Food Insecurity As A Longitudinal Predictor Of Diet, Sleep, Mental Health And College Academics
Open Access
- Author:
- Smith, Kiara
- Graduate Program:
- Nutritional Sciences
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- March 28, 2022
- Committee Members:
- A Ross, Program Head/Chair
Muzi Na, Thesis Advisor/Co-Advisor
Alison D Gernand, Committee Member
Jennifer Savage-Williams, Committee Member - Keywords:
- food insecurity
nutrition
diet quality
mental health
sleep
academic performance
college student - Abstract:
- Background: Food insecurity (FI) and its negative relationship with GPA has been thoroughly studied cross-sectionally in college students, but few have considered how FI changes over the course of a semester. Further, many previous studies have used the 6-item USDA short form, which is believed to overestimate food insecurity prevalence in this population. None of the other outcomes, besides academic performance measured via GPA alone, have been analyzed by food security status in college students over time. Aims: This study sought to assess the FI prevalence for 2nd and 3rd year undergraduate students at Penn State’s University Park campus and how that prevalence changed over a semester. It also sought to analyze associations between FI status and diet, sleep, mental health and academic outcomes. Methods: Participants (n=79 for phase 1, n=77 for phase 2) completed two surveys, one in October and another in December to reflect on the previous month during Fall 2021. FI was assessed using the 10-item Adult Food Security Survey Module. Diet, sleep, mental health and academic outcomes were assessed via validated questionnaires. Descriptive statistics were used to report food insecurity prevalence and participant demographics. Within-person differences from phase 1 to phase 2 were also analyzed. An unadjusted analysis revealed whether baseline food security status could predict any of the outcomes of interest. Afterwards an adjusted analysis was performed, controlling for potentially confounding variables identified in previous literature, including gender, race, first generation student status, BMI and family financial support. Results: Participants were primarily white (71.8%), non-Hispanic (96.1%) American young adult women (77.2%). FI prevalence was 26.6% at phase 1 and fell to 22.1% at phase 2, although only 1 participant changed status. An adjusted regression model showed that FI participants were more likely to consume ½ cup fewer fruits and vegetables, to have worse overall sleep quality, to have worse stress, depression and anxiety symptoms, and to have worse attention span in class. FI was not associated with self-report or university-report GPA, frequency of missing class or self-report ability to understand class material. In all cases, baseline food insecurity status did not predict outcome changes from phase 1 to phase 2. Conclusion: FI prevalence did not change significantly between phases, but this may be attributable to the timing of Fall break. FI was not significantly predictive of a change in GPA, potentially indicating GPA is not a sensitive enough predictor for change over one semester. Despite our sample being at decreased risk for FI and our use of the 10-item USDA Adult Food Security Survey Module, FI prevalence was still higher than the national household average and within the previously identified range of food insecurity prevalence seen in college students. Additionally, further research is needed to determine why FI students have lower GPAs on average, since there was no difference in missing class or self-reported ability to keep up with material.