Parental transmission of smoking among middle-aged and older populations in Russia and Belarus

According to the World Health Organization’s (WHO) estimates, post-Soviet countries, including Russia and Belarus, have some of the highest prevalence of smoking among adult men anywhere in the world, while rates have been increasing rapidly among women in the past 2 decades (McKee et al. 1998; Perlman et al. 2007; Roberts et al. 2012; Quirmbach and Gerry 2016; WHO 2016). The previous research has studied the macro-level effects of economic transition and market entry of transnational tobacco companies on these patterns (Gilmore and McKee 2005; Gilmore et al. 2011; Lillard and Dorofeeva 2015), as well as individual-level factors such as age, marital status, family disruption, education, urban residency, household economic situation, self-reported material deprivation, unemployment, occupation, and religious denomination (Gilmore et al. 2001; Pomerleau et al. 2004; Bobak et al. 2006; Perlman et al. 2007; Kislitsyna et al. 2010; Roberts et al. 2012). One area that has largely been overlooked in this research has been intergenerational transmission of propensity to smoke from parents to their offspring in the post-Soviet context.

Studies on social determinants of smoking have repeatedly demonstrated intergenerational transmission of smoking in various settings, populations, and times (Bailey et al. 1993; McGee et al. 2006; Brook et al. 2013; Christopoulou et al. 2013; Kandel et al. 2015). Nonetheless, one major limitation of this literature is that it is almost exclusively concerned with intergenerational transmission of smoking to adolescents and individuals in their 20s and 30s, while the role of parental smoking later in life is under-researched. The available research also suggests that, if they are to identify the net intergenerational transmission of smoking, empirical studies should adequately account for socio-economic status in each generation, which independently affects tobacco use (Conrad et al. 1992; Jefferis et al. 2003; Schori et al. 2014). Lower parental education and material deprivation in childhood, for instance, predicts an individuals’ increased risk of smoking in their adulthood and may explain spurious evidence of an intergenerational transmission of smoking (Fagan et al. 2005). Few existing studies account for social origin variables separately for fathers and mothers and so might overestimate the effect of parental smoking on their children’s smoking propensity.

The very high rates of smoking among men and the rapid changes among women in Russia and Belarus mean that these countries offer an opportunity to understand better the intergenerational role of parental influences on smoking. We are aware of only two studies in post-Soviet settings that investigate the link between parents and their offspring smoking (Kemppainen et al. 2006; Kislitsyna et al. 2010). These are limited in that they only include adolescents, cover limited geographical areas with small samples, do not account for effects with social origins, such as fathers’ and mothers’ education, and do not consider a wide array of confounding factors in an individual’s life, such as labour market characteristics, that might explain intergenerational transmission of smoking between parents and their children. The goal of the study is to exploit a unique data set from Russia and Belarus that can address many of the limitations of previous research by examining the association between fathers’ and mothers’ smoking and smoking in their middle-aged and older offspring, after taking account of those offspring’s social origin, socio-demographic, and socio-economic characteristics.

In this article, we examined intergenerational transmission of smoking among middle-aged and older populations in Russia and Belarus using the newly available data from the PrivMort retrospective cohort study. This data set is a unique source for understanding the association of parental smoking behaviour on their offspring’s tobacco use in post-communist countries. The validity of our analysis is strengthened by the fact that the prevalence of smoking in our sample for men, 53.3 and 51.9%, respectively, in Russia and Belarus, comes close to the latest available estimates for adult populations derived from the Russia Longitudinal Monitoring Survey of RLMS-HSE for 2014 (49.0%) (Quirmbach and Gerry 2016) and the Health in Times of Transition Study (HITT) for Belarus in 2010 (42.8%) (Roberts et al. 2012). The main reason why the prevalence of current smoking among women in our data set is lower than that reported in the above-mentioned nationally representative surveys (9.4 vs. 14.4% for Russia and 8.3 vs. 13.2% for Belarus) is that the PrivMort does not include groups aged 41 and below who have the highest prevalence of smoking.

Adjusting for social origin, socio-demographic, and socio-economic variables, we find that paternal smoking significantly increases the chances of smoking among sons, while having a regularly smoking mother significantly increases the chances of smoking among daughters. In absolute terms, this equates to about 9–12 percentage points higher prevalence of smoking in Russia and Belarus among women with smoking mothers in comparison to women with never smoking mothers, which is remarkable, considering that in our sample of middle-aged and older women, the rate of smoking in Russia and Belarus is less than 10%. These results are in line both with psycho-analytic theory (Boyd 1989) which claims that daughters tend to unconsciously internalise maternal values and behaviours as well as with social learning theory (Bandura 1977) that emphasises principles of intergenerational modelling and suggests that girls are consistently and positively reinforced when they learn to be like their mothers and imitate maternal behaviour. Daughters also tend to spend significantly more time with their mothers than sons do (Kislitsyna et al. 2010). Our results complement the earlier findings on the parental transmission of smoking that do not usually emphasise the varying roles of paternal and maternal tobacco use in middle-aged and older sons and daughters smoking behaviour (e.g., Brenner and Scharrer 1996; Bricker et al. 2006; Chassin et al. 1998; Kandel and Wu 1995; Melchior et al. 2010).

Our study has a number of limitations. Due to the survey design, we do not have information on smoking for those parents who died before 1982. Therefore, information on maternal smoking had to be imputed for 10 and 7% of male and 12 and 7% of female respondents, respectively, in Russia and Belarus. Furthermore, the data set is not a nationally representative survey of Russia and Belarus and the findings cannot be generalised to these countries’ entire populations. Another limitation of this study is recall bias and measurement error that can stem from asking questions about circumstances and events related to respondents’ parents and their childhood. Such misclassification can lead to misestimating of the strengths of associations between parental and offspring smoking. The design of the questionnaire we use, however, mitigates this limitation by introducing auxiliary sentences and incorporating active visualisation memory cards that can assist people in remembering various characteristics of their parents’ and their own childhood more easily.

Our findings not only contribute to the existing scholarship on intergenerational transmission of socio-economic disadvantage in post-communist contexts (Gugushvili 2015, 2016, 2017a, b), but are also relevant in interpreting trends and patterns of smoking among middle-aged and older Russian and Belarusian men and women, with implications for subsequent generations, an issue of importance given increasing smoking among women. Of course, we cannot assume that the strength of intergenerational transmission of smoking will be the same in the future, given the many other factors involved such as shrinking gender differences in smoking. Nonetheless, if parents are aware of the implications of smoking and how much they can influence their sons and daughters’ propensity to smoke even in the later stages of their life course, some might have an extra reason not to start smoking or to quit after smoking initiation (Bricker et al. 2006; Roberts et al. 2013).

However, the main implication of our results for research on prevention and control of tobacco use is that intergenerational transmission of smoking does not seem to be a methodological artefact of inadequately accounting for social origin variables and the many potentially confounding factors throughout individuals’ lives. Parental influence on smoking is a significant factor not only for adolescents but also for middle-aged and older populations in the considered post-communist countries. Therefore, to understand thoroughly the confounding factors of smoking in the region and, arguably, beyond, the existing longitudinal and cross-sectional surveys, such as Global Adult Tobacco Survey (GATS) and RLMS-HSE, should explicitly enquire into smoking behaviours of respondents’ parents. Intergenerational influences on smoking should also be taken into account of in studies seeking to monitor rates of smoking and the impact of tobacco control programmes.