Determinants of timely initiation of complementary feeding among mothers with children aged 6–23 months in Lalibela District, Northeast Ethiopia, 2015

Background Optimal complementary feeding alone prevents six percent of child mortality, but it has continued to be considered as sub-optimal in Ethiopia. Therefore, this study aimed to assess timely initiation of complementary feeding and associated factors among mothers with children aged 6–23 months in Lalibela District. Methods A community-based cross-sectional study was conducted from March 01 to April 29, 2015. Four hundred twenty-one mother-child pairs were selected by the systematic random sampling technique. An interviewer-administered questionnaire was used to collect data. A multivariable logistic regression analysis was employed to identify factors associated with timely initiation of complementary feeding. The Adjusted Odds Ratio (AOR) with a 95 % Confidence Interval (CI) was computed to assess the strength of association, and variables with a P-value of <0.05 were considered as statistically significant in the multivariable analysis. Results The study demonstrated that, the prevalence of timely initiation of complementary feeding was 63 %. In addition, mother’s education [AOR = 4.33, 95 % CI: 1.99, 9.40], antenatal care follow up [AOR = 5.90, 95 % CI: 2.45, 14.21], and institutional delivery [AOR = 2.54, 95 % CI: 1.33, 4.82] were found key determinants of timely initiation of complementary feeding. Conclusion In this community, timely initiation of complementary feeding was lower than the World Health Organization cut-off point for good practice of complementary feeding. Therefore, intensifying utilization of antenatal care and institutional delivery helps to improve the coverage of timely initiation of complementary feeding. Furthermore, the focus needs to be on uneducated women.


Background
Around the age of 6 months, infant needs for energy and micronutrients start to exceed what is provided by breast milk. They are developmentally ready to initiate additional (complementary) food, which is necessary to meet their extra energy and micronutrient requirement [1]. In addition, the transition period (6 months to 2 years) is part of the 'critical window of opportunity' to enhance the survival and optimal growth of the child [2]. Thus, the World Health Organization (WHO) recommends that mothers should initiate soft, semi-solid, or solid food to their infants at the age of 6 months [3].
Infants and young children bear the heaviest burden of undernutrition and continue to suffer from disability and death associated with it [4]. Globally, undernutrition results in 3 million child deaths annually and this amounts to 45 % of all causes of mortality. Over twothirds of these deaths are often associated with inappropriate feeding practice and occur during the first year of life [5,6]. Sub-optimal breastfeeding results in more than 800,000 deaths annually [6] and is also a significant determinant of childhood undernutrition [7,8]. On the other hand, optimal breastfeeding prevents 13 % of the deaths occurring in children under five, and appropriate complementary feeding results in an additional six percent reduction [9].
Nevertheless, global complementary feeding practice has been sub-optimal. Among South Asian countries, the rate of timely initiation of complementary feeding is lower than the WHO recommendation for good practice (80-94 %) [10]. In this regard, about 71 %, 70 %, 55 %, and 39 % of the infants in Bangladesh, Nepal, India, and Pakistan, respectively, are reported to have timely initiation of complementary feeding [8,[10][11][12]. On the other hand, there is a low rate of timely initiation of complementary feeding in Africa [1,13]. In Ethiopia, more than half (57 %) of the child mortality occurs mainly due to undernutrition [14], and the majority of the children have had suboptimal feeding practices. Only51% of the infants aged 6-9 months receive complementary food [15]. Furthermore, studies from different regions of the country show a low rate of timely initiation of complementary feeding (52.8-62.8 %) [16][17][18].
The determinants of timely initiation of complementary feeding vary between settings mainly depending on the level of health care utilization and sociodemographic characteristics. Reports from different countries reveal that, child sex, mothers wealth status, marital status, maternal and paternal education, maternal age (≥30 years), exposure to media, and knowledge about the right time for initiation of complementary feeding [11,[18][19][20][21][22][23], Antenatal Care (ANC) follow up, postnatal care, and institutional delivery [8,11,18,20,21,23] are the commonly reported determinants of timely initiation of complementary feeding.
In order to reduce the high burden of child malnutrition and mortality in Ethiopia [15], ensuring appropriate Infant and Young Child Feeding (IYCF) practices is of vital importance. The country has implemented the IYCF strategy for a decade [24], so, studies showing complementary feeding practices have a paramount significance in evaluating the progress of interventions aiming to address inappropriate child feeding practices. However, such studies are scarce in Ethiopia, particularly in the northeastern part of the country. Therefore, this study aimed to assess the timely initiation of complementary feeding and its determinants among mothers with children aged 6-23 months in Lalibela District, northeast Ethiopia.

Study design and setting
A community-based cross-sectional study was conducted from March 01 to April 29, 2015, in Lalibela District, northeast Ethiopia. The district is one of the historical tourist destinations registered by the United Nations Educational, Scientific, and Cultural Organization (UNESCO) as a world wonder site. Lalibela District, which is 700 km from Addis Ababa, the capital city of Ethiopia, has two urban and one rural kebeles (smallest administration units in Ethiopia). According to the 2014/15 District Finance and Economic Development Office annual statistical report, the total population of the district was 31,491, of which children aged 6-23 months comprised 4.39 % (1065). The total number of households in the district was estimated at 7323. At the moment, one district hospital, one health center, two urban and one rural health posts were providing health services to the community.

Sample size and sampling procedure
All mothers with children aged 6-23 months and have lived in the three kebeles of Lalibela District for at least 6 months were eligible for the study. The sample size was determined using single population proportion formula by considering the following assumptions: 52.8 % as a proportion of mothers practicing timely initiation of complementary feeding [18]; 5 % margin of error, and 95 % confidence level. In addition, a 10 % non-response rate which gives the final sample size of 421was anticipated. Information regarding the total number of households with eligible children was obtained from the District Health Office. The number of households sampled from the three kebeles of the district was proportionate-to-population size. A systematic random sampling technique was used to select individual households with eligible children. A sampling interval was calculated by dividing the total number of mothers who had children aged 6-23 months during the study period by the allocated sample size. For households with multiple children, one child was selected using the lottery method.

Data collection tools and procedures
A structured interviewer-administered questionnaire was used to collect data. To maintain consistency, the questionnaire was first translated from English to Amharic (the native language of the study area) and was backtranslated to English by professional translators. Six female clinical nurses and two health officers who were not working in the actual study area were recruited as data collectors and supervisors, respectively. Training was given to data collectors and supervisors for 2 days on basic skills of interview, ways of obtaining consent, and on how to maintain confidentiality of information. The investigators oversaw all data collection activities. The completion, accuracy, and clarity of the collected data were checked carefully on daily basis. A pre-test was administered on 5 % (21) of the sample out of the study area. During the pre-test, the acceptability and applicability of the procedures and tools were evaluated, but the result of the pre-test was not included for analysis.
Complementary feeding practices were assessed according to the key indicators recommended by WHO [3]. Accordingly, the outcome variable (timely initiation of complementary feeding) was confirmed by asking the mothers to recall the age at which they initiated additional food to [child's name], and were asked "When did you first introduce any solid, semi-solid, or fluid to [child's name] in addition to breast milk". Then, if the mother answered "at the sixth month", it was categorized as timely initiation of complementary feeding and coded as "1", but if she had initiated beyond (before or after) the sixth month, it was considered as untimely initiation of complementary feeding and was coded as "0". In addition, a single visit was made to gather data regarding complementary feeding practices and other characteristics. For a few mothers who faced difficulties in remembering the right time of initiation of complementary feeding, data collectors carried out different probing mechanisms to help them recall, thereby minimizing recall bias. Relating the time of initiation to known public events, occurrences of common childhood developmental milestones, and immunization schedules were some of the probing techniques used by data collectors. Moreover, to minimize the social desirability bias, data collectors created a conducive environment by keeping mothers apart and making them comfortable during data collection.
The standardized Dietary Diversity Score (DDS) tool with a 24 h recall was used to qualitatively assess the dietary intake of children. Mothers were asked to list the food items consumed by the children in the previous 24 h preceding the date of survey. The food items were categorized into seven food groups, such as starchy staples (grains, roots, and tubers), legumes and nuts, dairy products, flesh food (meat, fish, poultry, and organ meat), egg, vitamin-A rich fruits and vegetables, and other fruits and vegetables. By considering the minimum acceptable DDS, a child having a DDS of ≥4 was categorized as good dietary diversity, but if it had a DDS of <4, it was considered as poor dietary diversity [3].
The independent variables included in the study were mainly related to the parents' socio-demographic and economic characteristics (age, education, employment, household wealth, religion, household size, and marital status), and health-care related information (ANC, place of delivery, postnatal care, frequency of health extension visits, and mothers' knowledge and source of information about the right time for initiation of complementary feeding). Accordingly, to determine their knowledge about the right time for initiation of complementary feeding, the mothers were asked such a key question as "What is the appropriate time to start additional food for a child in addition to breastfeeding?". If the mother said "at the sixth month", she was considered as having accurate information about the right time for initiation of complementary feeding; otherwise, she was deemed to have inaccurate information. The Household Wealth Index was computed using a composite indicator by considering properties, such as livestock ownership, selected household assets, and size of agricultural land. Principal Component Analysis was performed to categorize household wealth status into poor, medium, and rich.

Data analysis
The collected data were entered into the EPI-Info version 7 software, and transferred, cleaned, coded, and analyzed using Statistical package for Social Science (SPSS) version 20. Descriptive statistics, including frequencies, proportions, means, and standard deviations were used to summarize the variables. A binary logistic regression was used to identify the determinants of timely initiation of complementary feeding. A bivariable analysis was done to show the crude effect of each independent variable on the outcome variable. Variables with a P-value of <0.2 in the bivariable analysis were entered to a multivariable logistic regression analysis. Stepwise backward Likelihood Ratio (LR) was used for multivariable logistic regression. The Adjusted Odds Ratio (AOR) with a 95 % Confidence Interval (CI) was computed to assess the strength of association, and a P-value of <0.05 was used to declare the statistical significance in the multivariable analysis.

Socio-demographic and economic characteristics
A total of 421 mother-child pairs were included in the study. The mean age (±Standard Deviation, SD) of the children was 14.87(±4.73) months. The majority of the respondents (95.2 %) were Orthodox Christians and 83.4 % of them were married. All of the participants were Amhara by ethnicity. More than half (52.3 %) of the children were male, and nearly one-third (27.8 %) were in the age range of 6 to 11 months. About 28.5 % of the mothers had no formal education, and more than half (59.9 %) were housewives (Table 1).  (Table 2).

Complementary feeding practices
The prevalence of timely initiation of complementary feeding was 63 % [95 % Cl: 58.0, 67.5 %]. One hundred eighteen children (28.0 %) started complementary feeding after sixth months, while twenty-eight (6.7 %) were initiated before they reached the sixth month. In addition, ten children (2.3 %) were offered no additional food since they have been born. Nearly two-thirds (58.9 %) of the mothers had accurate information about the right time of initiation of complementary feeding. The majority (87.6 %) of the children had poor dietary diversity, and below half (45.1 %) ate three times a day (Table 2).

Factors associated with timely initiation of complementary feeding
In the bivariable analysis, maternal education, occupation, marital status, wealth status, paternal education and occupation, possession of television, ANC follow up, place of delivery, postnatal care, and health extension visits in the past 6 months were significantly associated with timely initiation of complementary feeding.
However, the result of the multivariable analysis showed that maternal educational status, ANC followup, and place of delivery were significantly and independently associated with timely initiation of complementary feeding. Accordingly, the likelihood of timely initiation of complementary feeding among children whose mothers attended primary school [AOR = 3.39, 95 % CI: 1.52, 7.54] and secondary school and above [AOR = 4.33, 95 % CI: 1.99, 9.40] was higher compared  (Table 3).

Discussion
The result of this study revealed that the prevalence of timely initiation of complementary feeding was 63 %. This finding was lower than the WHO cut-off point (80 % to 94 %) for good practice of complementary feeding [10]. Similarly, it was lower than the reports from Abyi-Adi, Ethiopia (79.7 %) [25], India (77.5 %) [26], and Bangladesh (71 %) [27]. The higher prevalence of timely initiation of complementary feeding in the latter study areas could be related to the improvements in utilization of ANC and institutional delivery. Hence, nutrition education and counseling are components of maternal health care services; a higher utilization of these services will bring an added benefit to improve mothers' awareness on appropriate child feeding practices [25]. However, our finding was higher than that of the 2011 Ethiopian Demographic Health Survey report (51 %) [15] and the report from northern Ethiopia (52.8 %) [18]. This is probably related to the current improvements in the implementation of the Health Extension Program. Health extension workers are making home to home visits on regular bases to support families in accessing basic health services and to give home-based health education as well as other promotion services, including promotion of appropriate IYCF.
Compared to the current finding, a lower prevalence of timely initiation of complementary feeding was reported from Nigeria (41 %) [13] and India (55.7 %) [12]. This discrepancy could be attributed to differences in sample sizes between the two study areas, as more mothers were surveyed in India than in this study. Though the sample size was lower in Nigeria, a lower practice of timely initiation of complementary feeding could be related to the variation in mothers' working environments; in Nigeria, most of the mothers were working outside the home, while most mothers were housewives in Ethiopia. It   was affirmed that mothers working outside their home were more likely to return to work before they exclusively breastfed their infants for the recommended duration, 6 months [28][29][30]. As a result, they would be forced to initiate complementary feeding to their infants earlier than housewife mothers. The result of the adjusted analysis showed that the odds of timely initiation of complementary feeding were higher among educated mothers as compared to their uneducated counterparts. Parallel findings were reported by studies elsewhere [16-18, 25, 31]. This could be due to the fact that education has an immense benefit in promoting the optimal maternal nutrition and health seeking behavior [15]. In addition, other studies persistently claimed that mothers' education has a profound effect on children's nutritional status and caring practice [15,[32][33][34].
Higher odds of timely initiation of complementary feeding were observed among mothers who visited ANC as compared to those who did not. This finding was in agreement with other reports in Ethiopia [16][17][18]25]. Obviously, pregnancy has been considered as an important window of opportunity to deliver nutrition counseling and education on IYCF [35]. Nutrition education and counseling have a pivotal role in enhancing mothers' IYCF knowledge and practice and maternal self-efficacy in child care [36][37][38].
Mothers who delivered at health institutions were more likely to initiate complementary feeding at the right time (sixth month) as compared to those who gave birth at home. The finding was supported by reports elsewhere [16][17][18][19][20][21][22][23][24][25][26][27][28]. Mothers who give birth at health institutions have a better opportunity to access appropriate child feeding information, which ultimately improves their capacity to challenge unfavorable attitudes of the community [35,39]. The previous reports also claimed that institutional delivery was found to increase the likelihood of exclusive breastfeeding for 6 months, indicating timely initiation of complementary feeding [28,40,41]. On the other hand, home birthing is associated with inappropriate neonatal feeding practices, such as discarding colostrum and giving prelacteal feeds [42][43][44].
The investigators made a lot of effort to maintain the quality of the data, mainly through a pretest, frequent field supervisions, and training of data collectors. However, the study was not free from some limitations. First, since the estimation of some of the variables, including health extension visits, was made through recall (with the longest recall period of 6 months), there might have been a risk of recall bias. Secondly, the study did not consider the measurement of some of the independent variables, like type of nutrition services received during ANC and postnatal care visits, and mothers' attitude towards appropriate breastfeeding and complementary feeding which might confound the result of the study. Thirdly, the study was not free from the pitfalls of a cross-sectional study design.

Conclusion
This study revealed that, the prevalence of timely initiation of complementary feeding was low; it was by far lower than the WHO cut-off point for good practice of complementary feeding. Mother's educational status, ANC follow up, and institutional delivery were significantly associated with timely initiation of complementary feeding. Therefore, current efforts should be strengthened to improve optimal complementary feeding practices by stepping-up the utilization of ANC and institutional delivery. The focus needs to be on uneducated women. Furthermore, future studies should emphasize mixed methods, such as triangulating with a qualitative study design and prospective study designs to identify key barriers of timely initiation of complementary feeding.