Questionable study casts a cloud over the 75th anniversary of water fluoridation
Seventy-five years ago, on Jan. 25, 1945, public health officials began an experiment in Grand Rapids, Mich., to prevent the pain, misery, and cost of tooth decay: adjusting the level of fluoride in drinking water. Cavity rates plunged with fluoridation.
Today, 211 million Americans have access to fluoridated water. The Centers for Disease Control and Prevention hailed this strategy as one of 10 great public health achievements of the 20th century. Yet as we celebrate this milestone, community water fluoridation faces a renewed threat from a recent study.
Fluoride is a mineral found naturally in lakes, rivers, and other water sources, but the typical level in the U.S. is usually too low to prevent cavities. So numerous public water systems in cities and towns across the country add fluoride to achieve the optimal level of 0.7 milligrams per liter.
advertisement
Despite its proven success, critics often attack fluoridation by citing “medical freedom” and sharing conspiracy theories which unjustly blame all manner of ills — from acne to migraines — on fluoride.
Opponents of fluoridation gained fresh fodder for their criticism from a Canadian study, published last year in JAMA Pediatrics. It reported that the more fluoride children were exposed to in utero, the lower their IQ scores at ages 3 and 4.
advertisement
This study suffered from a host of weaknesses. Differences between the mothers other than fluoride, like parental IQ, socioeconomic factors, and the children’s early environments, could have influenced the results. The measures the researchers used to estimate maternal fluoride exposure have also been criticized by multiple experts.
The Canadian Agency for Drugs and Technologies in Health (CADTH), an independent organization that advises health care decision-makers in Canada, reviewed the JAMA Pediatrics study and called its evidence weak, pointing to potential errors, biases, and uncontrolled confounding, including how the researchers estimated maternal fluoride exposure. According to CADTH, the authors’ conclusion that maternal exposure to higher levels of fluoride during pregnancy was linked to lower IQ scores in children “was not supported by the data.”
In an interview with The Washington Post, John Ioannidis, a professor of medicine at Stanford University and an expert in statistical methods, called the study’s results “very borderline in terms of statistical significance” and said the gap between boys’ and girls’ IQ scores should raise eyebrows. “If you see a gender difference claim for this type of association, it’s far more likely to be a spurious finding rather than something true,” Ioannidis said.
Two months after the study was published, 30 international researchers and faculty members signed a letter raising questions about the study’s quality, including lack of data on multiple factors known to be key predictors of IQ scores. Their letter expressed the need to conduct an independent analysis of the study to determine if its findings could be validated.
Although the study relied on a Canadian birth cohort, it was funded by the U.S. National Institute of Environmental Health Sciences — a fact that it makes it entirely reasonable that the researchers should comply with NIH’s data-sharing requirements. The authors of the study eventually released some details of how they conducted their analyses, but not enough to determine how they reached their conclusions.
Despite its many weaknesses, this study is likely to confuse or scare parents who have heard about its result — many of whom may be unaware of the concerns about the research voiced by the scientific community. Sadly, even flawed research papers can have negative long-term impacts, especially when they are published by reputable journals. Take, for example, the now-infamous 1998 article by Andrew Wakefield in The Lancet that erroneously associated vaccinations with autism. Over the next 12 years, vaccination rates plummeted.
There are real consequences when unwarranted fear leads a community to end water fluoridation. In Alaska, where I live and work, the capital city of Juneau halted water fluoridation in January 2007, citing a presumed lack of evidence for its safety and effectiveness. Oral health worsened significantly after water fluoridation ended: Six years later, the average cavity-related treatment costs for Medicaid-enrolled children in Juneau increased by an inflation-adjusted 47%. For children born after fluoridation had been stopped, the increase was 111%.
Science is a discipline of objectivity, skepticism, and replication. Large studies and national research reviews from New Zealand, Sweden, Australia, and Canada have shown no association between fluoride at common levels in drinking water and IQ. While science welcomes studying old questions in new ways, when new results are at odds with previous research they should be interpreted with caution. Replication becomes even more important.
An independent re-analysis is needed of the Canadian cohort that was used for the article published in JAMA Pediatrics. Much is at stake for both public health and scientific integrity.
Jennifer Meyer, Ph.D., is an assistant professor of health sciences in the Division of Population Health Sciences at the University of Alaska Anchorage.
