Influence of psychosocial factors in improving breastfeeding Behavior of mothers in Kakamega central district, Kenya

Abstract

Health professionals currently use fact-based approach to promote breastfeeding behavior in Kenya. This practice imposes messages to breastfeeding mothers without considering their opinions and beliefs thus posing a challenge to the Ministry of Health. In Kakamega Central district 30.7% of mothers initiate breastfeeding within the 1st hour of birth,' 2'1.7% exclusively breastfeed for six months and 27.S% continue breastfeeding to the l" year; where these prevalence fall short of the national goal and desired target of achieving the fourth Millennium Development Goal aimed at reducing the under five mortality rates by 20 IS. This study was conducted to determine the predictive power of maternal attitude, subjective norm and perceived behavioral control on breastfeeding intention of mothers in Kakamega Central District using a modified Theory of Planned Behavior. A sequential exploratory mixed methods design was employed to execute the study. In this design qualitative study was initially conducted then followed by a qualitative study. Purposive sampling was used to select three postnatal clinics and one home based clinic from a total of 10 clinics in Kakamega Central District. Proportionate stratified sampling was used to select 230 respondents out of 3,873 breastfeeding mothers aged between 18-40 years. Qualitative data was collected using a Focus Group Discussion guide and analyzed using Constant Comparative Approach in grounded theory analysis. A structured questionnaire was used to collect Quantitative data where' Exploratory Factor Analysis was applied to test for the dimensionality of the questions. Confirmatory Factor Analysis was used to assess whether proposed constructs influenced measured variables. Skewness and kurtosis tests were used to assess the normality of data obtained. Structural Equation Modeling was used to show and determine the relationships between Latent and Observed variables. Findings indicated that the model fitted the data acceptably well within breastfeeding behavior. The predictive power indicated maternal attitude to a better predictor of breastfeeding intention (~=0.38, p<O.OI, n=230); followed by perceived behavioural control (~=0.3S, p<O.OI, n=230); then subjective norm (~=0.2S, p<O.OS, n=230). Intention in turn strongly predicted breastfeeding behaviour (~=O.9S, p<O.OO1, n=230) and its predictors accounted for 68% of the variance on breastfeeding intention. Findings of this research has provided information on the best approach to be employed by health professionals in promoting optimal breastfeeding through model based educational sessions that will in the long run increase initiation, exclusivity and breastfeeding continuation rates in Kakamega Central District.