Experience in our diabetes school showed that advice to
reduce fat and increase carbohydrates had a very limited
effect on long-term weight reduction in our obese diabe
tes patients (unpublished data). We therefore decided to
test a different approach in an observational study with a
control group. We were interested in seeing the effect o
the diet in compliant patients, who could be expected to
adhere to it 3–6 months. This enabled us to measure the
actual effect of a carbohydrate-restricted diet with little
contamination of non-compliant subjects. To this end, al
patients were well-informed of the diet and its rationale
before they started. We considered a weight reduction o
10% of bodyweight to be of clinical significance.