Evidence level: C
如果”外气时间窗”只在 NK 杀伤这一个终点上出现,它更可能是该测定协议自身的伪影;只有当增殖、凋亡、线粒体膜电位等多个终点在同一 3×2 梯度网格上呈现同构窗口,才谈得上跨终点泛化。本文证据等级为 C:该主题尚未进入 L1 对应网络,外气体外研究整体证据薄弱,且核心文献有撤稿记录。
正文
此前系列文章(见相关阅读)讨论过一个设计:把”治疗时长 × 检测时间”做成 3×2 梯度网格,画出 NK 杀伤活性的”窗口热图”——哪个处理时长、哪个检测时点效应最强、如何衰减。这一篇要问的是下一层问题:这个窗口形状能不能泛化到 NK 杀伤活性之外的细胞终点?
为什么要做这一步?因为单一终点的”窗口”天然可疑。NK 杀伤测定本身对实验条件极其敏感:效靶比、共孵时间、效应细胞状态都会显著改变读数,流式细胞术测定方案把孵育时长作为关键参数明确写进协议(PMID 30465199);冻存与体外扩增后的 NK 细胞,其杀伤读数随检测条件波动很大,需要标准化验证(PMID 35464440)。换句话说,NK 窗口热图里藏着大量”协议自身的时间结构”——如果不做跨终点对照,你无法区分窗口是”外气的作用节律”还是”测定的节律”。
协议设计上,关键是”同批次同步、逐终点独立”。同一批细胞、同一次处理,分到四个读数通道:增殖用 CCK-8 类代谢读数,早期/晚期凋亡用 Annexin V/PI 双染,线粒体膜电位用 TMRE/TMRM 荧光。增殖与细胞毒测定家族的原理与陷阱已有系统综述(PMID 27604355)。每个终点各自跑独立的 3×2 网格,然后参数化比较窗口形状:峰值位置(哪个处理时长×检测时点最强)、衰减斜率(离开峰值后读数掉多快),再用组内相关系数(ICC)或形状相似度给”跨终点一致性”打分(PMID 28904005)。
这里有个容易被忽略的深坑:四个终点并不是彼此独立的观察窗口。CCK-8 读的是线粒体代谢活性,TMRE 读的也是线粒体膜电位——两个”终点”共享同一条生化通路;而能量状态变化会同时拖动代谢读数和凋亡启动。所以如果四个窗口同构,存在一个平庸解释:不是外气在四个终点上同步起作用,而是处理在某个上游环节(比如膜电位)动了一下,下游读数自然一起动。要切开这条捷径,需要做通路中介分析:控制膜电位变化后,增殖/凋亡窗口是否仍然同构。反过来,如果窗口不同构,也不能直接宣布”无效应”——不同构可能恰恰说明 NK 杀伤窗口是 NK 测定协议的伪影,这正是本设计想抓的东西。
判决结构因此是三分支的:① 各终点窗口同构且中介分析后仍同构 → 支持存在一个上游共同靶点,值得追;② 仅 NK 杀伤有窗口、代谢类终点无 → 强烈提示伪影在 NK 测定协议里;③ 全部终点都无稳定窗口(多治疗师重复后 ICC 很低)→ 该现象在该协议下不可重复,管线应降级。无论落在哪支,都比”只报 NK 一个终点”的信息量大。
多治疗师重复是这套设计里最贵也最必要的一环。单治疗师画出的窗口热图,无法区分”外气的时间结构”与”这个人手部状态、节奏习惯甚至体温的季节波动”。最少三到五名治疗师各自独立跑完整 3×2 网格,再对”窗口形状参数”本身算 ICC:峰值位置在治疗师之间是否一致、衰减斜率是否同向。参考通用的一致性判读框架,ICC 低于 0.5 视为差、0.5–0.75 中等、高于 0.75 才算好——如果窗口形状本身跨治疗师不可重复,那么任何”跨终点同构”都只是同一个不可靠测量的两次投影,没有讨论价值。同时处理环节必须盲:操作者不知道哪盘细胞是真处理、哪盘是假处理,培养与读数人员也不知道分组,否则期望效应会顺着操作细节渗进读数。
必须写明的警示:外气体外研究文献的可信度问题严重。严新气功外气(YXQ-EQ)系列曾有报道宣称对培养肿瘤细胞诱导凋亡、抑制迁移(PMID 20110687、30184531),但同系列至少两篇已被正式撤稿(撤稿注记 PMID 33895982、39327376)。本系列引用这两篇仅作为”协议设计参考的历史案例”,不作为外气效应存在的证据。任何后续实验都必须以预注册、盲法处理(治疗师与细胞培养者互盲)、假处理对照为门槛。
证据与局限
已验证: 细胞增殖/毒性测定(CCK-8/MTT 家族)、Annexin V 凋亡检测、TMRE 膜电位检测均为成熟方法,有标准协议与已知陷阱(PMID 27604355、27683123、30465199);ICC 可用于跨终点/跨批次一致性量化(PMID 28904005)。
部分验证: “时间窗口”现象在药理学与 chronopharmacology 中有大量先例——处理与检测的时点是真实可测变量,窗口热图作为描述工具成立。
待验证: 外气处理在多终点上产生同构时间窗——零直接证据;现有外气体外文献核心部分已撤稿(PMID 33895982、39327376),证据基础薄弱。跨终点同构存在”共用通路”平庸解释,须中介分析排除。
相关阅读
L1 网络状态:该主题尚未进入 L1 对应网络。经核验,CORRESPONDENCE_NETWORK.md 与 EDGE_TABLE_science-yixue.md 中无「外气 ↔ NK/增殖/凋亡/线粒体膜电位跨终点窗口」对应边(EDGE_TABLE 仅含「气功/太极」泛节点;「凋亡」命中属 E-060 电针-心肌缺血语境,与本主题无关)。已登记至 science-yixue-l1-gaps.md。
证据等级: C
参考文献: PMID 20110687, PMID 30184531, PMID 33895982, PMID 39327376, PMID 27604355, PMID 27683123, PMID 30465199, PMID 35464440, PMID 28904005
证据等级:A=有PubMed直接支撑,B=科学类比/间接证据,C=传统文化共识/初步探索
免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。
Cross-endpoint generalization testing of NK window heatmaps: Is the window shape isomorphic across cell endpoints (proliferation/apoptosis/mitochondrial membrane potential) unde…
Evidence level: C
If the “external qi time window” appears only on the NK cytotoxicity endpoint, it is more likely an artifact of that specific assay protocol; cross-endpoint generalization is warranted only if proliferation, apoptosis, and mitochondrial membrane potential exhibit isomorphic windows across the same 3×2 gradient grid. This evidence level is C: the topic has not yet entered the L1 corresponding network, overall evidence for external qi studies is weak, and core literature includes retraction records.
Body
Previous articles in this series (see Related Reading) discussed a design that constructs a 3×2 gradient grid of “treatment duration × detection time” to generate NK cytotoxicity “window heatmaps”—identifying which treatment durations and detection timepoints yield the strongest effects, and how these effects decay. This article addresses the next level of inquiry: Can this window shape generalize to cellular endpoints beyond NK cytotoxicity?
Why take this step? Because windows based on a single endpoint are inherently suspect. NK cytotoxicity assays are extremely sensitive to experimental conditions: the effector-to-target ratio, co-incubation time, and the state of effector cells all significantly alter readings. Flow cytometry protocols explicitly list incubation duration as a critical parameter (PMID 30465199). Furthermore, NK cells thawed and expanded ex vivo exhibit large fluctuations in cytotoxicity readings depending on detection conditions, necessitating standardized validation (PMID 35464440). In short, the NK window heatmap conceals substantial “temporal structure inherent to the protocol itself.” Without cross-endpoint controls, one cannot distinguish whether the window reflects “the rhythm of external Qi action” or “the rhythm of the assay.”
Regarding Protocol Design, the key is “synchronous batch, independent endpoints.” The same cell batch, treated once, is distributed across four readout channels: proliferation via CCK-8-like metabolic readings; early/late apoptosis via Annexin V/PI dual staining; mitochondrial membrane potential via TMRE/TMRM fluorescence. Systematic reviews have already detailed the principles and pitfalls of proliferation and cytotoxicity assay families (PMID 27604355). Each endpoint runs its own independent 3×2 grid, followed by parameterized comparison of window shapes: peak position (which treatment duration × detection timepoint is strongest), decay slope (how fast readings drop after the peak), and then scoring “cross-endpoint consistency” using intraclass correlation coefficient (ICC) or shape similarity (PMID 28904005).
A deep pitfall often overlooked: the four endpoints are not independent observation windows. CCK-8 reads mitochondrial metabolic activity, and TMRE also reads mitochondrial membrane potential—these two “endpoints” share the same biochemical pathway. Changes in energy status simultaneously drive metabolic readings and apoptosis initiation. Therefore, if all four windows are isomorphic, a trivial explanation exists: it is not that external Qi acts synchronously on all four endpoints, but rather that the treatment perturbs an upstream step (e.g., membrane potential), causing downstream readings to shift together. To cut this shortcut, pathway mediation analysis is required: after controlling for membrane potential changes, do proliferation/apoptosis windows remain isomorphic? Conversely, if windows are not isomorphic, one cannot immediately declare “no effect”—non-isomorphism might precisely indicate that the NK cytotoxicity window is an artifact of the NK assay protocol, which is exactly what this design aims to capture.
The decision structure is thus three-branched: ① If all endpoint windows are isomorphic and remain so after mediation analysis → supports the existence of an upstream common target, worth pursuing; ② If only NK cytotoxicity shows a window while metabolic endpoints do not → strongly suggests the artifact lies within the NK assay protocol; ③ If all endpoints lack stable windows (ICC remains low after repetition by multiple therapists) → this phenomenon is irreproducible under this protocol, and the pipeline should be downgraded. Regardless of which branch is hit, the information content exceeds reporting only the NK endpoint.
Repetition by multiple therapists is the most expensive yet essential component of this design. A window heatmap drawn by a single therapist cannot distinguish “the temporal structure of external Qi” from “individual hand state, rhythm habits, or even seasonal body temperature fluctuations.” At least three to five therapists must independently run the full 3×2 grid, then calculate ICC on the “window shape parameters” themselves: whether peak positions are consistent across therapists and whether decay slopes share direction. Referencing general consistency interpretation frameworks, ICC below 0.5 is considered poor, 0.5–0.75 moderate, and above 0.75 good—if the window shape itself is irreproducible across therapists, then any “cross-endpoint isomorphism” is merely two projections of the same unreliable measurement and holds no discussion value.
Crucial warning to note: The credibility of external Qi literature is severely compromised. Previous reports on Yan Xin Qi Gong external Qi (YXQ-EQ) claimed induction of apoptosis and inhibition of migration in cultured tumor cells (PMID 20110687, 30184531), yet at least two of these papers have been formally retracted (retraction notes PMID 33895982, 39327376). This series cites these two papers solely as “historical case studies for protocol design reference,” not as evidence of external Qi effects. Any subsequent experiments must meet the threshold of pre-registration, blinded treatment (therapists and cell culturers mutually blinded), and sham control.
Evidence and Limitations
Verified: Cell proliferation/toxicity assays (CCK-8/MTT families), Annexin V apoptosis detection, and TMRE membrane potential measurement are established methods with standard protocols and known pitfalls (PMID 27604355, 27683123, 30465199); ICC can quantify cross-endpoint/cross-batch consistency (PMID 28904005).
Partially Verified: The “time window” phenomenon has numerous precedents in pharmacology and chronopharmacology—the timing of treatment and detection represents a real, measurable variable, making window heatmaps a valid descriptive tool.
Pending Verification: External Qi processing generates isomorphic time windows across multiple endpoints; there is zero direct evidence. Core portions of existing external Qi in vitro literature have been retracted (PMID 33895982, 39327376), leaving an evidence base that is weak. The existence of cross-endpoint isomorphism admits a mundane “shared pathway” explanation, which must be ruled out via mediation analysis.
Related Reading
- External Qi and NK Cells: Time Windows and Dose Response
- Directional Decomposition of “State-Effect Coupling”: EEG→Cells vs. Cells→EEG Bidirectional Granger
- Variance Explained by Gong State Markers: Benchmarking
- Research Pool: Biofrequency Code
L1 Network Status: This topic has not yet entered the corresponding L1 network. Verification of CORRESPONDENCE_NETWORK.md and EDGE_TABLE_science-yixue.md reveals no edge corresponding to “External Qi ↔ NK/proliferation/apoptosis/mitochondrial membrane potential cross-endpoint window” (the EDGE_TABLE contains only generic nodes for “Qigong/Taiji”; the “apoptosis” hit pertains to E-060 electroacupuncture-myocardial ischemia context and is unrelated to this topic). This gap has been registered in science-yixue-l1-gaps.md.
Evidence Level: C
References: PMID 20110687, PMID 30184531, PMID 33895982, PMID 39327376, PMID 27604355, PMID 27683123, PMID 30465199, PMID 35464440, PMID 28904005
Evidence Level: A = direct PubMed support; B = scientific analogy/indirect evidence; C = traditional consensus/preliminary exploration.
Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.
Leave a Reply