The vaccine being developed with Oxford University was 90% effective when a half-dose was given before a full-dose booster, the partners said on Monday. However, that regime was administered to participants in a group whose age was capped at 55, Warp Speed’s Moncef Slaoui said Tuesday in a phone call with reporters.

Researchers have been puzzling about the AstraZeneca report since it was released, wondering why a smaller dose of the vaccine might have appeared to be more effective than a larger one. Most of the people in the trial received a placebo or the regimen of two full doses, which was 62% effective. That group included people who were older than 55, Slaoui said.

“I don’t believe that the FDA will look positively at any trial where the dose, or the age cohorts, or any other variable were changed mid-trial, inadvertently or deliberately,” said Geoffrey Porges, an SVB Leerink analyst who predicted Monday that the U.S. Food and Drug administration would not clear the vaccine.

AstraZeneca reiterated that the data will be published in a peer-reviewed journal in due course. The drugmaker’s American depositary receipts fell as much as 2.2% as of 4 p.m. in New York.

There was no age breakdown given for the two groups from AstraZeneca when the company reported its results on Monday. The initial half-dose was used in some people because of an error in the quantity of vaccine put into some vials, Slaoui said.

“There are a number of variables that we need to understand, and what has been the role of each one of them in achieving the difference in efficacy,” Slaoui said.

Older people have a weaker response than younger people to some vaccines, like flu shots. Results from a phase 2 study of the Astra-Oxford vaccine published last week in The Lancet medical journal confirmed a strong response to the vaccine in older people.