Global Burden of Alcohol Use Disorders and Alcohol Liver Disease

1. Introduction

Alcohol use is a major risk factor for mortality and morbidity, playing a role in more than 200 diseases and injury outcomes [1,2]. However, for most regions, it can be shown that a few of the larger disease and injury categories capture more than 90% of alcohol-attributable mortality and burden of disease (the latter usually measured in disability-adjusted life-years DALYs [3]), namely alcohol use disorders (AUDs), cancer, cardiovascular disease, liver disease, and injury [4,5]. This contribution will focus on two of these major contributors to alcohol-attributable burden of disease: AUDs and alcoholic liver disease (ALD). Contrary to the other major contributors of alcohol-attributable disease, these two categories are fully alcohol attributable: there would be neither AUDs nor ALD without alcohol use [1]. Together, these two categories make up about a quarter of all alcohol-attributable mortality (26.3%) and one-third of alcohol-attributable burden of disease as measured in DALYs (31.5%; data based on [4]).

AUDs rank among the most prevalent mental disorders globally [6]. Individuals with an AUD have impaired control over their alcohol consumption and exhibit a chronic, heavy, and often escalating pattern of alcohol use despite significant detrimental consequences to their overall health, the lives of their family members and friends, and to society in general [7]; for definitions see DSM-5 [8], ICD-10 [9], or ICD-11 [10]. AUDs are very closely associated with a pattern of heavy alcohol use over time, and it has been suggested that this specific type of drinking behaviour should be part of the definition of AUDs [11,12,13].

ALD (for a definition and short overview see [14]) is the most prevalent type of chronic liver disease worldwide [15]. The International Classification of Diseases (ICD-10) recognizes several forms of ALD [14,16,17], which are usually considered as stages, that range from relatively mild and reversible alcoholic hepatic steatosis (fatty liver) and alcoholic hepatitis, to alcoholic fibrosis and sclerosis of the liver, and further to severe and irreversible stages, such as alcoholic liver cirrhosis and alcoholic hepatic failure (for a brief summary of the pathogenesis, see [18,19,20]). In some classifications, alcohol-attributable liver cancers have been listed as part of ALD as well [20,21], but we will not include this category in our calculations. Alcohol use and liver disease have an exponentially increasing dose-response relationship [22,23]. An additional impact may result from a daily heavy drinking pattern with some studies showing that the same overall amount of alcohol intake but with a non-daily drinking pattern is associated with less harm (“liver holidays” [24,25]).

AUDs and ALD are associated due to their shared etiology—alcohol use. A recent analysis of all French hospital patients between 2008 and 2012 showed that women with AUDs had 13 to 20-fold elevated risks for liver diseases (for liver cirrhosis: Hazard Ratio (HR): 19.3; 95% Confidence interval (CI): 19.1–19.5; for liver cancer: HR: 13.2; 95% CI: 12.8–13.7), and men with AUDs had 16-fold higher risks for liver cirrhosis (HR: 16.0; 95% CI: 15.9–16.1) and liver cancer (HR: 16.1; 95% CI: 15.9–16.4; [26]).

2. Materials and Methods

3. Results

4. Discussion

AUDs and ALD have both contributed markedly to alcohol-attributable burden of disease, with the higher contributor being ALD, especially with respect to mortality. To interpret these results, the following points and limitations should be considered:

• The definition of AUDs, and its operationalization, are both controversial, as each is based on relatively unspecific criteria (e.g., [43]), is culture-specific, and can therefore not be easily compared [44,45]. They do not necessarily coincide perfectly with heavy drinking levels and can therefore not be linked to harm [11,12] or with biological research [7,13]. While alternatives do exist to estimate AUDs [46], these are not commonly used and we therefore did not use them.

• The reported deaths resulting from AUDs only include deaths where either AUDs or alcohol poisoning was listed as the cause of death on the death certificate (see above). Similar to other mental disorders, addictive disorders such as AUDs have been associated with considerable excess mortality [47,48,49,50], even though most of these disorders do not appear in large numbers in cause-of-death statistics. The underlying cause of death in the WHO Global Health Estimates [42] is defined as the disease or injury that initiated the train of morbid events which led directly to the death, or the circumstances of the accident or violence that produced the injury. For AUDs, the cause of death would therefore be attributed to injuries such as suicides, or other chronic disorders such as ischemic heart disease or ALD [1,50,51]. However, to avoid double counting, these deaths are not counted as AUDs mortality.

• The last limitation of the estimation of AUDs prevalence and harm in this study is the reliance on the Global Health Estimates, which do not provide confidence intervals [35]. There is obviously uncertainty inherent in the estimates of AUDs and attributable harm which we could not capture.

• The burden for ALD is clearly larger than for AUDs in terms of DALYs and particularly for mortality (but see previous point). ALD is the single most important health harm caused by alcohol use [4,5]. The estimation methodology used is consistent with the definition of ALD (i.e., without alcohol use, these cases of ALD deaths or disease would not have existed). However, in contrast to other causes of death, ALD deaths were not estimated based on death certificates (in the countries where such certificates exist), or based on specific verbal autopsies [52,53], but were indirectly estimated via attributable fractions (see Materials and Methods above). The reason for this is the high stigmatization of all disorders which have the terms ‘alcohol’ or ‘alcoholic’ in their names (for general considerations, see [54]). Alcohol’s involvement in a death may be missed by those certifying or reporting the death (the latter in verbal autopsies), or may be deliberately not mentioned to protect the reputation of the deceased (see [55], for further conclusions). A landmark study of death recording practices in 12 cities in 10 countries found that the number of deaths assigned to the ICD category ‘liver cirrhosis with mention of alcoholism’ rose by 135% after taking into account additional information obtained from hospital records and interviews with attending physicians and family members. The majority of the new cases were previously coded in categories like cirrhosis without any mention of alcohol [56]. Underestimation of ALD and other 100% alcohol-attributable disease categories has persisted to this day (e.g., [57]; see short overview of a recent study in [58]).

• Finally, while there was uncertainty reported around the ALD estimates, it may be underestimated as it did not include any uncertainty for the cause of death or burden of disease (i.e., only for exposure and risk relations) [59].

Overall, AUDs and ALD have been shown to have a profound impact on the global burden of disease. Both disease categories are particularly tightly linked to heavy drinking occasions. In North America, both disease conditions have been contributing to decreasing life expectancies, especially in lower socioeconomic strata (US: [60,61,62]; Canada: [63,64,65,66]). For this region, disorders attributable to heavy alcohol consumption can, in effect, be used as an indicator for major increases in disease burden.

5. Conclusions

Policies to reduce the above-described burden of alcohol-attributable disease should be initiated [67]. While effective and cost-effective policies already do exist to reduce alcohol-attributable burden [67,68,69], two additional considerations should be focused on: first, each measure must be examined to ensure it does not widen the gap between socioeconomic strata and, second, the measures should target heavy drinking occasions. Thus, in addition to the raising taxes on alcohol, restricting its availability, and banning its advertisement, the current “best buys” for alcohol control—measures such as minimum pricing per unit—should be considered, as they have shown to have a strong effect on the drinking behaviour of heavy drinkers in lower socioeconomic strata [70].

Author Contributions

Both authors contributed equally to conceptualization, methodology, formal analysis, investigation, data curation, writing, and visualization.

Funding

This research was in part funded by the contributions of the World Health Organization to the WHO Collaborating Centre, and by from the Canadian Institutes of Health Research (CIHR) for the Ontario CRISM Node Team grant (#SMN-139150).

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.