Katikireddi SV, Whitley E, Lewsey J, Gray L, Leyland AH. Socioeconomic status as an effect modifier of alcohol consumption and harm: analysis of linked cohort data. Lancet Public Health 2017. Online publication May 10, 2017. http://dx.doi.org/10.1016/S2468-2667(17)30078-6
Background Alcohol-related mortality and morbidity are high in socioeconomically disadvantaged populations compared with individuals from advantaged areas. It is unclear if this increased harm reflects differences in alcohol consumption between these socioeconomic groups, reverse causation (ie, downward social selection for high-risk drinkers), or a greater risk of harm in individuals of low socioeconomic status compared with those of higher status after similar consumption. We aimed to investigate whether the harmful effects of alcohol differ by socioeconomic status, accounting for alcohol consumption and other health-related factors.
Methods The Scottish Health Surveys are record-linked cross-sectional surveys representative of the adult population of Scotland. We obtained baseline demographics and data for alcohol consumption (units per week and binge drinking) from Scottish Health Surveys done in 1995, 1998, 2003, 2008, 2009, 2010, 2011, and 2012. We matched these data to records for deaths, admissions, and prescriptions. The primary outcome was alcohol-attributable admission or death. The relation between alcohol-attributable harm and socioeconomic status was investigated for four measures (education level, social class, household income, and area-based deprivation) using Cox proportional hazards models. The potential for alcohol consumption and other risk factors (including smoking and body-mass index [BMI]) mediating social patterning was explored in separate regression models. Reverse causation was tested by comparing change in area deprivation over time.
Findings 50 236 participants (21 777 men and 28 459 women) were included in the analytical sample, with 429 986 person-years of follow-up. Low socioeconomic status was associated consistently with strikingly raised alcohol-attributable harms, including after adjustment for weekly consumption, binge drinking, BMI, and smoking. Evidence was noted of effect modification; for example, relative to light drinkers living in advantaged areas, the risk of alcohol-attributable admission or death for excessive drinkers was increased (hazard ratio 6·12, 95% CI 4·45–8·41 in advantaged areas; and 10·22, 7·73–13·53 in deprived areas). We found little support for reverse causation.
Interpretation Disadvantaged social groups have greater alcohol-attributable harms compared with individuals from advantaged areas for given levels of alcohol consumption, even after accounting for different drinking patterns, obesity, and smoking status at the individual level.
It has long been recognized from epidemiologic studies that both the positive and negative health effects of alcohol consumption are modified by the socio-economic status (SES) of individuals (Leyland et al, Mäkelä et al, McDonald et al, Probst et al, Mackenbach et al, Towers et al). Studies have suggested that moderate drinkers tend to have better health outcomes and fewer adverse effects from alcohol, and consistently show lower total mortality risks than non-drinkers or heavy drinkers. The usual explanation for differences in effect according to SES has been that higher SES subjects (those with higher education, income, job status, etc.) are more likely to be regular moderate drinkers (and more likely to be wine drinkers, mainly consumed with food), while lower SES subjects are more likely to binge drink (mainly beer or spirits), not with food, and may under-report their alcohol intake. However, scientific data on this relation are limited, and a key challenge to scientists is to determine the mechanisms by which these differences occur.
One potentially important factor is that the differing health outcomes according to SES relate to errors in judging the exposure, i.e., inadequate assessments of alcohol: not evaluating for binge drinking versus regular intake, consuming alcohol with or without food, type of beverage consumed, under-reporting of intake, etc. The authors of the present paper have made an attempt to evaluate a number of these factors as explanations for such findings; however, they still find large differences in health effects according to SES despite adjusting for other lifestyle factors, such as smoking and obesity, and conclude that the reasons remain unclear.
Specific comments from Forum members
Most Forum members stated that this appears to be a well-done study, although it must be pointed out that different subjects in this study were included in each of the surveys, and that only one assessment of alcohol intake for each subject was available. Still, each survey was population-based, and the assessment of outcome data was probably quite reasonable. Further, the authors had very good measures of SES, based on education, income, social class (professionals, skilled manual labor, unskilled labor, etc.), and a “measure of deprivation” of the geographic area of the subject (based on multiple measures of deprivation from government administrative data).
Forum member Ellison noted: “The “pattern of drinking” was estimated only from the weekly total amount of alcohol reported and a question on binge drinking; details on the number of days alcohol was consumed, the type of beverage, whether or not it was with food, etc., were not available. Thus, it was not possible to have a good determination of whether or not the reliability of reporting alcohol consumption may have varied by SES. The main outcome for the analyses was either alcohol-attributable admission to hospital or alcohol-attributable death, but data on these two outcomes separately or on total mortality were not presented.”
The authors conclude that smoking and obesity could not explain the marked differences in outcome for the different SES groups in terms of adverse effects of alcohol consumption, even though subjects in each group supposedly consumed similar amounts of alcohol. They conclude: “Our study findings show that the risk of alcohol-attributable harms among moderate alcohol consumers of low socioeconomic status is greater than for those who drink heavily but are socioeconomically advantaged. The lived experiences of poverty shape the emergence of health outcomes, not only through health-related behaviours but also as a result of poor material circumstances and psychosocial stresses. Poverty might, therefore, reduce resilience to disease, predisposing drinkers of low socioeconomic status to greater health harms despite exposure to similar levels of risk factors as drinkers of high socioeconomic status.” Reviewer Ellison stated: “I am not sure that these investigators were able to judge the intake or the pattern of drinking well enough, and they did not consider the type of beverage, so this conclusion may be premature. Differences in actual alcohol consumption, and the pattern of drinking, may still be factors affecting health outcomes.” Forum member Skovenborg considered that this was a good paper but added that “It is an example of the difficulties in separating the drink from the drinker in observational studies.”
Forum member Waterhouse noted that the risk ratios for adverse outcomes from lower SES in this study showed very little change when alcohol was added to the equation, perhaps further emphasizing the importance of low SES. He also commented on the lack of adjustment for beverage type in the present paper. “Considering the very close association in the questions studied, it is most surprising that the authors of the current paper do not cite or compare their data to those of Flensborg-Madsen et al from 2008. Those Danish authors found that individuals who include wine when they drink alcohol have lower risk of an alcohol-use disorder, even with adjustments for education, income, smoking, and cohabitation status. Those authors concluded that other lifestyle or personality factors, for which they could not adjust, were associated with beverage preference, resulting in preferable outcomes for consumers of wine.” Other Forum members also thought it unfortunate that the present authors could not separate results according to beverage type.
Forum member De Gaetano, with input from his colleagues Simona Costanzo and Marialaura Bonaccio, stated: “We agree that the findings in this paper from a large Scottish cohort showed that alcohol-attributable harms are much higher in disadvantaged social groups. Inequities still persisted after accounting for differences in consumption and binge drinking. The paper is of interest and add new insights on the pathways by which alcohol consumption affects a number of health outcomes. Unfortunately, owing to the low rate of events, the authors have had to combine the two categories of ‘never’ and ‘ex-drinkers’ into only one for the risk analysis (the authors acknowledged this limitation). Similarly, it was not possible to analyse separately the two outcomes: ‘alcohol-attributable admission’ and ‘death’.
“The Scottish findings appear to be in line with evidence from the “Moli-sani” study, an Italian cohort of 25,000 men and women aged >35 y randomized from the general population, showing a similar interaction between Mediterranean diet (MD) and SES factors in relation to risk of developing cardiovascular disease (Bonaccio et al). Basically, we found that the adherence to MD was significantly associated with lower CVD risk in higher but not in lower SES groups, with SES acting as an effect modifier of such association.
“Possible explanations were found from our study in diet-related differences across SES groups sharing similar adherence to the MD. Indeed, at a comparable adherence to the MD, individuals at different SES level (measured either as educational level or household income) still showed diet-related disparities, such as dietary antioxidant intake, fibre, minerals, dietary diversity, all being more favourable in the highest SES groups. Moreover, the most advantaged also showed dietary behaviours more in line with healthy recommendations, such as selecting whole grain foods or healthier cooking methods. Our findings indicated that, in terms of the net effects of adherence to a healthy dietary pattern, low SES individuals still presented substantial differences in diet than higher SES individuals.
“It might be that higher SES groups are more prone to select heathier products from the bundle of foods included in the MD and also the quality of foods possibly differs by SES. For example, subjects with lower SES may be less likely to choose foods rich in antioxidants, vitamins and nutrients, or even less likely to select Tuscany extra virgin olive oil over uncontrolled commercial types, which may ultimately have different health effects.
“As far as the Scottish cohort evaluated in the present study is concerned, one additional pathway to be explored in order to explain the differences documented may be the quality and the origin of the beverages consumed. One major limitation of the Scottish study may rely on the lack of information on dietary habits which may substantially vary across SES groups and are closely associated with alcohol drinking patterns. Moreover, we actually do not know whether, at similar alcohol amounts consumed, pattern of consumption may be different, e.g., drinking with meals or not.”
Reviewer Van Velden noted: “The most important factor is that the higher socio economic groups have a healthier lifestyle. It is important to note that alcohol must not be seen in isolation, but part of the healthy Mediterranean-type lifestyle. Lower SES groups tend to be more overweight, and do not consume a healthy diet, smoke more, and do not regularly participate in exercise. Scientists are unable to adjust adequately for all such lifestyle factors.”
Forum member Finkel had some interesting comments: “The vicissitudes of low-SES life are likely to render those so afflicted more vulnerable to the adversities that go with any level of alcohol consumption and less able to avail themselves of the healthful effects of moderate drinking. Might the sources of beverage alcohol for low-SES populations be less ‘pure,’ more likely adulterated, infected, unrefined than for more affluent portions of the population? Even the ice is likely less pristine. There is indeed an increasing number of studies indicating differential health effects of varying gastointestinal flora.”
Another potential factor affecting differences in alcohol effects by SES: In an attempt to explain why wine consumption, in particular, seems to decrease the risk of cognitive decline, Esteban-Fernández et al have recently shown that the exact composition of wine metabolites is important in the protective neuronal effect. “And this composition depends on gut microbiota composition, as the intestinal flora breaks down the wine into the different metabolites.” Thus, the gut microbiota is yet another variable in determining the risk and benefits on health of alcohol consumption. Fewer beneficial effects of moderate alcohol intake among lower SES subjects might even be related to their underlying gut microbiota. While many studies describe difference in gut microbiota and their effects on health, we are not aware of studies of the gut microbiome as being related to SES.
Forum member Thelle agreed with other reviewers that there were some important data missing from the analyses in this paper. “However, I don’t (and neither do most other reviewers) believe that the results are due to systematic errors or bias. Residual confounding is always an issue. Most chronic disorders do show a strong SES gradient (breast cancer being an exception) even if classical well known risk factors are accounted for. The idea that gut microbia might play a modifying role is interesting, especially as diet influences the microbia, and is strongly associated to SES. Thus a potential research avenue would be to assess the SES, diet, wine and gut microbia to see whether the latter changes in a potentially healthy direction. There are a number of papers on gut microbia and wine (phenols, flavonoids) out there, but I haven’t found any addressing SES as such.”
Bonaccio M, Di Castelnuovo A, Pounis G, Costanzo S, Persichillo M, Donati MB, de Gaetano G, Iacoviello L. Cardiovascular Protection by the Mediterranean Diet Differs Across Socioeconomic Groups: Prospective Findings From The MOLISANI Study. Abstract, American Heart Association. Circulation 2016;133:AMP20
Esteban-Fernández A, Rendeiro C, Spencer JP, Del Coso DG, de Llano MD, Bartolomé B, Moreno-Arribas MV. Neuroprotective effects of selected microbialderived phenolic metabolites and aroma compounds from wine in human sh-sy5y neuroblastoma cells and their putative mechanisms of action. Front Nutr 2017;4:3.
Flensborg-Madsen T, Knop J, Mortensen EL, Becker U, Makhija N, Sher L, Groenbeck M. Beverage preference and risk of alcohol-use disorders: A Danish prospective cohort study. J Stud Alcohol Drugs 2008;69:371-377.
Leyland A, Dundas R, McLoone P, Boddy FA. Cause-specific inequalities in mortality in Scotland: two decades of change—a population-based study. BMC Public Health 2007; 7: 172.
Mackenbach JP, Kulhánová I, Bopp M, et al. Inequalities in alcohol-related mortality in 17 European countries: a retrospective analysis of mortality registers. PLoS Med 2015; 12: e1001909.
Mäkelä P, Paljärvi T. 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 Community Health 2008; 62: 728–33.
McDonald SA, Hutchinson SJ, Bird SM, et al. Association of self-reported alcohol use and hospitalization for an alcohol-related cause in Scotland: a record-linkage study of 23 183 individuals. Addiction 2009; 104: 593–602.
Probst C, Roerecke M, Behrendt S, Rehm J. Socioeconomic differences in alcohol-attributable mortality compared with all-cause mortality: a systematic review and meta-analysis. Int J Epidemiol 2014; 43: 1314–27.
Towers A, Philipp M, Dulin P, Allen J. The “Health Benefits” of Moderate Drinking in Older Adults may be Better Explained by Socioeconomic Status. Pre-publication: J Gerontol B Psychol Sci Soc Sci 2016. doi:10.1093/geronb/gbw152
The majority of epidemiologic studies have shown that both the positive and negative health effects of alcohol consumption are modified by the socio-economic status (SES) of individuals, with truly moderate drinkers who are from higher SES strata having better health outcomes and fewer adverse effects from alcohol than lower-SES subjects supposedly consuming similar amounts. One potential explanation for this has been that higher SES subjects (those with higher education, income, job status, etc.) are more likely to be regular moderate drinkers, more likely to be wine drinkers, and to drink primarily with meals, while lower SES subjects are more likely to binge drink (mainly beer or spirits), not with food, and may tend to under-report their alcohol intake. However, there are limited scientific data on the specific mechanisms by which such differences occur.
The authors of the present paper have used data on alcohol intake collected in the Scottish Health Surveys from more than 50,000 subjects to investigate whether the harmful effects of alcohol differ by socioeconomic status when accounting for a number of SES factors. They had very good measures of SES based on education, income, social class (professionals, skilled manual labor, unskilled labor, etc.), and a “measure of deprivation” of the geographic area of the subject (based on multiple measures of deprivation from government administrative data). Their health outcome measures were based on public records for deaths, admissions, and prescriptions; their primary outcome was alcohol-attributable admission or death. Their data show that, even without considering alcohol consumption, lower SES subjects were at much greater risk for poor health outcomes. And, while risks associated with increasing amounts of alcohol were present in all SES groups, they were much greater in the lower groups.
Forum members considered this to be a well-done study, but pointed out that there was limited information on the pattern of drinking and no data on the type of alcoholic beverage consumed or whether or not it was consumed with food. There was no dietary information or genetic information available, and little data on access to health care. While the study showed very marked differences in adverse health outcomes according to SES, the authors found that the factors they considered explained little of the health effects seen. They concluded: “Disadvantaged social groups have greater alcohol-attributable harms compared with individuals from advantaged areas for given levels of alcohol consumption, even after accounting for different drinking patterns, obesity, and smoking status at the individual level.”
Forum members agree that there may be many other lifestyle factors associated with poverty that scientists are unaware of, or unable to adequately control for, in their analyses. Differences by SES in the assessment of alcohol intake remains as a potential factor (especially given that higher SES subjects are usually found to consume alcohol regularly and moderately, less in binges, more often consuming wine with food, etc.). The authors of the present paper have evaluated a number of factors (including smoking status and obesity) as potential explanations for the poorer outcomes among lower SES subjects, and still found large differences in health effects despite adjusting for a number of such lifestyle factors. Thus, the specific reasons why lower SES subjects have worse health overall and poorer effects after alcohol consumption remain unclear.
* * * * *
Comments on this critique by the International Scientific Forum on Alcohol Research have been provided by the following members:
Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy
R. Curtis Ellison, MD, Professor of Medicine, Boston University School of Medicine, Boston, MA, USA
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Dag S. Thelle, MD, PhD, Department of Biostatistics, Institute of Basic Medical Sciences, University of Oslo, Norway; Section for Epidemiology and Social Medicine, Sahlgrenska Academy, University of Gothenburg, Sweden
David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa
Andrew L. Waterhouse, PhD, Department of Viticulture and Enology, University of California, Davis, USA