“The alcohol-harm paradox”: Understanding socioeconomic inequalities in liver disease

Summary

The alcohol-harm paradox (AHP) refers to the fact that people from lower socioeconomic groups experience higher rates of alcohol-related illness despite consuming the same or even lower amounts of alcohol than their more affluent counterparts. While differences in drinking patterns and associations with other risky behaviours partially explain the paradox, they do not fully account for the disparities in morbidity and mortality across socioeconomic groups. The existence of an alcohol-harm paradox in liver disease has been demonstrated in many countries worldwide. Recently, the COVID-19 pandemic further exacerbated these differences and led to an increase in alcohol intake and alcohol-related mortality among racial and ethnic minorities in the United States. Approaches to limit alcohol sales, through introduction of minimum unit pricing or taxation, have led to reductions in alcohol-related liver disease, particularly in socioeconomically deprived areas. Disparities in access to treatment of alcohol use disorder, liver disease and liver transplantation further contribute to the AHP. This review focuses on the AHP, its impact on liver disease and the multi-level strategy that will be required to curb this phenomenon.

Graphical abstract

Keypoints.

• •The phenomenon that lower socioeconomic groups experience higher rates of alcohol-related illness despite consuming the same or even lower amounts of alcohol than their more affluent counterparts is called the “alcohol-harm paradox” (AHP).
• •The AHP and its impact on liver disease has been demonstrated in many European and American countries.
• •Differences in alcohol drinking pattern and association with other risky health behaviours partially explain the AHP.
• •Inconsistent access to treatment for alcohol use disorder, liver disease, and liver transplantation further contribute to the AHP.
• •Strategies to limit alcohol sales through minimum unit pricing or taxation, or reduce the density of alcohol outlets, have proven effective to reduce hospitalization and death due to alcohol, particularly in socioeconomically deprived areas.

Definition of alcohol-harm paradox and its potential causes

Alcohol intake is widely known to negatively affect physical and mental health, as well as social well-being. Increased consumption is associated with unemployment, financial instability, depression, medical comorbidities, and higher mortality rates.¹ Rising health inequality among socioeconomic groups presents growing challenges for individuals living in poverty, leading to poorer health outcomes and increased healthcare costs.² Poverty has also been shown to increase the likelihood of alcohol-related illness in multiple countries, demonstrating the universality of this social issue.³ The “alcohol-harm paradox” (AHP), which is the phenomenon that lower socioeconomic groups experience higher rates of alcohol-related illness despite consuming the same or even lower amounts of alcohol than their more affluent counterparts, describes an under-recognised aspect of our understanding of alcohol use.^4,5 This phenomenon has been extensively described in the United Kingdom, other European countries, as well as in Australia, New Zealand, North America, and Latin America.⁶ In the United States, a similar and likely related paradox seems to exist along ethnic/racial lines, with people from African American and Hispanic/Latino origins experiencing more alcohol-related harm, even after adjusting for alcohol consumption and demographic characteristics.⁷ The term “paradox” is somewhat controversial, as it may not be contradictory or entirely unexpected for patients at higher socioeconomic risk to experience escalating adverse health consequences. However, it is notable that the uneven burden of alcohol-related hospitalisation and mortality according to socio-economic level only seemed to arise in the last few decades.⁶ Prior to 1980, studies found that excess mortality from alcohol consumption was proportional to the amount of alcohol intake, regardless of social class or educational level.^8,9

Potential causes of the alcohol-harm paradox

Under-reporting actual alcohol consumption and frequency has been suggested as a possible explanation for the AHP, possibly due to methodological issues with alcohol consumption recording or missing data from certain groups, such as people experiencing homelessness.¹⁰ However, there is still debate as to whether more deprived or more affluent groups tend to underreport alcohol use.^4,11 Diagnostic coding bias may also play a role, as alcohol use can be underestimated or misclassified by providers in a differential manner. This may lead to underrepresentation of alcohol use disorder (AUD) and alcohol-related liver disease (ALD) among more privileged individuals, with liver disease potentially attributed to non-alcohol-related causes. In epidemiologic research, cases of ALD are often identified by codes for ALD or alcohol-associated hepatitis, or the combination of cirrhosis and AUD.¹² Studies have shown significant misclassification of AUD and ALD: >30% of AUD diagnoses may be missed using ICD coding, and >65% of patients classified as having non-alcohol-related cirrhosis have documented histories of heavy alcohol use.^13,14

Differences in drinking patterns rather than weekly amount of alcohol have also been implicated, with people from lower socio-economic status engaging in less frequent but heavier episodic drinking than their more privileged counterparts.^4,[15], [16], [17], [18] Binge drinking affects the liver more acutely than gradually paced alcohol intake, increasing the risk of alcohol intoxication, alcohol-associated hepatitis, and liver failure.^3,19 These differences in binge-drinking patterns could explain up to 25% of the socioeconomic inequalities in alcohol-related health problems in some countries.^3,20,21 Alcohol consumption without food intake can also lead to a higher incidence of cirrhosis.²² Whether this plays into the AHP is understudied. Yet, there is some evidence that people from low socio-economic status tend to drink more often without meals²³ and that the protective effect of drinking with food on all-cause mortality is mostly apparent in people with socioeconomic risk factors.²⁴ Additionally, the choice of alcoholic beverage may factor into the AHP. Deprived individuals are more likely to drink beer and spirits,⁴ while more affluent individuals tend to drink wine, which may be associated with cardioprotective qualities.[25], [26], [27]

Several studies have reported the importance of the association with other risky health-related behaviours among people with lower educational and social levels. Indeed, they are more likely to be overweight, smoke, and/or have a poor diet and exercise routine,⁴ which, taken together, potentially makes them more vulnerable to alcohol-related harm. Studies in Sweden and Finland found no interaction between BMI and low income, and adjusting for physical inactivity and high BMI did not reduce the socio-economic difference in all-cause²⁰ or alcohol-related¹¹ mortality. On the other hand, the interaction between low income and smoking led to 11.4 extra deaths per 10,000 person-years in Finland.¹¹ An English survey found the association with smoking, higher BMI and poor diet and exercise to be stronger in deprived individuals consuming what they called “increased risk levels” of alcohol (>168–400 g/week in males, >112–280 g/week in females) compared to higher risk drinkers (>400 g/week in males, >280 g/week in females).⁴ The updated nomenclature for steatotic liver disease (SLD) recognises the contributions of cardiometabolic risk factors to the liver disease, categorising patients with hepatic steatosis into metabolic dysfunction-associated steatotic liver disease (MASLD), ALD, and MetALD based on their reported alcohol consumption.²⁸ Cardiometabolic risk factors, which include hypertension, hyperlipidaemia, diabetes, and obesity, are more common in socioeconomically disadvantaged populations.^29,30 In the United States, the burden of unfavourable social determinants of health (SDOH) has been disproportionately found in those with ALD-predominant MetALD and ALD, with higher rates of food insecurity, limited healthcare access, and single living (not married nor living with a partner), relative to those with MASLD.³¹ In this study, the association of Hispanic ethnicity with SLD was attributable not only to metabolic risk factors and alcohol use but also to social factors including cumulative SDOH and foreign-born status. These findings suggest a negative synergistic contribution of these risk factors to the perceived AHP.

Furthermore, limited access to primary healthcare providers, specialty care, and mental health resources – due to associated costs – can increase the risk of disease progression, including of alcohol-related conditions.³² Treatment of ALD inherently requires treatment of the underlying AUD, yet access to these resources can be unreliable. Gender and racial/ethnic differences in access to treatment for AUD have been described, and admission to treatment is dependent on a variety of factors including sociodemographic characteristics, cultural expectations, time, transportation, social supports, and the availability of services.[33], [34], [35], [36] In the United States, Black patients seeking addiction treatment commonly report experiencing racial discrimination in the medical setting, which may lead to mistrust and decrease help-seeking.³⁷ Similarly, Black patients are less likely than White or Hispanic patients to receive pharmacotherapy for AUDs.³⁸ In addition, higher geographic density of gastroenterologists has been associated with lower ALD mortality, with physician-dependent variables influencing access to liver transplantation, yet there remains significant regional variability in the availability of such subspecialty care.^39,40 In this context, delayed presentations or referrals may also contribute to the observation that patients of lower socioeconomic status experience a higher burden of symptomatic disease.

Finally, more severe alcohol-consuming behaviour can lead to unemployment and adverse socioeconomic circumstances, and could thus contribute to the AHP via reverse causation.^6,20,41 Indeed, a causal risk pathway of socioeconomic deprivation leading to increased risk of aversive experience or trauma, followed by internalized symptoms, drinking to cope, and alcohol dependence, has been described.⁴²

These various factors likely contribute to this paradox but may not fully explain the difference in alcohol-related morbidity and mortality observed between people from different socio-economic backgrounds. Indeed, one study found that after adjusting for alcohol intake and other risky health behaviours such as smoking and high BMI, socioeconomically disadvantaged populations still experienced a three-fold increase in alcohol-related harm compared to people with more privileged status.⁵ It is also worth noting that the association of deprivation with alcohol-related harm seems stronger in males than in females.^10,43,44 Overall, most studies aiming to understand the AHP have focused on individualised behavioural causes, rather than the systemic social, economic and health inequalities that also deserve attention (Fig. 1).

Fig. 1.

Health gradient in alcohol-related liver disease.

The illustration is conceptual and the thickness of each layer does not reflect the weight of the associated factor.

Alcohol-harm paradox in liver disease

Alcohol has detrimental effects on health and increases the risks of depression, cardiovascular disease, cancer, and liver disease, which can be debilitating and life-threatening. ALD is a primary preventable cause of morbidity and mortality worldwide, encompassing a spectrum of disorders that include hepatic steatosis, alcohol-associated hepatitis, and alcohol-associated cirrhosis. SLD due to alcohol can be exacerbated by additional factors, including diet and exercise, obesity, metabolic diseases, genetics (e.g. PNPLA3 mutations) and medications.⁴⁵ Approximately 10-20% of people with AUD develop hepatitis or cirrhosis, and alcohol contributes to half of all cases of cirrhosis worldwide. Mortality from ALD, particularly alcohol-related cirrhosis, typically correlates directly with per capita alcohol consumption and income inequality.^46,47

Rates of ALD, as well as morbidity and mortality, vary by region and country. In the United Kingdom, an overall trend from drinking beers to stronger wines and spirits was correlated with a fourfold increase in mortality from chronic liver disease. Conversely, declines in overall alcohol consumption in France, Italy, Spain, and the Baltic countries were followed by falls in cirrhosis-related mortality.⁴⁸

Europe has the highest annual per capita alcohol consumption rate in the world, with Eastern Europe leading at 15.7 L per person. Rates of ALD in Europe have been declining, down to 3.0% by 2023, partly due to declining alcohol consumption.⁴⁹ This trend is especially evident in Eastern Europe, where minimum unit pricing and increased tax on alcoholic products in Baltic countries (Estonia, Latvia, and Lithuania) have reduced alcohol purchases.[50], [51], [52]

Asian countries such as China, India, South Korea, and Japan have slightly lower per capita alcohol consumption, ranging from 4.3 to 12.3 L per person, yet still experience high rates of deaths due to cirrhosis and disability-adjusted life years attributable to alcohol consumption (456.1 per 100,000), following the broader Asian region.⁴⁷ Rising rates of ALD-related mortality in Asia may be linked to several factors. Asian individuals are genetically predisposed to elevated acetaldehyde exposure, owing to variants in alcohol dehydrogenase, making them more susceptible to alcohol-related liver injury. From an epidemiological perspective, higher rates of comorbid viral hepatitis B infection along with ALD may contribute to higher rates of liver-related death. Additionally, growing economies such as China and India have shown rapid increases in per capita alcohol consumption.⁴⁷

Global impact

Most studies on the AHP have looked at non-liver-specific outcomes such as all-cause mortality or acute and chronic alcohol-related harm. Yet, specific evidence of the AHP and its impact on liver health has been demonstrated across multiple geographic settings, including the Americas and Europe³ (Table 1). In Europe, a socio-economic gradient was identified for most alcohol-associated hospital admissions in England² and for new diagnoses of ALD in Denmark.⁵³ In 2015, higher rates of alcohol-related mortality were found among lower educational and occupational groups in 17 European countries,⁵⁴ with the greatest disparities observed in Eastern Europe, Finland and Denmark. Interestingly, based on the WHO mortality database, Finland and Denmark also have the lowest proportions of unrecorded aetiology of liver disease using ICD coding: 9% in Finland compared to 62% in Italy.⁵⁵ It is thus conceivable that socioeconomic disparities in alcohol-related mortalities in other European countries may be missed by ICD coding misclassification.

Table 1.

Alcohol-harm paradox in liver-related studies.

Authors	Region	Study design	Main findings
	Europe
Petrovski et al., 201189,90	Eastern Europe (Hungary)	Case control study	•Increased risk of chronic liver disease in patients with low socio-economic status only partially explained by conventional behavioural risk factors (age, smoking, alcohol use and physical activity)
Sadler et al., 20172	United Kingdom (England)	Use of national health service hospital admissions (2010 – 2013)	•Socio-economic gradient for most alcohol-associated hospital admissions •Greatest inequalities in conditions associated with alcohol dependence, such as liver disease and mental and behavioural conditions •Steeper gradients for men than women
Askgaard et al., 202153	Northern Europe (Denmark)	Use of national registries (2009 – 2018)	•Among patients newly diagnosed with ALD: low or medium-low educational level in 86%; employment in 20% •Inverse correlation between incidence of ALD and educational level •Inverse correlation between incidence of ALD and employment status •Relative difference in incidence of ALD between educational levels larger in younger age groups (age 30–39)
	Latin America
Arab et al., 202056	Latin America	Regional Health Reports	•In Latin America, lower-income countries have higher mortality due to alcohol-related cirrhosis despite reporting lower alcohol consumption per capita
Oneto et al., 202191	Latin America (Chile)	National health survey (2016 -2017)	•In women: hazardous alcohol consumption only increased ALD among those with high income level who also presented obesity or metabolic syndrome in combination with type 2 diabetes •In men, hazardous alcohol consumption only increased ALD among those with low-income level (even those without comorbidities)
	North America
Major et al., 201464	North America (USA)	NIH American Association of Retired Persons Health study	•Area socio-economic deprivation is associated with increased risk of chronic liver disease after accounting for health risk factors, including alcohol consumption (HR 1.78)
Case and Deaton, 201792	North America (USA)	National surveys	•Increase in ALD-related mortality between 1998 and 2015 among White non-Hispanics aged 50-54 with a lower educational level (high school or less vs. bachelor or more)
Damjanovska et al., 202393	North America (USA)	Electronic records from health care systems (1999 - 2021)	•Highest relative increase in alcohol-related hepatitis admissions in African American patients (OR 2.63) during the COVID-19 pandemic
Ayares et al., 202494	North America (USA)	National databases (2011 – 2018)	•After accounting for key social and biological health determinants, the Hispanic population showed an increased risk of ALD, even with lower overall alcohol consumption. •Higher prevalence of heavy episode drinkers among this group

ALD, alcohol-related liver disease; HR, hazard ratio; OR, odds ratio.

Higher mortality due to alcohol-related cirrhosis was also reported in lower-income countries across Latin America.⁵⁶ In Brazil, where the most popular alcoholic drink is cachaça (with 40% alcohol by volume) and heavy alcohol consumption costs less than a 2,000 kcal/day staple diet, lower education levels were correlated with higher risk of alcohol-related illness.^57,58

In the United States, alcohol use has disproportionately affected communities of lower income. Alcohol outlets cluster in urban areas with higher levels of poverty and a higher proportion of ethnic minorities^59,60 – particularly in neighbourhoods with histories of redlining, a discriminatory practice related to financial resources that created geographically deprived areas where health inequities persist today.[61], [62], [63] Increased neighbourhood alcohol outlet density has been associated with adverse alcohol-related health outcomes including liver problems, likely contributing to the differences observed in liver-related mortality by socioeconomic status.^64,65

Impact of COVID-19

In the United States, the COVID-19 pandemic significantly impacted racial and ethnic minorities, immigrants, homeless individuals, inmates, and essential workers who often lived in tight and overpopulated housing units, high-density neighbourhoods, or areas with limited access to healthcare.⁶⁶ These communities that were hit hardest by COVID-19 also experienced surges in alcohol purchases and alcohol use,⁶⁷ partly because of social isolation and stay-at-home orders to “flatten the curve” and contain the pandemic – policies that led to disruptions in school and work, and higher rates of depression and anxiety.[68], [69], [70], [71] Overall alcohol intake rose significantly during this period, with national sales increasing by up to 54% and online sales surging by 262%.^72,73 Admissions for alcohol-associated hepatitis increased and were linked to higher mortality, indicating more severe disease than previously observed.⁷⁴ The number of liver transplants for alcohol-associated hepatitis increased by more than 50%, and ALD-related mortality increased by 20%, more so among Black and Hispanic communities.⁷⁵ Outside of the United States, data suggest that mortality rates from chronic liver disease in other countries were less affected by the pandemic.^76,77

Potential approaches to tackle the alcohol-harm paradox in liver disease

Alcohol sales have been closely linked to the burden of ALD. In 2018, a minimum unit pricing was introduced in Scotland on alcohol drinks sold to the public. The minimum unit pricing sets a legal minimum price below which alcohol cannot be sold.⁷⁸ Over a period of 3 years, this legislation led to a reduction of alcohol sales by 3%, which translated into reductions of hospitalisation and death attributable to alcohol of 4.1% and 13.4%, respectively. This was also associated with a significant reduction in ALD, with the greatest impact observed in the 40% most socioeconomically deprived areas.⁷⁸ This impact of the minimum pricing unit on alcohol-related hospitalisation and inequalities is consistent with what modelling studies or natural experiments conducted in Canada and the United Kingdom have observed.⁷⁹ Increased taxation of alcoholic beverages has also been implemented in several countries; such policies in Estonia, Latvia, and Lithuania have successfully curbed alcohol sales with reductions in all-cause and alcohol-attributable mortality.[50], [51], [52] By contrast, less restrictive alcohol policies in Poland were associated with higher rates of liver-related mortality.⁵¹ Some may view these strategies as a “tax on the poor,” as its impact is felt disproportionately by the poor; however, proponents suggest that this group experiences the most alcohol-related harms and that this is why such policies are effective.[80], [81], [82]

Other public health policies could help reduce the AHP, such as limiting access to alcohol by reducing the density of alcohol outlets. Interventions to address the co-morbid and compound risk of cardiometabolic risk factors on the development and progression of ALD may include taxes (e.g. soda tax), calorie labels on fast foods, restrictions on the sale of ultra-processed foods, and programmes aimed at reducing obesity and tobacco use. Population-level screening targeting at-risk populations can detect hazardous alcohol use and facilitate timely intervention to prevent adverse health consequences. Improved access to healthcare, including treatment for AUD, liver disease care and liver transplantation, would also help to address disparities in liver-related health outcomes arising from the AHP (Fig. 2).

Fig. 2.

Multiple factors involved in the alcohol-harm paradox.

AHP, alcohol-harm paradox; ALD, alcohol-related liver disease; AUD, alcohol use disorder; LT, liver transplantation.

Impact of the alcohol-harm paradox on liver transplantation

Severe ALD, including alcohol-associated hepatitis and decompensated cirrhosis, is associated with high mortality. In such cases, liver transplantation is often the only curative treatment option. Transplantation rates have historically been lower in individuals from lower socioeconomic groups, in part due to disparities seen in social support, access to health insurance, and education. These patients may find it more difficult to meet the strict selection criteria for liver transplantation, with lower health literacy, fewer social supports, and reduced access to consistent healthcare or health insurance, and may be disadvantaged when assessed on usual selection criteria, contributing to inequities in access to liver transplant.⁸³ This population also encounters upstream barriers to being considered for or referred to a liver transplant programme.^84,85 Increased area deprivation has been correlated with worse outcomes in decompensated cirrhosis, transplant waitlisting, and all-cause mortality. Factors such as lower income, lower levels of education, poor household conditions, and decreased social support are linked to a lower likelihood of being waitlisted for transplant and reduced survival.⁸⁵ There are relatively few contemporary studies investigating how SDOH may influence post-transplant disparities in outcome, but low annual income and limited literacy have been associated with post-transplant non-adherence.^86,87

Summary

In conclusion, tackling the AHP will require a multi-level strategy (Table 2). Individual behavioural measures aiming to reduce the prevalence or frequency of episodic heavy drinking have been shown to be effective. Understanding patterns of health and disease should integrate biologic and social factors.⁸⁸ In order to be effective and achieve equity in health outcomes, interventions will need to take into account the socioeconomic context and provide additional supports to at-risk groups. Future studies should focus on deepening our understanding of the mechanisms driving the AHP to address these complex and intricate eco-social factors.

Table 2.

Causes and potential solutions to improve liver-related outcomes related to the alcohol-harm paradox.

Influences of liver health on the AHP	Causes	Possible remedies
Alcohol consumption patterns	•Increased binging episodes have a more harmful impact on the liver than gradual alcohol intake •Lower income individuals more likely to choose spirits or beer over wine •Higher alcohol outlet density in urban areas contributes to higher alcohol intake	Strengthen alcohol policy •Introduce pricing policies that make cheap, high-alcohol-content beverages less accessible •Limit alcohol marketing, particularly in low-income areas and to vulnerable populations •Reduce the concentration of alcohol outlets in disadvantaged neighbourhoods
Education	•Insufficient public education and opportunities for individuals to seek help •Providers unaware of screening and treatment strategies for alcohol use disorder and ALD	Reduce stigma and enhance awareness •Educate about alcohol’s health risks, emphasizing liver disease, mental health, and chronic illnesses, e.g. US Surgeon General warning about alcohol and cancer risk •Train healthcare providers to address stigma, and deliver equitable care tailored to diverse socioeconomic backgrounds •Develop opportunities for education regarding treatment for alcohol use disorder and ALD
Healthcare access	•Lower income groups, immigrants, and individuals living in densely populated urban areas are less likely to have health insurance or easy access to preventive health •Limited access to healthcare leads to delayed diagnosis of ALD, leading to more severe illness and limited treatment options •Limited access and lower health literacy make treatment adherence more challenging, reducing the likelihood of transplant listing	Improve access to healthcare •Implement routine screening for alcohol use and liver disease, especially in underserved areas •Provide affordable access to addiction treatment programmes, counselling, and medications for AUD •Expand access to mental health care •Ensure timely referral to subspecialty care •Leverage use of virtual visits or consults
Psychosocial stress and stigmatization	•Chronic stress from unsafe living conditions, unstable housing, and/or financial insecurity may increase alcohol use as a coping mechanism •Progression of ALD from mild to severe illness and compounding symptoms and comorbidities may lead to further drinking	Reduce social determinants of poor health •Address financial insecurity through job training, housing assistance, and income support to reduce stress-related alcohol use •Improve housing access and living conditions •Reduce food insecurity and food deserts Community engagement •Involve disadvantaged communities in designing and implementing interventions to ensure they are relevant and culturally appropriate •Provide early interventions for children and teenagers in high-risk environments to prevent harmful drinking patterns from developing

Abbreviations

AHP, alcohol-harm paradox; ALD, alcohol-related liver disease; AUD, alcohol-use disorder; MASLD, metabolic dysfunction-associated steatotic liver disease; SDOH, social determinants of health; SLD, steatotic liver disease.

Financial support

The authors did not receive any financial support to produce this manuscript.

Authors' contributions

LW, AK and CM designed and coordinated the study; LW, JK, and AK wrote the manuscript; LW designed the figures; LW, JK, AK and CM revised and finalized the manuscript.

Figures were created in BioRender. Weichselbaum, L. (2025).

Conflict of interest

Laura Weichselbaum: no conflict of interest; Judah Kupferman: no conflict of interest; Allison Kwong: National Institute on Alcohol Abuse and Alcoholism (K23AA029197, Kwong); Christophe Moreno: Gilead, Roche, Ipsen, Echosens, Julius Clinical, Abbvie, Advarra.

Please refer to the accompanying ICMJE disclosure forms for further details.

Supplementary data

The following are the Supplementary data to this article: