Although troubling, the findings result from experiments conducted with blood samples in a laboratory, and may not reflect the vaccine’s performance in the real world. But the conclusions add to evidence that the 13 million people inoculated with the J.&J. vaccine may need to receive a second dose — ideally of one of the mRNA vaccines made by Pfizer-BioNTech or Moderna, the authors said.

The conclusions are at odds with those from smaller studies published by Johnson & Johnson earlier this month suggesting that a single dose of the vaccine is effective against the variant even eight months after inoculation.

The new study has not yet been peer reviewed nor published in a scientific journal. But it is consistent with observations that a single dose of the AstraZeneca vaccine — which has a similar architecture to the J.&J. vaccine — shows only about 33 percent efficacy against symptomatic disease caused by the Delta variant.