A Journal Promised to Retract a Flawed Autism Study. It's Still Online.

For years, researchers have been discussing what to do about "junk science." Junk science is a major problem resulting from medicine's "publish or perish" culture. It comes in different forms such as undeclared conflicts of interest (COIs), sneaking through peer review, or being published in a predatory journal. Once published, it is incredibly difficult to remove from the scientific record. Researchers have been aware of this issue for a long time. Most complain about it, but no one really knows what to do about it.

Junk science has infiltrated both academic discourse and public discussions. It impacts trust in science and scientific advances by hindering both. This is my story about trying to get just one article retracted, how long it's taken, and why it matters.

The Case Study: A 15-Month (and Counting) Saga

Since summer 2024, I've been watching a specific case report very closely. The case study looks at twins with autism and a specific treatment regime that included parental coaching, supplements, consulting a naturopath, and more.

The methodology of the article is important to consider. The authors made claims about how the study demonstrated that the twins' symptoms were reversed. But, again, this article is a case study. A case study cannot determine causality. It's an observational study of two children who were provided with myriad treatments, from supplements and homeopathy to parental coaching. According to the paper, the children showed dramatic improvement.

But there is reason to be suspicious of the results.

One of the main issues I initially raised with the journal was the lead author's perceived COI. According to the article, "A number of dietary supplements, including omega-3 fatty acids, a multivitamin, vitamin D, carnitine, 5-methyltetrahydrofolate, and bio-individualized homeopathic remedies, were taken by both girls." At the time of the study and publication, the lead author was jointly affiliated with a university as well as Documenting Hope.

Documenting Hope has many business partners, many of which are supplement companies. Given that the lead author had, at the time of study publication and at the time of this writing, affiliation with this organization that has close ties to supplement companies, one would expect to see acknowledgement of that in the COI section of the article. Yet when I looked, none were declared, so I wrote to the journal.

That first email began what has turned into a saga of the last 15 months and counting. I've emailed the journal regularly to check in on the article. I've been given a lot of promises, first in regards to updating the COI section (October 2024) and then around retraction (April 2025). Neither have happened to date.

The Emails

I've been emailing with the journal roughly every 2 months. In October 2024, they informed me:

"The Editorial Office has conducted a thorough investigation in consultation with our Editorial Board and Ethics Committee. We have determined that the identified potential conflict of interest did not affect the interpretation of the work or the recommendations made by the editors and peer reviewers. To ensure full transparency, we propose to publish a correction to the paper that will fully disclose the conflict of interest."

This was followed by a confirmation that they would publish a corrected version of the COI section as soon as they were finished confirming "some details." I emailed a few more times throughout the rest of the year and into the beginning of 2025. In March 2025, the journal emailed to say:

"After the investigation, we have drawn the conclusion that the paper will be retracted. We are preparing the materials for the retraction of this paper."

They couldn't say exactly when the retraction would happen, only that it was planned. They stated that they were "dealing" with the authors and hoping to have a retraction published "as soon as possible."

In May 2025, I was informed that a retraction file had been prepared.

"The retraction file has been prepared, it is expected to be published in this month."

I checked every day of May to see if the retraction had been published. Alas, May came and went and no retraction was published. I emailed in early June to inquire.

"Due to some legal issue, the retraction did not publish last month. We are actively solving the problem, and wish that there will be a result soon."

The retraction didn't publish in June either. Nor July. Nor August. The last correspondence I had with the journal was September 2025. At that time, I asked if the journal would consider publishing a notice or at least updating the COI section before the retraction.

"This case takes much more time to get a final result. And our journal editorial office is considering to publish a notice, after certain communications with authors, as well as MDPI Research Integrity and Publication Ethics Team."

As of the end of October 2025, the paper remains as published. No corrections have been made to the COI section. No retraction. According to public record, the paper is still in good standing and therefore fair game when considering the evidence of treatments for autism.

Systemic Failures and the Way Forward

Typically, articles like this are ignored by the mainstream press. However, in this case, the Daily Mail and The Telegraph wrote it up. This boosted its exposure and catapulted it into the world of social media yelling.

This tale is important for all researchers and every person to know. Once junk science has infiltrated the scientific record, it is extremely difficult to remove it. And once it has permeated the public consciousness, it's virtually impossible to change people's memory or opinion of it. Just think back to the retracted measles, mumps, and rubella (MMR) vaccine study from Andrew Wakefield, MBBS. The study was retracted, but a substantial number of people still believe there is a link between the MMR vaccine and autism.

It's not uncommon for retractions to take a long time. Just ask Retraction Watch. However, while I haven't given up hope that this retraction will happen, for now, it's still published and stands without even an updated COI section. As far as anyone is concerned, it's still a part of the scientific conversation about treatments for autism. The damage has been done. Even if it is eventually retracted, the Telegraph and the Daily Mail are almost certainly not going to run articles with splashy headlines covering the retraction. Retractions rarely make the news.

Meanwhile, in an official HHS press release about autism from September 2025, another article from the same journal was referenced as evidence for using leucovorin (folinic acid) to treat autism. Just one article, from the same journal, on the same topic. This is especially odd since the journal isn't specifically dedicated to autism research.

Junk science isn't just an ivory tower problem; it's everyone's problem. This is especially true when people with possible ulterior motives begin to cherry-pick articles to present as evidence. Science is difficult to conduct. Science is difficult to write. Science is difficult to understand. That is all exacerbated by the infiltration of junk science into the discourse. It's eroding public trust. It's making people doubt scientific consensus. It's made it seem like every argument has evidence to support it.

Evidence isn't just about the quantity; it's about the quality. The case report I have been following lacks rigorous methodology, isn't based on established treatment guidelines, and was conducted by at least one person with a perceived COI -- if not a real one. Yet, it can still be pointed to as evidence and that's the exact problem.

The failure to retract this single article -- and others like it -- directly contributes to misguided public health policy and allows the promotion of unproven, potentially harmful treatments.

This “Second Opinion” originally was posted on MedPage Today. It is used with permission by the author. You can find the original article here