Background Meal regularity is an elusive construct of eating patterns. There were various methods to assess meal regularity/irregularity, but no validation has been done so far.

Objectives To evaluate the validity of a clock hour-based meal regularity questionnaire.

Methods The study recruited 125 healthy young adults. Participants were interviewed on three separate days within 2–3 months. The participants provided three days of 24-hour dietary recall, recalling time of eating occasions on the previous day of interview. The three non-consecutive days of dietary recalls were used as reference method assessing meal regularity. At the third visit, the participants also administered the clock hour-based meal regularity questionnaire about meals (breakfast, lunch, dinner) and snacks (morning snack, afternoon snack, and nighttime snack). Participants recalled the number of days per week eating the indicated meal/snack in the past three months, and recalled all the hours they had consumed the indicated meal/snack. The dietary recall data and meal regularity questionnaire data were respectively converted into Chaotic Eating Index (CEI) by Annie Zimmerman et al. The algorithm of Meal Regularity (mReg) Score by Barbara Lohse et al. was also used to summarize the meal regularity based on 3-day recalls. Pearson correlation coefficient was used to evaluate the consistency between meal regularity measured by questionnaire and 3-day recalls.

Results The participants were on average 22.5 years-old, and 58% of them were women. The mean (SD) of meal CEI assessed by questionnaire was 3.6(1.1), and the mean(SD) of meal CEI assessed by 3-day dietary recall was 2.2(0.4). Correlation coefficient between meal CEI by questionnaire and meal CEI by 3-day dietary recalls was 0.20 (95% CI: 0.03, 0.36; p=0.024). Correlation coefficient between CEI assessed by questionnaire and mReg assessed by 3-day recalls was -0.35 (95% CI: -0.50, -0.19; p<0.001).

Conclusion We introduced a clock hour-based questionnaire to evaluate people’s meal regularity. Its comparative validity was fair. For health survey and nutrition surveillance that always need convenient tools to measure population’s health and dietary behaviors, this clock hour-based questionnaire may be applied for assessing the population’s meal regularity status.