Summary
Keywords: Childhood obesity, double breakfast, middle school students, school breakfast
Objectives:
The objectives are to identify breakfast location patterns (frequency and place of breakfast consumption) and explore the association between breakfast patterns and weight status over time among preadolescents.
Methods:
Surveys and physical measurements were completed among students from 12 randomly selected schools in a medium-sized urban school district. All students were followed from fifth (Fall, 2011) to seventh grade (Fall, 2013). Latent transition analysis and longitudinal analyses were used in the study.
Results:
Six distinct breakfast location patterns emerged at baseline (1) frequent skippers; (2) inconsistent school eaters; (3) inconsistent home eaters; (4) regular home eaters; (5) regular school eaters and (6) double breakfast eaters. Results from the longitudinal analyses revealed that there was an increased odds of overweight/obesity among frequent skippers compared with double breakfast eaters after adjusting for school, year and students’ race/ethnicity (AOR: 2.66, 95% CI: 1.67, 4.24). Weight changes from year to year were similar between double breakfast eaters and other students.
Conclusions:
Concerns that a second breakfast at school increases risk of excessive weight gain are unsupported. Students who regularly consumed breakfasts at school, including double breakfast eaters, were more likely to exhibit a healthy weight trajectory. Additional research is needed to understand the impact of universal school breakfast on students’ overall diets.
Introduction
Breakfast consumption has been associated among school children with improved cognitive performance, nutritional adequacy, bone and cardiovascular health (1,2), as well as healthy body weight (3,4). Conversely, skipping breakfast has been associated with obesity (2,5,6). The National School Breakfast Program (SBP) in the United States (US) is a federally funded meal programme designed to provide a nutritious meal to students in public and nonprofit private schools. The programme provides cash assistance to states to operate nonprofit breakfast programmes in >89000 schools and institutions nationwide, serving nearly 13 million children daily (7).
Currently, there are national advocacy efforts in the US from nonprofit organizations, foundations and corporations to promote higher participation in school breakfast, specifically, to make it easier for low-income communities to serve universal breakfast (i.e. daily breakfast to all students at no cost) (8). One of their recommendations is to use strategies such as serving breakfast in the classroom (BIC), or creating ‘grab and go’ breakfast boxes, to increase participation (9,10). However, a concern is that promoting school breakfast may inadvertently increase the likelihood of students consuming a double breakfast – by eating once before coming to school and once at school. The US Department of Agriculture School Breakfast Pilot Study of six school districts found that around 20% of students ate two or more breakfasts; among these students, 46% ate a ‘substantive breakfast’ at home in addition to the breakfast consumed at school (11). These findings raise the question of whether advocacy for maximum participation in the SBP may conflict with efforts to limit excess calories from school meals. Therefore, research is needed to determine the association between eating breakfast at school and body weight over time (12,13).
The aims of this study are to (1) identify breakfast location patterns (frequency and place of breakfast consumption) and changes in location patterns in a longitudinal sample of middle school students (from Grade 5 through 7); (2) explore predictors of breakfast location patterns; (3) assess the association between breakfast location patterns and weight over time and (4) examine whether students who consume a double breakfast have an increased risk of absolute weight gain compared with students with other location patterns.
Methods
Study design
Twelve schools were randomly selected from a total of 27 schools serving kindergarten through eighth grade in a medium-sized urban school district; all 12 schools agreed to participate. Students were eligible to participate if they were enrolled in fifth grade during the 2011–2012 school year. Participants were followed from fifth grade in 2011–2012 to seventh grade in 2013–2014. Student assent and parental consent were obtained prior to data collection.
Measures
Student surveys
Breakfast location patterns were defined by where students are eating and how frequently they are eating at these locations. Two items from the student surveys were used to describe breakfast location patterns (1) average number of days per week they eat breakfast (0–7) and (2) location where they ate breakfast the previous school day (home, school, both home and school or no breakfast).
Physical measurements
Trained research staff measured students’ height and weight using the World Health Organization Expanded STEPwise approach to Surveillance protocol (14). Height was measured in inches using a stadiometer (Charder Electronic, Taichung City, Taiwan), and weight was measured in pounds using an electronic flat scale (Seca, Hamburg, Germany). Measured heights and weights were then used to calculate body mass index (BMI) for each student, using the Centers for Disease Control and Prevention sex-specific and age-specific BMI percentile calculator (15).
Administrative data
Students’ sex, race/ethnicity and grade were obtained from school district records. Eligibility for free or reduced price school breakfast, set at 130% and 185%, respectively, of the federal poverty line (16), was included as a proxy for family socioeconomic status. Type of breakfast delivery model (BIC or serving in the cafeteria) was collected.
Statistical analysis
Latent transition analysis was used to identify unobserved breakfast location patterns underlying the observed data, and to estimate transition probabilities and movements between the identified patterns over time (17). Frequency and place of breakfast consumption from the two questions in the student surveys were treated as categorical variables and were used as indicators of breakfast location patterns. Akiake information criterion, Bayesian information criterion, likelihood ratio G2 statistic, model parsimony and interpretability criteria were considered when selecting the best model for the study
Latent transition analysis allowed estimation of baseline latent status membership probabilities, item-response probabilities conditional on time and latent status membership, transition probabilities and beta coefficients of logistic regression as published elsewhere (17).
Generalized estimating equations (GEE) models for categorical outcomes were used to examine whether the latent statuses membership predicts BMI trajectory over time. BMI values were collapsed into a binary variable: overweight/obese and normal/underweight for ease of interpretation. Students’ sex, race/ethnicity, school and study year were incorporated into GEE models. All covariates were chosen a priori and were significantly associated with the outcome. In addition, we examined whether latent statuses membership is associated with weight changes over time, adjusting for year using linear mixed models. Weight changes were calculated as the current year’s weight minus past year’s weight. All statistical analyses were performed using SAS software (version 9.2, SAS Institute, Cary, North Carolina, USA).
Results
Study participants
Complete data were available for 584 (85.4%) students in fifth grade in 2011, 602 (89.7%) students in sixth grade in 2012 and 539 (77.7%) students in seventh grade in 2013. The main reasons for non-participation were students absent during data collection (6.0% in 2011, 2.5% in 2012 and 15.9% in 2013), or no informed consent (8.6% in 2011, 7.7% in 2012 and 6.5% in 2013). Students who opted out or missed data collection in fifth grade were still eligible to participate the following year, as were students who transferred into participating schools. Students with data for only one of the three study years [11% {N =191}] were excluded.
The final analytic sample is described in Table 1. It included 513 fifth grade, 553 sixth grade and 468 seventh grade students. There was no significant difference in sex, age, BMI status and breakfast consumption between students in the final analytic sample and those excluded because of participation in only one study year. However, those excluded with data for only one study year had a lower proportion of Hispanics (32.3% vs. 46.8%) and a higher proportion of students who participated in BIC programme (40.6% vs. 22.7%).
Table 1.
Characteristics of the study sample
| Characteristic | Grade 5 (n = 513) | Grade 6 (n = 553) | Grade 7 (n = 468) | p for trend |
|---|---|---|---|---|
| N (%) | N (%) | N (%) | ||
| Covariates | ||||
| Sex | 0.575 | |||
| Male | 236 (46.0) | 248 (44.9) | 207 (44.2) | |
| Female | 277 (54.0) | 305 (55.2) | 261 (55.8) | |
| Race/ethnicity | 0.473 | |||
| Non-Hispanic White | 88 (17.2) | 99 (17.9) | 81 (17.4) | |
| Non-Hispanic Black | 182 (35.5) | 195 (35.3) | 153 (32.8) | |
| Hispanic | 239 (46.6) | 255 (46.1) | 229 (49.1) | |
| Other | 4 (0.8) | 4 (0.7) | 3 (0.6) | |
| Eligibility for SBP* | — | |||
| Free/reduced price | 415 (82.8) | — | — | |
| Full price | 86 (17.2) | — | — | |
| Participate in BIC programme# | <0.0001 | |||
| No | 397 (77.4) | 350 (63.3) | 393 (92.3) | |
| Yes | 116 (22.6) | 203 (36.7) | 33 (7.8) | |
| BMI | 0.221 | |||
| Underweight | 10 (2.0) | 9 (1.6) | 9 (1.9) | |
| Healthy weight | 227 (44.8) | 241 (43.6) | 224 (47.9) | |
| Overweight | 108 (21.3) | 119 (21.5) | 105 (22.4) | |
| Obese | 162 (32.0) | 184 (33.3) | 130 (27.8) | |
| BMI percentile, mean (SD) | 73.6 (29.3) | 75.0 (28.0) | 73.9 (27.7) | 0.845 |
| Indicators of latent statuses | ||||
| Breakfast frequency | <0.0001 | |||
| 0 day week | 14 (2.8) | 30 (5.4) | 28 (6.1) | |
| 1–3 days week | 72 (14.1) | 94 (17.0) | 101 (22.0) | |
| 4–5 days week | 78 (15.3) | 89 (16.1) | 78 (17.0) | |
| 6–7 days week | 346 (67.8) | 339 (61.4) | 253 (55.0) | |
| Yesterday’s breakfast location | 0.045 | |||
| Did not eat | 59 (11.6) | 97 (17.6) | 107 (23.1) | |
| Home | 302 (59.5) | 252 (45.7) | 216 (46.7) | |
| School | 85 (16.7) | 138 (25.1) | 86 (18.6) | |
| Both home and school | 62 (12.2) | 64 (11.6) | 54 (11.7) | |
Dashes indicate missing information.
Eligibility for free or reduced price school breakfast is set at 130% and 185%, respectively, of the federal poverty line.
BIC programme serves universal breakfast to students after the opening bell, in the classroom. BIC, breakfast in the classroom; BMI, body mass index; SBP, School Breakfast Program.
Breakfast location patterns
Using fit indices and model interpretability criteria, a six-class model was selected. Based on values of item-response probabilities (Table S1), the six latent status categories were (1) frequent skippers (71% reported eating breakfast 0–3 times a week and 100% reported not eating breakfast the day before); (2) inconsistent school eaters (97% reported eating breakfast 1–5 days a week and 77% ate at school the day before); (3) inconsistent home eaters (97% reported eating breakfast 1–5days per week and 100% ate at home the day before); (4) regular home eaters (100% reported eating breakfast 6–7 days a week and 100% ate at home the day before); (5) regular school eaters (100% reported eating breakfast 6–7 days a week and 100% ate at school the day before) and (6) double breakfast eaters (100% reported eating breakfast 6–7 days a week and 100% ate at school and home the day before).
Table 2 presents the prevalence of each status for each grade. Overall, breakfast frequency declined over time as students aged, and significantly more students skipped breakfast in seventh grade than earlier. At baseline, the most prevalent status was regular home eaters (43.7%), followed by inconsistent (home or school combined) eaters (22.6%). Notably, the proportion of students in the skippers group progressively increased over time, with 22.9% of the students in this group by seventh grade. Table 2 also presents the likelihood of students transitioning from one breakfast status to another over time. For instance, regular home eaters in fifth grade had 41.7% probability of being in the same status in sixth grade, and 15.3% chance of transitioning to skippers status. The highest probabilities of transitioning to the skippers status was among students in the inconsistent eaters groups. Similarly, skippers who changed statuses were most likely to transition to inconsistent eaters, and very unlikely to become double breakfast eaters over the study period. Overall, there was a higher probability for changes in status membership from fifth to sixth grade compared with sixth to seventh grade.
Table 2.
Class item-response probabilities, prevalence of latent statuses and transition probabilities
| Frequent skippers | Inconsistent school eaters | Inconsistent home eaters | Regular home eaters | Regular school eaters | Double breakfast eaters | |
|---|---|---|---|---|---|---|
| Prevalence of statuses at (%): | ||||||
| Grade 5 | 11.5 | 6.8 | 15.8 | 43.7 | 11.9 | 10.1 |
| Grade 6 | 17.5 | 9.4 | 14.4 | 31.6 | 16.9 | 10.3 |
| Grade 7 | 22.9 | 12.4 | 12.9 | 34.1 | 9.8 | 8.0 |
| Transitions from Grade 5 (rows) to Grade 6 (columns) (%)*: | ||||||
| Frequent skippers | 34.7 | 16.3 | 22.3 | 12.7 | 12.2 | 1.9 |
| Inconsistent school eaters | 14.1 | 18.5 | 11.2 | 20.6 | 14.0 | 21.6 |
| Inconsistent home eaters | 20.6 | 11.1 | 11.7 | 29.2 | 11.7 | 9.6 |
| Regular home eaters | 15.3 | 4.8 | 11.6 | 41.7 | 19.0 | 7.7 |
| Regular school eaters | 13.9 | 12.1 | 15.7 | 17.2 | 27.5 | 13.7 |
| double breakfast eaters | 9.6 | 9.8 | 13.4 | 37.3 | 11.3 | 18.6 |
| Transitions from Grade 6 (rows) to Grade 7 (columns) (%)*: | ||||||
| Frequent skippers | 50.1 | 15.1 | 15.7 | 11.9 | 3.7 | 3.5 |
| Inconsistent school eaters | 27.2 | 31.3 | 15.8 | 13.4 | 5.1 | 7.2 |
| Inconsistent home eaters | 28.0 | 12.0 | 24.1 | 21.5 | 6.6 | 7.9 |
| Regular home eaters | 11.1 | 7.3 | 11.2 | 62.0 | 5.0 | 3.5 |
| Regular school eaters | 13.9 | 11.2 | 8.4 | 29.9 | 30.1 | 6.5 |
| double breakfast eaters | 16.4 | 8.3 | 2.2 | 29.2 | 10.3 | 33.6 |
Transition probabilities are presented here as percentages (by multiplying by 100) for easier interpretation.
Predictors of breakfast location patterns
We conducted repeated measurement GEE models to examine predictors of breakfast location patterns. Using the double breakfast group as the reference group, significant sex differences in breakfast patterns were identified. Compared with boys, girls were more likely to belong in the skippers status (AOR: 3.00, 95% CI: 1.79, 5.02), inconsistent school eaters (AOR: 1.40, 95% CI: 0.81, 2.42), inconsistent home eaters (AOR: 3.03, 95% CI: 1.84, 5.01), regular home eaters (AOR: 1.98, 95% CI: 1.30, 3.03) and regular school eaters (AOR: 1.70, 95% CI: 1.03, 2.80) relative to double breakfast status, even after adjusting for year, weight status and BIC program.
In addition, significant differences emerged between overweight/obese students and normal weight students. Specifically, overweight and obese students were more likely to be skippers (AOR: 2.85, 95% CI: 1.74, 4.69), inconsistent school eaters (AOR: 2.51, 95% CI: 1.46, 4.32), inconsistent home eaters (AOR: 2.63, 95% CI: 1.61, 4.31) or regular home eaters (AOR: 1.74, 95% CI: 1.16, 2.62) than double breakfast eaters. We examined whether weight status predicted transition from one breakfast status to another at subsequent year, adjusting for race/ethnicity. We ran models including only baseline weight status and models including the status at each time point. In both models, being obese or overweight at any time did not predict transition from one status to another (all AORs = 1.00).
Association of breakfast location patterns with obesity status
The overall proportion of overweight and obese students in this cohort did not change significantly over time. Weight category was not proportionally distributed across the six latent breakfast statuses (Fig. 1). Notably, the proportion of students classified as double breakfast eaters who were identified as healthy weight increased over time (51.9% in Grade 5, 54.4% in Grade 6 and 79.5% in Grade 7).
Figure 1.
Weight status distribution of the six breakfast location patterns.
Because latent transition analysis allows examining the effect of weight status on transition between statuses but does not allow examining the significance of the association with latent status except at baseline, we conducted a longitudinal data analysis to examine whether breakfast patterns are significantly associated with BMI status over time. After accounting for clustering of students within schools and adjusting for year and race/ethnicity, a significant association between latent breakfast class membership and BMI category was revealed (p = 0.002) (Figure 1). Odds of being overweight or obese was significantly more likely for students in the skippers group compared with double breakfast eaters (AOR: 2.66, 95% CI: 1.67, 4.24). Similarly, inconsistent school eaters (AOR: 2.11, 95% CI: 1.29, 3.46), inconsistent home eaters (AOR: 2.02, 95% CI: 1.27, 3.21) and regular home eaters (AOR: 1.70, 95% CI: 1.13, 2.56) all were more likely to be overweight or obese compared with double breakfast eaters. Further, Hispanics (AOR: 1.78, 95% CI: 1.14, 2.78) and non-Hispanic black preadolescents (AOR: 1.75 95% CI: 1.10, 2.79) also had higher odds of obesity than non-Hispanic white preadolescents. Finally, to examine the specific question of whether students who consume a double breakfast have increased risk of excessive weight gain compared with students in the other breakfast categories, we tested the association between breakfast status and weight change (i.e. difference in BMI from past year) over time. We found that there was no difference between weight changes of double breakfast eaters over time compared with any of the other breakfast categories adjusting for year (F =0.67, p > 0.05). In other words, there was no evidence of greater weight gain over time among students who consume a double breakfast when compared with all other students.
Discussion
To our knowledge, this is the first longitudinal study to explore breakfast location patterns, including double breakfast, and obesity risk in the US, using a sample of middle school children in a diverse, urban district. The percentage of students eligible for free or reduced price meal in this study is 83%, which is much higher than the national average of 51% (18). The rate of overweight and obesity in this sample exceeds 50%, well above the national average of 35% for this age group (19). In this high-risk sample, six qualitatively unique patterns of breakfast consumption were identified prevalence of these patterns varied by sex, race/ethnicity and weight status.
Our observation that skipping breakfast increased over the 3-year time period and was more common in female students has been noted in other studies (20,21). The association we found at each time point between breakfast skipping and higher weight status is also consistent with previous cross-sectional studies (2,5,6,22,23). The reason why skipping breakfast is associated with higher weight is not well understood. It may reflect some degree of reverse causality if overweight and obese students think skipping breakfast will help them lower caloric consumption. Another theory is that skipping breakfast leads to overconsumption later in the day due to increased hunger; however, a recent review of the literature on breakfast and weight found that available evidence from randomized controlled trials is not sufficient to draw any causal connection between breakfast skipping and obesity (24). Nevertheless, even if breakfast skipping does not cause weight gain, eating breakfast is recommended because it is associated with a higher diet quality (22). In our sample, the largest increase in breakfast skipping was between fifth and sixth grades, especially among the inconsistent home eaters, suggesting this group may benefit from targeted breakfast promotion interventions.
Our study adds to the literature that has monitored weight and different breakfast location patterns longitudinally (25). Student’s weight changes from one school year to the next were similar across all breakfast groups, including the double breakfast eaters. A recent study by Vargas et al. (26) reported similar findings: although male adolescents in SBP were more likely to be double breakfast eaters, there was no association between SBP involvement and the probability of being overweight. The finding that students who eat two breakfasts do not gain significantly more weight than students who eat one breakfast appears paradoxical because they are eating an additional meal. It is possible that the double breakfast eaters may be more active and expend more energy during the day, particularly given the male predominance of double breakfast eaters found in Vargas et al. and in our study. Another possibility is that eating more calories earlier in the day is compensated by lower caloric consumption later in the day. Additional research is needed to examine energy intake over an entire day for children who eat one for children who eat one or two breakfasts, or skip breakfast on school days to better understand the link between morning meals, caloric intake and weight.
This study has several limitations. Our data are observational, not experimental, and the reasons why students have specific breakfast patterns are unknown; therefore, we cannot infer causal associations between breakfast consumption and weight outcomes. We also did not measure the quality or quantity of the breakfast consumed (e.g. did ‘double breakfast eaters’ have two small meals or eat twice as much?); thus, there is likely to be great heterogeneity in the caloric consumption among double breakfast eaters. Further research using direct observation such as through plate-waste data or accompanied by detailed 24-h recall is necessary to measure breakfast quality and to better understand the double breakfast eater consumption pattern. In our cohort, we identified around 10% of students who were double breakfast eaters during our study years. This is lower than the 20% reported in a US Department of Agriculture pilot study (11) and the 51% reported in a New York City study of double breakfast eaters when breakfast was served in the classroom (27). A possible reason is that students who were excluded from this study had a higher proportion of students who participated in BIC programme, which may have a positive association with double breakfast consumption. We did not explore reasons behind the double eating behavior; further qualitative research and detailed measurement of food security are needed in future research. We recognize that middle school is a period of rapid physical development where students are growing taller and gaining weight. Because growth spurts differ among adolescents and can lead to changes in dietary patterns, BMI may not be the most reliable measure of obesity. Further, research has shown that pubertal onset may differ by weight status, such that obese children enter puberty earlier than normal-weight children (28). Future research should assess students’ pubertal stage as well as their possible effects on eating behaviours. Finally, self-reported data are subject to reporting error and social desirability bias. The study sample represents an ethnic and racially diverse low-income school district; findings may not generalize to other types of school districts.
Despite these limitations, there are also several notable strengths. This is the first study to use longitudinal data and latent transition analyses to examine breakfast consumption and obesity risk in a sample of middle school students. Furthermore, this is the first study to examine double breakfast eaters and weight status over time. There has been concern about the impacts of promoting school breakfast, as it can lead to double breakfast consumption and potential risk of obesity. We found no evidence that this group of students had higher weight status compared with other groups. Given nearly four million households are unable to provide adequate and nutritious food for their children at times during the year, maximizing access to school breakfast is an important strategy to reduce the risk of child hunger (29).
Supplementary Material
Class item-response probabilities.
Acknowledgements
Research was supported by the National Institute of Child and Human Development (#R01-HD070740; Jeannette Ickovics and Marlene Schwartz, multiple principal investigators).
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors would like to thank Kathryn Gilstad-Hayden for providing data management, Susan Peters for project management and Dr Catherine McCaslin for her leadership and expertise. The authors are extremely grateful for the strong collaborative partnership with the New Haven Public Schools and to the children and families who participated in the study.
Footnotes
Supporting Information
Additional supporting information can be found in the online version of this article at the publisher’s website:
Conflict of interest statement
The authors have no conflicts of interest to report.
References
- 1.Rampersaud GC, Pereira MA, Girard BL, Adams J, Metzl JD. Breakfast habits, nutritional status, body weight, and academic performance in children and adolescents. J Am Diet Assoc 2005; 105: 743–60. [DOI] [PubMed] [Google Scholar]
- 2.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: 220–43. [DOI] [PubMed] [Google Scholar]
- 3.Sandercock GR, Voss C, Dye L. Associations between habitual school-day breakfast consumption, body mass index, physical activity and cardiorespiratory fitness in English schoolchildren. Eur J Clin Nutr 2010; 64: 1086–92. [DOI] [PubMed] [Google Scholar]
- 4.de la Hunty A, Gibson S, Ashwell M. Does regular breakfast cereal consumption help children and adolescents stay slimmer? A systematic review and meta-analysis. Obes Facts 2013; 6: 70–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.So HK, Nelson EA, Li AM, et al. Breakfast frequency inversely associated with BMI and body fatness in Hong Kong Chinese children aged 9–18 years. Br J Nutr 2011; 106: 742–51. [DOI] [PubMed] [Google Scholar]
- 6.Szajewska H, Ruszczynski M. Systematic review demonstrating that breakfast consumption influences body weight outcomes in children and adolescents in Europe. Crit Rev Food Sci Nutr 2010; 50: 113–9 (accessed: 01 March 2016). [DOI] [PubMed] [Google Scholar]
- 7.United States Department of Agriculture. (2013). The school breakfast program 2013. [WWW document].URL http://www.fns.usda.gov/sites/default/files/SBPfactsheet.pdf
- 8.The White House. (2014). The White House and USDA announce school wellness standards. [WWW document]. URL https://www.whitehouse.gov/the-press-office/2014/02/25/white-house-and-usda-announce-school-wellness-standards
- 9.No Kid Hungry Compaign. (2014). Alternative breakfast models. [WWW document]. URL https://bestpractices.nokidhungry.org/school-breakfast/program-details (accessed: 01 March 2016).
- 10.Anzman-Frasca S, Djang HC, Halmo MM, Dolan PR, Economos CD. Estimating impacts of a breakfast in the classroom program on school outcomes. JAMA Pediatr 2015; 169: 71–7 (accessed: 01 March 2016). [DOI] [PubMed] [Google Scholar]
- 11.Bernstein LS, McLaughlin JE, Crepinsek MK, Daft LM. Evaluation of the school breakfast program pilot project: final report. Nutrition Assistance Program Report Seires, No. CN-04-SBP; 2004. (accessed: 01 March 2016). [Google Scholar]
- 12.Paxton AE, Baxter SD, Tebbs JM, et al. Nonsignificant relationship between participation in school-provided meals and body mass index during the fourth-grade school year. J Acad Nutr Diet 2012; 112: 104–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Baxter SD, Hardin JW, Guinn CH, Royer JA, Mackelprang AJ, Devlin CM. Children’s body mass index, participation in school meals, and observed energy intake at school meals. Int J Behav Nutr Phys Act 2010; 7: 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.World Health Organization. (2015). The STEPS instrument and support materials. [WWW document]. URL http://www.who.int/chp/steps/instrument/en/
- 15.Centers for Disease Control and Prevention. (2014). BMI percentile calculator for child and teen. [WWW document]. URL http://nccd.cdc.gov/dnpabmi/Calculator.aspx
- 16.United States Department of Agriculture. (2015). School meals: income eligibility guidelines. [WWW document]. URL http://www.fns.usda.gov/school-meals/income-eligibility-guidelines (accessed: 01 March 2016).
- 17.Lanza ST, Collins LM. A new SAS procedure for latent transition analysis: transitions in dating and sexual risk behavior. Dev Psychol 2008; 44: 446–56 (accessed: 01 March 2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Southern Education Foundation. (2015). Low income students now a majority in the nation’s public schools. [WWW document]. URL http://www.southerneducation.org/getattachment/4ac62e27-5260-47a5-9d02-14896ec3a531/A-New-Majority-2015-Update-Low-Income-Students-Now.aspx (accessed: 01 March 2016).
- 19.Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA 2014; 311: 806–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Siega-Riz AM, Popkin BM, Carson T. Trends in breakfast consumption for children in the United States from 1965–1991. Am J Clin Nutr 1998; 67: 748S–56S (accessed: 01 March 2016). [DOI] [PubMed] [Google Scholar]
- 21.Brugman E, Meuleester JF, Spee-Van der Wekke A, Verloove-Vanhorick SP. Breakfast-skipping in children and young adolescents in The Netherlands. Eur J Public Health 1998; 8: 325–8. [Google Scholar]
- 22.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; 110:869–78. [DOI] [PubMed] [Google Scholar]
- 23.Antonogeorgos G, Panagiotakos DB, Papadimitriou A, Priftis KN, Anthracopoulos M, Nicolaidou P. Breakfast consumption and meal frequency interaction with childhood obesity. Pediatr Obes 2012; 7: 65–72. [DOI] [PubMed] [Google Scholar]
- 24.Casazza K, Brown A, Astrup A, et al. Weighing the evidence of common beliefs in obesity research. Crit Rev Food Sci Nutr 2015; 55: 2014–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Berkey CS, Rockett HR, Gillman MW, Field AE, Colditz GA. Longitudinal study of skipping breakfast and weight change in adolescents. Int J Obes Relat Metab Disord 2003. October; 27: 1258–66. [DOI] [PubMed] [Google Scholar]
- 26.Vargas AJ. Effect of school breakfast program on the prevalence of breakfast skipping, double-dipping, and obesity among adolescents: a time use perspective. American Social Science Association Anual Meeting 2015. Boston, US. [Google Scholar]
- 27.Van Wye G, Seoh H, Adjoian T, Dowell D. Evaluation of the New York City Breakfast in the Classroom Program. American Journal of Public Health 2013; 103: e59–e64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.De LC, Marcovecchio ML, Chiavaroli V, de Giorgis T, Chiarelli F, Mohn A. Timing of puberty and physical growth in obese children: a longitudinal study in boys and girls. Pediatr Obes 2014; 9: 292–9. [DOI] [PubMed] [Google Scholar]
- 29.Coleman-Jensen A, Nord M, Singh A. (2013). House-hold food security in the United States in 2012. U.S. Department of Agriculture, Economic Research Service; [WWW document]. URL http://www.ers.usda.gov/media/1183208/err-155.pdf [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Class item-response probabilities.