Diabetes status is associated with greater disease severity in hidradenitis suppurativa

Abstract

With the rising prevalence of endocrine conditions in the U.S., it is important to examine the relationship between diabetes mellitus and hidradenitis suppurativa (HS) severity. This retrospective, cross-sectional study evaluated whether diabetes status and HbA1c levels are associated with HS severity among 275 physician-diagnosed patients seen at an HS Specialty Clinic between June 2016 and January 2021. The primary outcome was HS severity measured by the International HS Severity Score System (IHS4); the secondary outcome was pain severity measured by the Numeric Pain Rating Scale (NRS).Among participants, 44.3% were non-diabetic, 35.3% pre-diabetic, and 20.4% diabetic. Median HbA1c levels were lowest in non-diabetic patients (5.3 [5.1, 5.5]) and highest in diabetic patients (6.9 [6.3, 8.4]). The NRS pain score was also highest in the diabetic cohort (5.0 [0.0–8.0]) compared to the non-diabetic (2.0 [0.0–5.0]) and pre-diabetic (4.0 [0.0–6.0]) groups (p = .005). A significant association was observed between HbA1c and IHS4 in univariate analysis (p = .004), with multivariable analysis approaching significance (p = .053) after adjusting for covariates. These findings suggest diabetes and higher HbA1c levels correlate with increased HS severity and pain. Prevention and management of diabetes may therefore modify the clinical course of HS.

Introduction

Hidradenitis suppurativa (HS) is a chronic inflammatory skin condition with disease activity characterized by abscesses, painful nodules, and draining fistulae predominantly in the skin folds. The global prevalence is 0.05–4%, with women affected more than men (Alikhan et al, 2009). Multiple studies in the United States have reported greater HS severity in African-American or Black patients (Byrd et al, 2023). Various comorbidities have been identified with HS, including endocrine, cardiac, and psychiatric conditions (Tzellos and Zouboulis, 2020). Endocrine comorbidities such as metabolic syndrome and diabetes mellitus have been observed in up to 50% and 25% of patients with HS, respectively (Garg et al, 2018; Gold et al., 2014). In this retrospective, cross-sectional observational study, we assessed whether there was an association between HS disease severity and diabetes status as well as hemoglobin A1c (HbA1c) levels.

Results

The cohort included 275 patients with HS: 122 (44.3%) without diabetes, 97 (35.3%) with prediabetes, and 56 (20.4%) with diabetes. Patients with diabetes showed the highest median (quartiles 1–3) International Hidradenitis Suppurativa Severity Score System (IHS4) score (12.0 [4.0–39.5] than nondiabetics (4.0 [1.0–11.0]) or prediabetics (6.0 [2.0–18.0]) (P < .001). An IHS4 score <3 signifies a mild disease, scores of 4–10 signify moderate disease, and severe disease receives a score >11.³ Numeric Pain Rating Scale score was the highest in the diabetic cohort (5.0 [0.0–8.0]), suggesting moderate pain, than that in nondiabetic (2.0 [0.0–5.0], mild pain) and prediabetic (4.0 [0.0–6.0]) cohorts (P = .005) (Table 1).

Table 1.

Demographic, Treatment Characteristics, and Disease Severity Stratified by Diabetes Status

Characteristics	No Diabetes (n = 122, 44.3%)	Prediabetes (n = 97, 35.3%)	Diabetes (n = 56, 20.4%)	Missing	P-Value
Age at visit, y	30.2 (23.3–39.6)	39.9 (31.9–50.8)	51.1 (42.8–60.6)	1	<.001
Male sex, n (%)	25 (20.5%)	17 (17.5%)	14 (25.0%)	0	.54
Race, n (%)				4	.19
African-American/Black	69 (56.6%)	66 (68.0%)	42 (75.0%)
Other/unknown	12 (9.8%)	8 (8.3%)	2 (3.6%)
White	41 (33.6%)	23 (23.7%)	12 (21.4%)
Hispanic/Latino, n (%)	6 (4.9%)	6 (6.2%)	2 (3.6%)	0	.77
Smoking status, n (%)				3	.064
Current smoker	20 (16.5%)	28 (29.2%)	19 (34.5%)
Former smoker	26 (21.5%)	21 (21.9%)	9 (16.4%)
Never smoker	75 (62.0%)	47 (49.0%)	27 (49.1%)
BMI (kg/m2)	31.4 (26.7–37.7)	35.0 (30.0–41.8)	36.0 (29.4–43.8)	4	.004
Hemoglobin A1c (%)	5.3 (5.1–5.5)	5.8 (5.7–6.1)	6.9 (6.3–8.4)	0	<.001
Duration of disease, y	9.0 (3.0–15.0)	15.0 (8.0–30.5)	22.5 (8.0–33.0)	4	<.001
C-reactive protein (mg/l)	0.6 (0.2–1.4)	0.9 (0.4–2.7)	1.0 (0.5–3.9)	55	.003
IHS4 scores	4.0 (1.0–11.0)	6.0 (2.0–18.0)	12.0 (4.0–39.5)	0	<.001
NRS pain scores	2.0 (0.0–5.0)	4.0 (0.0–6.0)	5.0 (0.0–8.0)	0	.005
Treatments, n (%)
Topical antibiotics	45 (36.9%)	29 (29.9%)	22 (39.3%)	0	.42
Oral antibiotics	33 (27.0%)	27 (27.8%)	18 (32.1%)	0	.77
Immunomodulatory therapy	19 (15.6%)	7 (7.2%)	10 (17.9%)	0	.094
Hormonal therapy	19 (15.6%)	18 (18.6%)	29 (51.8%)	0	<.001
Metformin	2 (1.6%)	6 (6.2%)	26 (46.4%)	0	<.001
Sulfonylurea	0 (0.0%)	0 (0.0%)	2 (3.6%)	0	.041
SGLT-2 antagonists	0 (0.0%)	0 (0.0%)	1 (1.8%)	0	.20
Insulin	0 (0.0%)	0 (0.0%)	18 (32.1%)	0	<.001
GLP-1 agonists	0 (0.0%)	0 (0.0%)	6 (10.7%)	0	<.001

Abbreviations: BMI, body mass index; GLP-1, glucagon-like peptide 1; HbA1c, hemoglobin A1c; IHS4, International Hidradenitis Suppurativa Severity Score System; NRS, numeric rating scale; Q1, quartile 1; Q3, quartile 3; SGLT-2, sodium glucose cotransporter 2.

Data are reported as median (Q1–Q3) or count (percentage) and compared using Kruskal–Wallis, chi-square, or Fisher’s exact test. The lowest value of HbA1c within 6 months of initial visit was recorded if multiple values were available. Race categories listed as other or unknown in the electronic medical record were documented as such in our dataset.

Univariable analysis showed a significant association between log(IHS4) and log(HbA1c) (β = 1.99, 95% confidence interval [CI] = 0.65–3.33, P = .004). Multivariable analysis demonstrated a 1.54-unit increase in log(IHS4), but the association was statistically insignificant (P = .053) after adjusting for other covariates (Table 2). Consistently, multivariable analysis also showed a significant association between log(IHS4) and diabetes status (P = .002). In comparison with individuals without diabetes, those with prediabetes exhibited an average increase of 0.56 units in log(IHS4) (95% CI = 0.02–1.10, P = .041), whereas individuals with diabetes demonstrated a 1.26-unit higher log(IHS4) (95% CI = 0.47–2.04, P = .002).

Table 2.

Association of Disease Severity as Measured by Log(IHS4) with Log(HbA1c) and Diabetes Status

Variable of Interest	Univariable Analysis		Multivariable Analysis1
	β (95% CI)	P-Value	β (95% CI)	P-Value
Log(HbA1c)2	1.99 (0.65–3.33)	.004	1.54 (−0.02 to 3.11)	.053
Diabetes status
No diabetes	Reference		Reference
Prediabetes	0.56 (0.05–1.07)	.032	0.56 (0.02–1.10)	.041
Diabetes	1.17 (0.56–1.77)	<.001	1.26 (0.47–2.04)	.002

Abbreviations: BMI, body mass index; CI, confidence interval; HbA1c, hemoglobin A1c; IHS4, International Hidradenitis Suppurativa Severity Score System.

β denotes regression coefficient. Missing data were imputed using the mean or most frequent category.

¹Covariates for each model included age, sex, race, smoking status, BMI, disease duration, and treatments (oral antibiotics, immunomodulators, hormonal therapy, and metformin).

²Because both IHS4 and HbA1c had skewed right distributions, the log-transformed IHS4 score and log-transformed HbA1c were used as the outcome and independent variables, respectively.

Discussion

In summary, we found that HbA1c and diabetes status were positively correlated with HS disease severity, as measured by IHS4. Secondary outcome Numeric Pain Rating Scale also demonstrated positive associations with diabetes status.

This adds to the expanding body of evidence demonstrating that patients with HS are at a significantly increased risk for developing diabetes and other metabolic disorders as well as cardiovascular morbidity and mortality. The pathophysiologic underpinnings of this association are multifactorial, encompassing chronic systemic inflammation, immune dysregulation, and a high prevalence of metabolic comorbidities.

In our cohort, the prevalence of diabetes mellitus was 20.4%, which is comparable with previously published data (Ahmad et al, 2021; Garg et al, 2018). Large-scale meta-analysis showed a 2–3-fold higher risk of diabetes mellitus in patients with HS than in non-HS controls after adjusting for body mass index (BMI) (Phan et al, 2019). Patients with diabetes have more severe HS disease as measured by Sartorius score (Bettoli et al, 2016). In a combined hospital- and population-based study, Miller et al (2014) found that patients with HS had significantly greater odds of diabetes, with an OR of 5.74 in a hospital cohort and 2.44 in a general population cohort, even after multivariable adjustment. Similarly, a large Israeli community-based analysis reported that HS was associated with diabetes, with an adjusted OR of 2.78 (95% CI = 2.22–3.47), along with increased prevalence of other metabolic comorbidities, including obesity, hypertension, and dyslipidemia (Shalom et al, 2015). A meta-analysis by Rodríguez-Zuñiga et al (2019) further confirmed this link, showing that HS was significantly associated with metabolic syndrome and its components, including diabetes, with a pooled OR of 1.82 (95% CI = 1.39–2.25). These estimates remain robust after adjustment for age and obesity, suggesting that systemic inflammation intrinsic to HS plays a central role in impairing glucose homeostasis. Specifically, HS is associated with an adipokine profile that promotes diabetes risk—specifically, lower adiponectin and higher leptin and resistin (Krajewski et al, 2023). Patients with HS frequently exhibit elevated serum levels of TNF-α, IL-1β, and IL-6, all of which interfere with insulin receptor signaling, exacerbate hepatic gluconeogenesis, and promote peripheral insulin resistance (Berbudi et al, 2025). Okdahl et al (2022) also found that patients with diabetes have low-grade inflammation with elevations in TNF-α compared with healthy controls, presumably promoting a feed-forward proinflammatory cycle in HS. Single-cell analysis shows a strong T helper 17 signature in HS, mainly driven by IL-1β, which links to IL-1/IL-17–responsive vascular injury pathways (Kim et al, 2023). In addition, the NLRP3 inflammasome is overexpressed in HS skin, and inhibiting it reduces inflammation, implicating this pathway in systemic inflammation and metabolic disease (Krajewski et al, 2024). Complement activation, with elevated C5a, further drives inflammation and cardiovascular risk in patients with HS (Kanni et al, 2018). Overall, these features support HS as a systemic inflammatory disease with important metabolic consequences. The coexistence of HS and diabetes may also influence HS disease severity: hyperglycemia and insulin resistance worsen systemic inflammation, impair wound healing, and predispose to recurrent infections, thereby aggravating HS activity and complicating its management. Conversely, poorly controlled HS, by sustaining chronic cytokine elevation, may perpetuate insulin resistance and worsen glycemic control, establishing a vicious cycle of disease amplification. Overall, these features support HS as a systemic inflammatory disease with important metabolic consequences and vice versa.

Limitations of this study include being single-center, cross-sectional, retrospective study with limited follow-up data and reliance on BMI without waist circumference measurements. However, the result of this study highlighted that prevention, early identification, and adequate control of diabetes may be clinically important in patients with HS.

For dermatologists, these data underscore the importance of screening patients with HS—particularly those with severe, longstanding, or early-onset disease—for diabetes and prediabetes. Simple measures such as fasting plasma glucose, HbA1c, or oral glucose tolerance testing can facilitate early detection. Integrating this metabolic surveillance into routine dermatology practice not only addresses HS as a skin disease but also as a systemic condition with multisystem consequences. Moreover, early lifestyle interventions such as dietary counseling, weight reduction, and smoking cessation can simultaneously benefit HS severity and diabetes risk. In patients with comorbid diabetes, collaboration with primary care and endocrinology is essential to optimize glycemic control, reduce systemic inflammation, and improve both dermatologic and cardiovascular outcomes.

In conclusion, the coexistence of diabetes complicates HS treatment and may necessitate more aggressive or tailored interventions. Elevated HbA1c levels, a marker of poor glycemic control, are associated with higher HS severity, highlighting the importance of effective diabetes management for controlling skin disease progression. Increased pain severity translates into reduced QOL, with physical discomfort, social stigma, and potential emotional or psychological distress. Patients with diabetes with HS may require multidisciplinary care to address these interconnected challenges. The observed positive correlations suggest that preventing diabetes or improving glycemic control in those at risk could mitigate HS severity and enhance long-term outcomes. Proactive metabolic health management may offer a pathway to reduce the burden of HS.

Materials and Methods

Study design and patient population

We conducted a retrospective, cross-sectional observational study of 275 patients with HS who were seen at an HS specialty clinic between June 2, 2016 and January 28, 2021. Patients included in this study were required to have a physician-confirmed diagnosis of HS and an available HbA1c measurement obtained within 6 months prior to their initial clinic visit. The Institutional Review Board reviewed the study and waived the informed consent requirement.

Data collection

Patient demographics (including age, sex, race, ethnicity, and current smoking status), BMI, HbA1c, physical examination findings, current treatments, physician-recorded outcomes, and patient-reported outcomes were collected during initial visits. Patient race and ethnicity were collected from the electronic health record, and these were self-reported. Race and ethnicity categories were collected because they are routinely documented in the electronic health record for all patients at our institution and are important demographic variables that may influence disease presentation and severity. Primary exposures were HbA1c and diabetes status. Diabetes status was determined by HbA1c and history of a diabetes diagnosis—HbA1c ≤5.6% is defined as no diabetes, 5.7–6.4% are defined as prediabetes, and ≥6.5% is defined as diabetes. Lowest value of HbA1c within 6 months of initial visit was recorded if multiple values were available. Primary outcome of the study is HS disease severity as defined by IHS4 (Zouboulis et al, 2017). Secondary outcome measure was Numeric Pain Rating Scale.

Statistical analysis

Descriptive statistics were performed to summarize patient demographics and clinical variables using the median with 25^th and 75^th percentiles (quartiles 1–3) for continuous variables and frequency counts/percentages for categorical variables. Groups were compared using the Kruskal–Wallis, chi-square, or Fisher’s exact test as applicable. The association between the level of HbA1c and IHS4 and diabetic status and IHS4 were examined using linear univariable and multivariable regression analyses, respectively. Because both IHS4 and HbA1c had skewed right distributions, the log-transformed IHS4 score and log-transformed HbA1c were used as the outcome and the independent variable, respectively. Other covariates included age at visit, sex, race, BMI, smoking status, oral antibiotics, immunomodulatory therapy, duration of disease, hormonal therapy, and metformin use. Missing data were imputed using the average value (for age, BMI, and disease duration) or most frequent category (for smoking status), given the small number of missing variables. Regression diagnostics were performed to assess the validity of the model. A P < .05 was considered statistically significant without adjustment, and all statistical analyses were performed using SAS, version 9.4 (SAS Institute, Cary, NC).

Ethics Statement

This study was approved by the institutional review board at Duke Health (protocol number Pro00107473). The Institutional Review Board waived the informed consent requirement.

Data Availability Statement

No large datasets were generated or analyzed during this study. Minimal datasets necessary to interpret and/or replicate data in this paper are available upon request to the corresponding author.

ORCIDs

Negar Foolad: http://orcid.org/0000-0002-2154-0562

Amy J. Petty: http://orcid.org/0000-0003-1413-8804

Winston Liu: http://orcid.org/0000-0001-9222-3452

Jeffery T. Kwock: http://orcid.org/0000-0002-0343-2453

Beiyu Liu: http://orcid.org/0000-0002-6706-9729

Cynthia L. Green: http://orcid.org/0000-0002-0186-5191

Andrea D. Coviello: http://orcid.org/0000-0003-0244-9787

Tarannum Jaleel: http://orcid.org/0000-0002-2362-5121

Conflict of Interest

T.J. is an investigator for Abbvie, Sonoma, Almirall, UCB and Eli Lilly. She reports consulting for Eli Lilly, Novartis and Chemocentryx and receiving honoraria. She has received funds from Pfizer and UCB for research fellow support. She also has received funds from Merck, Dermatology Foundation, Skin of Color Society, Duke Strong Start Award, and NIH K12 (grant number:K12 AR084231). The remaining authors state no conflict of interest.

accepted manuscript published online XXX; corrected proof published online XXX