- Email: [email protected]
Secretion of GIP in responders to acarbose in obese Type 2(NIDDM) patients
Journal of Diabetes and Its Complications 15 (2001) 245 – 249
Secretion of GIP in responders to acarbose in obese Type 2(NIDDM) patients Izumi Takei*, Kazunori Miyamoto, Osamu Funae, Norimi Ohashi, Syu Meguro, Mikiya Tokui, Takao Saruta Department of Internal Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan Received 11 July 2000; accepted 16 February 2001
Abstract Acarbose has been shown to reduce postprandial hyperglycemia and to improve lipid parameters in diabetics via its inhibitory effects on intestinal a-glucosidases. Response to acarbose may therefore be dependent upon gastric or pancreatic hormone function. To test this hypothesis, we treated 27 mild type 2 (NIDDM) Japanese diabetics who were mildly obese with low-dose acarbose (150 mg/day) for 3 months. We then performed a responder analysis to determine specific hormonal responses that may be associated with a good response to acarbose. At the end of the treatment period, a total of 15 evaluable patients was grouped as responders (n = 6) and nonresponders (n = 9) based on an effective decrease in postprandial glucose levels ( > 30 mg/day) and glycosylated hemoglobin (HbA1c) levels (>0.5%). There were no differences between the two groups in demographic variables or mean postprandial glucose levels at baseline. There was a small but significant increase in postprandial cholecystokinin (CCK) in responders, and fasting gastric inhibitory peptide (GIP) levels were significantly increased in responders and all patients after treatment. Serum leptin levels were reduced by treatment in our mildly obese responders and this was associated with a significant decrease in body weight. These results suggest that treatment with low-dose acarbose may reduce hyperglycemia in mild type 2 Japanese patients and may improve metabolic control by regulating hormones involved in glycemic control and digestive absorption. Acarbose may provide a safe adjunct to help treat insulin resistance in type 2 patients. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Type 2 diabetes; Acarbose; Obesity; Leptin; GIP
1. Introduction Acarbose potently inhibits intestinal brush border aglucosidases, thereby blocking the digestion of complex carbohydrates and disaccharide, providing a means of clinical reduction of postprandial hyperglycemia (Clissold & Edwards, 1988). Controlled clinical trials have shown that acarbose treatment decreased alimentary hyperglycemia and plasma insulin concentrations in subjects with both type 1 (IDDM) and type 2 (NIDDM) diabetes (Chiasson, Josse, Hunt, et al., 1994; Clissold & Edwards, 1988; Coniff, Shapiro, Seaton, & Bray, 1995; Hanefield, Fischer, Schulze, et al., 1991; Hoffman & Spengler, 1994; Sels, Verdonk, & Wolffenbuttel, 1998). Acarbose may
* Corresponding author. MD/AI Product Management and Scientific Information, Bayer Yakuhin, Ltd, 5-36 Miyahara 3-chome, Yodogawa-ku, Osaka 532-8577, Japan. Tel.: +81-6-6398-1131; fax: +81-6-6398-1129.
also reduce total serum triglycerides (TGs; Hoffman & Spengler, 1994; Maruhama & Goto, 1980) and in two Japanese studies has been shown to reduce total serum cholesterol (Maruhama & Goto, 1980; Shimoyama, Hori, Tamura, et al., 1982). The effects of acarbose on obesity have not been as encouraging, with it being inadequate when used alone (Atkinson, 1997). However, studies suggest that acarbose, unlike many oral antidiabetic agents, does not affect serum leptin levels and does not actually induce weight gain (Haffner, Hanefield, Fischer, Fucker, & Leonhardt, 1997). Obese humans have been shown to have increased leptin levels (Considine, Shinha, Heiman, et al., 1996), and insulin has been suggested to be a potential mediator of leptin expression. Recently, changes in leptin levels have been shown to parallel insulin changes (Haffner et al., 1997). Since acarbose affects gut mucosa function, it may potentially also influence the release of gut and pancreatic hormones. A number of studies has shown that acarbose
1056-8727/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 1 0 5 6 - 8 7 2 7 ( 0 1 ) 0 0 1 4 8 - 9
246
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249
may decrease gastric inhibitory peptide (GIP) levels (Groop, Groop, Totterman, & Fyhrquist, 1986; Uttenthal, Ukponmwan, Wood, et al., 1986); however, there is conflicting evidence on the effects of acarbose on other gastrointestinal hormones such as gastrin and cholecystokinin (CCK; Folch, 1995; Groop et al., 1986; Raptis, Dimitriades, & Hadjidkis, 1988). Based on these previous results, we hypothesized that response to acarbose may be influenced by gastric or pancreatic hormone function. To examine this hypothesis, we treated patients with mild obesity and mild type 2 with acarbose for 3 months and then performed a responder analysis to determine potential factors, in particular hormonal responses, which may be associated with a good response to acarbose.
2. Materials and methods We treated 27 mild type 2 patients, who were receiving no antihypertensive agents and who were all mildly obese [body mass index (BMI) over 24 kg/m2], with acarbose (150 mg/day) for a period of 3 months. At baseline and at 3 months of treatment, levels of glycosylated hemoglobin (HbA1c), fasting and postprandial plasma glucose and insulin, total cholesterol (TC), TG, free fatty acids (FFA), HDL-cholesterol (HDL-C), RLP-cholesterol, RLPTG, and fructosamin were measured. Postprandial blood samples were taken at 60 and 120 min after a test meal of Calorie Mate (Otsuka Pharmaceutical, Japan) with a total energy of 400 kcal, 52.8% of which was carbohydrate, 28.1% was fat, and 10.4% protein. Blood samples were stored in aliquots at 70C until assays. In addition, fasting and postprandial samples from all patients were commercially assayed for the hormones glucagon, gastrin, CCK, secretin, and GIP, and samples from eight patients were assayed for leptin. Based on these results, we performed a responder analysis, in which a responder was defined as a patient who, at the end of the 3-month acarbose treatment, had a decrease in postprandial plasma glucose level of more than 30 mg/day and a decrease in HbA1c levels of more than 0.5%. All other patients were classified as nonresponders.
Table 1 Baseline demographics of patients Parameter
All patients
Responders (n = 6)
Nonresponders (n = 9)
Sex (M/F) Age (years) Weight (kg) Height (cm) BMI Disease duration (years)
6:9 57.3 ± 10.2 68.0 ± 9.4 157.1 ± 9.8 27.5 ± 3.1 4.9 ± 4.5
1:5 56.7 ± 10.6 66.4 ± 10.7 153.2 ± 6.9 28.2 ± 3.8 6.5 ± 6.3
5:4 57.7 ± 10.0 69.1 ± 8.3 159.7 ± 10.5 27.1 ± 2.4 3.8 ± 2.1
Data are shown as mean ± S.D.
Fig. 1. Mean ± S.E.M. changes in blood glucose and HbA1c levels in mild type 2 patients after 3-month treatment with acarbose. * P < .05 vs. before treatment. ** P < .01 vs. before treatment. FSG, fasting glucose; PPG, postprandial glucose.
The paired t test was used to compare differences in patient variables before and after treatment and between responders and nonresponders. A P value of < .05 was considered significant.
3. Results A total of 15 patients was evaluable after the 3-month course of acarbose. Based on effective decrease in postprandial glucose levels (>30 mg/day) and decrease in HbA1c levels (>0.5%), patients were grouped as responders (n = 6; 1 male, 5 females; mean age 56.7 years) or nonresponders (n = 9; 5 males, 4 females; mean age 57.7 years). There were no significant differences in demographic variables between the two groups prior to treatment (Table 1). The mean postprandial glucose level in all patients at baseline before treatment was 196 mg/dl, and there were no differences between the two groups at this time. The mean fasting and postprandial blood glucose levels after the 12week treatment were both significantly decreased in the responder group (Fig. 1; P < .05 and .01, respectively). In the nonresponder group, there were no marked changes in either fasting or postprandial glucose levels after treatment. The mean HbA1c levels before and after the treatment period decreased from 7.7% to 6.9% in the responder group and increased from 7.3% to 7.5% in the nonresponder group, respectively (Fig. 1). There were no changes in mean fasting insulin levels after acarbose treatment in both the responder and nonresponder groups (Table 2). Postprandial insulin levels were decreased with treatment in both groups, but these decreases were not significant. Serum HDL-C was significantly decreased in responders after acarbose treatment ( P < .05) but was unchanged in nonresponders. TC, FFA, RLP-cholesterol, RLP-TG levels,
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249
247
Table 2 Mean changes in lipids and hormones after 3 months of treatment with acarbose All patients
Responder (n = 6)
Nonresponder (n = 9)
Parameter
Before
3 months
Before
3 months
Before
3 months
Lipids HDL-C (mg/dl) TC (mg/dl) TGs (mg/dl) FFA (mEq/l) RLP-cholesterol (mg/dl) RLP-TG (mg/dl)
49.8 ± 11.4 243.2 ± 27.6 156.5 ± 97.9 0.72 ± 0.18 8.7 ± 9.5 45.7 ± 51.5
46.9 ± 13.5 233.8 ± 28.1 156.4 ± 84.3 0.66 ± 0.17 7.6 ± 5.3 34.7 ± 31.0
54.3 ± 8.1 249.2 ± 24.6 129.0 ± 64.3 0.86 ± 0.10 6.0 ± 5.0 37.5 ± 39.0
47.0 ± 7.0 * 235.0 ± 27.0 115.8 ± 17.6 0.71 ± 0.19 4.8 ± 1.3 19.3 ± 6.0
47.0 ± 12.3 239.4 ± 28.7 173.6 ± 110.5 0.63 ± 0.16 10.4 ± 11.1 50.9 ± 57.3
46.9 ± 16.2 233.1 ± 28.8 181.8 ± 98.4 0.64 ± 0.15 9.3 ± 6.1 44.3 ± 36.0
10.5 ± 4.6 21.5 ± 8.0 82.9 ± 27.7 83.8 ± 32.1 103.2 ± 58.1 195.9 ± 173.6 10.6 ± 6.5 11.9 ± 8.1 87.4 ± 28.4 87.7 ± 26.9 169.8 ± 60.1 686.1 ± 247.0
10.7 ± 6.1 19.0 ± 8.3 82.5 ± 34.6 83.4 ± 36.1 91.0 ± 29.2 192.1 ± 183.5 12.0 ± 7.9 14.9 ± 9.8 81.3 ± 22.1 84.1 ± 21.6 213.5 ± 57.9 * 782.7 ± 324.7
11.6 ± 5.8 24.0 ± 7.6 72.0 ± 13.1 67.8 ± 11.0 117.5 ± 64.7 238.7 ± 170.2 11.3 ± 8.1 15.3 ± 11.6 92.2 ± 34.6 90.5 ± 24.8 158.5 ± 48.7 828.2 ± 279.2
11.9 ± 7.1 20.7 ± 7.1 67.5 ± 23.8 67.2 ± 24.7 97.5 ± 36.7 253.2 ± 250.2 14.9 ± 10.6 19.7 ± 12.5 * 83.7 ± 15.6 85.5 ± 20.9 233.0 ± 50.9 * 978.5 ± 420.8
9.7 ± 3.4 19.8 ± 7.7 90.1 ± 32.1 94.4 ± 36.7 93.7 ± 51.1 167.4 ± 170.0 10.2 ± 5.1 9.7 ± 2.5 84.2 ± 22.8 85.8 ± 28.0 177.3 ± 65.5 591.3 ± 165.2
9.9 ± 5.2 17.9 ± 8.8 92.4 ± 37.1 94.2 ± 38.4 86.7 ± 21.9 151.3 ± 101.1 10.1 ± 4.5 11.6 ± 5.4 79.8 ± 25.5 83.1 ± 22.0 200.6 ± 58.7 652.1 ± 122.5
Hormones Insulin (mU/ml) Glucagon (pg/ml) Gastrin (pg/ml) CCK (pg/ml) Secretin (pg/ml) GIP (pg/ml)
Fasting Postprandial Fasting Postprandial Fasting Postprandial Fasting Postprandial Fasting Postprandial Fasting Postprandial
Data are shown as mean ± S.D. * P < .05.
and fructosamin were all slightly decreased after treatment in both groups. TGs were slightly decreased in the responder group only. Postprandial CCK was significantly increased after treatment in the responders ( P = .02) but not in nonresponders (Table 2). Postprandial GIP significantly increased slightly after the treatment period in both responders and nonresponders. Postprandial gastrin, glucagon, and secretin levels were not changed after treatment in either group nor were there any significant intergroup differences in any of these hormones (Table 2). Serum leptin levels were only obtained from eight patients, three responders and five nonresponders. Mean postprandial serum leptin levels were reduced by acarbose treatment in both responders (8.2 –6.9 ng/ml) and nonresponders (7.3 – 5.6 ng/ml; P = .02). This trend was associated with a significant decrease in mean body weight in the responder group after treatment from 66.4 to 63.3 kg ( P < .001). There were no serious adverse effects as expected, considering the low dose of acarbose used. Some patients experienced mild abdominal distention, which resolved without treatment.
4. Discussion We treated mildly obese type 2 patients with acarbose for a 3-month period and then looked for determinants of a
good response to this treatment. A total of 6 out of 15 evaluable patients responded well to this low-dose treatment, showing a decrease in postprandial glucose levels of >30 mg/day by the end of the 3-month treatment period. Based on the results of previous reports, we hypothesized that response may be affected by hormones involved in glycemic control and digestive absorption. Acarbose treatment had a little effect on lipid metabolism in the responder group. HDL-C was significantly decreased in responders after 3 months, and there was also a mild but nonsignificant reduction in TG and TC in these patients. A number of studies in type 1 patients have found no effect of acarbose on TG and cholesterol (Hollander, Pi-Sunyer, & Connif, 1997; Sels et al., 1998). However, in a multicenter controlled trial of 290 type 2 patients (Coniff et al., 1995), a 24-week course of acarbose alone significantly reduced TG, TC, and HDL-C. Mertes (1998) also observed a decrease in TG and TC in type 2 patients who received acarbose for 2 years. These results suggest that the trend for decreased lipid parameters that we observed might have become significant with a longer course of treatment. The effects of acarbose on metabolic control may be more evident in patients with poor metabolic control to start with (Sels et al., 1998). In two studies of Japanese hyperlipidemic and diabetic patients, acarbose reduced TC (Maruhama & Goto, 1980; Shimoyama et al., 1982). A large study found that acarbose gave better metabolic control in patients in whom metabolic control was already
248
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249
poor (Spengler & Catagay, 1995). Our results of improved lipid metabolism in mildly obese patients on a low dose of acarbose further indicate that acarbose might be particularly suitable for metabolic control in type 2 patients. Previous studies in normal human subjects have observed decreases in insulin and gastric hormone levels, especially GIP, with acarbose treatment (Groop et al., 1986; Uttenthal et al., 1986). In our mild type 2 patients, we observed no marked changes in postprandial insulin, GIP, gastrin, glucagon, or secretin after treatment. Acarbose treatment did decrease postprandial insulin levels in both responders and nonresponders, but this effect was not significant. There was a small but significant increase in postprandial CCK in responders, and fasting GIP levels were significantly increased in responders and all patients after treatment. The former results about diabetic vascular complications and GIP show that visceral fat accumulation is related to high GIP secretion. Furthermore, obesity patients who secreted low CCK have restricted appetite because of improvement of CCK secretion by acarbose. We consider that especially obesity subjects are responders of acarbose. However, Uttenthal et al. (1986) saw a decrease in postprandial GIP in type 2 patients, but this was with long-term acarbose treatment. A potential limitation of our study was that we only measured three time points for postprandial hormone measurements. A more detailed study may identify differences in short-term rapid responses of some of these hormones that we failed to detect. Serum leptin levels were reduced in responders, and this was associated with a significant decrease in body weight in this group, which was not seen in nonresponders. While the majority of studies have failed to show that acarbose reduces body weight in diabetics, we saw a significant decrease in our mildly obese responder group after only 3 months of treatment. Insulin resistance is an important problem in mild diabetics. Our results suggest that treatment with acarbose may reduce hyperglycemia in mild type 2 patients, by effective reduction of postprandial insulin and glucose responses, and give improved metabolic control by reducing leptin, which may result in better maintenance of ideal body weight. These effects are mild but significant, which we believe indicates the use of this a-glucosidase inhibitor to help resolve insulin resistance in type 2 patients. Treatment with this a-glucosidase inhibitor therefore effectively decreased plasma glucose and increase GIP levels in the responder group. This effective control of postprandial glucose response by acarbose alone may be expected to be sustained for an extended period of treatment. Josse (1997) found that the maximal improvement in response, which occurred within 3 months of treatment of type 2 patients with acarbose, was sustained throughout the 12-month study period. The effect of acarbose on postprandial hormone responses is reversible upon cessation of
treatment (Uttenthal et al., 1986), supporting both the sustained effectiveness and safety of this agent.
References Atkinson, R. L. (1997). Recent advances in the pharmacological control of energy balance and body weight. Annals of the New York Academy of Sciences, 827, 449 – 460. Chiasson, J.-L., Josse, R. G., Hunt, J. A., Palmason, C., Rodger, N. W., Ross, S. A., Ryan, E. A., Tan, M. H., & Wolever, T. M. (1994). The efficacy of acarbose in the treatment of patients with non-insulindependent diabetes mellitus. Annals of Internal Medicine, 121, 928 – 935. Clissold, S. P., & Edwards, C. (1988). A preliminary review of its pharmacodynamic and pharmacokinetic properties and therapeutic potential. Drugs, 35, 214 – 243. Coniff, R. F., Shapiro, J. A., Seaton, T. B., & Bray, G. A. (1995). Multicenter: placebo-controlled trial comparing acarbose (BAY g 5421) with placebo, tolbutamide, and tolbutamide-plus-acarbose in non-insulin-dependent diabetes mellitus. American Journal of Medicine, 98, 443 – 451. Considine, R. V., Sinha, M. K., Heiman, M. L., Kriauciunas, A., Stephens, T. W., Nyce, M. R., Ohannesian, J. P., Marco, C. C., McKee, L. J., Bauer, T. L., & Caro, J. F. (1996). Serum immunoreactive leptin concentrations in normal weight and obese humans. New England Journal of Medicine, 334, 292 – 295. Folch, U. R. (1995). Clinical experience with acarbose as first line therapy in NIDDM. Clinical and Investigative Medicine, 18, 312 – 317. Groop, P. H., Groop, L., Totterman, K. J., & Fyhrquist, F. (1986). Effects of acarbose on the relationship between changes in GIP and insulin responses to meals in normal subjects. Acta Endocrinologica, 112, 361 – 366. Haffner, S. M., Hanefield, M., Fischer, S., Fucker, K., & Leonhardt, W. (1997). Glibenclamide, but not acarbose, increases leptin concentrations parallel to changes in insulin in subjects with NIDDM. Diabetes Care, 20, 1430 – 1434. Hanefield, M., Fischer, S., Schulze, J., Spengler, M., Wargenau, M., Schollberg, K., & Fucker, K. (1991). Therapeutic potential of acarbose as first line drug in NIDDM insufficiently treated with diet alone. Diabetes Care, 14, 732 – 737. Hoffman, J., & Spengler, M. (1994). Efficacy of 24-week monotherapy with acarbose, glibenclamide, or placebo in NIDDM patients. Diabetes Care, 17, 561 – 566. Hollander, P., Pi-Sunyer, X., & Coniff, R. F. (1997). Acarbose in the treatment of type 1 diabetes. Diabetes Care, 20, 248 – 253. Josse, R. G. (1997). Acarbose for treatment of type II diabetes: the results of a Canadian multi-center trial. Diabetes Research and Clinical Practice, 28, S139 – S145 (Suppl.). Maruhama, Y., & Goto, Y. (1980). Clinical application of a glucosidehydrolase inhibitor (Bay g 5421) to patients with hyperlipidemia and maturity onset type diabetes mellitus. Tohoku Journal of Experimental Medicine, 130, 243 – 252. Mertes, G. (1998). Efficacy and safety of acarbose in the treatment of type 2 diabetes: data from a 2-year surveillance study. Diabetes Research and Clinical Practice, 40, 63 – 70. Raptis, S., Dimitriades, G., & Hadjidkis, D. (1988). Acarbose treatment in insulin-dependent (type I) diabetes mellitus. In: Creutzfeldt W. (Ed.), Acarbose for the treatment of diabetes mellitus, 2nd international symposium on acarbose ( pp. 141 – 155). Berlin: Springer-Verlag. Sels, J. P. J. E., Verdonk, H. E. R., & Wolffenbuttel, B. H. R. (1998). Effects of acarbose (Glucobay) in persons with type 1 diabetes: a multicentre study. Diabetes Research and Clinical Practice, 41, 139 – 145. Shimoyama, T., Hori, S., Tamura, K., Tanida, N., Hosomi, M., & Wada, M. (1982). Effects of acarbose on fecal microflora of hyperlipidemic pa-
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249 tients. In: Creutzfeldt W. (Ed.), Proceedings of first international symposium on acarbose, Montreaux, October, 1981 ( pp. 123 – 136). Amsterdam: Elsevier Science (Excerpta Medica Amsterdam). Spengler, M., & Catagay, M. (1995). The use of acarbose in the primary-care setting: evaluation of efficacy and tolerability of acarbose by postmarking surveillance study. Clinical and Investigative Medicine, 18, 32 – 331.
249
Uttenthal, L. O., Ukponmwan, O. O., Wood, S. M., Ghiglione, M., Ghatei, M. A., Trayner, I. M., & Bloom, S. R. (1986). Long-term effects of intestinal alpha-glucosidase inhibition on postprandial glucose, pancreatic and gut hormone responses and fasting serum lipids in diabetics on sulphonylureas. Diabetes and Medicine, 3, 155 – 160.
Secretion of GIP in responders to acarbose in obese Type 2(NIDDM) patients Izumi Takei*, Kazunori Miyamoto, Osamu Funae, Norimi Ohashi, Syu Meguro, Mikiya Tokui, Takao Saruta Department of Internal Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan Received 11 July 2000; accepted 16 February 2001
Abstract Acarbose has been shown to reduce postprandial hyperglycemia and to improve lipid parameters in diabetics via its inhibitory effects on intestinal a-glucosidases. Response to acarbose may therefore be dependent upon gastric or pancreatic hormone function. To test this hypothesis, we treated 27 mild type 2 (NIDDM) Japanese diabetics who were mildly obese with low-dose acarbose (150 mg/day) for 3 months. We then performed a responder analysis to determine specific hormonal responses that may be associated with a good response to acarbose. At the end of the treatment period, a total of 15 evaluable patients was grouped as responders (n = 6) and nonresponders (n = 9) based on an effective decrease in postprandial glucose levels ( > 30 mg/day) and glycosylated hemoglobin (HbA1c) levels (>0.5%). There were no differences between the two groups in demographic variables or mean postprandial glucose levels at baseline. There was a small but significant increase in postprandial cholecystokinin (CCK) in responders, and fasting gastric inhibitory peptide (GIP) levels were significantly increased in responders and all patients after treatment. Serum leptin levels were reduced by treatment in our mildly obese responders and this was associated with a significant decrease in body weight. These results suggest that treatment with low-dose acarbose may reduce hyperglycemia in mild type 2 Japanese patients and may improve metabolic control by regulating hormones involved in glycemic control and digestive absorption. Acarbose may provide a safe adjunct to help treat insulin resistance in type 2 patients. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Type 2 diabetes; Acarbose; Obesity; Leptin; GIP
1. Introduction Acarbose potently inhibits intestinal brush border aglucosidases, thereby blocking the digestion of complex carbohydrates and disaccharide, providing a means of clinical reduction of postprandial hyperglycemia (Clissold & Edwards, 1988). Controlled clinical trials have shown that acarbose treatment decreased alimentary hyperglycemia and plasma insulin concentrations in subjects with both type 1 (IDDM) and type 2 (NIDDM) diabetes (Chiasson, Josse, Hunt, et al., 1994; Clissold & Edwards, 1988; Coniff, Shapiro, Seaton, & Bray, 1995; Hanefield, Fischer, Schulze, et al., 1991; Hoffman & Spengler, 1994; Sels, Verdonk, & Wolffenbuttel, 1998). Acarbose may
* Corresponding author. MD/AI Product Management and Scientific Information, Bayer Yakuhin, Ltd, 5-36 Miyahara 3-chome, Yodogawa-ku, Osaka 532-8577, Japan. Tel.: +81-6-6398-1131; fax: +81-6-6398-1129.
also reduce total serum triglycerides (TGs; Hoffman & Spengler, 1994; Maruhama & Goto, 1980) and in two Japanese studies has been shown to reduce total serum cholesterol (Maruhama & Goto, 1980; Shimoyama, Hori, Tamura, et al., 1982). The effects of acarbose on obesity have not been as encouraging, with it being inadequate when used alone (Atkinson, 1997). However, studies suggest that acarbose, unlike many oral antidiabetic agents, does not affect serum leptin levels and does not actually induce weight gain (Haffner, Hanefield, Fischer, Fucker, & Leonhardt, 1997). Obese humans have been shown to have increased leptin levels (Considine, Shinha, Heiman, et al., 1996), and insulin has been suggested to be a potential mediator of leptin expression. Recently, changes in leptin levels have been shown to parallel insulin changes (Haffner et al., 1997). Since acarbose affects gut mucosa function, it may potentially also influence the release of gut and pancreatic hormones. A number of studies has shown that acarbose
1056-8727/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 1 0 5 6 - 8 7 2 7 ( 0 1 ) 0 0 1 4 8 - 9
246
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249
may decrease gastric inhibitory peptide (GIP) levels (Groop, Groop, Totterman, & Fyhrquist, 1986; Uttenthal, Ukponmwan, Wood, et al., 1986); however, there is conflicting evidence on the effects of acarbose on other gastrointestinal hormones such as gastrin and cholecystokinin (CCK; Folch, 1995; Groop et al., 1986; Raptis, Dimitriades, & Hadjidkis, 1988). Based on these previous results, we hypothesized that response to acarbose may be influenced by gastric or pancreatic hormone function. To examine this hypothesis, we treated patients with mild obesity and mild type 2 with acarbose for 3 months and then performed a responder analysis to determine potential factors, in particular hormonal responses, which may be associated with a good response to acarbose.
2. Materials and methods We treated 27 mild type 2 patients, who were receiving no antihypertensive agents and who were all mildly obese [body mass index (BMI) over 24 kg/m2], with acarbose (150 mg/day) for a period of 3 months. At baseline and at 3 months of treatment, levels of glycosylated hemoglobin (HbA1c), fasting and postprandial plasma glucose and insulin, total cholesterol (TC), TG, free fatty acids (FFA), HDL-cholesterol (HDL-C), RLP-cholesterol, RLPTG, and fructosamin were measured. Postprandial blood samples were taken at 60 and 120 min after a test meal of Calorie Mate (Otsuka Pharmaceutical, Japan) with a total energy of 400 kcal, 52.8% of which was carbohydrate, 28.1% was fat, and 10.4% protein. Blood samples were stored in aliquots at 70C until assays. In addition, fasting and postprandial samples from all patients were commercially assayed for the hormones glucagon, gastrin, CCK, secretin, and GIP, and samples from eight patients were assayed for leptin. Based on these results, we performed a responder analysis, in which a responder was defined as a patient who, at the end of the 3-month acarbose treatment, had a decrease in postprandial plasma glucose level of more than 30 mg/day and a decrease in HbA1c levels of more than 0.5%. All other patients were classified as nonresponders.
Table 1 Baseline demographics of patients Parameter
All patients
Responders (n = 6)
Nonresponders (n = 9)
Sex (M/F) Age (years) Weight (kg) Height (cm) BMI Disease duration (years)
6:9 57.3 ± 10.2 68.0 ± 9.4 157.1 ± 9.8 27.5 ± 3.1 4.9 ± 4.5
1:5 56.7 ± 10.6 66.4 ± 10.7 153.2 ± 6.9 28.2 ± 3.8 6.5 ± 6.3
5:4 57.7 ± 10.0 69.1 ± 8.3 159.7 ± 10.5 27.1 ± 2.4 3.8 ± 2.1
Data are shown as mean ± S.D.
Fig. 1. Mean ± S.E.M. changes in blood glucose and HbA1c levels in mild type 2 patients after 3-month treatment with acarbose. * P < .05 vs. before treatment. ** P < .01 vs. before treatment. FSG, fasting glucose; PPG, postprandial glucose.
The paired t test was used to compare differences in patient variables before and after treatment and between responders and nonresponders. A P value of < .05 was considered significant.
3. Results A total of 15 patients was evaluable after the 3-month course of acarbose. Based on effective decrease in postprandial glucose levels (>30 mg/day) and decrease in HbA1c levels (>0.5%), patients were grouped as responders (n = 6; 1 male, 5 females; mean age 56.7 years) or nonresponders (n = 9; 5 males, 4 females; mean age 57.7 years). There were no significant differences in demographic variables between the two groups prior to treatment (Table 1). The mean postprandial glucose level in all patients at baseline before treatment was 196 mg/dl, and there were no differences between the two groups at this time. The mean fasting and postprandial blood glucose levels after the 12week treatment were both significantly decreased in the responder group (Fig. 1; P < .05 and .01, respectively). In the nonresponder group, there were no marked changes in either fasting or postprandial glucose levels after treatment. The mean HbA1c levels before and after the treatment period decreased from 7.7% to 6.9% in the responder group and increased from 7.3% to 7.5% in the nonresponder group, respectively (Fig. 1). There were no changes in mean fasting insulin levels after acarbose treatment in both the responder and nonresponder groups (Table 2). Postprandial insulin levels were decreased with treatment in both groups, but these decreases were not significant. Serum HDL-C was significantly decreased in responders after acarbose treatment ( P < .05) but was unchanged in nonresponders. TC, FFA, RLP-cholesterol, RLP-TG levels,
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249
247
Table 2 Mean changes in lipids and hormones after 3 months of treatment with acarbose All patients
Responder (n = 6)
Nonresponder (n = 9)
Parameter
Before
3 months
Before
3 months
Before
3 months
Lipids HDL-C (mg/dl) TC (mg/dl) TGs (mg/dl) FFA (mEq/l) RLP-cholesterol (mg/dl) RLP-TG (mg/dl)
49.8 ± 11.4 243.2 ± 27.6 156.5 ± 97.9 0.72 ± 0.18 8.7 ± 9.5 45.7 ± 51.5
46.9 ± 13.5 233.8 ± 28.1 156.4 ± 84.3 0.66 ± 0.17 7.6 ± 5.3 34.7 ± 31.0
54.3 ± 8.1 249.2 ± 24.6 129.0 ± 64.3 0.86 ± 0.10 6.0 ± 5.0 37.5 ± 39.0
47.0 ± 7.0 * 235.0 ± 27.0 115.8 ± 17.6 0.71 ± 0.19 4.8 ± 1.3 19.3 ± 6.0
47.0 ± 12.3 239.4 ± 28.7 173.6 ± 110.5 0.63 ± 0.16 10.4 ± 11.1 50.9 ± 57.3
46.9 ± 16.2 233.1 ± 28.8 181.8 ± 98.4 0.64 ± 0.15 9.3 ± 6.1 44.3 ± 36.0
10.5 ± 4.6 21.5 ± 8.0 82.9 ± 27.7 83.8 ± 32.1 103.2 ± 58.1 195.9 ± 173.6 10.6 ± 6.5 11.9 ± 8.1 87.4 ± 28.4 87.7 ± 26.9 169.8 ± 60.1 686.1 ± 247.0
10.7 ± 6.1 19.0 ± 8.3 82.5 ± 34.6 83.4 ± 36.1 91.0 ± 29.2 192.1 ± 183.5 12.0 ± 7.9 14.9 ± 9.8 81.3 ± 22.1 84.1 ± 21.6 213.5 ± 57.9 * 782.7 ± 324.7
11.6 ± 5.8 24.0 ± 7.6 72.0 ± 13.1 67.8 ± 11.0 117.5 ± 64.7 238.7 ± 170.2 11.3 ± 8.1 15.3 ± 11.6 92.2 ± 34.6 90.5 ± 24.8 158.5 ± 48.7 828.2 ± 279.2
11.9 ± 7.1 20.7 ± 7.1 67.5 ± 23.8 67.2 ± 24.7 97.5 ± 36.7 253.2 ± 250.2 14.9 ± 10.6 19.7 ± 12.5 * 83.7 ± 15.6 85.5 ± 20.9 233.0 ± 50.9 * 978.5 ± 420.8
9.7 ± 3.4 19.8 ± 7.7 90.1 ± 32.1 94.4 ± 36.7 93.7 ± 51.1 167.4 ± 170.0 10.2 ± 5.1 9.7 ± 2.5 84.2 ± 22.8 85.8 ± 28.0 177.3 ± 65.5 591.3 ± 165.2
9.9 ± 5.2 17.9 ± 8.8 92.4 ± 37.1 94.2 ± 38.4 86.7 ± 21.9 151.3 ± 101.1 10.1 ± 4.5 11.6 ± 5.4 79.8 ± 25.5 83.1 ± 22.0 200.6 ± 58.7 652.1 ± 122.5
Hormones Insulin (mU/ml) Glucagon (pg/ml) Gastrin (pg/ml) CCK (pg/ml) Secretin (pg/ml) GIP (pg/ml)
Fasting Postprandial Fasting Postprandial Fasting Postprandial Fasting Postprandial Fasting Postprandial Fasting Postprandial
Data are shown as mean ± S.D. * P < .05.
and fructosamin were all slightly decreased after treatment in both groups. TGs were slightly decreased in the responder group only. Postprandial CCK was significantly increased after treatment in the responders ( P = .02) but not in nonresponders (Table 2). Postprandial GIP significantly increased slightly after the treatment period in both responders and nonresponders. Postprandial gastrin, glucagon, and secretin levels were not changed after treatment in either group nor were there any significant intergroup differences in any of these hormones (Table 2). Serum leptin levels were only obtained from eight patients, three responders and five nonresponders. Mean postprandial serum leptin levels were reduced by acarbose treatment in both responders (8.2 –6.9 ng/ml) and nonresponders (7.3 – 5.6 ng/ml; P = .02). This trend was associated with a significant decrease in mean body weight in the responder group after treatment from 66.4 to 63.3 kg ( P < .001). There were no serious adverse effects as expected, considering the low dose of acarbose used. Some patients experienced mild abdominal distention, which resolved without treatment.
4. Discussion We treated mildly obese type 2 patients with acarbose for a 3-month period and then looked for determinants of a
good response to this treatment. A total of 6 out of 15 evaluable patients responded well to this low-dose treatment, showing a decrease in postprandial glucose levels of >30 mg/day by the end of the 3-month treatment period. Based on the results of previous reports, we hypothesized that response may be affected by hormones involved in glycemic control and digestive absorption. Acarbose treatment had a little effect on lipid metabolism in the responder group. HDL-C was significantly decreased in responders after 3 months, and there was also a mild but nonsignificant reduction in TG and TC in these patients. A number of studies in type 1 patients have found no effect of acarbose on TG and cholesterol (Hollander, Pi-Sunyer, & Connif, 1997; Sels et al., 1998). However, in a multicenter controlled trial of 290 type 2 patients (Coniff et al., 1995), a 24-week course of acarbose alone significantly reduced TG, TC, and HDL-C. Mertes (1998) also observed a decrease in TG and TC in type 2 patients who received acarbose for 2 years. These results suggest that the trend for decreased lipid parameters that we observed might have become significant with a longer course of treatment. The effects of acarbose on metabolic control may be more evident in patients with poor metabolic control to start with (Sels et al., 1998). In two studies of Japanese hyperlipidemic and diabetic patients, acarbose reduced TC (Maruhama & Goto, 1980; Shimoyama et al., 1982). A large study found that acarbose gave better metabolic control in patients in whom metabolic control was already
248
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249
poor (Spengler & Catagay, 1995). Our results of improved lipid metabolism in mildly obese patients on a low dose of acarbose further indicate that acarbose might be particularly suitable for metabolic control in type 2 patients. Previous studies in normal human subjects have observed decreases in insulin and gastric hormone levels, especially GIP, with acarbose treatment (Groop et al., 1986; Uttenthal et al., 1986). In our mild type 2 patients, we observed no marked changes in postprandial insulin, GIP, gastrin, glucagon, or secretin after treatment. Acarbose treatment did decrease postprandial insulin levels in both responders and nonresponders, but this effect was not significant. There was a small but significant increase in postprandial CCK in responders, and fasting GIP levels were significantly increased in responders and all patients after treatment. The former results about diabetic vascular complications and GIP show that visceral fat accumulation is related to high GIP secretion. Furthermore, obesity patients who secreted low CCK have restricted appetite because of improvement of CCK secretion by acarbose. We consider that especially obesity subjects are responders of acarbose. However, Uttenthal et al. (1986) saw a decrease in postprandial GIP in type 2 patients, but this was with long-term acarbose treatment. A potential limitation of our study was that we only measured three time points for postprandial hormone measurements. A more detailed study may identify differences in short-term rapid responses of some of these hormones that we failed to detect. Serum leptin levels were reduced in responders, and this was associated with a significant decrease in body weight in this group, which was not seen in nonresponders. While the majority of studies have failed to show that acarbose reduces body weight in diabetics, we saw a significant decrease in our mildly obese responder group after only 3 months of treatment. Insulin resistance is an important problem in mild diabetics. Our results suggest that treatment with acarbose may reduce hyperglycemia in mild type 2 patients, by effective reduction of postprandial insulin and glucose responses, and give improved metabolic control by reducing leptin, which may result in better maintenance of ideal body weight. These effects are mild but significant, which we believe indicates the use of this a-glucosidase inhibitor to help resolve insulin resistance in type 2 patients. Treatment with this a-glucosidase inhibitor therefore effectively decreased plasma glucose and increase GIP levels in the responder group. This effective control of postprandial glucose response by acarbose alone may be expected to be sustained for an extended period of treatment. Josse (1997) found that the maximal improvement in response, which occurred within 3 months of treatment of type 2 patients with acarbose, was sustained throughout the 12-month study period. The effect of acarbose on postprandial hormone responses is reversible upon cessation of
treatment (Uttenthal et al., 1986), supporting both the sustained effectiveness and safety of this agent.
References Atkinson, R. L. (1997). Recent advances in the pharmacological control of energy balance and body weight. Annals of the New York Academy of Sciences, 827, 449 – 460. Chiasson, J.-L., Josse, R. G., Hunt, J. A., Palmason, C., Rodger, N. W., Ross, S. A., Ryan, E. A., Tan, M. H., & Wolever, T. M. (1994). The efficacy of acarbose in the treatment of patients with non-insulindependent diabetes mellitus. Annals of Internal Medicine, 121, 928 – 935. Clissold, S. P., & Edwards, C. (1988). A preliminary review of its pharmacodynamic and pharmacokinetic properties and therapeutic potential. Drugs, 35, 214 – 243. Coniff, R. F., Shapiro, J. A., Seaton, T. B., & Bray, G. A. (1995). Multicenter: placebo-controlled trial comparing acarbose (BAY g 5421) with placebo, tolbutamide, and tolbutamide-plus-acarbose in non-insulin-dependent diabetes mellitus. American Journal of Medicine, 98, 443 – 451. Considine, R. V., Sinha, M. K., Heiman, M. L., Kriauciunas, A., Stephens, T. W., Nyce, M. R., Ohannesian, J. P., Marco, C. C., McKee, L. J., Bauer, T. L., & Caro, J. F. (1996). Serum immunoreactive leptin concentrations in normal weight and obese humans. New England Journal of Medicine, 334, 292 – 295. Folch, U. R. (1995). Clinical experience with acarbose as first line therapy in NIDDM. Clinical and Investigative Medicine, 18, 312 – 317. Groop, P. H., Groop, L., Totterman, K. J., & Fyhrquist, F. (1986). Effects of acarbose on the relationship between changes in GIP and insulin responses to meals in normal subjects. Acta Endocrinologica, 112, 361 – 366. Haffner, S. M., Hanefield, M., Fischer, S., Fucker, K., & Leonhardt, W. (1997). Glibenclamide, but not acarbose, increases leptin concentrations parallel to changes in insulin in subjects with NIDDM. Diabetes Care, 20, 1430 – 1434. Hanefield, M., Fischer, S., Schulze, J., Spengler, M., Wargenau, M., Schollberg, K., & Fucker, K. (1991). Therapeutic potential of acarbose as first line drug in NIDDM insufficiently treated with diet alone. Diabetes Care, 14, 732 – 737. Hoffman, J., & Spengler, M. (1994). Efficacy of 24-week monotherapy with acarbose, glibenclamide, or placebo in NIDDM patients. Diabetes Care, 17, 561 – 566. Hollander, P., Pi-Sunyer, X., & Coniff, R. F. (1997). Acarbose in the treatment of type 1 diabetes. Diabetes Care, 20, 248 – 253. Josse, R. G. (1997). Acarbose for treatment of type II diabetes: the results of a Canadian multi-center trial. Diabetes Research and Clinical Practice, 28, S139 – S145 (Suppl.). Maruhama, Y., & Goto, Y. (1980). Clinical application of a glucosidehydrolase inhibitor (Bay g 5421) to patients with hyperlipidemia and maturity onset type diabetes mellitus. Tohoku Journal of Experimental Medicine, 130, 243 – 252. Mertes, G. (1998). Efficacy and safety of acarbose in the treatment of type 2 diabetes: data from a 2-year surveillance study. Diabetes Research and Clinical Practice, 40, 63 – 70. Raptis, S., Dimitriades, G., & Hadjidkis, D. (1988). Acarbose treatment in insulin-dependent (type I) diabetes mellitus. In: Creutzfeldt W. (Ed.), Acarbose for the treatment of diabetes mellitus, 2nd international symposium on acarbose ( pp. 141 – 155). Berlin: Springer-Verlag. Sels, J. P. J. E., Verdonk, H. E. R., & Wolffenbuttel, B. H. R. (1998). Effects of acarbose (Glucobay) in persons with type 1 diabetes: a multicentre study. Diabetes Research and Clinical Practice, 41, 139 – 145. Shimoyama, T., Hori, S., Tamura, K., Tanida, N., Hosomi, M., & Wada, M. (1982). Effects of acarbose on fecal microflora of hyperlipidemic pa-
I. Takei et al. / Journal of Diabetes and Its Complications 15 (2001) 245–249 tients. In: Creutzfeldt W. (Ed.), Proceedings of first international symposium on acarbose, Montreaux, October, 1981 ( pp. 123 – 136). Amsterdam: Elsevier Science (Excerpta Medica Amsterdam). Spengler, M., & Catagay, M. (1995). The use of acarbose in the primary-care setting: evaluation of efficacy and tolerability of acarbose by postmarking surveillance study. Clinical and Investigative Medicine, 18, 32 – 331.
249
Uttenthal, L. O., Ukponmwan, O. O., Wood, S. M., Ghiglione, M., Ghatei, M. A., Trayner, I. M., & Bloom, S. R. (1986). Long-term effects of intestinal alpha-glucosidase inhibition on postprandial glucose, pancreatic and gut hormone responses and fasting serum lipids in diabetics on sulphonylureas. Diabetes and Medicine, 3, 155 – 160.