Allergic Rhinitis—Underrepresented Populations and Barriers to Healthcare Access

ABSTRACT

Objective

To establish the prevalence of allergic rhinitis (AR), categorized by demographics and barriers to healthcare, and the prevalence of antihistamine and nasal steroid use in these subgroups.

Methods

We performed a retrospective, cross‐sectional study utilizing the All of Us Database. Sociodemographic factors among AR patients were compared via Chi‐Square analysis and multivariable logistic regression (MLR). Subgroups of AR patients with or without nasal steroid spray or oral antihistamine listed in the electronic health record (EHR) were compared via chi‐square analysis and MLR.

Results

47,224 participants were identified with AR, an 11.6% estimated prevalence. AR patients were more commonly White (12.8% vs. 10.6%, p < 0.001), female (13.1% vs. 9.1%, p < 0.001), and older than 65 (14.7% vs. 7.6% vs. 11.6%, p < 0.001). MLR identified older age (OR 1.018, CI: 1.017–1.018), income > $35,000 (OR 1.035, CI: 1.021–1.049), finishing high school/college (OR 1.140, CI: 1.113–1.167; OR 1.113, CI: 1.085, 1.142), and health insurance coverage (OR 2.003, CI: 1.924–2.087) as predictive factors for AR. Patients with nasal steroid spray and/or oral antihistamine listed in the EHR were more commonly Black (OR 1.250, CI: 1.177–1.328; OR 1.491, CI: 1.398–1.590) and had an income < $35,000 (OR 0.856, CI: 0.814–0.900; OR 0.724, CI: 0.687–0.764).

Conclusion

AR patients were more likely to be insured, while oral antihistamines/nasal steroid spray listed in the EHR were associated with lower income and Black race. These results highlight the barriers to AR diagnoses and treatment and the need for providers to ensure that underserved patients are offered appropriate care.

Level of Evidence

3

Short abstract

Allergic rhinitis diagnosis is more likely in older patients, females, White patients, non‐Hispanic patients, those with higher education (both high school and college), those with higher socioeconomic status, and those with health insurance. At the same time, oral antihistamine and nasal steroid medications listed in the health record are more likely among Black patients and those with lower income levels. These results highlight the barriers to allergic rhinitis diagnoses and treatment, and need for providers to ensure that underserved patients are offered appropriate care.

1. Introduction

Allergic rhinitis (AR) is a common upper respiratory disorder affecting up to 60 million people in the United States and 10%–30% of the global population [1]. Symptoms include nasal congestion, sneezing, pruritis, and rhinorrhea [2, 3]. Approximately 75% of patients rate their symptoms as moderate to severe, and 35%–50% of adults reported that nasal allergies have at least a moderate effect on daily life [4, 5]. Risk factors associated with the development of AR include a family history of atopy, serum IgE > 100 IU/mL before age six, higher socioeconomic status (SES), and the presence of a positive allergy skin prick test [6, 7, 8, 9].

While the mechanism of allergic rhinitis remains the same in all population types, different environmental factors are heavily associated with disease burden. Outdoor and indoor air quality, specific antigen exposures, and access to clean air all affect the outcome of the disease [10]. First‐line treatment for all AR is avoidance of the triggering antigen, which may be more or less difficult based on the SES of each patient. Beyond allergen avoidance, intranasal corticosteroids such as fluticasone and oral antihistamines are used ubiquitously in treating AR. Newer therapies such as immunotherapy are beyond the scope of this study, but are quite expensive and cost‐prohibitive for most [3].

Even though AR affects a large portion of the population, there have been limited studies analyzing how patient demographics and SES affect the AR diagnosis rate. For example, there have been conflicting studies analyzing whether AR prevalence is associated with patient gender [11, 12] and age [13, 14]. While it is generally believed that there is a higher prevalence of AR among White patients [15], it remains unclear whether this is due to disease predilection or limited access to care among underserved groups. Moreover, there are clear barriers to care among Black AR patients, as they are less likely to receive AR treatment and more likely to be diagnosed in the emergency room [16]. While a Swedish study revealed that AR patients were more likely to have higher levels of education, there is a need for more comprehensive analysis of how SES affects AR prevalence in the US [17].

Given these gaps in knowledge, we aimed to better understand how AR prevalence differs by age, gender, SES status, and race by utilizing a database including patients who are traditionally underrepresented in medicine. The All of Us database, a National Institute of Health database, is a comprehensive and diverse health database of participants across the United States that encompasses a wide range of genetic, environmental, lifestyle, and clinical information. This database is a tool for researchers to unravel intricate relationships between these factors and health outcomes [18]. The All of Us Database defines underrepresented populations based on multiple factors including race/ethnicity, age, sexual orientation/gender identity, income, education, geography, and physical disability [19]. Leveraging the potential of the All of Us Database, this study aims to analyze allergic rhinitis in the underrepresented population defined by this novel database.

2. Materials and Methods

2.1. Database Information

This database study utilized the All of Us Database (Version 7, released 7/1/2022). The All of Us database was created by the National Institutes of Health to build a large biomedical database with a diverse group of participants [20]. The database includes information about US adults (18 years and older) from 2018 to July of 2022 and gathers information from health questionnaires, electronic health records, digital health technology, biospecimens, and physical measurements. Data is available at https://www.allofus.nih.gov. This study was considered nonhuman subjects research, and institutional review board review was waived by the University of Miami.

2.2. Inclusion Criteria and Study Variables

We conducted a population‐based cross‐sectional analysis to explore the relationship between AR and various sociodemographic variables. For our AR group, we included patients with the Systemized Nomenclature of Medicine (SNOMED) code 61582004 for “allergic rhinitis.” SNOMED connects various terminology, medical codes, and definitions from numerous electronic health records. Therefore, we excluded patients with diagnoses listed in the database as drug‐induced rhinitis, gustatory rhinitis, infectious rhinitis, atrophic rhinitis, etc. The All of Use Database reports demographic and socioeconomic data through many sources to form synthesized data sets including surveys, electronic health records, wearables (i.e., Fitbit), and biosamples [20].

We collected information among all patients listed in the database and those who met the inclusion criteria for the AR group. Demographic variables in the All of Us Database included sex, age, race, ethnicity, educational attainment, physical disability status, income, and insurance status. Those with a diagnosis of allergic rhinitis were also queried for intranasal nasal steroid sprays including fluticasone, mometasone, triamcinolone, budesonide, and beclomethasone spray, with all drug formulations reported in Supporting Information Table S1. There is no information in the database about patients receiving fluticasone by powder insufflation. AR patients were also queried for prescriptions of oral second‐generation antihistamines including cetirizine, loratadine, levocetirizine, desloratadine, and fexofenadine, with all drug formulations reported in Supporting Information Table S2. The database did not have substantial information about allergy testing or referral to otolaryngology, and therefore we could not analyze these variables. Patients were asked to disclose their own physical disability status, which is defined by the database as a condition that substantially limits one or more major life activities such as mobility, self‐care, and activities of daily living [19]. Income analysis was stratified as greater than or less than $35,000, as this was a limitation of the patient information listed in the database.

2.3. Statistical Analysis

Statistical analyses were performed using R (Version 4.4.0), embedded within the All of Us database. All data was stored on the All of Us Database portal. Multinomial logistic regressions were made with the glm function, predicting either the presence of allergic rhinitis or nasal steroid spray/oral antihistamine use with the categorical variables of income (reference category: < $35,000), sex (reference category: Female), education (reference category: Did Not Complete High School), race (reference category: White), ethnicity (reference category: Not Hispanic or Latino), disability (reference category: No Disability), insurance (reference category: No Insurance), and a continuous variable of age. T‐tests and Chi‐Squared tests were performed with the functions t.test and chisq.test, respectively. All analyses were 2‐sided, and p‐values of < 0.05 were considered significant.

3. Results

Among 407,236 total enrolled patients in the database, 47,224 (11.6%) were diagnosed with AR (Table 1). Patients diagnosed with AR were more likely to be female (13.1% vs. 9.1%, p < 0.001), older than 65 compared to 18–44 or 45–64 years old (14.7% vs. 7.6% vs. 11.6%, p < 0.001), White compared to Black (12.8% vs. 10.6%, p < 0.001), and non‐Hispanic compared to Hispanic (12.0% vs. 9.8%, p < 0.001). AR patients commonly utilized loratadine (59.7%), cetirizine (56.8%), fluticasone (93.4%), and/or mometasone (13.9%) therapy (Supporting Information Table S3). Among those with a diagnosis of AR, nasal steroid spray therapy was more commonly seen in the following groups: females (49.6% vs. 48.5%, p = 0.021), those older than 65 years old compared to those 18–44 or 45–64 years old (51.6% vs. 42.7% vs. 49.7%, p < 0.001), Black patients compared to White patients (53.6% vs. 47.3%, p < 0.001), and Hispanic patients (53.6% vs. 48.5%, p < 0.001) (Table 2). Likewise, oral antihistamine therapy was more commonly listed in the EHR among females (55.4% vs. 50.4%, p < 0.001), those 45–64 years old compared to 18–44 or older than 65 years old (56.6% vs. 50.5% vs. 53.3%, p < 0.001), Black patients compared to White patients (64.2% vs. 50.1%, p < 0.001), and Hispanic patients (59.1% vs. 52.9%, p < 0.001) (Table 3).

TABLE 1.

Demographic description of all patients vs. allergic rhinitis patients (AR) in the All of Us database.

Variable name	All enrolled n = 407,236	AR a diagnosis n = 47,224	No AR diagnosis n = 360,012	p
Sex n (%)
Male	151,966	13,842 (9.1%)	138,124 (90.9%)	< 0.001 *
Female	246,781	32,374 (13.1%)	214,407 (86.9%)
Unknown/not represented	8489	1008 (11.9%)	7481 (88.1%)
Age n (%)
18–44	117,642	8990 (7.6%)	108,652 (92.4%)	< 0.001 *
45–64	139,702	16,260 (11.6%)	123,442 (88.4%)
65 or older	149,892	21,974 (14.7%)	127,918 (85.3%)
Race n (%)
White	224,893	28,677 (12.8%)	196,216 (87.2%)	< 0.001 *
Black	77,594	8238 (10.6%)	69,356 (89.4%)
Asian	14,105	1076 (7.6%)	13,029 (92.4%)
Multiple races	7680	801 (10.4%)	6879 (89.6%)
Unknown/not represented	4315	504 (11.7%)	3811 (88.3%)
Another single population	2972	256 (8.6%)	2716 (91.4%)
Did not answer	75,677	7672 (10.1%)	68,005 (89.9%)
Ethnicity n (%)
Hispanic	73,551	7195 (9.8%)	66,356 (90.2%)	< 0.001 *
Non‐Hispanic or Latino	317,855	38,183 (12.0%)	279,672 (88.0%)
Unknown/not represented	4315	504 (11.7%)	3811 (88.3%)
Did not answer	11,515	1342 (11.7%)	10,173 (88.3%)

^a AR, allergic rhinitis.

^*Statistically significant (p ≤ 0.05).

TABLE 2.

Comparison of patients with allergic rhinitis (AR) and those receiving nasal steroid therapy.

Variable name	AR a diagnosis, n = 47,224	No nasal steroid therapy n = 23,922	Nasal steroid therapy n = 23,302	p
Sex n (%)
Male	13,842	7132 (51.5%)	6710 (48.5%)	0.021 *
Female	32,374	16,301 (50.4%)	16,073 (49.6%)
Unknown	1008	489 (48.5%)	519 (51.5%)
Age n (%)
18–44	8553	4900 (57.3%)	3653 (42.7%)	< 0.001 *
45–64	15,928	8004 (50.3%)	7924 (49.7%)
65 or older	22,743	11,018 (48.4%)	11,725 (51.6%)
Race (n, %)
White	28,677	15,121 (52.7%)	13,556 (47.3%)	< 0.001 *
Black	8238	3824 (46.4%)	4414 (53.6%)
Asian	1076	615 (57.2%)	461 (42.8%)
Multiple races	801	440 (54.9%)	361 (45.1%)
Not represented	504	251 (49.8%)	253 (50.2%)
Another single population	256	142 (55.5%)	114 (44.5%)
Did not answer	7672	3529 (46.0%)	4143 (54.0%)
Ethnicity (n, %)
Hispanic	7195	3340 (46.4%)	3855 (53.6%)	< 0.001 *
Non‐Hispanic or Latino	38,183	19,667 (51.5%)	18,516 (48.5%)
Unknown/not represented	504	251 (49.8%)	253 (50.2%)
Did not answer	1342	664 (49.5%)	678 (50.5%)

^a AR, allergic rhinitis.

^*Statistically significant (p ≤ 0.05).

TABLE 3.

Comparison of patients with allergic rhinitis (AR) and those receiving oral antihistamine therapy.

Variable name	AR a diagnosis n = 47,224	No oral antihistamine therapy n = 21,752	Oral antihistamine therapy n = 25,472	p
Sex n (%)
Male	13,842	6868 (49.6%)	6974 (50.4%)	< 0.001 *
Female	32,374	14,433 (44.6%)	17,941 (55.4%)
Unknown	1008	451 (44.7%)	557 (55.3%)
Age n (%)
18–44	8553	4236 (49.5%)	4317 (50.5%)	< 0.001 *
45–64	15,928	6905 (43.4%)	9023 (56.6%)
65 or older	22,743	10,611 (46.7%)	12,132 (53.3%)
Race (n, %)
White	28,677	14,321 (49.9%)	14,356 (50.1%)	< 0.001 *
Black	8238	2953 (35.8%)	5285 (64.2%)
Asian	1076	590 (54.8%)	486 (45.2%)
Multiple races	801	394 (49.2%)	407 (50.8%)
Not represented	504	231 (45.8%)	273 (54.2%)
Another single population	256	136 (52.7%)	120 (47.3%)
Did not answer	7672	3127 (40.8%)	4545 (59.2%)
Ethnicity (n, %)
Hispanic	7195	2946 (40.9%)	4249 (59.1%)	< 0.001 *
Non‐Hispanic or Latino	38,183	17,978 (47.1%)	20,205 (52.9%)
Unknown/not represented	504	231 (45.8%)	273 (54.2%)
Did not answer	1342	597 (44.5%)	745 (55.5%)

^a AR, allergic rhinitis.

^*Statistically significant (p ≤ 0.05).

Multivariable logistic regression showed AR diagnosis to be more common in older patients (OR 1.018, CI: 1.017–1.018, continuous) but less common in men (OR 0.642, CI: 0.635–0.650, vs. female), Black patients (OR 0.957, CI: 0.942–0.972, vs. White), Asian patients (OR 0.720, CI: 0.696–0.745, vs. White), and Hispanic patients (OR 0.886, CI: 0.852–0.920, vs. non‐Hispanic). AR patients were more likely to have completed High School (OR 1.140 CI: 1.113–1.167 vs. not completed High School), completed College (OR 1.113, CI: 1.085–1.142, vs. not completed High School), have a household income greater than $35,000 (OR 1.035, CI: 1.021–1.049, vs. less than $35,000), and have healthcare insurance (OR 2.003, CI: 1.924–2.087, vs. uninsured). There was no difference in physical disability status among those with or without AR diagnoses. These multivariate logistic regression results can be found in Table 4.

TABLE 4.

Logistic regression of patients with allergic rhinitis (AR).

Variable name	All enrolled n = 407,236	AR diagnosis n = 47,224	Prevalence estimate, OR (95% CI)	Multivariate‐adjusted regression, coefficient (95% CI)	p
Sex n (%)
Female	246,781	32,374 (13.1%)	Reference	—	—
Male	151,966	13,842 (9.1%)	0.642 (0.635, 0.650)	−0.443 (−0.455, −0.431)	< 0.001
Unknown	8489	1008 (11.9%)	NA	NA	—
Age n (%)
Continuous	407,236	47,224	1.018 (1.017, 1.018)	0.018 (0.017, 0.018)	< 0.001
Race n (%)
White	224,893	28,677 (12.8%)	Reference	—	—
Black	77,594	8238 (10.6%)	0.957 (0.942,0.972)	−0.044 (−0.060, −0.028)	< 0.001
Asian	14,105	1076 (7.6%)	0.720 (0.696, 0.745)	−0.328 (−0.362, −0.294)	< 0.001
Multiple races	7680	801 (10.4%)	0.977 (0.938, 1.018)	−0.023 (−0.064, 0.017)	0.264
Not represented	4315	504 (11.7%)	0.974 (0.922, 1.029)	−0.027 (−0.081, 0.028)	0.340
Another single population	2972	256 (8.6%)	0.762 (0.709, 0.819)	−0.272 (−0.345, −0.199)	< 0.001
Did not answer	75,677	7672 (10.1%)	1.082 (1.038, 1.128)	0.079 (0.038, 0.120)	< 0.001
Ethnicity
Non‐Hispanic or Latino	317,855	38,183 (12.0%)	Reference	—	—
Hispanic	73,551	7195 (9.8%)	0.886 (0.852, 0.920)	−0.122 (−0.160, −0.083)	< 0.001
Not represented	4315	504 (11.7%)	NA	NA	—
Did not answer	11,515	1342 (11.7%)	0.789 (0.736, 0.846)	−0.237 (−0.306, −0.168)	< 0.001
Education n (%)
Less than High School	35,871	3386 (9.4%)	Reference	—	—
Completed High School	178,509	20,432 (11.4%)	1.140 (1.113, 1.167)	0.131 (0.107, 0.155)	< 0.001
Completed College	179,593	22,055 (12.3%)	1.113 (1.085, 1.142)	0.107 (0.082, 0.133)	< 0.001
Unknown	13,263	1351 (10.2%)	—	—	—
Household income n (%)
< $35,000	130074	14,134	Reference	—	—
> $35,000	176,335	22,197 (12.6%)	1.035 (1.021, 1.049)	0.034 (0.021, 0.048)	< 0.001
Unknown	100,827	10,893 (10.8%)	—	—	—
Health insurance n (%)
No	26,626	1337 (5.0%)	Reference	—	—
Yes	365,787	44,624 (12.2%)	2.003 (1.924, 2.087)	0.695 (0.654, 0.736)	< 0.001
Unknown	14,823	1263 (8.5%)	—	—	—
Physical disability n (%)
No	374,525	43,230 (11.5%)	Reference	—	—
Yes	32,711	3994 (12.2%)	0.986 (0.958, 1.015)	−0.014 (−0.043, 0.015)	0.182

Note: Bold values indicate statistically significance p ≤ 0.05.

Additional multivariate logistic regression revealed that those with nasal steroid spray therapy listed in the EHR were more likely to be Black (OR 1.250, CI: 1.177–1.328, vs. White), older age (OR 1.009, CI: 1.008–1.010), have a physical disability (OR 1.116, CI: 1.033–1.206), and a lower household income (OR 0.856, CI: 0.814–0.900, vs. more than $35,000). These multivariate logistic regression results can be found in Table 5. Finally, MLR revealed that patients with oral antihistamine therapy listed in the EHR were more likely to be Black (OR 1.491, CI: 1.398–1.590, vs. White), older age (OR 1.003, CI: 1.001–1.004), and have a physical disability (OR 1.320, CI: 1.215–1.434, vs. no disability) (Table 6). On the other hand, patients with oral antihistamine therapy listed in the EHR were less likely to be male (OR 0.869 CI: 0.828–0.912, vs. female), have completed High School or College (OR 0.874, CI: 0.788–0.969; OR 0.726, CI: 0.651–0.810, vs. less than High School) and report an annual income more than $35,000 (OR 0.724 CI: 0.687–0.764, vs. < $35,000).

TABLE 5.

Logistic regression of AR patients with and without nasal steroid spray treatment.

Variable name	AR diagnosis n = 47,224	Nasal steroid spray use n = 23,302	Prevalence estimate, OR (95% CI)	Multivariate‐adjusted regression, coefficient (95% CI)	p
Sex n (%)
Female	32,374	16,073 (49.6%)	Reference	—	—
Male	13,842	6710 (48.5%)	0.965 (0.922, 1.011)	−0.035 (−0.081, 0.011)	0.133
Age n (%)
Continuous	47,224	23,302 (49.3%)	1.009 (1.008, 1.010)	0.009 (0.008, 0.010)	< 0.001
Race n (%)
White	28,677	13,556 (47.3%)	Reference	—	—
Black	8238	4414 (53.6%)	1.250 (1.177, 1.328)	0.224 (0.163, 0.284)	< 0.001
Asian	1076	461 (42.8%)	0.889 (0.769, 1.027)	−0.118 (−0.262, 0.026)	0.109
Multiple races	801	361 (45.1%)	0.949 (0.808, 1.113)	−0.053 (−0.213, 0.107)	0.518
Not represented	504	253 (50.2%)	1.104 (0.897, 1.359)	0.099 (−0.109, 0.307)	0.351
Another single population	256	114 (44.5%)	0.999 (0.738, 1.351)	−0.001 (−0.304, 0.301)	0.993
Did not answer	7672	4143 (54.0%)	1.467 (1.238, 1.737)	0.383 (0.214, 0.552)	< 0.001
Ethnicity
Non‐Hispanic or Latino	38,183	18,516 (48.5%)	Reference	—	—
Hispanic	7195	3855 (53.6%)	0.844 (0.721, 0.987)	−0.170 (−0.327, −0.014)	0.033
Not Represented	504	253 (50.2%)	NA	NA	NA
Did not answer	1342	678 (50.5%)	0.763 (0.597, 0.975)	−0.271 (−0.515, −0.026)	0.030
Education n (%)
Less than High School	3386	1870 (55.2%)	Reference	—	—
Completed High School	20,432	10,323 (50.5%)	0.970 (0.881, 1.067)	−0.031 (−0.126, 0.065)	0.532
Completed College	22,055	10,447 (47.4%)	0.938 (0.847, 1.039)	−0.064 (−0.165, 0.038)	0.221
Unknown	1351	662 (49.0%)	—	—	—
Household income n (%)
< $35,000	14,134	7420 (52.5%)	Reference	—	—
> $35,000	22,197	10,408 (46.9%)	0.856 (0.814, 0.900)	−0.156 (−0.206, −0.105)	< 0.001
Unknown	10,893	5474 (50.3%)	—	—	—
Health insurance n (%)
No	1337	677 (50.6%)	Reference	—	—
Yes	44,624	22,001 (49.3%)	0.978 (0.860, 1.112)	−0.023 (−0.151, 0.106)	0.066
Unknown	1263	624 (49.4%)	—	—	—
Physical disability n (%)
No	43,230	21,153 (48.9%)	Reference	—	—
Yes	3994	2149 (53.8%)	1.116 (1.033, 1.206)	0.110 (0.033, 0.187)	0.005

Note: Bold values indicate statistically significance p ≤ 0.05.

TABLE 6.

Logistic regression of AR patients with and without oral antihistamine treatment.

Variable name	AR diagnosis n = 47,224	Oral antihistamine use n = 25,472	Prevalence estimate, OR (95% CI)	Multivariate‐adjusted regression, coefficient (95% CI)	p
Sex n (%)
Female	32,374	17,941 (55.4%)	Reference	—	—
Male	13,842	6974 (50.4%)	0.869 (0.828, 0.912)	−0.141 (−0.189, −0.092)	< 0.001
Age n (%)
Continuous	47,224	25,472 (53.9%)	1.003 (1.001, 1.004)	0.003 (0.001, 0.004)	< 0.001
Race n (%)
White	28,677	14,356 (50.1%)	Reference	—	—
Black	8238	5285 (64.2%)	1.491 (1.398, 1.590)	0.399 (0.335, 0.464)	< 0.001
Asian	1076	486 (45.2%)	0.859 (0.738, 1.000)	−0.152 (−0.303, 0.000)	0.0494
Multiple races	801	407 (50.8%)	1.007 (0.851, 1.190)	0.007 (−0.161, 0.174)	0.939
Not represented	504	273 (54.2%)	1.069 (0.858, 1.330)	0.066 (−0.153, 0.285)	0.554
Another single population	256	120 (47.3%)	0.984 (0.717, 1.351)	−0.016 (−0.333, 0.301)	0.921
Did not answer	7672	4545 (59.2%)	1.189 (0.996, 1.420)	0.174 (−0.004, 0.351)	0.055
Ethnicity
Non‐Hispanic or Latino	38,183	20,205 (52.9%)	Reference	—	—
Hispanic	7195	4249 (59.1%)	0.917 (0.778, 1.080)	−0.087 (−0.250, 0.077)	0.298
Not represented	504	273 (54.2%)	NA	NA	NA
Did not answer	1342	745 (55.5%)	0.920 (0.711, 1.190)	−0.084 (−0.341, 0.174)	0.523
Education n (%)
Less than High School	3386	2232 (65.9%)	Reference	—	—
Completed High School	20,432	11,804 (57.8%)	0.874 (0.788, 0.969)	−0.135 (−0.239, −0.032)	0.011
Completed College	22,055	10,660 (48.3%)	0.726 (0.651, 0.810)	−0.320 (−0.430, −0.211)	< 0.001
Unknown	1351	776 (57.4%)	—	—	—
Household income n (%)
< $35,000	14,134	8734 (61.8%)	Reference	—	—
> $35,000	22,197	10,852 (48.9%)	0.724 (0.687, 0.764)	−0.323 (−0.376, −0.270)	< 0.001
Unknown	10,893	5886 (54.0%)	—	—	—
Health insurance n (%)
No	1337	833 (62.3%)	Reference	—	—
Yes	44,624	23,912 (53.6%)	0.894 (0.778, 1.027)	−0.113 (−0.251, 0.026)	0.112
Unknown	1263	727 (57.6%)	—	—	—
Physical disability n (%)
No	43,230	22,991 (53.2%)	Reference	—	—
Yes	3994	2481 (62.1%)	1.320 (1.215, 1.434)	0.277 (0.194, 0.360)	< 0.001

Note: Bold values indicate statistically significance p ≤ 0.05.

4. Discussion

Allergic rhinitis remains a common diagnosis affecting millions of people in the world with a large effect on both quality of life and work productivity [1]. Here, we present a large study of the All of Us database in order to better understand the reach of AR on typically less‐studied populations. This is the first study to our knowledge to include a broad patient population, including those traditionally under represented in research studies. The estimated prevalence of 11.6% found in this study aligns with previous studies citing 10%–15% [1]. This serves as a quality control in validating the use of the All of Us Database for population‐based studies. Multivariate logistic regression in our study showed AR diagnosis to be more likely in older patients, females, White patients, non‐Hispanic patients, those with higher education (both high school and college), those with higher SES, and those with health insurance.

While our results indicated that patients older than 65 years old were more likely to have AR, with a prevalence of 14.7%, there are limited studies investigating the prevalence of AR in the elderly. For example, select studies have reported an AR prevalence between 5% and 8% among the elderly that decreases with age [14, 21, 22]. However, national studies in Poland and Switzerland revealed an estimated AR prevalence among the elderly of 10.8%–14.6% [23] and 13%–15% [13], respectively, consistent with our results. Given conflicting information in the current literature, and the increasing prevalence of AR among the elderly as our population continues to age [24, 25], we believe our analysis is essential to better describe these trends.

In addition, our results indicating that Black patients are less likely than White patients to be diagnosed with AR are consistent with the results from a national U.S. survey study conducted from 1995 to 2007 by Mattos et al. that show disparities in the treatment of allergic rhinitis [16]. In this study, Black patients were significantly less likely to be treated in the office for AR compared to White patients (OR: 0.72, CI = 0.58–0.91), but over time this racial difference became less pronounced. Furthermore, Mattos et al. reported that Black patients were significantly more likely to receive a diagnosis of AR in the emergency room compared to White patients, presumably due to a lack of insurance or other barriers to healthcare access. Likewise, our study revealed that those with an AR diagnosis were significantly more likely to have health insurance and an overall higher SES. However, one limitation of our study was that we could not stratify patients treated in the emergency room vs. the clinic setting.

While our study found an increased prevalence of AR among female patients, there remains conflicting evidence in the literature for a gender‐based predominance of AR among adults. For example, a meta‐analysis revealed a female predominance for AR in children, but no sex predilection in adult females [11]. However, a literature review revealed that males were more likely to have atopic rhinitis after puberty that was likely related to hormonal changes [12]. This hypothesis is consistent with past studies revealing allergen sensitization is greater among males [26], and the fact that hormonal changes can induce worsened rhinitis among women during pregnancy [27]. AR is a clinical diagnosis that does not require positive allergen testing, and thus it is possible that hormonal changes may lead to increased AR diagnoses among women. Moreover, while our study excluded forms of nonallergic rhinitis, it is possible that providers are over‐diagnosing AR among women. Finally, a survey‐based study in Sweden revealed that women with AR report worsened quality‐of‐life scores than men [28], underscoring the importance of better characterizing gender‐based trends in diagnoses and pathophysiology in future research.

Interestingly, in this study, the multivariate logistic regression for nasal steroid and oral antihistamine therapy differed from expected by demographic groups. Both therapies were more commonly listed in the EHR among Black, older, disabled, and lower income patients. Specifically, oral antihistamine therapy was more common among patients who had less than a high school education. Given the large database population of nearly 500,000 patients, this is unlikely to be a statistical error. While many theories might explain this phenomenon, a leading hypothesis comes from the over‐the‐counter (OTC) availability of these medications. Although the diagnosis of AR was more common in the college‐educated with higher household income, oral antihistamine therapies listed in the EHR were less likely in these same groups. We believe this could be the result of these patients purchasing medical therapy OTC and the medication thus not being listed in the EHR and All of Us Database. Moreover, prescription medications automatically update in the EHR without requiring patients to report their medications or healthcare workers to manually input this information in the EHR. Similarly, those with higher education tend to have higher health literacy [29] and may purchase OTC antihistamines to self‐medicate. A review paper found that higher‐income and more educated consumers were more likely to purchase medications online, including OTC drugs [30]. This behavior reflects a tendency among these groups to seek independent solutions for their healthcare needs, thus reducing the frequency of receiving prescriptions for medications that are available OTC.

A final explanation for this finding may be that the stepwise treatment algorithm in allergic rhinitis includes additional nasal sprays (secondary intranasal steroid sprays, azelastine, ipratropium), allergen immunotherapy, and surgical therapies such as septoplasty and inferior turbinate reduction. It is possible that those of lower SES do not move past the first line of treatment with nasal steroid sprays or oral antihistamines to the next treatment options. Moreover, past research has shown that Black and Hispanic patients with AR are less likely than White patients to be prescribed subcutaneous and sublingual immunotherapy [31, 32]. While there is limited research investigating racial discrepancies among those undergoing nasal surgeries, Black patients are less likely than other ethnic groups to undergo septoplasty and turbinate reduction surgeries [33].

Additionally, the economic burden of allergic rhinitis cannot be overlooked. It is estimated that two to five billion dollars is spent on AR annually, half of which comes from prescription medications [34, 35]. Given that nasal steroid sprays can be purchased OTC in the United States, this number is likely an underestimate. A study by Nathan showed that those with AR had a 1.8‐fold increase in the number of healthcare visits and a 2‐fold increase in medication costs [36]. Additionally, the cost of most nasal sprays is increasing beyond the rate of inflation [37]. For patients of lower SES, these changes may be cost‐prohibitive, thus resulting in a physician or practitioner placing a prescription in the EHR with the hope that it may be covered by insurance. For those of higher SES, the increase may not be as cost prohibitive. In the allergic rhinitis logistic regression, the largest predictor of diagnosis was health insurance (OR: 2.003). This further underscores the need to establish health insurance for as many patients as possible in order to diagnose and treat patients effectively.

While a useful tool, the All of Us database does have limitations in both design and use. It relies heavily on data input and organization to be readily searchable. The specific model used in the All of Us Database is called the Observation Medical Outcomes Partnership Common Data Model. The goal of this model is to standardize vocabulary and medical terms. In this specific study, we chose to use “allergic rhinitis” as a diagnosis of inclusion but did not use “nasal congestion” as this would inadvertently include patients with other causes of nasal congestion. As a result, our prevalence estimate is likely underestimated if patients were miscategorized as “nasal congestion” when in fact the more accurate diagnosis was “allergic rhinitis.” This underscores the need to be accurate when placing diagnoses in the electronic medical record. Likewise, a diagnosis of AR in the medical record for elderly patients may have been overstated if providers did not remove the AR diagnosis from the EMR after dissipation of symptoms/treatment. While we aimed to characterize barriers to healthcare access among the uninsured, only 6.5% of subjects in the database were uninsured, which limited our analysis. Additionally, the authors would have liked to outline the number of patients who underwent allergy testing or referral to otolaryngology among patients with AR. Unfortunately, the All of Us Database had limited information about these variables, and they could not be analyzed with adequate statistical power.

5. Conclusion

Allergic rhinitis is more commonly diagnosed in females, older patients, White and non‐Hispanic patients, those with higher education and income, and those with health insurance. Simultaneously, Black race and lower income levels were significantly associated with nasal steroid spray and oral antihistamine prescriptions. Further research is needed to elucidate whether this finding reflects relatively increased utilization of OTC medications, under‐treatment of AR among those of lower SES, or a combination of the two factors. Our results reaffirm the need to increase healthcare access to all populations and prescribe nasal steroid sprays to insured patients of lower SES.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplemental Table 1. Nasal Steroid Spray Formulations and Counts

Supplemental Table 2. Oral Antihistamine Formulations and Counts

Supplemental Table 3. Percentage of AR Patients Utilizing Each Antihistamine and Nasal Steroid Formulation (Total number of AR patients utilizing nasal steroids = 23,302; Total number of AR patients utilizing oral antihistamines = 25,472).

Peifer S. J., Helmen Z. M., Duffield S., et al., “Allergic Rhinitis—Underrepresented Populations and Barriers to Healthcare Access,” The Laryngoscope 135, no. 9 (2025): 3071–3081, 10.1002/lary.32165.