Plasma adiponectin, visfatin, leptin, and resistin levels and the onset of colonic polyps in patients with prediabetes

Following the operation manual developed by the “Epidemiology Research on Tumor Risks in Type 2 Diabetes Patients in China (REACTION study)” group and using the cluster sampling method, from April 2012 to December 2012 we carried out a cross-sectional survey among residents who were Han (ethnicity) and aged 40 years or older living in four communities in Hefei, Anhui, China. We requested physical examinations and blood lipid and blood glucose tests to be conducted in 9986 subjects. Moreover, all the non-diabetic subjects were requested to undergo a 75 g oral glucose tolerance test (75 g OGTT); the results indicated 1912 prediabetes cases, which were considered in the survey. Then, on the basis of the provisions of the “Guidelines for Early Screening and Early Treatment of Chinese Cancer (trial edition)” which was developed by the Ministry of Health of China in 2005 and suggests a combined mathematical risk factor model assessment questionnaire and faecal occult blood test (FOBT), we evaluated the risk of colon cancer among the 1912 prediabetes subjects (excluding those who were diagnosed with colon cancer). After screening, those who met one of the following criteria were considered high risk for colon cancer: (1) those whose FOBT was positive; (2) those whose first-degree relatives had a history of colorectal cancer; (3) those who had a history of cancer or intestinal polyps; and (4) those who had two or more of the following: chronic constipation, chronic diarrhoea, mucous bloody stool, history of chronic appendicitis or appendectomy, and history of chronic cholecystitis or gallstones. This investigation was presented in detail in our previous study [14]. As a result, a total of 547 patients with high risk of colon cancer were screened, and we suggested that these patients undergo an electronic colonoscopy examination. Finally, 501 subjects underwent electronic colonoscopy examinations, colonic mucosal lesions were found in 281 subjects, and colonic polyps were detected in 248 subjects who were assigned to the polyp group with prediabetes. In addition, non-colon mucosal lesions were found in 220 subjects who were assigned to the polyp-free group with prediabetes. The total of 468 prediabetic with or without colonic polyps were the subjects of this study. (Fig. 1).

In the present study, the controls were 108 Chinese patients of Han nationality who were aged ≥40 years, who were frequency matched by sex and age with prediabetes subjects and confirmed to have normal glucose tolerance by a 75 g oral glucose tolerance test, and who agreed to a mathematical risk factor model assessment questionnaire and faecal occult blood test (FOBT) according to the “Guidelines for Early Screening and Early Treatment of Chinese Cancer (trial edition)” mentioned above during the same period and at the same hospital, with the same exclusion criteria. Then, 108 controls were received electronic colonoscopy examinations. And colonic polyps were detected in 46 subjects who were assigned to the polyp group with normal glucose tolerance. In addition, non-colon mucosal lesions were found in 62 subjects who were assigned to the polyp-free with normal glucose tolerance group (Fig. 1).

The criteria for inclusion were as follows: (1) those who were Han, aged 40 years or older, and male or female; and (2) those who provided written consent. The exclusion criteria were as follows: (1) those who received oral glucocorticoid treatment over the past 3 months; (2) those who previously had hypothalamic, pituitary, adrenal, thyroid disease; (3) those who had severe heart, liver, kidney dysfunction, or chronic wasting disease; (4) those who applied for any hypoglycaemic drugs for pre-diabetes intervention; (5) those who had taken statins, nonsteroidal anti-inflammatory analgesics, and aspirin for more than 1 month in the past 6 months; (6) those whose weight loss exceeded 5% over the past 3 months; (7) those who had inflammatory bowel disease (such as ulcerative colitis, Crohn’s disease, etc.); (8) those who had an evident diagnosis of colon cancer and colon polyp lesions; (9) those who experienced a history of gastrointestinal surgery; and (10) those who had incomplete coloscopy due to poor bowel preparation or failure to carry out caecal intubation. Prediabetes is a condition that meets the diagnostic criteria for the classification of glucose metabolism states, which were promulgated by the World Health Organization (WHO) in 1999. This study was approved by the Ethics Committee of the First Affiliated Hospital of Anhui Medical University.

The results of electronic colonoscopy examinations regarding colonic polyps were recorded by a single researcher. If necessary, a polyp biopsy was performed, and the specimens were sent for pathological examination. When measuring the diameter of multiple polyps, the largest number was considered. The number of polyps was divided into single polyps (number of polyps ≤2) and multiple polyps (number of polyps ≥3). According to the clinical characteristics of the polyps and the pathological results described in a previous study [6], polyps were divided into high-risk polyps (number of polyps ≥3 and/or high hyperplasia and/or villous structures and/or polyps diameter ≥ 1 cm) and low-risk polyps (number of polyps < 3 and/or without high hyperplasia and/or tubular structure and/or polyp diameter < 1 cm).

Each participant was asked to answer a questionnaire covering their lifestyle and personal and family medical history. Briefly, the questionnaires mainly included the subjects’ smoking habits, alcohol drinking, physical activity, vegetable consumption, medications (such as statins, nonsteroidal anti-inflammatory analgesics, and aspirin) and family history of colon cancer. The definition of these indicators was presented as described in our previous studies [13, 15]: current smoking was referred to as smoking at least five cigarettes per day on average and lasted for one year or more. Regular alcohol drinking was referred to as drinking an average of at least 25 g per day for more than 1 year. Regular exercise was referred to as an average of five exercise events per week lasting at least 30 min per event and a total of time at least 150 min per week. Regular vegetable consumption was referred to as intaking of more than 250 g of vegetables per day. In addition, the family history of colon cancer was referred to as colon cancer in first-degree relatives of the subjects.

The participants fasted overnight before their endoscopic examination; then, their height, weight, waist circumference, hip circumference, systolic blood pressure (SBP), and diastolic blood pressure (DBP) were measured by a trained investigator as routine practice, and their body mass index (BMI) and waist:hip ratio (WHR) were calculated. Moreover, the collected venous blood samples were stored at − 80 °C for unified detection. Glucose, insulin, adiponectin, visfatin, leptin and resistin levels were measured in plasma samples, and total cholesterol (TCH) and triglycerides (TGs) were measured in serum samples. Insulin resistance was evaluated by homeostatic model assessment of insulin resistance (HOMA-IR) using the equation HOMA-IR = fasting glucose (mmol/L) × fasting insulin (mIU/L)/22.5.

The plasma/serum level of each index was determined as described in our previous study [13, 15]: fasting plasma glucose (FPG) and postprandial 2 h plasma glucose (2hPG) were detected by the glucose oxidase method; and TGs and TCH were measured by the enzymatic method. Fasting insulin (FINS) was measured by chemiluminescence. Plasma adiponectin, visfatin, leptin, and resistin were analysed in one run by enzyme-linked immunosorbent assay (ELISA).

The SPSS statistical software package for Window version 17.0 (SPSS, Chicago, IL, USA) was used for all statistical analyses. The continuous variables are expressed as means±standard deviations, and the categorical variables are expressed as numbers (percentages). The x² test and One-way analysis of variance were used for comparisons of clinical parameters among polyp-free group with normal glucose tolerance, colonic polyps group with normal glucose tolerance, polyp-free group with prediabetes, and colonic polyps group with prediabetes, and least-significant difference (LSD) analysis was used for comparisons between two groups. For comparisons between single polyp group and multiple polyps group/low-risk polyps group and high-risk polyps group in prediabetes, normally distributed continuous variables were analysed using independent sample t-tests, non-normally distributed continuous variables were analysed by Mann-Whitney U tests, and count data were analysed using the x² test. Multivariate unconditional logistic regression was used to analyse the influencing factors of colonic polyps, including multiple colonic polyps and high-risk colonic polyps. All P-values are two-sided, and a value less than 0.05 was considered statistically significant.

Compared with the polyp-free group with normal glucose tolerance, the average age, plasma visfatin levels, and proportion of family history with colon cancer in the colonic polyps group with normal glucose tolerance were significantly higher (P = 0.032 for age; P = 0.028 for visfatin; P = 0.042 for family history with colon cancer, respectively); plasma adiponectin levels were significantly lower [(9.8 ± 4.8 vs 13.3 ± 3.9) mg/L, P = 0.013], and the other clinical parameters were not statistically significant between the two groups (P > 0.05) (Table 1, Fig. 2).

Compared with the polyp-free group with normal glucose tolerance, FPG, 2hPG, HOMA-IR, plasma visfatin, and resistin levels were significantly higher in the polyp-free group with prediabetes (P = 0.025 for FPG; P = 0.009 for 2hPG; P = 0.034 for HOMA-IR; P = 0.001 for visfatin; P = 0.041 for resistin, respectively); plasma adiponectin levels were significantly lower [(9.2 ± 4.4 vs 13.3 ± 3.9) mg/L, P = 0.028]; and the other clinical parameters were not statistically significant between the two groups (P > 0.05) (Table 1, Fig. 2).

Compared with the polyp-free group with normal glucose tolerance, the average age, BMI, WHR, FPG, 2hPG, TGs, FINS, HOMA-IR, visfatin, leptin and resistin levels, and the proportion of family history with colon cancer were significantly higher in the colonic polyps group with prediabetes (P = 0.017 for age; P = 0.006 for BMI; P = 0.024 for WHR; P = 0.011 for FPG; P < 0.001 for 2hPG; P = 0.042 for TGs; P = 0.047 for FINS; P < 0.001 for HOMA-IR; P < 0.001 for visfatin; P = 0.029 for leptin; P = 0.008 for resistin; P = 0.008 for family history with colon cancer, respectively), and the level of adiponectin was significantly lower [(5.6 ± 3.7 vs 13.3 ± 3.9) mg/L, P < 0.001). The other clinical parameters were not statistically significant between the two groups (P > 0.05) (Table 1, Fig. 2).

Compared with the colonic polyps group with normal glucose tolerance, the FPG and 2hPG levels in the polyp-free group with prediabetes were significantly higher (P = 0.001 for FPG; P < 0.001 for 2hPG, respectively), and age was lower (P = 0.015). The other clinical parameters were not statistically significant between the two groups (P > 0.05) (Table 1).

Compared with the colonic polyps group with normal glucose tolerance, the BMI, WHR, FPG, 2hPG, TGs, HOMA-IR, visfatin, leptin, and resistin levels were significantly higher in the colonic polyps group with prediabetes (P = 0.026 for BMI; P = 0.033 for WHR; P = 0.001 for FPG; P < 0.001 for 2hPG; P = 0.041 for TG; P = 0.046 for HOMA-IR; P = 0.011 for visfatin; P = 0.042 for leptin; P = 0.039 for resistin, respectively), but the adiponectin level was significantly lower [(5.6 ± 3.7 vs 9.8 ± 4.8) mg/L, P = 0.024). The other clinical parameters were not statistically significant between the two groups (P > 0.05) (Table 1, Fig. 2).

Compared with the polyp-free group with prediabetes, the male proportion, average age, BMI, WHR, 2hPG, TGs, visfatin level, the distribution of current smoking, alcohol drinking, and the proportion of family history with colon cancer were significantly higher in the colonic polyps group with prediabetes (P = 0.000 for male; P = 0.029 for BMI; P = 0.033 for WHR; P = 0.035 for 2hPG; P = 0.046 for TG; P = 0.041 for visfatin; P < 0.001 for current smoking; P < 0.001 for alcohol drinking; P = 0.034 for family history with colon cancer, respectively), but the plasma adiponectin level was significantly lower [(5.6 ± 3.7 vs 9.2 ± 4.4) mg/L, P = 0.007]. The other clinical parameters were not statistically significant between the two groups (P > 0.05) (Table 1, Fig. 2).

The average age, BMI, WHR, 2hPG, FINS, HOMA-IR, plasma visfatin level, and the proportion of family history with colon cancer were significantly higher in the multiple-polyps group than in the single-polyp group (P = 0.025 for age; P = 0.038 for BMI; P = 0.022 for WHR; P < 0.001 for 2hPG; P = 0.034 for FINS; P = 0.044 for HOMA-IR; P = 0.042 for visfatin; P = 0.012 for family history with colon cancer, respectively), and the level of plasma adiponectin was significantly lower than that in the single-polyp group [(4.3 ± 2.6 vs 6.7 ± 3.9) mg/L, P = 0.031). No significant differences were observed in sex ratio, SBP, DBP, FPG, TG, TCH, plasma leptin and resistin levels, as well the proportion of current smoking, regular alcohol drinking, regualer exercise, and regular vegetable consumption between the two groups (P > 0.05). (Table 2, Fig. 3).

The proportion of males, average age, BMI, WHR, 2hPG, FINS, HOMA-IR, distribution of current smoking, and the proportion of family history with colon cancer in the high-risk polyps group were significantly higher than those in the low-risk polyps group (P = 0.029 for male; P = 0.012 for age; P = 0.046 for BMI; P = 0.032 for WHR; P < 0.001 for 2hPG; P = 0.047 for FINS; P = 0.034 for HOMA-IR; P = 0.023 for smoking; P = 0.035 for family history with colon cancer, respectively), whereas the level of plasma adiponectin was significantly lower than that in the low-risk polyps group [(3.7 ± 2.9 vs 7.4 ± 3.5) mg/L, P < 0.001). No significant differences were observed in SBP, DBP, FPG, TGs, TCH, plasma visfatin, leptin, and resistin levels, as well the proportion of regular alcohol drinking, regualer exercise, and regular vegetable consumption between the two groups (P > 0.05). (Table 3, Fig. 4).

To identify the risk factors for colonic polyps in prediabetes subjects, the multivariate unconditional logistic regression analysis was performed after adjustment for sex, age, BMI, WHR, 2hPG, TGs, adiponectin, visfatin, leptin, resistin, current smoking, alcohol drinking, family history of colon cancer to determine the risk factors for colonic polyps in prediabetes subjects. At the same time, we carried out a multivariate unconditional logistic regression analysis adjusted for age, BMI, WHR, 2hPG, FINS, HOMA-IR, adiponectin, visfatin, and family history of colon cancer to determine the risk factors for multiple colonic polyps in prediabetes subjects. In addition, in order to determine the risk factors for prediabetes-related high-risk colonic polyps, We performed a multivariate unconditional logistic regression analysis adjusted for sex, age, BMI, WHR, 2hPG, FINS, HOMA-IR, adiponectin, current smoking, and family history of colon cancer. In these analyses, sex, WHR, 2hPG, and family history of colon cancer were found to be risk factors for colonic polyps in prediabetes subjects (OR values were 1.89, 2.98, 1.81, 1.36, respectively, P < 0.05) (Table 4). In addition, age, WHR, and HOMA-IR were risk factors for multiple colonic polyps in prediabetes subjects (OR values were 1.09, 2.12, 1.76, respectively, P < 0.05) (Table 5). Moreover, age and WHR were risk factors for high risk colonic polyps in prediabetes subjects (OR values were 1.45 and 2.11, respectively, P < 0.05) (Table 6). Tables 4, 5 and 6 show that lower levels of plasma adiponectin was a risk factor for colonic polyps, multiple colonic polyps, and high-risk colonic polyps in prediabetes subjects (OR values were 0.67, 0.79, and 0.61, respectively, P < 0.05).