Cannabis: when it becomes a problem and when it does not

“Every evening, yes. But that isn’t addiction. Addiction is alcohol. This is just how I get to sleep.”

I hear that sentence from people of all ages, and there are two claims inside it. One is probably true — plenty of people really do get to sleep that way. The other is a question that cannot be answered by comparison, only by looking at what is actually happening in that person’s life.

Cannabis is peculiar in conversations about addiction because the debate around it is ideological. One side insists it is an innocent plant, the other that it is the first step towards ruin. Both sides say something true and both exaggerate. I would like to write a piece that does neither.

What we actually know

Let us start with the figure most often misquoted in both directions.

The great majority of people who ever try cannabis do not develop a problem. Epidemiological research suggests that roughly one in ten of those who ever use it goes on to dependence. Among those who start in adolescence the proportion is considerably higher, as it is among those who use daily.

That means two things at once. First: the claim that cannabis is “as addictive as heroin” is not true, and such claims are actively harmful in prevention, because a young person who sees a different reality around him then stops trusting everything else you tell him. Second: the claim that cannabis dependence does not exist is not true either. It has been described, it appears in the diagnostic classifications, and it has its own withdrawal course.

One more thing gets overlooked in these conversations: today’s products are considerably stronger than those of twenty or thirty years ago. When an older person says “I did it in my youth too”, he is not talking about the same thing as his son. That difference in strength is precisely what research links to a large share of the risks.

Where the line runs between habit and problem

Now the part people expect, and where I am going to disappoint: the line is not about how often, or how much.

The diagnostic criteria for a substance use disorder are built, in every modern classification, around consequences and control, not around quantity. The reason is simple: identical use produces two completely different lives in two different people. One person uses at weekends and has orderly work, relationships, sleep. Another uses just as often and has let go of everything that used to matter to him.

The questions that actually tell you something run roughly like this. Have you tried to stop or cut down and not managed it? Have things you once cared about dropped away? Do you carry on even though you know it is spoiling something — sleep, a relationship, work, mood? Do you need more than before for the same effect? Can you not fall asleep without it? Have you noticed the day arranging itself around it?

If several answers are yes, that is not an accusation. It is information. And it is equally information if all six answers are no — in which case we probably are talking about a habit.

Why sleep is not so simple

The commonest reason people give is sleep. And it is true: you fall asleep faster.

The complication is what happens afterwards. Cannabis alters the architecture of sleep — above all it shortens the dreaming phase. You sleep, but you rest differently than you otherwise would. Many people notice this as mornings that feel heavy and as the disappearance of dreams.

More important still is what happens on stopping. The dreaming phase comes back with a swing: for a few nights sleep is broken, dreams are unusually vivid and unpleasant, and falling asleep is harder than ever. For someone who started precisely because of insomnia, that looks like proof that he cannot manage without — when in fact it is a transient state. How an aroused nervous system settles, and why predictability matters so much for sleep, is what I write about in the piece on insomnia and the nervous system.

So with people who use cannabis for sleep I generally advise tackling the sleep itself first — with methods that work in the long run — and only then reducing. The reverse order almost always fails.

The anxiety it calms and the anxiety it raises

A similar contradiction holds for anxiety. In the short term many people feel a settling; tension drops, thoughts slow. In the long run, research on regular users shows more anxiety, not less, and in some people anxiety intensifies markedly for a few weeks after stopping, which they read as evidence that they were fine before.

The mechanism is not mysterious. Whatever removes an unpleasant feeling quickly, the body learns to reach for earlier and earlier. And every time we remove an unpleasant feeling, we take from it the chance to settle on its own — so the system responds faster on the next round. It is the same mechanism I describe in the piece on what happens in the brain during anxiety.

There is one more curious point: in a proportion of people cannabis increases anxiety immediately rather than calming it. Same substance, opposite effect — which says a good deal about how badly these things generalise.

Why it is different for young people

Here the field argues comparatively little. The brain keeps reorganising itself into the mid-twenties, and it is precisely in that period that it is more sensitive to anything that regularly acts on it.

Research on those who start early and use often shows a higher risk of a use disorder, poorer educational outcomes and more difficulty with attention. Part of that difference is probably not the substance itself but the circumstances that lead to early use — it is only fair to say so. Yet even when researchers account for those circumstances, the difference does not disappear.

The most delicate subject is the link with psychosis. What we know is this: in people with a vulnerability — a family history, earlier unusual experiences — the risk of a psychotic disorder with regular use of stronger products is noticeably higher. Whether this is cause or coincidence is still debated, but the causal part shows up ever more clearly in recent European research. For an individual, something simple follows: if anyone in your family has had a psychotic disorder, that is information worth taking into account.

What a partner or a parent sees

People around a user often do not see the smoking. They see something else, and for a long time they cannot name it.

They see someone present but not really there. Conversations that end sooner. Plans for next month that are always soft. A circle of people that has narrowed. Irritability on days when it was not possible. None of this is evidence about a substance — it could equally be depression, burnout or simply a bad stretch. Which is why the only honest step is a question, not an accusation.

What does not work in such conversations is usually what people open with: statistics, articles, warnings about the brain. What works better is a description of your own experience — “I miss you in the evenings” says more than any study and does not put the other person on the defensive.

Medical use is a different subject

To be clear: there are conditions in which cannabis-derived products are used under medical supervision and with evidence behind them — some forms of epilepsy, spasticity, some chronic pain, nausea in chemotherapy.

That is not the same subject and is not an argument in either direction. That something is a medicine in a specific condition and under supervision does not make it harmless in everyday use — the same holds for plenty of things in the medicine cabinet. And conversely: the fact that someone smokes regularly does not mean that a person whose doctor prescribed something is doing the same thing.

What a therapist does and does not do

When someone comes to me with this, I do not start by asking how to stop. I start by asking what it does for him. The answer almost always makes sense: it calms, it closes the day, it interrupts churning thoughts, it makes solitude bearable, it lets someone switch off after work at all.

Until that job is assigned to someone else, stopping means a loss with nothing in its place — and that kind of stopping rarely holds. It is the same way of thinking I set out in the piece on the symptom as a message, and it is why, with people who smoke in order to stop their thoughts, I often work first on rumination itself.

What I do not do: I do not lecture, and I do not make abstinence a condition of being seen. Whoever walks into a service where he must first stop in order to be listened to generally does not come back.

Harm reduction, because abstinence is not always the first goal

For many people the goal is not complete abstinence, at least not immediately. That is not a defeat for my profession; it is a reality you can work with.

Useful intermediate goals are: not combined with alcohol or tranquillisers, never before driving, not during exam weeks or periods of heavy load, not as the only route to sleep, and at least some days each week without. Every one of those shifts reduces harm, even if nothing else changes.

And most concretely: if someone experiences severe anxiety or confusion after use and it does not pass, that is not something to sleep off — it is a reason to seek medical help. When a doctor is the first rather than the last step I wrote about here.

Where my own argument falls apart

Now the section against myself, which this subject genuinely requires.

First: a large share of people who use cannabis regularly have no problem that needs solving. They work, they have relationships, they sleep. If I am an honest therapist, I have to admit my profession has nothing to offer them, and that any persuading on my part would be moralising in professional dress.

Second: the evidence for therapy in cannabis use disorder is not thrilling. The best we have — a cognitive behavioural approach combined with motivational interviewing, sometimes with rewards for abstinence — helps, but the effects are moderate and fade over time. There is no medication that works so far. Anyone promising more is promising too much.

Third: scaring people does not work. Prevention that tells a young person something he can disprove in one evening loses its credibility for everything that comes after. It is more useful to say what holds: that most people will be fine, that a minority will not, and that you cannot know in advance which group you are in.

What to remember

Cannabis dependence exists, but most people do not get it. The line is not about how much or how often — it is about control, letting things go, and carrying on despite the damage.

Withdrawal is not life-threatening, but it is unpleasant enough that the first week is often the reason people return; sleep and anxiety get worse before they get better.

For young people, and for those with a family history of psychosis, the picture is different and caution is warranted. For everyone else a more modest conclusion holds: if it is not limiting you, it is probably not your main problem — and if it is, that is not measured in grams but in what you no longer do with your life.