In Reply We have carefully reviewed the letter from Wu and Leung regarding our recent JAMA Oncology article on fiber and yogurt consumption and lung cancer risk.¹ We believe that there were misunderstandings and misinterpretations of the study design and results.

Their main concern about the study conclusion was related to P values we presented in the Supplement (P = .059 in eFigure 1 and P = .055 in eFigure 2).¹ Wu and Leung considered these P values as evidence of statistical nonsignificance for overall findings. In fact, these P values represented significance of heterogeneity tests, indicating no statistically significant heterogeneities in lung cancer risk estimates associated with fiber and yogurt intakes across studies. The overall summary estimates for association (hazard ratios), presented in eFigures 1 and 2, were 0.83 (95% CI, 0.76-0.91) for fiber intake and 0.81 (95% CI, 0.76-0.87) for yogurt intake, which were statistically significant.¹

在答复中，我们仔细阅读了Wu和Leung关于我们最近的JAMA肿瘤学文章有关纤维和酸奶消费以及肺癌风险的信。¹我们认为研究设计和结果存在误解和误解。

他们对研究结论的主要关注与我们在补编中介绍的P值有关（ e图1中的P = .059，e图2中的P  = .055）。¹ Wu和Leung认为这些P值对总体发现没有统计学意义。实际上，这些P值代表异质性测试的重要性，表明在整个研究中，与纤维和酸奶摄入量相关的肺癌风险估计值在统计学上没有显着异质性。如图1和图2所示，纤维摄入量的关联性总概算（危害比）为0.83（95％CI，0.76-0.91），酸奶摄入量为0.81（95％CI，0.76-0.87），这在统计上是重大。^1个