Letters The idolatry of the surrogate

Type 2 diabetes is a poor example of the overuse of surrogate markers

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e735 (Published 31 January 2012) Cite this as: BMJ 2012;344:e735
  1. Ian Dickson, general practitioner1
  1. 1Springwell Medical Centre, Edinburgh EH11 2JL, UK
  1. iandickson{at}blueyonder.co.uk

As a general practitioner with a special interest in type 2 diabetes, I believe that Yudkin and colleagues have chosen the wrong disease to argue the dangers of the overuse of surrogate markers.1 They do not mention the UK Prospective Diabetes Study, which had hard clinical end points and emphasised the importance of lowering blood pressure to decrease hard cardiovascular end points such as myocardial infarction and death.2 The Heart Protection Study also showed that statin treatment decreased such events.3

Twenty five years ago I regularly saw patients with type 2 diabetes die in their 50s and 60s from cardiovascular complications. I see such events much less frequently in 2012, largely because of the interventions such studies have informed clinicians to make.

Microalbuminuria is another important example of a surrogate that if untreated has serious implications for renal replacement therapy in some patients. It is readily detectable and treatable, and its complications are largely preventable.

If risk factors (surrogate markers) have been used in studies of type 2 diabetes, then many patients have benefited. The above examples focus on the prevention of macrovascular diseases in type 2 diabetes in which surrogate markers have been crucial in advancing treatment.

In recent years improved glycaemic control measured by lowering glycated haemoglobin concentration is the surrogate marker that many studies have focused on, arguably because of new drug treatments. Many such treatments increase circulating insulin concentration, which is central to the metabolic syndrome that epitomises type 2 diabetes. Physiological lowering of blood glucose concentration without hypoglycaemia through improved insulin sensitivity and lower overall insulin concentrations should be the clinician’s goal, especially in the early stages of the disease.

Maintaining β cell function and lowering insulin resistance and insulin values are the important surrogates, but they are difficult and costly to measure and interpret. In glycaemic control in type 2 diabetes we are simply using the wrong surrogates out of convenience.


Cite this as: BMJ 2012;344:e735


  • Competing interests: None declared.