ASSOCIATIONS OF WEIGHT DISSATISFACTION ON DIET QUALITY, PERCENT BODY FAT, AND PHYSICAL ACTIVITY IN COLLEGE STUDENTS

.................................................................................................................. ii ACKNOWLEDGMENTS ............................................................................................. v PREFACE ..................................................................................................................... vi TABLE OF CONTENTS ............................................................................................ viii LIST OF TABLES AND FIGURE ............................................................................... ix THESIS: Associations of Weight Dissatisfaction on Diet Quality, Percent Body Fat, & Physical Activity in College Students Publication Status .................................................................................................. 1 Abstract ................................................................................................................. 2 Introduction.................................................................................4 Methods......................................................................................6 Results......................................................................................12 Discussion.................................................................................15 Future Implications and Conclusion...................................................21 Literature Cited...........................................................................23 Tables and Figure.........................................................................28


Introduction
Obesity rates are increasing nationwide in all age groups, with 42.8% of the U.S. population reported with overweight or obesity in 2017-2018. [1][2][3][4] As overweight and obesity rate continues to rise in the U.S., the social pressure to fit an impractical ideal body weight and image influences the satisfaction college-aged students have of themselves and can lead to higher body weight dissatisfaction (BWD). 5 The term BWD is defined as the absolute value of the difference between reported body weight in pounds and reported desired body weight, and has been shown to vary by sex, socioeconomic status, and race and ethnicity. [5][6][7][8][9][10][11][12][13][14][15][16] Previous studies has found that higher BWD is associated with negative health behaviors in college-aged students related to diet, such as excessive dieting and lower intake of nutrient-dense foods such as whole grains, fruits, and vegetables. 9,17 However, limited research has examined the association between BWD and overall dietary quality (DQ) of college-aged students, as measured by the 2015 Healthy Eating Index (2015 HEI), which aids in the evaluation and monitoring of particular dietary components to better understand dietary patterns in individuals. 18 There is minimal research on the association between BWD and DQ. Previous research has found those with lower BWD tend to consume more fruits and vegetables, compared to those with higher BWD. 7,17,19,20 However, the data were not evaluated through DQ indices that aim to evaluate the overall diet and assess dietary patterns of an individual. 7,17,19,20 One study analyzed BWD and DQ separately in a group of female university students. 21 The results indicated more than half of the sample (57.4%) had BWD, with a total DQ score of 38.5±6.7 measured by the Diet Quality Index. 21 Although this study found independent results for BWD and DQ in females, this study did not compare BWD with DQ, and therefore does not address the associations between the absolute value of BWD and DQ.
Higher BWD has also been associated with body mass index (BMI) and percent body fat (%BF). Research consistently shows a positive relationship between BMI, %BF, and BWD; higher BMI value and %BF, higher BWD. 10,[22][23][24] In general, BWD is influenced by BMI with a majority of individuals with overweight or obesity presenting with higher BWD. 7 However, males tend to be more satisfied with weight regardless of overweight status. 7 Although this relationship exists, there are differences observed by sex. Females tend to express greater BWD than males with a greater desire for a lower body weight and overall thinness. 7,10,11,13,14,19 Additionally, females of normal-weight status tend to express higher levels of BWD, regardless of normal BMI and %BF. 8,23 In contrast, males tend to express higher BWD with a greater desire for higher body weight with an increase in muscularity. 10,11 Much of the literature on the association of BWD to %BF has measured %BF through hydrostatic weighing or skinfold calipers. 17,23 However minimal research has examined %BF through air displacement plethysmography (Bod Pod) or multi-frequency bioelectrical impedance analysis (InBody 770).
Lastly, higher BWD has been associated with excessive or avoidance of physical activity (PA). Those with lower BWD tend to participate in more regular PA with higher levels of walking/jogging per week and higher cardiorespiratory fitness compared to those with higher BWD. 7,17,25,26 However, some literature suggests that more active individuals have higher BWD than inactive individuals, possibly due to their desire to change weight status. 19 Although literature has examined the association of BWD and PA, there is a lack of research in examining minutes of weekly moderateto-vigorous physical activity (MVPA). Since these research gaps remain within the BWD literature, this current cross-sectional, secondary data analysis was conducted to examine the associations BWD has with certain health behaviors in the population of college-aged students. This research will aid in better understanding the needs of the population and will assist in the formulation of future recommendations.
The purpose of the current study is to observe associations between BWD and total 2015 HEI score. The primary hypothesis is that participants with a lower BWD will have a higher total HEI score than participants with a higher BWD (n=434). The secondary hypothesis is that participants with a lower BWD will have a lower %BF than participants with higher BWD. The tertiary hypothesis is that participants with a lower BWD will report more minutes of weekly MVPA than participants with higher BWD.

Study Design
This is a cross-sectional, secondary data analysis using data from the Nutrition Assessment Study (NAS), an ongoing International Research Board approved study at the University of Rhode Island. This study aims to examine nutrition assessment data for research to increase the comprehension of relationships between diet and disease risk in college students in an Applied General Nutrition course (NFS 210). This study involves gathering anthropometrics, PA, dietary data, and blood values through assessments from semesters in Fall 2015 through Fall 2019 that are required as part of their coursework. To participate, potential participants were required to meet the following criteria: aged 18-24 years old, and enrollment in NFS 210 lab and course.
Four-hundred thirty-four consenting participants were utilized for data analyses.

Research Participants
A defined sample was utilized for data analyses from Fall 2015 to Fall 2019.
Students were excluded if they were <18 or > 24 years of age, pregnant, or had reported energy intakes of <400 and >7,000 kcal/day. This age group was selected to be consistent with other research conducted in college student populations and due to the lack of literature that addresses this age group in particular for BWD. 22,[27][28][29][30]

Data Collection
All data collection for this study occurred during the course lab sessions throughout the semester (Lab 2, 7, 9, and 10). Protocol guidelines were in place for all assessments including anthropometrics, blood values and survey data within the NAS Manual. During the first lab session, students were provided with verbal and written information about the research study, which was described in detail in the informed consent form (Appendix C). The NAS survey, known as the demographics survey, was administered during lab 2 and completed within one lab session. The IPAQ shortform was also administered in lab 2 and was to be completed within one lab session, which took approximately 10 minutes for the students to complete. The Diet History Questionnaire II (DHQ II) was administered during two lab sessions. Part 1 of the DHQ II was completed during lab 9 and part two was completed by lab 10. The anthropometric measurements were completed during lab 7 and included height weight, BMI, waist and hip circumference, and %BF using the air displacement plethysmography (Bod Pod) and multifrequency bioelectrical impedance analysis (InBody 770). 31 Anthropometric measures were collected by a trained member of the study. Additional assessments include blood values collected in lab 5 using Alere Cholestech® LDX System.

Demographic Data
The independent variable, body weight dissatisfaction (BWD), was evaluated by utilizing the NAS survey. The NAS survey is an electronic survey includes 26 questions varying in number of items per response with response formats including multiple choice, open-ended, and Likert scale. The overall NAS survey has not been validated but has been utilized in previous research as a tool to gather demographic data. This survey gathered pertinent information to help differentiate between the students' actual measured weight in lab 7 versus their reported desired body weight in lab 2 to categorize students as those lower or higher BWD. The question used for this differentiation was, "What would you like to weigh in pounds," with an open-ended response category. The NAS survey gathered pertinent information on demographics as potential covariates. These included multiple-choice questions on age, sex, race/ethnicity, current major, and year in school.

Dietary Quality Measures
The dietary intake was collected utilizing data from the DHQ II (Appendix F) and was defined as total HEI score utilizing the 2015 HEI. The DHQ II is the food frequency questionnaire (FFQ) that provides an estimation of total daily caloric intake and evaluated DQ by utilizing the 2015 HEI. 32,33 The 2015 HEI is a DQ index that measures the alignment with the 2015-2020 Dietary Guidelines for Americans (DGA).
The DHQ II was designed and tested by the National Cancer Institute. 32 The DHQ has been validated as a superior FFQ compared to the Block and Willett FFQs for estimating absolute intakes in participants 20-70 years of age. 34 The 2015 HEI score was derived from the DHQ II, an FFQ that includes questions on 134 food items and eight dietary supplements. 32 The DHQ II questioned the participant about food items and portion sizes that were consumed within the past year. 32,35 The 2015 HEI is an index ranging from zero to one-hundred, which is based on thirteen individual components with scores per item from zero to ten with nine adequacy components: total fruits, whole fruits, total vegetables, greens and beans, whole grains, milk/dairy, total protein foods, seafood and plant proteins, and fatty acids. It also includes four moderation components: refined grains, sodium, added sugars, and saturated fat. 33

See
Appendix I for 2015 HEI scoring guide. The 2015 HEI is updated every five years to reflect current federal dietary advice through a collaboration between the National Cancer Institute, and the US Department of Agriculture Center for Nutrition Policy and Promotion. 33 The output scores were calculated through the HEI-2015 algorithm within SAS software (SAS Institute Inc. version 9.4).

Anthropometric Measures
All measures for anthropometrics were taken according to standardized procedures. 36 Height was assessed using a wall-mounted stadiometer (SECA 240, Hamburg, Germany) and rounded to 0.1 cm, and weight was assessed using a digital scale (SECA 700, Hamburg, Germany) and rounded to 0.1 kg. 36 37 This device indirectly measures the volume of air displaced inside the chamber, "subtracting the volume of air remaining inside the chamber when the subject is inside to the volume of air when the chamber is empty." 37 The InBody 770 is a multi-frequency bioelectrical impedance analysis device that measures the body's resistance to flow of alternating electrical current at a designated frequency. 31 It has been found that the Bod Pod and InBody 770 are valid and reliable measures of body composition in relation to DEXA and to each other. 31,38,39 Physical Activity Assessment: IPAQ Short-Form PA was assessed using the IPAQ. The IPAQ is an electronic, seven item selfreport instrument with response format of open-ended questions. 40 The IPAQ is a selfadministered instrument that requires participants to report the frequency and duration of vigorous, moderate, and walking activities (10 minutes at minimum during the last seven days). 40 Weekly time spent in vigorous activity, moderate activity, and walking was determined by multiplying reported frequency and duration within each category of activity. This variable was calculated as minutes of weekly MVPA. 40

Statistical Analysis
The statistical analysis package SPSS (IBM version 26.0 SPSS Inc.) was used to perform statistical analyses. Skewness and kurtosis revealed data were non-normal when outliers were included. Outliers greater than three standard deviations from the mean were identified and excluded for this reason. 41 After exclusion of outliers, skewness and kurtosis were within normal ranges. A median split of BWD was used to categorize the independent variable into lower and higher BWD for the whole sample.
Likewise, the median split was also stratified by sex. To assess between group differences, independent t-tests were conducted for demographic data for the whole sample and stratified by sex. To assess statistical differences between lower and higher BWD, one-way analysis of variance (ANOVA) was conducted for the following main outcomes: mean total HEI score, %BF, and minutes of weekly MVPA. One-way ANOVAs were run to determine effect size and post-hoc power analysis for the main outcomes. Effect sizes are defined as small (0.01), medium (0.06), and large (0.14). 41 An additional one-way ANCOVA was run adjusting for energy for the primary objective. Both Bod Pod and InBody 770 are utilized similarly for their measurement of %BF, however, differ in methodology. For this reason, the two systems were combined for analysis and showed no statistically significant difference between the two systems (p=0.75). Likewise, previous literature are consistent with this finding showing relative agreement between Bod Pod and InBody, differing by less than 0.2%. 39 For these reasons, the two measurements were combined for the analysis of %BF. Pearson correlations were run with absolute value of BWD for 1) total HEI and component scores, 2) dietary components including total fat and dietary fiber in grams, 3) %BF and BMI. Additional Pearson correlations were run between %BF and BMI. Acceptance of significance was identified as p<0.05.

Subject Characteristics
Of the consenting participants (n=671), 237 were excluded based on the following criteria: non-consented students (n=170), age <18 (n=4), age >24 (n=30), intake <400 kilocalories (n=5), intake >7,000 kilocalories (n=3), missing data (n=172), and participation in multiple semesters (n=9). In addition, subjects were defined as outliers if BWD was greater than three standard deviations from the mean and these subjects were also excluded (n=14). 41 Four-hundred thirty-four participants were retained for the final sample. It is important to note that final sample size for the tertiary variable (n=307) minutes of weekly MVPA differs from primary and secondary sample sizes due to exclusion of participants with missing data from IPAQ (n=127). See Figure 1 for the flowchart on recruitment and retention of participants.
Mean subject characteristics for the whole sample are presented in Table 1.
Participants, aged 18-24, were assigned to lower BWD (n=217) or higher BWD (n=217) by median split. As shown in Table 1, participants were predominantly female (78.6%), Caucasian (83.8%), within their first year of college (60%), and with a mean age of 18.9 years. BMI for both lower BWD (21.9±3.2 kg/m 2 ) and higher BWD (24.6±3.1 kg/m 2 ) were within the normal range. Independent t-tests revealed significant differences between groups of lower and higher BWD for BMI (p=.001) and sex (p=.047) in the whole sample.
All models were conducted for the whole sample without stratification, and stratified by sex. Final results are presented as stratified by sex, and were analyzed through one-way ANOVAs. See Appendix J for additional results on whole sample data. As shown in Table 2, males (n=93) and females (n=341) were assigned to lower BWD and higher BWD by median split. See Table 5 in Appendix J for median split   criteria by whole sample and sex. Males had a mean age of 19 years, were predominantly Caucasian (76.3%) with a normal average BMI (24.4 kg/m 2 ). Females had a mean age of 18 years, were predominantly Caucasian (85.3%), with a normal average BMI (23.0 kg/m 2 ). Independent t-tests revealed no significant differences in BMI between lower and higher BWD in males. Significant differences were found in BMI between females with lower and higher BWD (p=.001), with lower BMI (21.5±2.9 kg/m 2 ) in the lower BWD compared to higher BMI (24.4±2.6 kg/m 2 ) present in the higher BWD group.

Total 2015 HEI score
A one-way ANOVA was conducted to determine if participants with a lower BWD have a higher total HEI score than participants with a higher BWD. As shown in Table 3, when stratified by sex, the hypothesis was not supported. The one-way ANOVA demonstrates a trend towards significance in males with small to moderate effect size for total HEI score (F=3.223, ηp 2 =.037, p=.076), suggesting a slightly higher total HEI score in male participants with lower BWD (64.3±10.3) compared to those males with higher BWD (60.2±10.8). There were no significant between group differences for total HEI score for females (F=0.161, ηp 2 =.001, p=.689). Even after adjusting for caloric intake, there was no statistical difference in males (p=0.088) or females (p=0.654).
Independent t-tests were run for all 2015 HEI adequacy and moderation components: total fruits, whole fruits, total vegetables, greens and beans, whole grains, total dairy, total protein, seafood and plant proteins, fatty acids, refined grains, added sugars, saturated fats, and sodium. In Table 4, results demonstrate significant between group differences for 2015 HEI components in males. Significance was yielded in adequacy components for total vegetables (t(85)=2.827, p=.006), greens and beans (t(85)=2.753, p=.007), and seafood and plant proteins (t(85)=2.209, p=.030) in males.
The results indicate males with lower BWD have a higher intake of total vegetables, greens and beans, and seafood and plant proteins compared to those with higher BWD.
No between-group differences were shown for 2015 HEI components in females.
Although results for adequacy components were significant in males, there is a chance of Type 1 error due to multiple comparisons increasing the likelihood of spurious results. 42

Percent Body Fat
In Table 3, the hypothesis that participants with a lower BWD will have a lower %BF than participants with a higher BWD was supported for females. A oneway ANOVA demonstrates no statistical significance for between group differences in %BF for males (F=.000, ηp 2 =.000, p=.988). There were between group differences in females (F=75.380, ηp 2 =.185, p=.001); females with lower BWD have a lower %BF (24.8±5.8%) compared to females with a higher BWD (30.3±5.8). Pearson correlations detected a significant, moderate correlation between %BF and BMI (p=.001, r 2 =.418) measured by BodPod or InBody 770. A one-way MANOVA reveals that there were no between subject effects in males for %BF (p=0.809) and total HEI score (p=0.137). However, the one-way MANOVA does reveal significant between subject effects in females for %BF (p=0.001), however no difference is revealed for total HEI score (p=0.744).

Minutes of Weekly Moderate-to-Vigorous Physical Activity
Within Table 3, a one-way ANOVA found no significant between group differences for minutes of weekly MVPA for males (F=.242, ηp 2 =.003, p=.625) or females (F=.453, ηp 2 =.002, p=.501). The hypothesis that participants with a lower BWD will report more minutes of weekly MVPA on the IPAQ than participants with higher BWD was not supported.

Discussion
This study demonstrated that higher BWD in males is associated with lower HEI adequacy components (greens and beans, total vegetables, seafood and plant proteins), whereas higher BWD in females is associated with higher %BF. The primary objective of this cross-sectional, secondary data analysis was to examine the association between BWD and total 2015 Healthy Eating Index (HEI) score. The secondary objective was to determine the association between BWD and %BF utilizing the InBody 770 or BodPod. The tertiary objective was to evaluate the association between BWD and minutes of weekly MVPA. in college-aged students.
BWD, the absolute value of the difference between reported body weight in pounds and reported desired body weight, was defined using a median split of to categorize the independent variable into lower and higher BWD. [5][6][7][8][9][10][11][12][13][14][15][16] Additionally, the median split was stratified by sex due to differences commonly seen by sex. Generally, the results for the primary and tertiary variables indicate BWD in males and females are not associated with total HEI score and minutes of weekly MVPA. However, the secondary variable, %BF, does support previous findings suggesting that those females with lower BWD will have a lower %BF compared to those with higher BWD. However, these findings are not seen in males. This is the first study to examine the association of absolute value of BWD as higher and lower values and total 2015 HEI score. We found no association between BWD and total HEI score in males or females. Although not statistically significant, males were trending towards significance with a small to moderate effect size and moderate power indicating that significant results may be possible with a moderate sample size. Females showed no statistical significance with small effect size and low power suggesting that even with a larger sample size there would still be no significant difference between groups. While the primary hypothesis has been rejected, the overall total HEI score in males (64.3±10.3 and 60.2±10.8) and females (65.4±10.9 and 64.9±11.2) is consistent with scores obtained by previous literature. Similar results were found in a cross-sectional survey of college students, 43 with diet intake gathered by the DHQ I and evaluated by 2015 HEI. Unlike Reedy et al. 33 and Amaral et al., 21 the present study shows a higher total HEI score for college-aged students than what has been previously found. One reason the total 2015 HEI score may be elevated compared to previous literature is because the respondents were enrolled in a nutrition course with a lab session. This could have increased the respondents' interest in what was being consumed. Likewise, a majority of respondents are health science or nutrition/dietetics majors. For this reason, these respondents are possibly more aware of their total caloric intake and therefore have a higher total HEI score.
BWD was found to be associated with adequacy components from the 2015 HEI among males in the present study, which is inconsistent with previous findings from Sunbul et al. 43 who found that males tend to have higher 2015 HEI component scores of total protein, while females have higher component scores of total fruits, total vegetables, whole grains, and greens and beans. 43  That study only analyzed female participants, therefore a comparison with HEI components in males was unattainable. A possible reason for significance in three adequacy components for males could be the higher caloric intake. However, even after adjusting for energy for total 2015 HEI score, there was no statistical significance between BWD groups in males (p=0.088) or females (p=0.654). See Table 11 for results on BWD and dietary HEI components. Although results for adequacy components were significant in males, there is a chance of Type 1 error due to multiple comparisons. 42 The significance level of 5% for alpha is set for single comparisons between groups. However, since the groups were compared multiple times, the probability of finding significance increases the possibility of spurious results. 42 The results obtained for BWD and %BF in the present study suggest that females with lower BWD have lower %BF compared to those with higher BWD. However, no association is shown in males for BWD and %BF given that the mean values were identical for both lower and higher BWD at 17.1±7.7 and 17.1±10.1 respectively. This suggests that the males with lower or higher BWD did not vary by %BF. It is important to note that %BF in males has a small effect size suggesting that even with a larger sample size, no significant would be obtained. The findings for females are similar to findings from Arroyo et al,. 23 who examined the predictors of the magnitude of BWD in undergraduate females and found that higher levels of %BF were associated with greater BWD. The relationship found in this study between BWD and BMI is consistent to that of previous literature. Females with lower BWD had lower BMI compared to those with higher BWD, whereas males with lower or higher BWD had the same BMI values indicating normal-weight status. These findings are similar to that of previous studies where those of overweight status expressed higher BWD compared to underweight and normal-weight counterparts. 10,14 Although relationships between BWD and weekly physical activity time were not statistically significant in the present study, the findings may still be of importance for future evaluation. Males who have lower BWD appeared to partake in more minutes of weekly MVPA compared to those with higher BWD, whereas females with lower BWD appeared to partake in less minutes of MVPA compared to those with higher BWD. These results are similar to one study, 46 however contrast with the previous literature by Blake et al., 7 a cross-sectional study with a large mixed gender cohort of adults (n=19,003) where physical activity was assessed through a leisuretime physical activity questionnaire and separated into three categories: Inactive (no regular activity), moderate (some participation in activity like walking, jogging, running 10 miles per week), and active (walking, jogging, or running more than 10 miles per week). 7 Weight satisfaction was associated with higher engagement in physical activity, whereas weight dissatisfaction was associated with lower physical activity. 7 The contrasting results in females is not consistent with a majority of the literature, however one article 19 presents similar results. In a large population-based study (n=18,156) of Swiss adults, PA was assessed by asking participants, "In your free time, do you exercise until you sweat, at least once per week?" 19 They were split into three categories: active, partly active, and active. 19 Results showed that more active individuals report higher BWD than inactive individuals, possibly due to their desire to change weight status. 19 These results indicate that females with higher BWD may desire to change weight status, and therefore have higher minutes of weekly MVPA. It is important to note that the insignificant results could be due to overreporting on the IPAQ, which is common in self-reported physical activity compared to objectively measured physical activity. 47 Although there is no significance in MVPA, the mean values can lead to a better understanding of activity behavior in the college-aged population. Overall, males present with higher minutes of weekly MVPA compared to females, which is consistent with findings from previous literature. 40,46,[48][49][50] Although the sample includes both lower and higher BWD, on average, both males and females are meeting and exceeding the 2018 Physical Activity Guidelines for MVPA. 48 While this study does make contributions to the existing literature for BWD, some limitations should be addressed. First, causation is not able to be inferred due to the cross-sectional design. Second, there is a lack of generalizability since the majority of the sample was 18-19 years of age, female (n=341), and Caucasian. Likewise, the sample gathered is a limitation since the majority of consented students were female and from a nutrition course with a lab session at a university campus, which presents a very select sample and not representative of the university population. Therefore, we are unable to generalize it to other university students, nor other age groups or populations. Additionally, BWD was measured indirectly within the current study through measured body weight and self-reported desired body weight. For this reason, we are only able to assume dissatisfaction based on a quantitative measure, rather than qualitative where the participant is questioned about their satisfaction. 5,7,14,19 Although this is a limitation, it is still an acceptable measure for absolute value of BWD and has been utilized in previous studies. 8,10,23,51 Moreover, when gathering BWD, the time difference upon gathering desired body weight and actual body weight increases probability of error. The desired body weight was recorded in lab 2, whereas the participants actual weight was recorded six weeks later in lab 7. For this reason, the participants may have been unaware of actual body weight, and may complete the survey response with an inaccurate reported desired body weight. Furthermore, the use of two different methods for %BF measures increase risk of error. The Bod Pod is considered air displacement, whereas InBody 770 is bioelectrical impedance.
Although previous literature states relative agreement between the two methods, there is still a chance of error but differs by less than 0.2%. 39 Additionally, self-report response bias, or social desirability bias, may lead to skewed results. 52 Lastly, incomplete data for IPAQ measurements decreases the overall sample size for MVPA, and thus the power of the analysis.
While there are limitations, this study does contain significant strengths. This is the first study to analyze BWD and the associations it has with dietary quality measured by the 2015 HEI as total HEI score. Prior literature collected dietary quality (DQ), but did not present an association with BWD, nor did they utilize the 2015 HEI to assess adherence with the 2015-2020 DGA. Additionally, this is the first study to analyze associations between BWD and %BF measured through air displacement plethysmography and multifrequency bioelectrical impedance analysis, which is a newly developed instrument. Previous literature assessed body image and BWD, however, analyzed %BF through skinfold calipers, which is a less accurate measurement. Further, this study utilizes multiple surveys and tools that have been validated, such as: DHQ II, 2015 HEI, IPAQ short-form, Bod Pod, and InBody 770.
Lastly, the sample size within this cross-sectional analysis is large compared to other studies that have analyzed BWD in college-aged students.

Future Implications and Conclusions
While the current study did not yield results for associations between BWD, and total HEI score and minutes of weekly MVPA, the results contribute to the existing literature on BWD by increasing our comprehension of health-related habits in college-aged students. Future research should analyze these variables to further increase our understanding of this university population. The focus should be on a more diverse population of varying majors on university campuses or in different settings, to better understand the habits of other college-aged students outside of the health field. Likewise, longitudinal studies should be conducted on BWD since there is minimal evidence exploring BWD over time. Additionally, total HEI and HEI components should be explored further with a larger sample size for males, as well as in other majors and populations to gather a better understanding of adherence to guidelines. Furthermore, associations between BWD and %BF in females increases our comprehension of the higher BWD that is apparent in females, but not males.
Females with lower BWD have lower %BF compared to those with higher BWD.
These findings add to the current literature on absolute value of BWD and may assist in understanding certain health behaviors, such as dietary intake and body composition, in college-aged students that will support the formulation of recommendations for this population.

Body Weight Dissatisfaction
The term BWD is a quantitative measure of the discrepancy between an individual's actual weight and desired body weight, and can be interpreted as a desire to weigh more or to weigh less. 8,14 BWD can be measured directly or indirectly through various methods including, but not limited to, absolute value, relative value, or as a polar question administered through survey or questionnaire. 8,9,11,13,14,18,[23][24][25] Although these methods measure the discrepancy of satisfaction in weight, they vary in methodology and interpretation of scores. The absolute value of BWD is an indirect measure that assesses the overall magnitude of weight discrepancy that is experienced by those who desire to weigh less or more. The score of zero indicates complete body weight satisfaction. 8,9,14 The further the individual is away from the score of zero, the more dissatisfaction they contain. This method of measurement is the most common found within the literature. The relative value, or the direction of BWD, scores the satisfaction based on their desire to lose or gain weight; a positive score indicates the desire to lose weight and a negative score indicates the desire to gain weight. 13,23 Lastly, a polar, or yes-no, question can be asked of the participant to directly measure if they are satisfied with their weight. 11,18,24,25 Although a polar question of whether the individual is satisfied with their weight is a better indication of their actual satisfaction, this method limits the ability to assess the magnitude of satisfaction, and is therefore used commonly in conjunction with the absolute or relative value.
BWD is found within all age groups, sexes, weight statuses, and racial and ethnic groups. The degree of BWD is highly variable between groups, but remains consistent among various studies. For instance, females tend to express greater BWD than males, with a greater desire for a lower body weight and overall thinness. 10,[13][14][15]17,18 In contrast, males tend to express higher BWD with a greater desire for higher body weight with an increase in muscularity. 14 18 Overall, females demonstrate higher BWD compared to males.
Regardless of differences in BWD by sex, it is commonly known that BWD is highly influenced weight status as overweight or obese having higher BWD. 8,9,13,26 In a cross-sectional study, Blake et al. 13  Although high BWD is seen in those with overweight or obese weight status, high BWD has also been observed in participants of normal-weight status. 8,9,13 A study examined the extent and predictors of BWD in a sample of female volunteers in nutrition and dietetics majors who were of normal-weight status according to World Health Organization BMI range (n=62). 9 To obtain the participants desired body weight, they were asked to respond to an open-ended question, "Ideally, how much would you like to weigh?" The discrepancy was assessed with measured weight and ideal weight to obtain an absolute value of BWD. Anthropometric measures were taken such as height, weight, and fat mass using skinfold calipers. Of the female sample, 67.7% of participants chose an ideal body weight lower than their actual body weight, indicating more than half of the sample of normal-weight females expressed high BWD regardless of normal-weight status. This BWD was highly associated with lower levels of muscle mass. These results were similar to an experimental study conducted by Harris et al. 8 who utilized similar methodology in female students majoring in nutrition and exercise science, and other majors not including nutrition and exercise science (n=89). Among all three groups measured, 83% of the participants expressed BWD with a desire to weigh less. Although 90% of the nutrition students were of normal BMI, 84% expressed a desire to weigh less than their actual weight. These results indicate that despite the participants' current major, "college students may experience pressures to weigh less and 'fit the image." 8 While BWD has been examined in multiple age groups and populations, considerable emphasis is on the adolescent population. 27 This can be due in part to the participants' transition from adolescence to young adulthood, a time of drastic developmental change and independent living that influences their health-behaviors that are carried into the future. [28][29][30] Although adolescent populations are a primary population of concern for adapting future health-behaviors, it has also been examined in young-adults and the adult population. Within these populations, BWD and body image have been identified as one of the behavioral patterns that are associated with eating disorders. 9 For this reason, it has been examined along with weight perception to gain a better understanding of the extent of these concepts and their influence on particular health-behaviors. 9 While BWD, body image, and weight perception differ in definition, they are each associated with dietary and PA behaviors. 27

Body Image & Weight Perception
Body image is defined as an individual's "perceptions, feelings, and thoughts about his/her own body," whereas weight perception is the way the individual views their body weight with no regard to appearance. 27,31,32 Although these concepts are defined in different ways, they depend on various factors, such a psychological components and sociocultural influences, that can either have a positive or negative influence on health-related behaviors. 31,32 The extent and associations of body image dissatisfaction and weight perception have been identified within previous literature. 6,9,11,15,17,33 An increase in body image dissatisfaction is associated with an increase in desirability for higher muscle mass, consumption of energy-dense foods, and in participation of weight control behaviors that include skipping meals, fasting, and restricting intake of food. 9,11,17,34 Likewise, those who misperceive their body weight tend to also partake in weight control behaviors, such as skipping breakfast and eating less than desired. 6,15,33 More emphasis is placed on gathering evidence of eating disorder risk and weight control behaviors in body image and weight perception studies. However, there is a lack of evidence in assessing the associations between BWD and overall DQ within the college-aged population.

Assessment of Dietary Intake and Quality
To obtain DQ, the common intake of the participant must be gathered and Approaches to Stop Hypertension (DASH). [40][41][42][43][44] The HEI was first created in 1995 by the U.S. Department of Agriculture to determine Americans' adherence to guidelines and the food pyramid. 42 It is not only a valuable tool to assess DQ in research, but also in population monitoring, evaluation of the food environment, food assistance packages, and nutrition interventions. 45 In 2002, the AHEI was created based on the original HEI and was constructed on food and nutrient intake predictive of chronic disease risk with a higher score indicating lower risk of major disease. 43,44,46 Although this is a valid tool to use in research populations, it is more appropriate for populations with increased disease risk such as cardiovascular disease, heart failure, diabetes, etc. 43,44 Since components expressed as ratios of a food group or nutrient intake to energy intake. 47 The components were scored as the following: "0 to M, where M is 5, 10, or 20." 47 Although the 2010 HEI remained with 12 components, the scoring of the HEI changed to a total score out of 100. This score is indicative of overall DQ, as well as separate scores of adequacy and moderation to reveal a pattern of quality. 48 The 2010 HEI components were reflective of the 2010-2015 DGA, with nine adequacy components including: total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and fatty acids. 45 The remaining three were refined grains, sodium, and empty calories (energy from alcohol, added sugars, and solid fat) known as moderation components. 45 The list of components remained the same for the 2015 HEI, however, the total components and moderation components were adjusted with the newly revised 2015-2020 DGA. Since quantified limits for added sugars and saturated fats were defined in the new 2015-2020 guidelines, empty calories from the 2010 HEI moderation components were replaced with added sugars and saturated fats making a total of 13 components. 48 Another change that was made for the 2015 HEI is the allocation of legumes. Previous versions of the HEI allocated legumes as either a vegetable or a protein food component, but not both through the algorithm. 48 In the 2015 HEI, legumes are now allocated in either total vegetables, greens and beans, total proteins, or seafood and plant proteins. 48 This development in the 2015 HEI may be beneficial in gathering more accurate DQ results for those who consume mainly plant-based diets such as vegetarians and vegans. 48 Although the HEI is valid and reliable DQ measure, it does possess some marked limitations. For instance, there are multiple ways to arrive at the same total score since it is based on the sum of adequacy and moderation components. 49 For this reason, examining component scores to assess what particular components led to a high or low score is necessary. Second, the HEI scores are truncated and are unable to capture excessive intakes which could be explored further. 49 Lastly, like many dietary intake data, it is based on self-reported behavioral variables which leaves much room for measurement error. 49 While the HEI does possess some limitations, the total score and components still remain in line with the DGA and gather pertinent dietary information for research.

Influences of Dietary Intake and Quality
The term DQ is defined as 'the consumption of a variety of food groups and nutrients that support bodily growth and maintenance of normal weight, physiological status, and PA. 50 According to the 2015-2020 DGA, DQ measured through the 2010 HEI continues to be low (mean total HEI score of 57.8) for all age groups as overweight and obesity continues to rise. 51 The DGA states that Americans continue to consume less nutrient-dense foods such as fruits and vegetables with an increase of highly processed foods. 51 Although DQ is low among all age groups, young adults have been shown to have lowest adherence to dietary guidelines compared to older counterparts. 52 In a cross-sectional study evaluating the validity and reliability of the 2015 HEI, the means across age groups were significantly different for total HEI score and dietary components. 49 Older adults (>60 years of age) presented with a higher mean total HEI score of 62.8±1.1 compared to younger adults (20-39 years of age) with a mean total score of 55.0±0.7. 49 Similarly, another study 45  showed that females had a significantly higher total HEI score (63.82±9.0) than males (61.24±8.7). The total score for females was weighted by higher intake of total fruits, total vegetables, whole grains, and greens and beans, whereas males tended to have a higher intake of total protein foods. 57 Similarly, another study 45 utilizing the 2010 HEI found men had a significantly lower mean total HEI score (49.8±0.6) than women (52.7±0.9). Like Sunbul et al., women had a higher intake of five of the 2010 HEI components including total vegetables, greens and beans, whole fruits, total fruits, and dairy. 45 Overall, females tend to exhibit higher DQ with greater intake of fruits, vegetables, dairy, and whole grains when compared to males.
As dietary intake and DQ has been shown to differ by sex, intake has also been shown to differ by BMI category within the college population. A cross-sectional study by Brunt et al. 58  were more likely to consume cheese, green leafy vegetables, and other vegetables. 58 Although these results identify associations with BMI categories and dietary intake, it does not assess the associations it has with overall DQ.
As previously stated, weight misperception, and dissatisfaction of weight and image can have a negative influence on young adults' dietary intake and DQ. 6,9,11,15,17,33,34 Similar findings have been found in previous literature in regard to associations between BWD and dietary intake. It has been identified that those who are satisfied with their weight tend to consume more fruits and vegetables, 24 compared to those who are more dissatisfied reporting more disordered eating behaviors, consuming fewer meals per day, and snacking more regardless of weight category. 10,13 Although these results have been identified, minimal research has identified associations between overall BWD and DQ measured by the DQ indices, specifically in college-aged students.
Although minimal, BWD and DQ has been examined in the adolescent population. However, it does not address their adherence to the recently updated DGA.
A 2018 cross-sectional study, conducted by Xu et al., 27  does not address the associations or differences between DQ and magnitude of BWD.
For this reason, more research must focus on the associations between magnitude of BWD and the overall DQ in college-aged students in accordance with the 2015-2020 DGA.

Body Mass Index vs. Percent Body Fat
Body composition encompasses body weight and the relative amounts of muscle, fat, bone, and other vital tissues of the body. 59 As body weight continues to rise within the U.S., the need for more accurate measures of body composition is pertinent in order to determine the individuals level of disease risk. 60 Various methods are utilized to obtain body composition and weight status in participants, whether in the clinical or research setting. BMI is a weight-to-height ratio that is commonly used in research as a predictor of weight status. 61,62 It has also been utilized in research when %BF data were unable to be assessed through validated measures. 61,62 However, BMI has been found to be less accurate in predicting health-related weight status due to its inability to differentiate between fat mass and fat-free mass. 60 70 The method utilizes the sum of 3-site or 7-site skinfolds on marked sites of the body (anterior thigh, anterior iliac crest, subscapular, chest, midaxillary, abdomen, and triceps) with use of predictive equations. 71 This measure is based on the principle that a relationship exists between measurement of subcutaneous fat and %BF. 72 However, this method of measurement has lower reliability due to high dependency on operator accuracy. 70,73 For this reason, skilled operators and multiple measures are necessary for increased accuracy. 70,73 To decrease the chance of operator error, more methods are available for obtaining accurate %BF results.
One measure that relies less on trained or skilled operators, as compared to the amount of training needed for skinfold measurements, includes the Bod Pod. This form of body composition is air displacement plethysmography and indirectly measures body density through the subtraction of the volume of air displaced by the participant in the chamber to the volume of air remaining in the empty chamber. 74 Some advantages to using this body composition measure includes quick analyzation of results, increased comfortably for the participant, is non-invasive, and is a safe measurement process. 74 Studies have assessed the reliability and validity of the Bod Pod measurement to DEXA and BIA, and have found excellent reliability with repeated measures differing by 0.2%. 69 However, another study 75 found it to not be interchangeable for those participants with morbid obesity (>40 kg/m 2 ).
Another measure of body composition commonly used in the research and clinical setting, and as a validation and reference tool, is DEXA. 60,75,76 The DEXA uses a 3-compartment model that separates body mass into bone mineral content, lean body mass, and %BF. 60 Although this is commonly utilized as a validation tool due to high accuracy, there are some disadvantages to using this technique. These disadvantages include high-cost which limits accessibility and high risk of radiation exposure. 60 For these reasons, other forms of body composition measurement such as the single or multi-frequency BIA can be utilized when the DEXA is inaccessible.
However, these devices must be validated against the gold standards (DEXA, UWW). 77,78 Since DEXA has been found to be a valid measure, it has been used in a multitude of BIA validation studies to gather accuracy of %BF and fat-free mass measurements.
BIA is an analyzer that indirectly measures %BF through "the body's resistance to flow (impedance) of alternating electrical current at a designated frequency between points of contact on the body." 60 Since fat-free mass is hydrated, the electrical current passes more easily through the tissue due to the high electrolyte content, with resistance to the electric current being inversely proportional to fat content. 78 BIA exists in methods of single frequency (hand-to-hand or foot-to foot), or multi-frequency. Each method is dependent on the tactile electrodes and frequencies that it contains, predictive equations, as well as under and over hydration of the participant. 79 Single frequency and multi-frequency BIA methods have been validated against the gold standard, DEXA, in previous literature. It remains clear that with an increase in electrodes and frequencies, there is more accuracy of the body composition analysis. The hand-to-hand or foot-to-foot, single frequency devices (e.g. Omron Body Fat Analyzer), utilize two electrodes. 64 Bipolar BIA is commonly used due to increased convenience, low cost, and less training needed to administer the test. 80 However, the results obtained are questionable due to the large variations that exist in the differences between DEXA and the single frequency devices. 80 Although the device is supposed to be representative of total body %BF, it tends to underestimate for those participants with higher overall muscle mass in the arms and higher muscle mass in one arm compared to the other. 66 Likewise, those with longer arms may have an overestimation of %BF. 66 Therefore, researchers must take these factors into consideration when using single frequency BIA.
Due to varying results in single frequency analysis, multi-frequency analysis should be utilized for increased accuracy of results. Multi-frequency BIA recognizes that the body includes five distinct cylinders rather than one (right arm, left arm, right leg, left leg, trunk), which allows for regional analysis of fat-free mass, %BF, and total body water. 60,67 Each cylinders contains a different resistivity and impedance which will alter the results for %BF and segmental water analysis. 60

Influences of Body Composition
Previous research consistently shows a positive relationship between BMI, and body image and BWD; the higher the BMI value, the higher the dissatisfaction. 9,14,32,86 A 2007 cross-sectional study examined body image and weight dissatisfaction in a sample of male and female undergraduate students (n=310) with self-reported BMI. 14 The results indicated that all overweight males and females (BMI >25 kg/m 2 ) expressed the highest BWD and body image dissatisfaction, whereas the underweight and normal-weight females expressed little BWD and body image dissatisfaction. 14 Another study 18

Physical Activity
PA is defined as any bodily movement that increases energy expenditure through muscular contraction, whereas exercise refers to "planned, structured, repetitive, and performed" movement that is a form of PA. 59 All exercise can be a form of PA, but not all PA is considered exercise. 59 In an effort to improve the health and fitness of the public, the first edition of the Physical Activity Guidelines for Americans was released in 2008 by the U.S. Department of Health and Human Services. 59,88 Since then, the second edition was released in 2018 due to emerging scientific evidence aspects of PA and fitness. 59 The guidelines provide all age groups and populations with minutes of weekly aerobic MVPA, and the number of days for muscular strengthening activity with intensity. 59,88 In a joint effort to improve overall health, these guidelines can be used in combination with the DGA to provide the public with science-based evidence on the benefits and importance of physical fitness and healthful eating. 59 The 2008 Physical Activity Guidelines for Americans recommend that adults partake in at least 150 minutes of weekly aerobic moderate-intensity activity, or 75 minutes of weekly aerobic vigorous-intensity activity. 88 Additional benefits can be obtained if moderate-intensity is increased to 300 minutes per week, and vigorousintensity to 150 minutes per week. 88 The aerobic activity can be performed in bouts of at least 10 minutes to obtain the goal of moderate-or vigorous-intensity activity. 88 The 2018 guidelines has similar recommendations: at least 150 to 300 minutes of weekly aerobic moderate-intensity PA, or 75 to 150 minutes of weekly aerobic vigorousintensity PA. 59 However, the bouts of 10 minutes to achieve PA recommendations was removed and replaced with the goal of increasing overall movement throughout the day. 59

Measurement of Physical Activity
To detect if age groups and populations are adhering to the guidelines, it is necessary to measure progress and activity though valid and reliable instruments.
Within the literature, there are various instruments used that are either objective or subjective tools that measure PA. Some include accelerometers or pedometers, which are objective tools, or surveys such as the National College Health Risk Behavior Survey, 7-Day Physical Activity Recall, or International Physical Activity Questionnaire (IPAQ), which are subjective tools. [89][90][91][92] Accelerometers and pedometers are considered quantitative measurements that directly measure the PA of the participant. 90 Since these tools give a direct measurement, they are often used to validate PA surveys, such as the IPAQ. 90 The IPAQ is self-reported assessment tool utilized to gather the amount of PA a participant completed over the past seven days. This assessment tool is offered in a long or short form, and gathers participation estimates in multiple domains of PA including transportation, occupation, house/lawn, and leisure time. 90,92,93 The long form provides detailed information about time spent in each domain, whereas the short form uses a sum of the scores to obtain a total score of PA. 93 Weekly vigorous activity, moderate activity, and walking are assessed and determined by multiplying the frequency and duration of each category of activity. 90,92 In order to utilize a survey such as the IPAQ, it should first be validated. A validation study was conducted Dinger et al., 90 examining the validity and reliability of the IPAQ short form in college students. The sample size included male and female undergraduate students (n=123) aged 18-30 years. The students were to wear an accelerometer and pedometer at their waists for seven consecutive days. At the end of the seven days, the participants would complete the IPAQ. The results indicated that the time spent in vigorous PA from the IPAQ was significantly correlated with steps per day from both the accelerometer and pedometer (p<0.01), whereas the time spent in moderate PA was significantly associated with the accelerometer (p<0.05). 90 The results of the study show the IPAQ responses were similar to that of the activity measured by the accelerometer and pedometer. Although it may be a reliable survey to use in place of a direct measurement tool, it is important to note the survey still may contain error due to selfreport response bias. Significant improvements are not shown for participation in PA for the whole population. However, differences have been found between sexes for participation in PA, where males tend to be more physically active than females. 27,59,90,94,95 In the NHIS, when adults were categorized by age group (18-24, 25-64, 65-74, and 75+), women in all age groups were less likely than men to meet the 2008 guidelines for aerobic activity. 94 It has also been found that when observing the college-aged population of adults (18-24 years of age), females participate in less PA (56.8%) than males (67.7%) with a total of 62.2% of males and females meeting the guidelines for aerobic PA. 94 Likewise, in the validation study by Dinger et al., 90 students overall reported engaging in 589.4±404.9 minutes of total PA per week with males reporting significantly more time in vigorous PA than females (p=.003). Overall, men tend to participate in more PA than females. However, U.S. adults are still not meeting the recommended frequencies and durations for aerobic PA.

Influences of Physical Activity
It is apparent that U.S. adults are not meeting the recommended amounts of aerobic PA. The reasoning for engagement in PA, or lack thereof, varies from person to person. Many individuals are active because it increases energy and health, whereas others are inactive due to body image dissatisfaction or their perception of their own ability. 59,62 Associations have been made between %BF, weight status, and BWD to participation in PA. In previous research, it has been shown that those with a lower %BF and are weight satisfied participate in more regular PA with higher levels of walking/jogging per week and higher cardiorespiratory fitness compared to those with higher %BF and BWD. 7,13,24,87 With regards to BWD, it has been shown that those who are dissatisfied with weight tend to be less physically active, compared to those who are satisfied with weight. 13,27 In a cross-sectional analysis conducted by Blake et al., 13

Conclusion
The increase in BWD has been identified as one of the behavioral patterns related to the development of eating disorders which can lead to detrimental behaviors that are associated with DQ, %BF, and participation in PA. 1,2,6,8,11,14,21,22

Study Design
The NAS is an ongoing IRB-approved study at URI that aims to examine nutrition assessment data for research to help us understand the relationship between diet and disease risk in college students in an Applied General Nutrition course (NFS 210). This study involved gathering anthropometrics, physical activity, dietary data, and blood values through assessments that are required as part of their coursework.
This cross-sectional, secondary data analysis investigated data that was collected in the NAS from college-aged students, aged 18-24 years old, during semesters in Fall 2015 to Fall 2019.
The independent variable was BWD. This was a quantitative independent variable defined as the absolute value of the difference between measured body weight in pounds and desired weight reported by the participant on the demographic survey.
A median split of BWD was used to categorize the independent variable into higher or lower BWD. The results include both the true value and the absolute value. A higher absolute value indicated a higher BWD and a greater desire to change weight, whereas a lower absolute value indicated lower BWD and a lesser desire to change weight. The dependent variables were total HEI score, body composition, and physical activity.
The primary dependent variable was total HEI score and examined the associations between BWD and DQ utilizing the total 2015 HEI score in college-aged students.  74 The InBody 770 utilized voice commands to guide the user through the InBody Test. 60 Students were to remove shoes, socks, heavy articles of clothing, and items in pockets. 60 They wiped their hands and feet with an InBody tissue or wipe. 60 They stood on the device barefoot and aligned their heel with the round silver electrodes and the rest of the feet with the foot electrode. 60 After weight was measured, the student input their age, height and sex. 60 When prompted, the student grabbed the hand electrodes, and kept arms relaxed and extended slightly away from the torso (roughly 15 degrees). 60 35,37 The HEI-2015 is an index ranging from zero to one-hundred, which is based on thirteen individual components with scores per item from zero to ten with nine adequacy components: total fruits, whole fruits, total vegetables, greens and beans, whole grains, milk/dairy, total protein foods, seafood and plant proteins, and fatty acids. It also includes four limited components: refined grains, sodium, added sugars, and saturated fat. 49 The HEI-2015 is updated every five years to reflect current federal dietary advice through a collaboration between the National Cancer Institute, and the US Department of Agriculture Center for Nutrition Policy and Promotion. 49 The output scores were calculated through the HEI-2015 algorithm within SAS software (SAS Institute Inc. version 9.4).

Percent Body Fat: Anthropometric Measures
The secondary dependent variable, percent body fat (%BF), was assessed using the Bod Pod or InBody 770, a wall-mounted stadiometer, and digital scale. Height was assessed using a wall-mounted stadiometer (SECA 240), and weight was assessed using a digital scale (SECA 700). 96 Subjects and undergraduate and undergraduate assistants were to follow proper protocol for accurate results. 96 Lastly %BF, which is the total mass of fat divided by total body mass, was measured using the Bod Pod or InBody 770. 60 The Bod Pod measures the volume of air displaced inside the chamber by the participant by subtracting the volume of air that remains inside the chamber to when then volunteer is not within the chamber. 74 The InBody 770 is a multifrequency BIA device that measures the body's resistance to flow of alternating electrical current at a designated frequency. 60 It has been found that the InBody 770 is a valid and reliable measure of body composition in relation to DEXA. 60,81 Physical Activity Assessment: IPAQ Short-Version The tertiary dependent variable, physical activity, was assessed using the IPAQ. The IPAQ is an electronic, seven item self-report instrument that with response format of open-ended questions. 90 The IPAQ is a self-administered instrument that required participants to report the frequency and duration of vigorous, moderate, and walking activities (10 minutes at minimum during the last seven days). 90 Weekly time spent in vigorous activity, moderate activity, and walking was determined by multiplying reported frequency and duration within each category of activity. This variable was calculated as minutes of moderate to vigorous physical activity per week. 90

Statistical Analysis
The statistical analysis package SPSS (IBM version 26.0 SPSS Inc.) was used to perform statistical analyses. Outliers were identified and excluded since there were significant differences in data when included. Skewness and kurtosis were used to assess normality of data distribution. A median split of BWD was used to categorize the independent variable into higher or lower BWD for the whole sample and stratified by sex. To assess between group differences, independent t-tests were conducted for demographic data for the whole sample and stratified by sex. To assess statistical differences between lower and higher BWD, analysis of variance (ANOVA) was conducted for the following main outcomes: mean HEI-2015 score, %BF, and IPAQ  (2) • Choose not to answer (3) Q7 Green Eating is: Eating locally grown foods, limited amounts of processed/fast foods, eating meatless meals at least one day per week, choosing organic foods as much as possible, and only taking what you plan on eating. Are you a green eater?
• No, and I do not intend to start within the next 6 months (1) • No, but I am thinking about becoming a green eater within the next 6 months (2) • No, but I am planning on becoming a green eater within the next 30 days (3) • Yes, I am a green eater and have been for less than 6 months (4) • Yes, I am a green eater and have been doing so for 6 months or more (5) • Choose not to answer (6) Q8 Which of the following best describes the MAJORITY of your meals during the academic year?
• I eat meals prepared at home. (1) • I purchase frozen or ready-to-eat meals. (2) • I eat at dining halls/restaurants. (3) • I get fast food/take-out. (4) • Choose not to answer (5) Q9 Do you have a campus meal plan?
Barely ever to never  Abbreviations: BWD-Body Weight Dissatisfaction, SD-Standard Deviation; 1 Median used to quantify high/low BWD in whole sample 2 Median used to quantify high/low BWD in males 3 Median used to quantify high/low BWD in females. Frequencies.