It’s almost impossible to get too excited about a drug that could potentially help millions of people safely lose a remarkable amount of weight, dramatically improving their lives and bringing an end to the obesity epidemic.
But people have somehow managed to get too excited anyway.
Don’t get me wrong. I’m incredibly impressed with what the research on the drug semaglutide—sold under the names Ozempic and Wegovy—shows in terms of weight loss (technically Ozempic is for diabetes and the stronger Wegovy is for weight loss – but very many people are using Ozempic “off label” for the latter purpose). Multiple trials have found that people injecting semaglutide lose about 15 per cent of their body weight compared to a placebo injection. It does this by making dieting, which can be torture for so many people, vastly easier. No wonder everyone’s talking about it (and even those who aren’t talking about it are probably secretly taking it).
Last week, an article appeared in The Atlantic entitled “Did Scientists Accidentally Invent an Anti-addiction Drug?”. The drug in question is, of course, semaglutide: the article noted how many people taking it have noticed effects that go far beyond their appetite. For example, “they have reported losing interest in a whole range of addictive and compulsive behaviors: drinking, smoking, shopping, biting nails, picking at skin”.
Could it be, the article asked, that the reason semaglutide works to reduce appetite isn’t just that it slows down your digestion, but also that it affects the brain, tamping down your appetite for food – and maybe your appetite for other things, too?
The article went viral, being shared across social media by people who were understandably fascinated that the drug could have so many different effects. “They should put Ozempic in the water, like fluoride,” said one almost-serious tweet.
It’s certainly refreshing to have a break from the relentless, and largely spurious, coverage of semaglutide’s side effects. And it’s interesting to consider unexpected positives for once (after all, the weight-loss effects of semaglutide were themselves discovered as a happy accident in the trials of the drug for diabetes).
But we’re getting a little ahead of ourselves here. The anecdotes from individual patients are very intriguing, but it’s not hard to think of reasons why people taking the drug for the first time might report effects that aren’t just to do with weight loss – but which don’t involve semaglutide actually affecting the parts of the brain responsible for addictions and compulsions.
Perhaps being on a new drug, and putting in effort to go on a diet or get more exercise (as you’ll likely have to, if you want to lose weight), makes people change their routines enough that they no longer find themselves in so many situations conducive to drinking or taking drugs. Maybe they become more confident and outgoing (or even just more physically mobile), and thus no longer stay at home drinking. Maybe the side effects – which include nausea and vomiting – are harsh enough for some that they’re distracted from concerns like nail-biting. Maybe people’s behaviour changed for a brief time, just due to the novelty of being on a drug, and they’re about to revert to their old habits.
Or maybe, for whatever reason, we’re just not hearing from the large majority of semaglutide patients who experience only food-related effects and nothing else. We can speculate all day long – and that’s the problem. Anecdotal, non-scientific accounts of the drug’s effects are really just an inspiration to do some proper research – if we try to draw conclusions from them alone, we risk major errors.
And unfortunately, there are absolutely no research studies that look at semaglutide and addictive behaviours specifically. There, however, are some that examine related drugs – not quite the same chemical but from the same subtype – and they’re mainly in non-human animals. For instance, a few studies have looked at how exenatide, a drug with some of the same characteristics as semaglutide, affects how rats self-administer cocaine (they can do so by pressing a lever in their cage). They’ve generally found promising results, with the rats on exenatide taking less cocaine overall. There are similar studies on amphetamines (also generally reporting positive results), nicotine (the same), and opioids (generally showing no effect of the semaglutide-like drug).
Aside from the obvious difficulties in extrapolating a study of rats to humans, these studies all tend to be very small (it’s difficult and expensive to do a rat study, let alone one where you have to administer them an illegal drug). Another important problem – common to a great deal of rodent research – is that for the vast majority of studies, only males were included, so even if we believe the effects are real and extend to humans, which is a big ask in and of itself, they skip over half of the population.
So we do need research in humans. One initial, tiny study in human cocaine addicts showed no effect of exenatide on their desire for, or taking of, cocaine (having said that, with only thirteen participants, the study would never have been particularly informative regardless of what they’d found). Another cocaine study had only three participants and thus isn’t a whole lot better than the anecdotes we’ve previously discussed.
The Atlantic article cites one study that showed no overall effects of exenatide on alcoholism – but says there was an effect found in obese participants, and an effect seen in brain scans on parts of the brain relevant for addiction. But “subgroup analysis” like this, where you run your statistical analysis again in a smaller subsection of your full dataset, is notoriously unreliable. If you cut up a dataset enough ways, you’ll eventually find something that looks like a real effect, even if it’s entirely due to random chance.
And as for the brain-imaging part of the study, a mere 22 participants got their brains scanned, making the conclusions very unreliable. Besides, what use is an effect seen on a brain scan if the thing we really want to impact – how much people drink – remains unaffected?
I’d say there’s very little to write home about, here. To be fair, the Atlantic article was itself very measured, noting that we’ll have to wait for studies to come in. Tantalisingly, it mentioned that several animal studies of semaglutide itself, rather than related drugs, are on the way – and it seems that several addiction studies in humans are in the works, too.
We can hold out hope that semaglutide might work not just for diabetes and obesity but also addictions – or at least gives us scientific clues we can use to design better addiction drugs. But we’ll need a lot of serious, high-quality research to get us there. Unlike the excellent, solid trials for the drug’s previous impacts, the addiction research has scarcely even begun.
If it turns out, when the research comes in, that semaglutide doesn’t make much of a difference to addiction, we’ll “only” have its dramatic effects on blood sugar and body weight to consider. That’s already more than enough.
