Table 2.
Association of BMI, WC and FMI with all-cause mortality in a random effects meta-analysis of Caucasian adults
| Phenotype | n | Deaths | HR (95% CI) | P-value |
|---|---|---|---|---|
| BMI (kg m−2)a | 169,551 | 27,100 | 1.02 (1.01-1.03) | 1.0 × 10−8 |
| BMI (kg m−2)IWC | 152,631 | 22,506 | 0.97 (0.96-0.98) | 4.5 × 10−6 |
| WC (cm)b | 152,631 | 22,506 | 1.01 (1.01-1.02) | 5.6 × 10−24 |
| WC (cm)IBMI | 152,631 | 22,506 | 1.02 (1.02-1.03) | 8.8 × 10−18 |
| FMI (kg m−2) | 48,192 | 6,436 | 1.05 (1.02-1.08) | 0.003 |
| FMI (kg m−2)IWC | 48,167 | 6,433 | 0.97 (0.95-0.99) | 0.009 |
BMI, body mass index; CI, confidence interval; FMI, fat mass index; HR, estimated hazard ratio of all-cause mortality per unit of the phenotype; n, number of individuals; WC, waist circumference; I, means adjusted for; e.g. BMI (kg m−2)IWC is BMI adjusted for waist circumference.
a
Due to the assumption of linearity of the BMI-mortality association for BMI ≥ 20 kg m−2, the association between high levels of BMI and mortality is potentially underestimated.
b
Due to the assumption of linearity of the WC-mortality association, the association between high levels of WC and mortality is potentially underestimated.