Validity of maternal recall of breastfeeding duration 8 to 9 years after childbirth

Abstract

In most studies, breastfeeding duration is reported using maternal recall. Some research has shown that over the short term mothers are able to accurately recall breastfeeding duration. However, there is a lack of evidence of the validity of long-term recall of breastfeeding as many existing studies had small sample sizes and few assessed factors influencing the accuracy of recall. The objectives of this study were to assess the accuracy of maternal recall of breastfeeding duration 8 to 9 years after childbirth and to examine the influential factors on recall accuracy. From 2006 to 2007, we recruited 1,320 mother-infant pairs who gave birth in four geographically distributed public hospitals in Hong Kong. Participants were followed-up for 12 months postpartum or until they stopped breastfeeding. We further followed up participants by telephone, mail or online questionnaire when the child was 8 to 9 years of age. Intraclass correlation (ICC), Bland-Altman plot and weighted Kappa statistics (κ) were used to assess the agreement between recalled and previously recorded duration of breastfeeding. The median of recorded breastfeeding duration was 8 weeks. The strength of agreement between recorded and recalled breastfeeding duration was good (ICC=0.67, p<.001). Mothers who were multiparous, younger and had shorter residence in Hong Kong were significantly more likely to overestimate breastfeeding duration. The weighted Kappa statistics showed almost perfect agreement (κ =0.84, 95% Confidence interval [CI] 0.80-0.87). Study finding suggest that mothers can accurately recall breastfeeding duration 8 to 9 years after birth. However, recall validity varies across maternal sociodemographic factors.