News Staff Writer Jon Cohen wrote the 2005 book Coming to Term: Uncovering the Truth About Miscarriage, so ScienceInsider asked him for his perspective on one of today’s hot stories.

Splashed across the web today were headlines connecting the flu vaccine to miscarriage. The “hint of a possible link” in a new study would likely lead to questions about the safety of the vaccine, The Washington Post wrote. Edward Belongia, the last author of the paper, is a widely respected influenza researcher at the Marshfield Clinic Research Institute in Wisconsin, who sits on the U.S. Advisory Committee on Immunization Practices. And the study’s place of publication, Vaccine, is a top-notch specialty journal.

But the small-scale study, which found a slightly higher chance of miscarriage in women who received a flu vaccine 2 years in a row that included a specific strain of the virus, is littered with “possible,” “may,” and “could.” And the researchers themselves stress the “important” limitations of the new work. Among them, in the authors’ own words:

• “This interaction effect was not an a priori hypothesis; the results were generated in a post hoc analysis with small numbers of women in the various subgroups.” The most convincing studies have a hypothesis and then follow large groups of people before and after an intervention to see whether it holds water. This study compared 485 women who miscarried with 485 who did not, and then broke them into subgroups by age.
•  The “confidence intervals were wide.” Statistical analyses include a “confidence interval” to show how much wobble there is in a finding. “Wide” means lots of wobble.
•  The “biological basis for our observations has not been established.”The researchers suggest possible mechanisms that could explain how the vaccine increased the risk of miscarriage, but they have no evidence to support the speculations.
• The “cases were significantly older than controls.” Miscarriage becomes more common in women over 35 years old, which researchers even refer to as “advanced maternal age.”
• One large previous study with the main component of the vaccine that’s linked to miscarriage found “a significantly reduced risk” of miscarriage. In that study of nearly 40,000 pregnant women—more than 9000 of whom were vaccinated before or during pregnancy—the risk of fetal loss was roughly half that of the unvaccinated group.

Helen Branswell of Stat, one of the world’s most astute reporters who covers influenza, appropriately raises the red flag high. “In fact, [the study authors] aren’t clear if the finding is actually real or the product of some unobserved statistical fluke.”

Scary miscarriage findings that did not prove true have a long history. Miscarriages are common, the causes are many and often mysterious, and confounding variables abound. One problem not mentioned by the study authors is that many miscarriages are not viable pregnancies from the start. We know this because of studies that have examined the miscarried embryos and fetuses and analyzed their DNA, or karyotype, and found that chromosomal abnormalities account for about half of all losses. Without karyotype comparisons of the cases and the control group, there’s no way to determine with confidence whether vaccination is linked to loss.

But then the study authors, to their credit, say as much: “It is important to note that this study does not and cannot confirm a causal association,” they conclude. “More research is needed regarding the immunologic effects of influenza vaccination during pregnancy.”

So a caveat to the reader. Interesting finding? Sure. Meaningful? Who knows.