In the emotionally charged debate over abortion, no one could accuse Professor David Fergusson of ideological bias.

He is "pro-choice" personally, but he admits his latest research - which suggests a strong link between abortion and mental illness - is liable to be used and misused as ammunition by the pro-life brigade.

Researchers found that at age 25, 42 per cent of women in the study group who had had an abortion also experienced major depression at some stage during the past four years.

This was nearly double the rate of those who had never been pregnant and 35 per cent higher than those who had chosen to continue a pregnancy.

"Those having an abortion had elevated rates of subsequent mental health problems, including depression, anxiety, suicidal behaviours and substance use disorders," said the researchers, whose study has been published in the Journal of Child Psychiatry and Psychology.

The study - the largest of its kind internationally - is the latest analysis to come out of the Canterbury Health and Development Study, which has followed the progress of 1265 children born in Christchurch from mid-1977.

Professor Fergusson, who leads the project, said the paper was declined by a number of journals, he suspects because of the "controversial" nature of the topic.

"We went to four journals, which is very unusual for us - we normally get accepted the first time."

He knew that by publishing the paper, he and the group would be throwing themselves into a vicious political arena, and their science would be claimed as "proof" of a certain ideology by one side, and damned by the other.

The draft paper was sent to a number of organisations for comment and analysis.

The Abortion Supervisory Committee, which supervises all legal and compliance aspects of terminations, concluded that it would be undesirable to publish the results in their "unclarified" state as they would be used as "a political football".

In a letter to the committee in June 2004, Professor Fergusson said he was fully aware of "the circus" that publication would provoke.

But it would be "scientifically irresponsible" not to publish the results just because they were controversial, he wrote in the letter, released under the Official Information Act.

"To provide a parallel to this situation, if we were to find evidence of an adverse reaction to medication, we would be obliged ethically to publish that fact," he told NZPA.

But he had no regrets about entering the debate because he believed that women's health was at stake.

"The fact is that abortions are the most common medical procedure that young women face - by the age of 25, one in seven have had an abortion - and the research into the costs and benefits have been very weak.

"This is because the debate between the pro-life and pro-choice has, in a sense, driven the science out.

"It verges on scandalous that a surgical procedure that is performed on over one in 10 women has been so poorly researched and evaluated, given the debates about the psychological consequences of abortion."

In his letter to the committee, Professor Fergusson goes even further, saying his reading of the literature on abortion suggested it was "one of the most methodologically flawed and illiterate" research areas he had ever encountered.

The scientists working in the area mainly came from a "pro-life" perspective, and they were often accused of making conclusions based on "weak evidence and strong faith".

But "no one can accuse me of that", he said this week with a laugh.

"I'm pro-choice but I've produced results which, if anything, favour a pro-life viewpoint.

"It's obvious I'm not acting out of any agenda except to do reasonable science about a difficult problem."

The study has already been criticised for "not being completely conclusive", which he finds irritating.

"Because it's not 'completely conclusive', then they say we know nothing.

"But no science is completely conclusive - it's cumulative.

"Our study is strongly suggestive of a link between abortion and developing mental illness.

"What people should be saying is, 'This is interesting ... we need to invest more to answer this important question'."

The researchers expected to find no evidence of harmful effects of abortion. But they found the opposite.

Professor Fergusson said the results could not be argued away by assuming the women involved already had mental health problems or "a pre-disposition".

"We know what people were like before they became pregnant. We take into account their social background, education, ethnicity, previous mental health, exposure to sexual abuse, and a whole mass of factors."

The study participants will be interviewed again next year, at age 30, and researchers plan to ask them more detailed questions about abortions.

"I want to ask whether they think it's had an effect on their mental health, and if so, how," said Professor Fergusson. "We also want to know more about the context of abortions and unwanted pregnancy, and add more to the picture."

The implications for the whole legal structure for abortion are huge.

Under the 1977 Contraception, Sterilisation and Abortion Act every abortion requires two "certifying consultants" to approve it on certain grounds, usually that a woman's mental health would be endangered by continuing with the pregnancy.

Of the 18,211 abortions performed in New Zealand in 2004, about 98 per cent were carried out for this reason.

Professor Fergusson said the idea behind the law that abortion was a mental health issue was "based on a conjecture". No one had examined the actual costs and benefits.

"If the legislation was based on health grounds, you would naturally think this would lead to monitoring of people who had had abortions."

But, he said, the health aspect was always secondary to personal choice.