https://orcid.org/0000-0002-6756-0104
Abstract
Inconsistent results have been reported regarding the prevalence of and factors associated with formula feeding in Ethiopia.
This study aimed to determine the pooled prevalence of and factors associated with formula feeding among mothers with infants 0–6 months of age in Ethiopia.
A comprehensive systematic search was conducted across 3 databases (PubMed, EMBASE, and ScienceDirect) and the Google Scholar search engine to identify relevant studies published up to April 2, 2024.
After assessing the quality of each study using the Joanna Briggs Institute Critical Appraisal Checklist for Prevalence Studies, data were independently extracted by 2 authors using pre-designed forms in an Excel spreadsheet. Any disagreements were resolved through discussion and consultation with additional authors.
Statistical heterogeneity across studies was assessed using the I2 statistic. A random-effects meta-analysis was used to pool the proportions, due to high heterogeneity, while a fixed-effect meta-analysis was used to pool associated factors with low heterogeneity. A sensitivity analysis was performed to assess the potential impact of outlier studies on the overall estimates. Five studies, with a combined sample size of 2344 participants, were included. The pooled prevalence of formula feeding was 34.0% (95% CI: 23.0%, 44.0%). Factors significantly associated with formula feeding included cesarean delivery (POR [pooled odds ratio] = 4.72, 95% CI: 3.32, 6.71) compared with vaginal delivery, a positive attitude toward formula feeding (POR = 2.26, 95% CI: 1.45, 3.53) compared with a negative attitude, initiation of breastfeeding more than 1 hour after delivery (POR = 2.27, 95% CI: 1.25, 4.13) compared with initiation within 1 hour, and receiving information about formula feeding from friends or family (POR = 2.47, 95% CI: 1.46, 4.20) compared with receiving information from health-care professionals.
The prevalence of formula feeding in Ethiopia is significant. Cesarean delivery, a positive attitude toward formula feeding, late initiation of breastfeeding, and receiving formula feeding-related information from friends and family were positively associated with formula feeding. Given these findings, the authors recommend that public health interventions in Ethiopia target these key determinants to decrease the high prevalence of formula feeding practices observed in the country. Strategies addressing factors such as promoting vaginal delivery, improving attitudes toward breastfeeding, encouraging early breastfeeding initiation, and limiting the influence of social networks on formula-feeding decisions may be beneficial.
BACKGROUND
Breast milk is an important source of nourishment and immunological protection for infants.1 Particularly in the first 6 months, only breast milk can provide infants with the perfect blend of nutrients, immunological factors, and other bioactive compounds necessary for achieving a rapid and sustained rate of growth and development.1–3 The World Health Organization and the United Nations Children’s Fund recommend exclusive breastfeeding for the first 6 months, followed by the introduction of complementary foods while continuing breastfeeding until the child is 2 years of age.4 Nevertheless, the use of formula milk is noticeably increasing in both developed and developing countries.5,6
Infant formula is a type of breast milk substitute, industrially produced to serve as an alternative to breast milk for infants under 6 months of age.7,8 It is typically a product of cow’s milk modification, which involves removing and diluting the excess protein, fat, and minerals, while other components such as vegetable oils, vitamins, minerals, and iron are added to compensate for the deficient and absent nutrients.9
It is undeniable that, over time, various forms of breast milk substitutes have progressively evolved to closely resemble human breast milk. Regardless, at this time, infant formula cannot replicate the full array of benefits provided by human breast milk. This is because infant formula lacks the immune cells, bioactive proteins, and enzymes necessary to ward off infections, as well as the dynamic nature observed in breast milk, which is crucial for meeting the frequently changing nutritional needs of children. It also lacks the probiotics and prebiotic oligosaccharides essential for maintaining a normal gut biome.9,10
In addition, formula feeding exposes infants to harmful pathogens, increasing their risk of contracting infections such as otitis media, diarrhea, and pneumonia, potentially leading to hospitalization.11,12 It also incurs additional costs for mothers and their families and harms the environment.13,14 Furthermore, formula feeding deprives mothers and their children of the many benefits of optimal breastfeeding, which include a lower risk of infectious and chronic diseases, protection against ovarian and breast cancer, appropriate birth spacing, improved intelligent quotient (IQ) and school performance, and a strong mother–child bond.14–16
Despite the growing prevalence of formula feeding globally, as well as efforts by the World Health Organization to address this issue; such as the introduction of the International Code of Marketing of Breast-milk Substitutes in 1981, the world sales of formula milk have continued to rise significantly. Between 2005 and 2019, global formula milk sales grew by 121.5%, indicating a persistent and increasing dependence on breast milk substitutes worldwide.5,7 Currently, the market for infant formula is estimated to be worth approximately 50 billion U.S. dollars, which is also expected to grow by 7% annually over the next 5 years, and in line with this growth, the number of children being fed formula is expected to increase.9
The reasons for the rise in the level of formula feeding are complex. However, important factors such as changes in birth rates, income growth, urbanization, shifts in women’s employment, the availability and price of breast milk substitutes, exposure to breast milk substitute marketing, and policies and regulations regarding breast milk substitutes are some of the factors that are believed to play significant roles.5,15
In 2018, one in 10 infants under 6 months of age was fed infant formula worldwide. This prevalence was significantly higher in Latin America and the Caribbean, where 37% of infants were formula-fed, making it the region with the highest global prevalence of infant formula feeding.6 In addition, a review that analyzed data from 85 countries revealed that a quarter of infants in low- and middle-income countries consumed infant formula between the years 2010 and 2019, and the highest and lowest percentages were recorded in Gabon (63.5%) and Burkina Faso (0.8%), respectively.17 Likewise, in Eastern and Southern Africa, 4% of infants received infant formula in 2018.6
In Ethiopia, approximately 3% of children under 2 years received infant formula in 2019; compared with the 2016 level, which was 1.4%, this indicates a significant increase.18,19 In addition, nearly 10% of children in the country received mixed feed (formula and/or animal milk in addition to breast milk) in 2019.19 Recognizing the alarming increase in the rate of formula feeding, the government introduced the Food Advertisement Directive and the Infant Formula and Follow-up Formula Directive in 2016.8 The Ethiopian Infant Formula and Follow-up Formula Directive regulate the marketing and composition of infant formulas, aiming to promote breastfeeding by restricting formula advertisements. It ensures that formula products meet safety and nutritional standards while limiting promotional practices that could undermine breastfeeding. The Food Advertisement Directive sets out regulations to prevent misleading claims in food marketing, especially those targeting children. It aims to encourage healthier eating habits and ensure that food advertisements provide accurate information, protecting consumers from deceptive marketing practices. Both directives collectively support public health by promoting accurate information and supporting breastfeeding.20
However, recent studies conducted in Mettu, Bahir Dar, Addis Ababa, Dire Dawa, and Jimma found a prevalence of infant formula feeding ranging from 21.4% to 47.2%, indicating infant formula feeding is still prevalent in the country. In addition, maternal age, educational status, occupation, number of births, mode of delivery, antenatal care utilization, pre-lacteal feeding, breastfeeding initiation time, and knowledge of and attitudes toward formula feeding were identified as important predictors of infant formula feeding in the country.21–25
Multiple primary studies have investigated the prevalence of formula-feeding practices and possible associated factors. However, the reported prevalence and contributing factors vary considerably across different regions and localities within the country. This variation in prevalence and associated factors may affect the implementation and enforcement of regulations on formula feeding. Therefore, determining the pooled prevalence of and factors associated with formula feeding is essential to addressing the increasing burden of formula feeding. This review aims to fill this gap by providing crucial information to help policy makers, program planners, and communities tackle the growing burden of formula feeding, ultimately benefiting children.
METHODS AND MATERIALS
Study Design and Setting
A comprehensive analysis of relevant research on formula-feeding practices and associated factors among women with children aged 0–6 months in Ethiopia was conducted. The review aimed to include diverse study designs, such as cross-sectional, case–control, and cohort studies. However, due to limitations in the available data, only cross-sectional studies were included. This restriction may impact the ability to draw definitive conclusions about the direction of associations between formula feeding and certain factors, such as attitudes toward formula feeding.
Eligibility Criteria
In this systematic review and meta-analysis, we specifically targeted observational studies conducted in Ethiopia, focusing on formula-feeding practices and associated factors among women with infants aged 0–6 months. We emphasized including studies published in English as part of our criteria, and we have included all relevant studies that were published and accessible online until April 2, 2024. It is worth noting that we did not identify any gray literature during our search. Qualitative studies and those lacking a clear and adequate definition of formula feeding were excluded from our analysis. We used CoCoPop (Condition, Context, and Population) for the prevalence study and PEO (Population, Exposure, and outcome mnemonics for the associated factors) to declare the inclusion criteria (Table 1).
CoCoPop (Condition, Context, and Population) and PEO (Population, Exposure, and Outcome) Criteria Were Used for the Inclusion of Studies
| Parameters | Inclusion criteria | Exclusion criteria |
|---|---|---|
| For the first objective (for the prevalence) | ||
| Condition | Practice of formula feeding among mothers with infants aged 0–6 months, including women of all age groups | Qualitative studies and those that lack an adequate definition of formula feeding were excluded |
| Context | Ethiopia | |
| Population | Mothers with infants aged 0–6 months living in Ethiopia | |
| For the second objective (for associated factors) | ||
| Population | Mothers with infants aged 0–6 months | |
| Comparator | Associated factors or determinates that influence formula feeding practice, which include mode of delivery, maternal education, maternal employment, media exposure, timely initiation of breast feeding and other associated factors | |
| Outcome | Practice of formula feeding | |
| Parameters | Inclusion criteria | Exclusion criteria |
|---|---|---|
| For the first objective (for the prevalence) | ||
| Condition | Practice of formula feeding among mothers with infants aged 0–6 months, including women of all age groups | Qualitative studies and those that lack an adequate definition of formula feeding were excluded |
| Context | Ethiopia | |
| Population | Mothers with infants aged 0–6 months living in Ethiopia | |
| For the second objective (for associated factors) | ||
| Population | Mothers with infants aged 0–6 months | |
| Comparator | Associated factors or determinates that influence formula feeding practice, which include mode of delivery, maternal education, maternal employment, media exposure, timely initiation of breast feeding and other associated factors | |
| Outcome | Practice of formula feeding | |
CoCoPop (Condition, Context, and Population) and PEO (Population, Exposure, and Outcome) Criteria Were Used for the Inclusion of Studies
| Parameters | Inclusion criteria | Exclusion criteria |
|---|---|---|
| For the first objective (for the prevalence) | ||
| Condition | Practice of formula feeding among mothers with infants aged 0–6 months, including women of all age groups | Qualitative studies and those that lack an adequate definition of formula feeding were excluded |
| Context | Ethiopia | |
| Population | Mothers with infants aged 0–6 months living in Ethiopia | |
| For the second objective (for associated factors) | ||
| Population | Mothers with infants aged 0–6 months | |
| Comparator | Associated factors or determinates that influence formula feeding practice, which include mode of delivery, maternal education, maternal employment, media exposure, timely initiation of breast feeding and other associated factors | |
| Outcome | Practice of formula feeding | |
| Parameters | Inclusion criteria | Exclusion criteria |
|---|---|---|
| For the first objective (for the prevalence) | ||
| Condition | Practice of formula feeding among mothers with infants aged 0–6 months, including women of all age groups | Qualitative studies and those that lack an adequate definition of formula feeding were excluded |
| Context | Ethiopia | |
| Population | Mothers with infants aged 0–6 months living in Ethiopia | |
| For the second objective (for associated factors) | ||
| Population | Mothers with infants aged 0–6 months | |
| Comparator | Associated factors or determinates that influence formula feeding practice, which include mode of delivery, maternal education, maternal employment, media exposure, timely initiation of breast feeding and other associated factors | |
| Outcome | Practice of formula feeding | |
Operational Definition
Formula-feeding practice was defined as feeding an infant under 6 months old with formula using a rubber nipple, cup, spoon, or other materials as a substitute for or supplement to breastfeeding. This practice was assessed with a “Yes” (1) or “No” (2) question. Mothers who answered “Yes” were classified as those who fed their child using formula.24
Late initiation of breast feeding was defined as starting breastfeeding more than an hour after birth.26
Search Strategies
All search strategies were independently reviewed by 2 authors (Z.A., B.E.). In the case of disagreement between these authors, a third (A.D.) and if necessary a fourth (W.M.) author was consulted. We attempted to apply a snowball search strategy by reviewing the references of the identified studies, but this did not yield any additional eligible studies for inclusion. The results from the database searches were exported to Endnote. To check the duplication of the research question an electronic search was performed using the following keywords: (formula feeding OR infant formula OR baby formula) AND (determinant* OR “associated factor*” OR predictor* Or “risk factor*”) AND (systemic review OR meta-analysis]) AND (Ethiopia) in Prospero, Epistemonikos, and the Cochrane library database. The findings of this systematic review were presented as per the guidelines provided by the Preferred Reporting Items for Systematic Reviews and Meta-Analysis, 2020. A thorough search was conducted across multiple electronic databases, including PubMed, EMBASE, Science Direct, and Google Scholar, to identify pertinent published studies. Furthermore, we explored unpublished papers on the official websites of Addis Ababa University and conducted preprint searches on medRxiv and bioRxiv. However, no gray literature was discovered. The search terms and keywords employed encompassed various variations of “formula feeding,” “infant formula,” and “baby formula,” combined with terms like “determinant,” “associated factors,” “predictor,” “risk factors,” and “Ethiopia.” The specific search strings used in PubMed, EMBASE, and Science Direct are provided in the Supplementary File S1.
Study Selection
The articles obtained from the various sources were imported into the citation manager EndNote X8, and duplicate studies were removed. Two authors (B.E. and Z.A.) independently reviewed the titles and abstracts of the studies based on predefined inclusion criteria. The same authors (B.E. and Z.A.) obtained and evaluated the full texts of the selected studies to determine their eligibility for final inclusion, considering factors such as study design, quality, and outcomes. In cases where disagreements arose between the authors, a third author (A.D.) and if necessary a fourth author (W.M.) acted as mediators. The selection process for the inclusion of articles in this systematic review and meta-analysis was visually presented using a Preferred Reporting Items for Systematic Reviews and Meta-Analysis flowchart.
Data Extraction Process
The process of extracting data from each included study was carried out independently by 2 authors (A.D., W.M.) using pre-designed forms prepared in an Excel spreadsheet. In the event of any disagreements during the extraction, they were resolved through discussion, and additional authors (Z.A., B.E.) were consulted for their input whenever necessary. The extracted information from each study included various details such as the author, year of publication, participant characteristics (eg, sample size, region, mode of delivery, attitude towards formula feeding, time of breastfeeding initiation, source of information about formula feeding), response rate, proportion of formula feeding practice, and the type of study design. Since all papers were available online, no effort was made to obtain them by contacting the respective authors via email or phone.
Risk of Bias Assessment
To assess the quality of the included studies, the Joanna Briggs Institute Critical Appraisal Checklist for Prevalence Studies was employed. This checklist includes 9 items that evaluate methodological quality, covering aspects such as the appropriateness of the sample frame, sample size adequacy, the method of analysis, and the response rate. Each item was rated with responses of “yes,” “no,” “unclear,” or “not applicable”.27 Studies were included if they reported a “yes” for at least 5 of the 9 items. All included studies met this criterion, with more than 5 “yes” responses. The assessment was performed independently by 2 reviewers, M.A. and W.T. Any disagreements were resolved through consultation with additional authors G.D. and F.Z (Supplementary File S2).
Data Management and Analysis
The relevant data were extracted from the source using a standardized format, created in Microsoft Excel, to ensure consistency in the data organization. After extraction and organization in Excel, the data were then exported to STATA version 16.0 for further analysis. The findings were presented using text, tables, and a forest plot. Proportions were combined through random-effects meta-analysis, and adjusted odds ratios (AORs) were considered for associated factors. For factors reported in multiple studies, the AORs were pooled to calculate a pooled odds ratio (POR), each with a corresponding 95% CI (CI). Statistical heterogeneity across studies was assessed using the I2 statistic, which quantifies the variability in results between studies.
The I2 values were classified into 4 groups: low (0–25%), moderate (26–50%), high (51–75%), and very high (76–100%), based on predetermined cut-offs.28 For studies with small I2 or insignificant Q-test values, indicating low or no statistical heterogeneity, a fixed-effects meta-analysis was conducted. However, for studies with a high I2 value or significant Q-test, indicating substantial heterogeneity, a DerSimonian and Laird random-effects meta-analysis model was used. Due to the limited number of studies included, subgroup analysis and publication bias assessments were not performed. A sensitivity analysis was performed to assess the potential impact of outlier studies on the overall estimation. Sensitivity analysis helps determine whether the results of a single study significantly influence the overall findings. If the point estimates fall within the 95% CI of the combined analysis, it indicates that there is no outlier or influential study.
RESULTS
Identification of Studies
This review involved the identification of a total of 10 613 studies from various electronic databases. Specifically, the studies were sourced from Pubmed (8173), EMBASE Research for Life (1583), Science Direct (817), and the Google Scholar search engine (40). During the study selection process, 76 duplicate studies were identified and removed. Additionally, 10 550 studies were excluded based on their titles, and an additional 23 studies were excluded based on their abstracts. Following these initial screening steps, a comprehensive evaluation of the full texts of 11 papers was conducted. Applying the predefined inclusion and exclusion criteria, 6 studies were excluded during this assessment, while 5 studies were considered suitable and included in the systematic review and meta-analysis (Figure 1).
A PRISMA Flow Diagram of Included Studies to Estimate the Pooled Prevalence and Factors Associated With Formula Feeding Among Mothers With Children Under the Age of 5 in Ethiopia
Characteristics of Included Studies
A total of 5 studies that met the eligibility criteria were included in this review. These studies were published between 2018 and 2023.22–25,29 Among them, 2 studies were conducted in the Oromia region, while 1 study each originated from the Amhara region, Addis Ababa, and Dire Dawa. The quality of all the included studies was assessed using the Joanna Briggs Institute quality assessment checklist, with quality scores ranging from 6 to 7 (Supplementary File S2). In total, 2344 participants were involved in these studies, with sample sizes ranging from 18625 to 71424 across the studies. The prevalence of formula feeding practice varied among the studies, ranging from 21.5%25 in Dire Dawa to 47.2% in Jimma town24 (Table 2).
Characteristics of the Studies Included in the Systematic Review and Meta-analysis of Infant Formula Feeding Among Mothers With Children Under 6 Months of Age in Ethiopia
| First author (year of publication) | Study design | Quality of article | Region | Sample size | Response rate n (%) | Prevalence of formula feeding n (%) |
|---|---|---|---|---|---|---|
| Alemu et al (2023) | Cross-sec | Good | Amhara | 630 | 593 (94.12) | 148 (25) |
| Taye et al (2021) | Cross-sec | Good | Addis Ababa | 500 | 494 (98.8) | 228 (46.2) |
| Kera et al (2023) | Cross-sec | Good | Oromia | 366 | 352 (96.3) | 97 (28.4) |
| Abebe et al (2019) | Cross-sec | Good | Oromia | 714 | 705 (98.7) | 333 (47.2) |
| Debebe et al (2018) | Cross-sec | Good | Dire Dawa | 186 | 186 (100) | 40 (21.5) |
| First author (year of publication) | Study design | Quality of article | Region | Sample size | Response rate n (%) | Prevalence of formula feeding n (%) |
|---|---|---|---|---|---|---|
| Alemu et al (2023) | Cross-sec | Good | Amhara | 630 | 593 (94.12) | 148 (25) |
| Taye et al (2021) | Cross-sec | Good | Addis Ababa | 500 | 494 (98.8) | 228 (46.2) |
| Kera et al (2023) | Cross-sec | Good | Oromia | 366 | 352 (96.3) | 97 (28.4) |
| Abebe et al (2019) | Cross-sec | Good | Oromia | 714 | 705 (98.7) | 333 (47.2) |
| Debebe et al (2018) | Cross-sec | Good | Dire Dawa | 186 | 186 (100) | 40 (21.5) |
Characteristics of the Studies Included in the Systematic Review and Meta-analysis of Infant Formula Feeding Among Mothers With Children Under 6 Months of Age in Ethiopia
| First author (year of publication) | Study design | Quality of article | Region | Sample size | Response rate n (%) | Prevalence of formula feeding n (%) |
|---|---|---|---|---|---|---|
| Alemu et al (2023) | Cross-sec | Good | Amhara | 630 | 593 (94.12) | 148 (25) |
| Taye et al (2021) | Cross-sec | Good | Addis Ababa | 500 | 494 (98.8) | 228 (46.2) |
| Kera et al (2023) | Cross-sec | Good | Oromia | 366 | 352 (96.3) | 97 (28.4) |
| Abebe et al (2019) | Cross-sec | Good | Oromia | 714 | 705 (98.7) | 333 (47.2) |
| Debebe et al (2018) | Cross-sec | Good | Dire Dawa | 186 | 186 (100) | 40 (21.5) |
| First author (year of publication) | Study design | Quality of article | Region | Sample size | Response rate n (%) | Prevalence of formula feeding n (%) |
|---|---|---|---|---|---|---|
| Alemu et al (2023) | Cross-sec | Good | Amhara | 630 | 593 (94.12) | 148 (25) |
| Taye et al (2021) | Cross-sec | Good | Addis Ababa | 500 | 494 (98.8) | 228 (46.2) |
| Kera et al (2023) | Cross-sec | Good | Oromia | 366 | 352 (96.3) | 97 (28.4) |
| Abebe et al (2019) | Cross-sec | Good | Oromia | 714 | 705 (98.7) | 333 (47.2) |
| Debebe et al (2018) | Cross-sec | Good | Dire Dawa | 186 | 186 (100) | 40 (21.5) |
Pooled Proportion of Formula Feeding Practice
All the studies included in this review provided data on the proportion of formula feeding practice. The forest plot (Figure 2) displays the pooled proportion of formula feeding, which was determined to be 0.34 (95% CI: 0.230, 0.440). It is noteworthy that there was considerable heterogeneity, as evidenced by I2 values exceeding 75% and a significant Q-test. Consequently, a random-effects meta-analysis was conducted; nevertheless, the heterogeneity persisted. Subsequently, attempts were made to address the heterogeneity through sensitivity analysis, but the issue remained unresolved. To gain further insights, a sensitivity analysis was performed.
Pooled Proportion of Infant Formula Feeding Practice Among Mothers Who Had Children Aged 0–6 Months of Age in Ethiopia
Sensitivity Analysis
A sensitivity analysis was carried out to examine the influence of individual studies on the pooled estimate of formula feeding practice in the random-effect model. Notably, no specific study was found to excessively impact the overall estimation (Figure 3).
Sensitivity Analysis for the Pooled Prevalence of Formula Feeding Practice Among Mothers Who Had Children Less Than 6 Months of Age in Ethiopia
Factors Associated with Formula Feeding Practice
Narrative Analysis
In this review, being within the age range of 25–35,23 being employed in the private or government sector, having 3 or more antenatal care visits,23 social pressure to formula feed, being attended by a traditional birth attendant,24 higher educational status,22,24 awareness about the health effects of formula feeding,21,24 lack of counseling about formula feeding practice during antenatal care follow-up, having a family income above 5000 per month,25 not engaging in pre-lacteal feeding practices,22 being a primiparous mother,21 being employed, cesarean delivery,21,22,25 positive attitude toward formula feeding practice, delayed initiation of breastfeeding,21,23 and receiving information from families or friends23,25 were identified as significant factors associated with formula feeding practice.
Meta-Analysis
In this review, we conducted pooled estimates of adjusted odds ratios for associated factors that were reported by at least 2 studies simultaneously. Specifically, we calculated pooled estimates for cesarean delivery (reported in 3 studies),21,22,25 positive attitude toward formula feeding practice (reported in 2 studies),21,23 delayed initiation of breastfeeding (reported in 2 studies),21,23 and receiving information from families or friends (reported in 2 studies).23,25 Positive association and significant association were observed between cesarean section, a positive attitude toward formula feeding, initiating breastfeeding within 1 hour after delivery, and receiving information about formula feeding from friends or family. The I2 values were within an acceptable range, indicating a lack of substantial heterogeneity. Furthermore, the P-values for the Q-test were not significant, providing further evidence of the absence of significant heterogeneity. As a result, a fixed-effects meta-analysis was conducted.
The meta-analysis findings indicated that women who underwent cesarean section had 4.72 times higher odds of practicing formula feeding compared with other delivery methods (pooled odds ratio [POR] = 4.72, 95% CI: 3.32, 6.71) (Figure 4). Additionally, women with a positive attitude toward formula feeding had 2.26 times higher odds of engaging in formula feeding (POR = 2.26, 95% CI: 1.45, 3.53) (Figure 5). Furthermore, mothers who initiated breastfeeding greater than 1 hour after delivery had 2.27 times higher odds of practicing formula feeding (POR = 2.27, 95% CI: 1.25, 4.13) than those who initiated breastfeeding within the first hour (Figure 6). Lastly, mothers who received information about formula feeding from friends or family had 2.47 times higher odds of clinical breast examination compared with those who obtained information from media sources (POR = 2.47, 95% CI: 1.46, 4.20) (Figure 7).
Forest Plot for the Pooled Adjusted Odds Ratio of Cesarean Delivery Among Mothers Who Had Children Under 6 Months of Age in Ethiopia
Forest Plot Showing the Pooled Adjusted Odds Ratio of Positive Attitude Toward Formula Feeding Practice Among Mothers Who Had Children Aged 0 to 6 Months of Age
Forest Plot That Shows the Pooled Odds Ratio of Initiation of Breastfeeding After 1 Hour After Delivery Among Mothers Who Had Children Less Than 6 Months of Age in Ethiopia
Forest Plot That Shows the Pooled Adjusted Odds Ratio of Receiving Information from Friends or Family Among Mothers Who Had Children Less Than 6 Months of Age in Ethiopia
DISCUSSION
In this review, the objective was to determine the pooled prevalence of formula feeding practice in Ethiopia and examine the factors associated with it. The review included a total of 2344 participants, with sample sizes ranging from 186 to 714 across different studies. The prevalence of formula feeding practice varied among the studies, ranging from 21.5% to 47.2%. The pooled prevalence of formula feeding practice was estimated to be 34.0%, with a 95% CI of 23.00% to 44.00%. Additionally, the study revealed significant associations between formula feeding practice in Ethiopia and factors such as cesarean delivery, a positive attitude towards formula feeding, delayed initiation of breastfeeding, and receiving information from friends and family.
The pooled prevalence of formula feeding in this review is consistent with the findings of a prior study carried out in Egypt.30 However, our result exceeds the global average of 10% reported in 2018,6 and indicates a level of disregard for the World Health Organization and United Nations Children’s Fund recommendations that infants under 6 months be exclusively breastfed.4 Despite this, the pooled prevalence in this analysis is lower than in previous research, such as findings of 88% in China,31 and 65% in Iceland.32 The discrepancy in formula feeding prevalence observed between this meta-analysis and other studies could be attributed to different factors. First, the degree of urbanization may play a role, as urban mothers may be more likely to initiate formula feeding due to greater availability and accessibility of formula products in urban areas, as well as potentially receiving more information about formula feeding options.33,34 Additionally, the mode of delivery could be a contributing factor, as areas with higher rates of cesarean births may also see higher formula feeding prevalence. Finally, differences in cultural attitudes towards and beliefs about formula feeding among the study populations may also help explain the variation in reported prevalence across the different settings.30
Those mothers who delivered by cesarean section had increased odds of formula feeding practice. This is consistent with the studies conducted in Egypt,35 and Indonesia.36 This could be attributed to the stress of the surgical procedure. Giving birth via cesarean section involves major abdominal surgery that creates a significant physiological stress reaction in the mother’s body. This physiologic stress might affect the formation and release of essential hormones involved in breast milk production, including oxytocin and prolactin, and this hormonal imbalance might make it more difficult for mothers to produce enough milk.37 Furthermore, the interruption of critical early skin-to-skin contact between the mother and newborn, which is frequently delayed after a cesarean birth, might impede the start of breastfeeding and increase the likelihood of formula feeding.
Mothers who had a positive attitude toward formula feeding had higher odds of formula feeding than those who had a negative attitude. This is consistent with the findings of the study conducted in Ireland.38 Furthermore, a mother’s favorable attitude towards formula feeding is typically associated with increased access to formula goods and the perception that formula feeding is a convenient option. For these mothers, formula feeding is a more viable infant feeding alternative due to its availability and perceived ease of use. Furthermore, pro-formula mothers may have limited access to breastfeeding education, resources, and social support. Because of a lack of breastfeeding support, many mothers may see formula feeding as an easier or more practical alternative.
Mothers who delay initiating breastfeeding beyond the first hour after birth are more likely to end up formula-feeding their infants. This could be explained by the fact that delayed initiation of breast milk is linked to pre-lacteal feeding, where the newborn receives other liquids of formula feed instead of the mother’s breast milk in those critical early hours. In addition, this disruption to timely breastfeeding establishment may hinder the development of a strong mother–child bond. Moreover, the failure to initiate breastfeeding promptly can also impair the mother’s milk production over time. The combination of pre-lacteal feeding and the detrimental effects on maternal–infant bonding and milk supply, stemming from delayed breastfeeding initiation, might increase the probability that the mother will ultimately resort to formula as the primary feeding method.
Receiving information regarding formula feeding from friends and family increased the likelihood of formula feeding as compared with receiving information from media and health professional sources. This is consistent with the findings of the study conducted in Hong Kong.39 This could be because, when delivering information, relatives and friends might focus on the benefits and convenience of formula feeding, rather than thoroughly outlining the potential health consequences and disadvantages of formula feeding. Health professionals, on the other hand, may provide mothers with a fuller overview of formula use, including both the benefits and risks. This selective framing by personal contacts, emphasizing the benefits of formula over the drawbacks, could contribute to increasing formula feeding prevalence, since mothers obtain a distorted perspective.
The findings of this meta-analysis have important implications for Ethiopia’s food and nutrition policy, particularly in addressing the high prevalence of formula feeding. The study highlights key factors associated with increased formula feeding, such as cesarean deliveries, positive attitudes towards formula, delayed initiation of breastfeeding, and reliance on information from friends and family. To align with Ethiopia’s existing food and nutrition policies, which emphasize the promotion of exclusive breastfeeding, policy makers should implement targeted interventions that address these specific factors. For instance, they could enhance postnatal support for mothers who undergo cesarean sections to facilitate early breastfeeding initiation and provide additional education on managing breastfeeding challenges. Furthermore, strengthening public health campaigns to counteract favorable attitudes toward formula feeding, and the promotion of accurate information through health-care professionals, could help shift perceptions and practices.
STRENGTHS AND LIMITATIONS OF THE REVIEW
To the authors’ knowledge, no previous systematic review or meta-analysis has been conducted specifically on formula feeding in Ethiopia. Therefore, this study provides comprehensive information on the pooled prevalence of and factors associated with formula feeding in the Ethiopian context. However, this review is not without limitations. First, significant heterogeneity was identified in the prevalence of formula feeding across the included studies. A random-effects model was used, and a sensitivity analysis was conducted to address this. However, it did not account for all the heterogeneity, so readers should consider this when applying these findings. Second, no studies were found from several regions of Ethiopia, including Afar, Somali, Gambela, South Nations, Nationalities, Sidama, and Harari. As a result, the available data may not fully represent the entire country, and researchers are recommended to conduct primary studies in those regions. The studies included in this review were limited to cross-sectional designs, which restricts the ability to establish causal relationships between identified factors, such as attitudes toward formula feeding and its practice. Further longitudinal research is needed to better understand the directionality of these associations. Additionally, the review was not registered in PROSPERO. Finally, all infants under 6 months old, including preterm infants and those who had been in the neonatal intensive care unit, were included, which could increase the prevalence of formula feeding.
CONCLUSION AND RECOMMENDATIONS
The pooled prevalence of formula feeding practice in Ethiopia was high. Cesarean delivery, positive attitude towards formula feeding, late initiation of breastfeeding, and receiving formula feeding–related information from friends and families were positively associated with formula feeding. Given these findings, the authors recommend that public health interventions in Ethiopia should target these key determinants to decrease the high prevalence of formula-feeding practices observed in the country. Strategies addressing factors like promoting vaginal delivery, improving attitudes towards breastfeeding, encouraging early breastfeeding initiation, and limiting the influence of social networks on formula-feeding decisions may be beneficial.
Ethical Approval and Consent to Participate
Because this review is based on primary studies, participants did not provide direct consent; nonetheless, informed consent was received from each individual participant during the primary study for all of the research included in this analysis. The participants’ privacy and confidentiality were safeguarded throughout the process. We did not need to get extra consent from the authors of the original research, since all relevant studies are freely available online.
Author Contributions
Conceptualization: Z.A.G. Data curation: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Formal analysis: Z.A.G. Investigation: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Methodology: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Project administration: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Resource: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Supervision: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Validation: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Visualization: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M. Writing: Original draft: Z.A.G. Rewriting: review and editing: Z.A.G., B.E.S., A.D.W., W.M.T., M.A.B., W.T.W., G.D.D., T.M.D., and F.Z.M.
Supplementary Material
Supplementary Material is available at Nutrition Reviews online.
Funding
No funding source was available for this review.
Conflict of Interest
The authors declared that there is a competing interest.
Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
