Theories

Whilst the dramatic fall in life expectancy and its association with cardiovascular disease and with accidental death and poisoning are well known, what is less well known are the causes for this in a peace time situation in the former Soviet countries (Eberstadt 1999). Alcohol and tobacco consumption are important but given that consumption of these two have always been high in the Soviet Union, why the dramatic fall in health over the last ten years? Explanations put forward by Russian authors tend to emphasise political and economic problems such as widening socio-economic inequalities (people cannot afford good food and medicaments) rising unemployment or the deterioration of the health services (Breev 1998, Shevaldina 1997). Others point to high levels of pollution or the particular climactic conditions in some parts of Russia, especially the North (Morozova 1994, Potapov, Ustyushin and Shushkova 1994, Feshbach and Friendly 1992). Others point to psycho-social factors such as depression and lack of control over one's life and inability to have or to fulfil life plans (Shilova 1994, Bobak et al 1998). Yet others are more likely to emphasise lifestyle habits such as heavy use of alcohol and tobacco, diet and especially binge drinking (Cockerham 1999, Eberstadt 1999). However, recent studies have emphasized that it is the way these things are combined through lifestyle, culture and "habitus" which is important (Cockerham 1999). This would also reflect current thinking about health inequalities and society more generally (Wilkinson 1999). This study would put together a range of sources of information at an individual and aggregate level in order to better understand the relative importance of these different factors and provide a better explanation of health outcomes.

Most studies have concentrated upon one or two countries, and the main concern has been with Russia. This study, by contrast, would compare several different post-Soviet countries in order to understand how ethnic, cultural, religious and other variations may have different health outcomes. By creating a pooled, multi-national data set, it would be possible to compare between country as well as within country differences according to socio-economic, demographic (gender, age),regional and psycho-social factors as well as behaviour. The project will be innovative in creating survey data with a variety of measures of life-styles. Instead of assuming that drinking or smoking is in itself a sufficient indicator of life-style characteristics, multivariate statistics will be used to determine whether or not it is part of a syndrome of attributes which are associated with below or above average health in different population, gender or lifestyle groups.

Most studies have tried to explain the deterioration in health. Yet this has not affected everyone in the same way. Women are not affected to the same extent as men. Indeed, in some of the countries that we are considering, health may have actually improved (see Table 1 page 6) and some have attributed this to widening ethnic differences leading Southern countries to leave vodka in favour of wine (Gerner 1995). Increased national pride in those countries who sought liberation in contrast to the general sense of post-imperial defeat in countries like Russia may play some part in this (Field 1995). This study would consider differences between countries and their association with other political and social indicators. It would consider what prevents bad health as well as what causes it.

While studies of the prevalence of smoking and drinking have been conducted in the countries concerned, this research will be distinctive as it will draw on recent insights into the relationship between risk factors and health and on new developments in international health policy. There is now an extensive body of evidence linking changes in alcohol consumption in the countries of the former Soviet Union to the large fluctuations in mortality since the mid 1980s (Leon et al, 1997). The apparent relationship between changes in alcohol consumption and cardiovascular deaths is, however, much more problematic and some have claimed that alcohol can even be cardioprotective. The vast majority of research on the link between cardiovascular disease and alcohol had failed to take account of the pattern of drinking and that, in those few studies that had done so, there was a clearly increased risk of, frequently sudden, cardiac death among those who drank in binges (McKee et al, 1998, Chenet et al, 1998a, McKee & Britton, 1998, Bobak et al, 1999). Although the pattern of drinking is important in determining mortality, the social context of drinking also plays a part. To take the example of death from injury, it is not sufficient for someone to get drunk. They must do so in a setting in which they are exposed to hazardous circumstances and where protective mechanisms are absent. There is a need to understand these issues much better. Finally, it is apparent that there are major socio-economic and gender differences in deaths from alcohol-related causes in the former Soviet Union although the mechanisms underlying these differences are not yet fully elucidated (Chenet et al, 1998b).They are, however, likely to reflect a complex mixture of health related knowledge, attitudes and practices - in other words, lifestyle.

New research is now required that will go beyond the traditional questions about weekly consumption levels at a population level and which will examine pattern of drinking, the social context of drinking (including the extent and nature of problem drinking), and the social determinants of knowledge and attitudes, as well as behaviour. Tobacco is also an important determinant of the high burden of disease in the former Soviet Union (Pudule et al, 1999, Shkolnikov et al, 1999). Research on tobacco consumption is especially timely because of the work by the World Health Organisation to develop a Framework Convention on Tobacco Control (WHO, 1998). Action to reduce smoking will require information on knowledge and attitudes as well as on practices which form part of living conditions and lifestyles more generally.

A major theoretical innovation is bridging the divide between epidemiological and sociological research that draws inferences about individuals from data about national populations and social structure, as against clinical researchers and micro-economists, who focus on individual-level attributes such as drinking or income. The project will analyse individuals in social contexts using applications of social capital which has been shown to influence physical and emotional health (cf. Kennedy et al., 1998; Bobak 1998). Its conclusions will be more robust because it will use multivariate statistics to test the combined influence of micro-level, network and large-scale contexts as well as health beliefs and behaviour collected at an individual level. It would therefore be able to identify which of the many factors described here might be most important in explaining health outcomes and thus offer better scope for intervention.