Percentages of problem drug use and their implications for policy making: A review of the literature

Abstract

This paper reviews the literature on problem drug use, aiming to answer the question ‘what percentages of drug users are problem drug users?’ We look at the conceptualisation of problem drug use, and related concepts such as addiction and dependence. While conceptually distinct, definitions vary, and terms are often utilised interchangeably in the literature. Because of the heightened controversy and recurring moral panics (often without factual basis) related to illicit drugs as compared to the licit drugs, we focus primarily on the use of the former. We summarise previous research findings on percentages of problem drug user in general as well as for the most commonly used illicit substances more specifically. Results show that the majority of drug use is episodic, transient and generally non-problematic. The majority of people who have used various drugs in their lifetime have not done so in the past year. Only a minority become problem drug users. Appreciating the limitations of current research, further research avenues are outlined and implications for policy making are provided. If most drug use is indeed non-problematic it implies the need for policy changes with a heightened focus on the development of improved risk communications and other harm reduction strategies. There is a need to move away from the narrative of ‘the problem drug user’ which is in many ways still prevalent in informing policy making. Incorrect perceptions of drugs and people who use drugs should be countered by providing evidence-based, transparent and easily accessible information. This is essential to change public attitudes towards and remove stigma of people who use drugs.

Keywords

problematic drug use,dependence,addiction,drug policy

Introduction

This paper provides an overview of percentages of problem drug use (PDU), based on findings of the peer-reviewed literature and ‘grey’ literature. Investigating problem drug use on such a broad scale is challenging. There are no universal definitions of PDU (if this is indeed the term used) and as such, there are no straightforward and globally agreed measurements or assessments which can be easily compared across different drugs. Hence concluding on a definite percentage of problem drug users globally is beyond the scope of this paper. Instead we look at what constitutes PDU before comparing current and past estimations of percentages of PDU in some depth.

A large proportion of people in society consume licit and/or illicit drugs. Only a minority of these use drugs in a problematic manner. Nevertheless, the effects of drugs are generally portrayed as negative in the media (Global Commission on Drug Policy, 2017), leading to the stigmatisation of people who use drugs as media representations can have a strong influence on public perceptions (Global Commission on Drug Policy, 2017). The relationship between stigma, discrimination and criminalisation is complex, as social and political spheres interplay with criminal law and the judiciary (UNAIDS, 2015). The negative representations and stigmatising terminology can directly impact clinical care: the more society stigmatises people who use drugs, the fewer harm reduction and treatment opportunities will be provided (Global Commission on Drug Policy, 2017).

What is problematic drug use?

There are a wide variety of approaches to addiction and/or problem drug use, from genetic approaches arguing that genes influence the risk of addiction (e.g. Kendler et al., 2003; Tsuang et al., 1996), to approaches of addiction as a chronic disease linked to pathological changes in neural circuitry (e.g. Hyman et al., 2006). Further, individuals’ liabilities to addiction have been researched in terms in terms of the factors that contribute to an individual’s risk of addiction (e.g. Tarter et al., 2012) as well as factors that contribute to transition to levels of addiction (Hasin and Grant, 2004). In parallel, there is also literature focusing on substances’ addictive liabilities (e.g. Ridenour et al., 2005).

PDU is a more recent classification than dependence and addiction. While early definitions of addiction were reliant on the role played by tolerance and withdrawal, recently more emphasis is placed on the psychological components of dependence. Today, the compulsive use of drugs is seen as central feature of dependence (Gossop, 1989; West and Gossop, 1994). Others argue that dependence is accepted as based on physiological drug responses, while addiction is often defined as compulsive drug seeking and use (e.g. Vanyukov et al., 2012).

The term PDU aims in many ways to replace these traditional concepts and to provide a more neutral perspective, that is less challenging than addiction, and less moralising than abuse and misuse (Seddon, 2011). It has become popular in the UK and Europe during the past 25 years as it appears to be a more pragmatic approach, focusing on managing drug related problems and leaving non-problematic use alone (Seddon, 2011).

There is still controversy surrounding the differentiation of concepts. Similar to the division between licit and illicit substance use, the terminology between addiction and dependence is frequently value-laden and fraught with stigmatisation (Vanyukov et al., 2012). Classifications are often subject to moral judgements regardless of the actual term used. While there is a conceptual distinction between dependence, addiction and PDU, classifications are often used interchangeably in the literature. Different definitions are still employed by different stakeholders and organisations highlighting the complexities of the issues as well as the need for clearer communications. For the purpose of clarity, this report will use the terms problem drug use (PDU) and problem drug users (PDUs).

While the terms high-risk drug use and PDU may be less moralistic than drug abuse, levels of stigma and prejudice against drug users have not diminished in recent years (Global Commission on Drug Policy, 2017). It is still a way of labelling ‘the other’, implying another, newer way of categorising fellow humans. Gootenberg (2009) highlights the importance to focus on discourse and language as essential for any critical analysis of drug control. These classifications shape the service and treatment provisions and the ways in which drug users are perceived and as such can have significant impacts on how people are dealt with by institutions and the public alike.

Although there seems to be agreement that opioids and cocaine are the drugs most frequently implicated in PDU, there is less agreement on whether (and if so, how) to include other substances. Definitions differ around key issues, such as inclusion/exclusion of particular drugs, different classifications of types of drug use and different views about cut-off points for problematic frequency. Many studies do not explicitly define problem drug use or what is problematic about drug use.

Measuring PDU

Drugs can cause various harms to the individual, society and environment but the relative harms differ. There is still a common perception that drugs are illegal based on a rational harm analysis, but this is not always the case. Drug harms are not based on science alone, but rather tend to incorporate a value judgement by society (Global Commission on Drug Policy, 2017). Many drug harms are also the result of illegality rather than being intrinsic to the drug. Potential harms can be exacerbated by repressive drug control policies, whereby drugs are produced, obtained and consumed illegally (Global Commission on Drug Policy, 2017).

While there are certain risks involved in all drug use, the legal status of a drug does not systematically correspond to the potential harms of that drug (Nutt et al., 2010). Understandings as to what is problematic in terms of drug use complex. A proportion of these problems are not attributable directly to the illicit drug per se but to the societal/political responses to drug use/users. Too narrow a focus on significant harms carries some drawbacks, as these are not the only dimensions which should be taken into account when addressing PDU.

Research on problem drug use ranges through a variety of scientific disciplines, from neuroscience and genetics to prevention, treatment and social services leading Conway et al. (2010) to characterise it as a ‘biobehavioural discipline’ (Conway et al., 2010: 2). Various attempts have been made to measure PDU. Consequently, in addition to the range of definitions and harms, there are a variety of assessments and measurements.

The range of methods for measuring PDU broadly fall under two categories: population size estimation and severity of addiction estimates. The former attempt to measure the number of PDUs, i.e. estimating the prevalence of the extent of PDU across whole populations (e.g. Gemmel et al., 2004; Johnston et al., 2008; SAMHSA, 2008; Vaissade and Legleye, 2009) and includes amongst other approaches, capture–recapture methods (e.g. Bouchard and Trembly, 2005; Mascoli and Rossi, 2008), and the triangulation of different measurement indicators (e.g. EMCDDA, 2018).

Capture-recapture methods involve ‘capturing’ a sample of a population, ‘marking’ them and then releasing them. Then a second sample will be captured, the proportion of marked individuals in this second sample is assumed to be equivalent to the proportion of individuals in the population that were first captured, hence the population size can be estimated. While not completely exact this method is useful to establish the number of ‘hidden’ drug users (Bouchard and Trembly, 2005; Mascoli and Rossi, 2008).

While capture-recapture studies are helpful to give a broad picture of PDU, their key limitation is often related to sampling in that drug users have to come forward for treatment in the first place. This likely leads to an underestimation of the target population (Payne, 2007).

The problem drug use indicator triangulates methods to investigate levels of problem drug use in Europe. Methods include for instance the frequency of drug use, the route of administration, and the number of users entering medical treatment (EMT), including both first time entrants and previously treated entrants, offering a breakdown by main drug as well as of demographics of the population (EMCDDA, 2018).

In addition to population size estimations, severity estimates investigate the degree of dependence as in the Severity of Dependence Scale (Gossop et al., 1995), the Severity of Alcohol Dependence Questionnaire (SADQ) or the Severity of Opioid Dependence Questionnaire (SODQ) (Sutherland et al., 1986, 1988).

In sum, a variety of methods can estimate both the prevalence of PDU and its severity. Although available estimates are improving in number and quality, many problems remain to be solved. The broad range of measurements for PDU can make establishing an exact percentage of PDUs challenging. Globally, there is no single method that can be applied in all countries to provide truly comparable results. Even if a standard method such as capture-recapture can be used at the local level, available datasets often differ between countries, making comparisons across countries challenging. In addition, the lack of exact definition of the target group is problematic, and drug use patterns vary considerably between countries. Indeed, Kilmer et al. (2015) found that cross-national comparisons of PDU are undermined by fundamental- and often unacknowledged- underlying methodological differences, highlighting that improved measurements of PDU across countries depend on the harmonization of measurement systems.

Percentages of PDU: Transience of drug use

In 2016, an estimated quarter of a billion people (age: 15–64 years) – around 5% of the global adult population – used currently illegal drugs (Global Commission on Drug Policy, 2017). Of these, about 11.6% are considered to suffer problematic drug use and/or addiction. In order of popularity the most commonly used illicit drugs tend to be cannabis, cocaine, MDMA, amphetamines and ketamine (e.g. European Monitoring Centre for Drugs and Drug Addiction, 2011; United Nations Office on Drugs and Crime, 2012).

Figure 1 shows the prevalence of drug users vs the prevalence of people with drug use disorders over the past decade, showing that the % of the latter is comparatively small at 0.6% and fairly stable, indicating that the vast majority of drug users likely do not have drug use disorders.

Figure 1.

Comparison of lifetime and past year drug usage (UNDOC, 2017).

Figure 2.

Percentages of PDU per drug.

The most common pattern of drug use is episodic and non-problematic (Global Commission on Drug Policy, 2017). The majority of people who have ever tried drugs have not used them in the past year. Figure 2 shows the huge differences in the number for global lifetime drug use vs. use in the past year for three of the most frequently consumed illicit drugs.

Comparison between lifetime, past year and past month drug use (%)

Taking a US perspective, SAMHSA’s (2017)¹ findings summarised in Table 1 clearly highlight the differences between lifetime usage, past year usage and last month usage, with a similar pattern across different drugs analysed. Whilst many people do use drugs in their lifetimes, many less use them during the past year, and less even the past month, indicating that there often is no continuation of initial and/or exploratory use to regular (and potentially problematic) drug use. Unsurprisingly, licit drugs such as alcohol and tobacco score far higher on all comparisons (lifetime, past year and past month usage) than any illicit drugs.

Table 1.

Lifetime, past year and past month drug use (%).

Drug	Lifetime	Past year	Past month
Cannabis	45.2	15.0	9.6
Cocaine	14.9	2.2	0.8
Crack	3.5	0.3	0.2
Heroin	1.9	0.3	0.2
Hallucinogens	15.5	1.9	0.5
LSD	9.6	0.8	0.2
PCP	2.2	0.0	0.0
Ecstasy	7.0	0.9	0.2
Methamphetamine	5.4	0.6	0.3

Source: SAMHSA (2017).

Highlighting that there is often a ‘natural’ course of (generally non-problematic) drug use Von Sydow et al. (2001) looked at the natural course of cannabis use and disorders amongst adolescents and young adults, estimating that the probability of developing cannabis abuse or dependence is relatively low at 8.8%. The majority of older participants had reduced their use at the follow up. Overall, the natural course of cannabis use is quite variable: while some users stopped completely in the 20s, others continued occasional or frequent use, but only few developed a cannabis disorder (Von Sydow et al., 2001).

Hence rather than being problematic, drug use is often transient, with many people ‘growing out of it’ and stopping use in their late 30s (Winstock and Shiner, 2015). In their longitudinal study, Von Sydow et al. (2002) examined the incidence and patterns of the natural course of ecstasy, stimulant and hallucinogen use in adolescents and young adults. For these drugs, 1% of users met the criteria for or abuse, and 0.6% for dependence (Von Sydow et al., 2002). The majority of lifetime users without dependence had stopped using these drugs during the following year. As such, the probability of developing disorders is fairly low at 1.6%. The majority (80%) stopped their use in their 20s but half of those who once had fulfilled criteria for dependence continued their use (Von Sydow et al., 2002). Thus while the majority of users of ecstasy and related drugs do not develop a DSM-IV disorder, there is a (small) high-risk group with long-term elevated use which raises clinical and public health concerns.

Percentages of PDU: PDU by drug

In 1994, Anthony et al. published their classic study on problem drug use, using data from the US National Comorbidity Survey. Findings present nationally representative survey-based population estimates for the lifetime prevalence of extra medical drug use and lifetime prevalence of drug dependence. The results are still cited frequently today. Their sample allowed comparisons of tobacco dependence, alcohol dependence and other psychoactive drugs, presenting (a) lifetime dependence, (b) lifetime prevalence drug use and c) the proportion of users who become dependent.

In 2016, the Substance Abuse and Mental Health Services Administration (SAMHSA) published their more recent findings which (despite some differences in measurement approaches) in many ways mirror Anthony et al.’s results over 20 years earlier. Both the order rankings of percentages of users who become addicted as well as the actual percentages are comparable, with heroin and cocaine having the highest percentages of users whose use becomes problematic (after tobacco). Figure 3 provides a summary of findings of both Anthony et al.’s (1994) and SAMHSA’s (2017) work, showing the analysed drugs from the lowest to the highest percentage of PDU.

Further details are provided in Table 2. After alcohol and tobacco, cannabis was the most frequently used drug, but it ranked much lower in terms of dependence. In Anthony et al.’s (1994) study, an estimated 46.3% had used cannabis but only 9.1% of users had developed cannabis dependence, i.e. for every user with a history of cannabis dependence, there were 10 users who had not become dependent. By comparison, an estimated 16.2% had tried cocaine at least once, and 16.7% of these users qualified as cocaine dependent (Anthony et al., 1994). For other stimulants (such as amphetamines) the prevalence rate was 15.3%, and 11.2% of these users had developed dependence (Anthony et al., 1994). Heroin was the least used drug at 1.5% but 23.1% of users had become dependent (Anthony et al., 1994). For psychedelics, the prevalence rate was 10.6%, and a relatively low 4.9% of users qualified for dependence diagnosis (Anthony et al., 1994).

Table 2.

Percentages of users with PDU (%).

	SAMHSA (2017)	Anthony et al. (1994)
Tobacco	33	32
Alcohol	10–15	15
Cannabis	<10	9
Cocaine	15–20	17
Heroin	23–30	23
Methamphetamine/stimulants	5–10 (Methamphetamine)	11 (Stimulants)
Psychedelics	9	5

Source: SAMHSA (2017). Results from the 2016 National Survey on Drug Use and Health; Anthony et al. (1994).

In terms of percentages of drug users who become addicted, similar findings (albeit in relation to a smaller selection of particular drugs) have also been reported by e.g. Von Sydow et al. (2001), SAMHSA (2003), Chen et al. (2005), RAND (2005), Wagner and Anthony (2002) and Lopez-Quintero et al. (2011), lending some validity to the above estimates despite measurement challenges.

These results indicate that a large majority of persons who have used drugs do not proceed to develop drug dependence and/or become PDUs. Even for drugs known for their dependence liability (such as heroin, cocaine and tobacco) the proportion of drug users who became dependent was in a range of 20–40% (Anthony et al., 1994; Hart, 2013). Highlighting the importance of the social context and environment in relation to the development of problem drug use, Helzer (1985) and Robins (1993) found that most Vietnam veterans who used heroin and other opioids in Vietnam did not become dependent taking these drugs while overseas and did not continue to do so upon their return when removed from that stressful setting. Some implications of these findings for drug policy are highlighted in the discussion.

In order be able to develop better drug policies and treatment options, it is important to examine these findings in detail, looking at variations within samples. There is considerable variation in the prevalence of drug use and in the transition from drug use to drug dependence, across individual drugs and drug groups (Anthony et al., 1994). For example, the ranking of drugs in relation to transitions from drug use to drug dependence shows variations between men and women and across age groups, depending on various drugs as well. But for both men and women and all age groups (apart from the oldest cohort), tobacco and heroin were top ranked, with psychedelic drugs and inhalants at the bottom in terms of their dependence risk (Anthony et al., 1994).

In relation to the distribution of PDU, Anthony et al. (1994) already noted that drug dependence is not distributed randomly, and that some population groups are more affected than others. For example, in relation to gender, males were more likely to become dependent on alcohol, controlled substances or inhalants but not in relation to tobacco dependence (Anthony et al., 1994). In relation to age, dependence on controlled substances was most likely to be found among young adults (age range 15–24 years) and least likely to be found among older cohorts (45-54 years old). In relation to socio-economic status (SES) compared with employed workers, unemployed cohorts were more likely to have dependence on tobacco, alcohol as well as other controlled substances/drugs. Lower annual income was associated with having been more frequently affected by drug dependence (Anthony et al., 1994). Similar risk factors contributing to the development of PDU have also been identified in the more recent literature (e.g. Breslau et al., 2001; Chen et al., 2005; O’Brian and Anthony, 2005; Wagner and Anthony, 2002).

As such, gender, age and SES may be important contributing factors when assessing PDU, indicating that they need to be taken into account when developing interventions and harm reduction approaches. In order to develop better treatment options, guide clinicians, and allocate resources effectively it is important to identify the risk of transition from drug use to dependence in further detail.

Limitations and avenues for further research

Certain limitations are shared between many of the studies cited in this review. Studies are often limited in that they have a self-selected and non-representative sample. Moreover, drug use tends to be self-reported, requiring users to reconstruct their histories of drug use. As such, findings may include some recall bias and possible over- or under-stating of drug use by cohorts. In particular, measuring the prevalence of the most problematic patterns of drug use, such as in the homeless and incarcerated populations, is challenging as these populations are unlikely to show up in household surveys or present for treatment. Additionally, because substance use may not be continuous, but rather intermittent, this may produce inexact estimates of the time between use and dependence development. Nevertheless, the fact that findings from various assessments are often in accordance with each other, lends some validity to the data.

Future research would benefit from less reliance of retrospective self-reporting and the use of more triangulation of methods to validate findings. To date, most studies are only able to address one particular dimension of PDU. Moreover, there are research gaps in terms of longitudinal studies, and cross-country comparisons. In light of the ongoing developments in drugs and drug markets over the past decades, the field might for example, benefit from a systematic replication of Anthony et al.’s (1994) classic study with newer comorbidity data. Despite these limitations, we can draw some general conclusions and highlight implications for policy making.

Conclusions

Drug use seemingly encompasses a wider range of substances all the time (EMCDDA, 2018). Individual patterns of use range from experimental to habitual to dependent consumption, on a continuum, from non-problematic to problematic use. The trajectory of drug use varies between different people and between different drugs. Current findings show that the majority of drug users are not doing so in a problematic manner. Only a relatively small percentage of individuals who use drugs develop dependence. In these cases, their drug use is problematic and likely to place a tremendous burden on the individual and on society (WHO, 2002).

Yet most people who use drugs do so only experimentally or moderately. Generally, people take drugs socially rather than living the life of a heroin addict or ‘spice zombie’ as presented in the media. This review highlights that a large majority of persons who have used drugs do not proceed to develop PDU, regardless of the particular substance used. Rather, most people’s drug use is episodic and transient.

Measuring PDU continues to be challenging because of the illicit nature of most drug use and the often hard to reach populations, as well as the absence of a concrete and widely accepted definition (Cave et al., 2009). In summarising the literature, we offer a ‘snapshot’ of the current state of the art, appreciating that there are many other dimensions related to PDU which are beyond the scope of this brief review.

Implications for policy making: Changing perceptions of ‘the drug user’

We have shown that the majority of drug users do so in a non-problematic manner. As such, it is essential to move away from the sole focus on problem drug use and users, and to provide a different narrative of ‘the drug user’. As long as drug use is being regarded as ‘morally wrong’ and always problematic it will be challenging to implement effective public policies which take both the individual and society into account (Global Commission on Drug Policy, 2017).

The understanding that most drug use is indeed non-problematic needs to go hand in hand with a change in perception of drug use and people who use drugs. This likely is a time-consuming process. Rather than a focus on the science per se, value and moral judgements are involved. In turn, this highlights the importance of improving communications about drugs, between scientists and policy makers and the public. However, when societal value judgements are involved, providing scientific information by itself might not be sufficient.

Drug reforms can change public opinion and contribute to reversing the cycle of discrimination, stigma and repression. Evidence indicates that the public might support more pragmatic and evidence-based drug polices when they are given credible and evidence-based information as in the case of Switzerland (Savary et al., 2009). Some countries (e.g. Portugal, Uruguay, the Netherlands, Canada and Mexico) have already changed their drug policies towards a more liberal stance and a change in public perception seems to be underway.² Governments need to correct false perceptions of drugs and of people who use drugs by providing evidence-based, easily accessible information. Countering incorrect perceptions is necessary to change attitudes and remove stigma.

Implications for policy making: Improving risk communications and focusing on harm reduction

In order to change public perceptions, it is essential to improve risk communications. The Global Commission on Drug Policy (2017) highlights that past risk communications about drugs have often failed because information provided was often incomplete and even incorrect. Such messages can be counter-productive. Indeed, children exposed to these messages might even be more likely to use drugs as they cannot trust the message and the sender. Consequently, this contributes to a decrease of trust between policy makes and the public. In line with the classic risk communication literature (e.g. Slovic, 1987; Renn and Levine, 1991), information provided has to be evidence-based and uncertainties have to be admitted. Risk communications need to be easily accessible and transparent and include what people want to and need to know in order to make informed decisions about drug use.

The fact that most drug use is non-problematic implies the need for policy changes involving a heightened focus on the development of harm reduction strategies. Even in the absence of notable PDU, users might nevertheless benefit from harm reduction approaches. Various harm reduction strategies can help to decrease and/or prevent drug harms and make drug use less problematic, ranging from drug testing of stimulants and hallucinogens³ to needle exchanges and safe injection facilities for opioid users. Harm reduction acknowledges that people may not be willing or able to abstain from illegal drug use and encompasses messages and communication as well as strategies that seek to reduce harm without requiring the person to cease drug use completely (Cave et al., 2009). Today, there is an increasing consensus on the need to focus resources on identifying and reducing harms rather than focusing simply on reducing drug supply and consumption (Cave et al., 2009).

In conclusion, following on from the understanding that the majority of drug use is non-problematic, the language about drug use and drug users needs to be changed, away from the prevalent narrative of PDU, in order to change public perceptions, Here, the media can play a positive role (The Global Commission on Drug Policy, 2017). Media and policy makers must be responsible in shaping public opinions, working together to promote the use of non-stigmatising language in order to improve the negative (and often incorrect) perceptions of people who use drugs.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Anthony

Warner

Kessler

(1994) Comparative epidemiology of dependence on tobacco, alcohol, controlled substances, and inhalants: Basic findings from the national comorbidity survey. Experimental and Clinical Psychopharmacology 2: 244–268.

Bouchard

Trembly

(2005) Risks of arrest across drug markets: A capture-recapture analysis of ‘hidden’ dealer and user populations. Journal of Drug Issues 35(4): 733–754.

Breslau

Johnson

Hiripi

, et al. (2001) Nicotine dependence in the United States: Prevalence, trends and smoking persistence. Archives of General Psychiatry 58: 810–816.

Cave

Hunt

Ismail

, et al. (2009) Tackling problem drug use. TR-795-NAO. Report prepared for the UK National Audit Office. Available at: www.nao.org.uk/wp-content/uploads/2010/03/0910297_reviews.pdf (accessed 23 January 2020).

Chen

O’Brian

Anthony

(2005) Who becomes cannabis dependent soon after onset of use? Epidemiological evidence from the United States: 2000–2001. Drug and Alcohol Dependence 79(1): 11–22.

Conway

Levy

Vanyukov

, et al. (2010) Measuring addiction propensity and severity: The need for a new instrument. Drug and Alcohol Dependence 111: 4–12.

European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) (2011) Annual Report 2011, The State of the Drugs Problem in Europe. Luxembourg: Publications Office of the European Union.

European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) (2018) European drug report: Trends and developments. Available at: www.emcdda.europa.eu/publications/edr/trends-developments/2018 (accessed 23 January 2020).

Gemmel

Millar

Hay

(2004) Capture-recapture estimates of problem drug use and the use of simulation based confidence intervals in a stratified analysis. Journal of Epidemiology and Community Health 58(9): 758–765.

10.

Global Commission on Drug Policy (2017) The world drug perception problem: Countering prejudices about people who use drugs. Report 2017. Available at: www.globalcommissionondrugs.org/reports/ (accessed 23 January 2020).

11.

Gootenberg

(2009) Talking about the flow: Drugs, border and the discourse of drug control. Cultural Critique 71: 13–46.

12.

Gossop

Darek

Griffiths

, et al. (1995) The severity of dependence scale (SDS): Psychometric properties of the SDS in English and Australian samples of heroin, cocaine and amphetamine users. Addiction 90: 607–614.

13.

Hart

(2013) High price: A neuroscientist’s journey of self-discovery that challenges everything you know about drugs and society. New York: Harper Collins Publishers.

14.

Hasin

Grant

(2004) The co-occurrence of DSM-IV alcohol abuse and DSM-IV alcohol dependence: Results of the national epidemiological survey on alcohol and related conditions on heterogeneity that differs between population subgroup. Archives of General Psychiatry 61(9): 891–196.

15.

Helzer

(1985) Specification of predictors of narcotic use versus addiction. In: Robins

(ed) Studying drug abuse. New Brunswick: Rutgers University Press, pp. 173–205.

16.

Hyman

Malenka

Nestler

(2006) Neural mechanisms of addiction: The role of reward-related learning and memory. Annual Review of Neuroscience 29: 565–598.

17.

Johnston

O’Malley

Bachman

, et al. (2008) Monitoring the future national survey results on drug use, 1975–2007. Bethesda: National Institute on Drug Abuse, p. 707.

18.

Kendler

Jacobson

Prescott

, et al. (2003) Specificity of genetic and environmental risk factors for use and abuse/dependence of cannabis, cocaine, hallucinogens, sedatives, stimulants, and opiates in male twins. American Journal of Psychiatry 160: 687–695.

19.

Kilmer

Reuter

Giommoni

(2015) What can be learned from cross-national comparisons of data on illegal drugs? Crime and Justice 44(1): 227–296.

20.

Lopez-Quintero

de los Cobos

Hasin

, et al. (2011) Probability and predictors of transition from first use to dependence on nicotine, alcohol, cannabis, and cocaine: Results of the National Epidemiological Survey on Alcohol and Related Conditions (NESARC). Drug and Alcohol Dependence 115: 120–130.

21.

Mascoli

Rossi

(2008) Capture-recapture methods to estimate prevalence indicators for the evaluation of drug policies. Bulletin on Narcotics LX(1/2): 5–25.

22.

Nutt

King

Phillips

(2010) Drug harms in the UK: A multicriteria decision analysis. The Lancet 376: 1558–1565.

23.

O’Brian

Anthony

(2005) Risk of becoming cocaine-dependent: Epidemiological estimates for the United States, 2000–2001. Neuropsychopharmacology 30: 1006–1018.

24.

Payne

(2007) Women drug users in North Cumbria: What influences initiation in to heroin in this non-urban setting? Sociology of Health and Illness 29(5): 633–655.

25.

RAND (2005) Assessing U.S. drug problems and policy: A synthesis of the evidence to date. Available at: www.rand.org/pubs/research_briefs/RB9110/index1.html (accessed 23 January 2020).

26.

Renn

Levine

(1991) Credibility and trust in risk communication. In: Kasperson

Stallen

(eds) Communicating risks to the public. Amsterdam: Kluwer Academic Publishers.

27.

Ridenour

Maldonado-Molina

Compton

, et al. (2005) Factors associated with the transition from abuse to dependence among substance abusers: Implications for a measure of addictive liability. Drug and Alcohol Dependence 80: 1–14.

28.

Robins

(1993) Vietnam veterans’ rapid recovery from heroin addiction: A fluke or normal expectation? Addiction 88: 1041–1054.

29.

Savary

Hallam

Bewley-Taylor

(2009) The Swiss four pillars policy: An evolution from local experimentation to federal law. Oxford: The Beckley Foundation.

30.

Seddon

(2011) What is a problem drug user? Addiction Research and Theory 19(4): 334–343.

31.

Slovic

(1987) Perception of risk. Science 236: 280–285.

32.

Substance Abuse and Mental Health Services Administration (SAMHSA) (2003) Results from the 2002 national studies on drug use and health: National findings. Rockville: Office of Applied Studies.

33.

Substance Abuse and Mental Health Services Administration (SAMHSA) (2008) Results from the 2007 national studies on drug use and health: National findings. Rockville: Office of Applied Studies.

34.

Substance Abuse and Mental Health Services Administration (SAMHSA) (2017) National survey on drug use and health 2016. Bethesda: Substance Abuse and Mental Health Services Administration.

35.

Sutherland

Edwards

Taylor

, et al. (1986) The measurement of opiate dependence. British Journal of Addiction 81: 485–494.

36.

Sutherland

Edwards

Taylor

, et al. (1988) The opiate dependence syndrome: Replication study using the SODQ in a New York clinic. British Journal of Addiction 83: 755–760.

37.

Tarter

Kirisci

Mezzich

, et al. (2012) Does the ‘gateway’ sequence increase prediction of cannabis use disorder beyond deviant socialization? Implications for prevention and policy. Drug and Alcohol Dependence 123: S72–S79.

38.

Tsuang

Lyons

Eisen

, et al. (1996) Genetic influences on DSM-III-R drug abuse and dependence: A study of 3372 twin pairs. American Journal of Medical Genetics 67: 473–477.

39.

UNAIDS (2015) Terminology guidelines . Geneva. Available at: www.unaids.org/sites/default/files/media_asset/JC2702_GARPR2015guidelines_en.pdf (accessed 23 January 2020).

40.

United Nations Office on Drugs and Crime (UNDOC) (2012) World Drug Report 2012. Vienna: United Nations Publications.

41.

United Nations Office on Drugs and Crime (UNDOC) (2017) World Drug Report 2017. Vienna: United Nations Publications.

42.

Vaissade

Legleye

(2009) Capture-recapture estimates of the local prevalence of problem drug use in six French cities. European Journal of Public Health 19(1): 32–37.

43.

Vanyukov

Tarter

Kirillove

, et al. (2012) Common liability to addition and ‘gateway hypothesis’: Theoretical, empirical and evolutionary perspective. Drug and Alcohol Dependence 123S: S3–S17.

44.

Von Sydow

Lieb

Pfister

, et al. (2001) The natural course of cannabis use, abuse and dependence over four years: A longitudinal community study of adolescents and young adults. Drug and Alcohol Dependence 64: 347–361.

45.

Von Sydow

Lieb

Pfister

, et al. (2002) Use, abuse and dependence of ecstasy and related drugs in adolescents and young adults – A transient phenomenon? Results from a longitudinal community sample. Drug and Alcohol Dependence 66: 147–159.

46.

Wagner

Anthony

(2002) From first drug use to drug dependence: Developmental periods of risk for dependent upon cannabis, cocaine and alcohol. Neuropsychopharmacology 26: 479–488.

47.

West

Gossop

(1994) Overview: A comparison of withdrawal symptoms from different drug classes. Addiction 89: 1483–1490.

48.

Winstock

Shiner

(2015) The best antidote to drug use isn’t tougher laws – It’s growing old. The Guardian, 9 June 2015. Available at: www.theguardian.com/commentisfree/2015/jun/09/global-drug-survey-antidote-drug-use-growing-old (accessed 23 January 2020).

49.

World Health Organization (WHO) (2002) World Health Report 2002: Reducing risk, promoting healthy life. Geneva: WHO.