Children who as babies received two defective formulas lacking in chloride, an essential nutrient, show evidence of lasting harm to brain development, according to a new study by National Institutes of Health researchers.

The study, published Saturday in the journal Pediatrics, is the first to examine long-term health effects on infants who received defective batches of Neo-Mull-Soy and Cho-free, both manufactured by Syntex Corp.

Before they were recalled, batches of the products deficient in chloride were fed to an estimated 20,000 children in 1978 and 1979. The incident led to the passage of a law in 1980 imposing tougher standards on baby-formula manufacturers.

In a small group of such children tested at 2 and 4 years of age, NIH researchers found what scientists call a dose-response relationship: The longer the children had received the chloride-deficient formulas as their only food, the lower they scored on certain measures of intellectual development.

"It is startling. We've been saying this all along," said Lynne Pilot, mother of one of the children and a cofounder of Formula, a consumer group. "It is vital that parents and educators become aware that these children are, in fact, at risk."

The scientists tested 20 children chosen randomly from a national registry of those who received one of the formulas as infants and who suffered a documented metabolic disturbance from the lack of chloride.

When the children were tested at age 2, exclusive use of one of the deficient formulas correlated strongly with below-average scores on overall mental development in a standard test. The longer the children had remained on the formula, the lower their scores.

When they were retested at age 4, use of the formulas correlated with below-average scores on tests of perception, overall muscle coordination and control of fine movements.

In the report, the scientists cautioned that the number of children tested was small and further studies should be done, but added that the findings "raise considerable concern about the outcome of children" exposed to the deficient formulas.

Syntex Corp., in a statement, said the NIH study was "not a well-controlled, definitive study" and that the numbers were too small to suggest broad conclusions. The corporation is involved in numerous lawsuits over the fomula, and a Justice Department investigation of the manufacture of the formula was reopened last year.

The findings were supported, however, by a second NIH pilot study that compared a group of Florida schoolchildren who received the deficient formulas as infants with a similar group who had not. Those fed the formulas scored somewhat lower on tests of intellectual functioning.

NIH researchers plan to conduct a larger study comparing children who received the formulas with others who did not, and perform more extensive tests on a group who suffered metabolic abnormalities while on the deficient formulas.

Pilot said Formula has conducted its own survey of about 800 parents whose children received the formulas. Of the parents who responded, 25 percent said that their children had been diagnosed as learning-disabled, 35 percent said their children were in special education classes, and 31 percent said their children had received speech or language therapy.