Iraq toll 655,000: report
However the figure has been rejected by both the US and Iraqi governments, who have dismissed the report as “not credible”.
The study in the British medical journal The Lancet estimated that one Iraqi in 40 had died as a result of the conflict by comparing the death rates from the period before the war to the period from March 2003 to June 2006.
According to the study just over 600 thousand of the deaths are due to violence, and about half of them are due to gunfire.
The study says the proportion of deaths ascribed to coalition forces has diminished in 2006, although the actual numbers have increased every year.
US President George W Bush told a White House media conference that he and his top military advisers believe “the methodology is pretty well discredited” in the study.
Mr Bush in the past has estimated the number of Iraqi deaths to be closer to 30,000, and reaffirmed that number today.
“I stand by the figure,” he said. “Six hundred thousand or whatever they guessed at … it’s not credible.”
“I do know that a lot of innocent people have died, and it troubles me and grieves me.”
The Iraqi government also dismissed the figure.
“This figure, which in reality has no basis, is exaggerated,” said Iraqi government spokesman Ali Debbagh.
“It is a figure which flies in the face of the most obvious truths,” he said, calling on research institutions to adopt precise and transparent criteria especially when the research concerns victim tolls.
The research by the Johns Hopkins University in Baltimore, Maryland will be published tomorrow.
In October 2004, a paper also published in The Lancet calculated that almost 100,000 deaths had occurred in Iraq between March 2003 and September 2004 as a result of violence and heart attack and aggravated health problems.
In updating the paper, a team led by Gilbert Burnham of Johns Hopkins Bloomberg School of Public Health sought to make an estimate of deaths in the post-invasion period.
They randomly selected 47 sites across Iraq, comprising 1,849 households and 12,801 people.
Interviewers asked householders about births, deaths and migration and if there had been a death since January 2002 and, if so, asked to see a death certificate to note the cause.
Of the 629 deaths recorded, 547, or 87 per cent, were in the post-invasion period.
Extrapolated across the country, 654,965 premature deaths — 2.5 percent of the population — have occurred since March 2003, the study says.
However the study acknowledges weaknesses in its data collection, saying that the “extreme insecurity” during the survey limited the number of teams that ventured out to interview families and the time they could spend interviewing.
But there was also the possibility that some deaths may have gone unrecorded, it says.