Abstract
Purpose
Gastrointestinal motility disorders are common in diabetes, though their prevalence and distribution have not been evaluated. The Wireless Motility Capsule (WMC) assesses transit times and pressures of the entire gastrointestinal (GI) tract in a single study. This study aimed to evaluate the prevalence and patterns of GI transit abnormalities using WMC in patients with diabetes.
Methods
This retrospective study included adult patients with Type 1 (T1D) and Type 2 Diabetes (T2D) who underwent WMC testing. Whole gut transit time (WGTT), gastric emptying time (GET), small bowel transit time (SBTT) and colon transit time (CTT) were analyzed. Univariate and multivariable analyses identified risk factors for delayed transit. The rates of diabetic complications were assessed. Kaplan–Meier estimates of mortality rates at 1, 3, and 5 years were compared for patients with T1D and T2D.
Results
A total of 475 patients (mean age of 54.4 years, 74% female) were included. Of these, 25.9% had T1D and 74.1% had T2D. Mean diabetes duration was 11.4 years. Patients with T1D had a higher mean disease duration of 21.7 years and those with T2D of 7.7 years. More than 35% of patients had HbA1c > 8 and 34% of patients had an HgbA1c between 6.5 and 7.99%. T1D patients had a higher mean HbA1c of 8.6% compared to 7.5% in T2D. Mean BMI was 32.3 in T2D and 26.5 in T1D. Mean BMI was higher in those with normal motility (32.9) than delayed motility (30.1). Delayed transit was observed in 75.8% with delayed GET (58.8%) being the most common abnormality. T2D had shorter median GET (4.4) than T1D (6.1) and longer CTT (95.9) than T1D (79.3). SBTT had an inverse correlation with HbA1c (-0.16). Microvascular complications were greater in T1D including neuropathy in 64.2%, retinopathy in 35% and nephropathy in 31.7% compared to T2D which were: neuropathy 50.3%, retinopathy 11.1% and nephropathy 19%. 7.3 and 1.1% of patients with T1D and T2D had kidney or pancreas transplants respectively. Kaplan–Meier analysis showed no significant difference in mortality in T1D vs T2D.
Conclusion
This study reports the prevalence of GI motility abnormalities in patients with longer duration of diabetes, higher HbA1c, lower BMI and associated diabetes microvascular complications. Delayed gastric emptying was the most common finding with greater prevalence of GET delay in T1D vs. T2D. T2D had shorter GET but longer CTT. Mean diabetes duration was 11.4 years and > 70% of patients with diabetes had suboptimal glucose control. Microvascular complications and pancreas/kidney transplant were higher in patients with T1D. Mortality rates were not different in T1D vs T2D. Identifying GI motility disorders in diabetes can guide management to improve outcomes.
Keywords: Motility disorder; Diabetes; Wireless motility capsule, Delayed gastric emptying
Introduction
Gastrointestinal (GI) motility disorders are common in patients with diabetes [1, 2], which often lead to impaired quality of life, increased hospitalizations and suboptimal glycemic control [1–3]. The Wireless Motility Capsule (WMC) is a useful tool for assessing motility throughout the entire GI tract, with the advantage of being a single noninvasive test that minimizes patient exposure to radiation [4, 5]. The WMC measures whole gut transit time (WGTT), as well as regional transit times, including gastric emptying time (GET), small bowel transit time (SBTT), and colonic transit time (CTT), by utilizing pH, temperature, and pressure sensors [4, 5]. In addition to its convenience and safety, WMC has demonstrated a greater diagnostic yield than other tests such as gastric emptying scintigraphy (GES) [6].
Given the emergence of novel therapeutics for obesity and T2D, such as glucagon-like peptide 1 receptor agonists (GLP-1 RA), which may impair gastric emptying and intestinal motility, understanding the prevalence of transit delays in patients with diabetes has become increasingly important. The aims of this study were to assess the prevalence of delay in GI transit and identify factors associated with these abnormalities in patients with T1D and T2D.
Methods
This was a retrospective, observational cohort study of patients with T1D and T2D referred for WMC at the Cleveland Clinic between 2002 and 2022. Patients with a diagnosis of diabetes greater than ≥ 18 years of age were identified by ICD codes. Exclusion criteria included patients with steroid-induced diabetes, those with a diabetes diagnosis following their motility testing, GI tract surgeries and those without transit time measurements due to testing malfunction (Fig. 1). The patient cohort was identified from an electronic medical record via natural language processing, and manual chart review was used to confirm variables of interest.
Fig. 1.
Inclusion and exclusion criteria flow chart. Patient flow diagram. Of 604 eligible patients, 475 met the inclusion criteria. There were 123 patients with T1D and 352 patients with T2D
Smartpill Motility Testing System (Wireless Motility Capsule, Medtronic Inc.)
We followed the protocol as outlined by Saad in Curr Gastroenterol Rep in 2016 [7]. The indication in all patients was to evaluate for generalized or multiregional motility disorders. Specific indications include (1) the assessment of gastric emptying (as well as regional and whole gut transit time) in suspected cases of gastroparesis and symptoms of upper gastrointestinal (GI) dysmotility, (2) the assessment of small bowel transit to specifically facilitate the detection of small bowel dysfunction in more generalized GI motility disorders, and (3) the assessment of colonic transit time in cases of chronic constipation. Opiates, laxatives, prokinetics, PPI, H2B, anticholinergics, antidiarrheals, antihistamines and antiemetics were held in all patients one week prior to testing. Normative values are determined to be 2–5 h for gastric emptying time (GET), 2–6 h for small bowel transit time (SBTT), and 10–59 h for colonic transit time (CTT). Hence delayed gastrointestinal transit times are defined as follows: gastric emptying is considered delayed if it exceeds 5 h, small bowel transit is delayed if it surpasses 6 h and colon transit time longer than 59 h is classified as delayed. Finally, whole gut transit is regarded as delayed when it exceeds 72 h. These delayed transit times serve as benchmarks for evaluating GI motility. Patients were required to have a serum glucose level of < 275 mg/dL prior to undergoing WMC.
Statistical Analyses
Categorical variables were summarized with frequencies and percentages, and compared using Pearson’s Chi-Square, Fisher’s exact, or Wilcoxon rank sum tests for ordered variables. Normally distributed continuous variables were described with means and standard deviations and compared with t-tests, while non-normally distributed continuous variables were described with medians and quartiles and compared using Wilcoxon rank sum tests. Multivariable logistic regression model was fitted utilizing significant risk factors that were either clinically relevant or identified on univariate analysis. Variables included in the model were age, BMI, gender, race, diabetes type, diabetes duration, and glycosylated hemoglobin (HbA1c). Analyses were performed using SAS software (version 9.4) and R software (version 4.2). Institutional Review Board approval was obtained prior to the initiation of the study.
Results
This study included 475 patients with diabetes who underwent WMC testing as seen in Fig. 1. The majority were female (73.9%) and white (73.7%), with a mean age of 54.4 ± 13.8 years (Table 1). Of the cohort, 25.9% had T1D and 74.1% had T2D. The mean diabetes duration was 11.4 ± 11.5 years. Compared to T2D patients, those with T1D were significantly younger (p ≤ 0.001), had a lower BMI (p ≤ 0.001), exhibited lower HbA1c values (p < 0.001), and had greater rates of overall transit delay (85.4% versus 72.4%; p = 0.004) (Table 1).
Table 1.
Baseline demographics by diabetes type
| Type I (N = 123) |
Type II (N = 352) |
||||
|---|---|---|---|---|---|
| Factor | N | Statistics | N | Statistics | p-value |
| Age, mean ± SD | 123 | 45.3 ± 12.8 | 352 | 57.6 ± 12.8 | < 0.001a1 |
| Patient sex, n (%) | 123 | 352 | 0.35c | ||
| Female | 87 (70.7) | 264 (75.0) | |||
| Male | 36 (29.3) | 88 (25.0) | |||
| Race, n (%) | 117 | 337 | 0.47c | ||
| Black | 18 (15.4) | 67 (19.9) | |||
| Other | 8 (6.8) | 17 (5.0) | |||
| White | 91 (77.8) | 253 (75.1) | |||
| Tobacco use, n (%) | 121 | 345 | 0.22c | ||
| Never | 56 (46.3) | 190 (55.1) | |||
| Passive/Quit | 41 (33.9) | 103 (29.9) | |||
| Yes | 24 (19.8) | 52 (15.1) | |||
| BMI, mean ± SD | 120 | 26.5 ± 6.1 | 348 | 32.3 ± 7.7 | < 0.001a2 |
| Diabetes duration (years), mean ± SD | 123 | 21.8 ± 13.9 | 352 | 7.7 ± 7.6 | * |
| Diabetes duration (years), median [Q1, Q3] | 123 | 20.2 [10.5, 31.4] | 352 | 5.4 [2.4, 10.3] | < 0.001b |
| Disease duration, n (%) | 105 | 331 | 0.90b | ||
| Less than 1 year | 42 (40.0) | 131 (39.6) | |||
| 1–4.99 years | 31 (29.5) | 101 (30.5) | |||
| 5–9.99 years | 20 (19.0) | 70 (21.1) | |||
| 10 + years | 12 (11.4) | 29 (8.8) | |||
| HgbA1c at Testing, mean ± SD | 123 | 8.6 ± 2.2 | 350 | 7.5 ± 2.1 | * |
| HgbA1c at Testing, median [Q1, Q3] | 123 | 8.1 [7.0, 9.7] | 350 | 6.9 [6.0, 8.3] | < 0.001b |
| HBA1C, n (%) | 123 | 350 | < 0.001b | ||
| < 6.5 | 18 (14.6) | 124 (35.4) | |||
| 6.5–7.99 | 40 (32.5) | 122 (34.9) | |||
| 8–9.99 | 40 (32.5) | 61 (17.4) | |||
| 10 + | 25 (20.3) | 43 (12.3) | |||
| Neuropathy, n (%) | 123 | 79 (64.2) | 352 | 177 (50.3) | 0.008c |
| Retinopathy, n (%) | 123 | 43 (35.0) | 352 | 39 (11.1) | < 0.001c |
| Nephropathy, n (%) | 123 | 39 (31.7) | 352 | 67 (19.0) | 0.004c |
| Transplant: Any Time, n (%) | 123 | 9 (7.3) | 352 | 4 (1.1) | 0.001d |
| Transplant: Prior to Test, n (%) | 123 | 6 (4.9) | 352 | 0 (0.00) | < 0.001d |
| Transplant: After Test, n (%) | 123 | 3 (2.4) | 352 | 4 (1.1) | 0.38d |
| Death, n (%) | 123 | 26 (21.1) | 352 | 48 (13.6) | 0.048c |
| Dysmotility, n (%) | 123 | 352 | 0.004c | ||
| Normal | 18 (14.6) | 97 (27.6) | |||
| Delayed | 105 (85.4) | 255 (72.4) | |||
| WGTT_DN, n (%) | 118 | 350 | 0.15c | ||
| Normal | 66 (55.9) | 222 (63.4) | |||
| Delayed | 52 (44.1) | 128 (36.6) | |||
| SLBTT_DN, n (%) | 116 | 343 | 0.72c | ||
| Normal | 76 (65.5) | 231 (67.3) | |||
| Delayed | 40 (34.5) | 112 (32.7) | |||
| GET_DN, n (%) | 123 | 348 | 0.002c | ||
| Normal | 36 (29.3) | 158 (45.4) | |||
| Delayed | 87 (70.7) | 190 (54.6) | |||
| CTT_DN, n (%) | 116 | 341 | 0.93c | ||
| Normal | 76 (65.5) | 225 (66.0) | |||
| Delayed | 40 (34.5) | 116 (34.0) | |||
| SBTT_DN, n (%) | 119 | 346 | 0.92c | ||
| Normal | 92 (77.3) | 266 (76.9) | |||
| Delayed | 27 (22.7) | 80 (23.1) | |||
Statistics presented as Mean ± SD, Median [P25, P75], N (column %)
p-values: a1 = t-test, a2 = Satterthwaite t-test, b = Wilcoxon Rank Sum test, c = Pearson's chi-square test
*P-value not shown due to non-normality of the data
Bold text is used to indicate p-values lower than 0.05, representing statistically significant values. This applies to all tables throughout the manuscript
75.8% (n = 360) of patients exhibited some form of GI transit delay. Delayed GET (58.8%) was the most common abnormality followed by delayed WGTT (38.5%), CTT (34.1%) and SBTT (23%). Rates of delayed transit were compared between T1D and T2D for WGTT (44.1 vs 36.6%; p = 0.15), GET (70.7 vs 54.6%; p = 0.002) SBTT (22.7 vs 23.1%; p = 0.92), and CTT (34.5 vs 34%; p = 0.93) (Tables 1, 2).
Table 2.
Intestinal transit times
| Measure | Type I | Type II | |||
|---|---|---|---|---|---|
| N | Statistics | N | Statistics | P-value | |
| Overall | |||||
| WGTT_Time_Decimal, median [Q1, Q3] | 118 | 65.0 [41.8, 110.0] | 350 | 50.8 [29.0, 100.7] | 0.056b |
| SLBTT_Time_Decimal, median [Q1, Q3] | 116 | 45.4 [21.3, 75.2] | 343 | 42.7 [20.9, 85.2] | 0.81b |
| GET_Time_Decimal, median [Q1, Q3] | 123 | 6.1 [3.2, 30.6] | 348 | 4.4 [2.9, 17.5] | 0.001b |
| CTT_Time_Decimal, median [Q1, Q3] | 116 | 39.5 [15.0, 69.4] | 341 | 37.8 [17.1, 78.5] | 0.75b |
| SBTT_Time_Decimal, median [Q1, Q3] | 119 | 4.6 [3.3, 6.0] | 346 | 4.4 [3.2, 5.8] | 0.68b |
| Normal emptying | |||||
| WGTT_Time_Decimal, median [Q1, Q3] | 66 | 42.9 [26.4, 54.5] | 222 | 33.4 [23.1, 49.3] | 0.100b |
| SLBTT_Time_Decimal, median [Q1, Q3] | 76 | 28.2 [14.2, 44.0] | 231 | 27.4 [17.8, 43.5] | 0.98b |
| GET_Time_Decimal, median [Q1, Q3] | 36 | 2.2 [0.77, 2.9] | 158 | 2.7 [2.0, 3.3] | 0.046b |
| CTT_Time_Decimal, median [Q1, Q3] | 76 | 23.8 [9.9, 38.6] | 225 | 21.5 [13.2, 37.7] | 0.93b |
| SBTT_Time_Decimal, median [Q1, Q3] | 92 | 4.2 [2.9, 4.9] | 266 | 3.9 [2.8, 4.8] | 0.40b |
| Delayed emptying | |||||
| WGTT_Time_Decimal, median [Q1, Q3] | 52 | 118.4 [95.1, 124.3] | 128 | 118.0 [95.3, 131.8] | 0.76b |
| SLBTT_Time_Decimal, median [Q1, Q3] | 40 | 85.6 [74.8, 106.5] | 112 | 104.7 [85.6, 120.7] | 0.005b |
| GET_Time_Decimal, median [Q1, Q3] | 87 | 16.8 [5.8, 45.7] | 190 | 16.8 [5.0, 22.9] | 0.025b |
| CTT_Time_Decimal, median [Q1, Q3] | 40 | 79.3 [68.8, 101.3] | 116 | 95.9 [76.6, 114.2] | 0.027b |
| SBTT_Time_Decimal, median [Q1, Q3] | 27 | 7.4 [6.7, 9.7] | 80 | 7.5 [6.8, 10.0] | 0.46b |
A total of 53.9% of patients had nephropathy: 64% in T1D and 50.3% in T2D. Retinopathy was present in 17.3% of patients: 35% in T1D and 11% in T2D. Additionally, 22.3% of patients had neuropathy: 31.7% in T1D and 19% in T2D (Table 1 and Fig. 2). Transplant rates were higher for T1D (7.3%) compared to 1.1% in T2D patients (Table 1 and Fig. 2). In multivariable models for select outcomes of delay (Table 3), BMI was inversely correlated with delay. No significant adjusted differences between diabetes types were observed across these measures.
Fig. 2.
Bar plots of complications by diabetes type. This figure presents the distribution of complications in patients with Type I and Type II diabetes. The rate of complications was higher in Type I diabetes, with all differences showing a significance level below 0.05
Table 3.
Multivariable analysis of motility factors
| WGTT delay | GET delay | SBTT delay | CTT delay | |||||
|---|---|---|---|---|---|---|---|---|
| Characteristic | OR (95% CI) | p-value | OR (95% CI) | p-value | OR (95% CI) | p-value | OR (95% CI) | p-value |
| Age | 1.01 (1.00, 1.03) | 0.11 | 1.00 (0.98, 1.01) | 0.70 | 1.03 (1.01, 1.05) | < 0.001 | 0.99 (0.97, 1.00) | 0.14 |
| BMI | 0.97 (0.94, 1.00) | 0.040 | 0.97 (0.94, 1.00) | 0.024 | 0.97 (0.94, 1.00) | 0.060 | 0.96 (0.93, 0.99) | 0.012 |
| Race | ||||||||
| White | – | – | – | – | ||||
| Black | 1.13 (0.68, 1.86) | 0.64 | 1.16 (0.69, 1.95) | 0.57 | 1.10 (0.65, 1.84) | 0.73 | 1.53 (0.87, 2.69) | 0.14 |
| Other | 0.79 (0.33, 1.89) | 0.59 | 1.08 (0.45, 2.57) | 0.86 | 0.87 (0.35, 2.17) | 0.76 | 0.52 (0.14, 1.99) | 0.34 |
| Gender | ||||||||
| Female | – | – | – | – | ||||
| Male | 0.65 (0.41, 1.02) | 0.063 | 1.02 (0.66, 1.58) | 0.93 | 0.49 (0.30, 0.81) | 0.005 | 1.12 (0.67, 1.88) | 0.66 |
| Diabetes type | ||||||||
| I | – | – | – | – | ||||
| II | 0.68 (0.40, 1.16) | 0.15 | 0.63 (0.37, 1.08) | 0.093 | 0.74 (0.42, 1.28) | 0.28 | 1.28 (0.69, 2.37) | 0.43 |
| DM duration | ||||||||
| Less than 1 year | – | – | – | – | ||||
| 1–4.99 years | 1.54 (0.93, 2.53) | 0.091 | 1.16 (0.73, 1.84) | 0.53 | 1.64 (0.99, 2.71) | 0.056 | 2.22 (1.26, 3.92) | 0.006 |
| 5–9.99 years | 0.73 (0.41, 1.29) | 0.28 | 1.49 (0.86, 2.58) | 0.15 | 0.80 (0.44, 1.46) | 0.46 | 1.96 (1.03, 3.73) | 0.042 |
| 10 + years | 0.62 (0.28, 1.39) | 0.25 | 0.93 (0.44, 1.99) | 0.86 | 0.56 (0.22, 1.47) | 0.24 | 1.42 (0.53, 3.84) | 0.49 |
| HbA1c Group | ||||||||
| < 6.5 | – | – | – | – | ||||
| 6.5–7.99 | 1.29 (0.78, 2.13) | 0.32 | 0.93 (0.58, 1.50) | 0.76 | 0.98 (0.58, 1.64) | 0.93 | 0.89 (0.51, 1.54) | 0.67 |
| 8–9.99 | 0.99 (0.56, 1.76) | 0.98 | 1.46 (0.83, 2.59) | 0.19 | 1.00 (0.54, 1.84) | > 0.99 | 0.59 (0.30, 1.17) | 0.13 |
| 10 + | 0.92 (0.48, 1.76) | 0.79 | 0.97 (0.52, 1.80) | 0.92 | 0.98 (0.50, 1.92) | 0.95 | 0.48 (0.22, 1.06) | 0.068 |
OR Odds Ratio, CI Confidence interval
Of those with a SBTT, GET, or CTT delay in T1D, 42% had GET delay alone, 7% had SBTT delay alone, 8% had CTT delay alone and 6% had delay in all 3 locations. In T2D, 38% had GET delay alone, 5% had SBTT delay alone, 14% had CTT delay alone, and 10% had delay in all 3 locations (Figs. 3, 4 and 5, Venn diagrams).
Fig. 3.
Total population transit delay distribution. Venn diagram shows the overlap of total individuals experiencing delays in Small Bowel Transit Time (SBTT), Gastric Emptying Time (GET), and Colonic Transit Time (CTT). Percentages indicate the proportion of the total population in each subset
Fig. 4.
Type 1 diabetes transit delay distribution. Venn diagram shows the overlap of T1D individuals experiencing delays in Small Bowel Transit Time (SBTT), Gastric Emptying Time (GET), and Colonic Transit Time (CTT). Percentages indicate the proportion of the total population in each subset
Fig. 5.
Type 2 diabetes transit delay distribution. Venn diagram shows the overlap of T2D individuals experiencing delays in Small Bowel Transit Time (SBTT), Gastric Emptying Time (GET), and Colonic Transit Time (CTT). Percentages indicate the proportion of the total population in each subset
A boxplot of HbA1c levels by diabetes type shows a higher HbA1c for T1D patients (Fig. 6). A Forest plot (Fig. 7) shows the results of the overall correlations between HbA1c and transit times from Tables 4 and 5. Analysis of HbA1c and GI transit times revealed no strong associations, with all correlations being very weak. While a statistically significant negative correlation between HbA1c and SBTT was observed, its magnitude was small. Group comparisons (Table 5) showed no significant differences in transit times across HbA1c levels (Table 4, Fig. 5).
Fig. 6.
Boxplots of HbA1c by diabetes type. This figure compares HbA1c levels between patients with Type I and Type II diabetes. The median HbA1c level is significantly higher in Type I diabetes (p < 0.001), as indicated by the boxplot distributions
Fig. 7.
Forest plot of correlations between HbA1c and transit times. This figure shows correlations between transit time factors (CTT, GET, SBTT, SBLTT, WGTT) and HbA1c levels, with 95% confidence intervals. SBTT shows a significant negative correlation (p < 0.001).
Table 4.
Correlations between HbA1c and transit times
| Factor | N | Rho (95% CI) | p-value |
|---|---|---|---|
| Overall | |||
| CTT time | 455 | − 0.09 (− 0.18, 0.01) | 0.065 |
| GET time | 469 | 0.06 (− 0.03, 0.15) | 0.18 |
| SBTT time | 463 | − 0.16 (− 0.25, − 0.07) | < 0.001 |
| SBLTT time | 457 | − 0.09 (− 0.18, 0.00) | 0.061 |
| WGTT time | 466 | − 0.03 (− 0.12, 0.06) | 0.47 |
| Type I | |||
| CTT time | 116 | − 0.06 (− 0.24, 0.12) | 0.51 |
| GET time | 123 | 0.10 (− 0.08, 0.27) | 0.28 |
| SBTT time | 119 | − 0.18 (− 0.35, 0.00) | 0.054 |
| SBLTT time | 116 | − 0.06 (− 0.24, 0.12) | 0.50 |
| WGTT time | 118 | − 0.01 (− 0.19, 0.17) | 0.93 |
| Type II | |||
| CTT time | 339 | − 0.09 (− 0.20, 0.02) | 0.093 |
| GET time | 346 | − 0.01 (− 0.11, 0.10) | 0.91 |
| SBTT time | 344 | − 0.17 (− 0.27, − 0.06) | 0.002 |
| SBLTT time | 341 | − 0.09 (− 0.20, 0.01) | 0.086 |
| WGTT time | 348 | − 0.07 (− 0.17, 0.04) | 0.20 |
rho Spearman's correlation; CI confidence interval
Table 5.
Associations between HbA1c groups and transit time
| < 6.5 | 6.5–7.99 | 8–9.99 | 10 + | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Factor | N | Statistics | N | Statistics | N | Statistics | N | Statistics | p-value |
| Overall | |||||||||
| WGTT Time, median [Q1, Q3] | 138 | 55.0 [30.9, 100.3] | 160 | 58.0 [30.1, 112.8] | 100 | 55.9 [31.8, 102.2] | 68 | 50.3 [25.9, 97.8] | 0.59b |
| SBLTT Time, median [Q1, Q3] | 133 | 48.4 [21.5, 92.9] | 158 | 43.6 [24.4, 83.2] | 98 | 41.2 [20.8, 74.8] | 68 | 35.6 [18.0, 70.2] | 0.28b |
| GET Time, median [Q1, Q3] | 139 | 4.9 [3.0, 17.6] | 161 | 4.5 [2.8, 18.1] | 101 | 5.2 [3.2, 20.9] | 68 | 4.4 [2.6, 20.7] | 0.29b |
| CTT Time, median [Q1, Q3] | 133 | 42.6 [17.8, 87.1] | 156 | 36.7 [19.2, 76.6] | 98 | 36.2 [16.7, 71.0] | 68 | 29.0 [13.8, 66.3] | 0.27b |
| SBTT Time, median [Q1, Q3] | 138 | 4.9 [3.7, 6.3]4 | 159 | 4.5 [3.2, 5.9] | 98 | 4.1 [3.0, 5.4] | 68 | 3.9 [3.1, 5.1]1 | 0.016b |
| Type I | |||||||||
| WGTT Time, median [Q1, Q3] | 15 | 59.3 [32.3, 93.2] | 39 | 60.7 [29.6, 109.0] | 39 | 81.3 [45.8, 120.0] | 25 | 54.4 [30.8, 116.0] | 0.47b |
| SBLTT Time, median [Q1, Q3] | 15 | 53.1 [17.4, 73.8] | 38 | 42.8 [21.2, 69.8] | 38 | 59.1 [33.4, 84.9] | 25 | 30.9 [12.7, 67.4] | 0.11b |
| GET Time, median [Q1, Q3] | 18 | 6.5 [0.80, 47.8] | 40 | 5.4 [3.7, 16.9] | 40 | 6.2 [3.6, 28.9] | 25 | 16.8 [4.3, 34.8] | 0.81b |
| CTT Time, median [Q1, Q3] | 15 | 46.2 [10.5, 64.1] | 37 | 36.7 [17.0, 67.3] | 39 | 51.8 [25.7, 76.5] | 25 | 25.5 [7.6, 61.1] | 0.13b |
| SBTT Time, median [Q1, Q3] | 16 | 5.5 [3.4, 7.2] | 39 | 4.9 [3.5, 6.0] | 39 | 4.3 [3.4, 5.6] | 25 | 4.2 [3.1, 5.3] | 0.22b |
| Type II | |||||||||
| WGTT Time, median [Q1, Q3] | 123 | 54.1 [29.5, 100.7] | 121 | 56.7 [30.1, 114.2] | 61 | 46.7 [27.9, 75.6] | 43 | 47.1 [25.0, 94.2] | 0.28b |
| SBLTT Time, median [Q1, Q3] | 118 | 48.0 [22.2, 95.2] | 120 | 44.5 [24.8, 84.6] | 60 | 33.9 [18.8, 64.2] | 43 | 39.1 [19.3, 73.1] | 0.25b |
| GET Time, median [Q1, Q3] | 121 | 4.8 [3.1, 17.1] | 121 | 4.0 [2.5, 18.8] | 61 | 4.9 [3.1, 17.5] | 43 | 3.4 [2.3, 16.5] | 0.16b |
| CTT Time, median [Q1, Q3] | 118 | 41.7 [18.1, 88.0] | 119 | 36.7 [19.5, 79.9] | 59 | 29.1 [14.6, 59.6] | 43 | 35.6 [14.6, 67.7] | 0.27b |
| SBTT Time, median [Q1, Q3] | 122 | 4.9 [3.7, 6.2] | 120 | 4.4 [3.1, 5.8] | 59 | 4.0 [2.8, 5.4] | 43 | 3.9 [3.1, 4.9] | 0.057b |
Statistics presented as Median [P25, P75]
p-values: b = Kruskal–Wallis test
1Significantly different from < 6.5
2Significantly different from 6.5 to 7.99
3Significantly different from 8 to 9.99
4Significantly different from 10 +
Post-hoc pairwise comparisons were done using Bonferroni adjustment
Kaplan–Meier estimates of risk for mortality over time show higher risk of mortality in patients with T1D at year 1, 3 and 5 compared to T2D patients (Table 6 and Fig. 8). However, the mortality difference in patients with T1D and T2D was not significant at 10 years.
Table 6.
Overall mortality rates
| Characteristic | Event (%) | 1 Years | 3 Years | 5 Years |
|---|---|---|---|---|
| Overall | 61 (13) | 1.73 (0.53, 2.91) | 8.40 (5.80, 10.9) | 12.1 (8.91, 15.2) |
| Type I | 21 (18) | 3.39 (0.07, 6.60) | 12.9 (6.57, 18.7) | 18.0 (10.5, 24.9) |
| Type II | 40 (12) | 1.16 (0.02, 2.29) | 6.80 (4.07, 9.45) | 10.0 (6.54, 13.3) |
Log-rank p-value: 0.089
Fig. 8.
Kaplan–Meier plot of mortality by diabetes type. This figure shows the Kaplan–Meier survival curves comparing mortality rates over a 10-year follow-up period for patients with Type I and Type II diabetes. Mortality was higher in Type I diabetes throughout the follow-up period. The number at risk for each time point is displayed below the plot
Discussion
We report that nearly three-quarters of patients with diabetes referred for motility tests have delayed GI transit which is higher than the general population as shown in other WMC studies [8]. Patients had a mean diabetes duration of 11.4 years (21.8 years in T1D and 7.7 years in T2D) and mean HbA1c of 7.8% (mean HbA1c 8.6% in T1D, mean HbA1c 7.5% in T2D) with > 70% had suboptimal glucose control with HbA1c > 6.5%. In addition, motility delay correlated with high rates of microvascular complications of diabetes in our cohort including neuropathy, retinopathy and nephropathy. Microvascular complications were greater in T1D including neuropathy was present in 64.2%, retinopathy in 35% and nephropathy in 31.7% compared to patients with T2D which were: neuropathy 50.3%, retinopathy 11.1% and nephropathy 19%. 7.3 and 1.1% of patients with T1D and T2D had kidney or pancreas transplants respectively. Parkman et al. found over 35% of patients with diabetic gastroparesis had retinopathy and > 60% had some other complication of diabetes [9] which is in line with our data. Another study found that gastroparesis was associated with diabetic neuropathy but not with glycemic control [10]. Additional studies correlating continuous glucose monitor (CGMs) metrics with the incidence of motility disorders may be helpful. The mortality analysis in this study indicates that while the raw percentage of deaths appears higher in patients with T1D (Table 1), this does not account for differences in follow-up duration or the timing of deaths. A more appropriate comparison is provided in Table 6 where the Kaplan–Meier analysis accounts for these factors. In this adjusted analysis, the difference in mortality rates between groups is no longer statistically significant (p = 0.089). This suggests that the initially observed difference in mortality rates may be partially explained by variations in follow-up rather than a true underlying difference in risk between groups.
Delayed GET was the most prevalent WMC abnormality. Patients with T1D and T2D showed comparable rates of delays in SBTT and CTT. Notably, higher BMI was associated with lower risk for small bowel and colon transit delay. Our finding of delayed GET in especially T1D is consistent with previous research. One study showed that 33.7% of T1D patients had delayed gastric emptying which correlated with female gender, increased BMI, symptoms of bloating and abdominal pain [11]. Other studies have also reported that gastroparesis is more common in T1D compared to T2D [12]. Similarly, studies have demonstrated that T1D patients have slower small bowel and colonic transit as compared to healthy controls [5, 13]. These studies also correlated testing results with GI symptoms however it has been shown that symptoms do not correlate with WMC testing results [8, 14].
We did not correlate symptoms to WMC findings. A previous publication at our center by Arora et al. showed that patient symptoms are poor predictors of GI dysmotility on WMC. There was no significant difference in past medical or past surgical history between patients with isolated regional versus multiregional involvement. This study was done in the general population in which 49% had WMC abnormality. In our current study in people with diabetes, we report 75% with dysmotility. Another study by Kuo et al. also found that symptoms did not predict the results of WMC testing. Despite lack of symptom correlation, WMC is a useful diagnostic test because of its ability to detect motility abnormalities in the entire GI tract that are often not found through other tests and which can change management decisions [4, 6, 8, 14].
An additional finding from our study is that higher BMI had a protective effect against delayed GI transit. This aligns with existing literature showing a positive association between obesity and increased gastric and GI transit times [15]. This finding may be explained by the differences in neuroendocrine molecule regulation in obese individuals compared to non-obese individuals. Studies have identified altered levels of neuroendocrine compounds such as ghrelin, substance P (SP), and vasoactive intestinal peptide (VIP) vary in those with obesity [15]. A recent study concluded that SP levels were increased in the gastric antrum of obese subjects, VIP levels were decreased in the antrum and fundus, and changes in ghrelin were insignificant compared to non-obese subjects [16]. SP is a neurotransmitter that is known to increase gastric and intestinal motility, and VIP is a polypeptide hormone that is known to decrease motility. An increase in gastric emptying and intestinal motility may lead to reduced satiety, which may further contribute to obesity. This observed neuroendocrine difference aligns with the results of our motility study, where subjects with higher BMI were less likely to have SBTT and CTT delays. The increased levels of SP and decreased levels of VIP in individuals with obesity may contribute to improved intestinal motility.
The strengths of our study include (1) large sample size, (2) the use of WMC to evaluate multi-regional transit times in patients with both T1D and T2D, (3) patients with longer duration of diabetes, (4) correlating with different microvascular complications. Limitations of our study include (a) retrospective nature of our study, (b) single center, (c) referred patient population, (d) lack of correlation of symptoms severity to WMC results.
Conclusion
We conclude that > 75% of patients with diabetes had a GI transit abnormality. Both patients with T1D and T2D had GET delay though it was more prevalent in T1D. Longer duration of diabetes and suboptimal glucose control played a significant role in motility delay. Microvascular complications and pancreas/kidney transplants were higher in T1D vs T2D. Kaplan–Meier analysis showed no significant difference in mortality in T1D vs T2D. Emerging therapies with GLP-1RA may exacerbate GI motility issues, making it critical for clinicians to recognize and address such complications early. Prospective randomized control trials are needed to further elucidate factors associated with GI motility disorders in patients with T1D and T2D.
Acknowledgments
Dr. Bret Lashner for his support and guidance.
Author Contributions
SG: Conceptualization, methodology, investigation, supervision, writing—original draft, writing—review and editing SV: writing—review and editing HS: writing of discussion BE: writing of discussion RS: writing—review and editing JB: Statistician, data analysis, table creation 1–5 SM: Statistician, data analysis, table creation 1–5 AL: supervision, writing—review and editing MC: supervision, writing—review and editing.
Data Availability
No datasets were generated or analysed during the current study.
Declarations
Conflict of interest
SG receives grant funding from Abbott Diabetes Care and is a consultant for NovoNordisk and Ardelyx. MC is a speaker for Evoke pharma and Kate Farms. AL is a consultant to Takeda, Ironwood, Ardelyx, Atmo, Vibrant, Allurion, Evoke, and owns stock in J&J and Bristol Myer Squibb.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Freeman R (2014) Diabetic Autonomic Neuropathy. Handb Clin Neurol 126:63–79. 10.1016/B978-0-444-53480-4.00006-0 [DOI] [PubMed] [Google Scholar]
- 2.Camilleri M. Gastrointestinal motility disorders in neurologic disease. Journal of Clinical Investigation. 2021;131. 10.1172/JCI143771 [DOI] [PMC free article] [PubMed]
- 3.Bharucha AE, Kudva YC, Prichard DO. Diabetic Gastroparesis. Endocr Rev. 2019;40:1318–1352. 10.1210/er.2018-00161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rouphael C, Arora Z, Thota PN, et al. Role of wireless motility capsule in the assessment and management of gastrointestinal dysmotility in patients with diabetes mellitus. Neurogastroenterology and Motility. 2017;29. 10.1111/nmo.13087 [DOI] [PubMed]
- 5.Klinge MW, Haase AM, Mark EB, et al. Colonic motility in patients with type 1 diabetes and gastrointestinal symptoms. Neurogastroenterology and Motility. 2020;32. 10.1111/nmo.13948 [DOI] [PubMed]
- 6.Lee AA, Rao S, Nguyen LA et al. Validation of diagnostic and performance characteristics of the wireless motility capsule in patients with suspected gastroparesis. Clinical Gastroenterology and Hepatology. 2019;17:1770–1779.e2. 10.1016/j.cgh.2018.11.063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Saad R. The wireless motility capsule: a one-stop shop for the evaluation of GI motility disorders. Feb 2016. Neurogastroenterology and Motility Disorders of the Gastrointestinal Tract (S. Rao, Section Editor). Curr Gastroenterol Rep 2016;18:14. https://www.naspghan.org/files/documents/pdfs/training/curriculum-resources/motility/Saad_The-Wireless-Motility-Capsule.pdf [DOI] [PubMed]
- 8.Arora Z, Parungao JM, Lopez R, Heinlein C, Santisi J, Birgisson S. Clinical utility of wireless motility capsule in patients with suspected multiregional gastrointestinal dysmotility. Dig Dis Sci. 2015;60:1350–1357. 10.1007/s10620-014-3431-9. [DOI] [PubMed] [Google Scholar]
- 9.Henry P. Parkman, Laura A. Wilson, Gianrico Farrugia, et al. Delayed gastric emptying associates with diabetic complications in diabetic patients with symptoms of gastroparesis. Am J Gastroenterol. 2019;114:1778–1794. 10.14309/ajg.0000000000000410 [DOI] [PMC free article] [PubMed]
- 10.Punkkinen J, Färkkilä M, Mätzke S et al. Upper abdominal symptoms in patients with Type 1 diabetes: unrelated to impairment in gastric emptying caused by autonomic neuropathy. Diabet Med J Br Diabet Assoc. 2008;25:570–577. 10.1111/j.1464-5491.2008.02428.x. [DOI] [PubMed] [Google Scholar]
- 11.Sfarti C, Trifan A, Hutanasu C et al. Prevalence of gastroparesis in type 1 diabetes mellitus and its relationship to dyspeptic symptoms. J Gastrointest Liver Dis JGLD. 2010;19:279–284. [PubMed] [Google Scholar]
- 12.Choung RS, Locke GR, Schleck CD et al. Risk of gastroparesis in subjects with type 1 and 2 diabetes in the general population. American Journal of Gastroenterology. 2012;107:82–88. 10.1038/ajg.2011.310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Faria M, Pavin EJ, Parisi MCR et al. Delayed small intestinal transit in patients with long-standing type 1 diabetes mellitus: investigation of the relationships with clinical features, gastric emptying, psychological distress, and nutritional parameters. Diabetes Technol Ther. 2013;15:32–38. 10.1089/dia.2012.0158. [DOI] [PubMed] [Google Scholar]
- 14.Kuo B, Maneerattanaporn M, Lee AA et al. Generalized transit delay on wireless motility capsule testing in patients with clinical suspicion of gastroparesis, small intestinal dysmotility, or slow transit constipation. Dig Dis Sci. 2011;56:2928–2938. 10.1007/s10620-011-1751-6. [DOI] [PubMed] [Google Scholar]
- 15.Al Mushref M, Srinivasan S. Effect of high fat-diet and obesity on gastrointestinal motility. Ann Transl Med. 2013;1:14. 10.3978/j.issn.2305-5839.2012.11.01. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Atas U, Erin N, Tazegul G et al. Changes in ghrelin, substance P and vasoactive intestinal peptide levels in the gastroduodenal mucosa of patients with morbid obesity. Neuropeptides. 2021;89:102164. 10.1016/j.npep.2021.102164. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
No datasets were generated or analysed during the current study.