Key Points
Question
Is granuloma annulare (GA) associated with type 2 diabetes, hyperlipidemia, autoimmune conditions, and hematologic malignant neoplasms?
Findings
In this population-based cohort study of 5137 individuals with granuloma annulare and 51 169 controls, individuals with GA were more likely than those without GA to have baseline diabetes (21% vs 13%) and hyperlipidemia (33% vs 28%) as well as baseline hypothyroidism (14% vs 11%) and incident hypothyroidism (0.8% vs 0.5%). There was no association between GA and an increased risk of hematologic malignant neoplasms.
Meaning
This findings of this study suggest that diabetes and hyperlipidemia may be risk factors for the development of GA and that autoimmunity may play a role in the pathogenesis of GA.
Abstract
Importance
Although granuloma annulare (GA) has been associated with several other conditions, these studies have been limited by single-center designs and small sample sizes.
Objective
To evaluate whether there is an association between GA and type 2 diabetes, hyperlipidemia, autoimmune conditions, and hematologic malignant neoplasms, using a large population-based cohort study.
Design, Setting, and Participants
This retrospective cohort study conducted between January 1, 2016, and June 30, 2019, used deidentified data from the US Optum Clinformatics Data Mart Database. A total of 5137 patients with GA were matched by age and sex with up to 10 randomly selected controls (n = 51 169) with a diagnosis of a nevus or seborrheic keratosis.
Main Outcomes and Measures
Logistic regression was used to evaluate for potential associations between GA and diabetes, hyperlipidemia, autoimmune conditions, and hematologic malignant neoplasms. All analyses were adjusted for race/ethnicity, income, and educational level.
Results
This study included 5137 individuals with GA (3760 women [73.2%]; mean [SD] age, 57.7 [19.0] years) and 51 169 controls (37 456 women [73.2%]; mean [SD] age, 57.7 [19.0] years). Those with GA were more likely than controls to have baseline diabetes (1086 [21.1%] vs 6780 [13.3%]; adjusted odds ratio [aOR], 1.67; 95% CI, 1.55-1.80), hyperlipidemia (1669 [32.5%] vs 14 553 [28.4%]; aOR, 1.15; 95% CI, 1.08-1.23), hypothyroidism (727 [14.2%] vs 5780 [11.3%]; aOR, 1.24; 95% CI, 1.15-1.36), and rheumatoid arthritis (62 [1.2%] vs 441 [0.9%]; aOR, 1.34; 95% CI, 1.02-1.75). Those with GA were more likely to have incident diabetes (144 [2.8%] vs 1061 [2.1%]; aOR, 1.31; 95% CI, 1.10-1.57), hypothyroidism (41 [0.8%] vs 252 [0.5%]; aOR, 1.59; 95% CI, 1.14-2.22), systemic lupus erythematosus (21 [0.4%] vs 65 [0.1%]; aOR, 3.06; 95% CI, 1.86-5.01), and rheumatoid arthritis (26 [0.5%] vs 122 [0.2%]; aOR, 2.05; 95% CI, 1.34-3.13). There was no association between GA and an increased risk of hematologic malignant neoplasms.
Conclusions and Relevance
This population-based cohort study identified associations between GA and baseline diabetes and hyperlipidemia as well as between GA and both baseline and incident autoimmune conditions. These findings suggest that diabetes and hyperlipidemia may be risk factors for the development of GA and that autoimmunity may be an important factor in the pathogenesis of GA.
This cohort study evaluates whether there is an association between granuloma annulare and type 2 diabetes, hyperlipidemia, autoimmune conditions, and hematologic malignant neoplasms.
Introduction
Granuloma annulare (GA) is characterized by ringed, erythematous plaques. Most patients have localized disease, but 15% to 25% will have more generalized skin lesions, and subcutaneous variants have also been described.1,2,3,4 The pathogenesis of GA is poorly understood.5 Several potential mechanisms have been proposed, including immune-mediated chronic microvascular damage and cell-mediated immune dysfunction.6,7,8,9,10,11,12
Although GA may occur as an isolated disease limited to the skin, several potential associations have been described, including diabetes and hyperlipidemia,4,13,14,15,16,17,18,19 autoimmune conditions such as thyroid disease,20,21 and hematologic malignant neoplasms.5,21 However, other studies have found no association between these conditions and GA.18,22,23,24 In addition, these studies have often been limited by small sample sizes, single-center designs, and suboptimal control groups such as patients with psoriasis, which is known to be associated with several of these conditions.5,22,25 These studies have also predominantly focused on associations present at the time of diagnosis of GA and have not explored associations with incident disease after the initial diagnosis of GA. The purpose of this study was to evaluate whether there is an association between GA and type 2 diabetes, hyperlipidemia, autoimmune conditions, and hematologic malignant neoplasms, using a large population-based cohort study.
Methods
Data Set Source
This retrospective cohort study conducted between January 1, 2016, and June 30, 2019, used deidentified data from the Optum Clinformatics Data Mart Database. The Optum Clinformatics Data Mart includes deidentified commercial claims data for approximately 12 million to 14 million covered individuals in the US annually. These data include both medical and pharmacy claims as well as patient demographic information, such as age, sex, race/ethnicity, income, and educational level. The demographic characteristics of the patient population available in the Optum Clinformatics Data Mart are similar to the those of the US population with respect to sex, age, and geographical distribution.26 Optum derives socioeconomic data from health information deterministically linked to data licensed from a consumer data vendor. A member’s ethnicity is derived by using their name and geographic location. Once the ethnicity is determined, the member is mapped to 1 of 5 race/ethnicity categories (Asian, Black, Hispanic, White, or unknown or other). This study was approved by the institutional review board of the University of Pennsylvania, and patient consent was deemed exempt because the data were deidentified. It was conducted in adherence with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline and the Reporting of Studies Conducted Using Observational Routinely Collected Health Data (RECORD) guideline.27
Study Population and Outcomes
Our study population included patients with at least 1 International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) code for GA (code L92.0). We have previously validated the use of ICD-10 codes to identify patients with GA, finding that a single code of L92.0 has a positive predictive value of 94% among dermatologists and 82% among internists.28 We defined the index date as the date of the first encounter with a diagnosis for GA.
We created a comparator cohort composed of patients with a diagnosis of a benign skin lesion (nevus or seborrheic keratosis), with the index date defined as the first encounter with this diagnosis. Because ICD-10 codes were implemented in October 2015 in the US, to allow for uptake, our study period was between January 1, 2016, and June 30, 2019.29 To enable accurate outcome identification, all patients were required to have at least 1 year of continuous enrollment in the database before and after the index date.
Based on prior putative associations with GA, we focused our analysis on whether there is an association between GA and disorders associated with metabolic syndrome, such as diabetes, hyperlipidemia, and hypertension; autoimmune conditions, such as hypothyroidism, alopecia areata, systemic lupus erythematosus and rheumatoid arthritis; and hematologic malignant neoplasms. We considered both baseline associations, defined as the presence of any encounter with a diagnosis of the comorbidity in the 1 year before the index date (does not need to be a new diagnosis), and incident associations, defined as the presence of a new diagnosis of the comorbidity in the first year after the index date.
Diabetes was defined by at least 1 ICD-10 code (E11.x and E13.x), which have previously been validated to accurately classify patients with diabetes using administrative data.30 Hyperlipidemia was defined by at least 1 ICD-10 code for hyperlipidemia (E78.5), which has been shown to have high positive predictive value to identify patients with hyperlipidemia.31 Hypothyroidism was defined by at least 1 ICD-10 code (E03.9) and a prescription for levothyroxine.32,33 Systemic lupus erythematosus (code M32.x), rheumatoid arthritis (codes M05.x and M06.x), alopecia areata (codes L63.1, L63.8, and L63.9), lymphoma (codes C81-85.x, C88.x, C90.x, and C96.x), and leukemia (codes C91-95.x) were identified as previously described.33,34
Statistical Analysis
Patients with GA were matched by age and sex with up to 10 randomly selected controls from the comparator cohort. Logistic regression was used to evaluate for potential associations between GA and the comorbidities of interest. All analyses were adjusted for race/ethnicity, income, and educational level. In addition, for comparisons with baseline and incident comorbidities, these analyses were adjusted for the presence of a visit to an internist in the year before the index date. For comparisons with incident comorbidities, analyses were also adjusted for the presence of a visit to an internist in the year after the index date. The presence of a visit to an internist was included to control for potential differences with respect to interaction with the health care system that may be associated with the likelihood of receiving a diagnosis of 1 of these comorbidities.
Because use of relevant testing (ie, hemoglobin A1c for diabetes, lipid panel for hyperlipidemia, thyroid-stimulating hormone for hypothyroidism, rheumatoid factor for rheumatoid arthritis, and complete blood cell count for hematologic malignant neoplasms) may be associated with the rates of detection of our outcomes of interest, we also conducted a sensitivity analysis, adjusting for the use of such testing. However, because this testing is likely to be highly correlated with our outcomes because it is often part of the diagnostic criteria for these conditions and used as part of treatment monitoring, we did not include these variables in our primary model. Statistical analyses were performed using Stata, version 16.1 (StataCorp). All P values were from 2-sided tests and results were deemed statistically significant at P < .05.
We considered a causal model with 3 pathways. In the first pathway, there is an underlying factor that is associated with both the development of GA and the comorbidity. In this pathway, we would expect to find an association between GA and both higher baseline prevalence and incidence of the comorbidity of interest. In the second pathway, GA increases the likelihood of developing the comorbidity of interest. In this pathway, we would expect to find an association only between GA and higher incidence, but not baseline prevalence, of the comorbidity of interest. In the third pathway, the comorbidity of interest increases the likelihood of developing GA. In this pathway, we would expect to find an association only between GA and higher baseline prevalence, but not incidence, of the comorbidity of interest (Figure).
Figure. Summary of Potential Associations Between Variables.
Pathway 1 (blue lines): an underlying factor predisposes individuals to developing both the comorbidity and granuloma annulare. In this scenario, we would expect to find an association between granuloma annulare and both higher baseline prevalence and incidence of the comorbidity of interest. Pathway 2 (red line): the comorbidity predisposes individuals to developing granuloma annulare. In this scenario, we would expect to find an association between granuloma annulare and higher incidence of the comorbidity of interest. Pathway 3 (green line): granuloma annulare predisposes individuals to developing the comorbidity. In this scenario, we would expect to find an association between granuloma annulare and higher baseline prevalence of the comorbidity of interest.
Results
We identified 5137 patients with GA (3760 women [73.2%]; mean [SD] age, 57.7 [19.0] years) who were matched exactly by age and sex with 51 169 controls (37 456 women [73.2%]; mean [SD] age, 57.7 [19.0] years) with an encounter with a diagnosis of nevus or seborrheic keratosis. The GA and control cohorts were also well matched with respect to race/ethnicity, educational level, income, and encounters with internists (Table 1).
Table 1. Demographic Characteristics of Study Cohorts.
| Characteristic | Participants, No. (%) | |
|---|---|---|
| Granuloma annulare (n = 5137) | Control (nevus or seborrheic keratosis; n = 51 169) | |
| Age, mean (SD), y | 57.7 (19.0) | 57.7 (19.0) |
| Female | 3760 (73.2) | 37 456 (73.2) |
| Male | 1377 (26.8) | 13 713 (26.8) |
| Race/ethnicity | ||
| White | 3842 (74.8) | 38 991 (76.2) |
| Black | 262 (5.1) | 2200 (4.3) |
| Asian | 82 (1.6) | 921 (1.8) |
| Hispanic | 324 (6.3) | 2814 (5.5) |
| Unknown or other | 632 (12.3) | 6243 (12.2) |
| Educational level | ||
| Less than 12th grade | 10 (0.2) | 51 (0.1) |
| High school diploma | 920 (17.9) | 7880 (15.4) |
| Less than bachelor degree | 2651 (51.6) | 25 994 (50.8) |
| Bachelor degree plus | 1125 (21.9) | 12 792 (25.0) |
| Unknown or other | 432 (8.4) | 4503 (8.8) |
| Income, $ | ||
| <40 000 | 627 (12.2) | 5270 (10.3) |
| 40 000-49 999 | 241 (4.7) | 1996 (3.9) |
| 50 000-59 999 | 308 (6.0) | 2558 (5.0) |
| 60 000-74 999 | 426 (8.3) | 4145 (8.1) |
| 75 000-99 999 | 714 (13.9) | 7112 (13.9) |
| ≥100 000 | 1767 (34.4) | 20 007 (39.1) |
| Unknown or other | 1058 (20.6) | 10 080 (19.7) |
| Had ≥1 internist encounter | ||
| Before the index date | 4474 (87.1) | 43 340 (84.7) |
| After the index date | 4541 (88.4) | 44 415 (86.8) |
Those with GA were more likely than controls to have baseline diabetes (1086 [21.1%] vs 6780 [13.3%]; adjusted odds ratio [aOR], 1.67; 95% CI, 1.55-1.80), hyperlipidemia (1669 [32.5%] vs 14 553 [28.4%]; aOR, 1.15; 95% CI, 1.08-1.23), and hypertension (2270 [44.2%] vs 20 183 [39.4%]; aOR, 1.14; 95% CI, 1.07-1.21) (Table 2). Those with GA were more likely than controls to have incident diabetes (144 [2.8%] vs 1061 [2.1%]; aOR, 1.31; 95% CI, 1.10-1.57), although the strength of this association was weaker than that of baseline diabetes. In contrast, there were no significant associations between GA and incident hyperlipidemia or hypertension. In addition, the association between GA and baseline hypertension was no longer significant after adjusting for the presence of baseline diabetes and hyperlipidemia (aOR, 1.00; 95% CI, 0.93-1.07).
Table 2. Association Between Granuloma Annulare and Diabetes, Hyperlipidemia, and Hypertension.
| Comorbidity | Control (nevus or seborrheic keratosis; n = 51 169), No. (%) | Granuloma annulare (n = 5137) | ||
|---|---|---|---|---|
| No. (%) | Odds ratio | |||
| Crude | Adjusteda | |||
| Diabetes | ||||
| Baselineb | 6780 (13.3) | 1086 (21.1) | 1.75 (1.63-1.88) | 1.67 (1.55-1.80) |
| Incidentc | 1061 (2.1) | 144 (2.8) | 1.37 (1.15-1.63) | 1.31 (1.10-1.57) |
| Hyperlipidemia | ||||
| Baseline | 14 553 (28.4) | 1669 (32.5) | 1.21 (1.14-1.29) | 1.15 (1.08-1.23) |
| Incident | 4135 (8.1) | 442 (8.6) | 1.07 (0.97-1.19) | 1.04 (0.94-1.15) |
| Hypertension | ||||
| Baseline | 20 183 (39.4) | 2270 (44.2) | 1.22 (1.15-1.29) | 1.14 (1.07-1.21) |
| Incident | 2839 (5.6) | 293 (5.7) | 1.03 (0.91-1.17) | 1.01 (0.90-1.15) |
Adjusted for race/ethnicity, income, and educational level. For comparisons with baseline and incident comorbidities, these analyses were adjusted for the presence of a visit to an internist. For comparisons with incident comorbidities, analyses were also adjusted for the presence of a visit to an internist in the year after the index date.
Baseline is defined as 1 year prior to index date.
Incident is defined as 1 year after index date.
With respect to autoimmune conditions, those with GA were more likely than controls to have baseline hypothyroidism (727 [14.2%] vs 5780 [11.3%]; aOR, 1.24; 95% CI, 1.15-1.36) and rheumatoid arthritis (62 [1.2%] vs 441 [0.9%]; aOR, 1.34; 95% CI, 1.02-1.75) (Table 3). Granuloma annulare was not associated with baseline systemic lupus erythematosus (aOR, 1.36; 95% CI, 0.94-1.98). In addition, those with GA were more likely than controls to have incident hypothyroidism (41 [0.8%] vs 252 [0.5%]; aOR, 1.59; 95% CI, 1.14-2.22), systemic lupus erythematosus (21 [0.4%] vs 65 [0.1%]; aOR, 3.06; 95% CI, 1.86-5.01), and rheumatoid arthritis (26 [0.5%] vs 122 [0.2%]; aOR, 2.05; 95% CI, 1.34-3.13). No significant associations were noted between GA and alopecia areata, although there were few events in either group.
Table 3. Association Between Granuloma Annulare and Autoimmune Disorders.
| Autoimmune disorder | Control (nevus or seborrheic keratosis; n = 51 169), No. (%) | Granuloma annulare (n = 5137) | ||
|---|---|---|---|---|
| No. (%) | Odds ratio | |||
| Crude | Adjusteda | |||
| Hypothyroidism | ||||
| Baselineb | 5780 (11.3) | 727 (14.2) | 1.29 (1.19-1.41) | 1.24 (1.15-1.36) |
| Incidentc | 252 (0.5) | 41 (0.8) | 1.63 (1.17-2.26) | 1.59 (1.14-2.22) |
| Alopecia areata | ||||
| Baseline | 131 (0.3) | 16 (0.3) | 1.22 (0.72-2.05) | 1.22 (0.73-2.06) |
| Incident | 49 (0.1) | 5 (0.1) | 1.02 (0.40-2.55) | 1.02 (0.41-2.57) |
| Systemic lupus erythematosus | ||||
| Baseline | 223 (0.4) | 32 (0.6) | 1.43 (0.99-2.08) | 1.36 (0.94-1.98) |
| Incident | 65 (0.1) | 21 (0.4) | 3.23 (1.97-5.28) | 3.06 (1.86-5.01) |
| Rheumatoid arthritis | ||||
| Baseline | 441 (0.9) | 62 (1.2) | 1.40 (1.08-1.84) | 1.34 (1.02-1.75) |
| Incident | 122 (0.2) | 26 (0.5) | 2.13 (1.39-3.25) | 2.05 (1.34-3.13) |
Adjusted for race/ethnicity, income, and educational level. For comparisons with baseline and incident comorbidities, these analyses were adjusted for the presence of a visit to an internist in the year prior to the index date. For comparisons with incident comorbidities, analyses were also adjusted for the presence of a visit to an internist in the year after the index date.
Baseline is defined as 1 year prior to index date.
Incident is defined as 1 year after index date.
With respect to hematologic malignant neoplasms, there was no significant association between GA and leukemia or lymphoma (Table 4). In our sensitivity analyses including relevant diagnostic tests as potential confounding variables, similar results were noted for associations between GA and diabetes, hyperlipidemia, autoimmune disorders, and hematologic malignant neoplasms (eTable in the Supplement).
Table 4. Association Between Granuloma Annulare and Hematologic Malignant Neoplasms.
| Malignant neoplasm | Control (nevus or seborrheic keratosis; n = 51 169), No. (%) | Granuloma annulare (n = 5137) | ||
|---|---|---|---|---|
| No. (%) | Odds ratio | |||
| Crude | Adjusteda | |||
| Leukemia | ||||
| Baselineb | 195 (0.4) | 24 (0.5) | 1.23 (0.80-1.88) | 1.19 (0.78-1.82) |
| Incidentc | 49 (0.1) | 4 (0.1) | 0.81 (0.29-2.25) | 0.75 (0.27-2.08) |
| Lymphoma | ||||
| Baseline | 304 (0.6) | 23 (0.5) | 0.75 (0.49-1.15) | 0.72 (0.47-1.10) |
| Incident | 106 (0.2) | 15 (0.3) | 1.41 (0.82-2.42) | 1.35 (0.78-2.32) |
| Any hematologic malignant neoplasm | ||||
| Baseline | 477 (0.9) | 41 (0.8) | 0.86 (0.62-1.18) | 0.82 (0.60-1.13) |
| Incident | 147 (0.3) | 19 (0.4) | 1.29 (0.80-2.08) | 1.21 (0.75-1.97) |
Adjusted for race/ethnicity, income, and educational level. For comparisons with baseline and incident comorbidities, these analyses were adjusted for the presence of a visit to an internist in the year prior to the index date. For comparisons with incident comorbidities, analyses were also adjusted for the presence of a visit to an internist in the year after the index date.
Baseline is defined as 1 year prior to index date.
Incident is defined as 1 year after index date.
Discussion
In this large, population-based cohort study, we identified associations between GA and diabetes and hyperlipidemia as well as between GA and autoimmune conditions. Consistent with several other studies, we did not identify any association between GA and increased risk of hematologic malignant neoplasms.18,35
Our findings add to the growing body of literature supporting an association between GA and diabetes and hyperlipidemia.4,13,14,15,16,17,18 Similar to a recent case-control study of patients presenting to the Johns Hopkins Hospital System, we found an association between GA and baseline diabetes.18 These findings are also consistent with prior small studies from Korea, Taiwan, and Iran.4,16,17 In addition, similar to a small, single-center case-control study, we also identified an association between GA and baseline hyperlipidemia.19 Given these consistent associations, it may be prudent to screen patients with GA for diabetes and hyperlipidemia.
We also identified an association between GA and baseline hypertension, which, to our knowledge, has not previously been reported. However, this finding was not significant after adjusting for the presence of baseline diabetes and hyperlipidemia, suggesting that diabetes and hyperlipidemia may be associated with both GA and hypertension and that hypertension is not independently associated with GA.
Because we identified that GA is more strongly associated with baseline diabetes and hyperlipidemia, rather than incident diabetes and hyperlipidemia, these findings suggest that diabetes and hyperlipidemia may be predisposing factors for the development of GA (Figure). Diabetes is associated with elevated inflammatory cytokines, such as interleukin 6, as well as T-cell and macrophage activation.36,37 In addition, hyperlipidemia can promote proliferative CD4+ T-cell responses and stronger antigen reactions.38 Dysregulation of lipid metabolism has also been associated with autoimmunity.39 Given that it has been suggested that GA may be a CD4+ T-cell–mediated disease, diabetes and hyperlipidemia may act as risk factors for GA through dysregulated T-cell activity.12 In addition, just as there has been interest in exploring the potential role of statins in the treatment of psoriasis, there may be value to considering statins as a potential treatment for GA.40
We identified several associations between GA and both baseline and incident autoimmune disorders, suggesting that there may be an underlying autoimmune predisposition in patients with GA (Figure). The association between GA and hypothyroidism has been inconsistent in prior studies owing to small sample sizes with few events in each group.17,18,20,21 The large sample size in our cohort study, with more than 500 events in each group, enables us to better characterize the potential association between GA and hypothyroidism.
In addition, we identified novel associations with other autoimmune disorders, such as systemic lupus erythematosus and rheumatoid arthritis, strengthening the association between autoimmune disorders and GA. Furthermore, GA has been reported in the setting of treatment with immune checkpoint inhibitors, which provides additional support for autoimmunity as a key component of the pathogenesis of GA.41,42 Analysis of GA lesions has identified increased Janus kinase-signal transducer activation, and tofacitinib has been used successfully in a case series of 5 patients with recalcitrant disease.11
Limitations
This study has some limitations, and the results should be interpreted in the context of its design. Despite our large sample size with more than 5000 patients with GA matched with more than 50 000 controls, for some of the rarer conditions evaluated, such as alopecia areata, systemic lupus erythematosus, and hematologic malignant neoplasms, we were limited by the frequency of events, reducing our power to detect potential associations. Even larger studies may be needed to fully characterize the potential association between GA and these conditions. Given the use of claims data without clinicohistopathologic confirmation of the diagnosis, there is the potential for misclassification bias, although the use of previously validated classification strategies for both GA and potential associated conditions reduces this risk.28,30,31,32,33,34 Although it is possible that disorders associated with systemic lupus erythematosus and rheumatoid arthritis, such as interstitial granulomatous dermatitis and palisaded neutrophilic granulomatous dermatitis, could be misclassified as GA, we did not identify any instances of this type of misclassification in our validation study, and these disorders are rare and unlikely to be able to account for the associations between GA and these autoimmune conditions. Because many of these associated conditions can be asymptomatic, differences in health care use between cohorts could result in differences in the identification of these comorbidities. This confounding is supported by the attenuation of the magnitude of our identified associations after controlling for health care use. However, we attempted to control for various factors that could be associated with the identification of comorbidities, including internist encounters, relevant testing use (in sensitivity analyses), and sociodemographic factors. Given the nature of claims data, we are unable to evaluate differences between subtypes of GA, and future studies are needed to evaluate whether associations with comorbidities may differ among subtypes of GA.
Conclusions
In this population-based cohort study, we identified associations between GA and baseline diabetes and hyperlipidemia as well as between GA and both baseline and incident autoimmune conditions. These findings suggest that diabetes and hyperlipidemia may be risk factors for the development of GA and that autoimmunity may be an important factor in the pathogenesis of GA.
eTable. Sensitivity Analysis Including Adjustment for Relevant Diagnostic Testing
References
- 1.Felner EI, Steinberg JB, Weinberg AG. Subcutaneous granuloma annulare: a review of 47 cases. Pediatrics. 1997;100(6):965-967. doi: 10.1542/peds.100.6.965 [DOI] [PubMed] [Google Scholar]
- 2.Piette EW, Rosenbach M. Granuloma annulare: clinical and histologic variants, epidemiology, and genetics. J Am Acad Dermatol. 2016;75(3):457-465. doi: 10.1016/j.jaad.2015.03.054 [DOI] [PubMed] [Google Scholar]
- 3.Dabski K, Winkelmann RK. Generalized granuloma annulare: clinical and laboratory findings in 100 patients. J Am Acad Dermatol. 1989;20(1):39-47. doi: 10.1016/S0190-9622(89)70005-0 [DOI] [PubMed] [Google Scholar]
- 4.Yun JH, Lee JY, Kim MK, et al. Clinical and pathological features of generalized granuloma annulare with their correlation: a retrospective multicenter study in Korea. Ann Dermatol. 2009;21(2):113-119. doi: 10.5021/ad.2009.21.2.113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Piette EW, Rosenbach M. Granuloma annulare: pathogenesis, disease associations and triggers, and therapeutic options. J Am Acad Dermatol. 2016;75(3):467-479. doi: 10.1016/j.jaad.2015.03.055 [DOI] [PubMed] [Google Scholar]
- 6.Dahl MV, Ullman S, Goltz RW. Vasculitis in granuloma annulare: histopathology and direct immunofluorescence. Arch Dermatol. 1977;113(4):463-467. doi: 10.1001/archderm.1977.01640040071010 [DOI] [PubMed] [Google Scholar]
- 7.Friedman-Birnbaum R, Ludatscher RM. Comparative ultrastructural study of generalized and localized granuloma annulare. Am J Dermatopathol. 1986;8(4):302-308. doi: 10.1097/00000372-198608000-00005 [DOI] [PubMed] [Google Scholar]
- 8.Umbert P, Winkelmann RK. Histologic, ultrastructural and histochemical studies of granuloma annulare. Arch Dermatol. 1977;113(12):1681-1686. doi: 10.1001/archderm.1977.01640120049007 [DOI] [PubMed] [Google Scholar]
- 9.Umbert P, Belcher RW, Winkelmann RK. Lymphokines (MIF) in the serum of patients with sarcoidosis and cutaneous granuloma annulare. Br J Dermatol. 1976;95(5):481-485. doi: 10.1111/j.1365-2133.1976.tb00857.x [DOI] [PubMed] [Google Scholar]
- 10.Bergman R, Pam Z, Lichtig C, Reiter I, Friedman-Birnbaum R. Localized granuloma annulare: histopathological and direct immunofluorescence study of early lesions, and the adjacent normal-looking skin of actively spreading lesions. Am J Dermatopathol. 1993;15(6):544-548. doi: 10.1097/00000372-199312000-00005 [DOI] [PubMed] [Google Scholar]
- 11.Wang A, Rahman N-T, McGeary MK, et al. Treatment of granuloma annulare and suppression of proinflammatory cytokine activity with tofacitinib. J Allergy Clin Immunol. Published online December 5, 2020. doi: 10.1016/j.jaci.2020.10.012 [DOI] [PubMed] [Google Scholar]
- 12.Min MS, Wu J, He H, et al. Granuloma annulare skin profile shows activation of T-helper cell type 1, T-helper cell type 2, and Janus kinase pathways. J Am Acad Dermatol. 2020;83(1):63-70. doi: 10.1016/j.jaad.2019.12.028 [DOI] [PubMed] [Google Scholar]
- 13.Haim S, Friedman-Birnbaum R, Haim N, Shafrir A, Ravina A. Carbohydrate tolerance in patients with granuloma annulare: study of fifty-two cases. Br J Dermatol. 1973;88(5):447-451. doi: 10.1111/j.1365-2133.1973.tb15448.x [DOI] [PubMed] [Google Scholar]
- 14.Muhlemann MF, Williams DR. Localized granuloma annulare is associated with insulin-dependent diabetes mellitus. Br J Dermatol. 1984;111(3):325-329. doi: 10.1111/j.1365-2133.1984.tb04730.x [DOI] [PubMed] [Google Scholar]
- 15.Veraldi S, Bencini PL, Drudi E, Caputo R. Laboratory abnormalities in granuloma annulare: a case-control study. Br J Dermatol. 1997;136(4):652-653. doi: 10.1111/j.1365-2133.1997.tb02180.x [DOI] [PubMed] [Google Scholar]
- 16.Cheng YW, Tsai WC, Chuang FC, et al. A retrospective analysis of 44 patients with granuloma annulare during an 11-year period from a tertiary medical center in south Taiwan. Dermatologica Sinica. 2016;34(3):121-125. doi: 10.1016/j.dsi.2015.11.002 [DOI] [Google Scholar]
- 17.Alirezaei P, Farshchian M. Granuloma annulare: relationship to diabetes mellitus, thyroid disorders and tuberculin skin test. Clin Cosmet Investig Dermatol. 2017;10:141-145. doi: 10.2147/CCID.S129187 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Almazan E, Roh YS, Belzberg M, et al. Comorbidities associated with granuloma annulare: a cross-sectional, case-control study. Medicines (Basel). 2020;7(9):E53. doi: 10.3390/medicines7090053 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wu W, Robinson-Bostom L, Kokkotou E, Jung H-Y, Kroumpouzos G. Dyslipidemia in granuloma annulare: a case-control study. Arch Dermatol. 2012;148(10):1131-1136. doi: 10.1001/archdermatol.2012.1381 [DOI] [PubMed] [Google Scholar]
- 20.Vázquez-López F, Pereiro M Jr, Manjón Haces JA, et al. Localized granuloma annulare and autoimmune thyroiditis in adult women: a case-control study. J Am Acad Dermatol. 2003;48(4):517-520. doi: 10.1067/mjd.2003.104 [DOI] [PubMed] [Google Scholar]
- 21.Rubin CB, Rosenbach M. Granuloma annulare: a retrospective series of 133 patients. Cutis. 2019;103(2):102-106. [PubMed] [Google Scholar]
- 22.Nebesio CL, Lewis C, Chuang TY. Lack of an association between granuloma annulare and type 2 diabetes mellitus. Br J Dermatol. 2002;146(1):122-124. doi: 10.1046/j.0007-0963.2001.04527.x [DOI] [PubMed] [Google Scholar]
- 23.Williamson DM, Dykes JR. Carbohydrate metabolism in granuloma annulare. J Invest Dermatol. 1972;58(6):400-404. doi: 10.1111/1523-1747.ep12540738 [DOI] [PubMed] [Google Scholar]
- 24.Li A, Hogan DJ, Sanusi ID, Smoller BR. Granuloma annulare and malignant neoplasms. Am J Dermatopathol. 2003;25(2):113-116. doi: 10.1097/00000372-200304000-00004 [DOI] [PubMed] [Google Scholar]
- 25.Armstrong AW, Harskamp CT, Armstrong EJ. Psoriasis and metabolic syndrome: a systematic review and meta-analysis of observational studies. J Am Acad Dermatol. 2013;68(4):654-662. doi: 10.1016/j.jaad.2012.08.015 [DOI] [PubMed] [Google Scholar]
- 26.Optum . Data assets chart sheet. Accessed September 5, 2018. https://www.optum.com/content/dam/optum/resources/productSheets/5302_Data_Assets_Chart_Sheet_ISPOR.pdf
- 27.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. doi: 10.1016/S0140-6736(07)61602-X [DOI] [PubMed] [Google Scholar]
- 28.Barbieri JS, Rodriguez O, Rosenbach M, Margolis D. Incidence and prevalence of granuloma annulare in the United States. JAMA Dermatol. Published online June 9, 2021. doi: 10.1001/jamadermatol.2021.1847 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Centers for Medicare & Medicaid Services . ICD-10. Accessed December 18, 2020. https://www.cms.gov/Medicare/Coding/ICD10 [PubMed]
- 30.Chen G, Khan N, Walker R, Quan H. Validating ICD coding algorithms for diabetes mellitus from administrative data. Diabetes Res Clin Pract. 2010;89(2):189-195. doi: 10.1016/j.diabres.2010.03.007 [DOI] [PubMed] [Google Scholar]
- 31.Oake J, Aref-Eshghi E, Godwin M, et al. Using electronic medical record to identify patients with dyslipidemia in primary care settings: International Classification of Disease code matters from one region to a national database. Biomed Inform Insights. 2017;9:1178222616685880. doi: 10.1177/1178222616685880 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Quan H, Li B, Saunders LD, et al. ; IMECCHI Investigators . Assessing validity of ICD-9-CM and ICD-10 administrative data in recording clinical conditions in a unique dually coded database. Health Serv Res. 2008;43(4):1424-1441. doi: 10.1111/j.1475-6773.2007.00822.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Tonelli M, Wiebe N, Fortin M, et al. ; Alberta Kidney Disease Network . Methods for identifying 30 chronic conditions: application to administrative data. BMC Med Inform Decis Mak. 2015;15:31. doi: 10.1186/s12911-015-0155-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Barbieri JS, Mostaghimi A, Noe MH, Margolis DJ, Gelfand JM. Use of primary care services among patients with chronic skin disease seen by dermatologists. JAAD Int. 2021;2:31-36. doi: 10.1016/j.jdin.2020.10.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Gabaldón VH, Haro-González-Vico V. Lack of an association between generalized granuloma annulare and malignancy: a case-control study. J Am Acad Dermatol. 2019;80(6):1799-1800. doi: 10.1016/j.jaad.2019.02.056 [DOI] [PubMed] [Google Scholar]
- 36.Devaraj S, Dasu MR, Jialal I. Diabetes is a proinflammatory state: a translational perspective. Expert Rev Endocrinol Metab. 2010;5(1):19-28. doi: 10.1586/eem.09.44 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Tsalamandris S, Antonopoulos AS, Oikonomou E, et al. The role of inflammation in diabetes: current concepts and future perspectives. Eur Cardiol. 2019;14(1):50-59. doi: 10.15420/ecr.2018.33.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Mailer RKW, Gisterå A, Polyzos KA, Ketelhuth DFJ, Hansson GK. Hypercholesterolemia enhances T cell receptor signaling and increases the regulatory T cell population. Sci Rep. 2017;7(1):15655. doi: 10.1038/s41598-017-15546-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Fessler MB. Regulation of adaptive immunity in health and disease by cholesterol metabolism. Curr Allergy Asthma Rep. 2015;15(8):48. doi: 10.1007/s11882-015-0548-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ramessur R, Gill D. The effect of statins on severity of psoriasis: a systematic review. Indian J Dermatol Venereol Leprol. 2017;83(2):154-161. doi: 10.4103/0378-6323.188655 [DOI] [PubMed] [Google Scholar]
- 41.Wu J, Kwong BY, Martires KJ, et al. Granuloma annulare associated with immune checkpoint inhibitors. J Eur Acad Dermatol Venereol. 2018;32(4):e124-e126. doi: 10.1111/jdv.14617 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Cornejo CM, Haun P, English J III, Rosenbach M. Immune checkpoint inhibitors and the development of granulomatous reactions. J Am Acad Dermatol. 2019;81(5):1165-1175. doi: 10.1016/j.jaad.2018.07.051 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable. Sensitivity Analysis Including Adjustment for Relevant Diagnostic Testing