Psychological Impact of COVID-19 Confinement and Its Relationship with Meditation

The objective of this study was to evaluate the psychological impact of confinement due to the COVID-19 pandemic, considering any protective factors, such as the practice of meditation or self-compassion, and their relationship with different lifestyles and circumstances of adults residing in Spain. A cross-sectional study was done using an anonymous online survey in which 412 participants filled out the Depression, Anxiety and Stress Scale-2; the Impact of Events Scale; and the Self-Compassion Scale-Short Form, reporting severe symptomatology of posttraumatic stress and mild anxiety and depression. Quality of cohabitation and age were found to be key variables in the psychological impact of confinement. The impact of confinement was more negative for those who reported very poor cohabitation as opposed to very good (F (3, 405) = 30.75, p ≤ 0.001, d = 2.44, r = 0.054) or for those under 35 years of age compared to those over 46 (F (2, 409) = 5.14, p = 0.006, d = 0.36). Practicing meditation was not revealed as a protective factor, but self-compassion was related to better cohabitation during confinement (F (3, 403) = 11.83, p ≤ 0.001, d = 1.05). These results could be relevant in designing psychological interventions to improve coping and mental health in other situations similar to confinement.


Introduction
In December 2019, an outbreak of pneumonia with an unknown etiology was announced in Wuhan (China), soon becoming a global pandemic and surpassing the severe acute respiratory syndrome (SARS) of 2003 [1]. The World Health Organization (WHO) identified this new pneumonia as a new coronavirus called SARS-COV-2 or as it is commonly called, COVID-19. Since the first cases were recorded in December 2019, the number has increased exponentially throughout the world and a public health emergency of international interest was declared on 30 January 2020 [2]. By March 11th, COVID-19 had already been declared a worldwide pandemic with 118,000 cases in 114 countries, although 90% of them were concentrated in China and the Republic of Korea [3]. At the time of writing, 28 June 2020, global coronavirus cases surpassed 10 million, with over half a million deaths. In Europe, over 2.5 million people have been infected, with almost 200,000 deaths, and Spain has had the third most cases in Europe, with 248,770, and the sixth most deaths, with 28,343 [4].
This rapid expansion of the virus is partly due to its high transmissivity, which has an estimated basic reproduction number of 2.24 to 3.58, meaning that an individual who is a transmitter of the virus could infect from two to four others [5], even though they do not show any symptomatology [6], and its incubation period varies from two to 14 days, with an average of seven days [7]. The symptomatology of those infected is fever, dry cough, sore throat, difficulties breathing, fatigue, nausea, vomiting, and diarrhea. The most commonly reported comorbidities are hypertension, diabetes, headache, dizziness, fatigue, myalgia, hypo/anosmia, and gustatory dysfunction [8]. Regarding psychiatric and neuropsychiatric presentations, it seems that in its acute stage, SARS-CoV-2 might cause delirium in a significant proportion of patients. It might also cause depression, anxiety, fatigue, posttraumatic stress disorder, and more rarely, in the long-term, neuropsychiatric syndromes [9].
COVID-19 can infect people of all ages, but men over 70 with preexisting conditions, such as asthma, diabetes, or heart disease, are the most susceptible, reporting an >8% mortality rate [5]. Women are less susceptible to infection and because of their distinctive immune system, severity and mortality are much lower in children than in adults [5].
In view of all of the above and especially because of the rapid expansion of the epidemic, many countries have decreed a state of emergency with mass quarantine of their citizens. In the particular case of Spain, this state of emergency has imposed a closure of establishments, ceasing of all except essential activity, and home confinement of the population during a period of time depending mainly on the evolution of the number of people infected and the expansion of the virus. In this respect, recent studies [10,11] have mentioned that understanding the effect this quarantine has on the population is of great importance, especially in countries as severely affected as Spain. Quarantine could be understood as an important source of stress as it is unpredictable and uncontrollable, which are core features of the psychobiological construct of stress [12].
A recent review [10] on the psychological impact of quarantining for SARS, H1N1 flue, and other causes reported fear in over 20% of the population, 18% sadness and 10% blame, and a high prevalence of distress, stress, depression, or bad mood. Some of the factors generating stress during the SARS quarantine were duration and fear of infection. Some studies showed a positive association between length of quarantine and increase in posttraumatic stress symptoms, avoidance behaviors, or anger [10]. In recent studies on the impact of COVID-19, women have manifested a greater psychological impact and worse mental health [13]. Young people and students are also more vulnerable because of schools and universities closing, as well as social isolation, altering the normal development of their education [14].
In such a situation, systematic psychological self-care should have a high priority in coping with the detrimental impacts of COVID-19 and social distancing. Thus, acknowledging one's own limitations with compassion as a part of the shared human experience could have a positive impact on mental health. In this respect, possible protective factors could be mindfulness or self-compassion, since, as mentioned in a recent review [15], it seems that mindfulness-based treatments are effective in reducing posttraumatic stress disorder (PTSD) symptoms and the Mindfulness-based Stress Reduction Program was the one that provided the most evidence [16]. In this vein, another review [17] emphasized the promising results associating an increase in self-compassion with reduction in PTSD symptoms, reducing the impact of exposure to trauma. Therefore, and since to date there are few studies reporting on the emotional impact the pandemic is having on the psychological health of those confined, the objective of this study was to explore the impact of confinement due to COVID-19 on the mental and emotional health of adult Spanish-speaking residents of Spain. In particular, we intended to inquire into the possible impact of confinement on anxiety, depression, and posttraumatic stress and whether practice of some type of meditation or self-compassion could be a protective factor against this impact. Then, any between-group differences in these variables would be analyzed by sample sociodemographic variables (age, cohabitation, etc.) and finally, the relationships between the abovementioned variables (anxiety, depression, stress, self-compassion) and the sociodemographic variables selected (e.g., age or treatment of those confined) would be explored.

Participants
A total of 412 adults from 63 Spanish provinces, with a mean age of 40.48 (SD = 10.79) participated in the survey out of a total of 420 (finalization rate: 98.09%). Most of the participants were Spanish (91%), women (84.5%), aged 18 to 73, with no changes in their employment because of COVID-19 (73.8%), a university education (42%), and did not practice any type of meditation (69.9%). The rest of the sociodemographic variables recorded are presented in Table 1.

Instruments
Data collected for the study were sociodemographic (gender, age, nationality, marital status, occupation, employment, education); health (whether under treatment or not and if so, what); quality of cohabitation during the pandemic; cohabitation with a case of COVID-19; whether religious or not and if so, practicing or not; and whether they practiced mediation and if so, what type.
Mental health was evaluated using the following scales: Depression, Anxiety and Stress Scale-21 (DASS-21). The DASS-21 consists of 21 items evaluating the frequency and severity of emotional symptoms experienced during the previous week, using a self-report scale with four choices from 0 (Never) to 3 (Almost always). It is made up of three subscales, Depression, Anxiety, and Stress, with seven items each. A high score on each subscale means stronger symptomatology ( [18,19] Spanish adaptation). Internal consistency of the Spanish adaptation has an α = 0.93 for the Depression scale, α = 0.86 for Anxiety, α = 0.91 for Stress, and α = 0.96 for the total scale [19].
Impact of Events Scale (IES). The IES consists of 15 items evaluating subjective stress from stressful and or traumatic experiences in the previous week, with two subscales, Intrusion and Avoidance, with seven and eight items, respectively ( [20,21] Spanish adaptation). It is a four-choice self-report scale varying from 0 (never) to 4 (often). Based on the sum of all the items, the cutoff points are 8.6 to 19 for mild symptoms and 19 or over for severe symptoms. The internal consistency of the Spanish adaptation has an α =.86 for the total scale, α = 0.76 for Intrusion and α = 0.82 for Avoidance [21].
Self-Compassion Scale-Short Form (SCS-SF). The short form of the scale is made up of 12 items that evaluate self-compassion ( [22,23] Spanish adaptation). It consists of three main subscales (Mindfulness, Self-Kindness, and Common Humanity) and their opposites (Over-identification, Self-Judgment, and Isolation). The scale has five answer choices from 1 (Almost never) to 5 (Almost always). A higher score is interpreted as more self-compassion. The internal consistency of the Spanish adaptation is α = 0.71 for Self-Judgment, α = 0.75 for Common Humanity, α = 0.77 for Isolation, α = 0.74 for Mindfulness, α = 0.76 for Over-Identification, α = 0.73 for Self-Kindness, and α = 0.85 for the total scale [23].

Procedure
A cross-sectional study was conducted from 4 April 2020 to 1 May 2020. The sample was recruited by non-probabilistic snowball sampling, since it was an exploratory study in which rapid access to data was needed. First, contacts of the research team were invited to participate and then, these contacts were asked to forward the invitation to their contact networks. An online survey form was prepared using Google Forms, which included information on the study; telling the participants that they would not receive any compensation for taking part in it; a series of sociodemographic and additional variables of interest; and the DASS-21, IES, and SCS scales. The study protocol was approved by the University of Almería Bioethics Committee (No. UALBIO2020/032) and the participants signed their consent after being informed of the implications of their participation and the confidentiality of the data provided, according to the Helsinki Declaration and the Organic Law on Data Protection 15/1999 of December 13th.

Data Analysis
All the analyses were performed with SPSS version 26.0 (IBM Corp., Armonk, NY, USA) and the descriptive results of the quantitative variables were reported with mean and standard deviation. An ANOVA was done to evaluate the between-group differences in the sociodemographic variables and then, a Student's t-test was applied for comparisons between each pair of groups. Further, the effect size was calculated for those that were statistically significant. Finally, the bivariate correlation was calculated to evaluate the relationship of the different sociodemographic ordinal variables and subscales. In all the analyses, p < 0.05 was considered statistically significant.

Impact of Confinement on Anxiety, Depression, Posttraumatic Stress, and Self-Compassion
The general mean on the DASS-21 scale for participant Stress was 6.67 (SD = 5.25), 3.88 (SD = 4.51) for Anxiety and 4.51 (SD = 4.54) for Depression, which reflects mild depression and anxiety, and the stress score was not symptomatic. On the IES scale, the mean for Intrusion was 12.61 (S = 7.39), 13.66 (SD = 7.62) for Avoidance, and on the total scale, it was 26.22 (SD = 13.78), which shows severe symptomatology. Finally, on the SCS scale, mean Self-Kindness was 6.46 (S = 1.67) and for Common Humanity, it was 6.44 (SD = 1.60) and in Mindfulness, it was 6.69 (SD = 1.69).

Between-Group Differences in Anxiety, Depression, Stress, and Self-Compassion by Sociodemographic Variables. Analysis of Each Questionnaire
An ANOVA was done with the most significant variables to test for differences between the sociodemographic variables and psychological scales. This section only shows the data for those sociodemographic variables that were statistically significant. To be able to compare the groups by age, a new variable was created a posteriori by grouping participants by age into three groups of equivalent size by the quartiles found from the cutoff points. This new variable, "Age Group", was divided into Young Adults (18 to 35), Adults (36 to 45), and Middle Aged (Over 46). In addition, and since one of the objectives of this study was to test the importance of meditation on the psychological impact of confinement, this variable was also included for testing.

DASS-21
After the ANOVA and meeting the assumptions of normality and homogeneity of variance, statistically significant differences were found between psychological variables and sociodemographic variables.
In Stress, significant differences were found between age groups [F (2, 409) = 5.14, p =0 The results show that the Total IES scores were lower in the groups with the best cohabitation, with a large effect size, except between Very Good and Good, which was small, and between Good and Poor, which was moderate. Differences were found with regard to occupation [F From the comparison between groups, it may be seen that the scores on Common Humanity were higher in those groups where cohabitation was better, with a large effect size. Finally, statistically significant between-group differences were also found between groups by type of meditation [F (4, 396) = 6.10, p ≤ 0.001], especially between None and Mindfulness (M = 6.23, SD = 1.54 vs. M = 7.02, SD = 1.64, d = 0.50, r = 0.001) and between None and Transcendental (M = 6.23, SD = 1.54 vs. M = 8.00, SD = 1.58, d = 1.14, r = 0.004). In both cases, the participants who showed more Common Humanity were those who practiced some type of meditation compared to those who did not, with a moderate effect size between those who did not and those who did practice Mindfulness and a large effect size between those who did not practice and those who practiced Transcendental meditation.
Concerning the last subscale, Mindfulness, statistically significant differences were observed between age groups [F ( In general, in those groups that practiced meditation, the score on the Mindfulness scale was higher than in those who did not practice any, with a large effect size, except between None and Mindfulness, which was small. Lastly, by type of occupation, statistically significant between-group differences [F (6, 398) = 2.96, p < 0.05] were also found between Unemployed and Housewife/Husband The rest of the sociodemographic variables with their mean and standard deviation are shown in Table 2.

Relationships between the Variables Mentioned (Anxiety, Depression, Stress, Self-Compassion) with the Selected Sociodemographic Variables (Ordinal)
After the correlation analysis, older age was observed to be related to higher frequency in practicing meditation, better cohabitation, more Common Humanity, and more Mindfulness. On the contrary, less Stress, Anxiety, and Depression were related with older age. Higher education was significantly related to more frequent practice of Meditation and an increase in the variables in the SCS scale. On the contrary, greater Anxiety, Intrusion, and Avoidance were related to a lower education. More frequent Meditation practice was significantly related to less Depression and an increase in SCS scale variables. Better cohabitation quality was significantly related to higher scores on SCS variables and a lower score on the rest of the scales. Finally, as expected, the psychological scales were significantly correlated with each other (see Table 3).

Discussion
After analyzing the survey results with regard to the objectives originally set, the participants showed mild depression and anxiety, not reporting symptoms of stress. However, they did report severe symptomatology on the IES scale as in previous studies, which related high scores in posttraumatic stress symptoms or avoidance behaviors after 10 days of quarantine [8]. This difference may be, in part, because the IES scale evaluates the psychological impact of an event after the fact and the participants could have understood that the event referred to COVID-19, but not on the DASS-21 scale.
Another of the objectives of the study proposed Self-compassion and Mindfulness as protective factors against the impact of confinement. It is worth mentioning that the participant scores on the various Self-Compassion Scale subscales (Mindfulness, Self-Kindness, and Common Humanity) were negatively related in both the IES and the DASS-21, suggesting that self-compassion might be an important protective factor, fostering emotional resilience [24]. The results show that higher scores of Self-kindness, Common Humanity, and Mindfulness were related to not being under treatment and in turn, higher scores of Self-kindness and Common Humanity were related to higher cohabitation quality. There was no significant relationship of mindfulness practice with lower scores on the IES or DASS-21 scales, but there was an increase in the Self-Compassion scale, as was to be expected, given the relationship of mindfulness with Self-compassion [25].
The sociodemographic variables that had the strongest relationship with psychological impact were being under treatment or not, age, and quality of cohabitation during the quarantine, indicating that not being under any treatment implied lower scores of Intrusion and Avoidance. Being young was related to higher scores of stress and depression, in agreement with recent studies on the psychological impact, relating age and being a student with higher vulnerability and psychological impact from confinement [14,26]. Furthermore, an improvement in quality of cohabitation was related to a lower Total IES score. On the contrary, worse cohabitation is related to greater Intrusion and Avoidance. These results would suggest that young people under some type of treatment and who report that the quality of their cohabitation is not good, would experience a negative psychological impact of confinement. Specifically, young people show more posttraumatic stress symptomatology, so it would be of interest for authorities to develop psychological interventions, providing a support network during quarantine as previous studies have related an increase in PTSS after similar events, such as Middle East Respiratory Syndrome [27]. It was also observed that cohabitation with a person who was COVID-19 positive implied more intrusive thoughts and avoidance, as well as a higher score on the Total IES, which would be related to experiencing more fear of becoming infected with the virus and in turn, infecting others [10,28,29]. Being Retired was also related to more Intrusion and a higher total score on the IES scale and on the contrary, being a Student was related to less Intrusion and a lower score on the Total IES scale.
This study has some limitations, especially due to the limited resources available and the urgency in its implementation during the state of emergency. We cannot establish any causal relationships between the variables studied. The sample was mostly comprised of women and only an adult population, so it is not a true representation of the entire population. Future studies should have a larger and more homogeneous sample and use a more complex sampling strategy or perform a prospective study that would enable a more specific result to support the need for health initiatives focused on improving the mental health of the participants, promoting attitudes or behaviors for coping with confinement.
In spite of these limitations, the results of this study provide important information on the psychological and social impact of the first three weeks of a COVID-19 quarantine period, including that self-compassion might be an important protective factor, and may be used as a reference in future epidemics or outbreaks of the virus that provoke similar short or long-term situations to generate hypotheses for improvement of prevention and intervention. The results could also provide a baseline for evaluating the psychological and social consequences during the rest of the COVID-19 epidemic, which was still underway at the time of writing.

Conclusions
During this first stage of confinement due to the COVID-19 outbreak, the participants reported mild anxiety and depression symptomatology, as well as severe posttraumatic stress symptomatology. The quality of cohabitation was shown to be a key variable in the psychological impact of the participants, since its poorer quality was related to higher scores of intrusion and avoidance, which were lower the better the quality of cohabitation was. Being young and being under treatment were also related to a greater psychological impact. Finally, self-compassion was related to better cohabitation during confinement. The findings shown in this study could be used for psychological interventions to improve mental health and coping with confinement during the COVID-19 epidemic.