The COVID-19 pandemic has not only caused almost 800 million cases and 7 million deaths globally, but also the long-lasting and profound changes at the personal level, such as social interactions, lifestyles, and mental-wellbeing. It also affected the population at the social level including the global economy, healthcare system, and social inequalities (1, 2). In October 2021, the World Health Organization (WHO) defined Long-COVID (LC) as the condition typically occurs in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually 3 months after onset, lasting for at least 2 months and cannot be explained by any alternative diagnosis (3). With global numbers of infections of COVID-19 exceeding 500 million, a conservative estimate of people being currently affected by LC worldwide could be more than 100 million (4). Typical acute COVID-19 cases are characterised by respiratory symptoms, fever, and neurological symptoms (5). The duration and severity of acute COVID-19 cases vary, with some patients being asymptomatic, while others require hospitalisation and mechanical ventilation. The average duration of a COVID-19 infection is typically shorter than 4 weeks (6). The new-onset conditions including cardiovascular, thrombotic/cerebrovascular disease, type 2 diabetes, myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and dysautonomia, especially postural orthostatic tachycardia syndrome (POTS), could last for a long time even years. These symptoms could be commonly called LC. It also have different labels, such as post-acute sequelae of COVID-19, post-acute COVID-19, or post-COVID-19 condition (7).

It has been widely reported that the risk of psychiatric symptoms was associated with COVID-19 infection. A recent meta-analysis revealed that up to one in four patients experienced neuropsychiatric symptoms spanning sleep disorders, fatigue, anxiety, and post-traumatic stress (PTSD) after the onset of COVID-19 with an approximate follow-up duration of 77 days (8). Studies have further explored potential correlations between the risk of LC and mental health problems. Specifically, LC was found to be associated with increased display of depression, anxiety, PTSD, and weakened life satisfaction (9), which might be partially attributed to the persistent physical symptoms of LC (10). In China, few studies have focused on evaluating the impact of LC on COVID survivors. Zhao et al.’s study (11) found that approximately 9.6% of hospitalised patients suffering from moderate impairment reported mainly mental health and cognitive symptoms 20 months after recovery. Several factors were reported to be predictive of long-term physical and cognitive symptoms, which included age, hospital stay, sex, and comorbidities. However, the sample was recruited in 2021 and was not based on a randomised sampling method. Another study conducted in China has found that having at least one LC symptom increased the risk of depression or anxiety by 3.44-fold (12). However, the study sample was limited to hospitalised patients recruited in 2020. Given the limitations of these studies, there is a need to further study the mental health effect of LC, particularly in patients who were infected in the later period of the pandemic using a random community-based sample.

The extent of suffering from chronic illness is not just determined by the severity of the illness itself, but rather moderated by external and subjective factors (13). As The Burden of Treatment Theory posits, management of chronic conditions involves routine work for patients to control their illness. As treatment burdens accumulate, some patients become overwhelmed leading to poorer outcomes, stress on caregivers, and increased healthcare costs (14). In addition, personality factors also play an important role when patients cope with the illness. According to Leventhal’s self-regulatory model, patients’ “illness representations” (beliefs about cause, timeline, consequences, etc.), reflecting their perceived burden, affect their self-management behaviors, psychological adjustment, and mental health (15). Therefore, the perceived burden of illness of patients suffering from Long-COVID, like any other chronic conditions, may also have a direct impact on their mental health.

In terms of the relationship between the perceived severity of LC symptoms and the mental health of patients, few studies have been conducted. In the study by Sivan and colleagues, it was found a positive correlation between the perceived severity of 12 types of LC symptoms and cognitive symptoms, PTSD, depression, and anxiety (16). However, in another study of patients with LC symptoms, the perceived severity of autonomic dysfunction and anxiety, as well as emotional well-being were not positively correlated (17).

Therefore, this study aims to describe the common symptoms related to LC and to quantify the prevalence of these symptoms among community-dwelling COVID patients. It also aims to examine the relationship between the LC symptomology and the mental health status of these patients, as well as the relationship between the perceived burden of the LC symptoms and the level of mental health problems. It is hypothesised that LC symptomatology is associated with the mental health of patients and that there is a positive dose-response relationship between the perceived burden of LC symptomatology and the severity of mental health problems.

In total, 781 patients tested positive with COVID-19 were recruited. Of these 482 respondents provided useful information for data analyses, after removing 299 questionnaires having inconsistent answers on geographical information or confirmation date of infection. A comparison between the respondents and non-respondents indicated no differences in demographic variables. Table 1 presents patients’ demographics, basic health information, and scores on mental health measures. Of the total 482 respondents, there were 195 (40.5%) males, with 374 (77.6%) receiving three or more doses of vaccine, and the most commonly received vaccine was the domestically produced inactivated vaccine (66.8%), followed by the domestically produced recombinant protein vaccine (30.5%). The most prevalent symptom of acute-phase infection reported by participants was fever (87.1%), followed by muscle and joint pain (56.0%), cough (47.9%), fatigue (42.7%), headache (41.9%), ENT abnormalities (25.1%), and finally shortness of breath (12.4%). Seven (n=7, 1.5%) of the respondents reported pre-morbid mental health problems, and 6 (1.2%) were hospitalised during the acute phase of the infection, with half (50%) of these hospitalised patients receiving treatment in the ICU. In terms of the outcome measures of the study, namely the mental health status, 61 (12.7%) exhibited symptoms of severe to extremely severe depression symptoms (mean=15.9, s.d=4.7), 26.8% (n=129) anxiety (mean=12.2, s.d=4.8), and 25.1% (n=121) stress (mean=17.2, s.d=3.8).

The results of LC symptomology were summarised in Table 2. As shown, among all respondents, 25.5% had symptoms that persisted for more than two months and could not be explained by causes other than COVID-19 infection. In addition, the top five common symptoms were postexertional malaise (22.2%), fatigue (21.2%), brain fog (14.3%), sleep (11.6%), shortness of breath (11.4%), and palpitations (10.2%).

The unadjusted bivariate relationships between LC symptomology, other study variables, and depression, anxiety, and stress were summarised in Table 3. As shown, a few variables were related to the mental health variables. LC symptomology, resilience, social support, and premorbid mental health problems were significantly related to all three aspects of mental health of interest. In addition, education level was significantly associated with stress and marginally related to anxiety. All other variables were insignificantly related to the mental health variables with a p-value much greater than 0.1. Hence, they were not included in further analyses. The results obtained from the multiple linear regression analyses were presented in Table 4. After adjusting for other variables in the model, the LC symptomology was still positively and significantly associated with depression (t=2.09, p=0.037) and anxiety (t=4.51, p<0.001), but not stress. The regression coefficient of LC symptomology was 1.13 (s.e. = 0.54) for depression suggesting a difference in the depression score of more than 1 unit between patients with LC symptomology and those without. The regression coefficient of LC symptomology for anxiety was 2.35 (s.e. = 0.52) indicating a difference of more than 2.3 units in the anxiety score between groups. In terms of the interaction terms between LC symptomology and resilience, as well as social support, the results indicated that none were significant even at the 5% types I error rate.

For the relationships between the perceived burden of LC symptoms and mental health, the results were presented in Table 5. As shown, after adjusting for variables including premorbid mental health problem, education level, resilience, and social support, positive and significant associations were found between the perceived burden of LC and depression, anxiety, and stress. The regression coefficients of the perceived burden of LC for depression and stress were 0.38 (s.e. = 0.07) and 0.35 (s.e. = 0.08) respectively, indicating, on average, for each increase in 1 unit of perceived burden measure there was a corresponding increase of 0.38 units in the depression and 0.35 in the stress scores. Among the three mental health variables, anxiety had the largest increase in the score with an increase of each unit of perceived burden of LC there was an increase of 0.54 units (s.e. = 0.07, t=8.24, p<0.001). These results suggested a positive dose-response relationship between the perceived burden of LC symptoms and mental health problems. As in the previous analyses, none of the interaction terms between perceived burden and resilience, as well as social support were significant.

This study aimed to investigate the prevalence of common symptoms of LC, and the propensity of LC in a community sample of COVID-19 patients. It also aimed to examine the relationship between LC symptomology and the mental health status, as well as the relationship between the perceived symptom burden of LC and mental health problems. The results suggested that patients exhibited the overall LC symptomology was about 25% among patients who had tested positive for COVID-19 after December 2022. Postexertional malaise and fatigue were the most prevalent LC symptoms, accounting for 22.2% and 21.2% of all LC patients, followed by cognitive impairment including brain frog (14.3%), and sleep problems (11.6%). The results further suggest a positive and significant association between LC symptomatology and poor mental health, particularly depression and anxiety. In terms of the perceived burden of LC symptoms, it is related to the mental health of patients also demonstrating a positive dose-response relationship with depression, anxiety, and stress.

The results on symptomatology, in general, are consistent with those reported in the literature, though some minor discrepancies are found in other less commonly reported symptoms, such as headaches and muscle aches that have been ranked less prevalent in the current study (29, 30). Furthermore, some differences in the prevalence of symptoms have also been identified compared to previous studies on COVID-19 patients one-year post-infection in China (12, 31). In Zhang’s study, the prevalence of muscle and joint pain was 11.8%, whereas in the current study, it was 8.3% (12). While the prevalence of brain fog and palpitations were 14.3% and 10.2% in this study, they are higher than that reported in the literature of 2.2% and 5.8% (31) This study also found that patients with LC symptomology scored significantly higher in depression and anxiety, but not stress. In terms of the propensity of these symptoms, these patients reported postexertional malaise, fatigue, brain fog, shortness of breath, and insomnia, as the most common symptoms, indicating that these symptoms had a greater impact on people’s lives and work. In the present study, postexertional malaise (PEM), a distinguishing feature of chronic fatigue syndrome (CFS) (32), was most prevalent and reported to be the most burdensome among other symptoms identified as post-COVID-19 sequelae. There could be many explanations for this result. One possible reason may relate to the potential physiological responses for the infection in different body parts including the brain, such as neuroinflammation (33). This has highlighted how the chronic and debilitating fatigue associated with CFS can significantly impact patients’ work productivity and quality of life, posing psychological distress (34). Moreover, symptoms related to CFS are relatively difficult to diagnose in hospitals to receive appropriate rehabilitation recommendations (35).

For the relationship of LC and mental health, the changes to routine happened due to pandemic, such as financial instability, social distance and mask wearing, have been challenging people’s psychological needs that previously fulfilled for about 3 years (36–38). Moreover, as for patients developed LC symptoms, they might experience additional symptoms related to brain function leading to a worse mental health outcome compared to those who recovered well from COVID-19 infection (39). The perceived burden of LC symptoms further exhibits the impact of the disease on the mental health of patients. There could be many reasons attributing to the interplay between the actual severity of the physical illness, treatment demands and individuals’ perceptions of their ability to adapt (40). According to the Burden of Treatment Theory (14), the perceived disease burden in LC patients may be aggregated by the difficulties of managing the symptoms. Overall, medical expenses on the treatment, which was estimated at $9,000 per person annually, could be a stress for many people (41). Moreover, the increased stringency of containment measures during the pandemic resulted in unemployment also exerted additional pressure on patients (42). As a result, patients who would like to retain their jobs are more likely to keep working and not take any medical leaves, thus becoming an additional layer of burden.

The findings highlight the need to provide mental health support for those severely affected by COVID-19 infection, particularly those who developed LC symptoms. In 2020, the National Health Commission of China issued “Rehabilitation Program for Discharged COVID-19 Patients” (43) recommending respiratory training, physical exercises, psychological support, and activities to regain daily living abilities to address residual respiratory, physical, and psychological dysfunctions like cough, fatigue, and anxiety that may persist post-hospitalisation. To date, the treatment strategies for LC in designated clinics or outpatient services mainly focused on symptom relief (44), with both Western and Chinese medicine focusing on physical rehabilitation (45, 46). In terms of mental health services in general, and particularly for patients with LC symptoms, it is a lacking area. China has very few numbers of mental health professionals relative to its large population (47). There is also a lack of mental health rehabilitation services, and the utilisation of telemedicine is also low mainly due to the limited access to facilities (47). With such limited resources, well-designed and validated preventive programs using a mHealth approach for early detection, intervention, treatment and management are urgently needed for the patient population with LC symptoms (48, 49).

The present study provides evidence that LC symptomology is associated with mental health problems of patients and that there is a dose-response relationship between the burden of LC symptoms and mental health problems. Long-term management and rehabilitation are essential in providing support to these patients. Continued investigation of how LC may influence mental well-being may also provide insights to guide targeted interventions and management strategies.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The informed consent and study protocol was confirmed by the Macau University of Science and Technology Medical Ethics Committee (MUST-HSS-20230505001).

YL: Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft. LL: Formal analysis, Supervision, Writing – review & editing. YX: Conceptualization, Supervision, Writing – review & editing. ZQ: Investigation, Validation, Writing – original draft. YZ: Investigation, Validation, Writing – original draft.