A new controversy over the nuclear accident at Three Mile Island appears likely to arise from the discovery that an abnormal number of children were born with serious thyroid defects in three Pennsylvania counties in the latter part of last year.

The condition is known as hypothyrodism, which arises when the thyroid gland is either absent or doesn't produce normal hormone levels. It can lead to grave mental retardation and stunted growth unless it is quickly treated.

State health officials confirmed yesterday that during the last nine months of 1979, 13 hypothyroid babies were born in three counties that might ordinarily expect three such births during that length of time. They said they are about to start an epidemiological investigation that "of course" will have to consider low-level radiation from the accident at Three Mile Island -- located adjacent to one of the counties -- as one possible cause.

But they -- as well as Dr. Thomas Foley of Pittsburgh Children's Hospital, an authority on hypothyroidism -- all said that the conditions could have many possible causes.

They said they know of no cases of hypothyroidism ever caused by radiation at the low level emitted by the crippled reactor, though there is a well-established association between high doses of radioactive iodine -- one chemical emitted by the disabled reactor -- and thyroid disease. Radioactive iodine tends to concentrate in the thyroid gland, with destructive effects when the dose is high enough.

Radiation specialists from the President's Commission on Three Mile Island and the Nuclear Regulatory Commission said flatly yesterday that iodine emissions from the March accident were far too low to have had any such effect.

"There cannot be any connection: I can say that unequivocally," said Dr. Victor P. Bond, associate director of the Brookhaven National Laboratories for biomedical and environmental sciences, a member of the presidential commission task force on radiation health effects. "For thyroid effects the doses would have to have been thousands of times higher than they were."

Harold Peterson of the NRC's office of standards said a total of 15 curies of Iodine 131 was released from the plant by the end of April, giving a maximum radiation dose to the thyroids of area residents of 8 to 20 millirems.

Background radiation provides 100 millirem per year. Tests of area residents revealed no iodine in their bodies, and none was detected in area animals or in cows' milk, Bond said. To affect fetuses born since the accident would have required a pickup of iodine.

"We would certainly not expect any effect on fetal thyroids from these levels," Peterson said.

A spokesman for General Public Utilities Inc., parent company of the utility that owns Three Mile Island, said no iodine measurements taken were ever high enough to cause fetal thyroid problems.

However, several local groups have challenged the official radiation readings, alleging that insufficient monitors were in place or operating at the time of the accident. Wind currents might have carried radioactive particles over nearby monitors and deposited them in faraway areas without the normal dispersal effect, these groups have said.

None of the hypothyroid cases were in areas that have been described as in the main "plume" or downwind direction of the Three Mile Island radiation.

Ssix cases occurred is Lancaster County, which is east of Dauphin County, the reactor site. Four were in Bucks County and three in Lehigh County.

Ordinarily one baby in 5,000 is born with hypothyroidism. In 1978 (the last year for which full birth staistics were available yesterday) Lancaster County had 5,500 live births, Bucks County, 6,439, and Lehigh, 3,206.

Unusual clusters, mere statistical aberrations, sometimes occur in many diseases, said Dr. Arnold Muller secretary of health in Harrisburg.

Also, said both Dr. Foley and Dr. Evelyn Bondin, a Pennsylvania health department pediatrician, a more logical explanation than radiation has been found in three and possibly four of the Lancaster County cases, the group most closely studed so far.

One had a familial or inherited condition and two had a misplaced thyroid gland, a condition not likely to becaused by radiation, Bodin said. The three other Lancaster County cases are still under study, but one was a twin whose twin did not get the disease "so it's unlikely," though not impossible, she said, that the cause in this case was environmental, since both babies were subjected to the same environment.

Another health authority said that many populations, such as the Amish, in Pennsylvania have a high concentration of genetically related diseases.

"I don't think there's any cause and effect" connected to Three Mile Island, Bodin said. Dr. Foley agreed, but called the timing peculiar and curious," and said "the fact that it did follow the accident raises an issue" that must be settled.

The cases' existence was disclosed in an interview yesterday by Dr. Gordon MacLeod, who was Pennsylvania health secretary at the time of the nuclear accident.

MacLeod became the state's chief health officer on March 16, only 12 days before the accident. Last Oct. 10, he said -- after criticizing the state's handling of the problem -- that he was asked by Gov. Richard Thornburgh to resign. He returned to his job as a well-regarded professor of public health administration at the University of Pittsburgh.

MacLeod, too, agreed that "it is impossible" to assign any common cause yet to the thyroid defects. But he said he was shocked that the health department had made no public announcement and had not started an investigation of possible causes. The first of the affected Lancaster County babies was born last June.Two were born in July, and one each in August, October and November.

MacLeod also said it is 'urgent" to look for any possibly undetected cases in babies born at home among the Amish and other Pennsylvania who often choose home deliveries.

Thyroid problems turned up among Marshall Islanders who were exposed to radiation from the fallout of a U.S. hydrogen bomb test in the Pacific on March 1, 1954.

The first cases discovered nine years later were two children, under 5 at the time of exposure, whose thyroid glands had disappeared.

Altogether, of 21 children under 12 who lived on Rongelap Island, some 110 miles from the test site, 19 developed thyroid problems or tumors beginning 10 years after exposure.

Their dosage according to measurements by the Atomic Energy Commission in 1954, was reported at 175 REM.