Table 3.
Sibling-matched analysis: monthly difference in timing of puberty per 5 kg/m2 higher BMI
| Model 1: Shared confoundersa | Model 2: Shared and non-shared confoundersb | ||
|---|---|---|---|
| Pubertal milestones | No. | Adjusted (95% CI) | Adjusted (95% CI) |
| Boys | |||
| Tanner Genital stage 2 | 796 | −5.2 (−12.4, 2.0) | −5.9 (−13.3, 1.5) |
| Tanner Genital stage 3 | 796 | −6.3 (−13.1, 0.6) | −4.3 (−11.3, 2.8) |
| Tanner Genital stage 4 | 796 | 0.3 (−8.2, 8.7) | 0.8 (−7.6, 9.2) |
| Tanner Genital stage 5 | 796 | −4.1 (−17.6, 9.4) | −5.5 (−19.8, 8.7) |
| Tanner Pubic Hair stage 2 | 796 | −8.9 (−15.5, −2.4) | −9.3 (−15.9, −2.6) |
| Tanner Pubic Hair stage 3 | 796 | −6.7 (−12.8, −0.6) | −4.1 (−10.5, 2.2) |
| Tanner Pubic Hair stage 4 | 796 | −1.3 (−8.2, 5.7) | −0.1 (−7.4, 7.3) |
| Tanner Pubic Hair stage 5 | 796 | −3.1 (−14.4, 8.2) | −3.3 (−14.8, 8.2) |
| Axillary hair | 796 | −9.9 (−18.8, −0.9) | −7.4 (−16.7, 1.9) |
| Acne | 796 | −8.9 (−15.4, −2.4) | −9.7 (−16.2, −3.3) |
| Voice break | 612 | −5.5 (−15.2, 4.2) | −7.9 (−17.5, 1.8) |
| First ejaculation | 796 | −2.3 (−10.4, 5.8) | −2.3 (−10.4, 5.7) |
| Girls | |||
| Tanner Breast stage 2 | 712 | −16.4 (−26.2, −6.6) | −14.1 (−24, −4.2) |
| Tanner Breast stage 3 | 712 | −12.2 (−18.0, −6.4) | −13.6 (−19.6, −7.6) |
| Tanner Breast stage 4 | 712 | −14.6 (−22.2, −7.0) | −12.5 (−20.1, −4.8) |
| Tanner Breast stage 5 | 712 | −20.0 (−35.2, −4.8) | −18.0 (−33.6, −2.4) |
| Tanner Pubic Hair stage 2 | 712 | −7.5 (−12.7, −2.3) | −7.3 (−12.7, −1.9) |
| Tanner Pubic Hair stage 3 | 712 | −7.1 (−13.0, −1.3) | −5.1 (−11.2, 0.9) |
| Tanner Pubic Hair stage 4 | 712 | −13.8 (−21.5, −6.1) | −10.6 (−18.6, −2.5) |
| Tanner Pubic Hair stage 5 | 712 | −13.4 (−25.1, −1.7) | −4.7 (−16.6, 7.2) |
| Axillary hair | 712 | −8.2 (−14.6, −1.7) | −8.6 (−15.1, −2.1) |
| Acne | 712 | −2.0 (−9.8, 5.8) | −3.1 (−11.2, 5.0) |
| Menarche | 712 | −11.4 (−17.2, −5.7) | −9.8 (−15.7, −4.0) |
BMI, body mass index; CI, confidence interval.
a
Model 1 is adjusted for shared confounders by sibling-matched design.
b
Model 2 is adjusted for shared confounders by sibling-matched design and adjusted for non-shared confounders (in regression model): birth date, maternal smoking in first trimester, maternal pre-pregnancy BMI, parity, birthweight, childhood diet, childhood physical activity and childhood sedentary behaviour.