Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Changes in breakfast frequency and composition during adolescence: The Adolescent Nutritional Assessment Longitudinal Study, a cohort from Brazil

  • Bruna Kulik Hassan ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Department of Epidemiology, Institute of Social Medicine, State University of Rio de Janeiro (UERJ), Rio de Janeiro, Rio de Janeiro, Brazil

  • Diana Barbosa Cunha,

    Roles Data curation, Formal analysis, Investigation, Supervision, Writing – review & editing

    Affiliation Department of Epidemiology, Institute of Social Medicine, State University of Rio de Janeiro (UERJ), Rio de Janeiro, Rio de Janeiro, Brazil

  • Gloria Valeria da Veiga,

    Roles Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Writing – review & editing

    Affiliation Department of Social and Applied Nutrition, Institute of Nutrition Josué de Castro, Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Rio de Janeiro, Brazil

  • Rosangela Alves Pereira,

    Roles Conceptualization, Formal analysis, Methodology, Project administration, Visualization, Writing – review & editing

    Affiliation Department of Social and Applied Nutrition, Institute of Nutrition Josué de Castro, Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Rio de Janeiro, Brazil

  • Rosely Sichieri

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation Department of Epidemiology, Institute of Social Medicine, State University of Rio de Janeiro (UERJ), Rio de Janeiro, Rio de Janeiro, Brazil

Changes in breakfast frequency and composition during adolescence: The Adolescent Nutritional Assessment Longitudinal Study, a cohort from Brazil

  • Bruna Kulik Hassan, 
  • Diana Barbosa Cunha, 
  • Gloria Valeria da Veiga, 
  • Rosangela Alves Pereira, 
  • Rosely Sichieri



To estimate changes over time of breakfast frequency and foods/food groups consumed at breakfast.


Cohort of 809 students aged 10–16 years old from Rio de Janeiro, Brazil, enrolled in 2010, and followed for three years. Breakfast skippers were those not eating breakfast. Those with breakfast frequency of 4 or less times per week were considered irregular breakfast eaters. Changes over time of breakfast frequency and breakfast foods/food groups were analyzed by generalized estimating equations.


At baseline, overweight/obese girls had higher prevalence of irregular breakfast than those non-overweight/obese (40% vs. 26%; p = 0.005); among boys, there was no significant difference in the prevalence of irregular breakfast according to weight status. After three years, among girls there was an increase in the risk of irregular breakfast consumption (RR = 1.29; 95% CI: 1.08; 1.54) and breakfast skipping (RR = 1.63; 95% CI: 1.12; 2.38). Also, overweight/obese boys (RR = 1.40, 95% CI: 1.03, 1.89) and non-overweight/obese girls (RR = 1.54, 95% CI: 1.17, 2.04) had higher risk of irregular breakfast in three years, compared to baseline. After three years, boys changed the consumption of foods/food groups at breakfast and presented higher risk of decreased intake of fruits (RR = 1.60, 95% CI: 1.20, 2.13), sandwiches and snacks (RR = 1.58, 95% CI: 1.12, 2.22), chips (RR = 1.43, 95% CI: 1.01, 2.13), and ham (RR = 1.52, 95% CI: 1.12, 2.07), and lower risk of cheese intake (RR = 0.73, 95% CI: 0.53, 1.00). Girls had higher risk of decreased intake of fruits (RR = 2.08, 95% CI: 1.47, 2.95), milk (RR = 1.49, 95% CI: 1.07, 2.08), chocolate powder (RR = 1.54, 95% CI: 1.11, 2.14) and ham (RR = 1.65, 95% CI: 1.16, 2.36).


Changes in breakfast patterns are different according to sex. Also, different changes in breakfast frequency according to BMI category were found. Consumption patterns of some foods/food groups have a tendency to become changed from initial to middle adolescence.


The practice of eating breakfast regularly among adolescents range from 51% to 95% worldwide [15], whilst 7% to 32% of them skip this meal or consume it irregularly [2,4,610]. This wide range in prevalence may be partly explained by how this meal is defined and analyzed. While some studies define breakfast consumption by time [8,11,12] and/or prior to performing other daily activities [2,12], others substitute or add to their definition some specific foods [2,5,9,13], or even a caloric percentage of the total daily energy requirements [12,14].

A frequency of at least five times a week is used to mark a regular frequency of breakfast eating [2,4], though lower [12,15] or higher cutoffs [3] have also been used. Also, cutoff points for skipping breakfast and irregular breakfast vary between studies. For example, Sjöberg et al. (2003) defined irregular breakfast intake as omission of this meal at least once a week [2]. Another study categorized as skippers those eating breakfast less than 3 days a week and “semi-skippers” those with breakfast frequency of 3 to 4 days a week [4]. Timlin et al. (2008) divided adolescent’s breakfast eating in three groups: (1) daily eaters, for those who responded “every day”; (2) irregular eaters, with breakfast frequency between 1 to 6 days per week and; (3) daily skippers, for those who responded “never” [16]. Therefore, despite being a consensus that breakfast is the first meal of the day [12,17], there is no clear and standardized definition or categorization when concerns breakfast frequency.

Breakfast frequency has been related to several health benefits and healthier lifestyle factors. Observational studies suggest frequent breakfast in adolescence to be associated with better academic and cognitive performance [6,18,19], regular family meal frequency [20], lower body fatness, lower fasting glucose levels, higher cardiorespiratory fitness and healthier cardiovascular profile [4,8,10,13]. On the other hand, obesity, the practice of dieting for weight control, inadequate physical activity, and sedentary lifestyle were shown to be related to irregular breakfast eating at this stage of life [2,3,9,10,1923]. Though, it is important to highlight that causal relationship is not well stablished and most studies have cross-sectional design, portraying these findings still as an open question. Also, few studies evaluated changes in breakfast habits during adolescence and there is no such data in Brazil.

Frequent breakfast eating may also improve diet quality, and maintenance of this behavior in adolescence is consistently associated with higher intake of fiber, calcium, potassium, zinc, iron and several vitamins [2,9,11,13,19], higher consumption of whole grains, fruits, vegetables and milk [4,24], higher adherence to Mediterranean diet [13], and lower consumption of sugar-sweetened beverages, sweets and chips [2,4].

Several factors have been related to breakfast eating. Girls [2,57,25] and adolescents from other non-white skin color [11] skip breakfast more often. Socioeconomic factors have been related to breakfast eating, such as type of school [25], parental education [16] material affluence [26,27] and men’s household income [28]. Also, evidence suggest breakfast frequency related to adiposity. Most longitudinal studies have focused on its influence on weight change [8,10,12,16,22], whereas the opposite direction is poorly studied. Despite cross-sectional findings showing obese students more likely to skip breakfast than overweight and non-overweight youth [22,29], as far as we know, our study is the first to evaluate weight status relationship with breakfast eating changes during adolescence.

Some studies showed a reduced frequency of breakfast eating as adolescents get older [5,8,11]. Considering all possible benefits related to breakfast and the lack of information on breakfast changes over time, it is fundamental to observe how this behavior develops during adolescence and factors associated to inform future public health programs and nutrition interventions for this age. Thus, the current study aimed to evaluate changes of breakfast frequency and foods/food groups consumed at breakfast during adolescence, according to factors usually associated with breakfast eating, such as weight status family breakfast and dieting.

Materials and methods

Study design and study population

Data of this prospective study were derived from the Adolescent Nutritional Assessment Longitudinal Study (ELANA), a cohort with adolescents from four private and two public schools located in Rio de Janeiro metropolitan area, Brazil. In Brazil, type of school attending students is considered a proxy of socioeconomic status [3032].

Sixth-graders from 26 classes attending morning or afternoon schools were invited to participate in the study. ELANA was designed to assess factors related to anthropometric trajectory and was approved by the Research Ethics Committee of Institute of Social Medicine of The State University of Rio de Janeiro (certificate number 0020.0.259.000–09). Participation of the students occurred after their legal guardian provided written informed consent for data publication of study participants.

Adolescents with physical or mental disabilities and who were pregnant or lactating were considered ineligible and not included in the study. Of 945 eligible middle school students, 36 refused to participate, 41 were not authorized by their parents and 59 were not found at school on the days of data collection, leading to a study population of 809 adolescents in 2010 (baseline). After three years, 39.6% of the students were lost to follow-up. Reasons for losses to follow-up included missing school on the days of data collection (n = 370), students refusals to participate (n = 99), non-authorization by parents of continuity in the research (n = 55) and pregnancies (n = 3). Thus, the final sample consisted of 488 adolescents in 2013.

Data collection and measurements

At all stages of the study, data collection of anthropometric data and a self-administered questionnaire were conducted by a previously trained team of research assistants and supervised by post-graduate students, who were responsible for clarifying questions with participants and reviewed all the questionnaires. Data collected were typed by two students and compared to rectify potential typing errors. Respondent’s errors or lack of information were returned to the field for data correction with each participant.

Sociodemographic, weight-related and behavioral data were obtained by application of a self-report questionnaire in each year of data collection. The self-reported questionnaire can be found available at Breakfast frequency was estimated at baseline and after three years. The respondents were asked how often they ate breakfast and five response options were given: never or almost never, 1 to 2 times a week, 3 to 4 times a week, 5 to 6 times a week, and daily. Response options were collapsed for analyses as less than 5 times a week, for irregular breakfast frequency, and 5 times a week to daily, for regular breakfast frequency. To analyze breakfast skipping, skippers were categorized as those who responded never or almost never, and non-skippers, those with breakfast frequency of 1 or more times a week. Similar categorizations for skipping breakfast were conducted in previous studies [12,33]. The same question and response options were provided for breakfast frequency with parents, categorized in regular and irregular family breakfast frequency.

At baseline and after three years of follow-up a qualitative food frequency questionnaire (FFQ) was applied with the students. The ELANA FFQ was adapted from a FFQ validated for adolescents from Rio de Janeiro metropolitan area, Brazil [34], though no differences were found in correlation coefficients of nutrient consumption between three-day food records and ELANA FFQ and the original FFQ. Pearson correlation coefficients for ELANA FFQ ranged from 0.41 to 0.23 for energy and nutrients adjusted for energy intake (carbohydrates, lipids, protein, calcium and iron), whilst the original FFQ ranged from 0.41 to 0.34.

The ELANA FFQ asks the frequency of consumption in the last three months of 72 foods and beverages. Eight response options for each item are available: less than once a month or never, 1 to 3 times a month, once a week, 2 to 4 times a week, 5 to 6 times a week, once a day, 2 to 3 times a day, 4 or more times a day. For our analyses, the response options were transformed into daily frequency and weights were given for each option, taking into account the average frequency of the response option and time, in days. According to the pattern of distribution of the food groups usually consumed at breakfast, response options were divided in three categories: monthly to no consumption, weekly and daily.

We assessed foods/food groups usually consumed at breakfast according to findings from Correa et al. (2016), who identified dietary patterns regarding breakfast from one 24-hour dietary recall (24hR) with adolescents aged 10 to 14 years old, from public schools of Rio de Janeiro metropolitan area, Brazil [35]. Thus, fourteen foods/food groups were designated from our FFQ: cheese (white cheese, yellow cheese, mozzarella cheese and others), milk, chocolate powders, fruits (banana, pineapple, guava, strawberry, orange, apple and papaya), bread, coffee, ham, chips, cookies, eggs, sweets (candies, chocolate, condensed milk, jelly, jam, peanut candy, ice cream), Sugar sweetened beverages (fruit drinks, guarana refreshment, sodas and tea), and snack and sandwiches (hamburger, hot dog, pizza, deep-fried and baked snacks).

Physical activity energy expenditure (PAEE) was estimated by the short form of the International Physical Activity Questionnaire (IPAQ), validated with Brazilian adolescents of at least 12 years old [36]. Since part of our study sample was younger at baseline, IPAQ was meticulously explained to students and response options were revised. Duration (time per day) and frequency (days per week) of walking, moderate-intensity activities and vigorous-intensity activities were obtained by the eight open-ended questions of IPAQ. Respective time spent per week with walking, moderate and vigorous activities were multiplied by metabolic equivalent of task (MET) intensities, and intensity-time products were summed to generate PAEE [37,38].

We asked the frequency of dieting for weight control in the last three months and response options were never, rarely, often and always. For analyses, we collapsed the options rarely, often and always as practice of dieting and never to represent no dieting. Skin color was obtained according to Brazilian Institute of Geography and Statistics classification [39], and students were divided in white and other skin color for analyses.

Along with the self-report questionnaire, anthropometrics were measured by a previously trained research team, following Lohman et al. (1988) protocols [40], and were standardized according to Habicht (1974) at baseline [41], and Norton and Olds for the follow-up occasions [42]. Adolescents’ weight and height allowed for the calculation of body mass index (BMI). Weight was measured with a portable electronic platform scale (Kratos-cas, Brazil) to the nearest 0.05 kg, and height was measured twice with a stadiometer to the nearest 0.1 cm (Alturexata, Brazil). Mean height was estimated for analyses, and when height variation exceeded 0.5 cm, the measurement procedure was repeated.

Adolescents’ weight status was assessed according to the age and sex-specific cut-off points for BMI, as published by the International Obesity Task Force [43]. We collapsed adolescents classified as adequate and thinness with a BMI-for-age z-score ≤ 1, who became the non-overweight/obese category, and those who had BMI-for-age z-score > 1 corresponded to the overweight/obese category. Thin adolescents joined those classified with adequate BMI-for-age z-score, considering that they were few (n = 22) and primary analyses showed the same direction of response between groups classified as thinness and those adequate. To calculate BMI-for-age z-scores we used WHO-Anthro Plus, version 3.2.2.

Statistical analyses

All statistical analyses were stratified by sex and were performed using Statistical Analysis System, version 9.4 (SAS Institute Inc., Cary, NC, USA). Cross-sectional analyses between breakfast frequency and sociodemographic, weight-related, and behavioral factors at baseline were estimated using Student t tests to compare means, and chi-square tests to compare proportions. To observe differences between adolescents that participated at baseline and those participating in both baseline and after three years we conducted chi-squared tests for baseline variables. Descriptive data were expressed in percentage, means and standard deviation values.

To address whether breakfast frequency changed after three years and the relationship with baseline characteristics, longitudinal analyses were applied using generalized estimating equations (GEE) with log-binomial models for repeated measures, with estimation of relative risks and 95% confidence intervals (PROC GENMOD from SAS). All analyses included the school cluster effect. The characteristics selected for longitudinal analyses were those statistically significant in cross-sectional analyses. Initially, each model included time, the baseline exposure variable and an interaction term time-exposure, which allowed the evaluation of potential differences between exposure groups in response change over time.

Models were also tested for effect modification by weight status, family breakfast and dieting. Since the interaction term was statistically significant only for weight status, an analysis stratified for weight status is shown. As models with the other exposure variables were not statistically significant for the interaction term, we withdraw the interaction term and conducted a multivariate model including the variable time and adjusted for the exposure variables that were significantly related to the outcomes at baseline. These analyses were conducted for regular/irregular breakfast and repeated for skipping breakfast (skippers vs. non-skippers).

To address changes in foods/food groups usually consumed at breakfast, we selected only the participants who reported breakfast for both time occasions, leading to a study sample of 380 adolescents. We also excluded 11 individuals at baseline with FFQ response inconsistencies. To verify how these foods/food groups changed in three years, time was modeled using generalized estimating equations with multinomial models, with estimation of relative risks and 95% confidence intervals (PROC GENMOD from SAS). All analyses included the school cluster effect.

The advantage of using generalized estimating equations is that analyses account for losses of follow-up for the outcome. This approach can handle time-unbalanced data as a function of missing values for the outcome and no data on individuals at any occasion is discarded [44,45].


At baseline, approximately a quarter of adolescents consumed breakfast irregularly. Breakfast was skipped by 9.6% of all adolescents. Half of study participants were having irregular breakfast with their parents and a high proportion of them were overweight or obese (42.5%). Almost half of students have been on a diet in a reference period of three months and 37.1% were from public schools. Adolescents were 11.8 years old, on average, aged 10–16 years old at baseline, and age ranged from 12.9 to 18.6 after three years There were no statistically significant differences between adolescents participating at baseline and those who continued participating of the study after three years for breakfast frequency, breakfast skipping, sex, skin color, weight status, family breakfast frequency and dieting. However, mean age and proportion of adolescents from public schools who participated of the study after three years of follow-up were lower than of those who participated at baseline (Table 1).

Table 1. Baseline characteristics of adolescents that participated at baseline and both at baseline and after three years of follow-up (3y fu).

Breakfast was consumed irregularly among 21.2% (n = 92) of boys and 31.8% (n = 118) of girls. As for skipping breakfast, 7.8% (n = 34) of boys and 11.6% (n = 43) of girls were skippers. Differences at baseline in irregular breakfast frequency and breakfast skipping according to socioeconomic, weight-related and behavioral characteristics, stratified by sex, are shown in Table 2. Both boys and girls eating breakfast irregularly and skipping breakfast were more likely to have irregular family breakfast. At baseline, no statistically significant differences of irregular breakfast eating and of breakfast skipping were found for type of school, skin color, physical activity energy expenditure and mean age for both boys and girls. Girls having irregular breakfast were more likely of being overweight/obese and practicing dieting than girls consuming breakfast regularly.

Table 2. Total sample size, frequency (%; n) and mean (sd) of irregular breakfast eating and skipping breakfast at baseline according to socioeconomic, weight-related and behavioral aspects of adolescents, by sex.

Table 3 shows change in irregular breakfast eating over time. Girls had about 30% higher risk of having irregular breakfast after three years (RR = 1.29, 95% CI: 1.08, 1.54). There was no significant effect of interaction between time vs. family breakfast frequency and time vs. dieting, meaning that changes over time were not influenced by family behavior or dieting at baseline. Thus, interaction terms were withdrawn from dieting and family breakfast since the smallest p-values for both irregular breakfast and skipping were 0.11. There was, however, an interaction effect between time and weight status for irregular breakfast. As there was a significant difference for girls that did not appear for boys in the interaction term for irregular breakfast (and not for skipping), we stratified the irregular breakfast for weight status.

Table 3. Relative risks (RR) of irregular breakfast after three years of follow-up.

Among those who were overweight/obese at baseline, breakfast change was influenced differentially according to sex. Thus, boys overweight/obese at baseline had a 40% increase of irregular breakfast, whereas among girls this increase occurred only among those not overweight/obese at baseline (RR = 1.54, 95% CI: 1.17, 2.04). When skipping breakfast was analyzed (Table 4), there was no significant effect of interaction between time and any of exposure variables. Girls had 63% higher risk of skipping breakfast after three years (RR = 1.63, 95% CI: 1.12, 2.38), without increase among boys.

Table 4. Relative risks (RR) of skipping breakfast change after three years of follow-up.

Cross-sectional analyses of frequencies of foods/food groups usually consumed at breakfast indicated that the consumption of most items were reduced after three years of follow-up, except for sugar-sweetened beverages, eggs and sweets, that had similar consumption after three years, and eggs, that increased (Table 5). Longitudinal analyses of foods/food groups change over time revealed that boys had higher risk of reducing consumption of fruits, chips, sandwiches and snacks, and ham after three years, lower of consumption of cheese, and change was not statistically different for milk, bread, eggs, cereals, cookies, sweets, sugar-sweetened beverages, coffee and chocolate powder (Table 6). Girls had higher risk of reducing consumption of fruits, milk, chocolate powder and ham after three years, while there were reductions of consumption for bread, chips, cereals, cookies, coffee and cheese, but estimates were not statistically significant. Risk of increased consumption after three years was observed, though not significant, of sandwiches and snacks, sugar-sweetened beverages and sweets (Table 6).

Table 5. Frequencies of foods/food groups usually consumed at breakfast at baseline and after three years of follow-up (3y fu) (n = 380).

Table 6. Relative risks (RR) of reducing intake of foods usually consumed at breakfast from baseline to three years of follow-up.


Our findings revealed that girls who skipped breakfast or had breakfast irregularly at the beginning of the study had higher risk of continue with these patterns after three years. Among boys, having breakfast irregularly or skipping this meal at baseline did not seem to be related to breakfast eating by the end of the study. Studies using data from the National Longitudinal Study of Adolescent Health found that Americans eating breakfast regularly at adolescence were more likely to maintain this pattern during young adulthood, in Merten et al. (2009) study [15], and breakfast eating decreased from adolescence to early adulthood, in Niemeier et al. (2006) study [46]. Although the analyses from Merten et al. (2009) were controlled for sex, no stratified analyses or modification effect were investigated [15]. Pedersen et al. (2013) reported that having low breakfast frequency (categorized as more seldom than daily) at 15 years predicted low breakfast frequency at 19 years that, in turn, predicted low breakfast frequency at 27 years [47]. Further, low breakfast frequency at 15 years predicted low breakfast frequency at 27 years. Same patterns were found in analyses stratified by sex, but the observed predictions between 15 and 27 years for men were not significant. To some extent, their findings are consistent with our observations.

Weight status was related to change in breakfast consumption over time. Overweight/obese boys and non-overweight/obese girls having breakfast irregularly at baseline had greater risk of having irregular breakfast after three years. As far as we know, no research of weight status modification effect on breakfast eating change over time was conducted. Though, longitudinal studies of breakfast eating influence on adiposity and weight status are better documented, despite no consistent evidence [22,33,48]. Berkey et al. (2003) investigated the modification effect of weight status and sex on the relation of breakfast history and BMI change among adolescents aged 9–14 years at baseline [33]. They found overweight children and breakfast skippers over the previous year with smaller BMI increase, relative to peers who were daily breakfast eaters. Though it was not significant, normal weight children gained weight relative to eaters. The National Heart, Lung, and Blood Institute Growth and Health Study (NGHS) observed that only among American girls with a high BMI-for-age at baseline, frequent breakfast eating was associated with decreased BMI after 10 years [49]. Other studies should further explore the combination of previous weight status, weight change and change in breakfast eating to clarify these associations.

When the outcome of interest was skipping breakfast, weight status did not appear as a modifying effect. Among girls, the risk of skipping breakfast in three years was of higher magnitude than the risk of having irregular breakfast in the same period. Dialektakou and Vranas (2008) compared 24 different definitions and categorizations of breakfast skipping among Greek high school students [50]. The authors observed that results changed depending on how breakfast skipping was defined and categorized. For example, when adolescents were asked if they had eaten any solid food in the morning of the interview, 39% responded negatively. By adding the consumption of any beverage to the question the prevalence of skipping breakfast fell by 12.5%. When argued if they did not eat breakfast weekly throughout the year, 8.4% were considered skippers, whilst 3.6% skipped the meal when the reference period was the last week. Additionally, the authors found some breakfast skipping definitions associated with BMI and weight status, but others were not. Given this, the findings can vary widely, which may compromise the comparison between studies.

It is evident that how one defines and categorizes breakfast can significantly influence findings and their interpretation. For this reason, further discussion on this matter and future research attempting to establish a standardized definition of what constitutes a breakfast in scientific literature is essential. This may allow better comparisons across studies and interpretation of the existing literature, and provide appropriate methods to advance the field that could lead to better grounded recommendations for population.

Among breakfast eaters we found higher risk of decreased consumption of fruits and ham for both sexes; chocolate powder and milk for girls; chips, sandwiches and snacks for boys. Cheese was the only food group with higher probability of lower consumption after three years, but only among boys. Also, we found no statistically significant changes in three years for consumption of bread, coffee, cereals, cookies, sugar-sweetened beverages and sweets for boys and girls; cheese, chips, and sandwiches and snacks for girls; milk, eggs and chocolate powder for boys. Few studies have tried to characterize breakfast over time. Lytle et al. (2000) followed eating patterns in students from Minnesota at the third, fifth, and eighth grades [1]. Their findings showed that the pattern of breakfast across three-time periods shifted to a less healthful dietary behavior, with declined variety of foods in this meal. Daily fruit, vegetable and salty snacks consumption fell by the end of the study. The proportion of beverage from milk and fruit juice consumption also had fallen, whilst sodas contribution nearly tripled.

It is important to highlight that we captured foods usually consumed at breakfast from Correa et al. (2016), who captured breakfast patterns from 24hR. As our data was collected from a FFQ, we had limited food options compared to those given by a 24hR and, thus, adaptions were necessary [35]. Besides, the FFQ captures daily food consumption, and foods/food groups chosen in the current study may be consumed in eating occasions other than breakfast. Therefore, these foods/food groups may not reflect foods actually consumed at breakfast, which reflects one limitation of the current study.

Despite this limitation, there is certain homogeneity between urban regions of western countries with regards to breakfast patterns, usually composed by some type of milk, bread and cereals [17]. Similar breakfast patterns were observed in studies with adults from São Paulo, Brazil [51,52]. Santos et al. (2015) revealed three eating patterns at breakfast: traditional, linked to coffee, sugar, butter/margarine and bread/toast; healthy, characterized by fruits, whole-grain bread, skimmed milk/semi skimmed milk and white cheese; and a snack pattern, correlated with consumption of cold cuts, bread/toast and yellow cheese in this meal [51]. Similar to these findings, Mattos and Martins (2000) showed breakfast composed mainly of coffee, bread, sugar, margarine/butter, whole milk, and also, fruits and cheese [52]. Despite these similarities, breakfast patterns of Brazilian adolescents are not well documented and should be further explored.

Studying adolescents from Rio de Janeiro metropolitan region, Brazil, Correa et al. (2016) reported that most consumed foods/food groups at breakfast were: breads (64%), sugar and sweets (58%), butter or margarine (50%), milk and milk-based beverages (49%), and coffee and tea (34%) [30], similar to breakfast patterns found in Brazilian adult’s studies [51,52]. In contrast to those studies, chocolate milk appeared as most consumed by adolescents (37%), while cheese (12%) and fruit (4%) were less consumed. Also, beverages marked differences in breakfast patterns, mainly by three patterns: sugar-sweetened beverages, milk and milk-based beverages consumption, and coffee and tea.

It is necessary to further understand the composition of breakfast to improve diet quality among adolescents, since they present unhealthy eating habits and higher nutritional requirements, as a result of intense physical growth at this stage of life [25,53,54]. Data from Brazilian studies revealed that approximately 20% of ninth grade middle school students did not consume milk, fruits and vegetables a week prior to data collection [25] and, compared to adults, Brazilian adolescents presented higher consumption of sodas, biscuits and sausages, and lower of vegetables and beans [55,56]. Another nationally representative study from Brazil showed that almost all adolescents had intake of calcium below the recommendation for the age group [54]. As breakfast eating is related to higher intake of several nutrients, such as calcium and vitamin D [9,11,19] and, thus, higher growth potential can be achieved, investments in policies that improve breakfast eating are needed.

One possible explanation for the changes we observed for breakfast frequency and composition may be due to the fact that adolescents get more autonomy and responsibilities as they grow, which includes changing lifestyle and dietary habits [57,58]. Along with that, parent’s participation in their breakfast also decreases as they grow older [59]. There is a growing body of evidence showing frequent family meal to be related with regular breakfast eating and overall improved diet quality [20,59,60]. We found a significant association between these behaviors in our cross-sectional analyses, and irregular family breakfast at baseline was related to increased risk of both skipping and irregular breakfast frequency. However, there was no effect of interaction between time and family breakfast frequency.

It is important to highlight that significant differences in type of school proportions between participants present at the baseline and those present in the two time points of our study can be explained by the fact that student’s evasion from schools was greater among adolescents from the public ones, as they started to work or entered military service. Also, students from public schools are older and are the ones who most left of our study, which explains the significant differences between baseline age averages of students participating at baseline and those who continued participating after three years. Therefore, differences between the losses according to type of school and age may not be related to breakfast eating pattern, which speaks against the possibility of a pattern of missing related to the outcome. Also, despite the high percentage of ELANA losses, all longitudinal modeling accounts for losses for the outcome, and the exposure variables were analyzed at baseline.

This study is limited in what concerns external validity, and thus, other longitudinal cohort with larger and more representative samples of adolescents from Brazil can corroborate to further explore this scenario. Also, more research is necessary with regards to explore how breakfast pattern change over time. Our findings showed that in three years student`s breakfast composition tend to decrease for some foods/food groups, while maintain a certain stability for others.

Our study represents advancement in the knowledge of breakfast frequency and composition and factors related to its consumption, especially the role of weight status on breakfast change. Given our findings of higher risk of breakfast to be less consumed among girls and that breakfast composition worsened as adolescents grow old, it is fundamental to encourage its consumption in this age group, once breakfast eating is recognized to be related to healthier food consumption and several other health aspects [6,911,13,18,19]. The promotion of heath programs considering the school environment could help changing this scenario. In Brazil, adolescents attending public schools have been receiving lunch as part of the Brazilian School Feeding Program, a successful public policy implemented in 1954 [61]. Objectively, considering breakfast inclusion in this program or, if not possible, planning intervention strategies through health education activities that stimulate breakfast eating could promote healthier food consumption among this age group.

In the public health perspective, our findings can also contribute to subsidize future dietary guidance for Brazilians, since the last Food Guideline for Brazilian Population, published in 2014, valorized eating and composition of the three main meals: breakfast, lunch and dinner. In Brazil, there is still lack of information on adolescent`s breakfast patterns and our findings contribute to fulfill this gap.

Supporting information

S1 File. Data set for foods/food groups analyses.


S2 File. Data set for breakfast frequency analyses.



We are thankful to all adolescents that agreed to participate of the study.


  1. 1. Lytle LA, Seifert S, Greenstein J, McGovern P. How do children’s eating patterns and food choices change over time? Results from a cohort study. Am J Health Promot. 2000; 14: 222–228. pmid:10915532
  2. 2. Sjöberg A, Hallberg L, Höglund D, Hulthén L. Meal pattern, food choice, nutrient intake and lifestyle factors in the Göteborg Adolescence Study. Eur J Clin Nutr. 2003; 57: 1569–1578. pmid:14647222
  3. 3. Keski-Rahkonen A, Kaprio J, Rissanen A, Virkkunen M, Rose RJ. Breakfast skipping and health-compromising behaviors in adolescents and adults. Eur J Clin Nutr. 2003; 57: 842–853. pmid:12821884
  4. 4. Ahadi Z, Qorbani M, Kelishadi R, Ardalan G, Motlagh ME, Asayesh H, et al. Association between breakfast intake with anthropometric measurements, blood pressure and food consumption behaviors among Iranian children and adolescents: the CASPIAN-IV study. Public Health. 2015; 129 (6): 740–747. pmid:26043966
  5. 5. Inchley J, Currie D, Young T, Samdal O, Torsheim T, Augustson L, et al. Growing up unequal: gender and socioeconomic differences in young people’s health and well-being. Health Behaviour in School-aged Children (HBSC) Study: International Report from the 2013/2014 Survey. Copenhagen: WHO Regional Office for Europe; 2016. Health Policy for Children and Adolescents, Report No.: 7.
  6. 6. Rampersaud GC, Pereira MA, Girard BL, Adams J, Metz JD, Breakfast habits, nutritional status, body weight, and academic performance in children and adolescents. J Am Diet Assoc. 2005; 105: 743–760. pmid:15883552
  7. 7. Estima CCP, Costa RS, Sichieri R, Pereira RA, Da Veiga GV. Meal consumption patterns and anthropometric measurements in adolescents from a low socioeconomic neighborhood in the metropolitan area of Rio de Janeiro, Brazil. Appetite. 2009; 52: 735–739. pmid:19501773
  8. 8. Smith KJ, Gall SL, Mcnaughton SA, Blizzard L, Dwyer T, Venn AJ. Skipping breakfast: longitudinal associations with cardiometabolic risk factors in the Childhood Determinants of Adult Health Study. Am J Clin Nutr. 2010; 92: 1316–1325. pmid:20926520
  9. 9. Deshmukh-Taskar PR, Nicklas TA, O’neil CE, Keast DR, Radcliffe JD, Cho S. The Relationship of Breakfast Skipping and Type of Breakfast Consumption with Nutrient Intake and Weight Status in Children and Adolescents: The National Health and Nutrition Examination Survey 1999–2006. J Am Diet Assoc. 2010; 110: 869–878. pmid:20497776
  10. 10. Hallström L, Labayen I, Ruiz JR, Patterson E, Vereecken CA, Breidenassel C, et al. Breakfast consumption and CVD risk factors in European adolescents: the HELENA (Healthy Lifestyle in Europe by Nutrition in Adolescence) study. Public Health Nutr. 2012; 12: 1–10.
  11. 11. Affenito SG, Thompson DR, Barton BA, Franko DL, Daniels SR, Obarzanek E, et al. Breakfast consumption by African- American and while adolescents girls correlates positively with calcium and fiber intake and negatively with body mass index. J Am Diet Assoc. 2005; 105: 938–945. pmid:15942545
  12. 12. Timlin MT, Pereira MA. Breakfast frequency and quality in the etiology of adult obesity and chronic diseases. Nutr Rev. 2007; 65 (6): 268–281.
  13. 13. Lazarou C, Matalas AL. Breakfast intake is associated with nutritional status, Mediterranean diet adherence, serum iron and fasting glucose: the CYFamilies study. Public Health Nutr. 2015; 18 (7): 1308–1316. pmid:25287356
  14. 14. Philippi S. T. Pirâmide dos alimentos: fundamentos básicos da nutrição (Food pyramid: basic fundamentals of nutrition). Barueri: Manole; 2008.
  15. 15. Merten MJ, Williams AL, Shriver LH. Breakfast consumption in adolescence and young adulthood: parental presence, community context, and obesity. J Am Diet Assoc. 2009; 109 (8): 1384–1391. pmid:19631044
  16. 16. Timlin MT, Pereira MA, Story M, Neumark-Sztainer D. Breakfast Eating and Weight Change in a 5-Year Prospective Analysis of Adolescents: Project EAT (Eating Among Teens). Am Acad Pediatrics. 2008; 21 (3): 638–645.
  17. 17. Trancoso SC, Cavalli SB, Proença RPC. Café da manhã: caracterização, consumo, e importância para a saúde (Breakfast: characterization, consumption, and importance to health). Rev Nutr. 2010; 23 (5): 859–869.
  18. 18. Hoyland A, Dye L, Lawton CL. A systematic review of the effect of breakfast on the cognitive performance of children and adolescents. Nutr Res Rev. 2009; 22: 220–243. pmid:19930787
  19. 19. Affinita A, Catalani L, Cecchetto G, Lorenzo G, Dilillo D, Donegani G, et al. Breakfast: a multidisciplinary approach. Ital J Pediatrics. 2013; 39 (44): 1–10.
  20. 20. Fulkerson JA, Larson N, Horning M, Neumark-Sztainer D. A review of associations between family or shared meal frequency and dietary and weight status outcomes across the lifespan. J Nutr Educ Behav. 2014; 46 (1): 1–19.
  21. 21. Neumark-Sztainer D, Wall M, Haines J, Story M, Eisenberg ME. Why does dieting predict weight gain in adolescents? Findings from Project EAT-II: A 5-year longitudinal study. J Am Diet Assoc. 2007; 107: 448–455. pmid:17324664
  22. 22. Brown AW, Brown MMB, Allison DB. Belief beyond the evidence: using the proposed effect of breakfast on obesity to show 2 practices that distort scientific evidence. Am J Clin Nutr. 2013; 98: 1298–1308. pmid:24004890
  23. 23. Safiri S, Kelishadi R, Qorbani M, Lotfi R, Djalalinia S, Salehifar D, et al. Association of Dietary Behaviors with Physical Activity in a Nationally Representative Sample of Children and Adolescents: the CASPIAN- IV Study. Int J Pediatr. 2016; 4: 1505–1517.
  24. 24. Pedersen PT, Meilstrup C, Holstein BE, Rasmussen M. Fruit and vegetable intake is associated with frequency of breakfast, lunch and evening meal: cross-sectional study of 11-, 13-, and 15-year-olds. Int J Behav Nutr Phys. 2012; 9 (9): 1–7.
  25. 25. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional de Saúde do Escolar, 2012 (National School Health Survey, 2012). Rio de Janeiro: IBGE; 2013.
  26. 26. David Doku D, Koivusilta L, Raisamo S, Rimpela A. Socio-economic differences in adolescents’ breakfast eating, fruit and vegetable consumption and physical activity in Ghana. Public Health Nutrition. 2011; 16(5): 864–872. pmid:22030213
  27. 27. Vereecken C, Dupuy M, Rasmussen M, Kelly C, Nansel TR, Sabbah HA, et al. Breakfast consumption and its socio-demographic and lifestyle correlates in schoolchildren in 41 countries participating in the HBSC study. Int J Public Health 54 (2009) S180–S190.
  28. 28. Nakamura S, Inayama T, Hata K, Matsushita M, Takahashi M, Harada K. Association of household income and education with eating behaviors in Japanese adults: a cross-sectional study. BMC Public Health. 2016; 16: 61. pmid:26800891
  29. 29. Lawman HG, Polonsky HM, Veur SSV, Abel ML, Sherman S, Bauer KW, et al. Breakfast patterns among low-income, ethnically- diverse 4th-6th grade children in an urban area. BMC Public Health. 2014; 14 (604): 1–9.
  30. 30. Brasil LMP, Fisberg M, Maranhão HS. Excess weight in children from Brazilian Northeast: difference between public and private schools. Rev Bras Saúde Mater Infant. 2007; 7: 405–412.
  31. 31. Sune FR, Dias-da-Costa JS, Olinto MT, Pattussi MP. Prevalence of overweight and obesity and associated factors among schoolchildren in a southern Brazilian city. Cad Saude Publica. 2007; 23: 1361–1371. pmid:17546327
  32. 32. Campos LA, Leite AJL, Almeida APC. Prevalence of overweight and obesity among adolescent students in the city of Fortaleza, Brazil. Rev Bras Saúde Mater Infant. 2007; 7: 183–190.
  33. 33. Berkey CS, Rockett HRH, Gilman MW, Field AE, Colditz GA. Longitudinal study of skipping breakfast and weight change in adolescents. Int J Obes. 2003; 27: 1258–1266.
  34. 34. Araújo MC, Yokoo EM, Pereira RA. Validation and calibration of a semiquantitative food frequency questionnaire designed for adolescents. J Am Diet Assoc. 2010; 110 (8): 1170–1177. pmid:20656092
  35. 35. Correa AS, Rodrigues PRM, Monteiro LS, Souza RAG, Sichieri R, Pereira RA. Beverages characterize the nutritional profile of Brazilian adolescents’ breakfast. Nutrire. 2016; 41 (3): 2–11.
  36. 36. Matsudo S, Araújo T, Matsudo V, Andrade D, Andrade E, Oliveira LC, et al. Questionário Internacional de Atividade Física (IPAQ): estudo de validade e reprodutibilidade no Brasil (International Physical Activity Questionnaire (lPAQ): study of validity and reliability in Brazil). Brazilian J Phys Act Health. 2001; 6 (3): 5–12.
  37. 37. Ainsworth BE, Haskell WL, Whitt MC, IRWIN ML, SWARTZ AM, STRATH SJ, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. 2000; 32 Suppl: S498–504.
  38. 38. Patterson E. Guidelines for data processing and analysis of the international physical activity questionnaire (GDPA-IPAQ); 2005.
  39. 39. Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos Familiares, 2008–2009: análise do consumo alimentar pessoal no Brasil (National Household Budget Survey: 2008/2009. Analysis of individual food consumption in Brazil). Rio de Janeiro: IBGE; 2010.
  40. 40. Lohman TGRA, Roche AF, Martorell R. Anthropometric Standardization Reference Manual. Illinois: Human Kinetics Books; 1988.
  41. 41. Habicht JP. Estandardización de metodos epidemiológicos cuantitativos sobre el terreno. (Standardization of quantitative epidemiological methods in the field). Bol Oficina Sanit Panam. 1974; 76 (5): 375–384.
  42. 42. Pederson D, Gore C. Anthropometry measurement error. In: Norton K, Olds T, editors. Antropometrica. Sidney: University of New South Wales Press; 1996. pp. 77–96.
  43. 43. De Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bull World Health Organ. 2007; 85 (9): 660–667. pmid:18026621
  44. 44. Fitzmaurice GM, Laird NM, Ware JH. Longitudinal and Clustered data. In: Fitzmaurice GM, Laird NM, Ware JH (eds.) Applied Longitudinal Analysis, 2nd edn. Wiley: Hoboken, NJ, USA, 2011, pp.1–18.
  45. 45. Fitzmaurice GM, Laird NM, Ware JH. Missing data and dropout: Multiple imputation and weighting methods. In: Fitzmaurice GM, Laird NM, Ware JH (eds.) Applied Longitudinal Analysis, 2nd edn. Wiley: Hoboken, NJ, USA, 2011, pp.515–550.
  46. 46. Niemeier HM, Raynor HA, Lloyd-Richardson EE, Rogers ML, Wing RR. Fast food consumption and breakfast skipping: predictors of weight gain from adolescence to adulthood in a nationally representative sample. J Adolesc Health. 2006; 39: 842–849. pmid:17116514
  47. 47. Pedersen PT, Holstein BE, Flachs EM, Rasmussen M. Meal frequencies in early adolescence predict meal frequencies in late adolescence and early adulthood. BMC Public Health. 2013; 13(445): 1–10.
  48. 48. Dhurandhar EJ. True, true, unrelated? A review of recent evidence for a causal influence of breakfast on obesity. Curr Opin Endocrinol Diabetes Obes. 2016; 23: 384–388. pmid:27584010
  49. 49. Albertson AM, Franko DL, Thompson D, Eldridge AL, Holschuh N, Affenito SG, et al. Longitudinal patterns of breakfast eating in black and white adolescent girls. Obesity. 2007; 15: 2282–2292. pmid:17890497
  50. 50. Dialektakou KD, Vranas PB. Breakfast skipping and body mass index among adolescents in Greece: whether an association exists depends on how breakfast skipping is defined. J Acad Nutr Diet. 2008; 108 (9): 1517–1525.
  51. 51. Santos RO, Fisberg RM, Marchioni DML, Baltar VT. Dietary patterns for meals of Brazilian adults. Br J Nutr. 2015; 114: 822–828. pmid:26220554
  52. 52. Mattos LL, Martins IS. Consumo de fibras alimentares em população adulta. Rev Saúde Pública. 2000; 34: 50–55.
  53. 53. Stang JS, Larson N. Nutrition in adolescence. In: Mahan LK, Escott-Stump S, Raymond JL. Krause’s Food and Nutrition Therapy. 13rd ed. St. Louis/Minneapolis: Elsevier; 2012. pp. 410–430.
  54. 54. Veiga GV, Costa RS, Araújo MC, Souza AM, Bezerra IN, Barbosa FS, et al. Inadequação do consumo de nutrientes entre adolescentes brasileiros. Rev Saúde Pública. 2013; 47 Suppl 1: S212–221.
  55. 55. Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos Familiares no Brasil, 2008/2009: Antropometria e estado nutricional de crianças, adolescentes e adultos no Brasil (National Household Budget Survey: 2008/2009. Anthropometry and nutritional status of children, adolescents and adults in Brazil). Rio de Janeiro: IBGE; 2010.
  56. 56. Souza AM, Pereira RA, Yokoo EM, Levy RB, Sichieri R. Alimentos mais consumidos no Brasil: Inquérito Nacional de Alimentação 2008–2009 (Most consumed foods in Brazil: National Food Survey 2008–2009). Rev Saúde Pública. 2013; 47 Suppl 1: S190–199, 2013.
  57. 57. Neumark-Sztainer D, Story M, Ackard D, Moe J, Perry C. The "Family Meal": Views of Adolescents. J Nutr Educ. 2000; 32 (6): 329–334.
  58. 58. Fismen AS, Smith ORF, Torsheim T, Samdal O. A school-based study of time trends in food habits and their relation to socio-economic status among Norwegian adolescents, 2001–2009. Int J Behav Nutr Phys Act. 2014; 11: 1–10.
  59. 59. Burgess-Champoux TL, Larson N, Neumark-Sztainer D, Hannan PJ, Story M. Are Family Meal Patterns Associated with Overall Diet Quality during the Transition from Early to Middle Adolescence? J Nutr Educ Behav. 2009; 41 (2): 79–86. pmid:19304252
  60. 60. Utter J, Scragg R, Schaaf D, Mhurchu CN. Relationships between frequency of family meals, BMI and nutritional aspects of the home food environment among New Zealand adolescents. Int J Behav Nutr Phys Act. 2008; 5: 50–57. pmid:18947431
  61. 61. Hawkes C, Brazil BG, Castro IRR, Jaime PC. How to engage across sectors: lessons from agriculture and nutrition in the Brazilian School Feeding Program. Rev Saúde Pública. 2016; 50.