Associations Between Nighttime Eating and Total Caloric Intake in College-Aged Students

................................................................................................................. ii ACKNOWLEDGMENTS .......................................................................................... iv PREFACE ......................................................................................................................v TABLE OF CONTENTS ............................................................................................ vi LIST OF TABLES .................................................................................................... viii MANUSCRIPT ..............................................................................................................1 APPENDICES .............................................................................................................32 A: Review of Literature.....................................................................34 B: Methods Timeline........................................................................43 C: Consent Forms .............................................................................44 D: Nutrition Assessment Survey.........................................................48 E: International Physical Activity Questionnaire....................................55 F: DHQ II Information ......................................................................58 G: Biochemical Measures Document ..................................................59 H: Anthropometric Instructional Document..........................................60 I: Anthropometric Assessment Document............................................61 J: HEI-2010 Total Score Export Information........................................62 K: HEI-2010 Scoring Guide.............................................................65 L: Table of Potential Pitfalls Identified in Thesis Proposal........................66 M: Table of Actions Taken to Manage Potential Pitfalls Identified in Thesis Proposal...............................................................................67


INTRODUCTION
Obesity is a nationwide concern across all age groups including young adults attending college. According to the American College Health Association-National College Health Assessment II in Spring 2016 approximately 35% of college students were reported to be overweight or obese in America 1 , as defined by a body mass index (BMI) over 24.9 kg/m 2 1-4 . Increased caloric intake has been found to be associated with an elevated body mass index (BMI) 5,6 . Nighttime eating may be a contributing factor to increased daily caloric intake; it has been associated with higher total caloric intake and weight gain in shift workers and older adults [7][8][9][10] . However, research has not examined this association in college-aged students [7][8][9][10] .
Nighttime eating has been classified by consumption of calories after a specific clock time or within a specified number of hours before sleep onset. Baron et al. classified nighttime eaters as those who consumed calories after 8:00PM or within 4 hours before sleep onset 7,9,10 . Nighttime eating has been identified by college students as a potential concern for weight gain, yet it has never been studied in this population 11 . Therefore, relationships between nighttime eating and caloric intake were examined.
Physical activity and gender affect total caloric intake, therefore both were examined in this analysis 8 .
Results from investigations between nighttime eating and diet quality are limited and contradicting, an association may exist for a later defined meal eating time. Reid and Baron examined nighttime eating using earlier parameters than Wang et al. and had found no difference in diet quality 3,32,33,38 . Breakfast consumption is often coupled with having a robust diet quality 37 , and in comparison breakfast skipping is associated with weight gain, poor diet quality, and shifting caloric intake to later in the day 4,15,38 .
Correlations have been found between higher caloric intake after 8:00PM and a higher total daily caloric intake, shorter sleep duration, and later sleep timing 3 . Repeated positive associations between nighttime eating and BMI have been found 3,32,38 , with the exception of one study in which no correlations were found between meal timing and BMI 33 . This may be related to the lower BMI of the sample, and two longitudinal studies reported that nighttime eaters gained more weight over time than non-nighttime eaters 2,17 . To our knowledge body composition and nighttime eating have not been examined using measured body fat percent.
The primary objective of this study was to examine possible relationships between nighttime eating and total calorie intake among undergraduate students at the University of Rhode Island (URI). The secondary objectives were to investigate associations between nighttime eating and dietary quality and sleep patterns. The exploratory objectives were to examine associations amongst nighttime eating and MetS risk and body composition.
This analysis aimed to determine if nighttime eating was related to total caloric intake, along with dietary quality, sleep patterns, metabolic syndrome (MetS) risk, and BMI status. This study identifies previously uninvestigated information regarding the prevalence of nighttime eating in college-aged students and correlations between nighttime eating and MetS risk for any aged population.

Study Design
Undergraduate students enrolled in a general nutrition course at URI completed a survey to identify nighttime eaters, and the validated Diet History Questionnaire II (DHQ II) was used to measure total caloric intake, the primary variable of interest 12 .
This was a supplemental study added to an ongoing Institutional Review Board approved research project examining the relationship between diet and chronic disease risk in college-aged students, (Nutrition Assessment Study [NAS]).
Nighttime eating was defined using data collected for the NAS in addition to assessing demographics, sleep and nighttime eating. This study examined anthropometric, biochemical, and survey data on students enrolled in the Applied General Nutrition course. This was an add on study that investigated dietary data collected during three semesters, spring 2016 through spring 2017 in college-aged students.
Nighttime eating was the independent variable, and total daily caloric intake, dietary quality examined by the Healthy Eating Index-2010 (HEI-2010) total score, total hours of sleep, number of metabolic syndrome risk criteria, and BMI were dependent variables. The primary hypothesis was that college-aged students defined as nighttime eaters would have a higher total caloric intake compared to those who do not participate in nighttime eating. The secondary hypotheses were that those who were identified as nighttime eaters would have a poorer diet quality compared to those who were not identified as nighttime eaters. Additionally, a higher proportion of identified nighttime eaters would report lower sleep quantities (<7 hours) compared those who were not identified as nighttime eaters. The exploratory hypotheses were that nighttime eaters would have more risk factors for MetS and a higher BMI status than those who were not identified as nighttime eaters.
All data collection occurred during five of the course lab sessions, with protocol guidelines in place for anthropometric, biochemical, and survey data. A timeline of data collection can be found in Appendix B. Undergraduate and graduate teaching assistants and research assistants were trained on proper protocol prior to the data collection process for anthropometric measures and other data collection techniques, and reliability of each measure was verified.

Data Collection
During the first lab, the study was explained; participating students read and signed consent forms to allow data to be used for research. Participants were required to be 18 years or older to participate; a copy of the consent form and surveys are presented in Appendices C through F. Participating students identification numbers were then entered into a password secure web portal, which provided students access where they virtually completed the NAS, DHQ II, and International Physical Activity Questionnaire (IPAQ) surveys 12,13 . Anthropometric and biochemical data were entered into the NAS web portal by research assistants.
The NAS survey and IPAQ were administered during lab two. Students logged into the NAS web portal to access the survey. Biochemical data and anthropometric measurements were collected during labs three and lab four respectively; Appendices G through I. During the last visit of the protocol the students completed the DHQ II.

Nighttime Eating Categorizations: NAS Survey
The NAS survey assessed multiple aspects of nutrition, including basic demographic information, weight satisfaction, and eating behaviors. It provided data to categorize students as nighttime eaters, and assess their sleep patterns. Lastly, significance was demonstrated in the remaining variables.

Total calorie intake and Diet Quality: DHQ II
The DHQ II was the food frequency questionnaire (FFQ) that provided an estimation of total daily caloric intake and evaluated dietary quality by providing the HEI-2010 total scores. Login to the web-based FFQ was overseen by a trained graduate teaching assistant and two undergraduate teaching assistants. Students took about one hour and 15 minutes to complete the DHQ II.
The DHQ II provided HEI-2010 total scores through an indirect statistical analysis software output, refer to Appendices J through K 15 . The HEI-2010 is a validated dietary quality index. Correlations between HEI-2010 total scores and BMI, nutrients in the diet, and individual's self-perception of diets have been identified 15,16 .
The DHQ II was designed and tested by the National Cancer Institute, the version utilized in this study asked about dietary patterns and intake for one year prior to completing the FFQ 17 .
Twenty-three participants were excluded from this analysis because they had not completed the DHQ-II, possibly due to: lab absence, error in submission, or logging off the website portal before DHQ II completion. Three participants were excluded from analyses due to reported intakes outside physiological possibilities, <400 or >7000kcals 18 .

Breakfast consumption: NAS Survey
The NAS provided breakfast consumption patterns with the question; 'How many days a week do you usually eat breakfast?'. Breakfast consumers were defined as those eating breakfast at least three times weekly 10 .

Physical activity assessment: IPAQ Short-Version
The IPAQ assesses physical activity level (PAL) in the students; this instrument was selected as it was previously used in a cross sectional analysis on nighttime eating 19 , and with this age group 20 .
There is high over reporting associated with the IPAQ and the IPAQ scoring guide provides protocol on data truncation and outlier exclusion which was utilized when analyzing the data 21 . First reported combined vigorous and moderate physical activity reported more than 960 minutes were excluded from physical activity analyses 21 . Next, reported vigorous and moderate physical activity minutes more than 180 minutes were truncated to equate to 180 minutes 21 .
A total of 23 participants were excluded from physical activity analyses. Per scoring guidelines 38 cases for moderate physical activity and 53 cases for vigorous physical activity were truncated to 180 minutes 21 . Fourteen participants were excluded from analyses for reporting >960 minutes, six participants did not complete the IPAQ due to lab absence, and three entered 'Choose not to answer.'

Body Composition and Metabolic Syndrome Risk Assessment: Anthropometric and Biochemical Measures
Height measures were taken after shoes and socks were removed, using a wall mounted stadiometer. Prior to weight measurements, using a voided digital scale, students removed any excess clothing, shoes, and socks. Height and weight measurements were taken twice and repeated until they were within .2cm and .1 kg of each other. These measures were used to calculate BMI to assess weight status 22,23 ; being overweight or obese is defined by a BMI over 24.9 kg/m 2 and 29.9 kg/m 2 , respectively 1-4 .
Body fat percent was privately measured using the Bod Pod; following protocol preparation guidelines (Body Composition System manufactured by Life Measurements (LMI) in California). The Bod Pod utilizes air displacement plethysmography to provide a calculated estimation of body density and total body fat percentage, which involves measuring body mass and volume, and thoracic lung volume 24 .
Height and weight measures (BMI) and BodPod measures were not completed on 57 participants; due to missed appointments. For one of the 57 participants information on BMI was not recorded on the form submitted to the instructor.
Hip and waist circumference were obtained using a Gulick measuring tape to determine abdominal obesity. Hip and waist measurements were taken in a private room with a separate check-in and waiting area. Waist circumference was measured at the slimmest part below the ribcage and above the iliac crest and two measures were taken; measures were repeated until they were within .5cm of each other 3,25 .
Fasting biochemical data included total cholesterol, low density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol (HDL-C), triacylglycerides (TAG), and blood glucose levels. These values were collected via a finger stick and assessed using the validated AlereCholestech® LDX System 26,27 . Before blood was drawn students were required to fast at least 8 to 12 hours overnight, during which time water was encouraged to ensure proper hydration 3 . Following the finger stick sample, participants rested for five-minutes before blood pressure was obtained with an automated computerized device (HEM-711DLX manufactured by Omron in Bannockburn, Illinois). The anthropometric and biochemical data collected allowed the researchers to determine MetS criteria. The clinical definition of MetS is having any three of the five criteria 2 . The criteria set forth by The American Heart Association and the National Heart, Lung, and Blood Institute include elevated blood pressure (≥ 130/85 mmHg), elevated fasting blood glucose (≥100mg/dL), elevated TAG (≥150mg/dL), increased waist circumference (males ≥40", females ≥35"), and suboptimal HDL-C levels (males ≤40mg/dL, females ≤50mg/dL) 2,3,28 .

Statistical Analysis
Statistical analyses were performed using SPSS (v24). Skewness and kurtosis were used to assess normality of all variables' distribution. Correlations were run for all variables to determine any covariate influence. Previously identified covariates from the reviewed literature included age, sex, physical activity, ethnicity, current major, and smoking status 7,9,10,19,[29][30][31] .
The primary hypothesis was assessed between groups, nighttime eaters and non-nighttime eaters; using an analysis of variance (ANOVA) followed by an analysis of covariance (ANCOVA) using co-variates previously identified. Secondary and exploratory hypotheses were assessed using the same approaches. Total caloric intake differed significantly between genders (p=.006, t=2.822), with males consuming more calories (+681.99kcals) ( Table 2). Thus, gender was used as a covariate when analyzing the primary variable using an ANCOVA. Smoking was used as a covariate when dietary quality was analyzed as previous nighttime eating and diet quality studies have used this methodology 3-6 , additionally smoking was negatively associated with diet quality in adults 7 .

Descriptive:
Population sample characteristics are displayed in Table 3 and and kurtosis demonstrated the variables were overall normally distributed; Table 5.
Correlations are reported in Table 6.

Primary outcome: Total daily caloric intake
An ANOVA showed significant difference in total daily caloric intake between nighttime and non-nighttime eaters; Table 7. Those who ate within 2 hours of sleep onset (p=.015, η 2 p=.034, r 2 =.034, F(1,170)=6.008) reported significantly higher daily caloric consumption (+476.26kcals). Even after controlling for gender and PAL (p=.034, η 2 p=.031, r 2 =.127) students who consumed calories within 2 hours of sleep onset reported a higher daily calorie intake. Both vigorous and moderate weekly activity minutes were controlled for; Table 8 depicts the ANCOVA outcomes. As noted earlier no association was found for intake after 10:00PM.

Sleep Quantity
Nighttime eaters were significantly more likely to sleep less than 7 hours.
Significance was found only when nighttime eating was categorized as caloric intake after 10:00 PM (x 2 =5.86, p=.015). A higher percentage of nighttime eaters (45.7%) reported less than seven hours of sleep each night, compared to non-nighttime eaters (27.8%). Significance was not found for low sleep quantity and any other nighttime eating categorization.

MetS
When the number of MetS risk factor criterion was analyzed in intervals (0 and 1-3 risk factors) no associations were found with any nighttime eating categorizations.

BMI and Body Fat Percent
No significant differences in group means were found between BMI and all categorizations of nighttime eating. However, nighttime eating, defined as caloric intake after 10:00PM was negatively associated with body fat percent in males (-  Table 10.

COMMENT Introduction
Research previously had not examined nighttime eating prevalence and whether an association existed between nighttime eating and increased caloric intake in college-aged students. This cross-sectional data analysis explored possible relationships between new definitions of nighttime eating and a range of caloric and metabolic outcomes. This study was of particular interest since traditional definitions of nighttime eating did not apply to this population.

Discussion
This perspective study found that nighttime eaters consumed more daily calories. Nighttime eating has been consistently found to be associated with increased total daily caloric intake in adults and shift-workers 7,9,10,30,31,34 . Nighttime eaters consumed an average of 475-550kcals more than non-nighttime eaters and had lower HEI-2010 total scores, supporting our primary and secondary hypotheses.
As researchers continued to use this cutoff in literature published after 7,9,10,14 . Later categorizations of nighttime eating may be attributed to later bed timings, as the most frequently reported (34.10%) usual bedtime was midnight. The later bed timings for college students may be due to range of studying and social events 33 .
This cross-sectional analysis contributes a more accurate definition of nighttime eating for college-aged students to the body of literature, with the intention to be used in further analysis of eating patterns for this population.
Nighttime eaters consuming calories within 2 hours of sleep reported significantly higher daily caloric consumption (+476.26kcals) in our sample. This is consistent with previous research examining total caloric intake and nighttime eating in the middle-and older-aged population and shift workers [6][7][8][9][10]15,19,30,31 , with the exception of the behavior occurring closer to sleep onset in our sample. Future studies examining associations between nighttime eating and college students should include 24-hour dietary recalls to assess the proportion of calories consumed during the nighttime period and quality of foods consumed to determine if the proportion of calories consumed is similar to findings in adults [6][7][8][9][10]15,19,30,31 .
This study assessed dietary quality based on HEI-2010 total score. Previously, relationships between diet quality and nighttime eating were examined using weekly amounts of fast food meals and sugar sweetened beverages as the measure of diet quality 7 . This study also found nighttime eating to be associated with a poorer diet quality 7 . Relationships between breakfast intake and a higher diet quality have been Our study also demonstrated that after controlling for covariates, gender, physical activity, and smoking status, relationships between nighttime eating (caloric intake within 2 hours) and total caloric intake remained significant.
To our knowledge there is no previous literature examining MetS risk and nighttime eating in any aged population. In this sample, there was no association found between having zero or one to three MetS criteria and any nighttime eating categorization. A separate study examining MetS and college students found a low proportion of MetS in men (9.9-12%) and women (3-6%) 2,3 . Additionally, 77% of the college-aged students had one criterion of MetS. In our sample 30.05% had one criterion of MetS 3 . The lower proportion in this study may be due to the fact our sample was only from one university and the other two studies examined three universities over a geographical span 2,3 . The lower proportion of MetS criterion may also be related to this sample being relatively healthy (BMI status and body fat percent), and being younger in age. Over time elevated caloric intake, may lead to development of more risk factors and/or accumulation of body weight 7,9,[36][37][38] , as demonstrated with nighttime eaters in this sample and others 7,9,10,19,31,32 .
Limited literature, with contradicting results, has explored the possible relationships between nighttime eating and BMI 10,19 . This study found no associations between BMI and nighttime eating, but when body composition was examined using body fat percent a negative association was found for nighttime eating after 10:00PM.
Similarly, other studies have found no correlations with meal timing and BMI 10 , this may be related to the lower BMI of the sample in previous studies 10,37, and in this analysis, with 19.6% being either overweight or obese. Even though no relationship with BMI was found, nighttime eating after 10:00PM was associated with a lower body fat percentage (-5.45%); over time higher caloric intake is likely to lead to weight gain unless offset by energy expenditure 10,37 . Future studies should include questions to assess timing of exercise and whether the participant is a student athlete.
Additionally, determining if nighttime eating is a new behavior adapted in college from adjusting to studying, social events, or psychological reasons 33 , or if the eating behavior was adapted over time may explain the negative association between body fat percentage and nighttime eating.
Nighttime eating and sleep duration are correlated and may be related to the mechanism for increased weight gain from increased caloric intake due to habitual nighttime eating 3 . This may be because decreased sleep duration is associated with hormonal disturbances; decreased leptin and increased ghrelin 14 . These hormonal changes may impact one's weight status, as ghrelin increases appetite and leptin reduces appetite 14 . Short sleep duration, linked to nighttime eating, was correlated with emotional disturbances such as stress and pessimism, which may decrease one's ability to adhere to dietary and physical activity regiments and thus impact one's weight status 3,14 .
An increase in total caloric intake related to nighttime calorie intake may be due to the decreased satiety associated with late night food consumption, this is known as the satiety ratio principle 9,30,31 . The satiety ratio is a decrease in satiety efficiency throughout the day 30 . The satiety ratio principle was demonstrated by de Castro, where an increase in time spent eating and meal size occurred over the progression of the day 30 . Additionally, the after-meal interval decreased, even though more calories had been consumed in meals 30 . This study showed that as the day progresses, individuals spent more time eating food and less time in between meals, even though more calories were consumed at those later meals 30 . This demonstrates that one's satiety level is not as greatly impacted with later meals compared to meals consumed earlier in the day 30 .

Limitations:
This was a cross-sectional data analysis therefore causation cannot be inferred from the results. Future research should utilize longitudinal trials. The sample used was a convenience sample, allowing us to examine nighttime eating in a the college population for the first time. Since the sample was primarily female, future studies should recruit a more diverse and representative college student population sample.
The independent variable, nighttime eating, was assessed by asking proximity of consumption to sleep onset and relative to clock-time. These questions have not been validated in this study or in any other study of nighttime eating. This is a limitation within the research as there is no validated measure to assess nighttime, eating apart from a questionnaire that serves as a psychiatric diagnostic tool for nighttime eating syndrome 8 . Thus, this is an area of future work.
Reliability of self-reported dietary data was a limitation to this study. However, since the nutrition course material includes portion estimation lessons, this population had more background to estimate portion sizes than a lay person. The DHQ II was not administered until the 10 th week of the academic semester, therefore participants had 10 weeks of exposure to general nutrition information at a college level, and short term portion size lessons have been shown to increase the accuracy of college-aged students' portion size estimates 39 . These students elected to enroll in a nutrition course and may have been a biased sample. The lower proportion of MetS criterion may also be related to this sample being relatively healthy (BMI status and body fat percent), and younger in age; potential pitfalls are discussed in Appendices L through M.

Strengths
The comparable nighttime eating studies primarily examine the middle-aged population and shift workers 7,9,10,19,31 , whereas this study examined the college-aged population. This is the first known study to explore the prevalence of nighttime eating in this population, and to explore relationships between MetS and nighttime eating.
Additionally, HEI-2010 total score was used as our assessment of diet quality rather than non-validated measures that have been used 40,41 . Further measures of body fat percent, BMI, and blood objectively reflected health status.
This is problematic as obesity earlier in life has been found to increase the risk for chronic diseases in adulthood, including cancers, diabetes, cardiovascular disease, mobility limitations, and arthritis 21,22,35 . College students have identified many barriers to healthful weight control, including nighttime eating, social stimulations such as going out to eat, alcohol consumption, and peer influences that hinder healthful behaviors.
Environmental barriers were also identified, including over availability of food in dining halls, large portion sizes offered, fast food accessibility, emotional eating, and lack of time and healthful food options 8,19,26,28 .
College-aged students are of interest due to their overall declining health profile; increased obesity rates in all gender and racial/ethnic groups, decreased fruit and vegetable intake, and higher levels of sedentary behavior 20,39 . In young adults (18-25 years old), obesity is overshadowed by health risks associated with sexual violence or substance abuse and the fact that most college-aged students do not view themselves at risk for chronic disease later in life 26 . Longitudinal studies have found college students can gain an average of 4.3kg over the traditional four-year length of college and thus are at risk for weight gain 21 .

Nighttime eating:
The literature based definition of nighttime eating is defined as any caloric intake after 8:00PM or within four hours of sleep onset, 8:00PM has been defined as the average time of maximum caloric intake for persons without an eating disorder 3,32,33 . Multiple mechanisms may explain nighttime calorie intake as a predictor for weight gain from increased total caloric intake 9,10,13,15,23,32,33,38 . In middle aged shift workers, nighttime calorie intake has shown to decrease total energy expenditure, with a reduction in resting metabolic rate (RMR) as the possible mechanism (15). Having a lower relative RMR may be a predictor to an elevated BMI or increase the risk of future weight gain 24 . During the evening fat oxidation is reduced, this may be another mechanism related to increased weight gain with nighttime eating 24 .
Nighttime eating and sleep duration are correlated 3 . Decreased sleep duration is associated with hormonal disturbances, including decreased leptin and increased ghrelin 14 . These hormonal changes may impact one's weight status, as ghrelin increases appetite and leptin reduces appetite 14 . Short sleep duration, linked to nighttime eating, was found to be correlated with emotional disturbances such as stress and pessimism 3,14 . These disturbances may decrease one's ability to adhere to dietary and physical activity regiments and thus impact one's weight status 3,14 .
An increase in total caloric intake related to nighttime calorie intake may be due to the decreased satiety of late night food consumption, which is known as the satiety ratio principle 9,10,32 . The satiety ratio principle is described as a decrease in satiety efficiency throughout the day 10 . The satiety ratio principle was demonstrated by de Castro, where an increase in meal time and size occurred over the progression of the day 10 . Additionally, the after-meal interval decreased, even though more calories had been consumed in the meals. This study showed that as the day progresses individuals spent more time eating food and less time in between meals, even though more calories were consumed at those later meals. This demonstrates that one's satiety level is not as greatly impacted with later meals compared to meals consumed earlier in the day 10 . It is also theorized that higher total caloric intake related to nighttime eating may have an evolutionary basis, in which there are increased opportunities to eat into the night hours due to technological advances 9,10 .

Nighttime eating and total calorie intake:
Nighttime eating has been consistently found to be associated with increased total daily caloric intake 3,9,10,25,32,33 . Over time, higher caloric intake related to nighttime eating was positively associated with weight gain 2,17 . The evidence for a higher BMI is variable amongst studies, with most finding a positive association 3,32,38 , with the exception of one 33 . All researchers had examined an adult population, with a mean age of 30; none targeted the collegeaged population 3,9,10,25,32,33 .
Wang et al. controlled for confounding variables (gender, age, total calorie intake, education level, body size, physical activity, and over/underreporting) and examined nighttime eating by using the proportion of calories consumed at meal times to analyze total caloric intake 38 . In the study a web based dietary recall system was used to assess total caloric intake and timing of intake 38 . It was found that those who consumed more than 33% of their total caloric intake in the evening were twice as likely to be obese 38 . It was not investigated whether the nighttime eaters (>33% total caloric intake between 5:00PM-12:00AM) had higher total daily caloric intakes, but it was found that participants consuming a higher proportion of their calories during the evening were more likely to be obese 38 .
Wang et al. was not the first to investigate nighttime eating in the middle-aged population using proportions of energy intake 38 , de Castro did as well 9,10 . In de Castro's studies, it was found that caloric intake was highest between the hours of 6:00PM-10:00PM after removing calories consumed from beverages 10 . These results were then replicated by de Castro and a positive association was found with and without removal of beverage calories 9 .
Additionally, the subjects who engaged in nighttime eating had a higher total caloric intake compared to those who did not eat at nighttime 9,10 .

Nighttime eating and BMI:
Repeated positive associations between nighttime eating and BMI have been found 3,32,38 , additionally nighttime eaters have been found to gain more weight over time than non-nighttime eaters 2,17 17,33 . Additionally, even though no relationship was found over time higher caloric intake may lead to weight gain 17,33 .
Other studies have previously found an association between nighttime eating, total calorie intake and BMI when controlling for sleep duration 3,32 .
Wang et al. found significance between total caloric intake and BMI, with the reliable reporters identified by doubly labeled water 38 . As weight gain is a concern for the college-aged population, and nighttime eating has been identified by students as a barrier to weight management this was a novel association to investigate in this age group.

Nighttime eating and diet quality:
Investigations into a relationship between nighttime eating and diet quality are limited and contradicting 3,38 . Diet quality is the total measure of high quality eating patterns in a diet, such as higher fruit and vegetable and lower processed food intake 3,36 . Total caloric intake has primarily been examined in nighttime eating studies, while research on diet quality has been overlooked 9,10 .Wang et al. found no association in the percentage of calories from fat, protein, or carbohydrate, grams of fiber, and servings of fruits or vegetables per a day between participants who consumed <30% of their total calories in the evening (or any meal period) and those who consumed >30% of kcals in evening 38 . Reid and Baron found that middle aged late sleepers had later meal timings for breakfast, lunch, dinner, and last meal, and poorer diet quality, which was defined as increased intake of sugar sweetened beverages and fast food 3 .
Further investigations into possible relationships between nighttime eating and diet quality are needed. An association may exist for a later defined meal eating time, as Reid and Baron examined nighttime eating in the parameters of caloric consumption after 8:00PM or within four hours before sleep, whereas Wang used between 5:00PM-12:00AM as cutoff points 3,32,33,38 .
Breakfast consumption is often coupled with having a robust diet quality and is a preventative factor for weight gain 37 . Breakfast skipping is associated with weight gain, poor diet quality, and shifting caloric intake to later in the day 4,15,38 . Baron et al. looked

Nighttime eating and sleep quality:
It has been continuously found that adolescent late sleepers, compared to non-late sleepers, had a poorer diet quality 11,16,18,34  Late sleepers had a higher caloric intake after 8:00PM compared to those who did not sleep as late, but not higher total daily caloric intake 3,32 . There were correlations found between higher caloric intake after 8:00PM and a higher total daily caloric intake, shorter sleep duration, and later sleep timing 3 .
These interactions between nighttime eating, higher total caloric intake and problematic sleep patterns in middle aged individuals, highlighted the importance to investigate this association in college-aged students.
Reid and Baron have found such novel associations related to sleep patterns that they examined nighttime eating based on closeness of consumption to sleep onset and total daily caloric intake, rather than using meal times 33 .
Sleep timing was measured by utilizing wrist actigraphy data 33 . Eating closer to sleep onset was found to be associated with higher total daily caloric intake 3,33 , this strengthens their previous findings that nighttime consumption and higher total caloric intake are associated in middle aged adults. Unfortunately, physical activity was not controlled for in either study, which impacts sleep patterns and calorie intake 3,33 . When examining adolescents (9 to 16-year-olds), Golley found correlations between sleep timing, diet quality, and BMI z-score 18 . These correlations were found with or without controlling for physical activity, energy intake, and sleep duration 18 .

Nighttime eating and MetS:
Due to high rates of overweight/obesity and MetS the college-aged population is at risk for developing chronic diseases in adulthood, including diabetes mellitus and cardiovascular risk 27
Alter definition to only capture those consuming calories within two or one hour(s) of sleep onset, or examine the cutoff point as 9:00, 10:00PM, or later as needed, for nighttime eating categorization.
Entire sample categorized as nonnighttime eaters, defined as no caloric intake within 4 hours of sleep onset or after 8:00PM.
Alter definition to capture those consuming calories within five or six hours of sleep onset, or examine the cutoff point as 8:00, 7:00PM, or earlier as needed, for nighttime eating categorization.
Accuracy of self-reported FFQ, the DHQ 38 .
The nutrition course material includes portion estimation lessons, this population enrolled in this general nutrition course has more background to estimate portion sizes than a lay person. The DHQ is not administered until the 10 th week of the academic semester, therefore they have had 10 weeks of exposure to general nutrition information at a college level, and short term portion size lessons have been shown to increase the accuracy of college-aged students' portion size estimates 7 . Measured variables (biochemical and body fat percentage) would be more heavily weighted if dietary data is deemed unreliable.
Accuracy of human measured height and weight to calculate BMI to assess body composition.
Height and weight measures are obtained by trained undergraduate teaching assistants under the supervision of a graduate teaching assistant. Body fat percentage was also measured using the BodPod and can be used as an alternative to BMI 29,31 .
Total sample are breakfast consumers, defined as more than or equal to 3 days of breakfast intake 33 .
Categorize breakfast consumers as those consuming breakfast as those consuming at least one day of breakfast intake Total sample are breakfast skippers, defined as less than or equal to 3 days of breakfast intake Categorize breakfast skippers as those consuming breakfast on 0 days 4 .