Type 2 diabetes is a major and rapidly increasing health problem worldwide. Keeping the blood glucose (sugar) from going too high helps prevent complications. Recently a number of new treatments (collectively called 'incretin based' treatments) to lower blood glucose have become available but response is very variable and it is difficult to predict which will work for an individual. The investigators want to see if we can identify whether the new treatments are likely to be effective for an individual patient. Identifying the right treatment would improve control and minimise the side-effects and costs from ineffective treatments. We will collect blood (for measures of blood glucose, insulin secretion and genetics information), urine (for a simple measurement of insulin secretion) and other clinical information (such as weight,age, duration of diabetes and medication) in people who are about to start these new 'incretin based' treatments and assess their response over the first 6 months of treatment. We will analyse this information to see if we can predict treatment response.

Study Hypothesis:

The investigators hypothesise that those who have low insulin secretion, as measured by post meal urine C-peptide Creatinine Ratio or blood C-peptide, will have poor blood glucose response to incretin based treatments.

Inclusion Criteria:

• A clinical diagnosis of type 2 diabetes mellitus where the patient's clinician has determined the need for a DPP-IV inhibitor or GLP-1 analogue as a result of inadequate glycaemic control
• HbA1c >= 58mmol/mol

Exclusion Criteria:

• Treatment with DPP-IV inhibitors or GLP-1 analogues prior to study initiation (within the previous 3 months)
• Renal failure as shown by a eGFR (estimated glomerular filtration rate) less than 30 mL/min/1.73m2

Patients with type 2 Diabetes commencing DPP-IV inhibitors or GLP-1 agonsists in primary or secondary care in England