+Bioline International Official Site (site up-dated regularly)

African Population Studies/Etude de la Population Africaine, Vol. 17, No. 1, May 2002, pp. 69-82

This paper examines the patterns and determinants of breast-feeding status in selected parts of Igboland (Imo State), Nigeria. Using the Multistage Stratified Sampling Design, a total of 819 ever married women were selected from specific areas of the State. Data management was done using Epi-Info version 5.0 while the SPSS PC+ was used for the analysis. The statistical analysis was done using Logistic Regression Model. Results identified the determinants of decision to breast-feed a baby to include education, age, working status, sex of the child, place of delivery and delivery assistants while predictors of specific duration of breast-feeding in the area were also identified. The paper notes that considering the declining trend in breast-feeding duration in the area under study there is need for policy to encourage longer maternity leave for mothers. Population education should be intensified to emphasis the benefits of breast-feeding for the infants.

Dans cet article, l'auteur étudie les types et les déterminants du statut de l'allaitement maternel dans certaines parties de Igboland (Etat d'Imo) au Nigeria. En utilisant un échantillonnage stratifié à plusieurs degrés, 819 femmes mariées ont été sélectionnées dans quelques régions spécifiques du Nigeria. Les données obtenues ont été gérées avec le logiciel Epi-Info, version 5, analysées avec le logiciel SPSS PC+. L'analyse statistique a été faite en utilisant le modèle de Régression Logistique. Les résultats obtenus ont montré que parmi les facteurs qui poussent les femmes à nourrir leur bébés au sein figurent l'éducation, l'âge, le type d'emploi, le sexe de l'enfant, le lieu d'accouchement. Compte tenu du fait que la tendance à allaiter les enfants au sein baisse dans la région considérée, l'auteur de l'article recommande que la période consacrée aux congés de maternité soit plus longue. L'éducation de la population doit être intensifiée pour montrer tous les avantages de l'allaitement maternel pour l'enfant.

Research has documented the benefits of breast-feeding not only for the health of infant but also as an inexpensive and appropriate source of nutrients and to stimulate strong mother-child emotional bond (UN,1984; Odimegwu, 1994, Afaf, 1987, Van Landingham, 1992).

The importance of breast-feeding in the prevention of diahrrhoea has been demonstrated in several studies (UN, 1984). The protection against deadly diseases has been greatest among infants who are exclusively breast-fed. Thus, research has demonstrated that giving young infants supplementary fluids such as water and or tea in addition to breast-milk is associated with a significant increase in the risk of diarrhoeal disease (WHO Update, 1991). The WHO Update reported that in Lima, Peru, the incidence and prevalence rates of diarrhoea in infants younger than 6 months were significantly higher among those who received water and tea in addition to breast-milk than among those who were exclusively breast-fed. The diarrhoea prevalence rates doubled with the addition of these supplementary fluids.

These studies and others support the view that the survival chances of the newly born child are enhanced when it is breast-fed, particularly during the first year of life (McCan et al, 1981). Naturally, breast milk provides the total nutrient requirements for the first four to six months of life, and, when combined with appropriate weaning foods is an invaluable source of nutrient until after the child's second birthday. Colostrum in breast-milk provides the baby's first immunization against disease and prevents diarrhoea, which kills millions of infants. Not only is breast-milk easily digestible, it is safe from contamination, readily available, and cheap (Jelliffe and Jelliffe, 1978). The importance of breast-feeding during this period is therefore widely recognised.

However, research surveys are indicating a declining trend in the pattern and duration of breast-feeding. Modernization is the principal factor that affects the pattern and duration of breast-feeding. The social transformation associated with the continuous process of development in some parts of Nigeria may affect the suckling patterns and consequently reduce the length of breast-feeding (Odimegwu, 1994).

It is in this context that this paper examines the socio-economic and demographic factors that affect the decision to breast-feed, the pattern and duration of breast-feeding among a sample of women in the Eastern part of Nigeria.

Talbot (1926) was the first to call attention to the prevalence of the custom of prolonged breast-feeding and post-partum sexual abstinence in his ethnographic account of Southern Nigeria. He reported that the traditional duration of breast-feeding and sexual abstinence varied widely from one ethnic group to another. For the Yoruba, the period of suckling was given as 2.5 years. Among the Ika Igbo, Delta State, the period was only seven months compared with two years among the Ngwa Igbo. By 1974, the breast-feeding period had apparently dropped to just over 12 months among the Ngwa Igbo women (Ukaegbu, 1977).

In a study to examine fertility differentials among Nigerian ethnic groups resident in Lagos, Adewuyi and Feyisetan (1988) reported that the mean normative duration of breast-feeding at the place of origin is almost 19 months but less than 10 months in Lagos. For the Igbo living in Lagos, the mean normative duration of breast-feeding is 16.3 months. From the Nigerian Fertility Survey, the mean duration of breast-feeding for Imo State is 13.3 months (Odimegwu, 1994). However, these studies did not address the issue of factors affecting breast-feeding status in the area.

Many authors have suggested the use of model in the study of determinants of infant feeding practices (West, 1980). Social class and previous breast-feeding experiences were found to be related to duration of breast-feeding. Popkin et al (1983) have also made a case for the use of explanatory model in the study of breast-feeding dynamics.

This study adapted the model suggested by Winikoff and her colleagues (Laukaran et al 1981). This identifies the social, economic, cultural and biomedical factors influencing the breast-feeding status of mothers. The model includes such underlying factors as mother's age, parity, sex of child. The relationship of these factors to the initiation and duration of breast-feeding has not been critically examined at least in Nigeria.

The fuller participation of women in society as a whole, and especially in the paid labour force has often been alleged to be detrimental to the fulfilment of maternal responsibilities. Because breast-feeding requires physical contact with the infant on a frequent basis, a conflict has been detected between the promotion of breast-feeding and increasing participation of women in non-traditional economic activities outside the home.

Information on the relationship of employment has been derived largely from studies of reasons for weaning as reported by mothers. Several authors have reviewed these studies and concluded that employment is seldom a determinant of early weaning (Van Esterick and Greiner 1980). In this model, women are defined according to their participation in the modern sector and traditional labour force. Employment itself is seen as a function of socioeconomic status and social mobility aspirations. These factors interact with others to determine breast-feeding practices of mothers. This model is therefore used to identify those factors that appear to have influence on breast-feeding status of a sample of ever married women in the Eastern part of Nigeria.

The study is based on data collected from a survey on lactation, birth spacing and fertility among the Igbos in Nigeria, between June and September, 1992. Information was collected on maternity history, family planning use, breast-feeding, work and fertility histories of the eligible women. The study was a single round survey of selected households in Imo State. Only women aged 15-49 years with at least one live-birth and who slept in the household in the night before the interviewer's visit were eligible for interview. A systematic multistage random sampling method was used to select the households and the eventual eligible respondent.

This involved dividing the area into four residential zones that reflect the residential patterns. A zone was randomly picked. Each of the locations selected was divided into blocks. Within each zone, a random sample of two blocks was done. The households in each block were listed. A systematic sampling procedure was used to select the households, where one eligible woman was interviewed.

A total of 1,000 households were selected but at the end of the interview, we were able to have interviewed only 819 ever married women. In a household where there were more than one eligible respondent, a simple lottery method was used to select the one to be interviewed. The relevant questions for this analysis are: Did you ever breast-feed this child? Are you still breast-feeding (Name of Child)? If stopped, how many months did you breast-feed (Name)? Why did you stop breast-feeding (Name)? Other questions were asked on prenatal care services and experience.

The technique for this analysis is the logistic regression. This is specified thus,

Where P is the probability of ever breast-fed, and ever breast-fed at specific terminal durations. to

are independent variables. Under this approach, the dependent variable for each observation in our study takes the value of 1 if the respondent ever breast-fed while a value of O is assigned to those individual, women who never breast-fed.

The independent variables are education, age, place of residence, marital separation (whether living together with spouse), marital status (type of marriage), working status (defined as whether working in a paid employment or not), occupational status (type of occupation), location of place of work, children ever born, type of marriages, sex of child, prenatal care, place of delivery and delivery assistant, living children and desire for more children. These variables are dichotomised.

The decision to breast-feed and to breast-feed for specific durations are the dependent variables. For the decision to breast-feed, the question was framed thus, "Did you ever breast-feed this child?". The decision to breast-feed is coded as Yes = 1 and Not breast-fed = 0. In the case of duration of breast-feeding, it is divided into 6 months, 9 months and 12 months. These are the points where breast-feeding durations peak. In each of the categories, the model is specified as one if she breast-feeds up to a specific duration and zero if not.

Table 1 shows that more than half of the interviewed women lived in the urban area (54 %) and 46 % in the rural areas at the time of the survey.

In the case of age, both urban and rural residents have larger concentration in age group 25-34, while in the case of education, 54 % have primary in the rural and in the urban 51 % have tertiary education. In other words about one out of every 5 of the respondents in the urban survey sites has a higher level of education. This is not surprising because rural-urban migration is selective of the youths who have finished their higher education and are looking for jobs. The urban centres are also the areas where there can be job openings for the high school graduates. This is also supported when we examine the distribution of the working respondents. Whereas in the urban segments, 52 % are working but in the rural area, 82 % are not working (that is in terms of paid jobs).

Also, 95.4% of the respondents are married while the remaining percentage is either widowed (4.5 %) or divorced/separated (1.9%). The type of marriage is monogamy (89.3%) with the urban area having the larger share of 51.2%.

Table 2: Differentials in Duration of Breast-feeding, Imo State, Nigeria
Characteristics	Rural	Urban	Total
Age
15-29	10.6	10.1	10.4
30-39	12.7	9.8	11.1
40-49	13.4	13.0	12.3
Education
None	16.0	12.0	15.6
Primary	13.4	11.1	12.7
Secondary	11.3	10.0	10.6
Tertiary	9.5	9.4	9.6
Husband's Education
None	14.0	-	14.0
Primary	13.1	9.5	13.1
Husband's Occupation
Prof/Tech..	13.4	9.3	9.8
Clerical	11.4	9.5	10.8
Sales	11.8	9.8	10.8
Transportation	13.7	10.8	10.9
Artisans	12.9	11.1	12.5
Farming	16.5	10.4	16.3
Teaching	10.8	10.8	10.7
Work Status
Working for Pay	11.1	9.5	9.9
Not working for Pay	12.9	10.5	12.0
Place of Work
Away	10.0	9.9	10.9
At Home	11.9	8.0	10.1
Occupational Type
Employee of Private Compagny	15.0	9.0	11.9
Governement Employee	9.9	7.0	10.2
Self-Employee	13.0	9.0	9.8
Farmer/Trader	11.5	8.6	11.2
Desire for Children
Yes	14.2	7.9	11.1
No	8.5	6.7	7.7
Children Ever Born
1-2	11.3	9.3	8.5
3-5	13.1	10.2	10.1
6+	13.0	11.2	12.2
Living Children
1-2	9.4	6.5	8.3
3-5	11.3	9.7	10.3
6+	13.4	10.8	12.7
Religion
Roman Catholic Mission	12.4	10.3	11.1
Protestant	12.7	9.1	11.3
Pentecostal	11.8	10.7	11.1
Other Sects	12.7	6.4	11.1
Total Mean	10.4	9.5	10.6
* Calculation based on current status technique.

Table 2 shows that differences in breast-feeding exist not only between urban and rural areas but also according to the age, education of the woman and her husband, the working status of the woman and the husband's occupation among others. It is also shown that women who had more than 6 children will breast-feed on the average 12 months plus (12.2 months) while those with smaller family size will tend to breast-feed for a shorter duration. This applies also to the number of living children categories. A woman with six children may not be in a hurry to breast-feed for longer time since she may not go for another baby. The duration of breast-feeding tends to decrease as the level of education increases. This is because the more educated women are more likely to work away from home, which may compete with childbearing. This variation is more pronounced in the rural than in the urban areas. Variations in breast-feeding duration and practice are also seen according to the mother's work status. Work status is a reflection of the value of time and the household socioeconomic status. Higher value of time is in competition with breast-feeding and thus the shorter the breast-feeding duration. It is thus shown that mothers who work away from home breast-feed for a shorter time in the rural area than those working at home. The pattern is not the same in the urban area where those who work away from home breast-feed longer than those at home. These women are more likely to have their babysitters with them and or their work places have creches where they can visit at break-time and feed their babies. It may also be that they expressed milks which can be given to the baby latter.

It should also be observed that in the rural area, working mothers at home breast-feed longer than those working away from the home and nearer home but in the urban area, women who are working at home breast-feed for a period of 7 months.

Whether a mother breast-feeds her baby or not is a function of a number of factors. The factors affecting the decision to breast-feed were examined. The decision to breast-feed is dichotomised into two: never breast-fed, and ever breast-fed. Table 3 presents the result.

The above result shows that for the entire population education, age 40-49, current working status, civil service status, polygyny, female child, place of delivery and delivery assistant influence the decision to breast-feed significantly.

When the level of education changes from none to primary, the log odds of ever breast feeding increases by the $ð-Coefficients for each of the category, which also shows that women aged 40-49 are less likely to ever breast-feed the last child than those aged 15-29 who are in the young age groups. While the result shows that for the entire sample, working mothers are more likely to breast-feed, they are less likely to initiate breast-feeding in both rural and urban areas. Those in the formal public sector are less likely to ever breast-feed. The identified variables are significant at different levels as indicated.

When the determinants are examined by place of residence, the analysis shows that the main factors that influence decisions to breast-fed in the rural place include tertiary education, widowhood, polygyny, place of delivery and delivery assistant.

In the case of the urban sectors, the predictive variables are age 40-49, place of delivery and delivery assistant. It should be noted that those other variables may be suffering from the effect of interaction. This analysis did not examine the net effect of the covariates (i.e. the categories effects of some of the variables). However, the non-significant effect of children ever born may be the problem of the confounding effect of age and number of living children.

There are three peak periods in the reporting of the duration of breast-feeding. This was divided into three groups for the purpose of this analysis: 1-6, 7-9, 12+ months. A logistic model was then fitted to the data using the selected variables.

From the above table, there are various determinants for breast-feeding duration in the different duration peaks. The determinants for six month-breast-feeding duration include marital separation (i.e. whether couples are living together or separated), employment status and delivery assistance. For duration of nine months, the factors are working status, marriage type (polygyny), sex of the child (female child), antenatal care experience and place of delivery. And for the probability of breast-feeding for 12 months or more, the factors of interest include education (particularly secondary education), age (age 30 - 39), working status and occupational status, marriage type (polygyny) and the place of delivery. Though the relationship between the other independent variables and the dependent variable is in the expected direction, yet they are not significant statistically.

It can be deduced from the result that relative to mothers whose husbands are living together, children born to mothers without husbands at home have a decreased probability of breast-feeding for six months. This indicates that if the husband is to be at home they will breast-feed for a shorter time than when he is not around. Employment status shows that children of mothers who worked before being married are likely to be breast-feeding for six months than otherwise. Mothers who were attended to by nurses and midwives during delivery have the greater probability to breast-feed for six months than the reference category. The various factors are significant at different levels of 5 % and 10 %.

Place of delivery is an important factor for durations of nine and twelve months. The result shows that children delivered in both private traditional birth homes have higher probability of breast-feeding for duration of 9 and 12 months than those who delivered in the government hospitals. This raises curiosity to call for the need to investigate the overall role of health services in the determination of breast-feeding duration and practice. Polygnous marriages have negative effects on both higher durations.

The relationship is also examined further controlling for place of residence. The result of the logistic analysis shows that the predictor factors in the rural place for duration of six months are education (tertiary), marital separation, location of work, higher number of children living, antenatal care services, place of delivery, desire for more children. There is no single statistically significant factor in duration of nine months, though they are in the expected direction. The partial correlation (R^o) values also support this, though their effects are small in the model. Also the determinants for breast-feeding for 12 or more months in the rural area are education, working status, employment status, place of delivery.

For the urban area, the predictor variables for six month breast-feeding duration include education, employment status, children ever born, delivery assistants. Those in education levels or primary, secondary and tertiary in the urban area are less likely to breast-feed for six months than those without education. In other words, those with no education will breast-feed for more than six months. Working status and delivery assistants have positive probability of breast-feeding for six months than those in the reference categories. In the urban segment, there are only six determinants for the nine months duration. These are employment status, children ever born, sex of the child, place of delivery, postnatal care and desire for more children.

This paper examines the factors that influence the decision to breast-feed and to breast-feed for particular peak durations of six, nine and twelve months. The analysis clearly identified that the best predictors of the decision of whether or not are education, age, working and occupational status, sex of the child, place of delivery and delivery assistants. The higher the level of education, the higher the probability of a mother breast-feeding the child. This has the reverse effect in the rural area where it is shown that education decreases the probability of ever breast-feeding the child.

In all the models, the effect of modernization variables and health care variables are noted. The result of our analysis shows that each of the factors identified exerts an independent effect on the duration of breast-feeding. The consistency of the health factors demand explanation. It has been explained that women who receive prenatal care from physicians and trained nurses and or experience labour and delivery in a hospital are less likely to initiate breast-feeding and more likely to terminate breast-feeding early. This may be because the nurses or hospital personnel do no explain to the nursing mothers the role of breast-feeding and for a longer time in the health status of their children. This finding has indeed supported an earlier one (Odimegwu, 1994).

This analysis presents supporting evidence of the effect of modernization (proxied by education and employment) on breast-feeding behaviour and duration. This tends to mean that with the increase in the level of female education and with it urban migration and employment, there is bound to be further decline in the duration of breast-feeding. So while an increase in female education and employment is a good indicator of economic development, they discourage women from upholding the tradition of breast-feeding. This implies that in the design and implementation of government policies, there should be a safeguard against undesirable consequences of otherwise positively valued programmes.

Because education is a single most important factor that influences the duration of breast-feeding, an increase in the number of females educated will continue to be a threat to the level of breast-feeding practice in the area. If the government policy-makers believe that breast-feeding has important implications for the health of the infants, efforts should be made to develop policies that will inform women especially the educated and working women about the benefits of breast-feeding. Population education should include increasing the social awareness of the benefits of breast-feeding.

The effect of employment of breast-feeding is also noted. This calls for the need to make policies that will encourage longer maternity leave for nursing mothers who are in the public sector. There is need also to develop cottage industries that will employ the women so that they will be nearer home. Alternatively, government should establish Day Care Centres and Creches very close to ministries and public places so that the babies can be kept in these places for the mothers to breast-feed them at work. For the working mothers, it is better to give longer maternity leave until supplementary feeding of non-milk products is advisable. While the immediate target of policies should be the home, the scope of these policies should be broadened to include other segments of the female population.

In the case of the role of health care personnel, studies should be conducted to fully investigate how delivery place and delivery assistants affect the duration of breast-feeding and the decision to breast-feed. The health sector occupies an important place in exclusive breast-feeding campaign. The health care personnel should be involved in the campaign. The current Baby Friendly Hospital Initiative is a good one and the extent of the support it is receiving from the hospitals should be investigated. The principal goal of the initiative is to mobilize health care systems and health workers to promote and support breast-feeding and to create a demand by women for hospitals that encourage breast-feeding. The WHO and UNICEF with the support of world leaders, health experts and NGOs undertook the Initiative to convince hospitals, health services and parents that breast-feeding gives babies the best possible start in life. The various women-in-development programmes in the country should incorporate aspects of breast-feeding. There should be an alternative to breast-feeding.

^["] I am very grateful to the Council for the Development of Social Research in Africa (CODESRIA), Dakar, Senegal for the grant that supported the project from which the data was extracted. I am also grateful to the reviewers who contributed to the improvement of the quality of the paper.