We thank Juan M Hincapie-Castillo and Philip F Daniels for their comments on our manuscript. They suggest that the positive outcome associated with lithium treatment might be partly attributable to patients who had bipolar disorder instead of unipolar depression. We discussed this issue in our original version of the manuscript as follows: “However, in our bipolar disorder cohort (n=18 018 for total cohort, n=2074 for incident cohort), the effect sizes for lithium were markedly weaker (hazard ratio >0·65) (Lähteenvuo and colleagues, unpublished), which suggests that, if anything, false negative bipolar patients included in the unipolar cohort have diluted the effect sizes for lithium.” This text was omitted from the proof because of journal policy, which specifies unpublished references cannot be cited.

中文翻译：

---

单相抑郁症患者中锂的使用-作者的回复

我们感谢Juan M Hincapie-Castillo和Philip F Daniels对我们的稿件的评论。他们认为与锂治疗有关的积极结果可能部分归因于患有双相情感障碍而不是单相抑郁的患者。我们在原始手稿中讨论了这个问题，如下所示：“但是，在我们的躁郁症队列中（总队列n = 18 018，入射队列n = 2074），锂的影响大小明显变弱（危险比> 0·65）（Lähteenvuo及其同事，未发表），这表明，单极队列中包括假阴性双极患者（如果有的话）已经稀释了锂的效应量。” 由于期刊政策，该文本从证明中被省略，该政策规定了未引用的参考文献不能被引用。