Using community volunteers to promote exclusive breastfeeding in Sokoto State, Nigeria

Background Exclusive Breastfeeding (EBF) refers to the practice of feeding breast milk only, (including expressed breast milk) to infants; and excluding water, other liquids, breast milk substitutes, and solid foods. Inadequately breastfed infants are likely to be undernourished and have childhood infections. EBF knowledge and infant feeding practices have not been studied sufficiently in Sokoto State, Nigeria. We describe the results of a randomized community trial to promote Exclusive Breastfeeding (EBF) in two local government areas Kware and Bodinga selected as intervention and control groups respectively. Methods During advocacy meetings with community leaders, a Committee was formed. Members of the Committee were consulted for informed consent and selection of ten female volunteers who would educate mothers about breastfeeding during home visits. Participants comprised mothers of infants who were breastfeeding at the time of the study. A total of 179 mothers were recruited through systematic random sampling from each community. Volunteers conducted in-person interviews using a structured questionnaire and counseled mothers in the intervention group only. Results At baseline, intervention and control groups differed significantly regarding maternal occupation (P=0.07), and age of the index child (P=0.07). 42% of infants in the intervention group were up to 6 months old and about 30% of them were exclusively breastfed. Intention to EBF was significantly associated with maternal age (P=0.01), education (P=0.00) and women who were exclusively breastfeeding (P=0.00). After counseling, all infants up to 6 months of age were exclusively breastfed. The proportion of mothers with intention to EBF increased significantly with maternal age (P=0.00), occupation (P=0.00) and women who were exclusively breastfeeding (P=0.01). Post-intervention surveys showed that source of information and late initiation of breastfeeding was not significantly associated with intention to EBF. Mothers who reported practicing EBF for 6 months, were older (P=0.00) multi-parous (P=0.05) and more educated (P=0.00) compared to those who did not practice EBF. Among them, significantly increased proportion of women agreed that EBF should be continued during the night (P=0.03), infant should be fed on demand (P=0.05), sick child could be given medication (P=0.02), EBF offered protection against childhood diarrhea (P=0.01), and helped mothers with birth spacing (P=0.00). Conclusion This study shows that there is a need for reaching women with reliable information about infant nutrition in Sokoto State. The results show decreased EBF practice among working mothers, young women, mothers with poor education and fewer than five children. Counseling is a useful strategy for promoting the duration of EBF for six months and for developing support systems for nursing mothers. Working mothers may need additional resources in this setting to enable them to practice EBF.

The following information is taken from DHS 2003 survey data [18]: Sokoto State is located in the North West region of Nigeria. 75% of women in the North West are uneducated and one-third of children are severely stunted. There is a positive relationship between antenatal care and educational level. Almost 89% of deliveries take place at home. Estimates show that 59% of women in Northern Nigeria compared to women in the South and almost half of teenage mothers compared to one-third of mothers over 20 years of age, do not receive ante-natal care. Exposure to media is low in Northern Nigeria, in rural areas and among older women while education and wealth are positively correlated with mass media exposure. Although radio is the most commonly used media source, approximately 27% of women in the North West have no exposure to newspaper, TV and radio. About 68% of women are traders and about 58% have no participation in decision making. Breastfeeding is universally accepted in Nigeria yet only one-third of babies are put to the breast within one hour of birth. In the North West, only half of infants are given breast milk within one hour of birth. Low maternal education and poverty are directly related to delayed initiation of breastfeeding and prelacteal feeding. 98% of infants in the North West are ever breastfed and among them, 79% of infants receive a prelacteal feed. Overall, median duration of any breastfeeding in Nigeria is 18.6 months but EBF lasts only half a month because of early introduction of water, breast milk substitutes and other liquids.

Culture
Studies from Zambia [20], Nigeria [21], Ghana [22] and Ethiopia [23] show that early introduction of water and other liquids to infants less than six months of age is acceptable and prelacteal feeds are the norm in these countries. In northeastern Nigeria the rationale for this practice is that a newborn child is thirsty [24]. There is a widely held belief that breast milk alone is insufficient for infants and breastfeeding during pregnancy can harm a developing fetus [24]. Grandmothers and traditional birth attendants often force mothers to discard colostrum [24]. Although fathers and grandmothers consider breast milk as "good" nutrition for the infant, breastfeeding in public is generally opposed by fathers [24]. Home visits and radio are the best methods for reaching women who are usually confined to their homes after marriage [24].
Among rural women in Zambia, fear of dying is reported as a major barrier to Exclusive Breastfeeding and these women rely on older relatives for information about breastfeeding [20]. In Nigeria, women report that colostrum is dirty "like pus" and expressed milk is susceptible to witchcraft and poisoning [21].
EBF knowledge and infant feeding practices have not been studied sufficiently in Sokoto State, Nigeria. We describe the results of a randomized community trial to promote Exclusive Breastfeeding (EBF) in two local government areas Kware and Bodinga selected as intervention and control groups respectively. Both communities share similar socio-demographic variables. The total population of Kware is 134,084 while that of Bodinga is 174,302 [19]. In Kware, maternal and child health services are provided by a Primary Health Centre as well as a Comprehensive Health Centre operated by Usmanu Danfodiyo University Teaching Hospital, Sokoto.

Study Design
Participants included biological mothers who were breastfeeding at the time of study. In the initial phase, a series of advocacy visits to community and opinion leaders were conducted. During these meetings, a Committee was formed, and the strategy was discussed. Ten female volunteers were nominated with these minimum qualifications: primary school certificate, previous breastfeeding experience, residence in the community and willingness to teach mothers about breastfeeding. Volunteers were trained at a four-day workshop held in Kware comprising lectures, role plays and demonstrations using posters and flip charts. Each session lasted two and a half hour. Training content covered counseling skills, the basics of nutrition, exclusive breastfeeding and the survey instrument.
In the second phase of the study the survey instrument was pre-tested in a similar community in neighboring Kebbi state. Results showed that 19% of women in the sample were exclusively breastfeeding and 45% had adequate knowledge of EBF after counseling. This pilot study was followed by baseline surveys in Kware and Bodinga, where volunteers conducted in-person interviews and collected data on socio-demographic characteristics, attitudes, maternal knowledge and infant feeding patterns including exclusive breastfeeding. Post-intervention surveys were conducted six months after counseling using the same questionnaire.
Using a combination of simple and systematic sampling methods, a one in eight sample of 179 mother-child pairs were recruited from each community. In homes without breastfeeding mothers at the time of counting, the next house was chosen. Given the existence of polygamy in Sokoto, whenever a situation was encountered in which there were more than one mother child pairs in a single house, one pair was selected by simple random sampling using the balloting technique.
Data were analyzed and preliminary findings were discussed with the Committee. An action plan was drawn up regarding implementation of counseling activities, and Principal Investigators discussed with Committee members, the advantages of breastfeeding, the myths and difficulties related to breastfeeding practice, importance of the "ten steps to successful breastfeeding", cultural beliefs about pre-lacteal feeding, essential foods that improve nutritional status of mothers, and significance of adequate rest and personal hygiene for nursing mothers. Additionally, community leaders were informed about the critical need for developing community support for breastfeeding. Follow-up visits were conducted by volunteers to investigate whether participants had further concerns regarding breastfeeding.

Ethics
Informed consent was obtained from all participants. The study was approved by Usmanu Danfodio University Institutional Review Board.

Measurement
Chi-square statistical test was used; and sample size was calculated using the two-sided test with a level of significance of 0.05; and approximately 90% power (β=0.1) for detecting the difference of 15% to 30% in knowledge of EBF by demographic characteristics [25].
Independent variables included socio-demographic characteristics of mothers, and knowledge of infant feeding practices.
Participants' knowledge and practice of EBF were assessed using a modified questionnaire based on Community Level Nutrition Information System for Action (COLNISA) [26]. The questionnaire was translated from English to Hausa and back translated to English. Local scholars participated in translation of the survey questionnaire to Hausa. The modified instrument consisted of four questions about attitudes related to EBF and eight questions about benefits of breastfeeding. The questionnaire also collected demographic information including age, parity, education, religion, occupation, husband's occupation, age and sex of the index child. Yes or no responses were coded by interviewers. Each correct response was given a score of 1 and each incorrect response was given a score of 0. Correct responses were based on information provided during counseling. Scores were graded as percentage with a cut-off at 50% to represent adequate knowledge. Participants were stratified into two groups based on their level of knowledge. "Intention to EBF" was defined as knowledge score of >50%. Mothers with knowledge score of <50% were classified as "No intention to EBF". Data was analyzed using EPI INFO 3.3. Table 3 shows that the majority of participating mothers were multiparous. The average number of children was 5. Approximately 20% of mothers in either group were under the age of 20 years and about 50% were between 23 -32 years old.

Characteristics of mothers and infants
Although the majority of women in either community were housewives, approximately 40% of women in Kware (intervention group) and 30% of women in Bodinga (control group) were working as civil servants and traders. This finding was statistically significant (P=0.07). Educational attainment was low in both communities; 34% of women in Kware compared to 43% of women from Bodinga had completed formal education. In both communities, approximately half of the participants initiated breastfeeding within thirty minutes and approximately 20% initiated breastfeeding after 24 hours.
The difference in age of the index child between both communities was statistically significant (P=0.07). 42% of infants in Kware and 37% of infants in Bodinga were up to six months old; however, only 30% of infants in Kware and 20% of infants in Bodinga were exclusively breastfed. In either community about 30% of infants were up to 12 months old; 18% of infants in Bodinga were up to 18 months old versus 15% of infants in Kware; and about 20% of infants in Bodinga were more than 18 months old versus 15% in Kware. In this sample, mother-child pairs residing in Bodinga (control group) were older than mother-child pairs in Kware (intervention group). No discontinuation in breastfeeding was noted in the first four months except for one mother in Kware who stopped breastfeeding due to a new pregnancy. The discontinuation rate in the first 24 months was statistically significantly higher in the control group compared to the intervention group (Bodinga: 50%; Kware: 29.6%; P=0.00). Table 4 shows the demographic characteristics of mothers with intention to EBF before and after counseling. Intention to EBF was defined as knowledge score of >50%. Pre-intervention assessment showed that the proportion of mothers with Intention to EBF increased significantly with maternal age (P=0.01), education (P=0.00) and women who were exclusively breastfeeding (P=0.00). After counseling, the proportion of mothers with intention to EBF increased significantly with maternal age (P=0.00), occupation (P=0.00) and women who were exclusively breastfeeding (P=0.01). The proportion of mothers with non-formal education who planned to breastfeed exclusively increased compared to those with formal education but the difference was not statistically significant. Table 5 show that source of information and reasons for late initiation of breastfeeding were not significantly associated with intention to EBF. Notably, none of the mothers in the sample reported husbands as their source of information.

Results from post-intervention surveys in
Benefit of counselling Table 6 shows demographic characteristics of mothers who reported breastfeeding up to six months after counseling. 75 infants (42%) in Kware were up to six months old and all were exclusively breastfed. The proportion of mothers who were practicing EBF increased significantly with parity (P= 0.05), maternal age (P=0.00), and education (P=0.00). In this setting, older mothers with more than five children who have completed formal education are more likely to practice EBF. Table 7 shows selected responses about perceived benefits of EBF among these mothers. Significantly increased proportion of women agreed that EBF should be continued during the night (P=0.03), infant should be fed on demand (P=0.05), sick child could be given medication (P=0.02), EBF offered protection against childhood diarrhea (P=0.01), and helped mothers with birth spacing (P=0.00).

Discussion
This study is the first randomized trial to promote Exclusive Breast Feeding (EBF) in Sokoto State. The objective is to describe whether counseling by community health volunteers is beneficial in improving maternal knowledge of EBF and infant feeding practices. It is assumed that participants are residents of the North West region and the sample is representative of all women living in this area.
The socio-demographic characteristics of study participants are fairly typical of women in the North West region where poor education, high fertility, and low socio-economic status are prevalent [18]. The study sample comprised housewives, working women, multi-parous women, and few women with formal education. Teen marriages are not uncommon in this setting; approximately 20% of the mothers in either group were under the age of 20 years.
Data from the study show that approximately 20% of mothers initiated breastfeeding after 24 hours suggesting the underlying perception that colostrum is dirty. Although, most women in Kware and Bodinga continued to breastfeed beyond 24 months, less than one third of participants in either community practiced Exclusive Breastfeeding. This indicates a potential role for appropriate education so that mothers can successfully learn about Exclusive Breastfeeding. In both groups, mothers continued to breastfeed their infants for four months. Only one mother reported discontinuing breastfeeding due to a new pregnancy possibly due to the belief that breastfeeding in pregnancy is harmful to the developing fetus.
Approximately 60% of mothers in Kware and 70% of mothers in Bodinga were housewives suggesting that they are not equal partners in decision making. A study on Healthy Timing and Spacing of Pregnancy conducted in Kano State showed that women perceive themselves as isolated, without access to information and unable to express their opinions [27].
Additionally, mothers in this sample are more likely to be confined to their homes and to have delivered at home, further compounding the problem of receiving EBF information from a health worker [28]. Results of this study show that counseling via home visits, increased the duration of EBF for six months. This indicates the existence of information asymmetry and the need for reaching women with reliable information about infant nutrition. None of the participants reported husbands as their source of information suggesting that husbands are unable to communicate factual information about EBF. Mothers who were exclusively breastfeeding agreed that birth spacing and protection against childhood diarrhea were benefits of EBF. This is important in poor communities where the prevalence of diarrhea is high and there are inequities in access to family planning/reproductive health services.
Lessons learned from research and program implementation worldwide show that counseling by community health volunteers is critical in improving EBF outcomes [40]. This finding underscores the importance of breastfeeding support groups in the community especially for young mothers, and those who are under stress and more likely to believe that breast milk is insufficient [40]. Data analysis shows decreased EBF practice among working mothers, young women, mothers with poor education and fewer than five children (occupation: P=0.25; age: P=0.00; education: P=0.00; parity: P=0.05). Involving the community leaders helps to shift cultural norms related to EBF practices in that community and may potentially influence women's empowerment [41][42][43][44].
Data were collected and analyzed in 2006. The timing of the study and characteristics of mother-child pairs are important limitations of this study. In Bodinga (control group), mother-child pairs were older than in Kware (intervention group). Data collection relied on interviews by community health volunteers. Their presence may have influenced responses or participants may have experienced recall bias. More rigorous data analysis would have enhanced the study.

Conclusion
This study shows that there is a need for reaching women with reliable information about infant nutrition in Sokoto State. The results show decreased EBF practice among working mothers, young women, mothers with poor education and fewer than five children. Counseling is a useful strategy for promoting the duration of EBF for six months and for developing support systems for nursing mothers. Working mothers may need additional resources in this setting to enable them to practice EBF.

Competing interests
The authors declare that they have no competing interests.

Authors contributions
Dr. Qureshi interpreted the data, prepared the final manuscript and is the corresponding author. Dr. Oche implemented the study and analyzed the data. Dr. Kabiru guided the design of the study and Dr. Sadiq assisted with data analysis. All authors have reviewed and approved the manuscript.