Place · Level 3
Multivitamin · #1 supplement worldwide
全球年销 $50 B+ · 30 亿粒 / 年 · USPSTF 2022: 健康人群 CV / 癌症预防证据不足 · 但特定人群确有获益
Last updated
Story path
Chapter 1
1747-1940s · vitamin era
1747-1940s · vitamin era
The discovery of vitamins is one of the most brilliant chapters of 19th-20th century nutrition science — in under two centuries, a handful of physicians and chemists walked from "why do sailors get scurvy?" all the way to "folate-fortified flour", and the classic deficiency diseases nearly vanished from developed countries. But this same chapter planted the seed of the extended expectation that "supplementing more prevents more disease" — which 50 subsequent years of RCTs had to disentangle:
1747 — James Lind ran the first human controlled trial aboard HMS Salisbury: 12 scurvy sailors split into 6 groups eating different diets → the citrus group recovered in 5 days. First evidence that 'something in food cures disease' (though he didn't isolate vitamin C — that waited for Szent-Györgyi in 1932).
1881 — Russian chemist Lunin showed that purified nutrients alone couldn't sustain animals — there had to be an extra factor in food.
1897 — Dutch physician Eijkman in a Java prison: inmates eating polished rice got beriberi, those eating brown rice didn't — he had discovered an anti-beriberi factor in rice husk.
1912 — Polish chemist Casimir Funk coined the word 'vitamine' (vital amine, later dropping the 'e'). Hypothesis: multiple diseases (scurvy, pellagra, beriberi, rickets) all stem from missing vital factors.
1920-1940s — vitamins were isolated, structurally characterized, and industrially synthesized one by one:
Vitamin C (1932), D (1922-1936), A (1931), B1 (1936), B2 (1933), K (1939), B12 (1948)The success of single vitamins curing specific deficiency diseases during this era sparked the extended hope that multivitamins could prevent multiple chronic diseases.
Post-1940s: Pauling and others pushed megadoses, commercial explosion → the modern multivitamin industry was born here.
The 'expectation' and 'evidence' gap existed from the start — Funk's hypothesis was correct (deficiency diseases are real), but the extended hypothesis that additional supplementation could prevent chronic disease has largely failed in 50 subsequent years of RCTs (see scenes 3 and 4).
1747 — James Lind ran the first human controlled trial aboard HMS Salisbury: 12 scurvy sailors split into 6 groups eating different diets → the citrus group recovered in 5 days. First evidence that 'something in food cures disease' (though he didn't isolate vitamin C — that waited for Szent-Györgyi in 1932).
1881 — Russian chemist Lunin showed that purified nutrients alone couldn't sustain animals — there had to be an extra factor in food.
1897 — Dutch physician Eijkman in a Java prison: inmates eating polished rice got beriberi, those eating brown rice didn't — he had discovered an anti-beriberi factor in rice husk.
1912 — Polish chemist Casimir Funk coined the word 'vitamine' (vital amine, later dropping the 'e'). Hypothesis: multiple diseases (scurvy, pellagra, beriberi, rickets) all stem from missing vital factors.
1920-1940s — vitamins were isolated, structurally characterized, and industrially synthesized one by one:
Vitamin C (1932), D (1922-1936), A (1931), B1 (1936), B2 (1933), K (1939), B12 (1948)The success of single vitamins curing specific deficiency diseases during this era sparked the extended hope that multivitamins could prevent multiple chronic diseases.
Post-1940s: Pauling and others pushed megadoses, commercial explosion → the modern multivitamin industry was born here.
The 'expectation' and 'evidence' gap existed from the start — Funk's hypothesis was correct (deficiency diseases are real), but the extended hypothesis that additional supplementation could prevent chronic disease has largely failed in 50 subsequent years of RCTs (see scenes 3 and 4).
够用之后为什么就平了
身体处理维生素的方式, 自带一个上限。 上一屏那条曲线为什么会平掉, 答案分成两半, 而这两半的走向刚好相反。水溶性的那一半 (B 族和维 C) 会撞到两道闸。 第一道在小肠: 每种维生素都靠自己那批转运蛋白被搬进肠壁细胞, 转运蛋白的数量是固定的, 一次涌进来太多, 排不上队的部分就原样往下走。第二道在肾脏: 血里的这些分子被肾小球滤出去之后, 肾小管会把身体还用得着的那部分捞回来——捞的能力同样有上限, 超过的部分直接留在尿里。所以大剂量的水溶性维生素, 相当一部分是从嘴进、从尿出, 中间只在血里晃了几个钟头。核黄素 (B2) 本身是亮黄色的, 吃完复合维生素尿会发黄, 看到的就是它被排掉的样子——那既不说明吸收得好, 也不说明全浪费了, 只说明你摄入的量已经越过了身体当下的需要。
脂溶性的那一半 (维 A、D、E、K) 没有这两道闸。 它们不溶于水, 走的是脂肪的路线: 在小肠里要先被胆汁酸包成一个个极小的油滴 (混合微团) 才能靠岸卸货, 进了肠壁细胞之后也不直接进血管, 而是被打包成乳糜微粒经淋巴绕一圈才汇入血液, 最后大部分存进肝脏和脂肪组织。存得下, 就排不快——这一句同时解释了两件看似无关的事: 为什么脂溶性维生素能让你撑过一段没有阳光、没有青菜的日子, 以及为什么它们是长期大剂量下唯一真会在体内累积到有害的那一类。
把两半合起来看, 就得到开关与旋钮的分别。一个真的有缺口的人, 补进去的每一份都被立刻用掉, 症状跟着改善; 一个本来就够用的人, 水溶性的多余部分几小时内进了马桶, 脂溶性的多余部分留在了肝里。前者是在修一条断掉的线, 后者只是在给一条已经通畅的线加压——而管道的流量, 不由压力决定, 由最窄的那一段决定。
维生素是怎么被找出来的
单一维生素治好一种特定缺乏病, 当年是实打实的奇迹。 复合维生素这门生意, 就是从这个奇迹的延长线上长出来的。二十世纪上半叶, 几位医生和化学家一个接一个把每种维生素找了出来。1747 年, James Lind 在英国海军船 HMS Salisbury 上做了人类第一个对照实验: 12 个坏血病水兵分 6 组吃不同食物, 吃橘子柠檬的那组 5 天就好了——这是第一次证明食物里有某种东西能治病 (但他没分离出维 C, 那要等到 1932 年 Szent-Györgyi)。1912 年, 波兰化学家 Casimir Funk 给这类生命必需的微量分子起名叫维生素 (vitamine), 并提出: 坏血病、糙皮病、脚气病、佝偻病这些病, 各自都是缺了某一种这样的因子。到 1920–1940 年代, 维 C、维 D、维 A、各种 B 族、维 K、维 B12 被逐一分离、定出结构、再工业合成出来。
问题出在后面那一步推广: 既然补一种能治一种病, 那一次补齐很多种, 是不是就能防住很多种慢病? 1940 年代起 Pauling 等人鼓吹大剂量, 现代复合维生素行业就是从这个期望里长出来的。
Funk 那一代人没说错——缺乏症是真的, 补上就好。错的是延伸出来的那句多补一点就能少生病: 这句话后续 50 年被一项项 RCT 反复检验, 大部分没站住。而回头看, 它错在哪里其实上一页已经写清楚了: 他们观察到的是曲线最左边那一段陡坡, 却以为整条曲线都是那个斜率。
The 'vitamine' coinage
Funk 1912's core insight + naming mistake:Isolated the anti-beriberi factor from rice bran (actually thiamine / vitamin B1)Measured it as nitrogen-containing + an amineCoined the term 'vital amine = vitamine' for this class of 'essential organic micromolecules for life'Later discovered: not all vitamins are amines (C isn't, D isn't) — so in 1920 Drummond proposed dropping the 'e' → 'vitamin'
Funk's other great insight: classifying the era's 'dietary deficiency diseases' as one category — this drove 20th-century public health programs:
Iodized salt (1924 Michigan, USA) → prevent goiterFortified flour (1941 USA) → add iron / niacin / thiamine / riboflavin to prevent beriberi / pellagra / iron-deficiency anemiaVitamin D-fortified milk (1933 USA) → prevent ricketsFolic acid-fortified grain (1998 USA) → prevent neural tube defects (NTD)
This is one of the greatest public health achievements of the 20th century: classic deficiency diseases were nearly eliminated in developed countries at minimal cost.
But 'fortified food + adequate diet = deficiency nearly extinct' also means: in 21st-century developed countries, the expected benefit of average adults taking multivitamins is actually very limited — this is the hidden reason for subsequent RCT failures. When most people's blood nutrient levels are already 'sufficient', adding more doesn't make them healthier.
Chapter 2
What's actually in the bottle
What's actually in the bottle
Typical 'Men's 50+ multivitamin' label — unpacking the truth:
A typical MVI capsule contains:
13 vitamins (A / C / D / E / K / B1-B12 / folate / biotin / pantothenic acid)10-20 minerals (Ca / Mg / Fe / Zn / Cu / Mn / Se / Cr / Mo / iodine / K / etc.)Added ingredients (lutein / lycopene / various 'protective blends')Excipients (magnesium stearate / microcrystalline cellulose / silicon dioxide / colorants / sugar coating)
Typical doses:
Most vitamins = 100% RDA (Recommended Daily Allowance)Some exceed RDA (especially B-complex + vitamin D, modern trend)Minerals are usually below RDA (because large mineral amounts make the pill huge to swallow + they interact with each other)Typical Ca only 200-300 mg (< 1/3 RDA)Mg only 50-100 mg (< 1/4 RDA)This is why MVI cannot replace targeted Ca + Mg supplementation
A few ingredient forms worth attention:
Folate: most use synthetic folic acid (cheap); carriers of MTHFR mutations (10-25% of population) have lower conversion efficiency — but clinical significance is limited, don't be misled by 'activated folate / 5-MTHF' marketingVitamin D: most use D3 (cholecalciferol); some still use D2 (ergocalciferol, lower potency)Vitamin K: K1 (phylloquinone) is default; K2 (menaquinone) appears in high-end brands — good thing but not worth paying a premium forIron: men's / 50+ formulas usually iron-free (adult men and postmenopausal women don't need supplementation; iron-containing MVI carries overdose risk)β-carotene: early MVIs commonly included it, modern mainstream removed (ATBC / CARET trials showed harm to smokers, see scene 4)
Misleading aspects of the '100% Daily Value' label:
DV is calculated based on outdated 1968 RDA — not equivalent to modern nutrition recommendations'Contains X% DV' does not equal 'your individual need'
A typical MVI capsule contains:
13 vitamins (A / C / D / E / K / B1-B12 / folate / biotin / pantothenic acid)10-20 minerals (Ca / Mg / Fe / Zn / Cu / Mn / Se / Cr / Mo / iodine / K / etc.)Added ingredients (lutein / lycopene / various 'protective blends')Excipients (magnesium stearate / microcrystalline cellulose / silicon dioxide / colorants / sugar coating)
Typical doses:
Most vitamins = 100% RDA (Recommended Daily Allowance)Some exceed RDA (especially B-complex + vitamin D, modern trend)Minerals are usually below RDA (because large mineral amounts make the pill huge to swallow + they interact with each other)Typical Ca only 200-300 mg (< 1/3 RDA)Mg only 50-100 mg (< 1/4 RDA)This is why MVI cannot replace targeted Ca + Mg supplementation
A few ingredient forms worth attention:
Folate: most use synthetic folic acid (cheap); carriers of MTHFR mutations (10-25% of population) have lower conversion efficiency — but clinical significance is limited, don't be misled by 'activated folate / 5-MTHF' marketingVitamin D: most use D3 (cholecalciferol); some still use D2 (ergocalciferol, lower potency)Vitamin K: K1 (phylloquinone) is default; K2 (menaquinone) appears in high-end brands — good thing but not worth paying a premium forIron: men's / 50+ formulas usually iron-free (adult men and postmenopausal women don't need supplementation; iron-containing MVI carries overdose risk)β-carotene: early MVIs commonly included it, modern mainstream removed (ATBC / CARET trials showed harm to smokers, see scene 4)
Misleading aspects of the '100% Daily Value' label:
DV is calculated based on outdated 1968 RDA — not equivalent to modern nutrition recommendations'Contains X% DV' does not equal 'your individual need'
同一片里谁在抢谁
二价金属共用一条通道, 这是复合维生素最被低估的设计难题。小肠上皮细胞面向肠腔的那一侧, 有一个叫二价金属转运体 (DMT1) 的蛋白, 负责把带两个正电荷的金属离子搬进细胞。铁 (二价形式)、锌、铜、锰、钴都能被它认出来——认得越杂, 就越容易堵车。你把这几样一起吞下去, 它们在同一段肠腔里同时溶解、同时到达同一个门口, 谁的浓度高谁被搬进去的次数就多, 其余的继续往下走, 最后随粪便离开。
这条通道解释了标签上几件看起来互相矛盾的事:
锌补高了会缺铜。 锌进入肠壁细胞后会诱导细胞里一种叫金属硫蛋白的粘手分子; 铜对它的亲和力比锌还强, 于是铜被粘在细胞里出不去, 等这个肠壁细胞几天后自然脱落进肠腔, 铜就跟着一起被扔掉了。长期高锌造成铜缺乏, 走的正是这条路——注意它是一个时间差: 你今天多补的锌, 是在几天后才把铜扣掉的。钙和铁不该同时上。 钙不走这条通道, 但它在肠腔里一多, 铁的吸收就是会掉。这是复合维生素把钙压得很低的原因之一, 也是钙片和铁剂错开吃这条老建议的出处。维 C 和铁反而是一对。 植物性食物里的铁多是三价的, 这个门搬不动; 维 C 在肠腔里把它还原成二价, 顺手拉住它不被别的东西沉淀掉, 铁这才进得去。同一片里维 C 和铁同框, 是少数真正互相帮忙的组合。茶、咖啡、全谷里的植酸和多酚会在肠腔里直接和金属离子抱成一团搬不动的东西, 谁都进不去。这跟片剂内部谁挤谁是两回事, 但结果是叠加的——一杯浓茶配一粒含铁复合维生素, 两种损失同时发生。
所以分开吃这条建议不是玄学。 如果你确实需要补铁, 把它和含钙的那一顿或那一片错开、再配一点维 C, 真正进到身体里的量会明显不同; 而一颗什么都有的复合维生素, 恰恰是把所有会互相干扰的东西压在了同一分钟。这也是它为什么替代不了针对性补充: 不是剂量不够写不上标签, 是同片共存这个前提, 从一开始就把每一样的剂量上限锁死了。
标签上值得看的几行
一瓶典型的男士 50+ 配方, 翻过来看背面, 大致是这几类东西凑在一起:十三种维生素: A、C、D、E、K, 加上整组 B 族 (B1 到 B12)、叶酸、生物素、泛酸十到二十种矿物质: 钙、镁、铁、锌、铜、锰、硒、铬、钼、碘、钾等一些额外添加物, 比如叶黄素 (lutein)、番茄红素 (lycopene), 以及标着各种护眼护心的混合配方一堆辅料: 硬脂酸镁、微晶纤维素、二氧化硅、着色剂和糖衣, 用来把上面这些压成一颗能吞的药片
剂量上有个容易被忽略的不对称。大多数维生素配到 100% 的每日推荐量 (recommended daily allowance, RDA), 个别 B 族和维 D 还会超出一些。但矿物质几乎总是配得偏低——因为矿物质又重又占地方, 真按 RDA 放进去药片会大到吞不下, 而且它们彼此还会在瓶子里反应。所以典型一粒里的钙常常只有 200-300 mg (不到 RDA 的三分之一), 镁只有 50-100 mg (不到四分之一)。这就是为什么复合维生素替代不了针对性的钙、镁补充。
标签上几个成分形式值得你瞄一眼:
叶酸多数用便宜的合成形式 (folic acid)。带 MTHFR 基因变异的人 (约占人群 10-25%) 把它转成活性形式的效率偏低, 但这件事临床意义有限, 不必为标着活性叶酸 (5-MTHF) 的版本多花钱。维 D 多数是 D3 (cholecalciferol), 也有少数还在用效力更低的 D2 (ergocalciferol)。维 K 默认是 K1 (phylloquinone); 高端品牌会换成 K2 (menaquinone), 是好事, 但不值得为它付溢价。铁在男士和 50+ 配方里通常不放——成年男性和绝经后女性本来就不缺, 而身体没有主动排铁的出口, 多出来的只能往肝、心、胰腺里存。这里还有两条比补过头更硬的提醒: 家里有小孩, 含铁的瓶子必须锁起来——儿童误服成人铁剂是幼儿急性中毒致死最常见的原因之一, 美国法规为此强制在所有含铁的固体剂型上印警示语; 另外, 遗传性血色病 (HFE 基因携带者) 的人本身就在超量吸收铁, 盲补等于往火上浇油, 该做的是先查铁蛋白和转铁蛋白饱和度。β-胡萝卜素早期配方常加, 现在主流都拿掉了, 因为对吸烟者有害 (原因在第四幕)。
最后, 瓶身上那个醒目的每日参考值 (Daily Value, DV) 100%其实会误导人: 这个 DV 是按过时的 1968 年 RDA 算的, 既不等于现代营养建议, 更不等于你这个人当下真正缺多少。它描述的是一个统计学上的平均人, 而那个人不存在。
nih-ods-iron
跟不跟餐, 差一半吸收
同一粒药片, 空腹吞和跟着一顿有油水的饭吞, 进到血里的量可以差出一大截——差别全在脂溶性的那四种上。维 A、D、E、K 不溶于水。它们在肠腔里必须先被胆汁酸包成一个个极小的油滴 (混合微团), 才能被送到肠壁细胞表面卸货。而胆汁不是一直在排的: 它是在你吃进脂肪、十二指肠感知到之后才被挤出来的。空腹吞一粒复合维生素, 胆汁没怎么出来, 微团组不起来, 这四种维生素相当一部分只是从你身体里穿过去了——你付了全价, 拿到的却是打折后的那一半。
水溶性的那一组正相反: B 族和维 C 溶在肠液里, 靠各自的转运蛋白被搬进去, 完全不需要油。它们的限制在另一头——转运蛋白一饱和, 多出来的到不了血里; 已经到血里而超过肾小管回收能力的那部分, 几小时内进尿。
这两条合起来, 给出一条极少被写在瓶身上的操作建议: 复合维生素跟着一天里最有油水的那一餐吃。同一颗药、同样的钱、同样的剂量, 只是换了个时间点, 脂溶性那四种才真的到得了。
顺带说清一个常见的反向误会。 有人因为怕累积而刻意空腹吃脂溶性维生素——那只是让它不被吸收, 并不能让它更安全地被吸收。真正决定累积风险的是你每天摄入多少, 不是你什么时候吞。要控风险就调剂量, 别拿吸收率当刹车。
Natural vs synthetic, expensive vs cheap
'Natural vitamins are better than synthetic' is marketing — mostly doesn't hold up:Vitamin C: ascorbic acid is ascorbic acid — the molecular structure is identical, whether from oranges or industrial synthesis. 'Contains rose hips' or 'contains acerola' is marketing; the actual vitamin C content from rose hips in one pill is tiny (mg-level)Vitamin E: here there is a difference —d-α-tocopherol (natural) vs dl-α-tocopherol (synthetic) do differ in biological activity (~1.4 vs 1.0)But more importantly: synthetic dl-α-tocopherol is only 1 of nature's 8 tocopherols → can't represent the entire vitamin E familyWhole foods (almonds / sunflower seeds / spinach) naturally contain α + β + γ + δ tocopherols + tocotrienols → more completeVitamin D: D3 (cholecalciferol), whether sourced from sheep lanolin (natural) or chemically synthesized, is clinically equivalentB-complex: synthetic vs natural difference is essentially negligible
'Expensive = good' is often wrong too:
Costco Kirkland / Walgreens / other cheap brands match premium brands in third-party testing for label accuracy'Organic' / 'natural' / 'whole-food' are often 5-10× more expensive with minimal nutritional difference
A few cases where quality premiums are genuinely worth it:
Third-party certification (USP / NSF / ConsumerLab) — verifies label accuracy + no heavy metal contaminationClearly need higher purity (pregnancy / specific sensitivity)Specific forms (rTG fish oil / methyl-B12 etc.), but most people don't need these
'Pharmacy generic vs international brand' — no significant difference for general use; investing the money in food diversification yields higher returns.
Chapter 3
Who actually benefits
Who actually benefits
'MVI is useless' and 'MVI saves lives' are both oversimplifications — the truth is population-specific:
Groups that genuinely benefit (evidence-based):
① Strict vegetarians / vegans
B12 supplementation is mandatory (animal-only source) — long-term deficiency causes irreversible neurological damageIron / zinc / calcium / vitamin D / DHA often need attention
② Pregnancy / lactation
Folic acid (400-800 µg) preconception + early pregnancy → reduces NTD ~72% (MRC 1991 *Lancet*)Iodine (220 µg) → fetal brain developmentIron (pregnancy ↑) → both maternal and fetal needs riseVitamin D → fetal bone + maternal calcium homeostasisMost guidelines recommend pregnancy-specific multivitamins rather than regular MVI
③ Post-bariatric surgery (gastric bypass / sleeve)
Long-term malabsorption (B12 / Fe / Ca / vitamin D etc.)Lifelong need for specialized post-op MVI
④ Elderly (75+)
B12: atrophic gastritis + PPI use → reduced absorption (~15-20% of those >60)Vitamin D: skin synthesis efficiency declinesFood intake declines
⑤ Chronic disease / long-term medications
Long-term PPI use: B12 + Ca + Mg ↓Long-term metformin: B12 ↓ (~10-30% of users >5 years)Diuretics: K + Mg + B1 ↓Antiepileptics: D + folate + Ca ↓Chemo / radiation / long-term steroids: broad nutritional support
⑥ Severe dietary restriction / food allergies / eating disorders
Extremely low-calorie diets → comprehensive micronutrient crisis
Groups that don't need MVI:
Healthy adults with varied food intake (DGAC 'My Plate' style diet)Large RCTs tested mainly the cardiovascular and cancer endpoints and found no clinically meaningful benefit from 'daily MVI' in healthy people: the cardiovascular endpoint was flat, and the closest thing to a positive result is PHS-II's very small 8% reduction in total cancer (see next scene)Claims like 'more energy' were never an endpoint in these trials at all — they have not been refuted so much as never seriously tested
Groups that genuinely benefit (evidence-based):
① Strict vegetarians / vegans
B12 supplementation is mandatory (animal-only source) — long-term deficiency causes irreversible neurological damageIron / zinc / calcium / vitamin D / DHA often need attention
② Pregnancy / lactation
Folic acid (400-800 µg) preconception + early pregnancy → reduces NTD ~72% (MRC 1991 *Lancet*)Iodine (220 µg) → fetal brain developmentIron (pregnancy ↑) → both maternal and fetal needs riseVitamin D → fetal bone + maternal calcium homeostasisMost guidelines recommend pregnancy-specific multivitamins rather than regular MVI
③ Post-bariatric surgery (gastric bypass / sleeve)
Long-term malabsorption (B12 / Fe / Ca / vitamin D etc.)Lifelong need for specialized post-op MVI
④ Elderly (75+)
B12: atrophic gastritis + PPI use → reduced absorption (~15-20% of those >60)Vitamin D: skin synthesis efficiency declinesFood intake declines
⑤ Chronic disease / long-term medications
Long-term PPI use: B12 + Ca + Mg ↓Long-term metformin: B12 ↓ (~10-30% of users >5 years)Diuretics: K + Mg + B1 ↓Antiepileptics: D + folate + Ca ↓Chemo / radiation / long-term steroids: broad nutritional support
⑥ Severe dietary restriction / food allergies / eating disorders
Extremely low-calorie diets → comprehensive micronutrient crisis
Groups that don't need MVI:
Healthy adults with varied food intake (DGAC 'My Plate' style diet)Large RCTs tested mainly the cardiovascular and cancer endpoints and found no clinically meaningful benefit from 'daily MVI' in healthy people: the cardiovascular endpoint was flat, and the closest thing to a positive result is PHS-II's very small 8% reduction in total cancer (see next scene)Claims like 'more energy' were never an endpoint in these trials at all — they have not been refuted so much as never seriously tested
吸收端是怎么断的
吃进去 ≠ 吸收到, 中间还有一整条流水线可以断。 B12 的那条线是全身最长的, 拿它当例子最清楚:① 在胃里松绑: 胃酸和胃蛋白酶先把 B12 从食物蛋白上撬下来。胃酸不足 (老年萎缩性胃炎、长期吃抑酸药) 时这一步就撬不动。这里有个反直觉的结论: 补剂里的 B12 是游离的, 不需要这一步——所以同样是胃酸低的人, 吃药片反而比吃肉拿得到更多。② 换护送员: 胃壁细胞分泌一种叫内因子的蛋白, 专门抓住 B12 一路护送。自身免疫把这些细胞打掉, 就是恶性贫血; 这时候口服再多也接不上, 因为没有护送员——所以这种情况要走注射, 绕开整条肠道。③ 在回肠末端交货: 小肠最后那一段的细胞表面有专门认内因子的受体, B12 到这里才真正进入身体。这一段被切掉或被炎症破坏 (克罗恩病、部分减重手术), 前两步做对了也白做。
减重手术为什么要终身补, 同样看流水线断在哪: 胃被缩小, 胃酸和内因子都少了; 十二指肠和近端空肠被绕过去, 而那里恰恰是铁和钙吸收最集中的一段。这不是吃得少的问题, 是路被改道的问题——所以术后要的是针对配方, 不是货架上那种通用款, 而且是终身的, 因为改道不会自己长回去。
药物把营养素拉低, 也各有各的位置。 抑酸药堵的是第一步; 二甲双胍干扰的是回肠那个受体对内因子复合物的摄取, 而这一步依赖钙离子; 利尿剂把电解质和水溶性维生素一起从尿里冲走; 抗癫痫药加快肝里的代谢酶, 把维 D 拆得比补进去的还快。四种药、四个完全不同的位置——这就是为什么吃药的人要补维生素这句笼统的话没什么用: 补哪一样, 由药卡在哪一步决定。
所以要不要补问错了方向。该问的是我这条线断在哪一段: 断在摄入端, 换食物就够; 断在吸收端, 要用对形式和剂量; 断在流失端, 补的同时还得看能不能少漏。
谁真的获益 · 完整清单
这些人补, 是在接一条断掉的线:严格素食 / 纯素: B12 必补 (只有动物性食物提供)——长期缺乏会造成不可逆的神经损害; 铁、锌、钙、维 D 和 DHA 也常需要盯一眼。孕期与哺乳期: 叶酸 (folic acid 400-800 µg) 从孕前吃到早孕, 降神经管缺陷 (NTD) 约 72% (MRC 1991 Lancet); 碘 (220 µg) 管胎儿脑发育; 铁的需求整个孕期都在升; 维 D 同时管胎儿骨骼和母体钙稳态。多数指南推荐的是孕期专用配方, 而不是普通复合维生素——差别不只在剂量, 还有一条硬禁忌: 普通复合维生素和鱼肝油里的预成型维生素 A (视黄醇 / retinyl palmitate), 在早孕期是已确立的致畸原; 前瞻队列里, 每天从补剂摄入预成型维 A 超过约 10 000 IU 的孕妇, 颅神经嵴来源的出生缺陷风险明显升高 (Rothman 1995 NEJM)。β-胡萝卜素没有这个问题: 它在肠壁细胞里按需转成维 A, 转多少由身体的需求决定, 转不动的部分就不转。所以孕期配方要么把维 A 换成 β-胡萝卜素, 要么把预成型的那部分压得很低——货架上那瓶男士全效和一勺鱼肝油, 恰恰是这条线最容易被踩到的地方。减重手术后 (胃旁路 / 袖状胃): 长期吸收障碍 (B12 / 铁 / 钙 / 维 D 等), 终身需要术后专用配方。老年人 (75+): 萎缩性胃炎叠加抑酸药 (PPI) 让 B12 吸收下降 (60 岁以上约 15-20%); 皮肤合成维 D 的效率也在掉; 再加上食量本身变小。慢性病与长期用药: 长期 PPI 拉低 B12、钙、镁; 长期二甲双胍拉低 B12 (服药超过 5 年者约 10-30%); 利尿剂冲走钾、镁、B1; 抗癫痫药拉低维 D、叶酸、钙; 化疗放疗和长期类固醇则需要全面的营养支持。严重饮食受限、食物过敏、进食障碍: 极低热量饮食会同时打出多个微量营养素的缺口。
这些人不需要:
吃得多样、消化正常的健康成年人 (大致就是 DGAC 的 My Plate 那种吃法)。日常复合维生素能让健康人活得更久、少得慢病、或者更有精力? 大型 RCT 主要检验了心血管和癌症这两条, 结论是看不到有临床意义的获益: 心血管终点上是零, 癌症这边最接近阳性的一次是 PHS-II 那个很小的 8% 下降 (下一幕细说)。更有精力这一类说法则从来没有进入过这些试验的终点——它没有被推翻, 是因为它从一开始就没被认真检验过。
这张名单有一个共同点值得停一秒: 上半部分的每一条, 你都能指出断在哪里——没有来源、吸收被绕过、被药拖低、摄入太少。下半部分之所以没有获益, 不是因为营养不重要, 而是因为线是通的, 找不到那个断点。
rothman-vitamina-1995nih-ods-vitamin-a
USPSTF 2022 official guidance
The official 2022 guidance from the US Preventive Services Task Force (USPSTF) — synthesizing 84 studies (N >700,000). Those 84 are mostly RCTs but not all of them: the evidence report also included observational cohort studies in order to assess serious harms:For general adults (non-pregnant / lactating / non-diagnosed nutritional deficiency):
Multivitamins / minerals for CV disease + cancer prevention: Insufficient evidence to recommend use (Grade I) — 'no evidence it prevents, no evidence it harms'β-carotene for CV / cancer prevention: Recommend against use (Grade D) — raises risk in smokers (CARET / ATBC)Vitamin E alone for CV / cancer prevention: Recommend against use (Grade D) — SELECT trial showed prostate cancer ↑
In other words: the US authoritative medical preventive body does NOT recommend healthy people take MVI to prevent chronic disease, and explicitly opposes high-dose β-carotene + vitamin E alone.
The contrast with commercial promotion:
Global annual MVI sales $50 B+About 1/3 of US adults + 50% of elderly take MVI'Insurance-style supplementation' psychology = 'no harm anyway'But RCT data on β-carotene + high-dose vitamin E + high iron + high calcium clearly show 'real harm exists'
Practical recommendations:
Healthy adults: spend MVI money on vegetables + fruit + whole grains + quality protein — more valuable; dietary variety beats any MVISpecific populations (list on previous page): choose targeted MVI or single-nutrient supplementation, not 'men's all-purpose' marketingAbsolute 'insurance-style MVI for life': unnecessary + occasionally mildly harmful
Chapter 4
RCT graveyard
RCT graveyard
The vitamin trial graveyard is nutrition medicine's heaviest lesson:
**ATBC trial (1985-1993, Finland, *NEJM* 1994)**
N = 29,133 male smokersβ-carotene 20 mg/day + vitamin E 50 mg/day × 5-8 yearsExpectation: reduce lung cancer / cardiovascular eventsResult: β-carotene group lung cancer ↑18% + all-cause mortality ↑8%This was the first hard evidence that 'smokers taking β-carotene increases cancer', overturning the prevailing expectation
**CARET trial (1985-1996, USA, *NEJM* 1996)**
N = 18,314 smokers + asbestos workersβ-carotene 30 mg/day + retinyl palmitate 25,000 IU/dayStopped 21 months ahead of schedule after a mean 4.0 years of follow-up: lung cancer ↑28% (RR 1.28, 95% CI 1.04-1.57) + all-cause mortality ↑17% (RR 1.17, 1.03-1.33)Mechanism: high-dose β-carotene becomes pro-oxidant in the high-ROS environment of smoker lungsFDA changed label recommendations; the MVI industry removed β-carotene
**HOPE trial (*NEJM* 2000)**
N = 9,541 high CV-risk patientsVitamin E 400 IU/day × 4.5 yearsResult: no benefit on primary CV endpoint
**SELECT trial (*JAMA* 2009, updated *JAMA* 2011)**
N = 35,533 menSelenium 200 µg + vitamin E 400 IU, median 5.46 years of follow-up at the 2009 reportExpectation: prevent prostate cancer2009 result: vitamin E alone, prostate cancer HR 1.13 (99% CI 0.95-1.35), P = .06 — not significant; the trial did not prevent prostate cancer2011 result, after extended follow-up: vitamin E alone prostate cancer ↑17%, now significant (HR 1.17, 99% CI 1.004-1.36, P = .008) — this is where the widely-quoted 17% comes from (covered in vitamin D story)
**PHS-II trial (*JAMA* 2012) — the largest RCT of long-term MVI in healthy people**:
N = 14,641 male physicians (50+)Daily MVI × 11.2 years medianThe results were published as two papers, one per endpoint — they are easy to merge into one and get backwards**Cancer paper (Gaziano et al., *JAMA* 2012): the primary endpoint is total cancer, and it fell by a small but statistically significant 8% (HR 0.92, 95% CI 0.86-0.998, P = .04). Prostate cancer was a secondary** endpoint and did not move (HR 0.98, P = .76)**Cardiovascular paper (Sesso et al., *JAMA* 2012, same cohort): major CV events — no benefit** (HR 1.01)So: no signal on cardiovascular disease, and a barely-significant small reduction on the cancer primary endpoint. Calling total cancer a 'secondary endpoint', or calling it negative, both misstate the trialThe real limits: an 8% relative reduction is small in absolute terms, and male physicians limit generalizability to women / younger populations
**Bjelakovic *JAMA* 2013** — a two-page evidence synopsis of the same authors' Cochrane review (78 RCTs, N = 296,707):
High-dose β-carotene + vitamin A + vitamin E alone → total mortality ↑3-7%Vitamin C + selenium → no effect on mortalityNote the interventions analysed are these single antioxidants — multivitamins are not one of its analysis categories, so the MVI conclusion has to come from PHS-II / USPSTF instead
So 'antioxidant supplements prevent aging / cancer': most RCTs failed, some caused harm. The mechanism sounds plausible, but 'in-vitro antioxidant ≠ in-vivo benefit' — this is nutrition medicine's deepest lesson.
**ATBC trial (1985-1993, Finland, *NEJM* 1994)**
N = 29,133 male smokersβ-carotene 20 mg/day + vitamin E 50 mg/day × 5-8 yearsExpectation: reduce lung cancer / cardiovascular eventsResult: β-carotene group lung cancer ↑18% + all-cause mortality ↑8%This was the first hard evidence that 'smokers taking β-carotene increases cancer', overturning the prevailing expectation
**CARET trial (1985-1996, USA, *NEJM* 1996)**
N = 18,314 smokers + asbestos workersβ-carotene 30 mg/day + retinyl palmitate 25,000 IU/dayStopped 21 months ahead of schedule after a mean 4.0 years of follow-up: lung cancer ↑28% (RR 1.28, 95% CI 1.04-1.57) + all-cause mortality ↑17% (RR 1.17, 1.03-1.33)Mechanism: high-dose β-carotene becomes pro-oxidant in the high-ROS environment of smoker lungsFDA changed label recommendations; the MVI industry removed β-carotene
**HOPE trial (*NEJM* 2000)**
N = 9,541 high CV-risk patientsVitamin E 400 IU/day × 4.5 yearsResult: no benefit on primary CV endpoint
**SELECT trial (*JAMA* 2009, updated *JAMA* 2011)**
N = 35,533 menSelenium 200 µg + vitamin E 400 IU, median 5.46 years of follow-up at the 2009 reportExpectation: prevent prostate cancer2009 result: vitamin E alone, prostate cancer HR 1.13 (99% CI 0.95-1.35), P = .06 — not significant; the trial did not prevent prostate cancer2011 result, after extended follow-up: vitamin E alone prostate cancer ↑17%, now significant (HR 1.17, 99% CI 1.004-1.36, P = .008) — this is where the widely-quoted 17% comes from (covered in vitamin D story)
**PHS-II trial (*JAMA* 2012) — the largest RCT of long-term MVI in healthy people**:
N = 14,641 male physicians (50+)Daily MVI × 11.2 years medianThe results were published as two papers, one per endpoint — they are easy to merge into one and get backwards**Cancer paper (Gaziano et al., *JAMA* 2012): the primary endpoint is total cancer, and it fell by a small but statistically significant 8% (HR 0.92, 95% CI 0.86-0.998, P = .04). Prostate cancer was a secondary** endpoint and did not move (HR 0.98, P = .76)**Cardiovascular paper (Sesso et al., *JAMA* 2012, same cohort): major CV events — no benefit** (HR 1.01)So: no signal on cardiovascular disease, and a barely-significant small reduction on the cancer primary endpoint. Calling total cancer a 'secondary endpoint', or calling it negative, both misstate the trialThe real limits: an 8% relative reduction is small in absolute terms, and male physicians limit generalizability to women / younger populations
**Bjelakovic *JAMA* 2013** — a two-page evidence synopsis of the same authors' Cochrane review (78 RCTs, N = 296,707):
High-dose β-carotene + vitamin A + vitamin E alone → total mortality ↑3-7%Vitamin C + selenium → no effect on mortalityNote the interventions analysed are these single antioxidants — multivitamins are not one of its analysis categories, so the MVI conclusion has to come from PHS-II / USPSTF instead
So 'antioxidant supplements prevent aging / cancer': most RCTs failed, some caused harm. The mechanism sounds plausible, but 'in-vitro antioxidant ≠ in-vivo benefit' — this is nutrition medicine's deepest lesson.
五项试验逐个看数据
把过去几十年的大型维生素试验放在一起看, 结论出奇地一致: 对一般人群, 抗氧化补剂没拿出明确益处, 个别剂量还把风险推高了。逐个看数据。先说两项最沉重的, 都栽在吸烟者身上。芬兰的 ATBC 试验 (1985-1993, NEJM 1994) 给 29 133 名男性吸烟者每天补 β-胡萝卜素 20 mg 加维 E 50 mg, 连续 5-8 年, 本来指望降肺癌和心血管事件, 结果 β-胡萝卜素组肺癌反而升 18%、全因死亡升 8%。这是吸烟者补 β-胡萝卜素反而增加癌症的第一个硬证据, 把当时的主流期待整个掀翻。美国的 CARET 试验 (1985-1996, NEJM 1996) 在 18 314 名吸烟者和石棉接触者身上重复了同样的方向: β-胡萝卜素 30 mg 加 retinyl palmitate 25 000 IU, 平均补了 4.0 年之后, 因为肺癌升 28% (RR 1.28, 95% CI 1.04-1.57)、全因死亡升 17% (RR 1.17, 1.03-1.33), 比原计划提前 21 个月终止。这里的时间尺度值得停一秒: 伤害不是第二年就跳出来的, 而是平均四年的暴露才把这个差异攒够——剂量乘以时间, 正是这一幕真正想教的东西。机制层面的解释就是上一屏那条: 高剂量 β-胡萝卜素在吸烟肺这种活性氧 (ROS) 很高的环境里, 会从抗氧化剂翻转成促氧化剂。此后 FDA 改了标签建议, 整个复合维生素行业也把 β-胡萝卜素拿掉了。
维 E 单独上场也没好到哪去。HOPE 试验 (NEJM 2000) 给 9 541 名高心血管风险患者每天 400 IU、连续 4.5 年, 主要心血管终点没有任何获益。SELECT 试验给 35 533 名男性补硒 200 µg 加维 E 400 IU, 本想防前列腺癌。它值得分两次读: 2009 年首次报告时 (JAMA), 中位随访 5.46 年, 维 E 组前列腺癌高出 13%, 但没到显著 (HR 1.13, 99% CI 0.95-1.35, P=0.06), 当时的结论只是没防住。坏消息在延长随访之后才落地: 2011 年的更新 (JAMA) 报的是维 E 组前列腺癌升 17% 且已经显著 (HR 1.17, 99% CI 1.004-1.36, P=0.008)。同一批人、同一个干预, 多看了几年, 一个不显著的苗头长成了确定的伤害——你在别处看到的那个 17%, 出处是 2011 年这一版, 不是 2009 年那一篇 (维 D 那一篇也讲过这件事)。
最贴近普通人问题的是 PHS-II 试验——它是健康人长期吃复合维生素的最大型 RCT, 让 14 641 名 50 岁以上男医生每天吃一颗、中位随访 11.2 年。它的结果分两篇论文发表, 各管一个终点, 很容易被混成一篇:
癌症那一篇 (JAMA 2012, Gaziano 等): 主要终点就是总癌症, 结果是显著但很小的下降——降 8% (HR 0.92, 95% CI 0.86-0.998, P=0.04)。前列腺癌是次要终点, 没有变化 (HR 0.98, P=0.76)。心血管那一篇 (同年 JAMA, Sesso 等, 同一队列): 主要心血管事件没有获益 (HR 1.01)。
所以准确的说法是: 复合维生素在这批人身上没碰到心血管, 在癌症这个主要终点上拿到了一个刚够到显著线的小幅下降。把总癌症说成次要终点、或者说成阴性, 两种说法都在替读者做结论。真正的限制在别处——8% 的相对下降在绝对量上很小, 而且全是男医生, 往女性和年轻人外推的能力很有限。
把这些汇总到一起的是 Bjelakovic 2013 (JAMA 上的一篇两页证据摘要, 概括的是同一批作者的 Cochrane 综述, 78 个 RCT、共 296 707 人): 单独高剂量的 β-胡萝卜素、维 A、维 E 让总死亡升了 3-7%; 维 C 和硒对死亡率没有影响。注意它分析的是这几种单一抗氧化剂, 复合维生素不是它的分析类别——复合维生素在健康人身上的结论要看 PHS-II 和 USPSTF 那一侧, 不能从这条推过来。
读完这一串数字, 有一个方向值得留意: 出事的全是单独一种、剂量远超食物能达到的水平的那几项; 而剂量贴着推荐量的复合维生素本身, 结论是没有明显坏处, 好处也小到要靠上万人随访十年才勉强量得出来。这两句话并不矛盾——它们描述的是同一条剂量曲线上的两段: 一段平, 一段已经拐下去了。
Why RCTs failed
Why did supplements fail RCTs against food? Mechanistic explanations:① The 'antioxidant paradox'
In vitro / test tube: vitamin E / carotene neutralize free radicals → 'should' prevent aging / cancerIn vivo: free radicals aren't 'enemies', they're cellular signaling molecules —Pro-proliferative signaling requires ROS suppression (e.g., exercise-induced hormesis)Bacterial killing requires ROS burstsMegadose supplementation → blunts these signalsThis is why the NRF2 pathway is suppressed + adaptation / immunity drops
② 'Food matrix effect' can't be replicated by a pill
A single carrot contains β-carotene + α-carotene + lycopene + lutein + vitamin C + fiber + polyphenols + potassium + hundreds of bioactivesThey work synergistically, not β-carotene acting aloneWhen you supplement only a single component, you lose the ensemble
③ High dose ≠ better
The 'U-curve' applies to almost all micronutrients'Supplemented to sufficient' ≠ 'supplemented to excess'RCTs often use pharmacological doses, far above achievable food levels → entering toxicology zone
④ Healthy people are hard to make 'healthier'
Almost all large MVI RCTs enrolled subjects whose diet is already near-sufficientFor deficient people, 'supplementing to sufficient' yields obvious benefit; for sufficient people, 'supplementing to excess' yields zero or negative marginal returns
⑤ 'Healthy-user paradox'
People who take supplements already have better health habits — non-supplementers may exercise more, eat more vegetablesIn observational studies, 'supplement users live longer' reflects habits, not supplementsRCT design breaks this confounding — and the result is 'oh, it turns out supplements aren't the cause'
These mechanisms explain: using 'drug' logic to study 'food edge cases' was flawed from the start — a deep methodological challenge in nutrition medicine.
Chapter 5
Better than MVI
Better than MVI
If not MVI, what to use instead? — evidence-based priority list:
① Improve dietary pattern (strongest evidence-backed 'health insurance'):
Mediterranean diet: multiple RCTs (PREDIMED etc.) show reduced CV / diabetes / dementia / depressionDASH diet: ironclad RCT evidence for lowering blood pressure30+ plants per week (American Gut Project): strongest single predictor of microbiome diversityLimit ultra-processed food: NOVA category 4 → multiple cohorts linked to all-cause mortality + multimorbidityAdequate protein 1.0-1.6 g/kg: maintain muscle + satiety2-3× weekly fatty fish: evidence stronger than fish oil
② Targeted single-nutrient supplementation (when there's a specific indication):
Vitamin D3 1000-2000 IU/day: northern regions / indoor workers / elderly / dark skin / obesity; test 25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. first to confirm real deficiency (<50 nmol/L = deficient)Vitamin B12 (vegans + atrophic gastritis + long-term PPI / metformin): 1000 µg sublingualIron (women / heavy menses / vegans): test ferritin (<30 µg/L = deficient)Calcium (postmenopausal women + non-dairy): food first, supplement 500-600 mg + vitamin DIodine (pregnancy / lactation + non-seafood eaters): 150-220 µgOmega-3 EPA + DHA (non-fish eaters / pregnancy): fish oil or algal oil 250-500 mgFolic acid (preconception / early pregnancy): 400 µg folic acidTrauma recovery / severe illness: protein + vitamin C + zinc short-term combo
③ Unnecessary 'trending supplements':
'Hair-nail-skin' MVI — mostly high-dose biotin (false-positive lab interference + no benefit)'Anti-aging NMN / NR' — animal evidence strong, human evidence weak + expensive'Whole-food multivitamins' — marketing, 5-10× more expensive than synthetic MVI with similar or even lower actual vitamin content'Capillary / detox / immune-boost' multi-component supplements — sound good, no RCTs
Key cognitive shift:
> The 'insurance' MVI mindset reduces health management to a single low-cost daily act, substituting one pill for the things that actually take time and habit (cooking, walking outside, sleeping well).
>
> 'Targeted + food-first' is the direction that actually moves blood markers and long-term risk — it asks not for more pills, but for connecting dietary pattern, lifestyle, and *specific evidence of deficiency* into one decision.
MVI isn't 'evil' — it suits a small subset of people (see the prior scene's 'real beneficiaries' list); most healthy adults get more out of redirecting MVI spend to an extra vegetable serving or fish portion — higher ROI, more directly.
① Improve dietary pattern (strongest evidence-backed 'health insurance'):
Mediterranean diet: multiple RCTs (PREDIMED etc.) show reduced CV / diabetes / dementia / depressionDASH diet: ironclad RCT evidence for lowering blood pressure30+ plants per week (American Gut Project): strongest single predictor of microbiome diversityLimit ultra-processed food: NOVA category 4 → multiple cohorts linked to all-cause mortality + multimorbidityAdequate protein 1.0-1.6 g/kg: maintain muscle + satiety2-3× weekly fatty fish: evidence stronger than fish oil
② Targeted single-nutrient supplementation (when there's a specific indication):
Vitamin D3 1000-2000 IU/day: northern regions / indoor workers / elderly / dark skin / obesity; test 25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. first to confirm real deficiency (<50 nmol/L = deficient)Vitamin B12 (vegans + atrophic gastritis + long-term PPI / metformin): 1000 µg sublingualIron (women / heavy menses / vegans): test ferritin (<30 µg/L = deficient)Calcium (postmenopausal women + non-dairy): food first, supplement 500-600 mg + vitamin DIodine (pregnancy / lactation + non-seafood eaters): 150-220 µgOmega-3 EPA + DHA (non-fish eaters / pregnancy): fish oil or algal oil 250-500 mgFolic acid (preconception / early pregnancy): 400 µg folic acidTrauma recovery / severe illness: protein + vitamin C + zinc short-term combo
③ Unnecessary 'trending supplements':
'Hair-nail-skin' MVI — mostly high-dose biotin (false-positive lab interference + no benefit)'Anti-aging NMN / NR' — animal evidence strong, human evidence weak + expensive'Whole-food multivitamins' — marketing, 5-10× more expensive than synthetic MVI with similar or even lower actual vitamin content'Capillary / detox / immune-boost' multi-component supplements — sound good, no RCTs
Key cognitive shift:
> The 'insurance' MVI mindset reduces health management to a single low-cost daily act, substituting one pill for the things that actually take time and habit (cooking, walking outside, sleeping well).
>
> 'Targeted + food-first' is the direction that actually moves blood markers and long-term risk — it asks not for more pills, but for connecting dietary pattern, lifestyle, and *specific evidence of deficiency* into one decision.
MVI isn't 'evil' — it suits a small subset of people (see the prior scene's 'real beneficiaries' list); most healthy adults get more out of redirecting MVI spend to an extra vegetable serving or fish portion — higher ROI, more directly.
有指征时才补的清单
只有当你能说出我缺的是哪一样、凭什么这么说时, 补剂才有事可做。常见的几个指征和对应做法:维 D3 1000-2000 IU/天: 北方地区、室内工作、老人、深肤色、肥胖人群; 抽血 25-hydroxyvitamin D: The storage form of vitamin D in blood — the number measured to check D status. 看是否真低 (< 50 nmol/L = 缺乏)维 B12 (素食者、萎缩性胃炎、长期 PPI 或二甲双胍): 1000 µg 舌下含片铁 (女性、经期量大、素食者): 抽血看 ferritin (< 30 µg/L 是缺铁)钙 (绝经后女性、不吃奶制品): 食物优先, 补片 500-600 mg 配维 D碘 (孕期、哺乳期、不吃海产): 150-220 µgomega-3 EPA + DHA (不吃鱼、孕期): 鱼油或藻油 250-500 mg叶酸 (备孕、早孕): 400 µg folic acid创伤恢复、重病: 蛋白 + 维 C + 锌 的短期组合
这张表和一颗什么都有的药片, 差别不在种类多少。 差别在两处: 每一行都是先有一个指征、才有一个剂量; 而且几乎每一行的剂量, 都比复合维生素里那一格高出一大截——因为复合维生素要把几十样塞进同一片, 每样只能给一点点 (第二幕讲过, 挤在同一片里的矿物质还会互相压)。所以想用它去补一个真实存在的缺口, 通常是补不动的。
还有一条顺序常被忽略: 先测再补, 补完再测。抽血这个动作把我大概缺变成我确实缺多少, 也让你知道什么时候可以停——而补到够就停这件事, 是复合维生素这种一年吃到头的产品结构上做不到的。
证据比补剂更硬的事
下面这些做法, 在 RCT 里拿到过复合维生素从来没拿到过的结果:地中海饮食: 多项 RCT (PREDIMED 等) 显示降低心血管事件、糖尿病、痴呆和抑郁的风险。DASH 饮食: 降血压有 RCT 级别的铁证。每周吃到 30+ 种植物 (American Gut Project): 目前对肠道菌群多样性预测力最强的单一指标。少吃超加工食品 (NOVA 4 类): 多个队列研究里与全因死亡和多病共存相关。蛋白吃够 1.0-1.6 g/kg: 维持肌肉, 顺带更容易饱。每周 2-3 次脂肪鱼: 证据比吃鱼油强。
为什么整套饮食能赢过单一成分? 一半的答案是上一幕的食物矩阵: 成分是绑在一起被吸收的, 拆开就不是原来那件事。另一半答案几乎从不被提起, 叫替换效应——你多吃一份蔬菜, 通常就少吃了一份别的东西。药片只会做加法, 而饮食同时做加法和减法; 在慢性病上, 减掉的那一项往往比加上的那一项更值钱。
这也解释了一个常被误读的现象: 为什么多吃蔬菜在队列研究里效果那么好, 而补蔬菜里的维生素在 RCT 里效果是零。两者的差别不只是矩阵, 还有那份蔬菜挤掉的东西——而药片不挤掉任何东西, 它只是又加了一样。
这些不用买
几类反复出现在广告里、但没有对应证据的产品:头发指甲皮肤配方: 主要卖点是大剂量生物素 (biotin)。它最确定的作用其实是干扰一批常用化验项目 (包括甲状腺功能和心肌损伤标志物), 让结果假高或假低; 对头发指甲本身, 在不缺生物素的人身上看不到获益。抗衰老 NMN / NR: 动物数据漂亮, 人的数据很薄, 价格很贵。全食物多维生素: 营销词。比合成配方贵 5-10 倍, 实测维生素含量差不多甚至更低。护毛细血管 / 排毒 / 增强免疫 一类的复方: 名字都很动人, 拿不出一项 RCT。
认出它们有一个通用办法: 看它承诺的是一个可以测量的缺口被补上, 还是一种说不清怎么验证的更好状态。前者你可以抽血证实或证伪, 后者永远不会有让你失望的那一天——因为它从一开始就没给出可以失败的条件。一个连没效长什么样都说不出来的承诺, 也就没法说有效。
The 'insurance supplement' mindset cost
Why do people keep taking MVI despite RCT failures?Psychological explanations:
① 'Insurance-style thinking'
'No harm anyway, just in case it helps'Overweighting low-probability events (CV events / cancer), granting disproportionate trust to the 'protection' supplements offerBehavioral economics calls this 'ambiguity aversion'
② 'I did something' psychology
Taking MVI provides cognitive comfort of 'I'm actively managing my health'The cost is extremely low compared to actually changing diet + exerciseBut the side effect of 'doing something' is 'I don't have to do the harder things'Multiple studies (Chiou 2011 etc.): supplement users are more likely to choose unhealthy behaviors — 'I already took my vitamins'
③ Marketing mechanisms:
60+ years of ads: 'full of vitality / energetic / hair shines / boosts immunity'None of these claims have been rigorously FDA-scrutinized (the DSHEA 1994 act made US supplement regulation very lax)'Structure/function claim' loophole — 'supports bone health' ≠ 'prevents osteoporosis', but consumers hear them as the same
④ 'Sunk cost'
'I've been taking it for 5 years, what if my body breaks if I stop'But the body won't suddenly break if you stop MVI — 'protection' was never proven to exist
Real costs:
Money: US households spend ~$370 per person per year on supplements on averageBehavioral substitution cost: lets people be satisfied with 'doing something' instead of making real changeRare but real harms: β-carotene / vitamin E / high iron / high calciumCognitive confusion: nutrition is reduced to 'taking pills', not 'eating food + lifestyle'
So 'saving the MVI money to buy vegetables at the market' isn't being stingy — it's a higher-ROI health investment.
References · 8
- US Preventive Services Task Force. (2022). Vitamin, mineral, and multivitamin supplementation to prevent cardiovascular disease and cancer: recommendation statement. JAMA, 327(23), 2326–2333. 10.1001/jama.2022.8970
- The ATBC Cancer Prevention Study Group. (1994). The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. NEJM, 330(15), 1029–1035. 10.1056/NEJM199404143301501
- Omenn, G. S., et al. (1996). Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease (CARET). NEJM, 334(18), 1150–1155. 10.1056/NEJM199605023341802
- Gaziano, J. M., et al. (2012). Multivitamins in the prevention of cancer in men: the Physicians' Health Study II randomized controlled trial. JAMA, 308(18), 1871–1880. 10.1001/jama.2012.14641
- Sesso, H. D., Christen, W. G., Bubes, V., Smith, J. P., MacFadyen, J., Schvartz, M., Manson, J. E., Glynn, R. J., Buring, J. E., & Gaziano, J. M. (2012). Multivitamins in the prevention of cardiovascular disease in men: the Physicians' Health Study II randomized controlled trial. JAMA, 308(17), 1751-1760. 14,641 male US physicians aged 50+, median 11.2 years. A daily multivitamin had no effect on major cardiovascular events (hazard ratio 1.01, 95% CI 0.91-1.10), nor on myocardial infarction, stroke or cardiovascular mortality. 10.1001/jama.2012.14805
- Bjelakovic, G., Nikolova, D., & Gluud, C. (2013). Antioxidant supplements to prevent mortality. JAMA, 310(11), 1178–1179. 10.1001/jama.2013.277028
- Lippman, S. M., et al. (2009). Effect of selenium and vitamin E on risk of prostate cancer (SELECT): a randomized controlled trial. JAMA, 301(1), 39–51. 10.1001/jama.2008.864
- Klein, E. A., Thompson, I. M., Tangen, C. M., Crowley, J. J., Lucia, M. S., Goodman, P. J., et al. (2011). Vitamin E and the risk of prostate cancer: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA, 306(14), 1549-1556. Among 35,533 men, vitamin E 400 IU/day significantly increased prostate cancer risk (about 17% relative increase; 76 vs 65 cases per 1,000). 10.1001/jama.2011.1437