Cohort studyindependent cross-checkabstract checked★ key study I-ODD-101 in-depth review
Metformin and blood acid build-up depend on kidney function
Association of Metformin Use With Risk of Lactic Acidosis Across the Range of Kidney Function: A Community-Based Cohort Study.
Overview
Lazarus B et al.. JAMA Intern Med 2018 · PMID 29868840 · DOI 10.1001/jamainternmed.2018.0292
In 75,413 people, metformin was not linked to acidosis overall (HR 0.98) but was linked when kidney function was severely reduced (eGFR <30: HR 2.07).
Does the study support the specific claim? Partly · Methodological quality: moderate
Plain-language summary
Study explained simply
- What did the researchers want to find out?
To measure acidosis risk with metformin at different levels of kidney function.
- How was the study done?
Health-record cohort with time-updated kidney function, replicated in a second large US data set.
- How many people, animals or samples were studied?
75,413 people with diabetes; replication with about 82,000 new users.
- What was actually found?
No higher acidosis risk at eGFR 30 or above; about twice the risk below eGFR 30.
- How large was the observed effect?
HR 0.98 overall; HR 2.07 (1.33-3.22) at eGFR <30.
- What do the results mean in everyday life?
Metformin is safe for most kidneys; doctors stop it when kidney function becomes very poor.
- What are the study's limitations?
Observational; acidosis code is not specific for metformin-related lactic acidosis.
- What does it say about Elawady's specific claim?
Contradicts 'automatic' lactic acidosis; confirms a risk in a contraindicated situation.
- Do the results apply to Germany and Europe?
High; same drug and thresholds.
- What can explicitly not be concluded from this study?
It cannot exclude rare cases during acute illness or dehydration.
Scientific in-depth review
1. Full reference
- Authors
- Lazarus B et al.
- Journal
- JAMA Intern Med
- Year
- 2018
- Identifier
- PMID 29868840 DOI 10.1001/jamainternmed.2018.0292
- Verification
- verified – PubMed efetch 2026-10-26; Crossref works API 2026-10-26
- Retracted
- no
- Corrections
- none
- Extent checked
- abstract only
- Conflicts of interest and funding
Research Support, N.I.H., Extramural (PubMed record); individual disclosures not reviewed (abstract only).
2. Research question
Is metformin use associated with hospitalisation with acidosis across the range of kidney function (eGFR)?
3. Study design and population
- Study design
- Cohort study – Community-based cohort with time-dependent exposure and eGFR stage, plus a replication cohort of new users
- Population
Adults with diabetes in the Geisinger Health System; replication in new metformin and sulfonylurea users from 350 US health systems
- Country
USA
- Period
2004-2017 (primary); 2010-2015 (replication)
- Sample size
75,413 (primary); 67,578 new metformin and 14,439 new sulfonylurea users (replication)
- Inclusion and exclusion criteria
Patients with diabetes; time-dependent eGFR stage
4. Methods and endpoints
- Exposure / intervention
Time-dependent metformin use
- Comparator
Alternative diabetes management; new sulfonylurea users in replication
- Endpoints
Hospitalisation with acidosis (ICD-9-CM 276.2)
- Statistical methods
Cox models with time-dependent exposure and eGFR; propensity-matched and new-user sensitivity analyses
5. Actual results
- Results
2,335 hospitalisations with acidosis over median 5.7 years. Overall adjusted HR 0.98 (95% CI 0.89-1.08); eGFR 45-59: HR 1.16 (0.95-1.41); eGFR 30-44: HR 1.09 (0.83-1.44); eGFR <30: HR 2.07 (1.33-3.22). Consistent in new-user comparison with sulfonylureas (eGFR 30-44: HR 0.77, 0.29-2.05), propensity-matched cohort (0.71, 0.45-1.12) and replication cohort (0.86, 0.37-2.01).
- Absolute risks
Not reported in the abstract (event count 2,335 across all users; group-specific rates not given).
- Authors' conclusion
Metformin use was associated with acidosis only at eGFR below 30 mL/min/1.73 m2; results support cautious use at eGFR of at least 30.
6. Strengths
- Large real-world cohort with time-updated kidney function
- Replication in an independent multi-system cohort
- Several sensitivity analyses (new users, propensity matching, exclusion of insulin users)
7. Weaknesses
- Outcome is an administrative code for acidosis of any cause, not adjudicated metformin-associated lactic acidosis
- Observational; residual confounding by indication possible
- Abstract only reviewed
8. Risk of bias
- Orienting tool
- ROBINS-E
- Type of assessment
- simplified assessment (abstract only)
- Full formal assessment
- no
- Overall risk of bias
- some concerns
- Rationale
Confounding by indication and outcome misclassification possible; direction of bias for the eGFR<30 subgroup unclear.
9. Assessment of the statistical interpretation
Hazard ratio CIs are log-symmetric (stats/C-ODD-01.py). The overall null result is precise (upper limit 1.08). The eGFR<30 signal is consistent with the pharmacology (renal clearance).
10. Relevance to the specific claim
Directly tests whether metformin causes acidosis 'automatically': it does not in normal or moderately reduced kidney function; risk rises only in severe kidney impairment, which is a contraindication.
11. Independent contrary findings
Inzucchi 2014 (I-ODD-102) and Cochrane 2010 (I-ODD-08) find incidence indistinguishable from background.
12. Transferability to Germany and Europe
Biology and eGFR thresholds are universal; EU labels use the same eGFR limits.
13. Overall methodological judgement
Good-quality evidence that the lactic acidosis risk is confined to situations with severely reduced kidney function.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | moderate |
| 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
Risk during acute kidney injury and dehydration (e.g. prolonged fasting) is not separately quantified.
15. Linked fact checks
- Metformin (Glucophage) disrupts normal metabolism and causes lactic acidosis (automatically, by blocking the Cori cycle). – Studies with mixed or unclear results