Association between mental illness and COVID-19 in South Korea: a post-hoc analysis

We are honoured to respond the letters by Hirofumi Hirakawa and colleagues1Hirakawa H Association between mental illness COVID-19 in South Korea.Lancet Psychiatry. 2021; (published online Feb 19.)https://doi.org/10.1016/S2215-0366(20)30539-3Summary Full Text PDF PubMed Scopus (2) Google Scholar Jewel Park colleagues2Park J Rhim HC 19.)https://doi.org/10.1016/S2215-0366(20)30465-XSummary (1) regarding our Article,3Lee SW Yang JM Moon SY et al.Association susceptibility clinical outcomes Korea: a nationwide cohort study.Lancet 2020; 7: 1025-1031Summary (136) which investigated potential association pre-existing positivity for SARS-CoV-2 of Korean cohort. The authors proposed need further analysis stratified subtype psychiatric disorders region residence. colleagues pointed out that study did not have subgroup subtype. Therefore, we post-hoc investigate specific (appendix pp 2–3). used propensity score matching 47 058 individuals without with illness, as previously described.3Lee risk infection was associated 95 (4·1%) 2321 patients who misused alcohol or drugs, compared 1391 (3·0%) any (table, appendix p 6; fully adjusted odds ratio [OR] 1·41, 95% CI 1·14–1·74), but having anxiety stress-related (951 [2·8%] 34 536), mood (707 [2·9%] 24 804), personality (13 [3·3%] 400), eating (four [1·3%] 313).TablePropensity score-matched ORs those testing positive disordersCOVID-19 event number total numberMinimally OR*Minimally age gender.Fully OR†Fully age; gender; residence; history diabetes, cardiovascular disease, cerebrovascular chronic obstructive pulmonary asthma, hypertension, kidney disease; Charlson comorbidity index.Original analysisNo illness1391/47 (3·0%)1 (ref)1 (ref)Any illness1383/47 (2·9%)0·99 (0·92–1·07)1·00 (0·93–1·08)Other illness1023/36 257 (2·8%)0·93 (0·85–1·01)0·94 (0·86–1·02)Severe illness360/10 801 (3·3%)1·11 (0·99–1·23)1·10 (0·99–1·22)Post-hoc (ref)Anxiety disorders951/34 536 (2·8%)0·92 (0·87–1·02)Mood (excluding people psychotic symptoms)707/24 804 (2·9%)0·96 (0·87–1·05)0·97 (0·89–1·06)Alcohol drug misuse95/2321 (4·1%)1·39 (1·12–1·71)‡Significant differences (p<0·05).1·41 (1·14–1·74)‡Significant (p<0·05).Personality disorders13/400 (3·3%)1·09 (0·62–1·90)1·10 (0·63–1·92)Eating disorders4/313 (1·3%)0·41 (0·14–1·13)0·43 (0·16–1·15)Data OR (95% CI) unless specified. OR=odds ratio.* Minimally gender.† Fully index.‡ Significant (p<0·05). Open table new tab Data ratio. Residents Daegu–Gyeongbuk had highest all Korea cases (special pandemic control area) two clusters at hospital.4Lee Yuh WT al.Nationwide results contact tracing individual participant data from an epidemiological survey.JMIR Med Inform. 8e20992Crossref (21) Although matched residence (urban rural area),3Lee suggested main should be interpreted carefully given dataset bias, could led unfavourable COVID-19. other areas severe outcomes. Among 216 418 tested SARS-CoV-2, identified 651 (16·0%) 181 767 (84·0%) areas. both groups 7, 10, 11; cohort, n=15 756; n=78 148; standardised mean difference each <0·08). test (fully 1·05, 0·90–1·23) (0·94, 0·76–1·17) is agreement original 4–6, 9). 7160 3827 (53·4%) 3333 (46·6%) 8, 12, 13; n=1718; n=848; <0·1). Patients high 2·55, 1·59–4·10), similiar (2·66, 1·08–6·57). Our showed region. According analysis, depression more susceptible infection,5Yang Chen W Hu Y al.Pre-pandemic COVID-19: UK Biobank analysis.Lancet Healthy Longev. 1: e69-e79Summary (55) inconsistent results. Mental influenced several environmental factors; thus, strict adjustment were required understand effect disorders. found novel relationships misuse, reported analysis. Biologically, ingestion increases alveolar permeability might facilitate viral entry lung decrease immune defence pathogens, consequently increasing rate infection.6Moss M Burnham EL Chronic abuse, acute respiratory distress syndrome, multiple organ dysfunction.Critical Care Med. 2003; 31: S207-S212Crossref Also, because drugs usually consumed during social activity talking, these highly likely violate rules distancing wearing masks.7Wang QQ Kaelber DC Xu R Volkow ND substance use disorders: analyses electronic health records United States.Mol 26: 30-39Crossref (312) Consistent reports, findings population increased COVID-19, showing screen treat this pandemic.7Wang there infections hospitals some areas, region-stratified similar results, suggesting slightly higher than no independent regional influential factors. This publication has been corrected. corrected version first appeared thelancet.com/psychiatry on February 26, 2021 declare competing interests. SWL JMY contributed equally work. Download .pdf (.8 MB) Help pdf files Supplementary KoreaWe read interest article Seung Won Lee Lancet Psychiatry,1 examined national register Korea. It surprising diagnosis likelihood opposed recent research findings. colleagues2 based Biobank. Full-Text KoreaIn their study, suggest COVID-19.1 classified into urban categories adjust confounding, Daegu (the fourth most populous city Korea) indicates bias remain due uncontrolled confounding result differences. studyDiagnosis SARS-CoV-2. illness. Clinicians treating aware Correction Psychiatry published 19. https://doi.org/10.1016/S2215-0366(21)00043-2Lee SW, JM, SY, al. https://doi.org/10.1016/S2215-0366(21)00043-2—In Correspondence, Namwoo Ki's name Kim. correction made 2021, will printed version.