The maker of Bendectin, for 26 years the only government-approved drug for the "morning sickness" of early pregnancy, will add a precaution to the prescribing instructions about two "preliminary and unconfirmed" 1981 studies which suggest that high dosages of the drug may cause certain birth deformities in some rats and in monkeys.

The Food and Drug Administration approved July 6, without public announcement, a revision of the labeling by Merrell Dow Pharmaceuticals, which makes Bendectin. The company is printing it and will supply it to physicians and others at an unspecified date.

Rep. Doug Walgren (D-Pa.), who researched Bendectin for more than two years, said this week that the studies merited a more forceful warning, and denounced the label revision as "misleading and totally ineffective."

One study, done in lab rats, involved diaphragmatic hernia, a life-threatening hole in the diaphragm that is often fatal, even in newborns who have successful surgery.

Given Bendectin's antihistamine ingredient, doxylamine succinate, 2 to 6 percent of the rats in the study produced fetuses with fatal diaphragm hernias. Dr. Reimer Roll, the researcher, termed this "striking," because the hernias had never been seen in untreated animals.

The labeling precaution also notes that 7 to 10 percent of the fetuses had a bone anomaly called wavy ribs, which is significant as a signal of a possible deforming agent.

The dosages in the lab rats were 125 to 375 times the maximum human dose (MHD) per unit of body weight. Scientists are uncertain whether any dosage of a particular drug will have greater or lesser effects on humans than on lab animals.

In the second study, done on monkeys, four of seven incompletely developed fetuses had a hole in the wall of the heart, called ventricular septal defect, which is relatively common in human babies but which often corrects itself.

Walgren said that the label revision submerged mention of the diaphragm hernias in the middle of the last of five paragraphs, after about 200 mainly reassuring new words on animal studies that demonstrated no birth deformities.

"I don't see how this will grab the attention of physicians who have prescribed Bendectin for as many as 400,000 women a year even though medical scientists say most women should get relief from eating soda crackers or dry toast or drinking hot or cold liquids," he said in an interview this week.

But Dr. Robert Temple, acting head of the FDA's office of new drug evaluation, defended the labeling revision, saying it "moves quite quickly to the bad news" and is in fair context, considering that the FDA is "quite uncertain as to how meaningful [the hernia study] is."

Temple also said that the FDA does not view the revision as that significant because the existing labeling already says, "Bendectin should be used only when clearly needed for the nausea and vomiting of early pregnancy not responsive to conservative non-drug measures." For the same reason, he said, the agency saw no need for a public announcement of the revision.

Merrell, "for the sake of completeness," offered June 22 to reflect the "preliminary" studies in the labeling and a patient leaflet and also to send a letter to doctors.

Several statistical surveys have been done on Bendectin's effects on babies. These, the prescribing instructions say, "do not demonstrate an association between Bendectin use and birth deformities."

As reported in June, however, recent reviews of the data in the surveys raised the possibility that babies exposed to the drug in the sensitive first trimester of pregnancy may develop the diaphragm hernia at a rate of about 40 per 10,000, and also that babies exposed either to Bendectin or to antinausea antihistamines may develop the hernia or less serious diaphragm defects. Temple termed the reviews inconclusive, but said the FDA is analyzing them.