Prevalence and Predictors of Breastfeeding Duration of 24 or More Months Free

The National Survey of Children’s Health is an annual survey designed to provide estimates of health of noninstitutionalized children age 0 to 17 years.⁸ Data are obtained from parents and caregivers through web-based and paper surveys.

Using data from 2018 to 2020, analyses were limited to children aged 2 to 5 years (n = 21 691). Respondents were excluded if they never initiated breastfeeding (n = 3498), were missing values for breastfeeding duration (n = 490), or missing ≥1 covariate (n = 997). Our final analytic sample included 16 706 respondents with all lengths of breastfeeding duration.

Breastfeeding duration was estimated by asking the age when the child stopped breastfeeding (and assuming initiation at birth). If a respondent reported still breastfeeding, breastfeeding duration was estimated as the child’s age.

Covariates include: maternal age at child’s birth (≤30 vs >30 years), marital status of first reported primary caregiver (single versus married or cohabitating), highest education of adults in household (≤high school, any college, ≥college), low birth weight (<2500 vs ≥2500 g), child race or ethnicity as characterized by respondent (non-Hispanic white, non-Hispanic Black or African American, non-Hispanic other, Hispanic), child’s current health insurance coverage (yes versus no), and federal poverty level ([FPL] ≤185% vs >185% FPL). FPL missing values were multiply imputed per National Survey of Children’s Health guidelines.⁹ 

The weighted prevalence of breastfeeding duration ≥24 months was analyzed for each covariate. Mother’s age, child’s race or ethnicity, and FPL were selected a priori to assess for possible interaction; no interactions were found. Adjusted odds ratios were calculated using logistic regression, controlling for all covariates. Breastfeeding rates were calculated at each month. Analyses account for complex sample design, adjusting for clustering, weighting, and stratification using SAS-callable SUDAAN (version 9.4). The Centers for Disease Control and Prevention determined that this secondary analysis of deidentified data was not human subjects research and did not require institutional review board review.

Among children aged 2 to 5 years who were ever breastfed, 11.5% were breastfed for ≥24 months. Breastfeeding rates declined rapidly by infant age, with a drop of ≥10 percentage points from 6 to 7 months and 12 to 13 months (Fig 1). Compared to mothers age >30 years, younger mothers were less likely to breastfeed for ≥24 months (aOR 0.57; 95% CI 0.45, 0.73) (Table 1). Non-married/non-cohabitating caregivers were less likely to breastfeed for ≥24 months than married or cohabitating caregivers (aOR 0.57; 0.40, 0.82). Non-Hispanic White children were less likely to breastfeed for ≥24 months than Hispanic children (aOR 0.70; 95% CI 0.51, 0.96). Compared to households with FPL ≤185%, households with FPL >185% were less likely to breastfeed for ≥24 months (aOR 0.67; 95% CI 0.48, 0.93).

Only 11.5% of US children breastfed for ≥24 months. Breastfeeding rates decline sharply by child age, especially at 6 and 12 months. Results indicated significant differences in breastfeeding ≥24 months by maternal age, caregiver marital status, child race, and household income level. The strength of this analysis is the ability to assess longer breastfeeding duration among a nationally representative sample. Limitations include potential recall bias among mothers of older children and the heterogeneity of the non-Hispanic, other race group. Programmatic interventions and policies aimed at supporting breastfeeding duration could help persons who desire to breastfeed for ≥24 months.

aOR

adjusted odds ratios

FPL

federal poverty level