「少照一会儿」还是「照暗一点」:时长与强度这两根杠杆,取决于你站在曲线的哪一段

Written by

in

Language

Evidence level: B

“时长比强度重要”这句话,只在强度已经饱和的那一小段区间里成立。互易律在两端都会断:阈值以下,时间买不到强度;饱和以上,强度买不到时间。文献之所以看起来互相矛盾,是因为它们各自站在曲线的不同段落上说话。

正文

月相-月经耦合被夜间光照慢慢磨掉,这一点前几篇已经立住。接下来必须回答一个更实际的问题:如果要保,先关哪一个开关——是”少照一会儿”,还是”照暗一点”? 这两个动作在生活中可以分开执行(睡前把灯调暗,或者早点关灯),但背后的生理含义完全不同:如果起作用的是剂量总量,两者可以互换;如果起作用的是剂量怎么切分,它们就不能互换。

文献给出的第一印象很干脆:时长赢,强度输。 Dewan 等人(PMID 21532952)把 56 名年轻受试者随机分进 9 个条件——强度取 2000/4000/8000 lux,时长取 1/2/3 小时,正交。双因素方差分析里只有时长主效应显著(F2,48=4.37,P=0.01),强度不显著(F2,48=0.88,P=0.42),而且两个因素之间没有交互。1 小时只延迟约 10 分钟,3 小时约 1.5 小时——从 1 到 3 小时剂量翻三倍,效果几乎全由”多久”带来。

但这个结论有一个被普遍忽略的前提。他们测的三档强度,全部落在反应的饱和区之上。 Zeitzer 等人(PMID 10922269)用 6.5 小时光照做出过标准的 S 形剂量-效应曲线:褪黑素抑制半饱和点约 105 lux、饱和点约 200 lux;相位延迟半饱和约 120 lux、饱和约 550 lux。也就是说,从 2000 lux 往上,强度这根杠杆已经推到底了——再加亮度,机器不再响。在这种条件下测出”强度不重要”,等于问一个已经满格的电池”再多充一点会不会更满”。那不是自然规律,那是天花板的形状。

把镜头移到饱和区以下,结论立刻翻面。McIntyre 等人(PMID 2761346)用 200/400/600 lux 三个低强度各照 3 小时,发现每一档强度都在 1 小时内就达到它自己的最大抑制,之后继续延长时间毫无作用——原文明确写”强度与时长之间不存在互易关系”。Reiter 等人在仓鼠上的结果更极端(PMID 6545810):0.019 µW/cm² 照 20 分钟不抑制褪黑素,把时长加到 30、60、120、180 分钟仍然不抑制。在阈值以下,时间买不到强度。 这是与 Dewan 方向完全相反的失效。

第三块证据来自另一头。Joyce 等人(PMID 35259981)用 2 毫秒闪光、把强度从 3 lux 拉到 9500 lux,测出清晰的剂量-效应,甚至带阈值样行为;而把闪光时长从 10 微秒拉到 10 秒、强度固定 2000 lux 时,时长几乎没有效果。同样,Rahman 等人(PMID 28405608)用 15 秒和 2 分钟的 9500 lux 脉冲,两组的相位延迟没有差别。在极短时程里,强度的杠杆还在陡坡上,时长的杠杆反而先平了。

三段证据拼起来,图像就唯一了。互易律(强度×时长=恒定剂量)不是普遍规律,它只在中间一段成立。 下界以下:剂量不足以触发反应,加时间无效(Reiter、McIntyre)。上界以上:反应已饱和,加强度无效(Dewan、Zeitzer)。只有夹在两端之间的那个”工作区间”里,两根杠杆才近似等价——对褪黑素抑制大约落在几十到两百 lux,对相位延迟大约落在十几到五百多 lux。这恰好是普通室内照明的亮度区间,也是我们每天真实生活的区间。

更麻烦的是,这个区间自己还在动。Sharma 等人(PMID 9919688)在实验鼠上测出,达到饱和所需的临界强度随昼夜相位不同而不同:CT20 约 100 lux,CT15 约 500 lux。“谁更重要”没有统一答案,因为饱和点本身随时间漂移。

所以真实生活的判决实验不能只做单因素。可执行的设计是纵向队列按”夜间光照时长×强度”四象限分层(短暗/短亮/长暗/长亮),以月相-月经同步度为结局,用双因素交互模型直接比较两个主效应;光照用佩戴式剂量计连续记录,剂量单位用黑视素等效照度而非 lux(PMID 32248548)。但必须同时处理两个混杂:一是光历史——一周昏暗环境会让人对后续光照的敏感性增强 60-70%(PMID 21224217、PMID 12390501),所以”今晚的光”不是独立事件;二是象限本身不独立——长时程曝光的人往往也是亮环境的人,观测性四象限天然相关,只有随机分配才拆得开。真实数据已证明这种拆分的必要:佩戴式剂量计的人群研究里,大多数人白天光照不足、睡前光照过量(PMID 37812713),两头的偏离是同时发生的。

对”天人相应”的意义,古人其实早写下了同一句话,只是写在了没有饱和区的地方。《素问·阴阳应象大论》说”少火之气壮,壮火之气衰……壮火食气,少火生气“——同一把火,小的是养,大的是耗,作用方向随剂量翻转。这正是倒 U 与饱和曲线的古典表述:不是”越多越坏”,而是存在一个转折点,过了它,同一个东西换了性质。而《素问·宣明五气》的五劳所伤”久视伤血,久卧伤气,久坐伤肉,久立伤骨,久行伤筋“,五条全部用”久”作自变量,且每一条”久”伤的是不同的东西。古人的病理模型把时长当成主变量——这不是巧合:在烛火与油灯的世界里,强度被天光封顶,室内照度根本够不到饱和点,所以唯一能拉动的杠杆就是时间。今天我们之所以要重算这笔账,是因为 LED 第一次让普通人把强度推到了饱和区之上——古人没见过的那个区间,正是”强度不重要”这个结论诞生的地方。

「少照一会儿」还是「照暗一点」,不是一个可以拍脑袋回答的问题——它取决于你现在站在曲线的哪一段。 而在两端的失效区里,任何”用另一边补回来”的直觉,都是错的。

证据与局限

已验证: 在强度处于饱和区(2000-8000 lux)时,光照时长显著影响相位延迟(1→3 小时,1 小时约 10 分钟、3 小时约 1.5 小时),而强度主效应不显著,两因素无交互(PMID 21532952);10,000 lux 下 0.2-4.0 小时的时长-效应呈非线性,每分钟效率以最短时长最高(短时长单位效率是 4 小时组的 5 倍以上,PMID 22526883);褪黑素抑制的 S 形剂量-效应曲线半饱和约 105 lux、饱和约 200 lux(PMID 10922269);极低照度下延长时间不能产生抑制(仓鼠实验,PMID 6545810);低强度(200-600 lux)各档均在 1 小时内达到自身最大抑制,延长无进一步作用(PMID 2761346);2 毫秒闪光下强度呈清晰剂量-效应而时长近乎惰性(PMID 35259981);15 秒与 2 分钟脉冲的相位延迟无差异(PMID 28405608)。

部分验证: 饱和点随昼夜相位漂移(CT20 约 100 lux、CT15 约 500 lux,PMID 9919688),故”哪根杠杆重要”具有相位依赖性;两周昏暗史可使后续光照的相位延迟效应增强 60-70%(PMID 21224217、PMID 12390501),说明单次曝光不独立;真实生活中白天不足、夜间过量的偏离同时发生(PMID 37812713)。

待验证: 上述全部为褪黑素抑制与相位延迟的剂量学结论,尚无任何研究以月相-月经同步度为结局比较时长与强度两个主效应;”工作区间”(两根杠杆近似等价的那一段)的上下边界只有褪黑素/相位层面估值,未在耦合结局上标定;四象限观测分层能否拆开时长与强度的天然相关(须随机分配验证);黑视素等效照度作为剂量单位在长期居家测量中的等效性(PMID 32248548 基于实验室短期刺激)。

相关阅读

证据等级: B
参考文献: PMID 21532952, PMID 22526883, PMID 28405608, PMID 35259981, PMID 2761346, PMID 6545810, PMID 10922269, PMID 9919688, PMID 21224217, PMID 12390501, PMID 37812713, PMID 32248548


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。

\”Shine for a shorter time\” or \”Dim the light\”: duration and intensity as levers depend on where you stand on the curve

Evidence level: B

“Duration is more important than intensity” holds true only within a small interval where intensity is already saturated. The law of reciprocity fails at both ends: below the threshold, time cannot purchase intensity; above saturation, intensity cannot purchase time. Apparent contradictions in literature arise because sources speak from different segments of the curve.

Main Text

The erosion of moon-phase-menstrual coupling by nocturnal illumination has been established in previous articles. The next practical question must be answered: If preservation is the goal, which switch should be turned off first—”Shao Zhao Yi Hui R” (reduce duration slightly) or “Zhao An Dian” (dim the light slightly)? These actions can be executed separately in daily life (dimming lights before bed or turning them off earlier), but their physiological implications differ fundamentally: if the mechanism relies on total dose, they are interchangeable; if it relies on dose partitioning, they are not.

The initial impression from the literature is straightforward: Duration wins, intensity loses. Dewan et al. (PMID 21532952) randomly assigned 56 young subjects to nine conditions—intensity levels of 2000/4000/8000 lux and durations of 1/2/3 hours, orthogonal. In the two-factor ANOVA, only the main effect of duration was significant (F2,48=4.37, P=0.01), while intensity was not (F2,48=0.88, P=0.42), and no interaction between the two factors was found. A 1-hour exposure delayed phase by approximately 10 minutes; 3 hours delayed it by about 1.5 hours—tripling the dose from 1 to 3 hours yielded effects driven almost entirely by “how long.”

However, this conclusion relies on a premise often overlooked. The three intensity levels tested all fell above the saturation zone of the response. Zeitzer et al. (PMID 10922269) established a standard S-shaped dose-response curve using 6.5 hours of illumination: melatonin suppression half-saturation at ~105 lux and saturation at ~200 lux; phase delay half-saturation at ~120 lux and saturation at ~550 lux. Thus, above 2000 lux, the intensity lever is already pushed to the bottom—adding more brightness yields no further response. Measuring “intensity doesn’t matter” under these conditions is akin to asking a fully charged battery if adding more charge will make it fuller. That is not a natural law; that is the shape of the ceiling.

Shifting the lens to below the saturation zone flips the conclusion immediately. McIntyre et al. (PMID 2761346) used three low intensities (200/400/600 lux) for 3 hours each and found each intensity level reached its maximum suppression within 1 hour; extending time further had no additional effect—the text explicitly states “no reciprocal relationship exists between intensity and duration.” Reiter et al. found even more extreme results in hamsters (PMID 6545810): 0.019 µW/cm² for 20 minutes failed to suppress melatonin; extending duration to 30, 60, 120, and 180 minutes still failed to suppress. Below the threshold, time cannot purchase intensity. This is a failure completely opposite to Dewan’s direction.

Third evidence comes from the other end. Joyce et al. (PMID 35259981) used 2-millisecond flashes, increasing intensity from 3 lux to 9500 lux, revealing a clear dose-response with threshold-like behavior; conversely, extending flash duration from 10 microseconds to 10 seconds at a fixed intensity of 2000 lux showed little to no effect of duration. Similarly, Rahman et al. (PMID 28405608) used 9500 lux pulses of 15 seconds and 2 minutes with no difference in phase delay between groups. In ultra-short durations, the intensity lever remains on a steep slope, while the duration lever flattens first.

Piecing together these three segments yields a unique image. The law of reciprocity (intensity × duration = constant dose) is not a universal rule; it holds only in the middle segment. Below the lower bound: insufficient dose to trigger response, adding time is ineffective (Reiter, McIntyre). Above the upper bound: response is saturated, increasing intensity is ineffective (Dewan, Zeitzer). Only within the “working interval” between these two bounds are the two levers approximately equivalent—melatonin suppression roughly falls in the tens to hundreds of lux range, while phase delay falls in the tens to over-hundreds of lux range. This coincides with the brightness range of ordinary indoor lighting and our daily lived reality.

Complicating matters further, this interval itself is dynamic. Sharma et al. (PMID 9919688) found in experimental mice that the critical intensity required to reach saturation varies with circadian phase: ~100 lux for CT20, ~500 lux for CT15. “Which lever is more important” has no unified answer because the saturation point drifts with time.

Therefore, a real-life adjudication experiment cannot rely on single-factor designs. A feasible design is a longitudinal cohort stratified by the four quadrants of “nocturnal illumination duration × intensity” (short/dim, short/bright, long/dim, long/bright), using moon-phase-menstrual synchronization as the outcome, employing a two-factor interaction model to directly compare the two main effects; illumination should be recorded continuously via wearable dosimeters, using melanopic equivalent lux rather than standard lux as the dose unit (PMID 32248548). However, two confounders must be addressed simultaneously: first, light history—a week in dim environments enhances sensitivity to subsequent illumination by 60-70% (PMID 21224217, PMID 12390501), so “tonight’s light” is not an independent event; second, the quadrants themselves are not independent—those with long-term exposure often also reside in bright environments, making observational quadrants naturally correlated, separable only through randomization. Real data confirms the necessity of this separation: population studies with wearable dosimeters show most individuals have insufficient daytime light but excessive pre-bedlight (PMID 37812713), with deviations at both ends occurring simultaneously.

Significance for “Tian Ren Xiang Ying” (Harmony between Heaven and Man): Ancient texts actually recorded the same sentence, but in a context without a saturation zone. Su Wen · Yin Yang Ying Xiang Da Lun states “Shao huo zhi qi zhuang, zhuang huo zhi qi shuai… Zhuang huo shi qi, shao huo sheng qi” (Weak fire strengthens qi; excessive fire depletes it… Excessive fire consumes qi, weak fire generates qi)—the same fire, small nourishes, large depletes, with the direction of action flipping with dose. This is the classical expression of the inverted U and saturation curves: not “more is worse,” but there exists a turning point where the same thing changes its nature. Meanwhile, Su Wen · Xuan Ming Wu Qi lists five injuries from prolonged activity: “Jiu shi shang xue, jiu wo shang qi, jiu zuo shang rou, jiu li shang gu, jiu xing shang jin” (Prolonged viewing harms blood, prolonged lying harms qi, prolonged sitting harms flesh, prolonged standing harms bone, prolonged walking harms tendons)—all five use “prolonged” as the independent variable, and each “prolonged” injury affects a different organ. The ancient pathological model treats duration as the primary variable—not coincidence: in a world of candles and oil lamps, intensity was capped by daylight, indoor illuminance never reached saturation, so the only lever to pull was time. We must recalculate this ledger today because LEDs first allowed ordinary people to push intensity above the saturation zone—the interval unseen by ancients is precisely where the conclusion “intensity doesn’t matter” was born.

“Shao Zhao Yi Hui R” or “Zhao An Dian” is not a question answerable by guesswork—it depends on which segment of the curve you currently stand on. In the failure zones at both ends, any intuition that “compensate with the other side” is incorrect.

Evidence and Limitations

Verified: At intensities in the saturation zone (2000-8000 lux), illumination duration significantly affects phase delay (1→3 hours: 1 hour ~10 mins, 3 hours ~1.5 hours), while the main effect of intensity is not significant with no interaction between factors (PMID 21532952); at 10,000 lux, duration-effectiveness from 0.2-4.0 hours is non-linear, with efficiency per minute highest at shortest duration (short-duration unit efficiency is over 5 times that of the 4-hour group, PMID 22526883); S-shaped melatonin suppression dose-response curve has half-saturation at ~105 lux and saturation at ~200 lux (PMID 10922269); extending time under very low illumination fails to produce suppression (hamster study, PMID 6545810); low intensities (200-600 lux) reach their own maximum suppression within 1 hour with no further effect from extension (PMID 2761346); under 2-millisecond flashes, intensity shows clear dose-response while duration is nearly inert (PMID 35259981); no difference in phase delay between 15-second and 2-minute pulses (PMID 28405608).

Partially Verified: Saturation point drifts with circadian phase (CT20 ~100 lux, CT15 ~500 lux, PMID 9919688), making “which lever matters” phase-dependent; two weeks of dim history can enhance phase delay effects from subsequent illumination by 60-70% (PMID 21224217, PMID 12390501), indicating single exposure is not independent; deviations in real life (insufficient daytime, excessive nighttime) occur simultaneously (PMID 37812713).

To Be Verified: All above are dose-response conclusions regarding melatonin suppression and phase delay; no study yet uses moon-phase-menstrual synchronization as an outcome to compare the main effects of duration and intensity; upper and lower bounds of the “working interval” (where levers are approximately equivalent) are estimated only at melatonin/phase levels, not calibrated at the coupled outcome; whether four-quadrant observational stratification can disentangle the natural correlation between duration and intensity (requires randomization verification); equivalence of melanopic equivalent lux as a dose unit in long-term home measurement (PMID 32248548 based on laboratory short-term stimulation).

Related Reading

Evidence Level: B
References: PMID 21532952, PMID 22526883, PMID 28405608, PMID 35259981, PMID 2761346, PMID 6545810, PMID 10922269, PMID 9919688, PMID 21224217, PMID 12390501, PMID 37812713, PMID 32248548


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.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *