The association of breastfeeding self-efficacy with breastfeeding duration and exclusivity: longitudinal assessment of the predictive validity of the Greek version of the BSES-SF tool

Abstract

Introduction: While breastfeeding self-efficacy (BSES) is an important modifiable determinant of breastfeeding, a structured assessment is not standard practice in Cyprus. We assessed the Greek version of the Breastfeeding Self-Efficacy Scale (BSES-SF), including its predictive validity in terms of Breastfeeding (BF) and Exclusive Breastfeeding (EBF) up to the sixth month. Methods: A methodological study with longitudinal design among 586 mother-infant dyads, as part of the “BrEaST Start in Life” project. BSES was assessed 24–48 h after birth and at the first month. Breastfeeding status was assessed at the clinic, the 1st, 4th and 6th month. The association between BSES and breastfeeding was estimated in logistic regression models and its diagnostic ability in ROC analysis. Results: With Mean = 3.55 (SD = 0.85), BSES was moderate, and lower among Cypriot women, primiparas and those who delivered by Cesarean Section (C/S). There was good internal consistency across the 14 items (Cronbach’s α = 0.94) while factor analysis revealed a two-factor structure. BSES scores were higher among mothers who initiated exclusive breastfeeding (M = 3.92, SD = 0.80) compared to breastfeeding not exclusively (M = 3.29, SD = 0.84) and not breastfeeding (M = 3.04, SD = 1.09; p-value < 0.001). There was a stepwise association with exclusivity (40.5% in the highest vs 7.9% lowest quartile of self-efficacy). The association between in-hospital BSES and long-term EBF persisted in multivariable models. Women in the upper quartile of BSES at 48 h were more likely to breastfeed exclusively by adjOR = 5.3 (95% CI 1.7–17.1) at the 1st and adjOR = 13.7 (95% CI 2.7–68.6) at the 4th month. Similar associations were observed between self-efficacy at the 1st month and BF at subsequent time-points. High first month BSES (> 3.96 as per ROC) had 58.9% positive and 79.6% negative predictive value for breastfeeding at 6 months which reflects higher sensitivity but lower specificity. Conclusions: The Greek version of BSES-SF showed good metric properties (construct, know-group, concurrent and predictive validity). In the absence of community support structures or programmes in Cyprus, prevalence of breastfeeding remains low. This suggests a need for policy, educational and community support interventions, including the systematic use of BSES scale as a screening tool to identify those at higher risk for premature BF discontinuation.