Characteristics and risk factors for colorectal polyps among children in an urban area of Wenzhou, China: a retrospective case control study

A total number of 81 children who were hospitalized from October 2020 to October 2022 and diagnosed with colorectal polyps by colonoscopy were enrolled as a case group. In the control group, we randomly selected and included 123 children who underwent colonoscopy during the same period and whose results showed no obvious abnormalities. If a patient had undergone multiple colonoscopies within the time frame selected, only the first colonoscopy was included in our research work to ensure per-patient analysis results have been obtained.

The following inclusion criteria were implemented: (1) Children aged 0–18 years who underwent colonoscopy; (2) Complete clinical data, including clear image data of colonoscopy and polyp histopathology; (3) Informed consent was acquired, and the respective forms were signed by the patient’s parents, including informed consent for colonoscopy, informed consent for biopsy tissue, and informed consent for endoscopic polypectomy.

The exclusion criteria applied were as follows: (1) Severe cardiopulmonary disease or other serious underlying diseases; (2) Acute active stage of intestinal inflammatory diseases; (3) Coagulation dysfunction, bleeding tendency; (4) Severe spinal deformity; (5) Poor compliance. Colonoscopy is not recommended in children with these conditions.

This clinical study was approved by the Ethics Committee of The Second Affiliated Hospital and Yuying Children’s Hospital of Wenzhou Medical University (Wenzhou, Zhejiang, China). All methods and procedures were performed in accordance with the ethical standards as laid down in the Declaration of Helsinki. Written informed consent was obtained from all participants’ legal guardian.

Clinical data were collected from electronic and paper medical records, including demographic characteristics and clinical manifestations. The residential addresses were divided into urban and rural. Based on clinical significance, we converted continuous data into categorical data. Age was classified into four categories:< 3 years, 3–6 years, 6–11 years, and ≥ 12 years; the course of disease was divided into a short course (< 6 months) and a long course (≥ 6 months).

The following blood test data were collected: hemoglobin (venous blood), sIgE, the white blood cell (WBC) count, platelet count, blood eosinophil count, C-reactive protein (CRP), albumin, alanine aminotransferase (ALT), AST, alkaline phosphatase (ALP), and gamma-glutamyl transferase (γ-GT). Based on the criteria of the World Health Organization (2011) [9], anemia (Hb < 110 g/L) in the age group 6–59 months was defined as mild (100–109 g/L), moderate (70–99 g/L), and severe (< 70 g/L). In the 5–11-year age group, anemia (Hb < 115 g/L) was classified into mild (110–114 g/L), moderate (80–109 g/L), and severe (< 80 g/L). Anemia (Hb < 120 g/L) in the age group of 12–14 years was categorized into mild (110–119 g/L), moderate (80–109 g/L), and severe (< 80 g/L).Based on current international standards, a concentration of sIgE < 0.35 kU/L was considered to indicate a negative result; otherwise, the result was considered positive [10].

Information was obtained by asking parents about their children’s defecation habits, dietary habits, and family aggregation. Diarrhea was considered to be present if their defecation habits met the Rome IV diagnostic criteria for diarrhea [11]. Similarly, we considered their defecation habits as constipation if they met the Rome IV diagnostic criteria for constipation [11]. We collected patients’ diet data through 3-day 24-hour dietary recall. We examined the consumption of three main food groups: meat (pork, chicken, duck, etc.), vegetables (green vegetables, cabbage, tomato, lettuce, etc.), and seafood (fish, shrimp, shell, etc.). The estimates of intake (g/day) were based on frequency and portion size information. If a relative of the patient had polyps or bowel cancer, it was considered that family aggregation is positive.

For the colonoscopy report of each patient, we extracted the following results: quality of intestinal preparation, depth and location of colonoscopy, presence or absence of polyps, location and characterizations of the polyps (such as number, size, and shape), and the pathologic type of the polyps (juvenile polyps/adenomatous polyps/hyperplastic polyps/inflammatory polyps/juvenile polyposis). To facilitate statistical data collection and analysis, the locations of polyps were grouped into rectal polyps, sigmoid polyps, descending colon polyps, pan-colorectal polyps, and polyps in other locations (including ileocecal, ascending, and transverse colon polyps), the number of polyps was grouped into solitary polyp (only one polyp) and multiple polyps (containing at least two polyps), the size of polyps was categorized into large polyps (maximum diameter of the polyps > 15 mm) and small polyps (maximum diameter of the polyps ≤ 15 mm). The shapes of polyps were grouped into sessile, pedunculated- sessile, and pedunculated polyps.

All statistical analyses were conducted using SPSS 22.0. Continuous variables were presented as mean with standard vision or medians with interquartile ranges. The difference between groups was compared by t-test or nonparametric test. Categorical variables were expressed as frequencies with percentages, and the differences between groups were tested by χ2 tests or Fisher’s exact tests and considerable covariates (P < 0.10) were included in multiple analyses conducted using logistic regression. A P < 0.05 (two tailed) was considered statistically significant. ROC curve analysis was used to evaluate the predictive ability and optimal cut-off values of statistically significant continuous variables.

A total number of 204 patients were enrolled in the study, of whom 81 had colorectal polyps. The mean age of the children with polyps was 6.77 ± 3.44 years, and the mean age of children without polyps was 9.27 ± 3.02 years. Compared with the normal control group, the patients who had colorectal polyps tended to have the following characteristics: male, younger, with anemia, with positive family aggregation and sIgE, with a higher platelet count and AST level, and with a higher intake of meat but a lower intake of vegetable (Table 1). There was no significant difference in the defecation habits, type of residence, WBC count, blood eosinophil count, CRP, albumin, ALT, ALP, γ-GT, and seafood intake (P > 0.05).

Children with colorectal polyps had different clinical manifestations, with incidences of clinical manifestations from high to low as follows: hematochezia (80.2%), abdominal pain (23.5%), and anal mass prolapse (8.6%). There were also a few patients without obvious clinical manifestations (6.2%) (including imaging suggesting the presence of polyps, CA199 found elevated, etc.). Long-course disease was observed in 67.9% of the cases, whereas short-course disease accounted for 32.1%. Comparing the incidence of each clinical manifestation in different age groups, the differences in abdominal pain was statistically significant (P = 0.007), whereas the differences in the remaining clinical manifestations were not statistically significant (P > 0.05) (Table 2). Comparing the incidence of various clinical manifestations in different gender groups, the difference in abdominal pain was statistically significant (P = 0.024), whereas the differences in the incidence of the remaining clinical manifestations were not statistically significant (P > 0.05) (Table 2).

The main features of colorectal polyps are presented in Table 3. The pathological type of polyps was predominantly juvenile (n = 72, 88.9%). Adenomatous polyps were found in five cases (6.2%), 4 of which were associated with low-grade intraepithelial neoplasia. There were two cases of inflammatory polyps (2.5%), one of which with adenomatous gland hyperplasia. Hyperplastic polyps were established in one case (1.2%) and juvenile polyposis in one case (1.2%). The majority of children (n = 71, 87.7%) had only one polyp, but in 10 cases (12.3%), multiple polyps were present. Small polyps were detected in 47 cases (58.0%), whereas big polyps were present in 31 cases (38.3%). Polyps were located mainly in the rectum (n = 41, 50.6%) and the sigmoid colon (n = 19, 23.5%). Descending colon polyps were found in 7 cases (8.6%); polyps in other locations of the colon (including ileocecal, ascending, and transverse colon) were detected in 10 cases (12.3%). In 4 cases (4.9%), the polyps were distributed in various parts of the colon. The majority of polyps were pedunculated (n = 51, 62.96%), followed by pedunculated-sessile (n = 22, 27.16%); sessile polyps were established in 6 cases (7.41%).

The potential influencing factors of colorectal polyps in children are listed in Table 4. Univariate analysis showed that sex, age, family aggregation, sIgE, platelet count, AST, and meat and vegetable intake were associated with the presence of colorectal polyps in children (P < 0.05). In contrast, type of residence, defecation habits, the WBC count, blood eosinophil count, CRP, albumin, ALT, ALP, γ-GT, and seafood intake were not associated with colorectal polyps (P > 0.05).

Considerable covariates (P < 0.10 or with important clinical relevance) were included in the logistic regression analysis. Multivariate analysis revealed that age ≤ 6 years, positive family aggregation, positive sIgE, and a higher intake of meat were risk factors for the presence of colorectal polyps in children. Meanwhile, a higher intake of vegetable was a protective factor for colorectal polyps in children. Compared to children who aged > 6 years, children with onset age at 6 or younger had a higher risk for colorectal polyps (OR:22.678,95% CI: 1.873–274.535) (OR:24.601,95% CI: 3.761–160.910) and the risk of colorectal polyps is highest in children aged 3–6 years. Children with positive family aggregation had a higher risk for colorectal polyps than those with negative family aggregation (OR:3.540,95% CI: 1.177–10.643). Children with positive sIgE had a higher risk for colorectal polyps than those with negative sIgE (OR:2.263,95% CI: 1.076–4.761). Children with higher intake of meat had a higher risk for colorectal polyps (OR:1.046,95% CI: 1.029–1.063) and those with higher intake of vegetable had a lower risk for colorectal polyps (OR:0.993,95% CI: 0.986–1.000) (Table 5). The predictive value of meat intake, vegetable intake, combination of meat intake and vegetable intake on colorectal polyps in children was analyzed by the ROC curve analysis (Fig. 1). The AUC of meat intake for colorectal polyps in children was 0.607, which had predictive value. The optimal cut-off value in this study was 92.14 g/d (P = 0.010,95% CI: 0.527–0.687) and the sensitivity and specificity were 66.67% and 69.92%, respectively. The AUC of vegetable intake for predicting colorectal polyps in children was 0.319, and the predictive value alone was low. The AUC of the combination of meat intake and vegetable intake for predicting colorectal polyps in children was 0.781. The predictive value was better than that of meat intake alone (P < 0.001,95% CI: 0.718–0.845).