Beyond composite scores in chronotype assessment: item-level predictive patterns in the Morningness–Eveningness Questionnaire

Abstract Chronotype represents individual differences in circadian preferences that influence sleep-wake patterns, cognitive performance, and clinical outcomes across the lifespan. Given its growing relevance for clinical research and application, efficient and reliable chronotype assessment is essential. However, widely used tools like the Morningness–Eveningness-Questionnaire (MEQ) rely on composite scores, despite methodological concerns about multidimensionality and unequal item contributions. Addressing this limitation, the current study applied machine learning techniques to enhance theoretical understanding and empirical precision in chronotype assessment. Using the German MEQ in the Dortmund Vital Study (ClinicalTrials.gov NCT05155397), a prospective cohort study on healthy cognitive aging, we identified item-level predictive hierarchies and response patterns distinguishing morning, neutral, and evening types. Item 19 (“Which chronotype do you think you are?“) demonstrated exceptional predictive utility, showing threefold greater importance than any other item. This metacognitive self-assessment appears to capture relatively accurate chronotype identification rooted in lived experience. Chronotype-specific analyses revealed distinct predictive patterns across types, each relying on different item combinations for optimal classification. Partial dependence analysis identified non-linear response patterns including sigmoid curves, threshold effects, and plateau regions. A six-item combination achieved robust overall classification, potentially reducing assessment burden by 70%. Overall, this research advances chronotype assessment by uncovering non-linear and heterogeneous classification mechanisms to circadian preference. The findings provide evidence-based foundations for developing abbreviated screening tools for primary care consultations, large-scale epidemiological studies, and clinical contexts. An open-access R tutorial ensures reproducibility and facilitates adaptation across diverse populations and assessment instruments, supporting widespread implementation in research and practice.