The first aim of this study was to assess the prevalence of neuropsychological impairments and psychiatric symptoms 69 months after SARS-CoV-2 infection. We compared patients’ performances with available normative data to identify the number of impaired scores per patient, group, and test. The prevalence of cognitive impairments in each group 236.51 ± 22.54 days after infection is set out in Table 2, and the prevalence of psychiatric symptoms in Table 3
The three groups differed significantly on i) long-term episodic memory in both the verbal (Grober and Buschke (RL/RI 16) delayed free recall, H = 10.75, p = .005) and visual (Rey Figure delayed free recall, H = 6.15, p = .046) modalities, ii) multimodal emotion recognition (GERT; H = 7.55, p = .023), iii) cognitive complaints (QPC; H = 6.38, p = .041) and anosognosia for memory dysfunction (SAD; H = 7.84, p = .020). The other effects were not significant (p > .05 for all comparisons).
Episodic memory. For Grober and Buschke delayed free recall, the mild patients scored significantly higher than the severe patients (z = 3.04, p = .002), but the other two pairwise comparisons were not significant after Bonferroni correction (moderate vs. severe: z = -1.47, p = .141; mild vs. moderate: z = 2.00, p = .046). Pairwise comparisons were not significant for visual episodic memory (mild vs. moderate: z = 2.26, p = .023; mild vs. severe: z = 0.48, p = .61; moderate vs. severe: z = 1.89, p = .059).
Emotion recognition. Mild patients scored significantly higher than moderate patients (z = 2.61, p = 0.009), but neither the difference between mild and severe patients (z = 1.97, p = 0.048), nor the difference between moderate and severe patients (z = .49, p = .620) reached significance after Bonferroni correction.
Cognitive complaints and anosognosia. Mild patients had more cognitive complaints than severe patients (z = -2.55, p = .010), but there were no differences between either the mild and moderate patients (z = -1.31, p = .191), or the moderate and severe patients (z = -0.93, p = .351). By contrast, severe patients exhibited more anosognosia for memory dysfunction than mild patients did (z = 2.97, p = .003), while there were no differences between either the mild and moderate patients (z = 1.41, p = .158) or the moderate and severe patients (z = − 0.76, p = .443).
ii) Psychiatric data
The three groups differed significantly on depression (H = 9.40, p = .009), state anxiety (H = 12.93, p = .002), emotional apathy (H = 7.10, p = .029), stress (H = 7.55, p = .023), and emotional contagion (H = 9.73, p = .002). The other effects were not significant (p > .05 for all comparisons). Pairwise comparisons for each of these group differences are described below.
Depression, stress, and state anxiety. The mild patients were more depressed, stressed and anxious than the severe patients (BDI-II: z = -2.99, p = .003; PSS: z = -2.55, p = .010; STAI-S: z = -3.57, p < .001), while there were no differences between either the severe and moderate patients (BDI-II: z = -1.38, p = .165; PSS: z = -1.08, p = .281; STAI-S: z = -1.76, p = .078) or the mild and moderate patients (BDI II: z = -1.66, p = .097; PSS: z = -1.08, p = .281; STAI-S: z = -1.72, p = .085).
Apathy. For the AMI emotional subscore, pairwise comparisons failed to reach significance after Bonferroni correction (severe vs. mild: z = 2.32, p = .020; severe vs. moderate: z = 2.20, p = .028, mild vs. moderate: z = 0.08, p = .933).
Emotional contagion. Pairwise comparisons failed to reach significance after Bonferroni correction (severe vs. moderate: z = -3.03, p = .017; severe vs. mild: z = -1.89, p = .059; moderate vs. mild: z = 1.18, p = .237).
iii) Fatigue and quality of life
Finally, the three groups differed on insomnia (H = 6.66, p = .036), fatigue (H = 6.45, p = .040), vitality (H = 6.50, p = .039), and emotional wellbeing (H = 9.18, p = .010). The other effects were not significant (p > .05 for all comparisons).
The mild patients reported more fatigue than the severe patients (z = -2.57, p = .010), while there were no differences between either the mild and moderate patients (z = -0.71, p = .481) or both moderate and severe patients (z = -1.52, p = .130). Pairwise comparisons did not reach significance after Bonferroni correction (mild vs. moderate: z = -1.99, p = .046; mild vs. severe: z = -2.28, p = .023; moderate vs. severe: z = -0.71, p = .481).
Conversely, severe patients reported more vitality, emotional wellbeing and social function than mild patients (vitality: z = 2.65, p = .008; wellbeing: z = 2.97, p = .003; social function: z = 2.99, p = .002). Pairwise comparisons between severe and moderate patients did not reach significance after Bonferroni correction (wellbeing: z = 2.01, p = .044; vitality: z = 1.06, p = .290; social function: z = 1.66, p = .097), nor did those between mild and moderate patients (wellbeing: z = 0.68, p = .494; vitality: z = 1.12, p = .263; social function: z = 1.06, p = .290) (see Table 4).
Table 4
Quality of life of patients with mild, moderate or severe COVID-19 69 months post-infection
Quality of life domains (SF-36) +
|
Mild (n = 15)
Mean (± SD)
|
Moderate (n = 15)
Mean (± SD)
|
Severe (n = 15)
Mean (± SD)
|
p value
|
Overall health
|
62.67 (± 16.89)
|
59.33 (± 27.31)
|
66.00 (± 24.14)
|
.808
|
Physical function
|
80.00 (± 17.22)
|
82.33 (± 19.44)
|
77.33 (± 24.41)
|
.806
|
Physical role
|
58.33 (± 30.86)
|
53.33 (± 43.16)
|
71.67 (± 36.43)
|
.353
|
Emotional role
|
64.45 (± 36.67)
|
73.34 (± 36.08)
|
80.00 (± 37.38)
|
.314
|
Social function
|
57.50 (± 23.05)
|
66.67 (± 31.93)
|
85.00 (± 18.42)
|
.011
|
Physical pain
|
57.83 (± 20.81)
|
72.00 (± 29.40)
|
71.83 (± 25.61)
|
.153
|
Emotional wellbeing
|
58.13 (± 17.75)
|
61.33 (± 24.96)
|
79.2 (± 17.90)
|
.010
|
Vitality score
|
38.66 (± 16.20)
|
49.00 (± 27.14)
|
56.00 (± 14.17)
|
.039
|
Health modification
|
30.00 (± 16.90)
|
35.00 (± 24.64)
|
43.33 (± 17.59)
|
.143
|
+ The higher the score, the better the quality of life.
Relationships between neuropsychological deficits, psychiatric symptoms, and other secondary variables
The third aim was to examine whether the presence of long-term neuropsychological deficits was correlated with psychiatric symptoms and/or other clinically relevant variables.
The results of the multiple regression performed on each cognitive variable are set out in Table 5. Interestingly, apathy, depression, anxiety, emotion regulation, emotion contagion, stress, PTSD, dissociative disorders, anosmia and diabetes all proved to be variables of interest when it came to explaining the neuropsychological sequelae. Therefore, both psychiatric and nonpsychiatric data correlated with neuropsychological deficits across the three groups. There were at least three patterns of results, depending on the neuropsychological domain: i) patterns in which neuropsychological sequelae did not correlate with any psychiatric variables, but did with other clinical variables, such as visuospatial long-term episodic memory (delayed recall of Rey-Osterrieth complex figure); ii) patterns in which neuropsychological sequelae correlated with both psychiatric and clinical variables, such as the object and action naming task scores (language); and iii) patterns in which neuropsychological sequelae only correlated with psychiatric variables, such as categorical verbal fluency. There was also a fourth possible pattern where the neuropsychological sequelae correlated neither with psychiatric variables nor with clinical ones, such as the score on the object decision task (object perception).
To reduce the dimensionality of the dataset, we computed a PCA. We selected the first three orthogonal components accounting for 43.67% of the total variance. The first component, accounting for 26.75% of the total variance, was difficult to interpret in terms of underlying cognitive processes, as it included language (semantic word and image matching), executive functions (mental flexibility), verbal episodic memory, and emotion recognition. Interestingly, these happened to be precisely the variables on which the three groups differed significantly (see Section “Neuropsychological and psychiatric symptoms as a function of disease severity”). We therefore labeled this component respiratory disease severity. The second component (9.79% of total variance) was labeled attention and anosognosia, as it included alertness, divided attention, and anosognosia for executive dysfunction. The third component (7.15% of total variance) was labeled instrumental functions, as it included language, visual perception, and ideomotor praxis.
For the respiratory disease severity component, the best fit was achieved with emotional apathy (R2 = .28, p = .007), stress (R2 = .19, p = .013), and anosmia (R2 = .11, p = .03). For the attention and anosognosia component, the multiple regression was not significant (p > .1). For the instrumental functions component, the best fit was achieved with anosmia (R2 = .23, p = .04), mania (R2 = .23, p = .006), and social apathy (R2 = .17, p = .004).
Table 5
Multiple regression results for each of the neuropsychological variables
|
|
Regressor
|
R2
|
p value
|
Memory functions
|
|
|
|
Verbal episodic memory
|
Grober & Buschke (RL/RI 16) - Immediate recall
|
ns
|
ns
|
ns
|
|
Grober & Buschke (RL/RI 16) - Delayed free recall
|
AMI – Emotional apathy
|
.45
|
.006
|
|
|
Epworth - Sleepiness
|
.20
|
.022
|
|
|
ERQ – Emotion regulation
|
.13
|
.034
|
|
|
ECS – Emotion contagion
|
.08
|
.034
|
|
Grober & Buschke (RL/RI 16) - Delayed total recall
|
AMI – Emotional apathy
|
.34
|
.022
|
|
|
Epworth - Sleepiness
|
.22
|
.031
|
|
|
AMI – Social apathy
|
.15
|
.034
|
Visuospatial episodic memory
|
Rey Figure - Copy time
|
ISI - Insomnia
|
.46
|
.005
|
|
|
ERQ – Emotional regulation
|
.18
|
.035
|
|
Rey Figure - Score
|
ns
|
ns
|
ns
|
|
Rey Figure - Immediate recall (3')
|
ns
|
ns
|
ns
|
|
Rey Figure - Delayed recall (20')
|
ISI – Insomnia
|
.30
|
.034
|
|
|
Days of hospitalization
|
.22
|
.039
|
|
|
Sniff test (anosmia)
|
.21
|
.013
|
Verbal short-term memory
|
MEM III - Spans
|
ns
|
ns
|
ns
|
Visuospatial short-term memory
|
WAIS IV - Spans
|
DES – Dissociation
|
.30
|
.035
|
Executive functions
|
|
|
|
Inhibition
|
Stroop (GREFEX) - Interference - Time
|
ns
|
ns
|
ns
|
|
Stroop (GREFEX)- Interference - Errors
|
AMI – Total apathy
|
.37
|
.015
|
|
|
ERQ – Emotion regulation
|
.21
|
.030
|
|
Stroop (GREFEX) - Interference/Naming - Score
|
ns
|
ns
|
ns
|
Working memory
|
MEM III – Verbal working memory
|
AMI – Behavioral apathy
|
.33
|
.026
|
|
WAIS IV - Visuospatial working memory
|
STAI-T Anxiety
|
.39
|
.013
|
|
|
Diabetes
|
.25
|
.014
|
|
|
STAI-S Anxiety
|
.13
|
.03
|
|
TAP - Working memory item omissions
|
AMI – Emotional apathy
|
.30
|
.035
|
|
TAP - Working memory false alarms
|
Diabetes
|
.47
|
.005
|
|
|
Mania – Goldberg Inventory
|
.16
|
.044
|
Mental flexibility
|
TMT A (GREFEX) - Time
|
ns
|
ns
|
ns
|
|
TMT A (GREFEX) - Errors
|
ns
|
ns
|
ns
|
|
TMT B (GREFEX) - Time
|
ns
|
ns
|
ns
|
|
TMT B (GREFEX) - Errors
|
AMI – Total apathy
|
.41
|
.010
|
|
|
ERQ – Emotion regulation
|
.14
|
.024
|
|
|
Gender
|
.13
|
.025
|
|
TMT B (GREFEX) - Perseverations
|
STAI-T Anxiety
|
.55
|
.002
|
|
|
BDI-II - Depression
|
.16
|
.026
|
|
|
ISI – Insomnia
|
.20
|
< .001
|
|
TMT B-A (GREFEX) - Score
|
ns
|
ns
|
ns
|
|
Verbal fluency (GREFEX) - Literal (2')
|
ns
|
ns
|
ns
|
|
Verbal fluency (GREFEX) - Categorical fluency (2')
|
DES - Dissociation
|
.28
|
.047
|
Incompatibility
|
TAP - Compatibility – Reaction time
|
AMI – Social apathy
|
.50
|
.003
|
|
TAP - Compatibility - False alarms
|
ns
|
ns
|
ns
|
|
TAP - Incompatibility - Reaction Time
|
Sniff test (anosmia)
|
.28
|
.043
|
|
|
DES - Dissociation
|
.25
|
.026
|
|
|
ERQ – Emotion regulation
|
.13
|
.028
|
|
|
Epworth - Sleepiness
|
.10
|
.017
|
|
TAP - Incompatibility - False alarms
|
Sniff test (anosmia)
|
.27
|
.045
|
|
TAP - Incompatibility - Visual field score
|
Days of hospitalization
|
.39
|
.013
|
|
|
Diabetes
|
.23
|
.007
|
|
|
STAI-T Anxiety
|
.08
|
.041
|
|
|
PCL-5 Posttraumatic stress disorder
|
.06
|
.041
|
|
TAP - Incompatibility task - Hands score
|
AMI – Social apathy
|
.44
|
.008
|
|
TAP - Incompatibility task - Visual fields * Hands score
|
ISI – Insomnia
|
.33
|
.025
|
|
|
STAI-T Anxiety
|
.24
|
.025
|
|
|
BDI-II - Depression
|
.18
|
.019
|
Attentional functions
|
|
|
|
Phasic alertness
|
TAP - Without warning sound - Reaction time
|
Gender
|
.35
|
.019
|
|
TAP - Without warning sound - SD of reaction time
|
AMI – Social apathy
|
.64
|
< .001
|
|
|
Diabetes
|
.11
|
.041
|
|
|
Sniff test (anosmia)
|
.10
|
.023
|
|
TAP - With warning sound - Reaction time
|
DES - Dissociation
|
.30
|
.033
|
|
TAP - With warning sound - SD of reaction time
|
ns
|
ns
|
ns
|
|
TAP - Alertness index
|
Gender
|
.28
|
.041
|
Sustained attention
|
TAP - Items Omissions
|
STAI-Trait Anxiety
|
.46
|
.005
|
|
TAP - False alarm
|
ns
|
ns
|
ns
|
Divided attention
|
TAP - Audio condition - Reaction time
|
DES - Dissociation
|
.41
|
.010
|
|
|
AMI – Behavioral apathy
|
.27
|
.008
|
|
TAP - Visual condition - Reaction time
|
Days of hospitalization
|
.38
|
.014
|
|
|
Sniff test (anosmia)
|
.27
|
.009
|
|
|
ISI - Insomnia
|
.23
|
< .001
|
|
|
AMI – Emotional apathy
|
.05
|
.016
|
|
TAP - Total omissions
|
ns
|
ns
|
ns
|
|
TAP - Total false alarms
|
AMI – Emotional apathy
|
.32
|
.029
|
|
|
Epworth - Sleepiness
|
.28
|
.015
|
|
|
AMI – Social apathy
|
.16
|
.023
|
Instrumental functions
|
|
|
|
Language
|
BECLA - Semantic image matching
|
ns
|
ns
|
ns
|
|
BECLA - Semantic word matching
|
ns
|
ns
|
ns
|
|
BECLA - Object and action image naming
|
ECS – Emotional contagion
|
.38
|
.014
|
|
|
STAI-State Anxiety
|
.29
|
.007
|
|
|
AMI – Emotional apathy
|
.10
|
.014
|
|
|
ISI - Insomnia
|
.08
|
.012
|
|
BECLA - Word repetition
|
NV
|
NV
|
NV
|
|
BECLA - Nonword repetition
|
ns
|
ns
|
ns
|
Ideomotor praxis
|
Symbolic gestures
|
ERQ – Emotion regulation
|
.35
|
.019
|
|
|
AMI – Behavioral apathy
|
.33
|
.004
|
|
Action pantomimes
|
BDI-II - Depression
|
.51
|
.003
|
|
Meaningless gestures
|
AMI – Total apathy
|
.30
|
.033
|
|
|
AMI – Social apathy
|
.18
|
.035
|
Object perception
|
VOSP - Fragmented letters
|
AMI – Total apathy
|
.26
|
.029
|
|
VOSP - Object decision
|
ns
|
ns
|
ns
|
Spatial perception
|
VOSP - Number localization
|
ns
|
ns
|
ns
|
|
VOSP - Cubic counting
|
Mania – Goldberg Inventory
|
.27
|
.047
|
|
|
PSS - Stress
|
.24
|
.034
|
Logical reasoning
|
WAIS IV - Puzzle
|
Diabetes
|
.34
|
.021
|
|
WAIS IV - Matrix
|
DES - Dissociation
|
.28
|
.041
|
|
|
Gender
|
.40
|
.002
|
Emotion recognition
|
GERT
|
ERQ – Emotion regulation
|
.33
|
.023
|
|
|
AMI – Emotional apathy
|
.28
|
.011
|
|
|
AMI – Behavioral apathy
|
.16
|
.004
|
|
|
Sniff test (anosmia)
|
.06
|
.008
|
|
|
AMI – Social apathy
|
.02
|
.027
|
Anosognosia
|
Memory dysfunctions
|
BDI-II - Depression
|
.62
|
< .001
|
|
|
ISI - Insomnia
|
.12
|
.038
|
|
|
AMI – Behavioral apathy
|
.09
|
.040
|
|
|
Epworth - Sleepiness
|
.05
|
.033
|
|
Executive functions - Inhibition
|
AMI – Total apathy
|
.40
|
.011
|
|
Executive functions - Flexibility
|
AMI – Behavioral score
|
.28
|
.043
|
|
Executive functions – Working memory
|
Epworth - Sleepiness
|
.38
|
.015
|
|
|
Sniff test (anosmia)
|
.25
|
.015
|
Abbreviations: AMI-behavioral: Apathy Motivation Index – behavioral score [58]; AMI-emotional: Apathy Motivation Index – emotional score [58]; AMI-social: Apathy Motivation Index – social score [58]; AMI-total: Apathy Motivation Index – total score [58]; BDI-II: Beck Depression Inventory-Second Edition [56]; BECLA: Batterie d’Evaluation Cognitive du Langage [45]; DES: Dissociative Experience Scale (Carlson & Putnam, 1986); ECS: Emotion Contagion Scale [64]; ERQ: Emotion Regulation Questionnaire [63]; GERT: Geneva Emotion Recognition Test [49]; Goldberg-Inventory: Goldberg Mania Inventory [60]; GREFEX: Groupe de Réflexion sur l'Evalutation des Fonctions Exécutives [38]; MEM III: Wechsler Memory Scale – Third Edition [39]; NV: no variance; ns: not significant; PCL-5: Posttraumatic Stress Disorder Checklist for DSM-5 [59]; PSS-14: Perceived Stress Scale – 14 items [62]; Rey figure: Rey-Osterrieth Complex Figure test [44]; RL/RI 16: free/cued recall 16 items (RL/RI-16) [42]; SD: standard deviation; STAI-S: State-Trait Anxiety Inventory [57] STAI-T: State-Trait Anxiety Inventory [57]; TAP: Test for Attentional Performance (TAP), Version 2.1 [41]; TMT: Trail Making Test; VOSP: Visual Object and Space Perception battery [47]; WAIS IV: Wechsler Adult Intelligence Scale–Fourth Edition [48].