Place · Level 3
Andropause · Late-Onset Hypogonadism
30 后 T 缓慢降 1-2%/年 · 真低 T < 300 + 症状 · 测量陷阱 + SHBG · 生活方式优先 · TRT 适应症
Last updated
Story path
Chapter 1
Not female menopause analog
Not female menopause analog
'Andropause' / 'male menopause' is a popular phrase but a poor clinical term: male hormone changes are fundamentally different from female menopause.
Female menopause (Atlas `perimenopause`):
A cliff: estrogen drops 90%+ over a few yearsClear endpoint: 12 months with no periodUniversal: all women experience it
Male hormones (LOH, late-onset hypogonadism):
Slow linear decline: testosterone falls 1–2% per year after age 30No clear endpoint; most men still have measurable testosterone at 80Not universal; only ~2% of men meet the true clinical definition (EMAS, Wu 2010 NEJM)
EMAS true LOH definition (European Male Aging Study, n=3,369):
Total T < 11 nmol/L (~320 ng/dL) + free T < 220 pmol/LPlus 3 sexual symptoms: reduced libido + reduced morning erections + erectile dysfunctionBoth criteria met → true LOH in ~2.1% of men aged 40–79Single criterion is not enough: low T without symptoms doesn't count; symptoms without low T don't either
Why does this matter?
'T clinics' + anti-aging marketing in China + US massively expand the LOH conceptA 'man tired / low energy / can't lose weight' bucket — 90% has nothing to do with testosteroneEven with low T, lifestyle factors (obesity / sleep apnea / chronic opioids / depression) account for ~90% and are reversible
Normal testosterone range (healthy adult male):
Total testosterone: 264–916 ng/dL (slight lab variation)Free T: 9–30 ng/dL (Vermeulen calculation)Morning draw (7–10 am) is the daily peak — a single low reading doesn't countRepeat at least twice, 4–6 weeks apart, to rule out transient swings
'Is a slow T decline a disease?'
Not necessarily. Compared to age 35, a normal 65-year-old man has T roughly 25–50% lower — that's physiology. The key is that symptoms + labs align before considering intervention:
Low T without symptoms: monitor, don't treatMid-range T (300–400 ng/dL) with symptoms: lifestyle first (next step)Clear LOH (low T + 3 sexual symptoms): endocrinology / urology evaluation
Female menopause (Atlas `perimenopause`):
A cliff: estrogen drops 90%+ over a few yearsClear endpoint: 12 months with no periodUniversal: all women experience it
Male hormones (LOH, late-onset hypogonadism):
Slow linear decline: testosterone falls 1–2% per year after age 30No clear endpoint; most men still have measurable testosterone at 80Not universal; only ~2% of men meet the true clinical definition (EMAS, Wu 2010 NEJM)
EMAS true LOH definition (European Male Aging Study, n=3,369):
Total T < 11 nmol/L (~320 ng/dL) + free T < 220 pmol/LPlus 3 sexual symptoms: reduced libido + reduced morning erections + erectile dysfunctionBoth criteria met → true LOH in ~2.1% of men aged 40–79Single criterion is not enough: low T without symptoms doesn't count; symptoms without low T don't either
Why does this matter?
'T clinics' + anti-aging marketing in China + US massively expand the LOH conceptA 'man tired / low energy / can't lose weight' bucket — 90% has nothing to do with testosteroneEven with low T, lifestyle factors (obesity / sleep apnea / chronic opioids / depression) account for ~90% and are reversible
Normal testosterone range (healthy adult male):
Total testosterone: 264–916 ng/dL (slight lab variation)Free T: 9–30 ng/dL (Vermeulen calculation)Morning draw (7–10 am) is the daily peak — a single low reading doesn't countRepeat at least twice, 4–6 weeks apart, to rule out transient swings
'Is a slow T decline a disease?'
Not necessarily. Compared to age 35, a normal 65-year-old man has T roughly 25–50% lower — that's physiology. The key is that symptoms + labs align before considering intervention:
Low T without symptoms: monitor, don't treatMid-range T (300–400 ng/dL) with symptoms: lifestyle first (next step)Clear LOH (low T + 3 sexual symptoms): endocrinology / urology evaluation
机制 · 睾酮从哪来, 又在身体里干什么
后面几幕的每一条, 都挂在同一条链子上。先把这条链子看清楚: 睾酮是谁下单造的, 造出来之后又去了哪里、干了什么。下单的是脑, 不是睾丸
脑底部的下丘脑每隔一阵子放出一小股信号 (GnRH), 像敲门一样敲垂体。垂体收到敲门, 往血里放两种激素 —— 黄体生成素 (LH) 和卵泡刺激素 (FSH)。LH 顺着血流到睾丸, 停在一类叫 Leydig 的细胞上; 这类细胞才是真正的车间, 它拿胆固醇当原料, 一步步改造成睾酮。FSH 走的是另一条线, 它作用在睾丸里的支持细胞 (Sertoli), 那是精子发育的苗床。
造出来的睾酮会回头按住上游
血里的睾酮, 以及它在身体各处被改造成的雌二醇, 会回头压住下丘脑放信号的频率、也压住垂体的响应。这就是负反馈。这条回路是理解这一整篇的钥匙 —— 后面那些看起来不相干的现象, 肥胖为什么压低睾酮、外源睾酮为什么关掉生精、为什么想生育的人不能直接打睾酮, 走的都是这同一条回路。
它是怎么进细胞干活的
睾酮是脂溶性的, 不用在细胞门口排队, 直接穿过细胞膜进去, 在细胞里遇到一个专门等着它的接头蛋白 (雄激素受体)。两者一结合, 受体形状变了, 带着睾酮钻进细胞核, 停在 DNA 的特定位置上, 打开一组基因。
所以睾酮不是给你能量, 它是改动一批细胞往后几天在生产什么。这也是为什么它的效果从来不是立刻的 —— 从基因被打开, 到蛋白造出来、组织发生变化, 中间隔着好几天到几周。
在肌肉里: 受体打开的是收缩蛋白的生产线, 同时唤醒贴在肌纤维旁边、平时处于待命状态的卫星细胞, 让它们并进肌纤维、贡献出新的细胞核。睾酮长期不足, 这条生产线的底噪就低一档: 同样的训练量, 肌肉守得更吃力, 掉得更快。
在骨里: 骨组织里有一种叫芳香化酶的酶, 会把送到的睾酮就地改造成雌二醇。雌二醇按住的是破骨细胞 —— 负责把旧骨拆掉的那一方。上游供货一少, 拆的一方就松了手刹, 骨量被一点点拆走。留意这个巧合: 同一个酶, 在脂肪里把睾酮改成雌二醇是坏消息, 在骨里恰恰是它在替你守骨量。位置不同, 同一个化学反应的意义就完全不同。
在脑里: 下丘脑和边缘系统那一带雄激素受体最密, 它们管的是性欲的中枢驱动, 也管夜里那一轮自发勃起。这正好解释了后面欧洲男性老化研究 (EMAS) 为什么把三条硬症状全挑在性相关的一侧 —— 不是研究者偏心, 而是脑内受体密度本来就分布在那里。至于没精力、脑子不清楚, 那几件事在脑里的经手人是睡眠、情绪和血糖, 不是睾酮。这一条先记下, 后面判断症状真假全靠它。
在骨髓里: 睾酮会推高造红细胞的信号, 男性血红蛋白平均高于女性有它一份功劳。记住这条, 后面 TRT 为什么要盯红细胞压积、为什么有人需要放血, 就不用死记了。
在皮肤和前列腺: 睾酮走到这些地方还会被一种还原酶再改造一次, 变成活性更强的雄激素 (DHT), 管的是胡须、体毛、头顶脱发和前列腺体积。同一个分子, 到了不同地方被改造成不同的样子, 干不同的活 —— 这也是为什么补睾酮从来不是只影响一件事。
对照 · 断崖与缓坡, 以及 EMAS 的双条件
女性绝经 (atlas `perimenopause`):断崖: 几年内雌激素跌 90%+明确终点: 12 月无月经普遍: 所有女性都经历
男性激素 (LOH, late-onset hypogonadism):
缓慢线性下降: 30 岁后睾酮 1-2%/年无明确终点, 大多数男性 80 岁仍有可测睾酮不普遍, 仅约 2% 男性满足真临床定义 (EMAS, Wu 2010 NEJM)
两条曲线的形状不同, 是因为坏掉的地方不同。女性那边是卵泡库存被用尽, 一个有限的、会耗光的库存; 男性这边没有库存概念, 睾丸是按订单现产现供的车间, 老化只是让车间的工人变少、对订单的反应变钝。耗光和变钝, 是两种完全不同的曲线。
真 LOH 怎么界定 (欧洲男性老化研究 EMAS): 必须同时满足两条——睾酮 (总睾酮和自由睾酮) 确实偏低, 而且有 3 种性相关症状 (性欲减退 + 晨勃减少 + 勃起功能下降)。两条都占, 40-79 岁男性里也只有约 2.1% 算真 LOH。关键就在同时二字: 睾酮低但没症状不算, 有症状但睾酮不低也不算。
为什么非要两条都占? 因为这两条线索单独拿出来都不可靠。只看数字: 睾酮一天之内自己就能上下晃一大截, 抽血抽在下午, 能把一个完全正常的人测成偏低。只看症状: 性欲和勃起同时还被血管、情绪、药物、睡眠管着, 其中任何一样出问题都能做出一模一样的表现。两条各自都会出错、而且出错原因互不相干的线索同时指向同一个方向, 才把巧合排掉。门槛一加上去, 人数就从看起来普遍塌到个位数百分比 —— 塌掉的那部分不是被漏诊了, 是本来就不该被算进来。
为什么这点重要?
中国 + 美国的T 诊所 + anti-aging 营销把 LOH 概念大幅扩大即使 T 偏低, 生活方式因素 (肥胖、睡眠呼吸暂停、长期阿片、抑郁) 90% 可逆
Symptoms — signal or noise?
The 'low T' symptom list is long, but most have weak association with testosterone.Truly T-related (EMAS + Wu 2010):
Reduced libidoReduced or absent morning erectionsReduced erectile function (ED)Muscle loss (also linked to training + protein intake)Reduced bone density (in T deficiency)
Weakly associated / multifactorial (often exaggerated by T clinics):
Fatigue: sleep / depression / chronic illness / iron deficiency / hypothyroidism are more likelyBrain fog: sleep / stress / nutritionDifficulty losing weight: diet + exercise + sleep + metabolic healthLow mood: depression + anxiety + sleepAttention problems: sleep / ADHD / stress
Key question: 'low T first, or these conditions first?'
Mostly the reverse: obesity / depression / sleep apnea directly lower testosterone, not the other way around.
Obesity: adipose-tissue aromatase converts T to estrogen + leptin feedback, T ↓ (after 5–10% weight loss, T ↑ ~15–25%)OSA (sleep apnea): hypoxia + fragmented sleep suppresses T synthesis (T rebounds after CPAP treatment)Depression + chronic stress: high cortisol suppresses the HPG axis, T ↓Long-term opioid analgesics: directly suppress the HPG axis → central low TMetabolic syndrome / T2D: SHBG ↓, so apparent total T is low (free T may still be normal)
In practice:
Symptom checklist + physical exam + lifestyle account for 80% of the evaluationLabs: morning, repeat, and concurrently measure LH + FSH + SHBG + estradiol + PSADon't jump to 'measure T then start TRT' — that's a T-clinic sales pitch, not evidence-based medicine
数字 · 正常范围, 以及缓慢下降算不算病
先说这几个数字该怎么用: 它们是拿来排除的, 不是拿来确诊的。一个落在区间里的值, 基本可以让你把睾酮这条线放下、去查别的原因; 一个落在区间外的值, 只是让你接着往下查, 它自己什么都还没证明。正常睾酮范围 (健康成年男性)
总睾酮 (Total T): 264-916 ng/dL (实验室略不同)自由 T: 9-30 ng/dL (Vermeulen 计算法)早晨 (7-10 am) 采血是一天中最高, 单次低不算重复测量至少 2 次, 间隔 4-6 周, 排除一过性波动
区间这么宽本身就是一条信息: 同一个年龄段的健康男性, 高的那头能是低的那头的好几倍, 而两个人都没病。所以我比别人低并不构成诊断; 有意义的问题是你有没有掉出自己原来的水平, 以及有没有对得上的症状。
T 缓慢下降是病吗?
不一定。与 35 岁相比, 65 岁正常男性 T 平均低 25-50%, 这是生理变化。关键是症状 + 实验室一致才考虑干预:
没症状的低 T: 监测, 不必治有症状的中等 T (300-400 ng/dL): 生活方式优先 (下一步)明确 LOH (低 T + 三性症状): 内分泌、泌尿评估
把这三行连起来看, 会发现一条统一的逻辑: 治疗针对的是症状, 化验值只是用来确认这个症状值不值得往激素上归因。一个没有症状的低数字, 治好了数字也没有东西可以变好; 而一个有症状但数字正常的人, 补上去只会让上游把自己的产量往下调。这就是为什么指南把两条一起写, 而不是只写一条阈值。
Chapter 2
Measure & confounders
Measure & confounders
Measuring testosterone looks simple but has multiple traps. Understanding them avoids 80% of 'low T misdiagnoses'.
Trap 1 · Time:
T has a strong circadian rhythm — morning (7–10 am) peak, afternoon and evening 30–40% lowerMust draw in the morningThis is a clinical-guideline requirement (Endocrine Society 2018 / AUA 2018)
Trap 2 · Single measurement:
T has high biological variation — the same man may show 600 one day, 350 the nextGuidelines: ≥ 2 draws, 4–6 weeks apart, both low to count
Trap 3 · Acute effects:
Acute illness (flu / cold / post-major-surgery): acute T dropIntense training + weight loss: transient T dropDon't diagnose during these periods
Trap 4 · Total T vs free T (SHBG decides everything):
Total T = tightly SHBG-bound (~44%) + loosely albumin-bound (~54%) + free (~2%)Free T is the biologically active fraction; free + albumin-bound = bioavailable TSHBG effects:SHBG ↑ (low insulin / estrogen / hyperthyroidism): total T high but free T may be normalSHBG ↓ (IR / T2D / obesity / hypothyroidism): total T low but free T may be normalThe PCOS mirror (Atlas `pcos/insulin-resistance` L4): female IR → SHBG ↓ → free T ↑ → androgen symptoms2025 consensus: borderline total T (200–400 ng/dL) should be paired with SHBG + calculated free T (Vermeulen)
Trap 5 · Lab method differences:
Direct immunoassay is cheap but inaccurate, especially at low valuesLC-MS/MS (liquid chromatography mass spectrometry) is the gold standard with good inter-lab consistencyUse LC-MS/MS preferentially; don't rely on a single immunoassay low result for diagnosis
5 reversible confounders (must be addressed before evaluation):
1. Obesity:
BMI up by 5 → T down ~100 ng/dLVisceral fat aromatase converts T → estradiolIntervention: 5–10% weight loss → T ↑ ~15–25% (Camacho 2013)
2. Sleep apnea (OSA):
Moderate–severe OSA: T 20–30% below age-matched peersIntervention: STOP-BANG questionnaire + polysomnography + CPAP, T rebounds within months
3. Long-term opioids:
Directly suppress GnRH, central low T75% of chronic opioid users have low TIntervention: discuss reduction or switching with the pain team
4. Depression + chronic stress:
Cortisol rises, suppressing the HPG axisIntervention: psychotherapy + SSRI / SNRI + CBT
5. Chronic disease (T2D / chronic liver disease / HIV / CKD)
Intervention: control the primary disease
In practice: before pursuing TRT, these 5 confounders should be addressed for 6–12 months. Most 'low T' returns to normal once the confounders are resolved.
Trap 1 · Time:
T has a strong circadian rhythm — morning (7–10 am) peak, afternoon and evening 30–40% lowerMust draw in the morningThis is a clinical-guideline requirement (Endocrine Society 2018 / AUA 2018)
Trap 2 · Single measurement:
T has high biological variation — the same man may show 600 one day, 350 the nextGuidelines: ≥ 2 draws, 4–6 weeks apart, both low to count
Trap 3 · Acute effects:
Acute illness (flu / cold / post-major-surgery): acute T dropIntense training + weight loss: transient T dropDon't diagnose during these periods
Trap 4 · Total T vs free T (SHBG decides everything):
Total T = tightly SHBG-bound (~44%) + loosely albumin-bound (~54%) + free (~2%)Free T is the biologically active fraction; free + albumin-bound = bioavailable TSHBG effects:SHBG ↑ (low insulin / estrogen / hyperthyroidism): total T high but free T may be normalSHBG ↓ (IR / T2D / obesity / hypothyroidism): total T low but free T may be normalThe PCOS mirror (Atlas `pcos/insulin-resistance` L4): female IR → SHBG ↓ → free T ↑ → androgen symptoms2025 consensus: borderline total T (200–400 ng/dL) should be paired with SHBG + calculated free T (Vermeulen)
Trap 5 · Lab method differences:
Direct immunoassay is cheap but inaccurate, especially at low valuesLC-MS/MS (liquid chromatography mass spectrometry) is the gold standard with good inter-lab consistencyUse LC-MS/MS preferentially; don't rely on a single immunoassay low result for diagnosis
5 reversible confounders (must be addressed before evaluation):
1. Obesity:
BMI up by 5 → T down ~100 ng/dLVisceral fat aromatase converts T → estradiolIntervention: 5–10% weight loss → T ↑ ~15–25% (Camacho 2013)
2. Sleep apnea (OSA):
Moderate–severe OSA: T 20–30% below age-matched peersIntervention: STOP-BANG questionnaire + polysomnography + CPAP, T rebounds within months
3. Long-term opioids:
Directly suppress GnRH, central low T75% of chronic opioid users have low TIntervention: discuss reduction or switching with the pain team
4. Depression + chronic stress:
Cortisol rises, suppressing the HPG axisIntervention: psychotherapy + SSRI / SNRI + CBT
5. Chronic disease (T2D / chronic liver disease / HIV / CKD)
Intervention: control the primary disease
In practice: before pursuing TRT, these 5 confounders should be addressed for 6–12 months. Most 'low T' returns to normal once the confounders are resolved.
机制 · 五个测量陷阱, 每一个背后的道理
测量 T 看起来简单, 实际有多个陷阱。理解它们能避免 80% 的低 T 误诊。陷阱 1 · 时间
T 有显著昼夜节律, 早晨 (7-10 am) 最高, 下午到晚上下降 30-40%必须早晨抽血这是临床指南要求 (Endocrine Society 2018 / AUA 2018)
昼夜节律不是玄学, 它是上游的节奏决定的: 下丘脑放信号是一小股一小股的脉冲, 这些脉冲在夜间睡眠时最密集, 垂体跟着一波波发 LH, 睾丸也就在后半夜到清晨这段产量最高。醒着活动一整天之后脉冲变稀, 数字自然回落。所以下午抽的血不是抽到了一个更真实的你, 是抽到了这条节律的谷底。 顺带也解释了为什么长期睡不好会直接把睾酮拉低 —— 一天里最高产的那个窗口被撕碎了。
陷阱 2 · 一次性测量
T 生物变异大, 同一男性两天可测出 600 / 350指南要求 ≥ 2 次抽血间隔 4-6 周, 都低才算
陷阱 3 · 急性影响
急性病 (流感、感冒、大手术后): T 急性下降大强度训练 + 体重减少: T 暂时下降这些时候不能做诊断
急性期下降是有道理的: 身体在应付感染或创伤时会把资源从长线建设上撤走, 而睾酮正是那个指挥合成的信号, 首当其冲被调低。换句话说, 那一刻的低值忠实地反映了你正在生病, 但它没法告诉你病好之后的基线在哪。
陷阱 4 · 总 T vs 自由 T (SHBG 决定一切)
总 T = SHBG 紧密结合 (~ 44%) + 白蛋白松散结合 (~ 54%) + 自由 (~ 2%)自由 T 是生物活性的那一份; 自由 + 白蛋白结合 = 生物可及睾酮 (bioavailable T)SHBG 影响:SHBG ↑ (低胰岛素、雌激素、甲亢): 总 T 高但自由 T 可能正常SHBG ↓ (IR / T2D / 肥胖、甲减): 总 T 低但自由 T 可能正常PCOS 的镜像 (atlas `pcos/insulin-resistance` L4): 女性 IR → SHBG ↓ → 自由 T ↑ → 雄激素症状2025 共识: 总 T 边缘 (200-400 ng/dL) 应测 SHBG + 计算自由 T (Vermeulen)
为什么被 SHBG 拽住的那份就不算数? 因为睾酮要干活, 必须穿过细胞膜、进到细胞里找它的受体, 而挂在 SHBG 上的那部分体积太大、结合太紧, 走不进去。白蛋白那一批不一样: 它抓得松, 血流过组织时就能松手, 所以它和自由的那 2% 一起被算作生物可及的部分 —— 这也是为什么除了自由的都不算数是个流传很广的误说。可以把 SHBG 想成血液里的一队搬运工: 搬运工手上的货不算库存, 只有卸到货架上的才能被取用。搬运工人数一变, 库存总量的读数就跟着变, 但货架上的货可能一直没动。这就是为什么总睾酮和自由睾酮能指向相反的结论, 也是为什么胖的人特别容易被总睾酮误判成低。
陷阱 5 · 实验室差异
直接免疫法便宜但准确度差, 尤其低值LC-MS/MS (液相质谱) 是金标准, 实验室间一致性好优先用 LC-MS/MS, 不应只凭单一免疫法低值结果做诊断
把五个陷阱排在一起看, 会发现它们指向同一件事: 一个睾酮数字只有在配套条件说清楚之后才有意义 —— 什么时候抽的、抽了几次、当时身体什么状态、用哪种方法测的、有没有同时测 SHBG。缺了这些, 那个数字就只是一个数字。
临床 · 五大可逆混杂, 各自压的是哪一环
上一页讲的是怎么把数字测准。这一页讲的是: 就算测准了, 那个低值多半也不是睾丸自己坏了, 而是下面这几件事之一在上游踩着刹车。每一件都能松开。5 大可逆混杂因素 (评估前应先排查)
1. 肥胖
BMI 升 5 单位, T 下降约 100 ng/dL内脏脂肪芳香化酶把 T 转成雌二醇干预: 减重 5-10% 后 T ↑ 约 15-25% (Camacho 2013)
这一条要看成一个圈, 不是一条线。脂肪组织里有芳香化酶, 它把路过的睾酮就地改造成雌二醇; 而雌二醇正是负反馈上最灵敏的那把刹车, 它回头按住下丘脑的信号频率、也按住垂体发 LH 的量。LH 一少, 睾丸接到的订单就少, 睾酮更低; 睾酮低又让人更难守住肌肉、更容易堆内脏脂肪, 脂肪多一点, 芳香化酶就多一点, 刹车踩得更深。所以减重在这里不是顺便有好处, 它是直接把踩着刹车的那只脚挪开。
2. 睡眠呼吸暂停 (OSA)
中度 - 重度 OSA, T 比同龄低 20-30%干预: STOP-BANG 问卷 + 多导睡眠图 + CPAP, T 几个月内回升
接着上一页那条节律讲: 睾酮的产量高峰挂在完整的深睡眠上。OSA 的人整夜被反复的气道塌陷唤醒, 睡眠被切成碎片, 再叠上一次次的缺氧 —— 一天里最高产的那个窗口等于被拆掉了。所以他们的低睾酮不是另外一个病, 是同一个睡眠问题在化验单上的投影; 把气道撑开, 窗口回来, 数字也跟着回来。
3. 长期阿片
直接抑制 GnRH, 中央性 T 低慢性 opioid 用户 75% T 低干预: 与疼痛科讨论减阿片或转其它
阿片作用的位置在最上游: 它按住的是下丘脑那台放脉冲的发生器。指令根本没发出去, 垂体不发 LH, 睾丸完好无损却一直没收到订单。这就是中央性三个字的意思 —— 车间没坏, 是订单没来。怎么分辨? 看 LH: 这种情况下 LH 不会升高, 最后一幕会把这条判据讲透。
4. 抑郁 + 慢性压力
皮质醇升, 抑制 HPG干预: 心理治疗 + SSRI / SNRI + CBT
压力走的是和阿片很近的一条路: 长期偏高的皮质醇同样压住下丘脑那台脉冲发生器。身体的逻辑其实很直白 —— 处在持续有威胁的状态里, 繁殖和长肌肉都不是当务之急, 于是整条轴被调低。这也解释了为什么这类人的睾酮往往在压力源移开之后自己回来, 不需要补。
5. 慢性病 (T2D / 慢性肝病 / HIV / CKD)
干预: 控制原发病
这一组多半是从 SHBG 那一头动读数的: 胰岛素抵抗和肝脏问题会让 SHBG 变少, 于是总睾酮显得低, 而自由睾酮未必低。测了 SHBG、算出自由睾酮, 这一类假警报大部分能当场排掉。
把五条并排看, 它们其实分布在同一条链子的不同环上: 肥胖动的是负反馈的刹车, OSA 动的是产量的时间窗, 阿片和压力动的是最上游的脉冲发生器, 慢病动的是血里的运输蛋白。没有一条是睾丸这个车间自己坏了 —— 这正是它们可逆的原因。
实操: 真要 TRT 之前, 这 5 个混杂建议先处理 6-12 个月。大多数低 T 在解决混杂后回到正常。
Chapter 3
Lifestyle — cheapest, strongest
Lifestyle — cheapest, strongest
Lifestyle interventions for testosterone are often more impactful than TRT, are free of side effects, and yet are systematically underweighted by T clinics.
1. Weight loss (especially abdominal):
5–10% body-weight loss → T ↑ ~15–25%Waist −5 cm matters more than a BMI dropMechanism: less visceral-fat aromatase, leptin resistance reverses, IR improvesMethod: any sustainable caloric deficit + protein preservation
2. Resistance training (strength training):
2–3× per week, big compound lifts, progressive loadAcute post-workout T rises 15–30% (short-term) + long-term muscle preservationDon't chase YouTube 'T-boosting workout' marketing — standard strength training is enoughSynergistic with Atlas `protein/muscle` L4 (Leu → mechanistic target of rapamycin: The cell's master 'grow / build' switch — turned on by enough protein and resistance training. → MPS)
3. Sleep:
< 5 h/night sustained for 1 week → T ↓ 10–15%OSA treatment (CPAP) is the single most powerful intervention in confirmed cases7–9 h/night is the baseline
4. Stress management:
Chronic high cortisol suppresses the HPG axisMindfulness, diaphragmatic breathing, exercise, and social connection all lower cortisolPsychotherapy (CBT) helps stress + depression, and T rebounds indirectly
5. Moderate alcohol + no smoking:
Chronic heavy alcohol (> 21 units/week): T ↓ + estrogen ↑Smoking acutely raises T but is overall harmful (CV + cancer)
Nutrition: there is no 'T-miracle food', but ensure:
Zinc adequacy (RDA 11 mg; only supplement if deficient; oysters / red meat / pumpkin seeds)Vitamin D adequacy (25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. > 30 ng/mL; correct deficiency)Healthy fats (olive oil / fish / nuts): an ultra-low-fat diet lowers TProtein 1.2–1.6 g/kg (Atlas `protein/muscle`)No need for keto — no evidence supports it
No benefit / marketing traps:
Ashwagandha: marginal effect in chronic-stress populations, not a T boosterMaca: subjective libido improvement, no notable effect on TTribulus: multiple RCTs negativeZMA (Zn + Mg + B6 combo): no benefit in healthy people'Testosterone booster' multi-ingredient supplements: D-level evidence overallDHEA: not recommended as a T booster (side effects / inconsistent conversion)
Bottom line: most 40–60-year-old men's 'low energy / can't lose weight' improves dramatically after 6 months of lifestyle intervention, without needing to head down the TRT road.
1. Weight loss (especially abdominal):
5–10% body-weight loss → T ↑ ~15–25%Waist −5 cm matters more than a BMI dropMechanism: less visceral-fat aromatase, leptin resistance reverses, IR improvesMethod: any sustainable caloric deficit + protein preservation
2. Resistance training (strength training):
2–3× per week, big compound lifts, progressive loadAcute post-workout T rises 15–30% (short-term) + long-term muscle preservationDon't chase YouTube 'T-boosting workout' marketing — standard strength training is enoughSynergistic with Atlas `protein/muscle` L4 (Leu → mechanistic target of rapamycin: The cell's master 'grow / build' switch — turned on by enough protein and resistance training. → MPS)
3. Sleep:
< 5 h/night sustained for 1 week → T ↓ 10–15%OSA treatment (CPAP) is the single most powerful intervention in confirmed cases7–9 h/night is the baseline
4. Stress management:
Chronic high cortisol suppresses the HPG axisMindfulness, diaphragmatic breathing, exercise, and social connection all lower cortisolPsychotherapy (CBT) helps stress + depression, and T rebounds indirectly
5. Moderate alcohol + no smoking:
Chronic heavy alcohol (> 21 units/week): T ↓ + estrogen ↑Smoking acutely raises T but is overall harmful (CV + cancer)
Nutrition: there is no 'T-miracle food', but ensure:
Zinc adequacy (RDA 11 mg; only supplement if deficient; oysters / red meat / pumpkin seeds)Vitamin D adequacy (25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. > 30 ng/mL; correct deficiency)Healthy fats (olive oil / fish / nuts): an ultra-low-fat diet lowers TProtein 1.2–1.6 g/kg (Atlas `protein/muscle`)No need for keto — no evidence supports it
No benefit / marketing traps:
Ashwagandha: marginal effect in chronic-stress populations, not a T boosterMaca: subjective libido improvement, no notable effect on TTribulus: multiple RCTs negativeZMA (Zn + Mg + B6 combo): no benefit in healthy people'Testosterone booster' multi-ingredient supplements: D-level evidence overallDHEA: not recommended as a T booster (side effects / inconsistent conversion)
Bottom line: most 40–60-year-old men's 'low energy / can't lose weight' improves dramatically after 6 months of lifestyle intervention, without needing to head down the TRT road.
清单 · 五条干预与各自的量级
1. 减重 (尤其腹脂)5-10% 体重减少, T 上升约 15-25%腹围 -5 cm 比 BMI 下降更重要机制: 内脏脂肪芳香化酶减少, leptin 抵抗回正, IR 改善方法: 任何可持续的热量赤字 + 蛋白维持
2. 阻力训练 (力量训练)
每周 2-3 次, 大复合动作, 进步性负荷训练后急性 T 上升 15-30% (短期), 长期肌量保护不必跟 youtube T 提升训练 营销, 标准力量训练就够与 atlas `protein/muscle` L4 协同 (Leu → mechanistic target of rapamycin: The cell's master 'grow / build' switch — turned on by enough protein and resistance training. → MPS)
3. 睡眠
每晚 < 5 h 持续 1 周, T 下降 10-15%OSA 治疗 (CPAP) 是确诊者最有力的单一干预每晚 7-9 h 是基线
4. 压力管理
慢性高皮质醇会抑制 HPG正念、横膈呼吸、运动、社交都能降皮质醇心理咨询 (CBT) 对压力 + 抑郁有帮助, T 间接回升
5. 节制酒 + 不抽烟
慢性高酒精 (> 21 单位/周): T 降 + 雌激素升抽烟急性会让 T 升, 但总体不利 (CV + 癌)
营养上: 没有T 神食物, 但建议确保
Zn 充足 (RDA 11 mg, 缺乏才补; 牡蛎、红肉、南瓜籽)D 充足 (25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. > 30 ng/mL, 缺乏纠正)健康脂肪 (橄榄油、鱼 / 坚果): 极低脂饮食会让 T 降蛋白质 1.2-1.6 g/kg (atlas `protein/muscle`)不必生酮, 没有循证支持
读这些数字时有几个地方值得停一下。
训练后那一阵上升不是你的日常水平。它维持得很短, 属于运动当下的应激反应; 真正值钱的是长期练下来守住的肌肉量 —— 肌肉多, 胰岛素敏感性好, 内脏脂肪少, 脂肪里那把芳香化酶刹车就轻。所以力量训练对睾酮的贡献是绕道走的, 不是那一阵峰值。
极低脂饮食为什么反而降睾酮, 也不神秘: 睾酮的起点就是胆固醇, Leydig 细胞是拿胆固醇一步步改造出睾酮的。原料长期供不上, 车间开工自然受限。但反过来不成立 —— 原料够用之后再往上加, 并不会多造出睾酮来, 就像面粉管够的面包房不会因为再多送一车面粉而多出面包。
酒精是两头一起压: 一头干扰睾丸里的合成步骤, 一头把更多睾酮推向被改造成雌激素的那条路。所以它降睾酮和升雌激素并不是两件事, 是同一件事的两面。
误区 · 提睾酮补剂为什么普遍无效
没用、营销陷阱Ashwagandha: 慢性压力人群有边缘效应, 不是 T boosterMaca: 性欲主观改善, 不显著影响 TTribulus: 多个 RCT 阴性ZMA (Zn + Mg + B6 组合): 健康人无意义Testosterone booster 复方补剂: D 级证据汇总DHEA: 不推荐做 T booster (副作用、转化不稳定)
这些产品有一个共同的话术破绽: 它们都跳过了从哪一环下手这个问题。前面那条链子上一共只有几个地方可以动 —— 下丘脑放脉冲的频率、睾丸里的合成步骤、血里被 SHBG 拽住的比例、以及细胞里的受体。一样东西如果真能提睾酮, 它必须说得清自己动的是哪一环, 以及为什么在你身上恰好是那一环卡住了。说不清动哪一环的, 通常是因为它哪一环都没动。
还有更深的一层: 睾酮被负反馈管着。就算某样东西真的把血里的睾酮往上推了一点, 上游读到浓度上升, 会立刻减少指令把它拉回去。这条回路的存在本身, 就限定了从外面往上加一点这条路能走多远 —— 除非加到大剂量药物的量级, 而那就是 TRT, 不是补剂。真正走得通的路是把踩着刹车的东西移开 (减重、治 OSA、降皮质醇、把睡眠还回来), 让身体自己把产量调回去。前者跟一条负反馈回路对着干, 后者是顺着它走。
底线: 大多数 40-60 岁男性精力下降、难减肥 在 6 个月生活方式干预下大幅改善, 不必走 TRT 之路。
Practical · the 6-month lifestyle plan
The scene above listed the lifestyle factors that affect testosterone; this page arranges them into an executable 6-month plan — most 'low T symptoms' improve markedly once this is done well, never reaching the TRT step.Prioritized by value:
First priority: lose abdominal fat. If your waist is over the threshold, this is the highest-return move. A 5-10% weight loss raises testosterone roughly 15-25%, via less visceral-fat aromatase and improved insulin resistance. Track waist rather than weight.Second priority: sleep well + screen for OSA. Under 5 hours/night for a week drops testosterone 10-15%. If snoring + daytime sleepiness are present, self-screen with STOP-BANG and get a sleep study if warranted; once OSA is confirmed, CPAP is the single most powerful intervention. Restoring 7-9 hours is the baseline.Third priority: strength training + adequate protein. 2-3 sessions/week of big compound lifts with progressive load, plus protein 1.2-1.6 g/kg, to preserve muscle and improve insulin sensitivity. Don't chase 'testosterone-boosting workout' marketing — standard strength training is enough.Fourth priority: lower cortisol. Chronic high cortisol suppresses the HPG axis. Mindfulness, diaphragmatic breathing, regular exercise, and psychotherapy (CBT) all help; as mood improves, testosterone often rebounds indirectly.Baseline habits: moderate alcohol, no smoking. Long-term heavy drinking (> 21 units/week) lowers testosterone and raises estrogen.
Don't get misled on nutrition: there is no 'testosterone-miracle food'. Ensuring zinc isn't deficient (supplement only if deficient), vitamin D is adequate (25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. > 30 ng/mL), and there's some healthy fat (an ultra-low-fat diet actually lowers testosterone) is enough. Ashwagandha / Maca / Tribulus / ZMA / assorted 'testosterone booster' blends are essentially D-level or no evidence in healthy men — don't spend money there.
How to evaluate: do this full set for 6 months before deciding whether to test testosterone or see a doctor. Addressing reversible confounders first is the evidence-based path; the reverse, 'test T then start TRT', is a T-clinic sales pitch, not medicine.
Chapter 4
TRT — indications + risks
TRT — indications + risks
TRT (testosterone replacement therapy) works in true indications, but fertility impact, cardiovascular, PSA, and erythrocyte risks are all real — the Atlas lays it out clearly.
True indications (Endocrine Society 2018):
Classical hypogonadism: structural hypothalamic-pituitary or testicular disease, congenital or acquired (Klinefelter / cryptorchidism / post-chemotherapy / pituitary tumor, etc.) — strong indicationLate-onset hypogonadism (LOH): ≥ 2 morning T < 300 ng/dL + 3 sexual symptoms (low libido + lost morning erections + ED) + 5 confounders ruled out — can considerHIV / chronic wasting / severe sarcopenia: case-by-case consideration
Real TRT effects (the TTrials programme — Snyder 2016 NEJM plus its companion papers):
Men 65+ with low T, n=790, 1-yr RCT:
Sexual function (libido + erections + satisfaction): significantly improved — the most certain benefitMood + depressive symptoms: slightly better than placebo. Note the direction: TTrials did not report 'no effect' here. The men were not selected for diagnosed depression, and the effect was smallVitality / fatigue: no significant benefit (FACIT-Fatigue)Walking distance: no significant difference within the Physical Function Trial itself; significant only when all three trials were pooled (20.5% vs 12.6% gained ≥ 50 m)Bone density: improved — but that is a separate paper (Snyder 2017 *JAMA Intern Med*, n=211: spine trabecular vBMD +7.5% vs +0.8% on placebo)Cognition / memory: no benefit — also a separate paper (Resnick 2017 *JAMA*, n=493)Cardiovascular / all-cause mortality: TTrials wasn't long enough to draw conclusions
TRAVERSE (Lincoff 2023 NEJM, n=5,246):
Middle-aged and older men with low T + high CV risk. Mean follow-up 33 months, mean 21.7 months actually on treatment — it is often miscalled a '4-year trial'. Primary MACE: TRT vs placebo, no significant difference (HR 0.96) — that is, in low-T + high-CV-risk men, previous TRT cardiovascular safety concerns are partly alleviated. But atrial fibrillation + acute kidney injury + pulmonary embolism rose slightly.
Fertility impact (often hidden by T clinics):
TRT suppresses the HPG axis: exogenous T feeds back, shutting down LH/FSH directly, and testicular spermatogenesis stops. Almost all TRT users see sperm counts drop > 90% within 4–6 months.
Men who want fertility should not start TRT directly:
Clomiphene (50 mg every other day): indirectly stimulates the HPG axis, T rises + spermatogenesis preservedhCG (human chorionic gonadotropin): mimics LH, preserves testicular functionEnclomiphene + AI (anastrozole) combination
Wanting fertility after TRT: stop + hCG + wait 6–12 months; some irreversibility.
Other risks:
Erythrocytosis (Hct > 54%): thrombosis risk ↑, monitor + therapeutic phlebotomy as neededPSA + prostate: known prostate cancer is an absolute contraindication; monitor PSA every 3–6 monthsSleep apnea worsening: caution in patients with existing OSABreast tenderness / gynecomastia: side effects of elevated estradiolSkin acne / male-pattern hair loss: cosmetic side effects
5 TRT formulations:
Gel (Androgel / Testim): daily application, risk of contact contamination of othersInjection (T cypionate / enanthate): every 1–2 weeks, peak-to-trough fluctuationSubcutaneous small frequent injection: steadier state, self-administeredImplant pellets: 3–6 months, surgical placementNasal spray / buccal tablet: less common
In practice:
Don't use T clinics / online TRT — most lack 5-confounder screeningFind an endocrinologist or urologist for full evaluation + long-term monitoringFertility first: consider Clomiphene / hCG before TRTTRT is lifelong (once started, HPG recovery is unreliable)
True indications (Endocrine Society 2018):
Classical hypogonadism: structural hypothalamic-pituitary or testicular disease, congenital or acquired (Klinefelter / cryptorchidism / post-chemotherapy / pituitary tumor, etc.) — strong indicationLate-onset hypogonadism (LOH): ≥ 2 morning T < 300 ng/dL + 3 sexual symptoms (low libido + lost morning erections + ED) + 5 confounders ruled out — can considerHIV / chronic wasting / severe sarcopenia: case-by-case consideration
Real TRT effects (the TTrials programme — Snyder 2016 NEJM plus its companion papers):
Men 65+ with low T, n=790, 1-yr RCT:
Sexual function (libido + erections + satisfaction): significantly improved — the most certain benefitMood + depressive symptoms: slightly better than placebo. Note the direction: TTrials did not report 'no effect' here. The men were not selected for diagnosed depression, and the effect was smallVitality / fatigue: no significant benefit (FACIT-Fatigue)Walking distance: no significant difference within the Physical Function Trial itself; significant only when all three trials were pooled (20.5% vs 12.6% gained ≥ 50 m)Bone density: improved — but that is a separate paper (Snyder 2017 *JAMA Intern Med*, n=211: spine trabecular vBMD +7.5% vs +0.8% on placebo)Cognition / memory: no benefit — also a separate paper (Resnick 2017 *JAMA*, n=493)Cardiovascular / all-cause mortality: TTrials wasn't long enough to draw conclusions
TRAVERSE (Lincoff 2023 NEJM, n=5,246):
Middle-aged and older men with low T + high CV risk. Mean follow-up 33 months, mean 21.7 months actually on treatment — it is often miscalled a '4-year trial'. Primary MACE: TRT vs placebo, no significant difference (HR 0.96) — that is, in low-T + high-CV-risk men, previous TRT cardiovascular safety concerns are partly alleviated. But atrial fibrillation + acute kidney injury + pulmonary embolism rose slightly.
Fertility impact (often hidden by T clinics):
TRT suppresses the HPG axis: exogenous T feeds back, shutting down LH/FSH directly, and testicular spermatogenesis stops. Almost all TRT users see sperm counts drop > 90% within 4–6 months.
Men who want fertility should not start TRT directly:
Clomiphene (50 mg every other day): indirectly stimulates the HPG axis, T rises + spermatogenesis preservedhCG (human chorionic gonadotropin): mimics LH, preserves testicular functionEnclomiphene + AI (anastrozole) combination
Wanting fertility after TRT: stop + hCG + wait 6–12 months; some irreversibility.
Other risks:
Erythrocytosis (Hct > 54%): thrombosis risk ↑, monitor + therapeutic phlebotomy as neededPSA + prostate: known prostate cancer is an absolute contraindication; monitor PSA every 3–6 monthsSleep apnea worsening: caution in patients with existing OSABreast tenderness / gynecomastia: side effects of elevated estradiolSkin acne / male-pattern hair loss: cosmetic side effects
5 TRT formulations:
Gel (Androgel / Testim): daily application, risk of contact contamination of othersInjection (T cypionate / enanthate): every 1–2 weeks, peak-to-trough fluctuationSubcutaneous small frequent injection: steadier state, self-administeredImplant pellets: 3–6 months, surgical placementNasal spray / buccal tablet: less common
In practice:
Don't use T clinics / online TRT — most lack 5-confounder screeningFind an endocrinologist or urologist for full evaluation + long-term monitoringFertility first: consider Clomiphene / hCG before TRTTRT is lifelong (once started, HPG recovery is unreliable)
机制 · 一针睾酮为什么关掉生精, 而 hCG 不会
把三种做法放在同一条链子上比一比, 差别就一目了然。从外面打睾酮 (TRT)
药进的是血。血里睾酮一高, 下丘脑和垂体读到的信号就是够了, 于是 GnRH 脉冲变稀、LH 和 FSH 掉下去。问题在于: 精子发育需要的不是血里那个浓度, 而是睾丸内部由 Leydig 细胞就地供给的、高得多的局部浓度; 而那个局部浓度完全依赖 LH 送来的订单。订单没了, 局部浓度塌了, 支持细胞 (Sertoli) 那边的精子发育链条就断在半路。这就是为什么血检显示睾酮充足, 精液检查却什么都没有 —— 两个数字量的根本不是同一个地方。
hCG
它的形状和 LH 很像, 像一把能配上同一把锁的钥匙, 直接停在 Leydig 细胞上, 越过被压住的上游把订单补上。睾丸继续自己生产, 局部高浓度维持住, 生精也就还在。所以它既能抬高血里的睾酮, 又不至于把车间关掉。
Clomiphene
走的是第三条路: 它挡住雌二醇在下丘脑和垂体上的受体。前面说过, 雌二醇是负反馈上最灵敏的那把刹车 —— 把刹车的接头堵住, 上游就误以为激素不够, 于是把 GnRH 脉冲和 LH 一起调高。整条轴从头到尾都还是自己的, 所以生精不受影响。
同一个道理也解释了停药之后为什么恢复慢
一条长期没被使用的通路, 从下丘脑的脉冲节律到睾丸细胞对 LH 的反应性, 都需要时间重新校准, 而且不保证回到原处。这就是为什么在没有明确适应症的时候贸然开始, 是一个不容易撤销的决定 —— 它不像停一片降压药那样第二天就恢复原状。
红细胞的事也在同一张图上
睾酮本来就会推高造红细胞的信号。外源睾酮把血里浓度顶到平时到不了的水平, 这个信号跟着被推高, 红细胞压积上去、血变稠, 血栓风险随之上升。所以监测红细胞不是例行公事, 它盯的是一条有明确机制的后果; 必要时的放血, 也不是保健手段, 而是直接把变稠的血稀回去。
临床 · 适应症、风险与剂型
适应症其实很窄 (Endocrine Society 2018)。最硬的指征是 classical hypogonadism: 下丘脑-垂体或睾丸有先天或后天的结构性病变 (如 Klinefelter、隐睾、化疗后、垂体瘤)。常被讨论的 late-onset hypogonadism (LOH) 门槛也不低: 要 ≥ 2 次早晨 T < 300 ng/dL, 加上三项性症状 (低性欲 + 晨勃消失 + 勃起功能下降), 而且前面讲的 5 大混杂 (肥胖、OSA、阿片、抑郁、慢病) 都已排除, 才可以考虑。达不到这些, 循证答案是先做满 6 个月生活方式, 而不是打针。TRT 真正被证据支持的回报, 主要落在性功能和骨密度上; 情绪只有很小的改善, 精力和记忆则没有被试验支持 —— 都远没有营销说的那么大。注意这两类适应症在机制上是两回事。classical hypogonadism 是链子上某一环真的坏了、而且修不回来 (睾丸被切除、被化疗损伤, 或者垂体长了东西), 替代进去的睾酮补的是一个确实缺失的环节。LOH 则是整条链子都还在, 只是运转得慢了、或者被别的东西压着 —— 补进去等于绕过那个压着它的原因, 而不是解决它。 这个区别, 就是为什么同一种药在前一类人身上是替代治疗, 在后一类人身上更接近增强剂。
生育冲击是最常被隐瞒的一条。外源睾酮会反馈关停 LH/FSH, 睾丸生精随之停摆——几乎所有 TRT 使用者的精子计数在 4-6 月内下降 90% 以上, 部分不可逆。所以想生育的男性不该直接上 TRT: 该先考虑 Clomiphene 或 hCG, 它们能在升睾酮的同时保住生精。已经用了 TRT 又想生育, 只能停药、加 hCG、再等 6-12 月, 且未必能完全恢复。
其它要监测的风险: 红细胞增多会升高血栓风险, 需要查血、必要时放血; 已知前列腺癌是绝对禁忌, 用药期间定期查 PSA; 原有睡眠呼吸暂停可能加重; 还有乳房压痛、痤疮、脱发等副作用。心血管方面, TRAVERSE (Lincoff 2023, n=5,246) 提示在低 T 且心血管风险偏高的男性中 TRT 并未显著增加主要心血管事件 (HR 0.96), 但房颤、急性肾损伤、肺栓塞略有上升。它常被转述成4 年试验, 实际平均随访 33 个月、平均真正用药 21.7 个月 —— 差别不小, 因为这决定了它能不能对更长期的风险说话。剂型有凝胶、每 1-2 周的肌注、皮下小剂量频繁注射、数月一次的植入颗粒等几种, 各有波动与便利上的取舍, 由医生按情况选。
剂型之间的差别, 说到底是血里浓度曲线的形状差别: 间隔越长的注射, 峰谷落差越大, 峰那几天人会觉得亢奋、谷那几天又打回原形, 而且峰值越高, 被芳香化酶改造成雌二醇的量也越多 —— 乳房压痛通常就出在峰的那一段。频繁小剂量和凝胶把曲线压平, 代价是操作频率和 (凝胶) 会蹭到别人身上的接触污染。这不是哪一种更好, 是拿平稳换方便。
实操: 别走 T 诊所或在线 TRT, 它们大多不做 5 大混杂筛查; 要做就找内分泌或泌尿专科, 全面评估加长期监测。想生育优先考虑 Clomiphene / hCG。记住 TRT 基本是终身治疗——一旦开始, HPG 轴不容易自行恢复。
Myth · 'TRT will make me 25 again'
What T clinics and anti-aging marketing love to sell is the picture: one testosterone shot and energy, muscle, libido, brainpower, and mood all rejuvenate. Set that against the actual trial data and the image turns out heavily airbrushed.What TRT is genuinely supported to do (the TTrials programme, 65+ low-T men, 1-year RCT, n=790):
Sexual function: libido, erections, and satisfaction improve significantly — TRT's most certain benefit (Snyder 2016, the main paper).Bone density: improved — but that result comes from a different TTrials paper (Snyder 2017 *JAMA Intern Med*, n=211: spine trabecular vBMD +7.5% vs +0.8%), not the main one.Memory / cognition: no improvement after a year — again a different paper (Resnick 2017 *JAMA*, n=493). This is what actually punctures 'TRT fixes brain fog'.Energy / vitality: no significant benefit on the fatigue scale in the main paper.Walking distance: not significant within the Physical Function Trial itself; significant only when all three trials were pooled.Mood / depressive symptoms: slightly better than placebo — mind the direction. The main paper reports a small improvement, not 'no effect'. Claiming otherwise is the most common misquotation of this trial, and it backfires: anyone who opens the paper can discredit the whole argument in one move. The honest version is that the effect is real but small, in men who were not selected for diagnosed depression — nowhere near 'a shot fixes low mood'.
Several costs the marketing quietly downplays:
Fertility impact: exogenous testosterone feeds back and shuts down LH/FSH, so almost all TRT users see sperm drop > 90% within 4-6 months, partly irreversibly. People who want fertility should not start TRT directly — consider Clomiphene / hCG first. T clinics often omit this.Erythrocytosis: rising hematocrit raises thrombosis risk, requiring monitoring and phlebotomy as needed.Prostate: known prostate cancer is an absolute contraindication; monitor PSA periodically on therapy.Lifelong dependence: once started, the HPG axis doesn't recover easily — it's essentially lifelong treatment.
So the accurate picture: TRT works in true indications (classical hypogonadism, or ≥ 2 morning T < 300 + 3 sexual symptoms + the 5 confounders ruled out), with the main payoff in sexual function and bone. But for someone who 'just wants more energy / to feel young again' yet doesn't meet guideline criteria, the evidence-based answer is to complete the 6-month lifestyle plan above first — not to get a shot at a T clinic. Treat TRT as a prescription drug for a specific deficit, not an anti-aging supplement.
Chapter 5
Decision tree + atlas loop
Decision tree + atlas loop
'Am I low on T?' self-check + decision tree
Step 1 · Self-check
Tick the following — at least 3 are needed before T testing is warranted:
Significantly reduced libido for ≥ 6 monthsSignificantly reduced or absent morning erectionsReduced erectile functionUnexplained loss of muscle strength (with training + diet unchanged)Chronic fatigue (after excluding sleep / depression / chronic illness)Unexplained weight gain (especially abdominal)
Fewer than 3 ticked: don't test T — focus on lifestyle + sleep + psychology.
Step 2 · Address the 5 confounders
BMI > 30 or waist > 102 cm (male): lose 5–10% firstSnoring + daytime sleepiness: STOP-BANG OSA evaluationLong-term opioids / sedatives: discuss with the prescribing physicianChronic stress + mood problems: psychotherapy + CBTT2D / chronic disease: control the primary disease
Resolve these for 6–12 months first, then evaluate T.
Step 3 · Labs (if you still want to measure)
Morning 7–10 amNot within 2 weeks of acute illness / surgery, or on a day of extreme trainingTotal T + free T (Vermeulen) + SHBG + LH + FSH + estradiol + PSAMethod: LC-MS/MSRepeat once, 4–6 weeks apart
Step 4 · Interpretation
Total T > 400 ng/dL: normal, look for other causesTotal T 300–400 + free T normal: borderline SHBG issue, no TRT neededTotal T < 300 on ≥ 2 occasions + sexual symptoms ≥ 3 + confounders addressed: endocrinology / urology evaluationTotal T < 200: abnormal regardless of confounders, evaluate
Step 5 · Treatment path
Classical hypogonadism: TRTLOH + wanting fertility: Clomiphene / hCG (not TRT)LOH + no fertility plans: discuss TRT formulation + monitoring with physician'Just want more energy' but doesn't meet guideline criteria: lifestyle first, not TRT
Situations to see a doctor promptly
Breast pain + lump (rule out breast cancer / gynecomastia)Visual field constriction + headache (rule out pituitary tumor)Unexplained PSA rise (prostate evaluation)Sudden severe fatigue + significant weight loss (chronic-disease screen)Infertility + low T (reproductive urology)
Atlas + Report loop
The report engine `fertility-prep-male` rule links back here. Atlas links out to:
`protein/muscle` L4 — sarcopenia + Leu/mechanistic target of rapamycin: The cell's master 'grow / build' switch — turned on by enough protein and resistance training.`bone/hormones` L4 — T maintains bone mass`endocrine/metabolic-syndrome` L4 — IR + weight loss restores T`cardiovascular/atherosclerosis` L4 — T + LDL`insomnia/what-types` L4 — sleep raises T
Bottom line: andropause is a real phenomenon (slow T decline) with a misleading name (not a menopause analog). True LOH affects only ~2%; most 'low T symptoms' improve dramatically with lifestyle intervention. The T clinic industry has systematically expanded indications; the Atlas brings it back to an evidence-based 'true indication + comprehensive evaluation + long-term monitoring' framework.
Step 1 · Self-check
Tick the following — at least 3 are needed before T testing is warranted:
Significantly reduced libido for ≥ 6 monthsSignificantly reduced or absent morning erectionsReduced erectile functionUnexplained loss of muscle strength (with training + diet unchanged)Chronic fatigue (after excluding sleep / depression / chronic illness)Unexplained weight gain (especially abdominal)
Fewer than 3 ticked: don't test T — focus on lifestyle + sleep + psychology.
Step 2 · Address the 5 confounders
BMI > 30 or waist > 102 cm (male): lose 5–10% firstSnoring + daytime sleepiness: STOP-BANG OSA evaluationLong-term opioids / sedatives: discuss with the prescribing physicianChronic stress + mood problems: psychotherapy + CBTT2D / chronic disease: control the primary disease
Resolve these for 6–12 months first, then evaluate T.
Step 3 · Labs (if you still want to measure)
Morning 7–10 amNot within 2 weeks of acute illness / surgery, or on a day of extreme trainingTotal T + free T (Vermeulen) + SHBG + LH + FSH + estradiol + PSAMethod: LC-MS/MSRepeat once, 4–6 weeks apart
Step 4 · Interpretation
Total T > 400 ng/dL: normal, look for other causesTotal T 300–400 + free T normal: borderline SHBG issue, no TRT neededTotal T < 300 on ≥ 2 occasions + sexual symptoms ≥ 3 + confounders addressed: endocrinology / urology evaluationTotal T < 200: abnormal regardless of confounders, evaluate
Step 5 · Treatment path
Classical hypogonadism: TRTLOH + wanting fertility: Clomiphene / hCG (not TRT)LOH + no fertility plans: discuss TRT formulation + monitoring with physician'Just want more energy' but doesn't meet guideline criteria: lifestyle first, not TRT
Situations to see a doctor promptly
Breast pain + lump (rule out breast cancer / gynecomastia)Visual field constriction + headache (rule out pituitary tumor)Unexplained PSA rise (prostate evaluation)Sudden severe fatigue + significant weight loss (chronic-disease screen)Infertility + low T (reproductive urology)
Atlas + Report loop
The report engine `fertility-prep-male` rule links back here. Atlas links out to:
`protein/muscle` L4 — sarcopenia + Leu/mechanistic target of rapamycin: The cell's master 'grow / build' switch — turned on by enough protein and resistance training.`bone/hormones` L4 — T maintains bone mass`endocrine/metabolic-syndrome` L4 — IR + weight loss restores T`cardiovascular/atherosclerosis` L4 — T + LDL`insomnia/what-types` L4 — sleep raises T
Bottom line: andropause is a real phenomenon (slow T decline) with a misleading name (not a menopause analog). True LOH affects only ~2%; most 'low T symptoms' improve dramatically with lifestyle intervention. The T clinic industry has systematically expanded indications; the Atlas brings it back to an evidence-based 'true indication + comprehensive evaluation + long-term monitoring' framework.
自查 · 五步决策树
我是不是 T 低了 自查 + 决策树第 1 步 · 自查清单
勾选下列, 至少 3 项才值得测 T:
性欲明显减退 ≥ 6 月晨勃明显减少、消失勃起功能下降不明原因肌力下降 (训练 + 饮食不变)慢性疲倦 (排除睡眠、抑郁、慢病后)不明原因体重增加 (尤其腹型)
没勾 3 项: 不必测 T, 关注生活方式 + 睡眠 + 心理。
第 2 步 · 看 5 大混杂
BMI > 30 或腰围 > 102 cm (男): 先减重 5-10%打鼾 + 白天嗜睡: STOP-BANG 评估 OSA长期阿片、镇静药: 与开药医生讨论慢性压力 + 情绪问题: 心理咨询 + CBTT2D / 慢病: 控制原发病
建议优先解决这些 6-12 月, 然后再评估 T。
第 3 步 · 抽血 (确实想测)
早晨 7-10 am不在急性病、手术后 2 周内、极度训练当天总 T + 自由 T (Vermeulen) + SHBG + LH + FSH + 雌二醇 + PSA方法: LC-MS/MS重复 1 次, 间隔 4-6 周
第 4 步 · 解读
总 T > 400 ng/dL: 正常, 寻找其它原因总 T 300-400 + 自由 T 正常: SHBG 边缘, 不必 TRT总 T < 300 ≥ 2 次 + 性症状 ≥ 3 + 混杂已排除: 内分泌、泌尿评估总 T < 200: 任何混杂时也异常, 应评估
第 5 步 · 治疗路径
classical hypogonadism: TRTLOH + 想生育: Clomiphene / hCG (不是 TRT)LOH + 不需要生育: 与医生讨论 TRT 形态 + 监测只是想体能好一点 但够不上指南: 生活方式优先, 不上 TRT
这张表的次序不是随便排的。它先问症状、再问混杂、最后才抽血, 是因为越往前的一步越便宜、越能解释更多人。 一个先抽血的流程会把大量本来只是睡不好或者腰围超标的人送进一条昂贵而且不解决问题的路径; 而先过前两关, 剩下真正需要往下走的人已经很少了 —— 那正好也是真 LOH 的比例。
机制 · 红旗清单上每一条各自在指什么
上面那五条不是随便挑的比较严重的症状。每一条都指向一个特定的位置, 而且都不是睾酮低了能解释的 —— 这才是它们必须单独列出来的原因。视野缩小 + 头痛 → 垂体
垂体坐在脑底部一个小骨窝里, 而管视野的神经在这里交叉, 正好从垂体正上方压过去。垂体上长了东西, 先被顶到的往往就是这束神经, 表现出来是两侧外周视野先变窄, 像戴上一副越来越深的眼罩。同时它还会把整条激素轴的上游一起搅乱, 于是睾酮低和视野问题会一起出现。这一条最值钱的地方在于: 它是少数几个把低睾酮和一个可以定位的病灶直接连起来的线索。
乳房疼痛 + 肿块 → 雄激素与雌激素的比例
男性的乳腺组织本来就在, 只是平时被雄激素压着不发育。当睾酮下降、或者被芳香化酶改造成雌二醇的比例上升 (肥胖、肝病、某些药物都会造成), 这个比例翻过来, 乳腺组织就开始增生, 摸得到、按上去疼。要紧的是: 单侧、质地硬、位置偏离乳头、伴皮肤改变的肿块, 和这种双侧对称的腺体增生不是一回事, 必须由医生分开判断。
不明 PSA 升高 → 前列腺
前列腺的生长一直受雄激素驱动 (更准确地说是被改造后的 DHT 驱动), 所以任何和睾酮有关的评估都会把它一起看。这不是说睾酮导致了什么, 而是说在动激素之前, 得先知道那里现在是什么状态。
突发严重疲倦 + 体重明显下降 → 去找别的病
这个组合和睾酮的典型表现方向相反: 睾酮慢慢降的人通常是慢慢地、以体重上升 (尤其腹型) 收场, 不是短时间内掉秤。所以出现这一组, 优先要排的是别的系统性问题, 而不是往低睾酮上套。方向相反的症状是一条很强的线索, 别把它当成同一件事的变体。
不育 + T 低 → 先分清是上游还是睾丸
这也是为什么抽血那一栏里要带上 LH 和 FSH。如果睾酮低, 而 LH、FSH 反而升高, 说明上游一直在加大催促、订单一直在发 —— 问题出在睾丸这个车间本身。如果睾酮低, 而 LH、FSH 也不高甚至偏低, 那说明订单根本没发出来, 问题在下丘脑或垂体。同样一个低睾酮, 这两种情况的病因、下一步该查什么、以及怎么治, 完全不同: 前面讲过的长期阿片、肥胖、垂体病变都属于后者, 而化疗后、隐睾、Klinefelter 属于前者。所以 LH 不是化验单上凑数的一项, 它是把两条完全不同的路分开的那个岔口。
闭环 · 报告与 atlas 链回
Atlas + 报告闭环报告引擎 `fertility-prep-male` 规则会链回这里。atlas 链回:
`protein/muscle` L4 — 肌少 + Leu/mechanistic target of rapamycin: The cell's master 'grow / build' switch — turned on by enough protein and resistance training.`bone/hormones` L4 — T 维持骨量`endocrine/metabolic-syndrome` L4 — IR + 减重恢复 T`cardiovascular/atherosclerosis` L4 — T + LDL`insomnia/what-types` L4 — 睡眠 ↑ T
这几条链回不是相关阅读的堆砌, 它们各自接的是这一篇里某一条断口的下半截: 肌少那条接的是睾酮在肌肉里开启的那条合成生产线; 骨量那条接的是芳香化酶在骨里把睾酮改成雌二醇、再按住破骨细胞的那一步; 代谢综合征那条接的是脂肪-芳香化酶-负反馈的那个圈; 睡眠那条接的是睾酮产量挂在深睡眠窗口上的那条节律。顺着任何一条走过去, 都会再一次撞上同一条 HPG 轴。
底线: andropause 是真现象 (T 缓慢下降) + 假命名 (不是绝经翻版)。真 LOH 仅 2%, 大多数低 T 症状 在生活方式干预下大幅改善。T 诊所产业系统性扩大了适应症; atlas 把它救回到真适应症 + 全面评估 + 长期监测 的循证框架。
References · 7
- Wu, F. C. W., Tajar, A., Beynon, J. M., Pye, S. R., Silman, A. J., Finn, J. D., et al. (2010). Identification of late-onset hypogonadism in middle-aged and elderly men. The New England Journal of Medicine, 363(2), 123-135. 10.1056/NEJMoa0911101
- Bhasin, S., Brito, J. P., Cunningham, G. R., Hayes, F. J., Hodis, H. N., Matsumoto, A. M., Snyder, P. J., Swerdloff, R. S., Wu, F. C., & Yialamas, M. A. (2018). Testosterone therapy in men with hypogonadism: an Endocrine Society clinical practice guideline. The Journal of Clinical Endocrinology & Metabolism, 103(5), 1715-1744. 10.1210/jc.2018-00229
- Benjafield, A. V., Ayas, N. T., Eastwood, P. R., Heinzer, R., Ip, M. S. M., Morrell, M. J., et al. (2019). Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis. The Lancet Respiratory Medicine, 7(8), 687-698. 10.1016/S2213-2600(19)30198-5
- Snyder, P. J., Bhasin, S., Cunningham, G. R., Matsumoto, A. M., Stephens-Shields, A. J., Cauley, J. A., et al. (2016). Effects of testosterone treatment in older men. The New England Journal of Medicine, 374(7), 611-624. Main TTrials paper: sexual function improved; mood and depressive symptoms slightly better than placebo; no significant benefit for vitality. Bone density and cognition are reported in separate TTrials papers, not this one. 10.1056/NEJMoa1506119
- Snyder, P. J., Kopperdahl, D. L., Stephens-Shields, A. J., Ellenberg, S. S., Cauley, J. A., Ensrud, K. E., et al. (2017). Effect of testosterone treatment on volumetric bone density and strength in older men with low testosterone: A controlled clinical trial. JAMA Internal Medicine, 177(4), 471-479. N=211; 1 year of treatment raised spine trabecular volumetric BMD 7.5% vs 0.8% on placebo. 10.1001/jamainternmed.2016.9539
- Resnick, S. M., Matsumoto, A. M., Stephens-Shields, A. J., Ellenberg, S. S., Gill, T. M., Shumaker, S. A., et al. (2017). Testosterone treatment and cognitive function in older men with low testosterone and age-associated memory impairment. JAMA, 317(7), 717-727. N=493; 1 year of testosterone was not associated with improved delayed paragraph recall, visual memory, executive function, or spatial ability. 10.1001/jama.2016.21044
- Lincoff, A. M., Bhasin, S., Flevaris, P., Mitchell, L. M., Basaria, S., Boden, W. E., et al. (2023). Cardiovascular safety of testosterone-replacement therapy (TRAVERSE). The New England Journal of Medicine, 389(2), 107-117. In 5,246 men 45-80 with low T and CV risk, TRT was non-inferior for MACE but showed more atrial fibrillation, pulmonary embolism, and acute kidney injury. 10.1056/NEJMoa2215025