Which Insulin to Use
Treating to Target in Type 2 Diabetes—the 4-T Study


In recent years, the necessity for introducing insulin as therapy in many individuals with Type 2 diabetes has been appreciated. More than 50% of these individuals will need insulin in order to achieve and maintain treatment targets. The process of initiating insulin therapy in Type 2 diabetes has, however, not been an easy one, not only due to the apprehension of patients but also to some reluctance from general practitioners, who handle most of this patient category. Thus, in many individuals with Type 2 diabetes, insulin has been introduced rather late, after an unnecessary long period with high HbA1c levels. This is in part explained by lack of experience, and therefore there is a need for studies elucidating the effect of relatively simple regimens for initiating insulin therapy.

Methods and Key Results

Holman and colleagues report the 1-year results from the 4-T (Treating to Target in Type 2 Diabetes) study, a large (708 adults with Type 2 diabetes) multicentre (58 centres) study in Ireland and the United Kingdom. The study was designed to evaluate the possible algorithms for initiating insulin therapy in patients with Type 2 diabetes who are unsatisfactorily controlled on a combination of metformin and sulfonylurea. According to the design, the oral antidiabetic therapy was continued, and the patients were randomized to 3 groups receiving insulin analogue therapy in the form of basal insulin detemir once or twice daily, prandial insulin aspart 3 times daily, or biphasic insulin aspart twice daily. A specific algorithm was developed for each type of insulin therapy in order to ensure individual dose titration based on the patients’ home blood glucose monitoring.

The outcome measures at 1 year were mean HbA1c, the proportion of patients reaching treatment target (HbA1c ≤6.5 %), the rate of hypoglycaemia, and gain in body weight.

The results reported demonstrated that the basal insulin regimen was inferior to both the biphasic insulin regimen and the prandial insulin regimen in terms of reducing HbA1c [mean, 7.6 %, 7.3 %, and 7.2 % for basal, biphasic, and prandial insulin, respectively (P<0.001 for both comparisons)]. This was also reflected in the respective proportions of patients reaching target in the 3 groups, 8.1%, 17.0%, and 23.9%, respectively (P≤0.0001 for the basal regimen compared with the biphasic and prandial regimens). However, the superiority in glycaemic control reported with the prandial and biphasic insulin regimens came at a cost of increased hypoglycaemic events, 2.3, 5.7, and 12.0 events per patient per year, respectively, in the basal, biphasic, and prandial groups. Also, weight gain was highest in the biphasic and the prandial groups.

Clinical Implications

This is an impressive multicentre trial skillfully executed in the best traditions and extending the legacy of the famous United Kingdom Prospective Diabetes Study. It is crucial to the understanding and interpretation of the study results to keep in mind that the patients were asked to continue with sulfonylurea as well as metformin.

In my opinion, the key messages of the 1-year data from the 4-T trial are the following:

  1. Insulin therapy added to oral antidiabetic therapy will significantly improve glycaemic control irrespective of the mode of insulin therapy chosen and should be initiated early, since this study also demonstrates how difficult it is to get patients down to target.
  2. An injection regimen with a long-acting insulin analogue alone is most often insufficient to bring patients to target (whether the target is chosen as being an HbA1c less than 7.0 % or 6.5 %).
  3. An insulin therapy regimen including some amount of rapid-acting insulin is superior to basal insulin alone in bringing patients to or below target. This is probably due to the fact that addressing the postprandial rise in glucose by means of prandial insulin delivery becomes more and more important as patients are approaching target (Monnier et al, 2003).
  4. The choice between biphasic and bolus insulin therapy depends on the extent to which the patients can accept a complex insulin regimen, but some patients certainly seem to need prandial insulin also at lunchtime.
  5. The use of biphasic or prandial insulin increases the frequency of hypoglycaemia. This, however, should be seen in the context of the 4-T study design, which asked for a continuation of metformin as well as sulfonylurea. It is likely that a more rational approach of stopping therapy with insulin secretagogues at the time when insulin therapy is initiated would prove more beneficial in terms of fewer hypoglycaemic events without sacrificing the impact upon HbA1c. It has recently been shown that infusion of analogue insulin into patients with Type 2 diabetes results in a “feed-back” reduction in endogenous insulin secretion when blood glucose starts to fall (Parkner et al, 2007). Hence, the use of sulfonylurea in patients where normal or near-normal glucose levels have been obtained through insulin therapy might be prone to elicit hypoglycaemic events.
  6. When insulin treatment is initiated in individuals with Type 2 diabetes, cessation of sulfonylurea therapy should be considered.


Monnier L, Lapinski H, Colette C. Contributions of fasting and postprandial plasma glucose increments to the overall diurnal hyperglycemia of type 2 diabetic patients: variations with increasing levels of HbA1c. Diabetes Care. 2003;26:881-885.

Parkner T, Møller MK, Chen JW, et al. Overnight CSII as supplement to oral antidiabetic drugs in Type 2 diabetes. Diabetes Obes Metab. 2007 June 26; [Epub ahead of print].