Maternal and child characteristics correlated with frequency of consuming ultra-processed food by children aged 6 to 24 months old

Objectives: to verify the correlation between the consumption of ultra-processed food among mothers and children under two years of age and the main characteristics related to this consumption. Methods: cross-sectional study conducted in public health services. Three 24-hour recalls were applied to assess food intake. The ultra-processed food was grouped into: sugary drinks; meat; sauces and creams; dairy products; snacks; pastas; and mucilage. Themother’s body mass index and waist/hip ratio, and the child’s weight/height, height/age, weight/age and body mass index/age were calculated. The children’s ultra-process frequency as correlated with: anthropometric dyadic variables; ultra-process frequency on breastfeeding. The children’s average ultra-process intake was compared to pacifier, bottle, breastfeeding and socioeconomic status. Linear regression models were conducted. Results: 172 pairs were evaluated. Similarity was found in the mothers and children’s consumption of ultra-processed products. The higher frequency of ultra-processed products was correlated with older child and the higher body mass/age index and weight/age index. Of the 39 ultra-processed food present in the mothers’ diet, 22 were correlated to child’s Conclusion: the consumption of ultra-processed food by children is similar to their mothers and correlates with higher z-score values of weight/age and body mass/age index.


Introduction
Over the first two years of life infant feeding, is characterized by progressive changes depending on the child's development.At six months of age, the introduction of other food in addition to breast milk is recommended and, the first year of life onwards, the diet has already progressed in terms of food consistency and presents similar characteristics to the family's diet. 1 During this period, the child's eating experiences, in addition to impacting health 2 , will also influence the development of their eating habits. 3,4t childhood, for the growth and development, the consumption of food rich in sugar and sodium, canned food, fried food, soft drinks, snacks and sweets are not recommended, as they are considered unhealthy food and compete with nutritious ones.However, the presence of these food in children under two years old's diet has been verified. 1The United Nations Children's Fund (UNICEF) report on the World's Children's situation indicating that infant feeding reflects the global "nutritional transition", due to the substitution of in natura food for high level processed food. 5ccording to UNICEF report, 44% of the children aged 6 to 23 months old in the world do not receive fruit or vegetables in their diet and only 29% have a diversified consumption of food groups. 5That is, from the first years of life, children are consuming little variety of healthy food and being exposed to ultra-processed food. 1 Often they are weaned before six months of age and receive ultra-processed food earlier. 6,7A study with children aged four to 24 months old found that 56.5% received some type of ultra-processed food before six months. 5Children and adolescents are more vulnerable in consuming these products due to their intense flavor, practicality and low cost. 8he poor quality of infant feeding has contributed to the increased risk of malnutrition and hidden hunger. 5everal factors are related to feeding practices, including socioeconomic and demographic conditions and parental behavior, 3 with emphasis on the maternal figure, since in most cases they are responsible for the selection and offering of food in this age group.In addition to these factors, we also call attention to the use of pacifiers and baby bottles, since such behaviors impair breastfeeding. 7,9ased on the above considerations, we sought to verify the mothers and children under two years old correlation between the consumption of ultra-processed food and the main characteristics related to this consumption.

Methods
A cross-sectional study, conducted in 2016 and 2017, with mothers and their children under two years of age treated in the public health network in the city of Viçosa (MG).The city has 18 Basic Health Units (UBS) and a Polyclinic.The collection was conducted at the Polyclinic, on vaccination days, and at five UBS, however the others did not develop puericulture activities and/or did not accept to participate in the research.All mothers with their children under 24 months of age who were attended at these collection sites were recruited.The inclusion criterion adopted was being a mother of a child up to two years of age.Pregnant mothers and those with gluten and lactose intolerance were excluded from the study, since such conditions imply changes in the diet.Data were collected by three previously trained researchers.
Sampling was by convenience and totaled 203 pairs, 15 of which were excluded because they did not respond to all food recalls and 17 mothers and/or children were due to the existence of food allergies.Therefore, we assessed 171 pairs.We used the OpenEpi online software to calculate the sample power and we considered the children's exposure to UPP food, the prevalence of overweight according to BMI/age (BMI/A) and a significance level of 5%.
We used questionnaires containing the following variables to obtain the data: socioeconomic (classified by the criteria of the Associação Brasileira de Empresas de Pesquisa -ABEP-2016(Association on Research Companies); demographic (age and sex); behavioral conditions (children using pacifier and baby bottle, mothers' use of tobacco and alcohol); health (presence of chronic diseases and postpartum depression); gestational weight gain; birth weight and length.These data were self-reported by the guardians and/or obtained from the child's health card.
We also assessed the dyad's nutritional status.In mothers, we measured the weight, height and waist and hip circumference.We measured weight in kilograms, using a Kratos® electronic scale, with a maximum capacity of 150 kilograms (kg).We measured the height in centimeters using an Altura Exata® stadiometer.We measured the waist circumference at the midpoint between the lower margin of the last rib and the iliac crest.We calculated the Body Mass Index (BMI) by dividing the weight by the height squared, and the waist-hip ratio (WHR) by dividing the waist circumference by the hip circumference.Both BMI and WHR were classified according to the recommendations of the World Health Organization (WHO). 10In children, we measured weight and length.,We used a pediatric scale, with a maximum capacity of 16 kg to measure weight.For children weighing more than this capacity, we used the same scale for mothers.We measured the children's length in centimeters using a children's anthropometer.We assessed nutritional status using the WHO Anthro 2011 version 3.2.2program, using the WHO recommendations as a reference standard. 11 We assessed the dyad's eating habits using the 24hour reminiscent (R24HR).The mothers answered three R24HRs regarding their own diet and their children's, on non-consecutive days, one referring to the weekend or holiday.The first R24HR was applied at the collection site, after applying the questionnaire, anthropometric measurements were taken.We conducted home visits to obtain the remaining R24h.At the end of each R24HR collection, we verified the presence of processed and ultraprocessed food, and if found, we questioned the brands and types of the product for further analysis and checked the ingredients on the labels.We verified the practice of breastfeeding, 12 identifying in the R24HR the offer of other food or formulas along with breast milk.The food present in the R24HR were classified according to the level of processing as: in natura or minimally processed, processed and ultra-processed (UPP).For this classification, we considered Monteiro et al. 13 proposal and when there was doubt about its classification, we consulted the list of ingredients presented on the product label to classify them correctly.Infant formulas, despite being considered UPP, were not counted in this study, as they are indicated to substitute breast milk.
UPP food s were identified in the reminiscent collected.Initially, we identified the presence of UPP food on each day.Subsequently, we calculated the mean consumption of UPP present in the three food surveys.For the analysis, we calculated the sum of the frequency of each food belonging to the assessed groups, according to the classification below: -Sugary drinks (soda, juice powder, juice and industrialized coconut water); -Ultra-processed meats (pepperoni, frankfurter, sirloin, hamburger, bologna, salami, ham, nugget, sausage); -Sauces and creams (margarine, mayonnaise, tomato sauce, sour cream); -Dairy products (fruit yogurt/fermented milk, cream cheese); -Snacks (cream-filled cookies, powder, cornstarch cookie/saltycracker, chips, cake mix, gelatin, popsicle, ice cream/milk-shake, chocolate, candies/sweets, cereal bar); -Pasta (instant noodles, loaf of bread, pizza); -Mucilage (chocolate milk, powdered milk).
We used Stata software version 13.0 for statistical analysis.We verified the normality of the variables by the Shapiro-Wilk test and asymmetry coefficient.We used Pearson's correlation test to assess the correlation of the frequency of UPP food in the children's diet with: the frequency of UPP in the mother's diet, the anthropometric and sociodemographic variables of the dyad.We used Student's t-test to compare the frequency of UPP in children's diet with: presence of husband/partner; use of baby bottle; use of pacifier; presence of breastfeeding; urban residence (yes/no); and socioeconomic condition (A, B or C, D and E).
We made bivariate analysis using a simple linear regression model, considering the frequency of UPP in children's diet as an outcome.We conducted multiple regression based on the explanatory anthropometric, behavioral, obstetric, breastfeeding, socioeconomic and demographic variables of the dyad and referring to the UPP present in the mothers' diet.We identified potential confounders based on data and literature.A directed acyclic graph (DAG) was constructed in the DAGitty program to assist in the selection of the covariates included in the analyses.This approach allowed to identify a minimal set of covariates sufficient to control the confounding situation.The variables identified as confounders were: ABEP economic classification, breastfeeding, mother's age, depression, and mothers' use of alcohol and tobacco; use of baby bottle.We assessed the adjusted quality of the model by the analysis of residuals, as well as by the analysis of multicollinearity among the variables included in the model.We adopted a significance level of 5%.
The study was approved by the Ethics Committee of the Universidade Federal de Viçosa (UFV), under registration number 1,833,627.

Results
A total of 171 pairs were assessed, a sample that obtained a statistical power of 96.6%.The children's mean age was 12 (±8.2) months old and the mothers' mean age was 27 (±6.7)years old.Approximately half of the children were male (53.8%) and the majority (59.1%) of the mothers had concluded high school or higher education.Table 1 shows the description of the sample and it 91.8% of the children had an adequate weight for their age and 64.3% had an adequate body mass index for their age.As for the mothers' nutritional status, 59.1% were eutrophic.Approximately half of the mothers (49.7%) had waist circumference with values indicating increased risk for cardiometabolic diseases (Table 1).
As shown in Figure 1(b), the children's diet contained food belonging to all the groups present in the maternal diet, revealing a diversified consumption of these food and a food profile similar to the maternal one.In all age groups, there was the presence of at least one UPP food from each assessed group, except for children under 6 months of age, where there was no presence of pasta and ultra-processed meats.Among the UPP food present in the children's diet, the dairy and mucilage group were the most frequent for those younger than six months old (15.5% in both groups) and snacks were the most frequent for children in the age groups of 6 to 12 months old (72.4%) and over 12 months (94.1%).
Of the 39 UPP food found in the maternal diet, 22 showed a positive correlation with food present in the children's diet (Table 2).Positive and significant agreements were found between the total frequency of UPP in the children's diet and the child's age, weight/age and BMI/age (Table 3).
The total frequency of UPP food present in the children's diet did not differ in terms of the presence of a husband/partner, type of urban residence (yes/no), use of pacifiers and socioeconomic condition (A and B or C, D and E).However, there was an increase in the frequency of UPP food in the children's diet who received preparations through a baby bottle (p=0.033) and who were not breastfed (p<0.001) In the bivariate analysis, was found that the frequency of UPP in the children's diet was positively associated with the children's weight, length, W/A, BMI/A and age, so the higher frequency of UPPwas accompanied by higher values of the referred parameters and indications.In addition, there was na association between the consumption of UPP with the presence of sugary drinks in the mother's diet, alcohol consumption, breastfeeding and the use of the baby bottle.In the gross regression analysis, there was na association between the UPP frequency in the children's diet, only with the child's age and consumption of sugary drinks were by mothers.After adjusting for confounding variables identified by the DAG, both associations were maintained (Table 4).

Discussion
As far as we know, this is one of the first studies that assesses the relation between the presence of ultraprocessed food in the mothers' diet and their children's first years of life.In it, we identified a similarity consumption of UPP food in the mother-child dyad diet, with a great diversity of UPP food in the dyad's diet.In addition, the study found that children consume this type of food even  in the first six months, and that the consumption of these food is associated with an increase in the values of the W/A and BMI/A indices.According to Martins et al. 14 there was a significant increase in the presence of industrialized food in the Brazilian population's diet, especially ultra-processed food Studies conducted in other countries have also verified this trend in food consumption. 15,16The presence of UPP food implies a reduction in the consumption of in natura or minimally processed food and culinary ingredients, 2,14 which should form the basis of the diet.Thus, both in the context of healthy eating and in public health, the increasing household consumption of industrially processed products has become a concern, 8,14 since it represents a risk for deficiency and chronic non-communicable diseases. 17In the present study, reveals in the found condition the mother not only includes UPP food in her diet, but also introduces it into her child's diet.

Table 4
Linear regression coefficients (gross and adjusted), respective confidence intervals and p-value for the association between the frequency of ultra-processed food in children's diet according to the final variables presented in the model, Viçosa, MG, 2016-2017 (n=171).The frequency of UPP food in children's diet who were breastfeeding was lower when compared to other children.Relvas et al., 9 identified that infants who were not breastfed had a higher frequency of ultra-processed food in their diet compared to those who were breastfed.

Variables
This result emphasizes the importance of the practice of exclusive breastfeeding, mainly because the harmful effects of the early offer of food to the child 18 are added to the risks of consuming UPP food.In addition, consumption of UPP food early in life can predispose children to the development of unhealthy eating patterns, which can also result in the development of chronic non-communicable diseases at childhood or adulthood. 17,19hen analyzing the UPP food groups, we found that pasta was not consumed by children younger than six months old.However, Longo-Silva et al. 20 found that 21.5% of the mothers offered instant noodles to their children before six months of age and, in the second study, 23% of children younger than 4 months also consumed this kind of food.In addition to noodles, Spinelli et al. 21ound consumption of hot dog bread in 37.9% of children under one year of age.
The consumption of cornstarch cookies, creamfilled cookies and salty crackers, and candies and sweets stood out in children's diet under six months of age.Longo-Silva et al. 20 also verified the consumption of food such as cream-filled cookies, chocolate and ice cream.Freiberg et al. 22 found 50% of the children aged between four and six months of age consuming thickeners together with milk in the baby bottle.This result is in line with the present study, since the frequency of UPP food in children's diet who were under six months of age was higher among those who used baby bottle.We believe that this relation is due to the fact that children are weaned before the recommended and early exposure to this type of food. 9 study conducted with children under 24 months of age found that about 31% of the children consumed artificial juice and 30% already consumed sugar and chocolate milk. 23Considering that the child is born with a predilection for the sweet taste, the offer of sugary food can get them used to the high consumption of sugars, with consequences for their current and future health. 24he frequency of UPP food increased with advancing age, similar to what was observed in the two studies. 2,25rom six months onwards, there was a greater similarity between the dyad regarding the presence of UPP food in the diet.This can be explained by the fact that complementary feeding favors the approximation of children to the family's eating habits, 3,4 which currently means a greater possibility of consuming UPP food. 14,15e found a relationship between maternal eating habits on the child's food profile by verifying a positive association between 22 UPP food present in the mother's and child's diet.Other studies have also found that family influence on children's food consumption, such as Araújo 26 when verifying that maternal consumption is a predictor factor for the consumption of healthy food at childhood and Jaime et al., 3 who found this association between children and adults.
The higher frequency of UPP food in children's diet was accompanied by higher values of weight, length and W/A and BMI/A indices.Considering that there was a relationship between the frequency of UPP and the child's age, this justifies its association with length, since at this stage of life growth occurs with advancing age.A study conducted with children aged from 4 to 7 years old found a positive association between the "Unhealthy" dietary pattern (artificial juice and soda, fried food, snacks and canned food, sweets, and cream-filled cookies) with indicators of total adiposity (BMI and total body fat) and central (WHtR and central fat). 27ositive associations between consumption of UPP food and body fat during childhood and adolescence 27,28 and obesity in adults and adolescents 16,29 have been verified.Canella et al. 30 found an association between overweight and the greater availability of these kinds of food in Brazilian households.This condition is due to the inadequate nutritional characteristics of these food, which contribute to the excess consumption of energy, total and saturated fat. 20,29Other characteristics also contribute, such as the fact that they are hyperpalatable and easy to transport, 8 which encourages the mother to include them in the children's diet and, sometimes, in an early and ascending way, which contributes to weight gain.
In the present study, there was no association between the frequency of UPP food in the children's diet with income and place of residence.This may be due to the fact that the entire population, regardless of their socioeconomic class, has UPP food in their diet.Martins et al. 7 corroborates this statement, as they found an increase in the consumption of UPP products in all economic strata.
Among the limitations of this study, we highlight those referring to the food survey used, such as the memory bias during the report of consumption referring to the previous day.Also, a study that considers the caloric contribution of these food could increase knowledge about their influence on the dyad's nutritional status.However, this is apparently the first study that assesses the relationship between the presence of ultra-processed food in the mothers' diet and their children in the first two years of life and that correlates the presence with the nutritional status.
We concluded that the presence of ultra-processed food in children's diet is similar to their mothers, so that mothers who included UPP food in their diet also introduced this kind of food to their child's diet, even before six months of life.The consumption of UPP food seems to have an impact on the child's nutritional condition, given its correlation with higher z-score values of the W/A and BMI/A anthropometric indices.Considering these findings, we emphasize the urgency of actions to take place to clarify and promote the awareness of parents and family members about the risks of introducing ultra-processed food s in children's diet under two years of age.

Figure 1 Frequency
Figure 1Frequency of mothers and children who had at least one ultra-processed food in their diet, according to age group (a) and groups assessed (b).Viçosa, MG, 2016-2017 (n=171).
L/A = Length for age; W/A = Weight for age; W/L = Weight for length; BMI/A = Body Mass Index by age; BMI = Body Mass Index; WC = Waist Circumference.*Socioeconomic condition according to the Brazilian Association in Research Companies (ABEP).