Abstract
Intestinal obstruction is a common and serious condition within the digestive system, with a higher incidence observed in the elderly population. This condition can lead to a series of significant complications. In recent years, there has been growing attention on the adverse gastrointestinal effects associated with antipsychotic medications such as clozapine, yet the relationship between these drugs and intestinal obstruction requires systematic investigation. This study is based on the U.S. National Health and Nutrition Examination Survey (NHANES) database, integrating participant health status, nutritional intake, medication use, and imaging diagnostic information. Weighted analysis, forest plots, and neural network models were employed to explore the risk factors associated with intestinal obstruction. A total of 5226 participants were included in the study. Statistically significant differences were observed in age, gender, marital status, income, and nutritional intake between the intestinal obstruction group and the control group. Individuals aged over 60 years, females, and those with low income exhibited a higher risk of developing the condition. The risk of intestinal obstruction increased by 78% among users of clozapine, and high protein and low fiber intake were also identified as risk factors. The predictive performance of the neural network model was moderate, with energy intake being the most important variable. In the intestinal obstruction group, levels of potassium and magnesium were significantly elevated, which may suggest their role in the pathophysiology of the condition. This study identified several important risk factors associated with intestinal obstruction, particularly highlighting the effects of clozapine use and nutritional elements. The findings provide a basis for the identification of high-risk populations and early interventions, as well as directions for future intelligent predictions that incorporate imaging characteristics.
Keywords: clozapine, intestinal imaging abnormalities, intestinal obstruction, NHANES database, nutrient intake, Radiological assessment
1. Introduction
Intestinal obstruction is a severe condition within the digestive system, characterized by a complex pathogenesis. It primarily manifests as a blockage of the passage of intestinal contents, leading to symptoms such as abdominal pain, vomiting, abdominal distension, and cessation of gas and bowel movements.[1] This disease not only significantly affects the quality of life of patients but can also result in severe complications, including intestinal necrosis, perforation, peritonitis, and even life-threatening situations.[2] Epidemiological data indicate a higher incidence of intestinal obstruction in the elderly population, with a notable increase in the risk of occurrence as age advances. This presents substantial economic and psychological burdens for patients and their families, while also posing challenges to public health resources.[3]
In clinical practice, imaging modalities such as abdominal X-rays, CT scans, and MRI are crucial for the diagnosis and assessment of intestinal obstruction. These techniques can identify key signs such as the typical “air-fluid levels” bowel dilation, intestinal volvulus, or obstruction, thereby assisting in clarifying the site and severity of the blockage. These imaging features are not only used for diagnosis but are also frequently employed for disease classification, treatment decisions, and monitoring therapeutic efficacy. Imaging plays a significant role in the diagnosis and classification of various types of intestinal obstruction. Previous studies have indicated that typical CT findings in cases of intestinal obstruction complicated by ischemia, such as bowel wall thickening, hypoperfusion, and pneumatosis, exhibit high specificity.[4] In the specific case of adult intussusception, a subtype of intestinal obstruction, the “target sign” observed on CT demonstrates extremely high diagnostic sensitivity and can accurately assess its etiology (such as tumors or structural lesions) and guide treatment decisions.[5] This further reinforces the position of imaging as a critical starting point in the diagnosis and management of intestinal obstruction. This study focuses on the analysis of risk factors for intestinal obstruction, with the use of clozapine as a key independent variable.
As an antipsychotic medication, clozapine demonstrates significant efficacy in treating psychiatric disorders; however, its potential gastrointestinal adverse effects, including intestinal obstruction, have garnered considerable attention in the medical community.[6] Research indicates a significant association between clozapine use and the occurrence of intestinal obstruction, suggesting that clinical practitioners should carefully assess the potential risks and explore preventive measures. Notably, some drug-related gastrointestinal side effects can be detected early through imaging studies, such as gas retention and diffuse intestinal dilation, which suggest an indirect link between the drug and imaging findings.[7–9]
In addition to clozapine, this study also included several other independent variables, including age, gender, marital status, income level, nutrient intake (such as protein, fiber, magnesium, potassium), alcohol consumption habits, and underlying conditions like diabetes and heart disease. These factors may have varying degrees of association with the occurrence of intestinal obstruction.
Intestinal obstruction is a common digestive system disorder with a complex pathogenesis, and early diagnosis is crucial.[10,11] In recent years, imaging studies have played an important role in the early diagnosis of gastrointestinal diseases.[12] Some drug-related gastrointestinal side effects, such as gas retention and diffuse intestinal dilation, can be detected early through imaging studies,[13,14] which provides new insights into exploring the potential association between drugs and intestinal obstruction. This study aims to investigate the risk factors for intestinal obstruction by comprehensively analyzing multiple factors, including drugs, demographic characteristics, lifestyle, and underlying diseases, and to provide new perspectives for clinical diagnosis and prevention.
2. Materials and methods
2.1. Clinical data
The data for this study were derived from the National Health and Nutrition Examination Survey (NHANES) database (https://www.cdc.gov/nhanes/index.htm), which collects nationally representative data covering health status, lifestyle habits, nutritional status, and other key health-related indicators. All participants provided written informed consent after obtaining approval from the NCHS Ethical Review Board, and this study uses only published data, so no additional informed consent is required. The study population included diverse participants across age groups, genders, and ethnic backgrounds. NHANES employs rigorously standardized data collection and processing protocols, including face-to-face interviews, physical examinations, and laboratory tests, all done by trained medical professionals. Data is cleaned and verified before being publicly released. This study utilized NHANES data to analyze the risk factors for intestinal obstruction and the differences in health status among different populations. It specifically divided the participants into clozapine users and non-clozapine users to investigate the characteristics of intestinal obstruction and the predictors of clozapine use. The findings will provide key evidence for strategies to prevent and control bowel obstruction and help improve health outcomes in the population.
2.2. Data collection
Based on previous research findings and available data, this study selected a series of variables potentially associated with intestinal obstruction, including demographic characteristics (e.g., age, sex), lifestyle factors, nutritional intake levels, and underlying medical conditions. The relevant data were obtained from the NHANES database through standardized questionnaires or laboratory tests, encompassing information on age, sex, body mass index, medication use, nutritional intake, chronic disease history, and health behaviors. To address missing values in the dataset, appropriate statistical methods such as multiple imputation were employed to ensure sample completeness and analytical accuracy. Through comprehensive analysis of these multidimensional variables, this study aims to systematically explore potential risk factors for intestinal obstruction.
2.2.1. Inclusion criteria
Age: participant age ≥18 years.
Imaging tests: The diameter of the small intestine dilation tube ≥2.5 cm; Colon dilation tube diameter ≥6 cm; The diameter of the cecum dilation tube ≥9 cm.
Clinical signs: Participants present with symptoms such as abdominal pain, bloating, inability to defecate, or gas.
Data integrity: Participants have a complete clinical medical record record.
Informed consent: All participants or their authorized person have provided informed consent.
2.2.2. Exclusion criteria
Incomplete imaging tests: No CT scan or no contrast on CT scan; Imaging data are missing.
Early postoperative intestinal obstruction: Intestinal obstruction that occurs within 6 weeks after surgery.
Other diseases: Active abdominal tumor (primary tumor, metastasis, peritoneal cancer); Hernia incarceration; Inflammatory bowel disease; Congenital bowel obstruction.
Special Circumstances: Acute gastrointestinal impairment; Cardiopulmonary, liver and kidney dysfunction; End-stage multiple organ failure; Psychiatric illness or cognitive impairment; Refusal to enroll.
2.3. Statistical analysis
All statistical analyses were performed in accordance with relevant guidelines, with sampling weights used to generate nationally representative prevalence estimates. For normally distributed continuous data, the mean ± standard deviation is reported, and the independent sample t-test is used for comparisons between groups if the variance is even, and the Welch t-test is used if the variance is uneven. Non-normally distributed data were described with median and interquartile spacing (M [P25, P75]), and the Mann–Whitney U test was used for comparisons between groups. Categorical variables were expressed as quantities and percentages, and the chi-square test was used for comparison between groups. Forest plots were examined by logistic regression analysis to examine the effects of nutrients and clozapine on intestinal obstruction. Spearman rank correlation analysis was used to investigate the association between nutritional factors and intestinal obstruction. Subgroup analysis and interaction testing were performed to verify the relationship between psychotropic drugs and intestinal obstruction. The difference was statistically significant in P < .05. Receiver operating characteristic (ROC) curve analysis was used to evaluate the prediction accuracy of the model containing psychotropic drugs and nutritional elements. In addition, a neural network model for classification prediction of intestinal obstruction and variable importance evaluation is constructed, and the model performance is evaluated by confusion matrix analysis.
2.4. Statistical tools
SPSS: for data management and statistical analysis. Version 28.0, provided by IBM Corporation (Armonk, NY).
Jupyter: For data processing and visualization. Version 1.0.0, provided by Project Jupyter (Berkeley, CA).
GraphPad Prism: For statistical analysis and graph drawing. Version 10.1.2, provided by GraphPad Software, Inc. (San Diego, CA).
3. Results
3.1. Baseline characteristics of participants
Baseline characteristics stratified by radiologically-confirmed intestinal obstruction status revealed significant differences between patients with intestinal obstruction (n = 1204) and controls (n = 4022) across multiple clinical variables. Age distribution showed 471 cases (39.1%) in individuals >60 years compared to 315 cases (26.2%) in the 18 to 39 age group, demonstrating a significant age-related risk gradient (P < .001). Gender differences were evident, with 725 female cases (60.2%) versus 479 male cases (47.7%) (P < .001). Marital status analysis revealed 671 cases (55.7%) among married/cohabitating individuals compared to 172 cases (14.3%) in unmarried participants (P < .001). Income disparities showed higher obstruction prevalence in low-income (44.4%) versus high-income groups (29.1%). Alcohol consumption patterns indicated 50.7% prevalence among occasional drinkers versus 6.4% in regular drinkers (P = .030).
Nutritional analysis demonstrated statistically significant differences in most nutrients (P < .05), with obstruction patients showing lower median protein intake (0.85 [IQR 0.015–4.47] g/day) versus controls (1.26 [IQR 0.2375–5.29] g/day) (P = .003). Similar patterns were observed for other nutrients except calcium, cholesterol and vitamin C (Table 1).
Table 1.
Baseline characteristics of included participants.
| Variable | Intestinal obstruction | P-value | |
|---|---|---|---|
| No (n = 4022) | Yes (n = 1204) | ||
| Age | |||
| 19–39 yr old | 1430 (35.6%) | 315 (26.2%) | <.001 |
| 39–60 yr old | 1457 (36.2%) | 418 (34.7%) | |
| >60 yr old | 1135 (28.2%) | 471 (39.1%) | |
| Gender | |||
| Male | 2103 (52.3%) | 479 (47.7%) | <.001 |
| Female | 1919 (39.8%) | 725 (60.2%) | |
| Race | |||
| Mexican American | 735 (18.3%) | 221 (18.4%) | .278 |
| Non-Hispanic White | 1980 (49.2%) | 584 (48.5%) | |
| Non-Hispanic Black | 697 (17.3%) | 218 (18.1%) | |
| Marital status | |||
| Married/Living with partner | 2473 (61.5%) | 671 (55.7%) | <.001 |
| Never married | 724 (18.0%) | 172 (14.3%) | |
| Widowed/Divorced/Separated | 821 (20.4%) | 351 (29.2%) | |
| Educational | |||
| <9th grade | 465 (11.6%) | 149 (12.4%) | .501 |
| 9th–11th grade or high school | 627 (15.6%) | 199 (16.5%) | |
| High school | 924 (23.0%) | 287 (23.8%) | |
| College or above | 1997 (49.7%) | 558 (46.3%) | |
| Income_level | |||
| Low income | 1627 (40.5%) | 535 (44.4%) | <.001 |
| Middle income | 1016 (25.3%) | 319 (26.5%) | |
| High income | 1379 (34.3%) | 350 (29.1%) | |
| BMI | 29.0734 ± 6.83123 | 29.56 ± 6.64 | .029 |
| Diabetes | |||
| No | 3564 (88.6%) | 975 (81.0%) | <.001 |
| Yes | 458 (11.4%) | 206 (17.1%) | |
| Heart disease | |||
| No | 3734 (92.8%) | 1046 (86.9%) | <.001 |
| Yes | 288 (7.2%) | 145 (12%) | |
| Clozapine | |||
| No | 2881 (71.6%) | 688 (57.1%) | <.001 |
| Yes | 1141 (28.4%) | 507 (42.1%) | |
| Energy intake | 1963 (1476–2612.25) | 1789.5 (1337–2350.5) | <.001 |
| Calcium | 28 (10–91) | 26 (8–84.5) | .054 |
| Protein | 1.26 (0.2375–5.29) | 0.85 (0.015–4.47) | .003 |
| Fiber | 0 (0–1.3) | 0 (0–1) | .005 |
| Magnesium | 17 (9–32) | 15 (8–29) | .008 |
| Sodium | 196.09 ± 361.364 | 172.16 ± 332.204 | .040 |
| Potassium | 128 (46–258.25) | 107 (17.25–238.5) | .002 |
| Total fat | 0.225 (0.01–4) | 0.155 (0–2.715) | .008 |
| Cholesterol | 0 (0–0) | 0 (0–0) | .090 |
| Iron | 0.22 (0.04–1.76) | 0.16 (0.03–1.47) | .020 |
| Zinc | 0.21 (0.675–0.95) | 0.17 (0.05–0.74) | .009 |
| Copper | 0.0923 ± 0.122 | 0.084 ± 0.0.111 | .034 |
| Carbohydrate | 18.1174 ± 30.58 | 15.89 ± 32.84 | .029 |
| Vitamin A | 0 (0–51.25) | 0 (0–21.75) | .035 |
| Vitamin B | 0.057 (0–0.194) | 0.0425 (0–0.154) | .004 |
| Vitamin C | 0 (0–0.3) | 0 (0–0.1) | .147 |
| Smoking | |||
| Never smoke | 2204 (54.8%) | 618 (51.3%) | .188 |
| Former smoke | 936 (23.3%) | 333 (27.7%) | |
| Current smoke | 882 (21.9%) | 251 (20.8%) | |
| Drinking | |||
| Never drank | 1034 (25.7%) | 349 (29%) | .030 |
| Occasionally drank | 1934 (48.1%) | 610 (50.7%) | |
| Regularly drank | 361 (9.0%) | 77 (6.4%) | |
| Heavy drinker | 693 (17.2%) | 156 (13.0%) | |
3.2. Forest diagram analysis of associations between imaging gastrointestinal obstruction and including variables
The forest plot illustrates the association between 7 variables (diabetes, heart disease, clozapine, fiber, and protein) and the risk of intestinal obstruction. Each variable is represented by an odds ratio (OR) and its 95% confidence interval, along with a corresponding P-value. The analysis reveals the relationship between diabetes, heart disease, clozapine, fiber, and protein and the risk of intestinal obstruction. Diabetes shows a significant trend (OR = 1.607, P < .001), indicating that individuals with diabetes have a 1.607 times higher risk of intestinal obstruction compared to those without diabetes. Heart disease also shows a significant trend (OR = 1.606, P < .001), suggesting that individuals with heart disease have a 1.606 times higher risk of intestinal obstruction compared to those without heart disease. Clozapine use shows a significant trend (OR = 1.783, P < .001), meaning that individuals taking clozapine have a 78% higher risk of intestinal obstruction compared to those not taking it. Fiber intake also shows a significant trend (OR = 0.991, P = .043), indicating that an increase of one unit in fiber intake reduces the risk of intestinal obstruction by 0.9%. Protein intake shows a significant trend (OR = 0.963, P = .039), meaning that an increase of one unit in protein intake increases the risk of intestinal obstruction by 3.7% (Fig. 1).
Figure 1.
The flowchart for building a neural network-based risk prediction model for intestinal obstruction shows the complete process from data preparation to model evaluation and feature analysis, providing clear guidance for constructing an effective risk prediction model for intestinal obstruction.
3.3. Construction of neural network prediction based on imaging intestinal obstruction–evaluation of the importance of independent variables
In this study, we employed neural networks, an advanced machine learning technique, to construct the predictive model. A neural network is a computational model that simulates the structure and function of neurons in the human brain. It can automatically learn complex patterns and relationships from a large amount of input data. Composed of multiple layers, including the input layer, hidden layers, and the output layer, each neuron processes the input signals through weights and activation functions and then passes the results to the next layer. The strength of neural networks lies in their ability to automatically extract features and perform complex nonlinear mappings, which makes them particularly effective in handling complex medical imaging data. The construction process of the intestinal obstruction risk prediction model based on neural networks (Fig. 2).
Figure 2.
Forest plot of the effects of variables on intestinal obstruction. It shows the strength of association and confidence intervals between multiple clinical and lifestyle-related variables and the risk of intestinal obstruction.
In our neural network predictive model, the analysis of feature importance produced the following findings: Energy intake stood out as the most critical feature, with a significance score of .407. This score unequivocally shows that, within the framework of our studied model and context, energy intake is considered the most important variable. In contrast, other features such as copper, protein, fiber, magnesium, sodium, potassium, zinc, and vitamin B demonstrated relatively low significance scores, indicating that they play a comparatively minor role in the model. Furthermore, the chart titled “The Importance of Standardization” suggests that the standardization process might have significantly influenced the weights assigned to these features.
Overall, apart from energy intake, there were no apparent significant differences in the importance of the other features, suggesting that they were less distinguishable within the model (Fig. 3).
Figure 3.
Important features displayed the importance scores of different nutrients or components in English.
3.4. Subgroup analysis and interaction analysis to verify the association between clozapine drugs and imaging-based intestinal obstruction disease
To carefully examine the reliability and robustness of the relationship between clozapine drugs and imaging-based intestinal obstruction, subgroup analyses were conducted. The summary results of these subgroup analyses are presented in Table 2. Consistent with previous findings in the entire population, taking clozapine drugs is associated with an increased OR for imaging-based intestinal obstruction, indicating consistent results across all subgroups (Table 2).
Table 2.
Subgroup and interaction analysis for the association between clozapine medication and intestinal obstruction disease.
| Variable | OR (95% CI) | P-value |
|---|---|---|
| Age | ||
| 19–39 yr old | 2.171 (1.689–2.789) | <.001 |
| 39–60 yr old | 1.760 (1.408–2.199) | |
| >60 yr old | 1.803 (1.435–2.265) | |
| Gender | ||
| Male | 1.712 (1.385–2.115) | <.001 |
| Female | 1.836 (1.541–2.187) | |
| Marital status | ||
| Married/living with partner | 1.880 (1.570–2.252) | <.001 |
| Never married | 2.071 (1.476–2.970) | |
| Widowed/divorced/separated | 1.557 (1.206–2.010) | |
| Income_level | ||
| Low income | 1.891 (1.551–2.306) | <.001 |
| Middle income | 1.933 (1.481–2.522) | .003 |
| High income | 1.637 (1.259–2.130) | .039 |
| Diabetes | ||
| No | 1.776 (1.523–2.048) | <.001 |
| Yes | 2.058 (1.473–2.877) | |
| Heart disease | ||
| No | 1.825 (1.582–2.105) | <.001 |
| Yes | 1.717 (1.141–2.584) | .010 |
| Drinking | ||
| Never drank | 1.734 (1.351–2.225) | <.001 |
| Occasionally drank | 2.063 (1.708–2.493) | <.001 |
| Regularly drank | 1.448 (0.849–2.468) | .174 |
| Heavy drinker | 1.472 (1.020–2.125) | .039 |
3.5. Correlation analysis between different variables and imaging-based intestinal obstruction
Through the correlation analysis of variables related to intestinal obstruction, we found that age is significantly positively correlated with the occurrence of intestinal obstruction (R = 0.107, P < .001). This suggests that the risk of developing intestinal obstruction may increase with age. Gender is also significantly positively correlated with the occurrence of intestinal obstruction (R = 0.105, P < .001), indicating that men have a slightly lower risk of developing intestinal obstruction compared to women. Marital status is also significantly positively correlated with the occurrence of intestinal obstruction (r = -0.070, P < .001), suggesting that those who are married or cohabiting have a lower risk of developing intestinal obstruction compared to other groups. Family income level is significantly negatively correlated with the occurrence of intestinal obstruction (r = −0.045, P < .001), indicating that higher family income is associated with a lower risk of developing intestinal obstruction. Among nutritional elements, appropriate intake of nutrients is significantly negatively correlated with the occurrence of intestinal obstruction (R < 0, P < .05), suggesting that higher nutrient intake is associated with a lower risk of developing intestinal obstruction. In terms of lifestyle habits, alcohol consumption is significantly negatively correlated with the occurrence of intestinal obstruction (r = −0.051, P < .001), indicating that higher alcohol consumption is associated with a lower risk of developing intestinal obstruction. Regarding diseases, having heart disease and diabetes is significantly positively correlated with the occurrence of intestinal obstruction (R > 0, P < .001), suggesting that people with heart disease and diabetes have a higher risk of developing intestinal obstruction. Taking clozapine drugs is significantly positively correlated with the occurrence of intestinal obstruction (R = 0.133, P < .001), indicating that taking clozapine drugs is associated with a higher risk of developing intestinal obstruction. These hypotheses need to be further studied and verified, and future studies can explore their mechanisms of action in depth (Table 3).
Table 3.
Correlation analysis between intestinal obstruction and different variables.
| Variable | Correlation coefficient | P-value |
|---|---|---|
| Age | 0.107** | <.01 |
| Gender | 0.105** | <.01 |
| Marital status | 0.070** | <.01 |
| Income_level | −0.045** | <.01 |
| Energy intake | −0.104** | <.01 |
| Protein | −0.053** | <.01 |
| Fiber | −0.037** | <.01 |
| Magnesium | −0.050** | <.01 |
| Sodium | −0.041** | <.01 |
| Potassium | −0.054** | <.01 |
| Total fat | −0.053** | <.01 |
| Iron | −0.047** | <.01 |
| Zinc | −0.045** | <.01 |
| Copper | −0.034* | <.05 |
| Carbohydrate | −0.049** | <.01 |
| Vitamin A | −0.035* | <.01 |
| Vitamin B | −0.056** | <.01 |
| Drinking | −0.051** | <.01 |
| Diabetes | 0.098** | <.01 |
| Heart disease | 0.091** | <.01 |
| Clozapine | 0.133** | <.01 |
When the confidence level (double test) is 0.05, the correlation is significant.
When the confidence level (double test) is 0.01, the correlation is significant.
3.6. Evaluating the predictive performance of risk factors for intestinal obstruction using ROC curves
In this study, we explored the relationship between intestinal obstruction and the included variables. By constructing a predictive model, we found that these factors can influence the risk of intestinal obstruction to some extent. Specifically, the figure illustrates the ROC curve, which is used to evaluate the performance of a classifier. The x-axis represents the false positive rate, and the y-axis represents the true positive rate. The orange solid line in the figure represents the ROC curve, with an area of 0.64, indicating that the classifier has good discrimination ability. The blue dashed line represents the random guess line, with an area of 0.5, serving as a reference benchmark. The closer the ROC curve is to the top-left corner, the better the classification effect (Fig. 4).
Figure 4.
ROC curve for classifier performance evaluation. The ROC curve shows the recognition ability of the classification model to predict intestinal obstruction, and the area under the curve is used to measure the overall discriminant performance of the model. ROC = receiver operating characteristic.
3.7. Confusion matrix analysis of clozapine drugs, nutritional elements and related factors for imaging-based intestinal obstruction diseases
In this study, we analyzed the confusion matrix is an important evaluation tool for comparing the differences between the predictions of a classification model and the actual labels. In this article, we will analyze the confusion matrix to gain insight into the performance of the model. Through the analysis of the confusion matrix, we draw the following conclusions: the accuracy of the model is about 66.63%, indicating that the prediction accuracy is generally good. The precision was about 77.88%, indicating that a good proportion of the samples predicted to be positive were indeed positive. The recall rate was about 78.86% and the F1 score was about 78.36%, and it can be seen from these indicators that the model has a high recall rate and accuracy, which indicates that the model performs well in identifying intestinal obstruction. The high recall rate means that the model is able to identify the majority of actual cases of intestinal obstruction (Fig. 5).
Figure 5.
Random Forest model’s confusion matrix. It is used to evaluate the performance of the classification algorithm.
3.8. Distribution of intestinal obstruction diseases and related influencing factors based on imaging
In the study of intestinal obstruction based on imaging, we analyzed the distribution of copper, potassium and magnesium (Fig. 6).
Figure 6.
Box plot showed the distribution of 2 sets of data (“intestine obstruction” and “no”).
As can be seen from the left figure (copper), the copper content in the intestinal obstruction group is generally at a low level, with a median close to 0.5. The copper content in the non-obstruction group is slightly higher than that in the intestinal obstruction group, with a median also close to 0.5, but its overall distribution range is a bit wider. Statistical analysis shows that there is no significant difference in copper content between the intestinal obstruction group and the non-obstruction group.
The middle part of figure (potassium) indicates that the potassium content in the intestinal obstruction group has a wide distribution range, with a median of approximately 2000. In contrast, the potassium content in the non - obstruction group is significantly lower than that in the intestinal obstruction group, with a median of about 1000. This suggests that there is a marked difference in potassium content between the state of intestinal obstruction and non-obstruction.
From the right part of figure (magnesium), we can observe that the magnesium content in the intestinal obstruction group has a relatively wide distribution, with a median of around 300. The magnesium content in the non-obstruction group is significantly lower than that in the intestinal obstruction group, with a median of approximately 150. This demonstrates that there is a significant difference in magnesium content between the intestinal obstruction group and the non - obstruction group.
In conclusion, among the samples included in this study, there is no significant difference in copper content between the intestinal obstruction group and the non - obstruction group. However, the potassium and magnesium contents in the intestinal obstruction group are significantly higher than those in the non-obstruction group, suggesting that these 2 elements may have higher accumulation or demand in the state of intestinal obstruction, but the specific mechanisms require further research.
4. Discussion
This study focused on potential risk factors for intestinal obstruction, particularly the effects of clozapine drug and nutrient intake, and analyzed the heterogeneity of health status in different populations. A total of 5226 participants were included in the study, who were divided into intestinal obstruction group and control group based on imaging diagnosis. The results showed significant differences in age, gender, marital status, income level, and nutritional intake between the 2 groups. Through forest plots, neural network models, subgroup analysis, correlation analysis, and confusion matrix, this study further clarified the significant association between clozapine drugs, nutritional intake, and underlying diseases and the risk of intestinal obstruction. This provides data support and theoretical basis for early identification and intervention of intestinal obstruction based on imaging diagnosis.
This study found that the use of clozapine is significantly associated with an increased risk of intestinal obstruction (OR = 1.783, P < .001), a trend consistent across all subgroups. Clozapine, by antagonizing dopamine and serotonin receptors, may inhibit the contraction of intestinal smooth muscle and neurotransmission, leading to a decline in intestinal motility.[15,16] Clozapine has a strong anticholinergic effect, and its main mechanism is to block acetylcholine M receptors, resulting in relaxation of gastrointestinal smooth muscles and slowing down of gastrointestinal peristalsis.[17] This effect not only directly affects the motor function of the gastrointestinal tract, but may also inhibit the secretion of the digestive glands, further increasing the risk of constipation and intestinal obstruction.[18] Acetylcholine plays a key role in regulating gastrointestinal motility, and the blockade of its receptors can significantly reduce the voluntary motility of the gastrointestinal tract, thereby increasing the probability of intestinal obstruction.[18] The risk of intestinal obstruction due to clozapine is positively correlated with drug dose and blood concentration. High doses and high blood concentrations may further exacerbate gastrointestinal motility dysfunction.[19,20] In addition, some patients may experience intestinal obstruction even at low doses, suggesting that individual differences also play an important role.[21] Individual differences in sensitivity to clozapine may be related to genetic background, underlying disease status, and other medications used concurrently.[22] Imaging studies often show diffuse dilation of the bowel, accumulation of fluid and gas planes, and even signs of pseudo-obstruction, indicating that imaging can serve as an important auxiliary tool for monitoring drug safety.[23]
The importance of nutrient intake in regulating intestinal function is increasingly recognized. This study found that fiber intake is negatively correlated with the risk of intestinal obstruction (OR = 0.991). Fiber increases the volume of intestinal contents, promotes intestinal peristalsis, and reduces the retention time of intestinal contents, thereby lowering the risk of intestinal obstruction. Additionally, fiber can absorb water, soften stools, and reduce the occurrence of constipation.[24] Protein intake was negatively correlated with the risk of intestinal obstruction (OR = 0.963, P = .039). Protein is an essential nutrient for the human body and is involved in a variety of physiological functions, including maintaining the integrity of the intestinal mucosa, regulating the balance of intestinal flora, and promoting intestinal peristalsis.[25] The results of this study suggest that increased protein intake may reduce the risk of intestinal obstruction through these mechanisms. This finding has important guiding implications for those at high risk of intestinal obstruction, such as the elderly, post-surgery patients, or people with chronic bowel disease.
This study found that potassium and magnesium levels were significantly lower in the intestinal obstruction group compared to the non-obstructive group. In addition, the study observed that serum potassium (r = −0.054, P = .043) and magnesium (r = −0.050, P = .043) levels were negatively correlated with intestinal obstruction. Potassium and magnesium are crucial for maintaining the electrical activity and rhythmic movement of the intestine because they play an important role in the excitation-contraction coupling process of smooth muscle.[26] A reduction in potassium and magnesium levels may lead to dysfunction of the intestinal smooth muscle, thereby increasing the risk of intestinal obstruction.[27] It is worth noting that, although a statistically significant correlation was observed between potassium and magnesium levels and intestinal obstruction, the correlation coefficient (R-value) was very close to 0, indicating that this correlation may be very weak. Future studies may need to further explore the exact relationship between these electrolyte levels and intestinal obstruction.
Beyond clozapine and nutrition, our study links other factors to intestinal obstruction risk.Elderly individuals (over 60) have the highest incidence (39.1%) due to declining intestinal function, slowed peristalsis, and chronic diseases like diabetes and heart disease, which worsen dysfunction and show clear imaging signs.[28] Women (60.2% of cases) are more susceptible, likely due to physiological changes during menstruation, pregnancy, and menopause, which affect motility.[29] Imaging shows higher structural abnormality rates in women. Unmarried people have a lower risk (55.7% vs 14.3% unmarried), which may be due to a stable lifestyle and better health awareness. Diabetic and heart disease patients face 1.607 and 1.606 times higher risks, respectively, due to autonomic neuropathy, ischemia, and inflammation. Moderate alcohol may protect, but excess harms gut function. Low-income groups have higher risks (44.4% vs 29.1%) from poor nutrition, limited healthcare access, and delayed imaging.
The neural network prediction model developed in this study demonstrates moderate predictive power in identifying the risk of intestinal obstruction (AUC = 0.64, accuracy 66.63%, F1 score 78.36%). Energy intake is identified as the most critical variable in the model, which may be closely related to its impact on overall digestive load and intestinal tension regulation. Notably, while the model is based on traditional variables, its structure and training mechanism can serve as a theoretical blueprint for future deep learning-assisted diagnostic approaches that integrate imaging data (such as CT image features), thereby achieving a precise evaluation path that combines imaging and clinical information.
This study has some limitations. data collection and analysis methods may have limitations that require further validation in future studies. Nevertheless, this study provides important insights into the risk factors for intestinal obstruction and provides a basis for future research.
5. Conclusion
This study systematically evaluated potential risk factors and analyzed the differences in health characteristics among various populations. The findings indicate that age, gender, marital status, income level, drinking habits, nutrient intake (such as fiber, protein, potassium, magnesium), and underlying conditions (such as diabetes and heart disease) are closely linked to the occurrence of intestinal obstruction. Particularly, the use of clozapine psychotropic drugs significantly increases the risk of intestinal obstruction, suggesting that their potential gastrointestinal side effects and possible imaging features should be carefully considered in clinical practice.
Furthermore, the neural network prediction model based on clinical variables developed in this study demonstrates excellent risk identification capabilities, providing a crucial basis for early warning and intervention strategies for intestinal obstruction. This model can also be integrated with imaging features such as abdominal CT scans in the future, promoting the development of intelligent diagnostic systems. Future research could further explore the mechanisms of nutritional and pharmacological factors and develop more personalized risk management plans.
Author contributions
Conceptualization: Jianfei Wang.
Data curation: Yunyun Yang, Zhiping He.
Formal analysis: Yunyun Yang, Zhiping He.
Methodology: Jianfei Wang, Yunyun Yang.
Writing – review & editing: Jianfei Wang.
Writing – original draft: Zhiping He.
Abbreviations:
- NHANES
- National Health and Nutrition Examination Survey
- OR
- odds ratio
- ROC
- receiver operating characteristic
The views expressed in the submitted article are the authors’ own and not an official position of the institution or funder.
The data in this article are from public databases and are exempt from ethical review.
The authors have no funding and conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
How to cite this article: Wang J, Yang Y, He Z. Imaging-based risk assessment of intestinal obstruction: The impact mechanisms of clozapine-class drugs and nutritional elements. Medicine 2025;104:37(e44591).
Contributor Information
Jianfei Wang, Email: 77819325@qq.com.
Yunyun Yang, Email: 304719893@qq.com.
References
- [1].Jackson P, Vigiola Cruz M. Intestinal obstruction: evaluation and management. Am Fam Physician. 2018;98:362–7. [PubMed] [Google Scholar]
- [2].He S, Jiang H, Wang H. Iatrogenic small intestinal obstruction. Rev Esp Enferm Dig. 2023;115:717–8. [DOI] [PubMed] [Google Scholar]
- [3].Dumic I, Nordin T, Jecmenica M, Stojkovic Lalosevic M, Milosavljevic T, Milovanovic T. Gastrointestinal tract disorders in older age. Can J Gastroenterol Hepatol. 2019;2019:6757524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Cox VL, Tahvildari AM, Johnson B, Wei W, Jeffrey RB. Bowel obstruction complicated by ischemia: analysis of CT findings. Abdom Radiol (NY). 2018;43:3227–32. [DOI] [PubMed] [Google Scholar]
- [5].González-Carreró Sixto C Baleato González S, García Palacios JD, et al. Intestinal intussusception in adults: location, causes, symptoms, and therapeutic management [published online ahead of print August 25, 2021]. Radiologia (Engl Ed). doi: 10.1016/j.rx.2021.06.008. [DOI] [PubMed] [Google Scholar]
- [6].Xu Y, Amdanee N, Zhang X. Antipsychotic-induced constipation: a review of the pathogenesis, clinical diagnosis, and treatment. CNS Drugs. 2021;35:1265–74. [DOI] [PubMed] [Google Scholar]
- [7].Perez AA, Pickhardt PJ. Intestinal malrotation in adults: prevalence and findings based on CT colonography. Abdom Radiol (NY). 2021;46:3002–10. [DOI] [PubMed] [Google Scholar]
- [8].Liang S, Cui JH, Hu LL. Abdominal cocoon sign: an unusual cause of intestinal obstruction. Rev Esp Enferm Dig. 2021;113:725–6. [DOI] [PubMed] [Google Scholar]
- [9].Yamaguchi Y, Miwa T, Murakami R, et al. A case report of olmesartan-associated sprue-like enteropathy: diagnosis and healing confirmed by capsule endoscopy. DEN Open. 2023;3:e142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Attallah O, Sharkas M. GASTRO-CADx: a three stages framework for diagnosing gastrointestinal diseases. PeerJ Comput Sci. 2021;7:e423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Peery AF, Crockett SD, Murphy CC, et al. Burden and cost of gastrointestinal, liver, and pancreatic diseases in the United States: update 2021. Gastroenterology. 2022;162:621–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Amzallag-Bellenger E, Oudjit A, Ruiz A, Cadiot G, Soyer PA, Hoeffel CC. Effectiveness of MR enterography for the assessment of small-bowel diseases beyond Crohn disease. Radiographics. 2012;32:1423–44. [DOI] [PubMed] [Google Scholar]
- [13].Zahnle M, McCormack SM, Tiesenga F, Jorge J. Unresectable mesenteric mass causing small bowel obstruction. Cureus. 2022;14:e29871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Kudaravalli P, Saleem SA, Goodman A, Pendela VS, Arif MO. Bouveret syndrome as a rare cause of gastric outlet obstruction. Proc (Bayl Univ Med Cent). 2020;33:235–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Spencer NJ, Hu H. Enteric nervous system: sensory transduction, neural circuits and gastrointestinal motility. Nat Rev Gastroenterol Hepatol. 2020;17:338–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Gershon MD. The Shaggy Dog Story Of Enteric Signaling: Serotonin, A Molecular Megillah. Adv Exp Med Biol. 2022;1383:307–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Spencer NJ, Hibberd TJ, Travis L, et al. Identification of a rhythmic firing pattern in the enteric nervous system that generates rhythmic electrical activity in smooth muscle. J Neurosci. 2018;38:5507–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Fung C, Vanden Berghe P. Functional circuits and signal processing in the enteric nervous system. Cell Mol Life Sci. 2020;77:4505–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Uyanikgil Y, Özkeşkek K, Çavuşoğlu T, Solmaz V, Tümer MK, Erbas O. Positive effects of ceftriaxone on pentylenetetrazol-induced convulsion model in rats. Int J Neurosci. 2016;126:70–5. [DOI] [PubMed] [Google Scholar]
- [20].Majczyński H, Cabaj AM, Jordan LM, Sławińska U. Contribution of 5-HT(2) receptors to the control of the spinal locomotor system in intact rats. Front Neural Circuits. 2020;14:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Wood JD. Motor behavior of mouse large intestine: a minireview. Neurogastroenterol Motil. 2021;33:e14206. [DOI] [PubMed] [Google Scholar]
- [22].Matos JF, Americo MF, Sinzato YK, et al. Role of sex hormones in gastrointestinal motility in pregnant and non-pregnant rats. World J Gastroenterol. 2016;22:5761–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Luetic K, Sucic M, Vlainic J, et al. Cyclophosphamide induced stomach and duodenal lesions as a NO-system disturbance in rats: L-NAME, L-arginine, stable gastric pentadecapeptide BPC 157. Inflammopharmacology. 2017;25:255–64. [DOI] [PubMed] [Google Scholar]
- [24].King AJ, Siegel M, He Y, et al. Inhibition of sodium/hydrogen exchanger 3 in the gastrointestinal tract by tenapanor reduces paracellular phosphate permeability. Sci Transl Med. 2018;10:eaam6474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Beaumont M, Blachier F. Amino acids in intestinal physiology and health. Adv Exp Med Biol. 2020;1265:1–20. [DOI] [PubMed] [Google Scholar]
- [26].Bayle D, Coudy-Gandilhon C, Gueugneau M, et al. Magnesium deficiency alters expression of genes critical for muscle magnesium homeostasis and physiology in mice. Nutrients. 2021;13:2169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Sanders KM, Kito Y, Hwang SJ, Ward SM. Regulation of gastrointestinal smooth muscle function by interstitial cells. Physiology (Bethesda). 2016;31:316–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].De Giorgio R, Ruggeri E, Stanghellini V, Eusebi LH, Bazzoli F, Chiarioni G. Chronic constipation in the elderly: a primer for the gastroenterologist. BMC Gastroenterol. 2015;15:130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Li H, Chen J, Ren X, et al. Gut microbiota composition changes in constipated women of reproductive age. Front Cell Infect Microbiol. 2020;10:557515. [DOI] [PMC free article] [PubMed] [Google Scholar]