BACKGROUND: Studies on prevalence and effects of breastfeeding call for reliable and precise data collection to optimize infant nutrition, growth, and health. Data on breastfeeding and infant nutrition are at risk of, for example, recall bias or social desirability bias.

OBJECTIVE: The aim of the present analysis was to compare data on infant nutrition, that is, breastfeeding, use of infant formula, and introduction to complementary foods, obtained by four different methods. We assumed that weekly short message service (SMS) questions were the most reliable method, to which the other methods were compared.

DESIGN: The study population was part of the Odense Child Cohort. The four methods used were: (a) self-administered questionnaire 3 months postpartum, (b) self-administered questionnaire 18 months postpartum, (c) registrations from health visitors visiting the families several times within the first year of life, and (d) weekly SMS questions introduced shortly after birth.

RESULTS: In total, 639 singleton mothers with data from all four methods were included. The proportion of mothers initiating breastfeeding varied from 86% to 97%, the mean duration of exclusive breastfeeding from 12 to 19 weeks, and the mean age when introduced to complementary foods from 19 to 21 weeks. The mean duration of any breastfeeding was 33 weeks across methods.

CONCLUSIONS: Compared with the weekly SMS questions, the self-administered questionnaires and the health visitors' reports resulted in a greater proportion of mothers with an unknown breastfeeding status, a longer duration of exclusive breastfeeding and later introduction to complementary foods, while the duration of any breastfeeding did not differ.