Arrhythmias & electrophysiology
Genetic Obesity and the Risk of Atrial Fibrillation: Causal Estimates from Mendelian Randomization
Authors' conclusion
Our data are consistent with a causal relationship between BMI and incident AF.
Study context
- Population
- Individuals of European ancestry without atrial fibrillation (AF) at baseline, 7 cohorts in the US, Iceland and the Netherlands4,178 incident AF cases
- Participants
- 51,646 individuals
- Comparison
- Genetically predicted body mass index (BMI) vs incident AF
- Outcome
- Causal association between BMI and incident AF
- Follow-up
- Mean 7.4 to 19.2 years
Methodological detailsmeta-analysis of 7 prospective population-based cohorts, instrumental variable analysis, random-effects, inverse variance-weighted meta-analysis
- Meta-analysis of 7 prospective population-based cohorts
- Instrumental variable analysis
- Random-effects, inverse variance-weighted meta-analysis
Pooled result
As reported in the abstract
Causal HR for incident AF per kg/m² BMI (instrumental variable)
Hazard ratio, per kg/m2 (CI)P value
FTO
1.15(1.04–1.26)
P = 0.005BMI gene score
1.11(1.05–1.17)
P < 0.001Per kg/m², as printed
Further pooled results
| Outcome | Difference (kg/m² per A-allele) (95% CI) | P value | |
|---|---|---|---|
| Association of genetic instruments with BMIFTO · age- and sex-adjusted | 0.43 (0.32–0.54) | P < 0.001 | |
| −0.500.51Difference, kg/m² per A-allele (linear scale) |
| Outcome | Difference (kg/m² per 1-U increase) (95% CI) | P value | |
|---|---|---|---|
| Association of genetic instruments with BMIBMI gene score · age- and sex-adjusted | 1.05 (0.90–1.20) | P < 0.001 | |
| −0.500.511.5Difference, kg/m² per 1-U increase (linear scale) |
| Outcome | Estimate | P value | |
|---|---|---|---|
| Association of genetic instruments with incident AFFTO · age- and sex-adjusted | HR 1.07 per A-allele (CI 1.02–1.11) | P = 0.004 | |
| Association of genetic instruments with incident AFBMI gene score · age- and sex-adjusted | HR 1.11 per 1-U increase (CI 1.05–1.18) | P < 0.001 | |
| Observed BMI and incident AF | HR 1.05 per kg/m² (CI 1.04–1.06) | P < 0.001 | |
| 0.512Hazard ratio (log scale) |
- Observed BMI and incident AF: age- and sex-adjusted; consistent with instrumental variable estimates
Original publication
Genetic Obesity and the Risk of Atrial Fibrillation: Causal Estimates from Mendelian Randomization. Circulation. Published 14 December 2016. doi:10.
Summarised from the published abstract; every number is checked against it