Yeast infections, which plague millions of American women, have been invoked as the cause of a complex of disorders popularly known as the "yeast connection," including depression, premenstrual syndrome and chronic fatigue. But a report published today in The New England Journal of Medicine casts doubt on the existence of the syndrome and on a drug regimen often prescribed for its symptoms.

As expected, nystatin, an antifungal drug frequently used to treat both vaginal yeast infections and the "yeast connection" syndrome, was clearly successful in relieving the infection among the study's subjects. But the drug proved to have no more significant effect on the other aspects of the supposed syndrome than a placebo.

The findings are expected to revive a long-standing debate in professional circles over the syndrome, which has little credibility in the mainstream medical community because it has no fixed definition and because research on its presumed causes and characteristics has not been published in the more prestigious medical journals.

Yet a popular book by Tennessee allergist William G. Crook, titled "The Yeast Connection," has sold three-quarters of a million copies, led to the creation of numerous support groups across the nation and prompted attention in the popular press.

The journal study, by a team of researchers at the University of Alabama Medical Center, examined 42 premenopausal women during 32 weeks of treatment. Each had a history of vaginal infections from a common gastrointestinal yeast called Candida albicans and also met three out of five diagnostic criteria for "candida hypersensitivity syndrome," which included symptoms of stomach or breathing problems, premenstrual distress, depression or memory disorders.

The subjects were divided into four groups, which were variously treated with four possible combinations of oral and/or vaginal nystatin and an oral or vaginal placebo. Women who received oral or vaginal nystatin, or both, showed substantial improvements in vaginal symptoms compared to the placebo-only group. But, the researchers report, "the regimens containing nystatin did not significantly reduce systemic symptoms more than the all-placebo regimen did." These negative results, they conclude, "provide additional evidence that the chronic candidiasis {yeast} syndrome is not a verifiable condition."

The researchers note, however, that they made no attempt to regulate or monitor the diet of their subjects. Proponents of the "yeast connection" theory, who rarely prescribe nystatin therapy alone, frequently advise their patients to avoid moldy environments, cut back on foods high in carbohydrates and molds and receive anti-allergy injections for the Candida yeast.

As a result, the controversy surrounding the syndrome, which the American Academy of Allergy and Immunology has deemed "speculative and unproven," is likely to persist.

"Few illnesses have sparked as much hostility between the medical community and a segment of the lay public as the chronic candidiasis syndrome," John Bennett of the National Institute of Allergy and Infectious Diseases commented in an editorial accompanying the Alabama group's article. The rancor, he wrote, is apt to continue until "additional scientifically sound studies" can determine "whether this syndrome does or does not exist, and if it does, what the optimal treatment is for patients."