Impact of oral anticholinergic on insulin response to oral glucose load in patients with impaired glucose tolerance.

BACKGROUND: Preliminary data indicates there is a cholinergic basis to insulin secretion. AIMS &
OBJECTIVE: To investigate the impact of oral anticholinergics on insulin secretion in subjects with impaired glucose tolerance (IGT), in comparison with volunteers having normal glucose tolerance (NGT). Material &
Methods: This prospective observational study recruited 10 IGT and 10 NGT subjects. An oral glucose tolerance test (OGTT) was conducted twice in the absence and presence of hyoscine butyl-bromide (HBB). The plasma glucose (PG) and insulin levels were serially estimated at 30-min increments for 2 h after the OGTT. Early (ΔI30/ΔPG30) & late (insulin/PGAUC 60-120) phase insulin activity were assessed subsequently.
RESULTS: The study constituted of 10 IGT (4M/6F, BMI: 28.80 ± 2.30) and 10 NGT (5M/5F, BMI: 23.00 ± 0.80) subjects. In the NGT group, the pre-HBB mean glucose levels (0-120 min) were comparable with those recorded after HBB intake. However, after HBBB, the mean insulin levels decreased significantly at t = 90 and 120min, confirmed by attenuated late phase insulin activity in IGT (P = 0.023) & NGT (P = 0.006) group. On the other hand, in the IGT group, however, HBB did not impact on the mean PG and insulin levels (0-120 min).
CONCLUSIONS: Our study findings indicate that insulin secretion is influenced by cholinergic system and that oral anticholinergics may attenuate the late phase insulin activity in varying degrees of glycemic status.

Introduction

Derangement of the entero-insulin axis is an early determinant of the development of glucose intolerance in type 2 diabetes mellitus.[1] Yet, the interaction between insulin, incretins, and vagal stimulation is complex and poorly understood.[1234] In the pancreatic islets, recent experimental data suggest that endogenous acetylcholine not only stimulates β-cell function by activation of M3 and M5 receptors but also mediates recruitment of δ-cells (by activating M1 receptors) and somatostatin secretion (that results in inhibition of β-cell function).[3] On the other hand, the L-cells of terminal ileum possess M1 receptors, which upon receiving vagal stimulation increase glucagon-like peptide 1 (GLP-1) secretion.[4] While both the cells of the islets and the L-cells of terminal ileum possess muscarinic receptors, the net result of vagal stimulation (or blockade) on the entero-insulin axis after ingestion of carbohydrate-rich meal in subjects with impaired glucose tolerance (IGT) remains unexplored.

A role of vagal mediation has been suggested in the incretin-insulin axis.[56] Only a tiny fraction of the GLP-1 secreted from the terminal ileum, after passing through the portal and pulmonary vessels, reaches beta-cells unaltered through the pancreatic arteries.[6] Hence, it is unlikely that the “insulinotropic” effect of GLP-1 is entirely due to its direct stimulation of the beta-cells.[567] Instead, a role of the parasympathetic system has been proposed: GLP-1, upon secretion, stimulates the local parasympathetic fibers that, in turn, stimulate the pancreatic beta-cells to secrete insulin.[789] The physiological relevance of the results, however, is not clear because GLP-1 infusion is not equivalent to meal-induced GLP-1 secretion. It is only in the latter situation that GLP-1 can be expected to act on local parasympathetic fibers to result in vagally mediated insulin secretion.

Carbohydrate-rich meal ingestion is the most important physiological stimulation for both secretion of insulin and incretins.[1011] Hence, we wanted to explore the role of vagal stimulation in the secretion of insulin after carbohydrate ingestion, and how this effect modifies in subjects with normal and IGT. To understand such neural basis to insulin secretion, we compared the excursions in meal-induced blood glucose and insulin responses in subjects with NGT and with IGT, before and after intake of an oral anticholinergic, hyoscine butyl-bromide (HBB).

Materials and Methods

Study protocol

This single-center, observational study conformed to Helsinki II Declaration and ICMR Guidelines for Biomedical Research Involving Human Subjects. It was registered with CTRI (www.ctri.gov.in: Identifier REF/2016/11/012554). The Institutional Ethics Committee at IPGMER, Kolkata, India, approved the study.

Participants

Subjects of either sex aged between 18 and 55 years were recruited. Subjects were excluded if they had Type-2 diabetes mellitus, body mass index (BMI) ≥30 kg/m2, had contraindications for anticholinergic agents, preexisting gastrointestinal disorders, derangement of renal or hepatic functions, history of substance abuse, or were being treated with medicines that could not be paused for 12 h. Pregnant or breastfeeding women or immunocompromised subjects were also excluded from the study. We recruited a total of 20 subjects, 10 who had NGT and 10 who had IGT (as per the World Health Organization definition of IGT [2 h PG <11.1 mmol/L, but >7.8 mmol/L]).

Experimental protocol

Subjects underwent an OGTT (glucose 75 g; t = 0 min) without (day 1) and with HBB 20 mg (day 3). Two tablets of Hyocimax® 10 mg was prescribed to all study participants on day 4. The drug had to be taken 30 min before start of 75 g OGTT (t = −30 min) as single dose. Blood samples were collected at 30-min interval (0–120 min) for serial PG and insulin estimations.

Statistical analysis

Baseline and biochemical characteristics were analyzed by descriptive statistics and presented as mean ± standard deviation or standard error of mean as indicated. Normality was tested using Kolmogorov–Smirnov Goodness-of-Fit. Selected data were described with a 95% confidence interval.

Student's t-test (for parametric data) and Mann–Whitney U-test (nonparametric data) were used to compare mean values. Data at individual time points were compared using one-way repeated measures analysis of variance, followed by Tukey–Kramer post hoc analysis. Area under the concentration–time curve (AUC) over 0–120 min was calculated using the linear trapezoidal rule. Inferences were made on the estimated treatment contrast with HBB (20 mg) versus baseline data. Correlation between beta-cell function (insulinogenic index; ΔI30/ΔPG30) and mean insulin levels was analyzed in relation to outcome. Simple regression was performed between pairs of indices in each group. Fisher's test was used to compare categorical data. Two-tailed P < 0.05 was significant. GraphPad Prism version 6.0 for Windows (GraphPad Software, La Jolla, CA) was used for statistical analysis.

Results

Baseline parameters

Following the screening of 25 subjects, 5 subjects were excluded; 20 subjects including 10 IGT (4 male/6 female, BMI: 28.80 ± 2.30) and 10 NGT (5 male/5 female, BMI: 23.00 ± 0.80) met the study requirements and completed the experimental protocol. Table 1 lists the baseline characteristics of study subjects in the NGT and IGT groups. In both the groups, a higher proportion belonged to the “overweight” category.

Table 1

Baseline characteristics of subjects

Characteristics	IGT subjects (n=10)	NGT subjects (n=10)
Age (years)
Range	29-77	28-62
Age, mean±SD	49.4±16.1	45.6±13.5
Median (IQR)	45.5 (35-72)	44.5 (35-55)
Sex, n (%)
Male	4 (40)	5 (50)
Female	6 (60)	5 (50)
BMI (kg/m2), n (%)
<18	0	0
18-22.9	1 (10)	3 (30)
23-24.9	6 (60)	7 (70)
≥25	3 (30)	0
Mean±SD	28.8±2.3	23±0.8*
Family history of diabetes, n (%)	5 (50)	5 (50)
Systolic BP (mmHg)±SD	129±2	110±2*
Diastolic BP (mm Hg)±SD	78±2	76±3
Resting heart rate (bpm)±SD	68±3	72±4
Body temperature (°C)±SD	37.2±0.3	37.1±0.2
Preexisting medication	None	None

*P<0.05. SD=Standard deviation, IGT=Impaired glucose tolerance, NGT=Normal glucose tolerance, BP=Blood pressure, IQR=Interquartile range, BMI=Body mass index

Net postprandial insulin and glucose excursion

The effect of HBB in the IGT group was examined in terms of pharmacodynamic parameters obtained during a 75 g OGTT (0–120 min), as shown in Table 2. The presence of HBB did not have an impact on their fasting PG (6.70 ± 0.08 vs. 6.80 ± 0.11 mM/L, P = 0.471) and PG Cmax values (11.10 ± 0.77 vs. 11.40 ± 0.73 mM/L, P = 0.780).

Table 2

Evaluation of pharmacodynamic parameters during oral glucose tolerance test

Parameter	NGT (n=10)		P	IGT (n=10)		P
	Hyoscine (−)	Hyoscine (+)		Hyoscine (−)	Hyoscine (+)
PG
Fasting PG (mM/L)	4.6±0.13	4.7±0.32	0.775	6.7±0.08	6.8±0.11	0.471
PG Cmax (mM/L; t=60 min)	6.5±0.30*	6.9±0.64**	0.578	11.1±0.77	11.4±0.73	0.780
PG AUC (0-120 min, mM/L ×120 min)	698.4±24.59	702.4±21.31	0.903	1367.7±47.67	1384.0±64.20	0.837
PI
Fasting PI (mIU/L)	6.5±1.32	5.7±0.87	0.618	11.8±1.82	13.4±2.21	0.583
PI Cmax (mIU/L; t=60 min)	30.5±15.42	31.0±7.11	0.976	39.4±9.49	34.3±7.18	0.673
PI AUC (0-120 min, mIU/L ×120 min)	3361±192.61	2772.5±161.76	0.031*	3995.4±222.48	3863.0±210.00	0.670

*P < 0.05; **P < 0.01. IGT=Impaired glucose tolerance, AUC=Area under the concentration-time curve, calculated using trapezoidal rule, Cmax=Maximum plasma concentration, SEM=Standard error of mean, PG=Plasma glucose, PI=Plasma insulin, NGT=Normal glucose tolerance, IGT=Impaired glucose tolerance, OGTT=Oral glucose tolerance test

The effect of HBB on plasma insulin is described in Table 2. In the IGT group, the presence of HBB had no effect on fasting insulin levels (11.80 ± 1.82 vs. 13.40 ± 2.21 mIU/L) and insulin Cmax at t = 60 min (39.40 ± 9.49 vs. 34.30 ± 7.17 mIU/L). The addition of HBB also did not impact on the insulin total AUC 0–120 min (3995.00 ± 222.48 vs. 3863.00 ± 210.00 mIU/L min).

In the NGT group, similar to the IGT group, the presence of HBB did not impact on the plasma glucose-based parameters, for example, fasting PG (4.60 ± 0.13 vs. 4.70 ± 0.32 mM/L) and PG Cmax (6.50 ± 0.30 vs. 6.90 ± 0.64 mM/L) [Table 3]. Furthermore, the presence of HBB had no effect on fasting insulin levels (6.50 ± 1.32 vs. 5.70 ± 0.87 mIU/L) and insulin Cmax at t = 60 min (30.50 ± 15.42 vs. 31.00 ± 7.11 mIU/L, P = 0.976) [Table 2]. However, the addition of HBB significantly decreased the insulin total AUC 0-120 min (3361.00 ± 192.61 vs. 2772.50 ± 161.76 mIU/L FNx01 min, P = 0.031) [Table 2].

Table 3

Indices of early (insulinogenic index 0-30) and late (I/glucose area under the curve 60-120) beta-cell function calculated from oral glucose tolerance test

Arms	IGI0-30 (pM/mM)		P	I/GAUC 60-120 (pM/mM)		P
	Hyoscine (−)	Hyoscine (+)		Hyoscine (−)	Hyoscine (+)
IGT	62.1±24.20	68.9±32.87	0.869	24.0±1.12	20.70±0.72	0.023*
NGT	100.2±34.05	126.8±78.09	0.758	39.4±1.79	30.27±2.34	0.006**

*P<0.05, **P<0.01. Data expressed as mean±SEM. IGI30=Insulinogenic index (IGI0-30=ΔI0-30/ΔPG0-30), I/GAUC 60-120 (ratio of the insulin to glucose AUC from 60 to 120 min). IGT=Impaired glucose tolerance, NGT=Normal glucose tolerance, AUC=Area under the concentration, GAUC=Glucose area under the curve, IGI=Insulinogenic index, PG=Plasma glucose, SEM=Standard error of mean

Beta-cell capacity

In both the groups, the early (IGI 0–30 min) phase insulin secretion remained largely unaffected in the presence of HBB [Table 3]. However, HBB impacted on I/G AUC 60–120 in both groups, as in the IGT group, HBB caused a 1.2-fold decrease in I/G AUC 60–120 min (24.00 ± 1.12 vs. 20.70 ± 0.72 pM/mM, P = 0.023). Similarly, in the NGT group, HBB caused a 1.3-fold decrease in I/G AUC 60–120 min (39.4 ± 1.79 vs. 30.27 ± 2.34 pM/mM, P = 0.006) [Table 3].

Discussion

In summary, the administration of an anticholinergic agent (hyoscine) attenuated late-phase insulin secretion in normal subjects. This effect was not observed in IGT subjects. To our knowledge, this is the first work examining the effect of an oral anticholinergic in IGT subjects.

The effect of intravenous atropinization has been examined in euglycemic subjects with conflicting findings. Ahrén B and Holst found that atropinization results in decreased early-phase postprandial insulin secretion;[8] Plamboeck et al. in 2015, on the other hand, observed a decrease in insulin and C-peptide levels after infusion of atropine, glucose, and GLP-1.[9] Our research with administration of an oral anticholinergic shows that the effect is modified depending on whether the subject is euglycemic or has IGT. In the euglycemic, the attenuation of insulin secretion by HBB strengthens the hypothesis that the postprandial insulin secretion is vagally mediated. In IGT subjects, on the other hand, diminution of the incretin effect is a well-known early feature.[1] It is likely that since it is already attenuated in the IGT subjects, an anticholinergic agent did result in further changes. It could be also argued that with a heightened parasympathetic tone in IGT, as opposed to euglycemic subjects, the attenuation of entero-insular axis may be difficult to attain at therapeutic doses of muscarinic antagonists. Taken together, the response to the anticholinergic effect in both the euglycemic and the IGT subjects fits the hypothesis of vagal mediation of the incretin effect.

An important difference from Ahrén and Holst in our results is that the anticholinergic attenuation of insulin secretion in their experiment is limited to the early-phase and does not extend to the late-phase insulin secretion. This can be explained by the relatively late onset of action of an oral agent like HBB as opposed to the IV atropine Ahrén and Holst use.[8] Importantly, however, it highlights that at least part of the postprandial late-phase secretion of insulin is affected by incretin and that this effect is vagally mediated as well. However, this may be due to delayed gastric emptying induced by hyoscine as incretin hormone release is dependent on the rate of entry of nutrients into the small intestine and may result in a deferred beginning of the incretin effect.

Despite several study limitations, these preliminary findings are interesting as it indicates that parasympathetic system influences the entero-insulin axis in early-phase diabetes mellitus and may provide a basis for exploring novel antidiabetic drug targets.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.