ROCHESTER, Minn. — Here's another reason to get vaccinated for COVID-19: New data suggests that having previously caught the virus is no guarantee you ever developed antibodies to defend against a future infection.

The latest news that not everyone who gets COVID-19 develops antibodies is first and foremost an argument for getting vaccinated, as getting the shot offers another opportunity to teach your body to recognize the virus.

Researchers cautioned that "persons who experienced COVID-19 symptoms might be less inclined to seek vaccination, believing they are protected," when "our results caution against this assumption."

The findings also offer an argument, however, that methods of studying the vaccines could have mismeasured their effectiveness by having missed some vaccinated persons who caught breakthrough infections but developed no antibodies to show for it.

"The fact that a considerable fraction of (COVID-19) positive persons fail to seroconvert (or develop antibodies) has practical implications," the authors explained.

Newsletter signup for email alerts

"Such persons remain undetected in seroprevalence studies, including in vaccine studies that assess protection from asymptomatic infection by measuring antibodies..."

Those are some of the takeaways from a new study examining the mysterious relationship between testing positive for COVID-19 and developing antibodies for coronavirus.

Researchers from the University of Pennsylvania and elsewhere published the paper in the September issue of the journal Emerging Infectious Disease, published by the Centers for Disease Control and Prevention.

After waiting for three weeks of no symptoms following diagnosis, the researchers took blood samples from 72 patients who had been diagnosed with COVID-19 and recovered.

Upon checking for antibodies to the spike protein, they learned that 26 of the 72 who clearly had gotten COVID-19 had no such antibodies. The authors stated these results indicated 36% of those studied did not develop antibodies to COVID-19.

It adds to a broad array of findings seeking to quantify how many people do not get antibodies after getting COVID-19, with some studies showing only a few patients who fail at so-called seroconversion, while others find that percentage to be high.

In the Pennsylvania study, neither the presence of symptoms, severity of illness, race, ethnicity or gender predicted those who were less likely to develop antibodies.

The only variables predictive of not getting antibodies were being younger, and having taken a COVID-19 test requiring higher sensitivity prior to diagnosis.

"It's a little difficult to know what to make of it," cautions Mayo Clinic vaccination expert Dr. Greg Poland. "It's a small study."

Citing "studies showing nonseroconversion from anywhere 5% to 80%," Poland says "these are studies are confounded by variables we may not understand or acknowledge."

The authors countered that earlier studies of the question were confounded in ways theirs was not. They asserted their findings "provide an explanation for the puzzling variability of seroconversion in different cohorts."

Still others have taken from the findings that any policy makers who propose herd immunity through infection are on shaky ground, given that a third of the infected may never get antibodies.

"Can nonseroconversion occur?" asks Poland. "Absolutely ... it particularly makes sense if somebody had very mild or asymptomatic infection."

As for what this all means to those who already had COVID-19, the official recommendation is for two doses of vaccine 90 days after recovery from infection, Poland says, adding that "the data show that perhaps only one dose is necessary."

But people who were previously infected "have a 2.3-fold increased risk of breakthrough infection," he adds, "compared to people previously infected who got at least one dose of vaccine."

On top of all this, the latest variants change the rules of protection as well.

"The variant you were exposed to and infected with a year ago," Poland says, "is almost like a different virus than the one circulating today."