Examining the Agreement Between Subjective and Objective Measures of Sleep: A Comparison of Munich Chronotype Questionnaire and Fitbit-Derived Sleep Metrics - PubMed (original) (raw)

Comparative Study

Examining the Agreement Between Subjective and Objective Measures of Sleep: A Comparison of Munich Chronotype Questionnaire and Fitbit-Derived Sleep Metrics

Kayla E Rohr et al. J Sleep Res. 2025 Oct.

Abstract

Understanding the relationship between subjective and objective sleep measures is essential for evaluating their agreement and utility. This study compared Munich Chronotype Questionnaire (MCTQ) and Fitbit metrics for sleep duration, sleep midpoint and social jetlag in 5252 participants from the Adolescent Brain Cognitive Development (ABCD) study. Linear and nonlinear models assessed relationships between Fitbit-derived and MCTQ-reported metrics, whilst moderation analyses examined the influence of age, sex, household income and BMI. A sensitivity analysis compared results pre- and post-COVID-19 to assess pandemic-related effects (pre-COVID n = 4451). Correlations were weak to moderate: r = 0.15-0.21 for sleep duration, r = 0.37-0.42 for sleep midpoint, and r = 0.12-0.16 for social jetlag. Quadratic and LOESS models confirmed nonlinear trends for sleep midpoint, with greater Fitbit-MCTQ divergence at extreme morningness or eveningness. Fitbit classified 63.2% of participants as having insufficient sleep, compared to 39.45% with MCTQ, suggesting Fitbit underestimates sleep duration. Bland-Altman plots confirmed MCTQ overestimation, especially for shorter sleepers. BMI was significantly associated with sleep duration and social jetlag, with higher BMI linked to shorter sleep and greater variability. Household income and BMI moderated specific sleep metrics, whilst age and sex did not significantly moderate any metric. Sensitivity analyses showed consistent results across pre- and post-COVID periods. Findings highlight stronger agreement for sleep midpoint than for sleep duration or social jetlag, with methodological differences driving discrepancies. The consistency across demographics and time periods supports the complementary use of Fitbit and MCTQ for adolescent sleep assessment.

Keywords: Fitbit; MCTQ; sleep metrics; sleep midpoint; social jetlag; subjective‐objective agreement.

Published 2025. This article is a U.S. Government work and is in the public domain in the USA.

PubMed Disclaimer

Conflict of interest statement

Declarations of Interest: The authors declare no conflicts of interest.

Figures

Figure 1.

Fitbit vs. MCTQ Sleep Duration. Scatterplot comparing Fitbit-derived and MCTQ-reported sleep duration (in hours). The blue solid line represents the linear regression fit, while the red dashed line represents the line of equality (y = x). Data points show individual observations, highlighting the overall weak positive relationship and variability between the two measures.

Figure 2.

Bland-Altman Plot: Sleep Duration. Bland-Altman plot showing the difference between Fitbit-derived and MCTQ-reported sleep duration (Fitbit - MCTQ) against the mean of the two measures. The blue dashed line represents the mean difference, while the red dotted lines indicate the limits of agreement (±1.96 SD). The plot highlights greater discrepancies at higher sleep durations, with subjective MCTQ values tending to overestimate compared to Fitbit measurements.

Figure 3.

Fitbit vs. MCTQ Sleep Midpoint. Scatterplot comparing Fitbit-derived and MCTQ-reported sleep midpoints. The blue solid line represents the linear regression fit, while the red dashed line indicates the LOESS curve, capturing potential nonlinear trends. The plot shows moderate agreement between the two measures, with divergence observed at extreme sleep midpoint values, suggesting that Fitbit and MCTQ may differ in their sensitivity to variations in sleep timing, particularly for individuals with later or earlier sleep midpoints. The MCTQ sleep midpoint was derived from self-reported sleep onset and wake times on free days, adjusted for accumulated sleep debt during the workweek. MCTQ sleep midpoint scores range from 16 to 40 hours, with higher values reflecting later sleep timing (e.g., a score of 25.3 corresponds to 1:20 a.m. the following day). The Fitbit-derived sleep midpoint was calculated using objectively measured sleep onset and offset on weekend days (i.e., the best approximation of free days) and aligned by centering with the MCTQ scale.

Figure 4.

Quadratic Fit for Sleep Midpoint. Scatterplot illustrating the relationship between Fitbit-derived and MCTQ-reported sleep midpoint with a quadratic regression fit (blue curve). The shaded area represents the confidence interval around the quadratic fit, highlighting a significant nonlinear association between the two measures. Divergences are evident at extreme sleep midpoint values.

Figure 5.

Fitbit vs. MCTQ Social Jetlag. Scatterplot depicting the relationship between Fitbit-derived and MCTQ-reported social jetlag. The blue solid line represents the linear regression fit, while the red dashed line represents the LOESS regression fit, capturing potential nonlinear trends. Weak alignment is observed between the two measures, with greater discrepancies at extreme values, suggesting that Fitbit and MCTQ may differ in their sensitivity to shifts in sleep timing across workdays and free days. The MCTQ social jetlag was calculated as the difference between the midpoint of sleep on workdays and free days, reflecting shifts in sleep timing due to external demands such as school schedules. Negative values indicate earlier sleep timing on free days, while positive values indicate delayed sleep on free days compared to workdays. The Fitbit-derived social jetlag was similarly defined but based on objectively measured sleep timing from weekend and weekday data, adjusted to align with MCTQ’s time scale. Differences become more pronounced at higher levels of social jetlag, where individuals with greater shifts in sleep timing may experience larger discrepancies between self-reported and device-based estimates.

Figure 6.

Moderation by age of (a) sleep duration, (b) sleep midpoint, and (c) social jetlag. This series of simple slope figures illustrates the moderation effect of age (9–13 years) on the relationship between different sleep measures derived from the Munich Chronotype Questionnaire (MCTQ) and Fitbit data. The first figure depicts the relationship between MCTQ Sleep Duration (hours) and Fitbit Sleep Duration (hours) moderated by age. The second figure shows the relationship between MCTQ Sleep Midpoint (hours) and Fitbit Sleep Midpoint (hours) moderated by age. The final figure presents the relationship between MCTQ Social Jetlag (hours) and Fitbit Social Jetlag (hours) moderated by age. No moderation effects are observed for sleep duration, sleep midpoint, or social jetlag, suggesting that the alignment between MCTQ and Fitbit measures remains stable across this age range.

Figure 6.

References

1. Anan YH, Kahn NF, Garrison MM, McCarty CA, & Richardson LP (2023). Associations between sleep duration and positive mental health screens during adolescent preventive visits in primary care. Academic Pediatrics, S1876–2859(23)00061-X. 10.1016/j.acap.2023.02.013 -DOI -PubMed
1. Armitage R, Trivedi M, Hoffmann R, & Rush AJ (1997). Relationship between objective and subjective sleep measures in depressed patients and healthy controls. Depression and Anxiety, 5(2), 97–102. 10.1002/(sici)1520-6394(1997)5:2<97::aid-da6>3.0.co;2-2 -DOI -PubMed
1. Baron KG, & Reid KJ (2014). Circadian misalignment and health. International Review of Psychiatry (Abingdon, England), 26(2), 139–154. 10.3109/09540261.2014.911149 -DOI -PMC -PubMed
1. Baseline Data Demographics 2.0. (n.d.). ABCD Study. Retrieved December 3, 2023, from https://abcdstudy.org/scientists/data-sharing/baseline-data-demographics...
1. Ben-Shachar MS, Lüdecke D, & Makowski D (2020). effectsize: Estimation of Effect Size Indices and Standardized Parameters. Journal of Open Source Software, 5(56), 2815. 10.21105/joss.02815 -DOI

Examining the Agreement Between Subjective and Objective Measures of Sleep: A Comparison of Munich Chronotype Questionnaire and Fitbit-Derived Sleep Metrics - PubMed (original) (raw)