Associations of Internalized Skin Bias With Age, Adverse Psychopathology, and Health-Related Quality of Life Among Patients With Hidradenitis Suppurativa: A Cross-sectional Analysis

Key Points

Question  To what extent does internalized skin bias (ISB) mediate the associations between depression and anxiety and health-related quality of life (HRQOL) in patients with hidradenitis suppurativa (HS)?

Findings  This cross-sectional study of 230 patients with HS found that ISB appeared to mediate the associations of depression and anxiety with HRQOL. After controlling for age, disease severity, and disease duration, both ISB and anxiety were substantial predictors of impaired HRQOL.

Meanings  The findings suggest that ISB contributes substantially to how adverse psychopathology affects patients with HS; interventions directed at ISB may mitigate the burden that depression and anxiety have among patients with HS.

Importance  Hidradenitis suppurativa (HS) is a chronic autoinflammatory disease that is highly associated with affective disorders such as depression and anxiety. Recent studies have shown that patients with HS may also endorse high levels of internalized skin bias (ISB). This redirection of skin disease−related social stigma toward oneself may also play an important role in the associations between adverse psychopathology (eg, depression, anxiety) and health-related quality of life (HRQOL).

Objectives  To evaluate the associations of ISB with other core outcomes in HS and to determine if ISB may mediate the associations between adverse psychopathological outcomes and impaired HRQOL.

Design, Setting, and Participants  A cross-sectional study of adult patients with HS recruited from 1 academic medical center as well as through virtual social and recruitment networks. Eligible participants completed an online survey comprised of 4 survey instruments along with demographic and disease history information from April 1, 2021, to July 15, 2021.

Main Outcomes and Measures  Measures included the Internalized Skin Bias Questionnaire (ISBQ), Burns Anxiety Inventory, the Beck Depression Inventory−II, the Hidradenitis Suppurativa Quality of Life (HiSQOL) instrument, along with demographic and disease history information. The primary outcome was HRQOL as measured by the HiSQOL. Data were analyzed in July through August 2021.

Results  The survey was completed by 230 of 279 patients (82.4%; mean [SD] age, 39.2 [11.2] years; 209 [90.9%] identified as female, 203 [88.7%] as not Hispanic, 178 [77.7%] as White). Depression and anxiety were shown to be a substantial burden in this sample, with 56.5% of participants’ scores qualifying for moderate or greater anxiety and 45.7% moderate or greater depression. The mean (SD) HRQOL scores were moderately high at 34.5 (16.7), indicating strong levels of impairment. There was a moderate correlation between ISBQ score and depression (r = 0.67); and fair correlations with HRQOL (r = 0.52) and anxiety (r = 0.56). Therefore, ISB appears to positively mediate the associations between depression and anxiety (estimated proportions of total effect that were mediated, 38.7% and 24.9%, respectively) and HRQOL. After controlling for age, disease severity, and disease duration, both ISB and anxiety were positive predictors of impaired HRQOL.

Conclusions and Relevance  This cross-sectional study found that ISB was associated with adverse psychopathology and impaired HRQOL in patients with HS. Furthermore, ISB appears to mediate the associations of depression and anxiety with HRQOL. Future studies are needed to design interventions targeted at addressing adverse psychopathology associated with ISB and improving HRQOL and well-being for patients with HS.

Hidradenitis suppurativa (HS) is an autoinflammatory dermatologic condition that can have substantial effects on health-related quality of life (HRQOL)¹ and the psychosocial health of patients.^2-6 A number of research teams have sought to characterize the trends and associations between adverse psychopathology and HS, specifically focusing on depression and anxiety. A recent meta-analysis estimated the overall prevalence of depression at 21% (95% CI, 17%-25%) and an overall odds ratio (OR) of 1.99 (95% CI, 1.63-2.43) among those with HS.² This study also estimated an overall prevalence of anxiety of approximately 12% (95% CI, 6%-17%) and OR of 1.97 (1.65-2.35).² Overall, patients with HS were nearly twice as likely to have depression or anxiety when compared with those without HS.

In recent years, researchers have begun to explore depression and anxiety further among individuals with skin diseases and to identify other psychological constructs that can contribute to negative affectivity (personal feelings, motivations, emotions [eg, anger, guilt, fear])⁷ and cognitions (thought processes used to learn and create understanding that, when misapplied, produce biases and stereotypes).⁸ One of the psychological constructs associated with negative affectivity and cognitions is internalized skin bias (ISB), also known as internalized skin stigma, which was recently evaluated in patients with acne⁹ and psoriasis.¹⁰ These studies’ findings suggest that ISB can be defined as the redirection of negative social stigmas and biases of skin disease toward how individuals feel about or perceive themselves.^9,10

The process of internalized bias has previously been studied among individuals with overweight and obesity,^11-13 mental illnesses,¹⁴ and dermatologic conditions.^9,10 To varying degrees, each of these health categories has incurred a notable level of social stigma. Specifically, because of the visible nature of skin diseases, patients with dermatologic conditions are often negatively perceived, being thought of as unattractive, contagious, and/or unhygienic and thereby, are stigmatized.¹⁵ Negative thought processes (ie, cognitions) from repeated exposure to external social stigma over time can lead patients to adopt these stigmatizing perceptions toward themselves; this is the final stage in the process of internalizing bias.¹⁴ Thus, perceptions of externalized stigma could in turn lead to negative feelings (ie, affectivity) associated with ISB at the individual level and may be notably correlated with impaired HRQOL and depressive symptoms.^7-9,16

Because of the potential relationship among ISB, depressive and anxious symptoms, and impaired HRQOL, a deeper understanding is needed of these particular psychosocial constructs and their associations with core outcomes in HS. Hidradenitis suppurativa-specific HRQOL, including emotional well-being, is a major domain in the core outcomes set needed in clinical trials,¹⁷ as well as in general patient care. However, a true understanding of the complex networks of adverse psychopathology, including depression and anxiety, is still limited and may negatively affect investigators’ and physicians’ abilities to evaluate emotional well-being. Thus, the purpose of this study is to evaluate the associations of ISB with other core outcomes in HS and determine if ISB may mediate the associations between adverse psychopathologic outcomes and impaired HRQOL.

This cross-sectional study was reviewed and approved by the Pennsylvania State University Human Research Protection Program. Informed consent was obtained from each respondent prior to their participation in accordance with the approved research protocols. An electronic summary of the research was provided at the start of the survey, and consent was implied for those who completed the questionnaire. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

Participants were recruited from a group of patients diagnosed with HS who had volunteered for research studies at Penn State Health Milton S. Hershey Medical Center. Participants were also recruited virtually through ResearchMatch StudyFinder at Penn State Health, as well as through network recruitment where participants shared the study's internet address with other potential participants. Eligible patients were 18 years of age or older, able to consent, and had either a positive result on a 2-item HS screening instrument¹⁸ or had active HS within the previous 2 years.

At the start of the survey, potential participants were able to review a summary explanation of the research and asked to affirm their interest by providing their age and email address and then responding to the screening questions.¹⁸ If the respondents’ screening result was positive, they proceeded to the survey. The survey, comprising the 4 instruments plus questions on demographic and disease history, was built using REDCap, a secure online web application.¹⁹ Data were collected from April 1, 2021, through July 15, 2021.

Four instruments were used in this analysis: (1) the Internalized Skin Bias Questionnaire (ISBQ),²⁰ (2) the Hidradenitis Suppurativa Quality of Life (HiSQOL) instrument,¹ (3) the Beck Depression Inventory−II (BDI−II),²¹ and the Burns Anxiety Inventory (BAI).²² The ISBQ was adapted from a previously validated instrument that measures internalized weight bias using a 7-point Likert scale (strongly disagree to strongly agree)^13,20; total score range was 0 to 54 points. The HiSQOL is a previously validated HRQOL measure that was specifically designed for patients with HS. The HiSQOL includes 3 subscales on symptoms, psychosocial factors, and activity adaptations; the total score range was from 0 to 68 points. The BDI-II is a 21-item instrument to screen for depressive symptoms and is consistent with the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition; DSM-IV) for identifying pathology associated with depression using a 4-point Likert scale;²³ and the total score can range from 0 to 63 points. The BAI is a 33-item screening instrument for symptoms of anxiety and has 3 subscales including anxious feelings, anxious thoughts, and physical symptoms. Similar to the BDI−II, the BAI is measured on a 4-point Likert scale and the total score can range from 0 to 99. For all 4 instruments, higher scores are correlated with higher pathology.

In addition to responding to these 4 instruments, participants were asked to provide demographic and health history information: age, sex, race, ethnicity, level of education, employment status, duration of HS disease, self-staging of HS severity, smoking status, and history of mental health concerns. Self-staging of HS used a combination of photographic examples²⁴ and descriptive qualifications of the various Hurley stages. Duration of disease was collected via self-report and patients were instructed to include the time prior to diagnoses in this estimation.

Data analyses were performed from July to August 2021 using SAS, version 9.4 (SAS Institute Inc). Statistical tests were 2-tailed with a level of significance set at α = .05. Given the voluntary nature of the survey, participants may have intentionally or accidentally missed items in the questions. Missing data on the items was minimal (0% for age to 7.83% for depression). To preserve sample size, multiple imputations were applied to missing data using the SAS−PROC MI function with the number of imputations set to 100. Multiple imputed data were included in the summary scores, correlation and regression models, and mediation analyses.

Descriptive statistics were used to characterize the study population. Correlations between assessments were calculated using Spearman correlation coefficients with Fisher z transformations to determine 95% CIs. Spearman correlation coefficients of 0 to 0.29 were considered poor; 0.30 to 0.60, fair; 0.61 to 0.80, moderate; and 0.81 to 0.99, very strong.²⁵ Univariable regression models were built to identify independent associations between covariates (ie, HRQOL and adverse psychopathology) and the ISBQ scores. A stepwise multivariable linear regression model was built predicting HRQOL as measured by the HiSQOL using a level of significance of 0.15 to be included in the final model and controlled for age, duration of disease, and disease severity. Mediation was evaluated using the SAS−PROC CAUSALMED function. Discrete models included both depression and anxiety as independent variables, HRQOL as the outcome, and ISB as a mediating variable. Mediation analyses were controlled for disease severity. Continuous predictor variables were standardized using the sample standard deviation. All outcome variables were included using the untransformed linear scores for regression and mediation models.

The authors hypothesized that there would be a negative association between ISB and HRQOL and that there would be a positive correlation between skin bias internalization and both depression and anxiety. Furthermore, ISB would partially mediate the association between depression and anxiety and HRQOL.

Of the 279 potential participants, 230 patients (82.4%; mean [SD] age, 39.2 [11.2] years; 209 [90.9%] women and 20 [8.7%] men) completed the survey; 49 (17.6%) respondents either elected not to participate (n = 3, 1.1%), were ineligible based on the screening criteria (n = 6, 2.2%), were duplicated (n = 2, 0.7%), or had incomplete data (n = 38, 13.6%), which was assumed to be a withdrawal of consent. Participant characteristics are presented in Table 1. Participants identified with the following race and ethnicity groups: 3 (1.3%) Asian/Pacific Islander, 27 (11.8%) Black/African-American, 3 (1.3%) Native American/Alaska Native, and 178 (77.7%) White/Caucasian, while 19 (8.3%) identified as Hispanic and 203 (88.7%) as non-Hispanic. Education level was evenly distributed with 115 (50.2%) participants reporting a bachelor degree or higher. Most patients self-identified as having HS Hurley stage 2 (n = 126; 55.0%) and 174 (76.0%) disclosed having had HS for 11 years or more. One-third of the sample (n = 86; 37.4%) reported being current smokers and 149 (67.73%) as having been diagnosed with a mental health condition (eg, depression, anxiety, bipolar disorder) by a health care practitioner.

Figure 1 illustrates the distribution of ordinal scores for both anxiety and depression with more than half of the participants qualifying for both moderate anxiety (n = 130; 56.5%) and depression (n = 105; 45.7%) or above. Estimated mean (SD) instrument scores with 95% CIs are presented in Table 2 and show a moderate association with HRQOL (34.48 [16.74]; 95% CI, 32.31-36.66) and moderate-to-strong endorsements of anxiety (31.05 [22.83]; 95% CI, 28.08-34.03) and depression (21.97 [13.68]; 95% CI, 20.17-23.77).

Correlation and regression data between the ISBQ and predictors of interest are also presented in Table 2. Spearman correlation coefficients and 95% CIs between ISBQ and outcomes of interest ranged from fair (HiSQOL, r = 0.52; 95% CI, 0.41 to 0.61; BAI, r = 0.56; 95% CI, 0.47 to 0.65) to moderate (BDI, r = 0.67; 95% CI, 0.59 to 0.74). Subscales regarding physical symptoms and functioning had the lowest correlations with ISBQ, including the symptoms (r = 0.32; 95% CI, 0.20 to 0.44) and activities (r = 0.45; 95% CI, 0.34 to 0.44) subscale of the HiSQOL and the somatic effects (r = 0.46; 95% CI, 0.35 to 0.56]) subscale of the BAI. Age had a poor negative correlation with ISBQ scores (r = −0.19; 95% CI, −0.31 to −0.06). The estimated regression coefficients (β) between ISBQ and the outcomes of interest demonstrated similar trends to the correlation coefficients and were all highly significant (supporting data in Table 2). Because predictor variables were standardized using the sample SD, regression coefficients can be interpreted as the estimated increase in the ISBQ per increase in 1 SD of the predictor. Predictors with the highest coefficients (95% CI) included depression (β, 7.70; 95% CI, 6.53 to 8.86), the psychosocial subscale of the HiSQOL (β, 7.56; 95% CI, 6.37 to 8.75), the anxious thoughts subscale of the BAI (β, 6.78; 95% CI, 5.53 to 8.04), and the total BAI score (β, 6.52; 95% CI, 5.25 to 7.79). The regression coefficient of age also showed significant association and demonstrated that ISBQ scores improved as patients grew older (β, −2.55; 95% CI, −4.05 to −1.05).

Table 3 displays results of the stepwise linear regression model using HRQOL as the outcome variable. Age, disease severity, and duration of disease were retained in the model as covariates. Of the 3 predictors, the ISBQ and BAI scores qualified for the significance level for entry and stay, and depression was removed from the model. Significant predictors in the final model included disease severity (supporting data reported in Table 3).

Figure 2A and B illustrate the results from the mediation analyses using depression and anxiety as predictors, respectively. The total effect of the association between depression and HRQOL was significant with a parameter estimate of 7.68 (95% CI, 5.92-9.43). The natural direct effect was lower and still significant with a parameter estimate of 4.71 (95% CI, 2.48-6.94). The estimated proportion of total effect that was mediated was 38.70%, indicating that ISB is a partial mediator in the association between depression and HRQOL. The total effect of the association between anxiety and HRQOL was significant was a parameter estimate of 8.70 (95% CI, 7.06-10.34). The natural direct effect was 6.53 (95% CI, 4.64-8.42). The estimated proportion of total effect that was mediated was 24.93% indicating that ISB also partially mediates the association between anxiety and HRQOL.

The findings of this study support previously published reports detailing elevated HiSQOL scores, indicating HS has a significant association with patient HRQOL.²⁶ Additionally, both depression and anxiety are prevalent comorbidities among patients with HS.⁶ Of the participants who completed the survey, 45.7% endorsed moderate symptoms of depression, and 56.52% endorsed moderate symptoms of anxiety. The total prevalence of participants with anxiety was markedly higher than estimated in previous studies. This discrepancy could be explained by the different measures used in previous studies as well as by the sampling strategies used in this study. However, it is important to note that both the BDI−II and BAI are intended to be screening tools for depressive and anxious symptoms, respectively. These instruments cannot clinically diagnose depressive and anxious disorders, and they are sensitive to change as the individuals’ moods change.

These findings also illustrate significant correlations between and associations of ISBQ scores with impaired HRQOL and adverse psychopathology, paralleling the findings of studies on acne and psoriasis.^9,10 Both mediation models revealed that ISB is an important mediator in the pathway from depression and anxiety to HRQOL. The purpose of mediation modeling is to identify a potential variable that could lie on the associative pathway between predictors and outcomes of interest. By understanding the percent to which a variable mediated the total effect of the association between the predictor and outcome, researchers can better understand the degree to which the mediator is contributing to the outcome of interest. Within the context of these study findings, the analyses show that the associations of depression and anxiety with HRQOL are partially mediated by ISB—38.7% and 24.9%, respectively. Because this was a cross-sectional study, it was not possible to capture the temporal relationship of the inclusive variables or their change over time; therefore, full mediation could not be evaluated.

Interestingly, age offered a protective effect against ISB in the univariable associations with a regression coefficient of −2.55 (95% CI, −4.05 to −1.05; P <.001). This association is consistent with other psychosocial analyses in which age offered a protective effect against adverse psychopathology. A recent study conducted in Canada found that older age groups reported lower rates of perceived stress, depression, and anxiety when compared with lower age groups during the COVID-19 pandemic.²⁷ Additional psychological studies have shown that affective disorders (eg, depression and anxiety) have the tendency to improve over time.^28,29 These trends are likely associated with changes in the psychological development of individuals over time, including learned resilience and the adaptation of various coping methods in response to social stressors. Because resilience and coping methods are learned behaviors, younger patients affected by ISB could benefit from psychological interventions, such as cognitive behavioral therapy (CBT), as a means to reduce the effect of ISB on patients.

Clinically, these findings highlight the degree to which depression and anxiety are associated with the psychological construct of ISB, as well as the fact that ISB plays an important role in the multidimensional psychopathologic network among patients with HS. Because depression and anxiety are major comorbidities associated with HS, identifying ISB as a contributing factor can help physicians and health care professionals understand better the mental health needs of this particular patient population. Consequently, mental health interventions that address the specific topic of ISB (eg, CBT) could mitigate the overall affects that depression and anxiety have on impaired HRQOL.⁸ Previous studies have shown that adjuvant psychotherapy, such as CBT, has proven effective for individuals with chronic health conditions because it can help address the mental health aspects of the disease, encourage the patient to engage with the health care team, and build skills to manage both their disease and their mental health concerns.^30-35 Patients with HS who present with symptoms consistent with affective mood disorders (eg, depression and anxiety) can be referred for mental health services such as adjuvant therapy along with standard dermatologic care.

A limitation of this research was the relatively small sample size of mostly female participants who identified as White—this could limit the generalizability of the findings. Furthermore, study participants were recruited from a single academic medical center and through network sampling on social media support platforms, which may have contributed to a potential selection bias. Additionally, disease severity was self-reported using a self-staging questionnaire, which may have contributed to misclassification of disease severity. Lastly, because the ISBQ scores were moderately correlated with depression and anxiety scores, there could be bidirectional applications to these models. However, given the known associations between affective disorders and impaired HRQOL, we sought to understand the role that ISB played in those associations. Because these associations were evaluated using cross-sectional data, the mediation models used in this analysis operated under the assumption that depression and anxiety acted as discrete causal influences on ISB.³⁶

This cross-sectional study supports previous studies that show impaired HRQOL, and affective disorders are prevalent in an HS population. Additionally, ISBQ scores are highly correlated and associated with impaired HRQOL, depression, and anxiety; however, age may offer a protective effect against ISB. Furthermore, ISB may function as an important partial mediator between the association of these common affective disorders and HRQOL. Future studies should develop psychosocial interventions aimed at addressing ISB as a contributing factor of depressive and anxious symptoms to improve overall adverse psychosocial outcomes and HRQOL and to promote wellness and well-being among patients with HS. Moreover, clinical trials should be conducted to evaluate the effectiveness of adjuvant psychotherapy on patient-centered and mental health outcomes among patients with HS. Systemically, advocacy initiatives to educate, raise awareness of, and eliminate social stigma in the general population as well as among health care can help counter the effects ISB can have on patients.

Accepted for Publication: January 25, 2022.

Published Online: March 16, 2022. doi:10.1001/jamadermatol.2022.0248

Corresponding Author: Melissa Butt, DrPH, Department of Family and Community Medicine, Penn State College of Medicine, 500 University Dr, Mailcode HU14, Hershey, PA 17033 (mbutt1@pennstatehealth.psu.edu).

Author Contributions: Dr Butt had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Kirby and Chinchilli contributed equally as senior authors.

Concept and design: Butt, Rigby.

Acquisition, analysis, or interpretation of data: Butt, Leslie, Foulke, Flamm, Kirby, Chinchilli.

Drafting of the manuscript: Butt, Rigby, Foulke, Chinchilli.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Butt, Chinchilli.

Administrative, technical, or material support: Butt, Rigby, Foulke.

Supervision: Leslie, Flamm, Kirby, Chinchilli.

Conflict of Interest Disclosures: Dr Butt reported support from HRSA (NIH funded T32 fellow) and grants from ACDS outside the submitted work; and being a co−designer of the Internalized Skin Bias Questionnaire during the conduct of the study. Dr Foulke reported an award from the Dermatology Foundation (Medical Dermatology Career Development Award) outside the submitted work. Dr Flamm reported grants from ACDS outside the submitted work. Dr Kirby reported grants from Incyte; serving on the advisory boards of AbbVie, Incyte, and Viela Bio; personal fees for speaking and consulting from AbbVie, ChemoCentryx, Janssen, Novartis AG, UCB Pharma, CSL Behring, and Incyte; and being the developer of the Hidradenitis Suppurativa Quality of Life (HiSQOL) instrument, all outside the submitted work. Dr Rigby reported being a codeveloper of the Internalized Skin Bias Questionnaire. No other disclosures were reported.

Funding/Support: The use of REDCap was supported by The Penn State Clinical & Translational Research Institute, Pennsylvania State University CTSA (NIH/NCATS No. UL1 TR002014). Dr Butt was supported by the Pennsylvania State University Primary Care Research Fellowship from the Health Resources and Services Administration (No. T32HP42015).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: The authors would like to thank the HS Patient Community, especially Hope for HS, HS Action Together, and HS Connect (HSConnect.org) for their support of this research. This research would not be possible without the support of the patients.