Some copies of this issue of the Journal of Diabetes are being distributed in China with a reprint describing the findings of a consensus conference organized by NovoNordisk on how best to use the combination of 30% insulin aspart (rapid-acting) and 70% protamine-based (intermediate-acting) insulin (BIAsp 30).1 The article offers guidance in approaches to intensifying treatment with this form of insulin, but does not compare BIAsp 30 with other insulin treatment regimens, particularly for people with Type 2 diabetes, a most interesting and important topic. It is not widely appreciated that the UK Prospective Diabetes Study (UKPDS) presented evidence that Type 2 diabetic people receiving either of the sulfonylureas (SU) used for initial treatment had progressively increasing need for addition of insulin over time, with nearly 10% needing this treatment after 1 year, and a linear increase to approximately half requiring this by 6 years.2 Insulin alone was inferior to combined treatment with insulin + SU in maintaining glycemic control.2 Although one may argue that in the current era of the availability of multiple oral agents the use of insulin is less necessary, comparison of maximal dose metformin and pioglitazone plus titrated SU treatment versus metformin plus BIAsp 30 showed that although either approach could successfully maintain excellent glycemic control, the former led to both a greater increase in weight and an increasing frequency of hypoglycemia.3 Therefore, insulin plus oral agents may be a preferable approach to thiazolidinedione/SU/metformin combinations for many Type 2 diabetic people. Multiple approaches have been shown to be effective with insulin treatment. The rapid-acting insulin analogs lispro, aspart, and glulisine lead to more physiologic replication of the normal postprandial insulin secretory pattern,4 with a reduction in postprandial glycemia.5-7 The use of such analogs in combination with a protamine-containing intermediate-acting insulins led to the development of BIAsp 30 and to the similar Eli Lilly product, namely protamine-lispro 75%/lispro 25% (BI-lispro 25), both of which have the advantage over both intermediate-acting insulin alone and combinations of regular human insulin plus intermediate-acting insulin of better reducing postprandial glycemia8, 9 while maintaining benefit in overnight glucose-lowering.10 Furthermore, these preparations have greater simplicity of administration, with the potential for a reduced number of injections. However, one problem with all such protamine-containing insulin preparations is their pronounced peak effect, often occurring at an undesirable time after evening administration,11 so that although effective in improving glycemia, many studies suggest that they do so at the expense of a greater likelihood of nocturnal hypoglycemia than seen with the ‘peakless’ insulin analogs detemir and glargine.12 (It should be noted that not all authors have reported this finding.13) Furthermore, intermediate-acting insulin given alone may not be quite as effective as insulin glargine in lowering A1c.14 Detemir appears to have less variability in overnight glucose-lowering action than seen either with protamine–intermediate-acting insulin or with insulin glargine,15 with studies comparing it with intermediate-acting insulin showing lesser degrees of weight gain.16 The question should then be asked, which is better: basal insulins, such as protamine-containing preparations, glargine, or detemir; bolus insulins, particularly the rapid-acting analogs lispro, aspart, and glulisine; or the combination insulin preparations BIAsp 30 and BI-lispro 25? This question was addressed by the Treating to Target in Type 2 Diabetes (4-T) study.17 A total of 708 Type 2 diabetic people were randomized to BIAsp 30, to aspart three times daily before meals, or to insulin detemir once or twice daily. At 1 year, the detemir group had a lesser improvement in A1c but less weight gain, less hypoglycemia, and a lower total daily insulin dose. After 3 years, when all patients were moved to a multiple daily insulin (MDI) approach, hypoglycemia rates remained significantly lower in those started on insulin detemir and, in addition, these patients continued to exhibit less weight gain even though the glycemic improvement was now similar across all groups. Furthermore, cardiovascular mortality was significantly lower in the group started on basal insulin initially and significantly more serious adverse events occurred in those people initially randomized to BIAsp 30.18 This context should be used in interpreting the consensus statement of Unnikrishnan et al.1 regarding intensification of insulin therapy with BIAsp 30. The authors appropriately suggest that this preparation, dosed up to three-times daily, may allow improved glycemic control over basal insulin preparations and present evidence supporting this concept. Although requiring considerable effort on the parts of both the clinician and the diabetic patient, initial use of basal insulin titrated to complex basal-bolus MDI regimens may be preferable to multiple doses of either BIAsp 30 or BI-Lispro 25, so that one should strongly consider MDI as the optimal treatment for management of insulin-requiring patients with Type 2 or Type 1 diabetes.