Vitiligo in Underrepresented Communities: An All of Us Database Analysis.

To the Editor: Vitiligo affects people of all ages, genders, and races.1Gandhi K. Ezzedine K. Anastassopoulos K.P. et al.Prevalence of vitiligo among adults in the United States.JAMA Dermatol. 2022; 158: 43-50https://doi.org/10.1001/jamadermatol.2021.4724Crossref PubMed Scopus (11) Google Scholar However, we are unaware of specific attempts to quantify vitiligo among underrepresented racial, sexual, and socioeconomic groups, who may experience increased disease burden.2Linthorst Homan M.W. Spuls P.I. de Korte J. Bos J.D. Sprangers M.A. van der Veen J.P.W. The burden of vitiligo: patient characteristics associated with quality of life.J Am Acad Dermatol. 2009; 61: 411-420https://doi.org/10.1016/j.jaad.2009.03.022Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar,3Takeshita J. Identifying disparities in dermatology: the importance of measuring differences that matter.JAMA Dermatol. 2018; 154: 1251-1253https://doi.org/10.1001/jamadermatol.2018.2938Crossref PubMed Scopus (13) Google Scholar We aimed to describe patterns of vitiligo diagnoses in groups historically underrepresented in medical research. We analyzed the All of Us Research Program, a prospective US cohort study that prioritizes the inclusion of these participants.4Denny J.C. Rutter J.L. Goldstein D.B. et al.All of Us Research Program InvestigatorsThe “All of Us” research program.N Engl J Med. 2019; 381: 668-676https://doi.org/10.1056/NEJMsr1809937Crossref PubMed Scopus (485) Google Scholar Vitiligo and other immune-mediated diagnoses were extracted from electronic health record data and sociodemographic characteristics from survey data. Characteristics of participants with and without vitiligo were compared using Pearson’s χ2 or Fisher’s exact tests, as appropriate. Adjusted odds ratios (ORs) for a vitiligo diagnosis were calculated using multivariable logistic regression. At the time of analysis, the database included 368,190 participants, of whom 284,004 (77.1%) had electronic health record data. A total of 926 (0.33%) participants had a vitiligo diagnosis (95% CI 0.31-0.35). Table I summarizes sociodemographic characteristics and comorbidities of participants. Black (OR 1.71, 95% CI 1.40-2.08), Hispanic (OR 2.25, 95% CI 1.87-2.71), and “Other” non-White participants had a higher adjusted odds of a vitiligo diagnosis compared with Whites (OR 1.60, 95% CI 1.22-2.08). Higher odds were also observed in participants aged >65 years (OR 1.87, 95% CI 1.62-2.15). Lower odds were observed in those without health insurance (OR 0.64, 95% CI 0.43-0.91) (Table II).Table IDistribution of sociodemographic characteristics and select comorbidities among patients with and without vitiligo in the All of Us electronic health record database (N = 284,004)CharacteristicVitiligo, No. (%), (n = 926)Non-vitiligo, No. (%) (n = 283,078)P value†P values based on χ2 test or Fisher’s exact test, as appropriate.Rate of vitiligo,% (95% CI)Race/Ethnicity<.001 White428 (46.2)155,333 (54.9)0.27 (0.25-0.30) Black or African American176 (19.0)51,440 (18.2)0.34 (0.29-0.40) Hispanic232 (25.1)49,104 (17.3)0.47 (0.41-0.53) Other∗Includes Asian, mixed-race, and other.64 (6.9)17,262 (6.1)0.37 (0.29-0.47) Unknown26 (2.8)9939 (3.5)0.26 (0.18-0.38)Age (y)<.001 ≤65581 (62.7)213,525 (75.4)0.27 (0.25-0.29) >65345 (37.3)69,553 (24.6)0.49 (0.44-0.55)Gender.045 Man334 (36.1)102,219 (36.1)0.33 (0.29-0.36) Woman572 (61.8)170,436 (60.2)0.33 (0.31-0.36) Other≤20‡Per All of Us data dissemination guidelines, values less than or equal to 20 are reported as "≤20". (≤2.2)10,423 (3.7)≤0.19LGBTQ identity<.001 No839 (90.6)245,836 (86.8)0.34 (0.32-0.36) Yes87 (9.4)37,242 (13.2)0.23 (0.19-0.29)Educational attainment.010 Completed college423 (45.7)125,778 (44.4)0.34 (0.30-0.37) Completed high school383 (41.4)122,222 (43.2)0.31 (0.28-0.35) Less than high school99 (10.7)24,372 (8.6)0.40 (0.33-0.49) Unknown21 (2.3)10,706 (3.8)0.20 (0.13-0.30)Annual household income ($).853 ≥25,000504 (54.4)155,964 (55.1)0.32 (0.30-0.35) <25,000228 (24.6)69,873 (24.7)0.33 (0.29-0.37) Unknown194 (21.0)57,241 (20.2)0.34 (0.29-0.39)Health insurance<.001 Yes873 (94.3)255,485 (90.3)0.34 (0.32-0.36) No30 (3.2)16,206 (5.7)0.18 (0.13-0.26) Unknown23 (2.5)11,387 (4.0)0.20 (0.13-0.30)Disability status.030 Without disability799 (86.3)251,934 (89.0)0.32 (0.29-0.34) With disability109 (11.8)26,974 (9.5)0.40 (0.33-0.49) Unknown≤20 (≤2.2)4170 (1.5)≤0.48Comorbidities Psoriasis75 (8.1)6272 (2.2)<.0011.18 (0.94-1.48) Atopic dermatitis68 (7.3)5384 (1.9)<.0011.25 (0.98-1.58) Localized scleroderma33 (3.6)1497 (0.5)<.0012.16 (1.54-3.02) Alopecia areata25 (2.7)835 (0.3)<.0012.91 (1.97-4.27) Cutaneous lupus≤20 (≤2.2)800 (0.3).006≤2.50 Rheumatoid arthritis73 (7.9)6775 (2.4)<.0011.07 (0.85-1.34) Type I diabetes mellitus65 (7.0)5484 (1.9)<.0011.17 (0.92-1.49) Thyroid conditions53 (5.7)4151 (1.5)<.0011.26 (0.96-1.65) Inflammatory bowel disease29 (3.1)4534 (1.6)<.0010.64 (0.44-0.91) Sjögren syndrome37 (4.0)3274 (1.2)<.0011.12 (0.81-1.54) Systemic lupus erythematosus39 (4.2)3238 (1.1)<.0011.19 (0.87-1.62) Addison disease30 (3.2)1990 (0.7)<.0011.49 (1.04-2.12) Celiac disease≤20 (≤2.2)1587 (0.6)<.001≤1.26 Pernicious anemia≤20 (≤2.2)830 (0.3)<.001≤2.41 Systemic scleroderma≤20 (≤2.2)655 (0.2)<.00113.6 Dermatomyositis≤20 (≤2.2)260 (0.1).211≤7.69LGBTQ, Lesbian, gay, bisexual, transgender, or queer.∗ Includes Asian, mixed-race, and other.† P values based on χ2 test or Fisher’s exact test, as appropriate.‡ Per All of Us data dissemination guidelines, values less than or equal to 20 are reported as "≤20". Open table in a new tab Table IIAge-adjusted and multivariable-adjusted odds ratios and 95% CI for vitiligo according to sociodemographic characteristic (N = 284,004)No. of cases/totalOR (95% CI)Age-adjustedMultivariable-adjusted†Adjusted for all listed sociodemographic characteristics in addition to select comorbidities (psoriasis, atopic dermatitis, localized scleroderma, alopecia areata, cutaneous lupus, rheumatoid arthritis, type I diabetes mellitus, thyroid conditions, inflammatory bowel disease, Sjögren syndrome, systemic lupus erythematosus, Addison disease, celiac disease, pernicious anemia, systemic scleroderma, and dermatomyositis).Race/Ethnicity White428/155,333refref Black or African American176/51,4401.47 (1.23-1.75)1.71 (1.40-2.08) Hispanic232/49,1042.32 (1.96-2.73)2.25 (1.87-2.71) Other∗Includes Asian, mixed-race, and other.64/17,2621.80 (1.37-2.33)1.60 (1.22-2.08)Age (y) ≤65581/213,525refref >65345/69,5531.82 (1.59-2.08)1.87 (1.62-2.15)Gender Man334/102,219refref Woman572/170,4361.12 (0.98-1.29)0.91 (0.79-1.04) Other<20‡Per All of Us data dissemination guidelines, values less than or equal to 20 are reported as "≤20"./10,4230.63 (0.39-0.97)0.79 (0.45-1.30)LGBTQ identity No839/245,836refref Yes87/37,2420.78 (0.62-0.97)0.82 (0.64-1.03)Educational attainment Completed college423/125,778refref Completed high school383/122,2221.01 (0.88-1.16)0.88 (0.76-1.03) Less than high school99/24,3721.32 (1.06-1.64)0.96 (0.74-1.24)Annual household income ($) ≥25,000504/155,964refref <25,000228/69,8731.13 (0.97-1.33)0.95 (0.79-1.14)Health insurance Yes873/255,485refref No30/16,2060.66 (0.45-0.94)0.64 (0.43-0.91)Disability status Without disability799/251,934refref With disability109/26,9741.25 (1.01-1.51)1.04 (0.84-1.29)LGBTQ, Lesbian, gay, bisexual, transgender, or queer; OR, odds ratio; ref, reference.∗ Includes Asian, mixed-race, and other.† Adjusted for all listed sociodemographic characteristics in addition to select comorbidities (psoriasis, atopic dermatitis, localized scleroderma, alopecia areata, cutaneous lupus, rheumatoid arthritis, type I diabetes mellitus, thyroid conditions, inflammatory bowel disease, Sjögren syndrome, systemic lupus erythematosus, Addison disease, celiac disease, pernicious anemia, systemic scleroderma, and dermatomyositis).‡ Per All of Us data dissemination guidelines, values less than or equal to 20 are reported as "≤20". Open table in a new tab LGBTQ, Lesbian, gay, bisexual, transgender, or queer. LGBTQ, Lesbian, gay, bisexual, transgender, or queer; OR, odds ratio; ref, reference. The 0.33% rate of clinician-diagnosed vitiligo in this study is substantially lower than the 0.76% rate recently obtained from a nationally-representative sample.1Gandhi K. Ezzedine K. Anastassopoulos K.P. et al.Prevalence of vitiligo among adults in the United States.JAMA Dermatol. 2022; 158: 43-50https://doi.org/10.1001/jamadermatol.2021.4724Crossref PubMed Scopus (11) Google Scholar This discrepancy is likely attributable to the underdiagnosis of vitiligo in the All of Us database, whose over-sampled, underrepresented populations may face challenges in accessing dermatologic care.3Takeshita J. Identifying disparities in dermatology: the importance of measuring differences that matter.JAMA Dermatol. 2018; 154: 1251-1253https://doi.org/10.1001/jamadermatol.2018.2938Crossref PubMed Scopus (13) Google Scholar Consistent with this hypothesis, we observed lower rates of diagnosed vitiligo among uninsured participants. Increased rates of vitiligo diagnosis in Black, Hispanic, and other people of color (POC) have not been previously described in detail. Possible explanations include increased psychosocial burden in POC leading to medical presentation,2Linthorst Homan M.W. Spuls P.I. de Korte J. Bos J.D. Sprangers M.A. van der Veen J.P.W. The burden of vitiligo: patient characteristics associated with quality of life.J Am Acad Dermatol. 2009; 61: 411-420https://doi.org/10.1016/j.jaad.2009.03.022Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar easier detection of depigmentation in darker skin, or some pathogenic mechanism by which vitiligo occurs more frequently in POC. These findings warrant future investigation into whether they are attributable to differences in clinical detection and diagnosis versus actual incidence of vitiligo in POC. Although vitiligo onset tends to be earlier in life,1Gandhi K. Ezzedine K. Anastassopoulos K.P. et al.Prevalence of vitiligo among adults in the United States.JAMA Dermatol. 2022; 158: 43-50https://doi.org/10.1001/jamadermatol.2021.4724Crossref PubMed Scopus (11) Google Scholar we observed increased rates in individuals aged >65 years. This likely reflects higher proportions of clinician-diagnosed to undiagnosed vitiligo in this group rather than true increased disease prevalence.1Gandhi K. Ezzedine K. Anastassopoulos K.P. et al.Prevalence of vitiligo among adults in the United States.JAMA Dermatol. 2022; 158: 43-50https://doi.org/10.1001/jamadermatol.2021.4724Crossref PubMed Scopus (11) Google Scholar This pattern may be accentuated in underrepresented communities by diagnostic delays related to difficulties in accessing healthcare, some of which are potentially alleviated upon Medicare eligibility.5Patel D.C. He H. Berry M.F. et al.Cancer diagnoses and survival rise as 65-year-olds become Medicare-eligible.Cancer. 2021; 127: 2302-2310https://doi.org/10.1002/cncr.33498Crossref PubMed Scopus (5) Google Scholar Limitations of the present study include reliance on electronic health record data. However, the diversity of All of Us cohort is an important strength. Our results suggest increased rates of vitiligo diagnosis in POC, who are also disproportionately affected by barriers to healthcare access. Such barriers may lead to delayed diagnosis, suggested by the increased rate of vitiligo we observed in participants aged >65 years. Clinicians should be mindful of possible underdiagnosis and delayed diagnosis of vitiligo in uninsured and underrepresented US communities. None disclosed. The All of Us Research Program is supported by the National Institutes of Health, Office of the Director: Regional Medical Centers: 1 OT2 OD026549; 1 OT2 OD026554; 1 OT2 OD026557; 1 OT2 OD026556; 1 OT2 OD026550; 1 OT2 OD 026552; 1 OT2 OD026553; 1 OT2 OD026548; 1 OT2 OD026551; 1 OT2 OD026555; IAA #: AOD 16037; Federally Qualified Health Centers: HHSN 263201600085U; Data and Research Center: 5 U2C OD023196; Biobank: 1 U24 OD023121; The Participant Center: U24 OD023176; Participant Technology Systems Center: 1 U24 OD023163; Communications and Engagement: 3 OT2 OD023205; 3 OT2 OD023206; and Community Partners: 1 OT2 OD025277; 3 OT2 OD025315; 1 OT2 OD025337; 1 OT2 OD025276. In addition, the All of Us Research Program would not be possible without the partnership of its participants.