SleepScore Labs, which is powered by ResMed and makes a non-contact sleep improvement system for consumers, has identified 8 distinct sub-groups of sleepers based on the analysis of sleep data coupled with demographic, lifestyle, and basic health information. To identify these sleeper profiles, sleep scientists from SleepScore Labs analyzed self-report data, and combined this with nearly 2 million nights of objectively measured sleep data collected unobtrusively in real home environments.

Roy Raymann, PhD, vice president of sleep science and scientific affairs at SleepScore Labs recently presented the findings of this study, “Two Million Nights to Characterize Sleep Heterogeneity: What Objective and Self Report Big Data Tell Us,” at ESRS 2018, the 24th Congress of the European Sleep Research Society, in Basel, Switzerland.

Demographic details, overall health, and lifestyle characteristics all affect sleep. However, little is known about the complexities associated with these factors as they relate to sleep outcomes in real-life settings, and the assessment of sleep is mostly limited to a single self-report sleep measure in many cases.

“Analyzing sleep and environmental data objectively collected by our devices helps us uncover more information about sleep diversity in the population in real life. It can help us define and target specific groups with different issues in a more personalized way,” says Raymann, in a release. “These findings will aid in the development of more personalized behavior change guidelines, which in turn might make it easier to be compliant to the suggested behavioral change. The ultimate end goal is to help people achieve better sleep and spur the development of new interventions to help improve sleep.”

Various categorical variables reflecting demographic, health, and lifestyle factors (eg, age, BMI, smoking, exercise, blood pressure, stress) were analyzed using multiple correspondence analyses (MCA). This analysis was complemented with data stemming from almost 2 million nights of S+ by ResMed sleep recordings.

As a result, eight distinct clusters of sleepers were revealed including:

  1. active and healthy;
  2. older, with healthy lifestyle;
  3. stressed, tired but active patients without sleep issues;
  4. active and healthy but with restless sleep and sleep breathing issues;
  5. otherwise healthy with sleep disorders;
  6. stressed smokers;
  7. overweight and older with health issues and restless sleep; and
  8. overweight older patients with health issues and sleep disorders.

Notably, the two clusters containing mostly smokers and individuals with high BMI (indicating obesity), experienced the poorest sleep, thus showing a strong association between poor sleep and unhealthy lifestyle factors.

Also at ESRS 2018, representatives from ResMed and SleepScore Labs presented, “Age Related Sleep Stage Trends as Measured Using Remote Sleep Sensing Hardware.” Findings demonstrate that non-contact bio-motion sleep sensing technology is capable of capturing the age-related sleep stage trends. SleepScore says this lays the foundations for further improvement in the performance of sleep monitoring technologies and shows their validity for use beyond consumer sleep tracking.