Processed and ultra-processed foods are associated with high prevalence of inadequate selenium intake and low prevalence of vitamin B1 and zinc inadequacy in adolescents from public schools in an urban area of northeastern Brazil.

Changes in eating behavior of adolescents are associated with high consumption of processed and ultra-processed foods. This study evaluated the association between these foods and the prevalence of inadequate micronutrient intake in adolescents. A cross-sectional study was conducted with 444 adolescents from public schools in the city of Natal, northeastern Brazil. The adolescents' habitual food consumption was evaluated using two 24-hour dietary recalls. Foods were categorized according to the degree of processing (processed and ultra-processed) and distributed into energy quartiles, using the NOVA classification system. Inadequacies in micronutrient intake were assessed using the estimated average requirement (EAR) as the cutoff point. Multivariate logistic regression models were used to estimate the relationship between energy percentage from processed and ultra-processed foods and prevalence of inadequate micronutrient intake. The mean (Standard Deviation (SD)) consumption of total energy from processed foods ranged from 5.8% (1.7%) in Q1 to 20.6% (2.9%) in Q4, while the mean consumption of total energy from ultra-processed foods ranged from 21.4% (4.9%) in Q1 to 61.5% (11.7%) in Q4. The rates of inadequate intake of vitamin D, vitamin E, folate, calcium, and selenium were above 80% for both sexes across all age groups. Energy consumption from processed foods was associated with higher prevalence of inadequate selenium intake (p < 0.01) and lower prevalence of inadequate vitamin B1 intake (p = 0.04). Energy consumption from ultra-processed foods was associated with lower prevalence of inadequate zinc and vitamin B1 intake (p < 0.01 and p = 0.03, respectively). An increase in the proportion of energy obtained from processed and ultra-processed foods may reflect higher prevalence of inadequate selenium intake and lower prevalence of vitamin B1 and zinc inadequacy.

Introduction

Adolescence is a period of high nutrient and energy demand and is thus a nutritionally critical period of life when lifestyle and dietary habits are changing [1]. This makes adolescents vulnerable to the consumption of heavily-processed, high-fat and high-sugar products [2, 3].

To study the effect of food processing on nutritional quality and health, we used the recently developed NOVA (which is not an acronym) classification system, which has been recognized as an appropriate framework for assessment of food processing levels [4]. NOVA is a system of grouping foods according to the nature, extent and purpose of the industrial processing used in their production. According to this system, all foods and beverages can be categorized into four groups: (i) unprocessed or minimally processed foods, (ii) processed culinary ingredients, (iii) processed foods, and (iv) ultra-processed foods. Processed foods are products manufactured by adding sugar, oil, salt, and other culinary ingredients to minimally processed foods to make them more durable and usually more palatable, for example, French bread, cheese, processed meats, canned fruits, and vegetables. Ultra-processed foods differ from less processed foods in that they fail to maintain their basic identity, undergoing various processing stages and techniques and including substances used exclusively in industry, such as cakes, pies, cookies, ready-to-eat and semi-ready-to-eat meals, bakery products, sugar-sweetened beverages, and snacks [4].

Meanwhile, among adolescents, some authors stress that the time spent in sedentary behaviors can lead to greater consumption of foods purchased in packaged form, ready to eat or heat [3]. The high consumption of processed and ultra-processed foods among adolescents has been associated with higher energy, saturated fat, or trans-fat intake, and a lower intake of dietary fiber and micronutrients [5, 6]. Inadequate food consumption patterns during childhood and adolescence are linked not only with the occurrence of obesity in youth, but also with the subsequent risk of developing diabetes, hypertension, and other noncommunicable diseases in adulthood [7]. In a healthy population of urban European adolescents, the HELENA (Healthy Lifestyle in Europe by Nutrition in Adolescence) Study showed that adolescents eat half of the recommended amount of fruit and vegetables and less than two-thirds of the recommended amount of milk (and milk products), but consume much more meat (and meat products), fats, and sweets than recommended [8].

In the United States (US), using 2009–2010 National Health and Nutrition Examination Survey (NHANES) data, an association was reported between consumption of ultra-processed foods and dietary content of added sugars [9]. Furthermore, the average content of protein, fiber, vitamins A, C, D, and E, zinc, potassium, phosphorus, magnesium, and calcium in the diet decreased significantly with an increase in the energy contribution of ultra-processed food [10]. Another study found that 7 to 20% of children and adolescents in the US miss lunch on both weekdays and weekends, and this practice was associated with lower micronutrient intake [11]. A systematic review points out that adolescents from Malaysia, a middle-income country in Asia, had lower diet quality, and Chinese adolescents spent less time in physical activity compared to other ethnicities [12]. These dietary patterns in adolescence are predicted to become global trends, and they have been associated with health risks for chronic diseases in adulthood, as this critical period of development forms the basis for establishing a lifetime of eating habits [13].

Latin American countries have experienced major shifts in intake of less-healthful low-nutrient-density foods and sugary beverages, changes in away-from-home eating and snacking, and rapid shifts towards high levels of overweight and obesity among all ages [14]. In a study of children from Colombia, there was a decline in intake of n-3 polyunsaturated fatty acids, calcium, zinc, and vitamins A, B12, C, and E, and this lower intake was related to an increased intake of processed and ultra-processed foods, as well as higher sodium, sugar, and trans-fatty acids intake. This study also observed that the amounts of some healthy nutrients, including folate and iron, were higher in processed and ultra-processed foods than that in unprocessed and minimally processed foods [15].

Brazilian household surveys conducted in the past three decades have shown a steady increase in ultra-processed food intake and a significant reduction in the intake of natural or minimally processed foods in all geographical regions [16, 17]. The 2013–2014 Brazilian Study of Cardiovascular Risks in Adolescents (ERICA) study of 71,791 adolescents aged from 12 to 17 years showed that rice, beans and other legumes, juice and fruit drinks, breads, and meat were the most commonly consumed foods. Saturated fat and free sugar intake were above the maximum recommended limit (< 10.0%). In short, the diets of Brazilian adolescents were characterized by the intake of traditional Brazilian food, such as rice and beans, as well as by high intake of sugar through sweetened beverages and processed foods. This food pattern was associated with an excessive intake of sodium, saturated fatty acids, and free sugar [18]. In addition, the Brazilian government National Adolescent School-based Health Survey (PeNSE 2015) points out that adolescents from the northeastern region of Brazil, whose mothers have a lower level of schooling, have a healthier dietary pattern compared to other more developed regions of Brazil [19,20].

To date, there are gaps in the literature of studies that focused on the association between intake of processed and ultra-processed foods and inadequate intake of micronutrients in adolescents. Only one study conducted among young people from southern Brazil reported that a higher consumption of ultra-processed foods was associated with a higher prevalence of inadequate intake of calcium, sodium, and iron [6]. Therefore, the purpose of this study was to evaluate the association of consumption of processed and ultra-processed foods with the prevalence of inadequate micronutrient intake in adolescents from public schools in an urban area of northeastern Brazil, given the urgent need to formulate dietary recommendations that favor the promotion of healthy eating.

Methods

Study population

This cross-sectional study was conducted among 444 adolescents from public schools in the city of Natal, northeastern Brazil. Data were collected in 2007 and 2008 as part of the “Risk factors for cardiovascular disease among adolescent beneficiaries of the Brazilian government National School Lunch Program (PNAE)” study [21-23].

Inclusion criteria corresponded to adolescents’ regular attendance at schools and the age group of interest (10 to 19 years) according to the date of birth recorded in the school’s enrollment document. Exclusion criteria corresponded to the presence of genetic syndromes associated with obesity (Down syndrome and muscular dystrophies), cerebral palsy, adolescent pregnancy, and adolescents with special needs or who used medications such as corticosteroids that could alter the results of biochemical tests. Adolescents were previously evaluated by a team of endocrinologists participating in the study, who identified the presence of these health conditions that excluded study participation.

The sampling plan was defined using a two-stage random stratified sampling strategy based on a population of 39,920 elementary and high school students from 66 municipal schools, and considering the four sanitary districts of the city: north, n = 19,270; south, n = 4,128; west, n = 3,728; and east, n = 12,794. A pilot study was developed in four schools, one in each district of the city, with the purpose of estimating the prevalence of altered lipid profiles in each district. A maximum error of 4% was considered, which resulted in a sample size of 483 adolescents. Stratified sampling using Neyman allocation was performed to define the sample size in each district, as follows: north, n = 285; south, n = 63; west, n = 34; and east, n = 101. However, 711 students were enrolled to compensate for losses of 30% in the sample (which corresponds to 628 students). To determine the number of schools, the average number of students per school was considered, assuming that the variance of that number in the four districts was approximately equal. The school sample size obtained was n = 21, and according to proportional allocation, the following numbers of schools were obtained for each stratum: north, n = 9; south, n = 3; east, n = 3; and west, n = 6.

The selection of schools was conducted by systematic draw, with the sample of students distributed randomly by district and in a manner proportional to the total number of students drawn from the schools. The sample units were selected from the student list, which was used to unify all series with sequential numbering and a systematic draw, without replacement, in case of refusal of the selected participants. For quality control and to reduce sample losses, there were up to two school visits when the number of absentees was higher than 30%. Fig 1 shows the number of adolescents eligible, adolescents enrolled, and details of sample losses. Of the 450 adolescents with complete information on dietary intake, data from six adolescents were excluded because of reported energy intakes <500 and >5000 kcal [24], resulting in a final population of 444 adolescents.

Fig 1

Flowchart of participant selection.

Ethical approval

The study was approved by the Research Ethics Committee of the Federal University of Rio Grande do Norte (No. 112/06). All adolescents and their parents or legal guardians provided informed consent for inclusion before participating in the study. The study is not a clinical trial and, therefore, it does not need to be registered.

Anthropometric assessment and sexual maturation

Weight and height were measured for calculation of body mass index (BMI), which was then classified according to World Health Organization criteria (2007) [25]. All anthropometric measurements (weight and height) were performed at the school in duplicate by trained staff, according to Habicht [26] standardization techniques. An analysis of the reproducibility of the anthropometric measurements showed no difference between the means of the first and second measurements of each trained evaluator. A medical team of pediatric endocrinologists assessed pubertal staging. Sexual maturation was classified according to the Tanner scale into: pre-pubertal (stage 1), initial pubertal (stages 2 and 3), and final pubertal (stages 4 and 5) [27, 28].

Food intake assessment

Food intake data were obtained using two 24-hour dietary recalls administered by trained staff to all study participants at an interval of 30–45 days. Recall surveys were performed according to the recommendation of Thompson and Byers [29], with emphasis on the following criteria: (1) alternate days, except Mondays, avoiding atypical data; (2) distinct periods in each month, considering the purchasing power of families; and (3) different periods of the year, considering the seasonal availability of food. Photographs of utensils and containers were used to identify food-serving items and quantify serving sizes, classified as small, average, or large. The quantity of each food or drink was converted into grams or milliliters using a measurement table that refers to food consumed in Brazil [30]. The foods were converted into energy and nutrients using the software Virtual Nutri Plus® 2.0 (São Paulo, Brazil) [31]. New preparations and foods were added to the software database as necessary, along with their nutritional composition, from the Brazilian food composition table [32] and the United States Department of Agriculture (USDA) [33] database, as appropriate. Nutritional information collected from industrial food labels was also added to the software database.

The nutritional information from both 24-hour recalls were inserted in the Multiple Source Method (MSM) https://msm.dife.de/ [34] to calculate usual dietary intake for individuals and then construct the population distribution based on these data. This method was used to correct dietary data for intra-personal variability in each group of processed and ultra-processed foods. Through the MSM, it is possible to estimate the habitual consumption of healthy individuals, using the probability of consumption and the quantity consumed, or the combination of both estimates, removing measurement error from the data [35].

Food classification according to NOVA and calculation of the prevalence of inadequate micronutrient intake

Foods consumed were classified according to extent and grade of processing using the NOVA classification system proposed by Monteiro et al [4]. This system categorizes food into four groups: (i) unprocessed or minimally processed foods, (ii) processed culinary ingredients designed to be combined with foods to make meals and dishes, (iii) processed foods, and (iv) ultra-processed foods [4,36]. Processed products are foods that have been altered to add substances such as salt, sugar, or oil that substantially change their nature or use. Ultra-processed products are food products formulated mainly or entirely from processed ingredients, typically including little or no whole food. They are durable, edible, drinkable, and palatable by themselves, and are made to be ready-to-consume or ready-to-heat. For the purposes of this study, only the third and fourth groups (processed and ultra-processed products) were analyzed. The foods were then divided into groups as described by Louzada et al [37]. (S1 Table).

Mean and standard deviation for the contribution of each processed or ultra-processed food group to the total energy intake in kilocalories (kcal) were then calculated as a percentage of total energy for each adolescent stratum, corresponding to quartiles of the distribution. The prevalence of inadequate micronutrient intake was estimated using the estimated average requirement (EAR) as the cutoff point for adequate/inadequate intake [38]. The prevalence of inadequate iron intake was calculated using the manually determined probabilistic approach method [39]. The percentage of individuals whose sodium intake exceeded the tolerable upper intake level (UL) was also calculated [40].

Statistical analysis

A descriptive analysis of the data was performed, and the data were presented as a frequency distribution and 95% confidence interval. The percentages of energy obtained from processed and ultra-processed foods in relation to total energy were transformed into ordinal quantitative variables, with cutoff points based on quartiles of the distribution. A dichotomous outcome variable was generated using the EAR cutoff point, according to sex and age group. for each micronutrient from processed and ultra-processed foods, with subjects categorized as having either adequate or inadequate intake. Inferential analysis to estimate the relationship between the energy percentage from processed and ultra-processed foods and the prevalence of inadequate micronutrient intake was developed using generalized linear models with an ordinal logistic distribution for the outcome variable. Initially, unadjusted models were constructed in a bivariate analysis to identify the association between consumption of processed and ultra-processed foods with the prevalence of inadequate micronutrient intake. Variables with p-values ≤ 0.30 in the unadjusted analysis were selected for inclusion in the adjusted model. The selected variables were then inserted into the model with adjustment using the enter method (all covariates entered into the model in a single step), with overall adjusted significance determined by the p-value associated with the omnibus test of the fitted adjusted model. The significance of each odds ratio (OR) was tested using a Wald chi-squared test, and the 95% confidence interval was constructed for both the unadjusted and adjusted OR. The estimated measure of effect (OR) was transformed to a prevalence ratio (PR) using the formula PR = OR(1—P1) / (1—P0), since PR is more appropriate than OR in cross-sectional studies. This methodological choice was made considering that consumption may not interfere alone in only one micronutrient, and this effect needs to be controlled. A significance level of 0.05 was adopted to minimize type I errors in the analysis of the unadjusted, adjusted, and PR models. The software package SPSS Statistics® version 22.0 was used for data storage and analysis.

Results

The adolescents had a mean age of 11.8 years. 36.9% of subjects were classified as pre-pubertal and 25.5% as initial puberty. 21.2% of subjects were overweight or obese, and 45.1% of subject mothers were illiterate or had only attended primary school. The general characteristics of the adolescents are shown in Table 1.

Table 1

General characteristics of adolescents from public schools.

Variables		95% CIa
Age, years (mean; SDb)	11.8 (1.4)	11.7–12.0
Sex (n [%])
Male	223 (50.2)	45.6–54.9
Female	221 (49.8)	45.2–54.4
Sexual maturation (n [%])
Pre-pubescent	164 (36.9)	32.6–41.5
Initial puberty	113 (25.5)	21.6–29.7
Final puberty	167 (37.6)	33.2–42.2
BMIc classification (n [%])
Eutrophy	332 (74.8)	70.5–78.6
Low weight	18 (4.1)	2.6–6.3
Overweight	62 (14.0)	11.1–17.5
Obese	32 (7.2)	5.2–10.0
Maternal Educationd (n [%])
Not literate	17 (4.6)	2.9–7.1
Primary school	152 (40.5)	35.7–45.6
High school or higher education	206 (54.9)	49.9–59.9

aCI = confidence interval

bSD = standard deviation

cBMI = body mass index

dValues may not correspond to the total number of subjects in each group because of missing data

The mean consumption of processed foods ranged from 5.8% (1.7%) of the total energy in Q1 to 20.6% (2.9%) in Q4, while the mean consumption of ultra-processed foods ranged from 21.4% (4.9%) of total energy in Q1 to 61.5% (11.7%) in Q4. The most consumed group of processed foods was French bread, whereas the most consumed ultra-processed foods were cakes, pies, and cookies (Table 2).

Table 2

Contribution (%) of processed and ultra-processed food groups to the total energy intake in adolescents from public schools.

Quartiles of percentage of energy from processed and ultra-processed foods
Food groups	Q1	Q2	Q3	Q4
	Mean (SDa)	Mean (SD)	Mean (SD)	Mean (SD)
Processed
French bread	1.7 (0.4)	4.0 (1.3)	8.6 (1.4)	16.4 (4.3)
Cheeses	0.1 (0.0)	0.2 (0.0)	0.5 (0.2)	4.0 (2.9)
Processed meats	1.5 (0.4)	3.0 (0.5)	5.0 (0.7)	9.9 (3.3)
Canned fruits and vegetables	0.0 (0.0)	0.0 (0.0)	0.0 (0.0)	0.2 (0.3)
Totalb	5.8 (1.7)	10.4 (1.2)	14.8 (1.5)	20.6 (2.9)
Ultra-processed
Cakes, pies, and cookies	7.5 (1.5)	11.4 (1.0)	14.7 (1.1)	19.7 (2.5)
Fast food dishes	0.4 (0.1)	0.7 (0.1)	1.3 (0.4)	8.2 (3.7)
Sugar-sweetened beverages	1.0 (0.2)	1.8 (0.2)	2.6 (0.3)	3.9 (0.8)
Sliced breads	0.9 (0.2)	1.5 (0.2)	5.9 (1.8)	12.7 (3.4)
Bakery products	0.8 (0.2)	1.7 (0.3)	3.1 (0.6)	6.5 (2.5)
Snacks	0.2 (0.1)	0.4 (0.1)	0.9 (0.3)	6.8 (3.7)
Ultra-processed meats	1.0 (0.3)	1.9 (0.3)	3.6 (0.5)	6.3 (1.8)
Ready-to-eat and semi-ready-to-eat meals	0.2 (0.0)	0.3 (0.0)	0.4 (0.0)	3.6 (3.4)
Sweetened milk drinks	0.2 (0.0)	0.2 (0.0)	0.3 (0.0)	3.3 (1.5)
Other ultra-processed foods	1.2 (0.4)	2.5 (0.3)	3.8 (0.4)	6.0 (1.2)
Totalb	21.4 (4.9)	31.5 (2.2)	41.4 (3.9)	61.5 (11.7)

aSD = standard deviation

bExpressed as a percentage of total energy intake.

One hundred percent of both sexes aged 10–13 years displayed inadequate intake of vitamin D, folate, vitamin E, and calcium. More than 80% of adolescents aged 10–13 years had an inadequate intake of phosphorus and selenium. For girls in the same age range, there was also a high prevalence of inadequate vitamin B6 intake (50.4%) (Table 3). The mean sodium intake was high (3549.9 mg (948.1 mg) in boys and 3316.9 mg (838.7 mg) in girls), indicating that more than 93.5% of adolescents had sodium intake that exceeded the upper limit.

Table 3

Daily nutritional recommendations, micronutrient intake, and prevalence of inadequacy of micronutrient intake (%IN) by sex and age group in adolescents from public schools.

Micronutrients	Malea						Femaleb
	EARc/ULd	Mean (SDe)	10th	50th	90th	%INf	EAR/UL	Mean (SD)	10th	50th	90th	%IN
Vitamin A (μg)
10 to 13 y	445	885.3 (539.9)	402.4	827.3	1275.3	20.9	420	822.5 (410.7)	334.5	766.6	1308.4	16.3
14 to 18 y	630	737.5 (248.7)	293.1	787.2	1064.0	33.4	485	1077.9 (1008.0)	410.9	731.9	2701.9	27.8
Vitamin C (mg)
10 to 13 y	39	162.9 (130.8)	55.6	118.0	338.7	17.1	39	161.5 (144.6)	42.5	106.5	383.3	19.8
14 to 18 y	63	166.4 (143.3)	49.8	134.3	359.0	23.6	56	161.1 (180.1)	34.7	100.8	523.1	28.1
Vitamin B1 (mg)
10 to 13 y	0.7	1.1 (0.4)	0.7	1.0	1.6	17.4	0.7	1.1 (0.4)	0.6	1.0	1.5	21.2
14 to 18 y	1.0	1.4 (0.8)	0.8	1.3	1.8	30.2	0.9	0.9 (0.2)	0.5	0.9	1.3	47.2
Vitamin B2 (mg)
10 to 13 y	0.8	1.1 (0.4)	0.7	1.1	1.7	21.5	0.8	1.1 (0.6)	0.6	1.0	1.6	30.1
14 to 18 y	1.1	1.3 (0.5)	0.7	1.1	2.1	37.5	0.9	1.1 (0.4)	0.6	0.9	1.6	33.4
Vitamin B6 (mg)
10 to 13 y	0.8	0.9 (0.3)	0.5	0.8	1.3	41.7	0.8	0.8 (0.3)	0.5	0.8	1.2	50.4
14 to 18 y	1.1	1.0 (0.3)	0.6	0.9	1.5	62.9	1.0	0.8 (0.3)	0.5	0.7	1.4	74.2
Vitamin B12 (μg)
10 to 13 y	1.5	1.7 (1.9)	0.4	1.2	3.5	46.4	1.5	1.5 (1.6)	0.3	1.1	3.3	49.2
14 to 18 y	2.0	1.3 (1.0)	0.1	1.0	3.4	75.5	2.0	1.9 (1.6)	0.3	1.7	5.8	51.6
Vitamin D (μg)
10 to 13 y	10	0.1 (0.1)	0.1	0.1	0.2	100	10	0.1 (0.1)	0.1	0.1	0.3	100
14 to 18 y	10	0.2 (0.1)	0.1	0.2	0.3	100	10	0.1 (0.1)	0.1	0.1	0.3	100
Folate (μg)
10 to 13 y	250	89.8 (48.2)	38.8	78.4	155.6	99.9	250	91.2 (57.2)	33.4	77.0	163.4	99.7
14 to 18 y	330	123.1 (87.5)	22.7	109.7	291.1	99.1	330	86.4 (35.3)	42.5	88.7	132.9	100
Vitamin E (mg)
10 to 13 y	9.0	2.9 (1.9)	1.0	2.4	5.6	99.9	9.0	3.1 (2.4)	0.8	2.4	6.0	99.2
14 to 18 y	12	3.3 (2.9)	0.8	2.6	7.0	99.9	12	4.1 (3.7)	1.1	2.7	8.4	98.4
Sodium (mg)g
10 to 13 y	2200	3504.6 (919.6)	2430.0	3369.8	4640.4	–	2200	3341.5 (850.4)	2333.9	3310.4	4529.0	–
14 to 18 y	2300	4067.1 (1132.5)	2506.1	3972.1	5779.0	–	2300	3039.4 (648.9)	2199.8	2927.2	4193.7	–
Calcium (mg)
10 to 13 y	1100	398.8 (140.1)	249.4	367.2	620.1	100	1100	380.4 (128.6)	221.0	353.0	561.0	100
14 to 18 y	1100	439.3 (117.2)	292.8	436.9	563.8	100	1100	371.3 (125.6)	188.9	385.8	532.7	100
Zinc (mg)
10 to 13 y	7.0	10.5 (3.4)	6.8	10.0	14.5	15.2	7.0	9.4 (2.7)	5.9	9.3	12.8	18.7
14 to 18 y	8.5	11.4 (2.8)	7.6	11.3	14.9	15.2	7.3	8.5 (1.4)	6.6	8.7	10.1	19.2
Phosphorus (mg)
10 to 13 y	1055	876.4 (198.0)	647.6	845.0	1165.1	81.6	1055	829.0 (189.5)	600.8	827.1	1072.7	88.3
14 to 18 y	1055	959.4 (180.8)	768.9	919.0	1293.2	70.2	1055	811.9 (175.9)	561.0	835.6	1053.3	91.6
Iron (mg)
10 to 13 y	5.9	11.9 (3.2)	8.1	11.5	16.3	1.3	5.7	11.1 (3.0)	7.4	10.9	15.1	3.6
14 to 18 y	7.7	13.9 (4.8)	7.4	14.6	19.8	11.1	7.9	10.1 (2.8)	7.4	9.2	14.0	29.2
Selenium (μg)
10 to 13 y	35	26.2 (9.9)	14.3	25.2	40.3	81.3	35	25.3 (9.9)	13.6	24.5	37.0	83.6
14 to 18 y	45	30.9 (15.1)	12.4	28.4	61.4	82.4	45	24.3 (7.4)	11.5	24.1	33.5	99.7

amale 10 to 13 years (n = 205); 14 to 18 years (n = 18)

bfemale 10 to 13 years (n = 203); 14 to 18 years (n = 18)

cEAR = estimated average requirement

dUL = tolerable upper intake level

eSD = standard deviation

fIN = inadequate

gsodium intake was analyzed using UL values.

The percentage of energy consumed from processed foods was significantly associated with inadequate selenium intake, suggesting that adolescents who consumed the highest levels of processed foods were almost twice as likely to have inadequate selenium intake than subjects who consumed the lowest levels of processed foods (p < 0.01, PR = 1.97, 95% CI: 1.22–3.1). However, consumption of processed foods was associated with a lower probability of vitamin B1 inadequacy (p = 0.04, PR = 0.55, 95% CI: 0.31–0.98; Table 4). The percentage of energy consumed from ultra-processed foods was also significantly associated with a lower probability of vitamin B1 inadequacy (p = 0.03, PR = 0.49, 95% CI: 0.25–0.94). In addition, a lower probability of inadequate zinc intake (p < 0.01) was observed for subjects with the highest levels of energy consumption from ultra-processed foods (p < 0.01, PR = 0.43 95% CI 0.23–0.78), while no association was noted between zinc intake and processed food levels (Table 4). No association was found between the proportion of energy obtained from processed or ultra-processed foods and the prevalence of inadequate intake of other micronutrients (all p > 0.05; Table 4).

Table 4

Models of the unadjusted and adjusted association between proportion of energy from processed and ultra-processed foods and the prevalence of inadequate micronutrient intake in adolescents from public schools.

Parametera	Unadjusted Model			Adjusted Modelb
	PRc	95% CId	P-value	χ2 of Wald	P-value	PR	95% CI
Processed foodse
1st quartile				4.09	0.04	12.96	(1.08–155.27)
2nd quartile				1.23	0.26	4.06	(0.34–48.38)
3rd quartile				0.04	0.84	1.28	(0.10–15.29)
4rd quartile	1
Vitamin A
inadequate	1.39	(0.86–2.24)	0.17	1.14	0.28	1.31	(0.79–2.17)
adequate	1					1
Vitamin B1
inadequate	0.58	(0.36–0.94)	0.02	4.06	0.04	0.55	(0.31–0.98)
adequate	1					1
Vitamin B2
inadequate	0.73	(0.50–1.07)	0.10	0.43	0.51	0.84	(0.51–1.38)
adequate	1					1
Vitamin B6
inadequate	0.79	(0.56–1.10)	0.17	1.53	0.21	0.76	(0.50–1.16)
adequate	1					1
Vitamin B12
inadequate	1.12	(0.86–1.74)	0.25	0.84	0.35	1.19	(0.81–1.73)
adequate	1					1
Folate
inadequate	2.87	(0.45–18.32)	0.26	0.45	0.50	1.88	(0.29–11.87)
adequate	1					1
Phosphorus
inadequate	1.70	(1.04–2.78)	0.03	2.53	0.11	1.56	(0.90–2.71)
adequate	1					1
Selenium
inadequate	2.05	(1.31–3.21)	<0.01	7.69	<0.01	1.97	(1.22–3.17)
adequate	1					1
Ultra-processed foodsf
1st quartile				4.53	0.033	0.149	(0.02–0.85)
2nd quartile				12.01	0.001	0.044	(0.008–0.25)
3rd quartile				22.12	0.001	0.014	(0.002–0.08)
4rd quartile	1
Vitamin C
inadequate	0.37	(0.18–0.75)	<0.001	2.40	0.12	0.54	(0.25–1.17)
adequate	1					1
Vitamin B1
inadequate	0.33	(0.19–0.57)	<0.0001	4.56	0.03	0.49	(0.25–0.94)
adequate	1					1
Vitamin B2
inadequate	0.69	(0.47–1.02)	0.06	2.47	0.11	1.50	(0.90–2.50)
adequate	1					1
Vitamin B6
inadequate	0.54	(0.38–0.75)	<0.001	3.04	0.08	0.68	(0.45–1.04)
adequate	1					1
Vitamin E
inadequate	0.45	(0.17–1.20)	0.11	2.51	0.11	0.44	(0.16–1.21)
adequate	1					1
Sodium
inadequate	0.27	(0.12–0.59)	0.001	0.60	0.43	0.69	(0.28–1.73)
adequate	1					1
Zinc
inadequate	0.29	(0.17–0.48)	<0.001	7.60	<0.01	0.43	(0.23–0.78)
adequate	1					1
Iron
inadequate	0.10	(0.02–0.49)	<0.01	0.73	0.39	0.47	(0.08–2.65)
adequate	1					1

aQ1 vs Q4 comparison for each micronutrient

bThe final model was adjusted with all the variables presented in the table.

cPR = prevalence ratio

dCI = confidence interval

eVitamins D, C, and E, and the minerals calcium, zinc, sodium, and iron were excluded from the adjusted model since they were not significant in the unadjusted analysis (p > 0.30).

fVitamins A, B12, D and E, folate, and the minerals calcium, phosphorous, and selenium were excluded from the adjusted model since they were not significant in the unadjusted analysis (p > 0.30).

Discussion

This study of students in public schools in an urban area in the city of Natal, northeastern Brazil, confirms an increase in the proportion of energy obtained from processed and ultra-processed foods in adolescent diets. This is consistent with the findings of other studies carried out in Brazil in the pediatric population [5, 10, 15]. The intake of processed foods showed significant associations with the prevalence of inadequacy of selenium intake. An increased proportion of energy from processed and ultra-processed foods was also associated with a lower probability of vitamin B1 inadequacy. In addition, energy consumption from ultra-processed foods was associated with a lower probability of inadequate zinc intake.

The Brazilian government National Adolescent School-based Health Survey (PeNSE 2015) reinforces that this new food behavior affects all of Brazil’s socioeconomic levels and regions, considering that 40% of schoolchildren reported daily consumption of at least one group of ultra-processed foods [19]. Accordingly, strategies to promote healthy eating and decrease sedentary behavior, as well as regulation of advertising for ultra-processed foods, are necessary to prevent unhealthy lifestyles from persisting into adulthood [3].

There were low percentages of energy intake from ready-to-eat and semi-ready-to-eat meals and sweetened milk drinks from the ultra-processed foods group. This finding could be attributed to the fact that most families living in northeastern Brazil are characterized as low-income families [41, 42], as an increase in the intake of ready-to-eat and semi-ready-to-eat meals and sweetened milk drinks has been associated with an increase in family income per capita [42].

Previous work points out that diets based on traditional Brazilian foods, characterized by high intake of beans and rice, are more frequent among adolescents [43]. The Pure Traditional Food System pattern, which consists of beans and a mixture of characteristic ingredients (e.g., onions, peppers, and tomatoes) was predominant in this study population of adolescents from northeastern Brazil [22]. Apart from beef, the foods comprising this pattern are of low cost, and few food items were associated with increased micronutrient intake, suggesting monotony and low nutritional quality in the diet of the less privileged socioeconomic classes [43].

Adolescents’ intake of these lower cost foods with low nutritional quality may be associated with a higher prevalence of vitamin D, calcium, and phosphorus intake inadequacy, with implications for bone health in this stage of life [44, 45]. The high prevalence of inadequate vitamin D intake observed in our study indicates an alarming situation, since one of the main actions of vitamin D is the regulation of calcium and phosphorus metabolism, controlling the processes of intestinal absorption and renal reabsorption of these ions [45].

The high prevalence of inadequate calcium intake observed in our study may be partly explained by the persistently low consumption of dairy products among adolescents. Other Brazilian studies have also shown low milk consumption and increased soft drinks or sugary drinks consumption [46, 47]. Similarly, Mexican adolescents aged 12–19 years showed high calcium inadequacy, reaching 88.1% in females and 71.8% in males [48]. During adolescence, there is accelerated growth, with greater anthropometric and body composition changes, which require an adequate supply of calcium, vitamin D, and phosphorus. These nutrients are essential in guaranteeing the achievement of growth and development [49].

The high prevalence of inadequate vitamin E and selenium intake was similar to those of the ERICA findings, reflecting the choice of ultra-processed foods with lower micronutrient content [18]. These micronutrients are recognized for their antioxidant roles in the body, and act synergistically to maintain glutathione peroxidase capacity. Importantly, an imbalance in these nutrients can favor oxidative stress, causing dysregulation in signaling and/or cell damage [50].

Food fortification with folate has been implemented in several countries. In Brazil, fortification of wheat and maize flours has been implemented since 2004, with the addition of 150 mg of folic acid and 4.2 mg of iron per 100 g of flour [51]. In this study, although adolescents presented with a high intake of processed and ultra-processed foods made with flour, a high prevalence of folate inadequacy was observed. Fiorentino et al. [52] also confirmed a high prevalence of folate intake inadequacy among Senegalese children and adolescents. In our study, this fact could be attributed to a lack of information on folate in Brazilian food composition tables [32], including fortified foods. Therefore, the prevalence of folate intake inadequacy has been overestimated.

The prevalence of iron intake inadequacy in girls is notable because, during menarche, dietary iron requirements are increased [53]. However, as observed with folate intake, the prevalence of iron inadequacy, while low, may be overestimated, considering the possible failures of Brazilian food composition tables [32].

Our results regarding excessive sodium intake confirm previous findings obtained in Brazil [18, 54] and Europe [8]. Although the adjusted model did not reach the level of significance to suggest an association between processed and ultra-processed food consumption and sodium intake, the high sodium intake observed among the adolescents in this study may be related to high intake of these types of foods. This is worrying because high sodium intake among adolescents may be associated with risk of cardiometabolic disease [55].

The most significant finding we observed pertained to the association between processed foods consumption and selenium intake inadequacy. Louzada et al. [56] also observed a relationship between ultra-processed food intake and low consumption of selenium in Brazilians aged over 10 years. Different authors have pointed out the difficulties in assessing selenium content in food because it depends on climatic conditions, cultivation and breeding methods. The concentration of this element in foods also varies depending on the species, plant part, concentration of selenium in the soil, and the ability of plants to accumulate this element [57]. In our study, we observed low intake of foods that are significant sources of selenium, such as fish, whole grains, and Brazil nuts. A previous study has shown that selenium inadequacy negatively affects the thyroid metabolism of iodine-replete children and may present a substantial public health concern. The authors reinforce the importance of selenium status for normal thyroid function, as well as others roles of this element in biological systems [58].

An unexpected finding in our study was the associations between the consumption of processed and ultra-processed food and lower probability of inadequate vitamin B1. In addition, a higher consumption of ultra-processed foods was also significantly associated with a lower probability of inadequate zinc intake. It is known that in the process of producing white flour, much of vitamin content is lost [59]. In the present study, a higher intake of foods produced with white flour, such as cakes, pies, and biscuits, as well as breads, was observed as the most commonly consumed food group among subjects from both groups. However, a possible explanation for this association is the fact that the loss of vitamins during food processing has led the industry to add vitamins and minerals to various foodstuffs, with the goal of reducing nutritional deficiencies in the population [60]. This fortification may explain the observed association between high consumption of processed foods and lower probability of inadequacy of vitamin B1 and zinc.

A previous study conducted with a group of Brazilian children aged 2–3 years has indicated a low prevalence of iron, vitamin C, vitamin A, calcium, and folate inadequacy. None of these children had intakes less than the EAR value for zinc. However, 4·0% of children exceeded the UL for vitamin A, 3.1% for zinc, 1.1% for folic acid, and 0.2% for iron. The authors emphasize that the consumption of ultra-processed foods, usually fortified foods, contributed to micronutrient supply, but may increase the risk of excessive micronutrient intake [61]. Authors point out that enriched or fortified foods in the US are an important source of micronutrients. According to this line of reasoning, variation in ultra-processed food consumption does not necessarily result in nutritional imbalances [62].

Previous studies have addressed the controversy over the association between consumption of ultra-processed foods and the proportion of individuals whose intakes are below the micronutrient requirements, as well as the difficulty in setting adequate and safe micronutrient levels in foods without increasing the risk of consuming amounts above the UL [63]. The bioavailability of micronutrients depends on food composition and possibly on its encapsulation method in some specific formulations. Furthermore, ultra-processed foods do not provide the same health benefits as natural foods [4]. Beyond the addition of synthetic micronutrients that can exceed the UL, ultra-processed foods are energy-dense, with high levels of fat, sugar, and/or sodium, contributing to the increased risk of obesity and noncommunicable diseases [64].

Our study has some limitations. We faced difficulties in assessing food consumption due to the incompleteness of the food composition tables of available foods. However, statistical techniques were used to minimize the probability of the occurrence of random errors. Another limitation is the number of days taken to evaluate the habitual intake of nutrients, since this number may vary between nutrients. Therefore, evaluation of the habitual intake of some of the micronutrients may be imprecise.

Conclusions

In conclusion, an increase in the proportion of energy obtained from processed and ultra-processed foods by adolescents from public schools, was associated with an increase in the prevalence of inadequate selenium intake, which could lead to short-term and long-term negative health consequences. However, energy consumption from processed and ultra-processed foods was associated with a lower probability of inadequate vitamin B1 intake. In addition, a lower probability of inadequate zinc intake was observed only for energy consumption from ultra-processed foods, possibly attributed to food fortification. Considering the risk of micronutrient intake exceeding the UL, and the fact that ultra-processed foods are energy-dense, with high levels of fat, sugar, and/or sodium, we believe that it is crucial to stress the importance of limiting personal consumption of processed and ultra-processed foods, especially among the pediatric population. The position of the Academy of Nutrition and Dietetics highlights that early care and education programs should recommended, from an early age, a high standard of nutritional quality for foods and beverages served. Efforts to promote healthy eating should be a national priority and include state and local policy changes [65]. Accordingly, the results of this study can help provide evidence for formulating global public policies that improve health in adolescence and prevent complications in adulthood.

Classification of foods and ingredients according to their industrial processing characteristics.

aAdapted from Louzada et al.[37].

(DOCX)

Click here for additional data file.

27 Aug 2019

PONE-D-19-18574

Processed and ultra-processed foods are associated with high prevalence of inadequate selenium intake and low prevalence of vitamin B1 and zinc inadequacy in adolescents from public schools

PLOS ONE

Dear Dr. Sena-Evangelista,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The manuscript addressed an important public health nutrition problem (consumption of processed and ultra-processed foods) in a vulnerable (adolescent) population and present the association between consumption of processed and ultra-processed foods and intake of selected micronutrients. The manuscript is also well written and properly organized. However, as stated below, I have serious concerns on the approach of analysis employed.

1. Line 242-3: “……the energy percentage from processed and ultra-processed foods and the prevalence of inadequate micronutrient intake was developed using generalized linear models with an ordinal logistic distribution”. It is not clear which one of the two is the dependent variable of the analysis (prevalence of inadequate micronutrient intake or percentages of energy obtained from processed and ultra-processed foods). As long as the purpose of the study is to see the association between consumption of processed and ultra-processed foods (kind of exposure) with the prevalence of inadequate micronutrient intake (kind of outcome); micronutrient intake should be the outcome/dependent variable. And what you should have used is binary not ordinary logistic regression analysis.

2. I really don’t see the purpose of adjusting one micronutrient nutrient intake for the other. Do you really expect confounding among them? What I propose is multiple bivariable models for each nutrient, by which the dependent variable the nutrient intake and the independent variable is the percentages of energy obtained from processed and ultra-processed foods.

We would appreciate receiving your revised manuscript by Oct 11 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Samson Gebremedhin Gebreselassie, PhD

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

- Kieliszek, Marek, and Stanisław Błażejak. "Current knowledge on the importance of selenium in food for living organisms: a review." Molecules 21.5 (2016): 609.

- Gashu, Dawd, et al. "Selenium inadequacy hampers thyroid response of young children after iodine repletion." Journal of Trace Elements in Medicine and Biology 50 (2018): 291-295.

- Sangalli, Caroline Nicola, Fernanda Rauber, and Márcia Regina Vitolo. "Low prevalence of inadequate micronutrient intake in young children in the south of Brazil: a new perspective." British Journal of Nutrition 116.5 (2016): 890-896.

The text that needs to be addressed involves some sentences of the Discussion.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

Additional Editor Comments:

Title: Please indicate the setting in the title

Abstract

• In the abstract please concisely indicate how foods were classified into processed and ultra-processed foods. You may simple indicate the “NOVA” classification system was used.

• Background: Line 69-73; can you provide some common examples for processed and ultra-processed foods so that readers can easily understand the differences between the two?

Methods:

• line 141 and 150: it is not clear why the study considered primary school students as source population while the age group of interest is 10-19 years of age.

• Line 153-154: ample size is calculated based on the assumption that the outcome of interested is altered lipid profiles in each district and this is not directly related with the current manuscript. How do you assure the adequacy of the sample size/power to address the objective of this specific manuscript?

• Line 189-90: the sentence “Recall surveys were performed at an interval of 30–45 days, according to the recommendation of Thompson and Byers” is not clear. Please provide additional description of this approach.

• Line 247: “Variables with p-values ≤ 0.30…….” can you please provide the list of variables considered for the bivariable analysis? As I stated above, I really question the importance of multivariable analysis here.

• Line 245-7: As I commented above, please also clearly indicate the dependent and independent variables of the model.

Results

• Please provide the operational definitions employed to classify the sexual maturation of adolescents.

• Line 303: “high prevalence of pyridoxine intake” confusing, high or low intake?

• Table 3: can you please explain why you limit yourself to the 15 micronutrients provided in the table?

• 97% more likely > almost two times more likely.

• Why it was not possible to present table 4 and 5 together in one table?

• Table 4 and 5: Please clearly describe the dependent and independent variables, I also propose for removing the multivariable analysis.

Table 4 and 5: as long as p values are there I don’t see the purpose of having the Wald statistic there.

Others

• Multivariate > Multivariable; bivariate > bivariable

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Processed and ultra-processed foods are associated with high prevalence of inadequate

selenium intake and low prevalence of vitamin B1 and zinc inadequacy in adolescents from

public schools.

The study addresses an interesting and relevant subject showing the association of consumption of processed and ultra-processed foods with micronutrient inadequacy in adolescents. However, the MS lacks logical presentation of information hence is difficult to understand.

The introduction doesn’t contain the relevant background information to guide readers to the rest of the manuscript.

The justification or motivation to do this research is not well formulated.

The language not clear.

The outcome of this study mainly depends on the quality of the food composition table used to convert into nutrient intake of the present study subjects. However, this may not work for some nutrients such as selenium. This is because; selenium concentration in foods is mainly determined by the amount of the element in the soil which is quite variable in short distances. Thus, data from food composition is not reliable.

Parts of the methodology lack clarity and are not detailed.

The authors in general tried to show the effect of level of processing on micronutrient intake inadequacy that consumption of processed foods was associated with lower intake of selenium but not Zn and B1. Processed foods in general are poor in micronutrient content. However, it is failed to identify whether foods in the present study were enriched with Zn and vitamin B1 concentration.

For example,

Line 48-50, Energy consumption from processed foods was associated with higher prevalence of inadequate selenium intake (p < 0.01) but lower prevalence of inadequate vitamin B1 intake (p = 0.04).

Line 50-54, Energy consumption from ultra-processed foods was associated with lower prevalence of inadequate zinc and vitamin B1 intake (p < 0.01 and p =0.03, respectively). An increase in the proportion of energy obtained from processed and ultra-processed foods may reflect higher prevalence of inadequate selenium intake and lower prevalence of vitamin B1 and zinc inadequacy.

Line 74-75. It is not always true to claim that sedentary behaviors among adolescents lead to greater consumption of packed or ready to eat foods.

Methodology

Line 136-140: Ethical consideration/approval has to be presented separately

Line 141-148: The inclusion and Inclusion criteria may work for the main study “Risk factors for cardiovascular disease among adolescent beneficiaries of the Brazilian government National School Lunch Program (PNAE)” but not for this specific MS.

The sample size was determined considering altered lipid profile as an outcome which is not consistent with the present study outcome.

Weight and height measurement procedure is not detailed

Line 179: ‘The morning was chosen to more easily enable methodological procedures’. This is not clear. Does this mean ‘anthropometric measurement was done during the morning? Why?

Line 182-185: It is not clear the importance of assessing pubertal stage given the age of study participants was known for this particular study.

Reviewer #2: 1) Abstract

2) Introduction - it would be good to also have some systematic review to be cited to reiterate how the unhealthy dietary intake from low middle income countries/ Asian as well to show this is a global burden.

3) - It is noted in methods the age of interest is 10-19. However, based on the results it appears only mean age 11.8 years. Why the age group seems to be specific (10-13)?

- Explain why just use 24 hours diet recall and repeated twice. Is it the most suitable for adolescents studies? Since your interest is micronutrient and ultra processed food, which method of dietary intake should be more robust to be used for this adolescents group? The inadequate number of days for this might influence why you discovered such results.

- what was the study response rate?

How reliable is the reported Tanner staging? Any confirmation by the paediatrician?

Results: The age group for 14 and above only 18? why is it so low? If this group is excluded, how different the results will be in table 3.

Figure - Out of 4457 subjects only 700+ enrolled. Why?

Since the two-stage random stratified sampling is used, why complex sample analysis not conducted?

It appears highlighted micronutrient intake was low but do you have biomarkers to associate with?

Discussion: If ultra processed will lead to reduce certain micronutrient content, will enrichment with that vitamins in the food minimise the problem of suboptimal nutrient?

Research has shown the association ultra-processed foods that contain added sugars, excess sodium, and unhealthful fats possibly lead to poor health outcome and is this pattern the same with yours?

In the conclusion - "stress the importance of limiting personal consumption of processed and ultra-processed", how and where at school, home or institution and what is the allowed value? If there is any? What would be the most suitable take home message?

Reviewer #3: Abstract

-Numerous sentences, such as the one highlighted below, starting with 'this':

"This was a cross-sectional study.”

-The term 'processed and ultra-processed foods’ is being used several times. It is very repetitive.

-Insert the acronym for 'estimated average requirement

Introduction

-The information stated in lines 96 and 97 has already been said in the introduction’s section elsewhere.

-The paper discusses that some results found might be due to differences of eating habits or economic status of northeastern Brazil, in comparison to the rest of the country, but it does not mention any of this in the introduction. The authors need to expand on this.

Methods

-Were rural schools included? The access to milk and dairy products could be different compared to the population from the urban ones.

-In Fig 1, why are there 700 adolescents enrolled when it is mentioned above only 11 were transferred from school?

-What about those participants who were taking dietary supplements? Was this type of information collected or evaluated?

-Was the apportioning and quantification of food reviewed by the 24-hour dietary recalls' interviewers ?

Discussion

-"We observed low percentages of energy intake from ready-to-eat and semi-ready-to- eat meals and sweetened milk drinks from the ultra-processed foods group.” I would prefer the passive voice.

-"We observed low percentages of energy intake from ready-to-eat and semi-ready-to- eat meals and sweetened milk drinks from the ultra-processed foods group. This finding could be attributed to the fact that most families living in northeastern Brazil are characterized as low-income families [39, 40], as an increase in the intake of ready-to-eat and semi-ready-to- eat meals and sweetened milk drinks has been associated with an increase in family income per capita [40].” Is it really different from "cakes, pies, and cookies"?

-“In our study, we observed low intake of foods that are significant sources of selenium, such as fish, whole grains, and Brazil nuts.” Can this information be drawn from table 2?

-“Brazilian children aged 2–3 years typically have a low prevalence of iron, vitamin C, vitamin A, calcium, and folate inadequacy. None of these children had intakes less than the EAR value for zinc. Fortified foods contributed to micronutrient supply. However, 4·0% of children exceeded the UL for vitamin A, 3.1% for zinc, 1.1% for folic acid, and 0.2% for iron. These results suggest a low prevalence of inadequate micronutrient intake among children, with the implication that this group could be at risk of excessive micronutrient intake provided by ultra-processed foods [59].” Isn't this paragraph deviating from the population targeted at this paper?

-According to results presented, processed and ultra-processed foods are associated with a low prevalence of vitamin B1 and zinc inadequacy. The authors need to expand on this.

-What are the limitations of the 24-hour dietary recall?

-The results need to be further compared with the literature.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PLOS ONE comments_PONE-D-19-18574.docx

Click here for additional data file.

14 Oct 2019

Editor

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The manuscript addressed an important public health nutrition problem (consumption of processed and ultra-processed foods) in a vulnerable (adolescent) population and present the association between consumption of processed and ultra-processed foods and intake of selected micronutrients. The manuscript is also well written and properly organized. However, as stated below, I have serious concerns on the approach of analysis employed.

Line 242-3: “……the energy percentage from processed and ultra-processed foods and the prevalence of inadequate micronutrient intake was developed using generalized linear models with an ordinal logistic distribution”. It is not clear which one of the two is the dependent variable of the analysis (prevalence of inadequate micronutrient intake or percentages of energy obtained from processed and ultra-processed foods). As long as the purpose of the study is to see the association between consumption of processed and ultra-processed foods (kind of exposure) with the prevalence of inadequate micronutrient intake (kind of outcome); micronutrient intake should be the outcome/dependent variable. And what you should have used is binary not ordinary logistic regression analysis.

Answer: We would like to thank the editor for this comment. The purpose of the study is to evaluate the association between consumption of processed and ultra-processed foods with the prevalence of inadequate micronutrient intake. Considering the cross-sectional design of this study, one of the disadvantages is that the researcher does not know temporally what happens first, in this case, inadequacy of nutrients or consumption of processed and ultra-processed foods. We examine only the association and not the cause-effect relationship. Even if there is the assumption of this cause-effect relationship, there is no way to prove which occurred first. This methodological choice was made considering that consumption possibly does not interfere alone with only one micronutrient, and this effect needs to be controlled. Thus, with a cross-sectional design, the variables are placed interchangeably on different sides of the equation.

In addition, the bivariate or crude analysis was performed and it is as shown in Table 4 and specified as “Unadjusted Model”. Initially, unadjusted models were constructed in a bivariate analysis to identify associations between the prevalence of inadequate micronutrient intake and food energy percentage of processed and ultra-processed foods. Cross-sectional studies with binary outcomes analyzed by logistic regression are frequent in the epidemiological literature1.

____________________________________________________________

1Barros AJ, Hirakata VN. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol. 2003 Oct 20;3:21.

2. I really don’t see the purpose of adjusting one micronutrient nutrient intake for the other. Do you really expect confounding among them? What I propose is multiple bivariable models for each nutrient, by which the dependent variable the nutrient intake and the independent variable is the percentages of energy obtained from processed and ultra-processed foods.

Answer: We are thankful to the editor for this comment. Multiple bivariate analyses cause a statistical problem called “type I error inflation” according to biomedical and biopsychosocial theoretical models, there is an interrelationship between health conditions. Thus, regarding food intake, the authors considered that it is necessary to control the covariance of inadequate micronutrient intake, as inadequate vitamin A intake may be possibly related to inadequate vitamin C intake, for example.

When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

- Kieliszek, Marek, and Stanisław Błażejak. "Current knowledge on the importance of selenium in food for living organisms: a review." Molecules 21.5 (2016): 609.

- Gashu, Dawd, et al. "Selenium inadequacy hampers thyroid response of young children after iodine repletion." Journal of Trace Elements in Medicine and Biology 50 (2018): 291-295.

- Sangalli, Caroline Nicola, Fernanda Rauber, and Márcia Regina Vitolo. "Low prevalence of inadequate micronutrient intake in young children in the south of Brazil: a new perspective." British Journal of Nutrition 116.5 (2016): 890-896.

The text that needs to be addressed involves some sentences of the Discussion.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

Answer: We are thankful to the editor for this comment. The overlapping text has already been rewritten in the Discussion section (Page 26, lines 451-455; lines 458-460; Page 26 and 27, lines 473-480).

Additional Editor Comments:

Title: Please indicate the setting in the title

Answer: We appreciate the editor’s comment. We have added this information in the “Title section “…in an urban area of northeastern Brazil”.

Abstract

In the abstract please concisely indicate how foods were classified into processed and ultra-processed foods. You may simple indicate the “NOVA” classification system was used.

Answer: We are thankful to the editor for this comment. The information has been added in the “Abstract section”.

Background

Line 69-73; can you provide some common examples for processed and ultra-processed foods so that readers can easily understand the differences between the two?

Answer: We are thankful to the editor for this comment. The examples have been provided in the” Introduction section” (Page 2, lines 69-70, and lines 73-74).

Methods:

line 141 and 150: it is not clear why the study considered primary school students as source population while the age group of interest is 10-19 years of age.

Answer: We are thankful the editor for this comment. We recognize this mistake. The text has already been rewritten correctly (Methods section- Study population, page 6, line 148).

Line 153-154: Sample size is calculated based on the assumption that the outcome of interested is altered lipid profiles in each district and this is not directly related with the current manuscript. How do you assure the adequacy of the sample size/power to address the objective of this specific manuscript?

Answer: We are thankful the editor for this comment. The sampling plan was defined for the study “Risk factors for cardiovascular disease among adolescents beneficiary of the Brazilian government National School Feeding Program (PNAE)”, which aimed to explore risk factors for cardiovascular diseases in adolescents. However, diet was included as one of the environmental variables of the study.

Line 189-90: the sentence “Recall surveys were performed at an interval of 30–45 days, according to the recommendation of Thompson and Byers” is not clear. Please provide additional description of this approach.

Answer: We would like to thank the editor for this comment. Since day-to-day variability in diet is high, the information from a single day cannot accurately reflect the usual diet of an individual. Other authors suggest collecting information on nonconsecutive days, since consecutive days suffer from the error in one day may be correlated with the error in the next day, and thus are not independent assessments1. Thus, we decided to apply the 24-hour dietary recall on 2 non-consecutive days, and set an interval of 30-45 days between the first and second 24-hour dietary recall collection. This information has already been rewritten for the purpose of clarification (Materials and Methods section, page 8, line 193-194).

______________________________________________________

1Frances E. Thompson, Amy F. Subar, Catherine M. Loria, Jill L. Reedy, Tom Baranowski Need for Technological Innovation in Dietary Assessment. J Am Diet Assoc. 2010 Jan; 110(1): 48–51. doi: 10.1016/j.jada.2009.10.008

Line 247: “Variables with p-values ≤ 0.30…….” can you please provide the list of variables considered for the bivariable analysis? As I stated above, I really question the importance of multivariable analysis here.

Answer: We appreciate the editor’s comment. The authors chose to use a relatively permissive p-value (0.3) to select candidate variables to be included in the multiple regression model to avoid the possible exclusion of important variables, based on an arbitrary statistical criterion1. Importantly, this p-value cutoff was only applied during the initial selection of variables to be included in the adjusted models. Standard p-value cutoffs (p = 0.05) were used for all final models. The variables considered for bivariable analysis are shown in Table 4 (vitamin A, B1, B2, B6, B12, folate, phosphorus, and selenium for processed foods, and vitamin B1, B2, B6, E, sodium, zinc, and iron for ultra-processed foods).

_________________________________________________

1Martínez-González, M.A., Sánchez-Villegas, A., Fajardo, J.F. Bioestadística Amigable. 3. ed. Madrid: Diaz de Santos. 2014. 612 p.

Line 245-7: As I commented above, please also clearly indicate the dependent and independent variables of the model.

Answer: We are grateful to the editor for this comment. The purpose of the study is to evaluate the association between consumption of processed and ultra-processed foods with the prevalence of inadequate micronutrient intake. Considering the cross-sectional design of this study, one of the disadvantages is that the researcher does not know temporally what happens first, in this case, the inadequacy of nutrients or consumption of processed and ultra-processed foods. We examine only association and not cause-effect. Even if there is the assumption of this cause-effect relationship, there is no way to prove which occurred first. This methodological choice was made considering that consumption possibly does not interfere alone in only one micronutrient, and this effect needs to be controlled. Thus, with a cross-sectional design, the variables are placed interchangeably on different sides of the equation. Therefore, we clarify this information on Methods section (page 10, lines 249-250; lines 259-261).

Results

Please provide the operational definitions employed to classify the sexual maturation of adolescents.

Answer: We are thankful the editor for this remark. The Tanner scale is widely used in studies with adolescents to classify sexual maturation. This assessment was performed by pediatric endocrinologists with expertise in this field. This information is described in the “Methods section” (page 8, line 186-189).

Line 303: “high prevalence of pyridoxine intake” confusing, high or low intake?

Answer: We would like to thank the editor for this comment. The information has already been rewritten correctly (section Results, page 15, line 322-323).

Table 3: can you please explain why you limit yourself to the 15 micronutrients provided in the table?

Answer: We appreciate the editor’s comment. We decided to explore these 15 micronutrients, considering the nutritional importance of these nutrients for the development and health care of adolescents. In addition, we are more certain about the nutrition quality information in terms of the composition of the foods described in the tables.

97% more likely > almost two times more likely.

Answer: We are thankful to the editor for this comment. The information has already been rewritten as suggested (Results section, page 19, line 343).

Why it was not possible to present table 4 and 5 together in one table?

Answer: We would like to thank the editor for this question. The Tables 4 and 5 were presented together as suggested (Results section, page 20,21 and 22).

Table 4 and 5: Please clearly describe the dependent and independent variables, I also propose for removing the multivariable analysis.

Answer: We are thankful to the editor for this comment. To clarify the results, we included, in Table 4, which associations were related to the consumption of processed foods, and which were related to the consumption of ultra-processed foods.

Table 4 and 5: as long as p values are there I don’t see the purpose of having the Wald statistic there.

Answer: We are thankful to the editor for this comment. The Wald statistic is the most appropriate significance test in generalized (non-general) linear models for effect measures (PR or OR or RR). We decided to include the Wald statistic because the reader can confirm the statistical findings.

Others

Multivariate > Multivariable; bivariate > bivariable

Answer:

Reviewer 1

Processed and ultra-processed foods are associated with high prevalence of inadequate selenium intake and low prevalence of vitamin B1 and zinc inadequacy in adolescents from public schools.

The study addresses an interesting and relevant subject showing the association of consumption of processed and ultra-processed foods with micronutrient inadequacy in adolescents. However, the MS lacks logical presentation of information hence is difficult to understand.

The introduction doesn’t contain the relevant background information to guide readers to the rest of the manuscript. The justification or motivation to do this research is not well formulated.

Answer: We appreciate the reviewer’s comment. The introduction of the manuscript describes information about adolescents and the impact of food processing on health. The changes in adolescent eating patterns are presented in several countries, such as the United States, and others from Latin America, including Brazil. This information makes the reader interested in reading the manuscript. As suggested, we reinforce the motivation for conducting this research in the

Introduction section (Page 5, lines 123-131).

The language not clear.

Answer: We would like to thank the reviewer for this comment. The manuscript was submitted for review of English language by a native speaker at Editage, as supported by the editing certificate provided below.

The outcome of this study mainly depends on the quality of the food composition table used to convert into nutrient intake of the present study subjects. However, this may not work for some nutrients such as selenium. This is because; selenium concentration in foods is mainly determined by the amount of the element in the soil which is quite variable in short distances. Thus, data from food composition is not reliable.

Answer: We are thankful to the reviewer for this comment. We are aware of the complexity of assessing selenium dietary intake. To minimize this problem, we use national tables. We checked the information on nutrients composition in special the amount of selenium in foods.

Parts of the methodology lack clarity and are not detailed. The authors in general tried to show the effect of level of processing on micronutrient intake inadequacy that consumption of processed foods was associated with lower intake of selenium but not Zn and B1. Processed foods in general are poor in micronutrient content. However, it is failed to identify whether foods in the present study were enriched with Zn and vitamin B1 concentration.

For example,

Line 48-50, Energy consumption from processed foods was associated with higher prevalence of inadequate selenium intake (p < 0.01) but lower prevalence of inadequate vitamin B1 intake (p = 0.04).

Line 50-54, Energy consumption from ultra-processed foods was associated with lower prevalence of inadequate zinc and vitamin B1 intake (p < 0.01 and p =0.03, respectively). An increase in the proportion of energy obtained from processed and ultra-processed foods may reflect higher prevalence of inadequate selenium intake and lower prevalence of vitamin B1 and zinc inadequacy.

Line 74-75. It is not always true to claim that sedentary behaviors among adolescents lead to greater consumption of packed or ready to eat foods.

Answer: We are thankful to the reviewer for this comment. Other authors have demonstrated the associations between sedentary behavior and consumption of ultra-processed foods. Costa et al (2015)1 observed that 40% of the schoolchildren reported daily consumption of at least one group of ultra-processed foods (39.7%; 95%CI: 39.2-40.3), while 68.1% (95%CI: 67.7-68.7) reported > 2 hours/day of sedentary behavior. Among schoolchildren with sedentary behavior > 2 hours/day, prevalence of daily consumption of ultra-processed foods was 42.8% (95%CI: 42.1-43.6%), higher than among those without a sedentary behavior (29.8%; 95%CI: 29.0-30.5%). Longer time spent in sedentary behavior was associated with higher prevalence of consumption of ultra-processed foods among Brazilian adolescents.

The information in the introduction section was rewritten for the purpose of clarification (Introduction section, page 3, line 75).

________________________________________________________________

1COSTA, Caroline dos Santos et al. Sedentary behavior and consumption of ultra-processed foods by Brazilian adolescents: Brazilian National School Health Survey (PeNSE), 2015. Cad. Saúde Pública. 2018, vol.34, n.3. :e00021017 http://dx.doi.org/10.1590/0102-311x00021017.

Methodology

Line 136-140: Ethical consideration/approval has to be presented separately

Answer: We would like to thank the reviewer for this comment. We presented the information about ethical approval in a separated topic (Methods section, page 7, line 174-178).

Line 141-148: The inclusion and Inclusion criteria may work for the main study “Risk factors for cardiovascular disease among adolescent beneficiaries of the Brazilian government National School Lunch Program (PNAE)” but not for this specific MS.

The sample size was determined considering altered lipid profile as an outcome which is not consistent with the present study outcome.

Answer: We are thankful the editor for this comment. The sampling plan was defined for the study “Risk factors for cardiovascular disease among adolescents beneficiary of the Brazilian government National School Feeding Program (PNAE)”, which aimed to explore risk factors for cardiovascular diseases in adolescents. However, diet was included as one of the environmental variables of the study.

Weight and height measurement procedure is not detailed

Answer: We are thankful to the reviewer for this comment. Weight and height were performed at the school in duplicate by trained staff. The measurement procedures were carried out according to Habicht (1974)1. We chose not to describe the procedures in detail and inserted the reference used for standardization techniques1. The information was described in the Methods section, page 7, lines 182-184.

_______________________________________________

1Habicht JP. Standardization of quantitative epidemiological methods in the field. Bol Oficina Sanit Panam. 1974;76(5):375-384.

Line 179: ‘The morning was chosen to more easily enable methodological procedures’. This is not clear. Does this mean ‘anthropometric measurement was done during the morning? Why?

Answer: We would like to thank the reviewer for this comment. We recognize this mistake. The information was excluded from the text.

Line 182-185: It is not clear the importance of assessing pubertal stage given the age of study participants was known for this particular study.

Answer: We are grateful to the reviewer for this comment. Pubertal staging is important in studies with adolescents to correct variations in body composition, since puberty is a physiological process of hormonal maturation and somatic growth, which makes the organism able to reproduce. Changes in body composition, mediated mainly by the actions of sex hormones and growth, are characteristics of pubertal maturation and result in physical differences between sexes (Rogol, 2002)1. So as to make the research more complete, we decided to include this variable, describing the general characteristics of the population.

____________________________________________________

1Rogol AD, Roemmich JN, Clark PA. Growth at puberty. Journal of Adolescent Health. 2002;31:192-200.

Reviewer 2

Introduction - it would be good to also have some systematic review to be cited to reiterate how the unhealthy dietary intake from low middle income countries/ Asian as well to show this is a global burden.

Answer: We would like to thank the reviewer for this comment. We have cited a systematic review about unhealthy dietary intake in Malaysia, a middle-income country in Asia (Introduction section, page 4, lines 94-96).

Methods - it is noted in methods the age of interest is 10-19. However, based on the results it appears only mean age 11.8 years. Why the age group seems to be specific (10-13)?

Answer: We are thankful to the reviewer for this comment. Data collection was performed with the students in the morning At this time of day, the age group of 10 to 13 years is more prevalent in schools in northeastern Brazil, and adolescents over 14 years of age are more prevalent in the evening period. This fact may explain the higher number of adolescents aged 10-13 years.

Explain why just use 24 hours diet recall and repeated twice. Is it the most suitable for adolescents studies? Since your interest is micronutrient and ultra processed food, which method of dietary intake should be more robust to be used for this adolescents group? The inadequate number of days for this might influence why you discovered such results.

Answer: We are thankful to the reviewer for this comment. As previously described, we followed the protocol of Thompson and Beyes, which establishes that for the evaluation of habitual food consumption, at least 2 recalls are required. In addition, we used the method AMPM To facilitate the accurate recall of dietary intake, AMPM employs five sequential passes in the interview: (1) asking the participant to start by quickly listing the foods consumed in the previous day (without the need for a time sequence); (2) asking about any foods they had forgotten to report in nine commonly forgotten categories of foods; (3) asking about the time and occasion of consumption for each food; (4) probing for specific details on foods, amounts consumed, and foods consumed between identified eating events; and (5) probing for whether any food that they had forgotten to report1.

_____________________________________________________

1Moshfegh AJ1, Rhodes DG, Baer DJ, Murayi T, Clemens JC, Rumpler WV, Paul DR, Sebastian RS, Kuczynski KJ, Ingwersen LA, Staples RC, Cleveland LE. The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. Am J Clin Nutr. 2008 Aug; 88(2): 324–332.

What was the study response rate? How reliable is the reported Tanner staging? Any confirmation by the paediatrician?

Answer: We are thankful the reviewer for this comment. As described in the methodology, “A medical team of pediatric endocrinologists assessed pubertal staging. Sexual maturation was classified according to the Tanner scale into: pre-pubertal (stage 1), initial pubertal (stages 2 and 3), and final pubertal (stages 4 and 5) [24, 25].”. Thus, all adolescents evaluated by the doctor for sexual maturation were included in the study.

Results - The age group for 14 and above only 18? why is it so low? If this group is excluded, how different will the results be in Table 3?

Answer: We appreciate the reviewer’s comment. Data collection was performed with the students in the morning. In this period, the age group of 10 to 13 years is more prevalent in schools in northeastern Brazil, and adolescents over 14 years of age are more prevalent in the evening period. This fact may explain why most adolescents were aged 10-13 years. Even with a small number of adolescents over 14 years of age, we decided to maintain this age range, according to the inclusion criteria of the study. In Table 3, we decided to keep the group of adolescents aged 14-18 years, as the nutrient recommendations for this age range are different from those for the age range 10-13 years.

Figure - Out of 4457 subjects only 700+ enrolled. Why?

Answer: We would like to thank the reviewer for this comment. The sample size calculation indicated 483 students. However, 711 students were enrolled to compensate for the sample losses of 30%. The information has already been rewritten for the purpose of clarification (Methods section, page 6, line 152-156).

Since the two-stage random stratified sampling is used, why complex sample analysis not conducted?

Answer: We are grateful to the reviewer for this comment. We agree that a complex analysis with consideration of the sampling strategy could be an appropriate alternative to the data analysis we performed, considering that the sample was performed in two stages. However, in a thesis carried out during the study, the effect of the design was found to be less than 2.5% for all estimates, which did not significantly alter the estimates or the associated conclusions (Lyra, CO, Anthropometric Nutritional Status, Body Composition and blood pressure in adolescents [Thesis] Natal / RN: UFRN, 2012 https://repositorio.ufrn.br/jspui/handle/123456789/13265). As such, the authors chose not to perform the complex sampling analysis.

It appears highlighted micronutrient intake was low but do you have biomarkers to associate with?

Answer: We are thankful to the reviewer for this comment. The protocol defined for the study “Risk factors for cardiovascular disease among adolescents beneficiary of the Brazilian government National School Feeding Program (PNAE)” focuses on risk factors for cardiovascular diseases in adolescents. The diet was elected as the study variable for micronutrient intake assessment. In this study, biochemical biomarkers of micronutrients were not included.

Discussion - If ultra processed will lead to reduce certain micronutrient content, will enrichment with that vitamins in the food minimise the problem of suboptimal nutrient?

Answer: We are thankful to the reviewer for this comment. The enrichment with vitamins in the food can minimize the problem of suboptimal nutrient; however, excessive consumption of certain vitamins and minerals could have deleterious consequences on health and development of individuals and populations. Simultaneous micronutrient-delivery interventions could be challenging in terms of safety as the target populations may overlap, posing a risk of excessive intake of certain micronutrients1. Additionally, the bioavailability of micronutrients depends on food composition and possibly on its encapsulation method in some specific formulations2. This discussion is included in the manuscript (Discussion section, page 27, lines 473-480).

________________________________________________

1Garcia-Casal MN, Mowson R, Rogers L1, Grajeda R; consultation working groups. Risk of excessive intake of vitamins and minerals delivered through public health interventions: objectives, results, conclusions of the meeting, and the way forward. Ann N Y Acad Sci. 2019 Jun;1446(1):5-20. doi: 10.1111/nyas.13975. Epub 2018 Oct 5.

2Monteiro CA, Cannon G, Moubarac JC, Levy RB, Louzada MLC, Jaime PC. The UN decade of nutrition, the NOVA food classification and the trouble with ultra-processing. Public Health Nutr. 2018; 21(1):5-17. doi: 10.1017/S1368980017000234.

Research has shown the association ultra-processed foods that contain added sugars, excess sodium, and unhealthful fats possibly lead to poor health outcome and is this pattern the same with yours?

Answer: We are thankful to the reviewer for this comment. Other authors concluded that there is higher caloric contribution of carbohydrates, lipids and saturated fatty acids of ultraprocessed foods, as well as a lower concentration of protein and fibers1. This pattern is also the same in Brazil, as demonstrated by Cunha et al (2018)2. This study confirmed that greater intake of ultra-processed foods is a marker of an unhealthy diet in Brazilian adolescents. Other studies confirmed the role of ultra-processed foods in the obesity epidemic among Brazilian adults and adolescents3.

____________________________________________________

1Monteles N, Larisse et al. The impact of consumption of ultra-processed foods on the nutritional status of adolescents. Rev. Chil. Nutr. 2019, vol.46, n.4, pp.429-435. http://dx.doi.org/10.4067/S0717-75182019000400429.

2Cunha DB, da Costa THM, da Veiga GV, Pereira RA, Sichieri R. Ultra-processed food consumption and adiposity trajectories in a Brazilian cohort of adolescents: ELANA study. Nutrition & Diabetes, 2018, v.8, n.28, p.1-9. DOI 10.1038/s41387-018-0043-z

3Louzada ML, Baraldi LG, Steele EM, Martins AP, Canella DS, Moubarac JC, Levy RB, Cannon G, Afshin A, Imamura F, Mozaffarian D, Monteiro CA. Consumption of ultra-processed foods and obesity in Brazilian adolescents and adults. Prev Med. 2015 Dec;81:9-15. doi: 10.1016/j.ypmed.2015.07.018. Epub 2015 Jul 29.

In the conclusion - "stress the importance of limiting personal consumption of processed and ultra-processed", how and where at school, home or institution and what is the allowed value? If there is any? What would be the most suitable take home message?

Answer: We are thankful to the reviewer for this comment. Healthy eating habits should be stimulated at home, at school or in institutions. The food guide for the Brazilian population1 recommends the consumption of fresh or minimally processed foods, in a large variety and predominantly of plant origin, because they are the basis for a nutritionally balanced, tasty, culturally appropriate diet for promotion of a socially and environmentally sustainable food system. This document recommends that the consumption of processed foods should be limited to small quantities, either as ingredients of culinary preparations, or as a meal complement. There is no recommendation for consumption of ultra-processed foods; the food guide for the Brazilian population recommends that people should avoid ultra-processed foods because of their ingredients – e.g., cookie sandwiches, packaged snacks, soft drinks and instant noodles. These products are nutritionally unbalanced. Because of their formulation and presentation, they tend to be consumed in excess and replace fresh or minimally processed foods. Their forms of production, distribution, commercialization and consumption adversely affect culture, social life and the environment.

__________________________________________________

1Brasil.Guia alimentar para a população brasileira.2nd ed. Brasília, Distrito Federal: Ministério da Saúde; 2014.

Reviewer 3

Abstract - Numerous sentences, such as the one highlighted below, starting with 'this':

"This was a cross-sectional study.”

Answer: We are thankful the reviewer for this comment. The sentences were rewritten according to your suggestion. (Abstract, page 2, lines 35-36).

-The term 'processed and ultra-processed foods’ is being used several times. It is very repetitive.

Answer: We are thankful the reviewer for this comment. We have edited the terms when appropriate. (Abstract, page 2, lines 34).

-Insert the acronym for 'estimated average requirement

Answer: We are thankful to the reviewer for this comment. The acronyms for 'estimated average requirement’ was added (Abstract, page 2, line 41).

Introduction

-The information stated in lines 96 and 97 has already been said in the introduction’s section elsewhere.

Answer: We are thankful to the reviewer for this comment. We reformulated the information in lines 96 and 97, and in the last paragraph of the introduction. (Introduction section, page 4, lines 96-99; page 5, lines 123-131).

-The paper discusses that some results found might be due to differences of eating habits or economic status of northeastern Brazil, in comparison to the rest of the country, but it does not mention any of this in the introduction. The authors need to expand on this.

Answer: We would like to thank the reviewer for this comment. We included information about eating habits or economic status of northeastern Brazil (Introduction section, page 5, lines 119-122).

Methods

-Were rural schools included? The access to milk and dairy products could be different compared to the population from the urban ones.

Answer: We are thankful to the reviewer for this comment. The schools included in this study are in the urban area of Natal city, northeastern Brazil.

-In Fig 1, why are there 700 adolescents enrolled when it is mentioned above only 11 were transferred from school?

Answer: We are thankful to the reviewer for this comment. We recognize this mistake. At first, 711 were enrolled instead of 771. Then, 11 were transferred from school, and 700 remained effectively enrolled. The information has already been rewritten correctly in Figure 1.

-What about those participants who were taking dietary supplements? Was this type of information collected or evaluated?

Answer: We are thankful to the reviewer for this comment. The information about the use of supplement was collected, but we did not find any adolescent using these products.

-Was the apportioning and quantification of food reviewed by the 24-hour dietary recalls' interviewers?

Answer: We are thankful to the reviewer for this comment. All the 24-hour dietary recalls were reviewed and the database was checked by more than one researcher.

Discussion

-"We observed low percentages of energy intake from ready-to-eat and semi-ready-to- eat meals and sweetened milk drinks from the ultra-processed foods group.” I would prefer the passive voice.

Answer: We thank the reviewer for this comment. The sentence was rewritten according to your suggestion. (Discussion section, page 23, lines 393).

-"We observed low percentages of energy intake from ready-to-eat and semi-ready-to- eat meals and sweetened milk drinks from the ultra-processed foods group. This finding could be attributed to the fact that most families living in northeastern Brazil are characterized as low-income families [39, 40], as an increase in the intake of ready-to-eat and semi-ready-to- eat meals and sweetened milk drinks has been associated with an increase in family income per capita [40].” Is it really different from "cakes, pies, and cookies"?

Answer: We are thankful to the reviewer for this question. In northeastern Brazil most of the cakes, pies, and cookies consumed are non-industrialized preparations, and therefore, they differ from the ready-to-eat and semi-ready-to-eat meals and sweetened milk drinks. The fact that they are homemade preparations, and therefore cheaper, does not necessarily make the food healthier, because ultra-processed culinary ingredients are included in their preparation.

-“In our study, we observed low intake of foods that are significant sources of selenium, such as fish, whole grains, and Brazil nuts.” Can this information be drawn from table 2?

Answer: We are thankful to the reviewer for this comment. This information cannot be drawn from Table 2, because the foods cited as selenium sources are not part of the processed and ultra-processed food group. This information was collected from the 24-hour dietary recall.

-“Brazilian children aged 2–3 years typically have a low prevalence of iron, vitamin C, vitamin A, calcium, and folate inadequacy. None of these children had intakes less than the EAR value for zinc. Fortified foods contributed to micronutrient supply. However, 4·0% of children exceeded the UL for vitamin A, 3.1% for zinc, 1.1% for folic acid, and 0.2% for iron. These results suggest a low prevalence of inadequate micronutrient intake among children, with the implication that this group could be at risk of excessive micronutrient intake provided by ultra-processed foods [59].” Isn't this paragraph deviating from the population targeted at this paper?

Answer: We are thankful to the reviewer for this comment. Although the population is different from the one addressed in the study, we decided to highlight this information because it is one of the few Brazilian studies that has addressed this theme. We rewrote the paragraph to improve the interpretation (Discussion section, pages 26-27, lines 473-480).

-According to results presented, processed and ultra-processed foods are associated with a low prevalence of vitamin B1 and zinc inadequacy. The authors need to expand on this.

Answer: We are thankful to the reviewer for this comment. The explanation about this topic is in the Discussion section, page 27, lines 461-472.

-What are the limitations of the 24-hour dietary recall?

Answer: We are thankful to the reviewer for this comment. The main challenges to accuracy with the 24HR are attention and memory. Many respondents are challenged with distinguishing between what they usually eat and what they ate yesterday, opening the possibility for omissions and intrusions (foods reported, but not actually eaten). Actual memory of distinct events decays with time, which has been documented to start within an hour of the meal. Portion size estimation is also challenging as the amount consumed has to be both remembered and accurately estimated. Since day-to-day variability in diet is high, the information from a single day cannot accurately reflect the usual diet of an individual. Multiple days of information are collected and statistical techniques employed to address this problem1.

________________________________________________

1Frances E. Thompson, Amy F. Subar, Catherine M. Loria, Jill L. Reedy, Tom Baranowski Need for Technological Innovation in Dietary Assessment. J Am Diet Assoc. 2010 Jan; 110(1): 48–51. doi: 10.1016/j.jada.2009.10.008

-The results need to be further compared with the literature.

Answer: We are thankful to the reviewer for this comment. The results are discussed on pages 23-25, lines 378-597.

Submitted filename: response_to_reviewers_plos_one.docx

Click here for additional data file.

28 Oct 2019

Processed and ultra-processed foods are associated with high prevalence of inadequate selenium intake and low prevalence of vitamin B1 and zinc inadequacy in adolescents from public schools in an urban area of northeastern Brazil

PONE-D-19-18574R1

Dear Dr. Sena-Evangelista,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Samson Gebremedhin, PhD

Academic Editor

PLOS ONE

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

11 Nov 2019

PONE-D-19-18574R1

Processed and ultra-processed foods are associated with high prevalence of inadequate selenium intake and low prevalence of vitamin B1 and zinc inadequacy in adolescents from public schools in an urban area of northeastern Brazil

Dear Dr. Sena-Evangelista:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Samson Gebremedhin

Academic Editor

PLOS ONE