Association between different patterns of obesity and the short-term outcomes of gastric cancer surgery.

Introduction
Gastric cancer is a global health problem, with more than 1 million newly diagnosed gastric cancer patients worldwide each year. Although the incidence and mortality of gastric cancer have declined globally over the past 50 years, it remains the fifth leading cause of cancer-related deaths (Thrift and El-Serag, 2020; Ajani et al., 2022). Holistic therapies for gastric cancer, including chemotherapy, targeted therapies, and immunotherapies, have evolved considerably over the past few years. For resectable gastric cancer, surgical treatment remains the mainstay (Sexton et al., 2020; Guan et al., 2023). However, postoperative adverse events continue to pose a challenge to surgeons. Meanwhile, the incidence of obesity-related cancer is persistently increasing globally. Obesity-related cancers account for approximately 11.9% in males and 13.1% in female (Avgerinos et al., 2019). The primary mechanism by which obesity leads to tumor development is through the chronic inflammatory state induced by obesity, as well as the release of certain cytokines and hormones that promote tumor initiation and growth. First, obesity can promote tumors locally via adipose inflammation and associated changes in the microenvironment, and systemically through circulating metabolism and inflammatory mediators linked to adipose inflammation (Iyengar et al., 2016). Then, adipokines, such as adiponectin, plays a protective role in cancers by balancing many of the pro-tumorigenic effects of leptin. Therefore, the decrease in adiponectin level under conditions of obesity may facilitate the growth of tumor cells and alter tumor signaling pathways and metabolism (Hopkins et al., 2016). Studies have indicated that higher BMI at the time of a cancer diagnosis has been associated with poorer survival for many cancer types, including breast cancer, prostate cancer, leukemia, and oral cancer (Greenlee et al., 2017). In some tumors, however, the opposite trend has been observed, such as in esophageal cancer, lung cancer, colorectal cancer, renal cancer, gastric cancer, leukemia, and diffuse large B cell lymphoma. Thus, some scholars explain this survival benefit brought about by overweight and slight obesity as the ‘obesity paradox’ (Lee and Giovannucci, 2019). The specific reasons for this phenomenon are not yet clear. To account for variability in fat distribution, some scholars argue that central obesity is more accurate than BMI. And there is substantial evidence suggesting a positive correlation between central obesity and both the risk of developing gastrointestinal tumors and adverse outcomes (Dong et al., 2017; Du et al., 2017; Choe, 2024). However, to our knowledge, no studies have combined these two indicators of obesity to explore their relationship with short-term outcomes after gastric cancer surgery.
In this study, we combined BMI and central obesity to form different obesity patterns to assess their relationship with postoperative complications and 30-day readmission after gastric cancer surgery.

Methods and materials

Patients and clinical parameters
This study was performed following the rules of Declaration of Helsinki, and the protocol was approved by the ethics of the second Hospital of Lanzhou University.
The patients who underwent gastrectomy with gastric cancer at Lanzhou University second hospital from May 2021 to July 2023 were studied. The inclusion criteria were as follows:
Diagnosed with gastric cancer

Aged between 18 and 80 years

Available body composition assessment by Inbody720 within 1-week preoperative.

The exclusion criteria were as follows:
Patients who have undergone palliative surgery

There is limb dysfunction

Taking medications that affect body composition (e.g. diuretics)

There are mental or central nervous system diseases

Concurrent tumors in other parts of the body

All clinical and laboratory data were obtained through the electronic health records in hospital. Basic patient information includes age, gender, length of hospital stay, surgical type, chemotherapy history, hypertension, diabetes, and abdominal surgery history. The tumor data includes tumor size, histological type, location, differentiation, staging, and Lauren type. Preoperative serological indicators include preoperative albumin, neutrophils, lymphocytes, platelets, and monocytes.

Body composition evaluation
The parameters measured for preoperative body composition include BMI, waist circumference, waist-to-hip ratio (WHR), visceral fat area (VFA), and muscle mass. All parameters were measured using the Inbody720 body composition analyzer (Korean Baisbais Body Composition, Seoul, South Korea) whose technical principle involves the use of bioelectrical impedance analysis to rapidly and noninvasively measure body composition (Myatchin et al., 2020). The cutoff values for each parameters were based on the cutoff values recommended by the Inbody720 body composition analyzer.
Cutoff values:
–waist circumference: 85 cm for men and 80 cm for women

–WHR: 0.90 for men and 0.85 for women

–VFA: 100 cm2 for men and 80 cm2 for women

–BMI: 23.9 for both men and women

Then, based on the cutoff values of these parameters, we divide the patients into four groups, which can represent different pattern of obesity
For men:
–non-obesity group: waist circumference < 85 cm and WHR < 0.9 and VFA < 100 cm2 and BMI < 23.9

–central-only obesity group: waist circumference ≥ 85 cm or WHR ≥ 0.9 or VFA ≥ 100 cm2 and BMI < 23.9

–central obesity plus high BMI group: waist circumference ≥ 85 cm or WHR ≥ 0.9 or VFA ≥ 100 cm2 and BMI ≥ 23.9

–high BMI-only group: waist circumference < 85 cm and WHR < 0.9 and VFA < 100 cm2 and BMI ≥ 23.9

For women:
–non-obesity group: waist circumference < 80 cm and WHR < 0.85 and VFA < 80 cm2 and BMI < 23.9

–central-only obesity group: waist circumference ≥ 80 cm or WHR ≥ 0.85 or VFA ≥ 80 cm2 and BMI < 23.9

–central obesity plus high BMI group: waist circumference ≥ 80 cm or WHR ≥ 0.85 or VFA ≥ 80 cm2 and BMI ≥ 23.9

–high BMI-only group: waist circumference < 80 cm and WHR < 0.85 and VFA < 80 cm2 and BMI ≥ 23.9

Short-term postoperative outcomes
In this study, the short-term outcomes we observed were primarily postoperative complications and readmission within 30 days. For postoperative complications, we mainly studied those that occurred during the hospitalization period. Postoperative complications were evaluated according to the Clavien–Dindo classification (Dindo et al., 2004). And we categorize them into five grades (I–V) based on the patients’ conditions.
Significant and severe perioperative complications were defined as events with Clavien–Dindo grade ≥II and ≥III, respectively.

Statistical analysis
Analysis of variance test for continuous variables and Chi-square tests for categorical variables were performed to compare the difference in baseline characteristic among different patterns of obesity groups. We coded missing data as a missing indicator category for categorical variables, and with mean values for continuous variables. Logistic regression model examined the association between different patterns of obesity and postoperative complications (significant and severe) and readmission within 30 days, after adjustment for hypertension, diabetes, history of abdominal surgery, preoperative chemotherapy, preoperative albumin, and type of surgery in Model 1. We further adjusted for sex and age in Model 2. To further explore whether the relationship of muscle mass was independent, we controlled it in Model 3.
Potential modification effect for postoperative complications were assessed through a stratified analysis by sex (male or female), American Joint Committee on Cancer (AJCC) stage (I & II or III & IV), albumin (<35 g/l or ≥35 g/l), preoperative chemotherapy (Yes or No), age (<65 or ≥65 years), surgical type (total gastrectomy or partial gastrectomy), hypertension (Yes or No), diabetes (Yes or No), history of abdominal surgery (Yes or No), and muscle mass (<45.05 or ≥45.05 kg). We evaluated the potential effect of modification by modeling the cross-product term of the stratifying variable with different pattern of obesity. P-value <0.05 was considered statistically significant. Statistical analysis was performed using SPSS version 26.0 (IBM, Armonk, New York, USA) and GraphPad Prism 10 software (San Diego, California, USA).

Results

Basic characteristics
Based on inclusion and exclusion criteria, 248 patients (200 males, 48 females) were enrolled in the study, with clinical baseline data available in Table 1. Patients were divided into four groups based on body composition: non-obese (n = 112), central obesity-only (n = 75), high BMI-only (n = 1), and central obesity plus high BMI (n = 60). Due to the small size of the high BMI-only group, it was combined with the non-obese group. We observed that the central-only obesity group is the oldest (P = 0.001). Central obesity plus high BMI has the highest prevalence among men, while non-obesity has the highest prevalence among women (P = 0.022). The average muscle mass is the lowest in the non-obesity group (P = 0.00). Hypertension is most common in central obesity plus high BMI group (P = 0.09). And, patients with central-only obesity are more likely to develop diabetes (P = 0.019).

Postoperative complications
According to Clavien–Dindo classification, 86 (37.07%), 113 (48.71%), 25 (10.78%), 6 (2.59%), and 2 (0.86%) were classified as grade I, grade II, grade III, grade IV, and grade V, respectively in these cohorts, and there are differences among the three groups (P = 0.02) (Fig. 1).
The most frequent postoperative were: pain (66.94%), pleural effusion (54.44%), fever (35.48%), and abdominal effusion (25.00%). Among the postoperative complications we included, we identified differences in the occurrence of fistula (P = 0.025), abdominal effusion (P = 0.049), bleeding (P = 0.042), and the need for intervention treatment (P = 0.054) among the three groups (Table 2).
The rate of severe complications (Clavien–Dindo ≥ III) was observed in non-obese (7.96%), central obesity-only (24.00%), and central obesity plus high BMI groups (10.00%) (P = 0.005). For significant complications (Clavien–Dindo ≥ II), rates were 53.98% in the non-obese group, 68.00% in the central obesity-only group, and 56.67% in the central obesity plus high BMI group (P = 0.148). For severe postoperative complications, compared with central-only obesity, non-obesity and central obesity plus high BMI were protective factors (0.284 [95% CI: 0.116–0.695], P = 0.006; 0.310 [95% CI: 0.109–0.883], P = 0.028, respectively) after adjustment for hypertension, diabetes, history of abdominal surgery, preoperative chemotherapy, preoperative albumin, and type of surgery (Model 1). These associations remained significant after further adjustment for age and gender (−0.319 [95% CI: 0.125–0.815], P = 0.017; 0.275 [95% CI: 0.094–0.803], P = 0.018, respectively) in Model 2; then adjustment for muscle mass, the result still remained (0.325 [95% CI: 0.127–0.833], P = 0.019; 0.245 [95% CI: 0.077–0.781], P = 0.017, respectively) (Model 3). However, no significant association was found between different obesity patterns and significant postoperative complications after adjustment for all covariates (Table 3).
Then we conducted a stratified analysis to investigate the impact of potential risk factors on the relationship between different obesity patterns and severe postoperative complications (Clavien–Dindo ≥ III), the association between severe postoperative complications and different patterns of obesity were not modified by sex, AJCC stage, albumin, preoperative chemotherapy, age, surgical type, hypertension, diabetes, history of abdominal surgery, and muscle mass (Figs. 2 and 3).
Then we further analyzed the association between subtype of central obesity and severe postoperative complications; we did not observe the difference (Supplementary Table 1, Supplemental digital content 1, http://links.lww.com/EJCP/A509).

Thirty-day readmission
In this study, we did not find the association between different obesity patterns and readmission within 30 days after surgery (Table 4).

Discussion
Gastric cancer ranks as the fifth most malignant tumor globally, with its mortality rate ranking fifth among all malignant tumors (Bray et al., 2024). Meanwhile, the global incidence of obesity is also showing a year-on-year upward trend. Obesity can alter lung, endocrine, and immune functions, adversely affecting health. Some complications of obesity include cardiovascular diseases, non-insulin-dependent diabetes mellitus, obstructive pulmonary disease, arthritis, and cancer (Conway and Rene, 2004). It is well known that obesity is a risk factor for postoperative complications and poor outcomes in malignancies. The main mechanisms by which obesity contributes to carcinogenesis include: (1) inducing a chronic state of inflammation in the body; (2) inducing metabolic-related endocrine changes; (3) production of adipokines, like leptin and adiponectin. It has been demonstrated that IGF1 promotes cancer development in obesity-related tumors through the PI3K/AKT and MAPK pathways (Friedenreich et al., 2021). Zhang et al. have shown that the high expression of hyperinsulinemia in obese individuals can indirectly lead to tumor development (Zhang et al., 2021). Therefore, accurately identifying the degree of obesity becomes very important. However, detailed studies on the state of obesity in relation to short-term outcomes after gastric cancer surgery remain very limited. The aim of this study was to combine central obesity and BMI to form a more detailed obesity patterns, and to investigate its relationship with postoperative complications and 30-day readmission in patients with gastric malignancies. Methods for measuring obesity-related body composition primarily include: weight scales, calipers, high-resolution CT scans, DXA (dual-energy X-ray absorptiometry), and others. Most studies focus on the relationship between the phase angle derived from bioelectrical impedance and cancer patients (Grundmann et al., 2015; Di Vincenzo et al., 2021; Han et al., 2022). There are also studies primarily focusing on using bioelectrical impedance technology to measure muscle mass in the body (Gonzalez and Heymsfield, 2017; Gort-van Dijk et al., 2021; Grossberg et al., 2021). Very few people evaluate the impact of other body composition parameters measured by bioelectrical impedance technology on postoperative outcomes in cancer patients, especially in gastric cancer. Shiomi et al. used this technology to measure back fat area and studied its relationship with postoperative complications in gastric cancer patients (Shiomi et al., 2021). Our study aims to further supplement the applicability of bioelectrical impedance technology in cancer patients. BMI is one of the primary indicators reflecting overall body obesity. Several studies have shown that overweight and obesity are associated with an increased risk and poor prognosis of malignant tumors (Bhaskaran et al., 2014). In research related to gastric cancer, Inagawa et al. reported that a high BMI is an independent risk factor for postoperative complications (Inagawa et al., 2000). While we did not solely focus on BMI as an obesity indicator in our study, we similarly found that, compared to non-obese individuals, a high BMI may potentially be a protective factor against severe postoperative complications. One possible reason could be that the BMI range of the study population falls within the overweight or class I obesity category. Hones et al. also observed a similar phenomenon following total hip arthroplasty (Hones et al., 2023). Simple BMI alone does not seem to reflect the distribution of fat, thus, the studies by Yoshiro Imai et al. and Nahid Hashemi Madani et al. suggested that VFA is a more accurate indicator than BMI for postoperative outcomes in gastric cancer (Yang et al., 2020; Imai et al., 2022). Meanwhile, Patel et al. have pointed out that central obesity, as opposed to BMI alone, more accurately reflects the distribution of body fat and holds higher predictive value for obesity-related tumors (Patel et al., 2023). In the present study, we found that central obesity alone was an independent risk factor for postoperative complications, especially leaks, after adjustment for hypertension, diabetes, history of abdominal surgery, preoperative chemotherapy, preoperative albumin, type of surgery, gender, age, and muscle. A possible reason is that central obesity increases intra-abdominal pressure, thereby becoming a high-risk factor for leakage. We categorized patients based on waist circumference, WHR, and VFA into groups: high WHR only, high WHR + visceral fat, and high WHR + high visceral fat + high waist circumference. However, we did not find any differences between these subtypes. To our knowledge, a previous prospective cohort study from the UK Biobank examined the relationship between six obesity metrics and the risk of 24 types of cancer, but there have been no studies reporting outcomes related to postoperative complications (Parra-Soto et al., 2021). Although Sloan et al. found that central obesity is a risk factor for 30-day readmission after surgery (Sloan et al., 2020), our study did not find any differences in 30-day readmission rates among the different obesity patterns. Our study had several limitations. First, due to the nature of observational research, we cannot control for all factors that might influence the results, so our conclusions may lack randomness. Second, factors affecting short-term postoperative outcomes, such as smoking, alcohol consumption, and perioperative diet, were not included in this study. Finally, as this is a single-center, small sample, retrospective study, it is inevitably subject to selection and statistical biases.

Conclusion
For cancer patients, simple obesity assessment models are insufficient. Individuals with central obesity alone may be at the highest risk for severe postoperative complications. Slightly high BMI may be beneficial for severe complications, therefore, we speculate that the fundamental cause of the obesity paradox is BMI rather than central fat. Therefore, the term ‘BMI paradox’ seems to be a more accurate expression.

Acknowledgements
We are grateful to everyone who contributed to this article.
This study was supported by The Second Hospital of Lanzhou University ‘Cuiying Scientific Innovation’ Program Project: Clinical Study on Microbiota Transplantation in the Treatment of Advanced Colorectal Cancer (CY2023-BJ-06).
J.M. involved in conceptualization; J.M. and P.D. involved in methodology; J.M. involved in validation; G.S. and T.G. involved in formal analysis; G.S. involved in investigation; J.M. involved in resources; G.S. involved in data curation; G.S. involved in writing—original draft preparation; G.S. involved in writing—review and editing; G.S., T.G., P.D., J.G., and Y.D. involved in visualization; G.S. and T.G. involved in supervision; J.M. involved in project administration and funding acquisition. All authors have read and agreed to the published version of the manuscript.
Due to this study being a retrospective study and the privacy and personal identity information of the patients be protected, the need for informed consent was waived by the Medical Ethics Committee of the second Hospital of Lanzhou University (Project Number: 2024A-697).
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of interest
There are no conflicts of interest.

Supplementary Material