Appropriate Complementary Feeding Practices and Associated Factors Among Caregivers of Children Age 6-23 Months in Rural Areas of Hunan Province, China: A Community-based Cross-sectional Study

Background This study aimed to evaluate the quality and associated factors that potentially inuence complementary feeding practices among Chinese caregivers living in Hunan, a rapidly developing province in China. Methods Three indicators( minimum dietary diversity(MDD), minimum meal frequency(MMF) and minimum acceptable diet(MAD)) were used to evaluate the rationality of complementary feeding. A multi-factor unconditional logistic regression model was employed to analyzeanalyse the relevant factors of complementary feeding.

2. This is a Well-conducted study on in uencing factors of complementary feeding for children aged 6-23 months in rural areas of China, including cultural and socio-economic factors.
3. It provides a basis for improving the situation of infant supplementary feeding in rural areas of China.

Background
Malnutrition in children is a global public health problem. About 144 million children under 5 suffer from stunting, and 47 million are diagnosed with wasting[1]. In low-and-middle-income countries, undernutrition is the main cause of death [ 2] among 45% of children under 5. For surviving children, undernutrition will also cause lasting damage to their neurodevelopment and cognition, increasing the risk of chronic diseases, and affecting their labour and income in their adulthood [3].
Undernutrition is also a major health problem among Chinese children. According to the Report of Nutrition and Development for Children in China, the rates of underweight, stunting and wasting among children under 5 are 3.6%, 9.9% and 2.3%, respectively, which are lower than in most developing countries and signi cantly lower than in Southeast Asian countries but higher than in developed countries, such as the United States.
In the complementary feeding period of 6-23 months of age, children can only eat a small amount of food due to their small stomach capacity. Thus, complementary food must be rich in energy and nutrient density, and reasonable feeding frequency is required to meet the needs of physical growth and brain development of children. Appropriate complementary feeding can promote the growth and development of children, while belated and unreasonable complementary feeding may lead to stunting [4] . Therefore, ensuring adequate complementary food for children aged 6-23 months is critical for achieving the global development goal of reducing the number of undernourished people from 171 million in 2010 to 100 million by 2025 [5] .
Research on factors related to complementary feeding in other regions has shown that highly educated parents, wealthy families, acceptance of media publicity, adequate pre-and post-natal education, gender and age of children, number of family members, deliveries at health facilities, low pregnancy frequency, occupations of gravida, urban areas, knowledge of complementary feeding and acceptance of feeding advice while immunisation are the decisive factors [6][7][8][9] .
Surveys were conducted in China, but the evidence on speci c predictive factors (including cultural and socioeconomic factors) of complementary feeding is still limited. Therefore,this work investigated the status of complementary feeding of rural children in Hunan Province. Indicators for Assessing Infants and Young Child Feeding Practices (WHO) was used for analysis to understand the feeding status and existing problems of rural children in Hunan Province. Results can provide basis for establishing feeding programs for children in this area and other similar areas.

Respondents
Multi-stage strati ed random sampling was used to select 24 sites (12 streets and 12 towns) in 6 cities of Hunan Province. Approximately 50 children aged 6-23 months and their caregivers were selected from each survey spot, with a total of 1,200 participants (the actual number was 1,220).

Contents and methods
A self-designed questionnaire was employed to conduct preliminary experiments to ensure high compliance. The contents mainly included children's basic information, condition of childbirth, feeding status, health states and diet retrospective investigation. The parents' feeding method, related knowledge and attitude towards nutrition and health were also included. The method of investigation was face-toface interview between investigators and caregivers.

Quality control
A uni ed questionnaire was used. Investigators with medical background were intensively trained to familiar with the questionnaire before formal investigation. The process was supervised and guided by the main staff from the beginning to the end to ensure timely correction for any problems.

Related indicators and their de nitions
According to Indicators for Assessing Infant and Young Child Feeding Practices, the main analysis indicators included the following: (1)Minimum Dietary Diversity (MDD): the proportion of children aged 6-23 months who receive foods from four or more food groups during the previous day; (2) Minimum Meal Frequency (MMF): the proportion of breastfed and non-breastfed children aged 6-23 months who receive solid, semi-solid or soft foods the minimum number of times or more (minimum is de ned as: two times for breastfed infants 6-8 months; three times for breastfed children 9-23 months; and four times for nonbreastfed children 6-23 months) in the previous day; (3) Minimum Acceptable Diet (MAD): the proportion of breastfed children aged 6-23 months who had at least the MDD and the MMF during the previous day, and non-breastfed children who received at least two milk feedings and had at least the MDD (not including milk feeds) and the MMF during the previous day.

Statistical analysis
Epi Data 3.0 was employed for double entry and error detection of data, and SPSS 19.0 software was used for statistical analysis. The results of count data were expressed as rate (%). The rate comparison between two or more groups was tested by χ2. Multivariate analysis was performed by unconditional logistic regression. If P<0.05, then the difference was statistically signi cant. The average age of surveyed mothers was 29.85 years. The educational background of them was mainly junior high school and below (45.3%), and they were chie y engaged in agriculture (43.4%). The education background of fathers was mainly junior high school and below (45.7%). Most of the caregivers (834, 67.4%) were mothers and primarily engaged in agriculture (65.8%). The education background of caregivers(59.3%) was mainly junior high school and below, followed by high school/technical secondary school (25.2%). Most of the families' annual total income was 60,000 yuan and above (746, 61.1%), and most of them spent 6000 yuan and above for food (920, 75.4%). Most (996, 81.6%) mothers did not go out for working; 58.3% of the families have two children and most (54.9%) of the investigated children were the second child in the family( Table 2). Third and above 109 8.9

Basic public health services
The utilisation of basic public health services for children was evaluated. The vast majority of children received free physical examination (90.6%), free height and weight measurement (90.5%) and free haemoglobin test (81.6%) ( Table 3).

Knowledge and attitude towards nutrition and health of parents
Results of the survey showed that most parents (67.7%) had scores of 80 or above on knowledge and attitude towards nutrition and health, and only very few (6.8%) failed. This nding indicated that most parents have certain knowledge of feeding and nutrition (Table 4).

Evaluation indicators
The evaluation indicators of complementary feeding were assessed. About 62.3% of the caregivers provided their children with four or more types of foods that met the MDD the day before the survey. Meanwhile, about three-quarters (75.1%) of the caregivers provided their children with food before the survey that achieved the MMF. Hence, the eligibility rate of the MAD of children was 44.0%.

Univariate and multivariate analyses
Children were grouped according to their basic conditions, family sociodemographic characteristics, basic public health services, knowledge and attitude of parents towards nutrition and health. The eligibility rates of MDD, MMF and MAD among the groups were compared. Statistically signi cant variables were selected through single-factor analysis, and they included month age of children, survey areas, mothers moving out to work or not, and scores of parents of knowledge of nutrition and health. With MDD, MMF and MAD as dependent variables (eligible=0; unquali ed=1), the statistically signi cant variables of the univariate analysis were incorporated into the multivariate logistic regression model for multivariate analysis. The results showed that young month-age of infants was a risk factor for ineligible MDD and MAD. Compared with the 6-11-month-old group, the probability of ineligible MDD of 12-17month-old group was about one-quarter (OR=0.238,95%:0.177-0.320), and the probability of ineligible MAD was one-half (OR=0.425,95%:0.317-0.517). The probability of ineligible MDD in the 18-23-monthold group was only one-tenth (OR=0.103,95%:0.074-0.145), and the probability of ineligible MAD was one-third (OR=0.376,95%:0.279-0.507). Low score of parents with regard to knowledge of nutrition and health was a risk factor for ineligible MDD, MMF and MAD for children. Compared with the group whose scores were <60, the probability of ineligible MDD, MMF and MAD in the group of 60-79 scores were about two-thirds, and that in the group of 80-100 scores were about one-half. The survey area was related to the disquali cation of MDD, MAD and MAD. The failure rate of MMF was higher in the group of mothers that did not go out for work compared with that in the group of migrant mothers (Table 6).

Discussion
Studies have reported no signi cant difference in the growth rate of height and weight of children within 6 months between China and developed countries. the growth curve of Chinese infants within 6 months is close to the standard curve recommended by the WHO. However, the growth curve after 6 months gradually lag behind primarily due to improper addition of complementary food [10] .In the present study, according to Programming Guide: Infant and Young Child Feeding [11] by the WHO, the eligibility rate of MDD, MMF and MAD were used to evaluate the status of complementary feeding of children aged 6-23 months in rural areas of Hunan Province.
Compared with the feeding indicators of 46 developing countries [12] reported by the WHO and UNICEF in 2010, the eligibility rate of MDD, MMF and MAD in the present study are 15th, 8th and 6th respectively, which are at the optimum level. The eligibility rate of MDD, MMF and MAD in this study are generally better than the data reported in all China [13] , central and western regions [14] and Xinjiang [15] , but lower than the data reported in Hubei Province [16] . Since 2012, some rural areas in Hunan Province have participated in the national Projects on Children Nutrition Improvement in Poor Areas [17] that provide children aged 6-23 months with a packet of nutrients rich in protein, vitamins and minerals. At the same time, health education on nutrition was carried out to improve family members' feeding knowledge in rural areas, and complementary feeding has been promoted.
This study shows that most children aged 6-23 months were fed with complementary foods, but the types of foods were too single, mainly cereals and grains, and other types of foods were rarely added, especially dairy products, legumes and nuts. Even for children aged 18-23 months, the addition rate was less than 50%, and the intake of high-quality protein and micronutrients was insu cient. Based on the recommendations of the American Academy of Paediatrics, the European Society for Paediatric Gastroenterology, Hepatology and Nutrition [18] and related food guide for infants and young children in British and Australia [19] , diverse food should be added. Complementary food in rural areas was mainly prepared by the family, mostly by grandparents, the dietary structure was unreasonable, which mainly included cereals, low amounts of fruits and vegetables and insu cient high-quality animal food due to the limited nutritional knowledge.
This study shows that young month-age, mothers' not going out to work and parents' lacking in knowledge about nutrition and health were high risk factors for ineligible complementary feeding.
The eligibility rates of MDD and MAD were the lowest in the group of infants aged 6-11 months. The addition of complementary foods for children is assumed to be a process that is gradually accepted. As the months of age of children increase, their digestive systems will be more perfectly developed, which means that more types of foods can be accepted, leading to a more reasonable dietary structure. Children aged 6-11 months required the highest nutrient density and was the most prone to lack trace elements.
During this period, consuming various iron-rich foods is essential to meet the high nutritional needs (especially iron and zinc, which are common de ciencies) [20] . A relatively small difference in complementary feeding was found between the groups of 12-17 months and 18-23 months, where the eligibility rate of MDD nearly doubled from 6-11months to 12-17 months, and that of MAD increased by 21.8%. Early dietary diversity of children is closely related to their subsequent growth and development [21] . Studies have shown that the risk of stunting in children with quali ed complementary feeding is signi cantly reduced [22] . The younger the children is, the greater their developmental potential will be. Therefore, attention should be paid to the complementary feeding of children at young month age and the types of food should be increased as soon as possible to meet the dietary requirements so children can achieve their maximum potential.
The eligibility rates of MDD, MMF and MAD in different surveyed areas were different. Differences in culture, economic status and knowledge level in different areas may affect the complementary feeding of infants and young children in the area. For example, the eligibility rate of various indicators of complementary feeding in Changsha was at a relatively high level, which was assumed to be due to the higher level of economic development and living standard of people in this area, leading to more money and energy invested in feeding of children.
The eligibility rate of MMF for children whose mothers were migrant workers was higher than that of children whose mothers were not migrant workers. It's speculated that children whose mothers were migrant workers were generally raised by their grandparents who have certain experience in raising children and are more familiar with the characteristics of each growth stage of children. Therefore, they can provide diverse and su cient complementary food. Moreover, grandparents tend to treat their grandchildren more lovingly and patiently, which is good for communication and a harmonious family atmosphere where children can take more suitable food.
This nding indicated that the higher the score of parents' knowledge and attitudes towards nutrition and health is, the higher the eligibility rates of MDD, MMF and MAD and the better complementary feeding for children. A study conducted in ve South Asian countries showed that caregivers' lacking knowledge in feeding was the main reason for inappropriate complementary feeding [23] . Other studies have shown similar results [24] .

Conclusion
The study shows that types of complementary food in rural areas of China were too single and children's dietary structure was unreasonable which needed to be improved. Young month-age, mothers' not going out to work and parents' lacking in knowledge about nutrition and health were high risk factors for ineligible complementary feeding.When analysing the results of this study, some design limitations should be considered. First of all, this study was a cross-sectional survey, it can only con rm the correlation between factors and results and not the causality. Secondly, this study assessed the complementary feeding at a certain time point while ignoring the possible changes in complementary feeding as the child grows up. That is to say, the current condition of complementary feeding can neither represent its previous condition nor predict the situation afterwards. Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.