Prevalence of exclusive breastfeeding practice in the first six months of life and its determinants in Iran: a systematic review and meta-analysis

We searched different scholarly electronic databases, namely PubMed/MEDLINE, Embase, Scopus, ISI/Web of Science, the Cochrane Library, Directory of Open Access Journals Directory (DOAJ) and Google Scholar as well as Iranian databases (Barakathns, MagIran and the Scientific Information Database, SID) up to November 2018. The search terms used were: (“exclusive breastfeeding” OR “breastfeeding” OR “breast-feeding” OR “breastfeeding patterns” OR “breastfeeding practices” OR “breastfeeding status” OR “feeding status”) AND (“frequency” OR “epidemiology” OR “prevalence” OR “patterns” OR “assessment” OR “investigation”) AND “Iran”. The reference list of included studies was also scanned in order to obtain relevant additional studies. A search strategy adapted to PubMed/MEDLINE, Scopus, ISI/Web of Science and Embase is reported in appendix (Additional file 1).

Inclusion criteria were: 1) studies in which mothers used their milk to feed their infants up to 6 months of age; 2) studies reporting the prevalence of EBF in the first 6 months of life; 3) studies in which babies were aged more than 6 months; 4) studies whose data were sufficient to calculate the prevalence; 5) studies published in peer-reviewed journals; and 6) studies written either in Persian or English.

Exclusion criteria were the following: 1) studies designed as case-series, case-reports, randomized clinical trials, or interventional investigations; 2) studies whose data were inadequate or insufficient to estimate the prevalence of EBF; and 3) studies unavailable in full-text.

The outcomes of interest of this study included: 1) the prevalence of EBF practice in the first 6 months of life in Iranian children; and 2) the determinants of EBF practice.

We extracted the following data from the studies included in the present systematic review and meta-analysis: first author, year of publication, location, sample size, number of breastfed children, mothers’ age, reported prevalence, determinants of EBF, study design, and language of study.

The Newcastle-Ottawa Scale (NOS) was used to assess the quality of studies. This tool consists of three major sections, concerning the methodological quality, the comparability and the outcomes and statistical analysis of each included study. Two authors independently critically appraised the quality of each original study using the NOS tool. Disagreements between the two authors were resolved by consensus. According to the stars assigned to each part, the studies with at least 5 stars out of 10 were considered of good quality [25].

To extract relevant data and to evaluate the quality of the studies, two authors independently performed these steps. In case of disagreement, consensus was reached through discussion. Kappa statistics was used to assess the agreement between the two authors. Kappa coefficient was 0.93 for data extraction and 0.81 for evaluation of study quality.

All analyses were conducted using Stata Version 12 (Stata Corp, College Station, TX, USA) utilizing the “metaprop” command [26]. Overall pooled estimates with inverse-variance weights analyses were performed by logistic-normal random-effects model using DerSimonian-Laird approach with 95% confidence interval (CI) [27]. In order to stabilize the variance the double arcsine transformation method according to Freeman and Tukey was used [28].

To test for heterogeneity, I² test was utilized [29]. Subgroup analyses were conducted based on the sample size, the geographical area, and other variables such as education concerning EBF received before/during pregnancy, mother’s job, education level, type of delivery, gender of child, birth weight, mother’s diseases/co-morbidities and location of delivery. The Egger’s linear regression test and the funnel plot were used to evaluate the publication bias [30]. To determine possible sources of heterogeneity, meta-regressions were carried out based on the year and the sample size of the studies. Also, sensitivity analysis was performed to check the stability of results. To assess the strength of the different determinants of EBF practice, odds Ratio (OR) with its 95% CI was calculated using the Mantel–Haenszel method.

In the initial search, 725 records were found. After removing duplicates, the title of 596 records was checked and 543 records were deleted. The abstract of 53 studies was then reviewed and, finally, the full text of 32 studies was selected based on inclusion/exclusion criteria. Figure 1 shows the process of the search and selection of studies.

Selected studies were conducted between 2003 and 2015 [31‐62]. The sample size of studies varied between 50 and 63,071 subjects. The main characteristics of retained studies are presented in Table 1.

According to the NOS tool, the quality assessment showed that 4 studies were scored 6 stars, 9 studies 7 stars, 14 studies 8 stars, and 5 studies 9 stars. No study was excluded after rating because the study quality was always above 5 stars. Result of assessment of risk of bias for each study are reported in Table 1.

Based on DerSimonian-Laird model, EBF prevalence in Iran was computed to be 53% (CI 95%; 44–62) (Fig. 2). Heterogeneity resulted statistically high, I² = 99.7%, P = 0.000. Sensitivity analysis also showed that the results did not change before and after the analysis and confirmed the stability of the results.

Table 2 shows the results of the subgroup analysis. According to the geographical area of the study, the highest prevalence of EBF was observed in the north (62% versus 61% in the west, 60% in the east, 48% in the south, and 47% in the center of Iran). In terms of sample size, in studies with a sample size comprising more than 500 subjects, the prevalence was 56%, and 52% in studies with less than or equal to 500 individuals. The prevalence of EBF in mothers who had been educated before and during pregnancy was 55% and 50%, respectively. The prevalence of EBF in unemployed and employed mothers was 58% and 55%, respectively. In terms of educational level, the prevalence of EBF in mothers without and with higher education was 58% and 56%, respectively. The prevalence of EBF in mothers who delivered vaginally was 58% and 49% in mothers who underwent cesarean section. The prevalence of EBF stratified according to the gender of baby was 60% and 50% in case of female and male, respectively. The prevalence of EBF in infants weighing less than 2500 g was 62% while was 60% in infants weighing more than 2500 g. In mothers with diabetes, kidney, and cardiovascular disease, the prevalence of EBF was 44%, while it was 50% in healthy mothers. The prevalence of EBF in mothers giving birth at government hospitals was 69% and 51% in mothers who delivered their babies in private hospitals.

Association between some variables and prevalence of EBF was considered in Table 3. In this table, the strengths of the determinants of EBF practice based on the OR computed according to the Mantel–Haenszel method are reported. More in detail, the OR for breastfeeding education received before pregnancy was 1.13 (0.94–1.36), for mother’s job 1.01 (0.81–1.27), for education level 1.12 (0.89–1.42), for type of delivery 1.16 (0.98–1.37), and for gender of child 1.03 (0.83–1.28). All of these predictors were not statistically significant, even though suggestive of a trend.

Meta-regressions were performed based on the year of publication and the sample size; the results are shown in Table 4. Based on the year of publication (P = 0.61) and the sample size (P = 0.26) of included studies, EBF exhibits a decreasing trend throughout the time, even though not statistically significant.

Using the Egger’s linear regression test, the publication bias of included studies was investigated, and resulted not statistically significant (P = 0.27), as pictorially shown in Fig. 3.

In this systematic review and meta-analysis, based on the data from included studies, we examined the determinants of EBF. Training received before and during pregnancy can have an impact on the three dimensions of knowledge, attitudes and behavior of the mothers and encourage them to practice EBF [63]. The findings of this study showed, indeed, that EBF in mothers who received training was higher than in untrained mothers.

Pregnant women and their spouses should be carefully informed about infant birth and breastfeeding, an integral part of prenatal care. Other members of the family who can support breastfeeding can be trained too [64]. Training can be done at health centers and clinics. Hospitals and other institutions can also provide training for pregnant women and their partners. Other health system staff, such as pediatricians, nurses and midwives, play an important role, as well as mother-to-mother education groups and other organizations [65].

Maternal occupation was one of the factors contributing to an increase in the prevalence of EBF. The findings of this study showed that the prevalence of EBF in Iranian housewife mothers was higher than that of employed mothers. This finding is consistent with the results of studies carried out in Ethiopia [66], Saudi Arabia [67], Canada [22] and Jordan [68].

Mothers who work suffer from lack of time, and fatigue, and may find difficulties in breastfeeding [69]. Employment regulations play an important role in promoting EBF practice, by giving mothers more time to breastfeed their babies [70]. On the other hand, it seems that postpartum leave is more likely to lead to an increase in EBF. As such, postpartum mothers need more support from their employers [71].

At present, women in Iran can use 6 months of maternity leave, and their husbands can use 2 weeks. This law is better enforced in governmental organizations but not in many nongovernmental organizations. Although policy- and decision-makers are making a lot of efforts to increase the application of this law, they still have problems such as lack of support from insurance organizations, from employers, and lack of sufficient funding [72].

The findings of this study showed that the prevalence of EBF in less literate women is higher than that of women with university education, which is consistent with findings from studies conducted in Bangladesh [5], United Arab Emirates [73] and Ethiopia [74]. Mothers with lower education appear to be more interested in EBF education. Maternal education is recognized as an important social component for promotion and health-care of children [75]. In a systematic review carried out in high-income countries, results showed that interventions designed and implemented for educational purposes significantly increased the practice of EBF [76].

In our study, the findings showed that the prevalence of EBF in women who gave birth vaginally was higher than that of mothers who had cesarean section, which is consistent with the results of studies performed in Ecuador [77], Saudi Arabia [78], and Jordan [68]. The results of a meta-analysis of 53 studies showed that EBF rates were lower in women with cesarean delivery than in women with vaginal delivery [79]. Health-care providers should increase the awareness of women concerning delivery. Cesarean section is, indeed, associated with special surgical procedures and the use of local anesthesia. It is characterized by a high probability of uterine or urinary infections, increased bleeding, constipation, increased hospitalization time, and higher economic costs [80]. The results of a study showed that women who had cesarean section had a greater tendency to do so in later pregnancies and, accordingly, increased their EBF levels compared to their previous one [81].

According to a meta-analysis, the prevalence of cesarean section in Iran was estimated to be 48% [82]. This rate is rather higher when compared to other countries. Since cesarean delivery can have negative effects on the mother and the baby, such as EBF reduction, health policy- and decision-makers in Iran should make a lot of effort to reduce the use of cesarean delivery [83]. In the Health System Transformation Plan (HSTP), which began in 2014 in Iran, much attention has been paid to reducing cesarean delivery in Iran, and health-care service providers pay for maternity welfare costs to mothers to reduce this kind of delivery [84].

In the present study, the prevalence of EBF in mothers with male children was higher compared to mother with female infants and this is consistent with the results of a study conducted in Saudi Arabia [78] and the findings of study performed in Ghana [85].

Based on the results of our study, the prevalence of EBF in infants whose birth weight was less than 2500 g was higher than that of infants above this weight. Infants with low birth weight are at risk for certain diseases. Breast milk can improve the function of the digestive system, reducing infections [86]. Studies have shown that EBF is a necessity for infants weighing less than 2500 g and should be taken seriously by mothers [87, 88].

Health policy- and decision-makers in Iran have always emphasized the importance and the benefits of EBF and, given its religious, social and economic implications, have implemented broad programs for education and promotion at the community level. Appropriate laws have been approved to promote EBF and supporting mothers during lactation in recent years. In 2011, the maternity leave law was approved for mothers who breastfeed their babies, and according to that, employers can extend maternity leave for a period of 9 months, and the period of maternity leave for triple or more childbirths (1 year receiving salaries). Also, these mothers can come to work 1 h later or leave the workplace 1 h earlier. All government agencies are required to provide female employees with appropriate facilities enabling EBF at the workplace. This law focuses on protecting working women and ensuring their job security, taking into account the specific circumstances of women in lactation, as well as on improving the condition of the growth of their infants.

Comprehensive search of various scholarly databases, sub-group analysis, meta-regressions and sensitivity analysis were among the strengths of this systematic review and meta-analysis. However, this study also had some limitations, which should be properly recognized. In some Iranian studies, there was no study on the prevalence of EBF. Methodological differences in studies may have led to a high, statistically significant heterogeneity. Low sample size of many studies (21 studies with a sample size less than 500) represents another limitation of the present investigation.