With the increasing number of patients with a high body mass index (BMI) worldwide, more attention is being focused on the clinical outcomes of patients with different BMI during the perioperative period1,2. The past few decades have witnessed the rapid development of noncardiac thoracic surgery and enhanced recovery after surgery (ERAS)3,4. However, early postoperative clinical outcomes, including postoperative nausea and vomiting (PONV), severe pain, delirium, and dizziness, lead to patient dissatisfaction and delayed hospital discharge5,6,7,8. Previous studies have shown a protective effect of a high BMI on the inflammatory response9,10. Early postoperative clinical outcomes are closely related to inflammatory response, and postoperative serum C-reactive protein (CRP), IL-6, IL-12, and IFN-γ levels can reflect the postoperative inflammatory status of patients. Studies have shown that the application of exogenous IL-12 in patients with cancer can stimulate IFN-γ production, along with an increased incidence of nausea and vomiting11,12.

It is estimated that there are 2.2 million new cases and 1.79 million deaths each year, lung cancer is one of the most commonly diagnosed cancers and the leading cause of cancer-related deaths worldwide13. Surgery is the primary treatment for patients with early stage lung cancer. With the development of minimally invasive technology, thoracoscopic surgery has become the most common surgical method in thoracic surgery because of its advantages of less trauma, faster recovery, less postoperative pain, and improved quality of life compared with thoracotomy. However, PONV remains an unsolved problem that can increase medical costs and affect the postoperative recovery. Studies have pointed out that the incidence of PONV within 48 h after thoracoscopic examination is 68.42–73.68%, and 57.89% of patients require drug intervention14. This study aimed to evaluate whether PONV following thoracic surgery for lung cancer correlates with different BMI values, and to explore whether serum inflammatory factors play a role in the early recovery of patients with different BMI.


Ethical approval/informed consent

This study was approved and the requirement for written informed consent was waived by the Medical Ethics Committee of Shanghai Pulmonary Hospital (#L21-307). This study conformed to the principles outlined in the Declaration of Helsinki.

Study participants

This retrospective cohort analysis included all patients aged 18–80 years who underwent thoracic surgery for lung cancer between January 2021 and June 2021 at Shanghai Pulmonary Hospital. This study was approved by the Shanghai Pulmonary Hospital Medical Ethics Committee, which waived the requirement for written informed consent. Patients were excluded if they underwent emergency surgery, had chronic pain, or had missing clinical information (e.g., BMI and postoperative pain scores). This study adhered to the STROBE guidelines.

Exposure and outcomes

Based on the World Health Organization classification, we defined high BMI as a preoperative BMI ≥ 25 kg/m2 and normal BMI as 18.5–25 kg/m2. Baseline variables, postoperative follow-up data, and laboratory test results were collected from the electronic medical records system of Shanghai Pulmonary Hospital. The primary outcome was the incidence of PONV during the first 48 h after surgery, defined as any episode of nausea, vomiting, or retching15. Nausea is a feeling of upset and discomfort in the stomach, which usually leads to an urge to vomit. Vomiting is the forced discharge of stomach contents into the mouth. Retching is an attempt to vomit, usually accompanied by dry heaves and gagging, spasmodic, and rhythmic contraction of the ventilator, without discharging stomach contents.

The secondary outcomes included nausea, retching, or vomiting during 0–12 h, 12–24 h, 24–36 h and 36–48 h after surgery using the simplified PONV impact scale by Myles et al.16 (Consisting of two questions: Q1. Have you vomited or retched? Q2. Have you experienced nausea? If yes, has your feeling of nausea interfering with activities of daily living, such as being able to get out of bed, being able to move freely in bed, being able to walk normally, or eating and drinking?), use of rescue antiemetics, incidence of severe pain, incidence of dizziness, and concentrations of related inflammatory factors, including CRP, IL-6, IL-12, and IFN-γ, on the first and second days after surgery (routine detection of inflammatory factors in this institution). Postoperative pain was evaluated using the visual analog scale (VAS), with scores ranging from 0 (no pain) to 10 (worst pain imaginable). Severe pain, which included severe static (at rest) and dynamic (at movement) pain, was defined as a VAS score of ≥ 7 points during the postoperative period within 48 h. The incidence of dizziness was defined as dizziness during the postoperative any time within 48 h.

Statistical analysis

Continuous variables conforming to normal distribution were presented as mean ± standard deviation and were analyzed using Student’s t-test. Continuous variables with non-normal distribution were expressed as median (interquartile range [IQR]) and analyzed using the Mann–Whitney U test. Categorical data were expressed as frequencies, and percentages were compared using the chi-square test.

Propensity score matching (PSM) was used to minimize the potential bias between 2 groups. A propensity score was calculated by logistic regression model based on 5 variables: sex, age, ASA, Charlson index and surgery type. A caliper of 0.05 was selected and the nearest neighboring method was adopted to match the high BMI group and the normal group with a 1:1 ratio using SPSS23.

A multivariate logistic regression analysis was used to test the association between BMI and PONV. We corrected the models using the following potential confounding variables: sex, age (≤ 40 and > 40 years), history of smoking, Charlson index, surgery type, history of surgery, mode of resection, and duration of anesthesia. Based on the results of multivariate logistic regression, receiver operating characteristic (ROC) curves and the corresponding area under the curve (AUC) were calculated for BMI and smoking history. All statistical analyses were performed using SPSS version 23 (IBM, Chicago, IL, USA) and R-studio 3.5.2, with significance defined as a two-tailed P < 0.05.


This study enrolled 891 patients aged 18–80 years who underwent thoracic surgery for lung cancer between January 2021 and June 2021 at Shanghai Pulmonary Hospital. After exclusion, 194 patients were eligible for the data analysis. We analyzed the baseline demographic and clinical characteristics using PSM. Before PSM, sex (P = 0.041) and ASA grade (P = 0.011) differed significantly between the two groups. After PSM for sex, age, ASA, age-adjusted Charlson comorbidity index (aCCI), and surgery type, 62 patients were included in the normal BMI group, which matched the 62 patients in the high BMI group (Fig. 1). The two groups were well-balanced for all baseline demographics after PSM (Table 1). The high BMI group had a median BMI of 26.57 kg/m2 with an interquartile range (IQR) of 1.97, whereas the normal BMI group had a median BMI of 22.56 kg/m2 with an IQR of 2.61. Almost all the patients underwent VATS, and there was no difference in the Apfel PONV risk score between the two groups before surgery. The duration of anesthesia for all patients was maintained within 3 h. Supplementary Table 1 provides the intraoperative data with details on the intraoperative anesthetic data.

Figure 1
figure 1

Types and numbers of excluded eligible patients aged between 18 and 80 years who had lung cancer surgeries at Shanghai Pulmonary Hospital between January 2021 and June 2021.

Table 1 Baseline characteristics after propensity score matching.

The higher BMI group had a lower incidence of PONV within 48 h after surgery (22% vs. 50%, p = 0.004). In addition, 43.5% of the 62 patients in the normal BMI group and 14.5% of the patients in the higher BMI group experienced nausea (p ≤ 0.001). The normal BMI group had a higher incidence of vomiting or retching (32.3% vs. 12.9%, p = 0.01). 37.1% in the normal BMI group and 14.5% in the higher BMI group of patients experienced nausea episodes within 12 h after surgery (p = 0.004), while 22.6% in the normal BMI group and 8.1% in the higher BMI group of patients experienced nausea episodes within 12–24 h after surgery (p = 0.025). 30.6% in the normal BMI group and 12.9% in higher BMI group of the patients experienced vomiting or retching episodes within 12 h after surgery (p = 0.017). There was no difference in the incidence of nausea, vomiting, or retching at other time points and in the use of rescue antiemetics.

There was no significant difference in the incidence of severe pain (severe static pain (p = 0.697) and severe dynamic pain (p = 0.158)). However, no significant difference was observed in the incidence of dizziness between the two groups (p = 0.571). Table 2 summarizes the data on postoperative clinical outcomes within the initial postoperative 48 h.

Table 2 Postoperative clinical outcomes.

The concentrations of postoperative serum inflammatory factors in the two groups within the first postoperative day 2 are plotted in Fig. 2. Notably, on the first day after surgery, the concentrations of IL-12 (2.24 ± 2.67 pg/ml vs. 1.48 ± 1.14 pg/ml, p = 0.048) and IFN-γ [1.55(1.00) pg/ml vs. 1.30 (0.89) pg/ml, p = 0.041] were higher in the normal BMI group than in the higher BMI group. On the second day, the concentrations of IL-12 and IFN-γ did not differ between the two groups. And there were no significant differences in the baseline concentrations of serum CRP (3.357 ± 0.68 mg/ml vs. 3.417 ± 0.78 mg/ml), IL-6 [3.50 (6.69) pg/ml vs. 4.13 (4.82) pg/ml], IL-12 (2.26 ± 3.10 pg/ml vs. 1.72 ± 1.23 pg/ml), and IFN-γ [1.64 (1.37) pg/ml vs. 1.34 (0.82) pg/ml] between the groups. There was no difference in the CRP and IL-6 levels between the groups on postoperative day 2.

Figure 2
figure 2figure 2

Inflammatory factors in the normal and high BMI groups on the initial 2 days after surgery. Abbreviations: BMI, body mass index; 1st, first; 2nd, second.

Age, sex, aCCI, duration of anesthesia, surgical history, surgery type, and resection mode were not significantly associated with the incidence of PONV. Smoking history was not significantly associated with the incidence of postoperative nausea; however, the incidence of postoperative vomiting or retching was 79% lower in patients who had smoked previously than in those without a smoking history (OR [odds ratio] = 0.21, 95% CI [0.05–0.95], p = 0.042). Additionally, BMI was associated with the incidence of PONV (Table 3). Multivariate logistic regression was conducted by including clinical variables to identify risk factors for PONV (Table 4). The OR for BMI was consistently significant in all three models, irrespective of whether BMI was analyzed as a continuous or categorical variable (nausea: OR = 0.16, p < 0.05; vomiting/retching: OR = 0.32, p < 0.05). When BMI was evaluated as a continuous variable in the fully adjusted model (model 3) the adjusted OR was 0.8 (95% CI, 0.69–0.94) for nausea and 0.8 (95% CI, 0.67–0.94) for vomiting or retching.

Table 3 Results of univariate analysis of PONV.
Table 4 Multivariate-adjusted ORs and 95% CI for BMI associated with nausea and vomiting.

According to the results of the multivariate analysis, two variables (BMI and smoking history) were selected to construct an ROC curve (Supplementary Fig. 1). The AUC for predicting nausea was 0.710 (95% CI, 0.611–0.809), with a sensitivity and specificity of 66.7 and 72.7%, respectively. Based on the ROC curves, the AUC for vomiting/retching according to BMI and history of smoking was 0.713 (95% CI, 0.614–0.812), with the best sensitivity and specificity of 64.3% and 68.7%, respectively.

Subgroup analyses were performed to assess the impact of BMI (per 1-unit increment) on PONV in distinct subgroups (Supplementary Fig. 2). Associations between BMI and PONV were coordinated in the following subgroups: postoperative nausea, sex (male vs. female; p-interaction = 0.539), smoking status (yes vs. no; p-interaction = 0.375), ASA classification (1 vs. 2; p-interaction = 0.789), history of surgery (yes vs. no; p-interaction = 0.657), postoperative vomiting/retching, sex (male vs. female; p-interaction = 0.619), smoking status (yes vs. no; p-interaction = 0.362), ASA classification (1 vs. 2; p-interaction = 0.978), and history of surgery (yes vs. no; p-interaction = 0.756).


PONV is one of the most common postoperative complications, occurring in approximately 30% of surgical patients and more than 70% of high-risk patients17,18. Our results demonstrated that patients with normal BMI had a higher incidence of PONV. PONV onset is a dynamic process. We observed the occurrence of PONV during different time periods. We found that postoperative nausea mainly occurred within 24 h after surgery, and postoperative vomiting or retching mainly occurred within 12 h after surgery. In addition, transiently elevated concentrations of IL-12 and IFN-γ were observed in the early postoperative period compared with those with higher BMI.

Currently, there is no consensus on the influence of BMI on PONV. Consistent with our results, some studies have revealed that higher BMI is associated with a lower risk of PONV in the early postoperative period19,20,21,22. This may be related to the decreased release of histamine and dopamine, neurotransmitters that mediate nausea and vomiting. During perioperative fasting, the amount of hypothalamic histamine in overweight patients is significantly reduced compared with that in patients with normal or lower body weight23,24. At the same time, the dopamine receptor level in overweight patients is lower than that in patients with normal BMI25,26. In our study, multivariate logistic regression analysis also showed that an increase in BMI reduced the risk of PONV. Univariate logistic regression and ROC curve analysis revealed that the incidence of PONV was lower in patients with a history of smoking than in non-smokers. This finding is consistent with the results of previous studies27,28,29. In Apfel's risk assessment system, the risk factors for PONV include female sex, nonsmoking status, postoperative use of opioids, and history of PONV or motion sickness30. In the univariate analysis in our study, non-smoking status was a possible risk factor for PONV. In addition, we successfully constructed a ROC curve for BMI and smoking history to predict PONV. Our results may identify the likelihood of PONV occurrence following thoracic surgery for lung cancer.

Obese patients have been reported to experience more postoperative pain than those with normal BMI31,32. However, we found no difference in the incidence of severe postoperative pain between the two groups. This may be related to the most popular surgical modality at our medical center. In our study, most surgeries were uniportal video-assisted thoracic surgeries, wherein a single incision was made in the chest wall. Improvement in the surgical technique is an important part of ERAS, which directly reduces the severity of pain in patients following thoracic surgery. In addition, the BMI of patients in our high BMI group rarely exceeded 30 kg/m2, which may explain the similar severe pain scores between the two groups.

There was no difference in CRP and IL-6 concentrations between the two groups. Previous studies have found that PONV changes with postoperative concentrations of CRP and IL-633. However, in the present study, the postoperative serum inflammatory concentrations of CRP and IL-6 showed no difference between the two groups within the first 2 days after surgery. A previous study indicated that IL-12 and IFN-γ levels were increased in people with higher BMI34,35; however, in this study, we found that serum IL-12 and IFN-γ levels were higher in patients with a normal BMI than in those with a high BMI on the first day after surgery. Consistent with our study, Huang et al. reported that the improvement in patients’ quality of life (including reduced nausea, vomiting, and pain) was associated with regulation of the inflammatory cascade comprising IL-12 and IFN-γ33. IL-12 promotes the production of IFN-γ and induces a Th1 response in various disease models36,37,38. More evidence is needed to determine whether patients with different BMI have different Th1 immunomodulatory mechanisms that affect their quality of life or long-term prognosis after surgery.

Our study had some limitations that warrant consideration. First, it is important to acknowledge that this was a retrospective study, which inherently carries the potential for bias. Second, it is pertinent to note that our study sample comprised individuals of the Chinese ethnicity. Thus, extrapolating our findings to other ethnic groups may require additional research for validation and generalization. Third, the sample size of individuals with BMI < 18 kg/m2 and > 30 kg/m2 in our study was limited. While the distribution of the population aligns with the demographic profile of China, caution should be exercised when generalizing the findings.


Patients with a normal BMI were more prone to PONV than those with a higher BMI following thoracic surgery for lung cancer. The elevated concentrations of pro-inflammatory factors, including IL-12 and IFN-γ, in patients with normal BMI indicate that inflammation may be associated with the pathological process of PONV and may affect the prognosis of patients with different BMI.