Regular exercise can reduce the long-term likelihood of diabetes, research suggests

A healthy diet and regular exercise can reduce the long-term likelihood of diabetes by a third in people at high risk of the disease, research has shown.

The benefits of lifestyle changes were greater than those of the diabetes drug metformin, which also helped prevent the disease developing.

Both effects were seen over a period of 10 years when researchers extended a previous study looking at ways to prevent type 2 diabetes.

The obesity-linked disease is the most common form of diabetes affecting more than two million people in the UK.

US researchers completed the original Diabetes Prevention Program (DPP) study in 2001. The trial involved more than 3,200 overweight or obese adults with elevated blood sugar levels who were at high risk of type 2 diabetes.

Over a period of three years, intensive lifestyle changes including exercise, reduced calorie consumption and fat intake, and frequent contact with health care professionals, cut the chances of developing the disease by 58%. Another branch of the trial found that patients assigned to two daily doses of metformin but no lifestyle changes had a 31% reduced risk of diabetes over the same period. Although the findings were striking, three years of data was not enough to indicate how long lasting the benefits were.

The new study answered this question by following the fortunes of a large majority of the DPP volunteers over a decade. The results, published online by The Lancet medical journal, showed that over 10 years lifestyle changes reduced diabetes rates by 34%. Patients treated with metformin lowered their long-term risk of the disease by 18%.

Patients undergoing the lifestyle programme received training in choosing a healthier diet and undergoing exercise. In the first year of the DPP study, volunteers lost 15 pounds on average but regained all but around five pounds over 10 years. The metformin group maintained a weight loss of around five pounds throughout the whole study period.

Between 5% and 6% of those in the lifestyle intervention group developed type 2 diabetes each year, an incidence rate that remained constant. When the DPP study ended in 2001, 8% of patients taking metformin were developing diabetes each year.

Untreated "placebo" patients who were given a dummy drug became ill with diabetes at a rate of 11% a year.