基本正确有充分科学证据支持证据确定性: 中初步 C-ODD-05

每一大类降糖药（二甲双胍、磺脲类、噻唑烷二酮类、DPP-4抑制剂、SGLT2抑制剂、GLP-1受体激动剂）在经同行评审的文献中都有严重危害的记录信号。

安全提示: 请不要因为这一说法而停用医生开具的糖尿病、心脏或肾脏药物；对心力衰竭、肾脏病或心血管疾病患者停用SGLT2抑制剂或GLP-1受体激动剂，会使他们失去一种已被证实的保护作用。如有顾虑，请与主治医生讨论。

简而言之: 作为一份清单，该说法基本正确：每一大类糖尿病药物都至少有一个有记录的严重危害信号，其中几个已在随机对照试验中得到证实。但它是片面的，因为有些信号并未得到证实，而同样的证据表明，SGLT2抑制剂和GLP-1受体激动剂能减少死亡。

结论: 所列出的每一个危害信号都确实存在，但这份清单是片面的，遗漏了已被证实的获益。

原始表述
科学评估
科学证据详解
通俗解释
与该说法相关的研究
地区背景
在我的国家适用什么？
参考文献

原始表述

كل فئة من فئات أدوية السكري ـ من الميتفورمين والسلفونيل يوريا والثيازوليدينديونات إلى مثبطات DPP-4 و SGLT-2 وناهضات GLP-1 ـ تحمل توقيعاً موثَّقاً من الأذى الجسيم في الأدبيات الطبية الرصينة

按来源为原话. 译文由本项目提供；以阿拉伯语原文为准。

译文

每一大类降糖药（二甲双胍、磺脲类、噻唑烷二酮类、DPP-4抑制剂、SGLT2抑制剂、GLP-1受体激动剂）在经同行评审的文献中都有严重危害的记录信号。

出处

来源状态: 档案网站编辑
归属: Archive editors (not Elawady)
出处: https://eltayebat.com/rule/oral-diabetes-drugs/
归属的不确定性: Editorial narrative and study selection written by the archive operators.

所声称的健康影响: Heart failure, MI, amputation, Fournier gangrene, bladder cancer, B12 deficiency, NAION
涉及人群: People with type 2 diabetes
Elawady 设定的地域限制: none found

科学评估

基本正确 – 核心内容正确，但有局限或细微不准确之处。

证据确定性: 中 – 进一步研究仍可能影响该评估。

1. 说法是什么？

All drug classes have documented serious harms.

2. 哪些内容可证实是正确的？

Rosiglitazone MI OR 1.43 (A-ODD-01) and heart failure HR 2.10 (A-ODD-02, I-ODD-13); pioglitazone bladder cancer HR 1.63 (A-ODD-03); metformin B12 deficiency (A-ODD-04, A-ODD-05, A-ODD-06); sulfonylureas vs metformin aHR 1.21 (A-ODD-07); saxagliptin heart-failure hospitalisation HR 1.27 (A-ODD-08); canagliflozin amputation HR 1.97 (A-ODD-09) and fractures in CANVAS (A-ODD-13); SGLT2 DKA (A-ODD-10, A-ODD-11), confirmed in the randomised CREDENCE trial (adjudicated DKA 11 vs 1; HR 10.80, 1.39-83.65; I-ODD-05), and Fournier reports (A-ODD-12); GLP-1 pancreatitis association (A-ODD-14), GI adverse events (A-ODD-15), NAION association (A-ODD-16), and in LEADER more acute gallstone disease (3.1% vs 1.9%), acute cholecystitis (0.8% vs 0.4%) and GI-related discontinuation (9.5% vs 7.3%) with liraglutide (I-ODD-06). All numbers were checked against the abstracts or full texts and are reported accurately.

3. 哪些内容是错误或夸大的？

The same CANVAS paper reports a 14% reduction of major cardiovascular events and possible kidney benefit (A-ODD-09); the archive mentions only the amputation finding. In CREDENCE canagliflozin reduced kidney failure and showed 'no significant differences in rates of amputation or fracture' (I-ODD-05). RECORD found rosiglitazone non-inferior for overall cardiovascular outcomes, with MI data 'inconclusive' (I-ODD-13). An RCT meta-analysis found no significant Fournier excess (I-ODD-12). Several of these drugs reduce death: empagliflozin -32% all-cause death (I-ODD-02), liraglutide HR 0.85 (I-ODD-06), dapagliflozin HR 0.69 in CKD (I-ODD-03), metformin -36% (I-ODD-01). Saxagliptin's primary ischaemic outcome was neutral (HR 1.00) (A-ODD-08). Listing harms without benefits gives a misleading net picture.

4. 哪些内容仍不明确？

Some signals (NAION, pioglitazone bladder cancer, sulfonylurea cardiovascular risk) remain debated.

5. 哪些研究支持该说法？

A-ODD-01 to A-ODD-16; I-ODD-05 (DKA) and I-ODD-06 (gallstone disease, cholecystitis) for the harm signals.

6. 哪些研究与该说法相矛盾？

I-ODD-01 to I-ODD-04, I-ODD-07, I-ODD-12, I-ODD-13 (benefits or non-confirmation); I-ODD-05 and I-ODD-06 also show net benefit (kidney failure, death).

7. 对德国和欧洲而言情况如何？

European (EASD) and German (DDG) guidance follow the same evidence base as the ADA/EASD consensus: individualised HbA1c targets, insulin as mandatory therapy in type 1 diabetes, and SGLT2 inhibitors/GLP-1 receptor agonists for people with cardiovascular or kidney disease. No country-specific difference changes the assessment.

8. 关于埃及的已知情况是什么？

Egypt has one of the highest diabetes prevalences in the region; the same drug classes and insulins are used. Access barriers and cost may make advice to stop medication attractive, which increases the practical risk of this claim. No Egypt-specific evidence changes the assessment.

9. 合理的实际结论是什么？

每种药物都有副作用，该档案网站的清单就每项研究而言在事实上是正确的。医学上的问题是针对具体某个人的获益与危害之间的权衡，而对于其中几类药物，在特定人群中这种权衡明显是有利的。

评估详情

证据确定性的依据: Each listed signal exists in the literature and is correctly quoted (verified for each study). However the strength differs greatly: some are confirmed in RCTs (rosiglitazone heart failure, metformin B12, canagliflozin amputation in CANVAS), some were not confirmed (canagliflozin amputation in CREDENCE, Fournier gangrene in RCT meta-analysis, rosiglitazone MI in RECORD), and the selection omits proven benefits in the same trials.
来源质量: Archive editorial with correctly cited, selectively chosen studies.
地区可推广性: Multinational data.
说法的地域范围: 普遍性（生物学/医学）
最近审核: 2026年10月26日初步

科学证据详解

有充分科学证据支持 – 该说法与一致的高质量证据相符。

研究问题: Does each major class of glucose-lowering drugs have a documented signal of serious harm in peer-reviewed literature, and how do these signals compare with documented benefits?

研究问题详解 (PICO)

人群

Adults with type 2 diabetes

干预或暴露

Metformin, sulfonylureas, thiazolidinediones, DPP-4 inhibitors, SGLT2 inhibitors, GLP-1 RAs

对照

Placebo/standard care

地域范围

universal biology

结局指标

Serious adverse events by class
All-cause mortality
Heart failure, kidney outcomes

研究结果按作者报告呈现。我们的评估见下文。

哪些证据支持该说法？

816 trials: what each diabetes drug class does and does not do

Shi Q et al.. BMJ 2023 · PMID 37024129 · DOI 10.1136/bmj-2022-074068 Meta分析已核查全文新增

816-trial NMA confirms class-specific harms.

实际结果
SGLT2i DKA OR 2.07, amputation OR 1.27; sulfonylurea severe hypoglycaemia OR 5.22; TZD heart-failure admission OR 1.54; GLP-1 RA severe GI events (low certainty).
方法学质量
high
相关性
direct
局限
Low certainty for older drugs.

原始文献
Rosiglitazone and heart failure (RECORD)

Komajda M et al.. Eur Heart J 2010 · PMID 20118174 · DOI 10.1093/eurheartj/ehp604 随机对照试验全文，含结构化偏倚风险评估

RECORD: rosiglitazone doubled heart failure.

实际结果
HR 2.10 (1.35-3.27).
方法学质量
high (RCT)
相关性
direct
局限
Open-label.

原始文献

哪些证据不支持该说法？

816 trials: what each diabetes drug class does and does not do

Shi Q et al.. BMJ 2023 · PMID 37024129 · DOI 10.1136/bmj-2022-074068 Meta分析已核查全文新增

The same NMA shows SGLT2i and GLP-1 RAs reduce death; harms are class-specific and small in absolute terms.

实际结果
All-cause death OR 0.88 for both classes (high certainty).
方法学质量
high
相关性
direct (limits the one-sided presentation)

原始文献
13 large trials: SGLT2 drugs protect kidneys and heart; ketoacidosis rare

Nuffield Department of Population Health Renal Studies Group; SMART-C Consortium. Lancet 2022 · PMID 36351458 · DOI 10.1016/S0140-6736(22)02074-8 Meta分析已核查全文新增

Amputation signal from CANVAS was not replicated in 12 other SGLT2i trials.

实际结果
RR 1.06 (0.93-1.21) excluding CANVAS.
方法学质量
high
相关性
direct
局限
Summary data.

原始文献

总体证据说明了什么？

基于上述研究的编辑评估——其本身并非研究结果。

哪种观点有更好的证据支持？

The literal statement is correct: each class has at least one documented serious-harm signal, several confirmed in RCTs (DKA with SGLT2i, heart failure with thiazolidinediones, severe hypoglycaemia with sulfonylureas, B12 deficiency with metformin). Some signals were not confirmed (class-wide amputation, Fournier gangrene in RCTs, rosiglitazone MI), and the list omits that SGLT2i and GLP-1 RAs reduce death with high certainty.

为什么有些研究权重更大

Network meta-analysis with GRADE and large blinded-trial meta-analyses outweigh single cohorts and case series.

因果关系是否已确立？ 很可能 Causal for several harms (RCT evidence); not established for others.

仍存在的不确定性

Rare harms (NAION, pancreatic) and long-term older-drug harms.

合理的结论

Mostly true as a list of harm signals; misleading if read as a net assessment.

为什么研究结果看起来互相矛盾？

Harm signals and mortality benefits come from the same trials; they answer different questions (specific adverse events vs overall outcome).

这是否取决于国家？

一般生物学效应: Drug pharmacology and physiology are universal.
饮食习惯: Prolonged fasting, Ramadan fasting and very-low-carbohydrate eating change the risk of hypoglycaemia and ketoacidosis with some drugs; relevant in Egypt and among Muslims in Europe.
食品原产地: Not applicable.
污染物: Not applicable.
本国生产条件: Not applicable (same active substances; generics widely used in Egypt).
法律规则: Drug labels and warnings are set by FDA/EMA and national regulators; Egyptian labelling not reviewed.
实际监测数据: No Egypt-specific pharmacovigilance data were found for this outcome.

足以改变结论的地区差异没有可靠数据支持。

统计复核

CI symmetry for mortality and harm ORs – Symmetric (see C-ODD-06.py and C-ODD-02.py). 已重新计算

总体证据的确定性（GRADE，简化版）

主要结局

Existence of serious-harm signals per drug class

起始等级

高（随机试验）

降级理由

inconsistency: several signals not replicated

结果

中

由单名审查者参照 GRADE 方法进行的简化评估；并非正式的 GRADE 专家组评估。GRADE 评价的是针对某一结局的整体证据，而非单项研究。

已核查的临床指南和官方声明

ADA: 9. Pharmacologic Approaches to Glycemic Treatment: Standards of Care in Diabetes-2026 (Diabetes Care 2026;49 Suppl 1) (2026) · 原始文献
Table 9.2 lists class-specific adverse effects (B12 deficiency, DKA, Fournier gangrene, pancreatitis, NAION, bladder cancer, heart failure) and simultaneously recommends SGLT2i/GLP-1 RA for cardiorenal benefit irrespective of A1C (recs. 9.7-9.10).
NICE: Type 2 diabetes in adults: management (NG28), last updated 18 February 2026 (2026) · 原始文献
Benefits of metformin plus an SGLT-2 inhibitor outweigh risks for most people; rare events at <=1 per 1,000.

检索策略

本次审查的文献检索记录（检索式采用原数据库语法）。

数据库	日期	检索式	检索目标	命中数
PubMed	2026年10月26日	`"type 2 diabetes"[ti] AND ("benefits and harms"[ti] OR "drug treatment"[ti] OR "glucose-lowering"[ti] OR "antidiabetic"[ti]) AND network meta-analysis[pt] AND 2020:2026[dp]`	双向	31
Europe PMC	2026年10月26日	`("adverse events" OR "safety") AND ("glucose-lowering drugs" OR "antidiabetic drugs") AND ("umbrella review" OR "network meta-analysis")`	支持性证据	554
Europe PMC	2026年10月26日	`("all-cause mortality") AND ("glucose-lowering" OR "antidiabetic") AND ("network meta-analysis") AND ("type 2 diabetes")`	反对性证据	525
PubMed	2026年10月26日	`(antidiabetic[ti] OR "glucose-lowering"[ti] OR hypoglycemic agents[mh]) AND ("serious adverse"[tiab] OR "safety signal*"[tiab]) AND (pharmacovigilance[tiab] OR meta-analysis[pt]) AND 2020:2026[dp]`	支持性证据	178
PubMed	2026年10月26日	`("SGLT2"[ti] OR "GLP-1"[ti]) AND ("all-cause mortality"[tiab]) AND (meta-analysis[pt]) AND ("type 2 diabetes"[ti]) AND 2021:2026[dp]`	反对性证据	22
ADA Standards of Care 2026, Section 9 (PubMed Central)	2026年10月26日	`Fetched https://pmc.ncbi.nlm.nih.gov/articles/PMC12690185/; Table 9.2 adverse effects per class. The diabetesjournals.org DOI landing pages returned HTTP 403; the identical article was read in PubMed Central.`	指南	1
NICE NG28 (2026)	2026年10月26日	`Fetched https://www.nice.org.uk/guidance/ng28/chapter/Initial-medicines; chapter Initial medicines`	指南	1

待解决的问题

None that would change the verdict.

核查深度如何？

28 项相关研究：12 项核查了全文（其中 10 项进行了结构化偏倚风险评估），16 项基于摘要核查，0 项仅核查了文献信息。尚未进行由两名独立审查者完成的正式评估，也尚未进行独立的人类专家审核。

最近一次科学审核: 2026年10月26日

评定结论的变更历史

2026年10月25日 · 第 1 阶段 · 初始评估: 基本正确 (中)
2026年10月26日 · 第 4 阶段（系统证据审查） · 在扩展证据后重新审查；评定结论不变: 基本正确 (中)
The new NMA (I-ODD-107) confirms both the class-specific harms and the mortality benefits; SMART-C limits the amputation signal. No change.
新来源: I-ODD-106 I-ODD-107 I-ODD-126

通俗解释

该档案网站列出了每一类糖尿病药物已报告的真实副作用，其数字也是正确的。但它遗漏了这样一个事实：在大型试验中，其中几种药物能预防心脏病发作、肾衰竭和死亡，有时就是在它所引用的同一项研究中。还有一些警示信号后来并未得到证实。公正的看法应针对每个人权衡利弊两方面。

与该说法相关的研究

支持该说法的研究 (16)

Rosiglitazone and heart attacks: pooled trials

Nissen SE et al.. N Engl J Med 2007 · PMID 17517853 · DOI 10.1056/NEJMoa072761

Meta分析被 Tayebat 档案网站引用已核查摘要★ 关键研究

Pooling 42 trials suggested 43% more heart attacks; the increase in heart deaths was not statistically certain.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Effect of rosiglitazone on MI and CV death
研究是如何开展的？
Meta-analysis of 42 randomised trials (fixed effects, trial-level data)
研究了多少人、动物或样本？
42 trials (116 screened); Type 2 diabetes in trials >24 weeks, mean age 56, HbA1c 8.2%
实际发现了什么？
MI OR 1.43; CV death OR 1.64 (n.s.).
观察到的效应有多大？
OR 1.43 (1.03-1.98)
研究结果对日常生活意味着什么？
A real warning signal that changed drug regulation.
该研究有哪些局限？
No patient-level data; Rare events, fixed-effect model; Short trials not designed for CV outcomes
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-09, C-ODD-05.
研究结果是否适用于德国和欧洲？
Rosiglitazone no longer marketed in EU (not verified).
从该研究中明确不能得出什么结论？
That all diabetes drugs raise heart attacks.
科学细节 →
Rosiglitazone and heart failure (RECORD)

Komajda M et al.. Eur Heart J 2010 · PMID 20118174 · DOI 10.1093/eurheartj/ehp604

随机对照试验被 Tayebat 档案网站引用全文，含结构化偏倚风险评估

Heart failure doubled with rosiglitazone (2.6 extra cases per 1,000 patient-years), but overall heart deaths did not increase.

该研究是否支持这一具体说法？是 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Heart failure risk with rosiglitazone
研究是如何开展的？
Analysis of heart failure events in an open-label RCT (RECORD)
研究了多少人、动物或样本？
4,447; Type 2 diabetes on metformin or SU, HbA1c 7.9%
实际发现了什么？
HF HR 2.10.
观察到的效应有多大？
HR 2.10 (1.35-3.27)
研究结果对日常生活意味着什么？
Avoid in heart failure.
该研究有哪些局限？
Open-label; sponsored and analysed by the manufacturer (GSK); relatively few HF events (90).
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-09.
研究结果是否适用于德国和欧洲？
European trial.
从该研究中明确不能得出什么结论？
Increased overall CV mortality.
科学细节 →
Pioglitazone and bladder cancer

Tuccori M et al.. BMJ 2016 · PMID 27029385 · DOI 10.1136/bmj.i1541

队列研究被 Tayebat 档案网站引用全文，含结构化偏倚风险评估

Pioglitazone users had 63% more bladder cancer, about 3 extra cases per 10,000 people per year.

该研究是否支持这一具体说法？是 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Pioglitazone and bladder cancer
研究是如何开展的？
Population-based cohort (CPRD), time-varying lagged exposure
研究了多少人、动物或样本？
145,806; New users of antidiabetic drugs
实际发现了什么？
HR 1.63 with dose/duration response.
观察到的效应有多大？
HR 1.63 (1.22-2.19)
研究结果对日常生活意味着什么？
Small absolute risk to consider.
该研究有哪些局限？
Observational; only 54 bladder cancers in pioglitazone users; possible residual confounding or detection (more urine tests).
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-12, C-ODD-05.
研究结果是否适用于德国和欧洲？
UK data.
从该研究中明确不能得出什么结论？
Certain causation.
科学细节 →
Metformin and vitamin B12

de Jager J et al.. BMJ 2010 · PMID 20488910 · DOI 10.1136/bmj.c2181

随机对照试验被 Tayebat 档案网站引用全文，含结构化偏倚风险评估★ 关键研究

Over 4 years metformin lowered B12 by 19%, causing deficiency in about 1 of every 14 users.

该研究是否支持这一具体说法？是 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Metformin and vitamin B12 deficiency
研究是如何开展的？
Multicentre double-blind placebo-controlled RCT (HOME trial follow-up)
研究了多少人、动物或样本？
390; Type 2 diabetes treated with insulin
实际发现了什么？
B12 -19%; NNH 13.8.
观察到的效应有多大？
-19% B12; +7.2 pp deficiency
研究结果对日常生活意味着什么？
Check B12 regularly on metformin.
该研究有哪些局限？
Secondary analysis; groups differed in age and smoking at baseline; one third without final blood sample; only blood levels, no clinical symptoms measured; insulin-treated patients.
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-07.
研究结果是否适用于德国和欧洲？
Dutch data.
从该研究中明确不能得出什么结论？
That metformin is net harmful.
科学细节 →
Metformin and B12 in US survey

Reinstatler L et al.. Diabetes Care 2012 · PMID 22179958 · DOI 10.2337/dc11-1582

横断面研究被 Tayebat 档案网站引用全文，含结构化偏倚风险评估

5.8% of metformin users had B12 deficiency vs 2.4% of other diabetics.

该研究是否支持这一具体说法？是 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
B12 deficiency prevalence with metformin; effect of supplements
研究是如何开展的？
Cross-sectional survey (NHANES)
研究了多少人、动物或样本？
8,488 (1,621 with T2D + 6,867 without); US adults >=50 with/without type 2 diabetes
实际发现了什么？
aOR 2.92.
观察到的效应有多大？
aOR 2.92 (1.26-6.78)
研究结果对日常生活意味着什么？
Consistent B12 effect.
该研究有哪些局限？
Snapshot survey (no time sequence); no duration-response; smoking/alcohol not adjusted; small inconsistencies between abstract and results; blood levels only.
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-07.
研究结果是否适用于德国和欧洲？
US data.
从该研究中明确不能得出什么结论？
Causation alone.
科学细节 →
Metformin and B12 over 13 years

Aroda VR et al.. J Clin Endocrinol Metab 2016 · PMID 26900641 · DOI 10.1210/jc.2015-3754

随机对照试验被 Tayebat 档案网站引用已核查摘要

Each year of metformin raised the odds of B12 deficiency by about 13%.

该研究是否支持这一具体说法？是 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Long-term metformin and B12 deficiency
研究是如何开展的？
Secondary analysis of RCT and open-label extension (DPP/DPPOS)
研究了多少人、动物或样本？
2,155 randomised (1,082 + 1,073); B12 measured in 1,715 at 5 years and 1,520 at 13 years; People with prediabetes and overweight
实际发现了什么？
OR 1.13 per year.
观察到的效应有多大？
OR 1.13 (1.06-1.20)
研究结果对日常生活意味着什么？
Monitor B12.
该研究有哪些局限？
Placebo group partly unblinded after DPP; Secondary analysis
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-07.
研究结果是否适用于德国和欧洲？
US data.
从该研究中明确不能得出什么结论？
Clinical harm magnitude.
科学细节 →
Sulfonylurea vs metformin as first drug

Roumie CL et al.. Ann Intern Med 2012 · PMID 23128859 · DOI 10.7326/0003-4819-157-9-201211060-00003

队列研究被 Tayebat 档案网站引用全文，含结构化偏倚风险评估★ 关键研究

Starting a sulfonylurea instead of metformin was linked to about 1-4 extra heart events or deaths per 1,000 people per year.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Sulfonylurea vs metformin monotherapy on CVD events/death
研究是如何开展的？
Retrospective cohort (VHA + Medicare) with propensity-matched sensitivity analysis
研究了多少人、动物或样本？
253,690; Veterans initiating metformin or SU, excluding CKD and serious illness
实际发现了什么？
aHR 1.21.
观察到的效应有多大？
aHR 1.21 (1.13-1.30)
研究结果对日常生活意味着什么？
Metformin is preferred first-line.
该研究有哪些局限？
Observational; mostly white men; under 1 year median follow-up; many stopped treatment; possible residual confounding.
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-11.
研究结果是否适用于德国和欧洲？
US veterans.
从该研究中明确不能得出什么结论？
That sulfonylureas cause death.
科学细节 →
Saxagliptin heart safety trial

Scirica BM et al.. N Engl J Med 2013 · PMID 23992601 · DOI 10.1056/NEJMoa1307684

随机对照试验被 Tayebat 档案网站引用已核查摘要

No effect on heart attacks or strokes, but more heart failure hospitalisations (3.5% vs 2.8%, i.e. 0.7 percentage points).

该研究是否支持这一具体说法？部分支持 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
CV safety/efficacy of saxagliptin
研究是如何开展的？
Double-blind placebo-controlled CV outcome RCT (SAVOR-TIMI 53)
研究了多少人、动物或样本？
16,492; Type 2 diabetes with or at risk for CV events
实际发现了什么？
HR 1.00 primary; HF HR 1.27.
观察到的效应有多大？
HR 1.27 HF
研究结果对日常生活意味着什么？
Avoid saxagliptin in heart failure risk.
该研究有哪些局限？
Short follow-up; Industry funded
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-05; limits C-ODD-04.
研究结果是否适用于德国和欧洲？
Multinational/US/UK data; transferable to EU practice.
从该研究中明确不能得出什么结论？
Class-wide harm.
科学细节 →
Canagliflozin heart and kidney trials (CANVAS)

Neal B et al.. N Engl J Med 2017 · PMID 28605608 · DOI 10.1056/NEJMoa1611925

随机对照试验被 Tayebat 档案网站引用全文，含结构化偏倚风险评估★ 关键研究

Fewer heart attacks, strokes and heart deaths (-14%), but more toe/foot amputations.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
CV, renal and safety outcomes of canagliflozin
研究是如何开展的？
Integrated analysis of two double-blind RCTs (CANVAS Program)
研究了多少人、动物或样本？
10,142; Type 2 diabetes with high CV risk, mean age 63.3
实际发现了什么？
14% fewer heart attacks, strokes or heart deaths; about twice as many amputations (6.3 vs 3.4 per 1,000 patient-years), mostly toes; more fractures and genital infections.
观察到的效应有多大？
HR 0.86 / 1.97
研究结果对日常生活意味着什么？
Both benefits and risks; foot care matters.
该研究有哪些局限？
Funded and co-analysed by the manufacturer; two trials combined; many secondary results only exploratory.
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-05 (signal); contradicts C-ODD-06 (benefit).
研究结果是否适用于德国和欧洲？
Multinational/US/UK data; transferable to EU practice.
从该研究中明确不能得出什么结论？
That SGLT2 inhibitors are net harmful.
科学细节 →
Ketoacidosis with canagliflozin

Erondu N et al.. Diabetes Care 2015 · PMID 26203064 · DOI 10.2337/dc15-1251

其他被 Tayebat 档案网站引用全文，含结构化偏倚风险评估

Ketoacidosis was rare (7 per 10,000 patients), often with insulin deficiency or triggers.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 低 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Incidence of DKA with canagliflozin
研究是如何开展的？
Pooled safety analysis of randomised trials
研究了多少人、动物或样本？
17,596; Type 2 diabetes in canagliflozin program
实际发现了什么？
DKA 0.07%.
观察到的效应有多大？
0.52-0.76 vs 0.24 per 1,000 PY
研究结果对日常生活意味着什么？
Rare but real risk.
该研究有哪些局限？
Analysed and written by manufacturer employees; only 12 cases; based on investigator reports (may miss cases).
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-02; mixed for C-INS-08.
研究结果是否适用于德国和欧洲？
Multinational/US/UK data; transferable to EU practice.
从该研究中明确不能得出什么结论？
Precise relative risk; that euglycaemic DKA is common (cases here had high glucose).
科学细节 →
Ketoacidosis after starting SGLT2 drugs (letter)

Fralick M et al.. N Engl J Med 2017 · PMID 28591538 · DOI 10.1056/NEJMc1701990

队列研究被 Tayebat 档案网站引用已核查摘要

Content not verified (no abstract).

该研究是否支持这一具体说法？不明确 · 方法学质量: 低 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Risk of DKA after initiating an SGLT2 inhibitor (per title)
研究是如何开展的？
Research letter (no abstract in PubMed; full text not accessed)
研究了多少人、动物或样本？
not verifiable; not verifiable
实际发现了什么？
Not verified.
观察到的效应有多大？
not verified
研究结果对日常生活意味着什么？
See I-ODD-09.
该研究有哪些局限？
Not read; letter format
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-02 (direction).
研究结果是否适用于德国和欧洲？
US data.
从该研究中明确不能得出什么结论？
Anything quantitative.
科学细节 →
Fournier gangrene reports with SGLT2 drugs

Bersoff-Matcha SJ et al.. Ann Intern Med 2019 · PMID 31060053 · DOI 10.7326/M19-0085

病例系列 / 病例报告被 Tayebat 档案网站引用已核查摘要★ 关键研究

FDA received 55 reports of this rare genital infection with SGLT2 drugs; reports cannot show how often it really happens.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 极低 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Describe Fournier gangrene cases with SGLT2 inhibitors vs other agents
研究是如何开展的？
Descriptive case series of FDA FAERS reports and published cases
研究了多少人、动物或样本？
55 SGLT2 cases; 19 comparator cases; Adults with diabetes reported to FAERS
实际发现了什么？
55 vs 19 reports.
观察到的效应有多大？
not estimable
研究结果对日常生活意味着什么？
Be alert to genital pain/swelling.
该研究有哪些局限？
No incidence/causality; Reporting bias (stimulated reporting after warnings); Different time windows
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-08 (signal).
研究结果是否适用于德国和欧洲？
FDA data; EMA issued similar warnings (not verified).
从该研究中明确不能得出什么结论？
Incidence or causality.
科学细节 →
Canagliflozin and fractures

Watts NB et al.. J Clin Endocrinol Metab 2016 · PMID 26580237 · DOI 10.1210/jc.2015-3167

随机对照试验被 Tayebat 档案网站引用已核查摘要

More fractures in one trial with older high-risk patients, not in others.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Canagliflozin and fracture risk
研究是如何开展的？
Pooled analysis of 9 phase 3 RCTs incl. CANVAS
研究了多少人、动物或样本？
10,194; Type 2 diabetes
实际发现了什么？
4.0 vs 2.6% in CANVAS.
观察到的效应有多大？
+1.4 pp in CANVAS
研究结果对日常生活意味着什么？
Fall prevention in frail patients.
该研究有哪些局限？
Pooled post-hoc
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-05 (signal).
研究结果是否适用于德国和欧洲？
Multinational/US/UK data; transferable to EU practice.
从该研究中明确不能得出什么结论？
Class effect.
科学细节 →
GLP-1 drugs and pancreatitis in US claims

Singh S et al.. JAMA Intern Med 2013 · PMID 23440284 · DOI 10.1001/jamainternmed.2013.2720

病例对照研究被 Tayebat 档案网站引用已核查摘要

Users had about twice the odds of pancreatitis, but they also had more risk factors.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 低 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
GLP-1-based therapies and acute pancreatitis
研究是如何开展的？
Population-based matched case-control (US claims)
研究了多少人、动物或样本？
1,269 cases, 1,269 controls; Adults 18-64 with type 2 diabetes
实际发现了什么？
aOR 2.24.
观察到的效应有多大？
aOR 2.24 (1.36-3.68)
研究结果对日常生活意味着什么？
Signal not confirmed in trials.
该研究有哪些局限？
Strong confounding; Claims data
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-04.
研究结果是否适用于德国和欧洲？
US data.
从该研究中明确不能得出什么结论？
Causation.
科学细节 →
Stomach and bowel side effects of weight-loss injections

Sodhi M et al.. JAMA 2023 · PMID 37796527 · DOI 10.1001/jama.2023.19574

队列研究被 Tayebat 档案网站引用已核查摘要

Not verified (no abstract).

该研究是否支持这一具体说法？不明确 · 方法学质量: 低 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
GI adverse events with GLP-1 RAs used for weight loss (per title)
研究是如何开展的？
Research letter; retrospective cohort (no abstract in PubMed; full text could not be retrieved)
研究了多少人、动物或样本？
not verifiable; not verifiable
实际发现了什么？
Not verified.
观察到的效应有多大？
not verified
研究结果对日常生活意味着什么？
GI side effects are common with GLP-1 RAs (RCTs).
该研究有哪些局限？
Not read
它对 Elawady 的具体说法说明了什么？
Background for C-ODD-04.
研究结果是否适用于德国和欧洲？
unknown
从该研究中明确不能得出什么结论？
Anything quantitative.
科学细节 →
Semaglutide and optic nerve stroke

Hathaway JT et al.. JAMA Ophthalmol 2024 · PMID 38958939 · DOI 10.1001/jamaophthalmol.2024.2296

队列研究被 Tayebat 档案网站引用已核查摘要★ 关键研究

In an eye-clinic registry, semaglutide users had about 4 times more NAION, but numbers were small.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 低 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Semaglutide and NAION risk
研究是如何开展的？
Retrospective matched cohort from a neuro-ophthalmology registry
研究了多少人、动物或样本？
16,827 screened; 710 T2D, 979 overweight/obese; Patients evaluated by neuro-ophthalmologists at one academic institution
实际发现了什么？
HR 4.28 (T2D).
观察到的效应有多大？
HR 4.28 (1.62-11.29)
研究结果对日常生活意味着什么？
Possible rare eye risk.
该研究有哪些局限？
Referral population (selection bias); Few events; Single centre
它对 Elawady 的具体说法说明了什么？
Supports C-ODD-10.
研究结果是否适用于德国和欧洲？
US data.
从该研究中明确不能得出什么结论？
Causation or population incidence.
科学细节 →

与该说法相矛盾或限制该说法的研究 (9)

Metformin in overweight type 2 diabetes (UKPDS 34)

UKPDS Group. Lancet 1998 · PMID 9742977

随机对照试验独立交叉核查已核查摘要★ 关键研究

Metformin reduced deaths from any cause by about a third over 10 years.

该研究是否支持这一具体说法？否 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Intensive control with metformin in overweight type 2 diabetes
研究是如何开展的？
RCT within UKPDS (UKPDS 34)
研究了多少人、动物或样本？
753 (main comparison); Overweight newly diagnosed type 2 diabetes, mean age 53
实际发现了什么？
Mortality -36%.
观察到的效应有多大？
RR reduction 36% (9-55)
研究结果对日常生活意味着什么？
Metformin is protective.
该研究有哪些局限？
Small arms; Open-label; Conflicting supplementary result
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-01, C-ODD-06; supports C-INS-11.
研究结果是否适用于德国和欧洲？
UK.
从该研究中明确不能得出什么结论？
Effects in non-overweight patients.
科学细节 →
Empagliflozin heart trial

Zinman B et al.. N Engl J Med 2015 · PMID 26378978 · DOI 10.1056/NEJMoa1504720

随机对照试验独立交叉核查已核查摘要★ 关键研究

Deaths fell from 8.3% to 5.7% over about 3 years.

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
CV outcomes of empagliflozin in type 2 diabetes at high CV risk
研究是如何开展的？
Double-blind placebo-controlled RCT (EMPA-REG OUTCOME)
研究了多少人、动物或样本？
7,020; Type 2 diabetes with high CV risk
实际发现了什么？
Primary 10.
观察到的效应有多大？
HR 0.86; death -32%
研究结果对日常生活意味着什么？
This drug class protects heart and/or kidneys in the studied patients.
该研究有哪些局限？
Industry funded; Selected high-risk population
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-06; context for C-ODD-02, C-ODD-08.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Benefit in low-risk people; absence of side effects.
科学细节 →
Dapagliflozin (Forxiga) in kidney disease

Heerspink HJL et al.. N Engl J Med 2020 · PMID 32970396 · DOI 10.1056/NEJMoa2024816

随机对照试验独立交叉核查全文，含结构化偏倚风险评估★ 关键研究

Forxiga reduced kidney failure and deaths; the trial was stopped early because the benefit was clear.

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Dapagliflozin in CKD with/without diabetes
研究是如何开展的？
Double-blind placebo-controlled RCT (DAPA-CKD)
研究了多少人、动物或样本？
4,304; CKD (eGFR 25-75, albuminuria)
实际发现了什么？
Kidney failure, kidney decline or kidney/heart death: 9.2% with dapagliflozin vs 14.5% with placebo; deaths 4.7% vs 6.8%. Amputations 1.6% vs 1.8%; no ketoacidosis cases with dapagliflozin (2 with placebo).
观察到的效应有多大？
HR 0.61; death HR 0.69
研究结果对日常生活意味着什么？
This drug class protects heart and/or kidneys in the studied patients.
该研究有哪些局限？
Industry funded; stopped early for benefit (can overstate effect); selected high-risk CKD population; type 1 diabetes excluded; median 2.4 years follow-up.
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-06 (Forxiga 'kills'); context for C-ODD-02.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Benefit in low-risk people; absence of side effects.
科学细节 →
Dapagliflozin in heart failure

McMurray JJV et al.. N Engl J Med 2019 · PMID 31535829 · DOI 10.1056/NEJMoa1911303

随机对照试验独立交叉核查已核查摘要★ 关键研究

Fewer heart failure worsenings and deaths with Forxiga.

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Dapagliflozin in heart failure with reduced EF
研究是如何开展的？
Double-blind placebo-controlled RCT (DAPA-HF)
研究了多少人、动物或样本？
4,744; HFrEF NYHA II-IV, EF <=40%, with/without diabetes
实际发现了什么？
Primary 16.
观察到的效应有多大？
HR 0.74; death HR 0.83
研究结果对日常生活意味着什么？
This drug class protects heart and/or kidneys in the studied patients.
该研究有哪些局限？
Industry funded; Selected high-risk population
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-06.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Benefit in low-risk people; absence of side effects.
科学细节 →
Semaglutide heart safety trial

Marso SP et al.. N Engl J Med 2016 · PMID 27633186 · DOI 10.1056/NEJMoa1607141

随机对照试验独立交叉核查已核查摘要★ 关键研究

Fewer heart attacks/strokes, but more diabetic eye complications and stomach side effects.

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
CV safety of semaglutide
研究是如何开展的？
Double-blind placebo-controlled RCT (SUSTAIN-6)
研究了多少人、动物或样本？
3,297; Type 2 diabetes, 83% with CVD/CKD
实际发现了什么？
Primary 6.
观察到的效应有多大？
HR 0.74
研究结果对日常生活意味着什么？
This drug class protects heart and/or kidneys in the studied patients.
该研究有哪些局限？
Industry funded; Selected high-risk population
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-06; GI and retinopathy harms relevant to C-ODD-05.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Benefit in low-risk people; absence of side effects.
科学细节 →
Fournier gangrene in SGLT2 trials

Silverii GA et al.. Diabetes Obes Metab 2020 · PMID 31637829 · DOI 10.1111/dom.13900

Meta分析独立交叉核查已核查摘要

Across 84 trials Fournier gangrene was not more frequent, but cases were too few to be sure.

该研究是否支持这一具体说法？否 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
SGLT2 inhibitors and Fournier gangrene
研究是如何开展的？
Meta-analysis of 84 RCTs
研究了多少人、动物或样本？
42,415 SGLT2i vs 27,158 comparators; Type 2 diabetes in RCTs
实际发现了什么？
OR 0.41 (n.s.).
观察到的效应有多大？
MH-OR 0.41 (0.09-1.82)
研究结果对日常生活意味着什么？
Very rare; no confirmed excess.
该研究有哪些局限？
Very few events; FG not systematically adjudicated
它对 Elawady 的具体说法说明了什么？
Limits C-ODD-08.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Exclude risk.
科学细节 →
RECORD: rosiglitazone heart outcomes

Home PD et al.. Lancet 2009 · PMID 19501900 · DOI 10.1016/S0140-6736(09)60953-3

随机对照试验独立交叉核查已核查摘要★ 关键研究

Overall heart outcomes were not worse, but heart failure doubled; heart-attack effect unclear.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 中 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
CV outcomes of add-on rosiglitazone
研究是如何开展的？
Multicentre open-label RCT (RECORD main results)
研究了多少人、动物或样本？
4,447; Type 2 diabetes on metformin or SU, HbA1c 7.9%
实际发现了什么？
Non-inferior CV; HF HR 2.10.
观察到的效应有多大？
HR 0.99 primary
研究结果对日常生活意味着什么？
Avoid in heart failure.
该研究有哪些局限？
Open-label; Manufacturer funded; Low event rates
它对 Elawady 的具体说法说明了什么？
Supports HF part, limits MI part of C-ODD-09; limits C-ODD-05.
研究结果是否适用于德国和欧洲？
European trial.
从该研究中明确不能得出什么结论？
Settle MI risk.
科学细节 →
13 large trials: SGLT2 drugs protect kidneys and heart; ketoacidosis rare

Nuffield Department of Population Health Renal Studies Group; SMART-C Consortium. Lancet 2022 · PMID 36351458 · DOI 10.1016/S0140-6736(22)02074-8

Meta分析独立交叉核查已核查全文★ 关键研究

Kidney progression -37%, CV death -14%; DKA risk doubled from about 0.2 per 1,000 patient-years.

Amputation signal from CANVAS not replicated in 12 other trials (RR 1.06).

该研究是否支持这一具体说法？部分支持 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Pool all large placebo-controlled SGLT2 inhibitor trials.
研究是如何开展的？
Collaborative meta-analysis of 13 blinded RCTs.
研究了多少人、动物或样本？
90,409 participants.
实际发现了什么？
Clear kidney and heart benefits; DKA doubled but rare; amputation signal mainly from one trial.
观察到的效应有多大？
RR 0.63 kidney; RR 0.86 CV death; DKA RR 2.12.
研究结果对日常生活意味着什么？
For 1,000 people with diabetes and kidney disease treated for a year, about 26 serious events are avoided and about 2 caused.
该研究有哪些局限？
Summary data; industry trials.
它对 Elawady 的具体说法说明了什么？
Confirms DKA risk, refutes net harm.
研究结果是否适用于德国和欧洲？
High.
从该研究中明确不能得出什么结论？
Effects in people following prolonged fasting regimens.
科学细节 →
NICE 2026: offer metformin plus an SGLT-2 inhibitor to most adults with type 2 diabetes

National Institute for Health and Care Excellence (NICE). NICE guideline 2026

指南 / 立场声明独立交叉核查已核查摘要★ 关键研究

Fewer heart events and heart-failure admissions; DKA and Fournier gangrene 1 per 1,000 or fewer.

Weighs harms against benefits rather than listing harms alone.

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Recommend first-line diabetes medicines.
研究是如何开展的？
Guideline based on meta-analyses.
研究了多少人、动物或样本？
Not applicable.
实际发现了什么？
Benefits outweigh risks for most people.
观察到的效应有多大？
HF hospitalisation 54 vs 26-40 per 1,000 over 3 years.
研究结果对日常生活意味着什么？
These medicines are recommended because they prevent heart problems.
该研究有哪些局限？
UK context.
它对 Elawady 的具体说法说明了什么？
Contradicts net harm.
研究结果是否适用于德国和欧洲？
High.
从该研究中明确不能得出什么结论？
Individual choices.
科学细节 →

结果不一或不明确的研究 (3)

Canagliflozin in diabetic kidney disease

Perkovic V et al.. N Engl J Med 2019 · PMID 30990260 · DOI 10.1056/NEJMoa1811744

随机对照试验独立交叉核查全文，含结构化偏倚风险评估★ 关键研究

Kidney failure and heart problems fell by about a third; amputations and fractures were not increased, but rare ketoacidosis was more common (11 vs 1 cases).

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Canagliflozin and renal outcomes in type 2 diabetes with nephropathy
研究是如何开展的？
Double-blind placebo-controlled RCT (CREDENCE)
研究了多少人、动物或样本？
4,401; Type 2 diabetes with albuminuric CKD
实际发现了什么？
Primary kidney/heart outcome 43.2 vs 61.2 per 1000 patient-years (30% lower); amputation 12.3 vs 11.2 per 1000 patient-years (not significant); diabetic ketoacidosis 2.2 vs 0.2 per 1000 patient-years.
观察到的效应有多大？
HR 0.70
研究结果对日常生活意味着什么？
This drug class protects heart and/or kidneys in the studied patients.
该研究有哪些局限？
Industry funded; stopped early for benefit; selected high-risk population; foot checks added during the trial may have lowered amputation risk; rare harms imprecise.
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-06; limits amputation/fracture signals in C-ODD-05.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Benefit in low-risk people; absence of side effects (ketoacidosis was increased); amputation safety in routine care without foot surveillance.
科学细节 →
Liraglutide heart trial

Marso SP et al.. N Engl J Med 2016 · PMID 27295427 · DOI 10.1056/NEJMoa1603827

随机对照试验独立交叉核查全文，含结构化偏倚风险评估★ 关键研究

Fewer heart deaths and deaths overall; pancreatitis not increased, but more gallstone disease and a small, non-significant excess of pancreatic cancer.

该研究是否支持这一具体说法？否 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
CV outcomes of liraglutide
研究是如何开展的？
Double-blind placebo-controlled RCT (LEADER)
研究了多少人、动物或样本？
9,340; Type 2 diabetes with high CV risk
实际发现了什么？
Heart attack, stroke or heart death 13.0% vs 14.9%; deaths 8.2% vs 9.6%; acute pancreatitis 18 vs 23 patients; pancreatic cancer 13 vs 5 (not significant); gallstone disease 3.1% vs 1.9%.
观察到的效应有多大？
HR 0.87; death HR 0.85
研究结果对日常生活意味着什么？
This drug class protects heart and/or kidneys in the studied patients.
该研究有哪些局限？
Industry funded with sponsor co-analysis; high-risk population; safety outcomes not powered and not adjusted for multiple testing; 3.5-5 years follow-up.
它对 Elawady 的具体说法说明了什么？
Contradicts C-ODD-06; limits C-ODD-04.
研究结果是否适用于德国和欧洲？
Multinational trial including European sites or directly applicable to EU practice.
从该研究中明确不能得出什么结论？
Benefit in low-risk people; that pancreatic cancer risk is absent or present (too few cases); anything about pancreatic cysts.
科学细节 →
816 trials: what each diabetes drug class does and does not do

Shi Q et al.. BMJ 2023 · PMID 37024129 · DOI 10.1136/bmj-2022-074068

Meta分析独立交叉核查已核查全文★ 关键研究

SGLT2 inhibitors and GLP-1 agonists reduce death by about 12%; each class has specific side effects.

Confirms class-specific harms (DKA, amputation, hypoglycaemia, heart failure with TZDs) and also mortality reduction.

该研究是否支持这一具体说法？部分支持 · 方法学质量: 高 · 证据确定性：见该说法的总体评估

原始文献 · 科学细节

研究通俗解读
研究人员想了解什么？
Compare benefits and harms of all diabetes drug classes.
研究是如何开展的？
Network meta-analysis of 816 RCTs with GRADE.
研究了多少人、动物或样本？
471,038 patients.
实际发现了什么？
Mortality benefit for SGLT2i and GLP-1 RA; class-specific harms such as DKA and genital infections (SGLT2i), hypoglycaemia (sulfonylureas, insulin), heart failure (TZDs).
观察到的效应有多大？
Death OR 0.88; DKA OR 2.07; amputation OR 1.27.
研究结果对日常生活意味着什么？
No drug is free of side effects; for several classes the benefits clearly outweigh them in people at risk.
该研究有哪些局限？
Low certainty for older drugs.
它对 Elawady 的具体说法说明了什么？
Supports 'every class has harms'; refutes 'net harmful'.
研究结果是否适用于德国和欧洲？
High.
从该研究中明确不能得出什么结论？
Individual treatment choice.
科学细节 →

地区背景

德国 / 欧盟

埃及

在我的国家适用什么？

一般科学评估

基本正确所列出的每一个危害信号都确实存在，但这份清单是片面的，遗漏了已被证实的获益。

该说法涉及生物学或医学关系。上述科学评估不论在哪个国家都适用；各国食品数据不会改变这一评估。

参考文献

Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes.. Nissen SE et al.. N Engl J Med 2007 · PMID 17517853 · DOI 10.1056/NEJMoa072761
Heart failure events with rosiglitazone in type 2 diabetes: data from the RECORD clinical trial.. Komajda M et al.. Eur Heart J 2010 · PMID 20118174 · DOI 10.1093/eurheartj/ehp604
Pioglitazone use and risk of bladder cancer: population based cohort study.. Tuccori M et al.. BMJ 2016 · PMID 27029385 · DOI 10.1136/bmj.i1541
Long term treatment with metformin in patients with type 2 diabetes and risk of vitamin B-12 deficiency: randomised placebo controlled trial.. de Jager J et al.. BMJ 2010 · PMID 20488910 · DOI 10.1136/bmj.c2181
Association of biochemical B₁₂ deficiency with metformin therapy and vitamin B₁₂ supplements: the National Health and Nutrition Examination Survey, 1999-2006.. Reinstatler L et al.. Diabetes Care 2012 · PMID 22179958 · DOI 10.2337/dc11-1582
Long-term Metformin Use and Vitamin B12 Deficiency in the Diabetes Prevention Program Outcomes Study.. Aroda VR et al.. J Clin Endocrinol Metab 2016 · PMID 26900641 · DOI 10.1210/jc.2015-3754
Comparative effectiveness of sulfonylurea and metformin monotherapy on cardiovascular events in type 2 diabetes mellitus: a cohort study.. Roumie CL et al.. Ann Intern Med 2012 · PMID 23128859 · DOI 10.7326/0003-4819-157-9-201211060-00003
Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus.. Scirica BM et al.. N Engl J Med 2013 · PMID 23992601 · DOI 10.1056/NEJMoa1307684
Canagliflozin and Cardiovascular and Renal Events in Type 2 Diabetes.. Neal B et al.. N Engl J Med 2017 · PMID 28605608 · DOI 10.1056/NEJMoa1611925
Diabetic Ketoacidosis and Related Events in the Canagliflozin Type 2 Diabetes Clinical Program.. Erondu N et al.. Diabetes Care 2015 · PMID 26203064 · DOI 10.2337/dc15-1251
Risk of Diabetic Ketoacidosis after Initiation of an SGLT2 Inhibitor.. Fralick M et al.. N Engl J Med 2017 · PMID 28591538 · DOI 10.1056/NEJMc1701990
Fournier Gangrene Associated With Sodium-Glucose Cotransporter-2 Inhibitors: A Review of Spontaneous Postmarketing Cases.. Bersoff-Matcha SJ et al.. Ann Intern Med 2019 · PMID 31060053 · DOI 10.7326/M19-0085
Effects of Canagliflozin on Fracture Risk in Patients With Type 2 Diabetes Mellitus.. Watts NB et al.. J Clin Endocrinol Metab 2016 · PMID 26580237 · DOI 10.1210/jc.2015-3167
Glucagonlike peptide 1-based therapies and risk of hospitalization for acute pancreatitis in type 2 diabetes mellitus: a population-based matched case-control study.. Singh S et al.. JAMA Intern Med 2013 · PMID 23440284 · DOI 10.1001/jamainternmed.2013.2720
Risk of Gastrointestinal Adverse Events Associated With Glucagon-Like Peptide-1 Receptor Agonists for Weight Loss.. Sodhi M et al.. JAMA 2023 · PMID 37796527 · DOI 10.1001/jama.2023.19574
Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients Prescribed Semaglutide.. Hathaway JT et al.. JAMA Ophthalmol 2024 · PMID 38958939 · DOI 10.1001/jamaophthalmol.2024.2296
Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group.. UKPDS Group. Lancet 1998 · PMID 9742977
Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes.. Zinman B et al.. N Engl J Med 2015 · PMID 26378978 · DOI 10.1056/NEJMoa1504720
Dapagliflozin in Patients with Chronic Kidney Disease.. Heerspink HJL et al.. N Engl J Med 2020 · PMID 32970396 · DOI 10.1056/NEJMoa2024816
Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction.. McMurray JJV et al.. N Engl J Med 2019 · PMID 31535829 · DOI 10.1056/NEJMoa1911303
Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes.. Marso SP et al.. N Engl J Med 2016 · PMID 27633186 · DOI 10.1056/NEJMoa1607141
Fournier's gangrene and sodium-glucose co-transporter-2 inhibitors: A meta-analysis of randomized controlled trials.. Silverii GA et al.. Diabetes Obes Metab 2020 · PMID 31637829 · DOI 10.1111/dom.13900
Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial.. Home PD et al.. Lancet 2009 · PMID 19501900 · DOI 10.1016/S0140-6736(09)60953-3
Impact of diabetes on the effects of sodium glucose co-transporter-2 inhibitors on kidney outcomes: collaborative meta-analysis of large placebo-controlled trials.. Nuffield Department of Population Health Renal Studies Group; SMART-C Consortium. Lancet 2022 · PMID 36351458 · DOI 10.1016/S0140-6736(22)02074-8
Type 2 diabetes in adults: management (NICE guideline NG28), published 2 December 2015, last updated 18 February 2026. National Institute for Health and Care Excellence (NICE). NICE guideline 2026
Canagliflozin and Renal Outcomes in Type 2 Diabetes and Nephropathy.. Perkovic V et al.. N Engl J Med 2019 · PMID 30990260 · DOI 10.1056/NEJMoa1811744
Liraglutide and Cardiovascular Outcomes in Type 2 Diabetes.. Marso SP et al.. N Engl J Med 2016 · PMID 27295427 · DOI 10.1056/NEJMoa1603827
Benefits and harms of drug treatment for type 2 diabetes: systematic review and network meta-analysis of randomised controlled trials.. Shi Q et al.. BMJ 2023 · PMID 37024129 · DOI 10.1136/bmj-2022-074068

← 返回概览

每一大类降糖药（二甲双胍、磺脲类、噻唑烷二酮类、DPP-4抑制剂、SGLT2抑制剂、GLP-1受体激动剂）在经同行评审的文献中都有严重危害的记录信号。

原始表述

译文

出处

科学评估

1. 说法是什么？

2. 哪些内容可证实是正确的？

3. 哪些内容是错误或夸大的？

4. 哪些内容仍不明确？

5. 哪些研究支持该说法？

6. 哪些研究与该说法相矛盾？

7. 对德国和欧洲而言情况如何？

8. 关于埃及的已知情况是什么？

9. 合理的实际结论是什么？

评估详情

科学证据详解

哪些证据支持该说法？

816 trials: what each diabetes drug class does and does not do

Rosiglitazone and heart failure (RECORD)

哪些证据不支持该说法？

816 trials: what each diabetes drug class does and does not do

13 large trials: SGLT2 drugs protect kidneys and heart; ketoacidosis rare

总体证据说明了什么？

哪种观点有更好的证据支持？

为什么有些研究权重更大

仍存在的不确定性

合理的结论

待解决的问题

核查深度如何？

通俗解释

与该说法相关的研究

支持该说法的研究 (16)

Rosiglitazone and heart attacks: pooled trials

Rosiglitazone and heart failure (RECORD)

Pioglitazone and bladder cancer

Metformin and vitamin B12

Metformin and B12 in US survey

Metformin and B12 over 13 years

Sulfonylurea vs metformin as first drug

Saxagliptin heart safety trial

Canagliflozin heart and kidney trials (CANVAS)

Ketoacidosis with canagliflozin

Ketoacidosis after starting SGLT2 drugs (letter)

Fournier gangrene reports with SGLT2 drugs

Canagliflozin and fractures

GLP-1 drugs and pancreatitis in US claims

Stomach and bowel side effects of weight-loss injections

Semaglutide and optic nerve stroke

与该说法相矛盾或限制该说法的研究 (9)

Metformin in overweight type 2 diabetes (UKPDS 34)

Empagliflozin heart trial

Dapagliflozin (Forxiga) in kidney disease

Dapagliflozin in heart failure

Semaglutide heart safety trial

Fournier gangrene in SGLT2 trials

RECORD: rosiglitazone heart outcomes

13 large trials: SGLT2 drugs protect kidneys and heart; ketoacidosis rare

NICE 2026: offer metformin plus an SGLT-2 inhibitor to most adults with type 2 diabetes

结果不一或不明确的研究 (3)

Canagliflozin in diabetic kidney disease

Liraglutide heart trial

816 trials: what each diabetes drug class does and does not do

地区背景

德国 / 欧盟

埃及

在我的国家适用什么？

一般科学评估

参考文献