Content area
Abstract
Purpose
Human appetite hormones are affected by various factors and there is a lack of detailed evidence on the impact of eating frequency on self-reported and serum markers of appetite. If self-reported measures of appetite and serum biomarkers are dysregulated, imbalanced caloric intake could occur, leading to related significant health problems. This study seeks to identify whether three or six meals a day has an impact on the relationships and differences of self-reported measures or serum biomarkers of appetite, along with the concordance of the relationships between the two, while the participants are in the fasted or semi-fasted state. This topic is a part of a data analysis of the Frequency of Eating and Satiety Hormones (FRESH) randomized cross-over clinical trial.
Methods
The FRESH randomized cross-over clinical trial enrolled 50 participants aged 18-50 who completed three weeks of either high eating frequency (6 eating sessions/day) or low eating frequency (3 eating sessions/day), and then swapped to the other eating frequency after a two-week wash-out period. Participants selected and prepared their own foods that followed a study protocol-directed meal and eating plan throughout the intervention. The protocol-directed eating plan offered individualized daily energy intakes to maintain baseline body weight throughout the study and both macronutrients and micronutrient intakes were consistent across both eating frequency periods. Self-reported measures of appetite were “Hunger”, “Full”, and “Desire” using a standardized 0-100 visual analog scale and plasma biomarkers of appetite were the hormones ghrelin (pg/mL), leptin (ng/mL), and peptide-YY (pg/mL). Paired t-tests were used to measure differences between continuous self-reported and plasma biomarker variables and Fisher Z transformation tests for differences between correlations. Appetite measurements were collected with blood draws and self-report surveys at the end of each eating frequency period in the morning in a fasted state for the perceived appetite variables, ghrelin, and leptin. Peptide-YY was then collected in the semi-fasted state one half-hour later after a standardized breakfast for those in the low eating frequency period, or a standardized breakfast that was one-half the size for those in the high eating frequency period. Results of the continuous variable analysis were also stratified into low vs high physical activity, split by the median, in order to determine whether physical activity had an effect modification on the data.
Results
Eating frequency had no effect on plasma ghrelin and leptin levels and had no effect on any of the self-reported measures of appetite (p>0.05 for all). High eating frequency led to significantly lower (9.7%) plasma peptide-YY (p<0.01) compared to the low eating frequency. No differences were discovered in concordance of self-reported and serum biomarkers of appetite across low vs high eating frequency, and all correlations found were of either weak or very weak strength. When the continuous variable differences were stratified by physical activity, an effect modification was found in that only those in the high physical activity stratum had significantly lower peptide-YY in the high eating frequency group compared to the low eating frequency group.
Conclusion
Three versus six periods of daily eating frequency did not play a significant role in overall appetite regulation based on the fasted plasma hormone values and perceived appetite values reported. While peptide-YY was significantly higher in the low eating frequency group, this was expected due to the sensitivity of peptide-YY to meal intakes with the blood draw taking place after eating a larger or smaller meal in the respective eating frequencies. Because all correlations between fasted plasma hormones and perceived appetite were either weak or very weak, currently commonly used perceived appetite scale measures may not accurately reflect biological appetite. However, some specific expected relationships were seen with a closer categorical analysis, demonstrating some coordination between the self-reported and plasma biomarker variables. Further research will be required to identify how eating frequency and appetite relationships could play a role in human health outcomes.
