This was a cross-sectional analytical study conducted over a 6-month period from 1 January 2022 to 30 June 2022. It took place in the consultation and weighing-vaccination units of the paediatrics department of the maternal and child protection centre (PMI) in the town of Cocody in the urban area of Côte d’Ivoire. The Cocody PMI is the closest public health facility to the Cocody hospital and university centre in Abidjan, the economic capital of Côte d’Ivoire. It provides curative, preventive, promotional and educational services for the health of mothers and children.

Our study population consisted of mother and child pairs admitted to the PMI paediatric units during the study period. Mothers who had newborns or infants under 6 months of age and who were breastfeeding, regardless of the type of breastfeeding, were included. Mothers of infants who were on replacement feeding were not included in the study.

Mothers were selected in order of arrival at the facility throughout the survey.

The size n of our sample was calculated using the Schwartz formula with

$n={\epsilon }^2\times pq/i^2$

ε = reduced risk gap α = 5%, is 1.96

p = prevalence of exclusive breastfeeding at national level, p = 0.23 according to UNICEF’s State of the World’s Children’s Health 2021 [9]

$q=1-p$

i = precision with which we wished to estimate the prevalence of exclusive breastfeeding is 5%.

Our minimum sample size was therefore 272 mothers and their newborns or infants.

The data were collected using a pre-established and pre-tested questionnaire administered to the mothers after the consultations. The data collection technique was a direct face-to-face interview with each respondent. Data from the mother-child health record was used to complete the information on the mother-child pair. An interpreter was used to collect responses from mothers who did not understand French.

The dependent variable was the practice of AME, i.e. feeding the child only with breast milk with no addition of water, decoction, fruit juice, milk, herbal tea or other foods with the exception of prescribed medication [6] . The independent variables were the mothers’ socio-demographic data (place of residence, age, level of education, occupation, religion, marital status), the infants’ clinical data (age, sex, birth weight, Apgar score at birth), data on access to healthcare and maternal behaviour (parity, inter-genital space, number of antenatal consultations (ANC), pregnancy monitoring staff, advice on the AME, decision to breastfeed before delivery, delivery facility, route of delivery, mothers’ knowledge of the AME).

The choice of these variables was based on the exhaustive review of the literature in West Africa carried out by Alive & Thrive and UNICEF [14] . In this study, we developed determinants at the individual and contextual levels.

Data was entered using Excel and analysed using SPSS version 23. Data analysis comprised a descriptive phase and an analytical phase.

In the descriptive phase, quantitative variables were described by means, standard deviation and extremes. The analysis of quantitative variables was completed by proportions for qualitative variables.

The variable exclusive breastfeeding practice was grouped into 2 classes coded yes for the dependent variable (AME) and no for the other breastfeeding practices reported by the mothers (breastfeeding with water, breastfeeding combined with formula, breastfeeding combined with foods other than water or formula).

Regression analyses were then performed to explore the association between the dependent variable (exclusive breastfeeding) and the independent variables. These variables were analysed in three phases:

Authorisation for the study was obtained from the PMI Cocody health authorities before the start of the study. Participation in the study was free and voluntary. Informed consent was obtained from the mothers before the start of the study after explanation of the protocol and the purpose of the study. Data were collected anonymously and confidentially.

A total of 308 mothers and their newborns or infants were enrolled in our study. The average age of the mothers was 31 years, with extremes of 17 and 45 years. More than half (52.92%) were aged between 30 and 39, and 4.2% were teenagers.

More than half the mothers were Christian (58.43%). Most of the mothers (89.29%) had attended school, and more than half (54.22%) had completed higher education. The majority of respondents lived with their spouse (90.6%) and 72.7% had an income-generating activity. The study found that 12.7% were students and 14.0% were housewives. The main characteristics of the mothers are shown in Table 1 .

The infants had a mean age of 3.1 +/− 0.05 months. There were 11.09% newborns and almost half of the infants were aged between 2 and 3 months (49.04). The sex ratio was 0.91. The majority of infants were eutrophic at birth (83.77%) and three quarters of them had a good APGAR score (76.62%). In 5.51% of cases, the APGAR score was not recorded in the diary. The characteristics of newborns and infants are shown in Table 2 .

NP*: not specified in the booklet.

The majority of mothers were poor and had given birth 2 or 3 times (67.20%). There was an average inter-generational space of 4.88 years between the last 2 children, with extremes of 11 months and 20 years. The mothers had an average of 6 antenatal visits, ranging from 0 to 12. The pregnancy was monitored by a midwife in the majority of cases (74.67%). Birth was by vaginal delivery in 58.44% of cases, and in a private health facility in half the cases (50.33%). Level 1, 2 or 3 publics health facilities were used by 49.35% of mothers to give birth.

More than half the mothers had received advice on AME (64.61%) and had decided to practise it before giving birth (67.53%). The respondents knew the definition of AME (84.74%) and its duration (96.42%) but practised AME in about ⅓ of cases, i.e. 36.4%. Breastmilk substitutes were used in 28.57% of cases and 16.53% of infants received other foods in addition to breastmilk in 14.9% of cases. Local cereal porridges were offered to 17.7% of children.

The characteristics of mothers regarding access to healthcare and their behaviour are shown in Table 3 .

ANC*: Antenatal care.

The overall breastfeeding rate was 36.4% ( Table 3 ). Less than half the mothers (42.53%) had started AME at birth. The AME rate was highest between 1 and 2 months (48.15%) and lowest between 5 and 6 months (21.73%). The EBF rate by age is shown in Figure 1 .

Bivariate analysis

The bivariate analysis revealed a statistically significant association between the EBF and the following maternal factors: place of residence, with 54.54% of mothers living outside the municipality of Cocody breastfeeding exclusively compared with 34.18% of mothers living in the municipality of Cocody (p = 0.0215), maternal age, with 67.25% of mothers aged 20 to 29 years breastfeeding exclusively compared with 46.15% of adolescents, 32, 51% of mothers aged 30 to 39 years and 37.5% of mothers aged 40 and over years (p < 0.0001), the level of education, with 51.51% of mothers with no education breastfeeding, 41.46% of mothers with primary education, 50.74% of mothers with secondary education and 26.34% of mothers with higher education (p = 0.0007), the occupation, with 48.89% of housewives having an EBF, 49.45% of informal sector workers having an EBF, 30.77% of schoolgirls and students having an EBF and 33.0% of civil servants and private sector employees having an EBF (p = 0.0004), the delivery facility, with 42, 38% EBF among mothers who had given birth in a public facility compared with 30.77% EBF among mothers who had given birth in a private health facility (p = 0.0312), and the mode of delivery, with 44.50% EBF in the case of vaginal delivery compared with 24.22% EBF among mothers who had undergone caesarean section (p = 0.0002). The decision to breastfeed prior to delivery was also a factor associated with EBF, with 47.32% EBF for mothers who had made this decision before giving birth, compared with 15.0% for those who had not (p < 0.0001).

The results of the bivariate analysis are shown in Table 4 .

In the multivariate analysis, the variables significantly associated with the practice of EBF were the mothers’ level of education with AOR = 2.48, 95%IC [1.04 - 5.93] in mothers with no education, OR = 2.06, IC95% [1.03 - 4.15] in mothers with secondary education, the decision to breastfeed prior to delivery with OR = 3.82, IC95% [1.94 - 7.49], vaginal delivery with OR = 2.33, IC95% [1.29 - 4.18]. The risk factors for not performing EBF were hypotrophy in the child (AOR = 0.27, CI95% [0.10 - 0.73]), higher education in the mother (AOR = 0.68, CI95% [0.6 - 0.92]) and whether the mother worked in the public (civil servant) or private sector (AOR = 0.361, CI95% [0.177 - 0.735]). The main results of the multivariate analysis are shown in Table 5 .

There are a number of limitations to this study which must be taken into account when interpreting the results.

The first is due to the difficulties in collecting responses to the questionnaire from mothers who did not attend school and were of foreign nationality, for whom we sometimes had to use interpreters, which could have an impact on the responses obtained. However, the number of these mothers remains negligible. The second limitation is linked to the fact that although the study was carried out in the city of Abidjan, the results are mainly valid only in the context of the commune of Cocody, which has its own realities. Thirdly, this is a self-reported survey, which presents limitations for establishing causal conclusions and is subject to recall and social desirability biases. Another constraint is the limitation of the study to two groups of determinants, thus neglecting national policies and social trends that could have been crucial for understanding mothers’ choices regarding breastfeeding.

Breast milk is the optimal food for newborns and young infants. It has no equal because it is tailored to their needs and abilities [19] . It has many advantages for both mother and baby [1] - [3] [19] [20] .

However, to benefit from these advantages, it is recommended that infants be exclusively breastfed for at least the first six months of life [5] [6] [20] . Our work shows that not all the children of mothers in our study area benefit from this preferential feeding method. The overall rate of EBF was 36.4%, more than a third of children. This prevalence is lower than that reported in the Mékébo study in Ethiopia (83%) [21] or the national prevalence of AME in Rwanda (81%) [9] . Nevertheless, it is similar to that of Ricci in Canada (35.6%) [22] and to the national prevalence in Côte d’Ivoire in 2021 (34%) and very close to the results of the study by Traoré in Mali (30.66%) [3] and Coulibaly in Côte d’Ivoire (33.51%) [20] . The prevalence of AME in the world varies from one country to another and within the same country from one region to another. A study carried out in 13 West African countries reported high rates of over 50% in Togo, Liberia and Ghana, while Côte d’Ivoire and Nigeria had the lowest rates, below 20% [23] .

However, the overall rate of EBF in our work attests to WHO data indicating that over the past decade countries as diverse as Côte d’Ivoire, Marshall Islands, Philippines, Somalia and Viet Nam have made considerable progress, demonstrating the effectiveness of initiatives to protect, promote and support breastfeeding [24] .

Sharing experiences from areas with high EBF prevalence can help EBF stakeholders in low-prevalence areas to identify and implement high-impact activities on EBF practice at their level.

In our study, a bivariate analysis revealed a statistically significant association between EBF and the following maternal factors: place of residence (p = 0.0215), age (p < 0.0001), level of education (p = 0.0007), occupation (p = 0.0004), delivery place (p = 0.0312), method of delivery (p = 0.0002) and decision to breastfeed before delivery (p < 0.0001).

Although no longer a significant determinant in multivariate analysis, place of residence was found in Coulibaly’s study to be a determining factor in the practice of breastfeeding among primiparous mothers selected in 3 different areas of the city of Abidjan, including an economically weak peri-urban area (Akéikoi urban health facility), an economically strong urban area to the north of Abidjan (Cocody PMI) and an economically average urban area to the south of Abidjan (Marcory General Hospital), an economically strong urban area in the north of Abidjan (Cocody PMI) and a medium-economic urban area in the south of Abidjan (Marcory General Hospital). In this study, primiparous mothers from the Cocody PMI, as in our study, practised exclusive breastfeeding less than the other respondents, i.e. 20.93% compared with 38.78% and 40.06%. According to the author, this could be justified by the fact that Cocody is a residential community where most managers employed in the public or private sector live. Their working environment is not conducive to breastfeeding [20] . Efforts to promote breastfeeding in our study area must take into account this particular feature of the place of residence.

As with place of residence, maternal age and delivery structure, which were factors significantly associated with EBF in bivariate analysis, were no longer so in multivariate analysis. The influence of the type of health structure on EBF has been noted by other authors. According to Ávila-Ortiz in Mexico, who reported 28% of EBF at 6 months among mothers from 2 private hospitals, women with a high level of education and actively participating in formal employment with economic autonomy, mainly use private medical services, which adhere less to the Baby-Friendly Hospital Initiative (BFHI) and where information on breastfeeding is limited [25] . These conditions not conducive to AME reported by Ávila-Ortiz are found within our study population, which resides in a municipality identified as economically strong, with 50.33% of mothers having given birth in a private health facility. Promotional activities of the Baby-Friendly Hospital Initiative (BFHI) in private health centers can help improve the EBF rate in our study area.

In Coulibaly’s work [20] , mothers who gave birth in private facilities also tended to practice EBF less and this would be due to the good socio-economic level which gave them the possibility of easy access to artificial milk and the marginalization of private sector agents from capacity building activities initiated by the Ministry of Health [20] . In order to address this situation, awareness campaigns and nutritional education focusing mainly on the practice of EBF should be carried out in private facilities and reinforced in public facilities in the city of Abidjan.

The determinants positively associated with the practice of EBF after bivariate and then multivariate analysis were the lack of schooling and the level of secondary education of the mothers with respectively 2.48 times and 2.06 times more chance of practicing EBF. Mothers who gave birth vaginally had 2.33 times more chance of practicing EBF than those who gave birth by caesarean section which is in line with the general trend in West Africa as reported in a study where 96.2% of mothers who did not have a caesarean section breastfed exclusively compared to 3.5% of mothers who gave birth by caesarean section [23] .

The decision to breastfeed when taken before delivery was associated with 3.82 times greater chance of practicing EBF among mothers. This finding was reported by Nabunya in Uganda (1.26 times) [26] , Ricci in Canada (2.5 times) [22] . Similarly, a report on EBF in West Africa reports that in Ghana, women who planned to breastfeed exclusively at the time of pregnancy were approximately 2.5 times more likely to exclusively breastfeed their infants, compared to mothers who did not plan to do so [14] , hence the interest in intensifying EBF promotion activities among young girls, women of childbearing age and pregnant women during ANC.

The risk factors for not practicing EBF were the mother’s higher level of schooling and the profession of civil servant or private sector employee with respectively 32% and 64% less chance of practicing EBF while hypotrophy in the child was associated with 73% less chance of being exclusively breastfed for 6 months. Our results on the mother’s profession as a determinant of EBF are reported by many authors [20] [23] [25] [27] .

As a palliative measure in Ivory Coast, the government adopted Decree No. 2025-120 of February 26, 2025, which increases maternity leave from 14 weeks (3 months) to 24 weeks, (6 months), including a prenatal period of 8 weeks and a postnatal period of 16 weeks [28] . This decision, which only concerns the civil service, could help increase the AME rate for our mothers as well as the prevalence at the national level in order to bring the country within the vision of 70% EBF by 2030 recommended by the WHO.