Association between age at symptom onset and disease severity in older patients with hidradenitis suppurativa

Dear Editor, Although approximately 50% of patients with hidradenitis suppurativa (HS) are over 40 years old, little is known regarding the factors associated with disease severity at an older age.¹ We observed active and severe disease in a substantial number of older patients (> 40 years) during their first clinic visit. Recent data indicate that self-reported age at HS symptom onset may be bimodally distributed, with most patients reporting onset of symptoms in late adolescence and a small portion reporting initial symptoms in their mid-40s.² We hypothesized that disease severity in older patients with HS was related to an older age at symptom onset.

In this cross-sectional study, we characterized the association between self-reported age at symptom onset and disease severity in older patients with HS. We included patients who were 40 years of age or older during their initial clinic visit, had a diagnosis of HS (ICD-10-CM code L73.2) and were seen at the HS Clinic at Duke University between January 2016 and March 2021. The primary outcome was the International Hidradenitis Suppurativa Severity Score System (IHS4).³ Age at symptom onset was defined as the age at which patients self-reported that ‘[their] HS boils first started’. Based on a previous study identifying an optimal separation at around 30 years of age, we classified age at symptom onset as common onset (age ≤ 30 years) or late onset (age > 30 years).⁴

The association between IHS4 and age at symptom onset was assessed using univariable and multivariable linear regression analyses. Covariates in the multivariable regression model included factors associated with HS severity and concomitant systemic therapy.⁵ As the raw score was skewed towards higher severity, IHS4 was log-transformed, and regression diagnostics were performed to assess the validity of the model. The results are presented as the geometric mean ratio (GMR) with 95% confidence interval (CI).

The study was reviewed and approved by Duke University institutional review board, protocol #00107473. Statistical analyses were performed using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA).

The final analysis consisted of 102 patients, including 77.5% female and 67.6% Black/African American. Their demographic characteristics are summarized in Table 1. As a sensitivity analysis, we used k-means to separate age at symptom onset, identifying groups with common onset (mean 18.0, SD 5.5; n = 64) and late onset (mean 47.8, SD 10.0; n = 38), with separation at around 30 years old. The categorization results are consistent with the kernel density plot for age at symptom onset. At their initial clinic visit, 28% of patients with common-onset HS had Hurley stage III (n = 18), while 58% with late-onset HS had Hurley stage III (n = 22, P = 0.010). A greater proportion of patients with late-onset HS (37%, n = 14) had a Physician’s Global Assessment score of ‘severe’ or ‘very severe’ compared with those with common-onset HS (8%, n = 5; P < 0.001).⁶ Univariable analysis demonstrated that the GMR of IHS4 between patients with late-onset HS and common-onset HS was 2.07 (95% CI 1.21–3.54, P = 0.009). Multivariable analysis also showed that patients with late-onset HS had a significantly higher IHS4 than patients with common-onset HS (GMR 3.35, 95% CI 1.24–9.03, P = 0.02), after adjusting for other risk factors associated with HS severity (Table 1).

Table 1.

Multivariable regression analysis of factors potentially associated with disease severity in older patients with hidradenitis suppurativa (HS). ‘Older patient’ is defined as age > 40 years at the initial clinic visit

Variable	Individual group comparisonsa			Multivariable regressionb
	Common onset	Late onset	P-value	GMR (95% CI)	P-value
Age at symptom onset, n (%)
Late (> 30 years)	–	38 (100)	< 0.001	3.35 (1.24–9.03)	0.02
Common (≤ 30 years)	64 (100)	–		Reference
Sex, n (%)
Male	8 (13)	15 (39)	0.002	0.99 (0.49–2.02)	0.98
Female	56 (88)	23 (61)		Reference
Race, n (%)
Black	38 (59)	31 (82)	0.03	1.57 (0.84–2.94)	0.15
Other	6 (9)	0 (0)		2.54 (0.74–8.74)	0.14
White	20 (31)	7 (18)		Reference
Body mass index (kg m−2)	34.2 (28.7–41.0)	31.6 (28.1–39.6)	0.38	0.98 (0.94–1.01)	0.19
Smoking status, n (%)
Formerly or currently	42 (66)	21 (55)	0.30	0.67 (0.38–1.18)	0.16
Never	22 (34)	17 (45)		Reference
Duration of HS (years)c	32.0 (26.0–36.5)	8.0 (3.0–11.0)	< 0.001	1.03 (0.99–1.06)	0.12
Prediabetes or diabetes, n (%)
Yes	26 (41)	13 (34)	0.93	0.92 (0.52–1.65)	0.79
No	38 (59)	25 (66)		Reference
Hormone modulatory therapy, n (%)d
Yes	16 (25)	12 (32)	0.47	1.18 (0.63–2.19)	0.60
No	48 (75)	26 (68)		Reference
Oral antibiotics, n (%)
Yes	18 (28)	12 (32)	0.71	1.19 (0.66–2.14)	0.57
No	46 (72)	26 (68)		Reference
Immunomodulatory therapy, n (%)e					0.49
Yes	5 (8)	5 (13)	0.38	1.38 (0.55–3.51)
No	59 (92)	33 (87)		Reference

The ‘common onset’ and ‘late onset’ columns report the statistics for each covariate in the model according to age at symptom onset (common onset ≤ 30 years, late onset > 30 years). The log-transformed International HS Severity Score System was used as the outcome in the model. CI, confidence interval; GMR, geometric mean ratio. ^aWilcoxon rank sum test or χ²-test used for individual group comparisons presented as the median (interquartile range) or count (percentage). ^bAll table variables included in the multivariable regression model. ^cDuration of HS defined as the difference between the age of first reported symptoms and the age at the initial clinic visit. ^dHormone modulatory therapy includes spironolactone, oral contraceptive pills, finasteride and metformin. ^eImmunomodulatory therapy includes adalimumab and infliximab.

We found that age at symptom onset in HS could be an important prognostic factor. Interestingly, a study examining the relationship between HS severity and age at symptom onset in an Italian population found that late-onset HS was associated with less severe disease, although these findings were restricted to age- and sex-adjusted analyses.⁴ Our study adjusted for other variables that impact disease severity including race, body mass index, smoking, duration of disease, insulin resistance, and concomitant medical therapy. Furthermore, we focused on an older patient population, and the majority of patients in our cohort were African American.

Study limitations include recall bias in terms of reported age of initial symptoms; however, the distribution of age at symptom onset was tightly clustered and did not overlap between the common-onset and late-onset groups. Our study is only generalizable to patients over age 40 years in the USA, and the limited sample of patients seen in our tertiary referral centre may restrict detection of significant factors in the multivariable regression and representation of the general population of patients with HS. Age and disease severity at first presentation to our clinic are variable, which may account for a large variance for ISH4 compared with the sample size in both symptom onset groups, and a wide 95% CI for the GMR comparing ISH4 between the late- and common-onset groups. Finally, the age and magnitude of maximal disease severity for both symptom onset groups are unclear, as our study lacked longitudinal data; prospective studies are needed to ascertain disease course.

In conclusion, age at symptom onset is an important factor associated with disease severity in older patients with HS, and could indicate a patient’s prognosis.

Funding sources

we wish to acknowledge support from the Biostatistics, Epidemiology and Research Design (BERD) Methods Core funded through grant award number UL1TR002553 from the National Center for Advancing Translational Sciences, a component of the National Institutes of Health.