A federally funded study published yesterday indicates that women who swallowed the original Bendectin tablets for "morning sickness" in early pregancy may have quadrupled the risk of a birth defect of the stomach that restricts an infant's ability to eat and can cause severe dehydration and malnutrition.

Brenda Eskenazi and Michael B. Bracken, Yale University epidemiologists, said their data "suggest that maternal Bendectin use is strongly associated" with the occurrence of the malformation, known as pyloric stenosis.

Over the quarter-century in which Bendectin has been the only medicine approved by the Food and Drug Administration for nausea during the first trimester of pregnancy, a series of conflicting, inconclusive studies either exonerated the drug or implicated it as a possible cause of a variety of congenital defects.

The Yale study is the first to point out a possible link between Bendectin and pyloric stenosis, which occurs in an estimated 3 per 1,000 live births.

Eskenazi and Bracken derived the new data from interviews of 1,427 mothers of malformed newborns and stillborns, most delivered at five central Connecticut hospitals in a two-year period ended in mid-November, 1976. The "controls" were 3,001 interviewed mothers of healthy newborns.

During that period, Bendectin contained a vitamin and an anti-spasm compound in addition to an antihistamine. In 1976, however, Merrell Dow Pharmaceuticals eliminated the anti-spasm ingredient to overcome the "possibly" effective rating given the three-part mixture by the National Academy of Sciences.

The company continues to sell the three-part mixture abroad under the trade names Debendox, Lenotan and Meribental. Last year, U.S. pharmacists dispensed 1.9 million new and refill prescriptions of the two-ingredient variety.

The original drug may account for more than one in 10 cases of the defect, Eskenazi and Bracken wrote in the American Journal of Obstetrics and Gynecology. They said it "is unclear" whether or not a direct causal relationship exists. They also said that cigarette smoking by Bendectin users may increase the risk of pyloric stenosis from four times to five times normal.

The researchers said relatively "short and safe" surgery can remedy pyloric stenosis, but they urged physicians to "carefully weigh the hazards of maternal nausea and vomiting during pregnancy against the risk of a malformation that can be treated." Left untreated, the defect can be fatal in four to six weeks.

They said they also found "a possible association" of Bendectin with a three-fold increase in the risk of defective heart valves. As to numerous other congential deformities, however, Eskenazi and Bracken said that, "We were unable to document . . . significantly increased risks." They noted that in some diagnostic categories the numbers of users and of nonuser controls were small.

The Yale scientists isolated 122 "cases" -- mothers who took no drug but Bendectin in the first trimester -- and matched them with 2,293 controls who took no drug at all in their entire pregnancies. They found pyloric stenosis in the infants of six of the 122 cases, or a rate 4.33 times higher than in the 2,293 controls, and heart-valve defects in the babies of four of the cases, triple the rate in the controls.

The study drew contrasting responses from two long-time Bendectin critics -- Rep. Doug Walgren (D-Pa.), who has researched the drug for 2 1/2 years, and the Public Citizen Health Research Group, which has petitioned the government to halt sales of the medicine on the ground that the evidence of effectiveness as well as safety is legally insufficient.

After reading an advance manuscript of the Yale study, Walgren termed it "alarming," and said it is the fourth study to associate Bendectin with gastrointestinal congenital defects.

He also said he thinks it "confirms the suspicions" raised by two 1981 animal studies, which are termed "preliminary and confirmed" in the company's new prescribing instructions. One of those studies shows a possible link between the drug's antihistamine ingredient and a potentially fatal hernia in rat fetuses; the other found a hole in the heart wall of four of seven fetuses of monkeys given the current version of the drug.

In a Dec. 1 letter to Surgeon General C. Everett Koop, Walgren wrote that the Yale study "contradicts the new Bendectin labeling, which states that 'a review of the results of epidemiological studies leads to the conclusion that the existing data do not demonstrate an association of Bendectin use with a specific congenital defect.' "

Sidney M. Wolfe, director of the Health Research Group, said only that the Yale data "add further to our concerns and demand an immediate follow-up to clarify this important hypothetis."

In Cincinnati, a spokesman for Merrell said that the stomach and heart-valve defects had not "popped up" in 18 previous epidemiological studies and are "probably a chance association."

The FDA, which said it is reviewing the Yale study, gave final approval to the new labeling Sept. 7, but did not approve a "Dear Doctor" letter until last Thursday. The letter and the labeling are to be mailed next Monday.