Cow's milk not linked to early diabetes.

New parents from diabetes-prone families take note: A study has found that feeding cow's milk to your infant may not raise his or her risk of diabetes after all.

The new study casts doubt on a flurry of recent reports indicating that cow's milk increases a child's chances of suffering from diabetes later in life-but only when fed to infants less than 3 months old who have a family history of the disease.

Jill M. Norris of the University of Colorado School of Medicine in Denver and her colleagues say they were surprised to find that their findings contradicted those of earlier studies. Their report appears in the Aug. 28 Journal of the American Medical Association.

Norris and her team approached the problem in a novel way. Rather than studying children diagnosed with diabetes, they tested 253 asymptomatic children for beta-cell autoimmunity, an early predictor of the disease. The children, age 9 months to 7 years, came from 171 diabetes-prone families. By selecting one child from each family years before the disease was likely to emerge, the researchers hoped to obtain fresher recollections of infant feedings.

Beta-cell autoimmunity results when a person's normally protective immune system goes awry, unleashing renegade antibodies that attack insulin- producing beta cells in the pancreas. Over the years, this onslaught destroys beta cells. Without insulin, the body can no longer supply glucose to energy-starved cells, so diabetics become dependent on daily injections of insulin for survival.

The researchers detected beta-cell autoimmunity in 18 of the children they tested. They then queried the children's parents about what they had fed the infants from birth to age 6 months. Their responses were compared to those of parents of the remaining 153 children, who did not have the autoimmune disorder.

The two groups of children did not differ significantly in their early feedings, the study found. This suggests that cow's milk "is not associated with beta-cell autoimmunity," the researchers say, adding that the result "calls into question the importance of cow's milk avoidance as a preventive measure for insulin-dependent diabetes mellitus."

Their assertion runs counter to a recommendation put forth 2 years ago by a working group of the American Academy of Pediatrics. The pediatricians were prompted to act by a study showing that a group of Finnish children with newly diagnosed diabetes all had antibodies to a protein found in cow's milk. No such antibodies were found in nondiabetic children.

That study, however, did not establish that early exposure to cow's milk causes diabetes in infants thought to be susceptible to the disease, say Desmond A. Schatz and Noel K. Maclaren of the University of Florida College of Medicine in Gainesville in a commentary accompanying the new report. Contrary to the academy's view, they argue, the evidence is "insufficient" to "recommend changes in infant feeding practices."

Indeed, Schatz and Maclaren note, alternative diets that lack cow's milk have led to a growing number of malnourished children, a trend that last year prompted Danish pediatricians to warn mothers against withholding milk from their babies.

Only by determining what triggers childhood diabetes and how this process works can doctors offer parents sound advice, Schatz and Maclaren say.

Answering this question also offers incalculable public health benefits and cost savings, they add.

Insulin-dependent diabetes afflicts an estimated 1 of every 300 people worldwide. The disease typically erodes blood vessels, retinas, and kidneys, causing disability and even death. Moreover, since people with diabetes run up an estimated $100 billion in medical bills annually, eliminating sources of risk would save money as well as lives.