Feeding practices patterns in Romanian infants 6-23 months old: findings from a national representative sample

Background. Besides breastfeeding, complementary feeding is necessary to meet the nutritional needs that foster healthy growth in children 6–23 months old. In 2007, WHO, UNICEF, USAID, UCDAVIS, IFPRI experts elaborated a set of indicators to evaluate infants and young feeding practices. These indicators are specific to children under 6 months and respectively 6-23 months. In this study three key indicators were used to measure complementary feeding practices: minimum acceptable diet (MAD), minimum meal frequency (MMF), and minimum dietary diversity (MDD). Aims. The aims of the present study were to determine the prevalence of adequate and inadequate minimum acceptable diet (MAD), minimum meal frequency (MMF), and minimum dietary diversity (MDD) among children 6-23 months of age, in relation to infant or mother’s risk of demographic factors. Methods. A nationally representative sample with 1,532 children (713 girls and 819 boys) 6 to 23 months of age ( M = 14.26; SD = 5.15) was included in our analyses. All eight macro-development regions within Romania were represented and participants were randomly recruited from at least two counties in each of these regions. Results. It was found that the rates of achieved minimum acceptable diet were relatively high (72.3%), as well as the minimum acceptable dietary diversity (76.1%), and minimum meal frequency (96.1%). Conclusions. The results of the study showed that the prevalence of children 6-23 months who have an adequate complementary diet in terms of the three indicators exceeds 72%. But there are also population groups that do not reach this prevalence. Adequate complementary nutrition is generally achieved around the age of 1 year and less than 6-9 months, when the development needs of children already require animal foods rich in micronutrients.


Introduction
The current study is the fourth of its kind to examine nutritional status and dietary practices in Romanian children under the age of two.The first was orchestrated in 1991, the second took place between 1993-2000 (Stănescu, 2002), through the analysis of the Nutrition Surveillance Data, and the third study was completed in 2004 (Nanu et al., 2006).
The results of these studies indicated that 91-92% of children aged 12 months were breastfed (BF).
Complementary feeding was introduced early, around 4 months, through additional liquid or semifluid food.According to various sources (Callen & Pinelli, 2004;Agostoni et al., 2009), it is recommended that children be exclusively breastfeed up to six months of age, and this practice should be accompanied by complementary foods up to two years or beyond this age (Lanigan et al., 2001;Wasser et al., 2011;Carlett et al., 2017;Jabri et al., 2020).
With stunting and wasting becoming a growing concern in many regions, such as South Asia, it is critical to increase not only breastfeeding practices, but also other complementary feeding factors.According to Harding and colleagues (Harding et al., 2018), the prevalence of child wasting in South Asia is 16%, and that of stunting is 36%.Thus, children under two years of age should have access to high nutrient food supporting their growth and development (Dewey, 2016).This focus on children aged two years and younger would provide a window of opportunity to prevent growth deficits such as wasting and stunting (Lutter, 2000).Besides breastfeeding, other complementary feeding practices are necessary to meet nutrient needs that foster healthy growth in children.As shown (Dewey, 2016), infants aged 6-8 months need nine times more iron and four times more zinc than an adult male.Micronutrient needs are high, while the caloric intake remains relatively low during these first 1000 days of life, thus presenting a challenge to parents in making sure that their child receives as many nutrients as possible.However, if adequate micronutrient intake can be attained, linear growth can take place (Onyango et al., 2014).

Hypothesis
The aim of the present study was to determine the prevalence of adequate and inadequate minimum acceptable diet (MAD), minimum meal frequency (MMF), and minimum dietary diversity (MDD) among children 6-23 months of age, in relation to infant or maternal risk of demographic factors.

Material and methods
All parents of the participants completed and submitted their written consent to the proper authorities before the study began, more specifically to the "Alessandrescu-Rusescu" National Institute for Mother and Child Health (Bucharest, Romania).

Research protocol a) Period and place of the research
The current study drew the data from a more exhaustive cross-sectional survey run during the latter half of 2010.The larger study, entitled Evaluating the effectiveness of interventions included in national programs on the nutrition of children under 2 years (Nanu et al., 2011), Part 2: Assessment of the Current Situation of Breastfeeding and Nutrition Practices for Children from Birth to Two Years Old (Stativa & Stoicescu, 2011), utilized a national family physician database and stratified random sampling to determine a sample of Romanian children aged 0-23 months, N = 2,117.This is a nationally representative sample, for it included all eight macro-development regions of Romania, and the participants were randomly recruited from at least two counties of each of these regions (16 total).Within each county, children from two urban districts (two sectors of the capital, Bucharest, and rural communities of the Bucharest region that was not the capital) and four to five villages were arbitrarily recruited.In congruence to the area of residence, the data are also nationally representative due to the age range of each dyad, a mother and her child, who were assigned an identification code and invited to their physician's office.Health-related inquiries were implemented by two or three of 17 trained and experienced investigators.Once the interview process was completed, the investigators drew blood samples from the children and measured their weight and height.

b) Subjects and groups
The study investigators were experienced individuals who were also trained for the specific demands of the interview process.The interviewers worked in dyads or triads and verbally queried participants and recorded data in a face-to-face, in situ interview process.All questions were pretested (N = 80) before official data collection started to affirm consistency among all researchers administering them.The mothers were questioned at their family doctor's office regarding their prenatal consultation and iron prophylaxis use during the current pregnancy; initiation, knowledge, and practices of breastfeeding; child feeding practices; maternal smoking; and iron and vitamin D prophylaxis (Nanu et al., 2011).Complete data were collected over a four-month period (Stativa et al., 2014).
In the current study, sampling data were ensured to be evidence of the most modern pattern in children's nutrition research, by only selecting children aged 6-23 months for this analysis.Therefore, the sample encompassed 1,532 children (713 girls and 819 boys) 6 to 23 months of age (M = 14.26;SD = 5.15).c) Applied tests -Complementary feeding indicators WHO's infant and young child feeding indicators were implemented for the current study ( 1); these indicators were measured based upon the interviewed mother's recollection of food given to her child within the last 24 hours before the survey.For this study, minimum acceptable diet (MAD) measures were estimated by taking into consideration the minimum dietary diversity (MDD) and minimum meal frequency (MMF).MDD and MMF are indicators which were based on the mother's report of food given to her child in the 24 hours before the survey.
-Minimum acceptable diet (MAD) This variable reflects the proportion of breastfed children aged 6-23 months who had at least the minimum dietary diversity and the minimum meal frequency during the previous day.To calculate the indicator, information on breastfed and non-breastfed children is combined by adding the following two fractions: Breastfed children 6-23 months of age who had at least the MDD and the minimum meal frequency during the previous day / Breastfed children 6-23 months of age and Non-breastfed children 6-23 months of age who received at least 2 milk feedings and had at least the MDD not including milk feeds and the minimum meal frequency during the previous day / Non-breastfed children 6-23 months of age.MAD was expressed as a dichotomous variable with category 1 for meeting the complementary feeding indicator (adequate) and category 2 for not meeting that indicator (inadequate).As presented above, MAD was computed by considering the minimum dietary diversity (MDD) and minimum meal frequency (MMF).
-Minimum dietary diversity (MDD) This variable is the proportion of children 6-23 months of age who received foods from four or more of the seven food groups during the previous day.The seven foods groups used for tabulation of this indicator were: (1) grains, roots, and tubers; (2) legumes and nuts; (3) dairy products (e.g., milk, yogurt, cheese); (4) flesh foods (e.g., meat, fish, poultry, and liver/organ meats); (5) eggs; (6) vitamin-A rich fruits and vegetables; and (7) other fruits and vegetables.MDD was expressed as a dichotomous variable with category 1 for meeting the complementary feeding indicator (yes) and category 2 for not meeting that indicator (no).
-Minimum meal frequency (MMF) This variable is the proportion of breastfed and nonbreastfed children aged 6-23 months who received solid, semi-solid or soft foods (but also includes 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) on the previous day.MMF was expressed as a dichotomous variable with category 1 for meeting the complementary feeding indicator (adequate) and category 2 for not meeting that indicator (inadequate).
-Explanatory factors The following demographic variables regarding mothers' characteristics were used as categorical explanatory variables (EVs): mother's age, domicile (urban vs. rural), ethnicity (Romanian vs. Hungarian vs. German vs. Roma), whether the mother was a smoker (yes vs. no), and education (no school/gymnasium -completed or not vs. high/technical school vs. college).Education-related categories were created based on the years of school attended: no school/gymnasium (0-8 years), high/technical school (9-12 years), and college (13 years or more).Additionally, mother's marital status was defined by two categories, married mothers vs. other (the "other" category included cohabitation, divorced/separated/ widow, and unmarried).Also, included in our analyses was socioeconomic status (SES) which was coded as low, medium, or high.To determine SES, participants were asked whether they owned 12 items: a stove, television, refrigerator, washing machine, mobile phone, flush toilet, central heat, private car, private housing, personal computer, video recorder, and vacation home.Participants who owned up to four goods or services were classified as low SES; those who owned at least five and up to eight were classified as medium SES; and those who owned at least nine were classified as high SES.
Additionally, infants' characteristics were used as categorical explanatory variables (EVs): infant's sex (male vs. female), age, birth order (first born vs. second born vs. third and up born), whether the infant was born at term or not (at term, 37-40 weeks vs. before term, 27-36 weeks), and hemoglobin level.Regarding the latter, blood samples were drawn from infants presumed to be healthy.A finger-prick sample of capillary blood was collected, and hemoglobin concentration was measured using a portable battery-powered hemoglobinometer (HemoCue).Afterward, parents were informed about their child's iron status.Initially, four levels were established to categorize hemoglobin concentration (Hb), namely Hb≥ 11.0 g/dL (no anemia); Hb 9-10.9 g/dL (mild anemia); Hb 7-8.9 g/ dL (moderate anemia), and Hb <7.0 g/dL (severe anemia) (Stativa et al., 2014).Logistic regression was utilized to adjust for complex sampling design and variable measurements.Models were built using stepwise backward regression to determine the variables that were significantly associated with unsuitable complementary feeding practices.The models formulated by backward elimination followed the following guidelines: (1) the variables used in the backward elimination models had a p-value <0.20 in univariate analysis; (2) possible confounding variables were included in the model and nonsignificant values (p >0.05) were removed step by step; and (3) collinearity was tested as well.Odds ratios with 95% confidence intervals were computed with the intent of assessing the adjusted risk of independent variables; those with p <0.05 remained in the final model.The relationships among all variables/indicators were analyzed regarding idiosyncratic child, parental health care and household characteristics within a multiple logistic regression model.

Table I
Demographic frequencies and percentages of mothers and their children (N = 1,532).Table I shows the frequencies and percentages of the characteristics of the entire sample of children and their mothers.Overall, out of 1,532 children, 819 (53.5%) were males, 593 (38.7%) were aged 6-11 months, and 789 (51.5%) resided with their mothers in urban locations.

Indicators
Additionally, most of the participants (mothers) were Romanian (84.3%), married (81.4%), had high/technical school level education (42.6%), were characterized as having a medium socioeconomic status (48.4%), and smoked in the last 30 days prior to data collection (71.7%).Furthermore, more than half of children were first-born (54.8%), as well as born at term (89.9%).
Prevalence of MAD, MDD, and MMF Table II displays the frequencies and percentages of MAD, MDD, and MMF.Specifically, in infants 6-23 months of age, it was found that the rates of achieved minimum acceptable diet were relatively high (72.3%),as well as the minimum dietary diversity (76.1%), and minimum meal frequency (96.1%).

MAD, MDD, and MMF model predictors
The odds ratios, confidence intervals (odds ratio scale) and p-values, along with other parameters for the MAD, MDD, and MFF model predictors are shown in Tables III, IV, and V.
As shown in Table III, those infants living with their mothers in rural areas were 1.3 times more likely (95% CI; 1.033, 1.655) to meet the complementary feeding indicators compared to infants who lived with their mothers in urban areas.Additionally, infants 12-17 months old and 18-23 months old were 1.7 (95% CI; 1.338, 2.298) and 1.5 (95% CI; 1.126, 1.973) times more likely to meet the complementary feeding indicators compared to infants 6-11 months old.Our results also showed that infants born second in their families were 0.7 (95% CI; 0.580, 0.957) times less likely to meet the complementary feeding indicators compared to their counterparts who were born first.Additionally, infants whose mothers identified themselves as Roma were 0.6 (95% CI; 0.414, 0.965) times less likely to meet the complementary feeding indicators compared to their Romanian counterparts.
As seen in Table IV, those infants living with their mothers in rural areas were 1.4 times more likely (95% CI; 1.070, 1.748) to meet the complementary feeding indicator MDD compared to infants who lived with their mothers in urban areas.Additionally, infants 12-17 months old and 18-23 months old were 2 (95% CI; 1.523, 2.674) and 2 (95% CI; 1.649, 3.037) times more likely to meet the complementary feeding indicators compared to infants 6-11 months old.Our results also showed that infants born at term were 0.6 (95% CI; 0.438, 0.900) times less likely to meet the complementary feeding indicator MDD compared to their counterparts who were born prematurely.Additionally, infants whose mothers identified themselves as Hungarian were 0.7 (95% CI; 0.422, 1.011), times less likely than their Romanian counterparts to meet the complementary feeding indicator MDD; this outcome was statistically marginally significant.In relation to this variable, mothers who identified themselves as Roma were 0.5 (95% CI; 0.347, 0.798) times less likely to meet the complementary feeding indicator MDD compared to their Romanian counterparts.
Finally, as seen in Table V, the results showed that infants with severe anemia were 0.2 (95% CI; 0.049, 1.170) times less likely to meet the complementary feeding indicator MMF compared to those with no anemia.Additionally, infants 12-17 months old and 18-23 months old were 0.4 (95% CI; 0.193, 0.982) and 0.2 (95% CI; 0.090, 0.393) times less likely to meet the complementary feeding indicators compared to infants 6-11 months old.The "other" category included cohabitation, divorced/separated/widow, and unmarried.
relation to this variable, mothers who identified themselves as Roma were less likely to meet the complementary feeding indicator MDD compared to those who identified themselves as Romanian.This result can be attributed to the fact that in rural areas, children generally eat from adult food much earlier than in urban areas.Traditionally, rural mothers have the culture of accustoming the child to adult food as early as possible so that they can carry out household activities.Additionally, there are still many pediatricians who do not recommend the introduction of foods such as beans, lentils, green peas before 3 years of age, and mothers in urban areas are more likely to "benefit" from such recommendations.
-MMF As seen in Table V, the results showed that infants with severe anemia were less likely to meet the complementary feeding indicator MMF compared to those without anemia.Additionally, infants 12-17 months old and 18-23 months old were less likely to meet the complementary feeding indicators compared to infants 6-11 months old.

Strengths and limitations
This cross-sectional survey utilized a national family physician database and stratified sampling across all eight macro-development regions of Romania to determine a large sample of children making the outcomes nationally representative.Although demographic factors may appear to have a causal relationship with children's nutritional patterns, the cross-sectional design of the present study does not allow such inferences to be made.Self-reporting bias could also be a factor that needs to be taken into consideration as a limitation of the study.

Conclusions
1.The present study provided detailed information on complementary feeding practices of infants aged 6-23 months in Romania.The assessment of infants' feeding practices based on the three indicators, MAD, MMF, MMD, provides for the first time in Romania information about the quality and quantity of food consumed by children after the first 6 months of life, at population level.
2. The results of the study showed that, overall, the prevalence of children who have an adequate complementary diet in terms of the three indicators exceeds 72%.However, there are also population groups that do not reach this prevalence.
3. Adequate complementary nutrition is generally achieved around the age of 1 year and less than 6-9 months, when the development needs of children already require animal foods rich in micronutrients.
4. Using these new indicators to measure infants' feeding practices allowed to more easily examine the influence of different social and demographic factors on adequate or inadequate dietary practices of children aged 6-23 months.
5. The results of this study are useful for developing policies to reduce health inequalities that affect vulnerable population groups.

Table II
Frequencies and percentages of MAD, MDD, and MMF (N = 1,532).

Table III
Multiple logistic regression analysis summary for various infant and mother variables predicting infant's minimum acceptable diet (MAD; N = 1,532).