In the last couple of years, there has been a healthcare revolution. Semaglutide, a drug initially prescribed for diabetes under the name Ozempic, has been shown to reduce obesity dramatically, and repackaged for that use as Wegovy. Other drugs have since been approved.

It’s a big deal. More than half of US and British adults are overweight or obese, and there are few really effective treatments: prescribing dieting is all but useless. As a result, sales have been astronomical. Ozempic’s manufacturer, Novo Nordisk, is now the most valuable company in Europe.

In Johann Hari’s latest book we learn that the author himself takes semaglutide to lose weight. But he is conflicted: in 2018’s Lost Connections, he wrote about what he considers the over-medicalisation of treatment for depression, so is wary of the idea of a “magic pill”. In any case, he discusses the drugs’ discovery and mechanisms, their health benefits, possible risks and how we ended up in a situation where so many of us are obese to begin with.

And if I were coming to it fresh, I’d think it was OK. It presents a vexed debate reasonably fairly. It never quite tips into scaremongering, despite a tendency to foreground low-probability risks. On the negative side, it’s content-light for a scientific book, and Hari’s breathless style grates somewhat – he’s always learning things. “The advantages of these drugs were now clear to me,” he writes, noticing with shock that obesity is bad for your health.

He relies on what seem to me to be convenient quotes from pseudonymous friends, who frequently provide the perfect foil at the perfect time – one he calls Judy, for example, helps him realise it’s no good telling desperate people that we should simply fix society, rather than providing them with an effective treatment.

My scepticism here, though, is a reminder that I am not coming to it fresh. Hari has a history. In 2012 he left his columnist job at the Independent after it was revealed he had stolen quotes and libelled rivals via online sockpuppet accounts. Hari has since admitted that he “failed badly”, but the facts remain. Magic Pill never mentions this backstory, or explains why we should trust him now.

So I went through the references even more carefully than I normally would, to see if the studies backed up his points. Not all of them did. He argues, for example, that “people who look at social media” can end up with “profoundly distorted” body images. The endnotes cite research from 2022. But in the passage itself, the study of 100 people he mentions in support of his point is from 1987.

Elsewhere he claims an educational intervention makes children “half as likely to become overweight or obese”. The result in question is not statistically significant, and by convention can be considered a fluke. He also seems to think the “glucagon gene” is produced in the pancreas, suggesting that he doesn’t know what genes are. He talks about it for long enough that I’m fairly sure it’s not a typo.

Obviously I couldn’t fact-check the anecdotes, so my notes are full of cynical little queries: Does Judy exist? Does the little boy at the Grand Canyon? Maybe they all do. Maybe Hari really did develop anhedonia as a side effect, a useful peg for the chapter on semaglutide and the brain. He may have changed his ways, but I can’t know for sure, and because of his record I find it hard to have much confidence in him.

It’s a shame: there is a book to be written about the rise of weight-loss drugs, and how they will change society. But Hari has not only failed to write it; the way he’s done so has undermined his already dwindling stock of credibility.

• Tom Chivers is a science writer with Semafor. Magic Pill by Johann Hari is published by Bloomsbury (£20). To support the Guardian and Observer, order your copy guardianbookshop.com. Delivery charges may apply.