Huang Yingdan, Zhang Wangjin, Chen Jinghui, Qiu Sihua, Xue Chang, Wu Huijing
Department of Lymphoma Medicine, Hubei Cancer Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, 430079, China.
Department of Ultrasound, Longyan First Affiliated Hospital of Fujian Medical University, Longyan, 364000, China.
Discov Oncol. 2024 Nov 2;15(1):612. doi: 10.1007/s12672-024-01448-5.
In prior observational investigations, it has been demonstrated that the consumption of milk is associated with the incidence of breast cancer (BC). Despite the existence of a two-sample Mendelian randomization (MR) that suggests a causal relationship between milk intake and breast cancer risk, the outcomes still lack a definitive conclusion. This ambiguity may be attributed to variables such as the variety of milk ingested, estrogen levels, the specific type of BC, and potential confounding factors. Therefore, our principal objective is to establish the causal association between the consumption of skimmed milk and full cream milk and the risk of different types of BC through the utilization of two-sample and two-step MR analyses.
In this study, we analyzed single nucleotide polymorphisms associated with skimmed and full-cream milk consumption in a cohort of 360,806 individuals from European populations through genome-wide association studies. We conducted a two-sample MR analysis using three different methods: inverse variance weighting (IVW) was used as the main analysis, and MR-Egger and Weighted median were used as supplementary analyses to IVW. We also performed sensitivity analysis, which included leave-one-out analysis, Cochran's Q test to detect heterogeneity, and MR-Egger intercept analysis to detect potential biases caused by pleiotropy. We used two-step MR analysis to evaluate potential mediators of associations.
In the two-sample MR analysis, IVW analysis suggests a potential inverse causal relationship between skimmed milk and BC [OR 0.34, 95% CI (0.12-1.00), P = 0.05]. Subgroup analysis revealed that skimmed milk reduces the risk of estrogen receptor-negative (ER-) breast cancer [OR 0.18, 95% CI (0.04-0.90), P = 0.04], but not estrogen receptor-positive (ER+) breast cancer [OR 0.42, 95% CI (0.15-1.22), P = 0.11]. MR Egger reached similar results, that is, skimmed milk reduces the risk of ER- breast cancer [OR 0.006, 95% CI (0.00-0.70), P = 0.04], but not BC [OR 0.16, 95% CI (0.01-4.66), P = 0.30] and ER+ breast cancer [OR 0.50, 95% CI (0.02-12.61), P = 0.65]. Additionally, we found no causal relationship between full cream milk and BC (P > 0.05). In two-step MR analysis, we found evidence for a mediating role of BMI in the relationship between skimmed milk intake and ER-breast cancer risk.
Our findings strengthen the evidence for a protective effect of skimmed milk consumption on ER-breast cancer risk. Further two-step MR analyses suggest that this protective effect may partly result from body mass index (BMI). There is no evidence that full cream milk consumption affects the risk of BC.
在先前的观察性研究中，已证明牛奶消费与乳腺癌（BC）的发病率相关。尽管存在一项双样本孟德尔随机化（MR）研究表明牛奶摄入量与乳腺癌风险之间存在因果关系，但其结果仍缺乏明确结论。这种不确定性可能归因于多种变量，如摄入的牛奶种类、雌激素水平、BC的具体类型以及潜在的混杂因素。因此，我们的主要目标是通过使用双样本和两步MR分析，确定脱脂牛奶和全脂牛奶消费与不同类型BC风险之间的因果关联。
在本研究中，我们通过全基因组关联研究分析了来自欧洲人群的360,806名个体队列中与脱脂和全脂牛奶消费相关的单核苷酸多态性。我们使用三种不同方法进行双样本MR分析：逆方差加权（IVW）用作主要分析，MR-Egger和加权中位数用作IVW的补充分析。我们还进行了敏感性分析，包括留一法分析、用于检测异质性的Cochran's Q检验以及用于检测多效性引起的潜在偏差的MR-Egger截距分析。我们使用两步MR分析来评估关联的潜在中介因素。
在双样本MR分析中，IVW分析表明脱脂牛奶与BC之间可能存在反向因果关系[比值比（OR）0.34，95%置信区间（CI）（0.12 - 1.00），P = 0.05]。亚组分析显示，脱脂牛奶可降低雌激素受体阴性（ER-）乳腺癌的风险[OR 0.18，95% CI（0.04 - 0.90），P = 0.04]，但对雌激素受体阳性（ER+）乳腺癌无影响[OR 0.42，95% CI（0.15 - 1.22），P = 0.11]。MR-Egger得出了类似的结果，即脱脂牛奶可降低ER-乳腺癌的风险[OR 0.006，95% CI（0.00 - 0.70），P = 0.04]，但对BC[OR 0.16，95% CI（0.01 - 4.66），P = 0.30]和ER+乳腺癌[OR 0.50，95% CI（0.02 - 12.61），P = 0.65]无影响。此外，我们发现全脂牛奶与BC之间无因果关系（P > 0.05）。在两步MR分析中，我们发现有证据表明体重指数（BMI）在脱脂牛奶摄入量与ER-乳腺癌风险之间的关系中起中介作用。
我们的研究结果进一步证明了饮用脱脂牛奶对ER-乳腺癌风险具有保护作用。进一步的两步MR分析表明，这种保护作用可能部分归因于体重指数（BMI）。没有证据表明饮用全脂牛奶会影响BC风险。
