Meta-analysisindependent cross-checkabstract checked★ key study I-INS-122 in-depth review
Insulin in type 2 diabetes: what randomised trials show
Efficacy and safety of insulin in type 2 diabetes: meta-analysis of randomised controlled trials.
Overview
Erpeldinger S et al.. BMC Endocr Disord 2016 · PMID 27391319 · DOI 10.1186/s12902-016-0120-z
In 20 trials with 18,599 people, insulin did not change the risk of death but caused more severe low blood sugar.
Does the study support the specific claim? Partly · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Summarise RCTs of insulin in T2D.
- How was the study done?
Meta-analysis of 20 RCTs.
- How many people, animals or samples were studied?
18,599 patients.
- What was actually found?
Mortality unchanged; severe hypoglycaemia +70%.
- How large was the observed effect?
RR 0.99 death; 1.70 severe hypoglycaemia.
- What do the results mean in everyday life?
Insulin does not kill, but low sugar is a real risk; use it carefully.
- What are the study's limitations?
Older trials; heterogeneity.
- What does it say about Elawady's specific claim?
Refutes 'insulin doubles mortality'; supports caution.
- Do the results apply to Germany and Europe?
Applicable.
- What can explicitly not be concluded from this study?
Cannot address insulin in type 1 diabetes, where it is essential.
Scientific in-depth review
1. Full reference
- Authors
- Erpeldinger S et al.
- Journal
- BMC Endocr Disord
- Year
- 2016
- Identifier
- PMID 27391319 DOI 10.1186/s12902-016-0120-z
- Verification
- verified – PubMed efetch 2026-10-26; Crossref 2026-10-26
- Retracted
- no
- Corrections
- none
- Extent checked
- abstract only
- Conflicts of interest and funding
Not stated in abstract.
2. Research question
What are the effects of insulin vs other glucose-lowering drugs or diet/placebo on clinical outcomes in T2D?
3. Study design and population
- Study design
- Meta-analysis – Systematic review and meta-analysis of RCTs (1950-2013)
- Population
Adults with T2D in RCTs
- Country
Multinational
- Period
Trials 1950-2013
- Sample size
20 RCTs; 18,599 patients
- Inclusion and exclusion criteria
RCTs of insulin vs drugs or diet/placebo reporting clinical outcomes
4. Methods and endpoints
- Exposure / intervention
Insulin
- Comparator
Other glucose-lowering drugs; diet/placebo
- Endpoints
All-cause and CV mortality; CV events; cancer death; microvascular; hypoglycaemia
- Statistical methods
Fixed-effect RR, random effects if heterogeneity; Cochrane risk of bias
5. Actual results
- Results
All-cause mortality vs drugs RR 0.99 (0.92-1.06); CV mortality 0.99 (0.90-1.09); vs diet/placebo 0.92 (0.80-1.07) and 0.95 (0.77-1.18); no effect on secondary outcomes; severe hypoglycaemia RR 1.70 (1.51-1.91) vs drugs.
- Absolute risks
not reported
- Authors' conclusion
No significant evidence of long-term efficacy of insulin on clinical outcomes in T2D; trend to harms such as hypoglycaemia and weight gain.
6. Strengths
- RCT-only
- Duplicate extraction
- Risk-of-bias assessment
7. Weaknesses
- Trials to 2013
- Heterogeneous comparators
- Dominated by UKPDS/ORIGIN
8. Risk of bias
- Orienting tool
- AMSTAR 2
- Type of assessment
- simplified assessment (abstract only)
- Full formal assessment
- no
- Overall risk of bias
- some concerns
- Rationale
Moderate confidence; search to 2013.
9. Assessment of the statistical interpretation
CIs exclude a doubling of mortality (stats/C-INS-05_erpeldinger.py).
10. Relevance to the specific claim
Randomised evidence: insulin in T2D does not increase (or reduce) mortality or CV death, but increases severe hypoglycaemia; supports caution with insulin in T2D (archive position) and refutes 'doubles mortality'.
11. Independent contrary findings
12. Transferability to Germany and Europe
Directly applicable; same biology and clinical standards in Europe.
13. Overall methodological judgement
Key randomised summary; balanced finding.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | some concerns |
| Precision of results | high |
| Directness | direct |
| Reproducibility | replicated |
| Consistency with the overall evidence | consistent |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
15. Linked fact checks
- Insulin therapy in type 2 diabetes roughly doubles all-cause mortality and raises cancer, renal and cardiovascular events, with a dose-response (each extra IU/kg/day +54% mortality), compared with metformin or sulfonylureas. – Studies contradicting or limiting the claim
- Insulin (and injected insulin) is a principal driver of atherosclerosis. – Studies contradicting or limiting the claim
- Archive conclusion: insulin is indispensable in type 1 diabetes; in type 2 diabetes, evidence supports caution with intensive dosing, prioritising metformin, weight loss and dietary change before insulin escalation, and aiming for moderate rather than near-normal HbA1c; this is not the same as rejecting insulin. – Studies supporting the claim
- Injected insulin halts the Krebs cycle, blocks protein and DNA synthesis, starves the brain of glucose and causes seizures and blood clots. – Studies with mixed or unclear results