Monitoring sleep in real-world conditions using low-cost technology tools

Monitoring sleep in real-world conditions requires bespoke equipment, e.g., Actigraphs or similar. However, such equipment is relatively expensive and not always available for large-scale field research. This study tests the possibility that sleep in real-world conditions can be monitored, sufficiently accurately, by tandems of commonly used smartphones (SP) and smartwatches (SW). 10 adult participants were asked to wear Actigraph, and SW for 30 consecutive nights. The accumulated records were analyzed using bi-variate statistics, mixed modeling and epoch-by-epoch analysis. A high degree of correspondence was found between Actigraph, SP/SW, and self-report records (R2 = 0.968–0.983). Although the mixed modeling also indicated high collinearity between Actigraphs and SP/SW tandems (b = 0.991; p < 0.001), it was also shown that SP/SW tandems add ⁓21.9 min to the Actigraph measurements while the self-reports were found to be even less accurate, adding ⁓39.9 min. Concurrently, the epoch-by-epoch analysis showed a good agreement between different types of measurements, varying between 81% and 100%. As we conclude, widely available and affordable SPs and SWs can help researchers to generate fairly reliable data for large-scale field studies, albeit measurement corrections need to be applied. Yet, estimates, obtained from sleep diaries, need to be treated with caution.