Medical Examiner

String of Lies

Our confusion over dental floss is the result of years of biased research, thanks to the war between floss and mouthwash.

dental floss.

All this controversy over some little white string?

Dmytro Panchenko/Thinkstock

My French roommate Julie had just come into the bathroom to get something.

“What are you doing with that string in your mouth,” she said, with a look of horrified surprise.

It turns out she had never seen floss. We were both in our 20s, and my dentist had spent over a decade instructing me to floss, so I was a little horrified, too. Was she sure? She hadn’t seen floss or even heard of it? Never, nada, nothing?

Non!” she said.

Julie had beautiful teeth—luminous, white kernels that pulled into a sultry smile. Did the French know something that we didn’t know about interdental hygiene?

This week, I learned that they might, thanks to an intrepid Associated Press reporter who noticed that flossing had been dropped from the United States Dietary Guidelines earlier this year and wondered why. The AP filed a Freedom of Information Act request pressing the government to say where the scientific evidence had come from for the original recommendation on floss. The government admitted to the AP, in a letter, that the evidence had been too weak, which is why it had taken out the recommendation.

How did we get into this tangle of silky string? A quick look at the scientific literature shows there are dozens of articles investigating floss. Most of them claimed to show a small reduction in plaque and a small reduction in gingivitis associated with flossing. But most of the studies were funded by pharmaceutical or consumer product companies, or written by authors who had received money from them: Johnson & Johnson, Procter & Gamble, GlaxoSmithKline, Colgate-Palmolive. It turns out there is such a thing as Big Floss, and we may all be in its pocket.

Apparently the American Dental Association, and most dentists around the country, took these studies at face value. But in recent years, several large review articles put together all the available data and found that when they considered the full picture, there was very little evidence that floss was effective at all. At this point, the government decided it had to walk back its recommendations.

One review article—written by Dutch researchers in 2008—found that regularly flossing had no effect on gingivitis or plaque. Another review—sponsored by the Cochrane Collaboration in the U.K. in 2011—found that floss did have a small effect on gingivitis and a possible very small effect on plaque. But the authors of the Cochrane review also concluded that the quality of the available evidence was “very low,” and the evidence on plaque was especially “weak and very unreliable.” For instance, many of the reviewed trials had only followed patients for one to three months, so the small benefit they observed could have been a “trial effect”—a bump in patient compliance and motivation because of the novelty of being in a clinical trial.

I emailed the authors of the reviews—and a dentist based in Amsterdam, G.A. van der Weijden, answered. I asked him about the American Dental Association recommendation that people floss every day.

“There is no scientific evidence to support this recommendation should be given to the general public,” Dr. van der Weijden said. As people age, if they get gum disease, then they should probably clean between their teeth, he said. “But based on our work … wood sticks would appear to be the most efficient tool, and in the case of open interdental spaces, interdental brushes are the first choice.”

Wood sticks? Like toothpicks? An internet search turned up a variety of “dental” wood sticks—hefty toothpicks with one oblong end.

Dr. van der Weijden added that floss may work “if you are able to use it at the level of a professional.” (Which, apparently, is like this.)

So Dr. van der Weijden is somewhat skeptical of floss. However, he is not particularly concerned by the supposed influence of Big Floss. In fact, he had no problem with dental floss studies being funded by dental floss companies.

“No researcher with a clear mind wakes up in the morning and says to himself: ‘you know what, let’s evaluate floss,’ ” he said. “Any study that is product-related needs a sponsor, simply because the government does not pay for this type of research. It depends on the researcher whether there is a conflict of interest and whether the industry has an influence on the outcome.”

Actually, even these industry-funded studies aren’t just straight-up evaluating whether floss works. As the Cochrane review pointed out, most scientific trials of floss were not designed to answer basic questions about the effectiveness of using the substance versus not using it and instead had trial designs (and short study periods) that focused more on comparisons and left huge gaping questions. Unfortunately, this lack of precision in industry-sponsored research is not limited to dental floss. In the United States, 58 percent of medical research is funded by industry sources, and industry research is notoriously difficult to interpret. This is largely because, unsurprisingly, most published industry studies have results that reflect favorably on the product under investigation—i.e., the product being championed by the sponsor. One study in 2010 found that 85 percent of industry-funded studies reported positive results, compared to only 50 percent of government-funded studies. A 2014 study of “head-to-head” clinical trials found that 97 percent favored the sponsor’s product.

John Ioannidis, author of the article “Why Most Published Research Findings are False,” has said that industry sources “ask the wrong questions with the wrong short-term surrogate outcomes, the wrong analyses, the wrong criteria for success … and the wrong inferences.” He has also warned that even the Cochrane Collaboration reviews, one of the most respected providers of scientific meta-analyses, “may cause harm by giving credibility to biased studies of vested interests through otherwise respected systematic reviews.”

So where does this leave the average home flosser like me, who sweeps a string around the mouth at night as if participating in a rhythmic gymnastics routine? Did the dental floss companies intentionally make their studies short in order to make sure that dental floss looked good in the end? Are they just bad at designing studies, or cheap? If these studies are supposed to be shilling for floss, and they can’t even produce positive results, isn’t that evidence enough that floss is totally worthless?

Actually, the studies aren’t all by Big Floss, in support of floss. Lots of flossing data results from investigations into whether there are floss alternatives. For example, in 2002 and 2003, Pfizer, the maker of Listerine, sponsored two studies evaluating Listerine vs. floss, and found that people who used Listerine were better off after six months. (Surprise, surprise.) After considering this data, the Council on Scientific Affairs of the American Dental Association said that Listerine could use the claim “ ‘now clinically proven as effective as flossing’ for patients with mild to moderate gingivitis.” Pfizer bought big ads, including a Super Bowl commercial, with Listerine balanced on a scale across from dental floss. Then McNeil-PPC, a subsidiary of Johnson & Johnson, the company with 40 percent of the American dental floss market, sued Pfizer for false advertising.

Johnson & Johnson took out a patent for dental floss in 1898 and was one of the first companies to market the product. Over a hundred years later, J&J still thinks flossing is important. (The AP report out this week noted that the dental floss industry is $2 billion strong.) “To the extent that, you know, we are the market leaders, I believe we are heavily associated with floss,” testified Susan Sweet, a Johnson & Johnson representative.

I read through the files for the 2004 court case at the United States District Court for the Southern District of New York, and it was interesting but not particularly clarifying. The pro-floss expert witness, David Paquette, an earnest dentist and associate dental professor from the University of North Carolina, spoke unreservedly of his belief in floss and the dental floss studies that he taught his residents. In addition, he reeled off a list of detailed criticisms of the Listerine studies. His certainty about floss and his ability to stick to his lawyers’ talking points made him a dream witness. When Listerine’s lawyers cross-examined him, questioning the basis of his beliefs, Dr. Paquette simply refused to engage in debate. As I read through the transcript of his testimony, I imagined him gazing at the opposing lawyers with the same round-eyed, patient stare that my dentist gives me when he describes how to wrap floss around a tooth.

Listerine’s expert witness, Philippe Hujoel, an academic researcher from the University of Washington with a Ph.D. in epidemiology in addition to his dental degree, had written a review article about floss. He was animated about the flaws of the “classic” dental floss studies. Most of these studies hadn’t even looked directly at dental floss in adult teeth, he said, but at children or oral hygiene more generally—including tooth-brushing and professional cleanings. The American dental community’s belief in floss, Dr. Hujoel said, seemed to be mostly based on biological plausibility—an untested theory that seems possible at first glance—and not the kind of proof that should guide dental recommendations in the 21st century. When a lawyer asked him to cite other medical theories based on biological plausibility rather than fact, Dr. Hujoel mentioned bloodletting.

As the transcript ran on into the hundreds of pages, Dr. Hujoel gave the courtroom an impromptu lesson in epidemiology. But when the judge tried to pin him down about his opinion on floss (“My dentist has been telling me for 25 years each and every time I am there to floss. Now, are you suggesting that this has been bad advice for 25 years?”) he refused to say yes or no. He would only say that for one of the most important questions—whether at-home oral hygiene prevents the serious gum disease periodontitis—there had only ever been three randomized controlled trials, and “all three studies basically came up with the same conclusion, that there was no impact over all [of] oral hygiene procedures on periodontitis.” But to really answer the judge’s question, he said, we’d need more evidence. For a scientist, it seemed like the morally correct answer. In the heat of courtroom exchange, Dr. Hujoel appeared evasive.

The judge was clearly more impressed by Dr. Paquette’s testimony. (Perhaps he begrudged his many years of cajoled flossing.) In his decision, he went as far as to say that he believed “the ADA was simply wrong in approving the message that ‘clinical studies prove’ that Listerine is as effective as floss …” He ordered Pfizer to pull its ads.

I’d like to think that Dr. Hujoel is getting his revenge this week. He has been quoted widely in the media in connection with the news about the Associated Press article on floss. “It is very surprising that you have two habits, flossing and tooth-brushing without fluoride, which are widely believed to prevent cavities and tooth loss, and yet we don’t have the randomized clinical trials to show they are effective,” he said to the New York Times.

Wait what? Tooth-brushing without fluoride isn’t effective, either? My god, where does it end?

At any rate, when it comes to the flossing debacle, it seems possible that mixing the Big Floss–funded research with the Big Mouthwash simply cancels each other out.

How can we get clarity? In their book Ending Medical Reversal, doctors Vinayak Prasad and Adam Cifu say that we need to create a system for conducting meaningful, independent clinical trials. They calculate that insufficient, flawed, and biased data have become so common in modern medicine that approximately 40 percent of the therapies doctors prescribe are later proved to be ineffective. They argue that “the ultimate solution to the problem of industry influence would be to remove industry from designing clinical trials altogether.” Ideally, they say, randomized controlled trials would be designed and run by independent researchers, perhaps supported by pooled industry fees, similar to the funding structure of the U.S. Food and Drug Administration.

As for me, as soon as I read the AP article, I started planning. I got out my calculator and figured out that I spend $30 and 10 hours every year on floss. If I spent that extra money and time on reading, I could probably get through an extra book a year. That would mean 60 or 70 books in a lifetime.

But when I tried to give the good news to my husband last night as we got ready for bed, he stopped me with a sweep of his floss-wrapped fingers.

“What about the food between your teeth?” he said.