Otherindependent cross-checkabstract checked★ key study I-SYS-110 in-depth review
'Good' HDL cholesterol may not itself protect the heart
Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study.
Overview
Voight BF et al.. Lancet 2012 · PMID 22607825 · DOI 10.1016/S0140-6736(12)60312-2
Genes that raise HDL did not lower heart attack risk, whereas genes that raise LDL did raise it.
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?
Test HDL causality.
- How was the study done?
Mendelian randomisation.
- How many people, animals or samples were studied?
Tens of thousands of cases.
- What was actually found?
HDL not causal; LDL causal.
- How large was the observed effect?
LDL OR 2.13 per SD.
- What do the results mean in everyday life?
Some blood values are signals, others real causes.
- What are the study's limitations?
Genetic assumptions.
- What does it say about Elawady's specific claim?
Partly supports, mainly contradicts.
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
Every lab value.
Scientific in-depth review
1. Full reference
- Authors
- Voight BF et al.
- Journal
- Lancet
- Year
- 2012
- Identifier
- PMID 22607825 DOI 10.1016/S0140-6736(12)60312-2
- Verification
- verified – PubMed efetch 2026-10-26 + Crossref
- Retracted
- no
- Corrections
ErratumIn: Lancet. 2012 Aug 11;380(9841):564
- Extent checked
- abstract only
- Conflicts of interest and funding
NIH, Wellcome, EU, BHF, German BMBF funding; erratum Lancet 2012;380:564 (PubMed)
2. Research question
Is plasma HDL cholesterol causally related to myocardial infarction?
3. Study design and population
- Study design
- Other – Mendelian randomisation study (meta-analysis of genetic association studies)
- Population
Up to 20,913 MI cases and 95,407 controls
- Country
International
- Period
Published 2012
- Sample size
20 studies (LIPG); 12,482 cases/41,331 controls (score)
- Inclusion and exclusion criteria
Genetic association studies
4. Methods and endpoints
- Exposure / intervention
Genetically raised HDL-C (LIPG Asn396Ser; 14-SNP score); LDL score as positive control
- Comparator
Non-carriers / lower score
- Endpoints
Myocardial infarction
- Statistical methods
Mendelian randomisation
5. Actual results
- Results
LIPG 396Ser raised HDL-C by 0.14 mmol/L (expected OR 0.87) but was not associated with MI (OR 0.99, 0.88-1.11). HDL genetic score per 1 SD: OR 0.93 (0.68-1.26). LDL genetic score per 1 SD: OR 2.13 (1.69-2.69), concordant with observational estimate.
- Absolute risks
not applicable
- Authors' conclusion
Some genetic mechanisms that raise HDL-C do not lower MI risk; raising HDL-C may not uniformly reduce risk.
6. Strengths
- Genetic instruments reduce confounding
- LDL positive control
7. Weaknesses
- Pleiotropy possible
- Instrument strength
8. Risk of bias
- Orienting tool
- none applicable
- Type of assessment
- simplified assessment (abstract only)
- Full formal assessment
- no
- Overall risk of bias
- low
- Rationale
MR assumptions plausible; erratum noted.
9. Assessment of the statistical interpretation
CIs mostly log-symmetric; LIPG null fairly informative.
10. Relevance to the specific claim
Fair kernel for C-SYS-07: HDL cholesterol behaves like a marker rather than a cause; but LDL behaves causally in the same study.
11. Independent contrary findings
Same study confirms LDL as causal.
12. Transferability to Germany and Europe
High.
13. Overall methodological judgement
Shows that some lab values are markers and others causes - so the claim's blanket statement is wrong both ways.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | low |
| Precision of results | moderate |
| Directness | indirect |
| Reproducibility | replicated |
| Consistency with the overall evidence | consistent |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | not applicable |