Exclusive breastfeeding and associated factors among HIV-infected mothers in sub-Saharan African countries: a systematic review and meta-analysis

Abstract

Introduction

The World Health Organization recommends exclusive breastfeeding for the first six months, along with maternal antiretroviral treatment, to reduce the risk of HIV transmission to infants and promote their health and survival. However, globally, this practice is not as widespread as it should be, especially in sub-Saharan Africa, as shown by several studies among HIV-infected mothers. However, no comprehensive review has examined the overall prevalence and the factors influencing it. This systematic review and meta-analysis aimed to determine the overall prevalence of exclusive breastfeeding and associated factors among HIV-infected mothers in sub-Saharan Africa.

Methods

International electronic databases including PubMed, Embase, Scopus, Cochrane, and Web of Science were systematically searched for observational studies on exclusive breastfeeding among HIV-infected mothers in sub-Saharan Africa published after 2010. Citation tracking revealed additional references. Two authors extracted the necessary data independently in a standardized format, which was analyzed with STATA version 17.

The Joanna Briggs Institute (JBI) critical appraisal tool was used to assess the study quality. Heterogeneity was evaluated using the Cochran Q test and quantified with the I² statistic. Publication bias was assessed using funnel plots and Egger’s test. Sensitivity analyses, subgroup analyses, and meta-regression were conducted to assess sources of heterogeneity. Finally, a random effects meta-analysis model was computed to estimate the pooled prevalence of exclusive breastfeeding. Associations between factors and exclusive breastfeeding were also examined via a random effects model.

Result

After 5305 studies were reviewed, 32 studies involving 9319 study participants fulfilled the inclusion criteria and were included in this meta-analysis. The findings of these 32 studies revealed a pooled prevalence of 55.9% (95% CI: 45.93, 68.23%), with substantial heterogeneity (I² = 99.68%). Breastfeeding counseling (AOR: 3.69, 95% CI: 2.41, 5.64), adequate antenatal care visits (AOR: 2, 95% CI: 1.49, 2.67), positive maternal attitudes toward infant feeding practices (AOR: 4.29, 95% CI: 2.32, 7.95) and good knowledge about infant feeding practices (AOR: 4.88, 95% CI: 3.3, 7.21) were associated with exclusive breastfeeding among HIV infected mothers.

Conclusion

This study revealed that slightly over half of HIV-infected mothers in sub-Saharan Africa engage in exclusive breastfeeding. The findings highlight the need for support for HIV-infected mothers to enhance exclusive breastfeeding and minimize transmission risks, recommending better breastfeeding counseling, antenatal care, maternal education on feeding, and consideration of sociocultural factors for tailored strategies.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12889-026-28418-6.

Introduction

Breast milk is the ideal food for infants. It is safe, clean, and contains antibodies that help protect against many childhood diseases. Breast milk provides all the energy and nutrients that the infant needs in the first months of life and continues to meet up to half or more of a child's nutritional needs in the second half of the first year and up to a third year of life. Therefore, breastfeeding is one of the most important ways to preserve child health and survival. The World Health Organization (WHO) recommends exclusive breastfeeding (EBF) for the first six months of life and continuing breastfeeding for up to two years or more, alongside appropriate complementary feeding practices [1].

Exclusively breastfed children for the first six months are 14 times more likely to survive than non-breast fed children. Sub-optimal breast feeding practices during the first six months is the most risk factor for the infant and child morbidity and mortality, including poor school performance, reduced productivity and impaired intellectual development. Annually, non-exclusive breastfeeding contributed to 804, 000 child deaths in 2011 [2–4].

In spite of WHO recommendations, only approximately 44% of infants under 6 months of age worldwide were exclusively breastfed from 2015–2020, regardless of their mothers' human immunodeficiency virus (HIV) status [5]. The prevalence of breastfeeding is particularly low in sub-Saharan African (SSA) countries, where only 36.0% of ever-breastfed children are exclusively breastfed. The highest prevalence was observed in Rwanda, whereas the lowest was observed in Gabon [6]. The practice of EBF among HIV infected mothers is affected by inadequate counseling on EBF practice [7], cultural and family influence [8], fear of stigmatization [7], maternal lack of decision power [8, 9], and fear of vertical transmission of HIV to infant [8, 10].

In order to prevent MTCT of HIV during breastfeeding while ensuring provision of appropriate nutrition to the child, several strategies have been identified by UNICEF and WHO to promote EBF practices. WHO updated its recommendations and advised HIV-infected mothers to exclusively breastfeed their infants for the first 6 months of life, under the cover of antiretroviral treatment (ART) in 2010. This updated recommendation is based on global evidence demonstrating that giving ART to mothers living with HIV significantly reduces the risk of HIV transmission through breastfeeding from 25% to less than 5% [11]. Exclusive breastfeeding for the first 6 months, combined with maternal ART, is now considered the optimal infant feeding practice for HIV-exposed infants [12, 13]. In line with WHO recommendation, several national ART and PMTCT guidelines, those emphasize integrated HIV care and maternal nutrition support developed in many SSA countries.

Although several studies have been conducted to assess the practice of EBF and its associated factors among HIV-infected mothers in SSA countries, there is no synthesized evidence evaluating exclusive breastfeeding practices in relation to the current WHO recommendations. Their findings are also inconsistent and context-specific. This limits the ability of policymakers, program planners, and clinicians to design evidence-based interventions. Therefore, this study aims to estimate the pooled prevalence and associated factors of exclusive breast feeding in sub-Saharan Africa countries.

Methods

Protocol registration

The protocol of this systematic review and meta-analysis is registered in the International Prospective Register of Systematic Review and Meta-analysis (Prospero) with a registration number of CRD42024563705.

Search strategy

The review followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 statement, which provide clear instructions for conducting and reporting systematic reviews (Table S4). The search strategy targeted to find both published and unpublished studies on the prevalence and factors associated with exclusive breastfeeding practices among HIV-infected mothers in sub-Saharan African countries started on January 1, 2010 (after the 2010 WHO recommendation on HIV and infant feeding practices). A step-by-step search strategy was used in this review. Primarily, the electronic databases that comprise health system research, including PubMed, Embase (Ovid), Scopus, Cochrane, and Web of Science, were used to search for relevant studies. Afterward, hand-searching of reference lists and citation tracking of included studies was performed, and additional studies were screened. The search was performed independently by the authors (DTA and MYT) using three main categories of keywords related to the review question, Medical Subject Headings (MeSH), and the Boolean search operators “OR” and “AND” were employed. The main keywords utilized for the article search included exclusive breastfeeding, HIV/AIDS, and sub-Saharan African countries. These keywords were selected because they align with the core concepts of the review objective. Various combinations and variations of these keywords were employed. We explored synonyms for each search term in the literature to create a comprehensive search strategy. The full search strategy employed and the search results identified across the 5 databases are available as (Table S2).

Selection criteria

For this systematic review, the populations were selected using PECO framework.

• Population (P): HIV-infected mothers with children aged < 24 months.

• Exposure (E): Mothers exposed to the respective factors.

• Comparison (C): Mothers not exposed to the respective factors.

• Outcome (O): Exclusive breastfeeding during the first six months of life.

In addition, studies were included if they met the following criteria: (1) the study area: studies which were exclusively conducted in sub-Saharan African countries. 2) Publication conditions: only articles published in peer-reviewed journals were considered. 3) Types of studies: observational study design (cross-sectional, case–control and cohort study designs) that reported exclusive breastfeeding and its associated factors among HIV-infected mothers. 4) Language: only English-language publications were included. 5) Study participants: HIV infected mothers with children aged less than 24 months, in order to minimize recall bias related to infant feeding practices. 6) Study period: studies conducted between 2010 and 2024 were included. The year 2010 was chosen as the baseline to capture studies published after the WHO’s recommendation on HIV and infant feeding [12].

Exclusion criteria: Studies that reported on breastfeeding alone, without including exclusive breastfeeding, were also excluded. Experimental and interventional studies were also excluded. Additionally, Studies with insufficient or unavailable data were excluded.

Data extraction

The search results were exported to Covidence (https://www.covidence.org) for evaluation in two distinct screening steps after duplicates were dropped. The first step is based on the title and abstract, filtering only relevant records and excluding irrelevant records on the basis of the review question and eligibility criteria. Each article was assessed independently by two reviewers (DTA and NKB). In the case of disagreements, the issues were thoroughly examined, and a third-party (MYT) opinion was consulted to resolve them. The same reviewers carried out the second step on the basis of the full texts. The studies agreed with the inclusion criteria, with any disagreements resolved through discussion and consensus to minimize bias. The full-text articles were extracted and assessed against the inclusion and exclusion criteria.

Data were independently extracted from the eligible studies by the authors (DTA and NKB) via a customized data extraction form. For the first outcome (prevalence), the data extraction form included the primary author, publication year, country where the study was conducted, sample size, study design, and prevalence with 95% confidence intervals. The association measurements with their confidence intervals were extracted for the secondary outcome (associated factors). In addition to this, the definitions and measurement approaches for key associated factors across the included studies are summarized (Table S1).

Risk of bias (quality) assessment

Two independent reviewers (NKB and KTW) assessed the quality of the included studies. Each article's quality was evaluated via the Joanna Briggs Institute (JBI) critical appraisal tool designed for cohort studies with 11 questions, case–control studies with 10 questions, analytical cross-sectional studies with 8 questions, and descriptive cross-sectional studies with 9 questions. Each question was scored as yes = 1, no = 0, or unclear or not applicable = 0. The total score for each study is reported as a percentage and classified as high risk if 20–50% of the items score yes, moderate risk if 50–80% score yes, and low risk if 80–100% score yes. Only studies that scored ≥ 50% were considered for the review (Table S3).

Data analysis

The data from the Covidence software were exported to a Microsoft Excel spreadsheet and processed. The processed data were then imported into STATA Version 17 statistical software for further analysis. The pooled prevalence of exclusive breastfeeding practices was estimated via a random effects model with the DerSimonian-Laird method. The presence of heterogeneity between studies was determined via the Cochran-Q test, and the degree of heterogeneity was quantified via I-square (I²) statistics. The level of heterogeneity was classified as low (I²: 0–25%), moderate (I²: 25–50%), or high (I²: ≥ 50%). On the basis of the results of the statistical test, significant heterogeneity was exhibited among the included studies (I² = 99.68%, p < 0.001); hence, a random effects meta-analysis model was computed to estimate the Der Simonian and Laird’s pooled effect [14, 15]. To identify potential publication bias, funnel plots were created [16], and Egger's test was applied. Publication bias was declared when Egger's test was significant (p value < 0.05) [17]. The trim-and-fill test was performed to identify the possible effects of the absence of studies related to the summary association measurement in the meta-analysis [18]. Sensitivity analysis, subgroup analyses, and meta-regressions were performed to explore the sources of heterogeneity in the studies included in the systematic review. Additionally, a Galbraith plot was created to locate any studies that had statistical outliers. For the secondary outcomes, a log odds ratio was used to determine the associations between EBF practices and associated factors.

Results

Study selection

In the first step of the search, 5305 potentially relevant articles concerning EBF and associated factors among HIV-infected mothers were retrieved from PubMed, Embase, Scopus, Cochrane, and Web of Science. After 2903 duplicates were removed by covidence, a total of 2402 articles were screened on the basis of title and abstract according to the inclusion criteria, as described in the methods. A full-text screening of the 94 potentially relevant articles resulted in 35 eligible studies. Among the eligible studies, 3 were excluded because they scored < 50% according to the JBI criteria used to assess the quality of the studies, and 32 studies were included in the review (Fig. 1).

Fig. 1.

PRISMA flow chart illustrating the study selection and screening process for systematic review and meta-analysis

Characteristics of the studies included in the meta-analysis

The systematic review included a total of 32 studies. Among these studies, 29 had a cross-sectional design [7, 9, 10, 19–44], whereas the remaining 3 had a cohort design. Among the 3 cohort studies, 2 were retrospective in nature [45, 46], and 1 was a prospective study [47]. In addition, almost all of the studies were conducted in health institutions. The studies included in the review were conducted across seven countries, with the majority occurring in Ethiopia (Table 1).

Table 1.

Characteristics of studies included in the meta-analysis for the pooled prevalence of EBF practice among HIV-infected mothers in SSA

Author, (Publication Year)	Country	Region	Study design	Study setting	Sample size	Infants' age	Number of women who practice EBF	Prevalence of EBF practice	Quality
Aguti et al. (2020) [45]	Uganda	East Africa	Retrospective cohort	Facility	345	0–24 months	100	29	Good
Aishat et al. (2015) [7]	Nigeria	West Africa	Cross-sectional	Facility	600	0–12 months	293	61	Good
Akpan et al. (2023) [19]	Nigeria	West Africa	Cross-sectional	Facility	246	0–12 months	181	73.6	Good
Andare et al. (2019) [20]	Kenya	East Africa	Cross-sectional	Facility	180	0–12 months	55	71.4	Good
Bultum et al. (2022) [21]	Ethiopia	East Africa	Cross-sectional	Facility	218	0–6 months	122	56	Good
Dagnew et al. (2019) [22]	Ethiopia	East Africa	Cross-sectional	Facility	213	0–12 months	77	36.2	Good
Ekubagewargies et al. (2019)	Ethiopia	East Africa	Cross-sectional	Facility	402	6–12 months	92	23.06	Good
Faustine and Moshe (2022) [24]	Tanzania	East Africa	Cross-sectional	Facility	372	6–12 months	216	58.1	Good
Flax et al. (2016) [25]	Malawi	East Africa	Cross-sectional	Facility	160	0–24 months	129	80.6	Good
Gebremariam et al. (2024) [26]	Ethiopia	East Africa	Cross-sectional	Facility	397	0–6 months	313	78.8	Good
Gejo et al. (2019) [27]	Ethiopia	East Africa	Cross-sectional	Facility	304	6–12 months	270	88.8	Good
Genetu et al. (2017) [28]	Ethiopia	East Africa	Cross-sectional	Facility	367	6–24 months	317	86.4	Good
Goon et al. (2021) [29]	South Africa	Southern Africa	Cross-sectional	Facility	469	6–24 months	150	32	Good
Iliyasu et al. (2019) [10]	Nigeria	West Africa	Cross-sectional	Facility	203	6–24 months	34	16.7	Good
Mebratu et al. (2020) [30]	Ethiopia	East Africa	Cross-sectional	Facility	207	0–12 months	169	81.6	Good
Mekonnen et al. (2022)	Ethiopia	East Africa	Cross-sectional	Facility	349	0–12 months	305	87.4	Good
Mihret et al. (2020) [32]	Ethiopia	East Africa	Cross-sectional	Facility	485	0–24 months	358	73.8	Good
Mnyani et al. (2017) [9]	South Africa	Southern Africa	Cross-sectional	Facility	180	0–12 months	23	63.9	Good
Moges et al. (2017) [46]	Ethiopia	East Africa	Retrospective cohort	Facility	305	0–24 months	294	96.4	Good
Mukerem and Haidar (2012) [33]	Ethiopia	East Africa	Cross-sectional	Facility	371	0–12 months	271	73	Good
Muluye et al. (2012) [34]	Ethiopia	East Africa	Cross-sectional	Facility	209	0–24 months	175	83.8	Good
Mutawulira et al. (2022) [35]	Uganda	East Africa	Cross-sectional	Facility	108	0–12 months	83	77	Good
Nabakwe et al. (2018) [36]	Kenya	East Africa	Cross-sectional	Facility	371	6–12 months	349	94	Good
Negash et al. (2019) [37]	Ethiopia	East Africa	Cross-sectional	Facility	348	0–24 months	246	70.7	Good
Oiye et al. (2017) [38]	Kenya	East Africa	Cross-sectional	Facility	60	6–12 months	45	75	Good
Okanda et al. (2018) [39]	Kenya	East Africa	Cross-sectional	Community	154	6–12 months	100	64.9	Good
Olorunfemi and Dudley (2018) [40]	Lesotho	Southern Africa	Cross-sectional	Facility	191	0–6 months	26	14	Good
Onubogu et al. (2015) [47]	Nigeria	West Africa	Prospective cohort	Facility	142	6–12 months	8	5.6	Good
Philemon et al. (2022) [41]	Tanzania	East Africa	Cross-sectional	Facility	524	0–24 months	425	81.1	Good
Remmert et al. (2020) [42]	South Africa	Southern Africa	Cross-sectional	Facility	156	0–6 months	44	28.2	Good
Wakwoya et al. (2016) [43]	Ethiopia	East Africa	Cross-sectional	Facility	260	0–24 months	201	77.3	Good
Zewdu et al. (2023) [44]	Ethiopia	East Africa	Cross-sectional	Facility	423	0–12 months	270	63.8	Good

Meta-analysis

The overall prevalence calculated from the 32 included sub-Saharan African studies revealed that the pooled prevalence of EBF among HIV-infected mothers was 55.9% (95% CI: 45.93, 68.23%). The analysis detected substantial heterogeneity across the included studies, as indicated by the high I2 statistic (I2 = 99.68, p value < 0.001). Owing to this considerable heterogeneity, a random-effect model was used to estimate the overall pooled prevalence of EBF among HIV-infected mothers, as this model provides a more conservative effect size estimate. The lowest reported prevalence of EBF was 5.6% [47], whereas the highest reported prevalence was 96.4% [46] among the included studies (Fig. 2). The Galbraith plot did not reveal any significant outliers (Figure S1).

Fig. 2.

Pooled prevalence of EBF practices among HIV-infected mothers in SSA

Subgroup analysis

A subgroup analysis was performed by classifying studies on the basis of study period, sampling frame, sample size, and region. With respect to study period, the studies were grouped into three categories: 2010–2015, 2016–2020, and 2021–2024. The subgroup analysis revealed that the highest prevalence (72.9%) of EBF was reported in studies conducted from 2021 to 2024, whereas the lowest prevalence (52.3%) was reported in studies done between 2010 and 2015 (Fig. 3).

Fig. 3.

Subgroup analysis using study year for EBF practices among HIV-infected mothers in SSA

Regarding to the subgroup analysis based on sampling frame, the studies that included women with infants aged 0–24 months reported the highest prevalence (70.98%) of EBF. In contrast, the studies that included women with infants aged 6–24 months reported the lowest prevalence (36.2%) of EBF (Fig. 4).

Fig. 4.

Subgroup analysis of EBF practices among HIV-infected mothers in SSA by sampling frame

With respect to the subgroup analysis based on sample size, the studies were categorized into two groups: those with sample sizes above the mean and those with sample sizes below the mean. The mean sample size across all included studies was 291. Studies with sample sizes above the mean of 291 reported the highest prevalence (63.9%) of EBF among HIV-infected mothers in sub-Saharan Africa. In contrast, the studies with sample sizes below the mean of 291 participants reported a lower prevalence (48.03%) of EBF (Fig. 5).

Fig. 5.

Subgroup analysis using sample size for EBF practices among HIV-infected mothers in SSA

Subgroup analysis based on study Region, the studies were categorized into four groups: East Africa, West Africa, Central Africa, and Southern Africa. The subgroup analysis showed that the highest prevalence (68.08%) of EBF was reported in studies conducted in East Africa., whereas the lowest prevalence (26.71%) was observed in studies conducted in West Africa (Fig. 6).

Fig. 6.

Subgroup analysis using region for EBF practices among HIV-infected mothers in SSA

Publication bias and sensitivity analysis

To assess the potential presence of publication bias, a graphical funnel plot and Egger's test at the 5% significance level were computed. The asymmetric appearance of the funnel plot, with most studies distributed on the upper-left side of the plot, indicated the existence of publication bias. This finding was further supported by Egger's test, which revealed a statistically significant presence of publication bias (β = − 9.50, SE = 0.83, z = − 11.43, p < 0.001). This pattern shows that smaller studies tended to report lower or more extreme prevalence estimates compared with larger studies (Fig. 7).

Fig. 7.

Funnel plot to assess publication bias in EBF practices among HIV-infected mothers in SSA

A trim-and-fill analysis was attempted to further assess the potential impact of publication bias. However, the trim-and-fill algorithm failed to converge, likely due to the extremely high between-study heterogeneity (I² = 99%). Therefore, adjusted pooled estimates could not be reliably generated using this method.

A sensitivity analysis was performed to evaluate the effect of each study on the pooled estimated prevalence of EBF by excluding each study step-by-step with a random effects model. The results revealed that the pooled EBF prevalence estimate did not significantly differ, regardless of which individual study was removed from the analysis (Figure S1).

Factors associated with exclusive breastfeeding

The systematic review and meta-analysis identified five factors that were reported to be significant for EBF among HIV-infected mothers on the basis of findings from at least two studies. One of these factors was HIV disclosure status to one’s spouse, which was found to be a significant determinant of EBF in two individual studies. However, the pooled effect size analysis did not reveal any significant association between HIV disclosure status and EBF. In contrast, the pooled estimates revealed that several other factors were significant factors for EBF. Specifically, the meta-analysis revealed that receiving breastfeeding counseling, having adequate ANC visits, having good knowledge about infant feeding and having a positive attitude toward infant feeding were all significant factors associated with higher rates of EBF among HIV-infected mothers.

Breastfeeding counseling

This systematic review and meta-analysis identified five studies that reported a significant association between breastfeeding counseling and EBF among HIV-infected mothers. Compared with those who did not receive such counseling, the lowest and highest reported adjusted odds ratio (AOR) for EBF among those who received breastfeeding counseling were 2.7 (95% CI: 1.3, 5.59) and 6.87 (95% CI: 2.62, 18.0), respectively. The pooled meta-analysis of these five studies revealed an overall (AOR: 3.69, 95% CI: 2.41, 5.64) of EBF among mothers who received breastfeeding counseling. The heterogeneity assessment revealed an I2 of 0.00% and a p value of 0.001, indicating a high degree of homogeneity across the included studies (Fig. 8).

Fig. 8.

Forest plot showing the pooled AOR for the associations between breastfeeding counseling and EBF among HIV-infected mothers

To examine potential publication bias, a funnel plot analysis was conducted, which revealed a symmetrical distribution of the study effects (Figure S2). Additionally, Egger's regression test (p = 0.36) did not detect any significant evidence of publication bias.

ANC visits

Four studies reported a significant association between ANC visits and EBF among HIV-infected mothers. Among these significant studies, the lowest and highest AOR for EBF among mothers who had adequate ANC visits were 1.76 (95% CI: 1.09, 2.83) and 4.20 (95% CI: 1.20, 14.74), respectively, compared with those who had inadequate ANC visits. The pooled meta-analysis of the three studies revealed an overall (AOR: 2.00, 95% CI: 1.49, 2.67) of EBF among HIV-infected mothers with adequate ANC visits. The heterogeneity assessment revealed an I2 of 0.00% and a p-value of 0.001, indicating a high degree of homogeneity across the included studies (Fig. 9).

Fig. 9.

Forest plot showing the pooled AOR for the associations between ANC visits and EBF among HIV-infected mothers

To assess potential publication bias, a funnel plot analysis was conducted, which revealed a symmetrical distribution of the study effects (Figure S3). Egger's regression test (p = 0.14) also did not detect any significant evidence of publication bias.

Attitude toward infant feeding practices

Three studies reported a significant association between maternal attitudes toward infant feeding practices and EBF practices among HIV-infected mothers. Among these significant studies, the lowest and highest AOR for EBF among mothers who had positive attitude towards infant feeding practices were 3.12 (95% CI: 3.03, 3.22) and 7.92 (95% CI: 2.96, 21.20), respectively, compared with those with negative attitude. The pooled meta-analysis of the three studies revealed an overall (AOR: 4.29, 95% CI: 2.32, 7.95) of EBF among HIV-infected mothers with a positive attitude. The heterogeneity assessment revealed an I2 of 53.23% and a p value of 0.001, indicating moderate heterogeneity across the included studies (Fig. 10).

Fig. 10.

Forest plot showing the pooled AOR for the associations between attitude towards EBF and EBF among HIV-infected mothers

To assess potential publication bias, funnel plot analysis was conducted, which revealed a symmetrical distribution of the study effects (Figure S4). Egger's regression test (p = 0.0514) also did not detect any significant evidence of publication bias. Additionally, leave-one-out sensitivity analysis was conducted to identify potential sources of heterogeneity in the pooled estimate for maternal attitude as a risk factor for EBF. The results of this analysis demonstrated that the overall pooled effect size was not unduly influenced by any single study, with the pooled estimate varying between 3.12 (95% CI: 3.03–3.22) and 6.77 (95% CI: 3.17–14.42) after the deletion of individual studies (Figure S5).

Knowledge about infant feeding practices

The meta-analysis identified three studies that reported a significant association between maternal knowledge of infant feeding practices and practices related to EBF among HIV-infected mothers. Among these significant studies, the lowest and highest AOR for EBF among mothers who had good knowledge of infant feeding practices were 3.32 (95% CI: 1.15, 9.57) and 5.50 (95% CI: 2.14, 14.13), respectively, compared with those with poor knowledge. The pooled meta-analysis of the three studies revealed an overall (AOR: 4.88, 95% CI: 3.30, 7.21) for EBF among HIV-infected mothers with good knowledge of infant feeding practices. The heterogeneity assessment revealed an I2 of 0.00% and a p value of 0.001, indicating a high degree of homogeneity across the included studies (Fig. 11).

Fig. 11.

Forest plot showing the pooled AOR for association knowledge about EBF and EBF among HIV-infected mothers

To assess potential publication bias, the researchers conducted a funnel plot analysis, which revealed a symmetrical distribution of the study effects (Figure S6). Additionally, Egger's regression test (p = 0.66) did not detect any significant evidence of publication bias.

Discussion

As no studies or reviews have yet estimated the pooled prevalence of exclusive breastfeeding and its associated factors in Sub-Saharan Africa, this systematic review and meta-analysis aimed to determine the pooled prevalence of exclusive breastfeeding among HIV-infected mothers in the region, along with its associated factors.

The overall pooled prevalence of exclusive breastfeeding among HIV-infected mothers was 55.98% (95% CI: 45.93, 68.23%). This result is higher than the global prevalence of exclusive breastfeeding reported over the period of 2015–2020, which was 44% of infants aged 0–6 months worldwide [1]. This difference might be due to several potential reasons. First, the global prevalence data include both HIV-infected and uninfected mothers, whereas this systematic review focused specifically on HIV-infected mothers in sub-Saharan Africa. Studies have shown that HIV-infected mothers may have higher rates of exclusive breastfeeding than general population, due to increased counseling and support services targeted at this population [19, 39, 48, 49]. The other reason might be time difference. This systematic review assessed studies conducted from 2010 to the present, and the global prevalence data covered the years 2015–2020. Exclusive breastfeeding rates may have improved over time, leading to a higher pooled prevalence in the more recent studies included in this review. This result is in line with the pooled prevalence of exclusive breastfeeding among HIV-infected mothers in Ethiopia (63.43%) [50] and East Africa [51].

The subgroup analysis revealed important trends in EBF among HIV-infected mothers in sub-Saharan Africa. The highest pooled prevalence of EBF occurred in studies conducted between 2021 and 2024, whereas the lowest prevalence was noted in studies done from 2010 to 2015. This suggests a positive trend over time, indicating that exclusive breastfeeding rates among HIV-infected mothers have likely increased in recent years compared with earlier periods [52–54].

The analysis also revealed a variation across subgroups explained in terms of sampling frame. The studies that include women with infants aged 0–24 months reported the highest prevalence of EBF. This inclusive sampling frame that includes a proportion of younger infants who are closer to the exclusive breastfeeding period (first 6 months), might enhance the likelihood of higher reported EBF and reducing recall error. In contrast, the studies that included women with infants aged 6–24 months reported the lowest prevalence of EBF. This subgroup excludes younger infants and relies entirely on retrospective reporting of feeding practices during the first 6 months, which may introduce recall bias and lead to underestimation [55, 56]. In terms of factors, in this systematic review and meta-analysis breastfeeding counseling, adequate ANC visits, having a positive attitude toward EBF and having good knowledge of EBF were significantly associated with EBF prevalence.

Compared with mothers who did not receive any breastfeeding counseling, HIV-positive mothers who had received breastfeeding counseling were 3.7 times more likely to practice EBF. This finding is consistent with the results of a systematic review of studies conducted in Ethiopia [50] and a scoping review on sub-Saharan Africa [57]. The provision of breastfeeding counseling to HIV-positive mothers appears to be a crucial factor in promoting EBF among this population. Counseling likely helps these mothers overcome barriers and concerns they may have about breastfeeding while living with HIV. This enables them to make informed decisions and adopt appropriate infant feeding practices that optimize their child's health outcomes while minimizing the risk of HIV transmission [58]. Incorporating and strengthening appropriate breastfeeding counseling within antenatal, postnatal, and HIV care services is important to enhance exclusive breastfeeding practices among HIV-infected mothers.

HIV-positive mothers who had an adequate number of ANC visits were 2 times more likely to practice EBF than those who did not have adequate ANC attendance. This finding is supported by another systematic review conducted in Ethiopia [22]. Regular ANC visits provide opportunities for healthcare providers to deliver targeted counseling, education, and support on optimal infant feeding options, including the benefits and techniques of EBF. During ANC, HIV-positive mothers can receive personalized guidance in navigating the challenges of breastfeeding while living with HIV. They may also be connected with relevant resources and peer support networks to help reinforce EBF practices [59]. Improving ANC utilization and integrating infant feeding counseling into ANC service is very essential to improve EBF practice.

Compared with those who had a negative attitude, HIV-positive mothers who had a positive attitude toward EBF were 4.3 times more likely to practice EBF. This result is in line with the result of a systematic review of qualitative research on barriers to and facilitators of EBF. A mother's attitude and beliefs about breastfeeding play a critical role in her infant’s feeding decisions and practices, especially in the context of HIV. Mothers with a positive, supportive attitude toward EBF are more likely to overcome any barriers or concerns they may have about breastfeeding while living with HIV. This positive mindset can motivate mothers to initiate EBF and persist with EBF, even in the face of potential challenges [60]. Interventions that address maternal attitude such as, counseling, behavioral change communication, and education are essential to improve EBF.

Mothers who had good knowledge of appropriate infant feeding practices were 4.9 times more likely to practice EBF than those with poor knowledge. This finding is in line with the scoping review on barriers to EBF in SSA [57]. Maternal knowledge and understanding of optimal infant feeding options emerge as crucial determinants of breastfeeding practices among HIV-infected women. Mothers with good knowledge are better equipped to make informed decisions that prioritize the health and nutritional needs of their infants, while also minimizing the risk of HIV transmission through breast milk [61]. Improving maternal knowledge through education and counseling is important to improve and strengthen the practice of EBF.

Strengths and limitations of the study

The strengths of this review include its comprehensive search strategy and the involvement of at least two reviewers at each stage of the review process. Additionally, it encompassed a significant number of relevant studies; however, the limited geographic scope, focused primarily on a few countries in Sub-Saharan Africa, reduces the overall representation of the target population. The search strategy was restricted to studies published in English, which may introduce reporting bias.

A key limitation of this study is that not all included studies reported the necessary information to conduct detailed subgroup analyses, such as variations in antiretroviral therapy regimens, clinical stage of HIV, viral load suppression status, comorbidities, health system characteristics, and sociocultural contexts, which could not be consistently extracted across all studies. This limited the ability to fully explain the sources of variability, despite the substantial heterogeneity observed among the included studies.

Even though publication bias was detected through funnel plot and Egger’s test, the trim-and-fill method failed to converge due to high heterogeneity, limiting our ability to adjust the pooled estimates. As a result, the pooled prevalence of exclusive breastfeeding may be over- or underestimated.

Conclusion and recommendation

In conclusion, this meta-analysis revealed that just over half of HIV-infected mothers in sub-Saharan Africa practice exclusive breastfeeding. The key factors associated with EBF included breastfeeding counseling, adequate ANC visits, positive maternal attitudes toward infant feeding, and good knowledge of infant feeding practices.

The findings highlight the importance of integrating breast feeding counseling in to maternal, newborn, and child health (MNCH) and HIV care services. Emphasizing early initiation and completion of recommended ANC visits is also important to improve the practice of EBF. Healthcare providers should use ANC, postnatal, and HIV care visits as critical opportunities to deliver consistent, individualized counseling. Additionally, promoting maternal knowledge and positive attitude towards EBF should be prioritized and the government and other concerned bodies should support behavior change communication strategies that promote knowledge and positive perceptions of EBF among HIV-infected mothers.

Abbreviations

ANC

Antenatal Care

AOR

Adjusted Odds Ratio

ART

Antiretroviral Therapy

EBF

Exclusive Breastfeeding

HIV

Human Immunodeficiency Virus

JBI

Joanna Briggs Institute

MTCT

Mother-to-Child Transmission

SDGs

Sustainable Development Goals

SSA

Sub-Saharan Africa

WHO

World Health Organization

Authors’ contributions

The review process was carried out in a collaborative and thorough manner by the authors. Based on this, the research idea was conceptualized by D.T.A. and N.K.B. The study selection process was also conducted independently and blindly by D.T.A. and N.K.B. after a thorough screening of the abstracts and full texts. The data extraction was also carried out by D.T.A. and N.K.B. The search was performed independently by the authors (DTA and MYT). D.T.A., N.K.B. and K.T.W. independently and meticulously appraised the contents of each of the identified articles. The quality assessment of the included studies was performed independently by N.K.B. and K.T.W. The data analysis and interpretation were conducted by D.T.A. and M.Y.T.

Funding

The authors did not receive any funding for publication.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Since the study did not use primary data, ethical approval was not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.