Abstract
The timing and types of food that people eat, along with associated daily fluctuations, can influence health and wellbeing. However, there are limited data on how eating patterns remain consistent over multiple days. Here we present an exploratory, cross-sectional analysis of over 20,000 adults who recorded more than 2.5 million food logs over 2 weeks using the myCircadianClock app. Our analysis reveals significant variability in food timing and diversity. The time window during which 95% of food and beverages were consumed ranged from 10 h 54 min for the lowest decile to over 16 h for the highest. The median number of unique food and beverage items consumed over the 2 weeks varied from 20 to 86, and only a subset was consistently eaten on multiple days. Many common foods were regularly consumed at specific times of the day, and factors like age, gender and work schedules influenced both eating patterns and food choices. These findings provide a foundation for using longitudinal food records in nutrition and lifestyle research to enhance our understanding of human behaviour and health.