Place · Level 3
Insomnia
腺苷 + SCN + HPA 三轴失调 · CBT-I A 级证据 · melatonin 是时钟信号不是安眠药· Z 药慎用
Last updated
Story path
- 1Insomnia types · ICSD-3Insomnia types · ICSD-3
- 2CBT-I — A-grade first lineCBT-I — A-grade first line
- 3Nutrient tools — Mg/Gly/L-theanine/melatoninNutrient tools — Mg/Gly/L-theanine/melatonin
- 4Z-drugs + benzos · long-term risksZ-drugs + benzos · long-term risks
- 5Decision tree + red flagsDecision tree + red flags
Chapter 1
Insomnia types · ICSD-3
Insomnia types · ICSD-3
Insomnia is any of three difficulties — falling asleep / staying asleep / waking too early — PLUS daytime functional impairment.
By duration:
Acute insomnia: < 3 months, usually with a trigger (stress / jet lag / illness)Chronic insomnia: ≥ 3 months / ≥ 3 nights per week / with daytime impairment → ICSD-3 diagnostic criteria30-35% of adults have occasional insomnia; 10-15% have chronic insomnia
By presentation:
Sleep-onset insomnia: > 30 minutes to fall asleep after going to bedSleep-maintenance insomnia: waking in the middle of the night and not getting back to sleep for > 30 minutesEarly morning awakening: waking ≥ 1 hour before the target wake time and unable to return to sleepNon-restorative sleep: 7-8 hours in bed but still tired
"I get 6 hours a night — is that insomnia?" Not necessarily — sleep need varies meaningfully between individuals:
Most adults need 7-9 hours, but a small subset (true "short sleepers", < 5%) do fine on 5-6 hoursThe judgment criterion: is there any daytime functional impairment (attention, reaction, mood, error rate)?No impairment + subjectively satisfied → 6 hours may be enough; impairment → you need more
Insomnia is one of the most badly-handled "symptoms" out there. Most people go straight to melatonin or an OTC sleep aid, but:
Z-drugs and benzodiazepines — the true sleep medications — carry serious long-term risks (falls, dementia, dependence)Melatonin is not a sleep aid; it is a clock signal. Wrong dose or timing makes it backfire.CBT-I (cognitive behavioral therapy for insomnia) is A-grade evidence, but adoption is low — most patients don't even know it exists
What this island is trying to do is upgrade the conversation from "what should I take" to "why can't I sleep + which levers actually work".
By duration:
Acute insomnia: < 3 months, usually with a trigger (stress / jet lag / illness)Chronic insomnia: ≥ 3 months / ≥ 3 nights per week / with daytime impairment → ICSD-3 diagnostic criteria30-35% of adults have occasional insomnia; 10-15% have chronic insomnia
By presentation:
Sleep-onset insomnia: > 30 minutes to fall asleep after going to bedSleep-maintenance insomnia: waking in the middle of the night and not getting back to sleep for > 30 minutesEarly morning awakening: waking ≥ 1 hour before the target wake time and unable to return to sleepNon-restorative sleep: 7-8 hours in bed but still tired
"I get 6 hours a night — is that insomnia?" Not necessarily — sleep need varies meaningfully between individuals:
Most adults need 7-9 hours, but a small subset (true "short sleepers", < 5%) do fine on 5-6 hoursThe judgment criterion: is there any daytime functional impairment (attention, reaction, mood, error rate)?No impairment + subjectively satisfied → 6 hours may be enough; impairment → you need more
Insomnia is one of the most badly-handled "symptoms" out there. Most people go straight to melatonin or an OTC sleep aid, but:
Z-drugs and benzodiazepines — the true sleep medications — carry serious long-term risks (falls, dementia, dependence)Melatonin is not a sleep aid; it is a clock signal. Wrong dose or timing makes it backfire.CBT-I (cognitive behavioral therapy for insomnia) is A-grade evidence, but adoption is low — most patients don't even know it exists
What this island is trying to do is upgrade the conversation from "what should I take" to "why can't I sleep + which levers actually work".
Sleep science · 3 key claims
Sleep is not a simple "power off" — it is an active neurochemical event.① Sleep is driven by three axes:
Adenosine pressure (homeostatic): the longer you are awake, the more tired you get (covered in detail in the atlas `caffeine-l-theanine/caffeine-pharm` scene)suprachiasmatic nucleus: The brain's master clock — set by light, it runs the body's day–night rhythm. clock (circadian): darkness triggers melatonin + morning light resets the clock (detailed in `melatonin/real-dose-vs-commercial`)hypothalamic–pituitary–adrenal axis: The body's stress-response chain (hypothalamus → pituitary → adrenal) that releases cortisol. / stress axis: chronic high cortisol → trouble falling asleep + light sleep (detailed in `ashwagandha/cortisol-mechanism`)
Disruption of any one of these three can produce insomnia:
16 hours awake but a late coffee → adenosine is masked → onset troubleBlue light at 22:00 → SCN didn't get the "time to sleep" signalHigh work stress + HPA won't shut down → waking at 3 amAll three dysregulated simultaneously: the typical chronic insomnia pattern
② Sleep stages (a full 7-8 h contains 4-5 cycles of about 90 minutes):
NREM 1: onset transition, a few minutesNREM 2: the bulk of sleep, 50-60%NREM 3 (slow-wave / deep sleep): physical recovery, immune activation, memory consolidation; concentrated in the first half of the nightREM: dreaming + emotion + procedural learning consolidation; concentrated in the second half of the night
Clinical implications:
Maintenance insomnia + early morning awakening → REM is lost → emotional and learning problemsOnset difficulty → deep sleep is lost → physical and immune problemsFragmented sleep (repeated brief awakenings) → both are impaired
③ Key facts:
"Catching up" by sleeping in on weekends: partially effective, but cannot fully compensate and scrambles the SCN, making Monday harder (social jet lag)The objective impact of short sleep: the widely quoted 'reaction speed equal to a blood alcohol of 0.05%' comes from a study of 17 hours of continuous wakefulness — it measures time awake, not 'I only got 6 hours last night', and the two don't convert into each other. The 'DUI level' framing also travels badly: China's drink-driving threshold is 20 mg/100 mL (0.02%), a different line from that figure. What is solid: reaction speed and attention genuinely degrade, and long-term risk of CV / T2D / Alzheimer / depression risesMemory consolidation lives in NREM3 + REM: studying and then sleeping well outperforms staying up to reviewNREM3 starts to decline from age 40+ (but should not disappear) — this is why light sleep is common in older adults, but recovery should not be given up
"Why can I lie in bed for an hour and not sleep, but doze off instantly on the couch?" This has a name — psychophysiological insomnia. After repeated failed nights, the bed itself is registered by the body as an "anxiety location", and the act of getting into bed starts triggering wakefulness. There is a CBT-I component called stimulus control designed for this: only enter bed when sleepy / leave immediately if not / the bed is only for sleep and sex.
分型与诊断 · 我这算失眠吗
按时间:急性失眠: < 3 月, 通常有诱因 (压力 / 时差、病)慢性失眠: ≥ 3 月 / 周 ≥ 3 次、白天功能受损 → ICSD-3 诊断标准30-35% 成人有偶发失眠; 10-15% 慢性
按表现:
入睡困难 (sleep-onset): 上床后 > 30 分钟仍睡不着维持困难 (sleep-maintenance): 半夜醒 + 难再睡 > 30 分钟早醒 (early morning): 比目标起床早 ≥ 1 h + 无法再睡非恢复性睡眠: 睡了 7-8 h 仍累
分型不是为了给自己贴一个标签, 而是为了知道该去拧哪一股力: 入睡困难通常指向腺苷压力不够 (咖啡因挡着、白天补过觉) 或者压力轴还没关掉; 维持困难与早醒更常是压力轴的后半夜回升来得太早; 非恢复性睡眠则要先怀疑深睡的量, 以及有没有被反复的微觉醒切碎 —— 后者往往根本不是失眠, 而是呼吸暂停之类的另一件事。
我每晚 6 小时, 算失眠吗? 答案是不一定, 因为睡眠需求个体差异不小:
多数成人 7-9 h, 但少数 5-6 h 也够 (短睡眠者, 真正占比 < 5%)。判断标准: 白天有没有功能受损 (注意力、反应、心情、错误率)?没有受损 + 主观满意, 6 小时可能就够; 有受损, 就需要更多。
失眠是被对待方式最错的症状 之一。多数人直接吃褪黑素或 OTC 安眠药, 但:
Z 药、苯二氮卓这类真正的安眠药, 长期使用风险很大 (跌倒、痴呆、依赖)。褪黑素不是安眠药, 它是时钟信号; 剂量或时机不对反而起反作用。CBT-I (认知行为治疗失眠) 是 A 级证据, 但普及度低, 大多数患者不知道。
这一岛想做的事, 是把失眠 从吃什么 的对话, 升级成为什么睡不着 + 哪些杠杆有用 的对话。
Chapter 2
CBT-I — A-grade first line
CBT-I — A-grade first line
CBT-I (Cognitive Behavioral Therapy for Insomnia) is, for chronic insomnia:
The first-line intervention (AASM 2021 / NICE / ACP)Supported by A-grade RCT evidenceEffect size: onset time ↓ 19 min · WASO ↓ 26 min · sleep efficiency ↑ 10%. Total sleep time moved only +7.6 min, and its confidence interval crosses zero (*Trauer 2015* *Ann Intern Med* meta-analysis)Long-term effect: maintained at 6 months — fundamentally different from sleep medications, which "rebound on stop"Side effects: essentially zero
The 5 components of CBT-I:
① Sleep Hygiene
Fixed wake + bed timesLimit caffeine (before noon)Limit alcoholBlue-light control for the hour before bedBedroom cold + dark + quietNo work / phone in bed
② Stimulus Control — the core:
Only get into bed when sleepyIf not asleep within 20 minutes → leave bed for another space + a quiet activity (reading, folding laundry)Return when sleepyBed only for sleep and sexRebuilds the "bed = sleep" conditioned association
③ Sleep Restriction — counterintuitive but the most powerful:
Estimate your actual sleep time (e.g. 5 h)Restrict time in bed to that number + 30 minutes (e.g. 5.5 h)No lounging in bed when not sleepingSleep efficiency ≥ 85% for 1 week → add 15 minutes per weekShort-term may feel more tired, but it rebuilds sleep driveTherapist-supervised — not recommended as DIY
④ Cognitive Restructuring
"If I can't sleep tonight, I'm done tomorrow" → reframe → "I might be tired but I'll get through""Normal people sleep 8 hours" → reframe → "my personal need may be 7 hours""This week the insomnia is going to get worse" → reframe → "acute insomnia mostly self-limits"
⑤ Relaxation Training
Progressive muscle relaxation / diaphragmatic breathing / mindfulness20-30 minutes pre-bed
How to access CBT-I:
In-person, 6-8 sessions: clinical psychology / sleep medicine centerDigital CBT-I apps:Sleepio (UK NHS-recommended), Somryst (US FDA-cleared prescription), CBT-I Coach (VA, free), Stellar, Sleep School, etc.Most are 6-week self-guided programsRCT evidence is close to in-person therapyBooks: "Quiet Your Mind and Get to Sleep" (Carney + Manber)
Why isn't CBT-I universal?
Doesn't sell like pills — industry incentives are misalignedPhysician training time is limitedPatients expect "fast" — CBT-I needs 6-8 weeksBut: most people diagnosed with insomnia are never referred for CBT-I at all (we have no source for a precise share, so no number here)
Drugs vs CBT-I — read the following as the guideline-level picture, not as a result from Trauer 2015: that meta-analysis compared CBT-I against waitlist / sleep hygiene / placebo, with no head-to-head drug arm, so it cannot support a tier-by-tier table.
Short-term (< 4 weeks): broadly comparableLong-term: CBT-I holds and drugs don't — tolerance, rebound and side effects accumulate. This is why every guideline puts CBT-I firstAfter CBT-I many patients no longer need medication (again, no sourced percentage, so none is given)
The first-line intervention (AASM 2021 / NICE / ACP)Supported by A-grade RCT evidenceEffect size: onset time ↓ 19 min · WASO ↓ 26 min · sleep efficiency ↑ 10%. Total sleep time moved only +7.6 min, and its confidence interval crosses zero (*Trauer 2015* *Ann Intern Med* meta-analysis)Long-term effect: maintained at 6 months — fundamentally different from sleep medications, which "rebound on stop"Side effects: essentially zero
The 5 components of CBT-I:
① Sleep Hygiene
Fixed wake + bed timesLimit caffeine (before noon)Limit alcoholBlue-light control for the hour before bedBedroom cold + dark + quietNo work / phone in bed
② Stimulus Control — the core:
Only get into bed when sleepyIf not asleep within 20 minutes → leave bed for another space + a quiet activity (reading, folding laundry)Return when sleepyBed only for sleep and sexRebuilds the "bed = sleep" conditioned association
③ Sleep Restriction — counterintuitive but the most powerful:
Estimate your actual sleep time (e.g. 5 h)Restrict time in bed to that number + 30 minutes (e.g. 5.5 h)No lounging in bed when not sleepingSleep efficiency ≥ 85% for 1 week → add 15 minutes per weekShort-term may feel more tired, but it rebuilds sleep driveTherapist-supervised — not recommended as DIY
④ Cognitive Restructuring
"If I can't sleep tonight, I'm done tomorrow" → reframe → "I might be tired but I'll get through""Normal people sleep 8 hours" → reframe → "my personal need may be 7 hours""This week the insomnia is going to get worse" → reframe → "acute insomnia mostly self-limits"
⑤ Relaxation Training
Progressive muscle relaxation / diaphragmatic breathing / mindfulness20-30 minutes pre-bed
How to access CBT-I:
In-person, 6-8 sessions: clinical psychology / sleep medicine centerDigital CBT-I apps:Sleepio (UK NHS-recommended), Somryst (US FDA-cleared prescription), CBT-I Coach (VA, free), Stellar, Sleep School, etc.Most are 6-week self-guided programsRCT evidence is close to in-person therapyBooks: "Quiet Your Mind and Get to Sleep" (Carney + Manber)
Why isn't CBT-I universal?
Doesn't sell like pills — industry incentives are misalignedPhysician training time is limitedPatients expect "fast" — CBT-I needs 6-8 weeksBut: most people diagnosed with insomnia are never referred for CBT-I at all (we have no source for a precise share, so no number here)
Drugs vs CBT-I — read the following as the guideline-level picture, not as a result from Trauer 2015: that meta-analysis compared CBT-I against waitlist / sleep hygiene / placebo, with no head-to-head drug arm, so it cannot support a tier-by-tier table.
Short-term (< 4 weeks): broadly comparableLong-term: CBT-I holds and drugs don't — tolerance, rebound and side effects accumulate. This is why every guideline puts CBT-I firstAfter CBT-I many patients no longer need medication (again, no sourced percentage, so none is given)
五个组件 · 效应大小 · 去哪里找
先看这五个组件里, 哪两个真的在搬动机制。刺激控制拆的是一条学出来的反射。睡眠压力来自一种叫腺苷的代谢残余: 醒着越久堆得越多, 堆够了你才困, 而只有深睡能把它清空。问题是, 当你连着很多个夜晚躺在床上却睡不着, 大脑会老老实实地把上床这个动作和开始警觉绑在一起 —— 到后来只是走进卧室、掀开被子, 心率就开始往上走。这不是想出来的, 是练出来的, 所以也只能重新练回去: 只在真困时上床, 躺了一阵没睡着就离开床去做点安静的事, 让床这个地点重新只跟很快睡着这一件事配对。它见效慢, 因为拆一条练了几年的反射本来就要重复很多次。
睡眠限制动的是腺苷本身。一个睡不好的人最自然的补救动作 —— 提早上床、早上多躺一会、下午补个觉 —— 每一样都在提前把当天攒的睡眠压力泄掉一部分, 于是晚上更不困; 更不困就更早上床, 在床上清醒得更久, 圈越收越紧。睡眠限制反过来做: 把在床时间压到接近你实际睡着的时长, 白天不许补, 于是到了晚上腺苷够高, 你倒下就睡; 等睡眠效率上来, 再一周一周把在床时间放回去。短期会更累是设计之内 —— 那份累正是压力被重新攒起来的证据。也正因为如此它必须由治疗师监督: 压得太狠时, 白天开车与操作机械的风险是真实的。
CBT-I (Cognitive Behavioral Therapy for Insomnia) 是慢性失眠公认的第一线治疗:
各大指南一致推荐 (AASM 2021 / NICE / ACP)A 级 RCT 证据效应大小: 入睡时间 ↓ 19 min · WASO ↓ 26 min · 睡眠效率 ↑ 10%; 总睡眠时间只动了 +7.6 min, 而且它的置信区间跨过 0 (Trauer 2015 Ann Intern Med meta-analysis)长期效应: 6 月后仍维持 —— 与安眠药停药即反弹根本不同副作用: 几乎零
CBT-I 5 个组件:
① 睡眠卫生 (Sleep Hygiene)
固定起床 + 睡眠时间限咖啡因 (中午前)限酒睡前 1 h 蓝光控制卧室冷 + 暗 + 安静不在床上工作、看手机
② 刺激控制 (Stimulus Control) —— 核心:
只在困时上床上床 20 分钟内没睡着 → 起床到另一空间 + 安静活动 (读书、折衣)困了再回床只用于睡眠 + 性重建床 = 睡眠条件反射
③ 睡眠限制 (Sleep Restriction) —— 反直觉但最有效:
估计你实际睡眠时间 (e.g. 5 h)把在床时间限制到这个数 + 30 分钟 (e.g. 5.5 h)不睡也不能赖在床上睡眠效率 ≥ 85% 持续 1 周 → 每周加 15 分钟短期可能更累, 但重建睡眠驱力由治疗师监督, 不建议自己做
④ 认知重构 (Cognitive Restructuring)
今晚再睡不着我明天就完了 → 转 → 我可能累, 但能撑过去正常人睡 8 小时 → 转 → 我个人需求可能 7 小时这周失眠就要严重了 → 转 → 急性失眠多数自限
⑤ 放松训练
渐进性肌肉放松、横膈呼吸、正念入睡前 20-30 分钟
获取 CBT-I 的途径:
面对面 6-8 次: 临床心理、睡眠中心数字 CBT-I 应用 (digital CBT-I): Sleepio (英国 NHS 推荐), Somryst (US FDA 批 prescription), CBT-I Coach (VA 免费), Stellar, Sleep School 等多数 6 周自助方案RCT 证据接近面对面治疗书籍: "Quiet Your Mind and Get to Sleep" (Carney + Manber)
为什么 CBT-I 不普及?:
不像药丸卖钱 —— 行业激励错位医生培训时间少患者期待快速, CBT-I 需 6-8 周但是: 拿到失眠诊断的人, 大多数根本没被转介去做 CBT-I (具体比例站内没有可靠来源, 所以这里不给数字)
药物 vs CBT-I 比较 —— 下面这几行请当作指南层面的总体图景读, 不要当成 Trauer 2015 的结果: 那份 meta 的对照臂是等待名单 / 只做睡眠卫生 / 安慰剂, 没有与药物的头对头比较, 撑不起一张分档表。
短期 (< 4 周): 大体相当长期: CBT-I 稳得住, 药物稳不住 —— 耐受、反弹、副作用会累积。这正是各大指南把 CBT-I 排在第一位的原因CBT-I 之后, 相当一部分人不再需要药物 (同样没有可靠来源支撑一个具体百分比, 所以不写)
Trauer 2015 + Edinger 2021 guideline
Trauer 2015 *Annals of Internal Medicine* meta-analysis (the gold citation for CBT-I):20 RCTs, N = 1162 chronic insomnia adultsComparing CBT-I (in-person, 4-9 sessions) vs control (waitlist / sleep hygiene only / placebo)Post-treatment results:Onset time: −19.0 minutesWake after sleep onset (WASO): −26 minutesTotal sleep time: +7.6 minutes, 95% CI −0.5 to 15.7 — the interval crosses zero, i.e. not statistically significant. Which fits: the sleep restriction phase deliberately tightens time in bed, so what CBT-I buys is sounder sleep in the same window, not more of itSleep efficiency: +10%Effects maintained at later follow-up time points
Edinger 2021 AASM Clinical Practice Guideline (*J Clin Sleep Med*):
STRONG recommendation — there is only one in the whole guideline:Multicomponent CBT-I (in-person + telehealth) for chronic adult insomniaCONDITIONAL recommendations (the guideline's wording is 'we suggest'):Brief Behavioral Treatment for Insomnia (BBT-I, 2-4 sessions)Stimulus control aloneSleep restriction aloneRelaxation training aloneRecommended AGAINST — the guideline's only negative recommendation: sleep hygiene as a single-component therapyOne more thing worth stating: digital CBT-I is not a standalone recommendation item in this guideline. It sits inside the evidence base for multicomponent CBT-I and cannot be labelled 'strong' on its own
Key: CBT-I is a multicomponent program — any single component is inferior to the full package.
Is CBT-I right for everyone?
Not appropriate for (these aim mainly at the sleep restriction component): acute insomnia (short-term, usually self-limiting); severe untreated depression (treat depression first); uncontrolled bipolar disorder (sleep restriction can trigger mania); sleep apnea (needs CPAP, not CBT-I); epilepsy or a history of seizures — sleep deprivation is a well-established threshold-lowering factor; parasomnias (sleepwalking, night terrors) — sleep deprivation worsens episodes; and anyone who drives or operates machinery daily: daytime sleepiness rises during the first week or two of restriction, the risk is real, and a therapist must run that phaseAppropriate for: chronic (≥ 3 months) primary insomnia and most comorbid insomniaCBT-I + comorbid management: when coexisting with depression / anxiety / chronic pain / GERD, CBT-I is still effective and synergizes with comorbidity management
Practical recommendations:
1. See your PCP first: rule out hypothyroidism / anemia / sleep apnea / depression
2. Find CBT-I resources:
US: psychologytoday.com to find a CBT-I clinical psychologistChina: tier-3 hospital sleep medicine center / psychiatryDigital apps: Sleepio / Somryst / CBT-I Coach3. At the same time, adjust lifestyle: caffeine cutoff / blue light / fixed wake time
4. Complete the 6-8 week program
5. If it fails: refer to sleep medicine specialty for evaluation of other causes
Chapter 3
Nutrient tools — Mg/Gly/L-theanine/melatonin
Nutrient tools — Mg/Gly/L-theanine/melatonin
Evidence-based "nutritional sleep tools" (ranked by evidence):
① Magnesium (Mg) — Grade B
Mechanism: SERCA calcium pump returns muscles to relaxation; physical NMDA-receptor Mg²⁺ block (atlas `magnesium/relax` L4); GABA modulationDose: 200-400 mg elemental Mg pre-bedForm: glycinate / threonate (crosses blood–brain barrier: The 'security gate' on brain vessels that blocks most substances in blood from entering the brain.) / citrate > magnesium oxide (poorly absorbed)Expect: onset time ↓ 10-15 minutes (RCT mixed); subjective relaxationSide effects: large doses cause diarrhea (a self-regulating mechanism); the glycinate form is gentleWho benefits most: people with known Mg deficiency + cramps + anxiety
② Glycine — Grade B
Mechanism: thermoregulation (drop in body temperature is a sleep trigger) + GABA modulation + atlas `glycine/metabolic-hub` L4Dose: 3 g, 30-60 minutes pre-bedExpect: improved onset + subjective sleep quality (*Yamadera 2007* RCT)Side effects: extremely rareWho benefits most: light sleepers + non-restorative sleep
③ L-Theanine — Grade B
Mechanism: GABA + α brain waves + reduced anxiety (detailed in atlas `caffeine-l-theanine/l-theanine`)Dose: 200-400 mg pre-bedExpect: subjective relaxation + improved onset (focus is anxiety-driven onset)Side effects: nearly zeroWho benefits most: anxiety-driven sleep onset difficulty
④ Melatonin — Grade A-B (only in specific use cases)
Mechanism: detailed in atlas `melatonin/real-dose-vs-commercial` L4True indications:Sleep-onset insomnia: 0.3-0.5 mg taken 30-60 minutes pre-bed (not "when getting into bed")Jet lag: 0.5-3 mg close to local bedtime (about 10 pm to midnight) after arrival, for 3-5 days — not moved up to dusk: the Cochrane review states plainly that taking it too early causes daytime sleepiness and delays adaptationShift work / DSPS: 0.3-0.5 mg taken 4-6 h before the target sleep time (phase advance)Total blindness (non-24): prescription dosingNot indicated for:Maintenance insomnia + early morning awakening: melatonin's half-life is too short to helpChronic primary insomnia: CBT-I is far superior to melatoninCommon mistakes:Taking 5-10 mg → receptor saturation + next-day grogginessSugar gummy children's melatoninTaking it at bedtime (too late)
① Magnesium (Mg) — Grade B
Mechanism: SERCA calcium pump returns muscles to relaxation; physical NMDA-receptor Mg²⁺ block (atlas `magnesium/relax` L4); GABA modulationDose: 200-400 mg elemental Mg pre-bedForm: glycinate / threonate (crosses blood–brain barrier: The 'security gate' on brain vessels that blocks most substances in blood from entering the brain.) / citrate > magnesium oxide (poorly absorbed)Expect: onset time ↓ 10-15 minutes (RCT mixed); subjective relaxationSide effects: large doses cause diarrhea (a self-regulating mechanism); the glycinate form is gentleWho benefits most: people with known Mg deficiency + cramps + anxiety
② Glycine — Grade B
Mechanism: thermoregulation (drop in body temperature is a sleep trigger) + GABA modulation + atlas `glycine/metabolic-hub` L4Dose: 3 g, 30-60 minutes pre-bedExpect: improved onset + subjective sleep quality (*Yamadera 2007* RCT)Side effects: extremely rareWho benefits most: light sleepers + non-restorative sleep
③ L-Theanine — Grade B
Mechanism: GABA + α brain waves + reduced anxiety (detailed in atlas `caffeine-l-theanine/l-theanine`)Dose: 200-400 mg pre-bedExpect: subjective relaxation + improved onset (focus is anxiety-driven onset)Side effects: nearly zeroWho benefits most: anxiety-driven sleep onset difficulty
④ Melatonin — Grade A-B (only in specific use cases)
Mechanism: detailed in atlas `melatonin/real-dose-vs-commercial` L4True indications:Sleep-onset insomnia: 0.3-0.5 mg taken 30-60 minutes pre-bed (not "when getting into bed")Jet lag: 0.5-3 mg close to local bedtime (about 10 pm to midnight) after arrival, for 3-5 days — not moved up to dusk: the Cochrane review states plainly that taking it too early causes daytime sleepiness and delays adaptationShift work / DSPS: 0.3-0.5 mg taken 4-6 h before the target sleep time (phase advance)Total blindness (non-24): prescription dosingNot indicated for:Maintenance insomnia + early morning awakening: melatonin's half-life is too short to helpChronic primary insomnia: CBT-I is far superior to melatoninCommon mistakes:Taking 5-10 mg → receptor saturation + next-day grogginessSugar gummy children's melatoninTaking it at bedtime (too late)
剂量 · 形态 · 谁最受益
为什么褪黑素的用法这么反直觉?它不是让你睡着, 而是告诉全身夜晚从此刻开始。松果体在天黑后才开始分泌它, 血里的浓度慢慢爬升, 停靠到主时钟 suprachiasmatic nucleus: The brain's master clock — set by light, it runs the body's day–night rhythm. 以及下丘脑视前区那些细胞表面的 MT1 / MT2 受体上 —— MT1 那一路把维持清醒的驱动往下压, MT2 那一路负责把整块时钟往前或往后挪。
所以它的效果主要取决于你在哪个时刻吃, 而不是吃了多少: 在身体自己该分泌的时刻之前吃, 时钟被往前拽, 你会更早开始困; 已经躺下了才吞, 相当于在夜晚早就开始之后再宣布一次夜晚, 时钟收到的是一条矛盾信息, 甚至可能被往后推。剂量上的反直觉是同一个道理 —— 受体在很低的浓度就已经坐满, 再往上加并不会让这则公告更响, 只会让血里的褪黑素在天亮之后还没清干净, 那就是第二天早上那股拖不动的钝感。
下面是四件工具的具体用法。
真正有循证的营养型睡眠工具 (按证据排序):
① Magnesium (Mg) — B 级
机制: SERCA 钙泵回放松肌肉; NMDA 受体 Mg²⁺ 物理塞 (atlas `magnesium/relax` L4); GABA 调节剂量: 200-400 mg 元素 Mg 睡前形态: glycinate / threonate (透血脑屏障) / citrate > 氧化镁 (吸收差)预期: 入睡时间 ↓ 10-15 分钟 (RCT 混杂); 主观放松副作用: 大剂量腹泻 (自调机制); glycinate 形态温和谁受益最大: 已知 Mg 缺乏 + 抽筋 + 焦虑
② Glycine (甘氨酸) — B 级
机制: 体温调节 (体温下降是入睡触发) + GABA 调节 + atlas `glycine/metabolic-hub` L4剂量: 3 g 睡前 30-60 分钟预期: 入睡改善 + 主观睡眠质量 (Yamadera 2007 RCT)副作用: 极罕见谁受益最大: 浅睡 + 非恢复性睡眠
③ L-Theanine — B 级
机制: GABA + α 脑波 + 减焦虑 (atlas `caffeine-l-theanine/l-theanine` 已详)剂量: 200-400 mg 睡前预期: 主观放松 + 改善入睡 (focus on anxiety-driven onset)副作用: 几乎零谁受益最大: 焦虑型入睡难
④ Melatonin — A-B 级 (but only in specific use cases)
机制: atlas `melatonin/real-dose-vs-commercial` L4 详真适应症:入睡延迟 (sleep onset insomnia): 0.3-0.5 mg 睡前 30-60 分钟 (不是上床时)时差: 0.5-3 mg, 抵达后在接近当地就寝的时间 (约 22:00-24:00) 服 ×3-5 天 —— 不要提前到傍晚: Cochrane 那份 meta 明写着吃得太早会引起白天困倦并拖慢适应倒班 / DSPS: 0.3-0.5 mg 目标睡眠前 4-6 h (相位提前)完全失明 (non-24): 处方剂量不适应症:维持型失眠 + 早醒: melatonin 半衰期短, 帮不上慢性原发失眠: CBT-I 远优于 melatonin常见错误:吃 5-10 mg → 受体饱和 + 次日嗜睡吃含糖软糖儿童 melatonin上床时才吃 (太晚)
Useless / caution list + protocol
Useless / use-with-caution list:5-HTP: interacts with antidepressants / MAOIs, can trigger serotonin syndrome; limited safety dataValerian: most RCTs negative; rare but real reports of hepatotoxicityKava: hepatotoxic, banned in several countriesChamomile tea: can soothe mild anxiety, but does not really treat insomnia; the ritual > the direct pharmacologyLemon balm: similar to chamomile — ritual > pharmacologyCBD: evidence mixed; some patients respond but product quality varies; prescription Epidiolex is the exceptionAshwagandha: may help in chronic stress + anxiety-driven insomnia (detailed in atlas `ashwagandha/cortisol-mechanism`), but carries DILI risk — not for long-term use
Practical:
First line: lifestyle + CBT-IIf adding supplements: Mg 300 mg + Gly 3 g pre-bed is the safest and best-value combinationL-theanine 200 mg: add if anxiety is significantMelatonin: only when matched to an indication, at low doseDon't expect supplements to "cure" insomnia — at best they raise subjective quality by 10-20%
Chapter 4
Z-drugs + benzos · long-term risks
Z-drugs + benzos · long-term risks
The truth about OTC and prescription sleep medications:
① Z-drugs (Zopiclone / Eszopiclone / Zolpidem [Ambien])
Mechanism: selective GABA-A α1 receptor agonistsEffect: onset time ↓ 5-15 min; total sleep ↑ 30-50 min (short-term)Tolerance: begins at 2-4 weeks, obvious by 6-12 weeksDependence: moderate; abrupt discontinuation can cause rebound insomnia + anxietySide effects:Complex sleep behaviors (Ambien sleepwalking, sleep eating, sleep driving) — FDA black-box warningNext-day grogginess + cognitive slowing (especially in the elderly)Increased fall and hip fracture riskDementia — the strong observational signal here belongs to benzodiazepines, not Z-drugs: a case-control study found the association from three months of cumulative use onward (*Billioti de Gage 2014* *BMJ*). Z-drug-specific evidence is weaker and more mixed; do not transplant the benzodiazepine finding onto themClinical recommendation: limit to short-term (< 4 weeks), intermittent use; avoid in the elderly and in those with cognitive impairment
② Benzodiazepines (Diazepam / Alprazolam / Lorazepam / Temazepam)
Mechanism: non-selective GABA-A agonistsEffect: strong sedation + anxiolysisDependence: high — begins at 4-6 weeks, withdrawal can be severe (seizure risk)Side effects:Falls + hip fracturesDementia — this is where that signal actually lives: 1,796 first Alzheimer's diagnoses matched to 7,184 controls, with risk rising from three months of cumulative use onward (*Billioti de Gage 2014* *BMJ* — a case-control study, not a cohort)Respiratory depression (especially with alcohol or opioids)Long-term personality changes (emotional blunting, memory issues)Clinical recommendation: avoid as first-line for insomnia; for acute anxiety, short-term only (≤ 2 weeks); always taper
③ Antihistamines (Doxylamine / Diphenhydramine — Benadryl)
OTC + common sleep aids (Unisom, NyQuil, etc.)Side effects and risks:Anticholinergic → dry mouth, constipation, blurred vision, urinary retentionStrong long-term association with dementia (*Gray 2015* *JAMA Intern Med*)Next-day grogginess + balance issuesTolerance develops quicklyClinical recommendation: not recommended for chronic insomnia — long-term risk exceeds short-term benefit
④ Tricyclic antidepressants (Trazodone / Doxepin / Mirtazapine)
Used at low dose for insomnia (off-label for trazodone)Effect: moderate; slower onset than Z-drugsSide effects: fewer, but possible cardiac / weight / sexual function effectsClinical: a reasonable choice when insomnia and depression coexist
⑤ New orexin receptor antagonists (Suvorexant / Lemborexant / Daridorexant)
Mechanism: block orexin (the wakefulness maintenance system) → sleepiness emerges naturallyAdvantages: low dependence + novel mechanism + less cognitive impactPrice: high (often requires insurance authorization)Side effects: next-day grogginess + very rare sleep paralysisClinical: a new option, a better fit for chronic insomnia
Each of these five drug classes has its own indications and red lines — turn to the next page for the clinical decision table + how to taper after years of use.
① Z-drugs (Zopiclone / Eszopiclone / Zolpidem [Ambien])
Mechanism: selective GABA-A α1 receptor agonistsEffect: onset time ↓ 5-15 min; total sleep ↑ 30-50 min (short-term)Tolerance: begins at 2-4 weeks, obvious by 6-12 weeksDependence: moderate; abrupt discontinuation can cause rebound insomnia + anxietySide effects:Complex sleep behaviors (Ambien sleepwalking, sleep eating, sleep driving) — FDA black-box warningNext-day grogginess + cognitive slowing (especially in the elderly)Increased fall and hip fracture riskDementia — the strong observational signal here belongs to benzodiazepines, not Z-drugs: a case-control study found the association from three months of cumulative use onward (*Billioti de Gage 2014* *BMJ*). Z-drug-specific evidence is weaker and more mixed; do not transplant the benzodiazepine finding onto themClinical recommendation: limit to short-term (< 4 weeks), intermittent use; avoid in the elderly and in those with cognitive impairment
② Benzodiazepines (Diazepam / Alprazolam / Lorazepam / Temazepam)
Mechanism: non-selective GABA-A agonistsEffect: strong sedation + anxiolysisDependence: high — begins at 4-6 weeks, withdrawal can be severe (seizure risk)Side effects:Falls + hip fracturesDementia — this is where that signal actually lives: 1,796 first Alzheimer's diagnoses matched to 7,184 controls, with risk rising from three months of cumulative use onward (*Billioti de Gage 2014* *BMJ* — a case-control study, not a cohort)Respiratory depression (especially with alcohol or opioids)Long-term personality changes (emotional blunting, memory issues)Clinical recommendation: avoid as first-line for insomnia; for acute anxiety, short-term only (≤ 2 weeks); always taper
③ Antihistamines (Doxylamine / Diphenhydramine — Benadryl)
OTC + common sleep aids (Unisom, NyQuil, etc.)Side effects and risks:Anticholinergic → dry mouth, constipation, blurred vision, urinary retentionStrong long-term association with dementia (*Gray 2015* *JAMA Intern Med*)Next-day grogginess + balance issuesTolerance develops quicklyClinical recommendation: not recommended for chronic insomnia — long-term risk exceeds short-term benefit
④ Tricyclic antidepressants (Trazodone / Doxepin / Mirtazapine)
Used at low dose for insomnia (off-label for trazodone)Effect: moderate; slower onset than Z-drugsSide effects: fewer, but possible cardiac / weight / sexual function effectsClinical: a reasonable choice when insomnia and depression coexist
⑤ New orexin receptor antagonists (Suvorexant / Lemborexant / Daridorexant)
Mechanism: block orexin (the wakefulness maintenance system) → sleepiness emerges naturallyAdvantages: low dependence + novel mechanism + less cognitive impactPrice: high (often requires insurance authorization)Side effects: next-day grogginess + very rare sleep paralysisClinical: a new option, a better fit for chronic insomnia
Each of these five drug classes has its own indications and red lines — turn to the next page for the clinical decision table + how to taper after years of use.
五类药 · 各自的红线
OTC + 处方安眠药的真相: 五大类各有适应症和红线, 先记住每类一句话——干嘛的 + 红线在哪, 机制、剂量、耐受这些细节都放在下方深入页。① Z 药 (Zopiclone / Eszopiclone / Zolpidem / Ambien): 目前最常开的助眠药, 短期确实能把你放倒。红线是只适合短期 (< 4 周)——会耐受、会依赖、骤停反弹, 还有 FDA 黑框警告的复杂睡眠行为 (梦游、梦食、梦驾)。这里要把两件事分开: 黑框只有这一个主题 (外加曾发生过这类事件的人禁用), 跌倒和痴呆都不在黑框里 —— 它们来自观察性研究, 而且痴呆那条信号主要属于苯二氮卓, 不是 Z 药。老人和已有认知障碍者尤其要避开 (Z 药在 AGS Beers 清单上)。
② 苯二氮卓 (Diazepam / Alprazolam / Lorazepam / Temazepam): 更老的强镇静 + 抗焦虑药。红线是依赖性高——几周就上瘾, 戒断可能诱发癫痫, 还有跌倒、呼吸抑制 (与酒或阿片同用更危险)。痴呆那条关联的证据也长在这一类药上: 一项病例对照研究里, 累计使用 ≥ 3 个月起风险升高 (Billioti de Gage 2014 BMJ)。不作失眠首选, 用也顶多急性期 (≤ 2 周) 短用并阶梯撤药。
③ 抗组胺 (Doxylamine / Diphenhydramine / Benadryl): 很多 OTC 助眠药 (Unisom、NyQuil 等) 的成分, 靠让你犯困起效。红线是不适合慢性失眠——一堆抗胆碱副作用 (口干、便秘、视物模糊、尿潴留)、耐受快, 而且累计用得越多, 与后来发生痴呆的关联越强 (Gray 2015 JAMA Intern Med: 3434 名 65 岁以上成人的前瞻队列, 中位随访 7.3 年; 苯海拉明这类第一代抗组胺正是该队列里最常用的强抗胆碱药之一)。
④ 三环类抗抑郁药 (Trazodone / Doxepin / Mirtazapine): 小剂量用于助眠 (trazodone 属 off-label)。起效不算快, 但失眠合并抑郁时是合理选择; 要留意心律、体重、性功能方面的影响。
⑤ Orexin 受体拮抗剂 (Suvorexant / Lemborexant / Daridorexant): 最新一类, 靠关掉清醒信号 (orexin) 让睡意自然发生。依赖性低、对认知影响小, 比较适合慢性失眠; 缺点是贵 (常需保险审批), 偶有次日嗜睡、极罕见睡瘫。
这 5 类药物各有适应症与红线 —— 翻一页看临床决策表 + 已经服用多年怎么减.
Decision table + tapering long-term Z-drugs
Clinical decision table (5 scenarios, which tier):Acute < 4 weeks + major stress / jet lag: short-term Z-drug is acceptable + start CBT in parallelChronic insomnia: CBT-I first, medications adjunctive + intermittentInsomnia + severe depression / anxiety: treat the underlying condition + short-term adjunctInsomnia + sleep apnea: no sleep medications (they worsen breathing); use CPAP — see atlas `sleep-apnea`Elderly (> 65): avoid Z-drugs + benzos + Benadryl — the AGS 2023 Beers Criteria list benzodiazepines, Z-drug hypnotics and first-generation antihistamines all as potentially inappropriate in adults 65 and older, on the basis of cognitive impairment, delirium, falls and fractures. Prioritize CBT-I + nutritional tools
"I've been on Ambien for 5 years — what should I do?"
This is not uncommon, and clinicians see it often. A few principles:
Don't stop abruptly. Rebound insomnia + anxiety can be worse than the original symptoms; benzos additionally carry seizure riskTaper over 2-3 months, reducing by ~25 % every 1-2 weeks, which gives GABA-receptor upregulation time to occur. This is a common clinical pace, not a schedule fixed by any one guideline — the actual taper has to be set by the prescriber from your dose, drug and duration of useStart CBT-I in parallel — that is the real replacement mechanism, not another drugSubstitute supplements: Mg 300 mg + Gly 3 g pre-bed, plus L-theanine 200 mg if anxiety is in the mix, can soften the subjective withdrawal (atlas `insomnia/nutrient-tools` covers this)Don't feel ashamed — long-term Z-drug use is more a failure of physicians and the medical system to deliver CBT-I than your personal failure. Starting now, you can change it.
Atlas + Report closure
Atlas links back to:`caffeine-l-theanine/caffeine-pharm` L4 — adenosine pressure + how it interferes with the deep-sleep threshold`melatonin/real-dose-vs-commercial` L4 — 0.3 mg receptor saturation + the *Erland 2017* label chaos`ashwagandha/cortisol-mechanism` L4 — chronic stress + hypothalamic–pituitary–adrenal axis: The body's stress-response chain (hypothalamus → pituitary → adrenal) that releases cortisol. + cortisol diurnal disruption`magnesium/relax` L4 — SERCA + NMDA Mg²⁺ block`glycine/metabolic-hub` L4 — thermoregulation + GABA`nervous/neurotransmitters` L4 — where GABA / benzodiazepines / Z-drugs act
Atlas + Report: the report engine's `sleep-mag` / `caffeine-pm` / `screen-evening` / `mindfulness-suggest` / `melatonin-skeptic` rules all link back to this story — taking users from "the report says I might have insomnia" to "I understand why, and here is what I can actually do".
The simplest 4-week insomnia improvement plan:
Week 1:
Fix the wake time (even on weekends)Caffeine cutoff by 12 pm noon (or earlier for slow metabolizers)Blue-light control after 9 pmDownload a CBT-I appNote: everything in this week is sleep hygiene, and the one negative recommendation in the AASM guideline is precisely 'do not use sleep hygiene as a single-component therapy' — it clears obstacles, it doesn't treat. So the CBT-I app in week 2 is not optional
Week 2:
Add a sleep diary (record actual sleep time)Implement stimulus control (bed only for sleep + sex)Supplement: Mg 300 mg + Gly 3 g pre-bed
Week 3:
Assess sleep efficiencyIf < 85% → take these two weeks of sleep diary to a CBT-I therapist, or use the sleep restriction module inside a digital program — don't compress time in bed on your own. Daytime sleepiness rises sharply in the first week or two, and the driving / machinery risk is real; anyone with epilepsy or a seizure history, uncontrolled bipolar disorder, untreated sleep apnea or a parasomnia should especially not self-administer this stepAdd progressive muscle relaxation
Week 4:
Review + adjustIf improvement is limited → CBT-I therapist / sleep medicine
"Perfect sleep" is not the goal: an occasional bad night is not catastrophic — chronic insomnia is the real problem. Accepting imperfection is one of CBT-I's core wisdoms — "tonight I might sleep poorly, but life continues, and so will tomorrow".
Chapter 5
Decision tree + red flags
Decision tree + red flags
"I can't sleep — what do I do?" Step by step:
Week 1 · Self-check + basics:
Is it acute or chronic?Acute < 3 months + a trigger (stress / jet lag / illness) → usually self-limited; simple sleep hygiene + short-term supplementsChronic ≥ 3 months → start CBT-IBasic workup:Caffeine (afternoon / evening)Blue light (evening phone / TV)Alcohol (pre-bed)Exercise (too late)Bedroom environment (noise / light / temperature)
Weeks 2-4 · Start CBT-I + nutritional tools:
CBT-I: digital app or scheduled therapistMg 300 mg + Gly 3 g pre-bedL-theanine 200 mg if anxiousSleep diary
Weeks 4-8 · Assess:
Sleep efficiency ≥ 85%?Subjective improvement?Improved daytime function?
No improvement after 8 weeks: refer to sleep medicine specialty:
Rule out other causes (apnea / RLS / parasomnia / psychiatric)Polysomnography (PSG) if apnea is suspected
Red flags (ER / see a doctor immediately):
Loud snoring + apneic pauses + extreme daytime sleepiness: sleep apnea → severe CV / neurological riskRestless legs + irresistible urge to move: restless legs syndrome (RLS)Sleep terrors / violent sleep behaviors: REM behavior disorder (RBD) — can be an early Parkinson's prodromeSudden episodes of falling asleep during the day: narcolepsy — rare but seriousInsomnia + depression + thoughts of self-harm: immediate psychiatric careInsomnia + palpitations / rapid weight loss / heat intolerance: rule out hyperthyroidism
"Accepting" long-term insomnia:
There is no promise of "always sleeping 8 hours": even with treatment, an occasional bad night is normalGoal: most nights are enough + daytime is functional + overall quality of lifePerfectionist thinking worsens insomnia — one goal of CBT-I is to loosen this tension
One last point: don't let insomnia define you. Insomnia is an experience, not an identity. While you are working on improvement, keep doing the things that make life rich — work, relationships, interests — they are themselves part of managing insomnia.
Week 1 · Self-check + basics:
Is it acute or chronic?Acute < 3 months + a trigger (stress / jet lag / illness) → usually self-limited; simple sleep hygiene + short-term supplementsChronic ≥ 3 months → start CBT-IBasic workup:Caffeine (afternoon / evening)Blue light (evening phone / TV)Alcohol (pre-bed)Exercise (too late)Bedroom environment (noise / light / temperature)
Weeks 2-4 · Start CBT-I + nutritional tools:
CBT-I: digital app or scheduled therapistMg 300 mg + Gly 3 g pre-bedL-theanine 200 mg if anxiousSleep diary
Weeks 4-8 · Assess:
Sleep efficiency ≥ 85%?Subjective improvement?Improved daytime function?
No improvement after 8 weeks: refer to sleep medicine specialty:
Rule out other causes (apnea / RLS / parasomnia / psychiatric)Polysomnography (PSG) if apnea is suspected
Red flags (ER / see a doctor immediately):
Loud snoring + apneic pauses + extreme daytime sleepiness: sleep apnea → severe CV / neurological riskRestless legs + irresistible urge to move: restless legs syndrome (RLS)Sleep terrors / violent sleep behaviors: REM behavior disorder (RBD) — can be an early Parkinson's prodromeSudden episodes of falling asleep during the day: narcolepsy — rare but seriousInsomnia + depression + thoughts of self-harm: immediate psychiatric careInsomnia + palpitations / rapid weight loss / heat intolerance: rule out hyperthyroidism
"Accepting" long-term insomnia:
There is no promise of "always sleeping 8 hours": even with treatment, an occasional bad night is normalGoal: most nights are enough + daytime is functional + overall quality of lifePerfectionist thinking worsens insomnia — one goal of CBT-I is to loosen this tension
One last point: don't let insomnia define you. Insomnia is an experience, not an identity. While you are working on improvement, keep doing the things that make life rich — work, relationships, interests — they are themselves part of managing insomnia.
四件最常背锅的事 · 各自打在哪里
咖啡因: 睡眠压力靠一种叫腺苷的代谢残余积累 —— 醒着越久堆得越多, 堆够了你才困。咖啡因长得像腺苷, 能坐进同一个受体口袋却什么也不做, 于是压力照堆, 你却读不到。它在体内被拆解得很慢, 下午那杯到你上床时往往还剩相当一部分, 挡着的正是最该被读出来的那段压力。睡前酒: 它一开始确实是踩刹车 —— 增强 GABA 这套抑制信号, 所以很多人喝完倒头就着。但肝脏把酒精拆完之后, 被压住的兴奋会一次性弹回来: 后半夜反复醒、心跳快、出汗、口渴。它还会把前半夜的 REM 挤掉, 逼到后半夜补偿性地反弹, 于是梦变多、睡得碎。用酒助眠等于拿后半夜换前半夜, 而后半夜正是情绪与记忆被整理的时段。
夜里的亮屏: 眼底有一小群只管测亮度的感光细胞, 它们把外面有多亮送进下丘脑的主时钟 suprachiasmatic nucleus: The brain's master clock — set by light, it runs the body's day–night rhythm.。SCN 认定天还没黑, 就继续按住松果体, 褪黑素的起点被整体往后推 —— 你的困意也跟着整体后移。所以问题不只是内容太刺激, 是你的夜晚被推迟了。
太晚的高强度运动: 入睡需要核心体温往下掉, 而剧烈运动把体温和交感神经一起顶上去, 得等它们降回来你才睡得着。白天运动反而是助眠的 —— 它把清醒时段的代谢负荷做实, 让当晚的睡眠压力更足。
四条各自动的是不同的链子, 所以只戒咖啡、不改亮屏常常一点感觉都没有 —— 你修的不是坏掉的那一条。
分步方案 · 何时转诊 · 怎么和它相处
分步走的话, 大致是这样一个节奏:第 1 周 · 自查 + 基础:
是急性还是慢性?急性 < 3 月 + 有诱因 (压力 / 时差、病) → 多数自限; 简单睡眠卫生 + 短期补剂慢性 ≥ 3 月 → CBT-I 启动
基础排查:咖啡因 (午后、晚)蓝光 (晚上手机、电视)酒精 (睡前)运动 (太晚)卧室环境 (噪声、光 / 温度)
第 2-4 周 · 启动 CBT-I + 营养工具:
CBT-I: 数字应用或预约治疗师Mg 300 mg + Gly 3 g 睡前L-theanine 200 mg if 焦虑睡眠日记
第 4-8 周 · 评估:
睡眠效率 ≥ 85%?主观改善?白天功能改善?
8 周后无改善: 转睡眠医学专科:
排其它病因 (apnea / RLS / parasomnia / 心理疾病)多导睡眠图 (PSG) 若怀疑 apnea
长期失眠的接受:
没有永远睡 8 小时承诺: 即使治疗好, 偶尔一夜不好也正常目标: 大多数夜晚足够 + 白天能用 + 总体生活质量完美主义思维加重失眠 — CBT-I 的一个目标是松开这个紧绷
最后一点: 别让失眠定义你。失眠是经历, 不是身份。在改善的同时, 继续做让你觉得充实的事——工作、关系、兴趣——这些本身就是失眠管理的一部分。
References · 8
- Edinger, J. D., Arnedt, J. T., Bertisch, S. M., et al. (2021). Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. Journal of Clinical Sleep Medicine, 17(2), 255-262. 10.5664/jcsm.8986
- Trauer, J. M., Qian, M. Y., Doyle, J. S., Rajaratnam, S. M. W., & Cunnington, D. (2015). Cognitive behavioral therapy for chronic insomnia: a systematic review and meta-analysis. Annals of Internal Medicine, 163(3), 191-204. 10.7326/M14-2841
- Brzezinski, A., et al. (2005). Effects of exogenous melatonin on sleep: a meta-analysis. Sleep Medicine Reviews, 9(1), 41–50. 10.1016/j.smrv.2004.06.004
- Yamadera, W., Inagawa, K., Chiba, S., Bannai, M., Takahashi, M., & Nakayama, K. (2007). Glycine ingestion improves subjective sleep quality in human volunteers, correlating with polysomnographic changes. Sleep and Biological Rhythms, 5(2), 126-131. 10.1111/j.1479-8425.2007.00262.x
- US Food and Drug Administration. (2019). FDA adds boxed warning for risk of serious injuries caused by sleepwalking with certain prescription insomnia medicines (Drug Safety Communication, April 30, 2019). The boxed warning added to eszopiclone, zaleplon and zolpidem covers COMPLEX SLEEP BEHAVIOURS ONLY - sleepwalking, sleep-driving and other activities while not fully awake - after 66 reported cases of serious injury or death over 26 years. The agency also contraindicated these drugs in patients who have previously experienced such an episode. Falls, dependence and dementia are NOT part of the boxed warning. www.fda.gov/drugs/drug-safety-and-availability/fda-adds-boxed-warning-risk-serious-injuries-caused-sleepwalking-certain-prescription-insomnia
- By the 2023 American Geriatrics Society Beers Criteria Update Expert Panel. (2023). American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. Journal of the American Geriatrics Society, 71(7), 2052-2081. Benzodiazepines, non-benzodiazepine 'Z-drug' hypnotics and first-generation antihistamines are all listed as potentially inappropriate in adults 65 and older, on the basis of cognitive impairment, delirium, falls and fractures. 10.1111/jgs.18372
- Billioti de Gage, S., Moride, Y., Ducruet, T., Kurth, T., Verdoux, H., Tournier, M., Pariente, A., & Begaud, B. (2014). Benzodiazepine use and risk of Alzheimer's disease: case-control study. BMJ, 349, g5205. Case-control study in the Quebec health insurance database: 1,796 people with a first diagnosis of Alzheimer's disease matched to 7,184 controls. Benzodiazepine use of three months or more was associated with increased risk, and the association strengthened with longer cumulative exposure. 10.1136/bmj.g5205
- Gray, S. L., Anderson, M. L., Dublin, S., Hanlon, J. T., Hubbard, R., Walker, R., Yu, O., Crane, P. K., & Larson, E. B. (2015). Cumulative use of strong anticholinergics and incident dementia: a prospective cohort study. JAMA Internal Medicine, 175(3), 401-407. Prospective cohort of 3,434 adults aged 65+ followed a median 7.3 years. Higher cumulative anticholinergic exposure was associated with incident dementia; first-generation antihistamines such as diphenhydramine were among the most commonly used strong anticholinergics in the cohort. 10.1001/jamainternmed.2014.7663