The unprecedented scale and complexity of modern public health emergencies, exemplified by the COVID-19 pandemic, have placed extraordinary demands on healthcare systems and, in particular, on frontline rescue and medical personnel. Rescue workers—encompassing medical staff, emergency responders, and logistical support teams—constitute the critical human infrastructure of any epidemic response, operating at the direct interface of disease containment and patient care (1). However, while substantial global attention has focused on the physical health risks these workers face (2), a growing body of evidence suggests that the psychological and behavioral toll may be equally profound and consequential (3).

Emerging research indicates that crisis environments fundamentally disrupt the lifestyle equilibrium of those working within them. Studies conducted during the COVID-19 pandemic have documented adverse changes in health behaviors among healthcare populations, including increased substance use (4), sleep disturbances (5), and physical inactivity (6). For rescue workers, these disruptions are compounded by unique occupational stressors: prolonged periods in restrictive environments, high-risk exposure, intense workload, and separation from social support networks (7). These conditions can create a “syndemic” of interacting health risks, where psychological distress and maladaptive coping behaviors converge, potentially compromising both individual well-being and operational effectiveness (8).

The downstream implications of these lifestyle and mental health deteriorations extend beyond the individual to impact systemic resilience. A rescue workforce experiencing burnout, anxiety, or depression may demonstrate reduced clinical performance and impaired decision-making (9). Furthermore, unhealthy lifestyle shifts can have long-term consequences for chronic disease burden within this essential workforce (10). Despite this recognized importance, significant research gaps remain. Existing studies often focus on healthcare workers in hospital settings, with less attention paid to the broader ecosystem of rescue personnel (11). Fewer studies integrate a comprehensive assessment of concurrent lifestyle changes across multiple domains (12). There is also limited evidence on how different “living statuses” modulate these risks (13), and the Chinese context remains under-explored (14, 15).

This study aims to address these gaps by conducting a comprehensive, cross-sectional assessment of lifestyle behaviors and mental health status among a large cohort of rescue workers in China during a public health emergency. Specifically, we seek to: (1) quantify and compare changes in key lifestyle domains; (2) assess the prevalence of anxiety and depressive symptoms and their correlation with lifestyle changes; (3) analyze how different living/working statuses modulate risks; and (4) identify independent predictors of psychological distress. By elucidating the interconnected patterns of behavioral and psychological adaptation under crisis conditions, this research aims to generate an evidence base for developing integrated support strategies to protect rescue workers and strengthen long-term biosecurity resilience.

This comprehensive study provides robust empirical evidence elucidating the profound and multi-layered impact of a major public health emergency on frontline rescue workers. Through systematic assessment of lifestyle behaviors and mental health status using validated instruments (16, 17), we document a concerning pattern of deterioration across multiple health domains. The simultaneous rise in substance use, digital overuse, sleep problems, physical inactivity, and clinical-level psychological distress reveals the heightened vulnerability of this essential workforce during crisis response. This underscores their critical role as the human infrastructure of epidemic containment, as emphasized by Page et al. (1). Moreover, our multivariate and stratified analyses go beyond identifying associations—they point to potential mechanisms and pinpoint high-risk subgroups, offering clear targets for intervention.

The observed increases in smoking and alcohol consumption among rescue workers reflect well-documented stress-response patterns (18, 19). The 13.98% increase in smoking prevalence aligns with global reports during COVID-19, where stress and anxiety were primary drivers for increased tobacco use among healthcare workers, a finding corroborated by Gaiha et al. (4). However, our stratified analysis reveals crucial nuance: those in solitary isolation showed the highest rates of increased smoking, followed by closed-off management. This gradient effect suggests that confinement intensity, rather than pandemic exposure per se, drives substance use escalation, aligning with research on the profound psychological impact of quarantine by Brooks et al. (7). This behavior reflects isolation-induced anxiety, driven neurobiologically by stress-induced dysregulation of the hypothalamic–pituitary–adrenal axis. Such dysregulation elevates cravings for substances that provide temporary relief, as supported by Keyes et al.’s (18) epidemiological study and McKee et al.’s (19) psychopharmacological research. The more modest but significant increase in alcohol consumption (6.02%) requires careful interpretation. While lower than some general population studies, this difference may reflect professional norms and access limitations. Consistent with prior studies, our results indicate that solitary isolation posed the highest risk. This aligns with findings by Koopmann et al. (20) and Neill et al. (21), which linked pandemic-related changes in drinking behavior to living conditions and heightened stress. This pattern of using substances to cope with acute occupational stress finds precedent in earlier outbreaks, as shown by Wu et al. (22) during the SARS epidemic.

The significant increase in mobile phone usage (median increase from 2 to 3 h daily) and its positive correlation with depressive symptoms (r = 0.25, p < 0.001) illuminates the complex role of technology during public health emergencies. Our findings support and extend prior research on media exposure and mental health by Zhao et al. (23) and Gao et al. (24), suggesting that for rescue workers, prolonged digital engagement may represent both a coping strategy and a stressor. This increase served multiple purposes: facilitating operational communication, seeking crisis-related information, maintaining social connections during isolation—a critical need noted by Brooks et al. (7)—and providing entertainment amid restricted mobility. These patterns align with technology use observed under confinement by Ni et al. (25).

However, the correlation with depression scores suggests potential negative consequences. Zhao et al. (23) found that excessive social media use was associated with higher anxiety and depression, mediated by COVID-19-related fear and information overload. This aligns with the concept of problematic digital media consumption introduced in the introduction (6). Our regression analysis, which showed mobile phone use as a marginal predictor when controlling for other factors, suggests its effect may be mediated through other pathways like sleep disruption or physical inactivity. This nuanced understanding, moving beyond simple screen time reduction, is crucial for developing targeted digital wellness interventions.

The sleep disturbances documented in our study—particularly the significant increase in nocturnal awakenings—provide important insights into the physiological impact of emergency response work under restrictive conditions. The 4.08% of rescuers experiencing 3–5 nocturnal awakenings per week post-outbreak represents a meaningful clinical concern, as sleep fragmentation of this magnitude is associated with impaired cognitive function and increased cardiovascular risk (26). Our reported rates are contextualized by the meta-analysis of Jahrami et al. (5), which found a 36% pooled prevalence of sleep problems among healthcare workers during the pandemic.

Our result offers novel visual evidence of how environmental restrictions exacerbate sleep problems. The gradient from normal work to solitary isolation strongly supports the hypothesis that confinement-related stressors directly impair sleep continuity. This may operate through several pathways, as identified by Pappa et al. (27) in their systematic review, where factors like longer duty hours were significantly associated with sleep disturbances. We observed a distinct pattern: sleep latency remained stable, but sleep maintenance worsened significantly. This aligns with Vgontzas et al.’s (28) classification of insomnia phenotypes, suggesting our cohort primarily experienced middle—rather than initial—insomnia.

The connection between sleep disruption and mental health is bidirectional and potent. Barros et al. (26) demonstrated that poor sleep quality increases depression risk by 61% after adjusting for confounders. The strong association suggests that sleep disturbances in our sample may both result from and contribute to psychological distress. This forms a vicious cycle, which is particularly relevant in high-stress occupational groups, as also noted in rescue-specific research by Xiao et al. (29).

The dramatic shift in physical activity patterns—with 17.68% engaging in no exercise post-outbreak compared to 6.56% pre-outbreak—represents one of the most concerning lifestyle changes documented. This 2.7-fold increase in sedentary behavior has significant implications for both physical and mental health (30). Mechanistically, reduced exercise decreases production of neurotrophic factors like brain-derived neurotrophic factor (BDNF), which supports neuronal health and regulates mood, a key pathway explored by Erickson et al. (31). The strong association between decreased activity and poorer mental health outcomes has been documented during COVID-19 restrictions by Lesser and Nienhuis (6) and Stanton et al. (12).

Our regression analysis identifying decreased physical activity as an independent predictor of depression scores (β = 1.54, p < 0.001) provides strong evidence for targeting exercise in interventions. This aligns with Kandola et al.’s (32) research on the antidepressant mechanisms of physical activity. The combination of increased sedentary behavior, adverse changes in substance use, and sleep disturbances creates a perfect storm for metabolic syndrome development. This may lead to long-term health consequences beyond the immediate emergency period, as discussed in reviews on inactivity and chronic disease by Booth et al. (33).

The doubling of clinically significant anxiety (from 8.5 to 18.6%) and depression (from 6.7 to 15.2%) represents the study’s most urgent finding. These rates, while concerning, are actually conservative compared to some healthcare worker studies during COVID-19, which reported anxiety up to 44.6% and depression up to 50.7% in a meta-analysis by Salari et al. (34). This discrepancy may reflect our sample’s military background, which typically emphasizes resilience training, or measurement timing differences.

Multivariate logistic regression showed that reduced physical activity (depression OR 2.34, anxiety OR 2.01), lower per capita monthly household income (<6,000 CNY), and being in home quarantine or isolated living status were significantly associated with a higher risk of moderate-to-severe depression and anxiety, whereas being married had a protective effect against depression (OR 0.68). Female rescue workers and those aged >28 years also exhibited greater susceptibility to depression. The younger average age of our sample (27.2 ± 7.0 years) warrants special consideration, as younger adults have shown particular vulnerability to pandemic-related mental health impacts (35, 36). This demographic factor may interact with the high-stress, high-responsibility nature of rescue work to create unique vulnerabilities requiring age-tailored support approaches.

The interconnectedness of our findings necessitates moving beyond single-issue interventions toward integrated systems approaches. We propose three concrete implications based on our findings and the broader literature. First, operational modifications must prioritize psychological safety alongside physical safety. Rotation schedules should minimize continuous isolation periods, as suggested by Greenberg et al. (9). Physical spaces in restricted settings should be designed to facilitate movement, social connection, and sleep hygiene. Second, mental health support must be proactive, normalized, and integrated. Regular screening using brief validated tools like the PHQ-9 and GAD-7 (16, 17) should be implemented during emergencies. Peer support programs have shown particular effectiveness, as evidenced in the systematic review by Pollock et al. (37). Third, biosecurity policy must explicitly include workforce psychological resilience. As nations develop strategic reserves for future pandemics, these should encompass not just physical protective equipment but also mental health resources—a concept akin to the “psychological personal protective equipment” emphasized by Holmes et al. (8). Investment in this “psychological infrastructure” is a strategic necessity for maintaining operational capacity during prolonged emergencies (38, 39).

While providing important insights, this study has limitations that future research should address. Although we rechecked the original dataset and excluded two questionnaires with logically inconsistent combinations of age and years of work, the possibility of residual reporting or recording errors cannot be completely ruled out. The cross-sectional retrospective design, with a single-timepoint assessment and no longitudinal follow-up, prevents causal inference about the observed associations and limits our ability to describe trajectories of lifestyle and mental health over time. Self-report measures, while practical for large-scale assessment, are subject to recall and social desirability biases. In addition, the sample consisted of Chinese military rescue workers from a single unit, which restricts generalizability to civilian rescue workers, other frontline occupations, and non-Chinese settings.

Future research should employ longitudinal designs with multiple assessments from pre-crisis through recovery. Objective measures like actigraphy for sleep, carbon monoxide monitoring for smoking verification, and mobile device usage tracking would strengthen findings. Qualitative components could illuminate the lived experience behind the quantitative changes. Comparative studies across different types of emergency responders and cultural contexts would clarify which findings are universal versus context-specific. Intervention trials testing the integrated strategies suggested by our findings are urgently needed.

This study systematically documents the significant toll that public health emergencies exact on rescue workers across behavioral, psychological, and physiological domains. We move beyond documenting problems to identifying specific risk factors and vulnerable subgroups, offering a roadmap for targeted intervention. The convergence of lifestyle deterioration and mental health decline creates a syndemic requiring integrated solutions that address the whole person in their environmental context. As the world prepares for future biological threats, protecting those on the front lines must extend beyond physical protective equipment to encompass psychological armor and systems that support sustainable resilience. In practical terms, our findings support the implementation of a scheduled mental health screening protocol (e.g., routine PHQ-9 and GAD-7 assessments) combined with ongoing monitoring of key lifestyle behaviors and timely, targeted interventions. Implementing these evidence-based, multifaceted strategies represents both an ethical imperative for those who serve and a strategic necessity for maintaining response capacity during prolonged crises.