Researchers at the National Institute of Environmental Health Sciences (http://www.niehs.nih.gov) said today that fertilized human eggs that implant late are less likely to survive.

The road to birth is a perilous one. These same NIEHS scientists showed in 1988 New England Journal of Medicine, 319, pp 189-94) that 25 percent of fertilized eggs fail to survive six weeks-so early that most mothers did not know they had been pregnant. Only daily urine testing for shifting traces of a hormone called hCG confirmed these pregnancies and their loss. The 25 percent early loss, when added to the clinical miscarriages that occur later, means that at least one-third of all embryos fail.

Today, the researchers reported they have found that the lost fetuses tend to be those that implant late. (NEJM, June 10.) The implantation of the fertilized egg in the wall of the uterus is necessary for the nourishment and growth of the fetus. Surviving pregnancies implanted only about one day earlier, on average, than the non-survivors: 10.5 days v. 9.1 days from fertilization to implantation. But the day-by-day trend was clear. If a fertilized egg implanted by the ninth day, it had only a 13 percent chance of loss. The risk rose to 26 percent if the implantation was on the tenth day, 52 percent on the eleventh day, and 82 percent thereafter. All three implantations after day 12 ended in early loss. There was no association in the NIEHS study between late implantation and the clinical miscarriages that occurred later in pregnancy.

NIEHS' Allen J. Wilcox, M.D., Ph.D., Donna Baird, Ph.D., and Clarice R. Weinberg, Ph.D., said their study could not determine the mechanisms at work. They speculate that embryos that implant more slowly may be imperfect in some way. Thus, they said, "the uterus may be receptive to pregnancy only during a limited time-window, shutting out defective embryos that get there too late. This would spare a mother the physiologic burden of supporting a non-viable embryo."

The current report is based on the careful testing of daily urine samples from about 200 North Carolina women. This relatively large study in normal, fertile women confirms more tentative data from two small reports in 1991 and 1993.