Smartwatch-derived sleep estimates in individuals with probable sleep bruxism: an exploratory pilot study

Sleep bruxism (SB) is sleep-related movement disorder potentially affecting sleep architecture. This exploratory pilot study evaluated smartwatch-derived sleep estimates in young adults with a probable SB phenotype versus controls.

In this cross-sectional study, 30 age- and sex-matched participants (18-50 years) were evaluated. The probable SB phenotype was defined by a triple-validations approach consisting of combined self report and clinical signs agreed upon by two clinicians. Controls were age- and sex-matched. Sleep estimates were recorded at home for 7 consecutive nights using a consumer-grade smartwatch (Apple Watch Series 8); nights with < 240 min of recorded sleep or substantial signal loss were excluded and repeated. Group comparisons used appropriate parametric/nonparametric tests and effect sizes (Cohen's d).

Participants were analyzed (15 in probable SB group, 15 controls; 86.7% female in each). Total sleep time and light sleep estimates did not differ between groups. Compared with controls, the probable SB group had significantly reduced estimated durations of REM sleep (82.7 ± 18.9 vs. 111.5 ± 17.2 min; p < 0.001; d = 1.59) and deep sleep (49.1 ± 9.8 vs. 60.7 ± 6.9 min; p < 0.001; d = 1.37), with greater device-estimated sleep fragmentation (WASO median 27.0 vs. 3.0 min; p < 0.001; awakenings median 5 vs. 1; p < 0.001). Awakenings correlated positively with wake duration (r = 0.898) and negatively with REM duration (r = - 0.573) (both p < 0.001).

Individuals with probable SB exhibited altered device-estimated sleep patterns, featuring reduced deep/REM sleep and increased fragmentation. Consumer wearables may offer exploratory, adjunctive data on sleep continuity in dental practice but cannot replace physiological sleep staging. Findings warrant confirmation with concurrent polysomnography.

ClinicalTrials.gov; Registration No: NCT07453121. 25/02/2026.