Obsessive-compulsive disorder and related symptoms amidst the COVID-19 outbreak: Results from the COLLATE project.

Studying obsessive–compulsive disorder (OCD) symptoms amidst the COVID‐19 pandemic is important for three major reasons. First, health anxiety can be prevalent in OCD, and exacerbated in the prevailing climate. Second, OCD can develop in response to traumatic events. Third, revised health guidelines have likely normalized certain compulsions (e.g., repeated handwashing). These behaviors may extend to the general population, including fears about becoming ill or infecting others with COVID‐19. Few studies have explored OCD amidst previous outbreaks of pandemics. Retrospective analysis of an electronic mental health database concluded that patients with OCD were overrepresented in those expressing moderate to severe swine flu concerns. Another study demonstrated that OCD symptoms significantly predicted swine flu fears in a student cohort. Pertaining to COVID‐19, an online survey found higher endorsement of OCD symptoms in medical, relative to non‐medical, workers. Having an organic disease was an overall risk factor for OCD symptoms, with being female, rural living, and potential COVID‐19 exposure as added risk factors for medical workers.

Our current study aimed to: (i) document COVID‐19 concerns in an OCD group relative to a matched general population (GNP) group; (ii) compare group members’ mental health status, including negative emotions; and (iii) explore endorsement of OCD‐related behaviors and associated predictors. We hypothesized that the OCD group would assign higher rankings for concerns related to becoming ill with COVID‐19, and be significantly more depressed, anxious, and stressed relative to the GNP group. These analyses utilized cross‐sectional data from Waves 1 (April) and 2 (May) of our COVID‐19 and You: Mental Health in Australia Now Survey (COLLATE). The project design has been published elsewhere. A description of methodology and data analyses, including participant matching (and psychiatric comorbidity in the OCD group in Table A), is summarized in Appendix A.

Table 1 shows the top 10 COVID‐19 concerns by group in Wave 1. The top two concerns were identical across groups and related to ‘a loved one dying or being infected with COVID‐19.’ Notably, ‘oneself dying or being infected with COVID‐19’ was ranked lower in the OCD group relative to the matched GNP sample. The OCD group did, however, place ‘implications for health and well‐being of self’ above that of their ‘family and loved ones,’ with the opposite pattern found for the GNP group. Group‐wise comparisons of mental health in Wave 1 revealed that the OCD group reported significantly increased depression, anxiety, and stress as well as poorer quality of life relative to the GNP group, with large effect sizes. Severe depression and anxiety, and moderate stress were more likely in the OCD group, with mild depression more likely in the GNP group (see Appendix B and Table B, which also characterizes groups by sociodemographic information and COVID‐19‐related lifestyle changes).

Table 1

Top 10 ranked COVID‐19‐related concerns by group (Wave 1)

	OCD group (n = 66)				GNP group (n = 198)
	Mean rank	Mean rating ± SD	n	%	Mean rank	Mean rating ± SD	n	%
Loved one dying from COVID‐19	1	2.35 ± 2.51	52	78.8	1	2.59 ± 2.44	162	81.8
Loved one being infected with COVID‐19	2	3.96 ± 2.45	51	77.3	2	3.56 ± 2.36	152	76.8
Not being able to attend regular place of worship	3	4.00 ± 2.94	4	6.1	23	7.64 ± 2.58	11	5.6
Dying of COVID‐19 myself	4	4.41 ± 2.79	37	56.1	3	4.67 ± 2.80	78	39.4
Implications for health and well‐being of self	5	4.54 ± 2.24	48	72.7	8	5.45 ± 2.57	119	60.1
Other	6	4.75 ± 3.28	8	12.1	9	5.69 ± 2.96	16	8.1
Implications for health and well‐being of family and loved ones	7	4.83 ± 2.04	52	78.8	4	4.68 ± 2.31	164	82.8
Risk of unemployment or reduced employment	8	4.87 ± 2.68	30	45.5	5	4.72 ± 2.67	122	61.6
Social isolation and social distancing	9	4.90 ± 2.64	31	47.0	11	6.04 ± 2.58	130	65.7
Oneself being infected with COVID‐19	10	5.44 ± 2.39	39	59.1	6	5.13 ± 2.75	118	59.6
Balancing work and caring for children or dependents	10	5.44 ± 3.28	9	13.6	7	5.30 ± 2.90	47	23.7
Implications for health and well‐being of society	12	5.96 ± 2.56	46	69.7	12	6.31 ± 2.70	131	66.2
Media coverage of the pandemic	13	6.50 ± 2.78	12	18.2	19	7.00 ± 2.67	33	16.7
Domestic violence	14	6.54 ± 2.07	13	19.7	20	7.04 ± 2.24	26	13.1
Access to appropriate medical care	15	6.59 ± 2.58	37	56.1	17	6.80 ± 2.39	85	42.9
Availability of food and medicines	16	6.95 ± 2.37	39	59.1	18	6.88 ± 2.16	95	48.0
The rapidly changing landscape	17	6.96 ± 2.43	26	39.4	13	6.53 ± 2.52	68	34.3
Government communication of key messages	18	7.05 ± 2.92	19	28.8	16	6.79 ± 1.77	34	17.2
Adapting to working from home	19	7.18 ± 2.18	11	16.7	21	7.46 ± 2.59	54	27.3
The Australian economy	20	7.20 ± 2.47	35	53.0	14	6.55 ± 2.59	105	53.0
Personal finances	21	7.29 ± 2.77	28	42.4	10	5.84 ± 3.03	95	48.0
The world economy	22	7.74 ± 2.51	19	28.8	22	7.57 ± 2.36	74	37.4
Travel restrictions	23	8.00 ± 2.11	14	21.2	15	6.64 ± 2.38	61	30.8

For current concerns relating to COVID‐19, rankings from 1 (greatest concern) to 5 (least concern) were computed (0 was assigned to options that were not endorsed).

GNP, general population; OCD, obsessive–compulsive disorder.

When these analyses were rerun for Wave 2, the top 10 themes remained largely similar, with a few exceptions (Appendix C and Table C). However, a trend towards poorer mental health was observed: extremely severe depression, anxiety, and stress were more likely in the OCD group, whereas moderate depression, mild anxiety, and moderate stress were more likely in the GNP group (Appendix D and Table D1). Notably, washing, checking, and obsession scores did not appear significantly elevated (relative to the original validation study ; Table D2). Regression analysis revealed that distal, F(5, 612) = 10.3, P < 0.001, r 2 = 0.078, and proximal, F(14, 612) = 35.9, P < 0.001, r 2 = 0.379, factors significantly predicted OCD symptoms across the entire Wave 2 cohort (Appendix E and Table E). Only age, education, and having an existing medical condition were significant predictors in Block 1 (distal), but were no longer significant in Block 2 (proximal); these were mediated by depression, anxiety, and stress, which served as unique predictors for OCD symptoms.

Our hypothesis was partly supported in that though the OCD group did not assign higher rankings for concerns related to becoming ill from COVID‐19, significantly increased negative emotions were reported relative to the GNP group. When coupled with the finding that negative emotions were significantly associated with OCD symptoms, this suggests that the mental health of persons with OCD may be more adversely affected in the longer run. Study limitations included relying on self‐reported OCD, and the inability to perform statistical comparisons between our two data waves (owing to unequal group sizes and few repeat respondents). Previous pandemic research has suggested delayed and prolonged mental health impacts, with time lags from pandemic onset to manifestation of psychopathology. Applying this reasoning, we infer that if effective interventions to address elevated negative emotions are not enacted in a timely manner, OCD symptoms may significantly worsen as the outbreak continues to unfold. This is the challenge that existing mental health‐care systems need to address.

Disclosure statement

The authors have declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.

Appendix A. Methods.

Click here for additional data file.

Appendix B. Sociodemographic information and mental health status by group.

Click here for additional data file.

Appendix C. Top 10 ranked COVID‐19 concerns by group.

Click here for additional data file.

Appendix D. Negative emotions and obsessive–compulsive symptoms.

Click here for additional data file.

Appendix E. Regression analyses predicting OCD symptoms.

Click here for additional data file.