Analysis of gastric electrical rhythm in patients with metabolic dysfunction-associated steatotic liver disease and type 2 diabetes mellitus

Abstract

BACKGROUND

Diabetic gastrointestinal autonomic neuropathy (DGAN) often affects the patients’ quality of life. Thus, research on influencing factors of DGAN may promote DGAN prevention and decrease the incidence rate. This study used electrogastrogram (EGG) to assess the relationship between metabolic dysfunction-associated steatotic liver disease (MASLD) and DGAN.

AIM

To analyze the changes of EGG in patients with MASLD and type 2 diabetes mellitus (T2DM), and to elucidate whether ultrasound-diagnosed MASLD is an independent risk factor for diabetic gastric motility disorders (DGMD).

METHODS

A total of 272 patients with T2DM hospitalized at the Second Affiliated Hospital of Soochow University from December 2020 to December 2021 and the Affiliated Suqian Hospital of Xuzhou Medical University from November 2023 to June 2024 were included in the cross-sectional study. General information, clinical data, and medical history of all study subjects, including name, age, gender, body mass index, and duration of diabetes were collected. Laboratory tests included biochemical parameters, glycosylated hemoglobin (HbA1c), fasting C-peptide, 2h postprandial C-peptide and 25-hydroxyvitamin D [25(OH)D]. EGG, fundus examination and carotid artery ultrasonography were performed and results were recorded. According to the results of EGG, the subjects were divided into the DGMD group and non-DGMD group.

RESULTS

The duration of diabetes, fasting blood glucose (FBG), HbA1c, 25(OH)D and the prevalence of MASLD were significantly higher in the DGMD group (P < 0.05). Multiple logistic regression analysis showed that the duration, FBG, 25(OH)D and the presence of MASLD were independent influencing factors.

CONCLUSION

MASLD is strongly associated with an increased incidence of DGMD. Timely treatment of MASLD is effective to prevent diabetic gastroparesis.

Key Words: Type 2 diabetes mellitus; Gastric electrical rhythm; Electrogastrogram; Metabolic dysfunction-associated steatotic liver disease; Diabetic gastric motility disorders

Core Tip: This study identifies metabolic dysfunction-associated steatotic liver disease (MASLD) as an independent risk factor for diabetic gastric motility disorders (DGMD) in type 2 diabetes patients, confirmed by electrogastrogram. Longer diabetes duration, elevated fasting blood glucose, and lower 25-hydroxyvitamin D levels synergistically increase DGMD risk. Early intervention for MASLD may prevent diabetic gastroparesis.

INTRODUCTION

Diabetic gastrointestinal autonomic neuropathy (DGAN) refers to the impairment of the autonomic nervous system that supplies the gastrointestinal tract[1]. It has received inappropriately little attention, despite the fact that the disease occurs frequently and often with serious sequelae. The manifestations of DGAN are varied, including symptoms such as nausea, vomiting, bloating, post-meal fullness, along with issues like constipation, diarrhea and fecal incontinence. DGAN can have a serious negative impact on life and also affect the regulation of glycemia. Diabetic gastric motility disorders (DGMD) represent the most important manifestation of DGAN.

Metabolic dysfunction-associated steatotic liver disease (MASLD) ranks as the prevalent chronic liver disorder globally, and the prevalence of MASLD in diabetic patients is much higher compared to the population without diabetes. To date, several studies have shown that MASLD is associated not only with a fairly high liver-related morbidity and mortality[2], but also with an increased risk of developing chronic kidney disease and cardiovascular ailments[3-6]. Some studies have shown that an increased prevalence of diabetic peripheral neuropathy (DPN) is associated with MASLD[7]. Type 2 diabetes mellitus (T2DM) patients with MASLD are more likely to develop neurological complications such as DPN[8-10] and diabetic cardiac autonomic neuropathy[11,12]. Some studies reported no association between MASLD diabetic retinopathy (DR)[13]. In addition, in a meta-analysis, subgroup analyses suggested that in China, T2DM patients with MASLD had a decreased risk of DR[13]. However, there are currently few studies on the relationship between MASLD and DGMD.

In the context of diabetes, the diagnosis of DGMD cannot rely solely on clinical symptoms and necessitates the assessment of gastric emptying (GE). Scintigraphy is recognized as the "gold standard" for evaluating GE[14,15]. However, scintigraphy has not been widely used in clinical practice because of radiation exposure, the need for specialized nuclear medicine-trained personnel and costly infrastructure[16]. Interstitial cells of Cajal generate electrical signals that can be recorded with electrogastrogram (EGG)[17]. T2DM patients with DGMD may have gastric electrical rhythm disturbances and decreased electric signal power postprandially[18]. Now more and more medical centers use EGG to evaluate DGMD[17]. In this study, we attempted to assess the relationship between MASLD and DGMD in T2DM patients based on the EGG parameters.

MATERIALS AND METHODS

Study design

According to the criteria for inclusion, the clinical data and results of laboratory tests and instrumental examination of T2DM patients were collected and analyzed. We divided the patients into the DGMD group and non-DGMD (NDGMD) group according to the results of EGG.

Study protocol

A total of 272 T2DM patients with good cognitive function and ages ranging from 18 to 80 were recruited from the endocrine department of the Second Affiliated Hospital of Soochow University and the Affiliated Suqian Hospital of Xuzhou Medical University. The diagnosis of T2DM was based on the diagnostic criteria of American Diabetes Association in 2023[19]. The exclusion criteria were as follows: (1) Pregnant or lactating female patients; (2) Patients with severe cardiovascular diseases, such as acute myocardial infarction, acute heart failure, chronic cardiac insufficiency decompensation, etc.; (3) Patients with acute cerebrovascular diseases such as cerebral infarction, acute cerebral hemorrhage, etc.; (4) Patients with previous gastric surgery, serious digestive system diseases, systemic diseases affecting digestive function or systemic diseases affecting digestive function or electrolyte disorders; (5) Patients with acute or chronic pancreatitis; (6) Patients who had used drugs affecting GE and autonomic nervous function in the past 2 weeks; (7) Patients with acute complications such as hypertonic hyperglycemia state, diabetic ketoacidosis, severe hepatic and renal insufficiency (transaminases level more than 2 times that of normal, estimated glomerular filtration rate < 45 mL/minutes/1.73 m²); (8) Those with a long history of heavy drinking not less than 5 years, ethanol amount 70 g/week for women and 140 g/week for man; (9) Neurological damage caused by other reasons, such as stroke, Guillain Barre syndrome, connective tissue disease, etc.; (10) Patients with severe cataract and unable to complete fundus examination; and (11) Those with autoimmune hepatitis, viral hepatitis, drug-induced liver disease, Wilson’s disease and other diseases that can cause fatty liver. During the interviews, physicians completed case report forms recording the disease profiles of diabetes and related treatment histories. The protocol of the present study was approved by the Institutional Review Board of the Second Affiliated Hospital of Soochow University and the Affiliated Suqian Hospital of Xuzhou Medical University, and the approving registration number is JD-HG-2022-07 and 2024036. The patients’ personal privacy was protected, the sample data was stored electronically in a special computer with a password, which was only available to researchers. The patients’ medical records were kept in the hospital.

Clinical and biochemical measurements

Each patient received anthropometric measurements, including body weight and height, measured in light clothes and bare feet. Body mass index (BMI) was calculated according to the formula: BMI (kg/m²) = weight (kg)/ height (m)².

Any medication that interferes with gastric myoelectric activity was discontinued 48 hours before EGG. The patient was asked to lie down and the original waveform was acquired for 30 minutes in the fasting state. Then the patient was asked to sit and had a test meal (Carbohydrates: 5 Daliyuan French buns, with a total weight of 100 g and energy content of 1422 kJ, Protein: 1 ham sausage, with a total weight of 60 g and energy content of 435 kJ, and drank water of less than 400 mL) within 5 minutes. This test meal may not be applicable for all populations due to regional dietary habits, potentially affecting the generalizability of EGG results. Future research should consider using more standardized or culturally adapted test meals to improve the reproducibility of gastric motility measurements. The patients then lied down and collection of the original waveform was continued for 30 minutes after the meal. EGG analysis software was used to analyze the data automatically. We recorded the proportion of slow waves, bradygastria, tachygasria, rhythm disturbance, main power before and after meals and power ratio (PR). PR reflects the power change of EGG before and after the meal. The PR value of healthy adults should be greater than 1, otherwise it will prompt the gastric motility disorders.

MASLD was diagnosed based on the 2022 guideline of the American Association of Clinical Endocrinology (AACE)[20]. The grading criteria for fatty liver severity are as follows: (1) Mild MASLD: Minimal diffuse increase in the fine echoes. The liver appears bright compared to the cortex of the kidney. Normal visualization of the diaphragm and intrahepatic vessel borders; (2) Moderate MASLD: Moderate diffuse increase in the fine echoes, and slightly impaired visualization of the intrahepatic vessels and diaphragm; and (3) Severe MASLD (S-MASLD): Marked increase in the fine echoes. Poor or no visualization of intrahepatic vessels and diaphragm, and poor penetration of the posterior segment of the right lobe of the liver[21]. While ultrasound is a widely used non-invasive method for detecting hepatic steatosis, it lacks diagnostic accuracy, which may lead to potential misclassification of disease severity. Carotid plaque detection was made using the Philips Color Doppler Ultrasound Diagnostic Instrument. Diagnosis of atherosclerosis was according to the 2009 Chinese Physicians Association Ultrasound Division Vascular Ultrasound Guidelines[22]. The diagnosis of diabetic nephropathy refers to the 2020 Expert Consensus on Diagnosis and Treatment of Diabetic Nephropathy. The diagnosis of DR refers to the DR grading standards established by the International Society of Ophthalmology in 2002. The diagnosis of DPN refers to the clinical diagnostic criteria recommended by the 2009 Toronto International Conference on DPN[23].

All patients fasted for at least 8 hours before blood samples were collected. The fully automated blood cell analyzer (Sysmex XN-2000, Japan) was used to determine blood routines. Biochemical parameters, such as fasting blood glucose (FBG), triglyceride, total cholesterol, high-density lipoprotein, low-density lipoprotein, alanine transaminase, aspartate aminotransferase and creatinine were measured with an automated biochemical instrument (AU5800). Fully automated chemiluminescence immunoassay analyzer (MAGLUMI X8) was used for fasting C-peptides, 2 hours postprandial C-peptides and 25-hydroxyvitamin D [25(OH)D]. The TOSOH HLC-723 G8 was used for the detection of hemoglobin A1c (HbA1c).

Statistical analysis

Software SPSS 26.0 was used for data analysis. The single sample K-S test and S-W test were used to determine whether the data conform to normal distribution, with the S-W test being more robust in small samples (n ≤ 50). Data conforming to normal distribution in continuous variables were expressed as means ± SD, while non-normal distribution data were presented by median, and categorical variables were expressed as frequency (n) or percentage (%). Comparisons between different groups were tested using two-sample t-test, Mann-Whitney U test, and one-way analysis of variance (Bonferroni correction) or χ² test. Logistic regression analysis was used to evaluate the risk factors of DGMD. A P value less than 0.05 was considered statistically significant.

RESULTS

EGG characteristics in patients with T2DM

A total of 272 patients with T2DM were enrolled in the study. As shown in Table 1, EGG results indicated that 90 patients (33.09%) had DGMD. The mean proportion of slow waves was below 65% in both fasting and postprandial states, with postprandial bradygastria being the predominant manifestation.

Table 1 Electrogastrogram characteristics in type 2 diabetes patients.

EGG parameters	Proportions (%)
Average pre-meal slow-wave ratio	52.19
Average postprandial slow wave ratio	53.57
Average proportion of pre-meal bradygastria	23.05
Average proportion of postprandial bradygastria	23.50
Average proportion of pre-meal tachygasria	18.82
Average proportion of postprandial tachygasria	19.48
Average proportion of pre-meal rhythm disturbance	8.62
Average proportion of postprandial rhythm disturbance	5.55

ECG: Electrogastrogram.

Risk factors for DGMD in patients with T2DM

As presented in Table 2, the patients were categorized into the DGMD group and the NDGMD group. There was no difference in gender or age between the two groups. The duration of diabetes, BMI, FBG, fasting C-peptide, C-peptide 2h postprandial and HbA1c were higher in the DGMD group (Table 2, ^aP < 0.05, ^bP < 0.01, compared with NDGMD, two-sample t-test and Mann-Whitney U test). The 25(OH)D was lower in the DGMD group (Table 2, ^bP < 0.01, compared with NDGMD, two-sample t-test). The presence of MASLD was associated with the presence of DGMD (Table 2, ^bP < 0.01, χ² test). In the logistic regression analysis, we employed stepwise regression with the prevalence or absence of DGMD as the dependent variable and 8 other variables as independent variables. The results showed that the duration, FBG, 25(OH)D and the presence of MASLD remained independently associated with the risk of DGMD (Table 3, ^aP < 0.05, ^bP < 0.01), while BMI, Fasting C-peptide, C-peptide 2h postprandial and HbA1c was no longer related. The area under the ROC curve (AUC) for the FBG was 0.708, FBG+Duration was 0.715, FBG+Duration+25(OH)D was 0.750, FBG+Duration+25(OH)D+MASLD was 0.777, as shown in Figure 1 and Table 4. The AUC for composite factors was greater than any single conventional risk factor.

Figure 1 Receiver operating characteristic curves comparing the potential of different variables to predict diabetic gastric motility disorders. Model 1 (Blue), fasting blood glucose (FBG), area under the curve (AUC) is 0.708; model 2 (Red), FBG+Duration, AUC is 0.715; model 3 (Green), FBG+Duration+25-hydroxyvitamin D [25(OH)D], AUC is 0.750; model 4 (Orange), FBG+Duration+25(OH)D+ metabolic dysfunction-associated steatotic liver disease, AUC is 0.777. FBG: Fasting blood glucose; AUC: Area under the curve; MASLD: Metabolic dysfunction-associated steatotic liver disease; 25(OH)D: 25-hydroxyvitamin D.

Table 2 Clinical and biochemical characteristics in the diabetic gastric motility disorders group and non-diabetic gastric motility disorders group, n (%)/mean ± SD/median (25^th-75^th percentiles).

Variables	DGMD (n = 90)	NDGMD (n = 182)	P value
Gender (male vs female)	52.22% vs 47.78%	54.40% vs 45.60%	0.735
Age (year)	52.77 ± 12.757	53.4 ± 11.493	0.683
Duration (year)	10 (5-15)	6 (2-10)	< 0.01b
BMI (kg/m2)	26.501 ± 3.851	24.754 ± 3.722	< 0.01b
FBG (mmol/L)	13.53 (8.1-16.04)	9.14 (6.69-12.48)	< 0.01b
Fasting C-peptide (ng/mL)	1.785 (1.188-2.51)	1.485 (0.89-2.023)	< 0.01b
C-peptide 2h postprandial (ng/mL)	3.56 (2.17- 6.213)	3.145 (1.86-4.73)	< 0.05a
HbA1c (%)	10.353 ± 2.201	9.401 ± 2.184	< 0.01b
TG (mmol/L)	4.85 (4.07-5.71)	4.9 (4.07-5.79)	0.768
TC (mmol/L)	1.6 (1.19-2.705)	1.43 (1.04-2.87)	0.461
LDL-C (mmol/L)	3.132 ± 1.055	3.183 ± 1.061	0.724
HDL-C (mmol/L)	1.14 (1.01-1.39)	1.16 (0.98-1.41)	0.552
Cr (μmol/L)	58.231 ± 18.477	59.277 ± 19.878	0.752
ALT (mmol/L)	20.4 (15.175-26.6)	18.75 (14.7-25.6)	0.092
AST (mmol/L)	18.55 (15.7-22.325)	17.4 (14-22.475)	0.199
25(OH)D (ng/mL)	17.140 ± 5.756	21.416 ± 7.234	< 0.01b
DR (Y:N)1	24.72% vs 75.28%	34.44% vs 65.56%	0.105
DPN (Y:N)	41.11% vs 58.89%	48.90% vs 51.10%	0.225
DN (Y:N)	10.11% vs 89.89%	11.60% vs 88.40%	0.714
CP (Y:N)	52.27% vs 47.73%	52.75% vs 47.25%	0.942
HBP (Y:N)	38.20% vs 61.80%	34.62% vs 65.38%	0.563
MASLD (Y:N)	62.92% vs 37.08%	21.98% vs 78.02%	< 0.01b

^aP < 0.05.

^bP < 0.01. P values for differences between two groups were obtained by two-sample t-test, Mann-Whitney U-test or χ² test.

¹Y: N: Indicates that the number of people with this disease vs those without.

BMI: Body mass index; FBG: Fasting blood glucose; HbA1c: Glycated hemoglobin A1c; TG: Triglycerides; TC: Total cholesterol; LDL-c: Low density lipoprotein cholesterol; HDL-c: High density lipoprotein cholesterol; Cr: Creatinine; ALT: Alanine transminase; AST: Aspertate aminotransferase; 25(OH)D: 25-hydroxyvitamin D; DR: Diabetic retinopathy; DPN: Diabetic peripheral neuropathy; DN: Diabetic nephropathy; DCAN: Diabetic cardiac autonomic neuropathy; CP: Carotid plaque; HBP: High blood pressure; DGMD: Diabetic gastric motility disorders; MASLD: Metabolic dysfunction-associated steatotic liver disease.

Table 3 Diabetic gastric motility disorders as dependent variable in multiple logistic regression analysis¹.

Variables	β	OR (95%CI)	P value
Duration	0.066	1.068 (1.012-1.126)	< 0.01b
FBG	0.131	1.14 (1.059-1.226)	< 0.01b
25(OH)D	-0.061	0.941 (0.893-0.991)	< 0.05a
MASLD	0.935	2.548 (1.204-5.396)	< 0.05a
BMI	0.097	1.102 (0.999-1.215)	0.051
HbA1c	0.193	1.212 (0.987-1.489)	0.066
Fasting C-peptide	0.244	1.276 (0.728-2.237)	0.395
C-peptide 2h postprandial	0.022	1.022 (0.839-1.244)	0.829

^aP < 0.05.

^bP < 0.01.

¹In logistic regression analysis, diabetic gastric motility disorders, as dependent variable, and the other 5 variables, duration, fasting blood glucose, glycated hemoglobin A1c, Vit-D, and metabolic dysfunction-associated steatotic liver disease, as independent variables, were included in the same model. Duration, fasting blood glucose, 25-hydroxyvitamin D and metabolic dysfunction-associated steatotic liver disease were the risk factors of diabetic gastric motility disorders.

BMI: Body mass index; FBG: Fasting blood glucose; HbA1c: Glycated hemoglobin A1c; 25(OH)D: 25-hydroxyvitamin D; MASLD: Metabolic dysfunction-associated steatotic liver disease.

Table 4 Receiver operating characteristic curves comparing the potential of different variables to predict diabetic gastric motility disorders.

Variables	AUC	95%CI	Sensitivity	Specificity
FBG	0.708	0.635-0.767	0.494	0.824
FBG+Duration	0.715	0.672-0.812	0.782	0.635
FBG+Duration+25(OH)D	0.750	0.697-0.820	0.765	0.648
FBG+Duration+25(OH)D+MASLD	0.777	0.710-0.843	0.588	0.82

AUC: Area under the curve; FBG: Fasting blood glucose; 25(OH)D: 25-hydroxyvitamin D; MASLD: Metabolic dysfunction-associated steatotic liver disease.

Prevalence of DGMD in different subgroups of patients with T2DM

As shown in Figure 2, when patients were divided into three groups according to the quartile of the duration of diabetes, which were less than 5 years, 5-10 years, and longer than 10 years, the presence rates of DGMD in the three groups were 24.37%, 26.58% and 54.05%, respectively. The results indicated that the presence of DGMD increases significantly as diabetes progresses (Figure 2, ^aP < 0.05, χ² test). Patients were divided into four groups according to the FBG levels, which were lower than 6.1 mmol/L, 6.1-7.8 mmol/L, 7.8-10.0 mmol/L, and higher than 10.0 mmol/L. The presence rates of DGMD for the four groups were 13.33%, 16.98%, 30% and 46.27%, respectively. The results indicated that the presence of DGMD increased significantly with FBG increase (Figure 2, ^bP < 0.05, χ² test). Patients were divided into three groups according to 25(OH)D, which were lower than 20 ng/mL, 20-30 ng/mL and higher than 30 ng/mL. The presence rates of DGMD in the three groups were 44.44%, 24.05% and 13.33%, respectively. The results indicated that the presence of DGMD increased significantly with 25(OH)D decrease (Figure 2, ^cP < 0.05, χ² test). In addition, the prevalence of DGMD was also found to correlate positively with MASLD. The general prevalence of DGMD in T2DM with and without MASLD was 58.33% and 18.86%, respectively (Figure 2, ^dP < 0.05, χ² test).

Figure 2 Prevalence of diabetic gastric motility disorders in different subgroups of patients with type 2 diabetes mellitus. Prevalence of diabetic gastric motility disorders (DGMD), according to the duration of diabetes, was 24.37%, 26.58% and 54.05% for < 5, 5-10 and > 10 years, respectively. Prevalence of DGMD increased significantly with prolonged course of diabetes (^aP < 0.05 vs group longer than 10 years). Prevalence of DGMD, according to the fasting blood glucose (FBG) of patients, was 13.33%, 16.98%, 30% and 46.27% for < 6.1, 6.1-7.8, 7.8-10 and > 10 mmol/L, respectively. Prevalence of DGMD increased significantly with increased FBG of patients (^bP < 0.05 vs group higher than 10.0 mmol/L). Prevalence of DGMD, according to the 25-hydroxyvitamin D [25(OH)D] of patients, was 13.33%, 24.05% and 44.44% for > 30, 20-30 and < 20 ng/mL, respectively. Prevalence of DGMD increased significantly with decreased 25(OH)D of patients (^cP < 0.05 vs group lower than 20 ng/mL). Prevalence of DGMD in type 2 diabetes mellitus with metabolic dysfunction-associated steatotic liver disease (MASLD) was 58.33% and in type 2 diabetes mellitus without MASLD was 19.32%. Incidence of DGMD positively correlated with MASLD (r = 0.402, ^dP < 0.05). FBG: Fasting blood glucose; MASLD: Metabolic dysfunction-associated steatotic liver disease; 25(OH)D: 25-hydroxyvitamin D.

Effects of MASLD on DGMD in patients with T2DM

According to the above results, DGMD occurred in more than 50% of patients with MASLD. To further analyze the influence of MASLD on DGMD, as presented in Table 5, the participants were divided into the MASLD group and the non-MASLD (NMASLD) group. The postprandial bradygastria ratio, postprandial tachygasria ratio and pre-meal rhythm disturbance ratio were significantly higher in the MASLD group (^aP < 0.05, ^bP < 0.01, compared with NMASLD using two-sample t-test). The pre-meal slow-wave ratio and postprandial slow wave ratio were significantly lower in the MASLD group (^aP < 0.05, ^bP < 0.01, compared with NMASLD using two-sample t-test).

Table 5 Electrogastrogram characteristics in the metabolic dysfunction-associated steatotic liver disease group and non-metabolic dysfunction-associated steatotic liver disease group, means ± SD.

Variables	MASLD (n = 96)	NMASLD (n = 176)	P value
Pre-meal slow-wave ratio	45.009 ± 10.364	56.106 ± 14.401	< 0.01b
Postprandial slow wave ratio	42.664 ± 11.705	59.520 ± 12.375	< 0.05a
Pre-meal bradygastria ratio	23.323 ± 9.476	22.902 ± 10.364	0.742
Postprandial bradygastria ratio	30.009 ± 8.788	19.945 ± 9.280	< 0.05a
Pre-meal tachygasria ratio	18.365 ± 8.367	19.063 ± 8.731	0.523
Postprandial tachygasria ratio	26.445 ± 9.284	15.673 ± 7.342	< 0.05a
Pre-meal rhythm disturbance ratio	12.186 ± 5.379	6.681 ± 3.812	< 0.05a
Postprandial rhythm disturbance ratio	5.898 ± 3.095	5.362 ± 3.361	0.198

^aP < 0.05.

^bP < 0.01. P values for differences between two groups were obtained by two-sample t-test.

MASLD: Metabolic dysfunction-associated steatotic liver disease; NMASLD: Non-metabolic dysfunction-associated steatotic liver disease.

As shown in Table 6, the patients were divided into NMASLD group, mild-NMASLD group, moderate-NMASLD group and severe-NMASLD group. There were significant differences in the postprandial slow wave, postprandial bradygastria ratio, postprandial tachygasria ratio and pre-meal rhythm disturbance ratio among the four groups while there were no significant differences in pre-meal slow-wave ratio, pre-meal bradygastria ratio, pre-meal tachygasria ratio and postprandial rhythm disturbance ratio (Table 6, ^a,b,cP < 0.0083, one-way analysis of variance, Bonferroni correction). As the degree of NMASLD increased, the proportion of postprandial slow wave decreased and the proportion of postprandial bradygastria, postprandial tachygasria and pre-meal rhythm disturbance increased. Notably, the small sample size in the S-MASLD group (n = 4) may limit the generalizability of the statistical findings. These results should be interpreted with caution, and larger studies are required to validate the observed association.

Table 6 Electrogastrogram characteristics in the non-metabolic dysfunction-associated steatotic liver disease group, mild metabolic dysfunction-associated steatotic liver disease group, moderate metabolic dysfunction-associated steatotic liver disease group and severe metabolic dysfunction-associated steatotic liver disease group, means ± SD.

Group	NMASLD, n = 176	Mi-MASLD, n = 80	Mo-MASLD, n = 12	S-MASLD, n = 4	F	P value
Pre-meal slow-wave ratio	56.106 ± 14.401	45.883 ± 10.458a	40.493 ± 8.186a	41.073 ± 12.732a	15.526	0.000
Postprandial slow wave ratio	59.520 ± 12.375	42.445 ± 11.414a	45.981 ± 14.767a	37.083 ± 4.608a	40.453	0.000
Pre-meal bradygastria ratio	22.902 ± 10.364	22.813 ± 9.6735	26.315 ± 8.793	24.555 ± 4.463	0.477	0.698
Pre-meal bradygastria ratio	19.945 ± 9.280	29.684 ± 9.053a	30.833 ± 8.066a	34.050 ± 4.623a	25.507	0.000
Pre-meal tachygasria ratio	19.063 ± 8.731b	17.148 ± 7.331b	24.415 ± 11.503	24.555 ± 8.915	3.434	0.018
Postprandial tachygasria ratio, pre-meal rhythm disturbance ratio	15.673 ± 7.342, 6.681 ± 3.812	26.160 ± 9.528a, 11.680 ± 4.743a	27.973 ± 9.408a, 14.223 ± 6.335a	27.560 ± 1.032a, 16.198 ± 11.482a,c	36.827, 34.941	0.000, 0.000
Postprandial rhythm disturbance ratio	5.362 ± 3.361	5.819 ± 3.044	5.692 ± 3.577	8.098 ± 2.351	1.191	0.314

^aP < 0.0083, vs non-metabolic dysfunction-associated steatotic liver disease (MASLD) group.

^bP < 0.0083 vs mild MASLD group.

^cP < 0.0083 vs moderate MASLD. P values for differences between four groups were obtained by one-way analysis of variance (Bonferroni correction).

NMASLD: Non-metabolic dysfunction-associated steatotic liver disease; Mi-MASLD: Mild metabolic dysfunction-associated steatotic liver disease; Mo-MASLD: Moderate metabolic dysfunction-associated steatotic liver disease; S-MASLD: Severe metabolic dysfunction-associated steatotic liver disease.

In this study, the proportion of missing values is shown in Table 7, and the missing data were processed by multiple imputation method, and the stability of the imputation results was ensured with multiple iterations.

Table 7 Proportion of missing values.

Variables	Proportion of missing values, %
HbA1c	2.6
FBG	5.5
25(OH)D	25.7
HDL-C	12.9
LDL-C	11.0
Fasting C-peptide	2.9
C-peptide 2h postprandial	4.0

FBG: Fasting blood glucose; HbA1c: Glycated hemoglobin A1c; 25(OH)D: 25-hydroxyvitamin D; HDL-C: High density lipoprotein cholesterol; LDL-C: Low density lipoprotein cholesterol.

DISCUSSION

In 1937, Ferroir found that X-ray in DM patients showed decreased gastric motility. In 1945, Rundles reported that delayed GE was associated with diabetic autonomic neuropathy, and in 1958, Kassander indicated using diabetic gastroparesis (DGP) that delayed GE in DM patients was common, including asymptomatic patients[24]. The clinical manifestations of DGP include nausea, vomiting, feeling of fullness in the upper abdomen, and so on. DGP not only causes various digestive symptoms, but also affects the pharmacokinetics of hypoglycemic drugs, delays the absorption of hypoglycemic drugs and affects the efficacy of drugs, leading to poor blood glucose control[25].

The myoelectric activity of the stomach controls the contractile activity of the stomach and the disturbance of the gastric electrical rhythm is the basis of gastric motility disorder. Gastric electrical rhythm is disturbed, the effective mechanical contraction is not produced, resulting in decreased gastric motility and delayed GE[26]. EGG can objectively record the rhythm and power of gastric electricity and has a wide range of clinical application because of its non-invasiveness and good reproducibility[27].

Hyperglycemia is a recognized risk factor for chronic complications of diabetes, including DGMD. In this study, the FBG of patients with DGMD was significantly higher than those with NDGMD, which is consistent with our previous study[18,28]. In addition, DGMD also affects the control of blood glucose levels, and the effect of GE disorders on blood glucose is generally underestimated in diabetic patients receiving insulin therapy, leading to a vicious circle between blood glucose fluctuations and GE[29]. Therefore, delayed blood glucose spikes caused by GE disorders should be considered when using insulin therapy.

25(OH)D is one of the indispensable nutrients in the human body. Deficiency of 25(OH)D will have significant harm to body[30]. Long-term deficiency of 25(OH)D will cause abnormalities in the skeletal system and recent researches have shown that it also had a significant impact on other tissues, increasing the risk of diseases such as breast cancer, diabetes, blood lipid metabolism and cardiovascular diseases[31,32]. The present study shows that 25(OH)D deficiency increases the risk of DGMD. However, the exact mechanism is not clear. Further in-depth research on the targets and specific mechanism of 25(OH)D affecting the pathogenesis of DGMD is of great clinical significance for the prevention, early diagnosis and precision treatment of DGMD. Tryptophan (Trp) is an essential amino acid that is a precursor to bioactive compounds such as serotonin (5-HT), kynurenine and ligands for aryl hydrocarbon receptor. 5-HT is closely related to gastrointestinal motility[33]. Studies have shown that 25(OH)D has regulatory effects in multiple parts of Trp metabolism. Therefore, we speculate that the effect of 25(OH)D in DGMD may be achieved through Trp metabolism, which needs to be further verified.

Interestingly, it has been also found that MASLD is a risk factor for DGMD. MASLD is one of the most common liver diseases in clinical practice, which is a genetic-environmental-metabolic related liver disease[34]. In recent years, along with the economic development and the other all population’s quality of life enhancement, the incidence of MASLD has increased year by year[35,36]. In this study, the incidence of MASLD in hospitalized Chinese T2DM patients was calculated at 40.57% which is lower than that reported by Zhou et al[37](42.1%).

MASLD is associated with metabolic risk factors such as obesity, diabetes and dyslipidemia[36]. Numerous studies have shown that MASLD is linked to a heightened risk of macrovascular and microvascular complications in diabetic patients. Currently, there are few studies on the association between MASLD and DGMD. Most findings suggested that chronic liver disease is also a major cause of autonomic neuropathy after excluding related factors such as alcohol consumption and diabetes[37]. However, there is no authoritative conclusion on whether non-alcoholic fatty liver disease can be used as an independent risk factor to affect the development of DGMD.

In the present study, MASLD group patients had obvious gastric electrical rhythm abnormalities. The decrease of the proportion of slow waves reflected the irregular movement of the slow waves of the stomach. These results indicated that the MASLD patients had obvious gastric motility abnormalities and gastric electrical disorders, especially after eating, mainly including the decline of gastric contractile motor function and the lack of gastric motility after meals. MASLD might be associated with DGMD in individuals with T2DM. The “multiple hit hypothesis” proposes that in T2DM patients, disrupted inter-organ crosstalk among the intestine, liver, pancreas, adipose tissue and skeletal system may trigger multiple damages. These damage synergistically promote the development and progression of MASLD[7]. During MASLD development, intestinal dysbiosis and excessive fat accumulation in adipose tissue, skeletal muscle and liver induce inflammation and insulin resistance, giving rise to hyperglycemia and hyperinsulinemia[36]. As MASLD progresses, glycolipid toxicity enhances reactive oxygen species production and endoplasmic reticulum stress, leading to cell death. These dead cells interact with infiltrating hepatic inflammatory cells, free fatty acids, gut-derived lipopolysaccharides, and transforming growth factor-β from Kupffer cells, activating hepatic stellate cells (HSCs). Activated HSCs lead to liver fibrosis by increasing the extracellular matrix. Among these damages associated with progression of MASLD, hyperglycemia, insulin resistance, oxidative stress and inflammation also play roles in the development of DGMD[38]. The present study suggests the need for special attention to DGMD in individuals with T2DM and MASLD. Although patients had no obvious gastrointestinal symptoms in the early stage of MASLD, abnormal gastric myoelectric activity existed. T2DM patients with MASLD should be alert to the development of DGMD, and EGG should be performed in clinical practice. Measures such as weight loss, dietary modification, medication and other interventions to stop the progression of MASLD should be actively taken. Future research is needed to explore the molecular mechanism underlying the association between MASLD and DGP.

In summary, patients with T2DM were more likely to develop MASLD, and patients with MASLD can develop abnormal gastric electrical rhythm at an early stage. As the MASLD progresses, gastric electrical rhythm disorders become more pronounced. Therefore, for T2DM patients with MASLD, it is necessary to be alert to the occurrence of DGMD. Early EGG examination can be performed, and the EGG can be dynamically monitored to prevent the occurrence of DGMD.

This study participants were primarily Chinese hospitalized patients, whose lifestyle, dietary patterns, disease spectrum, and healthcare resource allocation differ from those in other countries. Therefore, caution is needed when generalizing the study results to non-Chinese populations. The data in this study were mainly sourced from tertiary hospitals, where diagnostic and treatment models differ significantly from primary care settings. Further research is needed in community healthcare and primary care settings to validate the generalizability of our findings.

This study has several limitations. First, it is unable to establish a casual relationship because of its cross-sectional nature. Second, we did not perform liver biopsy, the gold standard method for the diagnosis of MASLD. In addition, although ultrasound classification of MASLD was performed by experienced radiologists, the assessment remains subjective and operator-dependent. Moreover, inter-rater reliability metrics (such as Kappa values) were not calculated, which may introduce classification bias. Future research should incorporate formal inter-observer agreement analyses to validate diagnostic consistency. Third, the hospitalized cohort may over-represent severe metabolic dysfunction, potentially inflating effect sizes. However, this design enabled rigorous confounder control and mechanistic exploration. We will validate our findings in the outpatient cohort in the future as a next step in the study. Forth, this study lacks comprehensive lifestyle data. Their absence may affect the observed associations. Future studies with detailed lifestyle assessments are needed to clarify the independent roles of gastric motility and lifestyle factors in the development of these conditions. Despite these limitations, this study provides valuable insights into DGMD and MASLD in T2DM.

CONCLUSION

DGMD in T2DM patients was strongly associated with a long duration of diabetes and FBG, low 25(OH)D and the presence of MASLD. We should attach great importance to the indicated risk factors and actively intervene MASLD in T2DM patients.