PO.PR01.04 · 预防研究
肥胖与癌症：纳入迄今最大证据基础的最新荟萃分析
Adiposity and cancer: Updated meta-analysis incorporating the largest evidence base to date
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景：全球范围内体重指数（BMI）的上升使肥胖成为世界性的重大公共卫生挑战。在2016年至2018年间，WCRF和IARC开展的里程碑式综述确立了BMI升高是至少12种癌症的危险因素，但在以下方面仍存在不确定性：其在欧洲和北美以外地区的普遍适用性、与较少见癌症或吸烟相关癌症的关联，以及腰围等其他肥胖测量指标的相关性。自这些综述发表以来，新的大规模队列研究（尤其是在东亚地区）和孟德尔随机化研究极大地扩展了现有证据，使得重新评估肥胖与癌症的关联成为可能。
方法：我们对前瞻性队列研究进行了系统综述和荟萃分析，考察BMI与25种常见癌症风险的关系，检索了PubMed、EMBASE和Scopus截至2025年4月23日的文献。采用随机效应荟萃分析计算汇总风险估计值。我们还对孟德尔随机化研究进行了荟萃分析，比较了BMI与腰围的关联，并综述了新兴的基于影像学的证据。
结果：在包含150万个癌症数据点的226篇文献中，大多数病例来自欧洲（59.0%），其次是东亚（30.1%）和北美（10.6%）。BMI与19种癌症类型呈正相关，与3种呈负相关。白血病（每5 kg/m²的RR = 1.09，95% CI 1.05-1.13）、非霍奇金淋巴瘤（1.05，1.01-1.08）、从不吸烟者中的膀胱癌（1.04，1.01-1.07）和胶质瘤（1.03，1.01-1.05）的正相关是既往WCRF或IARC综述中未曾报告的新发现，提示肥胖可能与比以往认识到的更广泛的癌症相关。对于吸烟为主要危险因素的癌症，我们观察到头颈癌（1.07，1.02-1.13）和膀胱癌风险升高，而在从不吸烟者中食管鳞状细胞癌（0.72，0.60-0.86）和肺癌（0.92，0.86-0.99）风险降低。区域差异明显，包括东亚地区绝经后乳腺癌和卵巢癌的关联更强，胆囊癌的关联较弱，尽管许多人群仍未被代表。孟德尔随机化分析总体上与因果效应一致。BMI与腰围的关联高度相似，可用的影像学数据有限。
结论：这项最新的荟萃分析代表了迄今最大规模的证据综合，凸显了肥胖对癌症风险的重大且不断扩大的影响，并强调了在全球范围内应对这一可改变危险因素的迫切需求。
查看英文原文 English abstract
Background: Global increases in body mass index (BMI) have made obesity a major public health challenge worldwide. Between 2016 and 2018, landmark reviews by the WCRF and IARC established elevated BMI as a risk factor for at least 12 cancers, yet uncertainties remain regarding generalizability beyond Europe and North America, associations with less common or smoking-related cancers, and the relevance of other adiposity measures such as waist circumference. Since these reviews, new large-scale cohorts, particularly in East Asia, and Mendelian randomization studies have greatly expanded available evidence, enabling reassessment of obesity-cancer associations.
Methods: We conducted a systematic review and meta-analysis of prospective cohorts examining BMI and risk of 25 common cancers, searching PubMed, EMBASE, and Scopus through April 23, 2025. Summary risk estimates were calculated using random-effects meta-analysis. We also meta-analysed Mendelian randomization studies, compared associations for BMI and waist circumference, and reviewed emerging imaging-based evidence.
Results: Across 226 articles comprising 1.5 million cancer data points, most cases originated from Europe (59.0%), followed by East Asia (30.1%) and North America (10.6%). BMI was positively associated with 19 cancer types and inversely associated with three. Positive associations for leukaemia (RR per 5 kg/m 2 = 1.09, 95% CI 1.05-1.13), non-Hodgkin lymphoma (1.05, 1.01-1.08), bladder cancer among never-smokers (1.04, 1.01-1.07), and glioma (1.03, 1.01-1.05) represent novel findings not reported in previous WCRF or IARC reviews, suggesting obesity may be linked to a broader range of cancers than previously recognised. For cancers where smoking is a dominant risk factor, we observed increased risks for head and neck cancer (1.07, 1.02-1.13) and bladder cancer, and decreased risks for oesophageal squamous cell carcinoma (0.72, 0.60-0.86) and lung cancer (0.92, 0.86-0.99) among never-smokers. Regional variation was evident, including stronger associations for postmenopausal breast and ovarian cancers in East Asia and weaker associations for gallbladder cancer, although many populations remain unrepresented. Mendelian randomization analyses were broadly consistent with a causal effect. Associations for BMI and waist circumference were highly similar, and available imaging data were limited.
Conclusions: This updated meta-analysis, representing the largest synthesis of evidence to date, highlights the substantial and expanding impact of obesity on cancer risk and underscores the urgent need to address this modifiable risk factor worldwide.
利益披露 Disclosure
E. Watts, None..
A. Gonzalez-Feliciano, None..
S. Moore, None.