From Passion to Abyss: The Mental Health of Athletes during COVID-19 Lockdown

The SARS-CoV-2 outbreak led to a global pandemic declared by WHO on 11 March 2020, prompting mandatory lockdowns to curb transmission. Such measures increased stress, anxiety, and sleep disturbances across populations, with athletes potentially at higher risk of mental health problems due to abrupt disruptions to training and competition. This study aimed to: (1) characterize athletes’ psychological health during lockdown (stress, subjective happiness, sleep disorders); (2) compare psychological health by remuneration status and presence/extent of pay cuts; and (3) test whether sleep disorders mediate the relationship between stress and subjective happiness. Hypotheses: H1 athletes show high stress; H2 athletes show high sleep disorders; H3 athletes show low subjective happiness; H4 athletes feel unhappier than before interruption; H5 pay cuts associate with higher stress and sleep disorders and lower happiness; H6 stress negatively influences happiness; H7 sleep disorders mediate the stress–happiness relationship. The study is important for informing athlete support, policy responses, and return-to-sport strategies during crises.

Prior research indicates lockdowns elevate stress, anxiety, depression, and sleep problems, and reduce well-being. WHO recommended maintaining physical activity, yet athletes faced severe restrictions, impairing routines and potentially worsening mental and sleep health. Evidence suggests athletes may be at higher risk of mental health problems than the general population. Sleep quality is closely linked to stress responsivity and well-being, with poor sleep associated with worse psychological outcomes. Financial insecurity threatens meaning in life and is associated with poorer mental health; in sport, pandemic-related wage cuts likely intensify stress and reduce happiness. These strands of literature support examining stress, sleep, happiness, and the role of financial factors during sport lockdowns.

Design and participants: Cross-sectional online survey administered between 1 April and 30 May to 1,492 Portuguese athletes (45.2% female), aged 13–48 years (M=20.32, SD=6.72), from eight team sports: roller hockey (n=319), basketball (n=294), volleyball (n=288), rugby (n=248), futsal (n=160), handball (n=66), football (n=64), water polo (n=53). Two levels: training (n=727; M age=15.45, SD=1.19) and elite/competition (n=765; M age=24.94, SD=6.55). In the elite group, 22.6% received salaries and 12.7% allowances; among the 270 athletes with income, 15.2% had no cuts, 32.6% partial cuts, and 52.2% total cuts. Procedures: Questionnaire distributed via social networks and sports federations; all items mandatory, resulting in no missing data. Measures: (1) Perceived Stress Scale (PSS-10; 0–4 Likert; α=0.83); (2) Subjective Happiness Scale (SHS; 4 items; 1–7 Likert; α=0.71); plus an independent single item assessing change in happiness since interruption (Happiness Change Item, HCI; 1=Much Less Happy to 7=Much Happier); (3) Sleep Disorders Scale (SDS; COPSOQ II subscale; 4 items scored 1–5; α=0.81). Additional items assessed club contact regularity and psychological support received. Statistical analysis: SPSS v27 used. Descriptive statistics (means, SDs, minima, maxima) and Pearson correlations (99% CI) computed. Group comparisons by remuneration status and pay-cut level used Kruskal–Wallis tests (non-normal distributions), with pairwise order mean comparisons at α=0.05. Mediation tested via linear regression (Enter method) and PROCESS macro Model 4 with 10,000 bootstrap samples to estimate indirect effects (significant if 95% CI excludes 0). Assumptions checked with Durbin–Watson and VIF.

- Stress: Sample mean PSS=22.26 (SD=6.38), above Portuguese pathological cut-offs (men>20, women>22). Female mean significantly above 22 (t(671)=6.584, p<0.001); male mean significantly above 20 (t(817)=5.377, p<0.001). Pathological stress proportions: training level—61.1% females, 41% males; competition level—61.1% females, 68.3% males. Supports H1. - Sleep disorders: SDS mean=3.39 (SD=0.75), higher than Portuguese reference (2.46±1.05); risk distribution: low 12.2% (1–2.33), moderate 44.5% (2.33–3.66), high 43.3% (>3.66). Supports H2. - Subjective happiness: SHS mean=4.77 (SD=1.04) lower than pre-lockdown Portuguese samples (general 5.12±1.02; athletes 5.41±0.09; adolescents 5.65±1.05). Training group M=4.94 (SD=1.06) vs adolescents 5.65 (t(726)=-17.989, p<0.001); competition group M=4.61 (SD=1.01) vs general 5.12 (t(765)=-13.989, p<0.001) and vs athletes 5.41 (t(765)=-21.976, p<0.001). HCI mean=2.65 (SD=1.17); 75.7% reported being less happy (scores 1–3). Supports H3 and H4. - Correlations: Stress–sleep r=0.416; stress–happiness r=-0.461; sleep–happiness r=-0.290 (all p<0.01). HCI correlations: stress r=-0.221; sleep r=-0.083; SHS r=0.241 (all p<0.01). - Remuneration status and cuts (H5): Stress differed by remuneration status (KW=22.246, p=0.001): non-remunerated 22.30±6.13 < remunerated 24.02±6.26 ≈ allowances 25.13±4.68. Pay cuts associated with higher stress: no cuts 19.88±5.93 < partial 22.73±5.37 < total 26.14±4.73. Sleep disorders differed by remuneration status (KW=7.001, p=0.030): allowances 3.70±0.62 > remunerated 3.51±0.68 > non-remunerated 3.44±0.85; sleep did not differ significantly by pay-cut level. Subjective happiness did not differ by remuneration status (KW=1.353, p=0.508) but was lower with cuts: total 4.29±1.03 < partial 4.76±0.84 < none 5.10±0.87. Overall, H5 partially supported. - Mediation (H6, H7): Stress negatively predicted happiness (β=-0.497; t(1490)=-15.793; p<0.001; R2=0.247), supporting H6. PROCESS mediation showed significant indirect effect of stress on happiness via sleep disorders (a*b β=-0.0411; 95% CI [-0.0697, -0.0141]); total effect standardized coefficient ≈ -0.4962; mediation model R2=0.2564, indicating partial mediation, supporting H7. - Club support/contact: Only 13.3% reported receiving psychological support from clubs; 58.8% reported no contact with clubs during the interruption.

The abrupt interruption of training and competition during COVID-19 lockdown was associated with pathological stress levels, substantial sleep problems, and reduced subjective happiness among Portuguese team-sport athletes. Elevated stress poses risks for physical and mental health, potentially impairing immune function and performance. The positive association between stress and sleep disorders indicates that stressed athletes are not recovering adequately during sleep. Sleep’s central role in metabolic, emotional, and cognitive functioning explains its partial mediation between stress and happiness: poor sleep exacerbates the negative impact of stress on well-being. Adolescents may be particularly vulnerable to sleep-related psychosocial and performance issues. Financial stressors, especially wage cuts, further worsened stress and diminished happiness, reflecting the detrimental effects of economic insecurity. The findings underscore the precarious work conditions faced by many athletes during the pandemic and the need for organizational support. Practically, stakeholders should prioritize sleep hygiene education, emotional regulation training, and structured psychosocial support, including maintaining regular communication with athletes during disruptions, to mitigate adverse mental health outcomes and facilitate safer and more effective returns to sport.

Sports lockdowns during COVID-19 negatively affected athletes’ psychological health: athletes experienced high stress, more sleep disorders, and lower subjective happiness, with a marked perceived decline compared to pre-lockdown. Sleep disorders partially mediated the negative effect of stress on happiness. Wage cuts aggravated stress and reduced happiness. Clubs largely failed to provide psychological support or maintain contact. Stakeholders should implement support systems during interruptions and returns to competition, enhance sleep hygiene and emotional competencies training, and address the precarious conditions of professional athletes. Future work should continue to examine mechanisms and interventions to protect athlete mental health in crisis contexts.

The study lacked pre-interruption baseline measurements for direct within-person comparisons; all data were self-reported via an online survey; the sample was limited to Portuguese athletes in team sports, which may affect generalizability; and the assessment of perceived change in happiness relied on a single added item (HCI).