Diabetes is not the most sexy of conditions, but it's an important one. Characterised by raised levels of sugar in the bloodstream, it can ultimately lead to diverse problems including blindness, gangrene, kidney disease, nerve damage and impotence, and is the third leading cause of death after cardiovascular disease and cancer. What is more, diabetes is turning into a bit of an epidemic in the UK, with the number of sufferers set to double over the next decade. But it's not all doom and gloom. The good news is that there's plenty of evidence that making informed dietary choices offers real potential for the treatment of diabetes.

The chief substance in the body responsible for keeping blood-sugar levels in check is the hormone insulin. In diabetes, insulin simply doesn't do its job. About one in 10 diabetics has what is known as type 1 diabetes, where the body fails to make sufficient quantities of insulin. In the more common type 2 diabetes, there is usually plenty of insulin around - the problem is that the body has become resistant to its effects.

Whatever the precise nature of the diabetes, eating a diet that helps to keep blood-sugar levels on an even keel is of obvious importance. Until recently, the traditional view has been that sugar, because it causes surges in blood-sugar levels, should be limited in the diet. On the other hand, starches such as bread, potato, rice and pasta are recommended by doctors and dieticians because of the long-held belief that they give slow, sustained releases of sugar into the bloodstream.

However, a growing number of nutritionists and nutritionally oriented doctors are beginning to question the conventional wisdom behind the standard diabetic diet. Over the last two decades scientists have been busily testing carbohydrate foods (those containing sugar and/or starch) for the speed and extent that they release sugar into the bloodstream. The sugar-releasing potential of a food is represented by a number known as its glycaemic index (GI). As expected, foodstuffs packed full of refined sugar such as soft drinks, biscuits and confectionery have been found to have high GIs. What comes as more of a surprise, however, is that many of the supposedly slow-releasing starches that pervade our diet turn out, in reality, to be fast sugar-releasing. Potatoes and most refined starches such as bread, breakfast cereals, crackers, corn chips, white rice and pasta are now known to have high GIs. In contrast, foods that are truly slow sugar-releasing include kidney beans, oats, chickpeas, lentils, vegetables other than potato, wholewheat pasta, pumpernickel bread, apples, pears and citrus fruits.

Because low-GI foods give less pronounced releases of sugar into the bloodstream, common sense dictates that these foods should form the basis of the diabetic diet. Despite this logic, many doctors and dieticians have been resistant to embracing the GI concept. In fact, a recent article in the British Medical Journal announced that the influential American Diabetes Association (ADA) has dismissed the concept of the low-GI diet. Moreover, it has also recommended a relaxation of restrictions on sugary foods for diabetics. The ADA's stance is based on its belief that there is little evidence for benefit from the low-GI approach. But is this really the case?

There have been more than a dozen studies that have compared the effects of low- and high-glycaemic diets in diabetics. Most treatment periods were between two and six weeks. One wonders how realistic it is to expect significant changes so quickly in individuals who may have had impaired blood-sugar control for many years or even decades. Also, many of the studies involved relatively minor dietary changes, and study participants were often still permitted to eat plenty of destabilising foods.

Yet despite these and other shortcomings, all but one of these studies showed improvements in some measure of blood-sugar control and/or the levels of the unhealthy blood fats cholesterol or triglyceride in the bloodstream (these tend to be raised in diabetes). Study after study concluded that eating a low-glycaemic diet appears to serve diabetics well, and that the approach has merit.

So while this may fly in the face of orthodoxy, there is a good case for the role of the low-GI diet in combating diabetes. The ADA's recent dismissal of the low-GI approach and its laissez-faire attitude to sugar is not only counter-intuitive, but appears to me to go against the grain of the current scientific evidence. My belief is this latest offering from the diabetes establishment is likely only to worsen the plight of diabetics and those at risk of this condition. Smack bang next to the British Medical Journal 's report trumpeting the ADA's recommendations was another article entitled 'Diabetic care must improve'. I couldn't have put it better myself.

Nutrition news: Going nuts over nuts

Oily and intensely calorific, nuts are often viewed as fattening and unhealthy. However, the fats found in them are generally healthy varieties such as monounsaturated fat (similar to olive oil) and omega-3 fatty acids (similar to fish oil). Eating plenty of these has been consistently linked with a reduced risk of heart disease, so it comes as no surprise that eating lots of nuts appears to protect us from the number-one killer. One study found that individuals eating five or more servings of nuts a week had significantly less risk of heart attack compared to those eating less than one serving. Another study found that women consuming at least 5oz of nuts a week appeared to reduce their risk of heart disease by a third. Nuts should be eaten in their fresh, raw state. Roasting damages the fats, reducing their health-giving properties.

· If you have any issues you would like Dr John Briffa to address in this column, please contact him by email on life@observer.co.uk. Please note that Dr Briffa cannot enter into any personal correspondence.