Hindawi

Depression Research and Treatment

Volume 2020, Article ID 3158954, 11 pages
https://doi.org/10.1155/2020/3158954

Research Article
Depression and Its Relationship with Coping Strategies and Illness
Perceptions during the COVID-19 Lockdown in Greece: A
Cross-Sectional Survey of the Population

Petros Skapinakis ,1 Stefanos Bellos,1 Achilleas Oikonomou,2 Georgios Dimitriadis,2

Paschalis Gkikas,2 Evridiki Perdikari,1 and Venetsanos Mavreas1
1
Department of Psychiatry, University of Ioannina School of Medicine, Ioannina, Greece
2
Private Psychiatric Practice, Greece

Correspondence should be addressed to Petros Skapinakis; p.skapinakis@gmail.com

Received 13 May 2020; Revised 18 July 2020; Accepted 28 July 2020; Published 26 August 2020

Academic Editor: Giovanni Galeoto

Copyright © 2020 Petros Skapinakis et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.

Objective. The COVID-19 epidemic has shown a more benign course in Greece possibly due to the early lockdown measures.
Mental health consequences of the lockdown however are unknown. In addition, illness perceptions and relevant strategies to
cope with the stress of the epidemic may have played a role in complying with the restrictions. We conducted a survey of the
Greek population with the aim to investigate the prevalence of anxiety and depression during the lockdown, the emotional
impact of the epidemic, and the effect of coping strategies and illness perceptions in mental health. Methods. Adult persons were
invited during the peak of the lockdown period through social media. Depressive and anxiety symptoms were assessed with the
PHQ-9 and GAD-2 scales, respectively. Coping strategies were assessed with selected items of the Brief COPE questionnaire,
while illness beliefs were assessed with items from the revised Illness Perception Questionnaire (IPQ-R). Results. A total of 3379
individuals took part. A strong emotional impact of the epidemic was more often in women and in those with severe financial
difficulties. Levels of depressive and anxiety symptoms were high but similar to past assessments. Participants showed high
levels of personal control and used more often positive strategies to cope with the stress of the epidemic. Depressive symptoms
were higher in the younger, in students, in those with a stronger emotional impact, in those isolated due to symptoms, and
those overexposed to media for COVID-19-related news. Lower levels of depression were seen in those using positive coping
strategies and showing high levels of personal and treatment control. Conclusions. The study shows that certain psychological
and social determinants were associated with increased depressive symptoms during the lockdown warranting the development
of public health guidelines to mitigate the effects of the epidemic to the mental health of the population.

1. Introduction with increased psychological symptoms as a result of the

COVID-19 pandemic would be useful from a clinical and
The COVID-19 pandemic may have important psychological public health perspective.
and social effects, but these have not been assessed at the pop- Previous research has shown that the psychological
ulation level [1]. Apart from the fear of a potentially severe impact of health-related stressful events, including chronic
disease with no specific treatment, there is also increasing diseases or infectious outbreaks, is influenced by the specific
concern that the mental health consequences of the social beliefs about the illness and the resulting coping behaviors
distancing measures and the lockdown may be comparable that people tend to use in order to adapt better and mitigate
or even exceed the burden of physical illness and mortality the mental health consequences [3, 4].
due to the virus itself [2]. Thus, a more thorough examina- “Coping” is the set of psychological responses to percep-
tion of the mechanisms and determinants that are associated tions of threat with the aim to prevent or diminish threat,
2 Depression Research and Treatment

harm, and loss or to reduce associated distress [5]. Meta- and we encouraged others to share the post (passive snowbal-
analyses indicate that engagement with a set of “functional” ling). In addition, we actively sought other health profes-
coping strategies is associated with better physical and men- sionals, journalists, and other individuals with strong social
tal health in individuals facing a range of stressors, such as media presence to share our invitation. In order to increase
traumatic events (social stress, bullying) or health-related the participation of men, younger, and elderly individuals
conditions (HIV, cancer, and diabetes) [6–11]. In addition, as well as individuals living in both major and small cities,
outbreaks of viral infections are considered among the condi- we also requested from individuals with strong social media
tions that coping strategies have been found to play an presence in these subgroups to share the links to our survey.
important role on emotional responses and the development We repeated the procedure on the third day to increase the
of psychological problems [4, 12–14]. sample size. We posted the first link to our survey at 9 am
Choice of coping behaviors may be influenced by specific on the 8th of April, and we closed the survey at 19.00 on the
illness perceptions defined as the set of beliefs that the indi- 12th of April. The percentage of participants recruited at each
viduals develop for a health condition [15]. Some studies day was as follows: 39% (8 April), 22% (9 April), 9% (10
have assessed Illness perceptions for a hypothetical physical April), 24% (11 April), and 6% (12 April). During this period,
health problem [16–18] or chronic infectious disease [19– the lockdown restrictions due to the COVID-19 epidemic
21] and found correlations between illness perceptions and were fully implemented in Greece.
attitudes or intentions towards the adoption of preventive The online survey was completely anonymous, and
behaviors or use of health services. informed consent was obtained from the participants at the
Illness perceptions and coping behaviors related to the beginning of the survey. All participants were free to leave
COVID-19 pandemic could be important as they may be the survey at any point before submitting the results. This
associated with the increased psychological burden of research was conducted in accordance with the Helsinki
depression and anxiety that has been reported mainly in Declaration and the ethical standards of our Institutional
health professionals [22]. In addition, specific perceptions Bioethics Committee. A second wave of data collection is also
and coping strategies may lead to more adaptive health planned in the future, after the easing of the lockdown
behaviors and rates of compliance with official recommenda- restrictions and the gradual opening up of the economy.
tions, as it was the case in previous viral outbreaks [23, 24].
Previous small surveys have studied the association between 2.2. Assessment of Current Mental Health Status and General
the perceptions about a viral infection and the psychological Health. Depressive symptoms were assessed with the PHQ-9,
responses to epidemic outbreaks, showing that perceptions the 9-item depression module of the Patient Health Ques-
about viral infections play an important role [25, 26]. tionnaire [28]. We had previously validated this version in
In Greece, the COVID-19 epidemic has shown a more chronic kidney disease patients in Greece [29]. A cut-off of
benign trajectory that has been portrayed in the international 10 had shown good sensitivity (82%) and specificity (93%)
news media as the “uniqueness of the Greek case,” at least against a diagnosis of depression using a structured diagnos-
compared with mortality rates of neighboring Italy or other tic interview. In the context of the current study in our anal-
Mediterranean countries such as France or Spain. Restric- yses, we used the total scores on the PHQ-9 (range 0 to 27),
tions were imposed early, and compliance to the lockdown and we also report two categories of depressive symptoms
measures was universal and successful with little opposition using the threshold of 10 for mild (or more) depression and
from the public. It is important to study any potential mental 15 for moderate/severe depression only [28].
health consequences of these restrictions. In addition, an Anxiety symptoms were assessed with the first 2 core
investigation of the main coping behaviors of the Greek pub- items of the Generalized Anxiety Disorder scale (GAD-2)
lic and the prevalent perceptions of the illness and the epi- [30]. This has shown high sensitivity and specificity to detect
demic in general may offer useful suggestions for public GAD in primary care [31]. For the analyses, we used the total
health officials and the better implementation of preventive scores (range 0 to 6) and two cut-offs at 3 and 4 to denote
measures elsewhere. Therefore, we undertook the current mild or moderate anxiety, respectively.
study with the aim to assess the mental health status of the Level of general health was assessed with the very well-
Greek population during the peak of the COVID-19 lock- validated single question regarding global health status [32].
down and its association with the main coping behaviors This assesses current general health using a 5-point Likert
used and the prevalent illness beliefs about the epidemic. scale. In the current paper, we derived a binary variable of
fair/poor general health versus good (or better) health.
2. Methods
2.3. Assessment of Coping Strategies. Coping strategies of the
2.1. Description of the Dataset. We conducted an electronic sample during the epidemic were assessed using selected
web survey using Google Forms. Participants were recruited items of the Brief COPE questionnaire [33]. This has been
through social media, mainly Facebook. All adults (18 or translated and validated in Greek using community samples
more) were eligible to participate with no other restrictions from Greece and Cyprus [34]. Brief COPE evaluates 14
other than the language barrier. In order to increase the sam- strategies to cope with stressful events. Each coping strategy
ple size and the representativeness of the sample, we used is represented by two items. In order to minimize the burden
both passive and active convenience snowball sampling of participants, two of the authors collected one item for each
[27]. We posted the invitation to the study on Facebook, strategy using a consensus procedure. One strategy (self-
Depression Research and Treatment 3

blame) was excluded as it was considered not applicable to 2.7. Statistical Analysis. We used Stata version 12.0 for all
the epidemic. Therefore, we finally included 13 items repre- analyses (StataCorp, College Station, Texas). Factor analyses
senting an equal number of coping strategies. Views about of the two instruments (Brief COPE and IPQ-R) were per-
coping styles are rated using a four-point Likert scale from formed with the command “factor” in Stata. Odd ratios for
“not at all” (coded 1) to “a lot” (coded 4). depression and anxiety and their 95% confidence intervals
were calculated with a series of adjusted logistic regression
2.4. Assessment of Illness Perceptions. Illness perceptions models. All evaluations of statistical significance are based
related to the COVID-19 epidemic were assessed using the on two-sided tests using the 5% level of significance.
revised “Illness Perception Questionnaire” (IPQ-R) [35].
IPQ-R has been widely used and validated in a range of 3. Results
chronic diseases as well as in healthy participants for hypo- 3.1. Description of the Sample. Three thousand three hundred
thetical disease scenarios or genetic predispositions [36, 37]. and seventy-nine (3379) adults took part in the study. The
The original IPQ-R includes 38 cognitive and emotional sample was predominantly female (73% of the final sample),
representations of illness grouped into the following dimen- mean age was 42 years old (SD: 12.6), 52% were married,
sions: timeline of the illness (acute/chronic), timeline (cycli- 69.7% were employed, and 78% were living in a major city.
cal), personal control, treatment control, illness coherence, The sample was well educated with 37% having an upper
emotional representations, and perceived consequences. We secondary or postsecondary education, 30% a university degree,
adapted the IPQ-R to be used during the COVID-19 epi- and 30% a postgraduate degree. (Table A1 in the supplementary
demic. Two of the authors initially selected 26 of the 38 items. appendix gives full details of the characteristics of the
We excluded all items of the illness coherence dimension as participants).
these were not applicable to the current epidemic. From the Almost 25% of the sample reported financial difficulties
remaining 26 items, we further eliminated 11 of them to of their household. Regarding exposure to media related to
reduce the burden of the participants and the total time of COVID-19 news, 17.37% of the sample reported a high/ex-
completion. The final 15 items included 1 item each for time- cessive use of several media with men slightly more likely
line (acute/chronic) and timeline (cyclical), 3 items for per- to report such use compared to women (19.6% vs. 16.52%,
sonal control, 2 items for treatment control, 4 items for p = 0:034). Time spent outside home, excluding time related
emotional representations, and 4 items for the consequences. to work, was small and only 10.36% of the sample reported
A Greek version of the IPQ-R adapted for breast cancer is being out of home for a high or excessive amount of time
available at the official IPQ site (http://ipq.h.uib.no). How- (15.29% of men vs. 8.5% of women, p < 0:001).
ever, we adapted the translation to better reflect the aims of
the study as it is suggested by the developers of the scale. 3.2. Level of Current General and Mental Health. Levels of
Views about illness are rated in a five-point Likert scale from general and mental health are shown in Table 1. It can be
strongly disagree (coded 1) to strongly agree (coded 5). We seen that there were high levels of mild depressive (at a cut-
reversed the coding of items appropriately as instructed by off of 10 or more for PHQ-9) and anxiety symptoms (at a
the developers. cut-off of 3 or more on GAD-2), although these were consid-
erably lower at a moderate level especially for depressive
2.5. Emotional Impact of the Epidemic. The emotional impact symptoms (PHQ‐9 ≥ 15). Fair or poor general health was
of the epidemic was assessed using the four items from the reported by 11.66% of the population. All health status vari-
IPQ-R related to emotional representations of the illness ables were statistically significantly higher in women except
(see Section 2.4). These four items were reworded to reflect for moderate depressive symptoms which were marginally
the epidemic as follows. (a) The current situation with the nonsignificant.
epidemic makes me feel angry. (b) I get depressed when I
think about the epidemic. (c) When I think about the epi- 3.3. Factor Analysis of the Illness Perception Questionnaire. A
demic I get upset. (d) This epidemic makes me feel afraid. factor analysis of the 15 items of the IPQ-R confirmed the
We combined these four items to derive a total score for hypothesized dimensions, namely, the “personal control”
the emotional impact of the epidemic (range from 4 to 20). dimension (related to beliefs about the ability of taking
In addition, in the analyses, we also present a binary variable measures to personally control the illness and self-efficacy),
corresponding to responses coded “agree or strongly agree” the “treatment control” dimension (beliefs about treatment
(i.e., coded at least 4 in each individual item) on average. expectancies), and the “emotional representation” dimension
of the epidemic (these represent the emotional impact of the
epidemic). The two items related to the timeline of the illness
2.6. Other Variables. We used direct questions in order to (acute/chronic and cyclical course) were more related to the
assess all remaining sociodemographic variables. Use of “treatment control” dimension. Due to their nature, the items
alcohol was also self-reported regarding frequency, quantity related to the consequences dimension were treated individ-
on a typical day, and binge drinking. We also asked two ually. (Details of the factor analysis are given in the supple-
COVID-19-related questions, the first concerning the amount mentary appendix, Table A2.)
of time spent out of home for reasons other than work and the
second regarding exposure to media for COVID-19-related 3.4. Emotional Impact of the Epidemic. The emotional impact
news in terms of time spent on a typical day. of the epidemic was assessed with the corresponding items of
4 Depression Research and Treatment

Table 1: Level of general and mental health of the sample during the COVID-19 epidemic in Greece (N = 3379).

Male Female Total sample p

(A) Mental health Continuous responses: mean (SD)
Depressive symptoms (scores on the PHQ-9) 5.95 (4.64) 7.25 (4.70) 6.90 (4.72) <0.001
Anxiety symptoms (scores on the GAD-2) 1.69 (1.40) 2.04 (1.40) 1.95 (1.41) <0.001
Binary responses: % of the sample
PHQ‐9 ≥ 10 (mild depressive symptoms) 18.11% 24.87% 23.02% <0.001
PHQ‐9 ≥ 15 (moderate depressive symptoms) 6.51% 8.55% 7.99% 0.051
GAD‐2 ≥ 3 (mild anxiety) 21.04% 30.24% 27.73% <0.001
GAD‐2 ≥ 4 (moderate anxiety) 11.28% 15.34% 14.23% 0.003
(B) General health
Fair or poor level of health (self-reported) 8.57% 12.82% 11.66% 0.001

the IPQ-R (see Methods). Approximately 25% of the sample Table 2: Intensity of perceptions (agree/strongly agree) about the
showed a strong emotional impact to the epidemic (Table 2). COVID-19 epidemic and its consequences in Greece (N = 3379).
This impact was larger in women (29% versus 14% in men),
in those with a lower education (31% versus 21% in those Dimension Male Female
Total
p
with a postgraduate degree) and in those with a lot of finan- sample
cial difficulties in their household (39% versus 18% in those High personal control 65.94% 66.46% 66.32% 0.776
without difficulties). All these differences were statistically High treatment control 26.25% 24.87% 25.24% 0.411
significant (p < 0:001, see Figure 1). Strong emotional impact 13.88% 28.65% 24.62% <0.001
There was a strong association with levels of depressive
Longer epidemic duration 47.40% 48.19% 47.97% 0.682
symptoms and emotional impact to the epidemic. Approxi-
mately 45% and 66% of those with mild or moderate depressive High severity of illness 26.03% 28.45% 27.79% 0.162
symptoms, respectively, showed a strong emotional impact to Severe financial consequences 69.52% 73.83% 72.65% 0.012
the epidemic (compared to 16% of those with low levels of Stigma related to
28.85% 25.40% 26.34% 0.042
depressive symptoms, p < 0:001). Similar figures were reported the illness
for anxiety symptoms. A similar association with a stronger
emotional impact was noted for levels of general health, with
52% of those who reported fair/poor general health showing a remaining 4 strategies were treated individually as they did not
stronger emotional impact compared to 21% of those with bet- load clearly to one of the two factors (details of the factor analysis
ter health (odds ratio: 3.84, p < 0:001 adjusted for age and sex). are given in the supplementary appendix, Tables A4 and A5).
Positive/active strategies were more often used in the study
3.5. Other Illness Perceptions. The sample showed high levels population compared to more supportive strategies. More
of personal control and self-efficacy (66.32%) as assessed dysfunctional strategies, such as denial, substance use, and
with the IPQ-R (Table 2). In contrast, only 25.24% of the giving up, were more rarely used. Supportive strategies and
population had an optimistic view about potential treatments religious coping were more likely to be used by women, while
of the illness, while 27.79% viewed the illness as potentially substance use by men. Denial and giving up did not differ
severe (Table 2). Sense of personal control increased with between genders (see Table A3 in the supplementary appendix).
age as was severity of illness (16.59% in the young versus Participants typically used several coping strategies at the
41.64% in the elderly). A higher educational status was also same time: overall, they reported the use of a mean number of
associated with higher personal control and lower levels of 3.85 positive/active strategies (SD: 1.24) versus 1.57 (SD: 1.2)
beliefs about the severity of illness. supportive/distractive strategies (see Figure 3).
Beliefs about future severe financial consequences were Regarding age and educational status, use of “planning”
also very prevalent (Table 3). Stigma related to the illness and religious coping increased with age. Instrumental support
was more prevalent in men compared to women and (getting advice from others) decreased with age. It is worth
decreased with a higher educational status. noting that two of the dysfunctional coping strategies, denial
and giving up, showed a U-type association with increased
3.6. Use of Coping Strategies. The percentage of the sample use at the two age extremes (the young and the elderly).
that used each one of the 13 coping strategies for at least a (Figure A1 in the supplementary appendix presents these
medium amount of time is shown in Figure 2. Acceptance, results in more detail.)
humor, and planning were the three more common strategies Positive coping strategies were more likely to be used in
used in both genders. (Table A3 in the supplementary those with high levels of personal and treatment control
appendix also presents the results in more detail.) (odds ratio: 1.91 and 1.53, respectively, p < 0:001 adjusted
Using factor analytical techniques, we derived two broad fac- for age and sex), while supportive coping strategies in those
tors, one positive/active comprised of 5 coping strategies and one with a stronger emotional impact of the epidemic (odds ratio:
supportive/distractive comprised of 4 coping strategies. The 2.45, p < 0:001 adjusted for age and sex). Denial was less
Depression Research and Treatment 5

31 31

26
24
21
29
25

14

Strong emotional impact Strong emotional impact

Total 9 years College

Male 12 years Postgraduate
Female Technical
(a) (b)

39 66

30 45
25

18

16

Strong emotional impact Strong emotional impact

1. Financial difficulties: none No depressive symptoms
2. Financial difficulties: few Mild depressive symptoms
3. Financial difficulties: some Moderate/severe depressive symptoms
4. Financial difficulties: a lot
(c) (d)

Figure 1: Associations of “strong emotional impact” of the COVID-19 epidemic with (a) gender, (b) educational status, (c) presence of
financial difficulties, and (d) depressive symptoms in Greece (N = 3379, all p < 0:001).

common in those with high levels of personal control (odds an increased prevalence. The number of positive strategies
ratio: 0.43, p < 0:001 adjusted for age and sex), while giving used was also negatively associated with depressive symp-
up was less common in those with high levels of personal tomatology independently. Illness perceptions related to high
and treatment control (odds ratio: 0.50 and 0.51, respectively, personal control or treatment control were also associated
p < 0:001 adjusted for age and sex). with fewer depressive symptoms. A stronger emotional
impact of the epidemic was associated with an increased
3.7. Regression Analysis of Depressive Symptoms. Table 3 pre- prevalence of depressive symptoms. It is also worth noting
sents the results of the logistic regression analysis for the that a positive association with depressive symptoms was also
association of depressive symptoms (binary variable) with found for a higher exposure to media for COVID-19-related
coping strategies adjusted for illness perceptions and other information and more time spent outside home for activities
COVID-19-related variables. It can be seen that a higher not related to work. Finally, individuals in isolation due to
score on the positive coping strategy dimension was associ- potential symptoms and students showed more depressive
ated with a lower prevalence of depressive symptoms, while symptoms. Similar analyses were obtained for anxiety scores
more supportive/distractive strategies were associated with (data on file).
6 Depression Research and Treatment

Table 3: Association of depressive symptoms (PHQ-9 scores ≥ 10) with coping strategies, illness perceptions, and other variables during the
COVID-19 epidemic in Greece (N = 3379).

Variable Adjusted OR1 95% CI2 p value

Gender
Men 1.00 Ref
Women 1.12 0.88–1.41 0.35
Age 0.98 0.96–0.99 <0.001
Married (compared to singles) 0.69 0.52–0.93 0.01
Being a student 1.72 1.16–2.54 0.006
Financial difficulties
No 1.00 Ref
Yes 2.05 1.66–2.53 <0.001
Alcohol consumption
Abstinent/small frequency 1.00 Ref
Moderate frequency 1.02 0.83–1.25 0.846
High frequency 1.79 1.26–2.53 0.001
Coping strategies
Coping: positive/active (scores) 0.88 0.81–0.95 0.001
Coping: supportive/distractive (scores) 1.15 1.05–1.27 0.004
Number of positive coping strategies used 0.80 0.69–0.94 0.005
Illness perceptions
Illness beliefs: high personal control
No 1.00 Ref
Yes 0.79 0.65–0.96 0.02
Illness perceptions: high treatment control
No 1.00 Ref
Yes 0.62 0.49–0.79 <0.001
Illness perceptions: strong emotional representations
No 1.00 Ref
Yes 4.09 3.31–5.04 <0.001
Illness perceptions: high severity of illness
No 1.00 Ref
Yes 2.08 1.66–2.60 <0.001
COVID-19 related
In isolation due to symptoms
No 1.00 Ref
Yes 2.79 1.42–5.49 0.003
Exposure to media for COVID-19-related news
Low–typical 1.00 Ref
High–excessive 1.76 1.38–2.24 <0.001
Time out of home (not work related)
Low–typical 1.00 Ref
High–excessive 2.68 1.18–6.10 0.02
1
OR: odds ratios adjusted for all other variables of the table plus educational status, employment status, locality, number of children, number of persons living at
home, number of supportive coping strategies used, being a person with a susceptible illness, being a carer of a susceptible person, having a business that stayed
open during the epidemic, and being a health professional. All variables omitted were not statistically significantly associated with depressive symptoms; 2CI:
confidence interval.

4. Discussion it was also associated with increasing depressive (and anxi-

ety) symptoms and a worse level of general health. The pres-
4.1. Main Findings. One in four of the participants in this sur- ence of financial difficulties in the household was associated
vey experienced a strong emotional impact due to the with a stronger emotional impact due to the epidemic.
COVID-19 epidemic. This effect was greater in women, and Despite this strong emotional impact, most participants used
Depression Research and Treatment 7

Coping mechanisms during the COVID-19 epidemic in Greece (N = 3379)

100
91
90

80 77
75
72
70
% of the population 70

60
54

50
42
40 37

30
24 23
20
12
10 5 5

Denial
Acceptance

Humor

Planning

Positive reframing

Active coping

Distraction

Venting

Emotional support

Instrumental support

Religion

Substance use

Giving up
Total
Male
Female

Figure 2: Coping strategies used during the COVID-19 epidemic in Greece (N = 3379). Positive/active strategies include acceptance, humor,
planning, positive reframing, and active coping. Supportive/distractive strategies include distraction, venting, emotional support, and
instrumental support.

positive/active strategies to cope with the stress of the epi- period of data collection which in our study was early April
demic and this was more likely in those who showed high while in the Papandreou et al. study was early May [45].
levels of personal control over the epidemic. In contrast, par-
ticipants with a stronger emotional impact turned to more 4.3. Coping Behaviors. Most of the participants adapted to the
supportive coping strategies. In the multivariate analysis, epidemic using positive/active coping strategies. Acceptance
increasing levels of current depressive symptoms were seen of the epidemic was very high and denial very small. In addi-
in the younger, in students, in those with a stronger emo- tion, participants used several positive coping behaviors at
tional impact to the epidemic, in those isolated due to the same time, therefore increasing their ability to adapt. At
COVID-19-related symptoms, and those overexposed to the same time, this strategy was associated with a better men-
media for COVID-19-related news. Lower levels of depres- tal health and fewer depressive and anxiety symptoms. It is
sion were seen in those using positive coping strategies and certainly possible that healthier individuals before the
showing high levels of personal and treatment control. epidemic were more likely to adapt using more functional
strategies, but given the strength of the association and the
dose-response effect on depression (since the number of
4.2. Prevalence of Depressive and Anxiety Symptoms. The positive strategies used was also negatively associated with
levels of depressive symptoms were higher compared to past depression), it is equally likely that these behavioral strategies
periods (that preceded the 2009 Greek financial crisis) but may have helped in mitigating the effects of the pandemic in
were comparable to the levels reported during the years of the mental health status of the population. More emotionally
financial crisis especially using the cut-off for moderate focused (supportive) strategies were used mainly by partici-
symptoms [38, 39]. A study in China [40] and another in pants who showed a strong emotional impact due to the epi-
Spain [41] also did not report higher than expected current demic. Although this was expected, this behavioral pattern
depressive symptoms, while a study in Italy [42] reported was also associated with a worsening of mental health. Chew
higher levels. It is difficult to compare these levels due to et al. [46] in their review of coping strategies during previous
the different methodologies used in these studies. Anxiety infectious disease outbreaks have reported similar results for
symptoms were higher compared to depression, a finding the effect of positive/active strategies versus supportive strat-
also confirmed in other surveys in China during this period egies in the mental health status of the population.
[40, 43, 44]. Compared to another web survey conducted in
Greece during this period [45], our study has shown 4.4. Associations of Depression. Although depressive symp-
increased depressive and anxiety levels probably due to the toms (using the typical PHQ-9 cut-off at 10) were more likely
8 Depression Research and Treatment

Number of ‘‘supportive-distractive’’ coping strategies used in the COVID-19

epidemic in Greece (N = 3379)
40

34 (p < 0.001)
35
31
30
31
% of the population

25 25

20 18
18 22
15
10
10 9

5
Mean number 1.57 (SD 1.2) 4
0
0 1 2 3 4
Number of strategies used by % of the population

Number of ‘‘positive-practical’’ coping strategies used in the COVID-19

epidemic in Greece (N = 3379)
45

40
41
35
29
% of the population

30 35

25
17 28
20

15 11
16
10
5
5 10
4 Mean number 3.85 (SD 1.24)
0
0 1 2 3 4
Number of strategies used by % of the population
Male
Female

Figure 3: Number of coping strategies used in the COVID-19 epidemic in Greece (n = 3379).

in women in the univariate analysis, this was not confirmed was associated with a higher depressive symptomatology.
in the multivariate analysis. This was mainly due to the inclu- Positive coping behaviors were negatively associated with
sion of the emotional impact variable which was the strongest depression independently of the illness perceptions.
correlate of depression in the analysis and was also stronger Overexposure to media for COVID-19-related news
in women. Students were more likely to report depression was also associated with a higher depressive symptomatol-
independently of age. Other studies from Spain and Greece ogy. This is a finding that has been reported by other stud-
also reported a similar finding [41, 47]. It is likely that the ies in China [43, 44, 48]. Although we cannot exclude
extra effort required for distance education and the uncer- reverse causality, it is equally likely that the way media
tainty regarding the progress of their studies may have played have presented the more severe extreme of the illness
a role in this finding. may have a negative impact in mental health, especially
Specific illness perceptions were also associated with for those overexposed to the news. On the other hand, time
depressive symptoms. A higher sense of personal control spent out of home was positively associated with depres-
over the epidemic and higher confidence in the efficacy of sion, a finding that more likely reflects an inability of par-
potential treatments were associated with fewer depressive ticipants with depression to follow the restrictive rules of
symptoms. In contrast, a higher belief in the severity of illness the lockdown.
Depression Research and Treatment 9

Subjects on isolation due to symptoms were also more Consent

likely to report depression. This is a finding confirmed else-
where as well [40] and highlights the need of mental health All potential participants gave their consent before entering the
assessment in those in quarantine or isolation as has been main survey page, by checking the appropriate box. The con-
suggested in a recent review [49]. sent form was also anonymous, and no email or other sign-in
procedure was asked from participants. The participants were
free to leave the survey at any point before submission.
4.5. Limitations. The convenient sample recruited through the
social media is not representative of the general population, and
we cannot exclude the possibility of selection bias. In addition, Conflicts of Interest
the cross-sectional nature of the study does not allow the inves-
tigation of the temporal association between variables. For this The authors declare that there is no conflict of interest in
reason, all reported associations could be bidirectional. relation to this paper.
Although we have used well-known and validated instruments
to assess the variables of interest, some of these measures have Acknowledgments
been adapted especially for the purposes of the current study
and their psychometric properties have not been studied in The authors would like to acknowledge the 3379 anonymous
detail. Therefore, issues of measurement bias are largely participants who took the time to complete the survey in this
unknown and may have influenced our findings. Depressive difficult period and helped in the realization of this research
symptoms were assessed with a well-validated instrument, but and all those who actively shared the link to the survey.
its reliability in different age groups is not known and results
might differ if age-specific instruments had been used [50]. Supplementary Materials

5. Conclusion This appendix has been supplied by the authors to give

readers additional information about the research reported
Our findings show that people in Greece adapted to the stress in this paper. Contents of the Appendix. Section 1: Table
caused by the epidemic using predominantly positive/active A1: basic description of the sample. Section 2: Table A2:
strategies. This behavioral pattern may have resulted in the factor analysis of IPQ-R items. Section 3: Table A3: use of
mitigation of the potential mental health effects of the epi- coping strategies by gender. Section 4: factor analysis of Brief
demic. Dysfunctional strategies were rarely adopted. One key COPE items. Table A4: initial factor analysis. Table A5: sec-
finding relevant from a public health perspective is the need ond factor analysis. Section 5: Figure A1: selected coping
to increase the sense of personal control over the epidemic mechanisms by age and educational status. (Supplementary
and enhance self-efficacy of the population. Public health Materials)
strategies often try to increase perceptions related to the vul-
nerability or the severity of the illness. Our findings however References
show that efforts to increase the perception of personal control
and to disseminate practical approaches towards the reduction [1] E. A. Holmes, R. C. O'Connor, V. H. Perry et al., “Multidisci-
plinary research priorities for the COVID-19 pandemic: a call
of risk may be more effective in protecting the mental health of
for action for mental health science,” The Lancet Psychiatry,
the population and increasing the compliance to imposed vol. 7, no. 6, pp. 547–560, 2020.
restrictions. It is likely that part of the success of Greece to
[2] J. P. Ioannidis, “A fiasco in the making? As the coronavirus
contain the spread of the outbreak can be attributed to the pandemic takes hold, we are making decisions without reliable
common sense illness perceptions and the functional behav- data,” Stat, vol. 17, 2020.
ioral patterns to cope with stress which were prevalent during [3] M. S. Hagger, S. Koch, N. L. D. Chatzisarantis, and S. Orbell,
this period. Finally, the role of media during the pandemic “The common-sense model of self-regulation: meta-analysis
should be reexamined and specific guidance regarding the and test of a process model,” Psychological Bulletin, vol. 143,
extent of reporting the extreme spectrum of the disease should no. 11, pp. 1117–1154, 2017.
be developed by health authorities and organizations. [4] K. Sim, Y. H. Chan, P. N. Chong, H. C. Chua, and S. W. Soon,
“Psychosocial and coping responses within the community
health care setting towards a national outbreak of an infectious
Data Availability disease,” Journal of Psychosomatic Research, vol. 68, no. 2,
pp. 195–202, 2010.
Data of this study are available from the authors upon
request. Please request data from the first author Petros Ska- [5] C. S. Carver, “You want to measure coping but your protocol’s
too long: consider the Brief COPE,” International Journal of
pinakis (p.skapinakis@gmail.com).
Behavioral Medicine, vol. 4, no. 1, pp. 92–100, 1997.
[6] K. M. Duangdao and S. C. Roesch, “Coping with diabetes in
Ethical Approval adulthood: a meta-analysis,” Journal of Behavioral Medicine,
vol. 31, no. 4, pp. 291–300, 2008.
This research was conducted in accordance with the Helsinki [7] H. Littleton, S. Horsley, S. John, and D. V. Nelson, “Trauma
Declaration and the ethical standards of our Institutional Bio- coping strategies and psychological distress: a meta-analysis,”
ethics Committee. The survey was completely anonymous. Journal of Traumatic Stress, vol. 20, no. 6, pp. 977–988, 2007.
10 Depression Research and Treatment

[8] J. T. Moskowitz, J. R. Hult, C. Bussolari, and M. Acree, “What flu outbreak: cross sectional telephone survey,” Bmj, vol. 339,
works in coping with HIV? A meta-analysis with implications no. jul02 3, p. b2651, 2009.
for coping with serious illness,” Psychological Bulletin, vol. 135, [24] J. T. F. Lau, X. Yang, H. Tsui, and J. H. Kim, “Monitoring com-
no. 1, pp. 121–141, 2009. munity responses to the SARS epidemic in Hong Kong: from
[9] S. C. Roesch, L. Adams, A. Hines et al., “Coping with prostate day 10 to day 62,” Journal of Epidemiology and Community
cancer: a meta-analytic review,” Journal of Behavioral Medi- Health, vol. 57, no. 11, pp. 864–870, 2003.
cine, vol. 28, no. 3, pp. 281–293, 2005. [25] L. Williams, A. Regagliolo, and S. Rasmussen, “Predicting psy-
[10] P. Y. Chen and H. C. Chang, “The coping process of patients chological responses to influenza A, H1N1 (“swine flu”): the
with cancer,” European Journal of Oncology Nursing, vol. 16, role of illness perceptions,” Psychology, Health & Medicine,
no. 1, pp. 10–16, 2012. vol. 17, no. 4, pp. 383–391, 2012.
[11] J. Raskauskas and A. Huynh, “The process of coping with [26] P. Wu, Y. Fang, Z. Guan et al., “The psychological impact of
cyberbullying: a systematic review,” Aggression and Violent the SARS epidemic on hospital employees in China: exposure,
Behavior, vol. 23, pp. 118–125, 2015. risk perception, and altruistic acceptance of risk,” Canadian
[12] C. Cheng, W. M. Wong, and K. W. Tsang, “Perception of ben- Journal of Psychiatry, vol. 54, no. 5, pp. 302–311, 2009.
efits and costs during SARS outbreak: an 18-month prospec- [27] M. Kosinski, S. C. Matz, S. D. Gosling, V. Popov, and
tive study,” Journal of Consulting and Clinical Psychology, D. Stillwell, “Facebook as a research tool for the social sciences:
vol. 74, no. 5, pp. 870–879, 2006. opportunities, challenges, ethical considerations, and practical
[13] D. Y. L. Yeung and H. H. Fung, “Age differences in coping and guidelines,” The American Psychologist, vol. 70, no. 6, pp. 543–
emotional responses toward SARS: a longitudinal study of 556, 2015.
Hong Kong Chinese,” Aging & Mental Health, vol. 11, no. 5, [28] K. Kroenke, R. L. Spitzer, and J. B. W. Williams, “The PHQ-9:
pp. 579–587, 2007. validity of a brief depression severity measure,” Journal of Gen-
[14] W. Zhong, Y. Kim, and M. Jehn, “Modeling dynamics of an eral Internal Medicine, vol. 16, no. 9, pp. 606–613, 2001.
influenza pandemic with heterogeneous coping behaviors: case [29] M. Ikonomou, P. Skapinakis, M. Eleftheriadou et al., “Routine
study of a 2009 H1N1 outbreak in Arizona,” Computational & assessment of depression in chronic kidney disease patients
Mathematical Organization Theory, vol. 19, no. 4, pp. 622– without the involvement of mental health professionals: a
645, 2013. two stage approach,” BANTAO Journal, vol. 9, Suppl.1, p. 18,
[15] K. Petrie and J. Weinman, “Why illness perceptions matter,” 2011, (http://bantao.org).
Clinical Medicine (London, England), vol. 6, no. 6, pp. 536– [30] K. Kroenke, R. L. Spitzer, J. B. W. Williams, P. O. Monahan,
539, 2006. and B. Löwe, “Anxiety disorders in primary care: prevalence,
[16] M. Prior, J. M. Burr, C. R. Ramsay et al., “Evidence base for an impairment, comorbidity, and detection,” Annals of Internal
intervention to maximise uptake of glaucoma testing: a theory- Medicine, vol. 146, no. 5, pp. 317–325, 2007.
based cross-sectional survey,” BMJ Open, vol. 2, no. 2, article [31] P. Skapinakis, “The 2-item Generalized Anxiety Disorder scale
e000710, 2012. had high sensitivity and specificity for detecting GAD in
[17] M. N. Raftery, K. Sarma, A. W. Murphy, D. De la Harpe, primary care,” Evidence-Based Medicine, vol. 12, no. 5,
C. Normand, and B. E. McGuire, “Chronic pain in the Repub- p. 149, 2007.
lic of Ireland-community prevalence, psychosocial profile and [32] E. L. Idler and Y. Benyamini, “Self-rated health and mortality:
predictors of pain-related disability: results from the preva- a review of twenty-seven community studies,” Journal of
lence, impact and cost of chronic pain (PRIME) study, part Health and Social Behavior, vol. 38, no. 1, pp. 21–37, 1997.
1,” Pain, vol. 152, no. 5, pp. 1096–1103, 2011. [33] C. S. Carver and J. Connor-Smith, “Personality and coping,”
[18] R. Lowe, A. Porter, H. Snooks, L. Button, and B. A. Evans, “The Annual Review of Psychology, vol. 61, no. 1, pp. 679–704,
association between illness representation profiles and use of 2010.
unscheduled urgent and emergency health care services,” Brit- [34] M. Kapsou, G. Panayiotou, C. M. Kokkinos, and A. G.
ish Journal of Health Psychology, vol. 16, no. 4, pp. 862–879, Demetriou, “Dimensionality of coping: an empirical contribu-
2011. tion to the construct validation of the Brief-COPE with a
[19] M. J. Figueiras and N. C. Alves, “Lay perceptions of serious ill- Greek-speaking sample,” Journal of Health Psychology,
nesses: an adapted version of the revised illness perception vol. 15, no. 2, pp. 215–229, 2010.
questionnaire (IPQ-R) for healthy people,” Psychology & [35] R. Moss-Morris, J. Weinman, K. Petrie, R. Horne,
Health, vol. 22, no. 2, pp. 143–158, 2007. L. Cameron, and D. Buick, “The revised illness perception
[20] J. R. Cates, N. T. Brewer, K. I. Fazekas, C. E. Mitchell, and J. S. questionnaire (IPQ-R),” Psychology & Health, vol. 17, no. 1,
Smith, “Racial differences in HPV knowledge, HPV vaccine pp. 1–16, 2002.
acceptability, and related beliefs among rural, southern women,” [36] E. Broadbent, C. Wilkes, H. Koschwanez, J. Weinman,
The Journal of Rural Health, vol. 25, no. 1, pp. 93–97, 2009. S. Norton, and K. J. Petrie, “A systematic review and meta-
[21] K. I. Fazekas, N. T. Brewer, and J. S. Smith, “HPV vaccine analysis of the brief illness perception questionnaire,” Psychol-
acceptability in a rural southern area,” Journal of Women's ogy & Health, vol. 30, no. 11, pp. 1361–1385, 2015.
Health (2002), vol. 17, no. 4, pp. 539–548, 2008. [37] A. A. Kaptein, I. M. van Korlaar, L. D. Cameron, C. Y. Vossen,
[22] J. Lai, S. Ma, Y. Wang et al., “Factors associated with mental F. J. M. van der Meer, and F. R. Rosendaal, “Using the common-
health outcomes among health care workers exposed to coro- sense model to predict risk perception and disease-related worry
navirus disease 2019,” JAMA Network Open, vol. 3, no. 3, arti- in individuals at increased risk for venous thrombosis,” Health
cle e203976, 2020. Psychology, vol. 26, no. 6, pp. 807–812, 2007.
[23] G. J. Rubin, R. Amlot, L. Page, and S. Wessely, “Public percep- [38] M. Economou, M. Madianos, L. E. Peppou, A. Patelakis, and
tions, anxiety, and behaviour change in relation to the swine C. N. Stefanis, “Major depression in the era of economic crisis:
Depression Research and Treatment 11

a replication of a cross-sectional study across Greece,” Journal

of Affective Disorders, vol. 145, no. 3, pp. 308–314, 2013.
[39] P. Skapinakis, S. Bellos, S. Koupidis, I. Grammatikopoulos,
P. N. Theodorakis, and V. Mavreas, “Prevalence and sociode-
mographic associations of common mental disorders in a
nationally representative sample of the general population of
Greece,” BMC Psychiatry, vol. 13, no. 1, 2013.
[40] C. Wang, R. Pan, X. Wan et al., “Immediate psychological
responses and associated factors during the initial stage of the
2019 coronavirus disease (COVID-19) epidemic among the
general population in China,” International Journal of Environ-
mental Research and Public Health, vol. 17, no. 5, p. 1729, 2020.
[41] N. Ozamiz-Etxebarria, M. Dosil-Santamaria, M. Picaza-Gor-
rochategui, and N. Idoiaga-Mondragon, “Niveles de estrés,
ansiedad y depresión en la primera fase del brote del
COVID-19 en una muestra recogida en el norte de España,”
Cadernos de Saúde Pública, vol. 36, no. 4, article e00054020,
2020.
[42] C. Mazza, E. Ricci, S. Biondi et al., “A nationwide survey of
psychological distress among Italian people during the
COVID-19 pandemic: immediate psychological responses
and associated factors,” International Journal of Environmen-
tal Research and Public Health, vol. 17, no. 9, p. 3165, 2020.
[43] Y. Huang and N. Zhao, “Generalized anxiety disorder, depres-
sive symptoms and sleep quality during COVID-19 outbreak
in China: a web-based cross-sectional survey,” Psychiatry
Research, vol. 288, article 112954, 2020.
[44] M. Y. Ni, L. Yang, C. M. C. Leung et al., “Mental health, risk
factors, and social media use during the COVID-19 epidemic
and cordon sanitaire among the community and health pro-
fessionals in Wuhan, China: Cross-Sectional Survey,” JMIR
Public Health Surveill, vol. 7, no. 5, article e19009, 2020.
[45] C. Papandreou, V. Arija, E. Aretouli, K. K. Tsilidis, and
M. Bulló, “Comparing eating behaviours, and symptoms of
depression and anxiety between Spain and Greece during the
COVID‐19 outbreak: Cross‐sectional analysis of two different
confinement strategies,” European Eating Disorders Review,
pp. 1–11, 2020.
[46] Q. H. Chew, K. C. Wei, S. Vasoo, H. C. Chua, and K. Sim,
“Narrative synthesis of psychological and coping responses
towards emerging infectious disease outbreaks in the general
population: practical considerations for the COVID-19 pan-
demic,” Singapore Medical Journal, vol. 61, no. 7, pp. 350–
356, 2020.
[47] M. E. Patsali, D. V. Mousa, E. V. K. Papadopoulou et al., “Uni-
versity students’ changes in mental health status and determi-
nants of behavior during the COVID-19 lockdown in Greece,”
Psychiatry Research, vol. 292, p. 113298, 2020.
[48] J. Gao, P. Zheng, Y. Jia et al., “Mental health problems and
social media exposure during COVID-19 outbreak,” PLoS
One, vol. 15, no. 4, article e0231924, 2020.
[49] S. K. Brooks, R. K. Webster, L. E. Smith et al., “The psycholog-
ical impact of quarantine and how to reduce it: rapid review of
the evidence,” Lancet, vol. 395, no. 10227, pp. 912–920, 2020.
[50] G. Galeoto, J. Sansoni, M. Scuccimarri et al., “A psychometric
properties evaluation of the Italian version of the geriatric
depression scale,” Depression Research and Treatment,
vol. 2018, Article ID 1797536, 7 pages, 2018.