- Email: [email protected]
Socioeconomic inequalities in alcohol consumption in Chile and Finland
Drug and Alcohol Dependence 173 (2017) 24–30
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Full length article
Socioeconomic inequalities in alcohol consumption in Chile and Finland ˜ a,∗ , Pia Mäkelä a , Gonzalo Valdivia b , Satu Helakorpi c , Niina Markkula d , Sebastián Pena Paula Margozzini b , Seppo Koskinen a a
Department of Health, National Institute for Health and Welfare, Mannerheimintie 168, PO 22671, Helsinki, Finland Department of Public Health, Pontifical Catholic University, Marcoleta 434, Casilla 114D, Santiago, Chile c Department of Welfare, National Institute for Health and Welfare, Mannerheimintie 168, PO 22671, Helsinki, Finland d Faculty of Medicine, University Diego Portales, Ejército 233, Santiago, Chile b
a r t i c l e
i n f o
Article history: Received 2 August 2016 Received in revised form 5 December 2016 Accepted 6 December 2016 Available online 30 January 2017 Keywords: Alcohol drinking Health equity Socioeconomic factors Concentration index
a b s t r a c t Background: Reasons for socioeconomic inequalities in alcohol harm are not sufficiently understood. One explanation relates to differential exposure to alcohol by socioeconomic status (SES). The present study investigated socioeconomic inequalities in alcohol use in two countries with high alcohol consumption and alcohol harm. Methods: Data from nationally representative surveys in 2009–2010 in Chile and in 2008–2011 in Finland were used. Surveys comprised 3477 participants in Chile and 9994 in Finland aged 30–64 years. Outcome measures included abstinence, weekly consumption of pure alcohol, heavy volume drinking and heavy episodic drinking (HED). We employed a novel method in alcohol research, the concentration index, to measure socioeconomic inequalities. Results: Alcohol abstinence showed a strong association with lower SES in Chile and Finland. These were largely driven by inequalities among women in Chile and older subgroups in Finland. In both countries, women aged 45–64 of higher SES showed higher weekly consumption of pure alcohol and heavy volume drinking. Heavy volume drinking among Chilean women aged 45–64 showed the highest inequality, favouring higher SES. HED was equally distributed among SES groups in Chile; in Finland HED disproportionally affected lower SES groups. Conclusions: Lower SES was associated with higher abstinence rates in both countries and heavy episodic drinking in Finland. Heavy volume drinking was more prevalent in middle-aged women of high SES. The results identified groups for targeted interventions, including middle-aged higher SES women, who traditionally have not been specifically targeted. The concentration index could be a useful measure of inequalities in alcohol use. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Harmful use of alcohol accounts for 3.3 million net deaths every year and is the fifth largest risk factor of death and disability worldwide (Lim et al., 2012; World Health Organization, 2014b). Harmful drinking is also associated with vast social harm, including domestic violence, low productivity, work absenteeism, social isolation and stigma (Babor et al., 2010; Rehm et al., 2010).
∗ Corresponding author. ˜ E-mail addresses: [email protected], [email protected] (S. Pena), pia.makela@thl.fi (P. Mäkelä), [email protected] (G. Valdivia), satu.helakorpi@thl.fi (S. Helakorpi), niina.markkula@helsinki.fi (N. Markkula), [email protected] (P. Margozzini), seppo.koskinen@thl.fi (S. Koskinen). http://dx.doi.org/10.1016/j.drugalcdep.2016.12.014 0376-8716/© 2017 Elsevier B.V. All rights reserved.
Within societies, populations at the lower socioeconomic spectrum experience greater alcohol-related harm. A steep social gradient has been consistently described in high-income countries, but similar patterns of socioeconomic inequalities have also been found in low and middle-income countries such as Brazil (Mackenbach et al., 2015; Probst et al., 2014; Silveira et al., 2014). Reasons for socioeconomic inequalities in alcohol-related harm are not sufficiently understood. Possible explanations include differential exposure due to differential alcohol use among socioeconomic groups; differential vulnerability, where the same amount of alcohol results in different consequences; selection or reversed causality or methodological shortcomings in survey measurements (Bellis et al., 2016; Grittner et al., 2013; Grittner et al., 2012; Jones et al., 2015; Makela and Paljarvi, 2008).
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
Evidence of differential exposure to alcohol by socioeconomic groups is inconsistent. The most recurrent pattern is that lower socioeconomic groups tend to have higher levels of abstinence (Bloomfield et al., 2006; van Oers et al., 1999). Men in disadvantaged groups tend to drink less frequently but report higher rates of heavy episodic drinking (Dzúrová et al., 2011; Harper and Lynch, 2007; Midanik and Clark, 1994), although some studies have described opposite patterns (Giskes et al., 2011). Women show inconsistent patterns: some studies report higher rates of mean consumption and heavy episodic drinking among higher socioeconomic groups (Bloomfield et al., 2000; Bloomfield et al., 2008; McKee et al., 2000), while others show the opposite (Bloomfield et al., 2008; Casswell et al., 2003; Helasoja et al., 2007), or no differences at all (Bloomfield et al., 2006). Studies in low- and middle-income countries suggest alcohol consumption might have different patterns than in high-income countries. A regional study from Brazil showed that higher socioeconomic status was associated with 3-fold odds of high-risk drinking (Almeida-Filho et al., 2005). A national study from India found lower abstinence rates in lower educational groups (Subramanian et al., 2005). Material affluence was positively associated with alcohol use in Ghanian adolescents, but self-reported drunkenness was higher among those with lower material affluence (Doku et al., 2012). A recent study in 50 countries found mixed results in heavy episodic drinking (Hosseinpoor et al., 2012). Much of the research literature available has measured inequalities in alcohol use in ways that fail to incorporate the whole socioeconomic spectrum (comparing, for example, only top and bottom levels of SES), are not able to compare all social categories at once or do not provide quantitative estimates of the severity of the inequality. The use of odds ratios is particularly problematic when the outcome has a high prevalence and varies with time, as it is often the case of alcohol use indicators (Khang et al., 2008). Several studies, particularly those in low- and middle-income countries, have not examined both levels and patterns of alcohol use and instead relied on a single indicator. These methodological shortcomings could partly explain the observed heterogeneity in the research literature. To overcome these limitations, we used the concentration index, a summative measure of inequality that has several advantages: (1) use information from the whole socioeconomic spectrum; (2) account for changes in the population distribution in social groups over time; (3) provides quantitative estimates of the severity of the inequality and (4) is adequate for international comparisons (Harper and Lynch, 2007; O’Donnell et al., 2008). We also examined socioeconomic inequalities using four comparable indicators of levels and patterns of alcohol use. The present study investigates the existence and patterns of socioeconomic inequalities in alcohol use in Chile and Finland using nationally representative data. The rationale for this comparison rests in the fact that Chile and Finland have the highest alcohol consumption in the Americas and Northern Europe and suffer high alcohol-related harm. The comparison of two economically, geographically and culturally different countries increases the potential for the results to be applicable to other study settings and samples (i.e., external validity). Given the social patterning of alcohol related-harm, we hypothesized that alcohol consumption would tend to be higher in lower socioeconomic groups, in particular those measures more closely linked to alcohol-related harm: heavy volume drinking and heavy episodic drinking. 1.1. Country settings 1.1.1. Chile. Chile, which during our observation period was an upper middle-income country, has the highest alcohol consump-
25
tion of the Americas, with a total alcohol consumption of 9.6 l of pure alcohol per capita (15+) in 2008–2010, 7.6 of which is recorded consumption. As the world’s tenth largest wine producer, its 16 million inhabitants drink predominantly wine (38%), followed by spirits (32%) and beer (30%) (World Health Organization, 2015). Chileans have a drinking culture of consuming alcohol a few days per week (1.6 days on average) and, on average, 55 g per occasion (Margozzini and Sapag, 2015). Harmful alcohol use is the leading risk factor of death and disability and accounts for 12.4% of DALYs lost (Ministerio de Salud, 2008). Prevalence of AUDIT score over 8 points is 10.9%, with a 8:1 ratio between men and women (Ministerio de Salud, 2014). 1.1.2. Finland. Finland has a total alcohol consumption of 12.3 l of pure alcohol per capita (15+) in 2008–2010, 10.0 l of which is recorded consumption. Finns drink predominantly beer (46%), followed by spirits (24%), and wine (17%) (World Health Organization, 2014a). Alcohol consumption has steadily increased since mid1990s and it is nowadays the highest of Nordic countries (World Health Organization, 2015). Finns have a similar drinking culture, concentrating consumption on few days accompanied by high rates of heavy episodic drinking (Makela et al., 2012). Trends in consumption have been mirrored by alcohol-related mortality, which has increased from 25.6 per 100.000 inhabitants in 1990–32.6 in 2014 (Statistics Finland, 2015). 2. Methods and materials 2.1. Data sources In Chile, we used data from the National Health Survey 2009–2010 (ENS09-10). ENS09-10 is a household, face-to-face health examination survey representative of the Chilean population aged 15 or over (n = 5293). The survey has a multi-stage clustered sampling stratified by region and by urban and rural areas, resulting in 29 sampling clusters. Participation rate was 70% (Ministerio de Salud, 2014). To ensure comparability with the Finnish data, we restricted our analysis to population below 65 years old. Similarly, respondents aged less than 25 years were not included in the analysis considering their education is still undergoing. The analytical sample of respondents aged 25–64 consisted of 3477 adults. In Finland, we used pooled data from the annual Surveys on Health Behaviour and Health among the Finnish Adult Population (AVTK) in 2008–2011. AVTK is a postal survey representative of the Finnish population aged 15–64 (total sample 2008–2011 = 11772). The study uses a random sampling of 5000 permanent residents in Finland; the response rate ranged from 64 to 57% (Helakorpi et al., 2012). The analytical sample after excluding respondents aged less than 25 years comprised 9994 adults. ENS09-10 was approved by the Ethics Committee of the Catholic University of Chile in 2009. The protocol of the AVTK-study had been accepted by the Ethical Review Board of the National Institute for Health and Welfare, Finland. Informed consents were obtained by participants in both studies. 2.2. Measures 2.2.1. Alcohol use. We used four measures of alcohol consumption: abstinence, weekly consumption of pure alcohol, heavy volume drinking, and heavy episodic drinking. Abstinence was assessed in both surveys by a question on whether respondents had consumed any alcoholic beverages in the last 12 months. A dichotomous variable was constructed. Weekly consumption of pure alcohol was measured in Finland with a question on the number of portions of beer or ready-mixed
26
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
long drink (0,3 l bottles), spirits (restaurant portion of 4 cl), wine (glass), or cider (glass) consumed during the past week. Number of drinks was converted into grams of pure alcohol by multiplying the number of drinks by 12 g (the approximate alcohol content of each of the categories). In Chile, weekly consumption of pure alcohol was measured by a question on the amount and frequency of consumption of beer, wine, or spirits for each day of the week during the last week. A card with pictures of portions (7 options for beer and spirits, four for wine) was shown for participants to assess the volume of alcohol consumed. Grams of pure alcohol were calculated, for each beverage type and day of the week, by multiplying the number of portions with its alcohol content. Weekly consumption was constructed by adding the grams consumed in each day and for each beverage type. Heavy volume drinking was derived from the alcohol consumed during the past week. In both countries, it was defined as a dichotomous variable of those drinking more than 30 grs in men and 20 grs in women per day as used in the multinational GENACIS study (Bloomfield et al., 2005). Heavy episodic drinking (HED) was measured by a question on how often respondents drank 6 or more drinks at a time in Finland and 5 drinks in Chile (72 and 78 total grams, respectively) (Ministerio de Salud, 2014). We constructed a dichotomous variable of those participants with heavy episodic drinking once a month or more often.
2.2.2. Socioeconomic status. The mathematical computation of the concentration index requires a ratio scale variable, which has three properties: (1) allows ranking individuals from lowest to highest, (2) the difference between the values is equal and (3) the scale starts from zero. The study uses years of education as proxy of socioeconomic status to ensure comparability between datasets.
2.2.3. Demographics. Data was analysed separately for men and women and for two age groups: 25–44 and 45–64 years. Respondents aged less than 25 years were not included in the analysis, considering their education is still undergoing.
2.3. Methods The study employs the concentration index (CI) as a summary measure of inequalities in alcohol use. Introduced by Wagstaff et al. in 1991, the CI has been extensively used in health economics and health services research, yet only very recently in analysing risk factors of non-communicable diseases (NCDs) (Combes et al., 2011; Harper and Lynch, 2007). Similar to the Gini index, the concentration index is calculated in reference to the concentration curve, which plots the cumulative proportion of individuals from lowest to highest socioeconomic level (i.e., years of education) against the cumulative proportion of the measure of health, in our case, four indicators of alcohol use. A diagonal line, representing that all individuals have equal share of the health variable, is used as a reference. By convention, the concentration index is expressed as a negative value when the curve lies above the diagonal, indicating that the health variable concentrates in individuals of lower socioeconomic level, whereas a positive value and a line below the diagonal means that the variable of interest concentrates in individuals with higher socioeconomic level. The CI is defined as twice the net area between the concentration curve and the line of equality (O’Donnell et al., 2008).
For computation, the concentration index can be expressed as a “convenient” covariance, where G(Y) is the cumulative distribution of the socioeconomic variable: CI = −2Cov
X
(X)
, (1 − G (Y )
(1)
When the health variable is continuous, the CI ranges from −1 to 1. A value of zero means that there is no socioeconomic inequality in the health variable. However, for binary health outcomes the CI is bounded by the prevalence of the outcome and no longer takes values between −1 and 1. To facilitate the comparison of inequalities, we follow the recommendation by Wagstaff (2011) and normalized the CI by dividing it by 1 minus the mean (Wagstaff, 2011). The CI for a binary variable is thus given as: CIN =
CI (1 − )
(2)
Repetitive values of the socioeconomic variable can lead to unstable point and variance estimates of the concentration index (Chen and Roy, 2009). In our case, the use of years of education as the socioeconomic variable results in incidental ties arising from respondents having the same number of years of education. To overcome this problem, the solution allowing for sampling weights we used was to correct the fractional rank by giving tied observations an identical fractional rank. This method ensures the sample mean of the fractional ranks is equal to 0.5 and can therefore be plugged in a standard sample covariance formula, allowing taking complex survey data into account (Van Kerm, 2009). Statistical inference was done by estimating the standard error and 95% confidence intervals. We used the jackknife method to calculate standard errors since it allows the use of sampling weights (Van Kerm, 2009). The user-written command SGINI in Stata was employed to calculate the concentration indices and their standard error in Stata (Van Kerm, 2009). We used Stata 11.2 for the analyses. 3. Results Table 1 shows the sociodemographic characteristics of Chilean and Finnish populations. Chileans were on average 3.8 years younger and had 2.9 years of education less than Finns on average. The Finnish population had a significantly lower proportion of abstinents and they reported drinking more than double the alcohol per week than Chileans. The proportion of population drinking in high volumes and the frequency of heavy episodic drinking was also significantly higher in Finland. 3.1. Socioeconomic inequalities in abstinence Results for abstinence from alcohol use are presented in Table 2. The overall CI for alcohol abstinence in Chile was −0.25 and in Finland −0.19, indicating that in both countries abstinence was higher among lower SES groups. Analysing by gender and age, in Chile men of both age groups showed no inequalities in alcohol abstinence, while abstinence in women of both age groups was markedly higher among lower SES groups. In Finland, alcohol abstinence was higher among lower SES groups in women aged 24–44 and in men and women aged 45–64. The degree of inequality was markedly higher in the older age group. 3.2. Socioeconomic inequalities in weekly volume of pure alcohol Table 3 shows the results for weekly volume of pure alcohol consumed. The overall CI in Chile was positive, indicating more drinking among those with higher SES. This result was largely driven by a positive CI for women aged 45–64; in the younger age group and in men aged 45–64 CIs were not statistically different
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
27
Table 1 Socio-demographic characteristics and alcohol use in Chile and Finland. Chilea
Socio-demographic characteristics Mean age (yrs) Mean education (yrs) Alcohol use Abstinence, % Mean weekly consumption of pure alcohol (grs/wk) Median weekly consumption of pure alcohol (grs/wk) Heavy volume drinking, % Heavy episodic drinking, % a
Finland
N
Total
Men
Women
N
Total
Men
Women
3477 3272
43.2 10.9
42.8 11.1
43.6 10.7
9994 9833
46.8 13.8
47.4 13.3
46.4 14.1
3221 3221 3221 3221 3221
22.7 30.1 0 2.7 16.2
13.7 47.6 0 4.1 28.3
31.5 13.0 0 1.4 4.3
9807 9940 9940 9940 9735
10.9 69.9 36 9.6 29.4
8.9 104.5 72 14.1 45.3
12.5 43.2 24 6.1 17.1
Population weighted.
Table 2 Concentration index (CIN ) of abstinence by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
All 25–44 25–44 45–64 45–64
Chile
Finland
N
CIN
95% conf. interval
Interpretation: higher rates among
N
CIN
95% conf. interval
Interpretation: higher rates among
3216 635 976 645 960
−0.25 −0.12 −0.22 −0.25 −0.26
(−0.33: −0.16) (−0.32; 0.08) (−0.39; −0.04) (−0.51; 0.02) (−0.37; −0.15)
Lower SES No inequality Lower SES No inequality Lower SES
9654 1669 2296 2251 3138
−0.19 −0.06 −0.08 −0.17 −0.28
(−0.23; −0.16) (−0.18; 0.06) (−0.16; −0.01) (−0.24; −0.01) (−0.33; −0.22)
Lower SES No inequality Lower SES Lower SES Lower SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
Table 3 Concentration index (CI) of weekly volume of pure alcohol by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
All 25–44 25–44 45–64 45–64
Chile
Finland
N
CI
95% conf. interval
Interpretation: higher rates among
N
CI
95% conf. interval
Interpretation: higher rates among
3216 635 976 645 960
0.13 0.05 0.03 0.17 0.26
(0.04; 0.23) (−0.04; 0.15) (−0.15; 0.21) (−0.05; 0.39) (0.14; 0.38)
Higher SES No inequality No inequality No inequality Higher SES
9781 1690 2328 2578 3185
−0.01 −0.04 0.003 0.05 0.06
(−0.03; 0.01) (−0.07; −0.01) (−0.03; 0.04) (0.02; 0.08) (0.03; 0.09)
No inequality Lower SES No inequality Higher SES Higher SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
from zero. In Finland, the overall CI was not statistically significant from zero. However, the CIs were positive and statistically significant for men and women aged 45–64 and negative for men aged 25–44.
equally across SES groups. In Finland, the overall CI as well as the CIs in each subgroup were negative; suggesting HED is consistently more prevalent in lower SES groups. Fig. 1 summarises the main results of the study.
3.3. Socioeconomic inequalities in heavy volume drinking
4. Discussion
Results for heavy volume drinking are shown in Table 4. In both countries, CIs for heavy volume drinking were statistically not significant from zero, at least partly because in Chile, the prevalence of heavy drinking was low in all subgroups, resulting in larger standard errors. The CIs were not statistically significant from zero for men and women aged 25–44 and men aged 45–64. On the other hand, women aged 45–64 showed a marked association between higher SES and heavy volume drinking, the highest found in the study. In Finland, we observed a negative CI among men aged 25–44, indicating that heavy volume drinking was more concentrated among those with lower SES status. Conversely, women aged 45–64 showed a positive CI, yet the magnitude of the inequality was modest.
Using the concentration index, we examined socioeconomic inequalities in patterns and volume of alcohol consumption in Chile and Finland. Our results show the existence of SES inequalities in several measures of alcohol use in both countries. A strong association was found between lower SES and alcohol abstinence in Chile and Finland. This was largely due to inequalities among women in Chile and in middle-age groups in Finland. The study found no socioeconomic differences in heavy volume drinking in either country in the whole working-age population, but we did observe a consistent pattern of higher weekly consumption of pure alcohol and heavy drinking among higher SES groups in Chilean and Finnish women aged 45–64. Finnish men aged 25–44 of lower SES men drank higher volumes of pure alcohol and showed higher rates of heavy drinking. Surprisingly, we did not observe socioeconomic inequalities in heavy episodic drinking in Chile, but in Finland there was a clear pattern of higher HED among lower SES groups. Therefore, and contrary to our expectations, we observed differential exposure to alcohol use in lower socioeconomic groups only in heavy episodic drinking in Finland; heavy volume drinking
3.4. Socioeconomic inequalities in heavy episodic drinking Table 5 shows the CIs for heavy episodic drinking (HED). In Chile, CIs for HED were not statistically significant in any subgroup, i.e., the “burden” of heavy episodic drinking appeared to be distributed
28
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
Table 4 Concentration index (CIN ) of heavy volume drinking by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
Chile
All 25–44 25–44 45–64 45–64
Finland
N
CIN
95% conf. interval
Interpretation: higher rates among
N
CIN
95% conf. interval
Interpretation: higher rates among
3216 635 976 645 960
0.04 0.07 −0.34 0.07 0.47
(−0.21; 0.3) (−0.43; 0.57) (−0.71; 0.03) (−0.34; 0.47) (0.18; 0.75)
No inequality No inequality No inequality No inequality Higher SES
9781 1690 2328 2578 3185
−0.01 −0.11 −0.04 0.06 0.11
(−0.05; 0.03) (−0.19; −0.03) (−0.14; 0.07) (−0.002; 0.12) (0.02; 0.19)
No inequality Lower SES No inequality No inequality Higher SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
Table 5 Concentration index (CIN ) of heavy episodic drinking by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
All 25–44 25–44 45–64 45–64
Chile
Finland
N
CIN
SE
95% conf. interval
Interpretation: higher rates among
N
CIN
SE
95% conf. interval
Interpretation: higher rates among
3221 635 976 645 960
−0.04 −0.12 −0.14 −0.07 −0.07
0.04 0.06 0.12 0.06 0.18
(−0.14; 0.07) (−0.31; 0.06) (−0.39; 0.11) (−0.22; 0.09) (−0.43; 0.28)
No inequality No inequality No inequality No inequality No inequality
9586 1664 2292 2512 3118
−0.10 −0.10 −0.10 −0.06 −0.10
0.01 0.01 0.03 0.01 0.02
(−0.13; −0.08) (−0.16; −0.05) (−0.16; −0.04) (−0.10; −0.01) (−0.16; −0.05)
Lower SES Lower SES Lower SES Lower SES Lower SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
Fig. 1. Forest plot of concentration index for four alcohol measures in Chile and Finland.
showed either no inequalities or the opposite pattern to what we hypothesized. These findings stress the importance of examining both levels and patterns of alcohol use in research and policy interventions, as the use of a single indicator might bias conclusions. In the context of examining inequalities on risk factors for NCDs, most studies have relied on a single measure of alcohol use (Combes et al., 2011; Harper and Lynch, 2007; Hosseinpoor et al., 2012; Marmot, 1997). The few studies carrying out a comprehensive assessment have shown a complex picture of socioeconomic inequalities in alcohol consumption, in line with our findings (Casswell et al., 2003; van Oers et al., 1999).
4.1. Comparison with previous studies In our analysis, alcohol abstinence was concentrated among the lower SES groups. This is consistent with previous research showing a corresponding socioeconomic gradient (Bloomfield et al., 2000; Bloomfield et al., 2006; Marmot, 1997; van Oers et al., 1999). Lower SES groups often face harder priority-choices and might have to prioritize their expenditure on basic needs. In Chilean women, social norms and stigma related to alcohol consumption could also explain higher abstinence in lower SES groups. Regarding volume of pure alcohol and heavy volume drinking, studies in Baltic Countries and Brazil have also observed no inequalities or higher consumption on higher SES groups (Almeida-Filho
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
et al., 2005; Helasoja et al., 2007). Interestingly, the GENACIS study found higher rates of heavy volume drinking among women of higher SES in several Central European countries, but no inequalities in Nordic and Latin American countries (Bloomfield et al., 2006). This is in line with our findings and strengthen the idea of higher SES middle-aged women as a high-risk group. This could potentially be helpful for targeting interventions (such as brief interventions). Harper et al. used the concentration index to examine SES inequalities in heavy episodic drinking in the United States, observing that it was increasingly concentrated among those with lower education, which is consistent with our findings in Finland (Harper and Lynch, 2007). Other studies in the United States and the Czech Republic have yielded similar results (Dzúrová et al., 2011; Midanik and Clark, 1994). This could be explained by improved affordability of alcohol, different drinking cultures, exposure to alcohol marketing, prevalence of mental health disorders or ability to assess risk, among other factors. 4.2. Limitations The study has limitations that need to be addressed. First, we used surveys not designed to be comparable, as they have differences in their sampling design and data collection modes. However, these were carefully selected since they provided nationally representative data using very similar instruments. The Finnish data had lower participation rates, so there is a risk of selection bias due to non-response of, for example, people with lower education and higher prevalence of alcohol-related problems. Second, the only comparable SES variable available was years of education, whereas income or wealth have been most extensively used in the literature. Educational status might vary by age cohort, as levels of education have normally increased over the years. Third, the use of past week alcohol use might not be representative of a drinker’s usual patterns at an individual level. It could be, however, representative of the drinking patterns at a group level. Fourth, the results from both countries might not be applicable to countries with lower levels of development, even though they might be helpful for countries with high alcohol-related harm. This is the first step in a broader research agenda seeking to examine patterns of socioeconomic inequalities across countries. Next steps should consist of examining socioeconomic inequalities in alcohol-related harm, exploring a wider range of countries and examining secular trends. 5. Conclusions In two countries with the highest alcohol consumption in their regions, lower socioeconomic status was associated with higher prevalence of abstinence in both countries and heavy episodic drinking in Finland. Heavy volume drinking was more prevalent in middle-age women of high socioeconomic status. The results highlight the importance of multidimensional assessments of alcohol use and identify middle-aged higher SES women as a group for targeted alcohol interventions. Conflict of interest No conflict declared Contributors All authors materially participated in the research. SP, SK and PMäkelä designed the study, analysed and interpreted the data and contributed in writing the report. NM contributed in analysing and interpreting the data and writing the report. GV, SH and PMar-
29
gozzini collected the data and contributed in writing the report. All authors approved the final manuscript before submission. Funding This study was supported in part by the Finnish Medical Foundation. Acknowledgments The authors are thankful to Zhuo Adam Chen, John Ataguba and Báltica Cabieses for their helpful methodological suggestions on earlier versions of the manuscript. References Almeida-Filho, N., Lessa, I., Magalhaes, L., Araujo, M.J., Aquino, E., James, S.A., Kawachi, I., 2005. Social inequality and alcohol consumption-abuse in Bahia, Brazil—interactions of gender, ethnicity and social class. Soc. Psychiatry Psychiatr. Epidemiol. 40, 214–222. Babor, T., Caetano, R., Casswell, S., Edwards, G., Giesbrecht, N., Graham, K., Grube, J., Hill, L., Holder, H., Homel, R., Livingston, M., Österberg, E., Rehm, J., Room, R., Rossow, I., 2010. Alcohol: No Ordinary Commodity: Research and Public Policy. Oxford University Press, Oxford. Bellis, M.A., Hughes, K., Nicholls, J., Sheron, N., Gilmore, I., Jones, L., 2016. The alcohol harm paradox: using a national survey to explore how alcohol may disproportionately impact health in deprived individuals. BMC Public Health 16, 111. Bloomfield, K., Augustin, R., Kraus, L., 2000. Social inequalities in alcohol use and misuse in the German general population. J. Public Health 8, 230–242. Bloomfield, K., Allamani, A., Beck, F., Bergmark, K., Csemy, L., Eisenbach-Stangl, I., Elekes, Z., Gmel, G., Kerr-Correa, F., Knibbe, R., Makela, P., Monteiro, M., Medina, M., Nordlund, S., Obot, I., Plant, M., Rahav, G., Romero, M., 2005. Gender, Culture and Alcohol Problems: A Multi-national Study Project Final Report. Charite-Campus Benjamin Franklin, Berlin. Bloomfield, K., Grittner, U., Kramer, S., Gmel, G., 2006. Social inequalities in alcohol consumption and alcohol-related problems in the study countries of the EU concerted action ‘Gender: culture and alcohol problems: a multi-national study’. Alcohol 41, i26–36. Bloomfield, K., Grittner, U., Rasmussen, H.B., Petersen, H.C., 2008. Socio-demographic correlates of alcohol consumption in the Danish general population. Scand. J. Public Health 36, 580–588. Casswell, S., Pledger, M., Hooper, R., 2003. Socioeconomic status and drinking patterns in young adults. Addiction 98, 601–610. Chen, Z., Roy, K., 2009. Calculating concentration index with repetitive values of indicators of economic welfare. J. Health Econ. 28, 169–175. Combes, J.B., Gerdtham, U.G., Jarl, J., 2011. Equalisation of alcohol participation among socioeconomic groups over time: an analysis based on the total differential approach and longitudinal data from Sweden. Int. J. Equity Health 10, 10. Doku, D., Koivusilta, L., Rimpela, A., 2012. Socioeconomic differences in alcohol and drug use among Ghanaian adolescents. Addict. Behav. 37, 357–360. Dzúrová, D., Spilková, J., Pikhart, H., 2011. Social inequalities in alcohol consumption in the Czech Republic: a multilevel analysis. Health Place 16, 590–597. Giskes, K., Turrell, G., Bentley, R., Kavanagh, A., 2011. Individual and household-level socioeconomic position is associated with harmful alcohol consumption behaviours among adults. Aust. N. Z. J. Public Health 35, 270–277. Grittner, U., Kuntsche, S., Graham, K., Bloomfield, K., 2012. Social inequalities and gender differences in the experience of alcohol-related problems. Alcohol 47, 597–605. Grittner, U., Kuntsche, S., Gmel, G., Bloomfield, K., 2013. Alcohol consumption and social inequality at the individual and country levels—results from an international study. Eur. J. Public Health 23, 332–339. Harper, S., Lynch, J., 2007. Trends in socioeconomic inequalities in adult health behaviors among U.S. states, 1990–2004. Public Health Rep. 122, 177–189. Helakorpi, S., Hostila, A.-L., Virtanen, S., Uutela, A., 2012. Suomalaisen aikuisväestön terveyskäyttäytyminen ja terveys: Kevät 2011 [Health behavior and health among the Finnish adult population: Spring 2011. National Institute for Health and Welfare (THL), Helsinki. Helasoja, V., Lahelma, E., Prattala, R., Petkeviciene, J., Pudule, I., Tekkel, M., 2007. The sociodemographic patterning of drinking and binge drinking in Estonia, Latvia, Lithuania and Finland, 1994–2002. BMC Public Health 7, 241. Hosseinpoor, A.R., Bergen, N., Kunst, A., Harper, S., Guthold, R., Rekve, D., d’Espaignet, E.T., Naidoo, N., Chatterji, S., 2012. Socioeconomic inequalities in risk factors for non communicable diseases in low-income and middle-income countries: results from the World Health Survey. BMC Public Health 12, 912. Jones, L., Bates, G., McCoy, E., Bellis, M.A., 2015. Relationship between alcohol-attributable disease and socioeconomic status, and the role of alcohol consumption in this relationship: a systematic review and meta-analysis. BMC Public Health 15, 400.
30
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
Khang, Y.H., Yun, S.C., Lynch, J.W., 2008. Monitoring trends in socioeconomic health inequalities: it matters how you measure. BMC Public Health 8, 66. Lim, S.S., Vos, T., Flaxman, A.D., Danaei, G., Shibuya, K., Adair-Rohani, H., Amann, M., Anderson, H.R., Andrews, K.G., Aryee, M., Atkinson, C., Bacchus, L.J., Bahalim, A.N., Balakrishnan, K., Balmes, J., Barker-Collo, S., Baxter, A., Bell, M.L., Blore, J.D., Blyth, F., Bonner, C., Borges, G., Bourne, R., Boussinesq, M., Brauer, M., Brooks, P., Bruce, N.G., Brunekreef, B., Bryan-Hancock, C., Bucello, C., Buchbinder, R., Bull, F., Burnett, R.T., Byers, T.E., Calabria, B., Carapetis, J., Carnahan, E., Chafe, Z., Charlson, F., Chen, H., Chen, J.S., Cheng, A.T., Child, J.C., Cohen, A., Colson, K.E., Cowie, B.C., Darby, S., Darling, S., Davis, A., Degenhardt, L., Dentener, F., Des Jarlais, D.C., Devries, K., Dherani, M., Ding, E.L., Dorsey, E.R., Driscoll, T., Edmond, K., Ali, S.E., Engell, R.E., Erwin, P.J., Fahimi, S., Falder, G., Farzadfar, F., Ferrari, A., Finucane, M.M., Flaxman, S., Fowkes, F.G., Freedman, G., Freeman, M.K., Gakidou, E., Ghosh, S., Giovannucci, E., Gmel, G., Graham, K., Grainger, R., Grant, B., Gunnell, D., Gutierrez, H.R., Hall, W., Hoek, H.W., Hogan, A., Hosgood 3rd, H.D., Hoy, D., Hu, H., Hubbell, B.J., Hutchings, S.J., Ibeanusi, S.E., Jacklyn, G.L., Jasrasaria, R., Jonas, J.B., Kan, H., Kanis, J.A., Kassebaum, N., Kawakami, N., Khang, Y.H., Khatibzadeh, S., Khoo, J.P., Kok, C., Laden, F., et al., 2012. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 380, 2224–2260. Mackenbach, J.P., Kulhanova, I., Bopp, M., Borrell, C., Deboosere, P., Kovacs, K., Looman, C.W., Leinsalu, M., Makela, P., Martikainen, P., Menvielle, G., Rodriguez-Sanz, M., Rychtarikova, J., de Gelder, R., 2015. Inequalities in alcohol-related mortality in 17 European countries: a retrospective analysis of mortality registers. PLoS Med. 12, e1001909. Makela, P., Paljarvi, T., 2008. Do consequences of a given pattern of drinking vary by socioeconomic status? A mortality and hospitalisation follow-up for alcohol-related causes of the Finnish drinking habits surveys. J. Epidemiol. Commun. Health 62, 728–733. Makela, P., Tigerstedt, C., Mustonen, H., 2012. The Finnish drinking culture: change and continuity in the past 40 years. Drug Alcohol Rev. 31, 831–840. Margozzini, P., Sapag, J., 2015. Consumo riesgoso de alcohol en Chile: Tareas pendientes y oportunidades para las políticas públicas [Risky alcohol drinking in Chile: pending issues and opportunities for public policies]. Temas de la Agenda Pública: Centro de Políticas Públicas UC. Marmot, M., 1997. Inequality, deprivation and alcohol use. Addiction 92 (Suppl. 1), S13–20. McKee, M., Pomerleau, J., Robertson, A., Pudule, I., Grinberga, D., Kadziauskiene, K., Abaravicius, A., Vaask, S., 2000. Alcohol consumption in the baltic republics. J. Epidemiol. Commun. Health 54, 361–366. Midanik, L.T., Clark, W.B., 1994. The demographic distribution of US drinking patterns in 1990: description and trends from 1984. Am. J. Public Health 84, 1218–1222.
Ministerio de Salud, 2008. Estudio de Carga de Enfermedad y Carga Atribuible. Ministerio de Salud, Santiago, Chile. Ministerio de Salud, 2014. Encuesta Nacional de Salud 2009–2010 (Online Available from: http://www.minsal.cl/estudios encuestas salud/ Accessed 4 December 2016). O’Donnell, O., Van Doorslaer, E., Wagstaff, A., Lindelow, M., 2008. Analyzing Health Equity Using Household Survey Data: A Guide to Techniques and Their Implementation. International Bank for Reconstruction and Development/The World Bank, Washington DC. Probst, C., Roerecke, M., Behrendt, S., Rehm, J., 2014. Socioeconomic differences in alcohol-attributable mortality compared with all-cause mortality: a systematic review and meta-analysis. Int. J. Epidemiol. 43, 1314–1327. Rehm, J., Baliunas, D., Borges, G.L.G., Graham, K., Irving, H., Kehoe, T., Parry, C.D., Patra, J., Popova, S., Poznyak, V., Roerecke, M., Room, R., Samokhvalov, A.V., Taylor, B., 2010. The relation between different dimensions of alcohol consumption and burden of disease: an overview. Addiction 105, 817–843. Silveira, C.M., Siu, E.R., Anthony, J.C., Saito, L.P., de Andrade, A.G., Kutschenko, A., Viana, M.C., Wang, Y.P., Martins, S.S., Andrade, L.H., 2014. Drinking patterns and alcohol use disorders in Sao Paulo, Brazil: the role of neighborhood social deprivation and socioeconomic status. PLoS One 9, e108355. Statistics Finland, 2015. Statistics Finland PX-Web databases (Online Available from: http://pxnet2.stat.fi/PXWeb/pxweb/en/StatFin Accessed 4 December 2016). Subramanian, S.V., Nandy, S., Irving, M., Gordon, D., Davey Smith, G., 2005. Role of socioeconomic markers and state prohibition policy in predicting alcohol consumption among men and women in India: a multilevel statistical analysis. Bull. World Health Organ. 83, 829–836. Van Kerm, P., 2009. ‘sgini − Generalized Gini and Concentration coefficients (with factor decomposition) in Stata’, v1.1 (revised February 2010). CEPS/INSTEAD, Differdange, Luxemburg. Wagstaff, A., Paci, P., van Doorslaer, E., 1991. On the measurement of inequalities in health. Soc. Sci. Med. 33, 545–557. Wagstaff, A., 2011. The concentration index of a binary outcome revisited. Health Econ. 20, 1155–1160. World Health Organization, 2014a. Finland − Alcohol Profile. World Health Organization, Geneva, Switzerland. World Health Organization, 2014b. Global Status Report on Alcohol and Health. World Health Organization, Geneva. World Health Organization, 2015. Global Information System on Alcohol and Health (GISAH) (Online Available from: http://www.who.int/gho/alcohol/en/ Accessed on 4 December 2016). van Oers, J.A., Bongers, I.M., van de Goor, L.A., Garretsen, H.F., 1999. Alcohol consumption, alcohol-related problems, problem drinking, and socioeconomic status. Alcohol 34, 78–88.
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Full length article
Socioeconomic inequalities in alcohol consumption in Chile and Finland ˜ a,∗ , Pia Mäkelä a , Gonzalo Valdivia b , Satu Helakorpi c , Niina Markkula d , Sebastián Pena Paula Margozzini b , Seppo Koskinen a a
Department of Health, National Institute for Health and Welfare, Mannerheimintie 168, PO 22671, Helsinki, Finland Department of Public Health, Pontifical Catholic University, Marcoleta 434, Casilla 114D, Santiago, Chile c Department of Welfare, National Institute for Health and Welfare, Mannerheimintie 168, PO 22671, Helsinki, Finland d Faculty of Medicine, University Diego Portales, Ejército 233, Santiago, Chile b
a r t i c l e
i n f o
Article history: Received 2 August 2016 Received in revised form 5 December 2016 Accepted 6 December 2016 Available online 30 January 2017 Keywords: Alcohol drinking Health equity Socioeconomic factors Concentration index
a b s t r a c t Background: Reasons for socioeconomic inequalities in alcohol harm are not sufficiently understood. One explanation relates to differential exposure to alcohol by socioeconomic status (SES). The present study investigated socioeconomic inequalities in alcohol use in two countries with high alcohol consumption and alcohol harm. Methods: Data from nationally representative surveys in 2009–2010 in Chile and in 2008–2011 in Finland were used. Surveys comprised 3477 participants in Chile and 9994 in Finland aged 30–64 years. Outcome measures included abstinence, weekly consumption of pure alcohol, heavy volume drinking and heavy episodic drinking (HED). We employed a novel method in alcohol research, the concentration index, to measure socioeconomic inequalities. Results: Alcohol abstinence showed a strong association with lower SES in Chile and Finland. These were largely driven by inequalities among women in Chile and older subgroups in Finland. In both countries, women aged 45–64 of higher SES showed higher weekly consumption of pure alcohol and heavy volume drinking. Heavy volume drinking among Chilean women aged 45–64 showed the highest inequality, favouring higher SES. HED was equally distributed among SES groups in Chile; in Finland HED disproportionally affected lower SES groups. Conclusions: Lower SES was associated with higher abstinence rates in both countries and heavy episodic drinking in Finland. Heavy volume drinking was more prevalent in middle-aged women of high SES. The results identified groups for targeted interventions, including middle-aged higher SES women, who traditionally have not been specifically targeted. The concentration index could be a useful measure of inequalities in alcohol use. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Harmful use of alcohol accounts for 3.3 million net deaths every year and is the fifth largest risk factor of death and disability worldwide (Lim et al., 2012; World Health Organization, 2014b). Harmful drinking is also associated with vast social harm, including domestic violence, low productivity, work absenteeism, social isolation and stigma (Babor et al., 2010; Rehm et al., 2010).
∗ Corresponding author. ˜ E-mail addresses: [email protected], [email protected] (S. Pena), pia.makela@thl.fi (P. Mäkelä), [email protected] (G. Valdivia), satu.helakorpi@thl.fi (S. Helakorpi), niina.markkula@helsinki.fi (N. Markkula), [email protected] (P. Margozzini), seppo.koskinen@thl.fi (S. Koskinen). http://dx.doi.org/10.1016/j.drugalcdep.2016.12.014 0376-8716/© 2017 Elsevier B.V. All rights reserved.
Within societies, populations at the lower socioeconomic spectrum experience greater alcohol-related harm. A steep social gradient has been consistently described in high-income countries, but similar patterns of socioeconomic inequalities have also been found in low and middle-income countries such as Brazil (Mackenbach et al., 2015; Probst et al., 2014; Silveira et al., 2014). Reasons for socioeconomic inequalities in alcohol-related harm are not sufficiently understood. Possible explanations include differential exposure due to differential alcohol use among socioeconomic groups; differential vulnerability, where the same amount of alcohol results in different consequences; selection or reversed causality or methodological shortcomings in survey measurements (Bellis et al., 2016; Grittner et al., 2013; Grittner et al., 2012; Jones et al., 2015; Makela and Paljarvi, 2008).
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
Evidence of differential exposure to alcohol by socioeconomic groups is inconsistent. The most recurrent pattern is that lower socioeconomic groups tend to have higher levels of abstinence (Bloomfield et al., 2006; van Oers et al., 1999). Men in disadvantaged groups tend to drink less frequently but report higher rates of heavy episodic drinking (Dzúrová et al., 2011; Harper and Lynch, 2007; Midanik and Clark, 1994), although some studies have described opposite patterns (Giskes et al., 2011). Women show inconsistent patterns: some studies report higher rates of mean consumption and heavy episodic drinking among higher socioeconomic groups (Bloomfield et al., 2000; Bloomfield et al., 2008; McKee et al., 2000), while others show the opposite (Bloomfield et al., 2008; Casswell et al., 2003; Helasoja et al., 2007), or no differences at all (Bloomfield et al., 2006). Studies in low- and middle-income countries suggest alcohol consumption might have different patterns than in high-income countries. A regional study from Brazil showed that higher socioeconomic status was associated with 3-fold odds of high-risk drinking (Almeida-Filho et al., 2005). A national study from India found lower abstinence rates in lower educational groups (Subramanian et al., 2005). Material affluence was positively associated with alcohol use in Ghanian adolescents, but self-reported drunkenness was higher among those with lower material affluence (Doku et al., 2012). A recent study in 50 countries found mixed results in heavy episodic drinking (Hosseinpoor et al., 2012). Much of the research literature available has measured inequalities in alcohol use in ways that fail to incorporate the whole socioeconomic spectrum (comparing, for example, only top and bottom levels of SES), are not able to compare all social categories at once or do not provide quantitative estimates of the severity of the inequality. The use of odds ratios is particularly problematic when the outcome has a high prevalence and varies with time, as it is often the case of alcohol use indicators (Khang et al., 2008). Several studies, particularly those in low- and middle-income countries, have not examined both levels and patterns of alcohol use and instead relied on a single indicator. These methodological shortcomings could partly explain the observed heterogeneity in the research literature. To overcome these limitations, we used the concentration index, a summative measure of inequality that has several advantages: (1) use information from the whole socioeconomic spectrum; (2) account for changes in the population distribution in social groups over time; (3) provides quantitative estimates of the severity of the inequality and (4) is adequate for international comparisons (Harper and Lynch, 2007; O’Donnell et al., 2008). We also examined socioeconomic inequalities using four comparable indicators of levels and patterns of alcohol use. The present study investigates the existence and patterns of socioeconomic inequalities in alcohol use in Chile and Finland using nationally representative data. The rationale for this comparison rests in the fact that Chile and Finland have the highest alcohol consumption in the Americas and Northern Europe and suffer high alcohol-related harm. The comparison of two economically, geographically and culturally different countries increases the potential for the results to be applicable to other study settings and samples (i.e., external validity). Given the social patterning of alcohol related-harm, we hypothesized that alcohol consumption would tend to be higher in lower socioeconomic groups, in particular those measures more closely linked to alcohol-related harm: heavy volume drinking and heavy episodic drinking. 1.1. Country settings 1.1.1. Chile. Chile, which during our observation period was an upper middle-income country, has the highest alcohol consump-
25
tion of the Americas, with a total alcohol consumption of 9.6 l of pure alcohol per capita (15+) in 2008–2010, 7.6 of which is recorded consumption. As the world’s tenth largest wine producer, its 16 million inhabitants drink predominantly wine (38%), followed by spirits (32%) and beer (30%) (World Health Organization, 2015). Chileans have a drinking culture of consuming alcohol a few days per week (1.6 days on average) and, on average, 55 g per occasion (Margozzini and Sapag, 2015). Harmful alcohol use is the leading risk factor of death and disability and accounts for 12.4% of DALYs lost (Ministerio de Salud, 2008). Prevalence of AUDIT score over 8 points is 10.9%, with a 8:1 ratio between men and women (Ministerio de Salud, 2014). 1.1.2. Finland. Finland has a total alcohol consumption of 12.3 l of pure alcohol per capita (15+) in 2008–2010, 10.0 l of which is recorded consumption. Finns drink predominantly beer (46%), followed by spirits (24%), and wine (17%) (World Health Organization, 2014a). Alcohol consumption has steadily increased since mid1990s and it is nowadays the highest of Nordic countries (World Health Organization, 2015). Finns have a similar drinking culture, concentrating consumption on few days accompanied by high rates of heavy episodic drinking (Makela et al., 2012). Trends in consumption have been mirrored by alcohol-related mortality, which has increased from 25.6 per 100.000 inhabitants in 1990–32.6 in 2014 (Statistics Finland, 2015). 2. Methods and materials 2.1. Data sources In Chile, we used data from the National Health Survey 2009–2010 (ENS09-10). ENS09-10 is a household, face-to-face health examination survey representative of the Chilean population aged 15 or over (n = 5293). The survey has a multi-stage clustered sampling stratified by region and by urban and rural areas, resulting in 29 sampling clusters. Participation rate was 70% (Ministerio de Salud, 2014). To ensure comparability with the Finnish data, we restricted our analysis to population below 65 years old. Similarly, respondents aged less than 25 years were not included in the analysis considering their education is still undergoing. The analytical sample of respondents aged 25–64 consisted of 3477 adults. In Finland, we used pooled data from the annual Surveys on Health Behaviour and Health among the Finnish Adult Population (AVTK) in 2008–2011. AVTK is a postal survey representative of the Finnish population aged 15–64 (total sample 2008–2011 = 11772). The study uses a random sampling of 5000 permanent residents in Finland; the response rate ranged from 64 to 57% (Helakorpi et al., 2012). The analytical sample after excluding respondents aged less than 25 years comprised 9994 adults. ENS09-10 was approved by the Ethics Committee of the Catholic University of Chile in 2009. The protocol of the AVTK-study had been accepted by the Ethical Review Board of the National Institute for Health and Welfare, Finland. Informed consents were obtained by participants in both studies. 2.2. Measures 2.2.1. Alcohol use. We used four measures of alcohol consumption: abstinence, weekly consumption of pure alcohol, heavy volume drinking, and heavy episodic drinking. Abstinence was assessed in both surveys by a question on whether respondents had consumed any alcoholic beverages in the last 12 months. A dichotomous variable was constructed. Weekly consumption of pure alcohol was measured in Finland with a question on the number of portions of beer or ready-mixed
26
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
long drink (0,3 l bottles), spirits (restaurant portion of 4 cl), wine (glass), or cider (glass) consumed during the past week. Number of drinks was converted into grams of pure alcohol by multiplying the number of drinks by 12 g (the approximate alcohol content of each of the categories). In Chile, weekly consumption of pure alcohol was measured by a question on the amount and frequency of consumption of beer, wine, or spirits for each day of the week during the last week. A card with pictures of portions (7 options for beer and spirits, four for wine) was shown for participants to assess the volume of alcohol consumed. Grams of pure alcohol were calculated, for each beverage type and day of the week, by multiplying the number of portions with its alcohol content. Weekly consumption was constructed by adding the grams consumed in each day and for each beverage type. Heavy volume drinking was derived from the alcohol consumed during the past week. In both countries, it was defined as a dichotomous variable of those drinking more than 30 grs in men and 20 grs in women per day as used in the multinational GENACIS study (Bloomfield et al., 2005). Heavy episodic drinking (HED) was measured by a question on how often respondents drank 6 or more drinks at a time in Finland and 5 drinks in Chile (72 and 78 total grams, respectively) (Ministerio de Salud, 2014). We constructed a dichotomous variable of those participants with heavy episodic drinking once a month or more often.
2.2.2. Socioeconomic status. The mathematical computation of the concentration index requires a ratio scale variable, which has three properties: (1) allows ranking individuals from lowest to highest, (2) the difference between the values is equal and (3) the scale starts from zero. The study uses years of education as proxy of socioeconomic status to ensure comparability between datasets.
2.2.3. Demographics. Data was analysed separately for men and women and for two age groups: 25–44 and 45–64 years. Respondents aged less than 25 years were not included in the analysis, considering their education is still undergoing.
2.3. Methods The study employs the concentration index (CI) as a summary measure of inequalities in alcohol use. Introduced by Wagstaff et al. in 1991, the CI has been extensively used in health economics and health services research, yet only very recently in analysing risk factors of non-communicable diseases (NCDs) (Combes et al., 2011; Harper and Lynch, 2007). Similar to the Gini index, the concentration index is calculated in reference to the concentration curve, which plots the cumulative proportion of individuals from lowest to highest socioeconomic level (i.e., years of education) against the cumulative proportion of the measure of health, in our case, four indicators of alcohol use. A diagonal line, representing that all individuals have equal share of the health variable, is used as a reference. By convention, the concentration index is expressed as a negative value when the curve lies above the diagonal, indicating that the health variable concentrates in individuals of lower socioeconomic level, whereas a positive value and a line below the diagonal means that the variable of interest concentrates in individuals with higher socioeconomic level. The CI is defined as twice the net area between the concentration curve and the line of equality (O’Donnell et al., 2008).
For computation, the concentration index can be expressed as a “convenient” covariance, where G(Y) is the cumulative distribution of the socioeconomic variable: CI = −2Cov
X
(X)
, (1 − G (Y )
(1)
When the health variable is continuous, the CI ranges from −1 to 1. A value of zero means that there is no socioeconomic inequality in the health variable. However, for binary health outcomes the CI is bounded by the prevalence of the outcome and no longer takes values between −1 and 1. To facilitate the comparison of inequalities, we follow the recommendation by Wagstaff (2011) and normalized the CI by dividing it by 1 minus the mean (Wagstaff, 2011). The CI for a binary variable is thus given as: CIN =
CI (1 − )
(2)
Repetitive values of the socioeconomic variable can lead to unstable point and variance estimates of the concentration index (Chen and Roy, 2009). In our case, the use of years of education as the socioeconomic variable results in incidental ties arising from respondents having the same number of years of education. To overcome this problem, the solution allowing for sampling weights we used was to correct the fractional rank by giving tied observations an identical fractional rank. This method ensures the sample mean of the fractional ranks is equal to 0.5 and can therefore be plugged in a standard sample covariance formula, allowing taking complex survey data into account (Van Kerm, 2009). Statistical inference was done by estimating the standard error and 95% confidence intervals. We used the jackknife method to calculate standard errors since it allows the use of sampling weights (Van Kerm, 2009). The user-written command SGINI in Stata was employed to calculate the concentration indices and their standard error in Stata (Van Kerm, 2009). We used Stata 11.2 for the analyses. 3. Results Table 1 shows the sociodemographic characteristics of Chilean and Finnish populations. Chileans were on average 3.8 years younger and had 2.9 years of education less than Finns on average. The Finnish population had a significantly lower proportion of abstinents and they reported drinking more than double the alcohol per week than Chileans. The proportion of population drinking in high volumes and the frequency of heavy episodic drinking was also significantly higher in Finland. 3.1. Socioeconomic inequalities in abstinence Results for abstinence from alcohol use are presented in Table 2. The overall CI for alcohol abstinence in Chile was −0.25 and in Finland −0.19, indicating that in both countries abstinence was higher among lower SES groups. Analysing by gender and age, in Chile men of both age groups showed no inequalities in alcohol abstinence, while abstinence in women of both age groups was markedly higher among lower SES groups. In Finland, alcohol abstinence was higher among lower SES groups in women aged 24–44 and in men and women aged 45–64. The degree of inequality was markedly higher in the older age group. 3.2. Socioeconomic inequalities in weekly volume of pure alcohol Table 3 shows the results for weekly volume of pure alcohol consumed. The overall CI in Chile was positive, indicating more drinking among those with higher SES. This result was largely driven by a positive CI for women aged 45–64; in the younger age group and in men aged 45–64 CIs were not statistically different
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
27
Table 1 Socio-demographic characteristics and alcohol use in Chile and Finland. Chilea
Socio-demographic characteristics Mean age (yrs) Mean education (yrs) Alcohol use Abstinence, % Mean weekly consumption of pure alcohol (grs/wk) Median weekly consumption of pure alcohol (grs/wk) Heavy volume drinking, % Heavy episodic drinking, % a
Finland
N
Total
Men
Women
N
Total
Men
Women
3477 3272
43.2 10.9
42.8 11.1
43.6 10.7
9994 9833
46.8 13.8
47.4 13.3
46.4 14.1
3221 3221 3221 3221 3221
22.7 30.1 0 2.7 16.2
13.7 47.6 0 4.1 28.3
31.5 13.0 0 1.4 4.3
9807 9940 9940 9940 9735
10.9 69.9 36 9.6 29.4
8.9 104.5 72 14.1 45.3
12.5 43.2 24 6.1 17.1
Population weighted.
Table 2 Concentration index (CIN ) of abstinence by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
All 25–44 25–44 45–64 45–64
Chile
Finland
N
CIN
95% conf. interval
Interpretation: higher rates among
N
CIN
95% conf. interval
Interpretation: higher rates among
3216 635 976 645 960
−0.25 −0.12 −0.22 −0.25 −0.26
(−0.33: −0.16) (−0.32; 0.08) (−0.39; −0.04) (−0.51; 0.02) (−0.37; −0.15)
Lower SES No inequality Lower SES No inequality Lower SES
9654 1669 2296 2251 3138
−0.19 −0.06 −0.08 −0.17 −0.28
(−0.23; −0.16) (−0.18; 0.06) (−0.16; −0.01) (−0.24; −0.01) (−0.33; −0.22)
Lower SES No inequality Lower SES Lower SES Lower SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
Table 3 Concentration index (CI) of weekly volume of pure alcohol by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
All 25–44 25–44 45–64 45–64
Chile
Finland
N
CI
95% conf. interval
Interpretation: higher rates among
N
CI
95% conf. interval
Interpretation: higher rates among
3216 635 976 645 960
0.13 0.05 0.03 0.17 0.26
(0.04; 0.23) (−0.04; 0.15) (−0.15; 0.21) (−0.05; 0.39) (0.14; 0.38)
Higher SES No inequality No inequality No inequality Higher SES
9781 1690 2328 2578 3185
−0.01 −0.04 0.003 0.05 0.06
(−0.03; 0.01) (−0.07; −0.01) (−0.03; 0.04) (0.02; 0.08) (0.03; 0.09)
No inequality Lower SES No inequality Higher SES Higher SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
from zero. In Finland, the overall CI was not statistically significant from zero. However, the CIs were positive and statistically significant for men and women aged 45–64 and negative for men aged 25–44.
equally across SES groups. In Finland, the overall CI as well as the CIs in each subgroup were negative; suggesting HED is consistently more prevalent in lower SES groups. Fig. 1 summarises the main results of the study.
3.3. Socioeconomic inequalities in heavy volume drinking
4. Discussion
Results for heavy volume drinking are shown in Table 4. In both countries, CIs for heavy volume drinking were statistically not significant from zero, at least partly because in Chile, the prevalence of heavy drinking was low in all subgroups, resulting in larger standard errors. The CIs were not statistically significant from zero for men and women aged 25–44 and men aged 45–64. On the other hand, women aged 45–64 showed a marked association between higher SES and heavy volume drinking, the highest found in the study. In Finland, we observed a negative CI among men aged 25–44, indicating that heavy volume drinking was more concentrated among those with lower SES status. Conversely, women aged 45–64 showed a positive CI, yet the magnitude of the inequality was modest.
Using the concentration index, we examined socioeconomic inequalities in patterns and volume of alcohol consumption in Chile and Finland. Our results show the existence of SES inequalities in several measures of alcohol use in both countries. A strong association was found between lower SES and alcohol abstinence in Chile and Finland. This was largely due to inequalities among women in Chile and in middle-age groups in Finland. The study found no socioeconomic differences in heavy volume drinking in either country in the whole working-age population, but we did observe a consistent pattern of higher weekly consumption of pure alcohol and heavy drinking among higher SES groups in Chilean and Finnish women aged 45–64. Finnish men aged 25–44 of lower SES men drank higher volumes of pure alcohol and showed higher rates of heavy drinking. Surprisingly, we did not observe socioeconomic inequalities in heavy episodic drinking in Chile, but in Finland there was a clear pattern of higher HED among lower SES groups. Therefore, and contrary to our expectations, we observed differential exposure to alcohol use in lower socioeconomic groups only in heavy episodic drinking in Finland; heavy volume drinking
3.4. Socioeconomic inequalities in heavy episodic drinking Table 5 shows the CIs for heavy episodic drinking (HED). In Chile, CIs for HED were not statistically significant in any subgroup, i.e., the “burden” of heavy episodic drinking appeared to be distributed
28
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
Table 4 Concentration index (CIN ) of heavy volume drinking by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
Chile
All 25–44 25–44 45–64 45–64
Finland
N
CIN
95% conf. interval
Interpretation: higher rates among
N
CIN
95% conf. interval
Interpretation: higher rates among
3216 635 976 645 960
0.04 0.07 −0.34 0.07 0.47
(−0.21; 0.3) (−0.43; 0.57) (−0.71; 0.03) (−0.34; 0.47) (0.18; 0.75)
No inequality No inequality No inequality No inequality Higher SES
9781 1690 2328 2578 3185
−0.01 −0.11 −0.04 0.06 0.11
(−0.05; 0.03) (−0.19; −0.03) (−0.14; 0.07) (−0.002; 0.12) (0.02; 0.19)
No inequality Lower SES No inequality No inequality Higher SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
Table 5 Concentration index (CIN ) of heavy episodic drinking by sex and age in Chile and Finland. Sex
Both Men Women Men Women
Age
All 25–44 25–44 45–64 45–64
Chile
Finland
N
CIN
SE
95% conf. interval
Interpretation: higher rates among
N
CIN
SE
95% conf. interval
Interpretation: higher rates among
3221 635 976 645 960
−0.04 −0.12 −0.14 −0.07 −0.07
0.04 0.06 0.12 0.06 0.18
(−0.14; 0.07) (−0.31; 0.06) (−0.39; 0.11) (−0.22; 0.09) (−0.43; 0.28)
No inequality No inequality No inequality No inequality No inequality
9586 1664 2292 2512 3118
−0.10 −0.10 −0.10 −0.06 −0.10
0.01 0.01 0.03 0.01 0.02
(−0.13; −0.08) (−0.16; −0.05) (−0.16; −0.04) (−0.10; −0.01) (−0.16; −0.05)
Lower SES Lower SES Lower SES Lower SES Lower SES
Note: Results with a p-value lower than 0.05 are highlighted in bold.
Fig. 1. Forest plot of concentration index for four alcohol measures in Chile and Finland.
showed either no inequalities or the opposite pattern to what we hypothesized. These findings stress the importance of examining both levels and patterns of alcohol use in research and policy interventions, as the use of a single indicator might bias conclusions. In the context of examining inequalities on risk factors for NCDs, most studies have relied on a single measure of alcohol use (Combes et al., 2011; Harper and Lynch, 2007; Hosseinpoor et al., 2012; Marmot, 1997). The few studies carrying out a comprehensive assessment have shown a complex picture of socioeconomic inequalities in alcohol consumption, in line with our findings (Casswell et al., 2003; van Oers et al., 1999).
4.1. Comparison with previous studies In our analysis, alcohol abstinence was concentrated among the lower SES groups. This is consistent with previous research showing a corresponding socioeconomic gradient (Bloomfield et al., 2000; Bloomfield et al., 2006; Marmot, 1997; van Oers et al., 1999). Lower SES groups often face harder priority-choices and might have to prioritize their expenditure on basic needs. In Chilean women, social norms and stigma related to alcohol consumption could also explain higher abstinence in lower SES groups. Regarding volume of pure alcohol and heavy volume drinking, studies in Baltic Countries and Brazil have also observed no inequalities or higher consumption on higher SES groups (Almeida-Filho
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
et al., 2005; Helasoja et al., 2007). Interestingly, the GENACIS study found higher rates of heavy volume drinking among women of higher SES in several Central European countries, but no inequalities in Nordic and Latin American countries (Bloomfield et al., 2006). This is in line with our findings and strengthen the idea of higher SES middle-aged women as a high-risk group. This could potentially be helpful for targeting interventions (such as brief interventions). Harper et al. used the concentration index to examine SES inequalities in heavy episodic drinking in the United States, observing that it was increasingly concentrated among those with lower education, which is consistent with our findings in Finland (Harper and Lynch, 2007). Other studies in the United States and the Czech Republic have yielded similar results (Dzúrová et al., 2011; Midanik and Clark, 1994). This could be explained by improved affordability of alcohol, different drinking cultures, exposure to alcohol marketing, prevalence of mental health disorders or ability to assess risk, among other factors. 4.2. Limitations The study has limitations that need to be addressed. First, we used surveys not designed to be comparable, as they have differences in their sampling design and data collection modes. However, these were carefully selected since they provided nationally representative data using very similar instruments. The Finnish data had lower participation rates, so there is a risk of selection bias due to non-response of, for example, people with lower education and higher prevalence of alcohol-related problems. Second, the only comparable SES variable available was years of education, whereas income or wealth have been most extensively used in the literature. Educational status might vary by age cohort, as levels of education have normally increased over the years. Third, the use of past week alcohol use might not be representative of a drinker’s usual patterns at an individual level. It could be, however, representative of the drinking patterns at a group level. Fourth, the results from both countries might not be applicable to countries with lower levels of development, even though they might be helpful for countries with high alcohol-related harm. This is the first step in a broader research agenda seeking to examine patterns of socioeconomic inequalities across countries. Next steps should consist of examining socioeconomic inequalities in alcohol-related harm, exploring a wider range of countries and examining secular trends. 5. Conclusions In two countries with the highest alcohol consumption in their regions, lower socioeconomic status was associated with higher prevalence of abstinence in both countries and heavy episodic drinking in Finland. Heavy volume drinking was more prevalent in middle-age women of high socioeconomic status. The results highlight the importance of multidimensional assessments of alcohol use and identify middle-aged higher SES women as a group for targeted alcohol interventions. Conflict of interest No conflict declared Contributors All authors materially participated in the research. SP, SK and PMäkelä designed the study, analysed and interpreted the data and contributed in writing the report. NM contributed in analysing and interpreting the data and writing the report. GV, SH and PMar-
29
gozzini collected the data and contributed in writing the report. All authors approved the final manuscript before submission. Funding This study was supported in part by the Finnish Medical Foundation. Acknowledgments The authors are thankful to Zhuo Adam Chen, John Ataguba and Báltica Cabieses for their helpful methodological suggestions on earlier versions of the manuscript. References Almeida-Filho, N., Lessa, I., Magalhaes, L., Araujo, M.J., Aquino, E., James, S.A., Kawachi, I., 2005. Social inequality and alcohol consumption-abuse in Bahia, Brazil—interactions of gender, ethnicity and social class. Soc. Psychiatry Psychiatr. Epidemiol. 40, 214–222. Babor, T., Caetano, R., Casswell, S., Edwards, G., Giesbrecht, N., Graham, K., Grube, J., Hill, L., Holder, H., Homel, R., Livingston, M., Österberg, E., Rehm, J., Room, R., Rossow, I., 2010. Alcohol: No Ordinary Commodity: Research and Public Policy. Oxford University Press, Oxford. Bellis, M.A., Hughes, K., Nicholls, J., Sheron, N., Gilmore, I., Jones, L., 2016. The alcohol harm paradox: using a national survey to explore how alcohol may disproportionately impact health in deprived individuals. BMC Public Health 16, 111. Bloomfield, K., Augustin, R., Kraus, L., 2000. Social inequalities in alcohol use and misuse in the German general population. J. Public Health 8, 230–242. Bloomfield, K., Allamani, A., Beck, F., Bergmark, K., Csemy, L., Eisenbach-Stangl, I., Elekes, Z., Gmel, G., Kerr-Correa, F., Knibbe, R., Makela, P., Monteiro, M., Medina, M., Nordlund, S., Obot, I., Plant, M., Rahav, G., Romero, M., 2005. Gender, Culture and Alcohol Problems: A Multi-national Study Project Final Report. Charite-Campus Benjamin Franklin, Berlin. Bloomfield, K., Grittner, U., Kramer, S., Gmel, G., 2006. Social inequalities in alcohol consumption and alcohol-related problems in the study countries of the EU concerted action ‘Gender: culture and alcohol problems: a multi-national study’. Alcohol 41, i26–36. Bloomfield, K., Grittner, U., Rasmussen, H.B., Petersen, H.C., 2008. Socio-demographic correlates of alcohol consumption in the Danish general population. Scand. J. Public Health 36, 580–588. Casswell, S., Pledger, M., Hooper, R., 2003. Socioeconomic status and drinking patterns in young adults. Addiction 98, 601–610. Chen, Z., Roy, K., 2009. Calculating concentration index with repetitive values of indicators of economic welfare. J. Health Econ. 28, 169–175. Combes, J.B., Gerdtham, U.G., Jarl, J., 2011. Equalisation of alcohol participation among socioeconomic groups over time: an analysis based on the total differential approach and longitudinal data from Sweden. Int. J. Equity Health 10, 10. Doku, D., Koivusilta, L., Rimpela, A., 2012. Socioeconomic differences in alcohol and drug use among Ghanaian adolescents. Addict. Behav. 37, 357–360. Dzúrová, D., Spilková, J., Pikhart, H., 2011. Social inequalities in alcohol consumption in the Czech Republic: a multilevel analysis. Health Place 16, 590–597. Giskes, K., Turrell, G., Bentley, R., Kavanagh, A., 2011. Individual and household-level socioeconomic position is associated with harmful alcohol consumption behaviours among adults. Aust. N. Z. J. Public Health 35, 270–277. Grittner, U., Kuntsche, S., Graham, K., Bloomfield, K., 2012. Social inequalities and gender differences in the experience of alcohol-related problems. Alcohol 47, 597–605. Grittner, U., Kuntsche, S., Gmel, G., Bloomfield, K., 2013. Alcohol consumption and social inequality at the individual and country levels—results from an international study. Eur. J. Public Health 23, 332–339. Harper, S., Lynch, J., 2007. Trends in socioeconomic inequalities in adult health behaviors among U.S. states, 1990–2004. Public Health Rep. 122, 177–189. Helakorpi, S., Hostila, A.-L., Virtanen, S., Uutela, A., 2012. Suomalaisen aikuisväestön terveyskäyttäytyminen ja terveys: Kevät 2011 [Health behavior and health among the Finnish adult population: Spring 2011. National Institute for Health and Welfare (THL), Helsinki. Helasoja, V., Lahelma, E., Prattala, R., Petkeviciene, J., Pudule, I., Tekkel, M., 2007. The sociodemographic patterning of drinking and binge drinking in Estonia, Latvia, Lithuania and Finland, 1994–2002. BMC Public Health 7, 241. Hosseinpoor, A.R., Bergen, N., Kunst, A., Harper, S., Guthold, R., Rekve, D., d’Espaignet, E.T., Naidoo, N., Chatterji, S., 2012. Socioeconomic inequalities in risk factors for non communicable diseases in low-income and middle-income countries: results from the World Health Survey. BMC Public Health 12, 912. Jones, L., Bates, G., McCoy, E., Bellis, M.A., 2015. Relationship between alcohol-attributable disease and socioeconomic status, and the role of alcohol consumption in this relationship: a systematic review and meta-analysis. BMC Public Health 15, 400.
30
S. Pe˜ na et al. / Drug and Alcohol Dependence 173 (2017) 24–30
Khang, Y.H., Yun, S.C., Lynch, J.W., 2008. Monitoring trends in socioeconomic health inequalities: it matters how you measure. BMC Public Health 8, 66. Lim, S.S., Vos, T., Flaxman, A.D., Danaei, G., Shibuya, K., Adair-Rohani, H., Amann, M., Anderson, H.R., Andrews, K.G., Aryee, M., Atkinson, C., Bacchus, L.J., Bahalim, A.N., Balakrishnan, K., Balmes, J., Barker-Collo, S., Baxter, A., Bell, M.L., Blore, J.D., Blyth, F., Bonner, C., Borges, G., Bourne, R., Boussinesq, M., Brauer, M., Brooks, P., Bruce, N.G., Brunekreef, B., Bryan-Hancock, C., Bucello, C., Buchbinder, R., Bull, F., Burnett, R.T., Byers, T.E., Calabria, B., Carapetis, J., Carnahan, E., Chafe, Z., Charlson, F., Chen, H., Chen, J.S., Cheng, A.T., Child, J.C., Cohen, A., Colson, K.E., Cowie, B.C., Darby, S., Darling, S., Davis, A., Degenhardt, L., Dentener, F., Des Jarlais, D.C., Devries, K., Dherani, M., Ding, E.L., Dorsey, E.R., Driscoll, T., Edmond, K., Ali, S.E., Engell, R.E., Erwin, P.J., Fahimi, S., Falder, G., Farzadfar, F., Ferrari, A., Finucane, M.M., Flaxman, S., Fowkes, F.G., Freedman, G., Freeman, M.K., Gakidou, E., Ghosh, S., Giovannucci, E., Gmel, G., Graham, K., Grainger, R., Grant, B., Gunnell, D., Gutierrez, H.R., Hall, W., Hoek, H.W., Hogan, A., Hosgood 3rd, H.D., Hoy, D., Hu, H., Hubbell, B.J., Hutchings, S.J., Ibeanusi, S.E., Jacklyn, G.L., Jasrasaria, R., Jonas, J.B., Kan, H., Kanis, J.A., Kassebaum, N., Kawakami, N., Khang, Y.H., Khatibzadeh, S., Khoo, J.P., Kok, C., Laden, F., et al., 2012. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 380, 2224–2260. Mackenbach, J.P., Kulhanova, I., Bopp, M., Borrell, C., Deboosere, P., Kovacs, K., Looman, C.W., Leinsalu, M., Makela, P., Martikainen, P., Menvielle, G., Rodriguez-Sanz, M., Rychtarikova, J., de Gelder, R., 2015. Inequalities in alcohol-related mortality in 17 European countries: a retrospective analysis of mortality registers. PLoS Med. 12, e1001909. Makela, P., Paljarvi, T., 2008. Do consequences of a given pattern of drinking vary by socioeconomic status? A mortality and hospitalisation follow-up for alcohol-related causes of the Finnish drinking habits surveys. J. Epidemiol. Commun. Health 62, 728–733. Makela, P., Tigerstedt, C., Mustonen, H., 2012. The Finnish drinking culture: change and continuity in the past 40 years. Drug Alcohol Rev. 31, 831–840. Margozzini, P., Sapag, J., 2015. Consumo riesgoso de alcohol en Chile: Tareas pendientes y oportunidades para las políticas públicas [Risky alcohol drinking in Chile: pending issues and opportunities for public policies]. Temas de la Agenda Pública: Centro de Políticas Públicas UC. Marmot, M., 1997. Inequality, deprivation and alcohol use. Addiction 92 (Suppl. 1), S13–20. McKee, M., Pomerleau, J., Robertson, A., Pudule, I., Grinberga, D., Kadziauskiene, K., Abaravicius, A., Vaask, S., 2000. Alcohol consumption in the baltic republics. J. Epidemiol. Commun. Health 54, 361–366. Midanik, L.T., Clark, W.B., 1994. The demographic distribution of US drinking patterns in 1990: description and trends from 1984. Am. J. Public Health 84, 1218–1222.
Ministerio de Salud, 2008. Estudio de Carga de Enfermedad y Carga Atribuible. Ministerio de Salud, Santiago, Chile. Ministerio de Salud, 2014. Encuesta Nacional de Salud 2009–2010 (Online Available from: http://www.minsal.cl/estudios encuestas salud/ Accessed 4 December 2016). O’Donnell, O., Van Doorslaer, E., Wagstaff, A., Lindelow, M., 2008. Analyzing Health Equity Using Household Survey Data: A Guide to Techniques and Their Implementation. International Bank for Reconstruction and Development/The World Bank, Washington DC. Probst, C., Roerecke, M., Behrendt, S., Rehm, J., 2014. Socioeconomic differences in alcohol-attributable mortality compared with all-cause mortality: a systematic review and meta-analysis. Int. J. Epidemiol. 43, 1314–1327. Rehm, J., Baliunas, D., Borges, G.L.G., Graham, K., Irving, H., Kehoe, T., Parry, C.D., Patra, J., Popova, S., Poznyak, V., Roerecke, M., Room, R., Samokhvalov, A.V., Taylor, B., 2010. The relation between different dimensions of alcohol consumption and burden of disease: an overview. Addiction 105, 817–843. Silveira, C.M., Siu, E.R., Anthony, J.C., Saito, L.P., de Andrade, A.G., Kutschenko, A., Viana, M.C., Wang, Y.P., Martins, S.S., Andrade, L.H., 2014. Drinking patterns and alcohol use disorders in Sao Paulo, Brazil: the role of neighborhood social deprivation and socioeconomic status. PLoS One 9, e108355. Statistics Finland, 2015. Statistics Finland PX-Web databases (Online Available from: http://pxnet2.stat.fi/PXWeb/pxweb/en/StatFin Accessed 4 December 2016). Subramanian, S.V., Nandy, S., Irving, M., Gordon, D., Davey Smith, G., 2005. Role of socioeconomic markers and state prohibition policy in predicting alcohol consumption among men and women in India: a multilevel statistical analysis. Bull. World Health Organ. 83, 829–836. Van Kerm, P., 2009. ‘sgini − Generalized Gini and Concentration coefficients (with factor decomposition) in Stata’, v1.1 (revised February 2010). CEPS/INSTEAD, Differdange, Luxemburg. Wagstaff, A., Paci, P., van Doorslaer, E., 1991. On the measurement of inequalities in health. Soc. Sci. Med. 33, 545–557. Wagstaff, A., 2011. The concentration index of a binary outcome revisited. Health Econ. 20, 1155–1160. World Health Organization, 2014a. Finland − Alcohol Profile. World Health Organization, Geneva, Switzerland. World Health Organization, 2014b. Global Status Report on Alcohol and Health. World Health Organization, Geneva. World Health Organization, 2015. Global Information System on Alcohol and Health (GISAH) (Online Available from: http://www.who.int/gho/alcohol/en/ Accessed on 4 December 2016). van Oers, J.A., Bongers, I.M., van de Goor, L.A., Garretsen, H.F., 1999. Alcohol consumption, alcohol-related problems, problem drinking, and socioeconomic status. Alcohol 34, 78–88.