Psychiatric co‐morbidity in patients with hidradenitis suppurativa: A cross‐sectional study of clinical characteristics and burden of disease

Hidradenitis suppurativa (HS) is a chronic inflammatory disorder of the hair follicles characterized by recurrent painful deep-seated nodules, abscesses, tunnels and scars primarily in the axillae, inguinal area, inframammary folds and anogenital regions.1-3 Estimates on prevalence vary greatly, and have been reported to range from 0.1% to 4%,2, 4 with an average diagnostic delay of approximately 7.2 years.5 The intractable nature of HS and its association with pain, suppuration, impaired daily activities and malodor results in both physical and psychological discomfort, but also significantly impaired quality of life.6-8 Patients with HS have a high occurrence of several co-morbidities, including hypertension, dyslipidemia and diabetes,9 but also inflammatory bowel disease (IBD), rheumatoid arthritis, spondyloarthropathy4, 8 and psychiatric disorders10-12 including suicide.13 Psychiatric co-morbidity (PCM) has already proven highly prevalent in dermatological patients,14 but the association is even greater in patients with HS with approximately one in four experiencing psychiatric disease.11, 15 While depression is considered the most prevalent psychiatric condition, other psychiatric co-morbidities including schizophrenia, anxiety and bipolar disorder have also been shown to have an increased prevalence in patients with HS,10, 11 with some studies suggesting that patients with HS have even more severe depression compared to patients with other skin diseases.16 The reasons as to why psychiatric disease is so prevalent in patients with HS remain elusive. Pro-inflammatory cytokines and chemokines (i.e., tumour necrosis factor-α and interleukin-10), which are involved in HS pathogenesis,17, 18 have previously been suggested to play a role in major depressive disorder.19, 20 Furthermore, quality of life and body image impairment21, 22 have been suggested as causes of PCM in patients with HS, but the association remains under investigation. Although the increased prevalence of PCM in patients with HS is already well established, the literature currently lacks details on demographic and clinical characteristics of this group of patients when compared to patients with HS without PCM. Identification of potential associations and predictors may aid in the multidisciplinary collaboration of treatment regimes in patients with HS and PCM. This study examined these characteristics, as well as burden of disease and temporal relationships of onset of disease in a well-characterized hospital-based cohort of HS outpatients. Consecutive newly referred patients with HS from a dermatological university hospital outpatient clinic (Department of Dermato-Venereology, Bispebjerg Hospital, Copenhagen, Denmark) between January 2016 and October 2022 were recruited for the study. Demographic and clinical data, including age, sex, employment status, ethnicity, age at HS onset, time to subspecialized care, presence of HS in a first-degree relative, weight, height, smoking status, current diagnosed co-morbidities, age at onset of co-morbidities, blood samples, blood pressure measurements, burden of disease and disease severity were obtained through patient interviews and clinical examination. Positive smoking status was considered as previous or current smoker, unemployment was considered as being between the age of 18 and 65 and having no current occupation or not under education, and obesity was considered as having a body mass index (BMI) above or equal to 30 kg / m 2 . Age at onset of HS was based on patient-reported first onset of symptoms of HS, and time to subspecialized care was calculated as time from HS onset to first visit at a dermatological hospital department. Disease severity was measured using Hurley staging, number of boils in the past month and International Hidradenitis Suppurativa Severity Score System (IHS4) scores calculated from clinical examination. Burden of disease was measured using the Dermatological Life Quality Index (DLQI) questionnaire and VAS-10 scores for pain and overall bother. All psychiatric conditions previously diagnosed by a physician were considered for this study and subgrouping of psychiatric disease was based on the International Classification of Diseases 10th Revision (ICD-10) groups for mental and behavioural disorders. In the overall group included PCMs were affective disorders, anxiety or other neurotic disorders, schizophrenia or psychosis, attention deficit hyperactivity disorder (ADHD) or Tourette's, personality disorders, eating disorders and autism spectrum disorders. Analyses were conducted on the overall group and the subgroups with affective disorders, anxiety or neurotic disorders, and with schizophrenia or psychosis, with patients with HS without PCM serving as controls. Only current, non-remitted PCM was considered for this study. Statistical analysis was computed using IBM SPSS statistics version 25 (SPSS, Inc.). Patient groups were defined as mentioned above. Continuous variables were presented with means and standard deviations and categorical variables with numbers and percentages. Continuous data were examined for normal distributions, and independent sample t-test, χ2 or Mann–Whitney U were used to examine relationships and differences in variables between groups. Box–Tidwell test was conducted to examine linearity of independent variables and log-odds before obtaining adjusted odds ratios (OR) using logistic regression with 95% confidence intervals (CI). Analysis of time to development of psychiatric disease was conducted using Kaplan–Meier curves and Cox proportional hazards regression with the time since patient reported onset of HS as the underlying time, and results reported as hazard ratios (HR) with 95% CI. All results were considered of statistical significance with a p value <0.05. A total of 667 patients were included in the study. Overall characteristics of the included patients are available in Table 1. Of the 667 patients with HS, 183 (27.4%) had a diagnosis of at least one PCM at the time of inclusion. 108 (16.2%) had an affective disorder, 71 (10.6%) had an anxiety or neurotic disorder, 22 (3.3%) had schizophrenia, schizotypal disorder or episodes of psychosis, 20 (3.0%) had ADHD or Tourette's syndrome, 11 (1.6%) had a personality disorder, 7 (1.0%) had a behavioural disorder, 3 (0.4%) had autism or Asperger's syndrome, and 6 (0.9%) had other psychiatric disorders. 39 patients reported having been diagnosed with two psychiatric conditions and 14 reported having three or more psychiatric conditions (Figure 1). Demographic data are available in Table 2. Overall, patients with PCM were more likely to be Caucasian (88.5% vs. 80.6%), OR: 1.86 (1.12–3.09), p < 0.05 and younger at HS onset (23.6 vs. 26.2 years), p < 0.01. When adjusted for age and sex, patients with PCM were more likely to be obese (44.0% vs. 34.5%), OR: 1.50 (1.06–2.13), p < 0.05; unemployed (52.5% vs. 18.7%), OR: 5.50 (3.73–8.10), p < 0.001 and current or previous smokers (89.6% vs. 73.2%), OR: 3.76 (2.20–6.40), p < 0.001. Specific for the different PCM subgroups, patients with affective disorders were more likely to be female (73.1% vs. 65.6%), OR: 1.67 (1.05–2.66), p < 0.05, whereas patients with anxiety or other neurotic disorders were younger (36.73 vs. 40.30 years), p < 0.05 when compared with patients without PCM. Without psychiatric comorbidity, n = 484 (72.6%) With psychiatric comorbidity, n = 183 (27.4%) Affective disorder, n = 108 (16.2%) Schizophrenia or psychosis, n = 22 (3.3%) ORs for specific co-morbidities are shown in Figure 2. Overall, patients with PCM and HS were more likely to have asthma or/and chronic obstructive pulmonary disease (COPD) (9.8% vs. 5.6%), OR: 1.94 (1.03–3.66), p < 0.05 when adjusting for age and sex, with asthma (6.0% vs. 3.5%), OR: 1.73 (0.79–3.78) and COPD (4.4% vs. 2.3%), OR: 2.35 (0.90–6.14) having increased ORs individually as well. Patients with PCM were less likely to have IBD (4.9% vs. 11.6%), OR: 0.39 (0.19–0.80), p < 0.05, specifically Crohn's disease (3.3% vs. 8.3%), OR: 0.38 (0.16–0.91), p < 0.05 when compared with patients without PCM adjusted for age and sex. When considering specific subgroups, patients with schizophrenia or psychosis were more likely to have acne (31.8% vs. 15.3%), OR: 2.59 (1.02–6.56), p < 0.05, though this did not remain statistically significant when adjusting for age and sex. No other examined co-morbidities were found to be significantly different between the groups, although type 2 diabetes seemed more prevalent in the subgroup with schizophrenia (18.2% vs. 8.2%) (Supporting Information: Table 1). No clinically significant differences in biochemical nor blood pressure measurements were found between the overall group with PCM, or the subgroups as compared with the patients without PCM when adjusting for age and sex (Supporting Information: Table 2). Within unadjusted analysis regarding disease severity, no significant differences in Hurley stages (I: 36.1% vs. 36.4%, II: 50.3% vs. 49.4%, III: 13.7% vs. 14.3%), IHS4 score groups (mild: 49.7% vs. 50.3%, moderate: 29.5% vs. 28.2%, severe: 20.8% vs. 21.5%) or boils in the past month (mean 2.67 vs. 2.38) between the group with PCM and the group without PCM were found (Supporting Information: Table 3). Regarding PCM subgroups, only the schizophrenia group had more patients with Hurley III (36.4% vs. 14.3%), p < 0.01 when compared to patients without PCM. Additionally, patients with anxiety were less likely to have Hurley stage III (4.2% vs. 14.3%), p < 0.05 and severe IHS4 score (11.3% vs. 21.5%), p < 0.05. When examining burden of disease, both overall and specific PCM subgroups had higher mean DLQI scores when compared to patients without PCM, but with p > 0.05. Overall, patients with PCM were less likely to have DLQI scores of 0–1 (5.0% vs. 10.5%), p < 0.05, a difference which persisted across both the affective disorder (3.7% vs. 10.5%), p < 0.05 and anxiety or other neurotic disorder subgroups (1.4% vs. 10.5%), p < 0.05. Specifically for the anxiety or other neurotic disorder subgroup, patients were more likely to have DLQI scores of 11–20 (48.6% vs. 35.2%). Overall bother scores were significantly higher across both the overall group with PCM (7.28 vs. 6.69), p < 0.05, the affective disorder group (7.43 vs. 6.69), p < 0.05 and the anxiety or other neurotic disorder group (7.34 vs. 6.69), p < 0.05. Mean (SD) years of PCM onset in relation to HS diagnosis was 6.2 (11.9) for depression, 3.5 (13.1) for anxiety, 2.5 (13.5) for schizophrenia or psychosis and −0.8 (10.6) for schizophrenia alone (Figure 3). Cox proportional hazards regression showed that patients with HS onset above the age of 30 years had the lowest risk of developing psychiatric disease within a 10-year follow-up (p < 0.01). With this group as reference, risk of PCM increased with HS onset age group 22–30 years, HR: 1.27 (0.56–2.89), p > 0.05; HS onset age group 16–21 years, HR 2.12 (1.02–4.40), p < 0.05 and HS onset age group <15 years, HR: 3.37 (1.56–7.31), p < 0.01 demonstrating a trend towards highest 10-year PCM risk at lowest age of HS onset (Figure 4). Results remained statistically significant when adjusting for sex. We found that psychiatric disease was present in 27.4% of patients with HS, with affective disorders being the most prevalent, followed by anxiety or other neurotic disorders. Furthermore, we found that 7.9% of the patients had more than one PCM, suggesting a high affliction of disease in this group of patients. We also found that patients with PCM had a high occurrence of smoking, obesity, unemployment and were predominantly Caucasian when compared to patients without PCM. Of the specific nonpsychiatric co-morbidities, COPD was more prevalent in the PCM group and IBD, especially Crohn's disease, was less prevalent. Apart from the schizophrenia subgroup, patients with PCM did not have more severe disease signified by Hurley stage and IHS4 scores. Nevertheless, a higher burden of disease shown by increased DLQI and increased overall bother score was found in the PCM group when compared to the group without PCM. Additionally, patients with PCM were less likely to be in the lowest scoring DLQI group. Patients with HS onset after the age of 30 years had the lowest risk of developing PCM within the following 10 years. Increasing risk with age groups 22–30, 16–21 and below 15 years was found, with the latter having the highest risk. As psychiatric disease in itself is associated with both smoking23, 24 and obesity,25, 26 the high occurrence of these factors in the PCM group is expected. It has been suggested that especially antipsychotic drugs could contribute to obesity,25 and that cigarette smoking could serve as a mean of self-medication of psychiatric symptoms,24, 27 with some studies arguing that smoking is actually a risk factor for developing psychiatric disease.27, 28 Of concern is that patients with HS have an increased risk of cardiovascular events and all course mortality,29, 30 which could suggest an even greater risk in patients with HS and PCM due to this increased occurrence of smoking and obesity. Even though other cardiovascular risk factors such as type 2 diabetes and hypertension did not occur more frequently in patients with PCM, the mean age was 39.2 years overall, which could suggest that these diseases were yet to ensue. Additionally, when examining other co-morbidities, the increased risk of COPD and/or asthma could likely be linked to the increased prevalence of smoking.31 Although HS has previously been reported to have a negative impact on work life,32 this is to our knowledge the first time a study has showed a higher risk of unemployment in patients with HS and PCM. Several studies on the effect of unemployment on physical and mental wellbeing have shown an increased risk of psychiatric disorders in unemployed persons,33-35 suggesting a possible causal explanation for the increased risk of unemployment in the PCM group. As unemployment to some extent increases both the chances of illness and mortality,36-39 this could pose an additional risk for the group of patients with both PCM and HS. Furthermore, unemployment serves as an additional factor in impairment of quality of life in patients with HS.40 The patients included in this study were predominantly Caucasian, but a significant difference in ethnicity was found between the study groups. Patients with PCM were more often Caucasian, which corresponds well to current literature suggesting PCM to be more prevalent in the white population than in people of non-Caucasian descent.41-43 Many factors could contribute to this difference, some of them being that people of non-Caucasian descent experience discrimination in contact with healthcare systems, especially mental healthcare services,44 resulting in reluctance to seek out aid. Another factor is that in some cultural and ethnical contexts, there is a considerable amount of stigma towards mental health issues, resulting in patients not seeing a healthcare professional for diagnosis and treatment.45 All these factors could contribute to an underrepresentation of PCM in the non-Caucasian population of this study. IBD was less prevalent in the PCM group. IBD has previously been established to be more prevalent in patients with HS,46, 47 with associations suggesting it as part of a systemic inflammation.8 A significant difference between the study groups was the heightened number of smokers within the PCM group. Inconsistent with the findings of this study, smoking has previously been shown to increase the risk of Crohn's disease.48 Furthermore, a recent review indicated that depression and anxiety is a common diagnosis in patients with IBD.49 The results of this study could consequently perhaps be due to a small sample size of IBD in the data set. The impact of HS on quality of life increased body image impairment and impact on sexual health have previously been demonstrated.21, 22, 50 This study found that patients with HS and concurrent PCM had an even more profound impact on quality of life, even though they are not experiencing more severe clinical disease. This discrepancy between the illness perception and clinical disease presentation of patients with HS has previously been described in literature.51 Furthermore, a previous study investigating the perception of pain in patients with HS found that patients with depression or anxiety report higher pain index scores than patients without.52 We believe the findings of this study warrant an even greater attention from the clinician to accommodate the needs and disease perceptions of this group of patients. As PCM is shown to be highly prevalent in patients with HS, one could suggest, in collaboration with psychiatrists, to include a specific tool for screening patients for psychiatric illness both upon first hospital visit and concurrent follow-up visits. As depression is the most common psychiatric illness within this group of patients with validated questionaries available, this could be easily integrated within dermatological practice. The greatest risk of developing PCM within 10 years of HS-onset was found in patients below the age of 15. With affective disorders and anxiety/neurotic disorder being the most common PCMs in this study with peak incidence rates in the early 20s for the background population,53, 54 it is possible that patients with HS are at even greater risk of developing early psychiatric disease in childhood or adolescence. This has also been suggested by a recent study by Wright et al.,55 although further studies investigating this specific association are needed. We found the prevalence of PCM in the HS population to be 27.4% which corresponds well to current literature,11, 15, 56, 57 strengthening the validity of the study. Additionally, this study included a large group of well-characterized patients with HS, most of them clinically diagnosed and assessed by the same physicians, providing continuity in data collection, and further adding to the strength of the study. Also, age of disease onset was based on patient-reported data, providing a better insight into temporal relationships. Nevertheless, as psychiatric disease is a broad spectrum of disorders exhibiting different pathogenetic and clinical manifestations, the reduction to one or more groups could reduce generalizability. Additionally, this study is limited to patients seen at one university hospital department, which could lead to an overrepresentation of more severe HS cases, and possibly patients having already initiated several different treatment regimens over a long period of time. The study is also limited to patients with HS, and to further characterize the PCM group, further studies including a healthy non-HS control group could prove valuable. Several factors could influence the diagnosis of psychiatric disease, and since allocation to the PCM group was based on physician-diagnosed psychiatric illness, it could lead to an underestimation of actual psychiatric disease in the patient population. Additionally, no systematic questionnaire assessing current symptoms of psychiatric disease was incorporated in the study, nor any HS-specific questionnaires on quality of life, which could prove valuable in future studies. Including the mentioned limitations of this study, additional further elaboration as to why PCM is more prevalent in patients with HS is needed. Potential clinical implications of the increased prevalence of risk factors would furthermore contribute to a better understanding of this group of patients and could help in future treatment strategies. This study showed a high prevalence of PCM in patients with HS and that patients with HS and concomitant PCM exhibit a variation in demographic and clinical characteristics, risk factors and burden of disease, which warrants an increased need for a multidisciplinary approach and awareness from the clinician. Nikolaj Holgersen: Conceptualization; methodology; investigation; data curation; formal analysis; writing—original draft; reviewing; editing. Simon Francis Thomsen: Conceptualization; methodology; supervision; writing—review and editing. Nana Aviaaja Lippert Rosenø: Formal analysis; writing—review and editing. Hans Christian Ring, Valdemar Wendelboe Nielsen, Alexander Egeberg, and Jacob Thyssen: Conceptualization; Supervision; writing—review and editing. This work did not receive any funding. The authors declare no conflict of interest. Ethical Approval: not applicable. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.