Targeted exercise interventions on stress, anxiety, depression, and sleep disorders in PhD students

Abstract

BACKGROUND

Doctoral students often encounter mental health challenges, including stress, anxiety, depression, and sleep disorders. It is important to explore effective intervention methods to enhance their overall physical and mental well-being. It is anticipated that targeted exercise will lead to a significant reduction in stress, anxiety, and depression levels as well as an improvement in sleep quality.

AIM

To assess the feasibility and potential benefits of both intervention models in enhancing the sleep quality of doctoral students while alleviating stress, anxiety, and depression.

METHODS

A retrospective analysis of health data from 64 doctoral students across three universities in Shenyang during the 2024-2025 academic year was conducted. The participants were divided into a targeted exercise group and a Tai Chi group. The study employed the Epworth Sleepiness Scale, Insomnia Severity Index, Pittsburgh Sleep Quality Index, Perceived Stress Scale-10, Generalized Anxiety Disorder Scale-7, and Patient Health Questionnaire-9 to evaluate the impact of the two interventions on reducing stress, anxiety, depression, and sleep disorders.

RESULTS

The primary results of the study indicated that targeted exercise interventions are significantly effective in alleviating symptoms of anxiety, stress, and depression, as well as in improving sleep quality. Compared to Tai Chi interventions, this approach demonstrates greater durability of effects. Although the efficacy of targeted interventions may gradually diminish over time, the overall research findings suggest that targeted exercise remains a more effective therapeutic approach than Tai Chi interventions.

CONCLUSION

The impact of targeted exercise on stress, anxiety, depression, and sleep disorders was greater than that of Tai Chi, confirming the potential benefits for psychological health intervention for doctoral students.

Key Words: Targeted exercise; Tai Chi; Pressure; Anxiety; Depression; Sleep disorders

Core Tip: This study compared the effectiveness of targeted exercise and Tai Chi in alleviating stress, anxiety, depression, and sleep disorders among doctoral students. While Tai Chi positively impacted mental health, targeted exercise was a more effective and sustainable intervention. This study addressed a gap in previous research that predominantly focused on the effects of a single exercise. Our study provided a new perspective and preliminary evidence for future mental health intervention strategies. Consequently, this study proposed more effective sports intervention strategies, thereby enriching the theoretical foundation of mental health interventions

INTRODUCTION

Stress, anxiety, depression, and sleep disorders are significant global public health challenges[1]. The prevalence rates of anxiety, depression, and sleep disorders among doctoral and postdoctoral researchers are 17%, 24%, and 64%, respectively[2,3], which is approximately three times higher than those observed in the general population[4]. A survey conducted in the United Kingdom revealed that the suicide risk among doctoral students was as high as 35%. Doctoral candidates encounter numerous sources of pressure, including uncertain training schedules, financial instability, social isolation, uncertain future prospects, and academic competition[5-9]. Furthermore, stringent assessment criteria (e.g., academic paper publishing requirements) is one of the factors contributing to the categorization of doctoral students as a high-risk group for mental health issues. The ongoing mental health crisis within academia has prompted extensive discussions to the underlying causes and potential remedial measures.

Previous research has utilized both pharmacological and non-pharmacological interventions to understand the mental health issues of doctoral students. Prior studies included cognitive-behavioral therapy[7], sertraline[8], and rumination[9]. All interventions have demonstrated varying degrees of therapeutic efficacy. However, barriers such as stigma, treatment costs, time constraints, and financial burdens have prevented many doctoral students from accessing effective support[10].

Exercise is a non-pharmacological intervention and alleviates stress, anxiety, depression, and sleep disturbances[11]. However, its effectiveness is contingent upon the type, intensity, and individual adaptability of the exercise regimen[12]. Previous studies on exercise interventions have primarily focused on singular forms of broad-spectrum activities, such as aerobic exercise[13], resistance training[14], Tai Chi[15], yoga[16], and fitness qigong[17]. These studies have often overlooked the distinct physiological and psychological needs of individuals, leading to a one-size-fits-all approach. Furthermore, the response of high-stress individuals to exercise is influenced by their preferences and biological rhythms. There are complex interactions between stress, anxiety, depression, and sleep disorders, complicating treatment efforts.

In response to these shortcomings, this study focused on the advantages of personalized menu-style targeted exercise intervention. Personalized menu-style targeted exercise intervention involves the customization of targeted exercise programs tailored to an individual’s physical condition, health goals, exercise capacity, interests, and other relevant factors. These interventions are presented in the format of a menu, offering selectable exercise options that cater to the unique needs of each individual[18]. Theoretically, the human nervous and endocrine systems respond differently depending on psychological states, and targeted exercise can modulate these systems through specific exercise stimuli, thereby improving psychological symptoms more effectively. Targeted exercise interventions have distinct advantages in the precise treatment of neurological and psychiatric disorders when applied in a real-world setting. Due to their low-risk and low-cost characteristics, exercise interventions are regarded as effective alternative treatment methods.

The primary objective of this study is to preliminarily explore a personalized menu-style targeted exercise intervention model and conduct a retrospective comparison with the well-known Tai Chi. This analysis aims to assess the feasibility and potential benefits of both intervention models in enhancing the sleep quality of doctoral students while alleviating stress, anxiety, and depression. We implemented targeted exercise interventions tailored to participants’ baseline levels and psychological needs and compared them with traditional Tai Chi interventions to evaluate differences in their effectiveness in alleviating specific psychological symptoms.

MATERIALS AND METHODS

General information

This study employs a randomized group comparison research design based on historical data, and according to the inclusion and exclusion criteria, and retrospectively analyzed the complete health data of 64 doctoral students from three universities in Shenyang from November 2024 to March 2025. The analysis was conducted in three stages: Initial (baseline); mid-term (8 weeks); and final (12 weeks). Based on prior intervention records, the 64 doctoral students were divided into a targeted exercise intervention group (n = 32) and a Tai Chi group (n = 32).

The study was reviewed and approved by the Martial Arts and Dance College, Shenyang Sport University Institutional Review Board, approval No. 20241012. All participants provided written informed consent.

Inclusion criteria

The following inclusion criteria were utilized: (1) Doctoral student aged between 27 and 33 years; (2) Met the diagnostic criteria for anxiety, depression, and sleep-wake disorders as outlined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; (3) No engagement in any sports activity in the prior 3 months; (4) Scores exceeding 13 on the Perceived Stress Scale (PSS)-10, 9 on the Generalized Anxiety Disorder 7-item scale (GAD-7), and 7 on the Pittsburgh Sleep Quality Index (PSQI) within the prior month; (5) Complete data records for the initial, mid-term, and final stages of the study; and (6) Basic physical activity skills.

Exclusion criteria

The following exclusion criteria were used: (1) Other sleep disorders, such as obstructive sleep apnea; (2) Non-doctoral research participants; (3) Serious hearing, visual, or language impairments; (4) Hypnotic drug, antidepressant, or similar substance use within the prior 3 months; (5) Pregnant or lactating females; (6) Engagement in other clinical trials within the previous year; or (7) Doctoral research that adhered to China’s regulations on confidential information management and incomplete data.

Intervention measures

Between July 23, 2024 and September 15, 2024, a total of 24 professors comprising 8 experts in higher education, 8 experts in sports science, and 8 experts in psychology conducted a feasibility assessment of the intervention plan for both the targeted exercise group and the Tai Chi group through five rounds of evaluation (Table 1). The detailed intervention measures are organized as follows below.

Table 1 Results of Kendall’s W test for inter-rater agreement.

Number of experts	Comparison of intervention methods	Number of evaluations	Kendall W	χ2	df	P value
24	Targeted exercise vs Tai Chi	7	0.859	123.726	6	0.000

Targeted exercise intervention group: The targeted exercise intervention group (OG) participated in the following exercises: (1) Morning run: Conducted 5 times/week from 7:00-7:50 am. This included a 10-minute warm-up, 30 minutes of jogging, and a 10-minute cool-down. Morning jogging served to invigorate bodily functions and effectively combat drowsiness; (2) Baduanjin: Held 3 times/week from 12:30-1:30 pm. This included a 10-minute warm-up and a 10-minute cool-down. Participants completed three rounds of practice within 40 minutes with peer assistance for body adjustments during 3-minute intervals. Baduanjin features gentle, slow movements that emphasize the harmonious integration of body, breath, and spirit to aid in relaxation and stress reduction; (3) Team sports game: Organized bi-weekly on Tuesdays for 60 min, including a 10-minute warm-up and a 10-minute cool-down. The 40-minute team sports games consisted of various activities such as eagle catching chick, three-person quadrupedal + dynamic five ring, star support moon + space shuttle, and obstacle crossing relay. The goals of collaborative sports games were enhanced social interaction and mitigation of depressive emotions; (4) Ball or swimming sports: On alternate Thursdays participants had the option to engage in various sports, including badminton, table tennis, basketball, tennis, football, or swimming. Each session lasted for 60 minutes, which encompassed a 10-minute warm-up and 10-minute cool-down. The core of the session consisted of 40 minutes dedicated to fundamental practice of the chosen sport, such as serving in tennis or receiving in badminton, with a focus on alleviating anxiety; (5) Resistance training: Resistance training sessions were conducted bi-weekly on Saturdays for 60 minutes. This included a 10-minute warm-up and a 10-minute cool-down. The resistance training component lasted for 40 minutes and featured exercises such as push-ups (20 repetitions for 4 sets), wall squats (10 minutes for 3 sets), and sit-ups (20 repetitions for 4 sets). This training was intended to enhance overall sleep quality in a targeted manner; and (6) Yoga: Yoga sessions were scheduled 5 days a week from 5:00-6:00 pm, with each session lasting 60 minutes. This included a 10-minute warm-up and a 10-minute cool-down. The main focus of the 40-minute yoga practice was to regulate breathing, enhance body balance, and facilitate stretching, thereby contributing to improvements in the insomnia index.

Tai Chi group: The Tai Chi group (CG) adopted the 24-style Tai Chi Chuan as promoted by the Wushu Sports Management Center of the General Administration of Sport of China. The practice sessions were conducted daily from 7:00-8:00 am, lasting for 60 minutes. Each session included a 10-minute warm-up and a 10-minute cool-down. Participants completed 6-8 rounds of practice within 40 minutes, with each round lasting 4 minutes and a 1-minute interval between rounds for body adjustment. This study evaluated the effects of 24-style Tai Chi on several indicators, including the Epworth Sleepiness Scale (ESS), Insomnia Severity Index (ISI), PSQI, PSS-10, GAD-7, and Patient Health Questionnaire-9 (PHQ-9).

Assessment scales

ESS: ESS is a widely utilized self-assessment tool to evaluate the propensity for daytime sleepiness in an individual’s daily life[19]. This scale comprises eight items that assess the likelihood of dozing off or falling asleep in various situations. It is suitable for screening pathological sleepiness and evaluating the effects of interventions on daytime alertness[20]. Each item is scored on a four-point scale, ranging from 0 (indicating no likelihood of dozing) to 3 (indicating a high chance of dozing). The total score can range from 0 to 24, with scores equal to or greater than 11 suggesting the presence of daytime sleepiness. Higher scores indicate more severe levels of daytime sleepiness. The scale has been validated for use in China[21].

ISI: ISI is a self-report scale comprising seven items to assess the severity and changes in insomnia symptoms among subjects[22]. This scale evaluates the impact of difficulties in falling asleep, maintaining sleep, and early morning awakening on quality of life. It also evaluates the level of concern a patient has regarding their insomnia. Each item is rated on a five-point scale, ranging from 0 to 4, yielding a total score that ranges from 0 to 28. Higher scores indicate more severe insomnia symptoms[23]. The scale has been validated for its reliability and validity within the Chinese population[24].

PSQI: PSQI was developed by Buysse et al[25] in 1989 to evaluate an individual’s sleep quality over the preceding month[25]. This scale comprises 19 self-rated items categorized into seven dimensions with each scored from 0 to 3. The total PSQI score, which ranges from 0 to 21, is derived from the cumulative scores of all dimensions, with higher scores indicating poorer sleep quality. Research conducted by Chinese scholars suggested that a PSQI score of 7 is an appropriate threshold for identifying sleep quality issues within the Chinese population[26].

PSS-10: PSS-10 is a widely utilized instrument for assessing an individual’s perceived stress levels[27]. This study utilized the validated Chinese version of the PSS-10. The scoring for this scale ranges from 0 (never) to 4 (very often), with items 4, 5, 7, and 8 reverse-scored[28]. The total possible score on the scale is 40, where scores below 13 indicate low stress, scores between 14 and 26 indicate moderate stress, and scores from 27 to 40 indicate high stress.

GAD-7: GAD-7 was employed to evaluate participant anxiety symptoms[29]. It consists of seven items, each rated on a four-point scale, yielding a total score range from 0 to 21. Scores ranging from 0 to 4 indicate no anxiety, scores from 5 to 9 indicate mild anxiety, scores from 10 to 14 indicate moderate anxiety, and scores from 15 to 21 indicate severe anxiety. This version was validated in China[30].

PHQ-9: The PHQ-9 is a validated scale designed to assess depressive symptoms in participants. It utilizes a four-point scoring system ranging from 0 to 3, resulting in a total score that can vary from 0 to 27. The scoring criteria are categorized as follows: Scores of 0-4 indicate no depression; scores of 5-9 indicate mild depression; scores of 10-14 indicate moderate depression; scores of 15-19 indicate moderately severe depression; and scores of 20-27 indicate severe depression[31]. The Chinese version of this scale was validated in China[32].

Physical Exercise Rating Scale-3: The Physical Exercise Rating Scale-3 was employed in this study to assess the physical activity levels of the doctoral students[33]. The overall physical activity of participants was calculated using the formula: Physical activity = intensity × time × frequency. The intensity and frequency ratings range from 1 point to 5 points, while the time rating ranges from 1 point to 4 points[34]. The overall Physical Exercise Rating Scale-3 rating spans from 0 point to 100 points, with higher scores reflecting greater physical activity. This scale categorizes physical activity into three levels: Low exercise (≤ 19 points); moderate exercise (> 19 points and ≤ 42 points); and high exercise (> 43 points)[35].

Statistical analysis

This study utilized SPSS 24.0 (IBM Corp., Armonk, NY, United States) and JASP 0.19 (https://jasp-stats.org/) for data analysis. The Delphi expert method was employed to assess the feasibility of the two intervention measures over five rounds. Baseline characteristics were assessed using the χ² test for categorical variables and the t-test for continuous variables. All data were subjected to the Shapiro-Wilk normality test. Generalized estimating equations were employed to analyze the data from the ESS, ISI, PSQI, PSS, GAD-7, and PHQ-9 at three stages: Baseline; post-intervention; and follow-up. They were used to investigate group effects, time effects, and interaction effects. Differences between the two groups were evaluated using the Bootstrap method of the independent samples t-test, and Cohen’s d effect size was calculated. The correlation and robustness between physical activity levels and ESS, ISI, PSQI, PSS, GAD-7, and PHQ-9 were assessed following targeted exercise interventions using Bayesian methods. The significance level for all analyses was set at P < 0.05.

RESULTS

Baseline characteristics

The baseline demographic characteristics indicated no significant differences between the OG and the CG across all baseline indicators (P > 0.05). The sex distribution was balanced in both groups (62.50% vs 3.13%, χ² = 0.577, P = 0.631), and the ages were comparable (30.16 ± 1.97 years vs 30.19 ± 2.05 years, t = -0.062, P = 0.951). There were no statistically significant differences in smoking rates (65.63% vs 56.25%, χ² = 0.591, P = 0.442) nor in drinking rates (68.75% vs 59.38%, χ² = 0.611, P = 0.434). The distribution of the years into the participant’s doctoral program was also similar (year 2: 28.13% vs 25.00%; year 3: 37.50% vs 34.37%; year 4: 34.37% vs 40.63%, χ² = 0.296, P = 0.874). The average body mass index of the two groups were nearly identical (24.38 ± 3.26 kg/m²vs 24.34 ± 2.84 kg/m², t = 0.041, P = 0.967). Key research indicators, including PSQI score (11.16 ± 1.46 vs 11.09 ± 1.51, t = 0.168, P = 0.867), GAD-7 score (14.03 ± 1.33 vs 14.00 ± 1.32, t = -0.094, P = 0.925), and PSS score (28.94 ± 1.85 vs 28.34 ± 2.03, t = 1.225, P = 0.225), also demonstrated no significant differences at baseline. The similarities in the baseline characteristics suggested that the two groups were well-matched prior to the interventions, thereby minimizing the potential influence of demographic characteristics and baseline indicators on the study results (Table 2).

Table 2 Baseline demographic characteristics, n (%).

Characteristic	Targeted exercise	Tai Chi	χ2/t	P value
Sex
Female	20 (62.50)	17 (53.13)	0.577	0.631
Male	12 (37.50)	15 (46.87)
Age (year), mean ± SD	30.16 ± 1.97	30.19 ± 2.05	-0.062	0.951
Smoking
Yes	21 (65.63)	18 (56.25)	0.591	0.442
No	11 (34.37)	14 (43.75)
Doctoral stage
Year 2	9 (28.13)	8 (25.00)	0.296	0.874
Year 3	12 (37.50)	11 (34.37)
Year 4	11 (34.37)	13 (40.63)
BMI, mean ± SD	24.38 ± 3.26	24.34 ± 2.84	0.041	0.967
Drinking
Drinking	22 (68.75)	19 (59.38)	0.611	0.434
Never	10 (31.25)	13 (40.62)
PSQI, mean ± SD	11.16 ± 1.46	11.09 ± 1.51	0.168	0.867
GAD-7, mean ± SD	14.03 ± 1.33	14.00 ± 1.32	0.094	0.925
PSS, mean ± SD	28.94 ± 1.85	28.34 ± 2.03	1.225	0.225

BMI: Body mass index; GAD-7: Generalized Anxiety Disorder 7-item scale; PSQI: Pittsburgh Sleep Quality Index; PSS: Perceived Stress Scale.

Changes in ESS, ISI, and PSQI

The results presented in Table 3 indicate that there were no significant differences in the ESS, ISI, and PSQI between the OG and CG at baseline (ESS: 11.81 ± 1.87 vs 11.50 ± 1.88, d = 0.16; ISI: 17.69 ± 1.33 vs 17.44 ± 1.72, d = 0.16; PSQI: 11.16 ± 1.46 vs 11.09 ± 1.51, d = 0.05; P > 0.05). At the 8-week intervention mark, all indicators in the OG exhibited a significant improvement trend [ESS: 6.81 ± 1.40 vs 10.38 ± 1.21, d = -2.73, mean difference (MD) = -3.56, 95% confidence interval (CI) = -4.20 to -2.94; ISI: 10.66 ± 1.29 vs 12.78 ± 1.58, d = -1.47, MD = -2.13, 95%CI = -2.87 to -1.42; PSQI: 5.97 ± 1.59 vs 8.97 ± 1.53, d = -1.92, MD = -3.00, 95%CI = -3.75 to -2.53; P < 0.05]. This improvement persisted marginally at the 12-week follow-up.

Table 3 Analysis of the Epworth Sleepiness Scale, Insomnia Severity Index, and Pittsburgh Sleep Quality Index.

Measurement	mean ± SD		Cohen’s d	Mean difference (95%CI)	P values
	Targeted exercise	Tai Chi			Group effect	Time effect	Interaction effect
ESS
Baseline	11.81 ± 1.87	11.50 ± 1.88	0.16	0.31 (-0.63 to 1.26)	0.001	0.001	0.001
8 weeks	6.81 ± 1.40	10.38 ± 1.21	-2.73	-3.56 (-4.20 to-2.94)
12 weeks	9.25 ± 1.37	10.93 ± 1.29	-1.26	-1.69 (-2.32 to -0.98)
ISI
Baseline	17.69 ± 1.33	17.44 ± 1.72	0.16	0.25 (-0.51 to 0.99)	0.001	0.001	0.009
8 weeks	10.66 ± 1.29	12.78 ± 1.58	-1.47	-2.13 (-2.87 to -1.42)
12 weeks	11.63 ± 1.49	15.03 ± 1.59	-2.21	-3.41 (-4.19 to -2.66)
PSQI
Baseline	11.16 ± 1.46	11.09 ± 1.51	0.05	0.06 (-0.65 to 0.87)	0.001	0.001	0.019
8 weeks	5.97 ± 1.59	8.97 ± 1.53	-1.92	-3.00 (-3.75 to -2.53)
12 weeks	6.56 ± 1.46	10.84 ± 1.63	-2.77	-4.28 (-5.06 to -3.59)

CI: Confidence interval; ESS: Epworth Sleepiness Scale; ISI: Insomnia Severity Index; PSQI: Pittsburgh Sleep Quality Index.

The group effect, time effect, and interaction effect of the related indicators all achieved statistical significance (P < 0.01), indicating that the improvement effect of targeted exercise was significantly superior to Tai Chi, with this difference maintained over time. Notably, PSQI demonstrated a robust intervention effect at 12 weeks, with sustained effects observed between the two groups in ISI (d = -2.21) and PSQI (d = -2.77), which further suggests that the related indicators may exhibit greater sensitivity to temporal persistence. However, it is evident from the changes in mean values that this trend of change gradually weakens over time, particularly as the intervention effect of Tai Chi diminished at a faster rate than targeted exercise (Table 3).

Changes in PSS, GAD-7, and PHQ-9

The outcomes presented in Table 4 indicate that there were no significant differences in the PSS, GAD-7, and PHQ-9 scores between the OG and CG at baseline (PSS: 28.93 ± 1.85 vs 28.34 ± 2.03, d = 0.30, MD = 0.59, 95%CI: -0.39 to 1.57; GAD-7: 14.03 ± 1.33 vs 14.00 ± 1.32, d = 0.02, MD = 0.31, 95%CI: -0.64 to 0.65; PHQ-9: 14.06 ± 1.56 vs 13.97 ± 1.60, d = 0.06, MD = 0.93, 95%CI: -0.69 to 0.86; P > 0.05). After 8 weeks of intervention, significant improvements were observed in all scores in the OG: PSS (13.13 ± 1.84 vs 16.09 ± 1.86, d = -1.60, MD = -2.96, 95%CI: -3.91 to -2.06, P < 0.01); GAD-7 (5.88 ± 1.43 vs 9.03 ± 1.59, d = -2.08, MD = -3.16, 95%CI: -3.93 to -2.46); and PHQ-9 (6.13 ± 1.56 vs 8.19 ± 1.63, d = -1.29, MD = -2.06, 95%CI: -2.86 to -1.26).

Table 4 Analysis of the Perceived Stress Scale, Generalized Anxiety Disorder 7-item scale, and Patient Health Questionnaire-9.

Measurement	mean ± SD		Cohen’s d	Mean difference (95%CI)	P values
	Targeted exercise	Tai Chi			Group effect	Time effect	Interaction effect
PSS
Baseline	28.93 ± 1.85	28.34 ± 2.03	0.30	0.59 (-0.39 to 1.57)	0.001	0.001	0.013
8 weeks	13.13 ± 1.84	16.09 ± 1.86	-1.60	-2.96 (-3.91 to -2.06)
12 weeks	14.31 ± 2.02	19.03 ± 1.82	-2.46	-4.72 (-5.62 to -3.76)
GAD-7
Baseline	14.03 ± 1.33	14.00 ± 1.32	0.02	0.31 (-0.64 to 0.65)	0.001	0.001	0.001
8 weeks	5.88 ± 1.43	9.03 ± 1.59	-2.08	-3.16 (-3.93 to -2.46)
12 weeks	6.31 ± 1.49	11.16 ± 1.41	-3.34	-4.84 (-5.54 to -4.15)
PHQ-9
Baseline	14.06 ± 1.56	13.97 ± 1.60	0.06	0.93 (-0.69 to 0.86)	0.001	0.001	0.019
8 weeks	6.13 ± 1.56	8.19 ± 1.63	-1.29	-2.06 (-2.86 to -1.26)
12 weeks	6.88 ± 1.45	10.06 ± 1.54	-2.12	-3.19 (-3.91 to -2.46)

CI: Confidence interval; GAD-7: Generalized Anxiety Disorder 7-item scale; PHQ-9: Patient Health Questionnaire-9; PSS: Perceived Stress Scale.

At the 12-week follow-up, both groups exhibited a persistence of the improvement. However, the OG demonstrated a more pronounced sustained effect compared with the CG as indicated by PSS (14.31 ± 2.02 vs 19.03 ± 1.82, MD = -4.72, 95%CI: -5.62 to -3.76), GAD-7 (6.31 ± 1.49 vs 11.16 ± 1.41, MD = -4.84, 95%CI: -5.54 to -4.15), and PHQ-9 (6.88 ± 1.45 vs 10.06 ± 1.54, MD = -3.19, 95%CI: -3.91 to -2.46). This analysis suggested that targeted exercise was superior to Tai Chi with a relatively sustained impact. However, over time the effects of this intervention may gradually rebound. Notably, at the 12-week mark, the difference between the two groups widened further (PSS: d = -2.46, r = -0.78; GAD-7: d = -3.34, r = -0.86; PHQ-9: d = -2.12, r = -0.73), demonstrating that the effect of targeted exercise delaying the rebound of PSS, GAD-7, and PHQ-9 scores was significantly superior to that of Tai Chi (Table 4).

Correlation between targeted exercise and assessment indicators

The results of the Bayesian correlation analysis indicate that targeted exercise exhibited significant correlations with the GAD-7, PHQ-9, PSS, ESS, ISI, and PSQI. All Bayes factor (BF₁₀) values exceeded the established threshold, thereby reinforcing the robustness of the relationships between targeted exercise and the assessment indicators. Notably, the correlation between sports activity value (SAV)-1 → GAD-7 (r = -0.968, BF₁₀ = 1.585 × 10¹⁶, 95%CI: -0.984 to -0.926) suggested that higher SAV-1 scores were associated with lower levels of anxiety. Similarly, the correlations between SAV-3 → PSS-10 (r = -0.942, BF₁₀ = 5.249 × 10¹⁵, 95%CI: -0.971 to -0.870) and SAV-4 → ESS (r = -0.957, BF₁₀ = 2.547 × 10¹⁴, 95%CI: -0.978 to -0.901) also demonstrated strong negative linear relationships. Among the remaining pairings, SAV-2 → PHQ-9 (r = -0.759, BF₁₀ = 4.029 × 10⁴, 95%CI: -0.868 to -0.531), SAV-5 → ISI (r = -0.869, BF₁₀ = 8.447 × 10⁷, 95%CI: -0.931 to -0.722, and SAV-6 → PSQI (r = -0.815, BF₁₀ = 1.010 × 10⁶, 95%CI: -0.900 to -0.624) exhibited slightly lower correlation coefficients. According to the criteria proposed by Kass and Raftery[36], the BF₁₀ values for PHQ-9, ISI, and PSQI still provided strong evidence, validating the robustness of these results. Overall, these findings revealed a negative correlation between anxiety, depression, stress, and sleep disorders and differentiated exercise intervention activity levels. This may reflect the protective effect of targeted exercise on doctoral students’ stress, anxiety, depression, and sleep disorders, suggesting an oppositional relationship (Table 5).

Table 5 Correlation between the Generalized Anxiety Disorder 7-item scale, Patient Health Questionnaire-9, Perceived Stress Scale, Epworth Sleepiness Scale, Insomnia Severity Index, and Pittsburgh Sleep Quality Index and targeted exercise intervention.

Index	r	BF10	95%CI
SAV-1 - GAD-7	-0.968	1.585 × 1016	-0.984 to -0.926
SAV-2 - PHQ-9	-0.759	4.029 × 104	-0.868 to -0.531
SAV-3 - PSS-10	-0.942	5.249 × 1012	-0.971 to -0.870
SAV-4 - ESS	-0.957	2.547 × 1014	-0.978 to -0.901
SAV-5 - ISI	-0.869	8.447 × 107	-0.931 to -0.722
SAV-6 - PSQI	-0.815	1.010 × 106	-0.900 to -0.624

BF₁₀: Bayes factor; CI: Confidence interval; ESS: Epworth Sleepiness Scale; GAD-7: Generalized Anxiety Disorder 7-item scale; ISI: Insomnia Severity Index; PSQI: Pittsburgh Sleep Quality Index; PHQ-9: Patient Health Questionnaire-9; PSS-10: Perceived Stress Scale; SAV: Sports activity volume.

DISCUSSION

We investigated the impact of targeted exercise and Tai Chi interventions on alleviating stress, anxiety, depression, and sleep disorders. The results indicated that although Tai Chi interventions yielded significant effects, targeted exercise interventions were more effective in enhancing GAD-7, PHQ-9, PSS-10, ESS, ISI, and PSQI scores among doctoral students. Further Bayesian correlation analysis revealed a significant negative correlation between the amount of targeted exercise and the GAD-7, PHQ-9, PSS-10, ESS, ISI, and PSQI scores in doctoral students, suggesting a strong association between various types of exercise and activity levels and psychological benefits. Additionally, we found that the long-term effects of exercise interventions gradually diminished over time, consistent with previous research findings[37]. Therefore, establishing exercise habits is crucial for improving stress, anxiety, depression, and sleep quality.

The stress experienced during the doctoral training period can disrupt the circadian rhythm[38]. Morning running exercises alleviated stress levels among doctoral students to some extent, aligning with earlier research findings[39]. The practice of Baduanjin effectively alleviated stress levels among doctoral students. This efficacy can be attributed to the interconnected trinity mechanism of Baduanjin, which encompasses the regulation of the body, breathing, and mind[40]. This approach restores cortisol levels in doctoral students[41] and enhances heart rate variability, thereby promoting balance within the autonomic nervous system[42]. Our findings align with previous research, highlighting the potential spillover effects of Baduanjin in stress management[43].

Participation in team sports can significantly reduce depressive symptoms experienced by doctoral students. This effect is attributed to enhanced social interactions, shared interests in sports, mutual support, and competitive dynamics inherent in team sports, all of which contribute to increased team cohesion. Consequently, increased cohesion appears to alleviate depression symptoms among doctoral students. These findings align with prior studies that explored the efficacy of sports in mitigating depression among the elderly[44].

Both swimming and ball sports effectively alleviated the anxiety levels of doctoral students. The sound produced during tennis matches contributes to this alleviation, as its appropriate loudness, stable rhythm, and orderly pitch changes align with human auditory characteristics, resulting in soothing auditory experiences[45]. Additionally, swimming in ice water significantly reduced anxiety, corroborating findings from previous research[46,47].

Resistance training effectively alleviated sleep disorders. This effect is attributed to the release of negative emotions during resistance training, particularly the enhancement of muscle strength and body composition, which can boost self-efficacy and improve psychological well-being. The moderate fatigue induced by muscle contraction and relaxation can promote sleep and enhance sleep quality[48]. The targeted intervention of yoga effectively reduced the ISI scores among doctoral students for three primary reasons. First, the regulation of breathing in yoga adjusts the frequency and rhythm of respiration, resulting in a more stable and even breathing pattern and contributing to emotional regulation. Second, the practice of body balance in yoga necessitates the elimination of external distractions. This concentration training can be transferred to daily life, aiding practitioners in sustaining calmness and composure. Third, doctoral students frequently experience muscle tension and stiffness due to prolonged periods of sitting and improper postures. Yoga stretching alleviates muscle tension, enhances flexibility and elasticity, and promotes sleep[49].

Previous studies have indicated that Tai Chi intervention positively affects stress relief, anxiety reduction, depression alleviation, and sleep quality improvement among doctoral students. Our research findings demonstrated that targeted exercise significantly outperformed Tai Chi in terms of both intervention effectiveness and sustainability. Our study effectively addressed the limitations of prior research that focused solely on the effects of a single broad-spectrum exercise on stress, anxiety, depression, and sleep disorders, thereby providing a new perspective and preliminary evidence to support future intervention strategies. Through comparative analysis, this study enriched the theoretical foundation of sports intervention and offered practical and effective solutions for psychological health interventions tailored to doctoral students.

The limitations of this study should be acknowledged. The sample size was relatively small and focused exclusively on a specific group of doctoral students, restricting the generalizability and applicability of the findings. This study also primarily utilized self-reported subjective scales that may introduce bias and compromise the accuracy of the results. The strict inclusion criteria and lack of control over intensity, duration, and frequency may have led to biases and risks in extrapolating the findings. Moreover, the absence of fatigue data may undermine the completeness and accuracy of the evaluation of intervention effectiveness. Finally, Due to the heterogeneity of the activities within the intervention group, it is challenging to identify which specific activity contributed to the observed benefits.

Future research plans will encompass several key aspects. First, we aim to expand the sample size and conduct multicenter prospective randomized controlled trials to analyze the precise mechanisms by which targeted exercise interventions affect stress, anxiety, depression, and sleep disturbances among doctoral students. This approach will reduce sample bias and mitigate interference from confounding factors, thereby enhancing the extrapolation and clinical guidance value of the research findings. Second, introducing objective measurement indicators, including polysomnography, heart rate variability, cortisol testing, brain imaging, and behavioral tasks will further elucidate these mechanisms. Additionally, we will adjust the statistical analysis for exercise intensity, duration, frequency, and other potential confounders and effect modifiers such as lifestyle, social support, smoking status, body mass index, sex, type of exercise, alcohol consumption, and coffee intake through stratification, subgroup, and sensitivity analyses to assess the robustness of the results and explore differences among subgroups. Finally, in future research, the team will implement a unified design to systematically verify and assess which specific activities contributed to the observed benefits.

CONCLUSION

This study systematically reviewed and compared the effects of targeted exercise and 24-style Tai Chi on stress, anxiety, depression, and sleep disorders among doctoral students. The results indicate that targeted exercise was more effective for alleviating these issues compared with Tai Chi. Although the effectiveness of the intervention gradually diminished over time, the sustained benefits of targeted exercise remained superior to those of Tai Chi. This finding provides empirical evidence for sports practitioners and school administrators to implement targeted exercise programs in mental health management of doctoral students. However, the study was limited by sample size, the lack of exercise intensity and fatigue data, and the use of subjective scales. Therefore, future research will focus on improvements in the following areas: Incorporating objective biological indicators; expanding sample sizes; conducting multicenter randomized controlled trials; and employing stricter statistical methods to evaluate intervention effects more comprehensively and accurately.

ACKNOWLEDGEMENTS

The authors would like to thank all the participants and workers who participated in this study. It was their participation that ensured the smooth progress of this study.