Resilience And Mental Health Among Young Adults In The Context of Lockdown Due To The COVID-19 Pandemic

Purpose: Psychological resilience is dened as the ability of an individual to adapt to adverse situations and in a certain way regulate mental health. Currently, the COVID-19 pandemic and lockdown measures implemented to avoid infection are considered risk agents for the development of mental disorders. The objective of this study is to explore the relationship between resilience capacity and the presence of depressive symptoms and poor sleep quality in the context of lockdown due to the pandemic. Methods: The sample consisted of n = 8,426 young adults who were recruited via convenience sampling. Sociodemographic and mental health data was collected through online surveys. Previously validated test were used to collect information on depression and sleep quality. Restriction of mobility was evaluated using information from Google Mobility Reports. Results: Most of participants were highly educated women. Independently from sex, age and other confounders, individuals with psychological resilience were less likely to develop depressive symptoms or sleep disorders during connement due to the COVID-19 pandemic. Moreover, higher mobility restriction was associated with depression but not with bad sleep quality. In this way, a longest stay at home and a higher restriction of mobility to groceries and pharmacies were related to depression. Conclusions: These results show evidence of the importance of psychological resilience on mental health, and can contribute to the development of preventive public health interventions to face the negative effects of connement on mental health.


Introduction
The impact of the COVID-19 pandemic, caused by the SARS-CoV-2 coronavirus, has generated an unprecedented damage to public health and therefore to the health of individuals. The rapid spread of the virus, and the and lack of knowledge about its prevention and treatment, have resulted in an alarming increase of infected patients, who in some cases have experienced long-lasting effects or even death [1].
While it is clear that the COVID-19 pandemic represents a serious threat to physical health, it has been suggested that it may also affect the mental health of the population [2].
For instance, recent literature indicates that individuals who have been affected by COVID-19 are likely to display psychological problems that in turn could trigger severe mental disorders [1][2][3][4]. Previous research also argues that at least one out of every ve new patients diagnosed with COVID-19 presents psychological alterations which result in poor mental health [5][6]. Similarly, it has been suggested that patients who have recovered from the disease might manifest some type of psychopathology after several years [7][8].
Individuals suffering from COVID-19, as well as those who have recovered from the virus, display depressive symptoms, anxiety disorders, post-traumatic stress disorder, panic attack, psychomotor agitation, and sleep disorders more frequently than the rest of the population [1,5,[9][10][11]. Despite the fact that most of the issues associated with mental health were detected in COVID-19 patients, there is empirical evidence that suggests an alteration of mental health in the general population. For example, it has been observed that non-infected individuals may show some type of psychopathological symptoms, largely caused by the lockdown measures established to reduce widespread infection [7].
In different countries around the world, ordinances were established that required citizens to use biohazard protection implements such as facemasks, practice social distancing, and to stay in con nement, avoiding unnecessary travel to reduce the rate of infection [12]. These guidelines aimed at protecting physical health overlooked the mental health of the population, becoming high risk factors for the development of mental disorders which can be aggravated by the presence of other risk factors such as low income, unemployment, student dropout, and interpersonal con icts [13].
In addition, it has been suggested that experiencing prolonged periods of self-isolation or going back to previous stages of isolation result in an increase in anxiety and depression symptoms, stress and loss of sleep [14]. Furthermore, research shows that mobility restrictions, as well as the prohibition of trips and outings during con nement can cause a decrease in social communication, less interaction with peers, and fewer relationships with the community, contributing to the appearance of feelings of loneliness, frustration, hopelessness, and emotional instability; characteristics that modulate the development of psychopathological disorders. [8, [15][16][17].
Although self-isolation itself plays an important role in the increment of mental health problems in the general population, there are other factors that can precipitate and trigger the psychopathological symptoms mentioned above. For instance, it has been found that people who have less interpersonal contact within their household, tend to report higher scores in anxiety, depression, sleep disorders and stress [18]. Likewise, recent studies suggest that being alone during lockdown has a greater impact on the development of depressive symptoms and sleep disturbances compared to being accompanied of relatives within the household [19][20].
Although these psychological reactions are considered as normal responses to the possible infection of COVID-19, their onset and development will also depend on the particular psychological characteristics of individuals [10]. In fact, several studies have shown that individuals who present higher psychological qualities such as self-con dence, self-esteem and optimism have better mental health [21]. Along the same line, it has also been shown that resilience is the main protective factor against the development of psychopathological symptoms [22]. Psychological resilience involves two peculiarities: the need to go through or be going through an adverse situation; and the execution of this dynamic process to overcome the threatening circumstance [23].
Psychological resilience implies more than just the individual's ability to overcome adversity, it comprises a dynamic construct, which is why different measuring instruments have been used to assess this capability. One of these instruments is Antonovsky's Sense of Coherence (SOC) scale that is intended to measure resilience through this construct, which is considered as a relevant factor within the notion of resilience [24]. In this way, SOC scale seeks to assess an individual's ability to perceive, understand and predict the events happening around them, as well as the manageability and the meaning that is given to said events. The original version of this scale has undergone some modi cations and adaptations [25], currently there is a short version of 15 items with a consistent factorial structure [26].
Although it is known that resilience constitutes a protective factor over mental health in times of adversity in general [27][28], few studies have investigated the relation between resilience and mental health in the context of the lockdown due to the COVID-19 pandemic [22]. Therefore, the objective of this study is to assess whether resilience capacity of an individual is related to depressive symptoms and poor sleep quality in a group of young adults of the general population in the context of lockdown due to the pandemic.

Design and context
The present is a cross-sectional study. Data was collected via an online, self-administered survey, between May and June 2020 when lockdown in Ecuador was mandatory.

Participants and sampling
Participants were young adults (18 to 35 years of age) who voluntarily decided to partake in the study. Participants were recruited by convenience sampling via invitations sent on social media sites, and institutional platforms of different universities in Ecuador. A total of 8426 participants were recruited.

Patient Health Questionnaire (PHQ-9)
This is a highly sensitive psychometric instrument consisting of nine items that identify symptoms of depressive disorder based on the DSM-5 criteria. This questionnaire assesses the symptoms of depression present during the two weeks prior to taking the test. Scores are measured using a Likert scale (0-3 points) that ranges from "not at all" to "nearly every day". The nal score ranges from 0 to 27. The higher the score, the more severe symptoms of depression. In this way, scores from 0 to 4 suggest none to minimal depression, from 5 to 9 mild, from 10 to 14 moderate, from 15 to 19 moderate/sever and > 20 severe depression [29].

Pittsburgh Sleep Quality Index (PSQI)
This is a 24-item questionnaire that assesses the quantitative and qualitative aspects of sleep quality for a time period of one month. For this evaluation, the report of an individual sleeping with the evaluated subject is considered in addition to the participant's self-report. This instrument consists of seven elements: sleep quality, sleep latency, sleep duration, sleep e ciency, sleep disturbances, use of sleep medication, and daytime dysfunction. Each of these items is measured with a score from zero to three, obtaining a maximum score of twenty-one points. Higher scores indicate a poorer quality of sleep, being a total score greater than ve points the cut-off point that indicates poor quality of sleep [30].

Sense of Coherence scale (SOC-15)
In order to determine resilience capacity, the SOC questionnaire was used, which measures the sense of coherence of an individual, who is considered the main element within this construct. It should be noted that the reduced version validated by Ortiz-Granja (2019), which consists of fteen items was used. This instrument uses a Likert-type model for scoring, with answers that range from "never" (one point) to "always" ( ve points). The minimum scores is 15 and the maximum 75, with the highest scores representing a greater sense of coherence [31]. Scores from 15 to 29 suggest very low, form 30 to 44 low, from 45 to 59 high, and > 59 very high sense of coherence.

Sociodemographic variables and compliance with the lockdown
Sociodemographic data of participants was collected as part of the on-line survey as previously described in our preprint publication [32]. Compliance with lockdown measures was measured indirectly analyzing information extracted from COVID-19 Community Mobility Reports [33]. In this way, we matched the day when the survey was answered with its corresponding percentage change of mobility according to COVID-19 Community Mobility Reports. Three aspects of mobility were evaluated: time spent on residency places, mobility towards groceries and pharmacies and to retail and recreation. Details on how percentage change was calculates is explained elsewhere [33].

Statistical analysis
Categorical variables were summarized by frequency and percentages. Median and interquartil range (IQR) were used to summarize numerical variables since all were non/normally distributed. Association between categorical variables and biological sex was tested using Chi2 test. Association between numerical variables and biological sex was tested using rank-sum test. We considered two main outcomes: PHQ-9 score depicting depression and PSQI depicting quality of sleep. Our main explanatory variable was SOC-15 depicting resilience. In order to study the association between the outcomes and explanatory variables we implemented two regression models with restrictive cubic splines (RCS), one for PHQ-9 and one for PSQI. RCS regression model determined the shape of the relationship between PHQ-9 and PSQI and SOC-15 without any prior assumption. RCSs tted a smooth continuous curve of adjusted means with 95% con dence intervals (95% CIs) across SOC-15 levels. RCSs allowed for changes in the function at de ned knot points and restricted the splines to linear relationships at the tail ends. Knot points were located at percentiles 5, 27.5, 50, 72.5, and 95 of SOC-15 values, as previously recommended to avoid forcing curvature or in ections [34] R version 3.6.3 (2020-02-29) and related packages including rms were used to analyze the data [34][35]. To check the in uence of mobility restrictions we also included the percentage change of mobility for retail and recreation, residential places and groceries and pharmacies as covariates in the models. Models were adjusted by age (years), biological sex (male/female), zone of residence (urban/rural), use of sleeping pills (yes/no), living with a partner (yes/no), occupation (student/have a job/unemployed) and people living at home (1/2-3/4-5/>5).

Sense of coherence, depression and sleep quality
Median SOC -15 score of the sample was 61 (IQR: 55.00, 67.00) with a minimum and maximum of 15.00 and 75, respectively. A high (n = 2961, 35.14%) or very high (n = 4944, 58.68%) sense of coherence was reported by nearly 94% of individuals, the remaining 6% reported low (n = 487, 5.78%) or very low (n = 34, 0.40%) sense of coherence. Median PHQ-9 score of the sample was 3 (IQR: 1.00, 7.00) with a minimum and maximum of 0.00 and 27, respectively. Symptoms of mild to moderate depression were reported by nearly 36% of the individuals (mild: 27.20%, n = 2292; moderate: 9.23%, n = 778). Symptoms of moderate/severe to severe depression were reported by nearly 6% of participants (moderate/sever: 3.96%, n = 334; sever: 1.95%, n = 164). Median PSQI of the sample was 4 (IQR: 2.00, 7.00) with a minimum and maximum of 0.00 and 17. respectively. Bad quality of sleep was reported by nearly 38.52% (n = 3246) of participants. A summary of SOC-15, PHQ-9, PSQI and corresponding groups given by respective cut-off points by biological sex is depicted in Table 1.

Resilience and depression in the context of the lockdown
An inverse non-linear relationship was found between depression and sense of coherence (F = 13.79, p < 0.001). As it is shown in Fig. 1 higher scores of PHQ-9 were associated with lower SOC-15 scores, suggesting that depression is associated with lower resilience. We also found that restriction of mobility was associated with depression. In this way, a longest stay at home was related with higher scores of PHQ-9 (F = 10.35, p = 0.001; Fig. 1, panel a). Similarly, a higher restriction of mobility towards groceries and pharmacies was related to higher scores of PHQ-9 (F = 5.85, p = 0.016; Fig. 1, panel b). Depression was not associated with restriction of mobility because of retail and recreation (F = 1.25, p = 0.263).

Resilience and sleep quality in the context of lockdown
An inverse non-linear relationship was found between quality of sleep and sense of coherence (F = 13.63, p < 0.001). As it is shown in Fig. 2 higher scores of PSQI were associated with lower SOC-15 scores, suggesting that bad sleep quality is associated with lower resilience. In contrast to depression, sleep quality was not signi cantly associated to time spend at residential places (F = 0.41, p = 0.524), neither to mobility towards groceries and pharmacies (F = 0.84, p = 0.358) or retail and recreation (F = 0.01, p = 0.905).

Discussion
The present study examined psychological resilience and two aspects of mental health (i.e. depression and sleep quality) in the context of restriction of mobility due the COVID-19 pandemic. The results obtained in this study indicate that an individual's capacity to adapt to adverse circumstances can modulate the appearance of mental disorders during con nement due to the COVID-19 pandemic. Correspondingly, a greater alteration in mental health was also evidenced in those individuals who experienced more restrictions regarding their mobility.
The results obtained in this study are in accord with those from previous studies, where it has been shown that psychological resilience is associated with the appearance of depressive disorders and sleep disorders during con nement, both for individuals who have been infected with the virus, and those who have not. Empirical evidence has demonstrated that non-infected people with low psychological resilience have a higher risk of presenting psychopathological alterations, especially associated with the presence of mood disorders [36][37]. In this same group it has been found that longer periods of con nement due to the pandemic, paired with a lower capacity to adapt, increase an individual's risk to develop depressive and sleep disorders [38 -39]. The same has been observed in individuals who have recovered from the virus, where patients with less psychological resilience are more likely to develop mental disorders [40][41]. Although resilience by itself has an effect on the psychological effects endured by COVID-19 patients, there are other variables that might intervene to preserve or alleviate the mental health of an individual. For instance, studies have found that patients who spent fewer days at the hospital, performed more daily exercise, and received sunlight out in the open, showed a greater resilience capacity and also had a better capacity to recover from the disease [19,36,42]. Additionally, it has been shown that resilience might be a protective factor that mitigates the impact of the pandemic on mental health and reduces the probability of developing psychiatric disorders for both COVID-19 patients and for the population in general [42][43][44][45][46][47]. A lower psychopathological presence was also found in frontline healthcare professionals who displayed better resilience. They also responded with greater assertiveness and were more e cient at work in comparison to their colleagues who had less adaptation capacity [36,48]. In any case, both low and high resilience capacity can modulate the presence of mental disorder.
This study also found an association between mobility restrictions and mental health. Essentially, individuals who engaged in trips and outings less frequently during lockdown, reported more depressive symptoms and poorer quality of sleep. These results are consistent with previous studies, which have also obtained relevant associations on this topic. For example, associations have been found between emotional eating and reduced trips to workplaces and parks [32]. The same is true for sleep quality, where a deterioration has been detected, due to restrictions of mobility and stress generated during lockdown [8, [49][50]. Other studies have shown that reduced exposure to daylight and the lack of external "timekeepers" to help regulate an individual's schedule affect their circadian rhythm and therefore their sleep pattern, resulting in poorer quality of sleep and a higher prevalence of sleep disorders such as insomnia and daytime sleepiness [14,38,[51][52][53][54]. Similarly, increasing symptoms of anxiety and depression have been identi ed on individuals who have not left their homes due to imposed restrictions or who have spent more time at home without traveling to workplaces, parks, or to carry out outdoor activities [7,42]. Although psychological resilience is described as an in uential agent in people's mental health, other variables such as social support, communication and good interpersonal relationships can also result in uential during self-isolation and lockdown. Along these lines, studies have also suggested that individuals who keep in contact with family and friends have an improved quality of life. Likewise, individuals who receive greater social support from their family and friends show better resilience [16,20]. While social support is linked to a better adaptation and coping with the pandemic, other factors such as good sleep quality, exercise, and religious beliefs can also lead to a better assimilation of adverse circumstances [7,38,42]. Although most variables have been positively associated with the ability to cope with problems and overcome adversity during lockdown, other factors such as substance abuse could precipitate and trigger the development of psychopathologies, [19,47].
Lastly, our results also indicated an association between greater psychopathological symptoms and resilience capacity, and staying at home, results that are consistent with previous studies [42]. These ndings can also be the result of lockdown and mobility restrictions, where there is evidence of a loss of personal autonomy, less social contact and support, a decrease in recreational activities, and greater interpersonal con ict in the household [47]. Essentially, and in a general manner, the expected results were obtained regarding the association between the elements discussed in this project.

Strengthens and limitations
One of the strengths of this study is the size of the sample, which included near 9,000 participants of the general population, which can provide a good picture of the mental health issues associated with the lockdown due to the pandemic. Moreover, we used previously validated instruments to collect information and objectively measured the impact of lockdown measures in the population. Nonetheless, the present work had certain limitations. For example, variables that could in uence a person's resilience capacity during lockdown were not considered, such as substance use and social support. Likewise, it has also been suggested that exercise can modulate and regulate an individual's capacity for psychological resilience [55]. This variable should be taken into consideration in future research projects, to determine its in uence as a protective factor to avoid the development of mental disorders or psychopathological symptoms. Moreover, although our sample included a good number of participants, they were recruited by convenience sampling, which could affect the external validity of our study. Still our ndings open the discussion regarding the effects of lockdown on mental health among young adults and the need of public health policies directed to enhance resilience in order provide psychological tools to resist to adversity.

Conclusion
In young Ecuadorian adults higher resilience is associated with less depression and better quality of sleep in the context of the lockdown due to the COVID-19 pandemic. Depression, but not sleep quality was associated with higher mobility restriction.
Declarations Acknowledgment: We thank to Dra. María Paulina Robalino Valdivieso and Ing. Hugo Moreno for their support on the realization of the project.
Funding: The authors did not receive support from any organization for the submitted work.
Financial interests: The authors have no relevant nancial or non-nancial interests to disclose.
Con icts of interest: the authors declare that they have no con ict of interest.
Availability of data and material: All data are available upon request to the corresponding author.
Code availability: Analysis was performed using R and related packages. Code is available upon request to the corresponding author.
Authors' contributions: All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Vinueza-Veloz. The rst draft of the manuscript was written by Guerrero-Vaca, and all authors commented on previous versions of the manuscript. All authors read and approved the nal manuscript.
Additional declarations: not applicable Ethics approval: The study was carried out following the Declaration of Helsinki to work with human beings and in accordance with the "Singapore Declaration on Integrity in Research". It was approved by the Scienti c Committee of the School of Medicine and Instituto de Investigación (IDI) of the Escuela Superior Politécnica de Chimborazo. The manuscript does not report the results of a clinical trial nor patient data.
Consent to participate: All participants accepted and lled an informed consent form included at the beginning of the survey. Participant's responses were anonymous and con dential.
Consent for publication: Not applicable.  Figure 1 Predicted means for PHQ-9 scores by levels of SOC-15. Time spent at residential places was categorized using as cut-point the median. Panel a) shows predicted scores for PHQ-9 by time spent at residential places. Panel b) shows predicted scores for PHQ-9 by mobility restriction to groceries and pharmacies.