Impact of diabetes mellitus history on postoperative infectious complications in gastric or colorectal cancer patients

Abstract

BACKGROUND

Postoperative infectious complications remain one of the major causes of morbidity in gastric or colorectal cancer patients who opt for surgical intervention. Diabetes mellitus has been previously shown to impair immune and metabolic functions, thereby increasing the potential risk of postoperative infections. However, very limited evidence is available to establish the same in gastrointestinal cancer patients.

AIM

To assess the effect of diabetes mellitus history on postoperative infectious complications observed in gastric or colorectal cancer patients.

METHODS

The study included 200 gastric or colorectal cancer patients (admitted from January, 2022 to January, 2025). The occurrence of infectious complications within 30 days of surgery was recorded, following which the patients were divided into two groups: (1) The postoperative infection group (n = 48); and (2) The non-infection group (n = 152). On the basis of diabetes history, the participants were further divided into the diabetes group (n = 62) and the non-diabetes group (n = 138). Risk factors for postoperative infectious complications were assessed, with odds ratios and 95% confidence intervals specifically calculated concerning diabetes history.

RESULTS

Among 200 subjects included in the study, 48 patients showed postoperative infectious complications. Out of these, only 25 (52.08%) patients had a previous history of diabetes. Interestingly, surgical site infection was found to be the most common complication [27 (56.3%) cases], followed by pulmonary infection [11 (22.9%) cases]. The univariate and multivariate logistic regression analyses further revealed that diabetes history with preoperative glycated hemoglobin A1c ≥ 7.0% and preoperative hypoalbuminemia (albumin < 35 g/L) acted as independent risk factors for postoperative infectious complications in gastric or colorectal cancer patients (P < 0.05).

CONCLUSION

Altogether, diabetes mellitus acted as a significant independent risk factor that increased the overall incidence of postoperative infectious complications in gastric/colorectal cancer patients. The observed risk might be attributed to certain mechanisms that affect different metabolic and immune functions. Strategies focusing on comprehensive assessment, intervention, and better management strategies might aid in reducing the incidence of postoperative infectious complications in gastric or colorectal cancer patients, thereby improving overall patient prognosis and quality of life.

Key Words: Diabetes mellitus; Gastric cancer; Colorectal cancer; Postoperative infections; Complications

Core Tip: Diabetes mellitus acted as a significant independent risk factor in postoperative infectious complications reported in patients with gastric or colorectal cancer. Importantly, the patients with a known history of diabetes, characterized by elevated preoperative glycated hemoglobin A1c ≥ 7.0%, and hypoalbuminaemia (albumin < 35 g/L) displayed a higher risk of postoperative infectious complications. Surgical site infection was the most common postoperative infection, followed by pulmonary infection. Early identification, preoperative optimization of blood glucose and nutritional status, and targeted perioperative management might help to reduce the incidence of postoperative infections and improve overall patient outcomes.

INTRODUCTION

Gastric cancer and colorectal cancer are the most prevalent malignant tumours that affect the digestive tract. In fact, they are among the leading causes of cancer incidence and mortality worldwide. Surgical resection remains the most commonly utilized clinical treatment modality[1]. However, postoperative infectious complications, such as surgical site infections (SSIs), pneumonia, urinary tract infections, and bacteraemia, severely impact patient prognosis. These complications often contribute to prolonged hospital stays and increased healthcare costs. In fact, these might lead to fatal outcomes in severe cases[2,3].

Diabetes mellitus has been widely recognized as a systemic metabolic disorder that is often associated with impaired neutrophil chemotaxis, reduced macrophage function, microvascular damage, delayed wound healing, and chronic low-grade inflammation[4]. These pathophysiological abnormalities are known to compromise host immune responses and may predispose cancer patients to serious postoperative infectious complications. However, limited studies have been previously conducted to evaluate the effect of diabetes history on the magnitude of infection risk in patients undergoing gastric or colorectal cancer surgery. In addition, the influence of preoperative glycaemic control on the patient outcome remains unclear[5].

Several previous studies provided evidence to support a connection between diabetes mellitus, hyperglycemia, and postoperative infection; however, studies specifically targeting patients with gastric or colorectal cancer remain limited. The available studies find limited applications owing to the combined use of various surgical populations or the predominant inclusion of Western cohorts. Consequently, the outcomes of these studies cannot be applied to Asian patients, primarily due to differences in the metabolic profiles and perioperative practices. Furthermore, a few studies also evaluated the combined predictive value of diabetes history, preoperative glycated hemoglobin A1c (HbA1c), and albumin levels within a unified risk model.

Postoperative infectious complications are among the most clinically significant adverse outcomes following gastric and colorectal cancer surgery. Beyond prolonging hospital stay and increasing healthcare costs, these complications can delay the initiation of adjuvant therapy and negatively affect long-term survival. Growing evidence indicates that even mild postoperative infections may provoke systemic inflammatory responses, hinder immune recovery, and ultimately compromise oncological outcomes[6]. Consequently, identifying modifiable risk factors for postoperative infectious complications remains a critical priority in the perioperative management of gastrointestinal malignancies.

From a pathophysiological perspective, diabetes mellitus is characterized by chronic hyperglycemia, insulin resistance, and a persistent low-grade inflammatory state, all of which predispose patients to postoperative infectious complications. Hyperglycemia impairs neutrophil chemotaxis and phagocytic activity, suppresses macrophage activation, and disrupts cytokine-signaling pathways, thereby weakening both innate and adaptive immune responses. Furthermore, diabetes-associated microvascular dysfunction and endothelial injury compromise tissue perfusion and oxygen delivery, resulting in delayed wound healing and an increased susceptibility to surgical site infections.

Despite these well-established biological mechanisms, the magnitude of infection risk attributable to diabetes mellitus, particularly in patients undergoing gastric or colorectal cancer surgery, remains insufficiently characterized, and the role of long-term glycemic control in this context remains controversial.

Therefore, the present study aimed to assess the association between diabetes history and postoperative infectious complications in gastric or colorectal cancer patients undergoing surgical intervention. Diabetes mellitus is known to be often associated with metabolic dysregulation, compromised immune function, and establishment of a chronic inflammatory microenvironment. Consequently, gastric or colorectal cancer patients with a history of diabetes might suffer from a heightened risk of postoperative infectious complications, demanding significant clinical attention. Thus, the present study investigated the specific impact of diabetes history on postoperative infectious complications in gastric or colorectal cancer patients, which might provide evidence-based guidance for clinical prevention and management in the future.

MATERIALS AND METHODS

Study population

The present study was conducted as per the guidelines of the hospital ethics committee. A total of 200 patients diagnosed with gastric cancer or colorectal cancer (from January 2022 to January 2025) were enrolled in the study. Sample size was calculated on the basis of previous studies, with a significance level of α = 0.05 (two-tailed) and power of 1-β = 0.80, wherein a 10% attrition or data loss rate was factored in to determine the total sample size.

Inclusion criteria: (1) Age ≥ 18 years; (2) Pathologically confirmed primary gastric or colorectal cancer; (3) Scheduled for radical surgical treatment; (4) Preoperative informed consent; and (5) Voluntary participation in this study.

Exclusion criteria: (1) Emergency surgery; (2) Confirmed active infection within 30 days before the surgery; (3) Concurrent occurrence of other malignant tumors; (4) Concurrent severe immunodeficiency disease; (5) Concurrent severe hepatic or renal insufficiency; and (6) Missing clinical data or missed follow-up.

All eligible patients were initially classified based on their diabetes mellitus history into diabetes and non-diabetes group. The primary outcome was the occurrence of postoperative infectious complications within 30 days of surgery. The incidence and distribution of these complications were then compared between patients without and with diabetes mellitus. For descriptive and analytical purposes, patients were also stratified by infection status to examine clinical characteristics associated with postoperative infectious complications. The target sample size was estimated a priori based on an anticipated difference in postoperative infection rates between patients without and with diabetes mellitus. Although the study ultimately employed a retrospective cohort design, the sample size calculation was informed by assumptions derived from past studies on postoperative infection rates in diabetic and non-diabetic populations. The following assumptions were taken from two recent studies that involved an Asian cohort, wherein baseline infection rates were reported after gastric or colorectal resection, with baseline infection rate in the non-diabetic patients = 15%[6] and expected rate in the diabetic patients = 32% [corresponding to odds ratio (OR) ≈ 2.5].

Following standard formula was used to compare two proportions; n = [Zα√(2p^-q^-) + Zβ√(p₁q₁ + p₂q₂)]²/(p₁ - p₂)², with α = 0.05 (two-tailed, Zα = 1.96), power = 0.80 (Zβ = 0.84), p₁ = 0.32, p₂ = 0.15, p^- = (p₁ + p₂)/2. The calculated size was n = 91 per group. Further, anticipating a 10% loss to follow-up, the minimally required sample size was calculated to be 100 diabetic and 100 non-diabetic patients. To improve the precision for multivariable logistic regression (≤ 10 events per covariate), the final cohort was increased to 200 patients, including 62 diabetic and 138 non-diabetic subjects, which satisfied the event requirement (48 infection events × 7 covariates). Postoperative infectious complications were strictly defined according to the Centers for Disease Control and Prevention/National Healthcare Safety Network 2022 guidelines. The patients were screened by dedicated infection control nurses for 30 days post-surgery, reviewed by the surgical team, and finally assessed by blinded infectious disease physicians. The diagnostic criteria for postoperative infection included: Pneumonia (new radiographic infiltrates + fever or purulent sputum + positive microbiology), urinary tract infection (≥ 10⁵ CFU/mL with pyuria and symptoms), bloodstream infection (positive blood culture + fever or hypotension), Clostridioides difficile colitis, and catheter-associated infections. Only cases meeting these criteria were included in the analysis.

Research methods

For all the patients included in the study, following data were collected: Gender, age, body mass index, American Society of Anesthesiologists (ASA) physical status classification, smoking history, alcohol consumption history, diabetes history, tumor location (gastric or colorectal), pathological staging [8^th edition American Joint Committee on Cancer Tumor-Node-Metastasis (TNM) staging system staging], differentiation grade, preoperative serum albumin level, surgical approach (open or minimally invasive), operative duration, intraoperative blood loss, intraoperative transfusion status, anastomosis technique (gastric or colorectal), prophylactic antibiotic use (type, timing, and duration), postoperative blood glucose monitoring frequency, postoperative insulin regimen, and postoperative nutritional support route (enteral or parenteral). In the present study, diabetes history was defined on the basis of following criteria: (1) Drug use: A history of intake of hypoglycemic medication or insulin for ≥ 90 days before the surgery; and (2) Glucose levels: Two fasting blood glucose readings ≥ 7.0 mmol/L or an Oral Glucose Tolerance Test 2-hour glucose level ≥ 11.1 mmol/L, recorded within 6 months before the surgery (American Diabetes Association 2022 criteria). The patients with isolated hyperglycemia readings, primarily at the time of initial hospitalization, were excluded.

HbA1c and serum albumin levels were measured from the venous blood samples, collected at the anesthesia clinic 1-7 days before the surgery, using high performance liquid chromatography and bromocresol green methods, respectively. These results served as baseline measurements, with albumin < 35 g/L indicating preoperative hypoalbuminemia. According to the American Diabetes Association guidelines and previously reported perioperative studies, HbA1c ≥ 7.0% was selected to define poor preoperative glycemic control. This threshold is commonly adopted to identify patients at elevated risk of postoperative complications.

To minimize information bias, all clinical variables were reported as per standardized definitions. Further, these were independently verified by two trained researchers. Postoperative infectious complications were identified according to Centers for Disease Control and Prevention/National Healthcare Safety Network criteria, and further adjudicated by infectious disease specialists, who were blinded to patient diabetes status.

Statistical analysis

The collected data were analyzed using SPSS 22.0. Quantitative data (mean ± SD) were assessed using t-tests, while categorical data, n (%) were evaluated using χ² tests. The variables demonstrating statistical significance in univariate analysis were incorporated into a multivariate logistic regression model, and ORs and their 95% confidence intervals were calculated. Here, P < 0.05 indicated a statistically significant difference.

Certain clinically relevant covariates, including frailty indices, perioperative steroid exposure, glycemic variability metrics, and validated nutritional scores, including the Prognostic Nutritional Index and Geriatric Nutritional Risk Index, were not included in the study due to incomplete documentation in the electronic medical record system. These variables were excluded from the primary model as their inclusion would have resulted in substantial missing data and reduced statistical power.

RESULTS

Distribution of postoperative infection complications

Among 200 gastric or colorectal patients included in the study, 48 patients displayed postoperative infectious complications. Out of these, 52.08% (25 patients) were diabetic, while 47.92% (23 patients) belonged to the non-diabetic group. SSI was found to be the most common postoperative complication, observed in 56.3% (27 patients). This was followed by pulmonary infection in 22.9% (11 patients). The distribution of postoperative infectious complications and SSI types is shown in Figure 1.

Figure 1 Distribution of postoperative infection complications and types of surgical site infections. SSI: Surgical site infection; DM: Diabetes mellitus.

Univariate analysis of postoperative infectious complications in gastric or colorectal cancer patients

Factors that exhibited statistically significant differences between the infected and the non-infected groups included age ≥ 65 years, history of diabetes mellitus, preoperative HbA1c ≥ 7.0%, preoperative hypoalbuminemia (Alb < 35 g/L), ASA grade ≥ III, tumor TNM stage III-IV, and operative time ≥ 180 min (P < 0.05), thus acting as potential risk factors for postoperative infectious complications. Comparatively, no significant differences were reported in the two groups for factors like sex, body mass index, tumor location, and surgical approach (P > 0.05, Table 1).

Table 1 Univariate analysis of postoperative infectious complications in gastric/colorectal cancer patients, n (%).

Factor	Postoperative infection group (n = 48)	Non-infection group (n = 152)	χ2 value	t value	P value
Age ≥ 65 years	32 (66.7)	56 (36.8)	13.607	-	< 0.001
Sex (male)	28 (58.3)	90 (59.2)	0.011	-	0.916
BMI (kg/m2), mean ± SD	23.54 ± 3.18	22.86 ± 3.43	-	1.318	0.189
History of diabetes	25 (52.1)	37 (24.3)	13.577	-	< 0.001
HbA1c ≥ 7.0%	22 (45.8)	30 (19.7)	13.261	-	< 0.001
Alb < 35 g/L	20 (41.7)	32 (21.1)	8.355	-	0.004
ASA score ≥ III	19 (39.6)	35 (23.0)	5.239	-	0.022
Tumor location (colorectal)	30 (62.5)	82 (53.9)	1.081	-	0.299
TNM stage (III-IV)	35 (72.9)	80 (52.6)	6.225	-	0.013
Surgical approach (open)	20 (41.7)	50 (32.9)	1.212	-	0.271
Operation time ≥ 180 minutes	36 (75.0)	78 (51.3)	8.379	-	0.004
Intraoperative blood loss ≥ 200 mL	18 (37.5)	45 (29.6)	1.052	-	0.305

BMI: Body mass index; HbA1c: Hemoglobin A1c; Alb: Albumin; ASA: American Society of Anesthesiologists; TNM: Tumor-node-metastasis.

Multivariate logistic regression analysis of postoperative infectious complications in gastric or colorectal cancer patients

For multivariate logistic regression analysis, age (≥ 65 = 1, < 65 = 0), history of diabetes mellitus (absent = 0, present = 1), HbA1c (≥ 7% = 1, < 7% = 0), albumin (≥ 35 g/L = 0, < 35 g/L = 1), ASA classification (I-II = 0, III-IV = 1), TNM stage (I-II = 0, III-IV = 1), and operative duration (< 180 minutes = 0, ≥ 180 minutes = 1) were scored and utilized. The results for the analysis indicated that a history of diabetes mellitus, preoperative HbA1c ≥ 7.0%, and preoperative hypoalbuminemia (Alb < 35 g/L) acted as independent risk factors for postoperative infectious complications in gastric or colorectal cancer patients (P < 0.05). These results further demonstrated that diabetes history (OR = 3.335, 95% confidence interval: 1.532-7.258), HbA1c ≥ 7.0% (OR = 6.524), and albumin < 35 g/L (OR = 2.678) remained significant independent predictors for postoperative infections following adjustment of confounders (Table 2).

Table 2 Multivariate logistic regression analysis of postoperative infectious complications in gastric/colorectal cancer patients.

Factor	β	Wald χ2	OR	95%CI	P value
History of diabetes	1.204	8.765	3.335	1.532-7.258	0.003
HbA1c ≥ 7.0%	1.876	12.341	6.524	2.456-17.329	< 0.001
Alb < 35 g/L	0.985	6.218	2.678	1.234-5.814	0.013
Age ≥ 65 years	0.672	3.112	1.958	0.912-4.205	0.078
ASA score ≥ III	0.501	1.876	1.651	0.754-3.614	0.171
TNM stage (III-IV)	0.587	2.654	1.799	0.823-3.931	0.103
Operation time ≥ 180 minutes	0.634	3.445	1.885	0.897-3.964	0.064

OR: Odds ratio; CI: Confidence interval; HbA1c: Hemoglobin A1c; Alb: Albumin; ASA: American Society of Anesthesiologists; TNM: Tumor-node-metastasis.

DISCUSSION

Globally, gastric cancer and colorectal cancer are among the most common malignant tumors that affect the digestive tract. Despite the continuous advancements in the surgical techniques and perioperative management strategies, postoperative infectious complications remain a major clinical challenge[6,7]. These complications significantly prolong the hospitalization stay, increase overall healthcare costs, delay adjuvant treatment, and affect long-term survival outcomes[9-13]. The findings of the present study showed that diabetes mellitus, inadequate preoperative glycemic control (HbA1c ≥ 7.0%), and hypoalbuminemia (Alb < 35 g/L) acted as independent risk factors contributing towards higher incidents of postoperative infectious complications in the patients undergoing gastric and colorectal cancer surgery. These findings were in concordance with several previous studies. The present study further demonstrated the combined impact of metabolic dysregulation and nutritional impairment on postoperative infection risk in a real-world Chinese gastrointestinal cancer population, thereby providing more refined risk stratification evidence.

Diabetes mellitus might influence these postoperative outcomes via a series of complex pathophysiological mechanisms. In particular, chronic hyperglycemia has been previously shown to impair innate and adaptive immune responses via reduction of neutrophil chemotaxis, inhibition of macrophage phagocytosis, and alteration of cytokine signaling[14-18]. Furthermore, long-term elevated glucose levels are known to promote the formation of advanced glycation end-products, which in turn damage vascular endothelial cells, affect microcirculatory perfusion, and hinder tissue oxygenation. These mechanisms collectively create a pro-inflammatory and immunocompromised state that results in predisposition of diabetic patients towards postoperative infections. Recent studies also highlighted the role of oxidative stress, mitochondrial dysfunction, and metabolic memory in perpetuating immune impairment, even in the presence of relatively stable perioperative glucose levels. These deeper insights into the metabolic and immunological effects of diabetes might further explain the occurrence of a higher incidence of the SSIs and pulmonary infections in the diabetic group reported in the present study.

Additionally, the patients with poor preoperative glycemic control, as indicated by HbA1c ≥ 7.0%, exhibited a markedly higher risk of postoperative infections, with the effect size exceeding that of the diabetes group alone. In general, HbA1c levels reflect cumulative glycemic levels over a period of the last 2-3 months, and thus serve as an indicator of chronic metabolic imbalance. It has been previously reported that uncontrolled hyperglycemia intensifies oxidative stress responses, increases the production of reactive oxygen species, and disrupts cellular repair mechanisms[23-25]. In the perioperative settings, the surgical trauma is often accompanied by the induction of various metabolic stress responses, including elevated cortisol levels, enhanced insulin resistance, and systemic inflammatory activation. In case of the patients that are already experiencing a hyperglycemic state, this stress-induced deterioration in the metabolic homeostasis might further lead to impaired immune defense, increased bacterial colonization, and delayed wound healing. Thus, the findings of the present study reinforced the importance of preoperative glycemic assessment and suggested that HbA1c levels might serve as a more reliable predictor for postoperative infectious risk assessment as compared to diabetes diagnosis alone.

The present study also identified hypoalbuminemia (Alb < 35 g/L) as an independent predictor for postoperative infectious complications. Albumin is known to hold a close association with the nutritional status of the body. Additionally, it is also related to immune competence, antioxidant capacity, and systemic inflammatory responses. Previous studies have shown that lower albumin levels are often associated with sarcopenia, frailty, impaired tissue repair, and poor tolerance to surgical trauma[26-30]. From a physiological perspective, hypoalbuminemia is known to act via reduction of plasma colloid osmotic pressure, further leading to interstitial edema and impaired tissue perfusion. This, in turn, hinders oxygen and nutrient supply to the surgical wounds. Furthermore, low albumin levels diminish the transport of hormones, fatty acids, and medications, weaken antioxidative buffering, and reflect underlying chronic inflammation. These biological alterations collectively weaken the ability of the body to fight against infections and support tissue regeneration, thereby making hypoalbuminemia a significant contributor to postoperative infectious complications. Thus, the results of the present study suggested that preoperative nutritional optimization should be considered as a key component of perioperative care for gastric and colorectal cancer patients opting for surgical intervention.

Importantly, the combined occurrence of diabetes, elevated HbA1c, and hypoalbuminemia indicated the presence of a cluster of high-risk metabolic and nutritional phenotypes in the patient cohort. These interconnected risk factors further created a cumulative effect on postoperative outcomes in the gastric or colorectal cancer patients. In particular, diabetes induces chronic metabolic derangements, HbA1c reflects the degree of sustained hyperglycemia, and albumin indicates overall nutritional reserve and inflammatory status. The coexistence of these abnormalities results in substantial impairment of the patient’s metabolic–immune resilience, further leading to increased vulnerability to surgical stress and infection. Therefore, the future clinical management strategies should consider the integration of metabolic and nutritional screening results to construct a more comprehensive preoperative risk assessment model, allowing for individualized intervention strategies.

The findings of the present study have strong clinical implications in the management of gastric or colorectal cancer patients opting for surgical intervention. Firstly, the diabetic patients undergoing gastric or colorectal cancer surgery should receive systematic preoperative evaluation. In addition to regular assessment of fasting blood glucose levels, HbA1c measurement should also be included. For patients with HbA1c ≥ 7.0%, short-term preoperative intervention measures, such as individualized glucose-lowering therapy, diet modification, structured exercise, and enhanced perioperative glucose monitoring, should be considered. Second, the nutritional optimization is equally essential in these patients. Particularly, incorporation of strategies like early nutritional risk assessment, dietitian consultation, consumption of a high-quality protein diet, oral nutritional supplementation, use of oral immunonutrition formulas, and short-duration preoperative parenteral nutrition might aid in improving overall albumin levels and physiological resilience of the body. Third, perioperative management strategies should be tailored according to the patient’s metabolic-nutritional profile. Enhanced monitoring of blood glucose levels, judicious use of antibiotics, early mobilization, respiratory exercises, wound care reinforcement, and closer postoperative observation might be considered essential for high-risk patients.

Despite these positive findings, the present study suffers from several limitations. Primarily, the present study involved a single-center retrospective design that might lead to inherent selection bias. This would further affect the generalization of these results to a larger population. Although the current study used stringent infection definitions and standardized evaluation procedures, the residual confounding cannot be eliminated. Also, certain clinically important variables, such as frailty status, perioperative steroid use, and glycemic variability, were not fully assessed due to the limitations linked to the medical record system. Particularly, the information regarding recognized nutritional indices, such as the Prognostic Nutritional Index and the Geriatric Nutritional Risk Index, was unavailable for many patients, further limiting a more comprehensive assessment. Clinically, HbA1c ≥ 7.0% is widely used as a clinical threshold for diabetes assessment; however, in the present study, it was not derived from data-driven statistical methods, such as receiver operating characteristic curve analysis or restricted cubic spline modeling. In the future, multicenter prospective studies should be conducted with a focus on the collection of more comprehensive metabolic and nutritional data profiles. Additionally, these studies should employ sophisticated modeling approaches to establish more accurate and individualized risk thresholds.

CONCLUSION

In conclusion, the present study demonstrated that diabetes mellitus, poor preoperative glycemic control, and hypoalbuminemia significantly increased the risk of postoperative infectious complications in patients undergoing gastric or colorectal cancer surgery. These findings highlighted the importance of integrated metabolic-nutritional assessment and the use of individualized perioperative optimization strategies in these patients. Importantly, the strengthening of the preoperative evaluation system, including HbA1c screening, proactive glucose regulation, and nutritional enhancement, might help to reduce the incidence of postoperative infectious complications and improve overall recovery outcomes. Future studies are required to validate these results in larger and multicenter settings. Additional studies focusing on more refined metabolic-nutritional risk prediction models might further improve the perioperative care in patients with gastrointestinal malignancies.