Alcohol and other drug related deaths among young people in CIS countries: Proximal and distal causes and implications for policy

Introduction 

The main elements of the collapse of the Soviet Union and the transformations that followed are well known—the widespread descent into poverty, accelerating inequalities, increased crime, the rise of lawlessness and corruption, the spread of armed conflict, mass migration of people to their ethnic ‘homelands’, and the mortality crisis that initially appeared to impact on all adults, but over the longer term became concentrated on middle aged men, particularly in Russia and Ukraine. Although social and economic indicators have generally improved since 2000, poverty, inequality, crime and corruption remain high. Moreover, widespread use of alcohol and other drugs (AOD) has emerged as a major issue affecting the health and well-being of people of all ages.

While there is now a large literature on AOD use in the region, the relationship between AOD use and mortality rates among young people has been somewhat neglected. Alcohol consumption, especially in Russia, by far the largest of the CIS countries, has been much studied in terms of its association with deaths among middle-aged and older men (Nemtsov, 2002, Nemtsov, 2005), and its association with homicide and suicide (Pridemore, 2004, Pridemore, 2006). Injecting drug use (IDU) has also been singled out for policy attention because of the human immunodeficiency virus (HIV) epidemic that has spread rapidly through many countries in the region since the late 1990s (Booth et al., 2006, Rhodes et al., 2006). The purpose of this paper is to examine how a range of risk and protective factors may have influenced AOD use and resulting mortality among young people (aged 15–29) across the 12 successor states of the Soviet Union collectively known as the Commonwealth of Independent States (CIS).

The question that this paper seeks to address is approached through two conceptualisations of the aetiology of AOD use and related problems. The first conceptualisation (the ‘proximal-causes model’) focuses attention on AOD-specific (e.g. AOD supply) or individual-level risk factors (e.g. ability to deal with stress). These factors are typically the focus of AOD research, policies and programmes. A recent example of this model has been presented by Birckmayer, Holder, Yacoubian, and Friend (2004). The second conceptualisation (the ‘distal-causes’ model) attends to broader social factors that can impact a range of problems across the life course, of which AOD problems are a subset (Spooner & Hetherington, 2005). Social factors include the economic, cultural, physical and policy environments of a society. The two conceptualisations are not in competition. Rather, attention to both is necessary for understanding the aetiology of AOD abuse and youth mortality, and for the development of policies to reduce AOD use and harm.

An ecological study of aggregate country-level administrative and survey-based data was used to explore the relationships between risk factors, AOD use and youth deaths across time and between countries. The greatest strength of this approach lies in its “low cost, convenience, and the simplicity of analysis and presentation” (Stevenson & McClure, 2005, p. 2). In the case of this analysis, moreover, the ecological approach offered a methodological framework when detailed individual or household level data were largely missing. Sources of data included the United Nations Children's Fund's (UNICEF's) TransMONEE database, the World Health Organization's (WHO) Health for All and Mortality databases, statistical data collected by the United Nations Office on Drugs and Crime (UNODC), and the World Values Survey. Ecological studies have weaknesses, including susceptibility to measurement error and ambiguity of cause and effect (Morgenstern, 1982, Susser, 1994), and the data for this study are not immune to this. Furthermore, systematic data on the range of risk and protective factors that might be relevant in the CIS were not available.

To compensate for this problem, qualitative English language research literature on AOD use in the region was used to supplement the statistical data. Key sources for this literature included academic databases (Medline, International Bibliography of the Social Sciences, Web of Science, JStor); websites of international organisations (WHO, European Monitoring Centre for Drugs and Drug Addiction (EMCDDA), the UNODC, UNAIDS, the United Nations Development Programme, and the World Bank); and materials referenced by research in these sources. One shortcoming was the exclusion of Russian language materials, compensated in part through support and advice from experts in the region.

In this paper the term ‘use’ denotes use which might or might not contribute to harm, while ‘abuse’ denotes use that is risky or harmful. This distinction originates in the aetiological literature that demonstrates different risk factors for AOD use and abuse (Spooner & Hetherington, 2005).

AOD use and youth mortality in CIS countries 

Official mortality rates among young people in most CIS countries are higher than in selected countries of Western Europe, but there is substantial variation between CIS countries (Table 1). In general they are highest in the western CIS, the wealthiest countries with the lowest poverty and under-5 mortality rates. They are lowest in the Caucasus, and the rate in Armenia is actually lower than Spain or the United Kingdom. Central Asia lies between these two groups, and includes the region's poorest and least developed countries with uniformly high poverty and under-5 mortality rates. Youth mortality rates may be less well recorded in Central Asian countries than in Russia, Ukraine or Belarus, but it seems unlikely that ‘true’ mortality rates in the former countries approach those in the latter. There are some exceptions to the neat geographical division of youth death rates. Kazakhstan in Central Asia has the second highest youth mortality rate after Russia; and the mortality rate in Moldova is more akin to Central Asia than to neighbouring countries of the Western CIS. Nonetheless the geographical division broadly captures divisions in mortality rates across the region.

Table 1. Mortality and other indicators for CIS countries

		Gross National Income per capita, $US (2005)	People in extreme poverty , per cent (2002–2003)	Under 5 mortality rate (2005)	Mortality rate among 15–29 year olds, per 100,000 (2004)	Change in mortality among 15–29 year olds 1990–2004 (1990=1.0)
	Belarus	2,750	4	12	142	1.10
	Moldova	880	43	16	93	0.66
	Russia	4,460	9	18	247	1.52
	Ukraine	1,520		17	161	1.22
	Kazakhstan	2,930	21	73	222	1.41
	Kyrgyzstan	440	70	67	126	0.86
	Tajikistan	330	74	71	106	1.04
	Turkmenistan	1,340		104	143	1.09
	Uzbekistan	510	47	68	98	0.87
	Armenia	1,470	50	29	47	0.56
	Azerbaijan	1,240		89	80	0.86
	Georgia	1,350	52	45	76	0.79
	Germany	34,580	0	5	44	0.62
	Spain	25,360	0	5	49	0.51
	UK	37,600	0	6	51	0.83

Source: Gross National Income (GNI) and under 5 mortality rate: UNICEF (2006a); People in extreme poverty: World Bank (2005); 15–29 mortality rate WHO Mortality Database, except for Turkmenistan, which are from UNICEF TransMONEE Database. Notes. GNI is calculated in US$ at market exchange rates. People in extreme poverty are those living in households with consumption valued at less than US$2.15 per person per day (calculated at Purchasing Power Parity exchange rates). The under-5 mortality rate is estimated for several countries, and is calculated from a mixture of official administrative data and survey data. The 15–29 mortality rate is calculated from official administrative data. This age range was chosen because it has been widely used for categorising AOD related mortality. Age 15–29 mortality rates for Tajikistan and Georgia refer to 2001, for Azerbaijan to 2002, and for Armenia to 2003, and change refers to the period 1990 to the latest year.

The collapse of the Soviet Union triggered an increase in mortality among young people. While increasing death rates among older people (particularly men) were driven by growth in ‘natural’ causes (illnesses and diseases), the rise in death rates among younger people was for the most part caused by increases in deaths from external causes: accidents, poisonings, suicides and homicides. Up to about 1990, youth mortality rates were in modest decline across most of the Soviet Union, but both natural and external deaths increased with the onset of the Transition. Rates were always higher in Russia, and they increased more steeply than elsewhere during this time (Table 1, final column). Trends in deaths in Kazakhstan tended to follow those in Russia. In 2004, the death rate for 15–29 year olds in Russia was almost twice that in Kyrgyzstan, and over five times the rate in Armenia.

Recent patterns of AOD use and youth mortality across most CIS countries suggests a relationship between the two. Table 2 shows recent estimates of consumption of alcohol and opiates, and of IDU. The Table includes two estimates of average alcohol consumption (Columns A and B), one estimate of the proportion of adults who used opiates over a year (Column C), and one estimate of the proportion of the population who were injecting drug users (Column D). Column E contains an estimate of the proportion of injecting drug users who were aged 29 or under.

Table 2. Estimates of alcohol and opiate abuse, and injecting drug use in CIS countries

		Pure alcohol consumption (litres) 2003	Pure alcohol consumption (litres) 2001–2003 among 18–29 year olds	Percent of adult population who have used opiates, latest year	Estimate of IDUs – percent population aged 15–64	Per cent IDUs aged ≤29	Overall death rate, 15–29 year olds, 2004 (proportion average for CIS)
		A	B	C	D	E	F
	Russia	8.9	4.1	2.0	2.22	86	1.92
	Kazakhstan	2.2	2.4	1.3	1.58	56	1.73
	Ukraine	5.2	2.8	0.8	1.19		1.25
	Turkmenistan	0.7		0.3	0.43		1.11
	Belarus	4.8	4.2	0.4	0.65	49	1.10
	Kyrgyzstan	2.4	1.0	2.3	1.63	53	0.98
	Tajikistan	0.3		1.0	1.53		0.83
	Uzbekistan	1.0		0.5	0.35		0.76
	Moldova	10.2	2.6	0.1	1.45	57	0.72
	Azerbaijan	3.1		0.2	0.39	27	0.62
	Georgia	1.3	2.9	0.6	0.37		0.59
	Armenia	1.1	1.3	0.3	0.4		0.37
	Correlation with death rate (R2)	0.14	0.25	0.37	0.47	0.60

Sources: Column A: WHO Health for All Database; Column B: Pomerleau et al. (2005); Column C: UNODC (2006); Columns D and E: Aceijas et al. (2005); Column F: WHO Mortality Database (except Turkmenistan – see notes to Table 1).

Countries are ordered in Table 2 according to death rates among 15–29 year olds in 2004, but similarities with the geographical ordering on Table 1 are notable. Russia, with the highest youth death rates, also has high levels of alcohol use, opiate use and IDU, while Armenia, Azerbaijan and Georgia have generally low rates of AOD use, and low youth mortality rates. In the middle are the countries of Central Asia plus Moldova, but not Kazakhstan, where death rates are second only to Russia. These countries display mostly low rates of alcohol use, but Kyrgyzstan and Tajikistan display high rates of opiate use and IDU. That Kazakhstan appears to have a similar pattern of AOD abuse to Kyrgyzstan, but almost double the youth mortality rate, suggests some important differences between these two countries (see Section ‘Aetiology of alcohol and other drug use’). While these estimates are very approximate, they nonetheless tell a fairly consistent story: the correlation coefficients in the Table show that all AOD-use indicators are positively correlated with youth mortality rates.

The patterns are not completely consistent. Data sources for Moldova give contradictory accounts of AOD use. Estimates of low use in Uzbekistan, and even more so in Turkmenistan, are inconsistent with their relatively high youth mortality, suggesting either that AOD abuse is not the only factor contributing to youth deaths here, or that AOD abuse is underestimated. Both suggestions are reasonable. Turkmenistan in particular continues to be ‘closed’ to researchers, and official information and statistics are often unreliable.

Evidence from rich countries suggests that a high proportion of deaths among young people can be either directly or indirectly attributed to AOD use. Ridolfo and Stevenson estimated that in Australia in 1998, 19% of deaths among people aged 15–34 years were directly or indirectly attributable to alcohol, and 15% to abuse of illicit drugs (Ridolfo & Stevenson, 2001, Table 6.3). EMCDDA (2006) suggested that illicit drugs were a significant contributory factor in the deaths of 10–23% of 15–49 year olds in six Western European cities. Degenhardt, Hall, and Warner-Smith (2006) concluded that the participants in cohort studies of injecting drug users in rich countries experienced an annual death rate from all causes of 1.08% before taking into account the impact of HIV, and 1.36% including HIV-related deaths.

There has been no comprehensive attempt to estimate the proportion of youth deaths attributable to AOD abuse in CIS countries. Studies of mortality associated with abuse of illicit drugs or IDU are lacking (Poznyak, Pelipas, Vievski, & Miroshnichenko, 2002). Nonetheless, the evidence cited above suggests that the number of deaths attributable to AOD abuse among young people in CIS countries could be high. Applying Degenhardt et al.'s (2006) non-AIDS mortality rate (1.08%) to the estimates of injecting drug users in CIS countries (Table 2) would suggest that IDU accounts for three in ten deaths among 15–29 year olds in Russia, and two in ten in Kyrgyzstan. These are likely to be underestimates, given the relatively poor state of public health care in both these countries. The fact that methadone treatment for heroin users is banned in most CIS countries is unlikely to help (Langendam, van Brussel, Coutinho, & van Ameijden, 2001). In addition to these deaths, alcohol may be associated with as many as a third of male deaths in Russia, Kazakhstan and Ukraine (Rehm & Gmel, 2002).

In contrast, the low rate of AOD use in Armenia and Azerbaijan suggests that AOD-related deaths are likely to constitute a relatively low proportion of the low total. This begs two important questions. First, what factors lie behind the dispersion of rates of AOD use across CIS countries, and second, what protective and risk factors have influenced the relationship between AOD abuse and youth mortality?

Aetiology of alcohol and other drug use 

Historically, drug prevention interventions have focussed largely on drug-specific and individual-level risk factors, called ‘proximal causes’ in this analysis. A review by Birckmayer et al. (2004) captured the risk factors of usual concern for AOD policy and interventions. They described research demonstrating that AOD abuse is influenced by availability, norms, and promotion, with each in turn influenced by enforcement practices. Availability includes economic availability (price), retail availability (e.g. liquor outlet density, minimum drinking age, access to illicit drug markets), and social availability (access via friends rather than retailers/dealers). Norms are informal (group and community norms) and formal (e.g. laws, policies). Promotions include advertisements and word-of-mouth. Individual-level risk factors operate in the context of the drug environment and include hereditary predisposition, level of strain, family dysfunction, low religiosity and school failure.

Such research suggests drug prevention mechanisms focusing on drug-specific environmental risk factors and individual risk factors. Environmental interventions can involve enforcement (e.g. random breath testing or imposing sanctions on users of illicit drugs), controls on availability (e.g. retail restrictions), efforts to change norms (e.g. mass media campaigns), and limits on promotion (e.g. alcohol advertising restrictions). Interventions to change individual risk factors include programmes that focus on social competency, improved academic performance, or family functioning.

Recently, some of the literature on the aetiology of drug use has adopted a broader approach, taking into account not only proximal causes, but also the social conditions that contribute to the proximal risk factors identified above (Rhodes, 2002; Spooner & Hetherington, 2005; Thomas, 2007). Spooner and Hetherington (2005) identified a range of factors in the social environment that might affect an individual's exposure to risk and protective factors, and susceptibility to the consequences of AOD use. These include economic factors (e.g. inequality in income and opportunity, job insecurity, concentrated disadvantage); social factors (e.g. social capital); the physical environment (e.g. community disorder, availability of transport, design of public spaces); cultural factors (e.g. levels of communitarianism, individualism, materialism and secularism); and policy factors (e.g. relating to income redistribution, welfare and education). Further, these researchers found that a developmental perspective is needed to understand the origins of AOD abuse. From conception to adulthood, humans must achieve each developmental task in order to proceed successfully to the next phase. Interruptions to youth development can have long-lasting impacts on youth resilience, resulting in a range of problem behaviours including AOD abuse. Friedman, Rossi, and Flom (2007) argued that “big events” such as wars, revolutions, or the economic and social turmoil that followed the collapse of the Soviet Union, could cause people to engage in risk-taking behaviour. Families are particularly important for shaping child and youth development, and the social and policy environment impacts upon parental capacity.

Spooner and Hetherington (2005) have acknowledged the difficulty of proving causal links to AOD use as the risk factors become more distal. However, the co-occurrence in wealthy countries of increases in drug use with social changes over the past half century (e.g. increase in working parents, reduced social capital) suggest associations between societal trends and health and social problems, and a need to consider how social structures (policies and programmes) can build resilience and reduce problems.

Both proximal and distal models have already been applied, albeit for the most part implicitly, to analysing AOD abuse in CIS countries, and particularly in Russia. From the ‘proximal causes’ perspective, Nemtsov (2005) emphasised the cultural acceptability, not only of alcohol use, but also of inebriation in the workplace. Pilkington (2004) showed that peer groups are a key influence on illicit drug consumption in three Russian regions. Pomerleau et al. (2005) argued that cultural norms are a strong influence, not only on overall alcohol consumption, but also on types of alcohol consumed, in different CIS countries. Rhodes et al. (2003) examined micro-environments influencing drug-injecting, risk-reduction and syringe-exchange practices among injecting drug users in a Russian city.

A considerable literature has also focused on the ‘distal causes’. Excessive alcohol consumption in Eastern Europe was found to be associated with increased mortality, the trigger being psychosocial stress (Cornia & Pannicià, 1996). Such stress arises in response to new and unexpected situations for which people do not know the appropriate coping behaviours; and stress-related mortality is associated with family breakdown, poverty, job insecurity, unemployment and migration, as well as emotional states such as anger, depression and hopelessness (p. 121). Cornia and Pannicià showed a positive relationship between an ‘economic stress index’ comprising changes in average real wages, consumer prices and employment, and mortality, across several post-communist countries over the early 1990s.

A sizeable literature has built on this early work and has increasingly focused on AOD abuse (particularly alcohol consumption) as the key link between social stress and mortality among adults in CIS countries. Rhodes et al. (1999) drew a connection between rapidly changing economic, welfare and health conditions in post-communist countries, and increases in IDU and susceptibility to HIV, citing many of the same indicators used by Cornia and Pannicià (1996). Walberg, McKee, Shkolnikov, Chenet, and Leon (1998) showed that areas of Russia where male life expectancy decreased the most experienced the greatest increases in income inequality and crime. Following Cornia and Pannicià, they argued that absolute levels of economic and social well-being are less significant than changes in well-being, and that the biggest increases in mortality occurred in some of the wealthiest regions of Russia. Pridemore (2002) and McKee and Leon (2005) used Durkheim's concept of anomie to analyse the association between rapid social change and AOD abuse in Transition countries:

“… recognition of the roles of alcohol and heroin are only the first step; what is needed is to understand why so many people seek solace in substances that will eventually kill them. This takes us back to Durkheim's analysis of the process of industrialization in western Europe. In both examples we can identify the role played by both economic and domestic anomie, with those affected most being those exposed to the most rapid change but with the fewest social and educational resources to help them cope.” (McKee & Leon, 2005, p. 1207).

Rhodes and Simic (2005) developed a model of the HIV risk environment in the context of Transition that directly relates to the literature on which Spooner and Hetherington (2005) and (to a lesser extent) Birckmayer et al. (2004) built their frameworks. They identified four domains of the micro and macro risk environments for risk behaviours: physical, social, economic and policy. The micro-environment includes drug-injecting and sex-work sites, peer and social norms, informal local economies, and policies for the distribution of syringes and condoms. The macro environment includes drug trade and trafficking routes, social and cultural norms, gender and social inequalities, and laws governing trade, trafficking and migration. While similar to the work of Cornia and Pannicià (1996) and McKee and Leon (2005), Rhodes and Simic placed more emphasis on the current situation, and less on change over time. In doing so, they implicitly identified a weakness in applying the literature on the developed countries to the Transition context: the developed country literature tends not to focus on economic and social change as an explanation for AOD abuse and concomitant risks. The Transition literature on the other hand focuses almost entirely on change, which fits well with Durkheim's conception of anomie as an explanator of increases in social stress, AOD abuse and mortality as responses to change and uncertainty. In summary, research suggests that a range of proximal and distal factors might be contributing to the increase in AOD abuse in the CIS.

Factors contributing to alcohol and other drug use and deaths 

This section examines information on proximal and distal risk factors for youth AOD abuse as identified by the two models discussed in Section ‘Aetiology of alcohol and other drug use’. To simplify the discussion, most attention is given to three CIS countries representing the three regions of the CIS and exhibiting different levels of AOD abuse and youth deaths – Russia (high mortality, Western CIS); Kyrgyzstan (mid-range mortality, Central Asia); and Armenia (low mortality, Caucasus) – although evidence for other countries in the region is also used. Proximal environmental risk factors for AOD abuse are discussed first, followed by social and economic conditions and wider societal changes that are likely to have influenced youth development, AOD abuse and AOD deaths.

Availability, promotion and acceptability of opiates 

The use of opioids in Russia and Kyrgyzstan is high by global standards, but appears to have remained low in Armenia. Across the CIS, IDU is socially and officially disapproved. Official responses tend towards sanctions (Burris et al., 2004), and there is little public investment in treatment facilities for users of illegal drugs. Where they exist, they tend to be harsh, reflecting social disapproval of, and a lack of compassion for, illicit drug users.

The main proximal factor that distinguishes the high-use and low-use countries is availability. Fig. 1 plots the relative price of heroin per gram and per capita gross domestic product (GDP) in the early 2000s for eight CIS countries. Heroin is cheap in Kyrgyzstan (about $2–$4 per gram) and expensive in Armenia (over $100 per gram). It costs about $20 per gram in Kazakhstan and $40 in Russia, but GDP per capita in these countries is higher than other CIS countries, making heroin affordable to many. Over the past decade the countries of Central Asia and the Western CIS have become the main route for transportation of heroin from Afghanistan, source of nearly nine-tenths of the world's total supply, to Western Europe (UNODC, 2006). Godinho et al. (2005) estimated that trafficking of drugs, mainly heroin, contributes between a third and half of Tajikistan's national income. A large portion of the supply is distributed and consumed along the way. Weak and ineffective governments, characteristic of some Central Asian countries such as Kyrgyzstan and Tajikistan, do nothing to block this supply, not least because “rulers do not have much power other than in their respective capitals” (Engvall, 2006, p. 833). Paoli (2001) reported that the supply of opiates throughout Russia is organised principally by migrants from Central Asian and Caucasian countries, millions of whom have gone there looking for work since the early 1990s. Many do not have proper work visas, and are therefore often engaged in informal and illegal activities, including drug trafficking and selling. On the other hand, some anecdotal evidence suggests that the role of these migrants in the illicit drug trade is greatly exaggerated by Russian authorities for political purposes. Moreover, drug supply is well diffused throughout Russia, and not confined to trafficking routes or the major cities (Pilkington, 2004)

View Large Image Download to PowerPoint

Fig. 1. Price of heroin and GDP per capita, 2002. Source: UNODC (2006); UNICEF TransMONEE Database. Notes. The horizontal axis shows GDP per capita in US dollars per annum at market exchange rates. The vertical axis shows the estimated cost of a gram of heroin in US dollars at market exchange rates.

Socio-economic environment 

The pace of change after 1992, with large fluctuations in GDP, wages and living standards, rising corruption, growing inequality, fewer employment opportunities and decreased quality of schooling, is likely to have influenced levels of social stress, AOD abuse, and adult mortality across most CIS countries (Cornia & Pannicià, 1996; Friedman et al., 2007; Leon & Shkolnikov, 1998; McKee & Leon, 2005; Pridemore, 2002, Walberg et al., 1998). Building on this literature, we suggest that variations in strength of family and community structures may be one important driver of the diverging trends in youth deaths associated with AOD abuse across CIS countries.

Countries such as Russia have high levels of AOD abuse and resulting youth mortality, not only because AOD are available, but also because these are more urbanised and industrialised societies, where bonds of family and community are generally weaker than in more traditional or agrarian societies. These bonds may have been further weakened by the rapid pace of socio-economic change, with little of the cushioning (however inadequate) that the institutions of a modern welfare state might provide to individuals in distress.

Many indicators suggest weaker family structures in Russia and other Western CIS countries, compared with Central Asian and Caucasus countries. Table 3 shows evidence on attitudes regarding attachment to families and friends taken from the World Values Survey, a wide-ranging survey undertaken in about 60 countries worldwide using nationally representative sampling methods. Responses to the four attitudes show significant differences between Russia and other high mortality countries on the one hand, and the countries of Central Asia, Caucasus and Moldova on the other. This is most notable in responses to the statement “One of my main goals in life has been to make my parents proud”; 15% of respondents in the high mortality countries agree with this statement, compared with 40% of respondents in the low-medium mortality countries. These differences are as strong, or stronger, for young people as they are for overall samples.

Table 3. Family values in CIS countries, 1990s (percent agreeing with statements)

		“Family is very important in life”	“Friends are important in life”	“One of my main goals in life has been to make my parents proud”	“Parents should do their best for their children”
	Russia 1999	75.0	27.3	14.8	53.9
	Ukraine 1999	81.4	38.4	12.9a	60.7
	Belarus 2000	77.7	26.7	16.6a
	Average percent	78.0	30.8	14.8	57.3
	Kyrgyzstan 2003	87.2	47.5	55.3	66.8
	Moldova 2002	85.0	34.9	31.6	73.4
	Azerbaijan 1997	84.7	35.2	37.1	57.7
	Georgia 1996	94.6	73.6	37.3	75.0
	Armenia 1997	86.0	44.4	38.6
	Average per cent	87.5	47.1	40.0	68.2

Source: World values surveys, various years. Notes. sample sizes for Russia: 2500; Ukraine: 1195; Belarus: 1000; Kyrgyzstan: 1043; Moldova: 1008; Azerbaijan: 2002; Georgia: 2008; Armenia: 2000.

a Data are for 1996.

Aggregate social statistics also provide evidence on differences in community and family cohesion between the high mortality countries and other countries in the CIS. For example, divorce rates have not only been historically higher in Russia and other Western CIS countries than in the Caucasus and Central Asia, but differences between Western CIS and the other countries have tended to increase since Transition (UNICEF, 2007). There are more single-parent families, as a proportion of all families, in Western CIS countries than in Caucasus and Central Asian countries, and more children in Western CIS countries are likely to be born out of wedlock (Menchini & Redmond, 2006; UNICEF, 2006b). Children in Western CIS countries are also more likely than children in most Caucasus and Central Asian countries to be taken into public care, through fostering or institutionalisation. About 1.5% of children in Russia were living in public institutions in 2004, compared with about 0.25% of children in Armenia and Georgia (UNICEF, 2007).

While economic change, poverty and the effects of corruption arguably hit households in Caucasus and Central Asian countries as hard as, or harder than households in Russia or Ukraine, the impact of rapid socio-economic change on AOD abuse and youth mortality may have been to some extent modulated by family and community strength – support that young people could rely on in times of difficulty. This is illustrated in Fig. 2, which shows data on the relationship between youth mortality and crude divorce rates for Russia, Kyrgyzstan and Armenia as representatives of high, medium and low mortality countries, and Kazakhstan as a high mortality country with a similar overall pattern of AOD abuse as Kyrgyzstan (see Table 2 above). The Figure shows not only that Kazakhstan and Russia have higher levels of divorce than Kyrgyzstan and Armenia, but also that the year-to-year relationship between divorce rates and youth mortality rates is stronger in the former two countries than in the latter two. This contrast in terms of social indicators between the high mortality and the low mortality countries requires more systematic study. Nonetheless, a similar differentiation between low mortality and the high mortality countries is evident in a number of other social indicators such as reported crime rates, the proportion of births out of wedlock, and the number of children in public care institutions.

View Large Image Download to PowerPoint

Fig. 2. The relationship between crude divorce rate and youth mortality, Armenia, Kyrgyzstan and Russia, 1990–2004. Source: WHO Mortality Database and UNICEF TransMONEE Database. Note. Mortality data are: mortality rate among 15–29 year olds (per 100,000 relevant population). Divorce data are: crude divorce rate per 1000 population. Each country is represented by separate series, and each point represents the relationship between the crude divorce rate and the youth mortality rate for one year in the period 1990–2004. Mortality data are matched to divorce data with a 1-year lag.

Qualitative literature has tended to contrast the importance of family and community in Caucasus and Central Asian countries with the more individualist or nuclear familist cultures that are more notable in Russia and its Western CIS neighbours (Harris, 2005; Kandiyoti & Azimova, 2004; Kurkchiyan, 2005). Kurkchiyan argued that in the case of Armenia, an informal network of kinship, friendship and mutual exchange of favours dominates everyday life. “These three social bonds together form a web of relationships that ties everyone together” (Kurkchiyan, 2005, pp. 211–212). Stachowiak et al. (2006) found that ethnic Tajik injecting drug users in Dushanbe, the capital of Tajikistan, were twice as likely to seek treatment as ethnic Russians, possibly because they were more likely to be put there by their families. Patterns of family and community strength in CIS countries remain under-researched, but the evidence presented here is not inconsistent with the proposition that this may be one explanatory factor for differences in AOD abuse, and resulting mortality among young people across CIS countries.

Discussion 

The proximal-causes model of AOD abuse explains well the variation in levels of consumption of both alcohol and opiates across CIS countries. While alcohol is freely available across the CIS, its consumption is particularly problematic in the ‘wet’ cultures of Russia and its neighbours (Nemtsov, 2005). Abuse of opiates, on the other hand appears to be driven in part by price, which may explain the very high levels of use in Central Asian countries such as Kyrgyzstan (where it is cheap), and the very low levels of abuse in Caucasus countries such as Armenia (where it is expensive). The relationship between availability, cultural norms and levels of abuse suggests the need for addressing norms around alcohol intoxication, alcohol availability and alcohol promotion in high consumption countries such as Russia; and addressing availability of opiates in countries of the Western CIS and Central Asia. These are no easy tasks, given the histories of these countries, the extent to which alcohol in particular is socially accepted, and the levels of corruption that can stymie efforts to address illicit drug supply (Engvall, 2006, Nemtsov, 2005, Paoli, 2001).

Following the distal causes model, we suggest that tackling availability is unlikely to either completely solve problems of AOD abuse in CIS countries, or to significantly reduce youth mortality rates, unless effort is also directed towards strengthening the role of families and communities in support for AOD abusers. The research that both Birckmayer et al. (2004) and Spooner and Hetherington (2005) drew upon suggests that parenting practices are crucial determinants in child and youth development, which in turn influences AOD use behaviours by young people. Young people who are detached from family and community support networks, or socially excluded in other ways, are more likely to abuse AOD and more likely to suffer deleterious consequences from AOD. People who live in communities with few social and health services (as is typical in CIS countries) may need to rely upon families to provide care and support when AOD problems arise. If families are separated or otherwise distressed, this support may be less available and the consequences of AOD abuse are more likely to be fatal.

The focus on ‘weak families’ indicators underlines what we believe to be an under-explored difference between Russia, Kazakhstan and other ‘high youth mortality’ Western CIS countries on the one hand, and remaining countries of Central Asia and the Caucasus on the other where youth mortality rates are lower. The strength of community, family and kinship bonds, and submission to parental authority is stressed throughout the literature on the latter countries (Harris, 2005, Kandiyoti and Azimova, 2004; Kurkchiyan, 2005). Writing of the mahalla system of community councils in Uzbekistan, Kandiyoti and Azimova (2004, p. 328) stated that “… it also enforces powerful communal norms and acts as an agency of socialization and mutual help.” It would clearly be mistaken to suggest that family and kinship bonds offer no support in countries such as Russia, Ukraine and Kazakhstan. Nonetheless, a number of writers point to the contrast between a more individualist or familist approach similar to that in Western countries inherent in Russian culture, as opposed to the wider communalism that are hallmarks of Caucasian and Central Asian cultures (e.g. see Harris, 2005). This communalism may not always prevent young people from abusing AOD, but it may act as a protective factor from the most serious risks associated with AOD abuse (Stachowiak et al., 2006).

The proximal and distal models of risk factors for AOD abuse and its consequences point towards quite different policy conclusions. For example, while both models highlight family situation as a crucial influence on AOD abuse, their arguments flow in somewhat different directions. Proximal models propose interventions for individuals and specific families (e.g. skills training programmes). Distal models suggest the need for systemic interventions to create environments that will support all individuals and families (e.g. universally available programmes to address youth unemployment and supports for sole-parent families). Further, interventions to create better opportunities might be most beneficial in those countries where social distress, AOD abuse and youth mortality rates are the highest. It seems unlikely that drug-specific policies alone will provide the answers in these countries. Our argument for creating health supporting environments is not new, but is consistent with principles of health promotion (World Health Organization, 1986) and recent research on human development and social determinants of health (Eckersley, Dixon, & Douglas, 2001; Spooner & Hetherington, 2005).

In arguing that there is enough evidence to warrant a more comprehensive examination of the link between family and community strength, AOD abuse, and youth mortality in CIS countries, we acknowledge the considerable evidence gaps that that this analysis has had to confront. Information on the level of AOD abuse in the region has improved greatly in recent years, and we use much of this recent evidence in this paper. However, epidemiological studies that trace the role of AOD abuse in mortality are needed. In particular, studies of mortality associated with opiate abuse are lacking. Second, we need more comprehensive evidence on how family and community act to protect vulnerable people who abuse AOD. Our argument in this regard has also been proposed by others, but has not been strongly tested. New data on AOD abusers and their families could greatly assist policymakers in understanding where interventions might be most effective in different countries of the CIS.

Finally, there may be something to be learned from the very low mortality countries in the region. Is it simply strong families and communities, coupled with distance from illicit drug markets that protect young people in countries such as Armenia from high levels of AOD abuse and mortality? More research on low AOD abuse and low mortality countries could reveal much that is useful for countries with higher levels of both. Research of this type is urgently needed, and the youth mortality crisis in the region needs to be moved higher up the policy agenda.