Comorbid Bronchial Asthma, Atopic Dermatitis and Hashimoto's Thyroiditis Are Risk Factors for Early‐Onset, Severe and Prolonged Alopecia Areata

To the Editor,

Alopecia areata (AA) is a common autoimmune disease characterised by sudden, non‐scarring scalp/body hair loss [1]. The AA‐clinical course is unpredictable, with wide inter‐individual variability. This may involve spontaneous remission of the initial hair loss episode without recurrence, complete lack of hair re‐growth or a chronic relapsing course involving episodes of varying extent and duration. Hair loss extent ranges from isolated and circumscribed bald patches (patchy AA), to a complete loss of the hair from the scalp (alopecia totalis; AT) or body (alopecia universalis; AU). The uncertain prognosis often has a major adverse psychological impact [2].

Comorbidity with chronic inflammatory disorders (CIDs) is common in AA and could be of relevance to prognosis and clinical management [3]. However, few studies have assessed associations between AA‐clinical features and CID comorbidity status. Furthermore, available studies show wide methodological variation, and many have involved small to moderate cohorts assessed for only one AA‐clinical feature and/or comorbidity profile. The clinically relevant hypothesis that distinct comorbidity profiles indicate distinct AA subtypes, as characterised by specific clinical features, thus warrants further investigation [4].

Here, we performed a comprehensive analysis of associations between comorbid CIDs and selected AA‐clinical features of relevance to prognosis using self‐report data from 2657 mainly Central European AA patients recruited via outpatient clinics, dermatology practitioners or AA self‐support groups in Germany and Belgium (Table S1) [5]. All participants, or legal guardians, provided written informed consent. Early‐onset AA was defined as AA age‐of‐onset ≤ 20 years. Severe AA was defined as a history of AT, AT/AU or AU during the most severe episode ever experienced. Prolonged AA was defined as a history of ≥ 1 AA episode of > 2 years' duration.

Overall, 53.7% of the AA cohort reported ≥ 1 comorbid CID of any type; 44.5% reported ≥ 1 of the atopic CIDs atopic dermatitis (AD; 26.7%), bronchial asthma (13.4%) and/or rhinitis (26.7%); and 17.4% reported ≥ 1 comorbid non‐atopic CID. The most frequent comorbid non‐atopic CIDs were Hashimoto's thyroiditis (6.1%), vitiligo (4.6%), psoriasis (2.7%) and rheumatoid arthritis (1.7%) (Data S1).

Table 1 highlights four key findings. Patients with comorbid AD, bronchial asthma or Hashimoto's thyroiditis were significantly more likely to report early‐onset, severe and prolonged AA compared to patients with no comorbid CIDs (i.e., AA‐only). Patients with comorbid rhinitis or vitiligo showed a significantly increased risk for prolonged AA. Comorbid bronchial asthma was associated with a higher risk for early‐onset, severe or prolonged AA than comorbid AD or rhinitis. CID comorbidity status showed a more pronounced association with AA duration than with age‐of‐onset or severity.

TABLE 1.

Association between comorbidity status for atopic or non‐atopic chronic inflammatory disorders and AA age‐of‐onset, severity and duration.

AA‐patient comorbidity subgroups (N = number of patients) a	Early‐onset AA (%) b	p, OR; 95 CI b	Severe AA (%) c	p, OR; 95% CI c	Prolonged AA (%) d	p, OR; 95% CI d
AA‐only (1229)	36.9%	—	49.0%	—	51.8%	—
AA + ≥ 1 CID of any type (1428)	44.5%	0.014; 1.246; 1.045–1.487	52.2%	0.125; 1.128; 0.967–1.316	57.5%	0.008, 1.255; 1.062–1.82
AA + ≥ 1 atopic CID and no non‐atopic CID (966)	47.3%	0.017, 1.269, 1.044–1.543	51.8%	0.256; 1.104; 0.931–1.310	56.4%	0.039, 1.217; 1.010–1.466
AA + ≥ 1 non‐atopic CID and no atopic CID (246)	37.4%	0.295; 1.186, 0.861–1.634	54.1%	0.267; 1.170; 0.887–1.543	58.5%	0.106, 1.272; 0.950–1.702
AA + ≥ 1 atopic and ≥ 1 non‐atopic CID (216)	39.8%	0.407, 1.151, 0.825–1.607	54.2%	0.237; 1.192; 0.891–1.596	60.9%	0.031, 1.411; 1.033–1.929
AA + AD (w/ or w/o other CIDs) (709)	49.5%	0.031, 1.270; 1.022–1.579	55.9%	0.005; 1.309; 1.083–1.582	58.6%	0.006, 1.335; 1.086–1.643
AA + AD (w/ or w/o other atopic CIDs) (587)	51.3%	0.019; 1.319; 1.046–1.663	54.9%	0.025; 1.259; 1.029–1.540	57.5%	0.030, 1.279; 1.025–1.598
AA + rhinitis (w/ or w/o other CIDs) (709)	44.5%	0.047; 1.238;1.003–1.528	51.6%	0.331; 1.097; 0.910–1.321	58.4%	0.008, 1.313; 1.073–1.606
AA + rhinitis (w/ or w/o other atopic CIDs) (581)	45.6%	0.051; 1.251; 0.999–1.566	50.8%	0.525; 1.067; 0.874–1.302	58.0%	0.017, 1.302; 1.049–1.616
AA + asthma e (w/ or w/o other CIDs) (357)	47.6%	0.007; 1.446; 1.105–1.892	58.5%	0.002; 1.470; 1.156–1.869	61.6%	0.002, 1.507; 1.160–1.957
AA + asthma e (w/ or w/o other atopic CIDs) (285)	51.6%	0.001; 1.606; 1.199–2.151	57.9%	0.006; 1.444; 1.111–1.878	62.8%	0.001, 1.604; 1.200–2.143
AA + AITD f (w/ or w/o other CIDs) (186)	44.6%	0.037; 1.452; 1.022–2.062	61.3%	0.007; 1.556; 1.131–2.141	68.2%	< 0.001, 1.911; 1.356–2.694
AA + AITD f (w/ or w/o other non‐atopic CIDs) (97)	44.3%	0.044; 1.628; 1.014–2.613	64.9%	0.008; 1.798; 1.163–2.778	65.2%	0.027, 1.661; 1.061–2.603
AA + HT (w/ or w/o other CIDs) (163)		0.031; 1.502; 1.038–2.173	60.7%	0.014; 1.524; 1.088–2.135	68.2%	< 0.001, 1.911; 1.328–2.748
AA + HT (w/ or w/o other non‐atopic CIDs) (83)	49.4%	0.015; 1.879; 1.132–3.119	65.1%	0.012; 1.818; 1.139–2.902	64.6%	0.049, 1.618; 1.002–2.614
AA + vitiligo (w/ or w/o other CIDs) (122)	38,5%	0.260; 1.285; 0.831–1.988	50.8%	0.776; 1.056; 0.726–1.534	63.2%	0.031, 1.578; 1.043–2.388
AA + vitiligo (w/ or w/o other non‐atopic CIDs) (51)	41.2%	0.076; 1.807; 0.941–3.473	47.1%	0.651; 0.878; 0.499–1.544	— a	— a
AA + psoriasis (w/ or w/o other CIDs) (71)	28.2%	0.073; 0.584; 0.324–1.052	39.4%	0.125; 0.681; 0.417–1.112	38.5%	0.039, 0.580; 0.347–0.972

Note: Analyses of the differences in AA age‐of‐onset, severity and episode duration between subgroups of AA patients with (specific) comorbid chronic inflammatory disorders (CIDs) and AA patients with no comorbid CIDs (i.e., AA‐only) using binary logistic regression, as adjusted for age‐at‐recruitment and sex. To ensure adequate statistical power, the analyses were restricted to subgroups containing at least 50 individuals. Significant results are given in bold.

Abbreviations: AA, alopecia areata; AD, atopic dermatitis; AITD, autoimmune thyroid disease; AT, alopecia totalis; AU, alopecia universalis; CI, confidence interval; CID, chronic inflammatory disorder; HT, Hashimoto's thyroiditis; OR, odds ratio; w/, with; w/o, without.

^a The patients were subgrouped according to self‐reported comorbidity status for CIDs. CIDs were divided into two main types: atopic CIDs and non‐atopic CIDs. The atopic CIDs comprised atopic dermatitis (AD), rhinitis and bronchial asthma. The non‐atopic CIDs encompassed CIDs with autoimmune, autoinflammatory or unclear aetiologies. The self‐reported comorbid non‐atopic CIDs in the present cohort are listed in the Data S1. The generalised comorbidity subgroups included AA patients with: (i) ≥ 1 comorbid CID of any type; (ii) ≥ 1 comorbid atopic CID and no non‐atopic CID; (iii) ≥ 1 comorbid non‐atopic CID and no atopic CID; and (iv) ≥ 1 comorbid atopic and ≥ 1 comorbid non‐atopic CID. Comorbidity subgroups for individual diseases (e.g., AA + AD) were defined by the following two means. First all AA patients with the respective disease, regardless of whether they reported additional comorbidities of any type, were represented by the phrase ‘w/ or w/o other CIDs’. Second, all AA patients with the respective disease with or without other CIDs of the same type and no reported CIDs of the other type were represented by the phrases ‘(w/ or w/o other atopic CIDs)’ or ‘(w/ or w/o other non‐atopic CIDs)’. For each comorbidity subgroup, the stated number of patients represents the total number of patients who had responded to at least one of the questionnaire items concerning age‐of‐onset, severity and duration of AA. Therefore, the proportion of patients given for each specific clinical feature do not necessarily have the same denominator. No binary logistic regression analysis was performed to compare AA episode duration between patients in ‘AA + vitiligo (w/ or w/o other non‐atopic CIDs)’ and those in ‘AA‐only’, since the number of patients who had responded to the questionnaire item on AA duration in the former subgroup was below 50.

^b Binary logistic regression analysis comparing the proportion of early (≤ 20 years) and late (> 20 years) onset AA cases in the respective comorbidity subgroup to those in patients with AA‐only, the latter being indicated in the first row.

^c Binary logistic regression analysis comparing the proportion of mild (patchy AA) and severe AA (AT, AT/AU, AU) cases in the respective comorbidity subgroup to those in patients with AA‐only, the latter being indicated in the first row.

^d Binary logistic regression analysis comparing the proportion of AA cases with and without prolonged AA (i.e., a positive and a negative history of ≥ 1 AA episodes of > 2 years' duration, respectively) in the respective comorbidity subgroup to those in patients with AA‐only, the latter being indicated in the first row.

^e Asthma = bronchial asthma.

^f Patients with autoimmune thyroid disease had Hashimoto's thyroiditis (HT) or Graves' disease; individual associations for Graves' disease were not assessed, since the number of individuals with comorbid Graves' disease was below the threshold of 50.

The number of self‐reported atopic comorbidities was significantly higher in early‐onset, severe and prolonged AA compared to late‐onset, mild and non‐prolonged disease, respectively. The odds for early‐onset, severe and prolonged AA increased by a factor of 1.179, 1.130 and 1.202, respectively, per additional atopic comorbidity (Table S2). Compared to patients with AA‐only, mean age‐of‐onset for AA was almost 10 years earlier in patients with AD + bronchial asthma + rhinitis, and around 5 years earlier in patients with one or two comorbid atopic diseases (Table 2).

TABLE 2.

Linear regression analysis of AA age‐of‐onset and number of comorbid chronic inflammatory disorders.

	Total number of comorbidities	Number of individuals	Mean age‐of‐AA‐onset ± SD	Regression coefficient	p	95% CI
AA‐only	0	1229	29.5 ± 17.7	—	—	—
AA + ≥ 1 CID of any type	1	793	26.5 ± 16.8	−1.373	0.021	−2.543–0.204
	2	415	25.9 ± 17.0	−2.069	0.005	−3.522–0.615
	≥ 3	220	24.2 ± 15.1	−3.260	< 0.001	−5.137–1.383
AA + ≥ 1 atopic CID and no non‐atopic CID	1	589	25.3 ± 16.4	−1.790	0.006	−3.058–0.521
	2	267	24.5 ± 16.1	−1.726	0.048	−3.433–0.019
	3	110	20.7 ± 13.2	−4.643	< 0.001	−7.157–2.130
AA + ≥ 1 non‐atopic CID and no atopic CID	1	204	30.0 ± 17.5	−0.231	0.819	−2.204–1.742
	2	35	31.8 ± 20.0	−3.115	0.172	−7.585–1.355
	≥ 3	7	28.6 ± 17.7	−7.707	0.126	−7.580–2.166

Note: Mean AA age‐of‐onset in cases with AA‐only was compared with that of each comorbidity subgroup. Significant results are given in bold.

Abbreviations: AA, alopecia areata; CI, confidence interval; CID, chronic inflammatory disorder; SD, standard deviation.

Although AA is considered a Th1‐mediated autoimmune disorder, research has implicated the Th2‐immune axis in AA pathobiology, and its involvement may be more pronounced in AA patients with comorbid atopic disorders [6]. A plausible hypothesis is that the extent of Th2 involvement in AA is an important modulator of disease course, which positively correlates with an increased risk for poor clinical outcomes.

Our findings suggest that distinct comorbid constellations may indicate AA subtypes with differing prognoses [4]. AA patients with comorbid CIDs, particularly AD, bronchial asthma or Hashimoto's thyroiditis, may benefit from increased clinical monitoring and earlier therapeutic intervention.

Author Contributions

A.F. performed the statistical analyses, was involved in the interpretation of data and contributed to the writing of the manuscript. M.‐T.S. supervised the statistical analyses, contributed to the design of the study and was involved in the interpretation of the data. Y.G., S.R., B.B., G.L., U.B.‐P. and R.C.B. enrolled patients and were responsible for data collection. M.M.N. and R.C.B. were involved in the study design and data interpretation and contributed to the supervision. F.B.B. conceived, designed and supervised the study, interpreted the data and wrote the manuscript. All authors have been involved in the critical revision of the manuscript for important intellectual content.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.