1
COVID-19 and depressive symptoms in students before and during lockdown
1
2
Nicola Meda
1
; Susanna Pardini
2
, MSc; Irene Slongo
2
, BSc; Luca Bodini
3
, MSc; Paolo Rigobello
1
;
3
Francesco Visioli
1,4*
, PhD; Caterina Novara
2
, PhD
4
5
1
Department of Molecular Medicine, University of Padova, Padova, Italy
6
2
Department of General Psychology, University of Padova, Padova, Italy
7
3
Department of Statistics and Quantitative Methods, University of Milano-Bicocca, Milan, Italy
8
4
IMDEA-Food, CEI UAM + CSIC, Madrid, Spain
9
10
*
Corresponding author: Francesco Visioli, PhD, Department of Molecular Medicine, University of
11
Padova, Viale G. Colombo 3, 35131 Padova, Italy (francesco.visioli@unipd.it)
12
13
14
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 30, 2020. ; https://doi.org/10.1101/2020.04.27.20081695doi: medRxiv preprint
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.
2
ABSTRACT
15
The lockdown due to coronavirus pandemic may exacerbate depressive symptoms, experts argue.
16
Here we report that students, a high-risk category for mental disorders, report on average worse
17
depressive symptoms than six months before isolation. The prospective data reported herein should
18
alert clinician of a possible aggravation as well as new-onsets of depressive symptoms in students.
19
20
The current coronavirus pandemic has been affecting Europe since late February 2020, forcing
21
governments to put citizens in lockdown. Among growing concerns of the effects of isolation on
22
mental health
1,2
, only retrospective data are available to assess if actual changes occur
3
. Here we
23
provide prospective evidence of a change in depressive symptomatology of Italian students during
24
COVID-19-related lockdown.
25
The study was approved by the University of Padova Ethical Committee of Psychology and
26
participants provided informed consent. Between October 3
rd
and October 23
rd
2019, we introduced
27
the study to approximately 1000 University of Padova students, 153 of which matched target
28
population characteristics (Italian native speaker students, age 18-30) and completed a demographic
29
questionnaire and the Italian version of Beck Depression Inventory-2
4
(BDI-2, a validated self-
30
report questionnaire for depressive symptoms evaluation, the score of which correlates with severity
31
of depressive symptomatology) online
5
, both in October and in April (between 3
rd
-23
rd
) 2020. We
32
implemented generalised linear mixed models to evince if BDI-2 score changed during isolation
33
with respect to the scores reported 6 months before. To assess a percentage change in BDI-2 score,
34
we defined %ΔBDI-2 as the difference between BDI-2 score during lockdown and before
35
lockdown, the whole divided by BDI-2 score before lockdown + 1 and analysed %ΔBDI-2 with
36
linear mixed-effects models. To assess clinically relevant changes in depressive symptoms, we
37
employed multinomial regression models. Sample characteristics and models employed are reported
38
in Tables A and B, respectively. Anonymised dataset, further details on data analysis, and script are
39
provided as Supplementary Material.
40
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 30, 2020. ; https://doi.org/10.1101/2020.04.27.20081695doi: medRxiv preprint
3
BDI-2 total score is slightly higher during lockdown than before (Figure, A and Table). We
41
recorded that the median percentage increase is higher in males (+36%; IQR = -12 – 91%) than in
42
females (+16%; -26 – 89%) and is independent from a history of mental disorder (Figure, B),
43
although students with such history report higher before and during lockdown BDI-2 scores than
44
students without any established diagnosis of psychopathology (Figure, C and Table). This increase
45
is not significantly linked to sex, familiarity for a mental disorder, worry for one’s economic
46
situation, or residence. Statistically, it is significantly linked to BDI-2 score before lockdown
47
(Figure, D) and age, evidencing that younger participants with lower BDI-2 score before lockdown
48
report higher percentage increases in BDI-2 score during lockdown. To assess if such increase
49
could be clinically relevant, we divided participants into three clinically useful categories according
50
to BDI-2 scores before lockdown (below 90
th
percentile, above 95
th
percentile, and between these
51
two ranges
4
) and tested how many participants switched from one category to another, or remained
52
in the same one during lockdown. We fit the observed data to a multinomial regression model and
53
found that a median increase of 22% in BDI-2 score (IQR= -21 – 90%) would not clinically affect
54
79,2% of our target population (IQR = 74,7 – 81,4%); 8,2% (6,9 – 9,8%) would progress to a more
55
serious clinical category (either from < 90
th
to 90
th
-95
th
range or from this latter to > 95
th
); and 6,2%
56
(5,3 – 7,2%) would directly progress from < 90
th
percentile category to the most severe clinical
57
category (Figure, E and F). Less than 5% of participants would improve.
58
As Italy was entirely put in lockdown, it is impossible to assess isolation-independent
59
changes in BDI-2 score. Students could be diversely affected by lockdowns: isolation may be
60
responsible of a median increase of 22% in BDI-2 score, which would be clinically relevant for up
61
to ≈ 15% of our target population. Our data should alert clinicians of possible aggravation of
62
depressive symptoms in students, independently from a history of mental disorder.
63
64
65
66
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 30, 2020. ; https://doi.org/10.1101/2020.04.27.20081695doi: medRxiv preprint
4
Author contributions: All authors designed the study protocol, interpreted data and critically
67
revised the manuscript; N.M. acquired data and analysed it and drafted the manuscript; P.R., F.V.
68
C.N., S.P. provided technical, material or administrative support to the study; F.V., C.N., S.P.
69
provided their supervision and expertise.
70
Competing interests: the authors declare no competing interests
71
Funding/Support: this study received no financial support
72
Additional Information: Dataset and R Script for analysis are provided as Supplementary Material
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 30, 2020. ; https://doi.org/10.1101/2020.04.27.20081695doi: medRxiv preprint
5
REFERENCES
93
1. Holmes, E. A. et al. Multidisciplinary research priorities for the COVID-19 pandemic: a call for
94
action for mental health science. The Lancet Psychiatry S2215036620301681 (2020)
95
doi:10.1016/S2215-0366(20)30168-1.
96
2. Reger, M. A., Stanley, I. H. & Joiner, T. E. Suicide Mortality and Coronavirus Disease 2019—A
97
Perfect Storm? JAMA Psychiatry (2020) doi:10.1001/jamapsychiatry.2020.1060.
98
3. Brooks, S. K. et al. The psychological impact of quarantine and how to reduce it: rapid review of
99
the evidence. The Lancet 395, 912–920 (2020).
100
4. Sica, C. & Ghisi, M. The Italian versions of the Beck Anxiety Inventory and the Beck
101
Depression Inventory-II: Psychometric properties and discriminant power. in Leading-edge
102
psychological tests and testing research 27–50 (Nova Science Publishers, 2007).
103
5. Harris, P. A. et al. Research electronic data capture (REDCap)—A metadata-driven methodology
104
and workflow process for providing translational research informatics support. Journal of
105
Biomedical Informatics 42, 377–381 (2009).
106
107
108
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 30, 2020. ; https://doi.org/10.1101/2020.04.27.20081695doi: medRxiv preprint