Effects of Duration of Type 2 Diabetes Mellitus on Insulin Secretion

Farhad Zangeneh, MD; Puneet S. Arora, MD; Peter J. Dyck, MD; Lynn Bekris; Ake Lernmark, PhD; Sara J. Achenbach; Ann L. Oberg, PhD; Robert A. Rizza, MD


Endocr Pract. 2006;12(4):388-393. 

In This Article


Several studies have reported that insulin secretion decreases during the first 5 to 10 years after the diagnosis of type 2 diabetes.[4,5,6,7,8,9] The current study extends those reports by demonstrating that, although insulin secretion decreases over time in some patients with type 2 diabetes, it persists in others who have had diabetes for in excess of 40 years. In contrast to the Belfast Diet Study[4] and the UKPDS,[5] however, glycemic control, as reflected by the A1C concentration, improved over time in the current study. Multiple factors likely contributed to this trend, including a change in practice after the Diabetes Control and Complications Trial reported that improved glycemic control is associated with a lower rate of diabetes-related complications.[14] This result presumably was due to the fact that the patients' therapeutic programs were progressively modified, as needed, to meet glycemic goals rather than waiting until the glucose concentration exceeded a prespecified threshold, as was done in the UKPDS[5] and the Belfast Diet Study.[4]

Although insulin secretion declined over time in the study group as a whole, the rate of decline varied both between individuals and within the same individual. Even though insulin secretion, whether measured as the fasting C-peptide, 6-minute C-peptide, or 6-minute postglucagon increment in C-peptide concentrations, decreased over time in approximately half of the study subjects, it either remained essentially unchanged or increased in the other half. This observation is consistent with the previous report by Borg et al,[7] who noted that when patients with evidence of autoimmune diabetes are carefully excluded from analysis, type 2 diabetes is not invariably associated with a decrease in insulin secretion. Age, sex, and duration of diabetes did not differ in patients in whom insulin secretion did or did not decrease. Of interest, the higher the initial C-peptide concentration, the greater the subsequent decline—perhaps suggesting the greater the insulin need, the more rapid the decrease in insulin secretion. Alternatively, this finding merely could be ascribed to regression toward the mean.

The observation that C-peptide concentrations considerably decreased in some patients from one visit to another, suggesting a substantial decrease in beta cell function only to return to baseline if not higher levels at the next visit, is particularly intriguing. This pattern was observed for the fasting, 6-minute, and postglucagon increment in C-peptide concentrations. The last finding argues against assay drift as an explanation, inasmuch as assay drift presumably would alter the absolute value but not the increment. Nevertheless, we believe that assay drift did occur during the years 1993 and 1994 because virtually all Cpeptide concentrations measured in those years were lower than those measured at the previous visit. If those data had not been excluded from analyses, the apparent variation in insulin secretion over time would have been even greater. Because insulin secretion is modulated by the prevailing glucose concentration, we speculated that the variability in plasma glucose levels at the time of testing might account for the variability in fasting C-peptide concentrations. Linear regression analyses, however, indicated a weak, albeit significant, relationship between the plasma glucose level and the fasting, 6-minute, or postglucagon increment in C-peptide concentration—perhaps attributable to the fact that antidiabetic therapy commonly was changed over time. The minimal influence of the prevailing plasma glucose concentration on the 6-minute postglucagon C-peptide increment also is consistent with the studies that originally evaluated intravenous glucagon injection as a means of assessing insulin secretion.[15,16]

The current study has certain limitations. Patients with a fasting C-peptide concentration and an increment in C-peptide level of less than 0.17 nmol/L 6 minutes after intravenous injection of 1 mg of glucagon were classified as having type 1 diabetes and, therefore, were excluded from the study. These criteria are similar to those used by the Diabetes Control and Complications Trial[14] and the Veterans Affairs Cooperative Study.[17] Although all patients with diabetes in Rochester, Minnesota, were invited to join, only half agreed to participate in the study.[10] Therefore, the results in the subset analyzed may have differed from those in the population as a whole. We doubt that this is the case because sex, type of therapy, and type of diabetes did not differ between those who did and those who did not participate in the Rochester Diabetic Neuropathy Study.[10] We do not know whether our conclusions would have been the same if a secretagogue other than glucagon had been used to assess insulin secretion. We chose glucagon because it has been widely used to measure insulin secretion in cross-sectional[15,16,18,19] and randomized clinical trials[17,20] and because the prevailing glucose concentration has a minimal effect on the response to glucagon.[15,16] Nevertheless, in future studies, it would be of considerable interest to determine whether the temporal pattern of change is the same as that observed in the current study when other secretagogues (for example, plasma glucose) are used to assess insulin secretion.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.