高性价比人生指南
record

核实记录

第 34 节第 14 条：肾功能不好的人，止痛药和杨桃

2026-11-04。起因是两个读者 issue，都没有给来源。

  • issue #146：肝或肾有问题的人，要分清常用药物的主要代谢途径，止痛药有的走肾、有的走肝，选在健康器官里代谢的药能保护受损的器官。
  • issue #153：列了一串会影响 CYP3A4 的植物：石榴、欧芹、贯叶连翘、杨桃、苦橙、黑胡椒。

落点怎么定的

issue #146 的原意是让读者自己按代谢途径挑药，本书不这么写，因为那就成了自己换药、自己定剂量。能成立、又是「一个动作避开重后果」的，是「肾不好的人别自己买布洛芬这类药长期吃，先问医生」。按说明书剂量、按肾功能调药是医生的事，KDIGO 2024 实践要点 4.2.1 也是写给开药的人的（Consider GFR when dosing medications cleared by the kidneys）。这句话放进了备注：「别自己按药在肝里还是在肾里代谢去挑」。

肝那一侧：对乙酰氨基酚伤肝、一日最多 2 克、别喝酒，本节第 1 条已有，不重复，只在备注里指路。美国 NSAID 非处方药标签那句「先问医生」本来就同时点了肝硬化，原样写进了收益栏。

issue #153 只评估了杨桃。杨桃对肾病病人的风险和药物相互作用无关，是杨桃本身的毒素和草酸盐，所以和「肾不好的人别自己吃止痛药」并成一条「肾功能不好的人」条目，追加在节末。

四问

  1. 普通读者真会碰上吗？中国成年人慢性肾病约 8.2%（第 16 节第 13 条，Wang 2023），高血压病人常同时吃利尿药和普利、沙坦类药。杨桃在南方常见。会碰上。
  2. 读者是不是普遍不知道？慢性肾病的知晓率只有 10.0%（同上）。KDIGO 写明 CKD G3–G5 的人每年有 18%–20% 至少用过一种可能不合适的伤肾药，主要是 NSAIDs。杨桃伤肾在国内知道的人不多。
  3. 动作是新的吗？第 3 条讲的是 NSAID 伤胃的六种高危人群，动作也是「先问医生」，但人群和后果都不同（胃出血对肾损伤），而且本条另有「三样一起吃」和杨桃两个动作。两条互相指路，没有并：第 3 条标题只许加字，把肾病塞进去会让标题过长，杨桃也没地方放。
  4. 有原始来源吗？有：系统综述两篇（Zhang 2017、Nderitu 2013）、大型巢式病例对照（Lapi 2013）、KDIGO 2024 指南原文、美国联邦法规原文、杨桃病例汇总三篇和病例系列两篇。

关键原文

NSAIDs 与肾

  • Zhang 2017, BMC Nephrology 18:256, https://doi.org/10.1186/s12882-017-0673-8（Europe PMC 摘要）：「In the general population, the pooled odds ratio (OR) of AKI for current NSAID exposure was 1.73 (95%CI 1.44 to 2.07), with somewhat higher risk observed in older people (OR 2.51, 95%CI 1.52 to 2.68). In people with CKD, individual study OR of AKI due to current NSAID exposure ranged from 1.12 to 5.25, with pooled estimate OR 1.63 (95% CI 1.22 to 2.19).」「No study reported baseline risk of AKI in different populations meaning absolute risks could not be estimated」。老年人那个可信区间（1.52 到 2.68，点估计 2.51 靠近上限）摘要原文就是这样，照抄。因为没有基础发生率，没按「结局少见」把比值比读成风险倍数，说人话只写方向。慢性肾病那个合并值是 5 项研究：「of which five additionally reported NSAID risk in people with CKD」。
  • Lapi 2013, BMJ 346:e8525, https://doi.org/10.1136/bmj.e8525：「A cohort of 487,372 users of antihypertensive drugs.」「During a mean follow-up of 5.9 (SD 3.4) years」「current use of a double therapy combination ... was not associated with an increased rate of acute kidney injury. In contrast, current use of a triple therapy combination was associated with an increased rate of acute kidney injury (rate ratio 1.31, 95% confidence interval 1.12 to 1.53). In secondary analyses, the highest risk was observed in the first 30 days of use (rate ratio 1.82, 1.35 to 2.46).」三样是「diuretics with angiotensin converting enzyme inhibitors or angiotensin receptor blockers and NSAIDs」。rate ratio 按 CLAUDE.md 的 IRR 译法写「发病率之比」，读法同 RR，说人话写「多约三成」。
  • Nderitu 2013, Family Practice 30(3):247-255, https://doi.org/10.1093/fampra/cms086：45 岁以上；「Regular-dose NSAID use did not significantly affect the risk of accelerated CKD progression; pooled odds ratio (OR) = 0.96 (95%CI: 0.86-1.07), but high-dose NSAID use significantly increased the risk ... pooled OR = 1.26 (95%CI: 1.06-1.50).」结论「The avoidance of NSAIDs in the medium term is unnecessary in patients with moderate to severe CKD, if not otherwise contraindicated.」这是反方，进了备注的「争议」。
  • KDIGO 2024（Kidney Int 105(4S):S117-S314，https://doi.org/10.1016/j.kint.2023.10.018；全文 PDF 取自 kdigo.org/wp-content/uploads/2024/03/KDIGO-2024-CKD-Guideline.pdf，pdftotext 逐字核对），第 4 章：
    • Practice Point 4.1.1：「People with CKD may be more susceptible to the nephrotoxic effects of medications.」正文写成「可能更容易被药物伤肾」，保留了 may。
    • Practice Point 4.1.3：「Review and limit the use of over-the-counter medicines and dietary or herbal remedies that may be harmful for people with CKD.」
    • 4.1.3 下的说明：「Indiscriminate chronic OTC NSAID use has been associated with a higher risks of kidney failure compared with nonuse and should be discouraged. However, judicious NSAID use, under careful supervision of a nephrologist, may be preferred to other pain medications such as opioids that have stronger associations with adverse events.」后半句进了备注的「争议」。
    • 4.1.1 下的说明：「Between 18%–20% of people with CKD G3–G5 receive at least one potentially inappropriate nephrotoxic medication annually, primarily NSAIDs」。只用在本记录的四问里，没进正文。
    • Figure 45（Selected herbal remedies and dietary supplements with evidence of potential nephrotoxicity）列了「Star fruit (Averrhoa carambola)」。
  • 21 CFR 201.326(a)(2)(iii)(B)，eCFR 现行版（versioner API 2026-10-01 全文，要带 --compressed）：「Ask a doctor before use if [bullet] stomach bleeding warning applies to you [bullet] you have a history of stomach problems, such as heartburn [bullet] you have high blood pressure, heart disease, liver cirrhosis, or kidney disease [bullet] you are taking a diuretic」。

杨桃

  • Chua 2017, Clinical Toxicology 55(7):624-628, https://doi.org/10.1080/15563650.2017.1314490：「We reviewed totally 126 patients ... from 33 articles」，检索 1993 到 2016 年；「The most common symptom was hiccups (65%), whereas confusion and seizure were the most common symptoms associated with mortality」；「there was no mortality in patients with normal renal function (NRF), the mortality of patients among reported cases with chronic renal insufficiency and end-stage renal disease undergoing dialysis were 36% and 27%, respectively」；「the number of SF consumed was substantially higher in the patients with NRF」。
  • Yasawardene 2021, Trans R Soc Trop Med Hyg 115(9):947-955, https://doi.org/10.1093/trstmh/trab026：「total number of individuals=136」「Ninety-four (69.1%) patients had prior renal impairment. Renal histology showed acute oxalate nephropathy」「Oxalate and caramboxin are considered the main substances causing nephrotoxicity and neurotoxicity.」caramboxin 正文译作「杨桃毒素」。
  • Chang 2000, Am J Kidney Dis 35(2):189-193, https://doi.org/10.1016/s0272-6386(00)70325-8：「In the past 10 years, 20 such patients were treated in our hospital when they developed clinical symptoms after eating the fruit or drinking star fruit juice.」「Eight patients died, including one patient with a serum creatinine level of 6.4 mg/dL who had not yet begun dialysis.」「Death occurred within 5 days」。摘要没写医院所在地，正文只写「一家医院」。
  • Neto 2003, Nephrol Dial Transplant 18(1):120-125, https://doi.org/10.1093/ndt/18.1.120：32 名尿毒症病人；「Patients who were promptly treated with haemodialysis, including those with severe intoxication, recovered without sequelae. Patients with severe intoxication who were not treated or treated with peritoneal dialysis did not survive.」作者单位在巴西（Ribeirão Preto），正文写「巴西一组」。
  • Aranguren 2017 系统综述（Saudi J Kidney Dis Transpl 28(4):709-715，PMID 28748871，Crossref 查不到 DOI）：123 人、8 个国家、28 人（22%）死亡。和 Chua 2017 的病例大量重叠，数字取 Chua 那篇（按肾功能分了死亡率），这篇没进来源栏。

没用的来源

  • 上海市卫健委网站 2021-09-02「颜值高、口感好，但这个水果却可能变身『肾脏杀手』」（wsjkw.sh.gov.cn/spwsyyy/20210902/c531581c5c1141cc99b7da451e17cc12.html）：页尾写明「摘自上海交通大学医学院附属瑞金医院微信公众号」，按引用规则不收。文中把杨桃伤肾的机制写成「肾小球毛细血管基底膜出现损伤……变态反应」，和上面综述的「草酸盐肾病、杨桃毒素」不是一回事，也是不收的理由之一。它建议高血压、糖尿病病人也尽量别吃、婴幼儿不适合吃，这两点没找到原始文献，没写。
  • 美国肾脏基金会（NKF）患者页「Always AVOID star fruit」：行业协会的患者教育材料，有 KDIGO 指南就够了，没引。
  • KDIGO 的 sick day rules（生病脱水时暂停 NSAIDs 等药）：指南自己写明「there is a paucity of evidence」，还提到病人搞错时可能有害，没写。

issue #153 其余植物不收的理由

  • 苦橙：第 13 条（西柚）已写酸橙（「常用来做橙子酱」）同样有呋喃香豆素作用。
  • 贯叶连翘：第 16 节第 4 条（别为了试偏方、保健品停掉正规治疗）的收益栏和备注已写它会减弱抗排异药、抗凝药等（NCCIH）。
  • 石榴：两项健康志愿者试验都没看出对 CYP3A 探针药咪达唑仑有影响。Farkas 2007（J Clin Pharmacol, doi:10.1177/0091270006298359）：「Pomegranate juice does not impair clearance of oral or intravenous midazolam」，同一试验里西柚汁有影响；Misaka 2011（Clin Ther, doi:10.1016/j.clinthera.2011.02.012）连喝 2 周，AUC 几何均数比 1.02（0.95–1.09）。体外有抑制、人体没看到。
  • 黑胡椒：能找到的是胡椒碱提纯物的小试验，比如 Bedada 2017 每天 20 毫克胡椒碱连吃 10 天，卡马西平 AUC 升 47.9%（12 人）。那是提纯物剂量，不是吃饭撒的胡椒，没找到日常食用量有临床意义的证据。
  • 欧芹：没找到人体药代或临床病例证据。

引用数

本节新增 4 处条目引用：第 14 条备注里的第 16 节第 13 条、本节第 1 条、本节第 3 条，以及第 3 条备注新加的本节第 14 条。用 HEAD 版本的本节在 scratchpad 副本里跑 check-refs --check 是 1844 处，改后 1848 处，涨 4，没有被当成法条吞掉的。

统计口径

本条成本标签：钱=0 时间=少 毅力=些 收益=中 口径=死亡率，不加人群标签（慢病没有对应值）。收益定「中」凭判断：急性肾损伤的数字是比值比和发病率之比，没有死亡降幅；杨桃的死亡比例来自病例报告，有发表偏倚，算不出照做能少几成。证据定 B，标「争议」（常规剂量是否伤肾，Nderitu 2013 与 KDIGO 后半句）。

搜索全书