The associated chronic organ damage exponentially increases in risk as alcohol consumption accumulates over time. Unmanaged heavy drinking is associated with subsequent heavy drinking, often culminating in brain damage, itself a consequence of heavy drinking but also a driver of future behaviour. Alcohol consumption itself is close to log-normally distributed in drinking populations, skewed towards heavy drinking. There is no natural cut-off point above which “alcohol use disorder” definitively exists and below which it does not. “Alcohol use disorder” is clinically defined as a score on a checklist of symptoms, and there is a smooth line exponential relationship between levels of alcohol consumption and the score on the checklist. Heavy drinking is a cause of the items on the checklist, including compulsion to drink more, which can also be a consequence of brain damage, itself caused by heavy drinking. Thus, “alcohol use disorder” is a diagnostic artefact. No more is needed to consider what is called “alcohol use disorder” other than heavy use over time. For alcohol , this approach does not imply that heavy use over time is the only cause of harm. There are other factors involved that that drive heavy alcohol use and harm that are independent of, or in interaction with, molecular and cellular levels , individual levels and environmental levels There is an ongoing discussion as to whether or not sugar is an ‘addictive’ substance that should be captured in the same category as drugs. Framing the problem as one of heavy use over time provides insight into this debate. As with alcohol and high blood pressure,roll bench chronic disease risk associated with plasma glucose levels has a continuous exponential relationship with sugar consumption.

The distribution of blood glucose levels is close to log-normally distributed in populations and skewed towards high consumption levels. There is no natural cut-off point above which diabetes linked to sugar definitively exists and below which it does not. Similar to the alcohol model where heavy use of alcohol over time leads to further heavy use of alcohol from the resulting brain damage, heavy use of sugar over time damages hippocampal function, which leads to further heavy use of sugar over time. Thus, in the ‘heavy use over time’ frame, sugar can be placed in the same category as alcohol and other drugs, and managed with similar governance approaches that promote public health.A core way to document the interference of drugs in human biology and functioning is to use quantitative risk assessment . QRA is a method applied in regulatory toxicology, for example, to evaluate water contaminants, and before safety approvals for food additives or pesticides. QRA has not been widely applied to drugs. Previous approaches for ranking harm have mostly been based on expert judgements which have been criticized as being arbitrary and biased. The advantage of QRA is that it provides a formal scientific method to rank the harm-potential of drugs, making optimum use of available data. There are several approaches for QRA available, with Margin of Exposure suggested by WHO as being most suitable for prioritizing risk management. In the alcohol field, MOE has been applied to evaluate the liver cirrhosis risk of ethanol, which is the single most important chronic disease condition attributable to alcohol globally.In a detailed study of the components in alcoholic beverages, ethanol was confirmed as the compound with highest risk. In a detailed comparison between ethanol and non-metabolically produced acetaldehyde contained in beverages, it was also judged that the risk of ethanol comprises more than 99% of the total risk.

It can be concluded that the risk of alcoholic beverages can be evaluated by looking at the effects of ethanol alone. The situation is less clear for tobacco, for which some industry MOE studies find toxicants other than nicotine. An MOE analysis of electronic cigarette liquids indicated that nicotine is the compound posing the highest risk. MOEs are calculated as the ratio of a toxic dose of the drug with the dose consumed either individually or on a population scale. The higher the MOE, the lower the level of risk, with low risk not implying safety. An MOE of 100 means that the drug is being consumed at one hundredth of the benchmark dose; an MOE of 1 means that the drug is being consumed at this toxic dose. The MOE for drugs can be calculated taking into account a range of hazard outcomes in health and other well-being domains, as long as suitable dose-response data are available . Therefore, analyses to date are primarily restricted to lethal outcomes based on animal studies. Results for European adults are summarized in Figure 1. The low MOE for alcohol is due to the high levels of consumption by European adults. The MOE results are consistent with the consensus of expert rankings in which cannabis is ranked with lower risk and alcohol with higher risk than current policies assume. The MOE is inherent to the drug itself; it does not account for the harms that arise from drug delivery systems, for example, smoked tobacco, or from secondary effects such as unclean syringes used for heroin intake. Of course, MOE, as presented here, focuses on the physical body of the adult user as the locus of harm. It does not take into account the sex and age of the user, or harm to individuals other than the user or at collective levels, which are a primary source of social differentiation between drugs. It also focuses on mortality, rather than intoxication in the moment. Differences between the intoxicating power of substances in the moment, and in the behavioural consequences of taking them, are primary reasons why, for example, societies have treated alcohol differently to tobacco.

Nevertheless, we believe that MOE should be applied at the current stage even when the underlying toxicological data are incomplete, to provide a better alignment of prioritization of policy to the drugs associated with higher risks, which in this case are nicotine, cocaine, heroin and alcohol.We have described three harmonizing approaches to reframe our understanding of addictions: biological predisposition to seek out psychoactive substances; heavy use over time as a fruitful characterisation; and quantitative risk assessment. Here, we propose two underlying pillars for a re-design of the governance of drug controls: embedding drugs governance within a comprehensive model of societal well-being; and creating a health footprint which, modelled on the carbon footprint,drying rack cannabis promotes accountability by identifying who causes what harm to whom from drugs.We propose that societal well-being should be our overarching frame for a more integrated governance and monitoring of drug control policies. Societal well-being, as captured by OECD, includes quality of life , material conditions and sustainability over time . Gross domestic product is included as a separate domain, recognizing that, while economic well-being is an important component of societal well-being, GDP has significant limitations. GDP excludes, for example, non-market household activity such as parenting, and activities such as conservation of natural resources. GDP also includes activities which do not contribute to well-being, such as pollution and crime, termed regrettables that are depicted within GDP but outside well-being. The use of and harm done by drugs are affected by and affect all well-being dimensions. Well-being analyses have found that, whilst some illegal drug policies may reduce health harms, they often come with adverse side effects including criminalization, social stigma and social exclusion, all of which exacerbate health harms. Humans are hard-wired to be social animals, with social networks strongly influencing tobacco use and alcohol intake. Punitive drug policies bring about the opposite: social exclusion due to stigma and social isolation. Engagement with illegal drugs conveys especially strong social meanings and can lead to stigma of marginalized heavy users, as opposed to the supposedly more responsible mainstream users. This can lead to punitive societal responses. Meanwhile, exclusion from the mainstream may allow harms to continue unchecked. If a user is caught using drugs in a country with “zero tolerance” to illegal drugs, the ensuing criminal sanctions will impede civic engagement and any improvements in quality of life and material living conditions.

For more detail, see ‘Well-being as a frame for understanding addictive substances’ by Stoll & Anderson. Changes in life expectancy in Mexico illustrate the negative consequences of criminalization. After six decades of gains in life expectancy in Mexico, the trend stagnated after 2000 for both men and women, and for men was reversed after 2005. This was largely due to an unprecedented rise in homicide rates, mostly as a result of drug policies promoting ‘gang wars’ and conflicts between gangs, the police and army. A well-being frame calls for whole-of-society approaches that progressively legalize illegal drugs to reduce violence and personal insecurity, while focusing on substances as drivers of harm. It balances the complex factors impacting drug use and related harm through the continuous monitoring of policy effects in a proactive way, with regulations embedded in international coordination. It calls for whole-of-society approaches that avoid criminalization where possible and where costs of addressing the problem are equally distributed across society. Governance strategies manage nicotine, illegal drugs and alcohol as a whole to avoid overlaps, contradictions, gaps and inequalities. The concern should be focused on harms, both to the user and to others, including family and friends, communities and society as a whole. The structures to support the strategies should be coordinated and multi-sectoral, involving high-level coordination of health, social welfare, and justice agencies in the context of international treaties, and, importantly, equitable across the lifespan, between genders and cultural groups. To increase the pace of policy change, regional and local public policies can create policy communities and networks within a common international framework. Managing ‘wicked problems’ requires clear rules of private sector engagement in policy making, particularly when private interests go against societal well-being. An evolved governance system must include measures to avoid industry co-optation, through transparency, checks and balances. Private sector stakeholders should operate within established rules.The ongoing monitoring of outcomes within a well-being framework would promote accountability. Modelled on the carbon footprint, we propose a health footprint as the accountability tool. Footprints were developed in the ecological field as a measure of human demand on ecosystems, including water footprints and carbon footprints that apportion greenhouse gas emissions to certain activities, products and populations. The central reason for estimating a carbon footprint is to help reduce the risk of climate change through enabling targeted and effective reductions in greenhouse gas emissions. The health footprint can be considered a measure of the total amount of risk factor attributable disability adjusted life years of a defined population, sector or action within a spatial or temporal boundary . It can be calculated using standard risk factor-related YLL and DALY methodologies of the Global Burden of Disease Study and of the World Health Organization. Health footprints are a starting point. To be accountable, we ultimately need to understand what drives the health footprint .Above the health footprint of Figure 3 are the structural drivers of harm that directly influence the size of the health footprint. Biological attributes and functions include, for example, the biological pre-disposition to seek out and use drugs. Genetic variants, for example, could be those that affect the function of alcohol dehydrogenase, influencing consumption levels and harm. Changes in global population size and structure can increase absolute numbers of drug-related DALYs, even though rates per person can decrease over the same time. As sociodemographic status improves in lower income countries, so do drug-related DALYs; yet, for the same amount of drug use, people with lower incomes suffer more drug-related DALYs than people with higher incomes.Above the structural drivers are the circumstantial drivers, those that can change. Related to drug potency and exposure, an MOE target for all drugs no greater than 10 has been argued6 . Policies could achieve such a result by either reducing drug exposure or by reducing the potency of the drug. Technological developments have led to electronic nicotine delivery systems as widespread alternatives to smoked tobacco, with current best estimates showing e-cigarettes to be considerably less harmful to health than smoked cigarettes. It may be that once e-cigarettes are heavily produced and marketed by the tobacco industry, that society will see cigarette-like levels of sustained heavy use of nicotine.