Lung cancer, as a malignant tumor with both high incidence and high mortality rates, is a leading cause of cancer-related deaths and imposes a significant burden on healthcare expenditure[25, 26]. Therefore, early screening of high-risk individuals for lung cancer and the development of effective strategies for prevention cannot be ignored. Previous research has explored the impact of behavioral risk factors on cancer incidence, including tobacco use, diet, physical activity, obesity, and energy balance, and effective interventions have been devised[27, 28]. Lung tumors exhibit distinct metabolic alterations compared to normal tissue, which provide a continuous energy supply for uncontrolled proliferation. As our understanding of metabolic gene expression and metabolomics deepens, it becomes apparent that modulating tumor cell metabolism might effectively inhibit tumor progression[29–32].
Earlier studies have found increased resting energy expenditure in certain subgroups of cancer, such as lung cancer and colon cancer. In these subgroups of cancer patients, there may be an absolute increase in metabolic rate. However, due to study design limitations, a definitive causal relationship between resting energy metabolism and cancer cannot be inferred [33, 34]. However, recent prospective investigations provide evidence that higher levels of BMR are positively associated with an increased risk of esophageal adenocarcinoma, colon cancer, postmenopausal breast cancer, endometrial cancer, and pancreatic cancer. This suggests that higher BMR may be an indicator of increased risk for certain cancer types and may serve as an independent predictor of cancer risk beyond obesity[15]. Given the limitations of the study design, the relationship between BMR levels and lung cancer risk is currently unknown. Additionally, considering that the relationship between BMR levels and lung cancer risk is an epidemiological topic, it is practically infeasible to conduct randomized controlled trials, which hinders the ability to establish a definitive causal association[35]. Thus, we included two-sample MR studies as evidence in the observational research framework, using genetic variants as instrumental variables for the exposure to improve causal inference in observational studies. This methodology provides a more robust approach to causal inference and is often free from the confounding factors typically associated with observational studies[36, 37].
Our MR study provided robust evidence supporting a potential causal relationship between higher levels of BMR and an increased risk of lung cancer. Further stability tests confirmed the reliability of our results. We also found no evidence of a reverse causal relationship between BMR levels and lung cancer, indicating that our results do not support such a relationship. Since lung cancer encompasses different histological types, directly studying the causal relationship between BMR levels and lung cancer may not provide clear insights into the relationship between BMR levels and different histological types of lung cancer. Therefore, we conducted stratified analyses using GWAS datasets for NSCLC and SCLC, respectively, and re-performed the two-sample MR analysis. Our analyses consistently supported a causal relationship between higher BMR levels and NSCLC. One potential explanation for the observed causal relationship could be through underlying biological mechanisms. Individuals with higher BMR may require more cellular energy production to meet their high energy metabolism demands. This could lead to the promotion of aerobic glycolysis and an increase in the generation of reactive oxygen species (ROS). Normally, ROS levels are well balanced; however, this balance is disrupted when there is an excess of ROS, leading to oxidative stress and the accumulation of potentially carcinogenic DNA defects, as well as the activation of potential carcinogenic signaling pathways[38, 39]. Furthermore, ROS can trigger potentially oncogenic signal transduction cascades, including mitogen-activated protein kinase (MAPK) and epidermal growth factor receptor (EGFR) signaling [40, 41]. Additionally, BMR is a composite function that is related to gender, age, weight, and height, all of which are risk factors for lung cancer. Therefore, BMR may capture some of the partial effects of these variables on cancer[42–44].
This study is the first to assess the causal relationship between BMR and lung cancer at the genetic level. Further stratified analysis using different GWAS datasets and five MR analysis methods evaluated the relationship between BMR and NSCLC and SCLC. All evidence supports a strong causal association between BMR and lung cancer, including NSCLC. However, we found no causal association between BMR and SCLC risk, even though both NSCLC and SCLC datasets are derived from the FinnGen cohort. This might be due to the limited SCLC dataset (GWAS ID: finn-b-C3_SCLC), which includes only 179 SCLC patients. Due to the small proportion of SCLC among all lung cancers and its distinct origin and evolution from NSCLC, as well as the scarcity of SCLC sequencing data, the current GWAS dataset for SCLC is insufficient. Therefore, we cannot further validate this conclusion[45–47]. However, in recent years, studies have proposed subtyping SCLC based on transcription factor dependencies, including SCLC-A, SCLC-N, SCLC-Y, and SCLC-P. It is believed that as more data on SCLC become available, we can clarify the causal relationship between exposure factors and SCLC, and even infer the causal relationship between exposure and different SCLC molecular subtypes[48, 49].
This study is the first bidirectional assessment of the relationship between BMR levels and lung cancer risk using a two-sample MR approach. While previous research has demonstrated a potential association between BMR levels and other cancers, to our knowledge, this is the first study to predict the association between BMR and lung cancer risk[15, 50]. Our findings indicate that high BMR levels are associated with an increased risk of lung cancer, which has important implications for prevention. However, our study has limitations. Firstly, our analysis was limited by the GWAS database, as all datasets were derived from European populations, and therefore, we cannot determine if our conclusions apply to other populations. Secondly, our analysis of the causal relationship between BMR and different histological subtypes of lung cancer was limited by the lack of data on SCLC patients. We cannot confirm if there is truly no causal association between BMR and SCLC, or if the lack of SCLC data led to biased results. Thirdly, because of the GWAS dataset limitations, we did not investigate the reverse causal relationship between BMR levels and lung cancer, even though we confirmed that lung cancer cannot alter BMR levels. Fourthly, while our MR analysis method initially confirmed the causal relationship between high BMR levels and increased lung cancer risk, the exact molecular mechanisms are unclear and require further mechanistic exploration to validate. Lastly, the exposure in this study is determined by SNPs in the human genome, and although they are representative, they cannot fully represent the exposure factors.