Impact of Stress Levels on Eating Behaviors among College Students

Weight gain is a common phenomenon among college students, especially those in their first year of university. Transitioning from high school to the college environment might increase perceived stress levels, thus affecting dietary behaviors and metabolism to promote overweight and obesity. The purpose of this study was to investigate the physical activities and dietary behaviors of college students in the context of their perceived stress levels. In addition, the demographic characteristics of the students were compared to ascertain their impact on dietary behaviors. Self-reported questionnaires were distributed to college students on campus in Korea. Perceived stress was measured by the 10-item Perceived Stress Scale (PSS-10), the scores for which were evaluated by independent t-tests to compare the dietary behaviors of the high- and low-perceived stress groups. Exploratory factor analysis was performed and Cronbach’s alphas were computed to assess the validity and internal consistency of the PSS-10 measurement items. Differences in the physical activities and dietary behaviors of the college students based on demographics such as sex, academic year, and residence type were found. Several dietary behaviors were significantly different between students with low and high perceived stress levels. Students with high perceived stress levels exhibited increased unhealthy dietary behaviors such as ready-prepared meal consumption (p < 0.001). These results suggest that stress management should be offered to college students. In addition, programs should be provided to help first-year students adjust to the college environment in order to promote healthy dietary behaviors.


Introduction
Being physically and psychologically healthy has been shown to improve quality of life. Despite the benefits of being healthy, the rate of obesity has increased two-to three-fold in the last 20 years [1]. Notably, data published in 2009 by the American College Health Association's National College Health Assessment indicated that 36.7% of college students were overweight or obese [2]. When investigating the relationship between sleep and obesity among U.S. and South Korean college students, Sa [3] observed higher overweight/obesity rates (59.4%) in Korean males than Black (51.5%) and White males (46.8%) and in Black women (53.8%) than in White (38.8%) and Korean women (24.8%) based on the standard body mass index (BMI) cut points for U.S. students and Asian BMI cut points for Korean students.
Dietary habits are influenced by various genetic, sociodemographic, and environmental factors. Food preferences developed and established in childhood may change during the college years because of decreased parental influence [4][5][6][7][8]. An abundance of literature demonstrates that most students gain weight during the first year of college [9][10][11]. During the transition from high school to university, the failure of college students to adjust to the new environment and associated stressors can lead to negative dietary habits such as excessive alcohol consumption and stress-related under-or over-eating [4]. factor) has been identified [35]. The perceived stress level was evaluated by the PSS-10 [34,38]. Each item in the PSS-10 is rated on a 5-point Likert scale ranging from never (0 points) to very often (4 points). The last part of the questionnaire sought to obtain demographic information, establish whether the participant was on a diet, and collect weight and height data to calculate the BMI of each participant.

Statistical Analysis
Data were coded in Excel and then analyzed using the Statistical Package for the Social Sciences (SPSS) [39]. Descriptive analyses were conducted to determine the patterns of college students' dietary behaviors and demographic information. Independent t-tests were run on dietary behaviors by sex (one-sided). Tukey's test was undertaken on dietary behaviors according to each respondent's academic year and residence type. Multiple regression was performed on BMI, with dietary behaviors as the independent variables. The construct validity of the PSS-10 was assessed by exploratory factor analysis using maximum likelihood estimation. Cronbach's alphas (>0.60) were computed to measure the internal consistency of the PSS-10 factors. Independent t-tests were performed to investigate dietary behavior differences between the low-and high-stress groups. In addition, stress differences were measured for sex, academic year, and residence type. Alpha was set at 0.05 for all statistical tests.

Results
The demographic characteristics of the respondents are shown in Table 1. Males represented 45.2% of the respondents and 54.8% were female. Approximately 35% of the respondents were freshmen, and 31% were sophomores. Most of the respondents answered that they lived alone (71.2%). Only one-third of the respondents (24.2%) were on a diet. BMI was calculated based on self-reported height (cm) and weight (kg). BMIs (kg/cm 2 ) < 18.5, 18.5-22.9, 23-24.9, and >25 were defined as low weight, normal weight, overweight, and obese, respectively [40]. About 50.3%, 18.3%, 13.6%, and 17.8% of the respondents were normal weight, overweight, low weight, and obese, respectively. The mean BMI was 20.854 (SD = 7.074), thereby falling in the normal weight range. Sex differences in BMI were significant (t = 7.722, p < 0.001). Males (M = 23.683, SD = 4.046) had a higher BMI than females (M = 18.462, SD = 8.134), who, in general, fell in the low-weight range. Students who were on a diet had a significantly (t = 3.369, p = 0.001) higher BMI (M = 22.894, SD = 6.501) than those not on a diet (M = 20.193, SD = 7.167). Significant differences were found among the academic years of the students (F = 4.186, p < 0.01). Freshmen (M = 22.470, SD = 5.094) had a higher BMI than sophomores (M = 19.488, SD = 6.784). Table 2 displays the physical activities, alcoholic beverage consumption, and dietary behaviors of the respondents. The majority of the respondents exercised 1-2 times a week (43.8%) for more than 30 min, whereas 42% did no exercise. A total of 46.1% of the respondents consumed alcoholic beverages 1-2 times a week, and more than half of the respondents were not taking dietary supplements (54.5%). Nearly 47% of the respondents answered that they ate when stressed. The respondents reported that they ate the following foods 1-2 times per week: over 40% ate more than two pieces of fresh fruits, nearly 44% ate fish and seafood, and over 43% ate dairy products such as milk, yogurt, and cheese (43.3%). Meats (39.7%) and bread or noodles made with wheat (43%) were consumed as a meal 3-4 times a week. Most of the respondents had snacks such as cakes, candies, and soft drinks (40.4%), fruit, milk, and nuts (41.1%), and ready-prepared meals (41.4%) 1-2 times a week. Fast food consumption 1-2 times a week had the highest frequency (51.7%), and 33.8% of the respondents ate out 3-4 times a week. Some 33.4% of the respondents skipped breakfast seven times a week, but most did not skip lunch (49%) or dinner (55.2%). The frequency of overeating 1-2 times a week was 49.4%. Dietary habits were compared between sexes (Table 3)  There were significant differences in the students' physical activities and dietary behaviors based on academic year. Exercising for more than 30  The PSS-10 items are shown in Table 4. Exploratory factor analysis was conducted and Cronbach's alphas were computed to assess the validity and reliability of the PSS-10 measurements. Two factors were extracted, of which one was positively coded (six items) and one was negatively coded (four items). All PSS-10 items were added and each total score was then divided by its mean value (M = 18.425, SD = 5.279). Results below the mean value were coded as "1" ("low stress") and results above the mean value were coded as "2" ("high stress"). Independent t-tests were conducted on the dietary behaviors of the low-and high-stress groups (Table 5), which revealed that the high-stress group (M = 1.81, SD = 0.854) performed less regular exercise (M = 2.02, SD = 0.981, t = 2.137, p < 0.05). In addition, the high-stress group (M = 2,43, SD = 0.933) had a higher frequency of eating to release stress than the low-stress group (M = 1.94, SD = 0.781, t = −5.556, p < 0.001) as well as a higher frequency of eating fast food, ready-prepared meals, and snacks (e.g., cakes, candies, and soft drinks); skipping meals (breakfast, lunch, and dinner); and overeating. However, meat consumption was higher in the low-stress group (M = 3.12, SD = 0.982) than the high-stress group (M = 2.93, SD = 0.956). A significant difference in stress was found based on sex (t = −4.049, p < 0.001), with a higher perceived stress level in females (M = 19.379, SD = 4.876) than males (M = 17.216, SD = 5.486) (not shown in table). The stress level did not differ with the respondents' academic year, residence type, or dieting.

Discussion
This study found several characteristics of college students' dietary behaviors including a strong reliance on meat consumption and eating out 3-4 times a week. The frequency meat intake was higher than that of fish/seafood and fresh fruit. The majority of the respondents did not exercise regularly or take dietary supplements. One-third of the respondents skipped breakfast almost every day, but few skipped lunch and dinner. Skipping meals has been associated with unhealthy dietary behaviors [41].
The observed dietary behaviors differed based on the students' sex, residence type, and academic year. More male students exercised regularly and for more than 30 min when compared with female students. Moreover, meat and fish/seafood consumption was more prevalent among males than females. Female students tended to eat to release stress, which supports the previous findings that females are more likely to be emotional eaters than males [21,42]. Interestingly, some studies have suggested that females eat more fresh fruit [12] and consume more snacks such as cakes, candies, and soft drinks than males [22]. Overall, the male students in this study tended to exhibit healthier dietary behaviors than the females, in contrast to a previous study [25].
Differences in the students' dietary behaviors based on academic year suggest the potential of academic institutional support and educational interventions. Some surveys have shown that first-year students display a low frequency of regular exercise, and it has been suggested that they fail to adjust to the new environment [4,20]. Hence, college administrations should provide assistance that supports first-year college students in their transition to the institution as well as implement recommended educational interventions regarding nutritional information, disseminated via leaflets or posters, to reinforce healthy dietary behaviors.
The majority of the respondents answered that they lived alone, which was the same as the general college population [43]. However, great differences were observed in the dietary behaviors of the students based on residence type [44]. Students who lived with their parents displayed much healthier dietary behaviors than those who lived on their own. For most students, this is their first time away from home, and so might lack the time, organization, or cooking skills to prepare meals [15]. Providing easy-cooking leaflets would encourage college students to cook for themselves and choose better food options by decreasing their reliance on eating out, fast food, and ready-prepared meals.
In this study, the average stress score was 18.43, which was higher than that reported by a group of infertile women (17.48) [35] and the general population (13.02) [36]. Adjusting to the college environment can be highly stressful, and stress greatly influences dietary behaviors. The high-stress group exhibited fewer healthy dietary behaviors as they tended to eat more sugar-based snacks, carbohydrate-rich food [21,27], fast food, and ready-prepared meals such as comfort food when compared with the low-stress group [26]. The results suggest that college students showed two opposing behaviors under stress including the more frequent intake of certain types of food (overeating) and less frequent eating such as skipping meals (undereating).
Two factors of stress are perceived helplessness and self-efficacy. These factors may increase eating in general as well as irregular eating, which can explain stress-induced overeating and undereating behaviors [45]. In addition, skipping meals is associated with stress and depressed mood [46], which may mean that skipping meals can be both the cause and result of stress. Regular breakfast consumption may increase physical activity [47], and so encouraging college students to eat breakfast is necessary for their health. Females have been shown to have a higher perceived stress level than males [25] and be more likely to diet. Moreover, distressed females are more likely to engage in binge eating [48] and be loss-of-control eaters [49]. Dietary restraint predicts overeating when under perceived stress [42] for both males and females [49]. Strict limitations on certain foods increase overeating in stressed individuals, so both female and male students should be informed about reasonable food consumption [48].
Dietary habits related to lifestyle are key elements of healthy living. Poorer mental and physical health-associated quality of life is related to body dissatisfaction, which is also associated with psychological distress [48,49]. For this reason, further studies should investigate the influence of internal and external lifestyle factors on dietary behaviors. Moreover, information on stress management should be provided to students, considering that high-stress students tend to eat when stressed. Emotional eating should be self-monitored and alternative stress-management strategies should be exercised for stress relief. University-affiliated clubs and societies that reflect the students' interests may help in this regard. The availability of a variety of healthy food options on campus including in cafes and vending machines, along with nutritional information, could prevent strict food limitations. In times of stress, food that is usually prohibited or limited is consumed, so a balanced intake of food from all food groups should be followed in daily life.
The BMI of the respondents in this study was normal and greatly affected by the intake of meat and fast food. A previous report suggested that students eat fast food because of its low cost and convenience [12]. A lack of cooking ability may explain the high consumption of fast food and meat [15]. Regarding cooking, meat can be cooked more easily compared with other food ingredients. In addition, a Westernized lifestyle means that college students eat more meat than fish/seafood and fruit. Unexpectedly, the higher consumption of snacks including sugar-based snacks was associated with a low BMI, implying that students eat snacks as replacements for proper meals, resulting in a relatively lower overall calorie intake.
This study has some limitations to be improved for future research. As this study was conducted with students attending a single college campus in Korea, a geographical limitation exists. Most students in this study lived alone, whereas different results would be expected among students living with family. Stress levels associated with long commutes, a factor not examined in this study, may also affect outcomes. More college students should be included for further analysis because the students' perceived stress levels may decrease with time spent at university as they adjust to the environment. This study used the frequency of intake of particular foods to determine dietary behaviors, whereas neither the amount of consumed food nor caloric information was considered. For further research, various food behaviors should be included to detect dietary behaviors more accurately. Moreover, the internal and external influences of dietary behaviors should be examined in the context of perceived stress levels. For example, body dissatisfaction, eating disordered behavior, and emotional eating should be assessed in future studies.

Conclusions
In this study, college students displayed a variety of dietary habits that differed between males and females. Moreover, differences were found in dietary habits based on the students' academic year and residence type. It is important to investigate and understand the environments that affect college students' dietary behaviors because eating patterns at this time in one's lifespan can affect health and food behaviors throughout adulthood. Support from college administrations such as educational interventions is necessary to improve unhealthy food behaviors. Moreover, measured stress was relatively high for college students, implying they were under stress due to the new academic environment. Students with high stress showed less healthy dietary behaviors compared to students with low stress. Thus, college students' stress should be managed properly to prevent unhealthy dietary behaviors related to stress. Stress management needs to start before college life, as stress should not hinder the college students' ability to engage in healthy dietary behaviors.