This one-year longitudinal study examined both subjective and objective adherence to oral appliance therapy in patients with obstructive sleep apnea, and identified several factors associated with adherence. Participants were not explicitly aware of the analytic comparison between objectively recorded and self-reported use, nor of the specific details of the one-year follow-up, which helped maintain a more natural pattern of MAD use.

Overall, adherence to OA therapy was suboptimal in this study, with fewer than half (49%) of participants achieving adequate adherence during the first three months, and 44% remaining adherent at the 12-month follow-up. Notably, distinct adherence patterns were observed between adherent and non-adherent users, and several factors influencing OA adherence were identified.

Adherence and usage patterns of OA therapy

Previous studies have reported a mean adherence of approximately 6–6.8 h per night for custom-made oral appliances [33]. However, in this study, the nightly use across all nights during the initial 3-months period was only 3.85 h/night (IQR: 2.38–6.20). This relatively low overall adherence appears to be driven primarily by infrequent usage (median: 3.7 days/week; IQR: 2.3–5.9) rather than by the duration of wear on nights used (median: 7.05 h; IQR: 6.45–7.55). Several factors may account for this difference. The absence of regular behavioral intervention in our study may partly contribute [34]. Kwon et al. have reported that using a mobile app for monitoring and providing daily reminders can significantly increase the number of wearing days for MADs [13]. Given the low wearing frequency observed in our study, we recommend incorporating reminders via alarms, mobile apps, or emails as part of routine care to prompt patients to use the appliance. Secondly, patient tolerance and adherence may vary depending on the specific OA design used. the relatively low adherence observed in this study should be cautiously generalized to other OA therapies. Future large-scale prospective studies are needed to more comprehensively evaluate OA adherence and its influencing factors. Besides, the generally shorter sleep duration among the Chinese population may also play a role in the relatively short nightly wear duration [35, 36].

A recent Canadian study reported fairly good adherence to OA early on but with a significant decrease over the first few months that seemed to stabilize around six months [14], suggesting that long-term, objective data are essential for monitoring OA adherence. Interestingly, in our study, patients who showed good adherence had already developed stable wearing habits during the first treatment week, characterized by higher use frequency and longer nightly duration. This suggests that, despite the multifactorial nature of adherence, a subset of patients experiences minimal early barriers. Similar findings have been widely reported in the CPAP literature. Evidence indicates that usage patterns established within the first week of therapy are highly predictive of long-term adherence, suggesting that adherence behavior is determined early in the course of treatment [7, 37, 38]. While non-adherent OA users appeared to experience “wearing fatigue”, exhibited lower initial use frequency, characterized by a gradual reduction in use when unsupervised, followed by a temporary increase shortly before clinic visits. This pattern highlights the importance of consistent education, scheduled follow-ups, and adequate social support.

Discrepancies between subjective and objective adherence

In this study, self-reported mean nightly use across all nights was overestimated by approximately 0.92 h at the three-month visit, which aligns with earlier studies reporting a 0.5–0.8-h overestimation [24, 39]. The discrepancy between self-reported adherence and objectively monitored adherence observed in this study suggests the presence of self-reporting bias. Participants were relatively accurate in reporting the duration of wear on nights used; but they tended to overestimate the frequency of use (i.e., the number of days worn). This was likely due to an intentional increase in device use shortly before the visit, followed by recall bias that reflected only recent behavior rather than the entire observation period.

This "white coat" adherence pattern and reliance on short-term memory warrants caution. Although most studies suggest that adherence to OA therapy is superior to CPAP [40, 41], raising hopes that it could address the lack of demonstrated cardiovascular benefits in CPAP studies [42,43,44,45], the widespread implementation of objective, long-term monitoring for oral appliances remains limited. To maximize cardiovascular benefits, OA management requires more comprehensive and accessible care and monitoring strategies.

Predictors of adherence to OA therapy in the first 3 months

Our study found that age was an important predictor of good adherence at both the 3-month and 12-month follow-ups. This suggests that OA should not be exclusively recommended to younger patients, as older individuals may also benefit from the therapy. Previous research on CPAP use has also suggested that older age may be associated with better treatment adherence [46,47,48]. The influence of age on adherence is complex. Younger individuals generally have fewer health concerns, greater access to alternative treatment information, and more variable daily schedules, which may reduce their motivation or ability to maintain regular OA use. While middle-aged and older adults may experience lower daily stress levels, have a greater awareness of health risks, and maintain more stable daily routines, all of which may contribute to better adherence. However, the role of age on OA adherence should be interpreted with caution. Most participants in our study were under 50 years old, an age range in which OSA prevalence is already high but oral health conditions and tolerance to OA therapy remain relatively favorable [49, 50]. With increasing age, dental conditions become more complex, with higher risks of caries, periodontitis, and tooth loss, which can compromise OA retention and comfort. In addition, systemic comorbidities and psychological symptoms may become more prominent as age increases and may further influence adherence. Therefore, studies that include a broader age range are needed to clarify how aging affects OA adherence across different stages of life.

In the first 3-month follow-up, patients with more severe daytime sleepiness before OA therapy tended to show poorer short-term adherence, although this difference was no longer evident at 12 months. This highlights the importance of providing targeted support for OSA patients with excessive daytime sleepiness during the early phase of OA use. Evidence from CPAP studies has been mixed, with some reports suggesting poorer adherence among patients with excessive daytime sleepiness, while others found no such association [48, 51]. Based on the adherence patterns observed in the sleepiness-comorbid OSA subtype, we speculate that patients with high ESS scores may initially experience limited symptomatic improvement, but as nightly use accumulates and symptoms gradually improve, they may receive positive feedback that enhances their motivation to continue using the device. In our study, mandibular advancement was not specifically adjusted according to the severity of daytime sleepiness. Future studies may consider increasing the advancement for this OSA–sleepiness subtype, combining this with enhanced education and encouragement, to further explore strategies to improve adherence.

Predictors of adherence to OA therapy at 12-month follow up

Our study indicated that patients with a hyperdivergent skeletal pattern tended to exhibit poorer adherence at 12-month follow up, providing a novel perspective on the potential role of craniofacial morphology in OA use. Since mandibular retrognathia and hyperdivergent facial types are known risk factors for OSA, the influence of craniofacial morphology on OA uses warrants attention [52]. Previous studies have found that the mandibular plane angle influences OA efficacy: patients with a smaller mandibular plane angle (hypodivergent) generally achieve better outcomes with fewer side effects [53], whereas hyperdivergent patients often require greater mandibular advancement for effective AHI reduction [22]. The prevalence of obstructive sleep apnea is among the highest worldwide in the Chinese population [54]. Moreover, craniofacial characteristics such as mandibular retrusion and hyperdivergent skeletal patterns are more commonly observed in Chinese patients with OSA. The present findings extend existing evidence by highlighting the potential influence of craniofacial morphology on adherence to oral appliance therapy, and suggest that a more comprehensive and cautious evaluation may be warranted before initiating OA treatment in patients with hyperdivergent skeletal patterns.

In this study, personality traits were not identified as predictors of adherence. Although personality influences behavioral tendencies through stable psychological patterns, adherence may be more strongly shaped by factors such as perceived disease risk, psychological state, and social support, and may require a longer observation period to fully evaluate [55, 56].

Strengths and limitations

To our knowledge, this is the first study to report objectively measured adherence to oral appliance therapy in a Chinese population. The use of embedded sensors enabled objective monitoring of OA use and allowed for direct comparison between subjective and objective adherence measures, providing a more comprehensive assessment of adherence behavior.

Several limitations should be acknowledged. First, objective adherence data were collected only during the initial three months, and sensor-based monitoring was not continued throughout the entire follow-up period. Adherence at 12-month was assessed via telephone follow-up, which may be less comprehensive than in-person clinical visits. Secondly, discomfort and other side effects associated with OA use were not recorded. These factors may affect overall adherence and should be comprehensively evaluated in future studies. In addition, adherence was not recorded using sleep diaries, which might have provided complementary contextual information. Furthermore, the study population consisted exclusively of Chinese patients, which may limit the generalizability of the findings to other ethnic groups. Finally, very elderly patients were underrepresented in this cohort, and adherence patterns in older populations warrant further investigation.