A review of 218 trials found that exercise eases depression, with walking and jogging ahead of the rest — and then graded its own certainty as low, for reasons that say a lot about how we test treatments patients must perform themselves
【1】Ask a psychiatrist what to do about a patient who cannot sleep, cannot concentrate and no longer enjoys anything, and the answer will usually contain a sentence that sounds almost too simple to matter: get some exercise.The advice is free, easy to follow and, according to a growing stack of trials, genuinely effective — which is exactly why it deserves a closer look.
【2】Clinical guidelines in Britain, Australia and the United States already recommend physical activity as part of treatment for depression.None of them, however, has ever agreed on what should be prescribed, how often, or at what intensity, which leaves doctors improvising and patients guessing.
【3】A network meta-analysis published in the British Medical Journal in February 2024 set out to settle the question.Michael Noetel, a psychologist at the University of Queensland, and seventeen colleagues pooled 218 randomised trials, containing 495 treatment arms and 14,170 participants who met the clinical threshold for major depression, and weighed every form of exercise against psychotherapy, antidepressants, usual care and placebo tablets.
【4】Walking or jogging came out strongest, with an effect size of −0.62 against active controls, followed by yoga at −0.55, strength training at −0.49, mixed aerobic exercise at −0.43 and tai chi or qigong at −0.42.More revealing than the ranking was the pattern beneath it: across almost every modality the benefit rose with the intensity prescribed, as though the body were answering a dose.
【5】Read plainly, those numbers put exercise in the same league as the treatments it is usually offered alongside.Yoga and strength training also proved the easiest to keep up, a detail that matters rather than decorates, because a therapy abandoned by half its participants delivers very little.
【6】Then comes the sentence that most coverage of the paper left out.The authors graded the certainty of their own findings with a standard instrument, and it came out low for walking or jogging and very low for everything else, because only one of the 218 trials met Cochrane's criteria for a low risk of bias.
【7】The obstacle is structural rather than careless.A drug trial can hand one group a sugar pill and keep both patients and assessors in the dark, whereas nobody can be blinded to the act of jogging, and volunteers for such studies almost always arrive hoping it will work.
【8】Expectancy of that kind inflates every self-reported outcome, and it cannot easily be removed when the treatment is something a patient must do rather than swallow.Nor is it obvious what a fair comparison would look like, since the control condition in many of these trials was ordinary care rather than an equally demanding alternative.What the field calls evidence, in other words, is a stack of trials that could not have been designed any other way.
【9】None of this makes the advice wrong; it makes it unfinished.Exercise appears to be a real treatment for depression whose dose is still unknown, and the honest message for a patient lies somewhere between the triumphant headline and the cynicaldismissal: this is worth trying, and we are still learning how much of it works.
原文出处：本文据 The BMJ 2024 年 2 月 14 日刊论文《Effect of exercise for depression: systematic review and network meta-analysis of randomised controlled trials》（Michael Noetel, Taren Sanders, Daniel Gallardo-Gómez, Paul Taylor, Borja del Pozo Cruz 等；University of Queensland / Australian Catholic University 等；BMJ 2024;384:e075847；DOI 10.1136/bmj-2023-075847）公开摘要与作者结论改写整理，供学习使用。
三、长难句分析（三步拆解法）
Michael Noetel, a psychologist at the University of Queensland, and seventeen colleagues pooled 218 randomised trials, containing 495 treatment arms and 14,170 participants who met the clinical threshold for major depression, and weighed every form of exercise against psychotherapy, antidepressants, usual care and placebo tablets.
主干：Michael Noetel ... and seventeen colleagues pooled 218 randomised trials ... and weighed every form of exercise against ...（两个并列谓语 pooled / weighed）
同位语：a psychologist at the University of Queensland（说明第一主语身份，插在主语与谓语之间，阅读时先跳过）
分词短语：containing 495 treatment arms and 14,170 participants（现在分词作后置定语，修饰 trials；内含 who 引导的定语从句修饰 participants）
搭配：weigh A against B"把 A 与 B 对照权衡"；arm 在此为"治疗组"，非"手臂"
Walking or jogging came out strongest, with an effect size of −0.62 against active controls, followed by yoga at −0.55, strength training at −0.49, mixed aerobic exercise at −0.43 and tai chi or qigong at −0.42.
主干：Walking or jogging came out strongest（主句；come out + 形容词，表示"结果是……、排名……"）
with 复合结构：with an effect size of −0.62 against active controls（补充说明"最强"的具体数值与比较基准）
过去分词短语：followed by yoga at −0.55 ...（"其后依次是……"，是学术写作列举排序的固定套路，at 后接数值）
列举格式：A, B, C, D and E 五项并列，注意每项后都用 at + 数字保持结构对称
考点：come out + adj.；with 复合结构；followed by 表顺序；并列列举的结构对称
The authors graded the certainty of their own findings with a standard instrument, and it came out low for walking or jogging and very low for everything else, because only one of the 218 trials met Cochrane's criteria for a low risk of bias.
主干：The authors graded the certainty ... and it came out low ...（并列分句，it 指代的是"确定性评级结果"而不是 findings）
原因状语从句：because only one of the 218 trials met Cochrane's criteria（说明"为何只有低"——这是全篇的转折支点）
介词短语作状语：for a low risk of bias（修饰 criteria，"达到……的标准"；risk of bias 是循证医学固定术语）
语气提示：their own findings（"他们自己的"）暗示作者主动给自己降级，是学术诚实的标志，也是本文论点的关键
A drug trial can hand one group a sugar pill and keep both patients and assessors in the dark, whereas nobody can be blinded to the act of jogging, and volunteers for such studies almost always arrive hoping it will work.
主干：A drug trial can hand ... and keep ..., whereas nobody can be blinded ..., and volunteers ... arrive ...（whereas 连接前后两个对比分句）
双宾结构：hand one group a sugar pill（hand sb sth 给某人某物；one group 为"一组人"）
对比连词：whereas"而、然而"，用于引出与前半句形成结构对照的事实，比 but 更书面
分词作伴随状语：hoping it will work（说明 volunteers 到来时的心理状态；it 指"运动这件事"）
考点：hand sb sth 双宾；keep sb in the dark；whereas 对比；be blinded to；现在分词作伴随状语
Nor is it obvious what a fair comparison would look like, since the control condition in many of these trials was ordinary care rather than an equally demanding alternative.
主干：Nor is it obvious what a fair comparison would look like（nor 置于句首引起完全倒装，形式主语 it，真正主语是 what 从句）
倒装还原：It is not obvious either what ... would look like（"……也并不显然"）
原因状语从句：since the control condition ... was ordinary care rather than an equally demanding alternative（since 表"因为"，语气比 because 轻，常引出双方已知的理由）
rather than：连接两个名词短语，表示"而不是"，此处对比"常规照护"与"同等费力的替代活动"——点出对照设计的先天不足
考点：nor 引起的倒装；形式主语 it + 主语从句；since 原因从句；rather than 的平行结构
