A study using actigraphy finds that sleep consistency offers better predictive power of preterm birth than average sleep duration.
Key takeaways:
- An interdisciplinary team analyzed sleep data from 665 pregnant patients using wrist-worn actigraphs to identify risk factors for preterm birth.
- Machine learning models revealed that variability in sleep patterns is a stronger predictor of preterm birth than average sleep metrics.
- The findings suggest that promoting consistent sleep schedules could serve as a potential intervention for patients at risk.
- The study focused on identifying signals in the first two trimesters, addressing a lack of predictive tools for early pregnancy.
An interdisciplinary research team at Washington University in St Louis has identified that variability in sleep patterns among pregnant individuals can effectively predict preterm birth, potentially offering a new avenue for early intervention.
Using machine learning models to analyze data from a clinically validated wearable device, the researchers found that inconsistent sleep schedules were more indicative of preterm birth risk than the total amount of sleep obtained. The findings were published in npj Women’s Health.
While disrupted sleep is a known predictor of delivery before 37 weeks, previous reliance on self-reported data obscured the specific reasons behind the correlation.
Ben Warner, a doctoral student in the McKelvey School of Engineering, and Peinan Zhao, assistant professor of obstetrics & gynecology at WashU Medicine, led the study. They analyzed data from 665 patients involved in a 2014 cohort study at Washington University in St Louis and BJC HealthCare. Approximately 14% of the cohort experienced a preterm birth.
Participants wore an actigraph wristwatch that measured body movements for roughly two-week periods during the first two trimesters. This allowed the team to extract daily patterns regarding sleep length, sleep and wake times, and movement during sleep.
“We found that measures of sleep are decently predictive of preterm birth,” says Warner in a release. “Variability in sleep patterns tends to be a stronger predictor of preterm birth than average sleep metrics, and getting consistent sleep is more important than getting good sleep on average.”
The researchers combined the actigraphy data with survey responses and utilized machine learning models to determine the impact of sleep patterns on birth outcomes. Chenyang Lu, the Fullgraf Professor in the Department of Computer Science & Engineering, noted the utility of AI in processing this type of information.
“Raw data from wearables can be very messy, but using a healthy combination of statistical methods, AI, and clinical knowledge, researchers can extract important clinical insights,” says Lu, who also serves as director of the university’s AI for Health Institute, in a release. “Then AI scientists and clinicians work together to extract the insights from these very complex data from the real world and get meaningful insights from it.”
The study aimed to identify clinically significant associations using intentionally simple models. According to Zhao, the machine learning approach outperformed traditional statistical models.
“We can look at the result on importance of individual variables: How much does this variable contribute to the predictive model, and how much does it affect the final result?” says Zhao in a release. “Based on that, a direction of a potential intervention is to promote a more consistent sleep schedule.”
Sarah England, the Alan A. and Edith L. Wolff Professor and vice chair for research of obstetrics and gynecology at WashU Medicine, emphasized the importance of finding signals before 20 weeks of pregnancy. While many pregnant individuals report disrupted sleep in the third trimester, early prediction is crucial for risk management.
“There is no intervention because we can’t predict who’s going to have a preterm birth,” says England in a release. “We’re hoping that this will be much more helpful in getting predictive power of women who are going to be at higher risk.”
The team plans to validate these results in other populations at different academic medical centers.
“This research highlights the collaborations between engineering and obstetrics and gynecology,” says England in a release. “Many people don’t see natural connections between engineering and the field of reproduction, and this is another example of a perfect way for engineers to interact with researchers in our field.”
