It is scientifically acknowledged that sleep and executive functions undergo changes in prefrontal circuits that underlie executive functioning as part of healthy aging [1]. According to the model proposed by Harrison and Horne [2] and recent findings, selective sleep loss affects prefrontal cortex (PFC) circuitry, leading to changes in brain metabolism and impairing executive function in healthy adults [3-5]. Based on these findings, several studies have debated the vulnerability of the PFC to sleep loss [6,7]. A study investigating sleep problems and executive function among university students found a significant relationship between poor sleep quality (SQ) and impaired executive function [8]. Medical students, in particular, are frequently affected by poor SQ during their time in medical school [9].

Medical students from various countries have shown differing prevalence rates of poor SQ, including 44.2% in Nepal, 41% in Iran, 51% in the United States [10], 59% in Lithuania [11], 90% in China, 70.0% in Hong Kong, 32.5% in Nigeria, 37% in Saudi Arabia [12], and 62.0% in Ethiopia [13]. Recently, in Malaysia, a prevalence of 59.6% was reported among clinical students at International Islamic University Malaysia [14].

Sleep-related problems can result in substantial medical costs and human burden [15,16]. Sleep disorders such as insomnia affect approximately 25 million people in the United States annually, contributing to an estimated health care cost of US $100 billion [17]. Therefore, improving SQ among medical students through nonpharmacological interventions may help prevent or minimize the consequences of poor sleep, such as burnout.

To date, only a few effective nonpharmacological interventions are known to treat sleep-related issues such as poor SQ [18]. In a systematic review of nonpharmacological interventions aimed at improving sleep among university students (including 11% of medical students), cognitive behavioral therapy (CBT) accounted for 33.3% of the interventions, followed by sleep hygiene at 33.1%; relaxation, mindfulness, and hypnotherapy at 22.2%; and Gestalt therapy or imagery rehearsal therapy at 11.1%. The researchers concluded that CBT was effective in treating sleep disorders among university students. Additionally, other outcome variables, such as sleep duration, SQ, sleep onset latency (SOL), sleep hygiene, and dysfunctional beliefs about sleep, were examined [19]. Based on these findings, CBT and sleep hygiene education (SHE) were identified as the most commonly used interventions to improve SQ.

CBT for insomnia (CBT-I) is a nonpharmacological intervention recommended as the first line of treatment for improving SQ [20] and is considered the gold standard for insomnia treatment [21,22]. CBT-I was developed based on the Spielman behavioral model of insomnia [23] and is an extension of the stimulus control theory of insomnia [24]. It is a multicomponent psychotherapy, comprising elements such as sleep education, stimulus control therapy, sleep restriction therapy, cognitive therapy, sleep hygiene, and relaxation training [25,26].

CBT-I focuses on dysfunctional cognition and maladaptive behaviors that contribute to sleep problems. However, access to CBT-I is limited due to the shortage of trained personnel and the high cost of treatment [27-30]. Internet-based CBT-I (ICBT-I), with or without therapist guidance, has been successfully used for managing chronic insomnia [27,28]. Similarly, SHE has been reported to be equally effective in improving SQ among individuals aged 12 years and older [29]. However, SHE is a single educational intervention that focuses on sleep hygiene, whereas CBT-I is a multicomponent intervention. Therefore, CBT-I is considered a potentially more effective intervention for improving SQ and executive function among individuals, but evidence-based data to support this claim are lacking. Furthermore, studies comparing the effectiveness of CBT-I and SHE interventions are scarce. Despite recognizing SQ as a foundation of a healthy lifestyle, there is limited emphasis on sleep hygiene and sleep education in medical schools. Hence, this study protocol outlines the methodology of a randomized controlled trial (RCT) designed to examine the effectiveness of CBT-I and SHE on SQ and executive function among medical students. The secondary outcomes to be investigated include presleep worry, attitudes toward sleep, sleep hygiene, and sleep parameters.

This study will use several theories, namely, the cognitive models of insomnia, the hyperarousal models of insomnia, and the PFC circuitry model, to develop the theoretical framework for the interventions to be investigated. First, according to the cognitive model of insomnia (Figure 1), individuals with sleep problems often experience excessive worry, which prevents them from falling asleep and consequently leads to poor SQ [30]. As a result of this worry, individuals with sleep problems tend to focus on sleep-related internal threats (eg, bodily sensations) and external threats (eg, environmental noises). This selective focus causes them to become hyperaware of minor cues (previously unnoticed), which leads to excessive worry and anxiety. Additionally, individuals with sleep problems develop negative cognitions that further increase their cognitive and arousal activity, preventing them from falling asleep [30]. The model also suggests that individuals with sleep problems often engage in safety behaviors, such as napping or consuming alcohol, to relieve this arousal. However, these safety behaviors eventually worsen the worry and prevent individuals from challenging their negative sleep beliefs [30].

Second, the hyperarousal model of insomnia incorporates both the cognitive and behavioral models of insomnia, while integrating principles of classical conditioning [24,25]. This model posits that conditioned arousal is a key factor in perpetuating sleep problems. It further explains that repeated associations between poor sleep and the bed can increase conditioned arousal, with the bed and sleep environment becoming stimuli for heightened arousal [31]. Cognitively, individuals with increased sleep arousal begin to perceive “learned sleep–preventing associations,” thereby perpetuating their sleep problems [25].

Third, it is scientifically acknowledged that sleep and executive functions undergo alterations in prefrontal circuits that support executive function as part of healthy aging [1]. According to the model proposed by Harrison and Horne [2] and recent findings [3-5], selective sleep loss affects PFC circuitry, potentially leading to changes in brain metabolism that impact executive function in healthy adults. Building on these findings, several studies have debated the vulnerability of the PFC to sleep loss [6,7]. In this study, improvements in SQ may affect the PFC circuitry, particularly executive function. Consistent with this, our study aims to use ICBT-I to alleviate poor SQ and enhance executive function among medical students. The delivery of ICBT-I overcomes common limitations associated with traditional face-to-face administration, such as the lack of trained personnel, high costs, and limited treatment availability due to time constraints [30].

This study aims to elucidate the effects of improved SQ on executive function. The following hypotheses will be investigated in the trial:

A 2-arm, parallel-group RCT will be conducted, comparing ICBT-I with internet-delivered SHE (ISHE). The trial design is summarized in Figure 2. The study population will consist of preclinical medical students who meet the established criteria for poor SQ, as indicated by self-report assessments, which are internationally recognized as standardized instruments. The medical students will be randomly assigned in a 1:1 ratio to either the ISHE or ICBT-I group. All interventions will be conducted online using a paid Zoom (Zoom Video Communications/Qumu Corporation) platform. As shown in Figure 2, participants will receive an electronic document containing the participant information statement, screening instruments, informed consent form, and randomized allocation to the intervention groups. Both baseline and follow-up assessments will be conducted either online or via mobile platforms. Participants will be able to print the participant information statement, and contact details (phone numbers and emails) will be provided for participants to reach the researchers before consenting and during the data collection process.

Ethical approval for this study was obtained from the Human Ethics Committees at Universiti Putra Malaysia (UPM; JKEUPM-2023-1446) and Universiti Kebangsaan Malaysia (UKM; JEP-2024-669). Informed consent will be obtained from each potential participant before their recruitment into the study. They will be briefed on their right to participate or decline, as well as their right to opt out of the study without penalty. Regarding privacy and confidentiality, data will be fully anonymized, and no detailed information about participants will be recorded. No incentives will be provided for participation in this study.

The RCT protocol aligns with the SPIRIT (Standard Protocol Items for Randomized Trials) guidelines [32], as shown in the recruitment and assessment procedures in Figure 2. The uploaded SPIRIT checklist can be found in Multimedia Appendix 1. Multimedia Appendix 2 presents the protocol approval obtained from the Australian New Zealand Clinical Trials Registry (ANZCTR; ACTRN12624000243516). Upon completion of the study, all data will be stored in the institutional database for a specified duration before being destroyed. The findings will be published as a research article in a journal, with no information that could reveal the participants’ identities.

As shown in Figure 3, the sample size for the RCT was calculated using G*Power software (Heinrich-Heine-Universität Düsseldorf), selecting the F tests for analysis of variance: repeated measures, within-between interaction. The calculation was based on a precision level of 0.05, a study power of 95%, an effect size of 0.20, and a correlation coefficient of 0.5 among the outcome measures. Therefore, the minimum required sample size for each group is 56. A low effect size of 0.20 was chosen based on previous interventions addressing SQ [33]. After adjusting for a 10% dropout rate, the sample size was increased to 60 medical students per group. Given that this is a 2-arm RCT, a total of 120 medical students will be required to complete the study.

The RCT will involve 120 medical students from 2 government-based medical universities: UPM and UKM. These institutions were selected for their convenience and proximity to the researchers. The inclusion of both universities is also intended to account for potential differences in medical curricula and academic schedules between government-based and private medical institutions. Participants will consist of volunteers, specifically undergraduate medical students in the preclinical years (ie, years 1-3). A list of potential participants will be compiled with permission from the dean. Recruitment advertisements will be posted across various platforms, including social media, campus spaces, and noticeboards, to facilitate participant engagement. Before the initial assessment, each participant will be required to complete the consent form and information sheet. Potential participants will undergo a screening process to assess their eligibility before being randomized into either the ISHE or ICBT-I group.

As participants will be recruited from 2 different universities, a stratified randomization method will be used to ensure balance between the groups from each university. Eligible students will be grouped by university and gender before random selection. Gender will serve as a matching factor during the stratified randomization process to ensure a comparable number of female and male students are recruited from both universities.

Data will be collected from volunteer medical students, and the researchers will have no influence on the randomization process or access to future allocations. Each participant will be assigned an identification number and listed before being randomized into either the ISHE or ICBT-I group using online randomization software [34]. A total of 60 students will be allocated to each group.

The initial assessment and experiment will be conducted using a paid Zoom platform. Participants will be interviewed by the researcher using a semistructured questionnaire to determine their eligibility based on the inclusion and exclusion criteria before randomization.

Upon expressing interest in participating in the RCT, medical students will be directed to a website with general information about the study and its procedures. They will then access the screening assessment document and consent form to complete the eligibility process. After providing informed consent and meeting the inclusion criteria, potential participants will be contacted by phone. During the call, the investigator will confirm the participant’s understanding of the RCT and the provisions for the internet-based interventions being investigated.

The inclusion criteria in Textbox 1 will be considered before recruiting a medical student for this study.

The criteria listed in Textbox 2 will be used to exclude individuals from this study.

Single blinding will be implemented in this trial. As an online self-report assessment will be conducted, the researchers will remain blinded to the outcomes during the RCT. An automated email will inform the participants of the randomization results, so the medical students will know which group they are assigned to. Limited contact will be maintained between the researchers and participants to prevent any bias in the allocation process or influence on the subsequent assessments. The assessors will remain blinded even after participants contact the research team and reveal their allocation. A researcher (VM; University Putra Malaysia), who has access to all data, will perform the statistical analyses.

Participants in both groups will be assessed at baseline (week 0 or pretreatment assessment), week 6, 1 month, 3 months, and 6 months after the intervention. T0 represents the pretreatment assessment, while the subsequent 4 assessment points are designated as T1, T2, T3, and T4, respectively.

Treatment adherence will be ensured through a standardized approach to delivering the intervention. The treatment manual outlines the key therapeutic information provided to the medical students, minimizing the influence of individual therapists. Additionally, the use of a therapist manual and supervision will further support adherence to the treatment in both groups.

The risk of severe or serious adverse events in this study is low, as no such events have been reported in previous studies using CBT-I for the management of insomnia [49].

Nevertheless, studies have shown that sleep restriction therapy, one of the components of CBT-I, may increase levels of vigilance impairment and daytime sleepiness among participants, potentially leading to adverse situations such as suicide attempts, incidents, and formal complaints about the online intervention [49]. As both the intervention and assessment will be conducted online, the researcher may not be made aware of all such events. The researcher will, however, instruct participants to complete a feedback form to report any self-reported adverse effects [50].

Participants can withdraw from the RCT at any time without penalty and are not required to provide reasons for their decision. In terms of crisis management, emergency psychiatric services will be available at UPM Teaching Hospital to assist participants in cases of crisis or distress that exacerbate their symptoms or lead to significant deterioration in functioning. All clinic attendees will have access to this service. The investigators will be notified if any medical student utilizes the emergency service, at which point a decision will be made regarding the RCT and whether the crisis qualifies as an adverse event related to the trial.

The primary researcher (VM) will collect all data under the supervision of a registered and licensed clinical psychologist (FM). The researcher holds a master’s degree in clinical psychology and meets the minimum qualifications to administer all questionnaires to participants. Online assessments will be conducted by the clinical psychologist, using an instrument with Yes or No responses to items related to the inclusion and exclusion criteria. All clinical interviews will be audio-recorded to enable subsequent reliability assessments (see Multimedia Appendices 3 and 4 for the study information sheet and consent form, respectively). Participants who are willing to participate will be provided with a consent form. Their consent will be confirmed once they sign the form. Participants who meet the inclusion criteria and have PSQI scores greater than 5 will be grouped and asked to complete the online Trail Making Test, DBAS-16, PSAS, and SHI.

The flowchart for the research activities is summarized in Figure 2. An SMS text message reminder will be sent by the Zoom administrator 1 day prior to and 30 minutes before the intervention. Follow-up assessments will be conducted at 1-month, 3-month, and 6-month intervals after the postintervention to reassess participants’ SQ, executive functioning, self-reported sleep problems, SE, sleep cognition, and presleep arousal. The same instruments described in the previous section will be used for the assessments. All participants will be monitored for adherence to the intervention, and potential dropouts will be identified and addressed accordingly.

Data will be analyzed using SPSS version 29 (IBM Corp.). Descriptive statistics will be used to obtain information on dropout rates, recruitment, and completeness. All participants will be included in the intention-to-treat analysis for assessing the main efficacy of the intervention, with no planned interim analysis for efficacy or futility. The baseline characteristics of participants will be presented according to their randomized group, with no inferential statistical tests performed at this stage.

For the RCT design, outcome measures will be assessed for normality and presented as either means with SD or medians with IQR, depending on the data distribution. Participants’ characteristics will be summarized using descriptive statistics, with categorical data presented as frequencies and percentages, and continuous data presented as means (SD) or medians (IQR). Normality will be tested using the Kolmogorov-Smirnov and Shapiro-Wilk tests.

Changes in all outcome measures (ie, SOL, TST, NOA, WASO, SQ, SE, dysfunctional beliefs about sleep, presleep arousal, and SHI) from pretreatment (T0) to the first (T1), second (T2), and third (T3) time points will be assessed for treatment effects using repeated measures analysis of variance. Both treatment and time interaction effects will be examined. Paired samples t tests will then be conducted within each condition to evaluate possible simple or main effects. Significant findings and statistical trends will be further analyzed using Cohen d effect size, which accounts for the correlation between pre- and posttreatment values.

Potential confounding factors, particularly class year and gender, will be addressed and analyzed as variables that may influence the research outcomes. These factors will be incorporated into the analysis through both univariate and multivariate linear regression analyses, allowing for the isolation of each factor and its association with the outcomes of interest.

Regarding the fourth hypothesis, mediation analyses will be conducted to explore whether changes in the outcome measures from pretreatment to posttreatment are significantly mediated by dysfunctional beliefs and attitudes about sleep, sleep hygiene, and presleep worry. The analysis will follow the traditional mediation model, where the direct, indirect, and total effects are defined in terms of the linear regression coefficients [51]. For example, when the mediation effect size is medium (ie, SM) and the intraclass correlation coefficient is 0.2, with 4 repeated measures, the Sobel method would require 191 participants to achieve 80% power. By contrast, the product-of-coefficients and bootstrap methods would require 188 and 185 participants, respectively, to achieve the same power [52]. This study will adopt the bootstrap method for mediation analysis due to its flexibility in terms of sample size requirements.

The results of this trial will be published in academic journals and presented at international and national conferences or clinical scientific meetings. Depending on the journal’s policies, the results will be made accessible online to interested readers.

CBT has long been established as the first-line treatment for individuals suffering from chronic insomnia, demonstrating positive effects and sustained improvements in sleep-related outcomes, as measured by both sleep diaries and objective indices of SQ [19-21]. A recent RCT with a well-defined placebo-controlled group found that dCBT produced effect sizes comparable to those of traditional face-to-face therapy [53]. However, it remains unclear whether CBT-I has a significant causal relationship with improving executive function among medical students compared with an SHE intervention. This knowledge gap has persisted for some time, and an investigation among medical students is therefore necessary. This study represents the first attempt to specifically examine SQ and executive function by comparing 2 established sleep interventions in this vulnerable group. The findings will be relevant to the care of medical students experiencing persistent SQ issues, particularly during their preclinical years. In particular, the results may highlight the potential for improving SQ as a novel therapeutic target, with benefits extending to generalized health outcomes and enhanced executive function. Furthermore, as an ICBT-I approach will be used in this trial, the study will demonstrate that addressing SQ among medical students is measurable, feasible, and effective.

Given the negative psychological and physiological consequences of poor SQ, such as burnout, suicidal ideation, and reduced empathy toward patients, this trial provides an opportunity to explore a nonpharmacological intervention for improving SQ and executive function among medical students. The findings will help bridge the important research gap identified in previous studies on CBT-I across diverse populations. These studies rarely focus on specific groups such as health professionals or young adults (eg, medical students). If poor SQ causally alters executive function, it is plausible to hypothesize that (1) CBT-I will improve both SQ and executive functioning and (2) the pathway through improved SQ will facilitate the treatment effects on executive functioning [54]. Furthermore, this trial provides a unique opportunity to expand existing knowledge, as only a few RCTs have investigated the roles of hyperarousal mediators (eg, presleep worry), cognitive mediators (eg, dysfunctional beliefs and attitudes about sleep), and behavioral mediators (eg, sleep hygiene) in CBT-I [20].

This study will be the first RCT in Malaysia to investigate SQ and executive function using a dTMT among medical students. Additionally, it is the first to examine potential mediators, such as presleep worry and sleep hygiene, in the relationship between beliefs about sleep and SQ in this population. A representative sample will be recruited through a simple random sampling technique, allowing the findings to be extrapolated to public medical students in Malaysia and potentially other countries with similar educational demographics.

Nevertheless, this study has important limitations. The sample will be drawn only from medical students at UPM and UKM, meaning it may not be representative of all government-based undergraduate medical students in Malaysia. Additionally, the risk of response biases cannot be ruled out, particularly regarding subjective sleep experiences, which may influence the findings. Participants may also be inclined to provide socially desirable answers. Studies on Asian cultures have shown that, due to the cultural emphasis on modesty, Asians are likely to suppress the expression of positive emotions, which could impact the accuracy of self-reported data [55]. This study will use only a quantitative research approach, which may limit a comprehensive understanding of medical students’ perspectives and views on poor SQ symptoms and the strategies they use to address the issue. Additionally, there is a risk that participants may provide socially desirable answers when responding to the research instruments.

This study is the first attempt to design a CBT intervention aimed at ameliorating poor SQ and its associated negative effects among medical students. It is also the first study to explore the relationship between improvements in SQ and executive function among medical students using CBT-I.

CBT-I focuses on addressing dysfunctional cognition and maladaptive behaviors that contribute to sleep problems. However, access to CBT-I is limited due to a shortage of trained professionals and the high cost of treatment [27-30]. ICBT-I, with or without therapist guidance, has proven effective in managing chronic insomnia [27,28]. Similarly, SHE has also been shown to improve SQ among individuals aged 12 years and older [29]. However, SHE is a single educational intervention that focuses on sleep hygiene, whereas CBT-I is a multicomponent intervention. Therefore, CBT-I is considered a potentially more effective intervention for improving SQ and executive function. However, evidence-based data to support this claim are lacking. Additionally, studies comparing the effectiveness of CBT-I and SHE are scarce. Despite recognizing SQ as a foundation of a healthy lifestyle, there is limited emphasis on sleep hygiene and sleep education in medical schools. Consequently, this study protocol outlines the methodology of an RCT aimed at evaluating the effectiveness of CBT-I and SHE on SQ and executive function among medical students.