生物与医学2019ncbi.nlm.nih.gov
饮用深焙咖啡混合物可减少人体DNA损伤
Consumption of a dark roast coffee blend reduces DNA damage in humans
进阶 · 约 25 分钟读完前置知识：生物医学统计、DNA 损伤检测、临床试验设计
导读
这篇论文研究饮用一种深烘焙拼配咖啡是否能减少人体DNA损伤，并报告在血液样本相关指标上观察到DNA损伤下降（评论提到 p=0.028）。但研究规模小、由咖啡公司 Tchibo 资助，结论稳健性受到质疑。
HN 关注点：是否只是小样本、p 值可疑且受行业资助的“咖啡有益”研究，以及这类营养干预证据的可信度。
HN 评论精选
这几条评论里：质疑 4
质疑指出研究由咖啡零售商资助，存在利益冲突和商业宣传动机。
资金说明：本研究由德国汉堡的 Tchibo GmbH 资助。Tchibo 是德国连锁咖啡零售商和咖啡馆，以每周更换的非咖啡类商品闻名。它有 700 家德国门店、300 家海外门店。他们可能想借此在店内宣传深烘焙咖啡……
1_over_n查看英文原文
"Funding This study has been supported by Tchibo GmbH, Hamburg, Germany." Tchibo is a German chain of coffee retailers and cafés known for its range of non-coffee products that change weekly. https://en.wikipedia.org/wiki/Tchibo Number of locations 700 shops (Germany) 300 (rest of world) They might want to make some in-store claims about Dark Roast...
质疑批评小样本、设计不佳的试验泛滥，并以自嘲调侃结论。
如果我们每 10 项这样的试验就有 1 项诚实的 meta 分析，我们的知识会比现在多。设计不一致或糟糕的小试验是医学科学的祸害。话虽如此，我喝深烘焙咖啡，所以这项研究肯定是真的。
dannykwells查看英文原文
If we had even 1 honest meta-analysis for every 10 trials like this, well, we'd know more than we do now. Small trials with inconsistent/poor designs are the bane of medical science. That said, I drink dark roast so this study must be true.
质疑质疑多重比较和 p 值，追问是否预注册以及是否只测了 DNA 损伤。
我几乎每天喝咖啡；此刻就在咖啡馆读这个。但 p=0.028？拜托。如果你测试几十个指标，至少会有一个得到这样的 p 值。这项研究是否预注册为只通过这个指标看 DNA 损伤？还是他们测了 100 种东西才找到一个 p=0.028？当然，如果预注册且理论就是如此，那会不同，但从文章很难判断。现在回去喝我的摩卡拿铁。
rossdavidh查看英文原文
I drink coffee nearly every day; I am reading this in a coffeeshop. But, p = 0.028? Come on. If you tested for a few dozen things, you would expect to find at least one that had a p-value like that. Was this pre-registered to look at only DNA damage, via this metric? Or did they look at 100 different things to find one with p=0.028? Of course, if they pre-registered and their theory was that it would reduce DNA damage as measured in this way, that would be a bit different, but from this article it's hard to say. Now, back to drinking my mocha latte.
质疑质疑作者资质、开放获取和方法，认为高方差或 p-hacking 可解释结果。
我讨厌糟糕的科学，这就是例子。Springer Science+Business Media（出版方）不像许多其他期刊那样要求作者把版权转让给出版方。那为什么这篇文章不是开放获取？第一作者 Dorothea Schipp 是德国小镇的统计顾问，不在大学，血液工作在斯洛伐克完成，团队中没有博士或医学博士。深入其研究方法：研究在四个月里比较咖啡组和非咖啡/咖啡因组的血液样本 DNA 断裂。我有几个问题，主要是时间段和测定总 DNA 稳定性的方法是否有效：前后测了多少单个细胞基因组？同一身体内细胞基因组变异的随机方差多大？我怀疑方差高到足以解释他们低得难以置信的 p。我猜统计师在玩 p-hacking，同时从行业拿钱。
max76查看英文原文
I really hate bad science, and this is an example. Springer Science+Business Media (the publisher) does not transfer copyright of the paper from the author to the publisher like many other journals. This begs the question, why isn't this article open access? The primary author, Dorothea Schipp, is a statistics consultant in small town Germany [0] not associated with a University and the blood work was done in Slovakia. No PHDs or MDs among them. [1]. Digging deeper into their research method [2] the study looked at DNA breakage over a four month period in blood sampling for a coffee and non-coffee/caffeine group. I have a few questions about this, mainly around how effective the time period and method for determining total DNA stability. How many individual cell gnomes were measured in the before/after? What's the random variance for genome variability among cells in the same body? I suspect the variance is high enough to explain their unbelievably low p. My guess is that the statistician is having fun with p-hacking [3] while collecting some funding from industry. [0] https://en.wikipedia.org/wiki/Rosenthal-Bielatal [1]https://link.springer.com/article/10.1007/s00394-018-1863-2 [2] https://www.mdlinx.com/journal-summaries/coffee-comet-assay-human-intervention-study/2018/11/20/7549520/ [3] http://psychology.wikia.com/wiki/P-hacking
相关资源
从这篇论文所有 HN 讨论帖的全部评论里整理出来的链接。
历次讨论
- 2019Hacker News 讨论帖266 分 · 160 评
同领域还值得读
CRISPR 在活体动物中消除 HIV-1 感染
CRISPR eliminates HIV-1 infection in live animals
报告利用 CRISPR/Cas9 基因编辑技术在活体动物中清除 HIV-1 感染：研究人员通过腺病毒载体递送特定的 Cas9 系统，从被感染细胞中切除嵌入的 HIV 病毒 DNA。
生物与医学2017 · 1,455 分 · 239 评入门 · 约 30 分钟编辑精选保护区飞行昆虫生物量27年下降75%
75% decline over 27 years in total flying insect biomass in protected areas
这篇发表于 PLOS ONE 的研究通过在德国保护区长达 27 年的昆虫生物量监测，发现飞行昆虫总量下降了约 75%。
生物与医学2017 · 901 分 · 432 评入门 · 约 45 分钟编辑精选为什么大多数已发表的研究发现是错的
Why Most Published Research Findings Are False (2005)
这篇 2005 年发表于 PLOS Medicine 的经典论文由 John Ioannidis 撰写，论证了在统计功效不足、研究偏倚和多重检验等因素下，大多数已发表的医学研究发…
生物与医学2022 · 388 分 · 351 评入门 · 约 60 分钟编辑精选人类免疫系统的物理接线图
A physical wiring diagram for the human immune system
这篇 Nature 论文系统绘制了人类免疫系统的“物理接线图”：通过大规模实验鉴定免疫细胞表面受体之间的相互作用（包括约 30 种此前未知的受体互作），并结合定量计算模型模拟免疫…
生物与医学2022 · 342 分 · 49 评专业 · 约 90 分钟编辑精选脂肪组织在减重后仍保留肥胖的表观遗传记忆
Adipose tissue retains an epigenetic memory of obesity after weight loss
这篇 Nature 论文研究了肥胖对脂肪组织的长期影响：即使减重成功，脂肪细胞仍保留着肥胖时期的表观遗传记忆，可能解释为什么减重后容易反弹。
生物与医学2025 · 263 分 · 327 评进阶 · 约 45 分钟编辑精选