Skip to main content
. 2022 Mar 16;18(3):e13347. doi: 10.1111/mcn.13347

Table 3.

Prevalence and RRs for nonverbal cognitive development, by BMF initiation and duration

Nonverbal cognitive development
Optimal Suboptimal Crude RRb Adjusted RRb (95% CI)
n (%)a n (%)a (95% CI) Model 1 Model 2
BMF initiation (n = 4075) (n = 3959) (n = 3592)
Never breastfed 713 (79.8) 170 (20.2) 1.48 (1.11–1.98) 1.39 (1.11–1.75) 1.29 (1.09–1.53)
Ever breastfed 2749 (85.4) 443 (14.6) Reference Reference Reference
BMF duration (n = 3870) (n = 3765) (n = 3422)
Never breastfed 713 (79.8) 170 (20.2) 1.58 (1.17–2.13) 1.52 (1.17–1.97) 1.42 (1.09–1.85)
>0 to <2 Months 379 (82.8) 72 (17.2) 1.26 (0.80–1.98) 1.49 (0.98–2.26) 1.44 (0.89–2.34)
2 to <4 Months 761 (86.4) 115 (13.6) 0.97 (0.74–1.27) 0.97 (0.75–1.26) 1.01 (0.74–1.37)
4 to <6 Months 485 (86.5) 73 (13.5) 0.99 (0.71–1.38) 0.94 (0.68–1.31) 0.95 (0.63–1.43)
≥6 Months 959 (86.4) 143 (13.6) Reference Reference Reference
p 0.007 <0.001 0.003

Note: Model 1: Adjusted for child sex, CA at assessment and perinatal risk. Model 2: Adjusted for child sex, CA at assessment, perinatal risk, maternal age, parity, type of pregnancy, mother's country of birth, mother's educational level and family situation.

Abbreviations: BMF, breast milk feeding (mother's own milk); CA, corrected age; CI, confidence interval; RR, risk ratio.

a

Proportions were calculated using inverse probability weights to account for nonresponse bias.

b

RRs were derived from the weighted sample (inverse probability of participating at the 2 years CA follow‐up), multilevel mixed‐effects generalized linear regression models with random intercepts at the country and mother level.