Prevalence and predictors of exclusive breastfeeding among women in Kilimanjaro region, Northern Tanzania: a population based cross-sectional study

Exclusive breastfeeding (EBF) is a simple, cheap and cost effective intervention in reducing child mortality and morbidity in low income countries [1, 2]. EBF means an infant receives breast milk from his or her mother or expressed breast milk or a wet nurse for the first six months of life and no other solids/semisolids are given with exception of vitamins, mineral supplements or medicine [3]. It has been estimated that EBF coverage of 90% will help to improve child survival [1]. EBF is recommended during the first six months of infants’ life because it confers many nutritional and health benefits to the child [3]. If well implemented it has been estimated to avert 15% of child deaths and in high HIV prevalence settings, it has been estimated to avert 13% of child deaths [1]. Exclusively breastfed children have shown to have lower risk of gastrointestinal infections and acute respiratory infection compared to children who were not exclusively breastfed [4–7]. EBF has also shown to reduce the rate of mothers to child transmission of HIV compared to mixed feeding practices [8].

WHO reported an overall prevalence of EBF of 36%, the highest rates of EBF were found in East Asia/Pacific (43%) and the lowest rates of EBF in West/Central Africa (20%) [9, 10]. In Sub Saharan Africa where there is high rates of mother to child HIV transmission, malnutrition, infant and child mortality rates, the overall prevalence of EBF was 33% [11]. Tanzania, a country with high infants’ and child mortality rates (51 and 81 per 1000 live births respectively), and high rate of stunting for children under age of 5 (42%), has high suboptimal breastfeeding practices and low prevalence of EBF for infants under 6 months (50%) [12]. According to the countdown report of 2012, Tanzania is among the countries which have made insufficient progress to meet the Millennium development goal 4. The target is to reduce under five mortality to 56/1000 live birth by 2015 (MDG 4) [13].

Information on local factors that influence EBF is vital in guiding strategies to improve EBF trend. The factors influencing EBF have shown to vary from country to country and within countries. Employment status, urban/rural differences, marital status, knowledge on breastfeeding, education status, place of delivery, HIV status, advice on breastfeeding, ant-natal care clinic (ANC) attendance and type of delivery have shown to have an influence on EBF [13–17]. While the Tanzania Demographic and Health Surveys (TDHS) collects data on national prevalence of EBF, it does not generate regional rates, nor does it provide detailed information on predictors of EBF. Further many studies on breastfeeding in Tanzania have only been concentrated in HIV positive women or prevention of mother to child transmission (PMTCT) of HIV [18, 19]. Among few studies on predictors of EBF conducted on general population of women, were either done in the late 90s or based in the urban settings [14, 17, 20]. This paper aimed to determine the prevalence and predictors of EBF in urban and rural areas in Kilimanjaro region, northern Tanzania.

The mean age of the infants was 8.75 (SD = 1.95) months and the median was 9 months. There was no significant difference in the mean age and sex of the urban and rural infants.

Seventy nine percent of the mothers (n = 493) initiated breastfeeding within one hour after delivery. Mothers who were assisted by the skilled health staff during delivery were less likely to initiate breastfeeding within one hour 78.2% (n = 445) compared to those assisted by traditional birth attendants and family members 94.0% (n = 47), p-value = < 0.01. Ninety one percent of mothers were still breastfeeding their infants during the study period. One percent of the mothers gave prelacteal feed to their infants. Twelve percent of the mothers (n = 80) gave their infant water during the first month of life at a median age of one month. At 3 months, 35% of the mothers gave their infants mtori (mashed cooked banana plus beef soup and/or other vegetables) and smooth porridge.

In multivariable logistic regression, advice on breastfeeding after delivery, [adjusted odds ratio (AOR) = 2.6; (95% CI: 1.5, 4.6)], single mothers [AOR = 0.4; (95% CI: 0.2, 0.9)] and alcohol intake [AOR = 0.4; (95% CI: 0.3, 0.7)] remained associated with EBF up to 6 months. Other factors like advice on breastfeeding during ANC visit, age of the child, area of residence, partner’s education were assessed but not associated with EBF up to six months in Kilimanjaro (Table 3).

Results showed a low prevalence of EBF practice in Kilimanjaro region. There was no significant difference in EBF rates between urban and rural areas. Advice on breastfeeding after delivery was an important factor for mothers to practice EBF. Single mothers and those who drank alcohol had decreased odds to practice EBF for six months.

High prevalence of EBF was found to be at one month in both urban and rural areas. The general prevalence of EBF in Kilimanjaro was 88.1% at one month, 65.5% at three months and 20.7% up to six months. Declining trend in EBF prevalence have also been shown by the TDHS where 81%, 51% and 23% of the infants at < 2 months, 2–3 months and 4–5 months respectively were exclusively breastfed [12]. In Kilimanjaro region, there is a traditional practice where mothers are relieved from their daily chores in the first three months after delivery. During this period mother in laws or grandmothers have to feed lactating mothers with special diet believed to enhance milk production. This could explain the higher rates of EBF at the first three months after delivery.

The prevalence of EBF is low compared to the national rate of 50% and is far from the estimated coverage of 90% [1]. In our study we have used recall since birth to estimate EBF prevalence this could have contributed our low prevalence compared to TDHS which used 24 hours recall. Study has shown that 24 hour recall tends to overestimate the EBF rates; for example in Uganda the EBF prevalence was 7% at three months when using recall since birth and 81% by using 24 hours recall [24]. Differences in study designs could explain the differences in prevalence between and within countries [25].

The introduction of prelacteal feed is discouraged because it limits the frequency of suckling and exposes the infant to risks of infections [12, 26]. About one percent of the mothers gave prelacteal feeds (mainly water) to their infants. The rate of prelacteal feeding was low compared to the national report of 31% [12]. In Kigoma higher rates (14%) of giving prelacteal feeds to the infants were also reported [14]. We might have got low proportion of prelacteal feeding compared to other researchers due to long recall bias. Or it might be due to socio desirability bias as the mothers might have reported the correct answer but not the real practice.

Complementary feeding started early; at the age of three months 35% of the mothers gave their infants smooth porridge and mtori. Most of the women in this study (about 80%) were engaged in income generating activity both in formal and informal sector. It may be that women are forced to introduce other foods before resuming work as maternity leave is 84 days in Tanzania. Another explanation may be that either women have little knowledge on the WHO recommendations of infants feeding, or do not believe EBF alone is enough for infants’ growth in the first six months [17, 24, 27, 28]. A qualitative study is needed to give an insight about women’s attitudes and beliefs about EBF in the region, given most of the demographic characteristics were not associated with EBF.

Mothers who drank alcohol had decreased odds to exclusively breastfeed their infants for six months compared to non drinkers. In this study 39% of lactating mothers were drinking alcohol. Studies in Brazil and Australia have shown alcohol intake was associated with reduced duration of EBF [29, 30]. It may be mothers who drink tend to spend less time with their infants and tend to leave them in care of others, who are forced to feed them with other foods when mothers are not around. Alcohol intake has also shown to be associated with other negative maternal and child health indicators like HIV and poor attendance for child immunization [31, 32]. In this setting the prevalence of alcohol intake among women ranges from 31% -39% [32, 33]. Traditionally, alcohol is used widely in Kilimanjaro, in various occasions or celebration such as weddings, burials, and payment of other disciplinary fines. There is a need to have community programs that will address the effects of alcohol intake in pregnant and lactating women. They should discourage women from alcohol use during pregnancy or breastfeeding as it negatively affects newborns, EBF, and general infant growth [30, 31].

Single mothers had significantly lower prevalence of EBF for six months compared to cohabiting or married women. It may be single mothers lack social support to continue practicing EBF as they have to earn for the family. Reports from Canada, Norway and Uganda showed that mothers living with their partners were more likely to exclusively breastfeed their infants for six months [16, 24, 34] while in Ethiopia single mothers were two times more likely to exclusively breastfeed their infants compared to married ones [35].

Advice on breastfeeding after delivery was a positive factor for mothers to exclusively breastfeed their infants for the first six months of infant’s life. Hospital delivery on its own did not increase the odds of EBF in this study, differing from others [17, 36]. Rather it is the breastfeeding counselling given by health providers to women during pregnancy, immediately after delivery or when they bring their children to the clinic for immunization. Studies from Malawi, Zambia and Morogoro, Tanzania showed that, women tend to believe and follow what they are taught by the health professionals and their advice was shown to positively influence EBF practice [17, 36, 37].

A difference in EBF rates for urban and rural mothers which was observed in other settings was not observed in this study [17, 38]. In those studies women from the rural areas were more likely to give prelacteal feeds and mix fed earlier compared to urban mothers. It may be that, in Kilimanjaro mothers in rural and urban areas are equally exposed to information/media, have equal access to health care and have similar demographic or reproductive health factors. There is a need for further studies to explore whether it is knowledge, cultural factors, or simply myths which affects EBF in the region.

The study had the following limitations. Recall bias could have affected the results. We studied mothers with infants aged 6–12 months and asked feeding practices of the infants when they were less than six months. Mothers might not remember the exact time when weaning started as the information collected was based on the mother’s report of breastfeeding history. Estimating EBF with a cross-sectional study using recall since birth may be difficult and inaccurate; a cohort study would have given reliable estimates as mothers with infants would be followed after delivery and their records of breastfeeding practices could be recorded prospectively. Despite the limitations, this study has its strengths. Women with infants who participated in the study were randomly selected from the whole region. Second the data for this study was collected from the whole of Kilimanjaro region and participation rate was high. Therefore, the results on breastfeeding practices observed in this study can be generalized to the whole region.

Prevalence of EBF for infants up to six months in Kilimanjaro was low (20.7%), with lack of variation between urban and rural areas. The differences in methodology estimating EBF prevalence could have contributed to our low prevalence compared to previous Tanzanian studies. Advice on breastfeeding after delivery, single mothers and alcohol intake influenced the EBF practice in the region. Since breastfeeding advice by providers increased the EBF prevalence, integration and/or strengthening of nutrition counseling at different reproductive and child health (RCH) clinics in the region is needed. We advice health education on the negative effects of alcohol needs to be given to lactating mothers at the RCH and the community at large, as it has shown to be associated with reduced duration of exclusive breastfeeding. Further, a study to explore cultural factors that influence EBF is needed in the region.