In this longitudinal study of measured weights and self-reported diet and health behaviors among 159 18-year-old men and women enrolled as freshmen at a mid-sized land grant University in the Western U.S., 20.4% entered college overweight or obese as indicated by a BMI ≥ 25. The average amount of weight gain experienced by these students during their first semester of college was modest (1.5 kg; 3.3 lbs), although 23% of participants gained an amount of weight ≥ 5% of their baseline body weight. The average amount of weight gained among those who gained ≥ 5% of their baseline body weight was 4.52 kg (9.9 lbs).
Like others studying the phenomenon of weight gain among freshmen, we observed that some but not all freshmen do gain a significant amount of weight during their first semester of college. Average weight gain in our study among all participants was 1.5 (SD 2.28) kg, an amount similar to the amount reported in other studies of weight gain among freshmen [4
]. Clinically significant weight gain was defined as an amount ≥ 5% of each individual's baseline body weight, a method not previously used in similar studies. This method of defining what is considered clinically significant weight gain may help to control for the large differences in the magnitude of weight gain experienced by men and women due to differences in body mass. The average amount of weight gain observed among men and women who gained ≥ 5% of body weight was 5.3 (± 1.9) among men, and 4.2 (± 1.4) kg among women.
In general, our findings are consistent with the findings of others who report the transition from high school to college promotes changes in behavior and environment that may support weight gain [23
]. Others have identified eating in dining halls as a significant risk factor for weight gain during the first semester of college [7
]. Levitsky [7
] hypothesized that the greater abundance and variety of food available in dining halls may promote intake in excess of energy needs.
In a community-controlled analysis of weight gain among university women, Hovel et al. [8
] found that by the junior year of college average weight returned to near baseline levels of the cohort as entering freshmen. The weight loss among women during their junior year was speculated to be associated with a move from mandatory dormitory housing and dining-hall-type dining experiences to other options. In a study examining weight change across years of college among a small cohort of men and women Racette et al. [25
] found that the accelerated rate of weight gain experienced during the freshmen year of college did not continue through senior year and that by senior year most students (85%) had moved from residence halls to off-campus housing. Approximately 65% (n = 102) of the participants in this study lived on campus and reported eating at least occasionally at all-you-can-eat dining facilities. In the present study, those who gained ≥ 5% of body weight ate an average of 2.1 more meals per week in on-campus dining facilities during fall semester (August – December) than did those who did not gain ≥ 5% of body weight, although the statistical significance of this difference is marginal (p-value = 0.06).
Somewhat surprising were the associations between more frequent breakfast consumption and greater amounts of sleep reported among those who gained ≥ 5% of body weight. These associations have not been previously reported in the literature regarding risk factors for weight changes during the transition to college. A substantial body of literature provides evidence that breakfast skipping in children, adolescents, and adults is associated with body weight [26
]. In this study of first-time freshmen, regular breakfast consumption was marginally associated with on-campus living (p = 0.057); on-campus living was associated with more frequent meals eaten in all-you-can-eat dining facilities (p-value = 0.009). The observed findings of a positive association between breakfast consumption and weight gain may reflect differences in access to all-you-can-eat dining facilities among college freshmen and nationally representative samples of adolescents and adults. The frequency of breakfast consumption in all-you-can-eat dining facilities was not quantified in the present study. In addition, regular breakfast consumption was defined as eating breakfast at least four times per week; additional details about the frequency and type of breakfast consumed may have helped to clarify the observed associations.
Short sleep duration has been associated with obesity in paediatric populations but the existing evidence regarding associations between habitual sleep duration and body weight among adults is not consistent [29
]. Our findings regarding associations between sleep patterns and risk for weight gain among college freshmen are novel. Relationships between sleep patterns among young-adults attending college and weight change and other indicators of health are important and deserves further study.
Our study has several limitations. First, participants with a BMI ≥ 25 at the baseline interview were more likely to drop out of the study than were those with BMI < 25. This may have biased our results to the null if those who dropped out were also those who gained weight; those who dropped out may have differed in other important ways from those who continued in the study. For example, participants who began the study with a BMI >25 had higher rates of drinking than did those with BMIs < 25. This may have contributed to differences in the prevalence of alcohol consumption at the August and December assessments.
Second, we assessed diet and physical activity using instruments that relied on the accurate memory recall of usual behaviors by participants. The baseline survey instruments asked participants to report their behaviors during a six-month period of time that occurred three to nine months prior to when the data were collected so as to capture usual behaviors during the period of time that included their last six months of high school. The first FFQ, being more retrospective than the second, may have provided a less accurate estimate of usual dietary intake and physical activity than did the more immediate assessment of behaviors collected at the second data collection period as it is known that reports of past behavior are influenced by current behavior [30
]. However, Maruti et al. [31
] found that FFQs may be used to provide a reasonable estimate of usual dietary intake in the distant past. In the Maruti et al. [31
] study young adults were able to use an FFQ to accurately report usual dietary intake from approximately 10 years in the past using a FFQ.
In addition, although total energy intake between the first and second assessments was correlated (r = 0.57, p-value < 0.001), there were significant differences between the absolute total energy intake reported at the first and second assessments that in general would not support weight gain among the population. Butler et al[4
], and Jung et al[9
] also found that energy intake decreased significantly during the first semester of college, despite overall increases in weight. Our finding of an association between being less physically active and weight gain during the first semester of college is consistent with the findings of both Butler et al[4
], and Jung et al[9
] who also found decreased physical activity associated with increased risk for weight gain despite overall decreases in energy consumption.
Third, because we did not assess body composition, we cannot determine whether the observed increases in body weight were associated with growth or increases in lean or non-lean body mass. However, averaged measured heights were not different between the baseline and follow-up assessments, indicating little change in stature during the 16-week study period among our participants.
Finally, the participants in this study were recruited from among first-time freshmen attending one university with a population of students who are predominately non-Hispance white (91%) and who report lower rates of smoking and drinking than reported nationally. This study population likely does not represent the diversity of first-time freshmen attending college campuses nationally.