Stigma perceptions in patients with eosinophilic gastrointestinal disorders

SUMMARY

This study aims to evaluate the presence of perceived stigma in people diagnosed (self-reported) with an eosinophilic gastrointestinal disorder and examine the relationship to the patient's health-related quality of life and additional psychosocial patient-reported outcomes. One hundred forty-nine patients diagnosed for a minimum of 6 months participated in the study. Eligible participants completed questionnaires to assess perceived stigma, psychological functioning, and health-related quality of life. Perceived stigma was moderately associated with a decrease in total health-related quality of life and perceived treatment efficacy. Additionally, greater perceived stigma was positively correlated with an increase in anxiety, depression, and healthcare utilization. The study demonstrates the influence of perceived stigma on several aspects of health-related quality of life in patients with these illnesses. As such, it is important for health professionals to be aware of stigma in patients diagnosed with an eosinophilic gastrointestinal disorder.

INTRODUCTION

Eosinophilic gastrointestinal disorders (EGIDs) are newly understood chronic inflammatory conditions of the gastrointestinal (GI) tract. EGIDs result in adverse physiological changes and are accompanied by symptoms of food impaction, vomiting, bleeding, diarrhea, and pain in the corresponding area of the GI tract affected.^1,2 Despite their recent increase in prevalence,³ underlying histopathology remains uncertain and diagnostic criteria lacks definition.¹ Thus, treatment options, which consist of both pharmacological and dietary management,^1,2,4,5 are limited and vary in effectiveness.

The aforementioned qualities of EGIDs, associated symptom burden, and limited treatment options can impact several psychosocial aspects within a patients’ health-related quality of life (HRQOL).^6,7 Patients with an EGID have reported an interference in social relationships and interactions, such as experiencing worry about eating in restaurants, scaring others, and hiding choking from others.^6,7 In addition, EGID patients have expressed concern associated with having a chronic illness, potential treatment options, and impact of their symptoms on others.⁶ In addition, several characteristics of EGIDs such as being concealable, incurable, and chronic are susceptible to stigma.⁸ Stigma, or the labeling of a person as abnormal or different because of a perceived characteristic,⁹ is frequently observed in relation to chronic illnesses. Chronic illness stigma is predictive of increased psychological distress and poorer disease outcomes,^10,11 as well as decreased quality of life.^12–14

Although research in chronic illness stigma primarily focuses on HIV/AIDS,^15,16 stigmatization in cancer,^17,18 epilepsy,^19,20 hepatitis,^21,22 and obesity^23–26 has also been recognized. Within the broad construct of stigma is perceived stigma, characterized as subjective awareness of stigma. Individuals with perceived illness stigma may feel others are viewing them in a negative way because of their illness. Examples of perceived illness stigma include feeling as though others do not regard their illness as serious, do not believe it is ‘real’, or that their illness is self-inflicted or a result of something they did.²⁷ Individuals have reported perceived stigmatization in relation to other gastrointestinal illnesses such as inflammatory bowel disease (IBD)^28,29 and irritable bowel syndrome (IBS).^27,29,30 However, there are very few studies that have investigated prevalence and impact of perceived stigma in patients with EGIDs.

We aim to evaluate presence of perceived stigma in a cohort of patients diagnosed with an EGID. Secondary aims are to determine what demographic or clinical characteristics may be associated with greater stigma perception, as well as if perceived stigma is associated with patient outcomes including psychological distress, HRQOL, healthcare utilization, and treatment adherence.

MATERIALS AND METHODS

Individuals diagnosed with an EGID for a minimum of 6 months were recruited from an outpatient university-based gastroenterology practice, via patient advocacy groups (American Partnership for Eosinophilic GI Disorders (APFED), CURED Foundation), and social media (Facebook, Twitter). After obtaining approval from the institutional review board and informed consent from all participants, a series of screening questions were presented to identify participants who met exclusion criteria which may confound the presence of stigma (other GI illness diagnosis, serious mental illness (SMI) diagnosis) and to verify an EGID diagnosis; GI illnesses included IBS and IBD, which are two conditions with identified illness specific stigma; SMI diagnoses included those with psychotic symptoms (schizophrenia, schizoaffective disorder, severe bipolar disorder, depression with psychotic features) and personality disorders. Participants meeting eligibility criteria then completed a series of questionnaires.

Demographic information

Age, gender, marital status, race, ethnicity, education level, employment status, household income, U.S. geographic area, and population of current place of residence were available.

Clinical information

Diagnosis duration (years), symptom duration prior to diagnosis (years), if ever misdiagnosed, number of outpatient appointments in past year, number of current EGID medications, current dietary treatment, number of endoscopic procedures in past year, number of emergency room visits in past year for EGID, dietitian treatment, psychotherapy treatment for EGID, self-rated symptom severity (0–10, 10 = very severe), expectations for symptom flare ups (0–10, 10 = extremely likely), symptom control in past month (0–10, 10 = extremely well), treatment efficacy (0–10, 10 = extremely well), quality of MD–patient relationship (0–10, 10 = excellent), medication adherence in past week (0–100, 100 = exactly as prescribed), dietary treatment adherence in past week (0–100, 100 = exactly as prescribed) were available.

Perceived stigma scale for IBS, PSS–IBS (Modified for EGID)

The PSS–IBS²⁷ is a 20-item measure of perceived stigma toward IBS. The measure was modified for this study by replacing the term ‘IBS’ with ‘EGID’. No other modifications were made. The PSS–IBS measures stigma perceptions for significant others (family, friends, and spouse) and healthcare providers (doctors, nurses, and therapists), with 10 questions under each subscale, on a five-point Likert scale (0 = Never to 5 = Always). Questions include ‘My EGID is not taken seriously’, ‘My EGID is believed to be caused by something I am doing or have done.’ Three scores are calculated by summing responses: significant others score, healthcare providers score, total score. Higher scores denote greater perceived stigma. The PSS–IBS shows excellent internal consistency, split-half reliability, and construct validity in both IBS and IBD studies of perceived stigma.²⁸

Health-related quality of life

The Eosinophilic Esophagitis Quality of Life scale for Adults (EoE-QOL-A)⁷ was used to gauge HRQOL in the study sample. The EoE-QOL-A is a validated measure of disease-specific HRQOL in EoE patients across five domains: eating impact, social impact, emotional impact, disease anxiety, swallowing anxiety. A total HRQOL score is also calculated. Higher scores denote better HRQOL. A modified version of the EoE-QOL-A was given to non-EoE patients with ‘EoE’ changed to ‘EGID’ throughout the questionnaire. Internal consistency for the modified version of the EoE-QOL-A was excellent (Cronbach α = 0.98, Guttman split-half reliability = 0.96).

Impact of illness scale (IIS)

The general effects of the EGID were measured via the IIS,³¹ a nine-item self-report measure rated on a four-point Likert scale. Questions address impacts in activities of daily living, social obligations, mobility, and concentration. The IIS is widely used in health outcomes research and demonstrates excellent reliability and validity.

NIH-PROMIS anxiety and depression scales

The NIH PROMIS³² scales for anxiety and depression evaluated presence of psychological distress in the study sample. The anxiety scale contains seven items and the depression scale has eight items, all ranked on a five-point Likert scale (0 = Never, 4 = Always). NIH PROMIS scales demonstrate excellent internal consistency, split-half reliability, temporal stability, and construct validity.

Data from the online system were exported into SPSS v.22 for analyses. Ineligible participants were removed from the sample prior to analysis. Descriptive statistics evaluated the study sample, including normal distribution of data to identify the need for nonparametric tests. Percentages and mean (SD) are calculated for categorical and continuous variables, respectively. Cronbach alpha and split-half reliability via the Guttman statistic evaluated the reliability of the modified version of the PSS–IBS for the entire sample, as well as by EoE and non-EoE participant samples to ensure no differences existed by diagnoses. Pearson's correlations are used to determine significant relationships between perceived stigma (total, significant other, healthcare provider) and continuous demographic and clinical variables. Additional correlations evaluated the relationship between the three perceived stigma scores and HRQOL, impact of illness, and psychological distress. Independent samples t-Test and one-way ANOVA evaluated differences between categorical demographic and clinical variables for perceived stigma scores. To control for Type 1 error due to multiple comparisons, P was set at ≤0.01 for statistical significance for t-Tests and ANOVAs based on Bonferroni correction.

A series of four linear regression analyses evaluated the predictive relationship between perceived stigma and the patient outcome variables: impact of illness, depression, anxiety, and HRQOL. Clinical and demographic variables that demonstrated significant relationships with perceived stigma were entered into the regression model to control for their influence on the relationship between stigma and patient outcomes. Standardized beta weight, 95% confidence interval, and adjusted R square are reported for each model.

RESULTS

A total of 209 patients consented to the study. Sixty were excluded from final data analyses: five were under 18 years of age, three had not been diagnosed for more than 6 months, 30 had co-morbid GI illness, 11 had co-morbid serious mental illness, and 11 dropped out before completing the surveys (72% completion rate). The final study sample consisted of 149 patients; demographic and clinical information is presented in Tables 1 and 2, respectively.

Table 1.

Demographic characteristics of study sample

Demographic variable	N = 149
Recruitment source
Clinic	9.4% (14)
Online	80.5% (120)
Patient organization	10.1% (15)
Age (Years)	37.3 ± 10.3†
Gender
Male	22.4% (33)
Female	77.6% (114)
Marital status
Married	56.1% (83)
Not married	43.9% (65)
Race
Caucasian	93.8% (137)
Non-Caucasian	6.2% (9)
Ethnicity
Hispanic	3.2% (4)
Non-Hispanic	96.8% (121)
Education
Some college or less	33.3% (49)
College degree or higher	66.7% (98)
Employment
Full time	61.6% (90)
Part time	13.0% (19)
Unemployed	15.8% (23)
Student	9.6% (14)
Household income
<50,000 USD	33.1% (48)
>50,000 USD	54.7% (81)
Did not disclose	12.2% (18)
Living environment
Urban	25.3% (37)
Suburban	52.1% (76)
<10,000 person town	15.1% (22)
< 5,000 person town	7.5% (11)

^†Mean ± standard deviation.

Table 2.

Clinical characteristics of study sample

Clinical variable	N = 149
EGID diagnosis
Esophagitis	77.6% (118)
Gastritis	1.3% (2)
Gastroenteritis	5.9% (9)
Colitis	1.3% (2)
More than 1 EGID	13.8% (21)
Diagnosis duration (Years)	4.0 ± 4.0†
Symptom duration prior to diagnosis (Years)	11.1 ± 11.2
Misdiagnosed in the past	71% (103)
Healthcare utilization
Outpatient visits	8.4 ± 13.1
Endoscopic procedures	2.0 ± 2.2
Emergency Department visits	0.7 ± 1.5
Dietitian (past or present)	45.2% (66)
Therapist (past or present)	15.8% (23)
# Current medications	2.1 ± 2.3
Current dietary treatment	53.4% (78)
Illness self-ratings (out of 10)
Severity	5.7 ± 2.1
Symptom control past month	5.7 ± 2.8
Treatment efficacy	5.5 ± 2.8
MD—patient relationship quality (out of 10)	6.1 ± 3.1
Treatment adherence past week (out of 100)
Medication	79.2 ± 31.4
Diet	72.8 ± 33.4

^†Mean ± standard deviation.

The sample is primarily female, Caucasian, non-Hispanic, with a college education. The mean age is 37.3 (10.3) years. Most participants were diagnosed with Eosinophilic Esophagitis (EoE) (78%), followed by gastroenteritis (6%), with equal representation of gastritis and colitis (1.3% each). Fourteen percent reported having more than one EGID diagnosis. The average time of diagnosis was 4.0 (4.0) years, with 11.1 (11.2) years of symptoms prior to diagnosis. Most participants were recruited from social media sources. Median values for psychological distress, HRQOL, illness impact, and stigma perception are outlined in Table 3.

Table 3.

Psychological distress, quality of life, impact of illness, and stigma perception

	EoE		Non-EoE†
	Median	IQR (range)	Median	IQR (range)
Depression	17.0	13.5 (8–37)	18.5	11.3 (8–32)
Anxiety	17.0	11.0 (7–35)	18.0	10.8 (7–28)
HRQOL	78.5	41.9 (30–147)	76.7	11.2 (37–127)
Impact of illness	16.0	12.0 (9–36)	20.5	6.8 (12–36)
Stigma perception	53.0	26.0 (20–99)	60.0	19.0 (20–85)

^†Eosinophilic gastritis, gastroenteritis, colitis, more than one EGID.

IQR, Interquartile range.

The PSS–IBS modified for EGID demonstrated excellent internal consistency and split-half reliability for each scale (all values ≥ 0.89). Significant, modest correlations exist between levels of perceived stigma and several clinical variables. Patients who report greater perceived stigma overall are more likely to be more recently diagnosed (r = −0.24), report poorer treatment efficacy (r = −0.40) and symptom control (r = −0.29), lower satisfaction with the MD–patient relationship (r = −0.36), and have higher expectations they will experience EGID symptoms regularly (r = 0.32) (all p < .01). Total stigma is also associated with a greater number of outpatient visits (r = 0.19, p < .05) and endoscopies (r = 0.25, p < .01). These relationships remain when stigma perceptions are evaluated by significant others and healthcare providers separately, with the exception that stigma from significant others is not associated with more outpatient appointments. No significant relationships existed for the remaining continuous demographic clinical variables including treatment adherence and EGID symptom severity.

Group differences existed by gender, with women more likely to report perceived stigma from both healthcare providers (t(135) = −3.55, p = .001) and significant others (t(131) = −3.02, p = .003). Participants recruited in the outpatient clinic reported significantly less perceived stigma from healthcare providers than those recruited from patient advocacy groups (F(133, 5) = 3.96, p = .002; Tukey post-hoc p = .008). While not meeting the more stringent significance cutoff applied to this study, participants recruited from social media also reported greater perceived stigma from healthcare providers (F(133, 5) = 3.96, p = .002; Tukey post-hoc p = .02). Other nonsignificant but important findings to consider are that patients diagnosed with non-EoE EGIDS (i.e., gastritis, gastroenteritis, colitis) reported greater perceived stigma from healthcare providers than their EoE counterparts (t(137) = −1.98, p = .05). Patients currently using a dietary therapy reported greater total stigma (t(129) = 2.17, p = .03) than those not on a dietary treatment, and those who had sought psychotherapy for their EGID reported less perceived stigma from significant others than those who had not (t(132) = 2.41, p = .02). No significant differences exist for the remaining categorical demographic or clinical variables. We were unable to compare race and ethnicity due to insufficient subject counts for non-Caucasian and non-Hispanic groups.

Perceived stigma was moderately and significantly (all P < .01) correlated with psychological distress, HRQOL, and illness impact (Figure 1). Greater perceived stigma positively correlated with greater illness impact (r = 0.58), increased depression (r = 0.55), and increased anxiety (r = 0.58). Conversely, greater stigma is associated with reduced HRQOL (r = −0.52). Specifically, emotional functioning is poorest in patients reporting higher stigma perception (r = −0.57), followed by disease-related anxiety (r = −0.45), eating/diet concerns (r = −0.40), swallowing anxiety (r = −0.33), and social function (r = −0.29).

Fig. 1.

Scatterplot matrix of perceived stigma and four patient outcome variables. (A) Anxiety (B) Depression (C) Impact of Illness (D) HRQOL.

Stepwise regression analyses demonstrated that perceived stigma was the strongest predictor of each patient outcome variable when controlling for gender, diagnosis duration, symptom control, treatment efficacy, MD relationship quality, and expectations for flare (Table 4). Standardized beta weights range from 0.40 (HRQOL) to 0.58 (Depression) and R² adjusted values suggest perceived stigma accounts for 29% of the variance in HRQOL, 31% of the variance in depression, 34% of the variance in anxiety, and 36% of the variance in illness impact. Flare expectations and gender demonstrated significant but very small predictive relationships with illness impact and HRQOL while diagnosis duration and MD relationship quality also demonstrated similar relationships with psychological distress; symptom control and treatment efficacy were nonsignificant and removed from the models.

Table 4.

Stepwise linear regression for perceived stigma as predictor of patient outcome variables

	Standardized β	95% CI	R 2adjusted	P
Impact of Illness
Perceived stigma	0.48	0.12–0.23	0.36	.000
Flare expectation	0.21	0.17–0.98	0.04	.005
Gender	0.16	0.22–4.77	0.02	.032
Depression
Perceived stigma	0.58	0.17 – 0.30	0.31	.000
Years diagnosed	−0.19	−0.65–−0.07	0.03	.030
MD relationship	0.16	0.01–0.83	0.02	.045
Anxiety
Perceived stigma	0.54	0.14–0.25	0.34	.000
Years diagnosed	−0.17	−0.53–−0.03	0.03	.027
HRQOL
Perceived stigma	−0.40	−0.72–−0.31	0.29	.000
Flare expectation	−0.22	−3.66–−0.68	0.05	.004
Gender	−0.18	−19.09–−1.68	0.03	.020

DISCUSSION

The current study demonstrates the presence of stigma in people with an EGID as well as the subsequent influence on the health-related and psychosocial aspects of their lives. The negative impact of perceived stigma on HRQOL and other psychosocial variables has been noted throughout literature in IBS and IBD.^27–30 Therefore, we anticipated the relationship between perceived stigma and HRQOL in people with an EGID to be similar. As expected, greater stigma is associated with, and may predict approximately 30% of the variance in, lower HRQOL, specifically emotional functioning, disease-related anxiety, and eating/diet concerns, as well as increased anxiety and depression. In addition, perceived impact of illness is positively associated with perceived stigma with a similar predictive relationship, indicating that as the perception of stigma increases so does the apparent impact of the EGID on their lives.

We examined if perceived stigma differs depending on the length of time since diagnosis. Results indicated those more recently diagnosed perceive greater stigma. Prior research in stigma and IBD is conflicted. One study examining stigma in patients with IBS and IBD found no relationship between stigma and duration of diagnosis in IBS patients.²⁹ Similar results were found for the IBD patients, excluding those diagnosed 11–20 years, in which stigma began to decrease.²⁹ Conversely, a study interested in perceived stigma and IBD found a positive relationship.²⁸ Based on our results, the first few years of illness may be critical to the development and perception of stigma in people with an EGID, underscoring the importance of healthcare providers evaluating stigma early on. However, we cannot conclude that perceived stigma subsides over time. Therefore, future longitudinal research should investigate the duration of stigma and long-term impact.

Interestingly, perceived stigma was independent of treatment adherence. Thus, people with an EGID remain on a treatment regimen regardless of the presence of stigma. These findings are contrary to literature in HIV^33–35 and epilepsy,³⁶ in which stigma is associated with poorer treatment adherence. Conflicting findings may be attributed to the difference in nature of EGIDs compared to HIV and epilepsy. However, even when evaluating this relationship in the scope of GI-related illnesses, perceived stigma was a predictor for decreased medication adherence in people with IBD.²⁸ Therefore, despite the occurrence of both EGIDs and IBD in the GI tract, the impact of stigma on treatment adherence differs. One potential explanation is people with an EGID have increased motivation that drives them to be different from people with IBD or other chronic illnesses where stigma does influence treatment adherence. Dietary modification is an effective component of the treatment regimen for people with an EGID. Despite the vigorous nature of the diet, patients report continued adherence. Perhaps the experience of food impaction caused by EGID is enough to motivate people to remain on the diet, regardless of stigma. Nevertheless, further research examining the relationship between stigma and treatment adherence and what is differentiating people with EGIDs from other chronic illnesses is warranted.

Greater stigma was associated with poorer treatment efficacy. EGID dietary treatments can be difficult, especially in social situations. This could encourage negative perceptions of treatment, which, in turn, may influence the way the person feels their treatment is working. Another explanation is the presence of depression and anxiety serving as a potential moderator between stigma and treatment efficacy. Psychological factors could influence cognitions, leading to distorted perceptions of treatment effectiveness, and amplified perceptions of stigma. As such, it is important for physicians to be aware of these psychosocial influences and the effects they can have on patients diagnosed with EGIDs. Additionally, psychological intervention should be utilized to address any psychosocial influences.

Greater perceived stigma was associated with an increase in healthcare utilization, more specifically, outpatient visits, and endoscopies. While there are similar findings in the IBS population,³⁰ one study investigating a variety of chronic illnesses found anticipated stigma resulted in less healthcare utilization.¹³ The nature and newness of EGIDs may explain the increase in healthcare utilization. A previous study assessing outcomes in patients with EoE found people report concern about the novelty of the disorder.⁶ Given that EGIDs are relatively new and rare, it is possible people seek out additional healthcare services in an effort to validate and gather more information about the disorder. Continuing education for both the patient and physician is important to promote understanding of the disorder as well as drive down costs associated with increased healthcare utilization.

Expectedly, non-EoE EGIDs patients report more stigma than EoE patients. EoE is the most common and most studied type of EGID.^1,2 Therefore, people with non-EoE EGIDs may feel further stigmatized due to the lack of attention and knowledge surrounding their disorder compared to their EoE counterparts. Additionally, our sample was heavily weighted toward respondents with EoE (77.6%), making the findings more generalizable to the EoE subset. As a result, future research on non-EoE EGIDs is essential to understanding the different disorders and patient experience. Individuals on a treatment diet reported greater total stigma than people who did not participate in the diet. The diet requires a significant modification of eating habits resulting in a substantial lifestyle change. Not only do individuals have to change their daily dietary consumption, social aspects of their lives such as going out to eat with friends may become difficult. Therefore, in addition to experiencing perceived stigma, individuals who are on a treatment diet may experience psychological distress. As a result, it is important for healthcare professionals to be aware individuals participating in the diet may need additional support. Finally, individuals who sought psychotherapy reported less perceived stigma from significant others than those who did not. There is strong evidence supporting the efficacy of psychological therapies for treatment in IBS,^37–39 including reducing stigma perception.³⁰ There is also evidence supporting the notion that psychological therapies can decrease healthcare utilization amongst patients with Crohn's Disease.⁴⁰ However, only 16% of participants in the current study reported they have utilized a therapist; suggesting future research should incorporate more people to acquire more balanced samples between the groups. The aforementioned findings, although significant, should be interpreted with caution due to possible type 1 error and further research is warranted.

There are several limitations to the current study. The cross-sectional design limits the ability to make cause and effect inferences about the relationships between variables. The majority of the sample was recruited online (80.5%) through social media. Although screening measures are in place, we are unable to confirm EGID diagnosis, the absence of other GI illness, or the absence of serious mental illness. It should also be noted that experimental control is decreased for those recruited online compared to those recruited from a clinic. Finally, patients recruited from the clinic reported less perceived stigma from physicians than those who were recruited through online sources, indicating that responses from clinic patients may have been influenced by the Hawthorne effect. As a result, future studies should attempt to obtain an even distribution of participations between clinic, online, and patient organizations. The current study did not include a control group to compare the results to. Including a control group that consists of a different GI illness, such as IBS or IBD, could strengthen the findings and provide the ability to make comparisons between GI illnesses. The sample was predominately Caucasian (93.8%) and female (77.6%), which is a limitation of generalizability across different racial and gender groups. As such, healthcare professionals should exercise caution when relating these findings to patients of non-Caucasian racial groups and to males. Further, while the majority of participants in the study were female, EoE predominately occurs in males. One explanation for this is that participants were recruited from online sources, which tends to have more female participation. In addition, females are more likely to seek support than males. Therefore, future research should attempt to recruit a more diverse sample. As previously discussed, our sample was biased toward patients with EoE. Therefore, an analysis with increased non-EoE respondents is needed to make the analysis more robust. Future research on EGID-related stigma is warranted.