Association of Complementary feeding practices and household food insecurity and anthropometric status of children aged 6-23 months old in Kabul city


 Introduction

During the first 1000 days of a child life, appropriate infant feeding practices are fundamental to growth, health and development of a child. Globally only one in six children receive a minimum acceptable diet. In Afghanistan, the status of minimum meal frequency, dietary diversity and acceptable diet were 55%, 23% and 18%, respectively among children aged 6-23 months. This study designed with the aim of determines the association of complementary feeding (CF) indicators and household food insecurity and anthropometric status of Afghan children aged 6-23 months.

Methods

We were selected 300 children aged 6-23 months old in this cross-sectional study through simple method. Three 24-hour dietary recall were collected for CF practices. Socioeconomic and United State Department of Agriculture (USDA) food security questionnaire were used from interview method and nutritional status was assessed using World Health Organization indicators. Chi-Square analysis were used to determined association between anthropometry of children among different age groups, complementary feeding indicators and household Food insecurity. Independent sample test was used to determined association between nutrient intake and complementary feeding indicators.

Results

Results showed that 59% of children received timely introduction of CF and Minimum Meal Frequency (MMF), Minimum Dietary Diversity (MDD) and Minimum Acceptable Diet were met by 87.7%, 44.7% and 42.3%, respectively. Prevalence rate of food insecurity was 90.7%. Result for odds ratio and 95% CI showed there is association between (Length for Age Z-score (LAZ) with MMF), (Weight for Age Z-score (WAZ), Weight for Length Z-score (WLZ) and Length for Age Z-score (LAZ) with MDD and MAD). Result for odds ratio and 95% CI also showed there are significant associations between WAZ, WLZ and LAZ with household food insecurity.

Conclusion

Household food insecurity and weak complementary feeding make more prevalence of malnutrition.

Introduction months old in our plan but, we were added those children who were came with their mothers and were came for anthropometric measurement in to our study. The samples were selected by simply method with the following inclusion criteria: 1) Children aged 6-23 months, 2) the mothers who were agreed to interview, 3) the children who were apparently healthy, not experienced chronic/congenital diseases, such as heart abnormality and not having experienced acute conditions such as fever, diarrhea and respiratory infection in the two previous weeks.
For sampling, we were selected our samples according to population size of those regions or hospitals/health centers which covered them.

Data collection
For data collection rst of all, we were collected information about nutrient intake, dietary diversity and meal frequency of children by using the 24-dietary recall from three days of a week and quantitative data, socioeconomic and demographic characteristics of children were collected using structured questionnaire in face to face interviews between mothers of children and interviewers. According to our objectives we were add 18-items USDA Food Security questionnaire at the end of general questionnaire too. We were included the following variables for a child: sex, age, birth order, birth interval, weight, recent vitamin A and iron supplementation status, vaccination record, and recent symptoms of diarrhea, fever and cough. For mothers we were included these variables: age, smoking status, use of reproductive health services: delivery at health facility, delivery with skilled birth assistance, caesarean delivery, number of antenatal care visits, timing of postnatal check-up for mother and for child, education level, occupation. For fathers, we were included age, education level, and occupation. For Household characteristics we were included: household head, household size, number of children under ve, type of cooking fuel, water source, toilet conditions and sharing, and household wealth quintile and time to get to water source.

Procedures
The four main complementary feeding indicators are: timely introduction of CF, MMF, MDD and MAD which we were got required information about timely introduction of CF via question in the questionnaire and MMF, MDD and MAD via 24-hour dietary recalls of food and liquid consumption during the previous day of the survey [11]. We were compared these CF indicators with four anthropometric measurements which they were Mid Upper Arm Circumference (MUAC), Weight for Age Z-score (WAZ), Length for Age Z-score (LAZ) and Weight for Length Z-score (WLZ) and with household food security. For prevalence of MDD we were used 24-hour dietary recall result, We were categorized dietary diversity as met when the child consumed 4 out of the bellow 7 food groups: cereals and tubers, legumes, dairy products (milk, yoghurt, and cheese), esh food (meat, sh, poultry, and organ meat), eggs, vitamin A-rich fruits and vegetables, and other fruits and vegetables [12,13]. We were got MMF information (frequency of meal and snack) through last 24-hour recall too. Recommended standard for infants who was only breastfeed, meal for children aged 6-8 months was 2 times/day or more, meal of children aged 9-23 months was 3 times/day or more and non-breastfeed children's meal was 4 times/day or more. MAD were good for those children aged 6-23 months who received the MDD and the MMF in the previous 24 hours (day or night). For breastfeed children, MAD were achieved if the child meets both the MMF and MDD criteria and for non-breastfeed children, the child was required to receive at least four food groups excluding dairy products, two milk feeds and MMF [12,13]. We were had some observation measuring like MUAC which were measured on the left arm, at the midpoint between the elbow and the shoulder. The arms were relaxed and hanging down the side of the body during measuring MUAC. A MUAC measuring tape were placed around the arm. The value was road from the window of the tape without pinching the arm or leaving the tape lose [14]. We were road value of MUAC tape and were categorized them as: ≥11.5 severe acute malnutrition, 11.5< to ≥12.5 moderate acute malnutrition and >12.5 were in normal range [15]. Body weight and length measurement were collected to determine the nutritional status of children, we were used Sec-a digital weight scale with a precision of 0.1 Kg for measuring of body weight and for measuring of length, we were used an infant length board with a precision of 0.1 cm. We were categorized nutritional status like this: underweight was de ned as a weight for age z-score less than -2 standard Deviation (SD), wasting as weight for length z-score less than -2 SD and stunting as length for age less than -2 SD. This analysis was done with using WHO Anthro 2005 v.2.0.4 software [16].

Data management and analyzed
First of all, we were change amount of received foods to portion size and after that we were change portion size to gram. Secondly, we had data entry and were transferred food grams in to Nutritionist IV (a diet analysis module) version 3.5.2 for analysis. This program was changed food grams to nutrient intake with particular size. Collected data like quantitative data, socioeconomic and demographic characteristics of children and 18-items USDA food security were analyzed after editing, coding and entry. We were used IBM SPSS Statistics version 24 for analysis of this data. Crosstabs and Chi-Square analysis were used to determined association among anthropometric measurement of children with different age groups and different complementary feeding indicators. This analysis was used to determined association among household Food Security with complementary feeding indicators too.
Independent sample test was used to determined association between nutrient intake and complementary feeding indicators. Logistic regression was used to determined association between anthropometric measurement of children with complementary feeding indicators and household food insecurity which output were presented as adjusted odds ratios (AOR) with 95% con dence intervals (CI). Table 1 present demographic and socioeconomic characteristics of subjects. After analyzing data of this survey about half 49.3% of children age group were between 12-23 months. More than half 54.3% of children were female. Less than half 38.3% of children were second to fourth births. More than half 62.3% of children had normal weight and 27.7% were low birth weight at birth. Birth interval of 27.3% of children were within a 24-months. More than twothird 92.3% of children had completed all the vaccinations scheduled for their age. More than half 56.7% of the mothers were aged 25-35 years and just 2% of them were smoker. More than two-third 83% of mother were used reproductive health services, 89.7% of them were delivered at health facility, 90% of them delivered their children with assistance of a midwife or a doctor and 15% of them had caesarean delivery. Our result was showed 43% of mothers were used four or more antenatal check-up services and only 11.7% of them were used two or more postnatal check-up services and 39% of children were used two or more postnatal check-up services. Among these 300 samples 60% of mothers were illiterate and jobs of most of them 88.7 were housewife and more than tree-fourth 76.7% of them were not exposure to none of medias. Results were showed 29.3% of the father were aged ≥36 years and 39.3% of them were illiterates and 29.3% of them were labor. Household head of 71.3% of participants were father. Mean of household size were 6.72 which 37% of them were ≥7 persons in a household and mean of children under ve were 1.59 which only 8% of household had ≥3 children under ve. Type of cooking fuel of 91.7% of household were LPG, 52.3% of household water source were shaft which, around half 48.3% of them were used from unimproved source of drinking water. Source of drinking water of 87.7% of them were in their own yard and 84.71 minutes were their mean of time to get water source. More than one-thirds 40.3% of the households did not have improved toilet facilities. More than half 59.0% of households were poor. Association of household food security with complementary feeding indicators Result of table 3 showed there is no association between food security (secure and total insecure) with MMF (P= 0.139) and introduction of complementary feeding practices at six months (P= 0.846) but there are associations between food security with MDD (P= 0.003) and MAD (P= 0.004). Table 4 showed the association of food security and insecurity with their sup groups (Food insecure without hunger, food insecure with moderate hunger and food insecure with severe hunger) among CF indicators. The results were the same as Table of 3 with different p-value. There is no association between household food security with MMF (P= 0.102) and timely introduction of CF (P= 0.198) but there are associations between household food security with MDD (P= 0.0001) and MAD (P= 0.0001).

Sample characteristics
Result for odds ratios and 95% con dence intervals for low MUAC, low WAZ, low WLZ and low LAZ among those who were food insecure are shown in Table 8. This Table showed  CF practices of children aged 6-23 months Results for CF practices of children aged 6-23 months are shown in Fig. 2. This gure and Table 1 show 87.7% of children met MMF, 44.7% of children met MDD and MAD of 42.3% of them were good. When children met MMF and MDD and start complementary feeding at sixth month of their birth so they had appropriate complementary feeding, in this study 43.5% of children had appropriate complementary feeding. Results show 53.7% of children were breastfeed until 1 hour after birth, 72.3% of them had exclusive breast milk until the age of six months, 78% of them were stilled breastfeed until the age of two years old or more and 59% of them were started CF at beginning of sixth months. The results of study also showed only 8.3% of children were received Iron or multivitamin supplement and 11.3% of them were consumed of forti ed food in the last three months. Food groups consumption were as follows: meat ≥ 3 time/week 2.7%, eggs ≥ 4 time/week 4%, sh ≥ 4 time/week 0%, legumes ≥ 3 time/week 19.7%, Vegetables ≥ 4 time/week 13%, fruits with MDD and also LAZ were not associated with MDD (P=0.106). In the end also MUAC were not associated with MAD (P= 0.136), but WAZ (P=0.0001) and WLZ (P=0.0001) were associated and LAZ were not associated with MAD (P=0.058) too.
Result for odds ratios and 95% con dence intervals for low MUAC, low WAZ, low WLZ and low LAZ among those who did not meet the complementary feeding criteria are shown in Table 9. This Table shows Table 2 shows, mean and SD of vitamin C intake among children aged 6-23m, 6-8m, 9-11m and 12-23m were 8.68±9.21, 9.62±14.00, 8.36±7.51 and 8.47±7.52 mg/d, respectively. Mean and SD of vitamin B1 intake among children aged 6-23m, 6-8m, 9- Food group consumption by age of children All food groups consumptions were calculated as time per week and they are: Mean and SD of red and chicken meat consumption among children aged 6-23m, 6-8m, 9-11m and 12-23m were 0.63 ± 0.75, 0.40 ± 0.57, 0.40 ± 0.59 and 0.87 ± 0.84, respectively. Mean and SD of Legumes consumption among children aged 6-23m, 6-8m, 9-11m and 12-23m were 1.69 ± 1.41, 0.58 ± 0.75, 1.47 ± 1.20 and 2.31 ± 1.44, respectively. Mean and SD of vegetables consumption among children aged 6-23m, 6-8m, 9-11m and 12-23m were 1.65 ± 1.66, 0.50 ± 1.03, 1.54 ± 1.52 and 2.20 ± 1.70, respectively. Mean and SD of Fruits consumption among children aged 6-23m, 6-8m, 9-11m and 12-23m were 2.07 ± 1.90, 1.51 ± 1.91, 1.95 ± 1.70 and 2.37 ± 1.97, respectively. Mean and SD of Grains consumption among children aged 6-23m, 6-8m, 9-11m and 12-23m were 16.01 ± 8.20, 11.53 ± 6.38, 15.80 ± 8.32 and 18.01 ± 8.10, respectively Conclusion As children have very fast growth in 6-23 months of their old, they need to having a good complementary feeding because complementary feeding period represents a window of opportunity for preventing all forms of malnutrition, including stunting, wasting, overweight and obesity. Our result found less than 50% of children are bene ting from MDD and MAD, beside this result showed very low consumption rate of fruits, vegetables and animal source foods. From other side most of the households live in food insecurity situation. Household food insecurity and weak complementary feeding make more prevalence of malnutrition. For prevent from these problems a long-term action plan is needed and should add these actions in to the plan: raising awareness and giving more nutrition education to mother, families, caregivers and health care worker, income generating activities to limit economic problems. Finally, we need to do more researches to have more accurate data or information that be comprehensive and covers the whole of Afghanistan on complementary feeding practices and its related factors.

Discussion
Results of odds ratios and 95% con dence intervals of low MUAC, low WAZ, low WLZ and low LAZ among those who did not meet complementary feeding criteria are shown as below: Results of Table 9 showed there was no association between MUAC, WAZ and WLZ with MMF expect LAZ which had association with MMF. This means Compared children who meet MMF, children who not meet MMF, marginally signi cant more likely to have low weight for age and low weight for length and signi cantly more likely to have low length for age.
Odds ratios and CI of (low MUAC, low WAZ, low WLZ and low LAZ) with MDD, which there is no association between MUAC and MDD but there is signi cant association between WAZ, WLZ, LAZ and MDD. This means Compared children who meet MDD, children who not meet MDD, signi cantly more likely to have low weight for age, low weight for length and low length for age. There is no association between MUAC and MAD too but there is association between WAZ, WLZ, LAZ and MAD which this means compared children who meet MAD, children who not meet MAD, signi cantly more likely to have low weight for age, low weight for length and low length for age. We were not found association between MUAC, WAZ, WLZ and LAZ with introduction of CF at six months old. This means there is no odds ratio for low MUAC, low WAZ, low WLZ and low LAZ because of children who not meet introduction of CF at six months old.
According to our nding from Table 5 and 6 expect regarding the variable intitled introduction of complementary feeding at six-month-old, none of CF indicators (MMF, MDD and MAD) associated with MUAC. MUAC is one of the methods for nding malnutrition children, which proper complementary feeding especially start CF at six months old of children may help to children for having good MUAC and prevent from malnutrition. So MUAC is very important and this part of study showed early or late start of complementary feeding may have effect on MUAC. Secondly all CF indicators (introduction of CF at six months old, MMF, MDD and MAD) associated with weight for age z-score (underweight) and all CF indicators associated with weight for length z-score (wasting) too. It is logical which infants need to extra feeding after a long time (after six month) so should start CF on that time and when infants grow after six months need to eat different foods frequently, so if a child start CF on time and receive MMF and MDD it is acceptable and would prevent from underweight and wasting. So, this part of study showed for having good WAZ and WLZ should meet all complementary feeding indicators. Expect MMF none of CF indicators (introduction of CF at six months old, MDD and MAD) associated with length for age z-score (stunting). Trend of children growth is very fast so they need to more energy for growing better. If a child be hungry and miss energy intake frequently, it would be suffered to chronic malnutrition (stunting). So, MMF is important for a child' growth and it would prevent from stunting.
In the study result of Udoh and Amodu, there was a signi cant association (p<0.05) between introduction of CF at six months old with wasting, the MDD was signi cantly associated with underweight and similarly, MAD was signi cantly associated with underweight too. Also, this study showed signi cant association between stunting and MMF, MDD and MAD [17].
In Saaka et al. study (which conducted for explore the relationship between IYCF indicators and child growth indicators in rural northern Ghana) the main nding was that three of the WHO core IYCF indicators were not associated with mean length for age z-score expect timely initiation of CF at 6 months and this study did not show any association between MDD or MMF and stunting. All WHO core IYCF indicators (introduction of CF at six months old, MMF, MDD and MAD) were not associated with mean weight for height/length z-score. Saaka et al. study presented signi cant positive association between introduction of CF at six months old and higher height for age z-score [18] but our study showed there is no association between introduction of CF at six months old and length for age z-score (stunting). In Ahmad et al. study there was no association on indicators of CF, namely MMF, MDD, MAD and timely introduction to CF with wasting, underweight and stunting [17]. Kimiywe et al. found that CF practices were signi cantly correlated with nutritional status, particularly MDD [19]. Korir JK's study also found that low MAD signi cantly correlated with wasting [20]. So, we can say a child have appropriate complementary feeding when he/she was breastfeeding at the time of study, met minimum meal frequency in the past 24 hours, met minimum dietary diversity in the past 24 hour and complementary feeding were introduced at six-month-old. As we discussed before most of previous studies found association between anthropometric measurements of children with complementary feeding indicators expect Ahmad et al' study which found no association between anthropometric measurement of children with complementary feeding indicators. Like most studies, the result of this study con rmed most of hypotheses and found associations between anthropometric measurement and complementary feeding indicators. But we need to more studies for con rm these associations like systematic review or meta analyze.
Result for odds ratios and 95% con dence intervals for low MUAC, low WAZ, low WLZ and low LAZ among food insecure subjects are shown in Table 8.
This Table shows there is no association between MUAC and household food insecurity, but there is signi cant association between WAZ, WLZ and LAZ with household food insecurity. This means compared children who were food secure, children who were not food secure, signi cantly more likely to have low Weight for age, low weight for length and low length for age. Table 3 is shown association of household food security with complementary feeding indicators with p value. This result showed there was no association between food security (secure and total insecure) and MMF, but there are associations between food security and MDD and food security and MAD. There is no association between food security and introduction of CF at six months old too. This means better household food security help children for having better MDD and MAD. There is no association between household food security (secure and insecure with their subgroups) and MMF, but there are associations between household food security (secure and insecure with their subgroups) and MDD and food security (secure and insecure with their subgroups) and MAD and also there is no association between food security (secure and insecure with their subgroups) and introduction of CF at six months old. The ndings con rm that most of Kabul households live in insecurity situation as only 9.3% of the households in the study were food secure. Low level of household food secure may be because of bias: In Afghanistan some health services providers or NGOs had some services to malnutrition mother/children like giving malnutrition materials (RUTF, RUSF, sprinkle powder, grains and etc.) to malnutrition child/mother they thought food security questionnaires are belong to these materials and from other side when we had interview about food security questionnaire, some of them asked: You asked me these questions for nancial or material helping? Or after this questionnaire you would give malnutrition material to me or my child? After asking these questions by two or three persons, to solve this problem, I rst explained to them that this questionnaire was not for any nancial or material help, this is just a research for nding nutrition or complementary feeding problems of Kabul children.

Result of
Our study result showed there is no association between food insecurity and MMF and introduction of CF at six months old, but showed association between household food insecurity and MDD and MAD. We categorized MDD as met when children consumed 4 groups out of 7 food groups and MAD were good when children met both MMF and MDD. A child for having a good growth and prevent from underweight, wasting and stunting should follow CF practices guideline. So those children who live in food insecurity situation cannot receive MDD or MAD and risk of malnutrition is more between these children. Macharia et al. study showed similar result: infants living in food secure household were signi cantly more likely to achieve appropriate infant feeding practices than those in food insecure household. Infants living in food secure household were signi cantly associate with introduction of CF at six months old, MDD and MAD which compared with infants from food insecure household without MMF [21]. Two other studies conducted in rural Bangladesh reported similar nding. Saha et al. study showed better household food security status was associated with better infant feeding practices for children aged 6-12 months [22] and Owais et al. showed that the odds of receiving a minimally acceptable diet for infants in most food secure households was higher than for infants living in least food secure households [23]. Table 2 showed receive of all nutrients and food groups were sub optimal and lower than recommended amount which can have different reasons. In our study were many children who were completed their six months old or may more, but were did not start complementary feeding and receive of their nutrients and food groups from way of diet were zero which these zeros had effect on total mean and SD of nutrient and food intake which this can be one of those reasons. The fruits of Afghanistan are seasonal, Fruits are cheap and more in their seasons and most of people can use from, but after end of their seasons fruits come from other countries which they are less and expensive because Afghanistan don't have especial refrigerator for saving seasonal fruits, So because of that most of people can't sue enough fruits and this study were done in season which fruits were less. This can be another reason for low level of nutrient and fruit intake. There is a high rate of illiteracy, poverty and food insecurity in the country which can affect people's nutrient and food intake. Finally, there is no previously study which explored complementary feeding practices and dietary intake of infants and children aged 6-23 months within the Afghanistan so, we need to do more researches to focus on nutrient intakes and its relationship with complementary feeding indicators and other variables.

Result of
According to de nition of complementary feeding indicators our result showed that all CF indicators (MDD, MAD, introduction of CF at start of sixth month, consumption of Iron or multivitamin supplementation and forti ed feeding) in this study were sub-optimal and still low , without MMF and Exclusive Breast Feeding (EBF) which they are better than other but they are not completely good. and 40% of the children, respectively [17]. In a previous study from Muzi Na and et al. in Afghanistan which conducted for predictors of CF practices and were used from analysis of the 2015 demographic and health survey data, the prevalence of introduction of CF at six months old between 6-8m aged, MMF, MDD and MAD among children 6-8 months were 56%, 55%, 23% and 18%, respectively [4]. Our study result showed better situation in part of CF indicators than study of Muzi Na and et al, surely this is result of health services and giving health education to people as we can see in community.
Globally, 64.5% of infants 6-8 months of age were received introduction of CF at six months old on 2017, and global rates of MMF, MDD and MAD were low at 50.3%, 28.2, and 15.9% respectively [1]. The present study nding was also higher than global and that reported by WHO found that less than onefourth of the children aged 6-23 months in developing countries had good consumption quality [24] but they still remained under the optimal level. I should add, our study was designed in capital of Afghanistan which most of Kabul people may have better economic situation, better accessibility to health services and may have higher education level than people of other provinces and from other side our sample size was very small. So, these four reasons or perhaps more, may show high level of CF indicators in all Afghanistan.

Limitations
Low rate cooperation of mothers for face to face interview was one of our limitation and for solving and decreasing it we were said to mothers which we would learn those mothers who have good cooperation, about proper children complementary feeding.
We were had a female interviewer because of Afghanistan culture which some of mothers didn't present for having interview with a male interviewer.
Most of the mothers were illiterate and data collection was done by interview method.
Start and opening late of Health facility or hospitals and early end and closing (daily limitation of time) which we called many times to responsible but we can't solve this problem and because of this problem data collection were completed late.             Figure 1 Prevalence of household food insecurity Complementary feeding practice indicators among children aged 6-23 months

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. CoverLetter.doc ReferencePDF.pdf