Rectal sensitivity correlated with gastrointestinal‐mediated glucose disposal, but not the incretin effect

Abstract Objective The mechanisms behind the diminished incretin effect in type 2 diabetes are uncertain, but impaired vagal transmission has been suggested. We aimed to investigate the association between the incretin effect and autonomic neuropathy, and the degree of dysglycaemia and duration of diabetes. Design and Methods For a cross‐sectional study, we included participants with either longstanding type 2 diabetes, recent onset, untreated diabetes and controls without diabetes matched for age, sex and body mass index. Autonomic nerve function was assessed with cardiovascular reflex tests, heart rate variability and sudomotor function. Visceral afferent nerves in the gut were tested performing rapid rectal balloon distention. An oral glucose tolerance test and an intravenous isoglycaemic glucose infusion were performed to calculate the incretin effect and gastrointestinal‐mediated glucose disposal (GIGD). Results Sixty‐five participants were recruited. Participants with diabetes had rectal hyposensitivity for earliest sensation (3.7 ± 1.1 kPa in longstanding, 4.0 ± 1.3 in early), compared to controls (3.0 ± 0.9 kPa), p = .005. Rectal hyposensitivity for earliest sensation was not associated with the incretin effect (rho = −0.204, p = .106), but an association was found with GIGD (rho −0.341, p = .005). Incretin effect and GIGD were correlated with all glucose values, HbA1c and duration of diabetes. Conclusions Rectal hyposensitivity was uncovered in both longstanding and early type 2 diabetes, and was not associated with the incretin effect, but with GIGD, implying a potential link between visceral neuropathy and gastrointestinal handling of glucose. Both the incretin effect and GIGD were associated with the degree of dysglycaemia and the duration of diabetes. Previously Published Some of the data have previously been published and presented as a poster on the American Diabetes Association 83rd Scientific Sessions: Meling et al; 1658‐P: Rectal Hyposensitivity, a Potential Marker of Enteric Autonomic Nerve Dysfunction, Is Significantly Associated with Gastrointestinally Mediated Glucose Disposal in Persons with Type 2 Diabetes. Diabetes 20 June 2023; 72 (Supplement_1): 1658–P. https://doi.org/10.2337/db23‐1658‐P.


| INTRODUC TI ON
The incretin effect refers to the amplified insulin secretion when glucose is ingested orally compared to administered intravenously. 1 The two main incretin hormones involved, glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) are secreted from enteroendocrine cells in response to food ingestion.
It is well established that the incretin effect plays an important role in normal glucose tolerance, and also that a diminished incretin effect is one of several pathomechanistic components of type 2 diabetes. 2Importantly, however, there is uncertainty as to whether the diminished incretin response in type 2 diabetes is a cause or a consequence of the condition. 3It is also unknown why the effect is diminished, with several studies having concluded that patients with type 2 diabetes have normal levels of total GLP-1 responses during oral glucose or liquid meal tests. 4,5Thus, other underlying mechanisms should explain the diminished incretin effect.
Bioactive GLP-1 has a half-life of less than 2 min, with ≥50% of secreted GLP-1 being inactivated before entering systemic circulation.This rapid degradation has led to the hypothesis that GLP-1 may act locally at vagal nerve terminals innervating the portal vein, before being degraded.There is also support for the activation of sensory vagal afferents expressing GLP-1 receptors found in the gut, with transmission to the nucleus of the solitary tract in the brain stem, further reaching the hypothalamus and thus influencing central autonomic regulation. 6Different lesions at these sites of vagal transmission have been shown to abolish satiation, increase meal size, induce postprandial hyperglycaemia and increase gastric emptying rate. 7,8Until recently, the effects of GLP-1 in human physiology were the most extensively studied; however, recently GIP has been indicated as perhaps the most important for postprandial glycaemic control. 9Evidence also support that endogenous GIP has a progressive loss of insulinotropic and glucose-lowering actions from health to obesity to type 2 diabetes. 10deeper understanding of the underlying mechanisms of the diminished incretin effect could lead to new treatment targets, and provide an understanding that could help in treatment individualisation.With an increased awareness of autonomic neuropathy in the early stages of diabetes, even in prediabetes, the main aim of the current study was to investigate if autonomic neuropathy was associated with the incretin effect. 11,12Secondary aims were to investigate associations between the incretin effect, degree of dysglycaemia and duration of diabetes.

| MATERIAL S AND ME THODS
A cross-sectional study was conducted including one group of people with type 2 diabetes for more than 10 years (longstanding diabetes group), one group with recently diagnosed (within 1 year), untreated type 2 diabetes (early diabetes group) and one group without diabetes (control group).Participants were recruited mainly through regional newspaper advertisements.Those with longstanding diabetes were classified based on self-reported duration, and an oral glucose tolerance test (OGTT) confirmed the compared to controls (3.0 ± 0.9 kPa), p = .005.Rectal hyposensitivity for earliest sensation was not associated with the incretin effect (rho = −0.204,p = .106),but an association was found with GIGD (rho −0.341, p = .005).Incretin effect and GIGD were correlated with all glucose values, HbA1c and duration of diabetes.

Conclusions:
Rectal hyposensitivity was uncovered in both longstanding and early type 2 diabetes, and was not associated with the incretin effect, but with GIGD, implying a potential link between visceral neuropathy and gastrointestinal handling of glucose.Both the incretin effect and GIGD were associated with the degree of dysglycaemia and the duration of diabetes.

K E Y W O R D S
diabetic autonomic neuropathy, evoked potentials, gastrointestinal-mediated glucose disposal, incretin effect, rectal balloon distention, rectal hyposensitivity, type 2 diabetes diagnosis.The two latter groups were recruited to match the longstanding diabetes group for age, sex and body mass index (BMI), and participants were further classified into early diabetes or controls according to the American Diabetes Association (ADA) criteria based on the OGTT. 13All investigations were performed at a single centre (Bergen, Norway).Exclusion criteria were major abdominal surgery, rectosigmoid disease interfering with sensitivity, chronic pancreatitis, uremic condition (eGFR<30 mL/min), atrial fibrillation or other major dysrhythmias, cardiac pacemaker, proliferative retinopathy or present use of a GLP-1 receptor agonist or insulin.
Participants were instructed to avoid alcohol, sedatives or stimulant drugs 24 h before the nerve tests.Other medications were taken as normal.Food, coffee, tea and nicotine were avoided 3 h before the examination.The participants fasted, including medications and nicotine products, for a minimum of 10 h before the OGTT and intravenous isoglycaemic glucose infusion (IIGI) with antidiabetic medication withdrawn 3 days ahead.

| Neuronal phenotyping
For cardiovascular reflex tests (CARTs) and heart rate variability (HRV) scores, we used the Vagus™ Device (Medicus Engineering, Aarhus, Denmark).Blood pressure was measured after 5 min rest, in response to standing, and after 1-and 3 min standing, using the WelchAllyn Connex ProBP 3400™ (EMEAI, Leiden, The Netherlands).
Orthostatic hypotension was defined as a drop in systolic blood pressure >20 mmHg or diastolic blood pressure >10 mmHg.Stages of cardiovascular autonomic neuropathy (CAN) based on CARTs were defined as borderline if one ratio was abnormal and as definite or confirmed if two or three ratios were abnormal.CAN was defined as severe or advanced if the latter was combined with orthostatic hypotension. 11e Sudoscan™ device (Impeto Medical, Paris, France) was used to test the sudomotor function of hands and feet.The device reports absolute values for electrochemical skin conductance and categorical interpretations into normal function, moderately reduced or severely reduced function.

| Oral glucose tolerance test
A cannula was placed in a cubital vein, with the forearm on the same side placed in a heating cuff to ensure arterialised blood.Fasting

| Intravenous isoglycaemic glucose infusion
Two cannulas were placed in cubital veins, one in each arm.Blood was sampled, and glucose was measured at the same intervals as the of OGTT (μmol/L × min at 70 min) was calculated as a proxy for gastric emptying. 14

| Visceral sensitivity: Evoked brain potentials following rapid rectal balloon distention
6][17] Figure 1 shows the setup for the examination.A rectal balloon was placed 15 cm above the anal verge.A distinct and short stimulus was used; 150 ms inflation, followed by an instant deflation.A random interstimulus interval of 8 ± 2 s was enforced.Recorded outcomes were pressure needed to induce the earliest sensation and unpleasant threshold, and latency and amplitude of the evoked potentials.

| Data analysis and statistics
In this explorative pilot project, a formal power analysis was unfeasible, mainly due to the unknown effect size for the rapid rectal balloon distention test.However, using effect sizes from previous studies on visceral sensitivity, the lowest number of participants needed to achieve a power of 0.8 would be 15 participants in each group. 18,19rmally distributed data are provided as means with SD and skewed data as median with IQR.If equal variance was present, we used TA B L E 1 Clinical characteristics at baseline.Note: Data are provided as percentage, mean ± standard deviation or median with (interquartile range).Two participants were excluded from the analysis of the incretin effect, one due to an OGGT/IIGI glucose AUC difference above 20%, and one due to errors in c-peptide measurements, both from the longstanding diabetes group.The reduction in the number of participants performing CARTs where mainly due to participants being unable to perform the Valsalva manoeuvre.RS ratio, the ratio between maximum HR within the first 15 s after standing up and minimum HR within the first 30 s after standing up; EI-ratio, the mean ratio between the longest and shortest RR-interval during deep respiration; VM ratio, the ratio between maximum heart rate at the end of forced respiration and minimal heart rate during inspiration-expiration in rest appr.30 s after releasing pressure.

| Subjects
We recruited a total of 65 participants with a mean age of 69 years, 52% women, all Caucasians.For clinical characteristics, see Table 1.

| Between-group differences
Neuronal phenotyping results have been previously published, but are summarily presented, as they are needed for further analyses in the present paper. 17rticipants with diabetes showed rectal hyposensitivity, evident as higher pressure needed to elicit earliest sensation: Longstanding (3.7 ± 1.1 kPa); early (4.0 ± 1.3 kPa) compared to controls (3.0 ± 0.9 kPa), p = .005.There was no difference in pressure required to reach an unpleasant threshold, nor in evoked potential latencies or amplitudes.
Neuronal phenotyping revealed two cases of definite CAN.
Proportions of definite or borderline CAN were as follows: longstanding diabetes 31%, early diabetes 23% and 17% of controls, p =0.537.No between-group differences were found for HRV or sudomotor function.
The fasting and 2-h glucose levels following the OGGT, and the HbA1c differed significantly between groups (Table 2).The mean differences between glucose in the OGTT and IIGI are shown in Figure 2A-C.
Paracetamol AUC concentration was significantly higher for those with longstanding diabetes compared to controls (Figure 4

| Clinical correlations
Rectal sensitivity for the earliest sensation was not correlated with the incretin effect (rho = −0.204,p = .106),but with GIGD (rho −0.341, p = .005).There was no difference in the incretin effect for those with rectal pressure for the earliest sensation <3.0 kPa

| DISCUSS ION
Rectal hyposensitivity, uncovered in both longstanding and early diabetes, was not associated with the incretin effect but was associated with GIGD.Additionally, an association was confirmed between both the incretin effect and GIGD, and dysglycaemia and duration of diabetes.The findings indicate a possible role for autonomic nerve dysfunction in the impact of gastrointestinal factors on glucose disposal other than the incretin effect.

| Gut hyposensitivity in early stages of diabetes
We have previously reported and discussed the findings of rectal hyposensitivity in early stages of diabetes, indicating early onset of visceral hyposensitivity in the gut wall, supporting the presence of distorted neuronal transmission. 17

| No association with the incretin effect, but with gastrointestinal-mediated glucose disposal
We did not find any association between visceral sensation and the incretin effect.This is indeed interesting as it could indicate that intact neuronal transmission is not a prerequisite for the incretin effect.However, reduced rectal sensation was associated with GIGD, which may be a highly relevant, but incompletely mapped, physiological parameter.GIGD is to an extent dependent on incretin hormones for insulin release, but also on factors such as glucagon secretion, first-pass hepatic glucose uptake from the portal vein, tissue glucose uptake, hepatic insulin uptake, gastric emptying and gut motility, and activity on the gut-brain or liver-brain axis, including local afferent nerves. 6,7,20,21Decoding which of the remaining factors mediated the correlation with visceral hyposensitivity is a potentially important area of future research and yields the possibility of discovering new relevant factors, that may serve as future therapeutical targets.
One possible explanation for the discrepancy between associations for rectal hyposensitivity with GIGD and the incretin effect could be that despite the limited amount of GLP-1 reaching circulation, it could be sufficient to elicit a proper incretin effect.This is supported by evidence of a preserved incretin effect after complete surgical denervation of the pancreas following pancreas transplantation. 22Interpolation of these results suggests that despite an adequate incretin effect, GIGD was reduced in the transplanted group. 6minished GIGD, but preserved incretin effect has earlier been reported in truncal vagotomised subjects, with higher early GIP response, higher peak intact GLP-1 levels and faster gastric emptying in those who were vagotomised. 6The fact that both GIP and GLP-1 are found to increase in vagotomised subjects opens a hypothesis of increased hormone secretion, compensating for a reduced effect through neural transmission.
In the vagotomised subjects, a lack of glucagon suppression was reported following OGTT, similar to early steps in the development of type 2 diabetes. 23This is supported by a study where lean people with well-controlled type 2 diabetes were found to have normal incretin effect but compromised GIGD, associated with hyperglucagonaemia after OGTT. 24Moreover, a study in pancreatectomised men has revealed that postprandial hyperglucagonaemia may be a gut-dependent phenomenon. 25possible association with GIGD, but not the incretin effect, could also point to a more important role for neuronal transmission in the central mechanisms of GLP-1.This is supported by the abovementioned cohort of vagotomised subjects, where exogenous GLP-1 did not have an effect on ad libitum eating or gastric emptying compared to controls, but showed the same reduction in postprandial glucose. 26other proposed mechanism for the reduction of GIGD related to parasympathetic withdrawal was shown in people with cirrhosis, characterised by decreased hepatic first-pass glucose uptake. 27In support of this, a review has suggested the presence of abnormal sympathetic dominance in prediabetes. 12her possible explanations for a lesser role of intact vagal transmission in the incretin effect include the fact that GIP does not exhibit the same dependence on neuronal transmission as GLP-1 and that GIP may be more central to the incretin effect than initially postulated. 7,9I G U R E 4 Paracetamol absorption test for the different groups.AUC at 70 min, μmol/L: Longstanding diabetes; 3868 ± 1320, early diabetes; 3256 ± 1151, controls; 2695 ± 1100, p = .004.Significant for longstanding diabetes compared to controls, p = .003.
To our knowledge, only one study has previously investigated whether a diminished incretin effect could be a consequence of diabetic autonomic neuropathy, reporting possible participation of autonomic neuropathy in the incretin effect through modulation of GIP secretion and hepatic insulin extraction. 28 noted with interest that both groups of diabetes had more rapid increase and higher paracetamol AUC concentration at 70 min, indicating more rapid gastric emptying than the control group.
Physiological hyperglycaemia is shown to slow gastric emptying in people without diabetes and in insulin dependent diabetes. 29Our results are supported by studies of similar cohorts in terms of diabetes duration and without severe complications, accelerated gastric emptying, regardless of glycaemic status. 30,31Several reasons for accelerated gastric emptying have been proposed including hyperglycaemia-induced oxidative stress leading to neural abnormalities, hyperplasia of interstitial cells of Cajal and smooth muscle hypercontractility, and loss of function of hormones responsible for slowing gastric emptying. 32The latter is proposed to be explained in part by a lack of small intestinal sensitivity to nutrients. 33It has previously been reported that a distal intestinal glucose exposure, compared to a proximal, was associated with an increased incretin effect and GIGD, in the same way as after surgical gastric bypass. 34 contrast to this, we report a negative correlation between gastric emptying and GIGD.This may strengthen the notion of a bidirectional relationship between gastric emptying and glycaemia, and could indicate that other factors that are associated with reduced GIGD contribute directly or indirectly to the gastric emptying rate.

| Association with the degree of dysglycaemia and duration of diabetes
We substantiate earlier findings of an association between the degree of dysglycaemia, duration of diabetes and the incretin effect.
The estimation of the incretin effect varies with the amount of glucose administered, gastric emptying rate, and whether the estimation is based on insulin, c-peptide or insulin secretion rate (ISR).Estimates based on ISR or insulin tend to be lower than those based on c-peptide, mainly due to hepatic insulin extraction.Compared to similar studies, estimating the incretin effect using c-peptide after 75 g of glucose ingestion, the incretin effect was reported at 59%-64% in healthy adults, 6% in type 2 diabetes and 37% in a group without diabetes but with BMI of 29. 9 Our rates of GIGD are also comparable to previous results, reported to be around 60% in healthy subjects, and in the range of 10%-30% in people with type 2 diabetes. 21The comparative similarity with our study strengthens the confidence in our results.
Our findings are novel in showing a continuum in the incretin effect, from healthy subjects to people with longstanding diabetes.
A previous study, including people with prediabetes, calculating the ISR has reported an incretin effect of 20% in type 2 diabetes, 45% in prediabetes and 59% in matched controls, but in a cohort with a mean age of 50 years and a BMI of 30. 35Hence, our results underscore the gradual progress of a reduced incretin effect, in synchronicity with the level of dysglycaemia, starting from well within normal HbA1c levels, towards frank diabetes.

| Strengths and weaknesses
The composition of our study population may explain the low prevalence of neuropathy and few between-group differences.Exclusion criteria, especially use of GLP-1 analogues and insulin, and proliferative retinopathy, may have led to selection of participants with longstanding diabetes with lean phenotype, an acceptable HbA1c and overall few microvascular complications.Also, in line with the results from the OGTT, some healthy volunteers had to be re-classified as early diabetes, and some controls satisfied the criteria for prediabetes, making our groups more homogenous than we would have anticipated.Consequently, the external validity was impaired as our groups were less representative of the general diabetes population.On the contrary, this highlights the gradual onset of hyperglycaemia in this age group, and the results contribute to new knowledge of the incretin effect in an age group that exhibits an increasing prevalence of type 2 diabetes worldwide, mainly due to a longer life expectancy.Finally, with regard to the impact of low BMI on external validity, this may also be seen as a strength, more certainly attributing the reduced incretin effect to diabetes, as obesity itself has also been shown to deteriorate the incretin effect. 23ucose was not measured immediately before or during rectal balloon distention or other neuropathy tests, and no examinations were done while performing a euglycemic clamp.Evidence concerning the effect of hyperglycaemia on rectal sensitivity is conflicting. 36,37We cannot exclude glucose levels may have affected our results on the day of examinations.
Investigating afferent nerves of the rectum can only be a surrogate for nerves relaying information on gut hormones.We also acknowledge that plasma levels of GLP-1 secretion in response to different nutrients have shown a substantial inter-individual variation. 38Still, there are similarities in both structural and physiological attributes between mucosal afferents in the rectum and vagal afferents found more proximally, providing a rationale for the rectal balloon distention test. 39,40e paracetamol absorption test is a feasible test evaluating gastric emptying, but still suboptimal, with several limitations as a surrogate marker. 41Hence, results on differences between groups on gastric emptying must be considered with caution.
Finally, our power estimates were uncertain due to the unknown effect size for the rapid rectal balloon distention test.With many non-significant tendencies in this study, it is possible that a larger sample size would have avoided type II errors.

| CON CLUS IONS
No associations were detected between CAN, sudomotor function or rectal sensitivity with the diminished incretin effect.However, rectal sensitivity, independent of diabetes duration, was associated with GIGD, alluding to a potential link between autonomic nerve function and gut-derived mechanisms governing glucose homeostasis.Finally, we confirm that the incretin effect deteriorates with a longer duration of diabetes and worsening of glycaemic control, in a cohort of older individuals.The novel results warrant further studies on the potential role of the autonomic nervous system in all aspects involved in gastrointestinal-mediated glucose disposal.
Published: Some of the data have previously been published and presented as a poster on the American Diabetes Association 83rd Scientific Sessions: Meling et al; 1658-P: Rectal Hyposensitivity, a Potential Marker of Enteric Autonomic Nerve Dysfunction, Is Significantly Associated with Gastrointestinally Mediated Glucose Disposal in Persons with Type 2 Diabetes.Diabetes 20 June 2023; 72 (Supplement_1): 1658-P.https:// doi.org/ 10. 2337/ db23-1658-P.
-way ANOVA, otherwise the Welch test for unequal variance when testing for statistical significance.For post hoc testing, we used the Bonferroni test where equal variance, and the Games-Howell test for unequal variance.Kruskal-Wallis test was used for statistical significance in the case of skewed data.For categorical data, the p-value was calculated using the chi-square test.For correlation analyses, we used Spearman's rank-order correlation test.The AUCs were calculated using the trapezoidal rule.Missing values for c-peptide and paracetamol were imputed from the average values before and after the missing values.For other missing data, subjects were removed from that analysis.Statistical significance was defined as a p-value ≤.05, and SPSS Version: 28.0.1.0.(IBM, New York, US) was used.Between-group differences.
Note: Data are presented as n, % or mean ± standard deviation.Diabetes duration, comorbidity, smoking status and drugs are self-reported.80% of early diabetes was newly discovered, and diabetes duration was therefore set to zero.All other significant for all groups.Abbreviations: ARB, angiotensin II receptor blocker; BMI, body mass index; DPP-4, dipeptidyl peptidase-4; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SGLT-2, sodium-glucose co-transporter 2. a Significant compared to controls.bSignificant compared to early diabetes.one