Eating Behavior in College-Students: TFEQ R-18 and Qualitative Perceptions of Cell-Phone Use for Recording Diet

Purpose: Worldwide, an estimated 600 million adults are obese and 1.9 billion are overweight (World Health Organization, 2014).  The etiology of obesity and overweight is complex, reflecting a combination of bio-behavioral, environmental, and social factors.  Individual variations in cognition regarding diet may be partially explained by personal tendencies toward visual food cues and the propensity to act on food cues when food is available. While advances in neuroscience are laying the foundation for a cognitive phenotype of obesity and overweight, a small body of evidence suggests a relationship between how individuals think and feel about food, personal characteristics, and the review of personal food images (Carnell, Benson, Ochner, Geliebeter, 2011; Long, et al., 2013; Doumit, et al, 2015).  The purpose of this project is to report phase one of a study designed to examine the relationship of cognitive restraint, emotional eating, and uncontrolled eating behavior with personal characteristics and perceptions of personal dietary images taken using cell phones to record diet in a college-age population.   

Methods:

This phase of the study used a hierarchical multiple regression, correlational, mixed-methods design.  The research question for this study is “What is the relationship of eating domains in the TFEQ R-18 with personal characteristics and the perception of diet after review of personalized images using cell phones?” Twenty-eight college students were recruited from a single, large public university in the southwestern U.S.  After study attrition, 27 subjects completed phase one of the study.  Following informed consent, subjects self-reported height, weight, and demographic data, and completed the TFEQ R-18 instrument and 30 minutes of training for three days of dietary recording using an online tracking website and personal cell phones to record digital images of diet. Phase two (neuroimaging) was completed within two to three-weeks after the three days of diet recording.  A private, 15-minute debriefing session was held with each participant after his/her neuroimaging session in which narrative responses were collected exploring their perception of use of the cell phone to record and review of food images.  Data were cleaned after entry into SPSS version 22 for descriptive analysis.  Hierarchal multiple regression analyses were conducted investigating the relationships between the cognitive restraint (CR), emotional eating (EE), and uncontrolled eating (UE) subscales from the TFEQ R-18 with gender, age, ethnicity, BMI, and weight category.  Associations between age, BMI and TFEQ R-18 subscales also were explored. Narrative responses from individual debriefing sessions were combined by question and entered into a Word Cloud to examine word frequencies and patterns, and to determine recurring themes.

Results:

Sample demographics indicated fifteen (55.6%) were male and 12 (44.4%) were female.  Seventeen (63%) were Caucasian, 3(1.1%) were African American, 5 (22.2%) were American Indian/Alaskan and one did not specify ethnicity.  Age ranged from 19-28 with a mean age of 22.3 years.  One (3.7%) was underweight, 16 (59.3%) were healthy weight, 6 (22.2%) were overweight, and 4 (14.8%) were classified as obese. 

The linear mixed-effects model found that personal attributes (weight status, gender, and BMI) and ethnicity explained a significant amount of the variance in the Cognitive Restraint subscale  (F(4, 20)=3.75, p=.020, R²=.314, R²_Adjusted=.315).  Statistically significant correlations were found with age and specific questions the EE subscale: items 4 – “Sometimes when I start eating, I just can’t seem to stop” and item 14 – “How often do you feel hungry” (r=.395, p=0.046; r=-.389, p=0.050).   No statistically significant associations were found with age or BMI and individual items in CR or UE domains or in CR, EE, or UE subscale totals. Recurring themes from the narrative analysis include “increased awareness” of food intake and “ease/easy” functionality of digital images from cell phones to record diet.   

Conclusion:

Narrative data obtained during the debriefing suggest visualization of personal diet heightened awareness of foods consumed.  Review of digital pictures from cell phone cameras was considered an easy and functional adjunct to dietary memory.  The TFEQ R-18 findings from phase one of this preliminary study are similar to those reported prior suggesting the limited relationships between subject characteristics and TFEQ R-18 domains to be sample-dependent.  In this study the hierarchical multiple regression analysis suggests that the significant combined effects of demographic factors and ethnicity explain the variability in the CR subscale, thus adding to the body of evidence investigating eating behavior in this population.  Prior work on the TFQE R-18 and the findings of phase one of this study suggest that the personal physiological mechanisms linked to how individuals view and feel about food remain enigmatic, emphasizing the need for continued inquiry at the biological level (phase two of this project).  Limitations in phase one of this preliminary study include the small sample size and exploratory nature of the associations being explored.