Adherence to complementary feeding guidelines among caregivers of children aged 6-23 months in Lamwo district, rural Uganda

Introduction Malnutrition contributes to half of all deaths among children under-five years in developing countries such as Uganda. Optimal complementary feeding is one of the crucial interventions that could prevent these deaths. This study measured adherence to complementary feeding guidelines and its associated factors among caregivers of children aged 6-23 months in Lamwo district, rural Uganda. Methods A household cross-sectional study was used to collect data on adherence to complementary feeding among 349 caregivers. A composite variable with 9 indicators of complementary feeding was used to measure adherence. Univariable and multivariable logistic regression was used for statistical analysis using STATA software. Results A household cross-sectional study was used to collect data on adherence to complementary feeding among 349 caregivers. A composite variable with 9 indicators of complementary feeding was used to measure adherence. Univariable and multivariable logistic regression was used for statistical analysis using STATA software.nearly all (97.7%, 341/349) children had ever been breastfed. Complementary feeding was initiated at six months for 47.0% (164/349) of the children. The number of complementary meals ranged from 1-4 meals per day with a mean of 3 meals per day (SD = 0.8). About half (55.8%, 195/349) of the children were given less than the recommended amount of food. Overall only 40.1% (140/349) of all study respondents were adherent to complementary feeding guidelines. The odds of adherence to complementary feeding were higher among caregivers with children aged 6-8 months (AOR = 4.68, 95% CI: 1.91-11.48), children whose fathers had attained 8 or more years of formal education (AOR = 2.27, 95% CI: 1.22-4.19), caregivers with two children under five years (AOR = 5.46, 95% CI: 1.46-20.36), those living in the poorest households (AOR = 3.00, 95% CI: 1.37-6.57) and those who showed willingness to recommend initiation of complementary feeding at six months to another mother (AOR = 1.34 95% CI: 1.06-1.70). Conclusion Adherence to complementary feeding guidelines was very low in this rural African setting indicating an urgent need for interventions such as health education to improve adherence with consequent reduction in rates of under nutrition. These interventions should target caregivers with older children, fathers with less than 8 years of formal education and those living in the wealthiest households.


Introduction
Globally child survival continues to be a priority since it was estimated in 2015 that 5.9 million children under five years died from preventable diseases [1]. Half of these deaths resulted from diseases related to under-nutrition. In Uganda, over 33% of children under-five years are stunted, a situation which is classified as a serious public health problem according to the World Health Organization (WHO) [2]. Adequate infant and young child feeding can improve child survival and development with evidence showing that applied breastfeeding and complementary feeding practices save the lives of more than three million children under-five years annually [3]. Nutritional interventions in the first two years of a child's life are particularly important, as adequate nutrition during this period leads to reduced morbidity, mortality and to overall better development. Under-nutrition at this early age also entraps societies in the "cycle of poverty" [1,3,4] with considerable financial loss to national economies of up to 3.5 trillion dollars per year globally and 25 million dollars every year in Sub-Saharan Africa alone [5]. Malnutrition is a complex phenomenon that stems from various underlying determinants, including a lack of optimal feeding practices for infants and young children [4,6]. According to the "Ugandan policy document on infant and young child feeding of 2004", complementary feeding should start at 6 months with continued breastfeeding until two years of age or beyond. In addition, complementary food should be of correct consistency and nutrient dense [7]. However, foods in the complementary feeding period are often introduced too late for most children or they receive less than the recommended meals per day [8,9]. This low adherence to complementary feeding guidelines means that many children continue to be vulnerable to under-nutrition including irreversible outcomes of stunting, poor cognitive development and significantly increased risk of infectious diseases [1,5]. Lamwo district was selected for this study because of its reportedly high level of stunting of 41%, which according to the WHO classification makes it a critical public health problem. According to a nutrition surveillance report 2011, Lamwo district reported the highest levels of severe acute malnutrition among five Northern districts in Uganda. In addition the district also reported the lowest prevalence for infant feeding indicators [10]. Interventions such as targeted supplementary feeding, deworming, Vitamin A supplementation and education on infant and young child feeding guidelines have been implemented however, high levels of malnutrition persist [6,7]. It is unclear why this is so, thus we measured adherence to complementary feeding guidelines and its associated factors among caregivers of children aged 6-23 months in Lamwo district so as to inform implementation of the policy on Infant and Young Child Feeding.

Methods
Study design and setting: We used a cross sectional study design to collect data on adherence to complementary feeding (IYCF) Guidelines among mothers and caregivers of children aged 6-23 months of age in Lamwo District in April 2015.
Sample size and sampling procedure: We estimated a sample size of 350 using the formula for cluster surveys by Bennett et al [11] with the following assumptions: prevalence of appropriate complementary feeding in Uganda 28% [12], precision of 0.03, design effect of 1.6, intra-cluster variability of 0.1 [11], estimated number of respondents in a cluster 7. We randomly sampled 50 villages out of a total of 358 villages in the district [13] and randomly selected seven households from each village using a sampling procedure adapted from the WHO reference manual for cluster surveys 2005. At the household level, we identified children aged 6-23 months using a child health card or from recall.

Measurements:
We measured Adherence to complementary feeding using nine indicators for adequate complementary feeding as adapted from the field manual by UNICEF and WHO [14].
Complementary feeding was defined using the following indicators: correct Time of start, correct amount of food, dietary diversity, correct meal frequency, meal consistency, hygiene practices, responsive feeding and correct feeding during illness. All the nine variables under complementary feeding were scored as either 1 or 0 to indicate adequate or inadequate practice respectively. We generated a composite variable with a maximum score of 9; any participant that scored ≥6 was categorized as adherent to the guidelines and any participant that scored ≤ 5 was considered as non-adherent [15][16][17]. We measured attitude towards complementary feeding on an ordinal scale using the following of the items as having little or no control. We developed a household wealth index by using principal components analysis with variables on asset ownership. Regression factor scores generated from the first principal component were ranked in ascending order and then categorized into quintiles (1) poorest, to (5) least poor.

Data collection and management:
We collected data using a structured questionnaire that was translated into Luo (the local language commonly spoken in Lamwo). We trained five data collectors for two days on sampling and interviewing techniques prior to the survey.

Data entry and analysis:
We entered data into excel spread sheets and analysed using STATA software version 12. Statistically significant response patterns were considered if a two-sided p-value was <0.05. We calculated means and proportions for socio demographic characteristics and feeding practices. All variables with a p-value < 0.2 at bivariable analysis were considered in multivariable analysis using the forward stepwise multivariable logistic regression model. We also used Adjusted odds ratios (AOR) and 95% Confidence intervals to identify independent predictors of adherence to feeding guidelines.  [6]. An earlier study indicated that although adherence to feeding guidelines was higher among HIV positive Ugandan mothers who had received infant feeding counselling, however, infant feeding counselling was rare in this setting [19]. A subsequent study showed continued low level of infant feeding counselling in this setting [20]. It is critical that effective interventions such as infant feeding counselling are rolled out across all Ugandan health facilities and communities in order to improve adherence to guidelines. In our study, the caregiver's age was not statistically associated with adherence to complementary feeding guidelines similar to findings from South India in a study done to establish complementary feeding practices among mothers of children aged 6months to 2 years [8]. In contrast, an Australian study that used a complementary feeding index found an association between older maternal age and adherence to complementary feeding [15]. The difference in findings could be due to the fact that Australian mothers were better educated than the Ugandan mothers. The highly prevalent lower education among our study participants may partly explain why we did not find a statistical difference in adherence to complementary feeding guidelines due to maternal level of education. In other settings, maternal education has been found to correspond to adherence to feeding guidelines and in these settings higher proportions of adherence to feeding guidelines are reported among mothers with more than 7 years of formal In our study the child's age (6-8 months) was associated with higher odds of adherence to complementary feeding guidelines similar to findings from urban slums in India [23]. However, a cohort study in a Nigerian hospital found that complementary feeding was highest in the age group 9-11 months in comparison to other age categories [24]. This difference in findings might be due to the fact that the Nigerian study was conducted in a hospital essentially introducing selection bias unlike our study, which was conducted at the household level. Furthermore, having two children less than five years was a predictor for adherence to complementary feeding in this study similar to findings from a high-income setting [25]. Although it may be difficult to completely decipher the reasons for this observation, it is possible that mothers with more than one child have high levels of infant feeding related knowledge since they have past experience with infant feeding [26]. Living in the poorest households was associated with higher odds of adherence to complementary feeding guidelines contrasting findings from India where higher income was associated with increased adherence to infant feeding guidelines [17,22]. Besides the differences in the feeding cultures between the two study populations, poverty was highly prevalent among our study participants and most caregivers reported that they were unemployed. Previous child health research in Uganda shows that living in the poorest households is a predictor of non-adherence to child health guidelines which is a reflection of the prevailing health system failure that may not deliver information adequately to all populations [27,28]. Furthermore, it is expected that those living in the less poor households should adhere to feeding guidelines since they can better afford to purchase the requirements for adequate nutrition of the infants and young children. However, our study population were mainly unemployed which means that mothers spent more time caring for their children unlike mothers who were employed, this could explain why those in the poorest households were more likely adhere to feeding guidelines. The index generated in our study is unlike other complementary feeding indices such as the Complementary Feeding Utility Index (CFUI) which provides a summary score that encompasses nutritional, developmental, and behavioural complementary feeding guidelines [15]. Our findings were lower than other study reports that used composite variables in Australia and Burkina Faso [15,16]. However, those studies used fewer indices in their measurements. In interpreting the findings from our study, some limitations of the study design should be considered.
First, a differential error in measurement could have arisen if the respondents that were not adherent to complementary feeding guidelines were less likely to report sub-optimal complementary feeding practices because of perceived unacceptability of this behaviour particularly consequent to previous advice on child feeding from health workers. Second, a study with a larger sample size would narrow the wide confidence intervals observed in our study. Lastly, we assessed feeding behaviour at one point in time.
This overlooks the varying nature of feeding behaviour as the child grows, current feeding practice does not necessarily predict previous or future feeding patterns. The survey was conducted