Factors associated with duration of breastfeeding in Bangladesh: evidence from Bangladesh demographic and health survey 2014

Over the last decade, scientific studies have substantiated the evidence of the integral role of breastfeeding in the survival, growth and development of children, as well as good health and wellbeing of mothers. The World Health Organization (WHO) and the United Nations Children’s Fund (UNICEF) recommended that optimum early breastfeeding particularly within one hour after birth should be encouraged by healthcare professionals [1]. According to WHO, only breast milk can ensure a complete nutritional requirements for growth and development of babies in first six months [2]. Infants should be exclusively breastfed to achieve optimum growth, development and maintenance of health [2]. Furthermore, it is safe and contains antibodies that help to protect infants and boost immunity. Consequently, optimum breastfeeding reduces the risk of diarrhea, respiratory or ear infections and other infectious diseases that increase infant mortality [3]. Furthermore, optimal breastfeeding is also identified as a protective factor for overweight and obesity in childhood [4]. A study clearly mentioned that sub-optimal breastfeeding can increase the risk of mortality in first two or more years of child life [5]. In addition, breastfeeding is inexpensive, easily available, and clean at the right temperature. Breastfeeding also acts as natural family planning method and reduces the risk of developing breast and ovarian cancers [6]. Many of the health benefits of human milk are closely related, that is, the longer the baby receives human milk, the greater are the benefits. For adequate growth and maintenance of health, infants should also receive nutritionally rich and safe complementary foods along with breastfeeding from six months to two years of age [7, 8]. However, knowledge and attitude towards duration of breastfeeding among mothers are influenced by sociocultural, demographic and physiological factors such as education, income, residence, tradition, belief, and parents age [9‐15].

This study was designed to work with the health related issues under the Sustainable Development Goals (SDGs). Breastfeeding practices for recommended duration are still sub- optimal in Bangladesh which would be a challenge to meet the SDGs by 2030. Subsequently, the benefits of breastfeeding would optimum when it continues for at least two years with complementary feeding [16]. To the best of our knowledge, there are a few studies on duration of breastfeeding practices in the context of Bangladesh by using BDHS-1999-2000 [17] and BDHS-2004 dataset [18]. None of these studies considered mothers’ BMI as a factor associated with duration of breastfeeding although some studies showed that BMI was significantly associated with breastfeeding status [19, 20]. Household wealth quintile, women education level and medical facilities have been increasing in Bangladesh during the last decades [21], which may have effect on knowledge, attitude and practice on duration of breastfeeding [22]. Therefore, it is important to investigate the duration of breastfeeding among mothers in Bangladesh considering the latest nationally representative data. The aim of the current study was to determine the effect of socio-economic, demographic and anthropometric variables on the duration of breastfeeding among mothers in Bangladesh.

A total of 3541 mothers having children aged 6–36 months were included in the study to investigate the socio-demographic determinants of DB in Bangladesh. The mean DB among Bangladeshi mothers was 18.91 months (95% CI: 18.65–19.17) and median was 19.00 months. The Kolmogorov–Smirnov non-parametric test exhibited our dependent variable (DB) was normally distributed. In addition, the Levene test showed that the data were homogeneous.

It was found, more than 78% of mothers received ANC visits during their pregnancy period, and independent sample t-test demonstrated that the mean DB was significantly (p < 0.01) lower (18.63 months) among the mothers who received ANC than mothers did not receive (18.63 months). The mean DB among rural mothers (19.00 months) was somewhat longer (p = 0.061) than mothers living in urban environment (18.71 months). Muslim mothers provided their breast milk for a shorter duration (18.82 months) than Non-Muslim mothers (20.10 months) (p < 0.05). The mean DB was longer (19.10 months) among vaginal delivered mothers compared to caesarean delivered mothers (18.27 months) (p < 0.01). Also, mean DB (19.18 months) was longer among mothers who delivered at home than mothers delivered at hospital or clinic (18.48 months) (p < 0.05). Mothers working outside of house provided their breast milk to their children for significantly (p < 0.01) longer time (20.14 months) than housewife mothers (18.51 months). It was found that the DB was decreased with increasing the education level of mothers, and ANOVA showed that the variation of DB among mothers’ education level was significant (p < 0.05). Almost same pattern of DB was observed among fathers’ education level (p < 0.01). Highest mean value of DB was found among mothers living in Rangpur division (20.00 months) followed by Khulna (19.38 months), Barisal (19.17 months), Sylhet (19.16 months), Rajshahi (18.97 months), Chittagong (18.33 months) and Dhaka (17.99 months). The variation of DB among divisions was statistically significant (p < 0.01). It was found that the mean value of DB was decreased with increasing household quintile index, and the variation was significant (p < 0.05). The increasing tendency of mean DB was found with increasing the nutritional status of mothers, however the ANOVA demonstrated that the variation was not significant (p > 0.05) (Table 1).

Pearson’s correlation coefficients demonstrated that the relationship between DB and mothers’ current age (p < 0.01) and their BMI (p < 0.05) was significantly positive. The relationship between DB and total children ever born (p < 0.01) and mothers’ age at first born (p < 0.05) was significantly negative (Table 2).

The significantly difference in mean DB between two and more than two groups of categorical variables provided by t-test and ANOVA respectively, and significantly correlated continuous variables provided by Pearson’s correlation coefficients were used as the predictors in multiple linear regression model. In Table 3, we observed that VIF values of all predictors lie between 0 and 5; there was no evidence of multicollinearity problem among the predictors. The model demonstrated that the mothers’ age had significant (p < 0.01) positive effect on predictors (DB). However, total number of children ever born and mothers’ age at first birth had a significant (p < 0.01) negative effect on duration of breastfeeding. Mothers did not visit ANC who provided DB to their children averagely 1.307 months longer compared to mothers who visited ANC at least one time (p < 0.01). Working mothers at the outside of house provided breast milk to their children averagely 1.240 months higher than housewife mothers (p < 0.01). Mothers living in Barisal, Khulna, Rangpur and Sylhet divisions provided 0.972 (p < 0.05), 1.305 (p < 0.01), 1.823 (p < 0.01) and 1.951 (p < 0.01) months longer respectively than mothers living in Dhaka division. Our selected model explained the variation of dependent variable (DBF) by about 78% (Table 3).

In this study, we found that the mean DB was 18.91 months among Bangladeshi mothers. One of the earlier studies with BDHS-1999-2000 data set reported that the mean DB among Bangladeshi mothers was 31.3 months [26]. Another study with BDHS-2004 dataset found that the mean DB of Bangladeshi mothers was 30.41 months [27]. It is observed that the mean DB is being decreased over time in Bangladesh. It may be occurred due to increase in the higher education level of women and number of caesarean delivery in Bangladesh [21]. Moreover, the average DB in Bangladesh was lower than that of other South Asian countries such as India (20.37 months) [28], Pakistan (21.8 months) [29], Sri Lanka (23.2 months) [30]. These were very old studies. One of the recent Indian studies reported that the median DB was 12 months, according to nationally representative data from the 2011–2012 Indian Human Development Survey II. They also found that the median DB had decreased by 50% from 1992 to 1993 to 2011–2012 [31]. We found that the median DB in Bangladesh was 19 months that was higher than that found in Indian study.

The mean DB was lower among the mothers who received ANC than who did not. Our results did not coincide with other studies [32] who found that DB was longer among the mothers who visited ANC. This dissimilarity was happened due to the fact that urban mothers received more ANC than rural mothers in Bangladesh [33, 34], and we found urban mothers provided their breast milk to their children averagely for a shorter period than rural mothers. Similar results were also found in India [28] that mothers residing in rural areas have longer DB compared to those living in urban areas. Most of the rural mothers delivered at home, and our study showed that the mean DB among home delivered mothers was longer than mothers who delivered at hospitals or clinics. Education, figure consciousness and availability of breast milk substitution in urban area might be the possible reason behind the shorter DB. It was also observed that the average breastfeeding period was shorter in younger mothers than older mothers. Similar results were also found in Brazil [35, 36], China [37], India [28] and Kuwait [38, 39]. This may be due to lack of experience and knowledge of younger mothers regarding breastfeeding. In addition, they might have received less counseling on benefits of breastfeeding. Our results indicated no significant difference in DB between male and female children. This finding was supported by a previous study [38]. The present study detected that average DB among educated mothers was comparatively shorter than low educated mothers. Our result coincided with other studies in Nigeria [32], Kuwait [38, 39] and India [28]. The higher educated women have more opportunities in the workforce and tend to choose their career over fertility-related matters [40]. Higher educated working mothers might not breastfeed their children for long time due to the demand of occupation [41, 42]. Educational status was one of the most important factors that influence breastfeeding practices which concords with the study conducted in Malaysia [43]. However, we found that exterior working mother breastfeed their children for long time than housewife, which was consistent with the findings of other study in Bangladesh [44]. In rural areas of Bangladesh, usually women involved in some casual works such as domestic work, jobs in cottage industries, small-scale marketing and so on. These types of working give them more time to take care of their baby and breastfeed for longer periods. Moreover, this result happened in our population due to working mothers’ education level; our data showed 11% of mothers were higher educated out of whom only 24.1% works outside at home. More research is required regarding this issue. In this study, mean DB was shorter among caesarian mothers. Similar findings were also observed in China [45] and Vietnam [46]. It is well known that mothers with C-section tend to experience longer recovery periods and more medical care [47]. Thus C-section mothers introduce solid foods for her baby and intend to stop breastfeeding earlier than mothers with normal delivery [48]. Therefore, mode of delivery can be stated as an important indicator for the DB. It was found that the mothers who delivered a large number of children had negative effect on DB in Bangladesh because their fertility returns early. In 2018, Al-Kandari also found the same results among Kuwaiti mothers [39]. The fathers’ educational level was also an important factor for DB discovered by the present study. Usually, educated male married educated female and educational level of female showed an inverse relationship with DB found in this study. This result is supported by other studies [39, 49].

It was observed that mothers who lived in Dhaka division breastfed averagely for a shorter period than other divisions in Bangladesh. Women living in Dhaka division, the Capital city of Bangladesh are comparatively more educated than women living in other divisions [50]. Our findings suggested that women who had completed at least primary education, breastfed their children for an averagely shorter period than illiterate women. Thus geographic factor can be mentioned as an important determinant for the DB.

In the present study, we tried to determine the factors which were related to the duration of breastfeeding among mothers in Bangladesh using nationally representative data collected by BDHS-2014. Our selected statistical technique/models provided that ANC, religion, mode of delivery, parents’ education, geographic location (division), mothers’ age, mothers’ BMI, total ever born children, mothers’ age at first birth and household wealth quintile were associated factors of duration of breastfeeding among Bangladeshi mothers. The socio-demographic factors related to overall duration of breastfeeding can be a valuable appliance when planning local actions and policies aimed at improving breastfeeding duration. The present study indicated that the breastfeeding-promotion programme such as a regular maternal, newborn, child and adolescent health (MNCAH) program, world breastfeeding week, national nutrition program (NNP) of Ministry of Health and Family Welfare (MOHFW) in Bangladesh should address our findings. Government should take proper care and more attention about the maternal health benefit of breastfeeding and encourage mothers to breastfeed their child for at least 6 months. Improving mothers’ knowledge and understanding of the breastfeeding was also recommended.