I entered this research space with the belief that as wearable technologies became more advanced and widespread, they could potentially replace self-reported measures in mental health research. Initially, I thought the focus would be straightforward—examining the relationship between sleep quality and depression. However, the data revealed a more complex narrative.

The Initial Hypothesis: Wearables as a Research Tool

At the outset, my goal was to utilize physiological sleep quality from passive sensor data on smartphones and smartwatches to predict self-reported sleep quality. Given the sophistication of these devices, I anticipated strong model performance. Surprisingly, this was not the case. This initial analysis eventually formed the basis of Figure 6 in our final paper.

What I expected to be an exploration of two primary constructs—sleep quality and depression—evolved into a more nuanced investigation. It became clear that self-reported sleep quality and physiological sleep quality are distinct constructs, each interacting differently with depression symptoms.

The Unexpected Finding: Divergence Between Self-Reported and Objective Sleep Measures

One of the most unexpected outcomes was the limited power of objective sleep measures to detect self-reported sleep quality. This was particularly perplexing because sleep quality is a critical domain in depression research and is often assessed through self-report in high-quality studies.

But if physiological sleep quality from wearable devices doesn’t predict self-reported sleep quality, they can’t be used interchangeably when studying mental health conditions. This discrepancy raises the important question: what does each sleep measure tell us about depression? To address this, we examined how self-reported and passively collected sleep data independently relate to depression symptoms. Our findings revealed a much stronger association between self-reported sleep quality and depression symptoms compared to physiological sleep data. It appears that self-reported measures capture subjective experiences closely linked to depression, which may not always be reflected in objective physiological data.

This discrepancy not only challenged our assumptions but also prompted us to consider other dimensions that could provide additional clarity. One such dimension was neurocognitive performance, a well-documented area affected by both sleep and depression.

Incorporating Neurocognitive Performance

To further explore the overlap between self-reported sleep quality and depression, we included neurocognitive performance as an additional variable. This was measured using TestMyBrain, providing an objective assessment distinct from both self-report and passive sensing. Given that neurocognitive performance often declines with depression and is influenced by sleep, we anticipated clear associations.

Interestingly, only self-reported sleep quality showed statistically significant correlations with neurocognitive performance, while physiological sleep data did not. This may be due to limitations in sample size or data quality. Nonetheless, the results underscore the importance of subjective sleep assessments in understanding depression-related outcomes such as cognitive performance.

The Role of Wearable Sleep Data

While consumer devices like smartwatches provide generally reliable sleep annotations, they are not without limitations. Issues related to the accuracy of detected sleep onset latency and awakenings could partially account for the weaker associations observed with physiological data.

Embracing the Complexity of Mental Health Research

This research journey highlighted the complexity inherent in mental health studies. Wearable technologies offer promising tools, but they are not a comprehensive replacement for self-reported measures. Both subjective experiences and objective data provide valuable, yet distinct, insights.

Understanding depression requires a multifaceted approach. Although self-reported sleep quality may not always align with physiological measures, it remains a critical component—capturing the nuanced interplay between psychological and physiological processes that objective data alone cannot fully elucidate.

Moving forward, integrating self-reported and physiological data holds promise for developing more comprehensive mental health assessment tools. As wearable technology continues to evolve, so too will our ability to refine these methods, potentially bridging the current gaps between subjective experience and objective measurement.

View the full article here: https://rdcu.be/d9b1b