Place · Level 3
Maca · Lepidium meyenii
秘鲁安第斯十字花科根茎 (不是人参) · 天然睾酮、激素平衡营销 · 但独立团队的 RCT 没有一项测到血清性激素变化 · 主观性欲、情绪有弱信号, 机制未知 · 全部人体证据小 + 短 · 最正面的几项集中在秘鲁同一个课题组
Last updated
Story path
- 1What it is · an Andean root + the 'Peruvian ginseng' tagWhat it is · an Andean root + the 'Peruvian ginseng' tag
- 2Marketing vs hormones · the 'natural T booster' refuted by its own trialsMarketing vs hormones · the 'natural T booster' refuted by its own trials
- 3RCT evidence · libido / ED / SSRI / athleticRCT evidence · libido / ED / SSRI / athletic
- 4Menopause + fertility · the non-hormonal flip sideMenopause + fertility · the non-hormonal flip side
- 5Safety + sourcing + decision · should you take itSafety + sourcing + decision · should you take it
Chapter 1
What it is · an Andean root + the 'Peruvian ginseng' tag
What it is · an Andean root + the 'Peruvian ginseng' tag
Maca = the swollen root-stem of Lepidium meyenii:
Latin name Lepidium meyenii Walp., family Brassicaceae — the same family as broccoli / cabbage / radishNot a ginseng: 'Peruvian ginseng' is pure marketing. Real ginseng is *Panax* (family Araliaceae) and is botanically unrelated to macaThe edible part is the hypocotyl (a swollen root-stem), not a true rootGrown almost entirely on the Junín plateau of central Peru, ~4000–4500 m — surviving intense UV, cold, and poor soilsIn the Andes it was a food staple long before it was a supplement — traditional uses center on energy / stamina / fertility / tolerating the harsh high-altitude environment
The color phenotypes are real, but 'black maca for the brain, red maca for the prostate' is mostly rodent data:
Yellow maca — most common / most commercialRed maca — reduced testosterone-induced prostate enlargement in rat models (Gasco 2007)Black maca — improved memory in ovariectomized female mice (Rubio 2011). Note: the 'black maca boosts sperm' line often attached to this citation is not from this study — it used no male animals and measured no reproductive endpointThe honest caveat: these color differences come almost entirely from rodent experiments and are not established in humans. Be wary of 'black maca specifically for the brain' sold as a human fact
Active compounds (maca's chemical fingerprint):
Macamides + macaenes — fatty-acid amides largely unique to maca; macamides structurally resemble the endocannabinoid anandamide and are hypothesized (not proven) to act as neuromodulatorsGlucosinolates / isothiocyanates — the cruciferous signature, the source of the pungent flavor and the thyroid / goitrogen discussion belowSterols + alkaloids (macaridine / β-carbolines / imidazole alkaloids, etc.)
Hold onto this one line, because every later judgment flows from it: maca has no known phytoestrogen and no established direct androgen-receptor agonism. How it actually works — if it works — is, to this day, unknown.
Latin name Lepidium meyenii Walp., family Brassicaceae — the same family as broccoli / cabbage / radishNot a ginseng: 'Peruvian ginseng' is pure marketing. Real ginseng is *Panax* (family Araliaceae) and is botanically unrelated to macaThe edible part is the hypocotyl (a swollen root-stem), not a true rootGrown almost entirely on the Junín plateau of central Peru, ~4000–4500 m — surviving intense UV, cold, and poor soilsIn the Andes it was a food staple long before it was a supplement — traditional uses center on energy / stamina / fertility / tolerating the harsh high-altitude environment
The color phenotypes are real, but 'black maca for the brain, red maca for the prostate' is mostly rodent data:
Yellow maca — most common / most commercialRed maca — reduced testosterone-induced prostate enlargement in rat models (Gasco 2007)Black maca — improved memory in ovariectomized female mice (Rubio 2011). Note: the 'black maca boosts sperm' line often attached to this citation is not from this study — it used no male animals and measured no reproductive endpointThe honest caveat: these color differences come almost entirely from rodent experiments and are not established in humans. Be wary of 'black maca specifically for the brain' sold as a human fact
Active compounds (maca's chemical fingerprint):
Macamides + macaenes — fatty-acid amides largely unique to maca; macamides structurally resemble the endocannabinoid anandamide and are hypothesized (not proven) to act as neuromodulatorsGlucosinolates / isothiocyanates — the cruciferous signature, the source of the pungent flavor and the thyroid / goitrogen discussion belowSterols + alkaloids (macaridine / β-carbolines / imidazole alkaloids, etc.)
Hold onto this one line, because every later judgment flows from it: maca has no known phytoestrogen and no established direct androgen-receptor agonism. How it actually works — if it works — is, to this day, unknown.
来历 · 它长在哪, 为什么长成这样
学名 Lepidium meyenii Walp., 十字花科 (Brassicaceae) —— 和西兰花、卷心菜、萝卜同科。食用部位是 hypocotyl (下胚轴, 一种膨大的根-茎), 不是真正的根。产地几乎全部在秘鲁中部胡宁高原 (Junín), 海拔 4000-4500 m —— 强紫外、严寒、贫瘠土壤里仍能长。在安第斯它先是主食, 后才是补品, 传统用途集中在精力、耐力、生育、适应高海拔严酷环境。
玛咖不是人参。秘鲁人参 (Peruvian ginseng) 纯属营销标签: 真人参是 *Panax* (五加科 Araliaceae), 和玛咖植物学上毫无亲缘。两者唯一的共同点, 是都长在地下、都被当滋补品卖。
顺手拆掉一条推理: 广告最爱讲的故事是, 它在那么严酷的地方都能活下来, 所以吃了你也能扛。这条推理是断的。植物顶住强紫外线靠的是自己细胞里合成的化学盾牌, 那套盾牌保护的是植物自己的细胞; 它被你嚼碎、被胃酸泡过、被肠壁挑拣、被肝脏改造之后, 到达你血液里的早已不是原来那个分子, 更不会顺带把耐受严酷环境这项本事转交给你。一种植物在野外多能扛, 和它进到你身体里能做什么, 是两个互不相干的问题。 这条规则对整座适应原岛都适用, 不只对玛咖。
颜色 · 黑玛咖补脑是老鼠说的
颜色表型是真的, 但黑玛咖补脑、红玛咖护前列腺基本来自鼠类实验:黄玛咖 —— 最常见、最商业化红玛咖 —— 大鼠模型里减轻睾酮诱导的前列腺增生 (Gasco 2007)黑玛咖 —— 在去卵巢的雌性小鼠里改善记忆 (Rubio 2011)。⚠️ 常和它一起被引用的黑玛咖提高精子, 不是这项研究做的 —— Rubio 2011 从头到尾没有雄性动物、没有任何生殖终点关键诚实: 这些颜色差异几乎全部来自啮齿动物实验, 在人身上没有确立。看到黑玛咖专补脑当人类事实卖, 就要警惕
为什么这些结果搬不到人身上? 三处断点值得记住:
那些动物本来就不健康。 红玛咖那组是先给大鼠注射睾酮把前列腺催大, 再看玛咖能不能压回去; 黑玛咖那组是先摘除小鼠卵巢造出记忆损伤, 再看玛咖能不能补。它们回答的是能不能把一个人为制造的损伤扳回一点, 而不是一个本来正常的人吃了会不会更好同一个中文词, 底下不是同一件事。 小鼠的记忆改善指的是它在一套固定任务里表现更好; 你说的脑子更清楚是一种自我感受。两者共用一个词, 测的东西却不重叠剂量和代谢都对不上。 动物实验的用量按体重折算过来, 常常远高于人的日常食用量; 而啮齿类的代谢速度和人差得很远 —— 同样一克粉末在两个物种体内走出的浓度曲线, 不是同一条
所以红玛咖护前列腺这句话, 现阶段最多只能说成: 在被人为诱导增生的大鼠身上观察到过。中间那几步, 没有人替你走过。
化学 · 它身上到底有哪三类分子
玛咖的化学指纹由三类分子组成, 每一类都直接对应后面的一场争论:Macamides + macaenes (玛咖酰胺 / 玛咖烯) —— 脂肪酸酰胺, 大体上是玛咖独有。所谓脂肪酸酰胺, 就是一条脂肪尾巴接上一个含氮的头。你自己的身体也造这种形状的分子, 内源大麻素 anandamide 就是其中之一, 它负责调节情绪、食欲、疼痛的音量。macamide 和它长得像, 于是有人假设玛咖是从神经这一端起作用的 —— 但长得像只是长得像: 连最偏向玛咖的那篇综述都把这条列为未证实Glucosinolates / isothiocyanates (硫代葡萄糖苷 / 异硫氰酸盐) —— 十字花科的家族标志。它们待在完整细胞里时是安静的; 你一咬碎, 植物自己的酶和它们碰面, 当场把它切成辛辣的异硫氰酸盐。玛咖那股冲鼻味就是这一刀的产物, 后面生玛咖会不会压甲状腺的讨论也全落在这一类分子身上甾醇 + 生物碱 (macaridine / β-carboline / 咪唑生物碱等) —— 含量低、人体研究极少, 目前还没有一条把它们和某个具体功效连起来的人体证据
把这三类合起来看, 有一件事很醒目: 没有一类长得像性激素。 大豆异黄酮之所以被叫作植物雌激素, 不是因为它天然, 而是因为它的骨架恰好能塞进雌激素受体那个口袋, 于是细胞真的会读到有雌激素来过。玛咖这三类分子, 没有一类具备那个形状 —— 这就是下一个场景里指针一动不动的化学理由。
Chapter 2
Marketing vs hormones · the 'natural T booster' refuted by its own trials
Marketing vs hormones · the 'natural T booster' refuted by its own trials
Maca's loudest marketing line is 'natural testosterone booster / hormone balancer'. There is an awkward problem with that line: maca's own most rigorous human trial directly refutes it.
The decisive trial (the core of this page):
Gonzales 2003 (Journal of Endocrinology) — a 12-week double-blind, placebo-controlled RCT designed specifically to test whether maca moves hormones56 healthy men, 1.5 g or 3.0 g/dayMeasured testosterone (T) / estradiol (E2) / luteinizing hormone (LH) / follicle-stimulating hormone (FSH) / prolactin / 17-OH-progesterone all at onceVerbatim conclusion: 'treatment with Maca does not affect serum reproductive hormone levels'
This isn't one lone study — it's a repeating pattern across independent teams:
Gonzales 2002 (healthy men): self-rated desire rose after week 8, but testosterone and estradiol did not move — the paper's whole thesis is 'the effect is unrelated to androgens'Brooks 2008 (postmenopausal women): mood / sexual-function scores improved, but estradiol / FSH / LH / SHBG were all unchanged, and the authors state the effect is 'not related to estrogen or androgen content'Stojanovska 2015 (postmenopausal women): depression score + diastolic BP dropped, but estradiol / FSH / SHBG / thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. unchangedMelnikovova 2015 (healthy men): hormones likewise unchanged
The exception, stated rather than hidden: a handful of postmenopausal studies did report estradiol rising and FSH falling — but they cluster in one author's set of small trials and have never been replicated by an independent group. Weighed on this page's own 'independence discount' scale, they get the same discount as the most maca-favorable Peruvian results.
So the defensible claim is not 'maca never moves hormones'; it is 'every trial run by an independent team measured no change'.
Why would that be? — because there is no plausible molecular mechanism:
Maca contains no phytoestrogen (unlike soy isoflavones)Maca does not agonize the androgen receptor (unlike anabolic steroids)It does not inhibit aromatase, and is not a 5α-reductase inhibitorThe 'macamide resembles anandamide' story is a hypothesis: the most maca-favorable 2024 review (Ulloa) writes only that macamides 'might act by similar mechanisms' and that they 'may act on the nervous system' by inhibiting FAAH. And the older idea that maca alkaloids balance hormones via the hypothalamic–pituitary–adrenal axis: The body's stress-response chain (hypothalamus → pituitary → adrenal) that releases cortisol. axis is the one that review flatly labels 'this hypothesis has not been confirmed'
So this layer's conclusion is firm: whatever some people 'feel', no independent trial finds it happening through a sex hormone you can measure. Buying it as 'natural testosterone' or 'hormone balance' is being contradicted by its own foundational research.
→ Open this scene's Level 4 micro-animation to see 'marketing dial vs measured dial': on one side the hormone needles get hyped upward, on the other side the real needles sit perfectly still in Gonzales 2003.
The decisive trial (the core of this page):
Gonzales 2003 (Journal of Endocrinology) — a 12-week double-blind, placebo-controlled RCT designed specifically to test whether maca moves hormones56 healthy men, 1.5 g or 3.0 g/dayMeasured testosterone (T) / estradiol (E2) / luteinizing hormone (LH) / follicle-stimulating hormone (FSH) / prolactin / 17-OH-progesterone all at onceVerbatim conclusion: 'treatment with Maca does not affect serum reproductive hormone levels'
This isn't one lone study — it's a repeating pattern across independent teams:
Gonzales 2002 (healthy men): self-rated desire rose after week 8, but testosterone and estradiol did not move — the paper's whole thesis is 'the effect is unrelated to androgens'Brooks 2008 (postmenopausal women): mood / sexual-function scores improved, but estradiol / FSH / LH / SHBG were all unchanged, and the authors state the effect is 'not related to estrogen or androgen content'Stojanovska 2015 (postmenopausal women): depression score + diastolic BP dropped, but estradiol / FSH / SHBG / thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. unchangedMelnikovova 2015 (healthy men): hormones likewise unchanged
The exception, stated rather than hidden: a handful of postmenopausal studies did report estradiol rising and FSH falling — but they cluster in one author's set of small trials and have never been replicated by an independent group. Weighed on this page's own 'independence discount' scale, they get the same discount as the most maca-favorable Peruvian results.
So the defensible claim is not 'maca never moves hormones'; it is 'every trial run by an independent team measured no change'.
Why would that be? — because there is no plausible molecular mechanism:
Maca contains no phytoestrogen (unlike soy isoflavones)Maca does not agonize the androgen receptor (unlike anabolic steroids)It does not inhibit aromatase, and is not a 5α-reductase inhibitorThe 'macamide resembles anandamide' story is a hypothesis: the most maca-favorable 2024 review (Ulloa) writes only that macamides 'might act by similar mechanisms' and that they 'may act on the nervous system' by inhibiting FAAH. And the older idea that maca alkaloids balance hormones via the hypothalamic–pituitary–adrenal axis: The body's stress-response chain (hypothalamus → pituitary → adrenal) that releases cortisol. axis is the one that review flatly labels 'this hypothesis has not been confirmed'
So this layer's conclusion is firm: whatever some people 'feel', no independent trial finds it happening through a sex hormone you can measure. Buying it as 'natural testosterone' or 'hormone balance' is being contradicted by its own foundational research.
→ Open this scene's Level 4 micro-animation to see 'marketing dial vs measured dial': on one side the hormone needles get hyped upward, on the other side the real needles sit perfectly still in Gonzales 2003.
机制 · 想推高睾酮, 一共只有三条路
要判断一个东西能不能助推睾酮, 得先看清睾酮是怎么被造出来的。这条产线是这样走的: 你脑子深处的下丘脑按节律放出一个信号, 脑垂体接到后往血里投放黄体生成素 (LH); LH 顺血流漂到睾丸, 停靠在睾丸间质细胞表面的受体上, 细胞这才启动那条把胆固醇一步步改造成睾酮的酶链。造出来的睾酮再顺血流回到脑子, 压住下丘脑和垂体的呼喊 —— 这是一台会自己回话的闭环机器。另一道指令卵泡刺激素 (FSH) 走隔壁那条线, 管的是精子车间。
要让睾酮读数上升, 通路只有三条:
自己冒充激素 —— 分子长得够像, 直接停进受体口袋 (合成代谢类固醇、大豆异黄酮走的是这一条)改动加工环节 —— 拦住把睾酮转成雌二醇的那个酶, 或者拦住把它转成更强形式的那一步, 于是同样的产量被读成不同的分配加大上游指令 —— 让垂体多放 LH, 睾丸就多开工
这就是那个决定性试验为什么要一次测那么多项。 它不是撒网, 是逐条堵路: 睾酮和雌二醇管成品, LH 和 FSH 管上游指令, 还有一项测的是合成路上的中间体 —— 连半成品有没有堆积或减少都看了。
结果是每一格都停在原处。不但成品没变, 连上游的呼喊声都没变。这比没测出效果要强得多: 它说明信号根本没在任何一个环节被送出去。换句话说, 玛咖不是推力太小, 而是压根没接在这条线上。
带走这个框架, 你以后看到任何天然助推激素的宣称, 都可以自己检查一遍: 它走的是哪一条? 有没有人去量那条路上的指标? 三条路都没有证据, 那句宣称就没有落脚点。
试验 · 决定性的那一个, 和它身后的模式
那个决定性的实验:Gonzales 2003 (Journal of Endocrinology) —— 一个专门为回答玛咖动不动激素而设计的 12 周双盲安慰剂对照 RCT56 名健康男性, 1.5 g 或 3.0 g/天一次性测了睾酮 (T) / 雌二醇 (E2) / 黄体生成素 (LH) / 卵泡刺激素 (FSH) / 催乳素 / 17-OH 孕酮原文结论 (逐字): treatment with Maca does not affect serum reproductive hormone levels (玛咖治疗不改变血清生殖激素水平)
这不是孤证, 是反复出现的模式:
Gonzales 2002 (健康男): 自评性欲在第 8 周后上升, 但睾酮和雌二醇没动 —— 论文的整个主题就是效应与雄激素无关Brooks 2008 (绝经后女性): 情绪、性功能评分改善, 但雌二醇 / FSH / LH / SHBG 全部不变, 作者明说效应与雌激素或雄激素含量无关Stojanovska 2015 (绝经后女性): 抑郁评分 + 舒张压下降, 但雌二醇 / FSH / SHBG / thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. 不变Melnikovova 2015 (健康男): 激素同样不变
请注意这四个试验的一致之处: 人群不同、性别不同、主要终点不同, 做的团队也分属秘鲁、澳大利亚和捷克, 唯独激素那一栏都是空的。阴性结果反复出现在不同的设计、不同的实验室上, 比同一个设计重复十次更有说服力。
但要诚实: 不是每一项测过激素的玛咖研究都是阴性的
本页挂着的那篇 2024 年综述 (Ulloa) 在绝经后女性一节里, 同时列了两类结果。报告没有变化的是 Brooks 2008 (雌二醇 / FSH / LH / SHBG 全不变); 而报告有变化的是另一组研究 —— Meissner 2005 与 2006 的几篇, 内容是雌二醇上升、FSH 下降。
所以每一个测了激素的试验都没动这句话是过头的, 我们把它收回来。真正站得住的版本是:
凡是独立团队做的(Brooks 澳大利亚 / Stojanovska 澳大利亚 + 香港 / Melnikovova 捷克 / Gonzales 秘鲁), 激素那一栏都是空的唯一报告激素变动的, 集中出自同一位作者的一批小型试验, 至今没有被任何独立团队复制出来
为什么这样写反而更强? 因为它用的是本篇下一幕就要教你的那把尺子 —— 当一个方向的结果集中在单一来源、又没有独立复制时, 它的可信度要打折。这把尺子在这里必须被用在对我们自己不利的方向上, 否则它就只是一件挑对手时才拿出来的武器。
⚠️ 一个边界: 上面对那批研究的判断 (同一作者、小样本、未被复制) 来自那篇综述的转述, 我们没有逐篇读过原文。所以正确的态度不是断言它们是错的, 而是: 在有独立复制之前, 它们撑不起激素平衡这个卖点。
误区 · 受体是锁, 分子是钥匙
为什么会这样? —— 因为没有合理的分子机制:玛咖不含植物雌激素 (不像大豆异黄酮)玛咖不激动雄激素受体 (不像合成代谢类固醇)它不抑制芳香化酶, 不是 5α-还原酶抑制剂macamide 长得像 anandamide、可能从神经这一端起作用, 这些都还是假设。连最偏向玛咖的 2024 年综述 (Ulloa) 用的也只是可能通过相似机制起作用 (might act by similar mechanisms) 这种措辞。而它明确判为未证实 (this hypothesis has not been confirmed) 的, 是另一条更老的假说 —— 玛咖生物碱通过 hypothalamic–pituitary–adrenal axis: The body's stress-response chain (hypothalamus → pituitary → adrenal) that releases cortisol. 轴平衡激素。请注意: 被这篇综述直接否掉的, 恰恰就是激素平衡这个卖点本身
把受体想成一把锁, 分子想成钥匙, 上面四条就好懂了。
大豆异黄酮被叫作植物雌激素, 不是因为它天然, 而是因为它的三维形状恰好能插进雌激素受体那个口袋, 把锁转开一点点 —— 细胞于是真的读到了有雌激素来过。合成代谢类固醇更直接: 它本来就是照着睾酮改出来的, 插进雄激素受体后能把一整套造肌肉的基因打开。这两类东西之所以能动激素, 是因为它们有对应的钥匙形状。
玛咖那三类分子, 没有一类具备这两种形状。它也不去动加工车间: 不拦芳香化酶 (那是把睾酮转成雌二醇的酶), 也不是那种拦住睾酮转成更强形式的药物。三把锁, 一把都没有对应的钥匙。
所以激素读数一动不动这件事, 本来就不该让你意外 —— 它恰恰是化学上早该预测到的结果。真正需要解释的是另外一半: 在化学上什么都没发生的情况下, 为什么有些人的自评分确实动了? 那半个答案在下一个场景, 它比营销故事诚实得多, 也有意思得多。
Chapter 3
RCT evidence · libido / ED / SSRI / athletic
RCT evidence · libido / ED / SSRI / athletic
If hormones don't move, what is maca actually good for? Lay the human RCTs out in tiers and the answer is 'a little, but far softer than the marketing claims'.
Libido / sexual function (Grade C — subjective endpoints + tiny samples):
Gonzales 2002 (healthy men, 12 wk): self-rated desire beat placebo after week 8. But the endpoint is subjective self-report, the sample is small, single-labTop-tier review Shin 2010 (BMC Complement Altern Med): found only 4 RCTs, none reporting allocation concealment / adverse events, concluding the evidence is 'limited' and insufficient for firm conclusions — the most honest one-liner
Erectile dysfunction (Grade C — only mild ED was ever tested):
Zenico 2009: 50 men with mild ED, 2400 mg maca dry extract (an extract, not the 1.5–3.5 g of powder used elsewhere). Both arms improved on IIEF-5, but maca's increment was 1.6 ± 1.1 vs placebo 0.5 ± 0.6, P < 0.001 — the cleanest statistical separation in maca's entire evidence base; do not read it as a trial swallowed by placebo responseIts real weaknesses are elsewhere: only 50 men, only mild ED, and every endpoint is a subjective scale. IIEF-5 runs to 25 points, so a 1.1-point gap is statistically clear but of uncertain size in lived experienceLee 2023 review: only 2 RCTs (n=79), one positive, one null; pooled, the result is significant (MD 1.13, 95% CI 0.64 to 1.61) — we don't hide that number, but read it with its base: two trials, 79 men, mild ED only, which is why the authors still conclude 'limited evidence'. Moderate/severe ED has never been studied
SSRI / antidepressant-induced sexual dysfunction (Grade C — a detail that's often misread):
Dording 2008: a dose-finding study with nitric oxide: A small signal molecule from the vessel lining that relaxes the vessel-wall muscle so the vessel widens. placebo arm. The high-dose group improved within-group, but no control = can't rule out placebo / regression to the meanDording 2015: this is the real placebo-controlled test (parallel design, ~42 women). Primary endpoints were largely null, with benefit only in a small 'postmenopausal subgroup' — the honest read is 'a largely negative trial dressed in a subgroup-positive headline'
Athletic performance (Grade C — fails the one hard head-to-head):
Stone 2009: 8 cyclists; the 40 km time-trial beat baseline (P=0.01) but not placebo (P>0.05). The endurance claim loses its only direct contest (self-rated desire, again, did beat placebo)
In one line: maca occasionally flashes a weak signal on soft, subjective libido / mood endpoints; the moment it faces objective, hard controls (endurance, hormones, sperm), it tends to revert to baseline.
Libido / sexual function (Grade C — subjective endpoints + tiny samples):
Gonzales 2002 (healthy men, 12 wk): self-rated desire beat placebo after week 8. But the endpoint is subjective self-report, the sample is small, single-labTop-tier review Shin 2010 (BMC Complement Altern Med): found only 4 RCTs, none reporting allocation concealment / adverse events, concluding the evidence is 'limited' and insufficient for firm conclusions — the most honest one-liner
Erectile dysfunction (Grade C — only mild ED was ever tested):
Zenico 2009: 50 men with mild ED, 2400 mg maca dry extract (an extract, not the 1.5–3.5 g of powder used elsewhere). Both arms improved on IIEF-5, but maca's increment was 1.6 ± 1.1 vs placebo 0.5 ± 0.6, P < 0.001 — the cleanest statistical separation in maca's entire evidence base; do not read it as a trial swallowed by placebo responseIts real weaknesses are elsewhere: only 50 men, only mild ED, and every endpoint is a subjective scale. IIEF-5 runs to 25 points, so a 1.1-point gap is statistically clear but of uncertain size in lived experienceLee 2023 review: only 2 RCTs (n=79), one positive, one null; pooled, the result is significant (MD 1.13, 95% CI 0.64 to 1.61) — we don't hide that number, but read it with its base: two trials, 79 men, mild ED only, which is why the authors still conclude 'limited evidence'. Moderate/severe ED has never been studied
SSRI / antidepressant-induced sexual dysfunction (Grade C — a detail that's often misread):
Dording 2008: a dose-finding study with nitric oxide: A small signal molecule from the vessel lining that relaxes the vessel-wall muscle so the vessel widens. placebo arm. The high-dose group improved within-group, but no control = can't rule out placebo / regression to the meanDording 2015: this is the real placebo-controlled test (parallel design, ~42 women). Primary endpoints were largely null, with benefit only in a small 'postmenopausal subgroup' — the honest read is 'a largely negative trial dressed in a subgroup-positive headline'
Athletic performance (Grade C — fails the one hard head-to-head):
Stone 2009: 8 cyclists; the 40 km time-trial beat baseline (P=0.01) but not placebo (P>0.05). The endurance claim loses its only direct contest (self-rated desire, again, did beat placebo)
In one line: maca occasionally flashes a weak signal on soft, subjective libido / mood endpoints; the moment it faces objective, hard controls (endurance, hormones, sperm), it tends to revert to baseline.
证据 · 性欲 / ED / SSRI / 运动, 逐条摊开
性欲、性功能 (C 级 —— 主观终点 + 小样本):Gonzales 2002 (健康男, 12 周): 自评性欲第 8 周后高于安慰剂。但终点是主观自评, 样本小, 来自单一实验室顶层综述 Shin 2010 (BMC Complement Altern Med): 只找到 4 个 RCT, 都没报告分配隐藏、不良事件, 结论是证据有限 (limited evidence), 不足以下定论 —— 这是最诚实的一句话
勃起功能障碍 (C 级 —— 只测过轻度 ED):
Zenico 2009: 50 名轻度 ED 男, 玛咖干提取物 2400 mg (注意: 是提取物, 不是 1.5-3.5 g 干粉)。玛咖和安慰剂都改善 IIEF-5, 但玛咖的增量是 1.6 ± 1.1 分 vs 安慰剂 0.5 ± 0.6 分, P < 0.001 —— 这是玛咖全部证据里统计分离最干净的一项, 别把它读成安慰剂效应吃掉了它它真正的软肋在别处: 样本只有 50 人、只招轻度 ED、终点全是主观量表; 而且 IIEF-5 是一个 25 分量表, 1.1 分的差距在统计上很清楚, 在你自己的体验上有多明显是另一回事Lee 2023 综述: 仅 2 个 RCT (n=79), 一个阳性一个阴性; 合并之后是显著的 (MD 1.13, 95% CI 0.64 到 1.61) —— 我们不藏这个数, 但请连它的底座一起看: 两项试验、79 人、全是轻度 ED, 所以作者自己的结论仍然是证据有限, 不足以下定论。中重度 ED 根本没研究过
SSRI / 抗抑郁药引起的性功能障碍 (C 级 —— 这里有个常被误读的细节):
Dording 2008: 这是个没有安慰剂臂的剂量探索试验。高剂量组组内改善, 但无对照 = 无法排除安慰剂、回归均值Dording 2015: 这才是真正的安慰剂对照试验 (平行设计, ~42 名女性)。主要终点基本阴性, 只有一个很小的绝经后亚组出现获益 —— 最诚实的读法是一个基本阴性的试验, 被包装成亚组阳性的标题
运动表现 (C 级 —— 唯一一次硬对照里直接失败):
Stone 2009: 8 名自行车手, 40 km 计时赛比基线快 (P=0.01), 但不比安慰剂快 (P>0.05)。耐力宣称在唯一一次正面较量里就败了 (倒是自评性欲又胜过安慰剂)
一句话: 玛咖在主观、软性的性欲、情绪终点上偶尔露出弱信号; 一旦面对客观、硬性的对照 (耐力、激素、精子), 就常常打回原形。
工具 · 比基线好, 和比安慰剂好
玛咖那个自行车试验里藏着一节值得单独学一遍的方法课 —— 同一份数据在里面得出了两个相反的标题。那八名车手吃完玛咖再骑, 成绩确实比自己开始前快, 差异在统计上还成立。如果论文只报到这里, 广告就可以印上实测提升耐力。可这份研究还带了一个安慰剂组: 同样的赛程、同样形状的粉末, 只是里面什么都没有。跟这一组比, 玛咖没有赢。
为什么两种比较能差出一个结论? 因为比开始前快这件事, 本来就有一堆和玛咖无关的力量在推:
你熟悉了赛道和设备 —— 同一个测试做第二次, 大多数人本来就会更快回归均值 —— 愿意来参加试验的, 往往是当下状态偏差的那一批 (状态好好的人不折腾), 而人的状态会自己往中间飘回来你知道自己在被观察 —— 用力程度和配速策略都会不一样主观打分那一栏更是敞开的 —— 你付了钱、投入了几周、盼着它有用, 分数自然往上走
安慰剂组的全部意义, 就是把上面这几股力量一次性吸走。 两组都经历熟悉过程、都回归均值、都被观察、都在期待; 剩下还高出来的那一截, 才是这个东西自己的效果。
所以以后看到吃了某某之后指标改善了, 你要问的第一个问题不是改善了多少, 而是跟谁比。只报比基线好的宣传, 缺的不是精度, 而是那一半足以推翻结论的信息。
这也解释了玛咖证据里那个反复出现的形状: 客观终点一上对照就塌, 而主观终点因为被期待推着, 有时还能守住一点点差距。
Why the evidence is so 'soft': lab + placebo
Maca's 'soft' evidence has two structural causes worth unpacking separately:① The most positive results cluster in one research group + a national industry interest
State it precisely first: maca's human trials are not almost all Peruvian. Zenico is Italian, Dording American, Stone British, Brooks and Stojanovska Australian, Melnikovova Czech, Shin and Lee's reviews Korean — independent teams are well representedWhat actually clusters are the most positive findings: libido, sperm, and the colour phenotypes, where Gonzales / Gasco / Rubio recur, all from one Peruvian research groupAnd the independent teams are the ones that came back null or trend-level (Stone 2009 endurance, Dording 2015 SSRI, Melnikovova 2015 sperm). That distribution is itself the information — it says more than a blanket 'it's all Peruvian'Maca is a national export crop of Peru, with a clear industry / national interest attached. This doesn't mean fraud, but in evidence terms it's an independence concern: when positive results are concentrated in a few interested parties and not replicated independently, their credibility is discountedThis ruler has to cut both ways. The studies in the previous scene that reported estradiol rising and FSH falling also cluster in one author's hands with no independent replication — same ruler, same discount. A standard you only reach for when it favours you is not a standardWhat would actually be reassuring is a third-party, large-sample, long-term, pre-registered trial — maca has none
② On subjective libido / mood endpoints, the placebo effect is naturally huge
'Desire', 'well-being', and 'mood' are subjective experiences highly shaped by expectationA ready example is Barton 2013 on the ginseng island: its placebo arm improved by 8.2 fatigue points at 4 weeks and 10.3 at 8 weeks — half the participants took nothing and still got substantially betterDo not use Zenico 2009 for this point: there the placebo arm gained only 0.5 points against maca's 1.6, P < 0.001 — that is a trial the placebo response failed to swallowIf you believe in an ancient Andean superfood + paid money + expect to feel better → your self-rated scores rise. That improvement is a real experience, but it belongs to your expectation, not maca's pharmacologyThe only way to separate the two is placebo control + objective measurement — and maca is remarkably consistent at 'getting weaker / disappearing once a control is added' (Stone 2009 endurance, Dording 2015 SSRI, Melnikovova 2015 sperm)
Put the two together: maca's entire evidence base ≈ a dozen small trials of n=8–50, ≤12 weeks, often single-lab, and three top-tier reviews (Shin 2010 / Lee 2011 / Lee 2023) say in unison: 'limited evidence, quality too low, can't conclude'. This doesn't mean 'it definitely does nothing' — it means anyone selling it as definitively effective is reaching beyond what the evidence allows.
lee-2023-maca-erectile-dysfunction-reviewdording-2008-maca-ssri-dose-findingdording-2015-maca-ssri-womenbarton-2013-jnci-ginseng-fatigue
Chapter 4
Menopause + fertility · the non-hormonal flip side
Menopause + fertility · the non-hormonal flip side
Maca carries high hopes in two specific groups: menopausal women and men wanting higher fertility. The data here are interesting — they offer a little hope while again proving 'nothing to do with hormones'.
Menopause / perimenopause (weak signal on mood / symptoms, but not hormone replacement):
Brooks 2008 (crossover RCT, 14 postmenopausal women, 3.5 g/day): anxiety / depression scores + sexual dysfunction improved beyond placebo — but the authors specifically note estradiol / FSH / LH / SHBG all unchanged, the effect 'unrelated to estrogen or androgen'Stojanovska 2015 (crossover RCT, 29 postmenopausal women, 3.3 g/day): depression score + diastolic BP dropped, hormones again unchangedLee 2011 review (Maturitas): 4 RCTs were consistently favorable on Kupperman / Greene menopausal scales, but 'limited evidence, methodological quality too low, safety not established'How to read it: this is a 'symptomatic / mood' weak benefit, not hormone replacement therapy (HRT/MHT). Someone with a genuine hormone gap and severe symptoms should be evaluating MHT (dive into postmenopausal health), not using maca as a substitute
Male fertility / sperm (only 'trend' level, not 'proven' level):
Gonzales 2001 (n=9, no placebo control, 4 months): semen volume / sperm count / motility rose, hormones unchanged — but open-label + single-lab + tiny sample = hypothesis-generating onlyMelnikovova 2015 (placebo-controlled, n=20, 12 wk): sperm concentration / motility showed only non-significant upward trends, hormones unchanged — underpowered to prove a fertility benefitHow to read it: there's a signal, but no adequate controlled trial confirms it. Infertility is a real problem for reproductive medicine; maca cannot replace proper evaluation and treatment
The shared throughline across both groups: even in its 'most-likely-useful' scenarios, maca's benefit is subjective / symptomatic, weak, and decoupled from hormones. It may be a gentle adjunct for some people, but it does not treat a hormone gap, and it does not treat infertility.
Menopause / perimenopause (weak signal on mood / symptoms, but not hormone replacement):
Brooks 2008 (crossover RCT, 14 postmenopausal women, 3.5 g/day): anxiety / depression scores + sexual dysfunction improved beyond placebo — but the authors specifically note estradiol / FSH / LH / SHBG all unchanged, the effect 'unrelated to estrogen or androgen'Stojanovska 2015 (crossover RCT, 29 postmenopausal women, 3.3 g/day): depression score + diastolic BP dropped, hormones again unchangedLee 2011 review (Maturitas): 4 RCTs were consistently favorable on Kupperman / Greene menopausal scales, but 'limited evidence, methodological quality too low, safety not established'How to read it: this is a 'symptomatic / mood' weak benefit, not hormone replacement therapy (HRT/MHT). Someone with a genuine hormone gap and severe symptoms should be evaluating MHT (dive into postmenopausal health), not using maca as a substitute
Male fertility / sperm (only 'trend' level, not 'proven' level):
Gonzales 2001 (n=9, no placebo control, 4 months): semen volume / sperm count / motility rose, hormones unchanged — but open-label + single-lab + tiny sample = hypothesis-generating onlyMelnikovova 2015 (placebo-controlled, n=20, 12 wk): sperm concentration / motility showed only non-significant upward trends, hormones unchanged — underpowered to prove a fertility benefitHow to read it: there's a signal, but no adequate controlled trial confirms it. Infertility is a real problem for reproductive medicine; maca cannot replace proper evaluation and treatment
The shared throughline across both groups: even in its 'most-likely-useful' scenarios, maca's benefit is subjective / symptomatic, weak, and decoupled from hormones. It may be a gentle adjunct for some people, but it does not treat a hormone gap, and it does not treat infertility.
临床 · 更年期: 三个研究说了什么
更年期、围绝经期 (情绪、症状有弱信号, 但不是激素替代):Brooks 2008 (交叉 RCT, 14 名绝经后女性, 3.5 g/天): 焦虑、抑郁评分 + 性功能障碍改善, 优于安慰剂 —— 但作者特意强调雌二醇 / FSH / LH / SHBG 全不变, 效应与雌、雄激素无关Stojanovska 2015 (交叉 RCT, 29 名绝经后女性, 3.3 g/天): 抑郁评分 + 舒张压下降, 激素同样不变Lee 2011 综述 (Maturitas): 4 个 RCT 在 Kupperman / Greene 更年期量表上方向一致地正面, 但证据有限、方法学质量太低、安全性未确立怎么理解: 这是一个症状性、情绪性的弱获益, 不是激素替代疗法 (HRT/MHT)。真正激素缺口大、症状重的人, 该评估的是 MHT (见 dive 到绝经后健康), 不是把玛咖当替代
机制 · 更年期难受的, 是哪一段供应断了
要看懂评分动了、激素不动这个组合有多干净, 得先知道绝经在身体里到底是哪一段断了。卵巢里的卵泡数量是有限的, 用完就不再回应上游的呼喊。于是雌激素的供应从卵巢这一端断档; 而下丘脑和垂体并不知道仓库空了, 它们只知道回话没了, 于是把指令一次比一次喊得响。这就是绝经后典型的那个组合: 雌二醇低、而 FSH 高 —— 一低一高本身就是这台闭环机器在空转的签名。
接下来这一步是关键。 如果一个东西真的补上了雌激素缺口 (哪怕只补上一部分), 垂体会比你自己更早察觉到: 它收到了回话, 呼喊就会降下来, FSH 读数跟着往下走。这是激素替代疗法在化验单上留下的指纹。
而在 Brooks 2008 和 Stojanovska 2015 这两项独立团队做的试验里, 玛咖吃了好几周, FSH 纹丝不动。这一条比它不含植物雌激素那句化学推断硬得多: 化学推断说的是它不该有用, 而 FSH 不动是在活人身上直接量到的 —— 这两群人的大脑从头到尾没有收到过任何一点雌激素信号。
⚠️ 一个必须写出来的反例: 另有一组绝经后女性研究报告过相反的方向 (FSH 下降、雌二醇上升)。它们集中出自同一位作者、样本都很小、至今没有独立复制 (细节见营销 vs 激素那一幕)。所以这一段的正确读法不是FSH 绝不会动, 而是: 在独立团队的手里, 它没有动过。
所以玛咖在更年期能做的事, 从一开始就被限定在另一条赛道上: 不是把断掉的那段供应接回去, 而是在你怎么感受这件事上做一点边际影响。这也正好解释了它成绩单的形状 —— 情绪量表上有弱信号, 激素栏里零信号。至于潮热这类直接由雌激素缺口驱动的症状, 量表上那点改善究竟是感受层面的, 还是真的减少了发作, 现有的小试验分不开。
给你一条可以自己推演的规则: 凡是宣称天然平衡激素的东西, 都可以拿这台闭环机器去检验 —— 真补进去了, 上游的呼喊必然减弱; 上游读数纹丝不动, 就说明什么也没补进去。
临床 · 男性生育: 从趋势到证明还差什么
男性生育力、精子 (只到趋势级别, 没到证明级别):Gonzales 2001 (n=9, 无安慰剂对照, 4 个月): 精液量、精子数、活力上升, 激素不变 —— 但开放标签 + 单一实验室 + 极小样本 = 只能算提出假设Melnikovova 2015 (安慰剂对照, n=20, 12 周): 精子浓度、活力只有不显著的上升趋势, 激素不变 —— 效力不足以证明生育获益怎么理解: 有信号, 但没有一个够格的对照试验确认。不育是要看生殖医学的真问题, 玛咖不能替代正规评估和治疗
为什么精子这个终点特别容易给出好看却站不住的数字? 三条原因:
你今天量到的精子, 不是今天造的。 一个精原细胞要走完整条生产线 —— 分裂、变形、长出尾巴、再在附睾里练会游泳 —— 才成为一颗能被计数的成熟精子, 这条线本身要跑上好几个月。一个只跑十几周的试验, 量到的相当一部分还是开工前就已经在管线里的旧库存同一个人两次送检, 本来就会差很多。 精液指标受禁欲天数、上一次发烧、季节、取样是否完整的影响都很大, 这种天然波动常常盖过一个补剂可能有的效果。所以在这个终点上, 没有对照组的前后对比几乎不含信息样本小的时候, 阴性不等于没效, 只等于看不清。 那个安慰剂对照试验里的不显著趋势就是这种状态 —— 它没有能力回答问题, 而不是回答了没有
把三条合起来: 现有数据允许你说这里有一个值得再查的信号, 不允许你说它提高了生育力。这两句话之间隔着的, 正是一个足够长、足够大、带对照的试验 —— 而它至今没有被做出来。
Chapter 5
Safety + sourcing + decision · should you take it
Safety + sourcing + decision · should you take it
Good news: as a food, maca is quite safe short-term.
LiverTox (NIH's authoritative drug-induced liver injury database) rates maca's hepatotoxicity likelihood as 'E (unlikely)', with only 1 case reported worldwide (a maca medicinal liquor, 2017)Short-term (≤12 wk) adverse effects are mostly mild, transient GI upset / headacheTraditional dietary dose is 1.5–3.5 g/day (dried powder)
But there are two real catches, both unrelated to 'natural' and entirely about the 'specific product':
① The thyroid / goitrogen concern of RAW maca
Maca is a crucifer containing glucosinolates — theoretically goitrogenicThe risk is mainly in people who eat it raw + high-dose + already iodine-deficientThis is exactly why Andean tradition boils maca, and commercial forms are often gelatinized: heat degrades raw glucosinolates and improves digestibility. (The frequently quoted '~90% removed by boiling and discarding the water' is a general cruciferous-vegetable figure, not a maca measurement — the safe statement is 'a substantial fraction')The honest boundary: there is no maca-specific long-term thyroid-outcome trial, so this is a 'precaution', not a 'documented harm'. The broader cruciferous evidence is reassuring (cooked is basically fine)
② Heavy metals from mining-affected Andean soils (the one to take seriously)
Mendoza 2021 (Toxicology Reports): across three mining-affected districts of Junín, maca measured cadmium 0.32 ± 0.23 mg/kg and lead 0.20 ± 0.12 mg/kg, both above the FAO/WHO limits. Modeled As / Cd cancer risk exceeded the 1 × 10⁻⁶ tolerable line in both children and adults, and in the Ondores district children's arsenic risk also exceeded 1 × 10⁻⁴The same paper's other half belongs here too: the bioconcentration factor was below 1, the estimated daily intake of each metal was below the oral reference dose, and the hazard quotient and hazard index were below 1, from which the authors judge non-cancer harm unlikelyBoth halves together sharpen rather than soften the point: the near-term non-cancer risk is low; what breaches the threshold is the long-run cancer model — because cadmium is a cumulative toxin. A little each day, and the bill arrives years laterThis is a product-quality / provenance issue, not a pharmacology oneCountermeasure: choose products with third-party heavy-metal testing (Cd / Pb / As 'not detected' or far below limits) + stated provenance + a reputable brand
Pregnancy / lactation: insufficient human data → standard 'avoid / insufficient evidence'.
Decision tree (honest version):
Want to 'boost testosterone / balance hormones' → maca cannot (proven on this page + the prior layer); don't buy it for this reasonMild menopausal symptoms + want a gentle, low-risk adjunct → a 4–8 week trial is reasonable, but first nail sleep / exercise / alcohol limits / stress management (effect size far larger than maca); for severe symptoms, evaluate MHTInfertility / moderate-severe ED → see reproductive medicine / urology; maca is not a treatmentIf you decide to try: 1.5–3.5 g/day gelatinized powder, pick a third-party heavy-metal-tested product, evaluate subjectively at 4–8 weeks, stop if you feel nothing — don't do faith-based daily use forever
Bottom line: maca is an Andean, basically safe food with a weak, possibly real, hormone-independent subjective effect on libido / mood. It is not a hormone booster, not a proven fertility / ED treatment, and not an ergogenic aid. Anyone selling it as 'natural testosterone / hormone balance' is refuted by maca's own foundational trials — which is exactly the rule of this adaptogen island: ask what the evidence says before you ask what the marketing says.
LiverTox (NIH's authoritative drug-induced liver injury database) rates maca's hepatotoxicity likelihood as 'E (unlikely)', with only 1 case reported worldwide (a maca medicinal liquor, 2017)Short-term (≤12 wk) adverse effects are mostly mild, transient GI upset / headacheTraditional dietary dose is 1.5–3.5 g/day (dried powder)
But there are two real catches, both unrelated to 'natural' and entirely about the 'specific product':
① The thyroid / goitrogen concern of RAW maca
Maca is a crucifer containing glucosinolates — theoretically goitrogenicThe risk is mainly in people who eat it raw + high-dose + already iodine-deficientThis is exactly why Andean tradition boils maca, and commercial forms are often gelatinized: heat degrades raw glucosinolates and improves digestibility. (The frequently quoted '~90% removed by boiling and discarding the water' is a general cruciferous-vegetable figure, not a maca measurement — the safe statement is 'a substantial fraction')The honest boundary: there is no maca-specific long-term thyroid-outcome trial, so this is a 'precaution', not a 'documented harm'. The broader cruciferous evidence is reassuring (cooked is basically fine)
② Heavy metals from mining-affected Andean soils (the one to take seriously)
Mendoza 2021 (Toxicology Reports): across three mining-affected districts of Junín, maca measured cadmium 0.32 ± 0.23 mg/kg and lead 0.20 ± 0.12 mg/kg, both above the FAO/WHO limits. Modeled As / Cd cancer risk exceeded the 1 × 10⁻⁶ tolerable line in both children and adults, and in the Ondores district children's arsenic risk also exceeded 1 × 10⁻⁴The same paper's other half belongs here too: the bioconcentration factor was below 1, the estimated daily intake of each metal was below the oral reference dose, and the hazard quotient and hazard index were below 1, from which the authors judge non-cancer harm unlikelyBoth halves together sharpen rather than soften the point: the near-term non-cancer risk is low; what breaches the threshold is the long-run cancer model — because cadmium is a cumulative toxin. A little each day, and the bill arrives years laterThis is a product-quality / provenance issue, not a pharmacology oneCountermeasure: choose products with third-party heavy-metal testing (Cd / Pb / As 'not detected' or far below limits) + stated provenance + a reputable brand
Pregnancy / lactation: insufficient human data → standard 'avoid / insufficient evidence'.
Decision tree (honest version):
Want to 'boost testosterone / balance hormones' → maca cannot (proven on this page + the prior layer); don't buy it for this reasonMild menopausal symptoms + want a gentle, low-risk adjunct → a 4–8 week trial is reasonable, but first nail sleep / exercise / alcohol limits / stress management (effect size far larger than maca); for severe symptoms, evaluate MHTInfertility / moderate-severe ED → see reproductive medicine / urology; maca is not a treatmentIf you decide to try: 1.5–3.5 g/day gelatinized powder, pick a third-party heavy-metal-tested product, evaluate subjectively at 4–8 weeks, stop if you feel nothing — don't do faith-based daily use forever
Bottom line: maca is an Andean, basically safe food with a weak, possibly real, hormone-independent subjective effect on libido / mood. It is not a hormone booster, not a proven fertility / ED treatment, and not an ergogenic aid. Anyone selling it as 'natural testosterone / hormone balance' is refuted by maca's own foundational trials — which is exactly the rule of this adaptogen island: ask what the evidence says before you ask what the marketing says.
安全 · 短期数据与传统剂量
好消息: 作为食物, 玛咖短期相当安全。LiverTox (NIH 权威药物肝损伤库) 把玛咖的肝毒可能性评为E (不太可能), 全球只有 1 例报告 (还是一种玛咖药酒, 2017)短期 (≤12 周) 不良反应多是轻微、一过性的胃肠不适、头痛传统膳食剂量 1.5-3.5 g/天 (干粉)
把这三行连起来读, 它说的是一件很具体的事: 在食物级的量和几个月的时间尺度上, 玛咖没有留下值得警惕的信号。这也是它和很多提取物类补剂真正的区别 —— 它本来就是一种被人类当饭吃了很久的作物, 而不是从植物里浓缩出来的单一化合物。
但短期安全不等于长期安全, 这两句话在证据上完全不是一回事: 前者有一批为期数周到十几周的试验撑着, 后者在玛咖身上根本没有数据。所以下面两页要谈的两个风险, 一个 (甲状腺) 属于时间够长才可能显形, 另一个 (重金属) 属于必须靠年计的累积才结账。它们都不会出现在一个十几周的不良反应表里。
机制 · 生玛咖为什么会挤到甲状腺
甲状腺要造激素, 第一步是把血里的碘搬进甲状腺细胞。细胞膜上有一道专门的门 (钠碘同向转运体), 它认的是碘离子的个头和电荷。问题出在这里: 十字花科的 glucosinolate 被咬碎、被酶切开之后, 产物里有一类叫硫氰酸盐的小离子, 个头和电荷恰好和碘接近。于是它跑去排那道门的队, 把碘挤在外面。碘进得少 → 造甲状腺激素的原料不够 → 血里的甲状腺激素往下走 → 垂体察觉后加大呼喊 (化验单上就是 thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. 升高) → 甲状腺被喊着长大。甲状腺肿这个词就是这么来的。
请注意这是一条抢位子的机制, 不是一条毒性机制。 抢位子的结果取决于两边的量: 碘足够多的时候, 硫氰酸盐排不过它; 碘本来就少的时候, 这一挤才挤得动。所以真正的风险条件是三件事同时成立: 生吃 (植物自己那把酶还活着) + 量大 + 本身缺碘。少了任何一条, 这条链就接不上。
这也让两个看起来只关乎口味和口感的做法, 露出了它们的安全含义:
安第斯传统先煮玛咖 —— 加热让那把酶失活, 也把一部分 glucosinolate 直接破坏掉。⚠️ 你常看到的弃水煮能去掉约 90%, 是十字花科蔬菜上的一个通用数字, 不是在玛咖身上量出来的 —— 保守的说法是: 煮过并弃水能去掉相当大一部分, 具体多少玛咖自己没有数据商业上常做成 gelatinized (糊化) 形态 —— 同样经过热处理, 顺带改善消化
诚实边界: 没有专门针对玛咖的长期甲状腺结局试验, 所以这是一个谨慎项, 不是已证实的伤害。十字花科整体的证据是安心的 —— 煮过基本无碍, 常规份量的西兰花、卷心菜从来不是甲状腺问题的原因。
你可以自己往下推的一步: 同样一勺粉, 对一个碘摄入充足的人几乎不构成问题, 对一个本身碘就紧、又天天生吞大量粉末的人, 风险才开始变得可谈。剂量、形态、和你自己的碘状态, 三者一起决定结果 —— 这比它有没有毒这种问法有用得多。
临床 · 重金属: 为什么偏偏找上根茎
Mendoza 2021 (Toxicology Reports): 受采矿影响的秘鲁胡宁地区三个 district, 玛咖测出镉 0.32 ± 0.23 mg/kg + 铅 0.20 ± 0.12 mg/kg, 两者都超过 FAO/WHO 设定的限值。砷和镉的致癌风险在儿童和成人身上都超过了 1×10⁻⁶ 这条可容忍线; 其中 Ondores 这个 district 的儿童砷风险还超过了 1×10⁻⁴。但同一篇论文自己的另一半结论也要写出来: 它算出的生物富集系数小于 1, 每种金属的估计每日摄入量都低于口服参考剂量, 危害商数 (HQ) 与危害指数 (HI) 也都小于 1 —— 作者据此判断不太可能造成非致癌的健康损害。
两半合起来才是完整的图景, 而且它反而让这一节真正的论点更清楚: 眼下的非致癌风险是低的, 被模型顶出阈值的是长期的致癌风险 —— 因为镉是一种累积毒物, 它的账不在今天结。下面这一段解释为什么。
为什么偏偏是这种作物? 两件事凑到了一起:
你吃的正是它埋在土里的储藏器官。 叶菜的可食部分长在空气里, 而玛咖的下胚轴整个泡在土壤溶液中, 又恰好是植物往里囤东西的地方 —— 土里有什么, 它就有条件囤什么根系的搬运通道分不清镉。 植物需要锌这类金属, 于是在根细胞上装了专门的转运蛋白; 镉的化学性质和它们太像, 会被这些通道顺手一起搬进来, 再随着向上的水流分配到储藏组织里
为什么镉的账要过很多年才结? 因为它不像急性毒物那样当场发作。进入身体的镉主要被肝和肾扣下, 和一种专门捆金属的小蛋白结合后, 随血流送到肾脏; 肾小管把这个复合物整个回收进细胞, 蛋白被拆掉, 镉就留了下来。它排出去的速度极慢, 于是一年一年往肾皮质里累。等到肾小管细胞被伤到, 表现是本该被回收的小分子蛋白开始漏进尿里 —— 而那时候, 累积早已完成。
所以对策不在剂量, 在来源。 这是产品质量和产地的问题, 不是药理问题: 两罐同样标着玛咖粉的东西, 镉含量可以差出很远。要选有第三方重金属检测 (镉、铅、砷未检出或远低于限值) + 标明产地 + 信誉品牌的产品。
这一条的适用面远比玛咖大。 凡是吃地下部分的作物 —— 根、块茎、膨大的茎 —— 都比吃地上部分更受土壤污染影响; 而凡是每天吃、吃很多年的东西, 都值得多花一点力气确认来源。吃得越久, 来源比剂量更重要。
决策 · 该不该吃 (诚实版)
决策树 (诚实版):想补睾酮、平衡激素 → 玛咖做不到 (本页 + 上一层已证), 别买这个理由更年期症状轻 + 想试温和、低风险的辅助 → 可以 4-8 周试, 但先把 睡眠、运动、限酒、压力管理 做到位 (effect size 远大于玛咖), 症状重则评估 MHT不育、中重度 ED → 看生殖科、泌尿科, 玛咖不是治疗如果决定试: 1.5-3.5 g/天 gelatinized 粉, 选第三方重金属检测产品, 4-8 周主观评估, 没感觉就停 —— 别长期信仰式每天吃
为什么要给自己定一个停下来的日子? 因为这一岛已经把玛咖的效应形状讲清楚了: 它落在主观感受那一端, 而主观感受最容易被期待推动。你花了钱、等了几周, 大脑在打分的那一刻会自动把这些一起读进去。定一个日子、定一个你在乎的具体感受 (不是整体状态这种模糊说法), 到日子诚实地回答一次 —— 这是你能给自己搭的、最接近安慰剂对照的东西。
底线: 玛咖是一种安第斯的、基本安全的食物, 在性欲、情绪上有弱的、可能真实的、与激素无关的主观影响。它不是激素助推器、不是已证实的生育 / ED 疗法、不是增强运动表现的 ergogenic。任何把它当天然睾酮、激素平衡卖的人, 都被玛咖自己的奠基试验否决了 —— 这正是这座适应原岛的规矩: 先问证据说什么, 再问营销说什么。
References · 15
- Ulloa del Carpio, N., Alvarado-Corella, D., Quiñones-Laveriano, D. M., Araya-Sibaja, A., Vega-Baudrit, J., Monagas-Juan, M., Navarro-Hoyos, M., & Villar-López, M. (2024). Exploring the chemical and pharmacological variability of Lepidium meyenii: A comprehensive review of the effects of maca. Frontiers in Pharmacology, 15, 1360422. 10.3389/fphar.2024.1360422
- Gasco, M., Villegas, L., Yucra, S., Rubio, J., & Gonzales, G. F. (2007). Dose–response effect of red maca (Lepidium meyenii) on benign prostatic hyperplasia induced by testosterone enanthate. Phytomedicine, 14(7–8), 460–464. 10.1016/j.phymed.2006.12.003
- Rubio, J., Qiong, W., Liu, X., Jiang, Z., Dang, H., Chen, S.-L., & Gonzales, G. F. (2011). Aqueous extract of black maca (Lepidium meyenii) on memory impairment induced by ovariectomy in mice. Evidence-Based Complementary and Alternative Medicine, 2011, 253958. 10.1093/ecam/nen063
- National Institute of Diabetes and Digestive and Kidney Diseases. (2019). Maca. In LiverTox: Clinical and Research Information on Drug-Induced Liver Injury. Bethesda, MD: NIDDK. Bookshelf ID NBK548552. www.ncbi.nlm.nih.gov/books/NBK548552
- Gonzales, G. F., Córdova, A., Vega, K., Chung, A., Villena, A., & Góñez, C. (2003). Effect of Lepidium meyenii (Maca), a root with aphrodisiac and fertility-enhancing properties, on serum reproductive hormone levels in adult healthy men. Journal of Endocrinology, 176(1), 163–168. 10.1677/joe.0.1760163
- Gonzales, G. F., Córdova, A., Vega, K., Chung, A., Villena, A., Góñez, C., & Castillo, S. (2002). Effect of Lepidium meyenii (MACA) on sexual desire and its absent relationship with serum testosterone levels in adult healthy men. Andrologia, 34(6), 367–372. 10.1046/j.1439-0272.2002.00519.x
- Shin, B.-C., Lee, M. S., Yang, E. J., Lim, H.-S., & Ernst, E. (2010). Maca (L. meyenii) for improving sexual function: A systematic review. BMC Complementary and Alternative Medicine, 10, 44. 10.1186/1472-6882-10-44
- Zenico, T., Cicero, A. F. G., Valmorri, L., Mercuriali, M., & Bercovich, E. (2009). Subjective effects of Lepidium meyenii (Maca) extract on well-being and sexual performances in patients with mild erectile dysfunction: A randomised, double-blind clinical trial. Andrologia, 41(2), 95–99. 10.1111/j.1439-0272.2008.00892.x
- Stone, M., Ibarra, A., Roller, M., Zangara, A., & Stevenson, E. (2009). A pilot investigation into the effect of maca supplementation on physical activity and sexual desire in sportsmen. Journal of Ethnopharmacology, 126(3), 574–576. 10.1016/j.jep.2009.09.012
- Brooks, N. A., Wilcox, G., Walker, K. Z., Ashton, J. F., Cox, M. B., & Stojanovska, L. (2008). Beneficial effects of Lepidium meyenii (Maca) on psychological symptoms and measures of sexual dysfunction in postmenopausal women are not related to estrogen or androgen content. Menopause, 15(6), 1157–1162. 10.1097/gme.0b013e3181732953
- Stojanovska, L., Law, C., Lai, B., Chung, T., Nelson, K., Day, S., Apostolopoulos, V., & Haines, C. (2015). Maca reduces blood pressure and depression, in a pilot study in postmenopausal women. Climacteric, 18(1), 69–78. 10.3109/13697137.2014.929649
- Lee, M. S., Shin, B.-C., Yang, E. J., Lim, H.-J., & Ernst, E. (2011). Maca (Lepidium meyenii) for treatment of menopausal symptoms: A systematic review. Maturitas, 70(3), 227–233. 10.1016/j.maturitas.2011.07.017
- Gonzales, G. F., Córdova, A., Gonzales, C., Chung, A., Vega, K., & Villena, A. (2001). Lepidium meyenii (Maca) improved semen parameters in adult men. Asian Journal of Andrology, 3(4), 301–303. pubmed.ncbi.nlm.nih.gov/11753476
- Melnikovova, I., Fait, T., Kolarova, M., Fernandez, E. C., & Milella, L. (2015). Effect of Lepidium meyenii Walp. on semen parameters and serum hormone levels in healthy adult men: A double-blind, randomized, placebo-controlled pilot study. Evidence-Based Complementary and Alternative Medicine, 2015, 324369. 10.1155/2015/324369
- Orellana Mendoza, E., Cuadrado, W., Yallico, L., Zárate, R., Quispe-Melgar, H. R., Limaymanta, C. H., Sarapura, V., & Bao-Cóndor, D. (2021). Heavy metals in soils and edible tissues of Lepidium meyenii (maca) and health risk assessment in areas influenced by mining activity in the Central region of Peru. Toxicology Reports, 8, 1461–1470. 10.1016/j.toxrep.2021.07.016