Sleep is a crucial physiological activity for humans, significantly contributing to the elimination of metabolites, growth and development, and the regulation of immune function; it also functions as a restorative neurobehavioral state (Buysse, 2014).Healthcare workers are susceptible to sleep disorders due to shift work and occupational stress, which significantly impacts the quality of healthcare services (Jaradat et al., 2020; Johnson et al., 2014; Sun et al., 2022).Numerous studies have indicated that the prevalence of sleep disorders among medical personnel ranges from 38.9 to 46.1% (Marvaldi et al., 2021; Pappa et al., 2020; Valente et al., 2019). Consequently, there is a need for further research and a comprehensive understanding of the sleep status and patterns of healthcare workers to furnish evidence for enhancing sleep quality.

Sleep patterns are defined as an individual’s sleep behaviors and habits over a specific period, encompassing the time of falling asleep and awakening, sleep duration, the regularity of sleep (e.g., circadian rhythms), and its quality (Kryger et al., 2011). Measurement of sleep patterns includes objective measures (e.g., polysomnography, activity monitors), sleep logs, and self-report questionnaires (e.g., Pittsburgh Sleep Quality Index (PSQI) and self-administered questionnaires). Polysomnography (PSG) is a comprehensive sleep monitoring technique that provides a detailed assessment of an individual’s sleep status by recording multiple physiological parameters. These parameters typically consist of electroencephalogram (EEG), eye movements (EOG), electromyogram (EMG), electrocardiogram (ECG), respiratory airflow, oxygen saturation, and other physiological data. PSG is regarded as the gold standard for the diagnosis of sleep disorders and the conduct of sleep research (Iber et al., 2007). In recent years, as technology has evolved, an increasing number of wearable devices have been utilized to measure sleep patterns and quality. In a study conducted in Spain, researchers assessed sleep patterns by utilizing the ActiGraph GT9X Link and Fitbit Charge HR worn on the non-dominant wrists of elementary school children, while parents recorded the times of going to bed and waking up (Brazendale et al., 2019). Sleep log methods are simple and cost-effective ways to track and assess sleep; however, because they rely on subjective self-ratings, they can often be inaccurate and incomplete. A Canadian cross-sectional study utilized a sleep log where stroke patients reported experiencing difficulty staying awake most or all of the time to measure sleep disturbances (Jeffers et al., 2023). Standardized self-administered questionnaires are also commonly used to measure sleep patterns. In a study involving Chinese patients with schizophrenia, sleep patterns were categorized into short, medium, and long sleep based on patients’ self-reported expected sleep duration and actual total sleep duration (Hou et al., 2017).

Sleep patterns can be characterized by various dimensions, including sleep latency, sleep duration, insomnia medication use, and daytime dysfunction. Sleep patterns evolve in various ways throughout the lifespan, leading to declines in both the quantity and quality of sleep. A study found that 9.3% of younger individuals reported falling asleep later, whereas 22.5% of older individuals reported less efficient or poorer quality sleep (Gadie et al., 2017). Significant gender differences in sleep patterns have been also observed, with women more likely to report disturbed sleep onset and men more likely to report nocturnal awakenings (Middelkoop et al., 1996). Furthermore, a British cohort study indicated that sleep proportion (the ratio of sleep duration to time in bed) is a useful indicator of sleep patterns in the general population (Leng et al., 2014). Nevertheless, limited research has focused on the underlying sleep patterns of healthcare workers. Therefore, there is a critical need for studies focused on healthcare workers, which are essential for developing targeted intervention strategies tailored to various sleep patterns to enhance sleep quality.

Established studies have demonstrated a potential causal relationship between sleep disorders and psychiatric symptoms in healthcare workers (Dong et al., 2017; Kalmbach et al., 2017; Liu et al., 2022; Sørengaard et al., 2019). A meta-analysis revealed that healthcare workers with sleep disorders faced a 3.74-fold increased risk of psychiatric symptoms compared to those without sleep disorders (Liu et al., 2022).These results may stem from the vital functions of sleep in the human body, which include maintaining consciousness and cognitive function, regulating biorhythms, enhancing immune defenses, and alleviating stress (Ding et al., 2022; Dinges et al., 1995). Therefore, understanding the complexities of sleep patterns is essential for elucidating their impact on psychiatric symptoms and overall wellbeing. Latent profile analysis (LPA) has been widely utilized in psychological and humanities research to identify individuals with varying personal attributes (Hensel, 2020).This approach can assist in exploring the sleep patterns of healthcare workers, further delineate distinct subgroups with unique sleep patterns, and facilitate the development of targeted intervention strategies for sleep management. The aims of this study were: (1) to identify the sleep patterns of healthcare workers through LPA, indexed by the seven sleep dimensions of the Pittsburgh Sleep Quality Index (PSQI); (2) to understand the predictors of varying sleep patterns in healthcare workers; and (3) to examine the relationship between sleep patterns and the psychiatric symptoms among healthcare workers.

In this study, 31.4% of healthcare workers reported poor sleep quality, which is lower than the findings of a systematic evaluation of sleep disorders among Chinese healthcare workers before the COVID-19 epidemic (39.2%) (Qiu et al., 2020),but higher than the prevalence reported in the general population (Adams et al., 2017; Morin and Jarrin, 2022). However, the prevalence of sleep disturbances among Chinese healthcare workers varied greatly across studies, ranging from 12.9 to 78% (Feng and Yang, 2001; Lin et al., 2012), which may be related to the sample sizes, different PSQI cut-offs, and the regions surveyed. The results of a meta-analysis that included 31,749 participants showed that the pooled prevalence of sleep disturbances among healthcare professionals in northern China was significantly higher than in southern China (Qiu et al., 2020).In addition, the prevalence of sleep disturbances among healthcare workers in urban areas was higher than in rural/community areas, which may be attributed to the fact that healthcare institutions in urban areas are required to manage more emergencies, engage in shift work, and endure greater work pressure (Danet, 2021; Lucchini et al., 2020). However, the results of studies on regional disparities were not entirely consistent, as gender ratios, the timing of studies, different departments, and hospital grades all influenced the findings.

Using LPA, we grouped the sleep patterns of healthcare workers and identified three profiles: “good sleepers,” “inefficient sleepers,” and “poor sleepers.” “Good sleepers” and “inefficient sleepers” exhibited better overall sleep quality and reported very similar sleep problems, except for sleep efficiency. “Poor sleepers” reported lower overall sleep quality, characterized by more frequent sleep latency, shorter sleep duration, lower sleep efficiency and poorer subjective quality, as well as increased disturbances and daytime dysfunction. To the best of our knowledge, this is the first study to identify sleep patterns among Chinese healthcare workers. The largest proportion of healthcare workers were classified as “good sleepers,” while the smallest proportion were classified as “poor sleepers.” An Australian longitudinal study similarly identified the “good sleepers” pattern as “untroubled sleepers,” exhibiting few problems across all sleep dimensions, while the “poor sleepers” pattern was characterized as “troubled sleepers,” with more problems across all sleep dimensions (Leigh et al., 2015).Although previous literature has demonstrated the existence of sleep patterns characterized by various sleep problems, such as the “multiple problems” group (Yu et al., 2017), “sleep medication use” group (Mou et al., 2023), and “trouble falling asleep” group (Leigh et al., 2015), our findings suggest that sleep efficiency is a crucial dimension in differentiating sleep patterns. This is consistent with the findings of a UK study (Leng et al., 2014),which demonstrated that sleep duration is significantly less than time in bed. Sleep efficiency, commonly defined as the ratio of total sleep time to time in bed (TIB), plays a critical role in insomnia research and practice. Difficulty falling asleep is associated with psychological distress at bedtime, during attempts to fall asleep, and throughout normal waking hours, all of which perpetuate insomnia (Baglioni et al., 2010; Soehner and Harvey, 2012). Additionally, age is an important factor affecting sleep efficiency. A meta-analysis of objective studies involving 3,500 individuals aged 5 to 102 years showed that sleep efficiency in adults declines significantly with age (Ohayon et al., 2004).

The difference between good sleepers and inefficient sleepers may be related to irregular work routines (e.g., shift work), work-related stress, and the use of electronic devices, etc. Because shift work often includes night work, the normal sleep–wake cycle (circadian rhythm) is disrupted, resulting in potential consequences for shift workers’ sleep quality. A systematic review indicated that approximately 20–30% of shift workers experience significant symptoms of insomnia and excessive daytime sleepiness, which, in turn, leads to decreased sleep duration and sleep inefficiency (Booker et al., 2018). Clinical frontline healthcare workers often face exhausting daily demands and longer working hours; greater work pressure is also an important factor affecting sleep efficiency (Tucker et al., 2015).The use of electronic devices is becoming increasingly common, particularly for activities such as using devices in bed and browsing social media. While the overall quality of sleep can be better, there is a relative lack of sleep efficiency. A survey conducted during COVID-19 pandemic showed that 60.9% of healthcare workers experienced sleep disruptions or nightmares at least once per week due to device use (Stewart et al., 2021).

Regarding the characteristics of the population with different sleep patterns, the results of this study indicated that men had higher odds of being classified as “good sleepers,” while married individuals had higher odds of being classified as “inefficient sleepers.” These results are consistent with previous studies that have identified predictors of sleep in healthcare workers (Lv et al., 2023; Okechukwu et al., 2022; Zhou et al., 2024). Generally, the reasons women experience poorer sleep quality than men include behavioral factors (e.g., greater susceptibility to negative messages and sensitivity to emotions) and biological factors (e.g., hormonal regulation) (Cheng et al., 2020; Guidozzi, 2015). Married healthcare workers exhibited lower sleep efficiency, which may stem from their struggle to balance work and family obligations, leading to increased stress and sleep disturbances (Liu et al., 2020).

Regarding the relationship between sleep and psychiatric symptoms, this study found that psychological problems were more prevalent among “poor sleepers,” consistent with the findings of previous studies (Liu et al., 2022; Mou et al., 2023; Zhang et al., 2024). The relationship between sleep and psychiatric symptoms may be reciprocal (Scott et al., 2017). On one hand, individuals with shorter sleep duration may feel less rested and experience elevated stress levels (Kim and Lee, 2018). Perceived stress has been suggested as a risk factor for depressive symptoms (Racic et al., 2017). Additionally, physical and mental fatigue during the day, resulting from poor nighttime sleep, may disrupt circadian rhythms and cause hormonal changes that can contribute to psychological problems. Conversely, the increased risk of infections, excessive work patterns, coping with negative patient emotions, and frequent high arousal states are factors that can significantly affect sleep (Yin et al., 2020).

The results of this study provide practical insights into managing sleep health for healthcare workers. This study identified differences in the sleep patterns of healthcare workers, suggesting a need to develop targeted intervention strategies in the future. Specifically, in addition to focusing on poor sleepers, it is vital to target those who are good sleepers but exhibit lower efficiency. A meta-analysis of sleep interventions aimed at reducing psychiatric symptoms indicated that greater improvements in sleep quality aligned with greater enhancements in psychiatric status (Scott et al., 2017). Specific sleep management modalities include medication, multi-component cognitive-behavioral therapy for insomnia (CBTi), acupuncture, sleep hygiene education, and environmental adjustments, among other options. Healthcare institutions should also offer sleep-related training for healthcare workers. Both the Royal Society for Public Health (RSPH) and the Mental Health Foundation (MHF) recommend that primary healthcare training include awareness of and skills in assessing sleep problems (Mental Health Foundation, 2011). Hospital management should adjust shift work schedules based on the sleep quality, duration, and drowsiness levels of shift workers (Hulsegge et al., 2023). Trade union organizations in hospitals can organize cultural and recreational activities to help employees relax and improve their mood.

This study has several limitations. First, the present study employed a cross-sectional design, which limited its ability to explore longitudinal sleep variability or identify causal relationships and potential mechanisms between sleep patterns and psychiatric symptoms. In the future, we will continue conducting annual sleep monitoring and analyze in-depth trends in the sleep patterns of healthcare workers by incorporating objective indicators (e.g., activity loggers). Second, certain socio-demographic variables (such as department, working hours, night shift, pregnancy, childbirth, etc.) were not captured in this study, these variables may positively predict sleep disorders. Third, as the sample was drawn from public hospitals, the findings may not be generalizable to all physicians in China, particularly those working in rural areas.

In this study, three sleep patterns were established using LPA among Chinese healthcare workers: (1) good sleepers (63.9%), (2) inefficient sleepers (30.3%), (3) poor sleepers (Class 3, 5.7%). The results of the multinomial logistic regression analysis indicated that gender and marital status were significant predictors of different sleep patterns. In comparison to “poor sleepers,” males exhibited higher odds of being classified as “good sleepers,” while married individuals were more likely to be categorized as “inefficient sleepers.” Significant differences in psychiatric symptom scores were observed across the three sleep patterns, with “poor sleepers” exhibiting the highest levels of psychiatric symptoms. Therefore, we recommend that hospital managers pay more attention to physicians classified as “poor sleepers” and “inefficient sleepers.” Targeted sleep management should be implemented to enhance sleep quality among healthcare workers and reduce their psychiatric symptoms.