mortality was analyzed. Standardized mortality rates (SMRs) were calculated, including SMRs for all causes and SMRs for alcohol poisoning. The European Standard Population age structure of 1976 was used for standardization. Mortality analyses were conducted separately for male and female populations. Key characteristics of the dataset are presented in Table 1.

The Table shows that the dominant diagnoses among all alcohol-related deaths during 1965–2020 were accidental alcohol poisoning and alcoholic cardiomyopathy (both 34.3%), followed by alcoholic liver disease (16.3%). The remaining categories did not exceed 5% each. Two other types of alcohol poisoning – poisoning with undetermined intent and deliberate self-poisoning – accounted for 3.5% and 0.02%, respectively.

Alcohol poisoning (X45) was selected as the primary indicator for analysis. Alcoholic cardiomyopathy was excluded due to the inaccuracy of the available data, caused by statistical reporting issues and regional variability (Samorodskaya et al. 2023), which likely exceeded the reporting inaccuracy of X45. Alcoholic liver disease was considered even less reliable for analysis (see below).

Time series analysis of annual general and alcohol-specific mortality indicators was conducted after transforming the series into stationary form using first-order differences. Stationarity was tested using the Augmented Dickey-Fuller Test (ADF Type 1) with lags 0–2. The conformity of the distributions to normality was assessed using the Shapiro-Wilk test. Correlation analysis was performed using the nonparametric Spearman method. Linear regression models were constructed for short quasi-linear segments of alcohol and overall mortality, while a quadratic polynomial was used to model overall mortality dynamics. Null hypotheses were rejected at p < 0.05.

Since 1965, the rate of increase in overall mortality in Russia rose sharply, from 1.0% per year during 1956–1965 (calculated according to (Kalabekov 2024)) to 3.4% per year in 1965–1975. The year 1965 marked a turning point in the subsequent mortality trend in Russia over the next half-century. Three main characteristics define this period:

1. Oscillatory nature and cyclicality – mortality rates displayed pronounced cycles of increase and decrease over time.

2. Unprecedented amplitude of fluctuations – the magnitude of mortality variation was historically high for Russia and unmatched in global practice.

3. Synchronicity of alcohol-related and general mortality fluctuations – peaks and troughs in alcohol-specific mortality correspond closely to those in overall mortality (Fig. 1).

Figure 1.

Trends in standardized mortality rates for overall mortality and alcohol-related mortality in men (A) and women (B), 1965–2020. Note: ACM – all-cause mortality; AAP – accidental alcohol poisoning mortality; ALD – alcoholic liver disease mortality; ChA – chronic alcoholism mortality (dependence syndrome); AP – alcoholic psychosis mortality. Dashed lines represent quadratic polynomial trends for all-cause mortality and alcohol poisoning.

Figure 1 shows that the combined dynamics of accidental alcohol poisoning and overall mortality in Russia exhibit three distinct cycles: 1965–1988, 1988–1998, and 1998–2019. Following 2019, mortality began to rise in parallel with the COVID-19 pandemic. Each cycle begins with a period of increasing mortality, while the cycle peaks in 1980, 1994, and 2003 mark the points at which mortality begins to decline. The durations of these cycles were 24, 10, and 21 years, respectively. The growth half-cycles lasted 20, 6, and 5 years (totaling 31 years), while the decline half-cycles lasted 4, 4, and 16 years (totaling 24 years). These observations suggest a trend in which the length of growth periods has decreased, whereas the length of decline periods has increased.

Over the 56-year study period, accidental alcohol poisoning was the leading cause of death among the four recorded alcohol-related categories (see Table 1). The three other causes – alcoholic psychosis (F10.2), chronic alcoholism (F10.4–F10.5), and alcoholic liver disease (K70) – largely mirrored the wave-like dynamics of poisoning (with alcoholic liver disease following a similar pattern before 1998), but with some notable differences. For example, in the 1990s, peaks in mortality from alcoholic liver disease and chronic alcoholism occurred in 1995, one year after the peak in accidental alcohol poisoning in 1994, which may be interpreted as a lag effect.

The sharp increase in mortality from alcoholic liver disease after 1998 – sevenfold over seven years – likely reflects an artifact of previously incomplete registration. Another unexpected observation is the rise in mortality from chronic alcoholism after 2003, occurring against a backdrop of declining trends in other alcohol-related causes. Finally, the increase in mortality in 2020 may be associated with the onset of the COVID-19 pandemic.

Another important characteristic of alcohol-related and overall mortality is the large amplitude of fluctuations associated with their cyclicality. From 1965 to 1994, overall mortality increased by 47.1% (accompanied by a 5.5% decrease in life expectancy at birth), while accidental alcohol poisoning rose by 358.5%. In the subsequent period, 1994–2019, both indicators declined by 42.4% (with life expectancy at birth increasing by 13.0%) and 78.2%, respectively. Significant fluctuations were also observed within certain sub-periods (Table 2). Since the durations of these periods differed, average annual rates of change are also provided, which can serve as an approximate measure of the contribution of alcohol to mortality at each stage.

For overall mortality, the magnitude of changes was generally greater in men than in women, both during periods of increasing and decreasing mortality (Table 3). This likely reflects gender differences in alcohol consumption patterns.

In the case of accidental alcohol poisoning, the fluctuations were particularly pronounced. During 1965–1980 and 1988–1994, the growth rate among women exceeded that of men. This phenomenon will be discussed further below (see Discussion). In other periods, the rates of change in standardized mortality rates (SMRs) were comparable between men and women.

Another feature of the dynamics of alcohol poisoning and overall mortality is the close direct relationship between these two indicators. This is confirmed by the correlation analysis performed after transforming the series to a stationary form. Since the distributions of the analyzed indicators for 1965–2020 were not normally distributed, Spearman’s nonparametric correlation was used (r_s = 0.84, 95% CI 0.73–0.91, p < 0.001). It should be noted that alcohol poisoning accounted for only a small share of overall mortality: 1.6% among men and 0.7% among women. The share of alcohol poisoning in overall mortality was linearly related to the SMR for alcohol poisoning, varying within the range of 0.7–2.7% for men and 0.3–1.5% for women, both overall and within individual time periods (Fig. 2).

Figure 2.

Linear regression of the share of alcohol poisoning in overall mortality versus the standardized mortality rate (SMR) from alcohol poisoning (solid lines). Note: 1 – 1965–1984; 2 – 1984–1992; 3 – 1994–2005; 4 – 2005–2019. Circles denote annual observations; the dotted line indicates the trend for 1992–1994.

Figure 2 shows that the values for 1993–1994 deviated from the regression for 1984–1992 and lay outside the 95% confidence interval. This deviation is explained by the sharp increase in somatic mortality during these two years, associated with the social consequences of market reforms. As a result, although SMR for alcohol poisoning increased, mortality from other causes grew even faster. Consequently, the share of alcohol poisoning in overall mortality declined, which was reflected in a decrease in the slope of the regression line.

The ratio of male to female mortality varied substantially between 1965 and 2020 and differed for alcohol poisoning and overall mortality (Fig. 3).

Figure 3.

Male-to-female ratio of mortality from alcohol poisoning and overall mortality, 1965–2020. Note: APO-M/F – male-to-female ratio for alcohol poisoning mortality; OM-M/F – male-to-female ratio for overall mortality; APO – alcohol poisoning mortality for both sexes combined.

For overall mortality, the almost twofold predominance of men was expected, as was the increase in their proportion during periods of rising alcohol poisoning mortality (Fig. 3). Less expected was the overall increase in the proportion of men over the 55-year period, although this increase was relatively small, from 1.7 to 2.0 times.

Mortality from alcohol poisoning among men occurred almost five times more often than among women. The dynamics of this indicator were unexpected: in 1965–1985 the male-to-female ratio of alcohol poisoning decreased substantially (Fig. 3). This is probably due to the fact that before the Great Patriotic War there was strong social condemnation of women’s alcoholism; therefore, heavy drinking among women, as well as alcohol poisoning, were relatively rare. Changes in the role of women during the war led to an increase in their social status. This facilitated women’s transition to heavier alcohol consumption, and they began to approach men in the frequency of alcohol poisoning. As a result, the male-to-female ratio of alcohol poisoning decreased from sevenfold in 1965 to fourfold in 1985. The question remains why a similar pattern was not observed among moderately drinking women (as reflected in overall mortality); we return to this issue in the Discussion.

During the anti-alcohol campaign of 1985, when the availability of alcohol was restricted, men were more active than women in obtaining access to alcohol. As a result, male mortality from alcohol poisoning increased relative to that of women, and the male-to-female ratio rose to 4.5. During the market reforms of the 1990s, men’s advantage in access to alcohol diminished, and the ratio returned to about four. In the 2000s, against the background of declining alcohol consumption and mortality, men – who were the main heavy alcohol consumers – reduced their consumption more slowly than women in the same group, probably due to greater behavioral rigidity. As a result, the male-to-female ratio of alcohol poisoning mortality began to increase again.

In 2020, mortality began to rise in connection with the COVID-19 pandemic. It cannot be ruled out that in 2020, or perhaps even earlier, a new cycle of alcohol consumption and alcohol-related mortality began.

It is noteworthy that mortality levels at the beginning and at the end of the study period were broadly comparable. Thus, overall mortality returned to the 1965 level in 2013, while mortality from alcohol poisoning (X45) reached the same level in 2017. In other words, after half a century the country had returned to approximately the same mortality levels from which it started.

The complexity of mortality dynamics in 1965–2020 requires analysis by individual periods. These time series can be decomposed into periods based on changes in trend, identifying quasi-linear segments (Fig. 4). For these selected periods, linear regressions of overall mortality on the SMR for alcohol poisoning were constructed.

Figure 4.

Linear regression of all-cause SMR on SMR for alcohol poisoning by selected periods, 1965–2020. Men (A) and women (B). Note: Selected periods are highlighted in color. Transition periods and 2019–2020 are indicated by dotted lines. Thick lines represent observed values; thin lines represent linear regressions for the corresponding periods.

Figure 4 shows seven consecutive periods from 1965 to 2020. The regression lines for most periods are approximately parallel, more clearly among men (Fig. 4A). There are four exceptions: 1965–1980 and 2004–2010 (both among women), and 2010–2019 (for both men and women). The 2019–2020 trend reflects the onset of the COVID-19 pandemic.

As noted earlier in Table 3, the rate of increase in alcohol poisoning among women was particularly high after 1965, which determined the slope of the regression in 1965–1980 and distinguished it from other periods (Fig. 4B). The increase in the regression slope after 2010, visible in Fig. 4, may be explained by the fact that the decline in overall mortality occurred more rapidly than the decline in mortality from alcohol poisoning.

Changes in the regression slope were observed throughout the analyzed period. This can be detected, although it is difficult to observe when short periods are analyzed (Fig. 4). The pattern becomes clearer when longer periods are analyzed using linear regression without stratification by sex (Table 4).

Table 4 shows that from period to period there is a progressive increase in the regression coefficient of overall mortality on mortality from alcohol poisoning. This means that over time overall mortality changes faster than mortality from alcohol poisoning (from approximately 10 to 48 per unit of SMR from alcohol poisoning). It is important to note that in our case the regression coefficient reflects the relative rate of increase or decrease in mortality.

If we assume that alcohol poisoning reflects the level of alcohol consumption, it can be inferred that over the years overall mortality has become progressively less dependent on alcohol consumption. One possible explanation is that the increase in life expectancy in the general population outpaces that among heavy alcohol consumers (see Discussion). Another factor may be advances in general medicine compared with developments in addiction medicine and resuscitation services.

More difficult to explain is the 2.5-fold increase in the regression coefficient after 2010 (from approximately 18 to 48 per 100,000; see Table 4). This may reflect a change in the relative rates of decline in the two types of mortality: either an acceleration in the decline of overall mortality or a slowdown in the decline in mortality from alcohol poisoning. The latter explanation appears more likely. For example, among men the decrease in mortality from alcohol poisoning was 7.3% per year in 2003–2010 but only 5.5% per year in 2010–2019, whereas overall mortality continued to decline at approximately the same rate (2.8% and 2.7%, respectively).

So far, the analysis has assumed a linear relationship between mortality from alcohol poisoning and mortality from all causes within individual periods. For short periods this assumption is acceptable (Fig. 4). However, the trends in the two mortality classes after 2004 (Figure 1), together with the slowing decline in alcohol-related mortality after 2010 (Table 4), suggest a more complex relationship between them during this period. This becomes evident if the period 2004–2019 is examined without dividing it into two subperiods. For ease of analysis, the mortality series were made comparable by aligning the extreme points (2005 and 2019; Fig. 5).

Figure 5.

Trends in overall mortality (OM) and mortality from alcohol poisoning (AP) in Russia, 2000–2020. Men (A) and women (B). Note: The indicators for 2005 and 2019 (marked with circles) are aligned to facilitate comparison. The solid line connects these points, and the dotted line represents the extrapolated trend using a cubic polynomial.

This approach allows clear visualization of both the similarities and differences in trends between overall mortality and alcohol poisoning mortality after 2004. Both series show a sharp decline in 2006–2007, followed by a slower decrease, largely in phase. An exception occurred in 2010, during an extreme summer heat wave. Unlike alcohol poisoning mortality, overall mortality temporarily slowed its decline, followed by a stronger decrease in 2011. The decline in 2006–2007 and the subsequent slowdown was more pronounced for alcohol poisoning.

Linear trends poorly fit the data (adjusted R² ≈ -0.018 for men and -0.054 for women), while a cubic polynomial provides an excellent approximation (R² = 0.991 for men, R² = 0.986 for women). This indicates a nonlinear relationship between the two mortality series during this period.

Russian history over the last 50 years has been characterized by a tense situation with high levels of alcohol consumption and associated excess mortality. During this period, Russia was among the global leaders in alcohol consumption, and in 1994 (Fig. 1) the country reached a peacetime record in mortality (Vishnevsky 2017). If the mortality trend is smoothed (dotted line in Fig. 1), an arc-shaped function emerges, with maximum mortality occurring in the middle of this period, in 1994.

After 50 years, overall mortality and mortality from alcohol poisoning returned to approximately the levels observed in 1965, having followed a difficult trajectory with enormous human losses. From 1965 to 1994, overall mortality increased by almost one and a half times (47%), while mortality from alcohol poisoning increased more than fourfold (359%), raising its share in overall mortality from 0.8% to 2.4%.

A distinctive feature of the Russian alcohol situation is its exceptional dynamism, with relatively short fluctuations (Fig. 1). Worldwide, abrupt changes in alcohol consumption have occurred rarely and usually only under extreme conditions. For example, in Denmark in 1914, when German troops blockaded the country, alcohol consumption declined sharply within a single month (Nielsen 1965). Another example occurred in Finland, where the liberal alcohol law of 1968 led to a 1.5-fold increase in alcohol consumption within a year (Lyalikov 1967). However, these were isolated jumps followed by a gradual return to the previous level.

In global practice, changes in alcohol consumption do occur, but they usually develop slowly while maintaining a long-term trend (Edwards et al. 1994), or take the form of waves with long periods (“long waves of consumption”) (Skog 1986). In this sense, alcohol consumption is generally considered a relatively conservative phenomenon.

In contrast, in Russia, over the 55-year period from 1965 to 2019, three cycles of overall mortality were observed, occurring in phase with mortality from alcohol poisoning and lasting 23, 10, and 21 years (Fig. 1). A hypothesis explaining these fluctuations in Russia has been proposed by (Grigoriev, Andreev 2015). The authors suggest that the anti-alcohol campaign of 1985 delayed the deaths of a large number of heavy drinkers. However, these individuals retained a high risk of premature death due to accumulated alcohol-related pathology, and this risk was realized during 1993–1995, when the social consequences of market reforms emerged. According to the authors, this mechanism – periodic accumulation of heavy drinkers followed by their subsequent mortality – determined the cyclical nature of mortality after 1985.

For the theory of mortality dynamics in Russia, it is tempting to identify a single mechanism underlying these fluctuations. Such a mechanism may indeed exist and appears to be related to fluctuations in alcohol consumption. As shown in the Introduction, the factors influencing alcohol consumption changed dramatically from phase to phase. In essence, the “working unit” of Russian alcohol history may be the half-cycle – periods of increasing or decreasing mortality – each determined by specific factors affecting alcohol consumption.

There is an example in which such a half-cycle (1988–1994) can be further divided into two subperiods according to the dominant factors. In 1988–1992, alcohol consumption and alcohol poisoning began to increase again after the anti-alcohol campaign of 1985, returning to a level determined by the population’s underlying demand for alcohol. In 1992–1994, the rapid growth of alcohol consumption and alcohol-related mortality was largely driven by the social and economic disruptions associated with market reforms.

In our study, two classes of mortality were examined. The first is mortality from alcohol poisoning, which occurs almost exclusively among heavy alcohol consumers (Martin McKee Leon; David Zaridze), many or most of whom suffer from alcoholism (Paukov & Ugryumov 1985; Paukov & Erokhin 2020). This indicator is widely used as a proxy for actual alcohol consumption (e.g., (Danilova et al. 2020)) because alcohol poisoning can result from both legally and illegally produced beverages, while the increased toxicity of illicit alcohol – except in cases involving methanol – has not been confirmed (Nuzhny & Savchuk 2006).

The second class is overall mortality, which can be conditionally divided into several groups: abstainers (zero alcohol consumption), moderate drinkers, individuals with problematic drinking (household alcoholics), and heavy drinkers, some of whom belong to the first group described above. The latter account for only about 1.4% of overall mortality and therefore have little influence on overall mortality trends.

These groups form a continuous population characterized by a unimodal lognormal distribution of alcohol consumption (Andrienko & Nemtsov 2005), the “tail” of which consists of heavy drinkers, most of whom eventually die from alcohol poisoning. The boundaries between these groups are not rigid; rather, they are closely interconnected through transitions of individuals from one drinking category to another, as described by the theory of collective drinking behavior (Skog 1985). The principal difference in drinking behavior between heavy drinkers and the rest of the population is the presence of mental and/or physical alcohol dependence among the former (alcoholism).

In our study, the overall mortality rate serves two functions. First, based on the synchronicity of trends and its correlation with mortality from alcohol poisoning, it illustrates the role of alcohol abuse in shaping mortality patterns in Russia over the past 55 years. Alcohol abuse is a significant, although not the main, cause of death in Russia. According to estimates based on (Kuznetsova 2020), alcohol-related mortality accounted for 10.7% of all deaths in Russia in 2018 (195.5 thousand deaths). At the same time, diseases of the circulatory system remain the leading cause of death (43.8%) (Healthcare in Russia 2023; Vishnevsky et al. 2016). However, it should be noted that a substantial proportion of this mortality is associated with alcohol abuse (Shkolnikov et al. 2013). The synchronicity and strong correlation between all-cause mortality and mortality from alcohol poisoning therefore suggest that alcohol consumption has been a major modulating factor of overall mortality in Russia since 1965.

The second function of the overall mortality rate is methodological. Since it covers virtually all deaths in the country, it represents the most comprehensive demographic indicator. For this reason, overall mortality can serve as a control indicator for mortality from alcohol poisoning, which is known to be affected by registration errors (Andreev 2016; Semenova et al. 2018). The regression lines for most periods show parallel relationships between overall mortality and mortality from alcohol poisoning (Fig. 4), which may indicate the absence of major systematic errors in recording alcohol-related mortality at the macro level. Two periods, however, represent exceptions.

The first occurred in 1965–1980 and was manifested by a divergence in the trends of overall mortality and alcohol poisoning among women (Fig. 1B and Fig. 4B), caused by a more rapid increase in alcohol poisoning mortality. This phenomenon may be related to the gender imbalance resulting from wartime male losses and widespread widowhood (Vishnevsky 2016), as well as to changes in the social status of women following the Great Patriotic War (Senyavskaya 1999). As a result, a growing number of women appear to have moved from the group of moderate drinkers to heavy drinkers. Consequently, women began to “catch up” with men in the prevalence of heavy drinking, which led to increased mortality from alcohol poisoning and a reduction in the male-to-female mortality ratio (Fig. 3).

At the same time, women who maintained moderate alcohol consumption preserved the usual gender mortality ratios relative to men. Their share in mortality tended to decrease when alcohol consumption increased – due to the rising contribution of male mortality – and to increase when alcohol consumption declined. This pattern is reflected in Fig. 3, where deaths from alcohol poisoning serve as an indicator of alcohol consumption levels.

The second period of divergence between the trends of the two mortality indicators occurred in the 2000s and was observed in both women and men (Fig. 1). In 2000, with the establishment of Rosspirtprom, the Russian government began implementing measures to regulate the alcohol market. However, due to difficulties in organizing the new system, the growth of alcohol consumption, mortality from alcohol poisoning, and overall mortality continued until 2002. Only in 2003–2004 did these indicators begin to decline slowly (Fig. 1).

In 2005, a consistent series of regulatory measures was introduced, and alcohol policy effectively took the form of an anti-alcohol campaign, although it was not officially declared (Appendix 1). Some implementation problems occurred. In particular, in 2006 the first, incomplete version of EGAIS was introduced. Together with other regulatory measures, this led to disruptions in logistics and temporary instability in the alcohol market, reducing alcohol availability. As a result, mortality from alcohol poisoning and overall mortality declined sharply in 2006–2007, at a rate comparable to the mortality decline observed during 1985–1987 (Table 2).

By 2008, the alcohol market had largely stabilized. Nevertheless, the anti-alcohol measures introduced in the mid-2000s and strengthened in subsequent years led to a sustained long-term decline in alcohol consumption (Alcohol policy … 2019; Salagai et al. 2021; Kolosnitsyna et al. 2017), as well as in mortality from alcohol poisoning and overall mortality (Fig. 1). This represents the main feature of the alcohol situation in Russia during 2000–2019.

Importantly, unlike the anti-alcohol campaign of 1985, regional authorities actively supported the alcohol control policies of the 2000s (Kolosnitsyna et al. 2014). Having long faced serious alcohol-related social problems, many regions introduced additional restrictions. At the same time, the population largely accepted these measures without resistance, since the policies were not radical and were implemented gradually.

The period after 2000 is also characterized by a gradual slowdown in the decline of mortality (Fig. 5). Several factors may have contributed to this slowdown. One of them is a fundamental property of population alcohol consumption: it can never fall to zero. As consumption declines, the rate of decline tends to slow and eventually approaches a non-zero plateau, or consumption may begin to increase again. Another possible factor is the decreasing effectiveness of previously introduced measures and the slowdown in the implementation of new anti-alcohol policies (Kolosnitsyna 2024), especially since the initial policy objective – replenishing government revenues – had largely been achieved.

The decline in mortality after 2000 also had another important feature, first described by Danilova et al. (Danilova et al. 2020): the trend in life expectancy ceased to correspond closely to the trend in mortality from alcohol poisoning, which the authors considered the best indicator of the level of alcohol consumption. A similar phenomenon was also observed in the present study: a divergence between the trends in overall mortality and alcohol poisoning (Fig. 1), particularly pronounced among men (Fig. 1A) after 2010 (Fig. 4).

The authors (Danilova et al. 2020) interpreted this phenomenon as evidence that life expectancy became less dependent on changes in alcohol consumption starting in the mid-2000s. They explained the divergence between trends in alcohol poisoning and overall mortality in 2003–2017 primarily using demographic factors. In particular, they linked this pattern to a reduction in mortality among the elderly and very old as a result of improvements in healthcare. According to their interpretation, this increased the proportion of older individuals in the population, a group in which alcohol consumption tends to decline due to age-related characteristics.

However, although mortality indeed declined in older age groups between 2000 and 2019, this decline occurred simultaneously across all ages above 15 years. Moreover, the reduction in mortality among people aged 60–74 and 75+ was smaller (-40% and -31%, respectively) than in the younger age groups of 15–29 and 30–44 (-62% and -45%). This suggests that the decline in mortality among the elderly and very old cannot fully explain the slowdown in the decline of overall mortality after 2010.

A more detailed analysis shows that the divergence between overall mortality and mortality from alcohol poisoning began earlier, as early as 2006 (Fig. 5), following radical changes in the alcohol market. It is important to note that, compared with overall mortality, mortality from alcohol poisoning declined more sharply in 2006–2010 but more slowly after 2011 (Fig. 5). Thus, the years 2010–2011 represent a turning point in the relationship between these indicators (Fig. 4).

Several factors may explain this shift. The first is the heat wave of 2010, which was accompanied by excess mortality – primarily cardiovascular – in both Europe and Russia (Revich 2011). As a result, unlike mortality from alcohol poisoning, the decline in overall mortality slowed in 2010.

This effect was reinforced by changes in alcohol policy. In 2010, a minimum retail price for alcoholic beverages was introduced, which led to the removal of part of the illegal alcohol supply from the market. In addition, in 2011 the government implemented a new tax policy regime (Yakovlev 2016). These measures may have had different effects on moderate and heavy drinkers, since the latter are generally less sensitive to price changes (Yakovlev 2018).

Another possible explanation for the slower decline in mortality from alcohol poisoning is the conservative behavior of heavy drinkers, many of whom suffer from alcoholism (Paukov & Ugryumov 1985; Paukov & Erokhin 2020). These individuals are disproportionately represented in older age groups and, due to alcohol dependence, have greater difficulty changing their drinking patterns. In contrast, younger cohorts showed a noticeable decline in vodka consumption (Denisova & Kartseva 2010; Kueng & Yakovlev 2021).

A similar phenomenon was described by Radaev and Roshchina (Radaev & Roshchina 2021). They documented a decline in alcohol consumption between 2006 and 2018 across all consumer groups and interpreted this as support for the collective consumption theory (Skog 1985). However, the magnitude of change differed across groups: the decline was smaller among heavy drinkers than among lighter drinkers. The authors referred to this pattern as the “soft collectivity hypothesis.”

A comparable situation was observed in Finland after the adoption of the alcohol law of 1968, which introduced price preferences for beer. As a result, the population as a whole increased beer consumption, whereas heavy drinkers continued to prefer strong spirits for several years. Consequently, total alcohol consumption increased for a certain period following the reform (Lyalikov 1987).

Due to the divergence in trends between overall mortality and mortality from alcohol poisoning, several considerations should be noted. In 2019, a methodology for estimating alcohol consumption at both national and regional levels was proposed (Appendix 1). This method relies on a combination of three indicators: mortality from alcohol poisoning, the incidence of alcoholic psychosis, and the difference in life expectancy between men and women.

In the 1980s–2000s, overall mortality and mortality from alcohol poisoning showed complete synchronicity, which made it reasonable to assume that mortality from alcohol poisoning adequately reflected the actual level of alcohol consumption in the population. However, after 2005, when a divergence between these two indicators became apparent, mortality from alcohol poisoning could introduce bias into estimates of total alcohol consumption.

Moreover, the incidence of alcoholic psychoses is known to be poorly recorded. This is evidenced by the substantial discrepancy between hospital statistics and dispensary records, the latter being the source used in official state reporting. Taken together, these limitations require a cautious interpretation of alcohol consumption estimates based on the Russian Ministry of Health methodology, particularly for the period after the mid-2000s.

The unprecedented decline in mortality from alcohol poisoning since 2004, associated with a reduction in alcohol consumption, should not lead to complacency. Since 2018, alcohol sales in Russia have been increasing, and in 2023 they reached the highest level observed in the previous six years. Demand for strong alcoholic beverages has grown particularly sharply, while Google Trends data indicate a substantial increase in internet searches related to illegal alcohol during 2024.

The adult mortality rate in Russia also remains extremely high, comparable to that observed in many developing countries (Eberstadt 2024). It is also alarming that the harm caused by alcohol in Russia is disproportionately large relative to the level of alcohol consumption. The country has long since left the group of global leaders in alcohol consumption, yet until 2020 it remained, together with Belarus, among the leaders in mortality from alcohol poisoning.

This discrepancy between alcohol consumption levels and alcohol-related mortality is not unique to Russia. A growing body of research shows that the negative consequences per unit of alcohol consumption – the so-called “harm per liter” of pure alcohol – tend to be higher in lower-income countries, whereas overall alcohol consumption is generally higher in wealthier countries (Boyd et al. 2022; Room & Rehm 2023). At the same time, international comparisons indicate that the role of alcohol in Russian mortality remains exceptionally large (Trias-Llimós et al. 2018).

The results of this study may be affected by the well-documented underreporting of post-mortem diagnoses of alcohol poisoning (Andreev 2016; Global Status Report… 2018). The existence of such underreporting is widely acknowledged; however, the magnitude of the resulting distortion and its potential influence on the conclusions require consideration. It should be noted that overall mortality represents the most reliable demographic indicator, and during 1980–2004 the trends in mortality from alcohol poisoning closely coincided with those in overall mortality. The deviations observed in 1965–1980 and 2004–2019 can be explained by substantive factors and are unlikely to be associated with systematic misclassification of causes of death. Although some misclassification undoubtedly occurred, the synchronicity between overall mortality and alcohol poisoning suggests that it did not substantially affect the long-term trend in alcohol poisoning mortality.

We did not adjust accidental alcohol poisonings (ICD-10 code X45) by including poisonings and exposures with undetermined intent (Y15), as is commonly done in recent studies. This decision was made because mortality data for Y15 have been available only since 2000, whereas the primary objective of the present study was to compare trends in overall mortality and alcohol poisoning mortality since 1965.

Moreover, it cannot be ruled out that the reassignment of X45 diagnoses to other categories also occurred before 2000. Local observations suggest that such reclassification did take place (Appendix 2). Including both X45 and Y15 would increase the divergence between trends in overall mortality and alcohol poisoning mortality after 2000, but only marginally. Expressed as shares of overall mortality, the difference between X45 alone and the combined indicator (X45 + Y15) amounts to 0.07 percentage points in 2000 and 0.18 percentage points in 2020.

We did not use alcohol sales data or estimates of alcohol consumption as controls for alcohol-related mortality. This decision was motivated by the known inaccuracy of sales statistics and the approximate nature of consumption estimates, both in Russia (Nemtsov 2004; Zaigraev 2004) and internationally (Kilian et al. 2020). A recent example illustrating this problem was the introduction of the Unified State Automated Information System (EGAIS), which resulted in a sharp increase in recorded alcohol sales between 2016 and 2018 due to the expansion of the reporting coverage rather than an actual rise in consumption.

Finally, substantial registration errors in mortality from alcoholic liver disease (ICD-10 code K70) and chronic alcoholism (ICD-10 code F10.2; Fig. 1) prevented their inclusion in the analysis. For the same reason, mortality associated with alcoholic psychoses (ICD-10 codes F10.4–F10.5) was also excluded.

This study was conducted as part of the research project “Population reproduction in the context of socio-economic development” by T.A. Fattakhov.

The authors express gratitude to Professor Olga Yuryevna Rebrova (Doctor of Medical Sciences, Federal State Autonomous Educational Institution of Higher Education N.I. Pirogov Russian National Research Medical University, Pirogov) for her assistance with calculations and valuable discussions of the results.

Alexander Vikentievich Nemtsov, Doctor of Medical Sciences, Federal Scientific Center for Psychological and Interdisciplinary Research, Moscow, 125009, Russia E-mail: nemtsov33@gmail.com

Timur Asfanovich Fattakhov, PhD in Economics, Senior Researcher, Department of Population, Faculty of Economics, Moscow State University named after M.V. Lomonosov, Russia, Moscow. E-mail: timur300385@mail.ru

One of the authors (A.V. Nemtsov) was sent to Petrozavodsk (then the Karelian ASSR) in 1988 in connection with a conflict between the republican forensic medical service and the republican authorities. The conflict was triggered by a threefold increase in deaths from alcohol poisoning recorded in 1988. The authorities considered this increase to be an artifact, rejected the forensic medical report, and insisted on the dismissal of the newly appointed head of the service.

During the subsequent investigation, which involved the police, it was revealed that in 1987 – and continuing in 1988 – there had been a large-scale import of illicit alcohol (moonshine), mainly from Belarus. Trucks carrying crates of moonshine would arrive in villages in the morning (often the alcohol was sold only in crates), and by the evening the first poisoning cases would begin to appear.

It became evident that the establishment of police checkpoints on the roads could have significantly reduced mortality from alcohol poisoning in Karelia.

The investigation also revealed that for many years prior to 1988 there had been a practice of artificially understating the number of deaths attributed to alcohol poisoning. This partly explains the unusually high figures recorded in 1988.