Characterizing eating patterns: a comparison of eating occasion definitions.Am J Clin Nutr. 2015 Nov; 102(5):1229-37.AJ
To date, many approaches have been used to define eating occasions (EOs). A standard definition for EOs is required to facilitate further research.
In this study, we examined the influence of differing definitions of EOs on the characterization of eating patterns.
Cross-sectional dietary data from two 24-h recalls collected during the 2011-12 Australian National Nutrition and Physical Activity Survey (n = 5242 adults, aged ≥19 y) were analyzed. Eight definitions were applied: participant-identified, time-of-day, and 6 neutral definitions (individual EOs separated by different time intervals and/or an additional energy criterion of 210 kJ). Frequency of and total energy intake from meals, snacks, and all EOs were estimated, as appropriate. Differences were tested by using F tests, stratified by sex and age group. Agreement between different definitions of meal and snack frequencies was assessed by using intraclass correlation coefficients (ICCs). For each definition, linear regression was used to estimate the proportion of variance in total energy intake (kJ) and amount of food intake (g) predicted by frequency of EOs and meals and snacks.
Among both sexes and across all age groups, mean frequencies of meals differed between the participant-identified and time-of-day definitions (mean difference range = 0.1-0.3; P < 0.001). There were statistically significant differences between mean frequencies of EOs across the 6 neutral definitions (P < 0.001). There was good agreement for snacks (men: ICC = 0.89; women: ICC = 0.87) but not meal frequencies (men: ICC = 0.38; women: ICC = 0.36) between the participant-identified and time-of-day definitions. The neutral definition (15-min time interval plus energy criterion) best predicted variance in total energy intake (R(2) range = 19.3-27.8).
Different approaches to the definition of EOs affect how eating patterns are characterized, with the neutral definition best predicting variance in total energy intake. Further research that examines how different EO definitions affect associations with health outcomes is needed to develop consensus on a standard EO definition.