The worries that tug at a pregnant woman’s mind tend to shift in time. As weeks and months go by, new concerns invariably crop up.

But for many women, the end of the first trimester brings with it a significant degree of relief: That’s the point at which the risk of miscarriage drops dramatically. It’s also a key milestone following a critical phase in fetal development.

On top of that, some diseases, like rubella, which can be catastrophic to fetuses whose mothers fall ill in the first trimester, are less serious if a woman is infected later in pregnancy. As a result, conventional wisdom focuses on the first trimester as the most vulnerable time for a fetus.

But that doesn’t seem to be the case with the Zika virus. A new study, published Wednesday in The Lancet, finds evidence that the virus can still cause devastating brain defects among newborns whose mothers were infected in their last trimester.

The finding is alarming for several reasons. For one, it underscores the risk of Zika to pregnant women at any stage. This builds on earlier research, published in the New England Journal of Medicine in March, that found Zika can cause “grave outcomes” for fetuses well past the 30-week mark.

The latest study also suggests that screening for microcephaly, a condition in which babies are born with abnormally small heads, isn’t enough to detect Zika-related birth defects. In other words, an infant whose head circumference appears normal might still have serious health problems from the virus. “Such children would be born with normal sized heads as cranial growth largely takes place up to 30 weeks, but yet present important brain damage,” the paper’s authors wrote. “In view of the huge interest in the epidemic, we believe that under-reporting of microcephaly cases is rare, but newborn babies affected late in pregnancy might fail to be reported as their heads will be in the normal range.”

The authors of the report suggest that hospitals should consider expanding screening standards to include testing all newborn babies in epidemic regions for the virus—rather than focusing primarily on microcephaly as an indicator. Looking at a group of 602 babies with probably or definite cases of congenital Zika virus syndrome, for instance, the researchers found one in five of them had head sizes in the normal range.

“There is no question,” they wrote, that “focusing on microcephaly alone will underestimate the true magnitude of this major epidemic.”