Hypertension, thyroid dysfunction, diabetes, dyslipidaemia, heart failure, chronic renal impairment—conditions like these lie at the centre of generalist medicine and often present together in the same patient. We seldom pause to reflect that not one of them is a disease in its own right, but that they are all arbitrarily defined by diagnostic tests within a continuously distributed spectrum. There is, for example, no determinable level of glycaemia that does not confer a risk of “diabetic” complications, including “diabetic” retinopathy; but in clinical practice we currently accept a fasting glucose concentration of 7 mmol/l as a reasonable starting point for surveillance and treatment. The beneficial interventions for which we have varying levels of evidence are exercise, weight loss, diet, and metformin. The first 3 came together as an intensive lifestyle intervention compared with diabetes support and education alone in a large randomised study (n = 5145, mean body mass index 36, mean age 58.7 y, Arch Intern Med 2009;169:163–71), which yielded good results at 1 year—mean weight loss of >8 kg and improvements …