Unrecorded alcohol: a threat to public health?

Abstract

Headlines such as this [1] appear in newspapers with certain regularity, more often than not referring to incidences and outbreaks in low- and middle-income parts of the world. These incendiary statements also often refer to death or fatal consequences through the usage of so-called 'unrecorded alcohol'. Unrecorded alcohol is a summary term for a number of categories not registered in the country where it was consumed. Four main categories can be distinguished (see [2], their Figure 1): Illegally produced or smuggled alcohol products (including illegal homemade alcohol). Surrogate alcohol: non-beverage alcohol products not intended officially for human consumption, such as industrial alcohol, mouthwash or eau de cologne [3]. Alcohol products that are recorded, but not in the jurisdiction of consumption (e.g. cross-border shopping). Legal but unrecorded alcohol products (homemade or other). In 2000 30% of global alcohol consumption was estimated to be unrecorded [4,5], with a high proportion in low-income countries and in the former Soviet Union. These estimates, based usually upon expert opinion or surveys [6], carry substantial uncertainty [5–7] and have many open questions. Thus, the regional distribution of the four subcategories cannot be quantified. As most of the unrecorded alcohol consumption is in countries such as India, China, Brazil, Russia or in the African continent, category 3, including mainly cross-border shopping, is clearly not very relevant on a global level, but it may still constitute a sizeable portion in some parts of the world, such as the Nordic countries [8]. More important, however, are questions about whether there are substantial effects from unrecorded consumption on public health over and above the effect of ethanol alone. Table 1 shows an overview of potential components which could pose a public health threat, and the level of evidence. In considering this issue two main questions arise: (1) what is the evidence on causality for the links to health; and (2) what is the prevalence of additional harmful components in unrecorded alcohol around the world? To date, neither of the questions has been answered conclusively. While it is beyond the scope of an editorial to review systematically the level of evidence by component, we will give some illustrations of the underlying problems using the example of a recent publication [9]. Given the global level of unrecorded consumption it is obvious that the alcohol industry, with the exception of cross-border trading or some minor categories, has a keen interest in its reduction. Thus, it is no surprise that the International Center for Alcohol Policies (ICAP), a not-for-profit organization funded by leading producers of beverage alcohol, has published repeatedly on unrecorded consumption and potential health problems [10]. In their most recent publication, Noncommercial Alcohol in Three Regions[9], ICAP pleads vehemently that unrecorded alcohol poses public health consequences beyond legal 'beverage alcohol'. This is exemplified when they state that 'although the production of many noncommercial beverages meets high quality standards, much[emphasis added] of what is included under this heading may be contaminated or toxic' ([9], p. 1). There is currently no scientific evidence for the conclusion that much of non-commercial alcohol is contaminated or toxic. The limited available literature is inconsistent and does not allow quantitative conclusions. In toxicology, dose–response relationships have to be taken into consideration: it is incorrect to assume that minor contaminations are associated with a major 'threat to public health' without providing a dose–response analysis, and such analyses are sorely lacking. Let us exemplify the problem with one component from Table 1: higher alcohols. In the publication cited above, Razvodovsky ([11], p. 20) claims that 'homemade beverages often contain aliphatic alcohols—toxic to liver cells'. This statement is in stark contradiction to the conclusion in a review [3] that the role of higher alcohols (e.g. propanol, isobutanol and isoamyl alcohol) in the aetiology of diseases associated with unrecorded alcohol is currently unclear. We also have to rebut Razvodovsky's claim that 1-propanol and isoamyl alcohol levels found in eastern European samples approached potentially hepatotoxic levels. The cited study on isolated perfused rat liver does not allow a risk assessment for humans [12]. We have reviewed recently the toxicology of higher alcohols and determined that alcoholic beverages seldom exceed acceptable daily intakes for these substances, calculated with a safety factor of 100 from animal experiments [13]. In addition, no substantial differences in the contents of higher alcohols between commercial alcoholic beverages and unrecorded alcohol were found in most of the available literature [3,13]. In sum, thus far there is no evidence for the impact of higher alcohols in unrecorded alcohol on public health. Examining the potential harmful components (Table 1) in detail, the evidence so far has supported only a potential impact of higher concentration of alcohol itself. All other components have not been found in the vast majority of unrecorded alcohol at levels known to cause health harm. This is not to belittle the tragedy of methanol outbreaks if they occur, but overall they occur relatively far too seldom to constitute a major public health threat, globally or in any region [3]. However, while the claims of the alcohol industry have been premature and remain unsubstantiated, given the sheer extent of unrecorded consumption and the global burden of alcohol consumption per se, there remains a dire need to examine the effects in greater detail. Currently, our knowledge is limited and systematic large-scale studies are lacking. For instance, a large-scale autopsy study [14] found that alcohol poisoning had been underlying recent mortality trends in Russia. However, little is known about the alcohol consumed which led to these poisonings, the role of ethanol concentration or the role of unrecorded consumption. We need to know these facts if we want to look seriously into interventions for reducing alcohol-attributable mortality in Russia. Going beyond this example, we propose to study systematically the impact of unrecorded consumption by conducting case–control studies with cases from alcohol poisoning entries to the emergency room, people treated for liver disease and alcohol dependence. These studies should include sampling and chemical analysis of alcohol consumed usually by these groups and matched controls. Given current and limited knowledge, the most important public health threat of unrecorded alcohol may, in fact, stem from an associated heavy drinking consumption pattern (i.e. 'binge drinking') combined with high alcoholic strength in beverages, thus leading to more pronounced effects of alcohol intoxication and poisoning [15]. As indicated above, the alcohol beverage industry has a marked commercial interest in reducing unrecorded consumption, independent of any potential health threat. Being sponsored by the alcohol beverage industry, it is understandable that ICAP has an interest in examining unrecorded consumption. However, overestimations of the magnitude of such consumption without sufficient empirical basis [16], or the publishing of reviews overstating the problem without empirical evidence or with misleading references, are problematic and disconcerting. On the other hand, there has not been enough interest from science for unrecorded consumption. It almost seems as if scientists are shying away from a systematic examination of unrecorded consumption due to fear of possibly 'supporting' the alcohol industry. In our view, this behaviour is equally problematic and disconcerting. We need to establish more accurate evidence on unrecorded consumption to inform policies more effectively, regardless of whether any single measure may or may not actually be congruent with some interests of the alcohol industry or any other. None.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 45 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.