## **Supplementary Online Content**

Kim J-S, Lee G-J, Jeong C-Y, et al. Risk of alopecia areata after COVID-19. *JAMA Dermatol*. Published online January 10, 2024. doi:10.1001/jamadermatol.2023.5559

## **eMethods. eReferences**

This supplementary material has been provided by the authors to give readers additional information about their work.

#### **eMethods.**

#### *Database and type of study*

This dataset consists of a National Health Insurance Service (NHIS) cohort of 8,528,533 individuals (20% of the national population), of whom 580,896 were diagnosed with COVID-19 between 2020 and 2021 (NHIS-2022-1-623). As approximately 97% of South Korea's population (50 million) is registered with the NHIS, the majority of COVID-19 patients are included in this dataset. In South Korea, RT-PCR tests are used to diagnose all COVID-19 cases, which are classified using the International Classification of Diseases, 10th revision (ICD-10) code U071.

#### *Study population with COVID-19 group*

The COVID-19 patients were recruited from October 08, 2020 (the initial date when medical information about Korean COVID-19 patients became available) to September 30, 2021. The exclusion criteria were as follows: (1) patients with confirmed COVID-19 between October 01, 2021 and December 31, 2021, and (2) patients diagnosed with alopecia areata (AA) before COVID-19 confirmation. The main outcome was newly diagnosed AA after COVID-19. The diagnosis of AA was defined as ICD-10 code L63, with its subtypes, patchy defined as "L63.8 and L63.9" and alopecia totalis and alopecia universalis as "L63.0 and L63.1". All patients were followed-up until December 31, 2021. The index date for the COVID-19 group was the date of the COVID-19 infection.

#### *Control (uninfected COVID-19) group and independent variables*

A control group was selected to prevent bias due to confounding variables. We recruited individuals with similar propensities to the study group based on the 1:1 'greedy nearest neighbor' algorithm propensity score (PS) matching, considering 13 independent variables. The demographic variables were classified as sex, age, economic status, and residential area. The comorbidities were classified as follows: hypertension, diabetes mellitus, chronic kidney disease, rheumatoid arthritis, systemic lupus erythematosus, ulcerative colitis, thyroid diseases (Graves' disease and Hashimoto's thyroiditis), vitiligo, and psoriasis. To reduce the immortal time bias, the index date of the control group was set to the same date as the 1:1 matched COVID-19 group. Both groups were followed up until the first diagnosis of AA, death, or end of follow-up, whichever occurred first.

*Positive/Negative outcome controls, subgroup analysis, and sensitivity analysis*  To mitigate the detection bias between COVID-19 and AA, we employed one positive outcome control (telogen effluvium) that has been known to be strongly associated with COVID-19 and four negative outcome controls that are not likely to be affected by COVID-19 (epidermal cyst, benign skin tumor, tympanic membrane perforation, and melanocytic nevus). 1-3 The definitions for each condition, based on the ICD-10 codes, are as follows: telogen effluvium (L650), epidermal cyst (L720), benign skin tumor (D23), tympanic membrane perforation (H72), and melanocytic nevus (D22).

We conducted subgroup analyses to determine the risk of AA in COVID-19 patients based on age, sex, and clinical subtype of AA. To prove the robustness of the results, we performed sensitivity analysis using the matching method, follow-up period, diagnostic criteria (number of encounters and type of diagnosing physician), and competing risk of death.

## *Statistical analysis*

We evaluated the balance of baseline characteristics in PS-matched populations using standardized mean differences (SMD). SMDs were calculated as the difference between groups in the number of standard deviations and is a more meaningful measure than *P* values from t-tests for large samples.<sup>4</sup> Any variable with a SMD <0.1 was considered well-matched. Hazard ratios (HR) and 95% confidence intervals were calculated using the Cox proportional hazards model that considers time variables. The adjusted HR was calculated through a multivariate analysis that considered all independent variables used in PS matching, and the unadjusted HR was calculated through a univariate analysis that considered only the exposure variables. The SAS statistical software (version 9.4; SAS Institute, Cary, NC, USA) and R 4.0.3 statistical program (R Foundation for Statistical Computing, Vienna, Austria) were used. Significance was set at 2 sided *P*<0.05.

# **eReferences**

- 1. Ju HJ, Lee JY, Han JH, Lee JH, Bae JM, Lee S. Risk of autoimmune skin and connective tissue disorders after mRNA-based COVID-19 vaccination. *J Am Acad Dermatol.* 2023;89(4):685-693.
- 2. Xie Y, Xu E, Bowe B, Al-Aly Z. Long-term cardiovascular outcomes of COVID-19. *Nat Med.* 2022;28(3):583-590.
- 3. Lipsitch M, Tchetgen Tchetgen E, Cohen T. Negative controls: a tool for detecting confounding and bias in observational studies. *Epidemiology.*  2010;21(3):383-388.
- 4. Roumie CL, Chipman J, Min JY, et al. Association of Treatment With Metformin vs Sulfonylurea With Major Adverse Cardiovascular Events Among Patients With Diabetes and Reduced Kidney Function. *JAMA.*  2019;322(12):1167-1177.
- © 2024 American Medical Association. All rights reserved.

JAMA Dermatology

<jats:p>This cohort study examines the incidence, prevalence, and risk of alopecia areata after COVID-19.</jats:p>

Risk of autoimmune diseases in patients with COVID-19: a retrospective cohort study.

EClinicalMedicine

Association between alopecia areata and COVID-19: a systematic review.

JAAD Int

Possible associations between alopecia areata and COVID-19 vaccination and infection.

Aesthet Surg J

Short-term stress-related increasing cases of alopecia areata during the COVID-19 pandemic.

J Dermatolog Treat

Lack of evidence of COVID-19 being a risk factor of alopecia areata: results of a national cohort study in South Korea.

Front Med (Lausanne)

Incidence and prevalence of alopecia areata according to subtype: a nationwide, population-based study in South Korea (2006-2015).

Br J Dermatol

This cohort study examines the incidence, prevalence, and risk of alopecia areata after COVID-19.

Risk of Alopecia Areata After COVID-19.

Risk of Alopecia Areata After COVID-19

Risk of Alopecia Areata After COVID-19

Clinical Summary

What was studied

Related Questions