Late Initiation Of Breastfeeding Is A Risk Factor For Early Interruption Of Exclusive Breast Feeding A Case -control Study At Dollo Ado Refuge Camp, Somali Region, Ethiopia.

Background: Globally , nearly three fourth of infants less than-six was not exclusively-brestfeed. While in Ethiopia, the proportion was only slightly above half. In refugee, setting life is not stable, due to lack of infrastructure and adequate health care service mother and infant face many health related problems, like interruption of exclusively breastfeed which intern lead them to malnutrition and impaired growth. Objectives: To assess the Determinants of Interruption of Exclusive Breastfeeding among mother who have infant less than six month at Dollo Ado Refugees camp, Ethiopia, 2017. Methods: A case -control study design was conduct at Dollo Ado refugee camp from April 05 to 25, 2017. Simple random sampling technique used to select 111 eligible case and 222 controls during the study period. A structured and pretested interviewer administered questionnaire used to collect the data. Logistic regression models tted to assess signicant factor, which determine interruption of Exclusive breast-feeding in the study area. Results: In this study not getting antenatal counseling about infant feeding (A-OR =5.87, 2.63-13.10), postnatal counseling about infant feeding (A-OR= 4.33, 2.7 - 10.78) and presence of problem of breastfeeding (AOR= 10.72, 4.55 - 25.23), late initiation of breastfeeding (AOR= 4.79, 2.28-10.0) were signicant determinants of interruption of exclusive breastfeeding. Conclusions : In this study, not getting antenatal and postnatal counseling about infant and young child feeding, problem of breastfeeding, late initiation of Exclusive breastfeeding were signicant determinants of interruption of EBF. Planer and manager should focus on provision of mother advice on Infant and Young Child Feeding during Antenatal and Postnatal Care. Health professional should be alert on treating problem of breastfeeding. This study shows that not getting antenatal and postnatal counseling on infant feeding, problem of breastfeeding, timely initiation of breastfeeding where found to be determinants of interruption of Exclusive breastfeeding. While marital status, educational, status occupational status, birth spacing, parity, ANC visit gender, age, health of the baby, knowledge mother on EBF were not associated with interruption of EBF.


Introduction
Exclusive breast-feeding is one of the elements of optimal infant and young child feeding which is de ned as giving infants less than six-month only human breast milk including expressed human milk (1)(2)(3). World Health Organization (WHO) suggest that infant less than six month should be Exclusively breast feed as it is safe and su cient mode of feeding to prevent infections, chronic disease and growth impairment it also helps to increase infant survival (1,(3)(4)(5)(6)(7)(8). Globally around 70 % of infants less than six month were not exclusively breastfeed In 2014 (9). In Africa, the Prevalence of interruption of exclusive breast-feeding is 68 % while in Ethiopia it was nearly 50%, according to Ethiopia Demographic and Health Survey (EDHS) 2011 (10,11). Among the 10.9 million worldwide child deaths 60 % of death is due to malnutrition directly or indirectly out of this 60 % of deaths are during the infancy period due to inappropriate feeding (1). Only in sub Saharan Africa 1.16 million infants die in their rst month of life, but if we do not miss the opportunities at home and at health facility for example EBF, we could have saved 800,000 infant deaths. In Africa the leading cause of infant mortality were 28 % infection, 3% diarrhea, 25 % prematurity, which are easily avertable by immunization and EBF (12). The global strategy of World Health Organization on infant and young child feeding (IYCF) was implementing program like Baby Friendly Hospital Initiative (BFHI) universal coding. However, the improvement was slight (13). In refugee camps, which were found in Ethiopia IYCF practices was implemented only in three of 24 camps (14,15).
In Refugee, setting social service were not established due to lack of infrastructure and adequate health care service mother and infant face many health related problems, like interruption of EBF (15). Hence, care full attention should have given to IYCF especially protecting and supporting breastfeeding, which is important not just for the time of emergency only, but it have lifelong impacts on child health and on women's future preference on type of feeding (16,17). Therefore, it is very important to assess factors that may leads to the interruption of exclusive breastfeeding in accordance with the context of population type, because, it is di cult to apply the nding from community-based studies in host population. In addition, there is no speci c study in Ethiopia on refugees setting. Moreover, the ndings of this study will help aid organization and policy makers to set priority intervention plan for the problem, and it will help other researchers, as base line data.

Determinant Factors for Interruption of Exclusive Breastfeeding
The determinant factors for interruption of EBF has been outlined in number of literature in stable population but this factors are not well studied in emergency situation like refugee camps. In emergencies, appropriate and safe IYCF practices are less likely than under stable conditions. Bottle feeding comes with increased risks; poor water quality, an inability to sterilize the bottle/nipple, arti cial ingredients in breast milk substitutes (BMS) (18). Evidences shows the a positive in uence of maternal socio demographic characteristics on EBF for example as the age of the mother increase the chance of the infant being exclusively breastfeed increased (19,20). Educational status affects early interruption of EBF positively (20). Mother working outside the home is inversely associated with interruption of EBF (21)(22)(23)(24)(25)(26). Income was also another determine interruption of breast-feeding (27).
From obstetric factors, signi cant determinants of interruption of EBF includes utilization of antenatal care (ANC) and postnatal care (PNC) service, it is associated with reduced interruption of EBF (19,20,27,28). Following ANC and PNC service, guidance on EBF given to the mother about appropriate type of feeding was another protective factor against interruption of EBF. Mother who give birth in health institution were less likely to interrupt EBF than those who give birth at home (21). Type and place of delivery was also another factor which determine early interruption of EBF (22). Timely initiation of breast feeding is also another determinant for interruption of EBF in infant who were not initiated EBF within rst one hour of delivery were at greater risk for interruption of EBF (28)(29)(30). Breast feeding problem of the mother like cracked nipple cause pain to the mother during breast-feeding which lead the mother to interrupt EBF early (20,21). Lack of prior breastfeeding experience in prim mothers risks them for early interruption of EBF (21). In addition to the obstetric factors, Knowledge of the mother on IYCF practice is also another determinant of interruption of EBF (22,28). Factors related to the child that is de antly important after the birth of the child like pre-lacteal that is giving infant anything before breast milk at any time after delivery favor interruption of EBF though how it does this is still debating (20,22,31). Male sex and increasing age of infant (19) was risk factor for interruption of EBF (20,22,23,25,27). Infant who are feed the rst breast milk were less likely to interrupt EBF (21,32,33). Mothers limited Night time breast feeding was also associated with increased risk with interruption of EBF (21).
Exclusive breast-feeding is one of the elements of optimal infant and young child feeding which is de ned as giving infants less than six-month only human breast milk including expressed human milk (1)(2)(3). World Health Organization (WHO) suggest that infant less than six month should be Exclusively breast feed as it is safe and su cient mode of feeding to prevent infections, chronic disease and growth impairment it also helps to increase infant survival (1,(3)(4)(5)(6)(7)(8). Globally around 70 % of infants less than six month were not exclusively breastfeed In 2014 (9). In Africa, the Prevalence of interruption of exclusive breast-feeding is 68 % while in Ethiopia it was nearly 50%, according to Ethiopia Demographic and Health Survey (EDHS) 2011 (10,11). Among the 10.9 million worldwide child deaths 60 % of death is due to malnutrition directly or indirectly out of this 60 % of deaths are during the infancy period due to inappropriate feeding (1). Only in sub Saharan Africa 1.16 million infants die in their rst month of life, but if we do not miss the opportunities at home and at health facility for example EBF, we could have saved 800,000 infant deaths. In Africa the leading cause of infant mortality were 28 % infection, 3% diarrhea, 25 % prematurity, which are easily avertable by immunization and EBF (12). The global strategy of World Health Organization on infant and young child feeding (IYCF) was implementing program like Baby Friendly Hospital Initiative (BFHI) universal coding. However, the improvement was slight (13). In refugee camps, which were found in Ethiopia IYCF practices was implemented only in three of 24 camps (14,15).
In Refugee, setting social service were not established due to lack of infrastructure and adequate health care service mother and infant face many health related problems, like interruption of EBF (15). Hence, care full attention should have given to IYCF especially protecting and supporting breastfeeding, which is important not just for the time of emergency only, but it have lifelong impacts on child health and on women's future preference on type of feeding (16,17). Therefore, it is very important to assess factors that may leads to the interruption of exclusive breastfeeding in accordance with the context of population type, because, it is di cult to apply the nding from community-based studies in host population. In addition, there is no speci c study in Ethiopia on refugees setting. Moreover, the ndings of this study will help aid organization and policy makers to set priority intervention plan for the problem, and it will help other researchers, as base line data.

Determinant Factors for Interruption of Exclusive Breastfeeding
The determinant factors for interruption of EBF has been outlined in number of literature in stable population but this factors are not well studied in emergency situation like refugee camps. In emergencies, appropriate and safe IYCF practices are less likely than under stable conditions. Bottle feeding comes with increased risks; poor water quality, an inability to sterilize the bottle/nipple, arti cial ingredients in breast milk substitutes (BMS) (18). Evidences shows the a positive in uence of maternal socio demographic characteristics on EBF for example as the age of the mother increase the chance of the infant being exclusively breastfeed increased (19,20). Educational status affects early interruption of EBF positively (20). Mother working outside the home is inversely associated with interruption of EBF (21)(22)(23)(24)(25)(26). Income was also another determine interruption of breast-feeding (27).
From obstetric factors, signi cant determinants of interruption of EBF includes utilization of antenatal care (ANC) and postnatal care (PNC) service, it is associated with reduced interruption of EBF (19,20,27,28). Following ANC and PNC service, guidance on EBF given to the mother about appropriate type of feeding was another protective factor against interruption of EBF. Mother who give birth in health institution were less likely to interrupt EBF than those who give birth at home (21). Type and place of delivery was also another factor which determine early interruption of EBF (22). Timely initiation of breast feeding is also another determinant for interruption of EBF in infant who were not initiated EBF within rst one hour of delivery were at greater risk for interruption of EBF (28)(29)(30). Breast feeding problem of the mother like cracked nipple cause pain to the mother during breast-feeding which lead the mother to interrupt EBF early (20,21). Lack of prior breastfeeding experience in prim mothers risks them for early interruption of EBF (21). In addition to the obstetric factors, Knowledge of the mother on IYCF practice is also another determinant of interruption of EBF (22,28). Factors related to the child that is de antly important after the birth of the child like pre-lacteal that is giving infant anything before breast milk at any time after delivery favor interruption of EBF though how it does this is still debating (20,22,31). Male sex and increasing age of infant (19) was risk factor for interruption of EBF (20,22,23,25,27). Infant who are feed the rst breast milk were less likely to interrupt EBF (21,32,33). Mothers limited Night time breast feeding was also associated with increased risk with interruption of EBF (21).

Study design and setting
A case -control study design was conduct at Dollo Ado refugee camp from April 05 to 25, 2017. Dollo Ado Refugee Camp found in Ethiopia Somali Region Liben Zone Dollo Ado Woreda, 935 km away from Addis Ababa in the southeast direction with around 111,000 host population. Dollo Ado Refugee Camp is one of the 24-Refugee Camps in the country. A total of 251,987 Somalia refugees' population hosted in the ve camps Bokolmayo, Melkadida, Kobe, heliwoyn and Bura amino. The Ethiopian Government provides protection for refugees in collaboration with national, international NGO's, and UN agencies (10,11,14,15).

Study population and sampling
All mothers who have infant less than six month at Dollo Ado Refugee Camp, who comes for collection of ration during the study period were given screening question using a 24 hour recall to classify their breast-feeding status. Mothers, who were severely ill and unable to respond excluded from the study. OpenEpi Version 3 statistical software for unmatched case control studies used to calculate the sample size considering the following assumption: two sided con dence level of 95 %, Power of 80 %, and Ratio of case to control one to two. By taking the proportion of exposure among cases from previous study using signi cant factors with Adjusted OR( A-OR) the maximum sample size was taken which is 302, adding 10 % for non-response rate the nal sample size were 333, (111 cases and 222 controls ). The sampling technique followed identi cation and coding of mothers whose index infant was less than six month from ration registration in Dollo Ado Refugee Camps. On the day of ration distribution, the breastfeeding status of infant less than six month were asked using a 24-hour recall method which is recommended by WHO, to identify case and control. Mother who have infant less than six month and responded yes for the screening question that is "have you given anything to eat or drink to your infant in the previous 24 hour or in the previous day and night" considered as case and who answered no considered as controls. After identifying eligible case and control, by using computer generated random number case and control where recruited with one to two propositi-ratio in the aforementioned refugee camp.

Data collection and analysis
Mothers identi ed as case and control interviewed using structured and pretested questionnaire adapted from EDHS 2016. The questionnaire was prepared in English language, translated to Somali language, and retranslated to original language to check for consistency. The questionnaire used to collect information on maternal socio demographic factors and maternal service use, which determine interruption of EBF from study population. Ten trained data collectors two to each camp and ve Bsc nurse supervisor one to each camp recruited. Descriptive and summary statistic carried out to describe study participants according to different characteristics. In binary logistic regressions model rst a bivariate logistic regression performed for each independent variable with the outcome variable interruption of EBF to identify factors associated with interruption EBF practice at ≤ 0.2. Finally, multiple logistic regression model implemented to determine independent predictors of interruption of EBF. Adjusted odds ratios (AOR) with 95% con dence interval (CI) and p values of less than 0.05 considered statistically signi cant association in the nal model.

Terms and operational de nitions
Case: Those mothers, who have infant less than six month, received any other feeding (animal milk, formula feeds, soups, water etc) besides mother's breast milk in the previous 24-hour hour.
Controls: Those mother who have infant less than six month whose infant were Exclusively Breastfed in the previous 24-hour.
Knowledge of mother is Mother's information on the advantages and recommended duration of exclusive breastfeeding.
Timely initiation of breastfeeding: infants who putted to the breast within one hour of birth Prelacteal feeding: children who have given something other than breast milk during the rst three days of life before they putted to the breast.

Socio-Demographic characteristics of study participant
All study participants were Muslim by religion and speaks Somali language. The mean (± SD) age of mothers was 25 (± 5). Rahan-wayn is the largest ethnic group, followed by Merihan and Hawiye. Majority of the mother were married, uneducated and all most all of the mother work inside home (Table 1).

Determinant Factors for Interruption of Exclusive Breastfeeding
Antenatal care service utilization of the study participant almost universal, Mother who has ANC advice on breast-feeding was 57 % and 92 % among case and control respectively. Place of delivery is 100 % institutional in all of case and control as it privileged them to have immediate birth certi cate. Concerning initiation of EBF 47 % of case and 86 % control initiated timely. All of the interviewed mothers responded that they had feed their infants the rst breast milk (colostrum) in both group, but prevalence of prelactel feeding is 96 % and 87 % in case and control respectively. The main food item used is water or tea, which is 83 % among case and 100 % among controls. PNC care service utilization of the mother in the study is 73 % in case and 94 % in control groups; out of this, 76 % of case and 88 % of control have PNC advice on IYCF. History of bottle-feeding reported in 36 % of case and 46 % of control. Mother who have breastfeeding problem were 35 % of case and 7 % of control. The main type of problem were pain in breast 50 % in case group followed by inadequate breast milk with 31 % in case and 86 % of control, the other problem were refusal of baby 19 % of case and in 14 % of control. The mean (± SD) age of infants were 3 (± 1) months 54 % case and 52 % of controls are male sex.
In the binary logistic regression model Not getting ANC advice on infant and young child feeding has signi cant association (C-OR = 8.02

Discussion
This study shows that not getting antenatal and postnatal counseling on infant feeding, problem of breastfeeding, timely initiation of breastfeeding where found to be determinants of interruption of Exclusive breastfeeding. While marital status, educational, status occupational status, birth spacing, parity, ANC visit gender, age, health of the baby, knowledge mother on EBF were not associated with interruption of EBF.
Not getting Antenatal counseling on Exclusive breastfeeding was a risk factor for interruption of exclusive breastfeeding. In this study, mother who did not receive antenatal counseling on Infant and Young Child Feeding practice while she was pregnant were 5.87 times more likely her infant to interrupt exclusive breastfeeding than mother who have ANC counseling about infant feeding. This could be explained to the fact that mother who have got IYCF counseling on EBF will have time to prepare her selves on appropriate type of infant feeding (19,20,27,28). This is may be the result of motivation by health professionals especially midwives who teach mothers about appropriate infant and young child feeding practices during ANC visit.
In this study mother who did not receive postnatal counseling on infant feeding while she was pregnant was 4 times more likely her infant to interrupt exclusive breastfeeding than those mother who have PNC counseling on infant feeding. This is in line with other studies done in Ethiopia (19,23).
Problem of breastfeeding put infant on greater risk for interruption of EBF, mother who have experienced breastfeeding problem were 10 times more likely her infant to interrupt exclusive breastfeeding compared to mothers who do not have breastfeeding problem on the previous 24-hour recall. This nding is in line with other study result multi center studies in eight different countries (20). This is may be due to in ammation of the breast, cracked nipple that cause pain to the mother or even lack of self-con dence on breast milk or the baby refuses to suck due to mouth ulcer.
Late initiation of breastfeeding was also a major determinant factor for interruption of exclusive breastfeeding. In this study mothers who had initiated breast-feeding lately or one hour after delivery of index infant were 7.7 times more likely to interrupt EBF when compared with those mother who had initiated breast-feeding immediately after birth or within the rst one hour of the birth of the index infant to their infant. This is may be infant may adapt what they have given immediately after birth and it increase their bondage with mother, this may affect infant feeding preference in other way mother who give breast immediately to their infant will continue feeding exclusively only their breast milk because they think that the infant cant adapt other food items. This is also consistent with several studies done in Ethiopia (28)(29)(30).

Strength And Limitation
This study tried to address interruption of EBF on most vulnerable population at most peripheries this may be one of the strength and it also used 24-recall method, which minimizes mothers recall bias. However, not addressing the qualitative aspect of the problem was its limitation. In addition, the study result was compared with the results from stable population due to shortage of literature in refugee population.

Conclusion
The nding of this study has shown that not getting antenatal and postnatal counseling about infant feeding, problem of breastfeeding timely initiation of breastfeeding was signi cant determinants of interruption of Exclusive breastfeeding.
The recommendation to halt the interruption of exclusive breastfeeding in the study areas are focusing on antenatal and postnatal counseling on infant and young child feeding. Health professional and community health worker should be alert on advising mother how to feed their infant especially timely initiation and to treat problem of breast-feeding, like breast pain, refusing baby due to oral trash, teach mothers to have self-con dence in case of inadequate breast milk. Managers and policy makers should have to prioritize there action plan not only on the increment proportion of EBF but also on reducing factors which risks interruption of EBF in the study area. Researchers recommended conducting similar study in the area with another design like longitudinal and tray to explore barriers of exclusive breastfeeding qualitatively.