Forget poverty, poor job prospects and stress. When it comes to suffering a heart attack in later life, being small at birth is a more important risk factor.

These findings, published in the 30 November issue of the The Lancet, support a controversial (ft) theory of heart disease proposed by the epidemiologist David Barker of the University of Southampton. Surprisingly, the paper comes from researchers who have been among Barker's toughest critics.

Since the late 1980s, Barker and his Southampton colleagues have argued that undernourishment in the uterus can permanently "reset" the physiology of a fetus so that it is more vulnerable to heart disease in later life.

George Davey Smith of the University of Bristol was not convinced. "I thought small babies were just born into a disadvantaged world—they'd get less education, worse jobs, and they'd smoke more," he says. Now he has changed his mind. "When we analyzed our data, we could not get rid of the effect of low birth weight. There's clearly something important going on."

Davey Smith, Stephen Frankel and others at Bristol, with their colleagues at the Medical Research Council's Epidemiology Unit at Landough Hospital in Penarth, Wales, and at Queen's University, Belfast, spent 10 years monitoring more than 1200 men in Caerphilly, Wales, aged between 45 and 59 whose birth weights had been recorded. They found heart disease was much more common in the men with low birth weight. Conventional risk factors for heart disease—such as lower social class, blood cholesterol levels, blood pressure or even smoking—could not explain away the effect of low birth weight.

But the team did find that people who were born small and the grow fat in middle age had the strongest risk of heart disease. Resetting the body's physiology a second time by becoming obese, they argue, may increase the risk of heart attack. Small babies who stayed slim as adults did not seem to be any more susceptible than those of normal birth weight.

Barker accepts the main finding of the study, but not the idea that later obesity is what harms underweight babies. His team's own investigation of heart disease in South India found no difference in body mass between those with heart disease and those without. The group with the highest rate of heart disease had weighed 2.5 kilograms or less at birth, and the group with the lowest rate had weighed 2.1 kilograms or more at birth, irrespective of adult weight. The people whose mothers were underweight during pregnancy also had an increased risk. "Heart disease in India is not about poor, impoverished men becoming fat," he says.

Researchers have known for a while that babies who are thin at birth face an increased risk of developing diabetes and having high blood pressure—both clear risk factors for heart disease. And pregnant rats fed a low-protein diet give birth to pups with high blood pressure. But Barker suspects that these risk factors are really just secondary consequences of a more general shift in metabolism caused by fetal undernourishment.