Table 2. Linear hazard ratios for all-cause and cause-specific mortality per 5 kg m−2 of BMI.
| Exposure, cause of death | Deaths | HR (95% CI) | Pquadratic | Shape | BMIvertex (95% CI) |
|---|---|---|---|---|---|
| BMI at age 20, follow-up from age 20 (N = 11,233): | |||||
| All cause | 2,438 | 1.04 (0.95, 1.14) | 0.021 | convex | 21.5 (16.7, 23.7) |
| Cardiovascular disease | 804 | 1.13 (0.96, 1.32) | 0.416 | convex | 18.6 (<11.8, >44.7) |
| Respiratory disease | 173 | 0.61 (0.42, 0.87) | 0.859 | concave | <11.8 (<11.8, >44.7) |
| Smoking-related cancer | 298 | 0.92 (0.71, 1.19) | 0.880 | concave | 14.3 (<11.8, >44.7) |
| Cancer not smoking-related | 647 | 1.22 (1.03, 1.45) | 0.377 | convex | 16.6 (<11.8, >44.7) |
| BMI in 2001, follow-up from 2001 (N = 4,841): | |||||
| All cause | 718 | 0.95 (0.85, 1.07) | 0.001 | convex | 27.7 (25.5, 30.4) |
| Cardiovascular disease | 231 | 1.10 (0.90, 1.35) | 0.031 | convex | 25.5 (15.1, 29.4) |
| Respiratory disease | 62 | 0.95 (0.64, 1.42) | 0.544 | convex | 27.9 (<11.8, >44.7) |
| Smoking-related cancer | 77 | 0.75 (0.52, 1.08) | 0.913 | concave | <11.8 (<11.8, >44.7) |
| Cancer not smoking-related | 220 | 1.11 (0.91, 1.36) | 0.840 | concave | 42.3 (<11.8, >44.7) |
| BMI at age 20, follow-up from 2001 (N = 4,841): | |||||
| All cause | 718 | 1.06 (0.90, 1.26) | 0.477 | convex | 20.4 (<11.8, >44.7) |
| Cardiovascular disease | 231 | 1.15 (0.86, 1.54) | 0.481 | convex | 19.7 (<11.8, >44.7) |
| Respiratory disease | 62 | 0.76 (0.41, 1.41) | 0.496 | concave | 20.0 (<11.8, 39.6) |
| Smoking-related cancer | 77 | 0.84 (0.49, 1.43) | 0.934 | convex | 36.5 (<11.8, >44.7) |
| Cancer not smoking-related | 220 | 1.07 (0.79, 1.44) | 0.952 | concave | 36.4 (<11.8, >44.7) |
| BMI at age 20, follow-up from age 20, healthy non-smokers at age 20 (N = 6,363): | |||||
| All cause | 1,108 | 1.18 (1.04, 1.35) | 0.051 | convex | 19.5 (<11.8, 26.1) |
| Cardiovascular disease | 350 | 1.34 (1.06, 1.70) | 0.315 | convex | 16.8 (<11.8, >44.7) |
| Respiratory disease | 61 | 0.52 (0.28, 0.97) | 0.906 | concave | <11.8 (<11.8, >44.7) |
| Smoking-related cancer | 104 | 1.64 (1.09, 2.46) | 0.824 | concave | >44.7 (<11.8, >44.7) |
| Cancer not smoking-related | 354 | 1.21 (0.97, 1.52) | 0.632 | convex | 14.8 (<11.8, >44.7) |
| BMI in 2001, follow-up from 2001, never-smokers in 2001 (N = 2,226): | |||||
| All cause | 256 | 1.02 (0.84, 1.25) | 0.091 | convex | 26.1 (14.5, 35.6) |
| Cardiovascular disease | 74 | 1.26 (0.88, 1.80) | 0.419 | convex | 21.9 (<11.8, >44.7) |
| Respiratory disease | 19 | 0.74 (0.34, 1.63) | 0.976 | concave | <11.8 (<11.8, >44.7) |
| Smoking-related cancer | 17 | 1.00 (0.46, 2.14) | 0.778 | concave | 26.0 (<11.8, >44.7) |
| Cancer not smoking-related | 99 | 1.18 (0.87, 1.58) | 0.648 | concave | 33.1 (<11.8, >44.7) |
BMI: body mass index, HR: hazard ratio, CI: confidence interval. BMI was measured at approximately age 20 and around 2001, when subjects were aged 49–78. Cox proportional hazards regression was used with age as the time axis. Models were adjusted for sex, date of birth (cubic splines) and smoking behaviour at the time of BMI reporting. Nonlinearity was assessed by adding BMI2 to the model and assessing its coefficient’s departure from the null (Pquadratic). BMIvertex was estimated as the BMI at which the tangent to the quadratic model was horizontal. In a convex quadratic curve (coefficient for BMI2 > 0), BMIvertex estimates the BMI at which mortality is minimised. BMIvertex values outside the observed range of BMI (11.8 to 44.7 kg m−2) indicate a monotonically increasing or decreasing association among the observed data and were abbreviated for ease of presentation.