1109 The Impact of Sleep Duration and Its Variability on the Human Phenome: Analysis of All of Us Wearables Data

Mason Manetta,Diego Mazzotti

DOI: https://doi.org/10.1093/sleep/zsae067.01109

IF: 6.313

2024-04-20

SLEEP

Abstract:Abstract Introduction Epidemiological evidence indicates that short, long and variable sleep duration are associated with cardiometabolic outcomes. However, due to the ubiquitous role of sleep, one would expect to see effects on a broad range of human diseases. Using electronic health records (EHR) and wearable sleep data from the All of Us Research Program, we conducted a cross-sectional analysis of sleep duration and variability with the risk of over 3530 chronic conditions across the human phenome. Methods All of Us participants (≥18 years) with available EHR data and at least 90 days of continuous monitoring using Fitbit devices under a ‘bring your own device’ model were included. Average (mean) and variability (standard deviation) of sleep duration were derived based on device-generated data. Chronic conditions were determined based on mapping of International Classification of Disease 10 codes available in the EHR to phecodes, an ontology representing clinically meaningful concepts. Associations between sleep traits and phecodes were assessed using multiple test corrected logistic regression adjusted for age, sex, and race. Results A total of 8,256 participants with available Fitbit data were included. Participants were 89% White, 72% female, 72% college educated and had median (IQR) age of 51 (36.6-52.0) years, BMI 28.3 (24.4-33.0) kg/m2, and 1,566 (860-2307) days with sleep measurements. The median average nightly sleep duration was 6.34 (5.67-6.91) hours and the median nightly sleep variability was 1.63 (1.33-1.95) hours. In adjusted analyses, participants with greater sleep duration had significantly lower odds of obesity (OR; 95%CI = 0.85; 0.80-0.91, obstructive sleep apnea (0.74; 0.74-0.88), type 2 diabetes (0.68, CI 0.58-0.80), and essential hypertension (0.86; 0.81-0.92). Conversely, participants with greater sleep duration variability had higher odds of major depressive disorder (2.15; 1.84-2.51), anxiety disorder (1.75; 1.51-2.03), gastroesophageal reflux disease (1.61; 1.39-1.86), and abnormal electrocardiogram (1.80; 1.47-2.19). Conclusion Our findings suggest important associations between sleep duration patterns and chronic conditions, emphasizing the importance of sleep health in disease prevention and management. Important limitations include analysis of a selected population with access to wearable devices, which may not represent the general population enrolled in All of Us. Support (if any) All of Us Research Scholars Program

neurosciences,clinical neurology

What problem does this paper attempt to address?

The proceedings explore multiple research questions related to sleep, specifically including the following aspects:

1. **Accurate Collection and Processing of Sleep Diary Data**:
- **Problem**: In clinical sleep science research, although sleep diaries are the gold standard measurement tool for self-reported sleep parameters, their effectiveness depends on accurate data input and consistency in standardized terminology. This paper aims to improve the collection and processing methods of sleep diary data by analyzing common participant errors.
- **Method**: The study is based on data from a randomized controlled trial of the effectiveness of cognitive behavioral therapy for insomnia in the RESTING project, analyzing consensus sleep diaries completed by 245 insomnia patients at baseline, with data collected via REDCap.
- **Conclusion**: Despite participant training and data validation, 5.7% of sleep diaries still contained errors that required post-hoc correction. Common errors included confusion between rise time (RT) and wake time (WT), bedtime (BT) and lights out time (LO), etc. These findings help in developing standardized data cleaning scripts and improving data validation processes.

2. **Impact of Previous Night's Sleep on Next Day's Symptoms**:
- **Problem**: Investigating how the quality of the previous night's sleep affects the subjective symptom performance of U.S. military personnel the next day.
- **Method**: 270 military personnel with sleep issues were recruited for a 10-day remote monitoring assessment, using standardized sleep diaries and Fitbit Inspire 2 devices to record sleep conditions, and twice-daily symptom surveys to assess next-day symptoms.
- **Conclusion**: Previous night's sleep parameters (such as total sleep time, sleep efficiency, and sleep quality) significantly impacted next-day cognition, energy levels, and emotional state. These findings support the use of remote monitoring methods to predict treatment response and provide personalized care.

3. **Impact of Sleep Duration and Variability on Human Diseases**:
- **Problem**: Exploring the relationship between sleep duration and its variability with various chronic diseases.
- **Method**: Analyzed electronic health records and Fitbit device data from over 8,256 participants in the All of Us Research Program, assessing the association between sleep characteristics and chronic diseases.
- **Conclusion**: Longer sleep duration was associated with lower risks of obesity, obstructive sleep apnea, type 2 diabetes, and hypertension; while greater sleep duration variability was associated with higher risks of depression, anxiety, gastroesophageal reflux disease, and abnormal electrocardiograms.

4. **Using Wearable Devices to Promote Non-PAP Treatment Engagement**:
- **Problem**: Exploring how data generated by consumer-grade wearable devices can be used to improve the management of non-positive airway pressure (non-PAP) treatment for obstructive sleep apnea (OSA) patients.
- **Method**: In a quality improvement project conducted at a VA medical center, patients' blood oxygen saturation (SpO2) data were collected via Fitbit devices and combined with patient-reported outcomes to create reports that facilitate shared decision-making.
- **Conclusion**: Preliminary results showed that most patients expressed willingness to retry PAP treatment after seeing low SpO2 values. This approach can enhance patient self-management and engagement, thereby improving health outcomes.

5. **Relationship Between Technology Use and Sleep**:
- **Problem**: Examining nationwide habits of evening technology device use and its relationship with sleep quality.
- **Method**: Analyzed representative sample data collected by the National Sleep Foundation, assessing the use of various technology devices within one hour before bedtime and its impact on sleep time and quality.
- **Conclusion**: 90% of respondents used technology devices before bedtime, with usage locations both inside and outside the bedroom. Technology device use was associated with poorer sleep quality, with specific associations requiring further statistical analysis to determine.