Network analysis identifies key variables of impaired sleep health
Excessive daytime sleepiness, nonrestorative sleep and sleep behaviors all play a significant role in sleep health, according to results of a network analysis of sleep variables published in Sleep Epidemiology.
“As with other health behaviors, sleep health was traditionally modeled using psychosocial techniques such as the Health Belief Model or the Theory of Planned Behavior,” Gauld Christophe, of the department of psychiatry at the University of Grenoble in France, and colleagues wrote. “These techniques are particularly interesting for sleep behavior change. However, such psychosocial modeling is unable to model the impact of the multiple symptoms related to sleep disorders, sleep disturbance, as well as comorbidities on sleep health.”
According to the researchers, network analysis has proved beneficial in other health domains for integrating multifactorial variables; however, it has yet to be used for sleep health. Recent data suggest network analysis has significant potential in this area.
Christophe and colleagues sought to examine sleep health using a multivariable network analysis approach. They analyzed 39 sleep variables, which they extracted via an online questionnaire, related to symptoms of sleep disorders, sleep disturbances, sleep behaviors and comorbidities. Data were obtained via a large French cohort (n = 35,808; mean age, 42.7 years) of individuals concerned about their sleep.
Of the 39 sleep variables, results showed nonrestorative sleep (categorized under sleep disturbances), excessive daytime sleepiness (categorized under symptoms of sleep disorders), circadian irregularity and chronic sleep deprivation (both categorized under sleep behaviors) represented the four most significant variables with a strong effect in the network.
“Even if a more systematic collection of sleep variables could have been used, the present study encourages further research: i) to promote the use of consensual and interculturally validated scale exploring a large set of sleep variables, ii) to provide food for thought for better systematizing and organizing sleep variables, iii) to better examine the interest of network analysis as an integrative sleep modeling and iv) to improve sleep prevention strategies,” Christophe and colleagues wrote.
The authors report no relevant financial disclosures.