BETHESDA, Md. (AP) _ Science, not political pressure, drove the government's decision to recommend that women in their 40s get regular mammograms, federal health officials say.

But some cancer experts said Congress was trying too hard to dictate science on this still controversial issue.

Congress ``stepped over the line, getting involved in science almost at the bedside instead of the Rotunda where they belong,'' said Ellen Stovall, a cancer survivor and member of the presidentially appointed National Cancer Advisory Board that issued the new recommendations Thursday.

``There should be no threat'' in science, she added.

The cancer board moved to end widespread confusion over when to begin breast cancer testing, declaring that for 40-somethings with an average cancer risk, ``it is prudent to have mammograms every one to two years.''

Women at higher cancer risk should consult a doctor for more specific advice, because a few may need the tests even before 40, and to determine how frequently to be tested in the 40s, the board decided.

The government and health groups long have agreed that mammograms starting at age 50 were vital. But the 40-something recommendation reverses the National Cancer Institute's 1993 decision that earlier testing was not scientifically justified.

The NCI immediately adopted Thursday's recommendations. NCI Director Richard Klausner acknowledged the pressure to do so was intense, but he insisted his decision was based on science, not politics.

``We did not waiver,'' Klausner said. Asked about reports that Sen. Arlen Specter, R-Pa., had threatened his job over the controversy, Klausner said: ``Private conversations are private conservations.''

Congress was under pressure from frightened women and the American Cancer Society to push for one national message on mammograms.

Rep. Jerrold Nadler, D-N.Y., has a bill pending that would require all private and public insurance to pay for annual mammograms starting at age 40. The Senate earlier this year unanimously passed a resolution urging mammograms for women in their 40s.

Specter, who heads a Senate subcommittee that oversees federal health spending, practically ordered the NCI to follow suit, and numerous lawmakers wrote or called the cancer board with the same request.

``We did not make this recommendation because of political pressure,'' the board's chairwoman, Dr. Barbara Rimer, insisted Thursday. ``We stepped back and looked at the science. We tried to come up with advice that would be useful to women.''

Specter, who held four congressional hearings on the subject, praised Thursday's change.

``The testimony of cancer-afflicted women at the hearings has helped lay a supplementary foundation of public and legislative opinion that women in their 40s would indeed benefit from regular mammograms,'' Specter said in a statement.

But he said he would investigate ``why it took so long to set the record straight.''

His office said the senator ``did not say that Dr. Klausner ought to resign or be replaced. But he did raise a question about whether they were using their funds properly, given overwhelming evidence supporting a clear recommendation in favor of regular mammograms.''

The NCI's 1993 decision had put the government at odds with groups like the American Cancer Society, which recommends yearly mammograms starting at 40. The outcry intensified in January when a ``scientific court'' the NCI convened to examine new research declared that women in their 40s should decide for themselves whether to be tested.

That research did show that regular mammograms in the 40s could cut breast cancer mortality by 17 percent. In contrast, mammograms starting at age 50 are proved to cut deaths by 30 percent.

The question was if the 40-something research was strong enough for a blanket recommendation that every middle-aged woman be routinely screened _ and Congress insisted it did.

Even though the cancer board voted 17-1 to recommend some regular mammography starting in the 40s, Rimer said the evidence was not overwhelming.

It was unclear whether mammograms in the early 40s or late 40s worked best, and no studies showed a clear benefit of yearly mammograms versus every other year, Rimer added.

So women and their doctors must decide whether the expense and the possibility of being frightened by benign tumors or lulled into a false sense of security is worth the possible benefit.

Breast cancer strikes about 180,000 American women each year, and is expected to kill 44,000 this year _ about 10 percent of them under age 50.