Tracking behavioural differences across chronotypes: A case study in Finland using Oura rings

Abstract

Non-invasive mobile wearables like fitness trackers, smartwatches and rings allow for an easier and relatively less expensive approach to study everyday human behaviour when compared to traditional longitudinal methods. Here we have utilised smart rings manufactured by Oura to obtain granular data from nineteen healthy participants over the time span of one year (October 2023 - September 2024) along with monthly surveys for nine months to track their subjective stress during the study. We have investigated longitudinal sleep and activity patterns of three chronotype groups of participating individuals: morning type (MT), neither type (NT) and evening type (ET). We find that while ET individuals do not seem to lead as healthy life as the MT or NT individuals in terms of overall sleep and activity, they seem to have significantly improved their habits during the duration of the study. The activity in all chronotype groups varies across the year with ET showing an increasing trend. Furthermore, we also show that the Daylight Saving Time changes affect the MT and ET chronotypes, oppositely. Finally, using a mixed-effects regression model, we show that an individual's perceived stress is significantly associated with their time spent in bed during the night time sleep, monthly survey response time, and chronotype, while accounting for individual variability.

Figures (10)

• Figure 1: Mid sleep time (MST) and sleep scores of the chronotypes. (a) Evening-types (ET) in blue have later MSTs while morning-types (MT) in red have earlier MSTs when compared with neither-types (NT) in yellow. We observe seasonal changes in all chronotypes for winter and summer holidays, showing later MSTs as can be seen from Fig. \ref{['fig:sesonal_var']} in the SI. (b) The sleep scores of the Oura ring takes into account several different sleep metrics for scoring as displayed here. ETs are observed to have poorer sleep scores when compared to other chronotypes, but their scores improve considerably with time. The shaded region around the line plots show the 95$\%$ confidence interval of the data.
• Figure 2: Sleep diversity index (SDI). The SDI is calculated using Shannon entropy from Eq.\ref{['eq:shannon_entropy']} and \ref{['eq:sdi']} utilising sleep hypnograms (see Fig. \ref{['fig:sleep_hypnogram']}) such that the number ranges from 1 to 100. Higher values indicate a more fragmented sleep during the night time. We find that NTs display lowermost values throughout the (a) year, as well as (b) on a weekly basis indicating better sleep quality than both MTs and ETs. The shaded regions represent $95\%$ confidence interval. The ETs had poorer values of SDI at the start of the study, but improved with time. We also observe that the ETs regularly have poorer quality sleep towards the later part of the week.
• Figure 3: Daylight saving time. Differences in SDI and sleep latency among the chronotypes before and after the time is changed in autumn and in spring. In autumn, the latency of the ETs decreases and that of the MTs, increases. This is also observed for SDI values before and after the DST change in the autumn. We do not find much differences in the latency during spring DST change, but, the SDI of MTs increases again during this time. The most significant changes ($p<0.05$) occur in sleep latency for MTs and NTs in the autumn DST.
• Figure 4: Activity scores and steps. The monthly variation of the activity scores throughout the study period is shown in (a) along with the weekly variation in the inset of the figure. The ETs have low scores in the beginning but their scores are observed to have increased with time. Peaks in activity scores of the MTs and ETs coincide with the annual summer and winter activity schedules of Finland. A similar trend is observed in the average monthly steps taken by the participants as shown in Fig. (b). NTs have have maintained higher number steps when compared with other chronotypes and MTs and ETs are observed to have improved their average of steps by the end of the study.
• Figure 5: Example of a sleep hypnogram provided by Oura ring. The different stages of sleep recorded during night time sleep is displayed. Each time point in the x-axis represents an interval of five minutes. The four stages of sleep are shown in the y-axis.