A study that links caffeine to birth defects in animals, and helped prompt a Food and Drug Administration warning to pregnant women, has been denounced as faulty.

The study, done in an FDA laboratory, is one of several in humans and animals that have made some scientists suspect heavy use of caffeine in coffee, tea or colas may trigger birth defects.

Based on this study and other evidence, then-FDA commissioner Jere Goyan last September warned pregnant women to stop or "minimize" caffeine consumption during their pregnancies.

The agency then formally proposed, but has not yet taken, several more steps: removing caffeine from its GRAS, or "generally regarded as safe," list, restricting caffeine use in prepared foods to present levels, and requiring disclosure of any caffeine on food labels. The FDA also proposed to change the law to allow soft-drink makers to sell "colas" and "pepper drinks" without caffeine, instead of requiring caffeine in drinks with these names as it now requires.

Goyan said at the time the evidence against caffeine was not conclusive. He said he did not think caffeine was producing an "epidemic" of birth defects, though it might be causing an unknown number of cases. But he said he did believe it "prudent" for pregnant women to reduce their caffeine intake.

The animal study that Goyan partly relied on showed that pregnant rats fed caffeine produced offspring with a high incidence of paw defects.

But a review panel headed by Dr. Robert Dixon of the government's National Institute of Environmental Health Sciences has said the pregnant rats may simply have been poisoned by the high doses of caffeine. This evidently made them lose weight, said the panel, and weight loss rather than the caffeine itself could have retarded their offspring's growth.

In short, said the panel, the rat study failed to determine whether or not caffeine is in itself a teratogen, or birth defect causing chemical.

An FDA spokesman said yesterday that FDA scientists still think the rat test results were valid. But he added that this will obviously be a point of contention as scientists, and the beverage industry, continue to argue about the need for FDA action.

For the present, he said, FDA is neither taking such action nor rescinding Goyan's caution to the "prudent."