Wearables in Pilot Studies

Table of Contents

  1. Sleep measures from wearables (e.g. REM, Wake, Total Sleep Time, etc)
  2. Cardiovascular measures from a smartwatch and smartring (HR, HRV, SpO2, etc)
  3. Raw sensor data from wearables for studies with Older Adults

Wearable devices can offer invaluable insights into physiological and behavioral patterns that are critical for understanding and managing AD/ADRD in the home setting. These devices, ranging from smartwatches to specialized health trackers, are increasingly used to monitor various health metrics such as physical activity, sleep patterns, heart rate, and heart rate variability. However, the effectiveness of these tools hinges on the accuracy and reliability of the data they capture.

One key gap is validation studies often don’t include older adults, leading to gaps in data applicability. Zinzuwadia and Singh, in The Lancet Digital Health [1] raised concerns about the bias and inequity in wearable device studies, notably lacking representation from diverse demographic backgrounds including older populations. These populations stand to benefit significantly from advancements in wearable technology for monitoring symptoms such as cognitive decline, changes in social behavior, and physical activity. However, validation studies frequently rely on data from younger, healthier cohorts. For example, in studies like the Apple Heart Study [2], the mean participant age was around 41, with less than 6% over the age of 65, resulting in diminished applicability of the findings to the elderly. This discrepancy is crucial since the device’s effectiveness, such as its sensitivity in detecting atrial fibrillation, notably decreased in a subsequent validation study with an older cohort with a mean age of 76. Thus, there is a critical need for research methodologies that directly address and adapt to the capabilities and needs of aging populations with cognitive impairments.

While there are limited studies validating wearable devices specifically for older adults, numerous validation studies exist focusing on younger cohorts. The results from these studies are still informative as they demonstrate the relative validity of different measures within a healthy population, providing a foundational basis for further research tailored to older adults. We describe some of this research in this section.

Sleep measures from wearables (e.g. REM, Wake, Total Sleep Time, etc)

Sleep wearables offer a non-intrusive method to monitor sleep patterns and disturbances, which are often prevalent in older adults, especially those with cognitive impairments. These devices range from specialized sleep trackers to multifunctional smartwatches that include sleep tracking as a feature. The accuracy of sleep wearables in detecting sleep stages, duration, and quality is crucial for their application in clinical research and patient care. This section summarizes current findings on the validity of sleep wearables and discusses their potential in enhancing the understanding of sleep-related disorders in aging populations. Validation studies on sleep wearables

Cardiovascular measures from a smartwatch and smartring (HR, HRV, SpO2, etc)

Smartwatches have gained popularity for their ability to monitor vital health metrics continuously. They offer potential benefits in tracking heart rate, physical activity, and other health indicators critical in managing common health conditions for older adults. However, their effectiveness and accuracy can vary widely between models and manufacturers. This section provides an overview of how smartwatches perform in clinical and everyday settings, particularly focusing on their precision in capturing heart rate (HR), heart rate variability (HRV) and other cardiovascular signals. Detailed exploration of smartwatch validation studies

Raw sensor data from wearables for studies with Older Adults

For pilots interested in improving the validity of health measures from wearable devices specifically for older adults, it is essential to prioritize the collection of raw data. Many popular devices, such as Fitbit, Oura, and Apple Watch, typically provide processed data that may already incorporate biases derived from their predominant use in younger, healthier populations. To ensure the accuracy and applicability of your research findings to older adults, consider utilizing devices like the Empatica E4. This device allows for the collection of raw physiological data, which is necessary when developing less biased algorithms tailored to older adult populations. Further information on Empatica E4

References

[1] Guu, Ta-Wei, Marijn Muurling, Zunera Khan, Chris Kalafatis, Dag Aarsland, and Anna-Katharine Brem. “Wearable devices: underrepresentation in the ageing society.” The Lancet Digital Health 5, no. 6 (2023): e336-e337.

[2] Perez MV, Mahaffey KW, Hedlin H, et al. Large-scale assessment of a smartwatch to identify atrial fibrillation. N Engl J Med 2019; 381: 1909–17.

Table of contents