The Coronavirus Disease 2019 (COVID-19) global pandemic has resulted in more than 650 million cases confirmed (more than 7 million deaths, and more than 28 thousand in Portugal, reaching its peak of deaths of more than 25 thousand in February 2021) by the end of June 2024. Many countries have implemented measures to contain the spread of the virus (from self-isolation to public health sanitary guidance) that have significantly impacted the lives of populations worldwide (Bedford et al., 2020), with long-term psychopathological symptoms (Hunt et al., 2022), mainly when psychopathology (e.g., depression and anxiety) was already a leading cause of disability and burden globally pre-pandemic (Vos et al., 2020). Indeed, COVID-19 survivors present an elevated risk of developing psychopathological symptoms (Ma et al., 2020), including post-traumatic stress symptoms (PTSD), with significant levels of distress and impairment (Tarsitani et al., 2021). Also, psychopathological symptoms (e.g., depression) were found in patients weeks (Renaud-Charest et al., 2021) and months after recovery (Guideline NG193 N, 2021), mainly, but not exclusively, in cases of severe COVID-19 (Weerahandi et al., 2021). It is suggested that psychopathological vulnerability during the pandemic may not only result from the interplay between inflammatory mediators and neurotransmitters directly resulting from infection (e.g., ‘cytokine storm’; Kempuraj et al., 2020), but also from elevated neuroinflammation indicators of stress (e.g., the psychologically mediated impact of lockdown, social distancing, and overall changes in lifestyle). However, some evidence suggests that symptoms of depression and anxiety decreased during lockdown (e.g., Rehman et al., 2023). This calls for a better understanding of the sociodemographic, interpersonal, and individual psychological factors that contribute to mitigating or exacerbating psychopathological risk during the pandemic.

Studies with multicultural samples found that the perceived threat of COVID-19 is a major contributor to depression, anxiety, and post-traumatic stress (Matos et al., 2021b), with contextual (e.g., financial stability) and psychological (e.g., fear of infection) factors impacting these symptoms (Di Crosta et al., 2020). On the other hand, some studies with samples of young adults suggest that fear of COVID-19 is associated with more adherence to preventive behaviors and thus with more satisfaction with life (Green and Yıldırım, 2022), which calls for the need to further understand the (mal)adaptiveness role of threat-based psychological processes. Parenting children with severe diseases, as well as experiencing stressful life events, seem to be risk factors for psychopathology during the pandemic (Guideline NG155 N, 2020), echoing decades-long psychological research (Kendler et al., 1999). Indeed, COVID-related anxiety is associated with the anticipation of life changes, job loss, and socioeconomic difficulties (Schonfeld et al., 2023), and thoughts about COVID-related death are associated with depression (Fairlamb, 2022). Notwithstanding, resilience (Rossi et al., 2021) and intrapersonal coping mechanisms (Wright et al., 2022) seem to be vital to understanding the vulnerability to and protection against the development of psychopathological symptoms during the pandemic.

Resilience is broadly conceptualized as the ability to bounce back from stressful events and overall adversity (e.g., loss, trauma), resulting in a stable trajectory of psychological health (Bonanno, 2004). During the COVID-19 pandemic, resilience has been brought to the center stage of discussions on mental health (Killgore et al., 2020), and was found to be a crucial factor in protecting against psychopathology and psychological distress and a buffer of the impact of demographic risk factors (e.g., age, gender, education), as well as health-related ones (e.g., chronic illness, SARS-Cov-2 infection, alcohol consumption) (Arslan and Yıldırım, 2021; Tudehope et al., 2022). Nevertheless, although a helpful construct, resilience only provides an overall snapshot rather than a multidimensional picture of the many factors that might contribute to resilience.

Psychological flexibility, defined as the ability to maintain or change behavior according to direct environmental cues and to adapt to situational demands to pursue a valued and meaningful life (Hayes et al., 2006), has been put forward as a critical resilience factor during the COVID-19 pandemic (McCracken et al., 2022), and was found a predictor of better mental health outcomes (Prudenzi et al., 2023) and of chronic anxiety and depression (Hemi et al., 2023) during the COVID-19 pandemic. Based on extensive research on its protective role against psychopathology (Cherry et al., 2021), psychological flexibility is especially relevant during global uncertainty. Indeed, psychological flexibility appears to be cross-sectionally associated with well-being outcomes during the pandemic, is associated with less peritraumatic distress (Kroska et al., 2020), and mitigates the nefarious impact of the COVID-19 lockdown-related risk factors on COVID-19 peritraumatic distress, depression, and anxiety (Pakenham et al., 2020). Although the protective role of psychological flexibility during the COVID-19 pandemic has been extensively studied, samples have been composed of a mix of infected and non-infected participants (see Yao et al., 2023), which may not rigorously grasp the specific impact of psychosocial factors in the mental health of COVID-19 survivors, such as the fear of infecting others, and the shame and guilt associated to ethically tricky decisions (e.g., providing care to loved ones in need) (Haller et al., 2020).

The detrimental impact of shame on mental health in the context of the COVID-19 pandemic has been suggested (Cavalera, 2020), but only a few studies have focused on COVID-related shame. Although clinicians recognize illness-related shame as a relevant health determinant, few studies have empirically explored its negative associations with psychological health in illness populations (Trindade et al., 2018). These associations are potentially more appropriate in an infectious condition such as COVID-19, where individual behaviour is directly accounted for its contraction. Narrative studies suggest that COVID-19 patients may experience internalized shame for having contracted and/or contaminated others (Sahoo et al., 2020), and shame seems to be a mechanism underlying threatening COVID-19 illness perception and psychological distress (Hamama and Levin-Dagan, 2022). These data suggest that the role of COVID-related shame should be duly explored.

The current study aimed to explore the role of contextual (major life events, previous psychiatric diagnoses, job loss) and psychological (COVID-related shame, resilience, psychological flexibility) factors as longitudinal predictors of symptoms of PTSD, anxiety, and depression over 1 year during COVID-19 pandemic, in a sample of Portuguese SARS-Cov-2 survivors.

This study examined the long-term impact of psychosocial factors on symptoms of PTSD, anxiety, and depression in a sample of Portuguese SARS-CoV-2 survivors. Results showed that SARS-CoV-2 survivors present high symptoms of depression and anxiety: approximately 60% present mild to severe symptoms of anxiety, and approximately 35% present symptoms of depression. Pre-pandemic normative data in community samples (Breeman et al., 2015) suggest that more than 80% present normal levels of anxiety and depression (see Zigmond and Snaith, 1983 for cut-off scores). The elevated symptoms of anxiety and depression seem to echo previous results on the nefarious impact of the COVID-19 pandemic on the population’s mental health (Penninx et al., 2022). Results of this study also showed that 42% of SARS-CoV-2 survivors present clinically significant symptoms that support a possible diagnosis of PTSD. This result is in accordance with many previous studies that found PTSD symptoms during the COVID-19 pandemic (Giannopoulou et al., 2021), including in multicultural samples (Matos et al., 2021a), although higher than found in other studies (Yuan et al., 2021). In addition to contextual stressors related to the pandemic and/or its mitigation measures (e.g., our sample was more than 2 weeks in isolation, 8.1% lost jobs during the pandemic, and 10.5% lost someone due to COVID-19), intrapersonal psychological processes may have contributed to the elevated symptoms of PTSD, anxiety, and depression. This seems to be the case in our sample, in which, although the majority of participants have recovered from the infection without prescribed medication (62.2%) (which suggests mild illness), they nonetheless showed regular concern for infecting others (65.9%).

Several regression models were tested according to the significance of the longitudinal correlations to test the contextual and psychological factors longitudinally predicting PTSD, anxiety, and depression. Having children, as well as having more recent major life events, were significant predictors of PTSD symptoms measured 1 year later. This goes in line with what cross-sectional literature has shown regarding the parental psychopathological symptoms during the COVID-19 pandemic, where parents reported struggles in balancing professional demands, homeschooling children, and overall domestic tasks (Davis et al., 2021). This seems to suggest that the traumatic effects of the pandemic are not exclusively experienced by parents of most vulnerable children (Guideline NG155 N, 2020) but rather by parents in general, who seem to experience long-term PTSD symptoms more than those without children. Although studies on the impact of major life events on mental health during the pandemic have been scarcely explored, our results seem to resonate with cross-sectional literature on the effects of daily hassles (e.g., unemployment, family discord) and stressful life events (e.g., economic, job, and housing difficulties) on mental health (Rossi et al., 2021).

In addition to these contextual factors, psychological flexibility was a significant predictor of PTSD symptoms. Consistent with previous studies (McCracken et al., 2022), this suggests that the ability to flexibly choose to maintain or change behavior according to environmental cues and personal values, particularly during the second lockdown (when our baseline assessment was conducted), was a protective factor against PTSD symptoms. This was also the case for anxiety symptoms: not only those who experienced fewer major life events during the pandemic, but also those who were more psychologically flexible, presented lower levels of anxiety 1 year later. These results seem to add to those findings by providing a longitudinal picture of the protective role of psychological flexibility against psychopathological symptoms during the pandemic. Interestingly enough, when it comes to depressive symptoms, major life events were the only significant predictor. While PTSD and anxiety, by their evolutionary programming (see Price, 2003), might involve overstimulation of threat/protection-focused systems (including psychological processes that activate “better safe than sorry” behavioral rules inherent in psychological inflexibility), depression is hypothesized to lay on other evolutionary routes, such as defeat and entrapment (Gilbert and Allan, 1998), which major life events such as unemployment, divorce, and loss may activate. In fact, major stressful life events are well-known predictors of depression (Tennant, 2002).

It should be noted that neither covid-related shame nor resilience had a significant long-term impact on symptoms of PTSD, anxiety, and depression. These results seem to indicate that although lower levels of shame and higher levels of resilience were associated with better mental health indicators during lockdown (T1), their long-term impact was not significant when other contextual and psychological variables were controlled. It may be the case that resilience is a more important factor in other symptomatology, perhaps more salient than PTSD, such as adjustment-related symptoms and disorder (for a discussion, see (Brunet et al., 2022)).

This study was not short of limitations, which should be considered when interpreting these results. Firstly, our sample was not balanced in terms of gender, given that the majority of participants were women. A different pattern of associations and predictions might be present when considering other genders. Also, the majority of our sample was entirely cisgender and binary, which precludes us from extrapolating to queer and gender non-conforming individuals whose mental health and financial resources seem to have been significantly impacted by the pandemic (Nowaskie and Roesler, 2022). We have also not accounted for specific socioeconomic variables, such as income, which might be a buffer of the family burden resulting from the pandemic mitigation measures. Additionally, sample selection bias due to online data collection should be acknowledged. Although studies suggest that self-report measures are psychometrically equivalent in online and paper-and-pencil data collections (Trindade et al., 2021), the generalizability of results is unwarranted, given that online data collection might yield bias due to unattentiveness and lack of sociodemographic representativeness. Also, the high attrition rate (70.83%) should be considered when interpreting results, as these might reflect a biased self-selected sample. Additionally, the fact that we have collected data in two-time points instead of multiple ones (>3) prevents us from conducting much-needed trajectory analyses that would provide a picture on changes of each variable over time. Future studies, with larger sample sizes that would ensure sufficient statistical power to test structural equation modeling, should conduct latent growth curve model analyses to explore the rate of changes over time (the slope factor). Also noteworthy is the fact that we have yet to explore indirect or interaction effects underlying the relationship between psychopathological symptoms at different time points. Future studies with sufficiently large sample sizes that would not compromise statistical power should explore these relationships, not only through mediational and moderation analyses, but also through network analyses that would provide evidence on the complex interconnection and inter-dependence between symptoms, contextual factors, and psychological processes. Finally, these results should consider the specificities of governmental measures before a cross-national generalization. The Portuguese government implemented strict measures to mitigate the pandemic compared to other European countries, which might have had a specific considerable impact on financial loss and mental health.

Overall, these results seem to suggest the importance of considering both contextual and psychological variables when developing comprehensive models, as well as mental health prevention and intervention programs, during periods of global pandemic. These results suggest the long-term detrimental effects of pandemic periods on PTSD, anxiety, and depression, as well as point out the predictive role of parenting, major life events, and psychological flexibility. This seems to suggest that measures promoting mental health during pandemic crises should focus on reducing parental burden and socioeconomic losses, as well as implementing evidence-based programs aiming to promote protective psychological skills, such as psychological flexibility (e.g., Acceptance and Commitment Therapy).