Abstract
Insulin resistance and stress response are relatively prevalent among patients with colorectal cancer before surgery. This study aimed to explore the effects of these two disorders on the risk of anastomotic leakage after colorectal cancer surgery. Briefly, 503 patients with this type of cancer scheduled for surgery were enrolled. The study used the HOMA-IR to evaluate the patients’ preoperative insulin resistance, collected blood samples to detect the preoperative levels of adrenaline and cortisol, and also adopted the State-Trait Anxiety Inventory and the Impact of Event Scale to assess the patients’ psychological stress status. After performing the surgery, the study monitored the onset of anastomotic leakage within one month. Multivariable logistic regression was used for data analysis. The results suggested that preoperative insulin resistance, elevation of the two hormone levels, and increased psychological stress scores were significantly associated with an increased risk of anastomotic leakage. When the levels of adrenaline and cortisol increased by one standard deviation, the increase in the risk of this postoperative complication was greater in patients with insulin resistance than in those without insulin resistance. In conclusion, both preoperative insulin resistance and stress response were potential risk factors for anastomotic leakage after colorectal cancer surgery.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-025-11462-4.
Keywords: Anastomotic leakage, Colorectal cancer, Insulin resistance, Stress hormone, Stress response
Subject terms: Cancer, Risk factors, Signs and symptoms
Introduction
According to the data from the International Agency for Research on Cancer, colorectal cancer is one of the most common malignant tumors worldwide, with over 1.9 million new cases annually, ranking third among all malignant tumors1. In China, the incidence of colorectal cancer also remains at a high level, with over 500,000 new cases each year2.
Surgery represents the most primary approach for treating colorectal cancer, with approximately 70–90% of patients with this disease requiring such treatment3. However, surgery is not without risks, and anastomotic leakage is one of the most important postoperative complications that cannot be ignored4.
The harms of anastomotic leakage are relatively severe4. Not only does it disrupt postoperative recovery, prolong hospital stays, and increase the risk of local infections that may escalate to life-threatening sepsis in severe cases, but it also delays subsequent treatment plans, thereby compromising the overall therapeutic efficacy and patient prognosis.
In recent years, the incidence rates of metabolic disorders such as type 2 diabetes, obesity, and hyperlipidemia have been increasing, and this is also the case in the population with colorectal cancer. Previous studies have confirmed that these metabolic disorders are associated with the development of colorectal cancer, and are also related to the risk of digestive tract anastomotic leakage5–7. It is worth emphasizing that insulin resistance holds a core position in these metabolic disorders8. And, malignant tumors, including colorectal cancer, can lead to insulin resistance, even in patients without a previous history of insulin resistance or diabetes9,10. Thus, it is of great necessity to explore the impact of insulin resistance on anastomotic leakage after colorectal cancer surgery. However, so far, no study has ventured into this area.
Suffering from colorectal cancer and being scheduled for surgery are likely to stimulate patients, thus triggering a preoperative stress response. This is a non-specific defensive reaction of the body, but can also exert adverse effects under pathological conditions11. Furthermore, although it has not been confirmed yet, we have reasons to speculate that there is an association between the preoperative stress response and anastomotic leakage after colorectal cancer surgery. The potential reasons are as follows12,13: Stress response may cause neuroendocrine changes, disrupt the metabolism of carbohydrates, fats, and proteins, which is detrimental to the healing of the anastomosis. Moreover, the interaction between stress response and metabolic abnormalities may exacerbate metabolic disorders, potentially affecting the body’s ability to respond to stress, and jointly increasing the risk of anastomotic leakage.
Therefore, we planned to recruit hundreds of colorectal cancer patients to evaluate the impact of preoperative insulin resistance and stress response on the risk of postoperative anastomotic leakage. Given that both metabolic disorders and stress response are quite common in colorectal cancer patients, the results of this study are highly relevant to clinical practice and of great guiding significance.
Materials and methods
Ethical requirements
This study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Xinxiang Medical University, China14. All methods were carried out in accordance with the Declaration of Helsinki of the World Medical Association and other relevant guidelines and regulations. All the participants gave their consent to take part in this study and signed the informed consent forms.
Participants
A total of 503 patients who were admitted to the Department of General Surgery at the First Affiliated Hospital of Xinxiang Medical University from January 1 st, 2020, to June 30th, 2024, were included in this study. The inclusion criteria were as follows: (1) The patients were pathologically diagnosed with colorectal cancer15. (2) They had not received any anti-tumor treatment previously. (3) They were planned to receive surgical resection during this hospitalization, and could also receive other anti-tumor treatments after the operation. (4) They may have suffered from metabolic disorders such as type 2 diabetes, insulin resistance, obesity, etc., or they may have never had these conditions. (5) They had no history of pheochromocytoma, Cushing’s syndrome, adrenocortical insufficiency, hyperthyroidism, or any type of psychological or cognitive diseases, so as to avoid affecting the determination of preoperative stress response. (6) They had no other types of malignant tumors. If the patients suffered from cardiovascular diseases or other chronic diseases, their conditions must be well controlled to reduce the health risks in the study. Patients who did not meet the above criteria were directly excluded from this study.
Preoperative data collection
The baseline characteristics, preoperative metabolic disorders, preoperative serological indicators, and characteristics of colorectal cancer were collected. The baseline characteristics included demographic data, personal history, chronic disease history, and genetic factors. The metabolic disorders included insulin indicators, history of type 2 diabetes, drugs that affect insulin resistance, and other metabolic indicators. The serological indicators consisted of complete blood count, coagulation profile, liver and kidney function, electrolytes, and inflammatory indicators. Tumor data included tumor markers and pathological characteristics.The specific items were shown in Tables 1, 2, 3, 4 and 5.
Table 1.
Baseline characteristics of patients in the two groups.
| Variable | AL (n = 71) |
NAL (n = 432) |
P value |
|---|---|---|---|
| Demographic data | |||
| Male (n) | 43 (60.6) | 235 (54.4) | 0.333 |
| Age (years) | 65 (10.0) | 63 (10.8) | 0.039 |
| Personal history | |||
| Tobacco smoking (n) | 17 (23.9) | 79 (18.3) | 0.261 |
| Alcohol drinking (n) | 18 (25.4) | 88 (20.4) | 0.340 |
| Lack of exercise (n) | 25 (35.2) | 134 (31.0) | 0.481 |
| Chronic disease history | |||
| Cardiovascular disease (n) | 29 (40.8) | 142 (32.9) | 0.189 |
| Cerebrovascular disease (n) | 12 (16.9) | 47 (10.9) | 0.144 |
| Hypertension (n) | 35 (49.3) | 134 (32.4) | 0.006 |
| Adenomatous polyp (n) | 30 (42.3) | 155 (35.9) | 0.302 |
| Inflammatory bowel disease (n) | 2 (2.8) | 8 (1.9) | 0.935 |
| Genetic factor | |||
| Family history of CRC (n) | 11 (15.5) | 53 (12.3) | 0.450 |
| FAP (n) | 1 (1.4) | 0 (0.0) | - |
Note: AL = Anastomotic leakage, NAL = Non-anastomotic leakage, CRC = Colorectal cancer, FAP = Familial adenomatous polyposis. Continuous variables are presented as median (interquartile range), and the differences between groups are evaluated using the Wilcoxon rank-sum test. Categorical variables are expressed as frequency (constituent ratio), and the differences between groups are assessed by the chi-square test. A P-value less than 0.05 indicates statistical significance.
Table 2.
Preoperative metabolic disorders of patients in the two groups.
| Variable | AL (n = 71) |
NAL (n = 432) |
P value |
|---|---|---|---|
| Insulin indicator | |||
| Fasting blood glucose (mmol/l) | 4.6 (0.6) | 4.7 (0.7) | 0.016 |
| Fasting insulin (mIU/L) | 13.0 (11.8) | 9.8 (8.2) | 0.008 |
| HOMA-IR | 2.6 (2.9) | 2.1 (1.9) | 0.048 |
| IR (n) | 38 (53.5) | 151 (35.0) | 0.003 |
| History of T2DM | |||
| Present (n) | 19 (26.8) | 59 (13.7) | 0.005 |
| Absent (n) | 52 (73.2) | 373 (86.3) | |
| Drugs that affect IR | |||
| Metformin (n) | 8 (11.3) | 21 (4.9) | 0.032 |
| TZDs (n) | 4 (5.6) | 7 (1.6) | 0.032 |
| GLP−1 receptor agonists (n) | 3 (4.2) | 9 (2.1) | 0.273 |
| DPP−4 inhibitors (n) | 3 (4.2) | 11 (2.5) | 0.425 |
| Sulfonylureas (n) | 7 (9.9) | 24 (5.6) | 0.162 |
| Meglitinides (n) | 4 (5.6) | 13 (3.0) | 0.257 |
| Other metabolic indicator | |||
| Triglyceride (mmol/l) | 1.7 (1.3) | 1.6 (1.2) | 0.224 |
| Total cholesterol (mmol/l) | 5.3 (2.1) | 5.1 (2.2) | 0.034 |
| LDL-C (mmol/l) | 3.0 (1.9) | 2.8 (1.9) | 0.028 |
| Body mass index (kg/m2) | 23.4 (3.4) | 22.2 (4.7) | 0.008 |
Note: AL = Anastomotic leakage, NAL = Non-anastomotic leakage, HOMA-IR = Homeostatic model assessment - insulin resistance, IR = Insulin resistance, T2DM = Type 2 diabetes mellitus, TZDs = Thiazolidinediones, GLP−1 = Glucagon-like peptide−1, DPP−4 = Dipeptidyl peptidase−4, LDL-C = Low-density lipoprotein cholesterol. Continuous variables are presented as median (interquartile range), and the differences between groups are evaluated using the Wilcoxon rank-sum test. Categorical variables are expressed as frequency (constituent ratio), and the differences between groups are assessed by the chi-square test. A P-value less than 0.05 indicates statistical significance.
Table 3.
Preoperative serological indicators of patients in the two groups.
| Variable | AL (n = 71) |
NAL (n = 432) |
P value |
|---|---|---|---|
| Complete blood count | |||
| Hemoglobin (g/l) | 105.5 (22.1) | 113.5 (26.7) | 0.069 |
| White blood cell count (×109/l) | 8.9 (3.6) | 8.5 (3.5) | 0.084 |
| Platelet count (×109/l) | 229.8 (137.2) | 229.3 (109.8) | 0.302 |
| Coagulation profile | |||
| Prothrombin time (s) | 12.9 (2.3) | 12.9 (1.9) | 0.218 |
| APTT (s) | 34.5 (5.7) | 33.6 (6.8) | 0.389 |
| Biochemical indicator | |||
| Albumin (g/l) | 37.6 (10.2) | 41.2 (10.6) | 0.041 |
| Alanine aminotransferase (U/l) | 32.5 (20.5) | 32.2 (22.8) | 0.513 |
| Aspartate aminotransferase (U/l) | 33.2 (21.1) | 34.1 (22.7) | 0.547 |
| Total bilirubin (µmol/l) | 14.5 (10.8) | 12.9 (9.5) | 0.208 |
| Creatinine (µmol/l) | 89.9 (28.9) | 87.3 (32.3) | 0.310 |
| Potassium (mmol/l) | 4.2 (1.3) | 4.2 (1.4) | 0.650 |
| Sodium (mmol/l) | 137.5 (7.6) | 138.9 (8.5) | 0.223 |
| Inflammatory indicator | |||
| C-reactive protein (mg/l) | 5.3 (3.1) | 4.8 (3.9) | 0.027 |
| Interleukin−6 (pg/ml) | 7.2 (3.5) | 6.2 (4.1) | 0.032 |
Note: AL = Anastomotic leakage, NAL = Non-anastomotic leakage, APTT = Activated partial thromboplastin time. Continuous variables are presented as median (interquartile range), and the differences between groups are evaluated using the Wilcoxon rank-sum test. Categorical variables are expressed as frequency (constituent ratio), and the differences between groups are assessed by the chi-square test. A P-value less than 0.05 indicates statistical significance.
Table 4.
Preoperative stress responses of patients in the two groups.
| Variable | AL (n = 71) |
NAL (n = 432) |
P value |
|---|---|---|---|
| Vital sign | |||
| Heart rate (bpm) | 82.0 (23.0) | 75.0 (23.0) | <0.001 |
| Systolic blood pressure (mmHg) | 128.0 (30.0) | 123.5 (27.0) | 0.001 |
| Diastolic blood pressure (mmHg) | 83.0 (12.0) | 76.5 (14.0) | <0.001 |
| Respiratory rate (bpm) | 19.0 (5.0) | 17.0 (4.0) | <0.001 |
| Hormone indicator | |||
| Epinephrine (pmol/l) | 507.0 (228.6) | 399.2 (206.4) | <0.001 |
| Cortisol at 8 am (nmol/l) | 481.8 (203.0) | 396.7 (144.2) | <0.001 |
| Cortisol at 8 pm (µg/dl) | 214.6 (88.2) | 160.9 (83.2) | <0.001 |
| Psychological indicator | |||
| State anxiety inventory score | 45.0 (11.0) | 39.0 (11.0) | <0.001 |
| Trait anxiety inventory score | 45.0 (8.0) | 38.0 (8.0) | <0.001 |
| Impact of event scale score | 31.0 (9.0) | 25.0 (8.0) | <0.001 |
Note: AL = Anastomotic leakage, NAL = Non-anastomotic leakage. Continuous variables are presented as median (interquartile range), and the differences between groups are evaluated using the Wilcoxon rank-sum test. Categorical variables are expressed as frequency (constituent ratio), and the differences between groups are assessed by the chi-square test. A P-value less than 0.05 indicates statistical significance.
Table 5.
Characteristics of colorectal cancer in the two groups.
| Variable | AL (n = 71) |
NAL (n = 432) |
P value |
|---|---|---|---|
| Preoperative indicator | |||
| CEA (µg/l) | 12.2 (13.1) | 11.8 (11.1) | 0.306 |
| CA199 (U/ml) | 62.8 (69.9) | 62.6 (68.5) | 0.337 |
| CA242 (kU/l) | 36.0 (41.2) | 29.4 (31.1) | 0.308 |
| Tumor location | |||
| Colon cancer | 35 (49.3) | 285 (66.0) | 0.007 |
| Rectal cancer | 36 (50.7) | 147 (34.0) | |
| Tumor size | |||
| ≥ 5 cm (n) | 31 (43.7) | 150 (34.7) | 0.146 |
| < 5 cm (n) | 40 (56.3) | 282 (65.3) | |
| Histological type | |||
| Adenocarcinoma (n) | 64 (90.1) | 410 (94.9) | 0.110 |
| Other (n) | 7 (9.9) | 22 (5.1) | |
| Differentiation grade | |||
| Low to undifferentiated (n) | 27 (38.0) | 129 (29.9) | 0.168 |
| High to moderate (n) | 44 (62.0) | 303 (70.1) | |
| Lymphatic metastasis | |||
| Present (n) | 31 (43.7) | 151 (35.0) | 0.157 |
| Absent (n) | 40 (56.3) | 281 (65.0) | |
| Distant metastasis | |||
| Present (n) | 10 (14.1) | 43 (10.0) | 0.293 |
| Absent (n) | 61 (85.9) | 389 (90.0) | |
| Surgical procedure | |||
| RHC/ERHC (n) | 7 (9.9) | 73 (16.9) | 0.133 |
| LHC (n) | 10 (14.1) | 91 (21.1) | 0.174 |
| SCCRR (n) | 18 (25.4) | 121 (28.0) | 0.643 |
| AR (n) | 23 (32.4) | 103 (23.8) | 0.123 |
| APR (n) | 13 (18.3) | 44 (10.2) | 0.045 |
Note: AL = Anastomotic leakage, NAL = Non-anastomotic leakage, CEA = Carcinoembryonic antigen, CA199 = Carbohydrate antigen 199, CA242 = Carbohydrate antigen 242, CA724 = Carbohydrate antigen 724, CYFRA21-1 = Cytokeratin Fragment 21 − 1, RHC = Right hemicolectomy, ERHC = Extended right hemicolectomy, LHC = Left hemicolectomy, SCCRR = Sigmoid colon cancer radical resection, AR = Anterior resection of the rectum, APR = Abdominoperineal resection of the rectum. Continuous variables are presented as median (interquartile range), and the differences between groups are evaluated using the Wilcoxon rank-sum test. Categorical variables are expressed as frequency (constituent ratio), and the differences between groups are assessed by the chi-square test. A P-value less than 0.05 indicates statistical significance.
Tobacco smoking and alcohol drinking were separately defined as regular smoking or drinking (continuous or cumulative) for more than 12 months during the patient’s lifetime. Lack of exercise was defined as the patient never having engaged in any form of regular and specialized exercise activities in the past 10 years. All of the patients’ previous disease histories were determined based on the patients’ medical records and treatment processes.
All patients had fasting venous blood collected before surgery, and all serological indicators (including fasting blood glucose and fasting insulin) were tested by the laboratory department of this hospital.
Preoperative insulin resistance determination
The insulin resistance of patients was determined based on the Homeostatic Model Assessment-Insulin Resistance (HOMA-IR) index16,17. HOMA-IR is a commonly used indicator for evaluating the degree of insulin resistance in the body. The formula for calculation is HOMA-IR = (fasting insulin×fasting blood glucose) ÷ 22.5. In this study, a HOMA-IR value of ≥ 2.5 was considered to indicate the presence of insulin resistance in patients.
Preoperative stress response determination
Gastrointestinal surgeons and psychologists jointly monitored the stress response of patients during the 24 h before surgery. The specific details were as follows: (1) Vital signs: Recorded the heart rate, blood pressure, and respiratory rate of patients at 8 a.m. in a quiet state. (2) Stress hormones: Blood samples were collected at 8 a.m. to detect the levels of peripheral adrenaline and cortisol. Then, blood samples were collected again at 8 p.m. to detect the levels of peripheral cortisol. The specific testing work was completed by the clinical laboratory of this hospital. (3) Psychological state: The State-Trait Anxiety Inventory (STAI) was used to evaluate the preoperative anxiety of patients, and the Impact of Event Scale (IES) was used to assess the degree of preoperative psychological stress response of patients.
The STAI is composed of the State Anxiety Inventory (SAI) and the Trait Anxiety Inventory (TAI), which are used to measure an individual’s transient state anxiety and more stable trait anxiety respectively18. Each subscale contains 20 items, and the theoretical highest score for both is 80 points. The higher the total score of the SAI, the higher the individual’s current situational anxiety level. Similarly, the higher the total score of the TAI, the more pronounced the individual’s personal trait of chronic anxiety tendency.
The IES is designed to assess the psychological impact of traumatic events on individuals19. It consists of 15 items and mainly measures post-traumatic intrusive symptoms (such as unwanted thoughts about the event) and avoidance behaviors (like making efforts to avoid thinking about the event). The total score ranges from 0 to 60. The higher the total score, the greater the impact of the traumatic event on the individual’s psychological state, indicating more pronounced post-traumatic stress reactions.
Surgical procedures
Gastrointestinal surgeons selected appropriate surgical procedures according to the specific conditions of patients with colorectal cancer20,21. When the cancer involved the cecum, ascending colon, and hepatic flexure of the colon, right hemicolectomy or extended right hemicolectomy was adopted. When the cancer involved the splenic flexure of the colon and descending colon, left hemicolectomy was carried out. When the cancer involved the sigmoid colon, radical resection of sigmoid colon cancer was performed. For rectal cancer involving a location 5 cm or more above the dentate line, anterior resection of the rectum was adopted to preserve the anal function as much as possible. For rectal cancer involving a location less than 5 cm from the dentate line, abdominoperineal resection of the rectum was carried out, and a permanent artificial anus was created.
During the case collection period of this study, no patients with colon cancer located in the middle segment of the transverse colon were included, and thus no transverse colon resections were performed. The concept of extended left hemicolectomy is not commonly used, and related cases were default classified into the category of left hemicolectomy.
Postoperative anastomotic leakage determination
The occurrence of anastomotic leakage was closely monitored by the attending physician within one month after the operation. The specific procedures were as follows22,23: (1) After the operation, the patient presented with symptoms such as fever (body temperature > 38℃), abdominal pain, abdominal distension, nausea, and vomiting. In severe cases, there were manifestations of septic shock, like a drop in blood pressure and an increase in heart rate. (2) There were signs of peritonitis, including abdominal tenderness, rebound tenderness, and muscular rigidity, especially in the area corresponding to the anastomosis. (3) An increase in the white blood cell count and a significant rise in C-reactive protein indicated the possible presence of an infection. (4) The CT scan showed fluid accumulation, gas accumulation around the anastomosis, or extravasation of the contrast medium. (5) The abdominal drainage tube drained turbid fluid, which suggested the presence of anastomotic leakage. The attending physician, in consultation with other gastrointestinal surgery specialists, made a comprehensive judgment on the occurrence of anastomotic leakage based on the above procedures.
It is important to emphasize that abdominoperineal resection of the rectum does not involve a strict intestinal anastomosis. Thus, in this study, complications related to pelvic floor wounds or rectal stumps after this surgery (including pelvic floor infection or abscess, perineal incision dehiscence with fecal exudation or infection, and rectal stump leakage) were defined as “anastomotic leakage” for analysis.
Finally, 71 patients developed anastomotic leakage of varying degrees and were assigned to the anastomotic leakage group. The remaining 432 patients were assigned to the non-anastomotic leakage group.
Statistical analysis
The normality distribution of continuous variables was evaluated using the Kolmogorov-Smirnov test and Shapiro-Wilk test. As shown in Supplemental Table 1, both methods indicated that the continuous variables in this study were not normally distributed. Thus, these variables were presented as medians (interquartile ranges), and the differences between groups were evaluated using the Wilcoxon rank-sum test. In addition, categorical variables were presented as frequencies (constituent ratios), and the differences between groups were assessed using the chi-square test.
The multivariable logistic regression was adopted to evaluate the independent association between preoperative insulin resistance/stress response and postoperative anastomotic leakage, and reported odds ratios (OR), 95% confidence intervals (95% CI), and P-values24. In this multivariable model, a series of potential confounding factors that may influence the outcomes were adjusted for, including demographic data, personal history, history of chronic diseases (excluding classic metabolic diseases), genetic diseases or family history, preoperative serological indicators (including blood routine, coagulation profile, liver and kidney function, electrolytes, inflammation), preoperative tumor characteristics (including tumor markers, pathological features), and surgical procedures for the tumor. However, insulin resistance-related factors, including fasting blood glucose, fasting insulin, HOMA-IR, the number of individuals with insulin resistance, history of type 2 diabetes, drugs affecting insulin resistance, blood lipids, and obesity indices, as well as stress response-related factors, including vital signs, hormone indicators, and psychological indicators, were not adjusted for to avoid interfering with the results of this study.
Additionally, the study divided all patients into two groups based on the presence of preoperative insulin resistance to separately investigate the association between preoperative stress response and postoperative anastomotic leakage, thereby evaluating the potential interaction between insulin resistance and stress response. Furthermore, the study grouped these patients according to tumor location and corresponding surgical procedures to explore the associations between preoperative insulin resistance/stress response and postoperative anastomotic leakage, respectively, in order to assess the impact of tumor location and corresponding surgical procedures on these correlations.
In all the statistical analyses, a P-value < 0.05 indicated statistical significance. SPSS 23.0 software (IBM Inc., USA) was used by this study for the above analyses.
Results
Baseline characteristics of patients
In Table 1, compared with the non-anastomotic leakage group, the patients in the anastomotic leakage group were older (P = 0.039), and a higher proportion of them had a history of hypertension (P = 0.006).
Preoperative metabolic disorders of patients
In Table 2, compared with the non-anastomotic leakage group, patients in the anastomotic leakage group had higher HOMA-IR levels (P = 0.048), a higher proportion of patients had insulin resistance (P = 0.003), a higher proportion of patients had type 2 diabetes (P = 0.005), and a higher proportion of patients were taking metformin and thiazolidinediones (P = 0.032, P = 0.032). In addition, compared with the non-anastomotic leakage group, patients in the anastomotic leakage group had higher levels of total cholesterol, low-density lipoprotein cholesterol, and body mass index (P = 0.034, P = 0.028, P = 0.008).
Preoperative serological indicators of patients
In Table 3, compared with the non-anastomotic leakage group, the patients in the anastomotic leakage group had lower level of albumin (P = 0.041), and higher levels of C-reactive protein and interleukin-6 (P = 0.027, P = 0.032).
Preoperative stress responses of patients
In Table 4, compared with the non-anastomotic leakage group, the patients in the anastomotic leakage group had significantly higher heart rates, systolic blood pressures, diastolic blood pressures, respiratory rates, and epinephrine levels, as well as cortisol levels at 8 a.m. and 8 p.m. (P < 0.001, P = 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001). The scores of the SAI, TAI and IES were also higher in the anastomotic leakage group than in the non-anastomotic leakage group (P < 0.001, P < 0.001, P < 0.001).
Characteristics of colorectal cancer
In Table 5, compared with the non-anastomotic leakage group, the anastomotic leakage group had a higher proportion of patients with rectal cancer (P = 0.007) and a higher proportion of those who underwent abdominoperineal resection of the rectum (P = 0.045).
Correlation between preoperative stress response/insulin resistance and the risk of colorectal cancer anastomotic leakage
In Table 6, the multivariable model showed that the presence of insulin resistance or increases in the levels of HOMA-IR, adrenaline, cortisol (at 8 a.m. and 8 p.m.), SAI, TAI, and IES were significantly associated with an increased risk of anastomotic leakage (P < 0.05, Table 6).
Table 6.
Correlation between preoperative stress response/insulin resistance and the risk of colorectal cancer anastomotic leakage in the study.
| Variable | Multivariable model | |
|---|---|---|
| P value | OR (95%CI) | |
| All of the patients | ||
| IR | 0.005 | 2.075 (1.246 ~ 3.453) |
| HOMA-IR | 0.031 | 1.189 (1.016 ~ 1.392) |
| Epinephrine | <0.001 | 2.301 (1.818 ~ 2.911) |
| Cortisol at 8 am | <0.001 | 3.085 (2.189 ~ 4.349) |
| Cortisol at 8 pm | <0.001 | 3.291 (2.184 ~ 4.957) |
| State anxiety inventory score | <0.001 | 1.194 (1.137 ~ 1.253) |
| Trait anxiety inventory score | <0.001 | 1.366 (1.266 ~ 1.472) |
| Impact of event scale score | <0.001 | 1.269 (1.192 ~ 1.351) |
| Patients with IR | ||
| Epinephrine | <0.001 | 4.577 (1.931 ~ 6.994) |
| Cortisol at 8 am | <0.001 | 5.901 (2.798 ~ 8.902) |
| Cortisol at 8 pm | <0.001 | 6.565 (2.673 ~ 9.793) |
| Patients without IR | ||
| Epinephrine | <0.001 | 1.155 (1.017 ~ 1.595) |
| Cortisol at 8 am | <0.001 | 1.560 (1.094 ~ 2.769) |
| Cortisol at 8 pm | <0.001 | 1.878 (1.095 ~ 3.054) |
| Patients with colon cancer | ||
| IR | 0.036 | 1.674 (1.052 ~ 3.411) |
| Epinephrine | <0.001 | 2.546 (1.806 ~ 3.589) |
| Cortisol at 8 am | <0.001 | 3.947 (2.356 ~ 6.612) |
| Cortisol at 8 pm | <0.001 | 3.222 (1.777 ~ 5.845) |
| Patients with rectal cancer | ||
| IR | 0.013 | 2.576 (1.220 ~ 5.439) |
| Epinephrine | <0.001 | 2.045 (1.478 ~ 2.828) |
| Cortisol at 8 am | <0.001 | 2.416 (1.494 ~ 3.909) |
| Cortisol at 8 pm | <0.001 | 3.463 (1.955 ~ 6.134) |
Note: IR = Insulin resistance, HOMA-IR = Homeostatic model assessment - insulin resistance, OR = Odds ratio, 95%CI = 95% confidence interval. The multivariable model adjusted for the patients’ baseline characteristics, serological indicators, tumor characteristics, and surgical characteristics. A P-value less than 0.05 indicates statistical significance.
The subgroup analysis reported that the correlations between adrenaline, cortisol (8 a.m. and 8 p.m.) and the risk of anastomotic leakage remained statistically significant in both patients with and without insulin resistance (P < 0.05, Table 6). More importantly, these results also showed that when the levels of adrenaline or cortisol increased, the risk of anastomotic leakage increased more significantly in patients with insulin resistance than in those without insulin resistance.
The subgroup analyses were also conducted in patients with colon cancer or rectal cancer. The results suggested that the correlations between insulin resistance, adrenaline levels, and cortisol levels (at 8 a.m. and 8 p.m.) and the risk of anastomotic leakage were statistically significant in both subgroups (P < 0.05, Table 6). Additionally, the magnitude of the increase in the risk of anastomotic leakage caused by the exposure factors showed little difference between the two subgroups.
Discussion
According to the results of this study, both the categorical variable of insulin resistance and the continuous variable of HOMA-IR were significantly associated with anastomotic leakage risk following colorectal cancer surgery. Specifically, the presence of insulin resistance doubled the risk of anastomotic leakage, while each standard deviation increase in HOMA-IR levels correlated with a approximately 19% elevation in anastomotic leakage risk. These findings appear to confirm a dose-response relationship between insulin resistance and anastomotic leakage risk. Previous studies have already identified associations between various metabolic disorders and anastomotic leakage risk, which partially corroborates the results of this study5–7.
The results of this study also confirmed that increases in physiological and psychological indicators related to the stress response were all associated with the occurrence of anastomotic leakage to varying degrees. Among them, the correlation between the two stress hormones, adrenaline and cortisol, and this postoperative complication was the most striking. Specifically, for each standard deviation increase in the levels of adrenaline and cortisol, the risk of anastomotic leakage increased by approximately 1 to 2 times. In contrast, for every one-standard-deviation increase in the scores of the two psychological scales, the increase in the risk of anastomotic leakage was only 20-40%. A possible explanation was that stress hormones might be involved in the pathophysiological processes related to anastomotic leakage, while psychological factors might only play an indirect role.
Additionally, this study found that insulin resistance appeared to influence the association between preoperative stress response and postoperative anastomotic leakage. In the subgroup analyses, the increase in postoperative anastomotic leakage risk induced by preoperative stress response was more pronounced in patients with insulin resistance; conversely, the risk increase was less substantial in patients without insulin resistance. Thus, we hypothesized that a synergistic effect may exist between insulin resistance and stress response in influencing postoperative anastomotic leakage.
Furthermore, the subgroup analyses also confirmed that tumor location and corresponding surgical procedures had no obvious impact on the association between preoperative insulin resistance/stress response and postoperative anastomotic leakage. This suggests that the influence of preoperative insulin resistance/stress response on this postoperative complication is generalizable or stable across different anatomical and procedural contexts.
The focus of this study on preoperative stress response rather than postoperative stress response stems primarily from the aim to explore the prognostic impact of preoperatively modifiable factors (including stress response and insulin resistance). The findings can assist physicians in completing risk stratification before surgery and making necessary adjustments to clinical decisions such as surgical plans and anesthesia methods. Although the impact of postoperative stress response on anastomotic leakage cannot be ignored, it overlaps significantly in timing with the anastomotic healing process, which substantially diminishes its utility as an independent predictive factor.
At present, there is no unified diagnostic standard for stress response. In view of this situation, based on previous experience, this study screened and included as many as 10 relevant indicators. These indicators covered three aspects: vital signs, stress hormones, and psychology. Through comprehensive consideration and analysis, it was possible to accurately determine whether a stress response occurred. This enhanced the reliability and persuasiveness of the results of this study.
In this study, the incidence of anastomotic leakage was approximately 14%. While previous literature reports a wide range of incidences (from 2.8 to 30%), the majority of studies report incidences below 10%4,25. Thus, the incidence in this study is relatively high. In reality, numerous factors influence the risk of anastomotic leakage. However, for the hospital in this study, the specific reasons are as follows: As a regional authoritative institution for colorectal cancer treatment, the hospital admits a high proportion of transferred patients. These patients often have complex medical conditions or underlying comorbidities, which objectively increase the risk of anastomotic leakage. This scenario not only motivated the conduct of this study but also provided a unique subject pool for its implementation.
The rationale for using logistic regression instead of Cox regression in this study to investigate the impact of preoperative stress response and insulin resistance on the risk of anastomotic leakage within 1 month after surgery is as follows: (1) This study focuses on preliminary risk factor identification, with the observation endpoint defined as the occurrence or non-occurrence of anastomotic leakage within 1 month postoperatively. As a binary outcome within a fixed time window, this endpoint is adequately addressed by logistic regression. (2) The vast majority of anastomotic leakages in colorectal cancer patients occur within 1–2 weeks postoperatively, a very short time frame that calls into question the practical value of time-to-event analysis. (3) The anastomotic leakage incidence in this study is approximately 14%, a rate that does not qualify as a “rare event” in statistical terms. In such cases, the OR introduces minimal bias in estimating the true risk ratio and is unlikely to significantly overestimate risk. However, future studies with longer postoperative follow-up periods that focus on the impact of preoperative factors on early, intermediate, and late anastomotic leakage risk will necessitate Cox regression.
In the multivariable logistic regression model, the selection of variables for adjustment aligned with the core objectives of this study. As previously noted, the model included a series of factors such as demographic data and tumor characteristics for adjustment, as these variables are widely recognized as potential influencers of anastomotic leakage risk, probably operating through pathways independent of insulin resistance or stress response and thus capable of introducing significant confounding. Meanwhile, insulin resistance-related factors (e.g., fasting blood glucose, insulin, and HOMA-IR) and stress response-related metrics (vital signs, hormonal/psychological markers) were not adjusted for, primarily due to two reasons: (1) These factors serve as direct measurement indicators or mechanistic intermediate variables of the exposure variables, and their adjustment would erroneously attenuate or mask the direct effects of insulin resistance/stress response on anastomotic leakage; (2) Strong correlations exist between certain indicators (e.g., fasting blood glucose and HOMA-IR, systolic blood pressure and cortisol) due to physiological mechanisms or calculation logic, and their incorrect inclusion in the model could introduce multicollinearity risks. This strategy aimed to preserve the integrity of the exposure variables to isolate their independent effects on anastomotic leakage risk and avoid potential bias.
Insulin resistance and stress response may affect the risk of anastomotic leakage in colorectal cancer through multiple potential mechanisms. During insulin resistance, the reduced insulin sensitivity triggers disorders in glucose and lipid metabolism26. The high-glucose environment is conducive to the growth of bacteria, increasing the risk of infection, and the abnormal lipid metabolism affects the transportation of nutrients. Insulin resistance can also lead to endothelial dysfunction, reducing local blood flow and impeding the healing of the anastomosis27. The activation of the inflammatory signaling pathway causes chronic inflammation, disrupting the healing microenvironment28. When a stress response occurs, the body secretes stress hormones such as adrenaline and cortisol. The former causes blood vessels to constrict, reducing the blood supply to the anastomosis, while the latter suppresses the immune function, increasing the risk of infection. Stress can also promote an increase in catabolism within the body, resulting in a negative nitrogen balance. This, in turn, affects protein synthesis and cell proliferation, which is detrimental to the healing of the anastomosis29.
However, this study has several limitations. First, mechanistic insights were not explored through animal experiments or in vitro models. Second, the bidirectional relationship between metabolic disorders and stress responses remained uninvestigated, primarily due to limited access to specialized research resources and data. Future studies incorporating experimental models and long-term follow-up are needed to address these gaps.
In conclusion, this study confirmed that both preoperative insulin resistance and stress response were potential risk factors for anastomotic leakage after colorectal cancer surgery. This study also speculated that there seemed to be a synergistic effect between preoperative insulin resistance and stress response in influencing this postoperative complication. If further confirmed by future studies, the above findings and speculations may contribute to a more comprehensive understanding of the pathogenesis of anastomotic leakage after colorectal cancer surgery, thus strengthening relevant prevention and control efforts.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Author contributions
Chen Binghe were responsible for the overall study design. Li Shuaichao, Gao Zhengjie, Fan Longxin, Meng Tao and Chen Binghe carried out data collection, processing, and analysis. Li Shuaichao, Gao Zhengjie, Fan Longxin, Meng Tao and Chen Binghe drafted the initial version of the article. Li Shuaichao, Gao Zhengjie, Fan Longxin, Meng Tao and Chen Binghe participated in the review and finalization of the entire text.
Data availability
The datasets of this study are not publicly available for the time being because sufficient authorization from all the participants has not been obtained. However, these datasets can be obtained from the corresponding author upon a reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Ethics approval and consent to participate
This study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Xinxiang Medical University, China. All methods were carried out in accordance with the Declaration of Helsinki of the World Medical Association and other relevant guidelines and regulations. All the participants gave their consent to take part in this study and signed the informed consent forms.
Consent for publication
All patients consented to the publication of this study.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Morgan, E. et al. Global burden of colorectal cancer in 2020 and 2040: incidence and mortality estimates from GLOBOCAN. Gut72, 338–344 (2023). [DOI] [PubMed] [Google Scholar]
- 2.Qu, R., Ma, Y., Zhang, Z. & Fu, W. Increasing burden of colorectal cancer in China. Lancet Gastroenterol. Hepatol.7, 700 (2022). [DOI] [PubMed] [Google Scholar]
- 3.Yau, T. O. Precision treatment in colorectal cancer: now and the future. JGH Open.3, 361–369 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chiarello, M. M. et al. Anastomotic leakage in colorectal cancer surgery. Surg. Oncol.40, 101708 (2022). [DOI] [PubMed] [Google Scholar]
- 5.Li, S. J. et al. Diabetes mellitus and risk of anastomotic leakage after esophagectomy: a systematic review and meta-analysis. Dis. Esophagus. 30, 1–12 (2017). [DOI] [PubMed] [Google Scholar]
- 6.Yu, L., Wu, W., Xia, S., Li, Y. & Xu, Z. Visceral obesity and anastomotic leakage rates in colorectal cancer: a systematic review and meta-analysis. Front. Oncol.13, 1224196 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fan, Y. et al. Predictive value of early postoperative blood lipid metabolism for anastomotic leakage after esophageal cancer surgery. Am. J. Transl Res.16, 3794–3800 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rao, G. Insulin resistance syndrome. Am. Fam Physician. 63, 1159–1163 (2001). [PubMed] [Google Scholar]
- 9.Bruce, W. R., Wolever, T. M. & Giacca, A. Mechanisms linking diet and colorectal cancer: the possible role of insulin resistance. Nutr. Cancer. 37, 19–26 (2000). [DOI] [PubMed] [Google Scholar]
- 10.Dev, R., Bruera, E. & Dalal, S. Insulin resistance and body composition in cancer patients. Ann. Oncol.29, ii18–ii26 (2018). [DOI] [PubMed] [Google Scholar]
- 11.Behrenbruch, C. et al. Surgical stress response and promotion of metastasis in colorectal cancer: a complex and heterogeneous process. Clin. Exp. Metastasis. 35, 333–345 (2018). [DOI] [PubMed] [Google Scholar]
- 12.Russell, G. & Lightman, S. The human stress response. Nat. Rev. Endocrinol.15, 525–534 (2019). [DOI] [PubMed] [Google Scholar]
- 13.Charmandari, E., Tsigos, C. & Chrousos, G. Endocrinology of the stress response. Annu. Rev. Physiol.67, 259–284 (2005). [DOI] [PubMed] [Google Scholar]
- 14.World Medical Association. World medical association declaration of helsinki: ethical principles for medical research involving human subjects. JAMA310, 2191–2194 (2013). [DOI] [PubMed] [Google Scholar]
- 15.Cubiella, J. et al. Clinical practice guideline. Diagnosis and prevention of colorectal cancer. 2018 Update Gastroenterol. Hepatol.41, 585–596 (2018). [DOI] [PubMed] [Google Scholar]
- 16.Tahapary, D. L. et al. Challenges in the diagnosis of insulin resistance: focusing on the role of HOMA-IR and tryglyceride/glucose index. Diabetes Metab. Syndr.16, 102581 (2022). [DOI] [PubMed] [Google Scholar]
- 17.Tang, Q., Li, X., Song, P. & Xu, L. Optimal cut-off values for the homeostasis model assessment of insulin resistance (HOMA-IR) and pre-diabetes screening: developments in research and prospects for the future. Drug Discov Ther.9, 380–385 (2015). [DOI] [PubMed] [Google Scholar]
- 18.Julian, L. J. Measures of anxiety: State-Trait anxiety inventory (STAI), Beck anxiety inventory (BAI), and hospital anxiety and depression Scale-Anxiety (HADS-A). Arthritis Care Res. (Hoboken). 63, S467–S472 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Horowitz, M., Wilner, N. & Alvarez, W. Impact of event scale: a measure of subjective stress. Psychosom. Med.41, 209–218 (1979). [DOI] [PubMed] [Google Scholar]
- 20.Smith, A. J. et al. Guideline for optimization of colorectal cancer surgery and pathology. J. Surg. Oncol.101, 5–12 (2010). [DOI] [PubMed] [Google Scholar]
- 21.Rentsch, M., Schiergens, T., Khandoga, A. & Werner, J. Surgery for colorectal Cancer - Trends, developments, and future perspectives. Visc. Med.32, 184–191 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Denicu, M. M. et al. Anastomotic leakage after colorectal surgery: risk factors, diagnosis and therapeutic options. Curr. Health Sci. J.49, 333–342 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kauv, P. et al. Anastomotic leakage after colorectal surgery: diagnostic accuracy of CT. Eur. Radiol.25, 3543–3551 (2015). [DOI] [PubMed] [Google Scholar]
- 24.Bosch-Gaya, A., Matute-Blanco, L., Fernández-Rodríguez, D. & Worner, F. Comparative statistical analysis in observational epidemiological studies. Med. Intensiva (Engl Ed). 42, 134–135 (2018). [DOI] [PubMed] [Google Scholar]
- 25.Li, Y. W. et al. Very early colorectal anastomotic leakage within 5 Post-operative days: a more severe subtype needs relaparatomy. Sci. Rep.7, 39936 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Col, V. New insights into insulin resistance pathophysiology: how it affects glucose and lipid metabolism. Acta Clin. Belg.56, 155–162 (2001). [DOI] [PubMed] [Google Scholar]
- 27.Jansson, P. A. Endothelial dysfunction in insulin resistance and type 2 diabetes. J. Intern. Med.262, 173–183 (2007). [DOI] [PubMed] [Google Scholar]
- 28.Matulewicz, N. & Karczewska-Kupczewska, M. Insulin resistance and chronic inflammation. Postepy Hig Med. Dosw (Online). 70, 1245–1258 (2016). [PubMed] [Google Scholar]
- 29.Larsen, V. & Brockner, J. Nitrogen balance and operative stress. Effect of early postoperative nutrition on nitrogen balance following major surgery. Acta Chir. Scand. Suppl.343, 191–202 (1965). [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets of this study are not publicly available for the time being because sufficient authorization from all the participants has not been obtained. However, these datasets can be obtained from the corresponding author upon a reasonable request.