| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Departments of Internal Medicine I and Neuroendocrinology, University of Luebeck, D-23538 Luebeck, Germany
Address all correspondence and requests for reprints to: Bernd Fruehwald-Schultes, M.D., Medical Department of Internal Medicine I, University Luebeck, Ratzeburger Allee 160, D-23538 Luebeck, Germany. E-mail: fruehwal{at}kfg.mu-luebeck.de
Abstract
Body weight gain and severe hypoglycemia are the major adverse effects of insulin therapy in type 2 diabetic patients. Metformin has been shown to prevent insulin therapy-induced body weight gain when used in combination with insulin. However, the effects of metformin on hormonal and symptomatic responses to hypoglycemia mediating hypoglycemia awareness have not been assessed to date. Fifteen young healthy men were treated with 850 mg metformin and placebo twice daily for a 16-d period in a double blind, cross-over design. On the last 2 d of the treatment period, the subjects underwent three hypoglycemic clamp experiments, with the first and the last performed with identical patterns of plasma glucose decrease. Differences between the effects of metformin and placebo (effect of metformin) as well as between first and last hypoglycemic clamps (effect of antecedent hypoglycemia) were assessed. Antecedent hypoglycemia significantly reduced epinephrine, ACTH, cortisol, glucagon, GH, and symptomatic responses to hypoglycemia (P < 0.05 for all variables). There was no detectable effect of metformin on epinephrine, norepinephrine, ACTH, cortisol, glucagon, or autonomic symptomatic response to hypoglycemia (P > 0.05 for all comparisons), except that metformin slightly increased the response of GH to hypoglycemia (P = 0.039). The latter finding may be due to an IGF-I-reducing effect of metformin, as after 14 d of metformin treatment baseline levels of IGF-I were significantly lower than in the placebo condition (236.9 ± 13.9 vs. 263.2 ± 14.4 µg/liter; P = 0.015). The data indicate that metformin does not adversely affect hormonal and symptomatic responses to hypoglycemia. This finding appears to be relevant with regard to the safety of the combination of metformin with insulin therapy.
HYPOGLYCEMIA IS A major clinical problem of insulin therapy, which is often required in type 2 diabetes patients to achieve good glycemic control. Although the risk of hypoglycemia appears to be much greater in patients with type 1 diabetes (1) than in patients with type 2 diabetes (2), it can be anticipated that with the increasing efforts to tighten glycemic control in type 2 diabetic patients, incidents of hypoglycemia will increase (3).
Body weight gain constitutes another major problem of insulin therapy in type 2 diabetic patients (2). Sole treatment with metformin has been shown to improve glycemic control without increasing body weight in patients with type 2 diabetes (4, 5). Furthermore, metformin decreases diabetes-related death and all-cause mortality in type 2 diabetic patients (4). The latter effect is probably a consequence of beneficial effects of metformin on several cardiovascular risk factors (5, 6, 7, 8). In type 2 diabetic patients requiring insulin therapy, two recent studies (9, 10) have indicated that adding metformin to insulin therapy further improves glycemic control without causing the body weight gain that is normally associated with insulin therapy. Although a decrease in mortality after additive metformin therapy has not yet been proven, the combination of metformin and insulin represents an attractive therapeutic approach for many patients with type 2 diabetes (8, 11).
Although metformin itself rarely causes hypoglycemia (4, 12), it may increase the risk of hypoglycemia when used in combination with insulin therapy because of its putative adverse effects on hypoglycemic counterregulation. Several animal studies (13, 14, 15, 16, 17) indicated an inhibiting effect of metformin on the sympathetic nervous system, which may represent a direct central nervous action of the drug. Thus, iv (16) and, similarly, intracerebroventricular (17) administration of metformin decreased the sympathoexcitatory response to various stressors. As the sympathoadrenal system constitutes an important part of the hypoglycemia counterregulatory system precipitating awareness of hypoglycemia (18, 19), a sympathoinhibitory effect of metformin would be of clinical relevance by increasing the risk of hypoglycemia. In light of the attractiveness of metformin in combination with insulin therapy, the question arises as to whether and to what extent metformin adversely affects the physiological responses to hypoglycemia, especially the sympathoadrenal and symptomatic responses. On this background, the present study assessed the effects of metformin on hormonal and symptomatic responses to insulin-induced hypoglycemia in humans. Moreover, the interaction of metformin with the effect of antecedent hypoglycemia was assessed given the fact that repeated occurrence of hypoglycemia is an established and clinically relevant factor that reduces hormonal and symptomatic responses to hypoglycemia (20, 21, 22, 23, 24, 25).
Subjects and Methods
Subjects
Fifteen young healthy men participated in the experiments (mean age, 26.7 yr; range, 21–33 yr). Exclusion criteria were chronic or acute illness, current medication of any kind, smoking, alcohol or drug abuse, obesity, diabetes, and hypertension in first degree relatives. Each volunteer gave written informed consent, and the study was approved by the local ethics committee.
Study design
The study was performed in a placebo-controlled, double blind, cross-over design with the order of conditions balanced across subjects. On one condition, subjects received 850 mg metformin twice daily for a 16-d period; on the other, placebo was administered with a 4-wk interval between the two treatment periods. On the last 2 d of the treatment periods (d 15 and 16) subjects underwent a total of three hypoglycemic clamps. Hormonal and symptomatic responses to hypoglycemia were assessed during the first and third clamps, inducing identical patterns of decrease in plasma glucose levels. The second hypoglycemic clamp was introduced to assure hypoglycemia unawareness on the third clamp, which took place in the morning of the following day (26).
Hypoglycemic clamp experiments
On d 15 of the treatment period (antecedent clamps), subjects
reported to the medical research unit at 0800 h and then underwent
two hypoglycemic clamps, one in the morning and another in the
afternoon. Subjects were instructed not to have breakfast on this day
and to abstain from eating until the end of the clamps. The experiments
took place in a sound-attenuated room with the subjects sitting with
the trunk in an almost upright position (
60°) and the legs in a
horizontal position on a bed. A cannula was inserted into a vein on the
back of the hand, which was placed in a heated box (50-55 C) to obtain
arterialized venous blood. A second cannula was inserted into an
antecubital vein of the contralateral arm. Both cannulas were connected
to long thin tubes that enabled blood sampling and adjustment of the
rate of dextrose infusion from an adjacent room without being noticed
by the subject. After a 30-min baseline period starting at 0900 h,
plasma glucose was lowered to a plateau of 2.8 mmol/liter by infusion
of insulin (H-insulin, Hoechst Marion Roussel, Inc.,
Frankfurt, Germany) at a rate of 1.5 mU/min·kg. Ninety minutes after
the start of insulin infusion (1100 h), plasma glucose was further
decreased to a plateau of 2.5 mmol/liter for another 60 min by
increasing the rate of insulin infusion to 2.0 mU/min·kg. Plasma
glucose levels were measured (Glucose Analyzer, Beckman Coulter, Inc., Munich, Germany) every 5 min, and a variable infusion of
20% dextrose solution was adjusted to maintain plasma glucose
plateaus.
At 1200 h, insulin infusion was stopped, and plasma glucose levels were returned to normal by dextrose infusion. At 1300 h, insulin infusion was started again at a rate of 2.0 mU/min·kg to induce the second hypoglycemia. During this hypoglycemic clamp, a plasma glucose plateau of 2.5 mmol/liter was achieved for 90 min (until 1500 h). After the clamp, the subject went home. Eating was allowed until 2200 h.
On d 16 of the treatment period subjects again reported to the research unit at 0800 h. This third clamp was performed in the same way as the first hypoglycemic clamp (starting at 0900 h).
During the baseline period and the first 120 min of the first and third hypoglycemic clamps, blood samples were drawn every 15 min for determination of insulin, epinephrine, norepinephrine, ACTH, cortisol, GH, and glucagon in plasma and serum, respectively. The hormone concentrations were measured as previously described (27). Serum IGF-I concentrations were measured by ELISA (Active IGF-I, Diagnostics Systems Laboratories, Inc., Sinsheim, Germany) with an intraassay coefficient of variance of 6.5% and an interassay coefficient of variation of 4.8%. A semiquantitative symptom questionnaire was presented every 15 min during the hypoglycemic clamps. Subjects rated from 0 (none) to 9 (severe) each of the following symptoms: tremor, hunger, palpitations, sweating, irritability, anxiety, dizziness, blurred vision, difficulty in thinking, tingling, and faintness. Consistent with the categories used by previous investigators (22), the first six symptoms were considered autonomic, and the other five were considered neuroglycopenic.
Statistical methods
All values are presented as the mean ± SEM. Statistical analysis was based on ANOVA for repeated measures, including the factors drug for metformin vs. placebo treatment and ante-hypo for the effects of antecedent hypoglycemia (first vs. third hypoglycemia). To compare the hormonal and symptomatic responses between the treatment conditions and the first and third hypoglycemic clamps, areas under the curve with reference to the baseline values were calculated. For pairwise comparisons, paired t test was applied. P < 0.05 was considered statistically significant.
Results
Plasma glucose concentrations on both first and third hypoglycemic
clamps were very similar between the metformin and placebo conditions
(Fig. 1
). Likewise, the decrease in
plasma glucose during the third and third hypoglycemic clamps was
almost identical. Serum insulin concentrations as well as the rate of
dextrose infusion required to control plasma glucose did not differ
between the metformin and placebo conditions for all hypoglycemic
clamps (data not shown). On d 15, before starting the clamp
experiments, baseline levels of serum insulin and plasma glucose were
significantly lower in the metformin than in the placebo condition
[insulin, 29.7 ± 3.1 vs. 36.3 ± 2.9 pmol/liter
(P = 0.041); glucose, 4.9 ± 0.1 vs.
5.2 ± 0.1 mmol/liter (P = 0.002)].
|
. Compared with the first hypoglycemic
clamp, counterregulatory responses of these hormones were attenuated
during the third hypoglycemic clamp. However, metformin had no effect
on the responses of epinephrine, norepinephrine, ACTH, cortisol, and
glucagon (Table 1). Serum glucagon levels
were distinctly higher in the metformin than in the placebo condition
throughout the experiments (Fig. 2), i.e. before and during
the hypoglycemic clamps. However, the hypoglycemic response of glucagon
itself was not changed by metformin (Table 1). Metformin slightly
enhanced the GH response to hypoglycemia (Fig. 2), but, this effect did
not interact with the attenuating effect of antecedent hypoglycemia
(Table 1). Supplementary determinations of serum IGF-I revealed
significantly lower levels in the metformin than the placebo condition
at baseline (236.9 ± 13.9 vs. 263.2 ± 14.4
µg/liter; P = 0.015) and the end (232.6 ± 13.4
vs. 278.4 ± 17.1 µg/liter; P <
0.001) of the first hypoglycemic clamp, whereas the difference did not
reach statistically significance during the third hypoglycemic clamp
[baseline, 265.2 ± 15.9 vs. 289.7 ± 20.5
µg/liter (P = 0.119); end, 272.3 ± 16.9
vs. 289.1 ± 22.8 µg/liter (P =
0.221)].
|
|
). Metformin slightly increased
awareness of neuroglycopenic symptoms, but had no influence on
autonomic symptoms (Table 1
).
|
The novel finding reported here is that metformin does not adversely affect hormonal and symptomatic responses to hypoglycemia. The present data confirm previous results (20, 21, 22, 23, 24, 25) showing an attenuated hormonal and symptomatic response to hypoglycemia after antecedent hypoglycemia. However, metformin does not interact with the effect of antecedent hypoglycemia on subsequent counterregulation.
Animal studies (13, 14, 15, 16, 17) have indicated a sympathoinhibitory effect of metformin with possible relevance for hypoglycemic counterregulation. Activation of the sympathoadrenal system essentially contributes to the normalization of glucose levels and also to awareness of hypoglycemia by promoting typical hypoglycemic symptoms such as sweating, tremor, and palpitations often referred to as autonomic symptoms (18, 19). However, the present data indicate that metformin in healthy men neither inhibits the sympathoadrenal nor impairs the symptomatic, especially autonomic, response to hypoglycemia. It should be pointed out that the metformin dose studied here, i.e. 1700 mg daily, was not the maximum dose used in type 2 diabetic patients, i.e. 2250 mg daily. Although it appears unlikely that a higher dose of metformin will adversely affect hypoglycemic hormonal and symptomatic responses to hypoglycemia, it should be emphasized that this possibility cannot be excluded by the present data.
Metformin enhanced the GH response to hypoglycemia. This effect is probably a consequence of a tonic reducing influence of metformin on IGF-I concentration, as IGF-I is an important inhibitory feedback signal to the pituitary in the control of GH release (28). A decreased feedback on the pituitary caused by lower IGF-I levels will increase the responsiveness of GH release on stimuli such as hypoglycemia. However, although remarkable and new, this finding appears to be of minor clinical relevance with regard to hypoglycemic counterregulation, where GH does not play a central role (29).
Serum glucagon concentrations were higher in the metformin than in the placebo condition throughout the clamps, indicating that metformin increased glucagon levels regardless of the glycemic state. There may be several explanations for this finding. Metformin could increase the glucagon concentration via its reducing influence on insulin and IGF-I release. Both insulin (30, 31) and IGF-I (28) directly suppress pancreatic glucagon secretion. Alternatively, metformin could increase circulating glucagon levels simply by impairing glucagon degradation. In the context of hypoglycemic counterregulation, it is important to note that in the present study the response of glucagon to hypoglycemia was unaffected by metformin. However, it should be pointed out that the present data do not exclude an adverse effect of metformin on glucose recovery after hypoglycemia. There is some evidence that metformin decreases hepatic glucose output by inhibiting the effects of glucagon on hepatic glucose metabolism (32, 33).
Metformin slightly, but significantly, decreased the fasting plasma glucose concentration, most likely as a consequence of diminished endogenous glucose production (8). Reduced insulin secretion and increased glucagon secretion probably represent physiological responses to decreased plasma glucose concentration, preventing a further decline in plasma glucose. The fact that the decrease in plasma glucose concentration during metformin treatment was small suggests that these mechanisms that prevent hypoglycemia, i.e. low insulin and high glucagon levels, were operative and effective. One may speculate that in advanced type 2 diabetes these mechanisms may be less operative, which could explain the occurrence of hypoglycemic episodes during metformin monotherapy in diabetic patients (4, 12). This view is supported by the finding of unchanged glucagon levels during metformin therapy in type 2 diabetic patients (34).
The present results were obtained in healthy subjects, which limits conclusions with regard to the conditions in diabetic patients. Healthy subjects were used because they show a distinctly lower variability in physiological responses to hypoglycemia than diabetic patients. In diabetic patients, the magnitude of hormonal and symptomatic responses to hypoglycemia depends on many factors, such as previous glycemic control (35, 36), recent hypoglycemic episodes (22, 23, 24, 25), and prevalence of neuropathy (37, 38, 39). These factors are difficult to sufficiently control experimentally to allow clear-cut conclusions with regard to the effects of metformin on hormonal and symptomatic responses to hypoglycemia. On the other hand, there is no evidence to date against the assumption that the influence of metformin on the clinically most important variables, i.e. sympathoadrenal and symptomatic responses to hypoglycemia, will not differ between healthy and diabetic subjects.
From a clinical point of view, the important finding here is that metformin does not impair the sympathoadrenal or symptomatic response to hypoglycemia, thereby leaving the awareness of hypoglycemia unaffected. Previous studies (40, 41, 42) have clearly indicated that hypoglycemia unawareness is an important risk factor for the experience of severe hypoglycemic episodes. The present data lend themselves to predict that metformin in combination with insulin therapy will not increase the frequency and severity of hypoglycemic episodes. Recent observations, in fact, suggest that adding metformin to insulin therapy decreases rather than increases the frequency of symptomatic and biochemical hypoglycemic episodes in patients with type 2 diabetes (9).
Acknowledgments
We thank Christiane Zinke for her expert and invaluable laboratory assistance and Anja Otterbein for her organizational work.
Received March 27, 2001.
Accepted May 17, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  S. M. Schmid, M. Hallschmid, K. Jauch-Chara, N. Bandorf, J. Born, and B. Schultes Sleep Loss Alters Basal Metabolic Hormone Secretion and Modulates the Dynamic Counterregulatory Response to Hypoglycemia J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3044 - 3051. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Schultes, K. M. Oltmanns, W. Kern, H. L. Fehm, J. Born, and A. Peters Modulation of Hunger by Plasma Glucose and Metformin J. Clin. Endocrinol. Metab., March 1, 2003; 88(3): 1133 - 1141. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) or metformin (•).
