While this is an exciting unifying hypothesis to explain the relationship between insulin resistance, vessel stiffness and hypertension, there are different data which would challenge this hypothesis. Thus several studies have shown that the hypertension of obese subjects is a relatively benign condition. Barrett Connor and colleagues have studied coronary heart disease mortality in subjects with raised and with normal blood pressure according to deciles of body mass index. They have shown that thin hypertensive subjects have a substantial excess mortality, while those in the top 4 deciles of blood pressure distribution show no increased mortality risk with higher blood pressures. Thus the most marked excess risk of hypertension occurs in thin hypertensive subjects. Why might this be?