Journal Logo

Research Article: Observational Study

Effect of short-term intensive insulin therapy on α-cell function in patients with newly diagnosed type 2 diabetes

Zheng, Hai-Lan MMa,b; Xing, Yan MMc; Li, Fan MMb; Ding, Wei MMb; Ye, Shan-Dong MDa,c,∗

Editor(s): Barzilay., Joshua

Author Information
doi: 10.1097/MD.0000000000019685


1 Introduction

Type 2 diabetes mellitus (T2DM) is a global problem, and its prevalence is rapidly increasing. T2DM is characterized fundamentally by a disorder in the metabolism of carbohydrates, resulting from the relative or absolute lack of insulin secretion.[1] Excess secretion of glucagon may underlie the important manifestations of diabetes, since chronic sustained hyperglucagonemia appears in T2DM patients.[2,3]

Islets are clusters of cells in the pancreas that regulate blood glucose by producing insulin and glucagon, which together are part of a feedback system ensuring a stable blood glucose level. The glucagon-secreting α-cells are scattered among the β-cells, functioning reciprocally. About 40 years ago, Unger et al[4] proposed a bi-hormonal theory to explain the pathophysiology of diabetes, that is, insulin deficiency or resistance accompanied by an absolute or relative excess of glucagon. Thus, T2DM is a complex disorder resulting from impaired function of α- and β-cells, as well as abnormal secretion patterns of insulin and glucagon. However, a controversy remains regarding the role of α-cell dysfunction in the pathogenesis of diabetes, and especially whether α-cell hyperactivity is secondary to, or concomitant with, the decline of β-cell secretion.

During the early course of diabetes, insulin secretion defects are considered reversible. Recent studies have shown that in patients with newly diagnosed T2DM, short-term intensive insulin therapy can improve β-cell function, decrease glycemic variability, induce glycemic remission, and even maintain glucose homeostasis after the therapy is stopped.[5–9] However, few studies have focused on the function of α-cells. Questions, such as whether α-cell hyperfunction is a primary defect or secondary to the decline of β-cell function, remain unanswered.

In addition, treatment durations have varied in most studies, from 2 to 4 weeks[5–7,10,11]; in others, patients with diabetes were given 4 to 8 weeks of intensive insulin therapy.[12] These latter focused on the effects of short-term intensive therapy on β-cell function as well as the heterogeneity of response to insulin intensive therapy in different patients.

The China guidelines recommend courses of intensive insulin therapy of 2 weeks to 3 months for the prevention and treatment of T2DM.[13] Yet, it is unknown whether patients may not benefit more by prolonging the duration of intensive insulin therapy, and there is no consensus regarding the proper indicator (eg, hemoglobin A1c, or glycated hemoglobin [HbA1c]) for initiating treatment.[14,15]

This study compared the effect of 12 weeks of intensive insulin therapy, relative to 12 weeks of oral hypoglycemic agents (OHAs), on hyperglucagonemia in newly diagnosed T2DM. The function of β and α cells after intensive insulin therapy and their correlations were evaluated, and the glycemic and lipid outcomes were compared in a 1-year follow-up.

2 Materials and methods

2.1 Subjects and grouping

Patients with newly diagnosed T2DM (n = 108, defined as diagnosis within the previous 6 months) were consecutively enrolled for this prospective study from January 2015 to December 2015. The inclusion criteria were the following: age >18 years; newly diagnosed T2DM; and HbA1c between 9% and 11%. Patients with acute diabetic complications such as diabetic ketoacidosis or severe heart, liver, or kidney dysfunction were excluded. All patients provided written informed consent before participation.

The 108 patients were assigned to 2 groups to receive either intensive insulin therapy or OHAs (IIT and OHA, respectively) by using computer-generated random numbers. All patients were initially hospitalized for 2 weeks, and then discharged and followed up in outpatient clinics. The blood glucose was checked 7 to 8 times during hospitalization and 4 to 5 times after discharge by self-monitoring blood glucose.

Patients in the IIT group were administered intensive BIT (rapid-acting insulin analogs at 3 meals + insulin glargine at bedtime) + metformin 0.5 g tid, for 12 weeks. Patients in the OHA group received metformin 0.5 g tid + gliclazide sustained-release tablets (60–120 mg daily), with acarbose (50–100 mg tid), if the postprandial glucose reached >10 mmol/L, or pioglitazone 15 to 30 mg qd for body mass index >25 kg/m2, for 12 weeks. The dose of OHAs or insulin was adjusted every 3 days to achieve or maintain within the following glycemic ranges: fasting glucose 4.4 to 6.1 mmol/L and postprandial glucose level 4.4 to 7.8 mmol/L. Remission was defined as achievement of target glycemic goal (HbA1c <7%) without use of hypoglycemic agents, lasting >6 months.

After 12 weeks, patients in the IIT group discontinued the insulin and were maintained on metformin; the patients were treated with other hypoglycemic agents (gliclazide sustained-release tablets 60–120 mg daily, with or without acarbose 50–100 mg tid, pioglitazone 15 mg qd). Patients in the OHA group remained with the original regimen. No lipid-lowering agents were used in either group. In the course of follow-up, OHAs were adjusted to maintain the glycemic goal. The patients discontinued the hypoglycemic agents and were encouraged with dietary and physical activity intervention for glycemic control, if they achieved the glycemic goal for more than 1 month on 2 or fewer hypoglycemic agents.

Thirty normal healthy volunteers were enrolled as the control group.

2.2 Laboratory tests

The level of proinsulin was determined via enzyme-linked immunosorbent assay with a microplate reader (Biocell Ht-1, R&D system). The glucagon and C-peptide (CP) were measured by radioimmunoassay (Roche Diagnostics). Blood glucose was measured using the glucose oxidation method with a Roche Diagnostics Biochemistry Analyzer (Roche Diagnostics). HbA1c was detected with high-performance liquid chromatography (Bio-Rad Laboratories). Immunoturbidimetry and Benedict–Behre creatinine colorimetry were used to determine the concentrations of albumin and creatinine in the urine samples, respectively. The urinary albumin-to-creatinine ratio was calculated as urinary albumin (mg/L) divided by urinary creatinine (g/L).

All patients were routinely examined with urine/blood tests, liver and renal function, blood lipids, electrolytes, and electrocardiogram. The blood samples were collected at the start of therapy (baseline) and at 12 weeks and 12 months after treatment.

2.3 Meal tolerance test

The subjects in all 3 groups underwent a meal tolerance test, with 100 g of standardized steamed bread. During the testing period, levels of glucagon, proinsulin, CP, and blood glucose were measured at timepoints 0 (fasting), 30, and 120 minutes.

The area under the receiver operating characteristic curve (AUC) of glucagon (AUCg) was calculated using the following formula: (fasting glucagon value, ng/L × 15) + (postprandial glucagon value at 30 minutes × 60) + (postprandial glucagon value at 120 minutes × 45).

The early insulin secretion index consisted of the CP/glucose (Glu) ratio response over the first 30 minutes of the test (ΔCP30/ΔG30) as follows: ΔCP30/ΔG30 = CP30 – CP0, pmol/L/Glu30 – Glu0, mmol/L.

The overall insulin secretion index was defined as the CP/Glu ratio response over the 120 minutes of the test (ΔCP120/ΔG120) as follows: ΔCP120/ΔG120 = CP120 – CP0, pmol/L/Glu120 – Glu0, mmol/L.

2.4 Statistical analysis

All statistical analyses were performed using SPSS 13.0. The continuous and discrete variables are expressed as mean ± standard deviation or the number of patients (percentage). Continuous variables were compared using the paired t test to evaluate intragroup differences before and after treatment, and the independent-samples t test to evaluate intergroup differences. Measurements at multiple timepoints were analyzed by the repeated measures analysis of variance method. Categorical variables were compared using the Chi-squared test. Correlations between data were analyzed using Spearman correlation analysis. P < .05 was considered statistically significant.

3 Results

3.1 Baseline characteristics of the subjects

The study included 108 patients with newly diagnosed T2DM, each given 12 weeks of therapy: 56 in the IIT group (intensive BIT + metformin), and 52 in the OHA group (OHAs + metformin). Another 30 healthy volunteers were enrolled as the control group. The diabetic patients exhibited significantly higher levels of fasting blood glucose, HbA1c, total cholesterol (TC), and triglyceride (TG) compared with the control group (Table 1). However, the IIT and OHA groups were similar with regard to age, gender ratio, duration of disease, body mass index, fasting blood glucose, HbA1c, urinary albumin-to-creatinine ratio, systolic and diastolic blood pressure, TC, and TG (all, P > .05).

Table 1
Table 1:
Baseline characteristics of subjects in the ITT, OHA, and control groups.

3.2 Changes in glycemic parameters at 12 weeks and 12 months after treatment

Before treatment, the diabetic patients in both the IIT and OHA groups exhibited higher glucagon levels at 0, 30, and 120 minutes of the meal tolerance test, as well as AUCg, compared with the control group (Fig. 1 and Table 2, and Supplementary Table 1, In addition, the ratio of fasting proinsulin-to-CP (proinsulin/CP) was significantly higher, and ΔCP30/ΔG30 and ΔCP120/ΔG120 were significantly lower, compared with the control group.

Figure 1
Figure 1:
Changes in glycemic parameters at 12 wk and 12 mo after treatment in the ITT, OHA, and control groups. (A) Glucagon levels at 0, 30, and 120 min of the meal tolerance test. (B) Ratio of fasting proinsulin-to-C-peptide (proinsulin/CP); (C) ΔCP30/ΔG30; (D) ΔCP120/ΔG120. The data are expressed as mean ± SD. Glucagon, ng/L; proinsulin, pmol/L; CP, pmol/L. CP = C-peptide, IIT = intensive insulin therapy, OHA = oral hypoglycemic agent, SD = standard deviation.
Table 2
Table 2:
Comparison of glycemic parameters.

Between the OHA and IIT groups, there were no significant differences in any of the analyzed glycemic parameters at baseline. In the IIT group, after intensive BIT for 12 weeks, compared with baseline the glucagon levels (0, 30 minutes and 120 minutes) and proinsulin/CP were significantly lower, whereas ΔCP30/ΔG30 and ΔCP120/ΔG120 were significantly higher (all, P < .01). In contrast, in the OHA group at 12 weeks, ΔCP30/ΔG30 and ΔCP120/ΔG120 were significantly higher compared with the baseline; while glucagon levels (0, 30 minutes, and 120 minutes) and proinsulin/CP were slightly lower, which failed to achieve statistical significance.

At 12 weeks, patients in the IIT group exhibited significantly lower glucagon levels and proinsulin/CP, and higher ΔCP30/ΔG30 and ΔCP120/ΔG120, compared with the OHA group, indicating that intensive BIT was superior to OHAs in glycemic control. On the other hand, although the proinsulin/CP, ΔCP30/ΔG30, and ΔCP120/ΔG120 of the IIT group were comparable to that of the control group at 12 weeks, the glucagon levels were still significantly higher than the normal levels.

In the OHA group, all the analyzed glycemic parameters differed significantly from the normal levels of the control group.

A 12 months, in the IIT group there were no significant changes in any of the glycemic parameters compared with their respective levels at 12 weeks. However, in the OHA group the levels of glucagon and proinsulin/CP were significantly higher at 12 months compared with the corresponding parameters at 12 weeks.

3.3 Spearman correlations of glycemic variables with the percentage change in total AUCg in the ITT group

There were no correlations between the percentage change in total AUCg (change between baseline cf. 12 weeks) and other glycemic parameters (proinsulin/CP, ΔCP30/ΔG30, and ΔCP120/ΔG120; Table 3).

Table 3
Table 3:
Spearman correlations of glycemic variables with the percentage change in total AUC (glucagon) in the ITT group.

3.4 Glycemic and lipid outcomes at 12 weeks and 12 months

In the IIT group, 97.9% and 80.9% of patients achieved the target glycemic goal (defined as HbA1c <7%) at 12 weeks and 12 months, respectively; whereas in the OHA group, only 62.5% and 44.4% achieved the target glycemic goal (both P < .05 compared with the IIT group at the same timepoint; Table 4). Compared with the OHA group, patients in the IIT group had lower TC and TG levels at 12 weeks and 12 months. Interestingly, patients in the IIT group used significantly fewer types of OHAs (1.5 ± 0.8 at 12 months), but achieved a significantly higher remission rate (31.0%, defined as achievement of target glycemic goal without use of hypoglycemic agents lasting >6 months) compared with the OHA group (2.4 ± 0.9, 5.6%, respectively).

Table 4
Table 4:
Glycemic and lipid outcomes at 12 wk and 12 mo in the IIT and OHA groups.

4 Discussion

Proinsulin is the precursor of insulin and CP. In the secretory granules of β-cells, proinsulin can be cleaved by a combination of pancreatic trypsin and carboxypeptidase, which leads to a conversion of proinsulin into equimolar concentrations of insulin and CP. Proinsulin-like material represents only 15% of serum immunoreactive insulin in fasting healthy individuals.[16] However, circulating proinsulin levels and proinsulin/CP ratios are significantly higher than normal in T2DM patients, and may serve as highly specific markers of insulin synthesis dysfunction as well as insulin resistance.[17]

In our present study, the basal proinsulin/CP ratio of the newly diagnosed patients was ∼28%, which was significantly higher than that of the normal glucose tolerance group (14%). In addition, ΔCP30/ΔG30, and ΔCP120/ΔG120, the indices reflecting early and overall insulin secretion, were remarkably lower than normal. This suggests impairment of β-cell function in the newly diagnosed T2DM patients.

The above data revealed abnormalities in both synthesis and secretion of insulin. On the other hand, the newly diagnosed T2DM patients exhibited hyperglucagonemia compared with the subjects with normal glucose metabolism, suggesting functional hyperactivity of α-cells. The ratio of proinsulin/CP in the T2DM patients of the IIT group was significantly lower after a short term of intensive BIT compared with the baseline, accompanied by higher ΔCP30/ΔG30 and ΔCP120/ΔG120. This indicates that intensive BIT could improve insulin synthesis and secretion of β-cells. Intensive BIT was also associated with a reduction in the amount of glucagon at all timepoints of the meal tolerance test, although it still failed to normalize glucagon levels. This result is consistent with other studies.[5,7]

There is no consensus regarding whether α-cell hyperactivity is secondary to, or concomitant with, a decline of β-cell secretion. It has been reported that the elevations in glucagon levels observed in diabetic patients develop due to diminished islet β-cell function.[2,18] However, there is also a view that α-cell hyperfunction develops concomitantly with the onset of β-cell dysfunction in T2DM patients, and the inhibitory effect of hyperglycemia and insulin on glucagon secretion are both weakened.[5,19] In the present study, the proinsulin/CP ratio was significantly lower at 12 weeks after intensive BIT, which were close to the normal levels, suggesting great improvement in β-cell function. However, the AUCg in diabetic patients after intensive treatment was still significantly higher than that of the normal subjects. Especially, there were no correlations between the percentage change in total AUCg and various glycemic parameters, such as proinsulin/CP, ΔCP30/ΔG30, and ΔCP120/ΔG120. These data indicated a possibility of primary α-cell defect which did not improve in parallel with the recovery of β-cell function.

Similarly, Kramer et al[5] found a significant reduction in AUCg after intensive insulin therapy. However, the decrease in AUCg was not associated with the change in either β-cell function or in glycemic variability. Thus, it is presumed that the partial reversibility of α-cell function may be independent of the recovery of insulin secretion. The exact mechanism by which intensive insulin therapy ameliorates reversible α-cell dysfunction at the early stage of the course of T2DM remains unclear. One possible reason is that, in addition to the known adverse effects of chronic hyperglycemia on β-cells, there may be one (or more) unknown or unmeasured factors that affect α-cell function or the secretion of glucagon. Such unfavorable factor(s) could be at least partially eliminated by intensive insulin therapy, and as a result hyperglucagonemia was alleviated.[20,21]

There is a view that for newly diagnosed T2DM patients with HbA1c >9.0%, intensive insulin therapy can facilitate remission of diabetes in some patients, not only achieving acute correction of hyperglycemia and avoiding glucotoxicity, but also successfully laying a foundation for prolonged good glycemic control after discontinuation of insulin.[15,22] Short-term intensive insulin therapy has been reported to improve the underlying pathophysiological defects in early T2DM, and maintain normoglycemia without use of anti-diabetic medication after discontinuation of insulin.[6,23] In the present study, the newly diagnosed T2DM patients were treated with intensive BIT for 12 weeks and then followed for 1 year. A higher percentage of patients with intensive BIT achieved the target glycemic goal (HbA1c <7%), and had lower levels of TC and TG at 12 weeks and 12 months, compared with those given OHAs. In addition, patients treated with short-term intensive BIT used fewer types of OHAs after discontinuation of insulin, but achieved a high remission rate without the use of hypoglycemic agents. These data indicate that better, longer metabolic improvements may be achieved in patients with newly diagnosed T2DM after short-term intensive insulin therapy, as compared with the use of OHAs.[1,9,14]

This study had several limitations. First, it is based on a single-center with a relatively small sample size. Second, the role of some unmeasured confounding factors that could have possibly influenced the observed association cannot be entirely ruled out. For example, the extent to which lifestyle and dietary habits affected our study population was not known.

In conclusion, in this study short-term intensive BIT was associated with improvements in both β-cell and α-cell function in patients with newly diagnosed T2DM, allowing better, long-term metabolic control with minimal OHAs. A decline of β-cell secretion and concomitant α-cell dysfunction may together be involved in the pathogenesis of T2DM.

Author contributions

Conceptualization: Hai-Lan Zheng, Yan Xing, Fan Li, Shan-Dong Ye.

Data curation: Hai-Lan Zheng, Yan Xing, Fan Li, Wei Ding.

Formal analysis: Shan-Dong Ye.

Funding acquisition: Shan-Dong Ye.

Investigation: Hai-Lan Zheng, Yan Xing, Fan Li, Wei Ding, Shan-Dong Ye.

Methodology: Hai-Lan Zheng, Yan Xing, Fan Li.

Project administration: Shan-Dong Ye.

Resources: Hai-Lan Zheng, Wei Ding, Shan-Dong Ye.

Software: Fan Li.

Supervision: Shan-Dong Ye.

Validation: Hai-Lan Zheng, Yan Xing, Fan Li, Wei Ding, Shan-Dong Ye.

Visualization: Hai-Lan Zheng, Yan Xing, Fan Li, Wei Ding, Shan-Dong Ye.

Writing – original draft: Hai-Lan Zheng, Yan Xing, Fan Li.

Writing – review & editing: Hai-Lan Zheng, Yan Xing, Fan Li, Wei Ding, Shan-Dong Ye.


[1]. Acloque H, Adams MS, Fishwick K, et al. Epithelial-mesenchymal transitions: the importance of changing cell state in development and disease. J Clin Invest 2009;119:1438–49.
[2]. Quesada I, Tuduri E, Ripoll C, et al. Physiology of the pancreatic alpha-cell and glucagon secretion: role in glucose homeostasis and diabetes. J Endocrinol 2008;199:5–19.
[3]. Lund A, Bagger JI, Christensen M, et al. Glucagon and type 2 diabetes: the return of the alpha cell. Curr Diab Rep 2014;14:555.
[4]. Unger RH, Orci L. The essential role of glucagon in the pathogenesis of diabetes mellitus. Lancet 1975;1:14–6.
[5]. Kramer CK, Zinman B, Choi H, et al. Effect of short-term intensive insulin therapy on post-challenge hyperglucagonemia in early type 2 diabetes. J Clin Endocrinol Metab 2015;100:2987–95.
[6]. Kramer CK, Zinman B, Retnakaran R. Short-term intensive insulin therapy in type 2 diabetes mellitus: a systematic review and meta-analysis. Lancet Diabetes Endocrinol 2013;1:28–34.
[7]. Ryan EA, Imes S, Wallace C. Short-term intensive insulin therapy in newly diagnosed type 2 diabetes. Diabetes Care 2004;27:1028–32.
[8]. Retnakaran R, Kramer CK, Choi H, et al. Liraglutide and the preservation of pancreatic beta-cell function in early type 2 diabetes: the LIBRA trial. Diabetes Care 2014;37:3270–8.
[9]. Xu W, Weng J. Current role of short-term intensive insulin strategies in newly diagnosed type 2 diabetes. J Diabetes 2013;5:268–74.
[10]. Kramer CK, Zinman B, Choi H, et al. Predictors of sustained drug-free diabetes remission over 48 weeks following short-term intensive insulin therapy in early type 2 diabetes. BMJ Open Diabetes Res Care 2016;4:e000270.
[11]. Cheng L, Xu M, Lin X, et al. The intriguing effects of time to glycemic goal in newly diagnosed type 2 diabetes after short-term intensive insulin therapy. Endocr J 2016;63:739–46.
[12]. Retnakaran R, Yakubovich N, Qi Y, et al. The response to short-term intensive insulin therapy in type 2 diabetes. Diabetes Obes Metab 2010;12:65–71.
[13]. Diabetes Branch of Chinese Medical Association. China guidelines for the prevention and treatment of type 2 diabetes (2013). Chin J Endocrinol Metab 2014;30:893–942.
[14]. Stein CM, Kramer CK, Zinman B, et al. Clinical predictors and time course of the improvement in beta-cell function with short-term intensive insulin therapy in patients with Type 2 diabetes. Diabet Med 2015;32:645–52.
[15]. Weng JP. Short-term intensive insulin therapy could be the preferred option for new onset type 2 diabetes mellitus patients with HbA1c>9. J Diabetes 2017;9:890–3.
[16]. Kahn SE, Halban PA. Release of incompletely processed proinsulin is the cause of the disproportionate proinsulinemia of NIDDM. Diabetes 1997;46:1725–32.
[17]. Pfutzner A, Standl E, Hohberg C, et al. IRIS II study: intact proinsulin is confirmed as a highly specific indicator for insulin resistance in a large cross-sectional study design. Diabetes Technol Ther 2005;7:478–86.
[18]. Dunning BE, Gerich JE. The role of alpha-cell dysregulation in fasting and postprandial hyperglycemia in type 2 diabetes and therapeutic implications. Endocr Rev 2007;28:253–83.
[19]. Gromada J, Franklin I, Wollheim CB. Alpha-cells of the endocrine pancreas: 35 years of research but the enigma remains. Endocr Rev 2007;28:84–116.
[20]. Jamison RA, Stark R, Dong J, et al. Hyperglucagonemia precedes a decline in insulin secretion and causes hyperglycemia in chronically glucose-infused rats. Am J Physiol Endocrinol Metab 2011;301:E1174–83.
[21]. Zhang B, Chen YY, Yang ZJ, et al. Improvement in insulin sensitivity following intensive insulin therapy and association of glucagon with long-term diabetes remission. J Int Med Res 2016;44:1543–50.
[22]. Hanefeld M. Use of insulin in type 2 diabetes: what we learned from recent clinical trials on the benefits of early insulin initiation. Diabetes Metab 2014;40:391–9.
[23]. Chon S, Oh S, Kim SW, et al. The effect of early insulin therapy on pancreatic beta-cell function and long-term glycemic control in newly diagnosed type 2 diabetic patients. Korean J Intern Med 2010;25:273–81.

glucagon; intensive insulin therapy; type 2 diabetes mellitus

Supplemental Digital Content

Copyright © 2020 the Author(s). Published by Wolters Kluwer Health, Inc.