Understanding the Enablers and Barriers to Appropriate Infants and Young Child Feeding Practices in India: A Systematic Review

Despite efforts to promote infant and young child feeding (IYCF) practices, there is no collective review of evidence on IYCF enablers and barriers in India. This review was conducted using 2015 Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidelines. Six computerized bibliographic databases, Scopus, PubMed, PsycINFO, CINAHL, Embase, and Ovid MEDLINE, were searched for published studies on factors associated with IYCF practices in India from 1 January 1993, to 30 April 2020. IYCF practices examined were early initiation of breastfeeding, exclusive breastfeeding, continued breastfeeding at one year, introduction to solid semi-solid or soft foods, minimum dietary diversity, minimum meal frequency, minimum acceptable diet, continued breastfeeding at two years, predominant breastfeeding, and bottle feeding. In total, 6968 articles were retrieved, and 46 studies met the inclusion criteria. The common enablers of IYCF were higher maternal socioeconomic status (SES) and more frequent antenatal care visits (ANC) (≥3). Common barriers to IYCF practices were low SES and less frequent ANC. The review showed that the factors associated with IYCF practices in India are largely modifiable and multi-factorial. Improving IYCF practices would require the adoption of both facilities- and community-based policy interventions at the subnational and national levels in India.


Introduction
Appropriate infant and young child feeding (IYCF, comprising of breastfeeding and complementary feeding) play important roles in optimal child growth and development. This is because appropriate breastfeeding is associated with a lower prevalence of childhood diarrhea [1,2], upper respiratory tract infection and obesity, and maternal diseases like diabetes mellitus [3]. Additionally, appropriate complementary feeding is associated with a reduced risk of undernutrition (i.e., underweight, stunting, and/or wasting) [4][5][6]. Despite the benefits of appropriate IYCF, many low-and middle-income countries (LMICs) still report higher prevalence of inappropriate IYCF [1,[7][8][9][10][11][12]. In India, inappropriate IYCF practices have contributed to childhood malnutrition contributing to about 68% of the • (Child* or Preschool* or Pediatric* or Infant* or Bab* or Newborn* or Neonate*) AND • (Feed* or Breastfeed* or Complementary Feed* or Food*) AND • (Factor* or Determinant* or Correlate* or Cause* or Influence* or Enabler* or Barrier* or Promoter*) AND • (India)

Eligibility Criteria
Studies were included in the review if they meet the following criteria: (i) focused on children under two years of age, (ii) were conducted in India, (iii) analyzed factors associated with IYCF indicators (EIBF; EBF; continued breastfeeding at one year; introduction of solid, semi-solid, or soft foods; minimum dietary diversity (MDD); minimum meal frequency (MMF); minimum acceptable diet (MAD); continued breastfeeding at two years; predominant breastfeeding; and bottle feeding), (iv) were published between 1993 and 2020, (v) were observational studies (qualitative studies, case studies, books, policy briefs, or theses were excluded), (vi) were published in a peer-reviewed journal (non-peer-reviewed research, review, or commentaries were excluded), and (vii) were written in English. Eight WHO/UNICEF IYCF indicators were selected for this review based on the available published literature at the regional and national level of India [18,[22][23][24][25]. These indicators were defined using the WHO/UNICEF definitions for assessing IYCF practices [5]: • EIBF was defined as the proportion of children born in the last 24 months who were put to the breast within one hour of birth. • EBF was defined as the proportion of infants 0-5 months of age who receive breast milk as the only source of nourishment but are allowed oral rehydration solution, drops or syrups of vitamins, and medicines. • Continued breastfeeding at one year was defined as the proportion of children 12-15 months of age who are fed breast milk. • Introduction of solid, semi-solid, or soft foods was defined as the proportion of infants 6-8 months of age who receive solid, semi-solid, or soft foods. • MDD was defined as the proportion of children 6-23 months of age who receive foods from four or more food groups. The seven foods groups used for this indicator are: grains, roots and tubers, legumes and nuts, dairy products (milk, yogurt, cheese), flesh foods (meat, fish, poultry, and liver/organ meats), eggs, vitamin-A rich fruits and vegetables, as well as other fruits and vegetables. • MMF was defined as the proportion of breastfed and non-breastfed children 6-23 months of age who receive solid, semi-solid, or soft foods (but also including milk feeds for non-breastfed children) the minimum number of times or more (Minimum is defined as: two times for breastfed infants 6-8 months, three times for breastfed children 9-23 months, and four times for non-breastfed children 6-23 months). • MAD was defined as the proportion of children 6-23 months of age who receive a minimum acceptable diet (apart from breast milk).

•
Continued breastfeeding at two years was defined as the proportion of children 20-23 months of age who are fed breast milk.
• Predominant breastfeeding was defined as the proportion of infants 0-5 months of age who receive breast milk as the main source of nourishment but are allowed water, water-based drinks, fruit juice, oral rehydration solution, drops or syrups of vitamins, and medicines. • Bottle feeding was defined as the proportion of children 0-23 months of age who are fed with a bottle during the previous day.

Data Collection Process and Data Items
All articles identified in the search were exported into EndNote X8 and used for removing duplicates, screening, and selection. A three-step screening process was then employed. In the first screening phase, the first author (MVD) screened all publications by reading the titles. The second screening phase involved reading the abstracts of studies retained from the first screening phase, and eligible articles were retained. In the final screening phase, MVD read the full text of the remaining articles and retained studies that met the inclusion/exclusion criteria. All data extraction and appraisals of retrieved studies were independently reviewed by MVD and BA, and all disagreements between the two reviewers were resolved through discussion and consensus. A third reviewer FAO adjudicated the differences that emerged in the selection of the final studies for inclusion. The summary of the selected studies was recorded, and this included: author, year of publication, number of children/number of mothers, age of children, factors associated with IYCF indicators, and quality assessment score.

Quality Assessment
The quality assessment of the review was based on the assessment tools of the National Heart, Lung, and Blood Institute of the National Institutes of Health (NIH) for quality assessment of Observational Cohort and Cross-Sectional Studies and Controlled Intervention Studies [26]. The checklist consists of 14 items that evaluate the external validity (based on potential selection bias) and internal validity (based on potential measurement biases and confounding) of observational studies. After the initial assessment of all reviewed studies, the items were further collapsed into eight quality-appraisal criteria: sample size, sampling methodology, responses rate, outcome measures, statistical analyses, study limitation, ethical consideration, and control for confounding. Scores assigned to each reviewed study range from zero to 14 points (zero if none of the criteria were met and 14 points if all criteria were met). The sum of points awarded represented the overall quality of the study. Studies were rated as good (≥11), medium (6)(7)(8)(9)(10), and poor (≤5). A low-quality rating implies a high risk of bias in the study and vice versa. Research has indicated that the NIH checklist is a comprehensive tool for assessing the risk of bias in observational and experimental studies [27][28][29].

Results
A total of 6968 articles were retrieved from the six databases. A manual search of the bibliographic references of the retained articles identified 10 additional articles. After the removal of duplicates, 4537 articles were retained. The screening of the titles in the first screening phase resulted in the exclusion of 4323 articles. Further screening of the resulting 214 abstracts led to the exclusion of another 134 articles. In the final screening phase, the full texts of the remaining 80 articles were reviewed, and a further 39 articles were excluded. After the entire screening, 41 articles met the inclusion criteria and were thereby retained, as shown in Figure 1.
removal of duplicates, 4537 articles were retained. The screening of the titles in the first screening phase resulted in the exclusion of 4323 articles. Further screening of the resulting 214 abstracts led to the exclusion of another 134 articles. In the final screening phase, the full texts of the remaining 80 articles were reviewed, and a further 39 articles were excluded. After the entire screening, 41 articles met the inclusion criteria and were thereby retained, as shown in Figure 1.

Characteristics of the Study
Tables A1 and A2 demonstrates the summary of the studies included in this review. Of the studies conducted, nine studies were conducted at the national level, and 32 studies were conducted at the regional level. Sample sizes ranged from 77 mothers/children to 94,401 mothers/children. The criteria used to evaluate the quality of the included studies demonstrated that all the 41 studies were of medium quality. The details of the specific scores are provided in Supplementary Table S2. The extensive details of the studies have  been provided in Supplementary Tables S3-S12. During the search, three randomized control trials (RCTs) were found. However, the RCTs were not included in the review, as their sampling procedures, study design, methodology, implementation, and the quality assessment criteria were different from observational studies.

Characteristics of the Study
Tables A1 and A2 demonstrates the summary of the studies included in this review. Of the studies conducted, nine studies were conducted at the national level, and 32 studies were conducted at the regional level. Sample sizes ranged from 77 mothers/children to 94,401 mothers/children. The criteria used to evaluate the quality of the included studies demonstrated that all the 41 studies were of medium quality. The details of the specific scores are provided in Supplementary Table S2. The extensive details of the studies have  been provided in Supplementary Tables S3-S12. During the search, three randomized control trials (RCTs) were found. However, the RCTs were not included in the review, as their sampling procedures, study design, methodology, implementation, and the quality assessment criteria were different from observational studies.

Evidence from the Reviewed Studies
As shown in Tables A1 and A2, the most consistent factors associated with the IYCF indicators included (i) socioeconomic factors such as family characteristics such as marital status, socioeconomic status/standard of living, family type/size and access to media sources like newspapers, radio, and television; (ii) child characteristics such as sex, birth status (pre-term, term, post-term), perceived size of the baby at birth, preceding birth interval, and birth order of the child; (iii) maternal characteristics such as maternal age, maternal age at marriage, education/literacy level, employment status, power over earnings, power over household purchases, type of caste or tribe, and religion and parity; (iv) community-level characteristics such as place of residence (urban or rural); and (v) health service factors such as breastfeeding counselling, registration for antenatal care (ANC), number of ANC visits, place of birthing, type of birthing assistance, and mode of birthing.

Factors Associated with Predominant Breastfeeding Less than Six Months of Age
Srivastava et al. [25] demonstrated that factors associated with predominant breastfeeding included lower socioeconomic status, lower maternal and paternal education, fewer ANC visits (<3), and fewer TT vaccinations.

Factors Associated with Bottle Feeding, 0-23 Months
Patel et al. [24] found that the factors associated with bottle feeding included birthing assisted by non-health professionals, smaller birth size, higher socioeconomic status, higher media exposure, maternal employment, higher maternal education, and urban residence.

Discussion
The review showed that the most common factors associated with appropriate IYCF indicators were middle/higher socioeconomic status, frequent exposure to media, child gender (male), and higher birth order (≥2). Other common factors included maternal age, higher maternal education, employment status (housewife and employed), multiparity and a higher number of ANC visits (≥3), health facility birthing, and vaginal birthing.
The association between higher socioeconomic status and IYCF practices reported in this review is consistent with studies conducted in Bangladesh [65,66] and Pakistan [67,68]. Higher socioeconomic status and better media exposure [69] may translate into better awareness about appropriate IYCF practices, which may in turn influence a mother's decision to improve child-related health outcomes, including nutrition [70]. Our review also showed that male children and those with higher birth order (≥2) were more likely to be appropriately fed. Similarly, a study from Bangladesh [71] has found significant association between birth order and IYCF practices, while studies from Pakistan [72][73][74] and Nepal [75] have reported that higher birth order was associated with inappropriate IYCF practices, whereas a study from Sri Lanka [76] has found no significant association between birth order and IYCF practices. Higher birth order may reflect a more experienced mother in relation to appropriate infant feeding, as the mother may be more aware of what type of food to give to the child at every stage of growth and development [77]. In India, evidence suggests that there is an increasing desire for male children, and this male preference may have an impact on child health and development [78]. It is uncertain to what extent this cultural practice may be affecting infant and young child nutrition in the population, and future research may consider the impact of male preference on infant and young child nutrition in the country.
Previous studies from Pakistan, Bangladesh, and Nepal have reported that maternal characteristics such as higher maternal age (≥25 years) [67], higher maternal education [67,79], employment status [80][81][82], parity (≥2), and maternal autonomy in finance [83] were associated with appropriate IYCF practices. Similarly, our review showed that maternal characteristics (maternal age, maternal education, employment status, parity, and maternal autonomy) were associated with appropriate IYCF indicators. Globally, women empowerment indicators have been shown as a major determining factor for optimal child growth and development [80,81,84,85]. Higher maternal education increases women's opportunities for employment, household earnings, and autonomy; empowers the woman to make informed child health-related decisions such as the uptake of appropriate IYCF information; and improves the woman's attitude towards seeking appropriate child health support for appropriate IYCF [82]. While some studies have shown that women in employment have advantages of improving earnings/confidence and subsequent health-related decisions for IYCF [67,[86][87][88], other studies have indicated that "stay-at-home" mothers (housewives) also have advantages for appropriate IYCF [89,90]. Being a housewife allows the mother to have enough time and support for careful consideration of appropriate IYCF practices, and the mother is not distracted by external work activities compared to the mother in employment [89,90].
Globally, numerous studies from LMICs have shown that health service factors (including breastfeeding counselling in health facilities, ANC visits, and health facility birthing and normal vaginal births) are strongly associated with appropriate IYCF [91][92][93][94]. Consistent with past studies, the review showed that health service factors were associated with appropriate IYCF in India. Empirical evidence suggests that increased access to health services for women can improve many maternal and child health domains. These areas include access to relevant information and support for behavior change for both mother and child; better opportunities for making informed decisions about preventive maternal and child health measures; and possibly providing an entry point for better household decision-making process about child health [95].

Policy Implications of the Study Findings
In India, information relating to IYCF policy initiation and implementation has been extensively documented in previously published studies [2,4,17,18,35]. Briefly, we highlight various national policies and programs that are being implemented in India to improve Nutrients 2021, 13, 825 9 of 34 child nutrition. These policies are in support of the major IYCF programs introduced nationally to support the Integrated Child Development Scheme (ICDS) [96]. Efforts have been made to improve the coordination between the national and state level ICDS implementation through recommendations in the Twelfth five-year plans (2012-2017), such as organization "Village Health and Nutrition Days" to promote the uptake of appropriate IYCF practices [97]. However, the strategic decision to restructure the focus of ICDS to children under three years of age to target the increased uptake of key IYCF indicators to guide policies and laws still requires revisions to document any major impact on the nutritional status of children. Additional policies to support ICDS have been implemented, which include the Pradhan Mantri Surakshit Matritva Abhiyan (PMSMA) [98], Pradhan Mantri Matru Vandana Yojana (PMMVY) [99], LaQshya programme [100], HealthPhone [101], cash transfer schemes such as Janani Suraksha Yojna (JSY) [102], and recent amendments to the Maternity Benefit Act, 1961 (now known as the Maternity Benefit Amendment Act, 2017) [103]. The impacts of these initiatives across regional and national India are yet to be publicly documented, as current estimates of IYCF showed that these indicators are below expected levels in many regional areas [17,18]. Similarly, employed mothers still face significant resistance around workplace breastfeeding [104]. Streamlining India's IYCF policies and programs may likely have maximal impacts due to the positive impacts of policy-driven and targeted approaches for IYCF interventions employed in other contexts. For example, Sri Lanka now has a strong focus on policy directives on mass communication for maternal education and special consideration for working mothers around IYCF. This initiative has been shown to significantly increase the uptake of IYCF practices across all levels of the population [105].
Moreover, the improvement of breastfeeding and complementary feeding participation of Indian women would also require an increase in female education as articulated in the Sustainable Development Goals [106] as well as the pragmatic implementation of the World Breastfeeding Trends Initiative (WBTi) recommendations [107]. One of the recommendations includes the establishment of standard BFHI centers for promoting, protecting, and supporting breastfeeding. While the Government of India has established programs, including Mother's Absolute Affection (MAA) [108] and the Prime Minister's Overarching Scheme for Holistic Nourishment (POSHAN) Abhiyaan [109] to promote, protect, and support breastfeeding, intensifying efforts to improve the uptake of these programs is still needed [108].
Worldwide, husband/partner and family (e.g., grandmother) support for breastfeeding is essential for appropriate IYCF practices in the household [110]. With the appropriate knowledge, these key family members provide an emotional, psychological, and physical support system for new mothers [111]. A recent systematic review indicated that appropriate partner breastfeeding support (in terms of verbal encouragement to new mothers from their partners, assistance in preventing and managing breastfeeding difficulties, and/or assistance with household/child care duties) influenced new mothers' decision to initiate, continue, or cease breastfeeding in the early postnatal period [112]. The important role of grandmothers in influencing maternal IYCF decisions has also been documented in Malawi [113], Nigeria [114], and internationally [115]. In India, the new Home-Based Care for Young Child initiative introduced in 2018 aims to strengthen the nutritional levels of children through structured home visit counselling of the caregivers and mothers [116]. This initiative is focused on a community-based approach with family units as the center of care to improve the IYCF uptake of children. However, these policy interventions that seek to improve IYCF practices in India should also consider maximizing the aspects of a family support system to improve the IYCF uptake.

Strengths and Limitations
This systematic review is a comprehensive search of existing literature on the association between socioeconomic, demographic, and health service factors and IYCF practices in India to inform targeted policy interventions. The strengths of our systematic review lie in the exhaustive search through extensive databases utilizing broad search strings and having two independent reviewers undertake the study selection, reanalyzing the studies to be included in the review based on the inclusion and exclusion criteria through discussion and consensus, as well as quality assessment. However, the study has limitations. First, the review was limited to quantitative studies with the exclusion of qualitative studies to appropriately answer the research question. The inclusion of the qualitative studies would allow for triangulation of results and provide alternative explanations for the findings [117]. Future studies should be conducted to highlight the in-depth reasons for the varied patterns of IYCF feeding practices in India. Second, the exclusion of studies not written in English and those published in multiple centers across different countries (including India) would have limited our evidence, as those excluded studies may have had additional information. Third, there may have been a publication bias, given that grey literature was excluded. Fourth, most of the included studies were cross-sectional studies, and recall bias is potentially inherent in the findings due to the nature of the data collection. Finally, the evidence from this study may be limited, given the study design. Future experimental studies that investigate the association between the socioeconomic, demographic, and health service factors and IYCF practices in India may be needed.

Conclusions
In India, our review has shown that the factors associated with IYCF practices are multi-factorial. There is a need for a multi-sectorial strategy that hinges on both facilityand community-based approaches at the sub-national and national levels to improve IYCF practice in India. These public health measures should not only include IYCF education or counselling sessions for mothers, but should also involve other important hierarchical (socioeconomic, demographic, and health service) factors in the households and community to improve childhood feeding.     A clear temporal association between the study factors and EIBF cannot be established due to the cross-sectional study. There could also have been recall bias in the study. There could also have been measurement bias leading to an overestimation or underestimation of factors. The information on the study factors and outcome variable were based on self-reporting, and this is a source of recall or measurement bias, which could result in an overestimation or underestimation of the association between the study factors and EIBF.
Additionally, lack of assessment of unmeasured confounding factors could have also influenced the outcomes. A large number of demographic and clinical data were collected, which can influence the association of duration of EBF. The same interviewer collected all information, which reduces the inter observer bias. Advanced statistical methods were employed to analyze the association of socio-demographic and clinical correlates with EBF.
A causal relationship could not be established due to the cross-sectional nature of the study. There is a possibility of recall bias due to the nature of the reporting. Additionally, the population is hospital-based and does not represent the national population.   Temporal association between outcome and the study factors may not be established due to the nature of the study. There is a possibility of recall bias due to the nature of data collection. There could have been differential misclassification bias due to the nature of the factors.   A predesigned, pretested questionnaire was used The study findings could not be generalized to the national Indian population. The causality cannot be established due to the cross-sectional nature of the study; recall bias and misclassification bias could also be there. A predesigned, pretested questionnaire was used The study was done in public hospitals among mothers from low socio-economic groups and therefore cannot be generalized for all institutional deliveries.
The breastfeeding patterns can differ for home-delivered infants and institution delivered infants and hence cannot be generalized. The study could also not address the reason behind the low prevalence of exclusive breastfeeding in the study population. Pretested standardized questionnaire based on NFHS-3 was used to collect information on the mothers The study findings could not be generalized to the national population of India. The causality cannot be established due to the cross-sectional nature of the study; recall bias and misclassification bias could also be there.       Women with child of 6-17 months, low education, did not read newspaper, less power in household decision making, less frequent ANC visits (<6 to none), lower socio-economic status were negatively associated. West and North were negatively associated.
Women with child of 6-17 months, low education, did not read newspaper, less power in household decision making, less frequent ANC visits (<6 to none), lower socio-economic status were negatively associated. East was negatively associated.
The ability to determine the most susceptible age group and the modifiable factors that affect inappropriate practices in a large sample size, which allows for control of confounders. The sample is nationally representative. A pre-validated questionnaire was used.
The temporality cannot be established due to the cross-sectional nature of the study; recall bias and misclassification bias could also be there due to the method of data collection. Hospital delivery was positively associated Pretested, pre-validated questionnaire was used.
The temporality cannot be established due to the cross-sectional nature of the study; recall bias and misclassification bias could also be there due to the method of data collection. The study findings are not representative of the national population of India. Feed consistency was not taken into consideration for complementary feeding practices. Some of the questions asked were not open-ended. Additionally, the timescale over which the study was conducted was also a limitation.