The quality of sleep serves as a crucial physiological metric that evaluates an individual’s sleep status, incorporating subjective perception and objective sleep parameters (Ohayon et al., 2017). Adequate sleep is a fundamental component of a healthy lifestyle, as it can bolster memory consolidation, enhance attention and immune function (Moosavi-Movahedi et al., 2021), efficiently replenish expended energy from daily activities (Chandrasekaran, 2020), and promote overall well-being (Vandekerckhove and Wang, 2018). Furthermore, it plays a critical role in safeguarding cardiovascular health (Plante, 2006; Holfinger et al., 2021); regulating bodily processes, and sustaining optimal brain function (Plante, 2006; Pickersgill et al., 2022). Due to the crucial role of sleep in regulating and preserving human health, the quality of sleep has emerged as a key indicator for evaluating health status and overall quality of life (Dzierzewski et al., 2021). Presently, concerns regarding sleep quality have garnered widespread attention in society.

Sleep issues are present across all age groups, from adolescents to middle-aged and older adults. Relevant survey data show that among adolescents, about two-thirds are observed to have less than 6 h of sleep (Ohida et al., 2004; Paruthi et al., 2016). Furthermore, globally, over one-third of adults face varying degrees of sleep disturbances (Panel et al., 2015; Liu, 2016; Galinsky et al., 2018). Among the elderly population, the problem is even more pronounced, with up to 50% of older adults experiencing difficulty falling asleep or maintaining sleep (Paruthi et al., 2016; Schroeck et al., 2016). Epidemiological studies consistently indicate that sleep problems tend to intensify with increasing age (Almondes et al., 2021), a decline in sleep quality not only hampers individual learning capabilities but also compromises immune function (Dzierzewski et al., 2021). In addition, poor sleep quality is not only associated with a heightened susceptibility to chronic illnesses and an elevated likelihood of suicidal behaviors (Mikula et al., 2020; Bjorvatn et al., 2021), but also places a substantial financial strain on both families and society. For instance, in Canada, sleep-related issues lead to an estimated loss of $484 million (Streatfeild et al., 2021), while in Australia, the economic impact of sleep disorders is calculated at $35.4 billion (Chaput et al., 2022). This underscores the importance of addressing sleep quality as a critical public health concern that requires collective action on a global scale.

Pharmacological treatment, as a conventional means of improving sleep, although effective, is associated with potential dependency, a higher risk of drug-related mortality (especially in cases of overdose) (Bushnell et al., 2022), and uncertainty in therapeutic outcomes (Sys et al., 2020). These factors have driven the search for safer and more efficient non-pharmacological alternatives (Morrison et al., 2022). Non-pharmacological interventions, particularly exercise interventions, are gradually becoming the preferred choice for improving sleep quality due to their low risk and significant effects (Rios et al., 2019). Physical activity, sport, and exercise are terms often used interchangeably in research (Chennaoui et al., 2015). Physical activity refers to any form of body movement that results in energy expenditure, encompassing all activities of daily life, whether professional, household, or recreational (Bull et al., 2020). Sport typically refers to competitive physical activities. Exercise is a component of physical activity; it is planned, structured, and characterized by its frequency, intensity, and duration (Chennaoui et al., 2015). It includes aerobic exercises, resistance training, jogging, dancing, Tai Chi, Baduanjin, etc., with specific types of exercise depending on an individual’s physical capacity (Hasan et al., 2022). Research has shown that sustained engagement in exercise has a notable positive impact on sleep quality. For instance, a 12-week moderate-intensity exercise program implemented by Tseng et al. demonstrated significant improvements in sleep quality. Furthermore, the study demonstrated a notable impact on enhancing sleep quality among elderly individuals (Tseng et al., 2020). Additionally, Siu’s investigation revealed that both Tai Chi and traditional sports activities were effective in improving sleep quality, with lasting effects for up to 24 months (Siu et al., 2021). Although meta-analyses have assessed the therapeutic effects of exercise on sleep quality and generally recognized the positive impact of exercise in improving sleep quality (Banno et al., 2018), there are still certain limitations in existing studies. Firstly, previous research has focused on the immediate effects of exercise on short-term sleep (such as the following day), and there is a lack of in-depth research on the impact of long-term exercise programs on sleep quality (Stutz et al., 2019). Secondly, studies have mostly concentrated on the assessment of a single or a few exercise doses (such as exercise type, duration of intervention), lacking a comprehensive evaluation of the integrated effects of exercise types and dosages. For example, Xie et al. only analyzed the impacts of exercise type and intervention duration on sleep quality (Xie et al., 2021), which limits our comprehensive understanding of the effects of exercise types and dosages. Furthermore, due to the methodological limitations of traditional meta-analyses, while they can confirm the positive impact of exercise on sleep quality (Banno et al., 2018), there is still a lack of evidence to prove which types of exercise and dosages are most effective in enhancing sleep quality. There is a need for systematic reviews and meta-analyses of randomized controlled trials (RCTs) in this field (Kreutz et al., 2019).

Network meta-analysis, implemented within the Bayesian framework with the aid of Markov Chain Monte Carlo simulation technology (Moher et al., 2015; Vats et al., 2019), is an advanced statistical method that can assess and compare the efficacy differences among multiple interventions (Lei et al., 2022). One of the advantages of network meta-analysis is its ability to integrate direct and indirect comparative evidence from randomized controlled trials into a network evidence map, allowing not only direct but also indirect evaluations of intervention effects (Tonin et al., 2017). Furthermore, it can compare and rank the effects of different interventions to identify the optimal intervention (Kodama et al., 2021). Compared to traditional meta-analysis, network meta-analysis demonstrates significant advantages in assessing precision (Cipriani et al., 2013). To our knowledge, no study has yet explicitly identified which combination of exercise type and dosage can most effectively improve sleep quality. Therefore, this study utilizes the network meta-analysis method, combined with data from randomized controlled trials, to analyze the differences in the effects of exercise on improving sleep quality from the perspective of exercise type and dosage, thus providing evidence-based support for the development of scientific and effective exercise programs.

This study, adhering to the PRISMA guidelines for reporting systematic reviews and meta-analyses (Hutton et al., 2015), has been registered on the international platform PROSPERO with the registration number CRD42024555428.

This study compares the effects of different interventions on sleep quality through a network meta-analysis of existing research. We conducted a comprehensive and in-depth synthesis analysis, including direct and indirect evidence from 58 randomized controlled trials involving 5,008 participants who underwent various exercise programs to analyze their impact on sleep quality. The purpose of this study is to address the limitations of previous meta-analyses that were often confined to a single aspect or immediate exercise effects by exploring the impact of exercise frequency, duration, period, intensity, and type on multiple dimensions of sleep quality.

In recent years, exercise has been widely recognized for its role in improving sleep quality. Specifically, aerobic exercise, traditional Chinese exercises, resistance exercise, and combined exercise have shown unique therapeutic effects, each promoting the enhancement of sleep quality through different mechanisms. Aerobic exercise can increase aerobic capacity parameters, reduce REM latency, and decrease wakefulness (Santos et al., 2012). Traditional Chinese exercises, such as Qigong and Tai Chi, significantly enhance the tone of the vagus nerve by slowing down bodily movement and respiratory frequency, thereby promoting the balance of the autonomic nervous system (ANS). This physiological regulatory process not only adjusts the reactivity of the hypothalamic–pituitary–adrenal axis (HPA axis) but also leads to a significant reduction in serum cortisol levels. This effectively alleviates stress responses and significantly improves sleep quality (Mudano et al., 2019). Resistance exercise is an enhancer of brain-derived neurotrophic factor (BDNF), with increased BDNF concentration levels in the brain after exercise (Babaei et al., 2013), promoting the connection and reorganization of neurons in the brain, further optimizing neural networks related to sleep regulation, and thus facilitating sleep initiation and maintenance (Chennaoui et al., 2015). Combined exercise has been found to be more effective than single exercise modalities in improving sleep quality (Zhou et al., 2022). The network meta-analysis concluded that: a frequency of exercise four times per week is more effective in improving sleep quality; a single exercise session duration of ≤30 min is more effective in improving sleep quality; a 9–10 weeks exercise intervention period is more effective in improving sleep quality; high-intensity exercise is more effective in improving sleep quality; and combined exercise is more effective in improving sleep quality.

This study found that combined exercise is the best type of exercise to improve sleep quality, which is inconsistent with some existing research results. Previous meta-analysis results have shown that aerobic exercise has the most significant effect on improving sleep quality (Fang et al., 2019). The reason may be that combined exercise can leverage the advantages of various types of exercise. Research has indicated that a combined program of aerobic and resistance exercise demonstrates superior effects in improving sleep quality (Brellenthin and Lee, 2022). This is due to the fact that aerobic exercise has been shown to positively impact the release of serotonin and noradrenaline, essential neurotransmitters for enhancing sleep quality (Dolezal et al., 2017). Additionally, aerobic exercise can increase melatonin levels, which aid in improving sleep propensity (Kline et al., 2011), regulating the sleep–wake cycle, and facilitating better and faster sleep at night (Chen et al., 2024). On the other hand, resistance exercise has been found to reduce sleep onset latency, promote faster sleep onset, decrease the number of awakenings after sleep onset, and improve sleep efficiency (Zhou et al., 2022). Consequently, the combined exercises demonstrate enhanced effects on improving sleep quality.

A frequency of exercise four times per week is most effective for improving sleep quality, a finding that is consistent with previous research conclusions (Feitl et al., 2009). Ezati’s investigation proposes that exercise regimens consisting of four to seven sessions weekly are linked to improved sleep quality (Ezati et al., 2020), while other studies have demonstrated that exercising at least three times per week yields more significant enhancements in sleep quality (Hasan et al., 2022). This phenomenon may be attributed to the modulatory impact of exercise frequency on mood, as evidenced by studies indicating that regular exercise at a frequency of four times per week can enhance the release of endorphins (Huang et al., 2023). Elevated levels of endorphins have been shown to mitigate stress and anxiety, induce relaxation, and regulate circadian rhythms, ultimately enhancing the quality of sleep (Ye et al., 2022). Additionally, Wu’s study suggests that exercising less than three times per week does not significantly improve sleep quality (Wu et al., 2015). Therefore, engaging in physical activity four times per week may yield more favorable results in terms of sleep quality.

This study demonstrates that an exercise duration of ≤30 min is most effective for enhancing sleep quality, aligning with previous research (Wu et al., 2015). Izci-Balserak suggests that, prolonged exercise sessions exceeding 60 min may result in significant muscle fatigue and elevated lactate levels (Izci-Balserak et al., 2022), hastening lymphocyte apoptosis and potentially causing discomfort and lengthening recovery periods, ultimately disrupting nighttime sleep patterns. As exercise duration extends beyond 90 minutes, it may adversely affect psychological and physical health, leading to excessive arousal at night, which can impact sleep quality. (Medic et al., 2017; Chekroud et al., 2018). On the other hand, brief bouts of exercise lasting between 10–30 min have been shown to stimulate the secretion of hormones, including endorphins, which can enhance the quality of sleep by increasing the amount of deep and rapid eye movement sleep (Chan et al., 2019). Other studies have demonstrated that engaging in a 30 min exercise regimen can lead to notable improvements in mental well-being (Tanaka et al., 2002). Maintaining a positive mental state before bedtime has been linked to reduced cognitive arousal prior to sleep onset (Dressle et al., 2023), facilitating quicker and more restful sleep (Lowe et al., 2019; Fank et al., 2022). Moreover, engaging in physical activity for less than 30 min can prevent the prolonged elevation of cortisol levels that may result from extended exercise sessions (Chan et al., 2019). This is significant as heightened cortisol secretion has been linked to disturbances in sleep patterns, including reduced deep sleep stages and disrupted sleep continuity and depth (Athanasiou et al., 2023). Consequently, limiting exercise duration to 30 min or less is shown to be more beneficial in enhancing the quality of sleep.

The study concluded that an exercise intervention lasting 9–10 weeks is most effective in enhancing sleep quality, aligning with previous research findings. A meta-analysis further supports the notion that short-term (<3 months) exercise interventions can notably improve sleep quality (Xie et al., 2021). Hasan’s study suggests that an exercise intervention lasting a minimum of 8 weeks is necessary for significant improvements in sleep quality (Hasan et al., 2022). Moreover, research findings indicate that a 9-week exercise regimen has been linked to notable enhancements in sleep quality, subjective sleep perception, and reduced sleep onset latency. These improvements in subjective sleep quality have been observed to persist for up to 3 months following the conclusion of the exercise intervention (Chan et al., 2014). Additionally, a 10-week exercise program has been demonstrated to effectively decrease sleep fragmentation and elevate melatonin levels, thereby promoting quicker sleep onset (Cai et al., 2014; Bonardi et al., 2016). It is important to note that engaging in exercise for a year may actually lead to difficulties in falling asleep (Tworoger et al., 2003). Thus, a moderate exercise intervention period is more beneficial for enhancing sleep quality.

The results of this study indicate that high-intensity exercise has the best effect on sleep quality. This is inconsistent with existing research results, where meta-analysis results have shown that moderate-intensity exercise interventions have the most significant effect on improving sleep quality (Zhao et al., 2023). This variation may arise from the distinct physiological effects of exercise intensity on the body. Lower exercise intensities have a limited stimulating impact on the sympathetic nervous system (SNS) (Daniela et al., 2022), which is insufficient to induce a significant transition from the SNS to the parasympathetic nervous system (PNS), a shift that is advantageous for enhancing deep sleep and overall sleep quality during the night (Garbarino et al., 2019). Conversely, high-intensity exercise elicits a more pronounced activation of the SNS; following exercise, the heightened SNS activation prompts a robust shift toward PNS activity (Druică et al., 2018), a mechanism that significantly promotes the onset of deep sleep (Trinder, 2012). Additionally, high-intensity exercise has been found to reduce daytime sleepiness (Chekroud et al., 2018), facilitate falling asleep at night (Xie et al., 2022), and decrease the number of awakenings after sleep onset (Larsen et al., 2019), ultimately leading to an overall improvement in sleep quality (Frimpong et al., 2021). Thus, high-intensity exercise is more conducive to the improvement of sleep quality.

At present, systematic comparative studies on the impact of exercise type and dosage combinations on sleep quality effects are still insufficient. In addition, previous studies have often used passive control methods in the setting of control groups, which may to some extent limit the comprehensiveness and accuracy of the results. In light of this, future research should focus on constructing an analytical framework based on different combinations of exercise types and dosages, delving into the specific effects of various types of exercise and their dosage changes on sleep quality, in order to scientifically evaluate the effectiveness of exercise interventions in enhancing sleep quality and to identify the optimal exercise type and dosage combination strategies. At the same time, to strengthen the rigor of research design, it is recommended to use active control methods to effectively reduce the potential impact of the placebo effect, ensuring the reliability and universality of research conclusions. By precisely analyzing the effects of exercise types and dosages on sleep quality, a basis can be provided for developing personalized and precise exercise intervention programs.

This network meta-analysis further supports the efficacy of exercise interventions in enhancing sleep quality. Based on the findings, we recommend the continuation of exercise routines. The research indicates that combined exercises performed four times a week, at a high intensity, over a period of 9–10 weeks, with each session lasting ≤30 min, are more effective in improving sleep quality. We hope that the results of this study will provide evidence for healthcare professionals, educators, and policymakers to make informed decisions. Subsequent research should prioritize the design of exercise types and dosages based on individual attributes, including gender and health status, to enhance the effectiveness of strategies for improving sleep quality.

Several limitations exist for this NMA: ① This study encompassed 58 randomized controlled trials with a total of 5,008 participants; however, the distribution of sample sizes across various types of exercise, exercise periods, durations, intensities, and frequencies were uneven. This disparity somewhat hinders the validity of the statistical analysis. ② The study predominantly evaluated sleep quality using the Pittsburgh Sleep Quality Index, a subjective self-assessment method that may introduce bias into the measurement results. ③ Our research centered on assessing the impact of exercise on sleep quality and did not involve a comparison with alternative intervention measures, such as cognitive behavioral therapy. ④ Furthermore, numerous randomized controlled trials lacked comprehensive descriptions of their randomization procedures, potentially compromising the generalizability and reliability of the study’s conclusions. ⑤ The control groups in the randomized controlled trials selected for this study all employed passive control, which is insufficient to completely rule out the presence of Hawthorne/Placebo effects.

This study systematically integrates 58 RCTs and analyzes the effectiveness of various exercise programs in improving sleep quality through a network meta-analysis approach. The research focuses on the perspective of exercise type and dosage combinations, comprehensively analyzing these elements of exercise programs to identify the optimal exercise program to improving sleep quality. The results show that a combined exercise program with a frequency of four times per week, a duration of ≤30 min, a period of 9–10 weeks, and a high-intensity level exhibits superior effectiveness in improving sleep quality. This finding transcends the limitations of previous studies that often focused on a single exercise element or short-term exercise effects, and by thoroughly analyzing all five exercise dimensions, it helps to determine a more effective exercise program for improving sleep quality. However, due to the limitations in the number and quality of included literature, as well as potential selection biases, caution should be exercised in interpreting the results of this study. Therefore, future research should further expand the sample size, optimize study designs, and notably in the selection of control groups, active controls should be adopted to effectively control the interference of placebo effects, thereby enhancing the reliability of the research findings.