Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jan 28.
Published in final edited form as: Eat Behav. 2020 Sep 23;39:101435. doi: 10.1016/j.eatbeh.2020.101435

Impulsivity as a risk factor for weight gain and body roundness change among college freshmen

Kayla Bjorlie a,b, Tera L Fazzino a,b,*
PMCID: PMC8796869  NIHMSID: NIHMS1631956  PMID: 33022473

Abstract

Background:

The college setting is considered an obesogenic environment due to high availability of palatable foods. However, only a minority of freshmen gain weight. Individual-level risk factors, such as impulsivity, may hold utility in predicting weight and adiposity changes. Individuals with an impulsive decision-making style may favor immediately rewarding foods at the expense of long-term health. Individuals who seek rewarding foods during strong emotions may also be at risk. The study tested decisional (delay-discounting; DD) and emotion-driven impulsivity (urgency) as risk factors for 1) weight and 2) body roundness change during freshman year.

Methods:

Freshmen (N = 103) completed questionnaires assessing DD, positive urgency (PU), and negative urgency (NU). Weight and body roundness index (BRI) were collected at the beginning and end of the academic year. Four repeated measures regression models examined impulsivity factors predicting change in 1) weight and 2) BRI. Models included baseline weight and height or BRI, respectively. Covariates included average daily caloric intake, energy expenditure from walking, and sex.

Results:

In models examining weight, neither DD nor NU were significantly associated with weight at follow-up (b = 0.008, p = .977; b = 0.280, p = .075) when holding covariates constant. In contrast, PU was significantly associated with weight at follow-up (b = 0.303, p = .033). In models examining BRI, DD (b = −0.039, p = .511) and PU (b = 0.049, p = .072) were not associated with BRI at follow-up. In contrast, NU was significantly associated with BRI at follow-up (b = 0.068, p = .017).

Conclusions:

Emotion-driven impulsivity may be a risk factor for weight gain or change in body roundness during freshman year.

Keywords: Impulsivity, Weight gain, Body roundness index, Young adult, College

1. Introduction

Young adulthood is considered a critical risk period for development of obesity (Barbour-Tuck, Erlandson, Muhajarine, Foulds, & Baxter-Jones, 2018; Lanoye, Brown, & LaRose, 2017). For the majority (70%) of individuals, young adulthood involves the transition from high school to college (McFarland et al., 2018), which may present a substantial challenge to energy balance. Broadly, factors that have been associated with weight gain at college entry have included characteristics of the food environment, changes in eating behavior, changes in physical activity (Crombie, Ilich, Dutton, Panton, & Abood, 2009; Vadeboncoeur, Townsend, & Foster, 2015; Vella-Zarb & Elgar, 2009), and alcohol consumption (de Vos et al., 2015; Deforche, Van Dyck, Deliens, & De Bourdeaudhuij, 2015).

Although research has pointed to a variety of predictors of weight gain during college, an obesogenic food environment on college campuses has been identified as a key contributing factor. Specifically, all-you-can-eat style cafeterias are common on many college campuses and provide unlimited access to palatable foods. In addition, campus stores and vending machines also provide easy access to palatable foods. Thus, the college food environment may make healthy eating a challenge for most, if not all students (Horacek et al., 2020; Leischner, McCormack, Britt, Heiberger, & Kattelmann, 2018; Mongiello, Freudenberg, & Spark, 2015; Roy et al., 2017) and may influence food choice and energy balance over time (Byrd-Bredbenner et al., 2012).

Although navigating the college food environment may be challenging, research using an evidence-supported definition of weight gain (> 2 kg) has demonstrated that only a minority of students (~30%) gain weight in their first year of college (Fazzino, Serwatka, Schneider, & Sullivan, 2019). Thus, individual-level factors may play a role in distinguishing students at risk for weight gain. For example, college food environments may be especially challenging for young adults with impulsive traits. Specifically, young adults with a propensity toward impulsive decision-making may be more likely to choose immediately rewarding palatable foods over long-term goals such as health. An individual's degree of preference for immediate gratification from reinforcers such as food, relative to larger delayed rewards, has been conceptualized as a trait-based factor called delay discounting (DD) (Ainslie, 1975; Bickel & Marsch, 2001; Rachlin, 2000). High levels of DD reflect an impulsive decisional style, and in an obesogenic college environment may be expressed as greater consumption of highly rewarding, energy-dense foods, which may result in excess energy intake over time.

Additionally, individuals with high levels of urgency, or emotion-driven impulsivity, may seek out the rewarding effects of palatable foods in response to intense emotional states and may also be susceptible to overeating in college cafeterias (Heatherton & Baumeister, 1991). For example, in response to negative affective states such as stress, some individuals may seek out the rewarding effects of palatable foods to relieve negative emotions (Tannenbaum, Anisman, & Abizaid, 2010). Similarly, during positive affective states such as excitement, individuals may be motivated to consume palatable foods in order to further enhance positive emotions, as has been found previously among college students (Fazzino et al., 2018). Thus, food reward-seeking behavior may be common among individuals with high levels of emotion-driven impulsivity, formally referred to as urgency. Urgency may be conceptualized by valence of the driving emotion, with positive urgency (PU) referring to impulsive reward-seeking behavior during positive affective states and negative urgency (NU) referring to reward-seeking behavior during negative affective states (Lynam, Smith, Whiteside, Lafayette, & Purdue, 2006). Akin to high levels of DD, high levels of urgency may influence how an individual navigates a college cafeteria setting during heightened emotional states. Reliance on reward-driven eating may lead to excess energy intake and weight gain.

Overall, DD and urgency, although discrete dimensions of impulsivity, may both contribute to eating behavior, which may influence energy balance over time. In this regard, evidence from a variety of sources supports the premise that high levels of DD and urgency may play a role in weight gain and obesity risk (Barlow, Reeves, McKee, Galea, & Stuckler, 2016; Murphy, Stojek, & MacKillop, 2014; Steward et al., 2018). However, DD and urgency have been examined in separate studies and their relative contributions to weight gain/obesity risk are unclear. Regarding DD, several reviews have identified a positive correlation between DD and obesity or BMI (Amlung, Petker, Jackson, Balodis, & Mackillop, 2016; Barlow et al., 2016; McClelland et al., 2016). In addition, among initial studies using longitudinal designs, results have shown that among samples of women or children, high DD predicted weight gain over one to three years (Duckworth, Tsukayama, & Geier, 2010; Francis & Susman, 2009; Kishinevsky et al., 2012). However, more research is needed to clarify the role of DD in predicting anthropometric changes over time, particularly in young adults and samples that include men.

Similar to findings with DD, initial evidence examining urgency has reported that both PU and NU have been positively associated with BMI among adolescents (Coumans et al., 2018; Delgado-Rico, Río-Valle, González-Jiménez, Campoy, & Verdejo-García, 2012; Moreno-Padilla, Fernández-Serrano, & Reyes del Paso, 2018; Nazarboland & Pharmacology, 2015). However, no research has directly tested whether PU or NU predict weight change longitudinally. Furthermore, no studies have examined the associations of PU and NU with anthropometric outcomes among young adult or college samples.

Taken together, research suggests that decisional and emotion-driven impulsivity may predict weight gain. However, no studies have examined the relative utility of DD and urgency in predicting anthropometric changes. In addition, most research has been conducted cross-sectionally and in select populations (i.e., children, women). Further, most studies relied on BMI as an indicator of adiposity and obesity status; however there are limitations to this approach (Adab, Pallan, & Whincup, 2018; Okorodudu et al., 2010; Romero-Corral et al., 2008). BMI does not distinguish fat mass from fat-free mass, such as muscle mass. Among college athletes and non-athletes, BMI cut-points have demonstrated substantial inaccuracy in estimating excess adiposity (Ode, Pivarnik, Reeves, & Knous, 2007). Furthermore, evidence suggests that weight gain during college may be partially attributable to gain in fat-free mass, due to continued growth (increase in height) or gain in muscle mass (Hoffman, Policastro, Quick, & Lee, 2006; Morrow et al., 2006; Zahavich, 2010). Thus, BMI may have limitations in predicting obesity-related factors among young adults. Recently, the body roundness index (BRI) measure was created as a proxy for adiposity and has demonstrated superiority in predicting adiposity compared to other measures including BMI (Thomas et al., 2013). Thus, BRI may have utility in predicting changes in adiposity among young adults that may not be detected by traditional measures such as BMI or weight change. However, change in BRI has not previously been investigated among young adults.

The purpose of the current study was to test whether DD, PU, and NU at the start of college predicted weight and BRI change across the freshman year of college. We hypothesized that higher levels of DD and urgency would predict both weight and BRI change across the first year of college.

2. Methods

2.1. Data and parent study

The current study analyzed data from a parent trial that prospectively tested the role of reward-seeking behaviors involving caloric consumption, primarily alcohol use, in weight change during the first year of college (Fazzino, Forbush, Sullivan, & Befort, 2019). The study was also designed to examine common mechanisms underlying reward-driven consumption behaviors and obesity-related outcomes. Therefore, hypotheses regarding the constructs examined in the current study were identified a priori. Study details have previously been reported (Fazzino, Forbush, et al., 2019) and are presented herein as relevant to the current study.

2.2. Sampling and enrollment

All incoming college freshmen at a large midwestern university (N = 4116) received an email invitation to complete a brief online study screening. Inclusion criteria were: (a) incoming freshmen and (b) age 18+. Individuals were excluded if they reported “daily or almost daily” heavy episodic drinking, defined as 4+ drinks for females, or 5+ drinks for males on one occasion. Stratified sampling was used to obtain a representative sample of the freshman class based on sex (male/female), race/ethnicity (non-Hispanic white/racial or ethnic minority), and at-risk drinking status (at-risk drinking/not at-risk drinking). At-risk drinking status was identified by the Alcohol Use Disorder Identification Test- Consumption Questions, as normed for college students (score of 5+ for females or 7+ for males (DeMartini & Carey, 2012)). Eligible students from the eight groups were randomly selected to participate at the beginning of the academic year. Fig. 1 presents details of recruitment and enrollment. Enrollment rates across the eight groups ranged between 41 and 75%, with the lowest enrollment rate being among White males not engaged in at-risk drinking (Fazzino, Forbush, et al., 2019). The final sample consisted of 52% males, 46% individuals identifying as racial/ethnic minority, and 45% at-risk drinking.

Fig. 1.

Fig. 1.

Open in a new tab

Recruitment and enrollment.

2.3. Procedures

Study procedures were approved by the Institutional Review Board and all individuals provided written informed consent. Participants (N = 103) attended assessment visits at the beginning (September) and end (April) of the academic year. Participants completed surveys about health behaviors (eating habits, physical activity) and provided anthropometric measurements (height, weight, and waist circumference). Three days of dietary recall data (one weekday, two weekend days) were collected following each visit. Participants were compensated $25 for completing the first assessment visit, $45 for the final visit, and $15 for each completed dietary recall.

2.4. Measures

2.4.1. Impulsivity

2.4.1.1. Delay discounting.

Decisional impulsivity was assessed using the Monetary Choice Questionnaire (MCQ) (Kirby & Maraković, 1996). The 27-item questionnaire prompts individuals to choose between an immediate, smaller monetary reward, or a larger delayed reward. The choices are hypothetical and vary in length of delay, ranging from 7 to 186 days, and monetary amounts, ranging from $11 to $85. Evidence suggests that discounting outcomes for hypothetical versus real rewards are strongly correlated (Madden et al., 2004; Robertson & Rasmussen, 2018). Participant discounting rates were calculated using the Excel-based Kaplan automated calculator (Kaplan et al., 2016).

2.4.1.2. Positive and negative urgency.

Positive and negative urgency were assed using the Short Urgency-Premeditation-Perseverance-Sensation Seeking-Positive Urgency Impulsive Behavior Scale (SUPPS-P) (Cyders, Littlefield, Coffey, & Karyadi, 2014). This 20-item self-report measure assesses five facets of impulsivity, including positive and negative urgency. The original positive and negative urgency subscales have evidence for validity in first year college students, and the short form has shown high correlation with original subscales (Cyders et al., 2014; Cyders & Smith, 2010). Items corresponding to each facet of impulsivity were summed to obtain subscale scores, ranging from 4 to 16. Cronbach's alpha estimates for positive and negative urgency in our sample were 0.81 and 0.66 respectively. Although negative urgency had a lower reliability compared to positive urgency, it fell within a similar range of estimates for SUPPS-P subscales in prior studies (Doran & Trim, 2015; Mirhashem et al., 2017).

2.4.2. Anthropometric measurements and body roundness index

2.4.2.1. Height, weight, and waist circumference.

Height was obtained using a stadiometer. Weight was measured using a calibrated digital scale (Befour PS5700; Befour, Inc., Saukvile, WI). Participants were weighed in light clothing with shoes removed. Participants were weighed at the same time of day across visits to account for daily weight fluctuation (Fazzino, Forbush, et al., 2019). Waist circumference was measured twice within 2 cm using standardized procedures (Lohman, 1988). All measures were taken twice and averaged to obtain final values.

2.4.2.2. Body roundness index.

Waist circumference and height were used to estimate body roundness via the BRI calculator provided online by the authors along with the source article (Thomas et al., 2013). The BRI calculator also used demographic variables (age, sex, and race) to improve the accuracy of BRI estimates. A standardized double data entry process was used to enter anthropometric data into the BRI calculator, and discrepant values were resolved by a third entry conducted by a third researcher. BRI values range from 1 (long lean body) to 20 (round circular body).

2.4.3. Dietary intake

The Automated Self-Administered 24-Hour Diet Recall (ASA24) (Subar et al., 2012) was used to assess dietary intake. The ASA24 asks individuals report all food and beverages consumed during the past 24 h using an online platform. Online dietary recalls have convergent validity in estimating energy intake when compared to interview-administered recalls (Thompson et al., 2015). A photography-assisted method was used to enhance recall (detailed previously (Fazzino, Forbush, et al., 2019)). Average daily total caloric intake was calculated and included as a covariate in analyses.

2.4.4. Physical activity

The International Physical Activity Questionnaire (IPAQ) Short Form, was used to measure physical activity, consisting of moderate and vigorous physical activity, and walking. The measure has evidence supporting its validity when compared to accelerometer data (van der Ploeg et al., 2010). Estimates from moderate and vigorous activity were not used in analyses due to a substantial percentage of values being outside of plausible ranges (described in (Fazzino, Forbush, et al., 2019)). However, walking data appeared to be within a plausible range. Research has indicated that walking is the most commonly reported physical activity among adults (Rafferty, Reeves, et al., 2002) and self-reported weekly walking estimates have been associated with total moderate-intensity activity as assessed by accelerometers (van der Ploeg et al., 2010). Thus, average metabolic equivalent task (MET) minutes per week for walking was included as a covariate in analyses.

2.5. Data analysis

Analyses were conducted using Mplus version 8 (Muthén & Muthén, 2020). Multiple imputation was used to account for missing data at baseline (n = 25 diet recalls; n = 4 SUPPS-P; n = 11 MCQ) and at follow up (n = 23 participants missing at follow-up). The use of imputed datasets (N = 20) allowed for the inclusion of all data from the full study sample (Graham, Olchowski, & Gilreath, 2007). As is typical for discounting parameters, the distribution of the discounting rate variable was positively skewed (Green, Myerson, & McFadden, 1997). Therefore, a natural log transformation of k values was completed, and the geometric means of the natural log transformed values were used in analyses.

Repeated measures regression models were used to test DD, PU, and NU as predictors of weight and BRI change from the beginning to end of the academic year. Both weight and BRI were used as continuous outcome variables. Due to high correlations between PU and NU (r = 0.56, p < .001), each variable was analyzed in a separate model. First, four models were run to test associations with predictor and outcomes variables without covariates. Thus, two models included DD and PU as predicting change in 1) weight and 2) BRI, and two models included NU as predicting change in 3) weight and 4) BRI. Next, the four models were run with the addition of the following covariates to account for variation in the outcomes related to energy balance: average daily caloric intake (kcal intake), walking MET minutes per week (walking METs), and sex. Race/ethnicity and age were initially tested as covariates in analyses; however, they were not correlated with outcome variables and did not substantially impact beta estimates of the main variables of interest; thus race/ethnicity and age were not included in final models. Regression effects were estimated while holding covariates constant in the models.

3. Results

3.1. Demographics

Sample characteristics are displayed in Table 1. Of the 80 participants who completed the study, 28% (22/80) gained > 2.3 kg. Average weight change among those who gained weight was 4.60 kg (SD = 1.80). BRI increased an average of 0.08 points (SD = 1.09). On average, participants reported a mean daily kcal intake of 2100.93 (SD = 641.21), and 892.44 (SD = 929.60) mean walking MET minutes per week. Descriptive statistics for predictor and outcome variables are presented in Table 2. DD was not significantly correlated with either PU (r = −0.085, p = .422) or NU (r = −0.114, p = .286).

Table 1.

Sample characteristics (N = 103).

Variable Baseline
N (%) or M (SD)
Sex (female) 49 (48%)
Age 18.2 (0.6)
Race/ethnicity
 White 58 (56%)
 African American 6 (6%)
 Asian/Asian American 14 (14%)
 Native American 2 (1%)
 Hispanic American 15 (15%)
 Multiracial 8 (8%)
Body mass index 24.9 (4.4)
Open in a new tab

Table 2.

Descriptive statistics for predictor and outcome variable.

Variables Baseline
(N = 103)		 Visit 2
(N = 80)
M (SD) M (SD)
Positive urgency (n = 102)a 8.05 (2.93) –
Negative urgency (n = 100)a 8.80 (2.73) –
Delay discounting (k value; n = 92)b −4.82 (1.27) –
Weight (kg) 71.83 (14.96) 72.76 (16.53)
Body roundness index (n = 102)c 2.66 (1.40) 2.77 (1.21)
Open in a new tab
a

Scores could not be computed for n = 1 PU and n = 3 NU scales due to missing items on the SUPPS-P.

b

Score could not be computed for n = 11 DD values due to missing items on the MCQ.

c

Baseline BRI missing for one individual due to documentation error.

3.2. Weight change models

Results from the initial repeated measures regression models showed that DD was not significantly associated with weight at follow-up when accounting for baseline weight and height (b = −0.046, p = .876). However, PU was significantly associated with weight at follow-up when accounting for baseline weight and height (b = 0.295, p = .046). In contrast, NU (b = 0.249, p = .103) was not associated with weight at follow-up when accounting for baseline weight and height.

Results of the repeated measures regression models including covariates of kcal intake, walking METs, and sex are presented in Table 3. Findings were similar to the initial models without covariates; PU was significantly associated with weight at follow-up when holding the aforementioned covariates constant (b = 0.303, p = .033; Table 3). For every one-unit increase in PU, weight was 0.303 kg greater at follow up.

Table 3.

Repeated measures regression models for impulsivity variables and weight during the first year of college.

Beta SE p-Value 95% CI
Model 1: weight
Delay discounting (k) 0.008 0.298 0.977 −0.575 to 0.592
Positive urgency 0.303 0.142 0.033 0.024 to 0.581
Baseline weight 0.991 0.030 < 0.001 0.932 to 1.050
Baseline height 0.195 0.158 0.217 −0.115 to 0.504
Daily caloric intake 0.000 0.001 0.864 −0.002 to 0.001
Walking MET minutes −6.530 2.789 0.019 −11.996 to −1.063
Sex 0.668 1.199 0.578 −1.682 to 3.017
Model 2: weight
Negative urgency 0.280 0.158 0.075 −0.029 to 0.589
Baseline weight 0.986 0.030 < 0.001 0.927 to 1.046
Baseline height 0.153 0.159 0.335 −0.158 to 0.465
Daily caloric intake 0.000 0.001 0.864 −0.002 to 0.001
Walking MET minutes −7.359 2.618 0.005 −12.490 to −2.229
Sex −0.075 1.290 0.954 −2.603 to 2.453
Open in a new tab

Note: Unstandardized estimates are presented. SE = standard error; CI = confidence interval; MET = metabolic equivalent task.

3.3. BRI change

In the initial repeated measures regression model examining DD, findings indicated that DD was not significantly associated with BRI at follow-up, when accounting for baseline BRI (DD, b = −0.035, p = .585). However, PU (b = 0.061, p = .033) and NU (b = 0.062, p = .047) were both significantly associated with BRI at follow-up when accounting for baseline BRI.

Results of the BRI models including covariates are presented in Table 4. DD was not significantly associated with BRI when holding covariates constant (p = .511; Table 4). In contrast to the initial models, PU was not significantly associated with BRI when holding covariates constant (p = .072; Table 4). NU was significantly associated with BRI at follow-up when holding covariates constant (b = 0.068, p = .017; Table 4). For every one-unit increase in NU, BRI was 0.068 points greater at follow-up.

Table 4.

Repeated measures regression models for impulsivity variables and body roundness during the first year of college.

Beta SE p-Value 95% CI
Model 3: BRI
Delay discounting (k) −0.039 0.060 0.511 −0.157 to 0.078
Positive urgency 0.049 0.027 0.072 −0.004 to 0.103
Baseline BRI 0.625 0.072 < 0.001 0.484 to 0.765
Daily caloric intake 0.000 0.000 0.003 −0.001 to 0.000
Walking MET minutes −1.169 0.575 0.042 −2.295 to −0.043
Sex −0.381 0.161 0.018 −0.698 to −0.065
Model 4: BRI
Negative urgency 0.068 0.028 0.017 0.012 to 0.123
Baseline BRI 0.618 0.071 < 0.001 0.479 to 0.756
Daily caloric intake 0.000 0.000 < 0.001 −0.001 to 0.000
Walking MET minutes −1.335 0.520 0.010 −2.355 to −0.316
Sex −0.447 0.154 0.004 −0.749 to −0.145
Open in a new tab

Note: Unstandardized estimates are presented. SE = standard error; CI = confidence interval; MET = metabolic equivalent task.

4. Discussion

Young adulthood signifies a high-risk period for weight gain as many individuals transition from home to independent living in a college setting. The current study investigated the role of decisional impulsivity (delay discounting) and emotion-driven impulsivity (urgency), in predicting change in weight and body roundness index across the first year of college. Results indicated that DD was not significantly associated with change in weight or BRI, in contrast to our hypothesis. Findings also suggested that positive urgency and negative urgency differentially predicted weight and BRI change over the first year of college. Specifically, PU was significantly associated with weight change, while NU was significantly associated with change in BRI. However, the magnitude of the effects was similar across PU and NU in predicting both weight and BRI change, suggesting that differences in significance across constructs may have been due to limited power. Overall, our results suggest that urgency traits but not DD may play an important role in weight and BRI change during the freshman year.

Our findings that DD was not significantly associated with weight or BRI change are in contrast to the majority of prior research on DD and obesity (Amlung et al., 2016; Barlow et al., 2016; Garza, Ding, Owensby, & Zizza, 2016; McClelland et al., 2016). However, differences between our study and prior research may have contributed to this discrepancy. First, most of the research reporting positive associations between DD and obesity has been conducted using cross-sectional samples with dichotomous obesity outcomes. In contrast, we used a longitudinal study design with weight change as a continuous measure. Additionally, our study was representative of the incoming class of college freshmen with regard to sex and race/ethnicity. The few studies that found positive longitudinal associations between DD and weight gain were conducted among samples of women or children (Duckworth et al., 2010; Francis & Susman, 2009; Kishinevsky et al., 2012). Thus, our methodology and representative sample of young adult college students may have contributed to the differences in our findings compared to existing literature.

The current findings suggest that individuals who tend to act impulsively during strong emotions, both positive and negative, may be at an increased risk for weight or BRI gain during the freshman year. Our findings are in line with previous research that reported positive correlations between urgency and BMI among adolescents using cross-sectional study designs (Delgado-Rico et al., 2012; Nazarboland & Pharmacology, 2015). Our findings build upon the existing literature by indicating that urgency may play a role in anthropometric changes observed longitudinally in a sample of young adults.

The study also extends the literature by reporting the relative utility in examining multiple, distinct aspects of impulsivity as predictors of anthropometric changes. Consistent with the literature, our findings indicated that discounting and urgency were not significantly correlated with each other, supporting the premise that they are distinct facets of impulsivity (Reynolds, Ortengren, Richards, & de Wit, 2006). Accordingly, in this first empirical test of the relative associations between impulsivity constructs and weight/BRI change, findings indicated the unique predictive utility of emotion-driven impulsivity in weight/BRI change, suggesting that students who enter college with higher emotion-driven impulsivity may be at risk for weight and body roundness change during their freshman year. In contrast, delay discounting alone does not appear to be a risk factor for weight gain among college freshmen.

The study had several limitations. First, there was potential for selection bias given that 15.5% of freshmen completed the eligibility screen, and 58% of eligible students enrolled in the study. Enrollment rates among the eight sampling groups were relatively high (mostly 60–70%) and resulted in a representative sample based on sex and race/ethnicity. However, it is possible that individual-level factors, such as personality characteristics, may have influenced both the likelihood of participation in the study and variables of interest, such as impulsivity characteristics. In this regard, the study also sampled a greater percentage of at-risk drinking students than is representative of typical college samples, which may have increased the rate of high delay discounting in the sample, as discounting is strongly correlated with risky drinking behavior (Kollins, 2003). Thus, results should be replicated in a college sample that did not over-sample at-risk drinking students. Second, the amount of missing data for energy intake (n = 25) and follow-up outcomes (n = 23) was notable. However, multiple imputation, a robust missing data estimation method, was used to address missing data (Imputation, Fichman, & Cummings, 2003; Newgard, 2006). Another limitation was we were unable to use moderate to vigorous physical activity data. However, walking data were included as a proxy measure of physical activity in analyses. Additionally, NU demonstrated lower than ideal reliability (alpha = 0.66) within our sample, which may increase the risk for Type II errors. Finally, both energy intake and walking data were self-reported, and thus may be subject to limitations of recall and reporting biases.

Strengths of the study included the evaluation of multiple facets of impulsivity in assessing weight and BRI change during the first year of college. In addition, the study was the first to assess change in BRI among young adults. Furthermore, our sample was balanced across males and females and included 46% of participants who identified as racial/ethnic minority, which may facilitate generalizability to other young adult college samples.

5. Conclusions

Findings from the present study suggest that emotion-driven impulsivity may constitute an important individual-level risk factor for weight gain and body roundness change during the first year of college. Future obesity prevention work may consider targeting individuals high in positive and negative urgency during the first year of college.

Role of funding sources

Funding for this study was provided by the National Institute on Alcohol Abuse and Alcoholism, grant F32 AA024669-01 (PI: Fazzino). The funding source had no role in the study design, analysis, interpretation, or writing and submission of the manuscript for publication.

Footnotes

CRediT authorship contribution statement

Kayla Bjorlie: Formal analysis, Original draft preparation, Writing – Review & Editing

Tera Fazzino: Conceptualization, Methodology, Investigation, Writing – Review & Editing, Funding acquisition

Declaration of competing interest

All authors declare they have no conflicts of interest.

References

  1. Adab P, Pallan M, Whincup PH. Is BMI the best measure of obesity? Vol. 360, BMJ (Online). BMJ Publishing Group; 2018. [DOI] [PubMed] [Google Scholar]
  2. Ainslie G (1975. Jul). Specious reward: A behavioral theory of impulsiveness and impulse control. Psychological Bulletin, 82(4), 463–496. [DOI] [PubMed] [Google Scholar]
  3. Amlung M, Petker T, Jackson J, Balodis I, Mackillop J. Steep discounting of delayed monetary and food rewards in obesity: a meta-analysis. 2016. [cited 2018 Nov 25]; Available from: https://search-proquest-com.www2.lib.ku.edu/docview/1804904170/fulltextPDF/3D421588053B4263PQ/1?accountid=14556. [DOI] [PubMed] [Google Scholar]
  4. Barbour-Tuck E, Erlandson M, Muhajarine N, Foulds H, & Baxter-Jones A (2018. Mar). Influence of childhood and adolescent fat development on fat mass accrual during emerging adulthood: A 20-year longitudinal study. Obesity [Internet], 26(3), 613–620. (cited 2020 Jan 27). Available from: http://doi.wiley.com/10.1002/oby.22111. [DOI] [PubMed] [Google Scholar]
  5. Barlow P, Reeves A, McKee M, Galea G, & Stuckler D (2016). Unhealthy diets, obesity and time discounting: a systematic literature review and network analysis. Obes Rev [Internet], 17(9), 810–819. (cited 2019 Apr 6). Available from http://www.ncbi.nlm.nih.gov/pubmed/27256685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bickel WK, Marsch LA. Toward a behavioral economic understanding of drug dependence: delay discounting processes. Addiction [Internet]. 2001. Jan [cited 2020 Jan 27];96(1):73–86. Available from: http://doi.wiley.com/10.1046/j.1360-0443.2001.961736.x. [DOI] [PubMed] [Google Scholar]
  7. Byrd-Bredbenner C, Johnson M, Quick VM, Walsh J, Greene GW, Hoerr S, et al. (2012. Jun). Sweet and salty. An assessment of the snacks and beverages sold in vending machines on US post-secondary institution campuses. Appetite. 58(3), 1143–1151. [DOI] [PubMed] [Google Scholar]
  8. Coumans JMJ, Danner UN, Intemann T, De Decker A, Hadjigeorgiou C, Hunsberger M, et al. (2018. Apr 1). Emotion-driven impulsiveness and snack food consumption of European adolescents: Results from the I. Family study. Appetite. 123, 152–159. [DOI] [PubMed] [Google Scholar]
  9. Crombie AP, Ilich JZ, Dutton GR, Panton LB, Abood DA. The freshman weight gain phenomenon revisited. Nutr Rev [Internet]. 2009. Feb 1 [cited 2018 Nov 21];67(2):83–94. Available from: https://academic.oup.com/nutritionreviews/article-lookup/doi/10.1111/j.1753-4887.2008.00143.x. [DOI] [PubMed] [Google Scholar]
  10. Cyders MA, Littlefield AK, Coffey S, & Karyadi KA (2014). Examination of a short English version of the UPPS-P Impulsive Behavior Scale. Addictive Behaviors, 39(9), 1372–1376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cyders MA, & Smith GT (2010. Jan). Longitudinal validation of the urgency traits over the first year of college. J Pers Assess [Internet], 92(1), 63–69. (cited 2020 Jan 29). Available from: http://www.tandfonline.com/doi/abs/10.1080/00223890903381825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Deforche B, Van Dyck D, Deliens T, & De Bourdeaudhuij I (2015. Feb 15). Changes in weight, physical activity, sedentary behaviour and dietary intake during the transition to higher education: A prospective study. Int J Behav Nutr Phys Act [Internet], 12(1), 16. (cited 2020 May 26). Available from: https://ijbnpa.biomedcentral.com/articles/10.1186/s12966-015-0173-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Delgado-Rico E, Río-Valle JS, González-Jiménez E, Campoy C, Verdejo-García A. BMI predicts emotion-driven impulsivity and cognitive inflexibility in adolescents with excess weight. Obesity [Internet]. 2012. Aug 1 [cited 2020 May 18];20(8):1604–10. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1038/oby.2012.47. [DOI] [PubMed] [Google Scholar]
  14. DeMartini KS, Carey KB. Optimizing the use of the AUDIT for alcohol screening in college students. Psychol Assess [Internet]. 2012. Dec [cited 2020 Sep 8];24(4):954–63. Available from: /pmc/articles/PMC3869562/?report=abstract. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Doran N, & Trim RS (2015. Dec 1). Correlates of other tobacco use in a community sample of young adults. Addictive Behaviors, 51, 131–135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Duckworth AL, Tsukayama E, & Geier AB (2010. Apr). Self-controlled children stay leaner in the transition to adolescence. Appetite. 54(2), 304–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fazzino TL, Forbush K, Sullivan D, & Befort CA (2019. May 8). A prospective study of alcohol use patterns and short-term weight change in college freshmen. Alcohol Clin Exp Res [Internet], 43(5), 1016–1026. (cited 2020 Jan 27). Available from: https://onlinelibrary.wiley.com/doi/abs/10.1111/acer.14025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fazzino TL, Raheel A, Peppercorn N, Forbush K, Kirby T, Sher KJ, Motives for drinking alcohol and eating palatable foods: An evaluation of shared mechanisms and associations with drinking and binge eating. Addict Behav [Internet], et al. (2018). [cited 2018 Aug 1];85:113–9. Available from 10.1016/j.addbeh.2018.04.025. [DOI] [PubMed] [Google Scholar]
  19. Fazzino TL, Serwatka C, Schneider H, Sullivan D. A systematic review of the methodology used to study weight change among young adults attending college. Vol. 35, Eating Behaviors. Elsevier Ltd; 2019. p. 101333. [DOI] [PubMed] [Google Scholar]
  20. Francis LA, & Susman EJ (2009. Apr). Self-regulation and rapid weight gain in children from age 3 to 12 years. Archives of Pediatrics & Adolescent Medicine, 163(4), 297–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Garza KB, Ding M, Owensby JK, & Zizza CA (2016. Jan 1). Impulsivity and fast-food consumption: A cross-sectional study among working adults. Journal of the Academy of Nutrition and Dietetics, 116(1), 61–68. [DOI] [PubMed] [Google Scholar]
  22. Graham JW, Olchowski AE, & Gilreath TD (2007. Sep). How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prevention Science, 8(3), 206–213. [DOI] [PubMed] [Google Scholar]
  23. Green L, Myerson J, & McFadden E (1997). Rate of temporal discounting decreases with amount of reward. Memory and Cognition, 25(5), 715–723. [DOI] [PubMed] [Google Scholar]
  24. Heatherton TE, Baumeister RF. Binge eating as escape from self-awareness. Vol. 110, Psychological Bulletin. 1991. [DOI] [PubMed] [Google Scholar]
  25. Hoffman DJ, Policastro P, Quick V, Lee SK. Changes in body weight and fat mass of men and women in the first year of college: A study of the “freshman 15.” Journal of American College Health 2006;55(1):41–6. [DOI] [PubMed] [Google Scholar]
  26. Horacek TM, Erdman MB, Reznar MM, Olfert M, Onikia D; Brown-Esters N, et al. Nutrition/health promoting community design, [cited 2020 Jan 27]; Available from: http://www. [Google Scholar]
  27. Imputation M, Fichman M, & Cummings JN (2003). ARTICLE ORGANIZATIONAL RESEARCH METHODS multiple imputation for missing data: Making the most of what you know. [Google Scholar]
  28. Kaplan BA, Amlung M, Reed DD, Jarmolowicz DP, McKerchar TL, & Lemley SM (2016. Oct 1). Automating scoring of delay discounting for the 21- and 27-item monetary choice questionnaires. Behavior Analyst, 39(2), 293–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kirby KN, & Maraković NN (1996). Delay-discounting probabilistic rewards: Rates decrease as amounts increase. Psychonomic Bulletin & Review, 3(1), 100–104. [DOI] [PubMed] [Google Scholar]
  30. Kishinevsky FI, Cox JE, Murdaugh DL, Stoeckel LE, Cook EW, Weller RE. fMRI reactivity on a delay discounting task predicts weight gain in obese women. Appetite [Internet]. 2012. Apr 1 [cited 2018 Nov 21];58(2):582–92. Available from: https://www-sciencedirect-com.www2.lib.ku.edu/science/article/pii/S0195666311006581. [DOI] [PubMed] [Google Scholar]
  31. Kollins SH (2003. Aug 1). Delay discounting is associated with substance use in college students. Addictive Behaviors, 28(6), 1167–1173. [DOI] [PubMed] [Google Scholar]
  32. Lanoye A, Brown KL, LaRose JG. The transition into young adulthood: A critical period for weight control. Vol. 17, Current diabetes reports. Current Medicine Group LLC; 1; 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Leischner K, McCormack L, Britt B, Heiberger G, Kattelmann K. The healthfulness of entrees and students' purchases in a university campus dining environment. Healthcare [Internet]. 2018. Mar 22 [cited 2020 Jan 27];6(2):28. Available from: http://www.mdpi.com/2227-9032/6/2/28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lohman TRA (1988). Anthropometric standardization reference manual. [Google Scholar]
  35. Lynam D, Smith G, Whiteside S, Lafayette MC-W, & Purdue I (2006). The UPPS-P: Assessing five personality pathways to impulsive behavior. 2006. [Google Scholar]
  36. Madden GJ, Raiff BR, Lagorio CH, Begotka AM, Mueller AM, Hehli DJ, et al. (2004. Nov). Delay discounting of potentially real and hypothetical rewards: II. Between- and within-subject comparisons. Experimental and Clinical Psychopharmacology, 12(4), 251–261. [DOI] [PubMed] [Google Scholar]
  37. McClelland J, Dalton B, Kekic M, Bartholdy S, Campbell IC, Schmidt U. A systematic review of temporal discounting in eating disorders and obesity: Behavioural and neuroimaging findings. Vol. 71, Neuroscience and Biobehavioral Reviews. Elsevier Ltd; 2016. p. 506–28. [DOI] [PubMed] [Google Scholar]
  38. McFarland J, Hussar B, Wang X, Zhang J, … Wang K, National Center for Education Statistics (2018). The condition of education 2018. Institute of Educational Sciences. [Google Scholar]
  39. Mirhashem R, Allen HC, Adams ZW, van Stolk-Cooke K, Legrand A, & Price M (2017. Jun 1). The intervening role of urgency on the association between childhood maltreatment, PTSD, and substance-related problems. Addictive Behaviors, 69, 98–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Mongiello LL, Freudenberg N, & Spark A (2015. Feb 13). Making the healthy choice the easy choice on campus: A qualitative study. Heal Behav Policy Rev. 2(2), 110–121. [Google Scholar]
  41. Moreno-Padilla M, Fernández-Serrano MJ, & Reyes del Paso GA (2018. Aug). Risky decision-making after exposure to a food-choice task in excess weight adolescents: Relationships with reward-related impulsivity and hunger. PLoS One. 1, 13(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Morrow ML, Heesch KC, Dinger MK, Hull HR, Kneehans AW, & Fields DA (2006). Descriptive epidemiology freshman 15: Fact or fiction? . [DOI] [PubMed] [Google Scholar]
  43. Murphy CM, Stojek MK, & MacKillop J (2014. Feb 1). Interrelationships among impulsive personality traits, food addiction, and body mass index. Appetite. 73, 45–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Muthén L, Muthén B Mplus version 8 [Internet]. [cited 2020 Jan 30]. Available from: http://www.statmodel.com/virg_nov_course.shtml. [Google Scholar]
  45. Nazarboland N, Pharmacology NF-B and, 2015. undefined. The role of BMI in predicting emotion-driven impulsivity and sensitivity to reward/punishment in over-obese adolescents. biomedpharmajournal.org [Internet]. [cited 2020 May 19]; Available from: http://biomedpharmajournal.org/vol8octoberspledition/the-role-of-bmi-in-predicting-emotion-driven-impulsivity-and-sensitivity-to-rewardpunishment-in-over-obese-adolescents/. [Google Scholar]
  46. Newgard CD. The validity of using multiple imputation for missing out-of-hospital data in a state trauma registry. Acad Emerg Med [Internet]. 2006. Feb 22 [cited 2020 May 16];13(3):314–24. Available from: http://doi.wiley.com/10.1197/j.aem.2005.09.011. [DOI] [PubMed] [Google Scholar]
  47. Ode JJ, Pivarnik JM, Reeves MJ, & Knous JL (2007. Mar). Body mass index as a predictor of percent fat in college athletes and nonathletes. Med Sci Sport Exerc [Internet], 39(3), 403–409. (cited 2020 Jan 29). Available from: https://insights.ovid.com/crossref?an=00005768-200703000-00002. [DOI] [PubMed] [Google Scholar]
  48. Okorodudu DO, Jumean MF, Montori VM, Romero-Corral A, Somers VK, Erwin PJ, et al. Diagnostic performance of body mass index to identify obesity as defined by body adiposity: A systematic review and meta-analysis. Vol. 34, International Journal of Obesity. 2010. p. 791–9. [DOI] [PubMed] [Google Scholar]
  49. van der Ploeg HP, Tudor-Locke C, Marshall AL, Craig C, Hagströmer M, Sjöström M, et al. (2010). Reliability and validity of the international physical activity questionnaire for assessing walking. Research Quarterly for Exercise and Sport, 81(1), 97–101. [DOI] [PubMed] [Google Scholar]
  50. Rachlin H The science of self-control [Internet]. Harvard University Press. 2000. [cited 2020 Jan 30]. Available from: https://books.google.co.uk/books?hl=en&lr=&id=fx-lW2Tp4yAC&oi=fnd&pg=PP10&dq=self+control+during+sleep&ots=C_EcfSRFeA&sig=xgKwVl50thQeUTB4Vg3KBpXFbzY#v=onepage&q=self. [Google Scholar]
  51. Rafferty A, Reeves M, … HM-M& S in, 2002. undefined. Physical activity patterns among walkers and compliance with public health recommendations. journals.lww.com [Internet]. [cited 2020 May 16]; Available from: https://journals.lww.com/amjdermatopathology/00005768-200208000-00005.fulltext. [DOI] [PubMed] [Google Scholar]
  52. Reynolds B, Ortengren A, Richards JB, & de Wit H (2006. Jan 1). Dimensions of impulsive behavior: Personality and behavioral measures. Pers Individ Dif. 40(2), 305–315. [Google Scholar]
  53. Robertson SH, & Rasmussen EB (2018. Apr 1). Comparison of potentially real versus hypothetical food outcomes in delay and probability discounting tasks. Behavioural Processes, 149, 8–15. [DOI] [PubMed] [Google Scholar]
  54. Romero-Corral A, Somers VK, Sierra-Johnson J, Thomas RJ, Collazo-Clavell ML, Korinek J, et al. (2008. Jun). Accuracy of body mass index in diagnosing obesity in the adult general population. International Journal of Obesity, 32(6), 959–966. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Roy R, Rangan A, Hebden L, Yu Louie JC, Tang LM, Kay J, et al. (2017. Feb 1). Dietary contribution of foods and beverages sold within a university campus and its effect on diet quality of young adults. Nutrition. 34, 118–123. [DOI] [PubMed] [Google Scholar]
  56. Steward T, Mestre-Bach G, Vintró-Alcaraz C, Lozano-Madrid M, Agüera Z, Fernández-Formoso JA, et al. Food addiction and impaired executive functions in women with obesity. Eur Eat Disord Rev [Internet]. 2018. Nov [cited 2020 Jan 29];26(6):574–84. Available from: http://doi.wiley.com/10.1002/erv.2636. [DOI] [PubMed] [Google Scholar]
  57. Subar AF, Kirkpatrick SI, Mittl B, Zimmerman TP, Thompson FE, Bingley C, et al. (2012. Aug). The Automated Self-Administered 24-Hour Dietary Recall (ASA24): A resource for researchers, clinicians, and educators from the National Cancer Institute. Journal of the Academy of Nutrition and Dietetics, 112(8), 1134–1137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Tannenbaum BM, Anisman H, & Abizaid A (2010). Neuroendocrine stress response and its impact on eating behavior and body weight In: Obesity Prevention. Elsevier Inc; 261–271. [Google Scholar]
  59. Thomas DM, Bredlau C, Bosy-Westphal A, Mueller M, Shen W, Gallagher D, et al. Relationships between body roundness with body fat and visceral adipose tissue emerging from a new geometrical model. Obesity [Internet]. 2013. Nov [cited 2020 Jan 29];21(11):2264–71. Available from: http://doi.wiley.com/10.1002/oby.20408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Thompson FE, Dixit-Joshi S, Potischman N, Dodd KW, Kirkpatrick SI, Kushi LH, et al. Comparison of interviewer-administered and automated self-administered 24-hour dietary recalls in 3 diverse integrated health systems. Am J Epidemiol [Internet]. 2015. Jun 15 [cited 2020 Jan 30];181(12):970–8. Available from: https://academic.oup.com/aje/article-lookup/doi/10.1093/aje/kwu467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Vadeboncoeur C, Townsend N, Foster C. A meta-analysis of weight gain in first year university students: is freshman 15 a myth? BMC Obes [Internet]. 2015. Dec 28 [cited 2018 Nov 21];2(1):22. Available from: http://bmcobes.biomedcentral.com/articles/10.1186/s40608-015-0051-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Vella-Zarb R, & Elgar F (2009). The “Freshman 5”: A meta-analysis of weight gain in the freshman year of college. Journal of American College Health, 58(2), 161–166. [DOI] [PubMed] [Google Scholar]
  63. de Vos P, Hanck C, Neisingh M, Prak D, Groen H, & Faas MM (2015. Apr 1). Weight gain in freshman college students and perceived health. Prev Med Reports. 2, 229–234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Zahavich JBL. Examining the freshman weight gain phenomenon with measures of fat-free mass and fat mass. Available from ProQuest Diss Theses Glob; (847250228). 2010. [Google Scholar]

RESOURCES