A study by government researchers provides strong new evidence that the virus thought to cause AIDS can be spread sexually between husbands and wives.

But the research also provides reassurance that the deadly disease does not seem to be spread by "close household contact" between parents and their children.

"It is . . . evident that heterosexual activity has the potential of playing a significant role in human-to-human transmission" of the virus for acquired immune deficiency syndrome, reports Dr. Robert R. Redfield of the Walter Reed Army Institute of Research in today's issue of the Journal of the American Medical Association.

Redfield and his colleagues looked at seven active-duty military men with AIDS or an illness that sometimes precedes it and found that five of their spouses were infected.

But, despite family contact, only one of the couples' 11 children -- a 14-month child -- showed signs of exposure to the virus, known as Human T-Cell Leukemia Virus III, or HTLV-III.

"Clearly they kissed their children, shared glasses and coughed on them," Redfield said in an interview, but this did not cause infection among the other children, who ranged in age from 8 to 14 years.

In the case of the infant, it is "probable" that transmission occurred from mother to child during pregnancy or shortly thereafter, the researchers said.

The study provides "more substantial evidence" that heterosexual transmission occurs but that "the risk to other members of the household is very small," said Dr. Harold Jaffe, one of the leaders of the AIDS effort at the Centers for Disease Control.

As of March 8, the CDC reported, 8,797 adults and children in this country had been infected with AIDS, and 4,230 had died. About three-fourths of the cases were among sexually active homosexual men, with the rest largely among intravenous drug abusers, hemophiliacs and recent Haitian immigrants.

The disease, which destroys the immune system, is thought to be spread through contact with body fluids. Evidence of heterosexual transmission first appeared among sexual contacts of high-risk groups, such as intravenous drug abusers. Only recently has evidence begun to grow that the disease may be spread sexually into the general population.

Jaffe said the CDC has documented 70 instances of AIDS thought to be acquired through heterosexual contact -- eight men and 62 women -- but mostly through contacts with drug addicts. Recent data from Africa suggests that AIDS there is transmitted primarily through heterosexual contact.

In the Army study, done in cooperation with scientists from the National Cancer Institute and two commercial laboratories, the seven married men studied included three who reported intravenous drug abuse and three who reported "heterosexual promiscuity," including contact with prostitutes. None of the men reported homosexual contact.

Redfield said unpublished data indicates that about 40 percent of 38 cases of AIDS and AIDS-related illness among U.S. military personnel might be related to heterosexual promiscuity, with either prostitutes or bisexual men.

In the Walter Reed study, four of the men had diagnosed AIDS cases and three were affected with the less severe condition involving lymphadenopathy, enlarged lymph nodes and fatigue as well as white blood cell depletion. Of the five wives with signs of virus infection, none had AIDS but three had symptoms of lymphadenopathy and two appeared healthy. There is no way to tell if the infected wives will contract the disease.

There is no effective treatment for AIDS, but a study in the British journal Lancet says that in the laboratory the antiviral drug interferon inhibited AIDS virus growth, suggesting that it might be useful in recently infected humans.