Association of alcohol control policies with adolescent alcohol consumption and with social inequality in adolescent alcohol consumption_ A multilevel study in 33 countries and regions

Background Previous research found inconsistent associations between alcohol control policies and socioeconomic inequality with adolescent drinking outcomes. This study expands the focus beyond individual associations to examine whether a combination of policies is related to socioeconomic inequality in adolescent drinking outcomes and whether this relationship varies across survey years. Methods Multilevel modelling of 4 waves of repeat cross-sectional survey data (2001/02, 2005/06, 2009/10, and 2013/14) from the Health Behaviour in School-aged Children (HBSC) study was carried out. The sample was composed of 671,084 adolescents (51% girls) aged 11, 13, and 15 (mean age=13.58; SD=1.65) from 33 European and North American countries/regions. The dependent variables were lifetime alcohol consumption, weekly alcohol consumption, and lifetime drunkenness. Independent variables were of three types: individuallevel variables (age, sex, Family Affluence Scale, and the Perceived Family Wealth), time-level variable (survey year), and context-level variables (minimum legal drinking age, physical availability, advertising restrictions, a total alcohol policy index, and affordability of alcohol). Results The total alcohol policy index showed a negative relationship with both lifetime and weekly consumption. Higher affordability of alcohol was related to higher lifetime and weekly consumption and higher lifetime drunkenness. Family Affluence Scale was positively related to all three alcohol measures and Perceived Family Wealth was negatively related to lifetime drunkenness, with these associations increasing across survey years. The total alcohol policy index buffered the associations of Family Affluence Scale and Perceived Family Wealth with adolescent drinking outcomes. Conclusion A combination of alcohol control policies is more effective in reducing adolescent drinking outcomes than single policy measures. Reducing the affordability of alcohol stood out as the most successful single measure. Socioeconomic inequalities (i.e. higher alcohol consumption and drunkenness in adolescents with higher family affluence and higher drunkenness in adolescents perceiving their families to be poor) have persisted and even increased across survey years. A combined alcohol control policy can help in tackling them. The harmful use of alcohol is one of the most important risk factors for population health worldwide, causing more than 200 disease and injury conditions, and being responsible for 3 million deaths every year (5.3% of all deaths) (World Health Organization, 2018). Special attention needs to be paid to adolescent drinking. First, consuming alcohol in adolescence has been shown to be related to significant differences in brain structure and functioning (Feldstein-Ewing, Sakhardande & Blakemore, 2014) as well as to different physical and mental health https://doi.org/10.1016/j.drugpo.2020.102854 ⁎ Corresponding author. E-mail address: eva.leal-lopez@hbsc.org (E. Leal-López). International Journal of Drug Policy 84 (2020) 102854 0955-3959/ © 2020 Published by Elsevier B.V. T problems, and other risk behaviours such as delinquency and sexual risktaking behaviour (Lavikainen, Salmi, Aaltonen & Lintonen, 2011; Newbury-Birch et al., 2009). Second, an association between early initiation and alcohol use disorders in adulthood has been found (Waller, Murray, Shaw, Forbes & Hyde, 2018). Marked decreases in adolescent alcohol consumption have been observed across many countries in recent years, including Europe (Inchley et al., 2018, with HBSC data) and the USA (Miech et al., 2018). However, prevalence still remains higher than desired owing to its adverse impact on adolescent development and future health. Given the severity of the situation, a decrease of 10% in the volume of alcohol use by 2025 was established by the World Health Organization (WHO) as one of nine voluntary targets for non-communicable diseases. However, there is no international public health treaty on alcohol, and policy initiatives are recommended only in general terms. In an attempt to address the problem, a number of policy measures have been implemented by national governments. These policy initiatives can be divided into three major groups: restricting alcohol availability, regulating alcohol advertising, and controlling alcohol pricing. The most commonly used measure to restrict alcohol availability is to impose a minimum legal drinking age (MLDA). Evidence suggests this can have a positive impact on public health outcomes such as a decrease in alcohol-related traffic accidents (Wagenaar & Toomey, 2002) and reduced mortality and morbidity rates in young people (Zhang & Caine, 2011). However, mixed results have been found regarding alcohol consumption. While most studies concluded that MLDA was related to decreases in adolescent drinking (Subbaraman & Kerr, 2013; Wagenaar & Toomey, 2002), some reported only a temporary effect (Miron & Tetelbaum, 2009) or an impact on only a specific drinking behavior such as binge drinking (Plunk, Cavazaos-Rehg, Bierut & Grucza, 2013). Other policies targeting alcohol availability include restrictions on outlet density, retail monopoly, and the hours and days of alcohol sales. These three measures are usually cited as being effective at reducing alcohol consumption and related harms (Burton et al., 2017; Holm, Veerman, Cobiac, Ekholm & Diderichsen, 2014; World Health Organization, 2018). The second group of initiatives includes policies regulating alcohol advertising. Some studies indicate that these are an effective way of reducing alcohol consumption (Holm et al., 2014), while others found a lack of robust evidence for or against such measures (Siegfried et al., 2014). The third group of initiatives refers to policies controlling the price of alcohol. An overall negative relationship has been observed between price and alcohol consumption (Wagenaar, Salois & Komro, 2009) although mixed results have been found regarding adolescent alcohol use, especially binge drinking (Nelson, 2015). Inequalities in alcohol consumption should be considered when developing policy interventions. Alcohol consumption is typically influenced by socioeconomic status (SES) in the sense that rates of drinking are related to higher income, both at an individual and at a population level (Collins, 2016; World Health Organization, 2018). However, with respect to adolescent drinking, evidence is inconsistent. Possible explanations are that the association is dependent on the alcohol measure used or that it differs according to the indicator employed to assess socioeconomic position. For this reason, we use three different measures of adolescent drinking and two different measures of socioeconomic status to provide a more comprehensive analyses of inequalities in adolescent alcohol use. The family affluence scale (FAS) is one of the most commonly used indicators of socioeconomic status among adolescents. This scale is used to evaluate material assets within the home (e.g., the number of cars and computers). Results on the association between FAS and alcohol consumption are mixed, with some studies showing higher alcohol use related to higher FAS, others to lower FAS, while other studies concluded that there was no association (Hanson & Chen, 2007). Furthermore, there is growing evidence suggesting that relative deprivation –measured by indicators such as perceived family wealth (PFW)– is strongly related to adolescent health and lifestyles, even after taking into account the effect of other socioeconomic indicators (Goodman, Huang, Schafer-Kalkhoff & Adler, 2007). In line with FAS, results are inconsistent. Some studies found that a higher PFW was a protective factor for alcohol consumption (Liu et al., 2018), whereas others found the opposite (Zaborskis, Sumskas, Maser & Pudule, 2006). Finally, differences have also been reported in relation to socioeconomic trends in alcohol use. Whereas some studies have reported an overall decrease in adolescent drinking in all SES groups – for example, in Australia (Livingston, 2014), Germany (Richter, Kuntsche, de Looze & Pfoertner, 2013), and the United States (Twenge & Park, 2017) – others found that the decrease was not the same for all SES groups, with higher levels of drinking being maintained among adolescents from lower SES groups, for example, in Finland (Liu et al., 2018) and New Zealand (Jackson et al., 2017). Policies and interventions aimed at promoting healthy habits and reducing risk behaviours such as alcohol consumption might have different effect on adolescents from different socioeconomic backgrounds, and interventions can narrow, widen, or have no effect in the existing socioeconomic inequalities (Moore, McDonald, Carlon & O'Rourke, 2015). In fact, there is a concern about universal public health interventions having the potential to increase social inequality in the population (Babones, 2009). To the best of our knowledge, there is no international study investigating the association between alcohol control policies and social inequality and trends in social inequality in adolescent alcohol consumption. The present study aims to analyze the association between (i) alcohol control policies and adolescent drinking outcomes (ii) socioeconomic inequality and adolescent drinking outcomes and (iii) a combination of policies and trends in socioeconomic inequality in adolescent drinking outcomes.