A Prospective Cohort Study on Soy Product Intake and the Risk of Lung Cancer Based on Shanghai Suburban Adult Cohort and Biobank

Abstract

肺癌是全世界发病率最高的恶性肿瘤之一，探索肺癌发病的影响因素对其防治具有重要意义。尽管生活习惯和遗传因素已被证实与肺癌的发生密切相关，但饮食因素，如豆制品摄入对肺癌发病风险的影响仍不完全明确。本研究旨在探讨豆制品摄入、遗传风险与肺癌发病风险之间的关联，验证豆制品摄入在欧洲人群中的效应一致性，为肺癌防治提供新的思路。方法 本研究首先基于上海郊区自然人群队列（Shanghai Suburban Adult Cohort and Biobank, SSACB）（n=66,311）使用Cox比例风险回归模型分析豆制品摄入与肺癌发病风险之间的关联，并根据性别、吸烟及肺癌病理类型对人群进行分层分析。采用英国生物银行（UK Biobank, UKB）人群验证豆制品对于肺癌发病的影响。在探究遗传因素与肺癌发病的关联中，在整合既往研究报道的中国人群遗传变异的基础上新增了东南地区两大独立人群的新发现位点进行补充，包括SSACB（433例病例/650例对照）及肿瘤医院-泰州队列（1359例病例/1359例对照）。通过meta分析及连锁不平衡聚类分析（linkage disequilibrium clumping, LD clumping）后，最终筛选出23个显著位点，用于构建多基因风险评分（polygenic risk score, PRS）模型。随后，利用条件Logistic回归模型评估PRS对肺癌发病风险的影响。结果 在SSACB人群中，调整年龄、性别、吸烟、慢性支气管炎、身体质量指数（body mass index, BMI）、蔬菜及红肉摄入后发现，摄入足量豆制品与肺癌发病风险降低有显著关联［风险比（hazard ratio, HR）=0.60，95%CI: 0.47-0.77，Padj=6.69E-05］，且该效应在男性、女性、吸烟和非吸烟人群中均保持一致。在UKB人群中豆制品与肺癌发病的关系虽然没有统计学差异，但也呈现出对肺癌的保护趋势（HR=0.76, 95% CI: 0.55-1.06, Padj=0.10）。在SSACB的巢式病例对照人群中，基于中国人群构建的PRS评分与肺癌发病显著相关，调整年龄、性别、吸烟、慢性支气管炎及豆制品摄入后，与低风险人群相比，高PRS评分人群的发病风险是低评分人群的1.88倍（Padj=1.84E-03）。结论 通过前瞻性队列研究发现，摄入足量豆制品可显著降低肺癌发病风险，而高PRS是肺癌发病的重要危险因素。将豆制品摄入量和PRS作为肺癌发病的影响因素进行综合评估，可为肺癌高危人群的个体化分层和精准预防提供一定指导作用。

Background and objective Lung cancer is one of the malignant cancers with the highest incidence rate, and it is important to identify the factors contributing to lung cancer carcinogenesis for prevention. Lifestyle and genetic factors play important roles in cancer development, however the impact of dietary factors, such as soy product intake, on lung cancer risk remains inadequately understood. This study aims to explore the associations between soy product intake, genetic risk, and lung cancer incidence, and validate the consistent effects of soy product intake in European populations, thereby providing new insights for lung cancer prevention. Methods Utilizing the Shanghai Suburban Adult Cohort and Biobank (SSACB) (n=66,311), Cox proportional hazards model was adopted to assess the association between soy product intake and lung cancer incidents, followed by subgroup analyses stratified by gender, smoking status, and pathological types of lung cancer. The UK Biobank (UKB) was used for validation of the effect of soy product intake on lung cancer. To investigate the association between genetic factors and lung cancer, in addition to previously reported loci, we incorporated newly identified loci from two independent studies in Southeast China: a nested case-control population from the SSACB cohort (433 cases/650 controls) and a case-control study from the Shanghai Cancer Center-Taizhou cohort (1359 cases/1359 controls). Meta-analysis and Linkage disequilibrium clumping (LD clumping) of the association results identified 23 loci for polygenic risk score (PRS) construction. Subsequently, conditional Logistic regression model was used to assess the association between genetic risk and lung cancer. Results In SSACB cohort, after adjusting for age, gender, smoking, chronic bronchitis, body mass index (BMI), vegetable intake and red meat intake, sufficient soy product intake was significantly associated with a reduced risk of lung cancer [hazard ratio (HR)=0.60, 95%CI: 0.47-0.77, Padj=6.69E-05], an effect that was consistent in males and females, smokers and non-smokers. In UKB, although the association did not reach statistical significance, a protective trend against lung cancer was also observed (HR=0.76, 95%CI: 0.55-1.06, Padj=0.10). In the nested case-control population within SSACB, a PRS score generated in the Chinese population was significantly correlated with lung cancer risk. After adjustment of age, gender, smoking, chronic bronchitis, and soy product intake, the high-PRS group had a 1.88 times higher risk of lung cancer compared to the low-PRS group (Padj=1.84E-03). Conclusion The prospective cohort study found that adequate intake of soy products was significantly associated with a reduced risk of lung cancer, while a high PRS is a risk factor for lung cancer development. Integrating soy product intake and PRS into traditional epidemiological risk factor prediction will guide personalized lung cancer prevention and high-risk population stratification.