Skip to main content
. 2020 May 20;47(2):106–115. doi: 10.1080/03014460.2020.1751286

Table 4.

Association between breastfed status and age at onset of puberty, as measured as age at menarche in girls and whether voice first changed above the median in boys.

Outcome: Age at menarche (months)
Breastfed type Model 1 (univariable)
Model 2 (confounders*)
Model 3 (conf.+BMI)
Betaa (95% CI) p Value Betaa (95% CI) p Value Betaa (95% CI) p Value
Not breastfed (reference) (reference) (reference)
Breastfed <6m 0.57 (−1.32, 2.46) 0.6 0.35 (−1.55, 2.25) 0.7 0.50 (−1.33, 2.33) 0.6
Breastfed ≥6m 2.76 (0.9, 4.63) 0.004 1.91 (−0.06, 3.88) 0.06 1.51 (−0.38, 3.40) 0.1
Outcome: Age at first voice change (above median)
  Model 1 (univariable)
Model 2 (confounders*)
Model 3 (conf.+BMI)
Breastfed type
Odds ratioa (95% CI)
p Value
Odds ratioa (95% CI)
p Value
Odds ratioa (95% CI)
p Value
Not breastfed (reference) (reference) (reference)
Breastfed <6m 1.12 0.87, 1.45) 0.4 1.09 (0.83, 1.42) 0.6 1.08 (0.83, 1.41) 0.5
Breastfed ≥6m 1.12 (0.87, 1.46) 0.4 1.01 (0.76, 1.33) 0.9 1.02 (0.77, 1.34) 0.9

aFor age at menarche: mean difference in months.

*Potential confounding variables: maternal prenatal anxiety, maternal smoking, maternal age, maternal parity, maternal education, and maternal BMI (model 2). We also considered all variables and a potential mediator of BMI at 10 years (model 3). Note that ‘Effect’ refers to the regression coefficient for the model on age at menarche, and the odds ratio for the model on age at voice change.

For age at voice change: odds ratio between above and below median.