Every day, research published in highly respected medical journals informs life-changing health decisions – from which vaccines to give our children, to ways to reduce our dementia risk.
And we trust that medical advice is based on good evidence. But could that trust be under threat?
The worrying truth, as Good Health can reveal, is that some top journals are refusing to retract influential research that’s since been proven wrong.
Such fundamental flaws are often discovered by fellow scientists who seek to repeat the study experiments, only to find that the data simply doesn’t stand up.
A retraction can be simply done, by declaring in print and online that the research has been withdrawn, yet journal editors may be reluctant to retract discredited studies because they fear it may lower their publications’ prestige. Sometimes the authors of misleading research threaten the journals with expensive legal action.
Even when journals do eventually retract bad science, this can often be a year or more after publication – with a fraction of the impact that the original research had.
At best, this leads to countless people receiving baseless treatments or advice – at worst, it could prove lethal. Indeed MPs on the science, innovation and technology committee have warned of the ‘serious harm, including loss of life’, highlighting papers that proclaim quack cures for infections such as Covid-19.
In a 2023 report, they urged journal editors to retract faulty research within two months of problems being identified.
In a 2023 report, MPs on the science, innovation and technology committee urged journal editors to retract faulty evidence within two months of the problem being identified
One major danger is that when discredited results are not retracted, they get cited by other studies in other journals, which in turn get cited by further studies, and so on, poisoning the body of modern medicine.
One example is the now commonly held belief that people who develop hearing loss and don’t wear hearing aids raise their risk of dementia.
An internet search of ‘hearing aids and dementia’ brings up a plethora of results confirming the dangers. High up is this statement from the Alzheimer’s Society: ‘Studies have shown that people who use hearing aids to manage their hearing loss are less likely to develop dementia.’
This was based on a 2023 study in The Lancet Public Health, which concluded wearing hearing aids when needed could cut dementia risk by nearly 10 per cent.
Understandably, the study made international headlines – and was subsequently cited in more than 70 further studies.
But what very few people appear to have noticed is that The Lancet retracted the original study in December 2023.
This was after Jure Mur, a postdoctoral scientist at the University of Edinburgh, tried to repeat the study – but found the numbers would not add up. In fact, he found that, among people with hearing loss, the dementia rate was higher for those using hearing aids.
Jure Mur emailed the study’s authors in China several times but received no reply. He then contacted the journal’s editors, who issued a retraction – albeit eight months after publication.
After discovering that figures in a paper which concluded wearing hearing aids could cut the risk of dementia did not add up, Jure Mur got the study retracted
One of the most controversial current cases over retraction concerns four studies published in the early 2000s, which concluded that women who have abortions raise their risk of mental health problems, such as depression.
Despite leading academics repeatedly critiquing the methods and results – and other research showing otherwise – these studies are being used in US legal cases to limit women’s access to abortion.
The studies were all authored by Priscilla Coleman, who retired recently as a professor of human development and family studies at Bowling Green State University, Ohio.
Her research was used in the 2022 US Supreme Court judgment that ended American women’s constitutional right to abortion.
One of her controversial studies was published in the BMJ in 2002, another in the British Journal of Psychiatry in 2011.
In February last year Julia Littell, a research professor at Bryn Mawr College, Pennsylvania, along with 14 fellow professors, wrote an analysis – also published in the BMJ – highlighting flaws in the interpretation of the data, arguing that ‘after correction of these errors abortion was not associated with increased risk of subsequent maternal depression’.
Professor Littell told Good Health that although the BMJ published criticism of the original article and partial corrections, ‘serious methodological problems remain uncorrected’.
A spokesman for the BMJ told Good Health: ‘This matter is still under review.’
Professor Littell also said: ‘In 2022, the British Journal of Psychiatry’s owner, the Royal College of Psychiatrists, overturned the recommendations of an independent panel that the article should be retracted after the author and her lawyers threatened legal action.’
Few people appear to have noticed that The Lancet retracted the hearing aids and dementia study in December 2023
Two members of the journal’s editorial board, as well as independent panel members, resigned in protest at the decision, a BBC Newsnight investigation reported in 2023.
In response to the Newsnight claims, Professor Coleman told the programme she had published 63 peer-reviewed journal articles, with the majority related to the psychology of abortion. She pointed out that her career had spanned three decades with publications in highly reputable academic journals and that criticism of her work was driven by the political nature of the topic.
The editors of the British Journal of Psychiatry told Good Health: ‘After careful consideration, given the time since the original article was published, the widely available public debate on the paper, including the letters of complaint available alongside the article online, and the fact that the article has been subject to a full investigation, it has been decided to reject the request for the article to be retracted. We regard this matter closed.’
According to Nature, more than 10,000 research papers from medical and science journals were retracted in 2023 – a record figure.
But Ivan Oransky, co-founder of Retraction Watch, told Good Health: ‘We think that there are ten times as many articles that should be retracted.
‘A major problem is that people who work in medical journals are rushing to publish – as there is tremendous commercial pressure on them, so they may not have adequate resources to check crucial things properly.’
Yet they are often in no hurry to retract them.
Julia Littell, a research professor at Bryn Mawr College, Pennsylvania, wrote an analysis of the ‘serious methodological problems’ of a study linking abortion and depression
Take research published by two journals early in the pandemic that claimed the anti-parasite drug ivermectin could treat Covid-19.
It was hailed as a miracle drug, mainly by anti-vaxxers. However, the studies supporting these claims were later shown to be unreliable.
The journal Toxicology Reports published one such paper in March 2021 – it took 14 months to retract it. The Journal of Antibiotics published similar research in June 2021, retracting it eight months later.
Such delays, though, pale by comparison with the 12 years it took for The Lancet to retract the debunked and dishonest research that the gastroenterologist Dr Andrew Wakefield published in 1998, claiming a link between the measles, mumps and rubella (MMR) vaccine and an increased risk of autism.
Research published last year found that a quarter of US adults still believe that the MMR vaccine causes autism.
And it’s not only the general public. A 2022 study in the Journal of Clinical Epidemiology found 127 examples of discredited studies that had been included in clinical practice guidelines and systematic reviews (which compile previous research to produce ‘best of’ results).
Even when debunked research is retracted, the process may actually strengthen the influence of bad research – this is partly because we prefer to believe information once we’re told it, rather than go through the mental effort of reversing our beliefs, suggested psychologists at Bristol University in 2017.
They recommended that when editors publish retractions they avoid repeating what the faulty study originally said.
