Comorbidities of Keloid and Hypertrophic Scars Among Participants in UK Biobank

Abstract

Importance: Keloids and hypertrophic scars (excessive scarring) are relatively understudied disfiguring chronic skin conditions with high treatment resistance.

Objective: To evaluate established comorbidities of excessive scarring in European individuals, with comparisons across ethnic groups, and to identify novel comorbidities via a phenome-wide association study (PheWAS).

Design, setting, and participants: This multicenter cross-sectional population-based cohort study used UK Biobank (UKB) data and fitted logistic regression models for testing associations between excessive scarring and a variety of outcomes, including previously studied comorbidities and 1518 systematically defined disease categories. Additional modeling was performed within subgroups of participants defined by self-reported ethnicity (as defined in UK Biobank). Of 502 701 UKB participants, analyses were restricted to 230078 individuals with linked primary care records.

Exposures: Keloid or hypertrophic scar diagnoses.

Main outcomes and measures: Previously studied disease associations (hypertension, uterine leiomyoma, vitamin D deficiency, atopic eczema) and phenotypes defined in the PheWAS Catalog.

Results: Of the 972 people with excessive scarring, there was a higher proportion of female participants compared with the 229 106 controls (65% vs 55%) and a lower proportion of White ethnicity (86% vs 95%); mean (SD) age of the total cohort was 64 (8) years. Associations were identified with hypertension and atopic eczema in models accounting for age, sex, and ethnicity, and the association with atopic eczema (odds ratio [OR], 1.68; 95% CI, 1.36-2.07; P < .001) remained statistically significant after accounting for additional potential confounders. Fully adjusted analyses within ethnic groups revealed associations with hypertension in Black participants (OR, 2.05; 95% CI, 1.13-3.72; P = .02) and with vitamin D deficiency in Asian participants (OR, 2.24; 95% CI, 1.26-3.97; P = .006). The association with uterine leiomyoma was borderline significant in Black women (OR, 1.93; 95% CI, 1.00-3.71; P = .05), whereas the association with atopic eczema was significant in White participants (OR, 1.68; 95% CI, 1.34-2.12; P < .001) and showed a similar trend in Asian (OR, 2.17; 95% CI, 1.01-4.67; P = .048) and Black participants (OR, 1.89; 95% CI, 0.83-4.28; P = .13). The PheWAS identified 110 significant associations across disease systems; of the nondermatological, musculoskeletal disease and pain symptoms were prominent.

Conclusions and relevance: This cross-sectional study validated comorbidities of excessive scarring in UKB with comprehensive coverage of health outcomes. It also documented additional phenome-wide associations that will serve as a reference for future studies to investigate common underlying pathophysiologic mechanisms.

Figure.. Multivariate Logistic Regressions to Estimate the Association of Excessive Scarring Status With the Risk of Each Phecode Diagnosis, Adjusting for Age, Sex, Ethnicity, Smoking Status, Body Mass Index, and Townsend Deprivation Index

Dots represent phecodes, grouped into systemic categories denoted by alternating colors. Statistical significance was set at P < .015/1518 (based on the number of phecodes tested), indicated by the horizontal hatched line. HPV indicates human papillomavirus. An interactive version of this Figure can be found at https://cyu06.gitlab.io/ukb_excessive_scarring_phewas/.