Admitting where our knowledge of cannabis is limited is as important as acknowledging what strong research evidence can make us certain of, says Ian Hamilton

In 2017 an estimated 91.2 million adults in the EU reported using cannabis in their lifetime, with eleven countries reporting an increase in young adults’ (15-24 years old) use of cannabis from the previous year. The picture in America is less clear, with some reports suggesting that young people’s use has stabilised or declined, while adult use is rising. It is probably too early to tell what impact changes in cannabis regulation in America and other countries will have on use of cannabis. However, it is likely that as more countries allow access to cannabis for health reasons or just for recreation, cannabis will become easier to obtain and in a range of more potent formats, such as cannabis concentrates. As the cannabis industry develops and begins to market the benefits of its product, it’s therefore important that we are able to accurately assess the potential for harm.

Understanding the risks of developing dependence to such a widely used drug is clearly important, but there are still gaps in our knowledge that need to be addressed. We know that some individuals who use cannabis will develop dependence, but some people are unaware that addiction to cannabis is even possible—perhaps its status as a relatively benign drug has helped obscure this possibility. This image of cannabis as innocuous is likely promoted by claims (many of which have still to be evidenced) that it can cure or alleviate all manner of ills, including cancer. It’s even been studied by some researchers for its ability to alleviate addiction to other drugs, such as opiates.

The attention given to the health problems associated with cannabis tend to focus on the risk of psychosis rather than the risk of addiction. However, a number of largely observational studies have reported rates of cannabis dependence that vary from 0.5% to 42% in those that have used cannabis in the past year. Such wide ranging estimates would suggest that there are problems in trying to count the number of people with cannabis dependence, and consequently, in determining the risk of becoming addicted to cannabis.

If you look at how cannabis dependency is diagnosed and studied, it reveals a number of  problems that might explain these wide ranging estimates.

One example is the Diagnostic and Statistical Manual of Mental Disorders (DSM–5) used by researchers to assess cannabis dependence. DSM5 states that cannabis withdrawal symptoms should not be due to withdrawal or intoxication from other substances, but establishing this is easier said than done. Several symptoms of cannabis withdrawal are not unique to just cannabis. For example, tobacco is frequently combined with cannabis in a joint, so if someone stops smoking a joint they could experience tobacco withdrawal as well as withdrawal from cannabis. Patient and clinician could wrongly attribute the withdrawal symptoms to cannabis rather than tobacco and vice versa. As withdrawal symptoms are used as criteria to indicate dependence, some people will have potentially been assessed as cannabis dependent when it was actually tobacco dependence and withdrawal they were experiencing. The reverse may also be true. Unfortunately, some studies do not consider the part that tobacco may play in presenting symptoms of dependence when it is combined with cannabis, therefore overlooking what could be a crucial factor.

Another difficulty with estimating cannabis dependence has been our failure to fully understand the differences in rates of dependence between men and women. At a population level, generally twice as many men as women report using drugs like cannabis. Estimates as to the prevalence of cannabis dependence is also roughly twice as high for men as women, reflecting the usage ratio and suggesting that men and women are equally affected. However, there are clear differences in the way men and women are treated in society, and some academics have argued that these gender differences have been reflected in the diagnostic criteria used in the various editions of DSM.

Although such studies are rare, research that does attend to the role of sex in cannabis dependence has revealed some intriguing differences. The journey from cannabis exposure to dependence seems to be a shorter one for women compared to men, a phenomenon known as telescoping. Recent preclinical biological investigation suggests that women develop a physical tolerance to the psychoactive ingredient in cannabis Δ9-tetrahydrocannabinol (THC) more quickly than men. Sex differences in the pharmacokinetics of cannabis are still to be fully revealed; if we understood them better it could mean that dependence is assessed and defined in a more sensitive way.

Our understanding of cannabis dependence could also be skewed by the fact that much of the research so far tends to be drawn from the populations of higher income countries, despite exposure to cannabis also occurring in lower and middle income countries.

So diagnostic criteria, sex and gender, and sampling bias are just three examples that demonstrate our limited knowledge of the potential that cannabis use has in facilitating dependence. Yet there are others, such as study variance in the proxy measures used to establish cannabis dependence. Some studies, for example, use daily cannabis use to define dependence, while others use less frequent exposure. Little attention is also given to how cannabis potency has changed over time. Most reports of cannabis potency in recent years have pointed to the increasing strength of all varieties of cannabis available to users, particularly as producers of cannabis resin compete with producers of hybrid forms often referred to as “skunk.”

Our inability to accurately estimate the prevalence of cannabis dependence has several consequences. Policy makers who are considering opening up access to cannabis for recreational or medicinal reasons need to have confidence in the estimates of risk, as do those using cannabis or considering using it. The public health messages that governments and scientists disseminate need to have credibility.

Admitting where our knowledge is limited is as important as acknowledging what strong research evidence can make us certain of. Given the problems inherent in previous attempts to measure cannabis dependence, it’s time to confess that there’s more we don’t know than we do.

Ian Hamilton is an academic at the University of York with an interest in addiction and mental health. He previously worked as a mental health nurse with people who had combined mental health and substance use problems. Twitter @ian_hamilton_

Competing interests: I am affiliated with Alcohol Research UK.