Exercise and sleep health in college students: Efficacy, mechanisms, and implications for practice

Table 1 Characteristics of the included trials

Ref.	Research type	Sample size	Intervention duration	Session duration/frequency	Measurement	Types of exercise	Exercise intensity	Exercise effects	Study quality
Liu et al[31], 2023	RCT	EG: 29; CG: 12	8 weeks	55 minutes/session, 3 times/week	PSQI, portable sleep monitor, Actigraph-GT3X, Inbody BSM330	AE	Moderate to high	Subjective: Improved PSQI total score and subscales (sleep quality, efficiency, disorders, daytime function). Objective: Reduced sleep-wake ratio, increased deep sleep ratio and SE. Energy metabolism: Increased lean body mass and basal metabolic rate	High
Miller et al[33], 2020	Crossover study	12	AE: 1 day; RE: 1 day; control: 1 day	30 minutes/session, only one session of each exercise	KSS, PSG, ingestible temperature capsules	AE, RE	Moderate	Sleep: Moderate-intensity evening aerobic/resistance exercise ending 90 minutes before bedtime did not impair sleep in healthy young males. Core body temperature: Returned to pre-exercise levels by bedtime, with no differences in sleep metrics compared to control	High
Takemura et al[38], 2024	RCT	AE group: 75; Tai chi group: 76; CG: 75	16 weeks	AE: 60 minutes/session, 2 times/month for group exercise and 150 minutes/week for home-based exercise; Tai chi: 60 minutes/session, 2 times/week	PSQI, BFI, HADS, EORTC QLQ-C30, actigraphy, physical function test	AE, mind-body exercise	AE: Moderate; Tai chi: Low	Sleep quality: Both AE and Tai chi showed significant PSQI reduction vs CG at 16 weeks and 1 year; Tai chi had greater improvement than AE. Secondary outcomes: Both interventions improved anxiety, depression, physical function, step count, and diurnal cortisol slope; Tai chi showed greater reduction in fatigue and better balance vs AE/CG	High
Martinez Aguirre-Betolaza et al[39], 2020	RCT	EG: 109; CG: 37	16 weeks	60 minutes/session, 2 times/week	IPAQ, STOP-Bang Questionnaire, Actigraph-GT3X+	AE	Individually tailored	Both groups improved SE and TST; EG had longer bedtime on weekdays and better weekend SE	High
Zhao et al[40], 2022	RCT	EG: 43; CG: 45	8 weeks	30 minutes/session,7 times/week	PSQI, PSG	AE	Moderate to high	PSG: EG showed greater improvements in TST, SE, NREM III/IV stages, and reductions in SOL, WASO vs control. Inflammatory factors: Post-intervention TNF-α, IL-1β, IL-6 Lower in EG (P < 0.01)	High
Liu and Zhang[41], 2022	Crossover study	30	-	30 minutes, single acute session	PSG	AE	Moderate	Cognitive control ability improved immediately, and at 30 minutes and 1-hour post-exercise; blood 5-HT levels increased (P < 0.01) at all time points; no change in glucose levels was observed	High
Xu et al[46], 2022	RCT	EG: 43; CG: 43	12 weeks	3 sets/session, 3 times/week	PSQI, SCL-90	RE	Low	PSQI: Both groups improved post-intervention, with EG showing greater reduction. SCL-90: Both groups reduced total scores, with EG showing better improvement	High
Kowalsky et al[47], 2023	Crossover study	24	CON: 7 days; REX: 7 days	REX: 8 hours/day, 1 break/hour	KSS, PDFQ, ActivPAL Micro 3 device	RE	Low to moderate	Significant improvement in overall discomfort and sleepiness, not significant improvement in mental fatigue and physical fatigue	High
Zhang[49], 2022	RCT	EG: 24; CG: 24	10 weeks	80 minutes/session, 2 times/week	PSQI, SDS	Mind-body exercise	Low to moderate	Sleep quality: EG’s PSQI score decreased (P < 0.01), with significant improvements in sleep duration, sleep latency, and daytime dysfunction (P < 0.05). Depression: Experimental group SDS score decreased from 60.52 ± 4.08 to 44.63 ± 4.78 (P < 0.01), significant difference vs control group (P < 0.01)	High
Hu et al[48], 2025	Crossover study	18	3 days for each exercise type	40 minutes/session	PSQI, MEQ, SAS, SDS, PARS-3, Actigraph-GT3X+, sleep watch actigraph, Cosmed K5 Metabolic Gas Analyser	AE, RE	AE: Moderate to high; RE: Moderate	Sleep: RE improved SOT, SOL, TST, WASO and SE; AE improved SOT and SE. RE had better TST and SE than AE. Melatonin: Both increased aMT6s, with RE more effective. Inflammation: Both reduced IL-6 and increased IL-10; RE had greater IL-6 reduction. Mood: Both reduced SAS scores; AE improved SDS scores, while RE showed no significant SDS change	High
Turmel et al[50], 2022	Prospective single-group pre-post study	21	14 weeks	5-30 minutes/session, 2-3 times/day	PSQI, ESS, HADS, PS, PSG, Actigraphy	Mind-body exercise	Low to moderate	Subjective: Significant improvements in all scales (PSQI: P < 0.01; HADS-A: P < 0.05; HADS-D: P < 0.01; ESS: P < 0.05; PS: P < 0.01). Physiological: Actigraphy showed reduced nighttime arousals (P < 0.01); PSG showed no changes in most parameters, but PSQI improvement correlated with increased N3 sleep stage (P < 0.01)	High
Chang et al[51], 2024	RCT	SG: 31; MG: 30; LG: 32; CG: 31	24 weeks	SG: 30 minutes/session; MG: 45 minutes/session; LG: 60 minutes/session; all groups: 5 times/week	PSQI, BDI	Mind-body exercise	Moderate	Within-group: All exercise groups showed reduced TNF-α/IL-6 and increased 5-HT at 12/24 weeks vs baseline; LG had significant BDI/PSQI reductions at 24 weeks. Between-group: At 24 weeks, LG had lower BDI/PSQI and TNF-α than CG/SG/MG; MG had lower TNF-α than SG. Sleep duration improved most in LG	High
Siu et al[52], 2021	RCT	Exercise group: 105; Tai chi group: 105; CG: 110	12 weeks	1 hour/session, 3 times/week	PSQI, ISI, Actigraph-wGT3X-BT	Conventional exercise: Aerobic + resistance training; Tai chi: Mind-body exercise	Moderate	Actigraphy outcomes: SE increased, with effects sustained at the 24-month follow-up; WASO and number of awakenings decreased. Subjective outcomes: PSQI and ISI scores reduced in both intervention groups; insomnia remission rate: 34.4% (Tai chi) vs 19.4% (exercise) post-intervention (P < 0.05)	High
Sun et al[53], 2024	RCT	EG: 18; CG: 19	12 weeks	60 minutes/session, 5 times/week	PSQI, HAM-A, FS-14, SF-36	Mind-body exercise	Low to moderate	The intervention reduced total and mental anxiety, improved sleep quality, and alleviated bodily pain	High
An et al[55], 2020	RCT	STAN: 96; HIGH: 101; COMB: 104	12-18 weeks, median 17 weeks	STAN: 25-30 minutes/session; HIGH: 50-60 minutes/session; COMB: 25-30 minutes/session AE + 25-30 minutes/session RE; all groups: 3 times/week	PSQI, SF-36, FACT-B, FACT-F, FACT-ES, FACT-T, PSS, HM, RSES, CES-D, SSAI	STAN/HIGH: AE; COMB: AE + RE	AE: High; RE: Low to moderate	Subjective: COMB had better sleep quality (vs STAN, P < 0.05) at 6 months; group-by-time interactions for happiness, anxiety, fatigue, stress, sleep. Physiological: Compared to the HIGH group, the COMB group demonstrated superior upper-body muscular endurance at 12 months (P < 0.05) and lower-body strength (P < 0.05)	High
Yuan et al[56], 2022	RCT	AR group: 10; AM group: 8; RM group: 9; CG: 8	8 weeks	60 minutes/session, 3 times/week	PSQI, Actigraph-GT3X+	AR: AE + RE; AM: AE + mind-body; RM: RE + mind-body	Aerobic: Moderate to high; resistance/mind-body: Low to moderate	Subjective: All exercise groups showed significant reductions in PSQI total score, sleep quality score; AR group reduced sleep latency, AM group reduced daytime dysfunction, RM group reduced sleep duration. Objective: All exercise groups increased TST and SE; AM and RM groups reduced number of awakenings, SOL and WASO. Group differences: AM group had greater subjective improvements; RM group had greater objective improvements	High
Lin Latt et al[57], 2024	RCT	MSH + RE: 11; RE: 8; CG: 11	5 weeks	RE: 45-60 minutes/session, MSH + RE: 50-65 minutes/session, both groups: 3 times/week	ISI, PSS, PDP, FFMQ, WHO-5, HES-7, EHS	RE, mind-body exercise	High	Perceived stress: Significant reduction in MSH + RE (P < 0.01) and RE (P < 0.05); MSH + RE > CG (P < 0.05). Mindfulness/well-being/sleep: Significant improvements in MSH + RE only (FFMQ, WHO-5, ISI, P < 0.05). Strength: Significant increases in MSH + RE (P < 0.01) and RE (P < 0.01); MSH + RE > RE (P < 0.01) and CG (P < 0.01)	High
Park et al[59], 2021	Crossover study	9	-	60 minutes/session, single session per trial	PSG, EEG, metabolic chamber, core body temperature sensor	AE	Moderate	Metabolic: Increased energy expenditure during post-exercise sleep (P < 0.05); no significant change in mean core body temperature. Subjective: Worse “refreshness” (P < 0.05) and more “frequent dreaming/nightmares” (P < 0.05) in exercise trial; no other differences. Objective sleep: Shorter REM latency (P < 0.05) and reduced SWS duration (P < 0.01); increased δ power in SWS (P < 0.05) and improved SWS stability (P < 0.05)	High
Goldberg et al[65], 2024	Crossover study	13	-	45-60 minutes/session, single session per condition (morning/evening/rest)	SSI, EEG, Actigraph-wGT3X-BT	AE	Moderate to high	Objective sleep: Both morning and evening exercise increased NREM sleep duration compared to rest, primarily due to extended N2 sleep. No significant effects on SOL, WASO, N1, N3, REM, or TST. Subjective sleep: No differences in SSI scores between conditions	High
Barrett et al[62], 2020	RCT	AE group: 137; MBSR group: 138; CG: 138	8 weeks	MBSR: Weekly 2.5-hour classes + 5-hour weekend retreat + daily home practice (20-45 minutes). AE: Weekly supervised sessions (matched to MBSR in contact hours)	PSQI, SF-12, PHQ-9, PSS	AE, mind-body exercise	Moderate	PSQI scores: Both AE and MBSR decreased, with AE showing significant benefit. Perceived sleep quality: Improved in both groups. Daily disturbances: MBSR improved more than AE. SE: No significant improvements in either group. Sustainability: Benefits persisted over 7-month follow-up	High
Saidi et al[64], 2023	Crossover study	M-type: 12; I-type: 14; E-type: 16	3 days	2 hours/session	PSQI, KSS, POMS-A, Hooper questionnaire, PSG, Actigraph-GT3X+	AEX: AE; EEX: AE	High	Whole sample: EEX showed lower SE, longer SOL, higher N1/N2 stages, lower N3, and more cortical arousals vs AEX. Chronotype differences: M-type and I-type: Lower SE, longer SOL/WASO, more cortical arousals and higher mood disturbances in EEX vs AEX. E-type: No significant sleep or mood changes between AEX and EEX. Next-day wellness: M-type reported higher stress/sleep disturbances after EEX; E-type had lower pre-sleep sleepiness	High
Gong et al[68], 2019	RCT	EG: 34; CG: 36	8 weeks	40 minutes/session, 3 times/week	PSQI, PANAS, SAS	AE, mind-body exercise	Moderate	After intervention, EG’s sleep quality, anxiety, and negative emotion scores were improved. Negative emotions played a partial mediating role in the effect of exercise on sleep quality	High

RCT: Randomized controlled trial; EG: Experimental group; CG: Control group; PSQI: Pittsburgh Sleep Quality Index; AE: Aerobic exercise; SE: Sleep efficiency; RE: Resistance exercise; KSS: Karolinska Sleepiness Scale; PSG: Polysomnography; BFI: Brief Fatigue Inventory; HADS: Hospital Anxiety and Depression Scale; EORTC QLQ-C30: European Organization for Research and Treatment of Cancer Quality of Life Core Questionnaire; TST: Total sleep time; NREM: Non-rapid eye movement; SOL: Sleep onset latency; WASO: Wakefulness after sleep onset; 5-HT: 5-hydroxytryptamine; SCL-90: Symptom Checklist-90; CON: Control week; REX: Resistance exercise week; PDFQ: Physical Discomfort and Fatigue Questionnaire; SDS: Self-Rating Depression Scale; MEQ: Morning-Evening Questionnaire; SAS: Self-Rating Anxiety Scale; PARS-3: Physical Activity Rating Scale-3; SOT: Sleep onset time; IL: Interleukin; ESS: Epworth Sleepiness Scale; PS: Pichot Fatigue Scale; N3: Deep sleep stage; SG: Short-time group; MG: Medium-time group; LG: Long-time group; BDI: Beck Depression Inventory; TNF: Tumor necrosis factor; ISI: Insomnia Severity Index; HAM-A: Hamilton Anxiety Rating Scale; FS-14: Fatigue Scale 14; SF-36: Short Form-36 Health Survey; STAN: Standard aerobic; HIGH: High-dose aerobic; FACT-B: Functional Assessment of Cancer Therapy-Breast; FACT-ES: Functional Assessment of Cancer Therapy-Endocrine Symptoms; FACT-F: Functional Assessment of Cancer Therapy-Fatigue; FACT-T: Functional Assessment of Cancer Therapy-Taxane; PSS: Perceived Stress Scale; HM: Happiness Measure; RSES: Rosenberg Self-Esteem Scale; CES-D: Center for Epidemiological Studies-Depression Scale short form version; SSAI: Spielberger State Anxiety Inventory; COMB: Combined aerobic and resistance; AR: Aerobic exercise + resistance exercise; AM: Aerobic exercise + mind-body; RM: Resistance exercise + mind-body; MSH: Mindful self-hypnosis; PDP: Psychological Distress Profile; FFMQ: Five Facet Mindfulness Questionnaire; WHO-5: World Health Organization-5 Well-Being Index; HES-7: Healthy Eating Score-7; EHS: Elkins Hypnotizability Scale; EEG: Electroencephalographic; REM: Rapid eye movement; SWS: Slow wave sleep; SSI: Spiegel Sleep Inventory; N2: Light sleep stage; N1: Sleep onset stage; MBSR: Mindfulness-Based Stress Reduction; PHQ-9: Patient Health Questionnaire-9; M-type: Morning chronotype; I-type: Intermediate chronotype; E-type: Evening chronotype; POMS-A: Profile of Mood States-Adolescents; AEX: Afternoon exercise; EEX: Evening exercise; PANAS: Positive and Negative Affect Schedule.