Effect of the Α-Glucosidase Inhibitor Miglitol on the Glucose Profile in Japanese Type 2 Diabetic Patients Receiving Multiple Daily Insulin Injections

Endocrine Journal 2012, 59 (4), 345-352

Or i g i n a l Effect of the α-glucosidase inhibitor miglitol on the glucose profile in Japanese type 2 diabetic patients receiving multiple daily insulin injections

Hiroyuki Kato, Akio Ohta, Suzuko Kobayashi, Satoshi Ishii, Yukiyoshi Sada, Hidetoshi Kobayashi, Shintaro Ohmori, Akihiko Kondo, Takuyuki Katabami, Junro Fuse, Hisashi Fukuda, Yoshio Nagai and Yasushi Tanaka

Department of Internal Medicine, Division of Metabolism and Endocrinology, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan

Abstract. Strict postprandial glycemic control may have a preventive effect on atherogenesis in patients with type 2 diabetes. The α-glucosidase inhibitor (α-GI) miglitol is useful for controlling the early postprandial increase of glucose, but the combined effect of miglitol and multiple daily insulin injections (MDI) on glucose excursion has not been evaluated. First, we retrospectively compared the daily glucose profile, evaluated by self-monitoring of blood glucose (SMBG) at nine times on the day before discharge from hospital, between type 2 diabetic patients receiving MDI (n=81) or MDI plus miglitol at 150 mg daily (n=24). Second, we prospectively examined the effect of adding miglitol to MDI on the daily glucose profile (SMBG) in 19 other type 2 diabetic patients. Although the daily insulin dosage and the glucose level before meals did not differ between the two groups, the 1-h postprandial glucose level after each meal, 2-h glucose level after lunch and dinner, mean and standard deviation of glucose, and amplitude of glucose excursion were significantly lower or smaller in the MDI plus miglitol group than in the MDI group. All of these glucose parameters were significantly improved by adding miglitol to MDI in the prospective cohort of 19 patients. In conclusion, adding miglitol to MDI reduces postprandial glucose levels and attenuates daily glucose fluctuation in type 2 diabetic patients. This trial was registered with UMIN (no. UMIN000005383).

Key words: Daily glucose profile, Miglitol, Multiple daily insulin injections

The postprandial plasma glucose level is patients undergoing prospective follow up for 5 years thought to be a risk factor for the cardiovascular com- [5]. Recently, Esposito et al. reported that the post- plications of diabetes independently of fasting plasma prandial peak glucose level determined by self-moni- glucose (FPG) [1-3]. Hanefeld et al. found that the 1-h toring of blood glucose (SMBG) at home was associ- postprandial glucose level after breakfast, but not FPG, ated with carotid artery intima-media thickness (IMT), was a risk factor for myocardial infarction in an 11-year a surrogate marker of atherosclerosis, independently of prospective study of patients with newly diagnosed hemoglobin A1c (HbA1c) in type 2 diabetic patients [6], non-insulin dependent diabetes mellitus (NIDDM) and that there was an incremental glucose peak within [4]. Another study showed that the 2-h postprandial 1 hour of starting either main meal (lunch or dinner). glucose level after lunch, but not FPG, was a strong Similarly, Hu et al. showed that the postchallenge glu- predictor of cardiovascular events in type 2 diabetic cose spike (the difference between the maximal plasma glucose level during a 75 g oral glucose tolerance test Submitted Oct. 6, 2011; Accepted Jan. 19, 2012 as EJ11-0283 and the baseline fasting plasma glucose level) was Released online in J-STAGE as advance publication Feb. 2, 2012 associated with carotid artery IMT independently of Correspondence to: Akio Ohta, M.D., Department of Internal HbA1c in type 2 diabetic patients [7]. Although post- Medicine, Division of Metabolism and Endocrinology, St. Marianna University School of Medicine, 2-16-1 Sugao, Miyamae- prandial glucose levels were not measured after every ku, Kawasaki, Kanagawa 216-8511, Japan meal in these studies, the findings imply that control- E-mail: [email protected] ling excessive early postprandial hyperglycemia could

©The Japan Endocrine Society 346 Kato et al. help to prevent cardiovascular complications. as the National Glycohemoglobin Standardizaion Rapid-acting insulin preparations, such as Insulin Program (NGSP) equivalent value, which was calcu- Aspart or Insulin Lyspro, are useful to improve post- lated as HbA1C (NGSP) (%) = HbA1C (JDS) (%) + 0.4%, prandial hyperglycemia, since these insulins have sub- based on the relationship of HbA1C (NGSP) values to stituted amino acids at position 28 and/or 29 and are HbA1C (JDS) (%) values measured with the Japanese absorbed more quickly to act more rapidly than reg- standard and measurement method [12]. Serum GA ular human insulin [8, 9]. The α-glucosidase inhibi- was measured by an enzymatic method using an albu- tor (α-GI) miglitol is a unique oral antidiabetic agent min-specific protease (ketoamine oxidase) and an albu- that suppresses the early postprandial increase of glu- min assay reagent (Lucica GA-L, Asahi Kasei Pharma, cose more potently than other α-GIs, since miglitol is Tokyo, Japan). All patients were given instructions absorbed from the small intestine after oral administra- about how to perform SMBG with a One Touch Ultra tion and has a stronger action especially in the proxi- (Johnson and Johnson KK, Tokyo, Japan) by educa- mal small bowel [10, 11]. However, it remains unclear tional nurses. All patients started MDI after admis- whether the combination of miglitol and a rapid-act- sion, and their insulin dosage was adjusted to achieve ing insulin analog can further improve the early post- a glucose level <140 mg/dL, <200 mg/dL, and <180 prandial glucose profile compared with insulin alone. mg/dL before, 1 h after, and 2 h after meals, respec- In the present study, we retrospectively compared tively, in the absence of hypoglycemia (<70 mg/dL). the daily glucose profile obtained by SMBG between When the 1-h or 2-h postprandial glucose target could type 2 diabetic patients receiving multiple daily insu- not be achieved by MDI within one week after hospi- lin injections (MDI), i.e., rapid-acting insulin (Insulin talization despite reaching a preprandial glucose <140 Aspart or Insulin Lyspro) before each meal and long mg/dL, miglitol was added at 150 mg daily (50 mg - acting insulin (Insulin Glargine or Insulin Detemir) just before each meal). Other oral antidiabetic agents before bedtime, without an α-GI versus MDI plus were not changed during admission. All patients per- miglitol. In addition, we prospectively evaluated the formed SMBG at nine times (before each meal and 1 effect of adding miglitol to MDI on the daily glucose h and 2 h after each meal) on the day before discharge, profile (SMBG) in other type 2 diabetic patients. and the daily glucose profile of the patients treated by MDI alone (MDI group, n=81) was compared with that Materials and Methods of patients receiving MDI plus miglitol (MDI + Mig group, n=24). Subjects A total of 124 patients with type 2 diabetes admit- Effect of combined therapy on postprandial glucose ted to St. Marianna University Hospital (Kanagawa, In 19 patients receiving MDI who had stable pre- Japan) in 2009-2010 were studied, including 71 men prandial glucose levels <140 mg/dL, but inadequate and 53 women aged 60 ± 12 years (mean ± SD). They postprandial glycemic control, miglitol was added at were admitted to hospital for diabetes education and 150 mg daily (50 mg just before each meal) without improvement of glycemic control, and achieved good changing the insulin dosage. Other oral antidiabetic control by MDI with a glucose level of <140 mg/dL agents were not changed. To evaluate the effect of add- before each meal. The inclusion for this study crite- ing miglitol on the glucose profile, SMBG was done at rion was a glucose level <140 mg/dL before each meal nine time points on the day before and 7 or 8 days after without any symptoms of hypoglycemia throughout starting miglitol. the day. The exclusion criteria were pregnancy, any severe illness, chronic liver disease, thyroid disease, Statistical analysis malignancy, and a history of abdominal surgery. All Results are presented as the mean ± SD. Comparison patients gave written informed consent and the study of mean values was done by the paired or unpaired was approved by the Ethics Committee of St. Marianna t-test. All analyses were performed with Stat-View University School of Medicine (No. 1226). software (Abacus Concepts, Berkeley, CA) and statistical significance was accepted atp <0.05. Comparison of MDI with MDI plus miglitol On the day of admission, HbA1C (%) was estimated Miglitol with multiple insulin injections 347

Results of glucose excursion were lower or smaller in patients treated with MDI plus miglitol than in patients receiv- Retrospective study ing MDI alone, although there was no difference of The clinical profile on admission and a compari- their daily insulin dosage. Adding miglitol to MDI son of the daily insulin dose and glucose excursion decreased the early postprandial glucose surge after between the two treatment groups on the day before each meal, even in patients who had not achieved good discharge are shown in Table 1. The age, gender, BMI, postprandial glucose levels with MDI, resulting in HbA1c, and GA did not differ between the two groups reduction of the mean, SD, and amplitude of glucose. on admission. The daily dosages of rapid-acting and Our results suggested that combined therapy with MDI long - acting insulin, as well as the total insulin dose, plus miglitol achieves better control of glucose excur- also did not differ between the two groups. However, sion compared with MDI alone. Physicians some- both the mean and SD of the daily glucose level were times encounter patients without a decrease of the 1-h significantly lower in the MDI + Mig group than in the postprandial glucose level despite an increase of their MDI group on the day before discharge, and the ampli- insulin dose. In this study, we found that the 1-h post- tude of daily glucose excursion (the difference between prandial glucose levels were significantly reduced by maximum and minimum glucose levels) was smaller adding miglitol to MDI. Also, the percent reduction of in the former group on the day before discharge. As glucose at 1 h after lunch and dinner (-41.3∼-42.7%) shown in Fig. 1, while the mean glucose level before achieved by adding miglitol was significantly greater each meal and at 2 h after breakfast did not differ than at 2 h after those meals. Our results indicate that between the two groups, the mean glucose levels at 1 h adding miglitol to MDI was effective in Japanese dia- after each meal and at 2 h after lunch and dinner were betics with impaired early insulin secretion receiving a significantly lower in the MDI + Mig group than in the high carbohydrate load. MDI group. Since α-GIs inhibit disaccharidases in the small intestine and thus attenuate the postprandial elevation Prospective study of plasma glucose by delaying the digestion and absorp- The clinical profile of the 19 patients investigated pro- tion of carbohydrates, these drugs are usually taken spectively and the effect of miglitol on glucose excur- just before meals. However, a recent study showed sion are shown in Table 2. Despite the daily insulin that administration of miglitol up to 30 min after the dosage being unchanged, the mean and SD of the daily start of a meal was effective in patients with type 2 dia- glucose level and the amplitude of daily glucose excur- betes [13]. Although these patients were not receiv- sion showed a significant decrease after starting migli- ing insulin, their plasma glucose levels at 120 and 180 tol. As shown in Fig. 2 and Table 3, the mean glucose min postprandially were lower than when miglitol was level before each meal was unchanged, but the mean given just before meals. Therefore, miglitol was given glucose levels at 1 h and 2 h after meals were signifi- just before each meal in the present study to avoid the cantly reduced by miglitol. The percent improvement risk of late postprandial hypoglycemia. of daily glucose excursion in the 19 patients after adding Recently, Nemoto et al. assessed the effect on glu- miglitol to MDI is shown in Table 3. The percent reduc- cose excursion of adding miglitol (50 mg per dose) in tion of the 1-h postprandial glucose level after lunch a placebo-controlled double-blind study of type 2 dia- and dinner was significantly greater than that of the 2-h betic patients receiving various types of insulin [14]. glucose level (lunch: -41.3±14.9% vs. -28.4±17.7%, They evaluated plasma glucose during a morning meal P=0.0174; dinner: -42.7±16.9% vs. -29.1±18.7%, tolerance test at baseline and after 12 weeks of migli- P=0.0011). None of the patients had hypoglycemic tol or placebo therapy, and found lower plasma glu- symptoms before or after starting miglitol. cose levels at 0.5 h, 1 h, and 2 h after the test meal in the miglitol group compared with the placebo group, Discussion as well as a greater decrease of the mean 1-h and 2-h postprandial plasma glucose levels from baseline The present study demonstrated that the early post- in the miglitol group than the placebo group (60.3 ± prandial increase of glucose after each meal, the mean 70.1 vs. -5.1 ± 68.2 mg/dL, mean ± SD). The results and SD of the daily glucose level, and the amplitude of the present study are consistent with their findings, 348 Kato et al.

Table 1 Clinical profile and comparison of glucose excursion between the two groups MDI group MDI + Mig group p value On admission N (Male/Female) 81 (42/39) 24 (16/8) 0.200 Age (yrs) 61 ± 12 58 ± 12 0.309 BMI (kg/m2) 26.3 ± 5.6 26.8 ± 5.2 0.486 HbA1c (%) 9.5 ± 1.9 10.3 ± 2.4 0.156 GA (%) 30.7 ± 25.9 28.2 ± 8.2 0.592 GA/HbA1c 3.17 ± 3.67 2.63 ± 0.69 0.480 U-CPR(μg/day) 39.6 ± 41.5 53.9 ± 44.5 0.155 At discharge Daily insulin dose (U/day) rapid-acting insulin analog (U/day) 24 ± 11 21 ± 12 0.097 long-acting insulion analog (U/day) 15 ± 10 11 ± 7 0.106 total daily insulin dose (U/day) 39 ± 18 32 ± 17 0.061 Mean daily glucose level (mg/dL) 155 ± 30 130 ± 20 0.003 SD of daily glucose (mg/dL) 42 ± 15 32 ± 11 0.002 Amplitude of daily glucose excursion 122 ± 44 94 ± 40 0.006 (maximum-minimum) (mg/dL) Other medications, n (%) Pioglitazone 12 (14.8) 3 (12.5) 0.804 Metoformin 31 (38.3) 13 (54.2) 0.389 Statins 42 (51.9) 13 (54.2) 0.912 ACE I and/or ARB 35 (43.2) 10 (41.7) 0.932 Calcium channel blocker 14 (17.3) 6 (7.4) 0.932 Diuretics 6 (7.4) 1 (4.2) 0.598 β-blocker 2 (2.5) 0 (0.0) 0.447 Date are means ± SE or % (95% confidence interval). BMI, body mass index; ACE I, angiotensin-converting enzyme inhibitor; ARB, angiotensin II type 1 receptor blocker, GA, glycoalbumin; HbA1c, hemoglobin A1c; U-CPR, urinary C-peptide

Fig. 1 Comparison of daily glucose excursion in the two groups Data are the mean ± SE. *p<0.01. **p<0.001 vs. the MDI group. BB, before breakfast; 1HAB, 1 h after breakfast; 2HAB, 2 h after breakfast; BL, before lunch; 1HAL, 1 h after lunch; 2HAL, 2 h after lunch; BD, before dinner; 1HAD, 1 h after dinner; 2HAD, 2 h after dinner. Miglitol with multiple insulin injections 349

Table 2 Clinical profile and effect of adding miglitol to MDI on glucose excursion in 19 patients After starting Baseline p value miglitol N (Male/Female) 19 (13/6) Age (yrs) 58 ± 13 BMI (kg/m2) 26.3 ± 4.6 HbA1c (%) 9.3 ± 1.2 GA (%) 27.2 ± 6.1 GA/HbA1c 2.9 ± 0.34 U-CPR(μg/day) 61.7 ± 38.1 Daily insulin dose rapid-acting insulin analog (U/day) 19.1 ± 7.4 long-acting insulion analog (U/day) 9.9 ± 7.5 total daily insulin dose (U/day) 29.1 ± 10.3 Mean daily glucose level (mg/dL) 176 ± 46 124 ± 18 <0.0001 SD of daily glucose (mg/dL) 46 ± 45 26 ± 9 <0.0001 Amplitude of daily glucose excursion 128 ± 35 71 ± 28 <0.0001 (maximum-minimum) (mg/dL) Other medications for diabetes, n (%) Pioglitazone 3 (12.5) Metoformin 13 (54.2) Statins 13 (54.2) ACE I and/or ARB 10 (41.7) Calcium channel blocker 6 (7.4) Diuretics 1 (4.2) β-blocker 0 (0.0) Date are means ± SE or % (95% confidence interval). BMI, body mass index; ACE I, angiotensin-converting enzyme inhibitor; ARB, angiotensin II type 1 receptor blocker; GA, glycoalbumin; HbA1c, hemoglobin A1c; U-CPR, urinary C-peptide

Fig. 2 Effect of adding miglitol to MDI on daily glucose excursion in 19 patients Data are the mean ± SE. *p<0.01. **p<0.001 vs. baseline. BB, before breakfast; 1HAB, 1 h after breakfast; 2HAB, 2 h after breakfast; BL, before lunch; 1HAL, 1 h after lunch; 2HAL, 2 h after lunch; BD, before dinner; 1HAD, 1 h after dinner; 2HAD, 2 h after dinner. 350 Kato et al.

Table 3 Improvement rate of adding miglitol to MDI on daily glucose excursion in 19 patients Before meal 1-h after meal 2-h after meal IR (%) -11.7 ± 12.3 -39.5 ± 11.6 -32.5 ± 24.2 breakfast p value* p<0.01 p<0.001 p<0.001 p value** ns IR (%) -5.9 ± 23.9 -41.3 ± 14.9 -28.4 ± 17.7 lunch p value* ns p<0.001 p<0.001 p value** p=0.0174 IR (%) 7.9 ± 27.5 -42.7 ± 16.9 -29.1 ± 18.7 dinner p value* ns p<0.001 p<0.001 p value** p=0.0011 Data are the mean ± SE. IR: Improvement Rate from base line p value*: Improvement rate of adding miglitol to MDI vs. base line p value**: Improvement rate of glucose level 1-h after meal vs. 2-h after meal ns, not significant

although Nemoto’s study included various insulin reg- postprandially, with the decrease of plasma insulin at 1 imens and did not involve assessment of glucose lev- and 2 h being comparable to that after administration of els after lunch and dinner. Osonoi et al. previously placebo or voglibose (0.3 mg) [18]. They also exam- examined the effect of switching from another α-GI ined the reactive hyperemic response of forearm blood (voglibose at 0.3 mg per meal or acarbose at 100 mg flow as an index of endothelium-dependent vasodila- per dose) to miglitol (50 mg per dose) on daily glucose tation after a test meal. While patients given placebo excursion (SMBG) in type 2 diabetic patients receiving or voglibose showed a significant decrease of both oral anti-diabetic agents or insulin [15]. While switch- the absolute duration and percent duration of reactive ing to miglitol for 3 months did not affect HbA1c or hyperemia, these indices did not decrease in patients the plasma lipid profile, the glucose levels at 1 h after receiving miglitol, suggesting amelioration of vascu- lunch and dinner and the M-value [16] (a marker of lar endothelial dysfunction by miglitol therapy. In the daily glucose variation) were significantly reduced, present study, the mean and SD of the daily glucose and the incidence of hypoglycemia or hypoglycemic level and the amplitude of daily glucose excursion were symptoms were also decreased. These results sug- lower in the MDI + Mig group than in the MDI group gest the usefulness of combining miglitol with insulin (Tables 1 and 2). These results indicate that miglitol to achieve greater reduction of the early postprandial can have a beneficial effect for preventing the onset glucose spike compared with insulin therapy alone. and/or progression of atherosclerosis when it is added Recently, Nagai et al. reported that combining miglitol to basal insulin therapy. Since hyperinsulinemia itself with MDI was also effective for stabilizing the glucose is associated with atherogenesis and carcinogenesis profile in type 1 diabetic patients [17]. They observed [19, 20], it may be better to limit the dose of exogenous a significant increase of active GLP-1 after meals when insulin by adding an oral agent that does not promote miglitol was added, suggesting that its efficacy may not insulin secretion, such as an α-GI. Addition of migli- be entirely due to delaying carbohydrate absorption but tol markedly reduced postprandial glucose levels in our also to modifying the response of GLP-1. Although patients who had preprandial glucose levels <140 mg/ we did not measure incretin, C-peptide, and gluca- dL and had not achieved their postprandial glucose tar- gon concentrations in the present study, the marked get with MDI alone, and did so without an increase of effect of miglitol on the postprandial glucose level that the insulin dosage. However, direct evidence of this we observed may have partly been associated with has not been obtained, and the optimum glucose level enhancement of incretin activity. where we should consider adding miglitol has still not Recently, Hiki et al. reported that a single dose of been established. Thus, further prospective studies are miglitol (50 mg) improved the plasma glucose level at needed for clarification of these points. 1 h postprandially in diabetic patients with coronary The present study had several limitations. First, the artery disease, as well as the area under the concen- number of patients enrolled was small. Second, the tration vs. time curve of plasma glucose from 0 to 2 h glucose profile was only examined for 1 day by SMBG Miglitol with multiple insulin injections 351 and not by continuous glucose monitoring (CGM), present results. so the reliability and reproducibility of our data were In conclusion, adding miglitol to MDI reduced the not evaluated. Third, we did not use a test meal, but early postprandial glucose spike and decreased daily glu- instead used the standard meals planned by our hospi- cose excursion in Japanese patients with type diabetes. tal dieticians. Although the total energy and nutrient intake were similar in each patient, the contents of the Acknowledgments meals varied somewhat before and after starting miglitol. Fourth, our data were obtained from hospitalized No sources of funding were used to assist in the patients, and we did not assess the chronic effect of preparation of this manuscript. The authors have no miglitol. Therefore, a large-scale prospective evalua- conflicts of interest that are relevant to the content of tion in outpatients should be performed to confirm the this manuscript.

References

1. The DECODE study group on behalf of the European 9. Rys P, Pankiewicz O, Lach K, Kwaskowski A, Diabetes Epidemiology Group (2001) Glucose toler- Skrzekowska-Baran I, Malecki MT (2011) Efficacy and ance and cardiovascular mortality. Comparison of fast- safety comparison of rapid-acting insulin aspart and ing and 2-hour diagnostic criteria. Arch Intern Med 161: regular human insulin in treatment of type 1 and type 2 397-404. diabetes mellitus: A systematic review. Diabetes Metab 2. Coutinho M, Gerstein HC, Wang Y, Yusuf S (1999) The 37: 190-200. relationship between glucose and incident cardiovascu- 10. Ahr HJ, Boberg M, Brendel E, Krause HP, Steinke W lar events. A metaregression analysis of published data (1997) Pharmacokinetics of miglitol,. Absorption, dis- from 20 studies of 95, 783 individuals followed for 12.4 tribution, metabolism, and excretion following admin- years. Diabetes Care 22: 233-240. istration to rats, dogs, and man. Arzneimittelforschung 3. Bonora E (2002) Postprandial peaks as a risk factor for 47: 734-745. cardiovascular disease: epidemiological perspectives. 11. Scott LJ, Spencer CM (2000) Miglitol; A review of its Int J Clin Pract Suppl 129: 5-11. therapeutic potential in type 2 diabetes mellitus. Drugs 4. Hanefeld M, Fischer S, Julius U, Schulze J, Schwanebeck 59: 521-549. U, Schmechel H, Ziegelasch HJ, Lindner J (1996) Risk 12. The Committee of Japan Diabetes Society on the diag- factors for myocardial infarction and death in newly nostic criteria of diabetes mellitus (2010) Report of the detected NIDDM: the Diabetes Intervention Study, committee on the classification and diagnostic criteria 11-year follow-up. Diabetologia 39: 1577-1583. of diabetes mellitus. J Jpn Diabetes Soc 53: 450-467 (in 5. Cavalot F, Petrelli M, Traversa M, Bonomo K, Fiora Japanese). E, Conti M, Anfossi G, Costa G, Trovati M (2006) 13. Aoki K, Nakayama A, Ito S, Nezu U, Iwasaki T, Takahashi Postprandial blood glucose is a stronger predictor of M, Kimura M, Terauchi Y (2007) Administration of cardiovascular events than fasting blood glucose in type miglitol until 30 min after the start of a meal is effective 2 diabetes mellitus, particularly in women: Lessons in type 2 diabetic patients. Diabetes Res Clin Pract 78: from the Dan Luigi Gonzaga Diabetes Study. J Clin 30-33. Endocrinol Metab 91: 813-819. 14. Nemoto M, Tajima N, Kawamori R (2011) Efficacy 6. Esposito K, Ciotola M, Carleo D, Schisano B, Sardelli of combined use of miglitol in type 2 diabetic patients L, Di Tommaso D, Misso L, Saccomanno F, Ceriello A, receiving insulin therapy-placebo-controlled double- Giugliano D (2008) Post-meal glucose peaks at home blinded comparative study. Acta Diabetol 48: 15-20. associated with carotid intima-media thickness in type 15. Osonoi T, Saito M, Mochizuki K, Fukaya N, 2 diabetes. J Clin Endocrinol Metab 93: 1345-1350. Muramatsu T, Inoue S, Fuchigami M, Goda T (2010) 7. Hu Y, Liu W, Huang R, Zhang X (2010) Postchallenge The α-glucosidase inhibitor miglitol decreases glucose plasma glucose excursions, carotid intima-media thick- fluctuation and inflammatory cytokine gene expression ness, and risk factors for atherosclerosis in Chinese pop- in peripheral leukocytes of Japanese patients with type ulation with type 2 diabetes. Atherosclerosis 210: 302- 2 diabetes mellitus. Metabolism 59: 1816-1822. 306. 16. Schlichtkrull J, Munck O, Jersild M (1964) M-value, 8. American Diabetes Association (2010) Standards of an index for blood sugar control in diabetics. Ugeskr medical care in diabetes--2010. Diabetes Care 33: S11- Laeger 126: 815-820. S61. 17. Nagai E, Katsuno T, Miyagawa J, Konishi K, Miuchi M, 352 Kato et al.

Ochi F, Kusunoki Y, Tokuda M, Murai K, Hamaguchi inhibitors on endothelial function and incretin secretion T, Namba M (2011) Effects of miglitol in combination in diabetic patients with coronary artery disease. Circ J with intensive insulin therapy on blood glucose control 74: 1471-1478. with special reference to incretin responses in type 1 19. Breen DM, Giacca A (2011) Effects of insulin on the diabetes mellitus. Endocr J 58: 869-877. vasculature. Curr Vasc Pharmacol 9: 321-332. 18. Hiki M, Shimada K, Kiyanagi T, Fukao K, Hirose 20. Giovannucci E, Harlan DM, Archer MC, Bergenstal K, Ohsaka H, Fukushima Y, Kume A, Matsumori R, RM, Gapstur SM, Habel LA, Pollak M, Regensteiner Sumiyoshi K, Miyazaki T, Ohmura H, Kurata T, Miida T, JG, Yee D (2010) Diabetes and cancer: a consensus Daida H (2010) Single administration of α-glucosidase report. Diabetes Care 33: 1674-1685.