How a person's daily activity pattern affects performance in China's medical residency training, with loneliness playing a role

Considering the contextual framework, our institution, the First Hospital Affiliated to Soochow University, operates across two campuses in Suzhou and hosts residents from all over the country who participate in standardized training programs. The residency program is divided into two categories. The first is a combined training program integrated with postgraduate studies, where students enroll directly into the residency program and must fulfill both postgraduate coursework requirements and residency training requirements upon graduation. The second category is a socialized residency training program, designed for residents who have not previously participated in standardized training. Both types of programs span three years, with clinical training being largely consistent between them.

Each program is tailored to the resident's specialty, with rotational training lasting approximately 1–3 months per specialty. Postgraduate students also have additional coursework and research training requirements. The primary language of instruction is Chinese, though some components are delivered in English. In the clinical setting, feedback is mainly provided verbally, supplemented by mandatory written monthly evaluations for each rotation and exit evaluations at each site. These evaluations are accessible to residents for their final assessments. Each clinical site has two co-supervisors who offer feedback to the residents participating in the program.

Between March 2024 and June 2024, we distributed questionnaires to residents who completed standardized training from the 2019–2021 cohort at each campus. The questionnaires were collected between September 28, 2024, and October 28, 2024. They were disseminated through lecture announcements, email, and an applet. Our goal was to achieve a diverse sample, ensuring a balanced representation of participants across gender, study location, program type, and training specialty. All residents routinely wrote electronic medical records for both the inpatient beds under their management in the corresponding wards and for outpatient/emergency encounters. The number of inpatient beds managed by each resident remained relatively consistent (3–4 beds). The frequency of electronic classroom use was generally fixed, and residents were required to attend multiple academic lectures each month (commonly around five). The number of monthly night shifts was broadly consistent across participants, usually ranging from 4 + 2 to 5 + 2 shifts, with only minor variations between individuals. These comparable schedules indicate that participants were likely to experience similar levels of sleep disruption and evening exposure to blue light. Furthermore, all residents lived either in hospital staff dormitories or in apartments within approximately 1 km of the hospital, suggesting that their levels of exposure to outdoor nighttime lighting were also similar. All residents used smartphones; however, only two participants occasionally used the night (dark) mode feature, which was insufficient to conduct stratified analysis. We recorded the time taken to complete the questionnaire and excluded all responses from participants who took less than ninety seconds to complete it. Subjects with known emotional abnormalities, severe sleep disorders, or orofacial disorders, as well as those who did not wish to participate in the study, were also excluded.

The study was approved by the Ethics Committee of the First Affiliated Hospital of Soochow University before the questionnaire collection began. The experiments comply with the current laws of the country in which they were performed. At the beginning of each electronic questionnaire, participants were presented with a detailed informed consent form outlining the study's purpose, procedures, potential risks and benefits, as well as their right to withdraw at any time without penalty. Participants provided explicit written informed consent by agreeing to this form. All collected data were anonymized and treated with strict confidentiality to ensure privacy.

In addition to gathering basic demographic information, the survey incorporated two key assessment tools. Chronotype was evaluated using the Morningness-Eveningness Questionnaire-Self-Assessment (MEQ-SA), which comprises 19 items aimed at determining an individual's inclination towards morning or evening activities [13]. Responses are scored on either a 4- or 5-point Likert scale, depending on the specific question, resulting in a total score ranging from 16 to 86. Higher scores reflect a stronger morning preference, whereas lower scores indicate a tendency towards eveningness. The University of California, Los Angeles Loneliness Scale (UCLA-LS), third edition (1988), was used to assess the overall level of loneliness among participants. This scale consists of 20 items, divided equally between positive and negative statements, with responses recorded on a 4-point scale, where 1 represents "never" and 4 represents "frequently." The total score can range from 20 to 80, with higher scores correlating with greater perceived loneliness. The Chinese version of the UCLA-LS has been previously validated and demonstrated consistent internal reliability with Chinese college students [14,15]. Each participant was also asked a single question: "What is the one goal you most want to achieve during the three-year residency training?" They were instructed to answer with a single sentence or phrase. Furthermore, data on normative training completion scores were retrieved and matched using an electronic system.

Data normality was assessed using the Kolmogorov-Smirnov test. Since continuous variables showed non-normal distributions, non-parametric tests were employed for all analyses. For descriptive analysis, differences between continuous variables were evaluated using the Mann-Whitney U-test, while differences in categorical variables were assessed with the chi-square test.

Spearman correlation coefficients were computed to examine the relationships among the variables. The relationship between chronotype, loneliness levels, and final training completion scores was analyzed using a generalized linear model with binomial family and logit link function to analyze the binary outcome variable (high vs. low grades), adjusting for general demographic factors (gender and age).

To further investigate whether loneliness acts as a mediating factor in the relationship between chronotype and training outcomes, mediation analysis was conducted using the R "mediation" package with 1,000 bootstrap iterations to explore the indirect role of loneliness. Average causal mediation effect (ACME), average direct effect (ADE), and proportion mediated were calculated for pathways.

All statistical analyses were performed using Origin 2024, Vosviewer 1.6.11 and R version 4.3.2, with a two-sided significance level set at 0.05. To address multiple testing concerns, p-values from multiple comparisons in the GLM were adjusted using the False Discovery Rate (FDR) correction method.

A total of 198 participants who had completed their residency training were included in our study. They were divided into two groups based on a median cutoff score of 429, with 93 participants achieving relatively higher final scores. No significant differences were observed between the two groups regarding gender and age. However, there were significant differences in chronotype and loneliness levels (p < 0.001). Among all participants, 79 (39.9%) identified themselves as having an intermediate chronotype. Notably, in the higher-performing group, 41 (44.1%) were categorized as evening-type, compared to only 2 (1.9%) in the lower-performing group (p < 0.05; Table 1).

In addition, correlation coefficients were calculated between grade level, gender, and scores on each question of the MEQ-SA scale and the UCLA Loneliness Scale. The results showed that there was a significant correlation between the standardized training closing score and the type of time and the level of loneliness (Fig 1).

A bubble chart was utilized to visually represent the distribution of chronotype, loneliness levels, and grades (Fig 2). The effects of chronotype and loneliness levels on standardized training grades were further analyzed by using generalized linear models. Further multivariate analysis was performed to evaluate the effects of chronotype and loneliness levels on standardized training grades. All reported p-values were adjusted for multiple comparisons using the false discovery rate (FDR) correction method. For the overall sample, in multivariate analyses, we found that the MEQ-SA scale (OR: 0.811, 95% CI: 0.770–0.855, p < 0.001) and UCLA Loneliness scale (OR: 1.193, 95% CI: 1.076–1.323, p = 0.001) was significantly associated with the standardized training grades after adjusting for age and gender (Table 2). Chronotypes were categorized based on scale scores, with scores ranging from 42 to 58 classified as "intermediate type," while the remaining participants were classified as either "evening type" or "morning type." Our analysis revealed that the evening type (OR: 15.431, 95% CI: 3.431–69.395, p < 0.001) and morning type (OR: 0.057, 95% CI: 0.022–0.149, p < 0.001) were significantly associated with standardized training grades after adjusting for age and gender (Table 2).

Our findings indicated that lower scores on the MEQ-SA scale were associated with a higher likelihood of achieving better exit grades in residency training, while higher scores on the UCLA Loneliness Scale also correlated with improved grades (Fig 3A). To further explore this relationship, a mediation analysis was conducted to determine whether loneliness level mediates the link between chronotype and success in standardized residency training (Fig 3B). As anticipated, loneliness level was identified as a significant mediator in this association. The indirect effect of loneliness level on the association between chronotypes and training achievements implied a 0.0186-point increase in the likelihood of Better grades (95% CI, 0.0077–0.0279), accounting for approximately 55.66% of the association between chronotypes and achievements of resident standardisation training (Fig 3B).

The study included responses from 198 participants to the question, "What is the one goal you most want to achieve during the three-year residency training?" Keywords were derived from participants' answers and displayed through a text clustering model (Fig 4). In this model, terms such as "standard training," "graduation," "future," "exam," "work," "stay," and "article" appeared frequently in the responses, with "graduation" being the most common keyword, apart from "standard training."