Asthma Morbidity Measures Across Black Ethnic Subgroups

Abstract

Background:

Black adults are disproportionately affected by asthma but are often considered a homogenous group in research studies despite cultural and ancestral differences.

Objective:

We sought to determine if asthma morbidity differs across adults in Black ethnic subgroups.

Methods:

Adults with moderate-severe asthma were recruited across the continental US and Puerto Rico for the PREPARE trial. Using self-identifications, we categorized “Multiethnic Blacks” (ME/B; n=226) as “Black Latinx” (n=146) or “Caribbean, Continental African, or Other Blacks” (CAO/B; n=80). “African Americans” (AA/B; n=518) were categorized as Black participants who identified their ethnicity as being American. Baseline characteristics and retrospective asthma morbidity measures (self-reported exacerbations requiring systemic corticosteroids (SC), emergency department/urgent care (ED/UC) visits, hospitalizations) were compared across subgroups using multivariable regression.

Results:

Compared with AA/B participants, ME/B participants were more likely to be younger and residing in the US Northeast and Spanish speaking; to have lower body mass index, health literacy, and <1 comorbidity but higher blood eosinophil counts. In a multivariable analysis, ME/B participants were significantly more likely to have ED/UC visits (incident rate ratio [IRR]=1.34, 95% confidence interval [CI]=1.04-1.72) and SC use (IRR=1.27, 95% CI=1.00-1.62) for asthma than AA/B participants. Of the ME/B subgroups, Puerto Rican Black Latinx participants (n=120) were significantly more likely to have ED/UC visits (IRR=1.64, 95% CI=1.22-2.21) and SC use for asthma (IRR=1.43, 95%CI=1.06-1.92) than AA/B participants. There were no significant differences in hospitalizations for asthma among subgroups.

Conclusions:

ME/B adults, specifically Puerto Rican Black Latinx, have higher risk of ED/UC visits and SC use for asthma than other Black subgroups.

Graphical Abstract

Capsule Summary

Multi-ethnic Black adults, especially Puerto Ricans, with moderate-severe asthma exhibit greater asthma morbidity than adults in other Black ethnic subgroups. Healthcare policies that acknowledge diversity within Black ethnic subgroups might improve asthma morbidity outcomes.

INTRODUCTION

Asthma is a common chronic respiratory disease that affects approximately 8.0% of adults in the United States (US) with a higher disease burden in certain racial or ethnic populations (1). Compared with White adults, Black adults have a higher prevalence of asthma (10.1% vs 8.1%) and fourfold higher rates of emergency department (ED) visits for asthma exacerbations (1). Several factors may underlie the greater burden of asthma in Black adults. These include lower socioeconomic status (2), higher eosinophilic airway inflammation (3), depression (4), and inequities in healthcare and other social determinants of health (5). However, few studies attempted to address risk factors for the greater asthma burden within the Black population.

Prior studies regarded Black Americans as a homogenous group despite significant cultural, social, and ancestral differences within this population (6). Black Americans comprise various unique ethnic subgroups, each with its distinct culture, beliefs and behaviors (7), which are often learned and passed from generation to generation. This heterogeneity within the Black population may play an important role in health disparities. For example, among Black adults residing in the US, non-Latinx Black adults born in the US have a higher predicted 10-year risk of developing coronary heart disease than Latinx Black adults and non-Latinx Black adults born outside the US. (8). Similarly, a study of cancer deaths in Black adults reported that cancer mortality rates were highest in African American and lowest in African immigrants, with intermediate values in Afro-Caribbean immigrants (9).

Whether asthma morbidity differs among different Black ethnic subgroups in the US is unknown. A higher proportion of African genetic ancestry, as opposed to European and Native American, has been shown to be a risk factor for asthma prevalence among mixed populations of African descent, including adults from Brazil, Colombia, and New York City (10). However, that study focused on asthma prevalence and not asthma morbidity, and the proportion of genetic ancestry is not the same as ethnic differences. In a separate study focused on Puerto Rican children, higher percentages of African ancestry were associated with lower lung function (11). With regard to other aspects of health including metabolic conditions, cardiovascular disease, and substance use, foreign-born Black adults have better outcomes than US-born Black adults (12). Sociocultural factors are a suggested mechanism for these differences, whereby US-born Black adults, being from a majority-White context, experience more racism and perceive racism differently than Black immigrants from outside the US and a cause for higher levels of psychological distress and worse health outcomes (13).

Based on prior findings from other conditions, we hypothesized that self-identified African American (AA/B) adults have worse asthma morbidity than Multi-ethnic Black (ME/B) adults. In this study, we examined differences in asthma morbidity amongst adults from various Black ethnic subgroups (African American, Black Latinx, Caribbean Black, Continental African, and Other Black) using data from participants in a clinical trial conducted in the continental US and Puerto Rico.

METHODS

Study population

From November 2017 through March 2021, adults with moderate-severe asthma were recruited from clinical practices across the continental US and Puerto Rico for enrollment in the Person Empowered Asthma Relief (PREPARE) trial (NCT02995733). The PREPARE trial was a randomized, open-label, pragmatic study, which found that an as-needed inhaled corticosteroid supplementation strategy named PARTICS plus usual care successfully reduced asthma exacerbations when compared to usual care alone among Black and Latinx adults with asthma (14). This study was approved by the institutional review board of Partners HealthCare and local institutional review boards. Full details on the study design and methods of the PREPARE trial have been published (15).

Black ethnic subgroups

In this secondary analysis, we included baseline data from 744 Black adult participants from the PREPARE trial (Figure 1). Subgroups were defined based on participant self-identification as AA/B (n=518), participants who identified their ethnicity as being only from the US, or ME/B (n=226). ME/B was further divided into two subgroups. The first (n=80), was comprised of participants who self-identified as Caribbean Black (non-Latinx; n=42), Continental African (origin identified as being from Africa; n=17), or Other Black (n=21); CAO/B. The second was comprised as participants who self-identified as Black Latinx (n=146).

Figure 1:

Flow diagram of the selection of study participants

Participant characteristics

At the baseline visit, data were collected on characteristics including demographics (age, sex, number of persons living in the household, preferred language, income, poverty status, educational attainment, geographic region, employment status and health literacy), clinical characteristics (body mass index [BMI], number of co-morbidities, smoking status and living in a smoking environment, asthma controller therapy regimen, perceived stress and discrimination and depressive symptoms) and phenotypic characteristics (fractional exhaled nitric oxide [FeNO] and blood eosinophil count).

Participant incomes were recorded as categories in $10,000/year increments, from “less than $10,000/year” to “more than $75,000/year.” To determine poverty status, we used income and household size, referenced to the federal poverty income guidelines (16). While federal poverty guidelines were established for the 48 contiguous states, they are usually applied to Puerto Rico (17). Data on medical comorbidities included heart disease, cancer (other than non-melanoma skin cancer), stroke, diabetes, chronic kidney disease, chronic obstructive pulmonary disease, HIV/AIDS, hypertension, depression, and sleep disorders were based on self-reporting. Perceived stress, which refers to an individual’s feelings about their level of stress and their ability to handle it (18), was determined using the 4-item Perceived Stress Scale (PSS) (19). Depressive symptoms were assessed using the Patient Health Questionnaire-2 (PHQ-2) depression screen (20). Perceived discrimination was assessed by the Everyday Discrimination Scale (21). Health literacy was measured using the Brief Health Literacy Screen (BHLS) (22).

FeNO was measured at enrollment using NIOX devices provided by Circassia Limited (Morrisville, NC). Peripheral blood eosinophil levels were either measured at baseline among participants who consented to a blood draw or a historical value from up to a year prior to enrollment, if available, was provided.

Asthma morbidity measures

Our primary outcome was self-reported asthma exacerbations in the year prior to enrollment. In accordance with the 2009 American Thoracic Society definition, a severe asthma exacerbation is one that leads to use of systemic corticosteroids (SC) for ≥3 days or to an asthma-related hospitalization or emergency department/urgent care visit requiring systemic corticosteroids(23). Participants self-reported during enrollment the number of SC courses for asthma, emergency department (ED) or urgent care (UC) visits for asthma, and asthma hospitalizations that occurred in the year prior to enrollment (24).

Secondary outcomes included asthma control and asthma-related quality of life. Asthma control at study enrollment was assessed with the Asthma Control Test (ACT) (25) and the Asthma Activities, Persistent Triggers, Asthma Medication and Response to Therapy (Asthma APGAR) surveys (26), while asthma-related quality of life was assessed with the Asthma Symptom Utility Index (ASUI) score (27).

Statistical Analysis

Self-reported clinical and phenotypic characteristics were compared between AA/B and ME/B, and between their subgroups in exploratory analyses, using Chi-square test for dichotomous data and student’s t-tests for continuous data. Differences in self-reported exacerbations were compared using negative binomial regression for SC bursts, ED/UC visits, and asthma hospitalizations, with adjustment for significant covariates.

Comparisons among subgroups were adjusted for characteristics found to be statistically significantly different among the Black ethnic subgroups at p<0.10 (Supplementary Table 1). Language and geographical region were not included in final multivariable models since they were colinear with Black ethnicity and caused model non-convergence. Similarly, sparsely populated covariates were not included in the final models. All final models were thus adjusted for age, BMI, number of medical comorbidities, educational attainment (high school diploma or higher versus no high school diploma), depressive symptoms (PHQ-2), health literacy (BHLS), perceived stress (PSS), smoking status (current, former, or non), and blood eosinophil count.

RESULTS

Baseline characteristics

Table 1 displays the baseline characteristics of the 744 self-identified Black adult participants in the PREPARE trial. Of the 744 Black study participants, 518 self-identified as AA/B and 226 as ME/B (with 146 as Black Latinx, 42 as Caribbean Black, 17 as Continental African, and 21 as Other Black). Compared with AA/B participants, ME/B participants were significantly likely to be younger and residing in the Northeast US, to prefer Spanish as spoken language and to have a lower BMI, fewer medical comorbidities, to use more baseline controller medications, to have more depressive symptoms, to have lower health care literacy, and to have a higher blood eosinophil count.

Table 1:

Demographics for AA/B compared with ME/B

					Multi-ethnic Black n=226
	N	(%) or ± SD	N	(%) or ± SD	N	(%) or ± SD	P-value
Age	48	±13.5	49.2	±13.2	45.3	±13.8	0.0003
Sex	626	(84.1)	435	(84)	191	(84.5)	0.8539
Female
Male	118	(15.9)	83	(16)	35	(15.5)
Preferred Language	669	(89.9)	518	(100)	151	(66.8)	<.0001
English
Spanish	75	(10.1)	0	0	75	(33.2)
Region	254	(34.1)	137	(26.4)	117	(51.8)	<.0001
Northeast
Ohio Valley Central	133	(17.9)	124	(23.9)	9	(4)
Puerto Rico	51	(6.9)	0	0	51	(22.6)
Southeast	305	(41)	257	(49.6)	48	(21.2)
Southwest	1	(0.1)	0	0	1	(0.4)
Low income	121	(16.3)	78	(15.1)	43	(19)	0.7561
Missing
Low income ($20000 or less)	314	(42.2)	220	(42.5)	94	(41.6)
Not low income	309	(41.5)	220	(42.5)	89	(39.4)
Federal Poverty Line	121	(16.3)	78	(15.1)	43	(19)	0.5077
Missing
Above federal poverty line	309	(41.5)	222	(42.9)	87	(38.5)
Below federal poverty line	314	(42.2)	218	(42.1)	96	(42.5)
Education level – High School (HS) or higher	671	(90.2)	464	(89.6)	207	(91.6)	0.3949
HS diploma or higher
No HS diploma	73	(9.8)	54	(10.4)	19	(8.4)
Employment Status	309	(41.5)	223	(43.1)	86	(38.1)	0.2034
Employed
Unemployed, Disabled, Retired, Other,	435	(58.5)	295	(56.9)	140	(61.9)
Prefer not to answer
Body Mass Index (BMI) (kg/m2)	36.3	±9.7	37.3	±9.8	34	±8.8	<.0001
Comorbidity Count	209	(28.1)	125	(24.1)	84	(37.2)	0.0007
0
1	174	(23.4)	120	(23.2)	54	(23.9)
2	170	(22.8)	136	(26.3)	34	(15)
3	97	(13)	67	(12.9)	30	(13.3)
4+	94	(12.6)	70	(13.5)	24	(10.6)
Asthma Control Test (ACT)	14.9	±4.4	15	±4.4	14.6	±4.4	0.248
Asthma Symptom Utility Index (ASUI)	0.7	±0.2	0.7	±0.2	0.7	±0.2	0.3896
Asthma APGAR	3.4	±1.8	3.4	±1.8	3.5	±1.8	0.6843
Depression: PHQ2	168	(22.6)	102	(19.7)	66	(29.2)	0.0043
Depressed
Not depressed	576	(77.4)	416	(80.3)	160	(70.8)
Health literacy: BHLS	650	(87.4)	463	(89.4)	187	(82.7)	0.0122
High
Low/marginal	94	(12.6)	55	(10.6)	39	(17.3)
Everyday Discrimination Scale score	9.2	±4.7	9.1	±4.6	9.5	±4.9	0.2337
Perceived Stress Scale score	5.4	±3.2	5.3	±3.1	5.6	±3.3	0.2114
Baseline Controller Medication	122	(16.4)	86	(16.6)	36	(15.9)	0.08
ICS only
ICS+1 other controller therapy	309	(41.5)	229	(44.2)	80	(35.4)
ICS+2+ controller therapy	293	(39.4)	192	(37.1)	101	(44.7)
No ICS reported	1	(0.1)	1	(0.2)	0	0
Regimen includes biologic	19	(2.6)	10	(1.9)	9	(4)
FeNO (N=622) (ppb)	28.5	±31.8	28.3	±32.6	28.9	±30	0.8264
Absolute eosinophil count (N=614) (cells/μL)	236	±200.7	220.4	±176.3	268.7	±241.5	0.0125
Smoking status	71	(9.5)	48	(9.3)	23	(10.2)	0.9193
Current smoker
Former smoker	168	(22.6)	118	(22.8)	50	(22.1)
Non-smoker	505	(67.9)	352	(68)	153	(67.7)
Smoking environment	561	(75.4)	385	(74.3)	176	(77.9)	0.3009
No
Yes	183	(24.6)	133	(25.7)	50	(22.1)

Values are mean. Bold values are statistically sianificant.

AA/B: African American/Black

ACT: Asthma Control Score

ASUI: Asthma Symptom Utility Index

APGAR: Activities, Persistent, triGGers, Asthma medications, Response to therapy

BMI: Body Mass Index

PHQ-2: Patient Questionnaire-2

BHLS: Brief Health Literacy Scale

FeNO: Fractional Exhaled nitric oxide

Asthma outcomes across Black ethnic subgroups

ME/B participants experienced higher rates of ED/UC visits (incident rate ratio [IRR]=1.34, 95% confidence interval [CI] 1.04-1.72) and more SC use (IRR= 1.27, 95% CI=1.00-1.62) for asthma, but not hospitalizations than AA/B participants (Figure 2), even after adjustment for covariates (age, BMI, medical comorbidities, educational attainment, depressive symptoms, health literacy, perceived stress, smoking status, and blood eosinophil counts). Further, ME/B and AA/B participants had similar levels of asthma control (based on ACT and Asthma APGAR scores) and asthma-related quality of life (based on ASUI scores).

Figure 2:

Panel A: Forest plot comparing ME/B (Black Latinx, Caribbean Black, Continental African, and Other Black) with AA/B using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for ME/B.-Panel B: Forest plot comparing Black Latinx with AA/B using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for Black Latinx. Panel C: Forest plot comparing AA/B with CAO/B (Caribbean Black, Continental African, Other Black) using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for AA/B. Panel D: Forest plot comparing Black Latinx with CAO/B (Caribbean Black, Continental African, Other Black) using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for Black Latinx.

Exploratory analyses

We then set out to determine whether self-identifying with a specific subgroup (i.e., Black Latinx participants (n=146) or CAO/B [i.e., Caribbean Black, Continental African, and Other Black participants; n=80]) accounted for the greater rates of ED/UC asthma visits and SC use in the ME/B group compared to the AA/B group.

Compared with AA/B, Black Latinx adults were significantly more likely to be younger and residing in the Northeast US or Puerto Rico, to prefer Spanish as their spoken language, have lower BMI, fewer medical comorbidities, use more controller medications, have greater perceived stress levels and be depressed, and have lower health literacy and higher blood eosinophil counts (Supplementary Table 2). Black Latinx adults had significantly greater rates of ED/UC visits (IRR=1.53, 95%CI=1.15-2.04) and SC use (IRR=1.33, 95%CI=1.00-1.77) for asthma than AA/B (Figure 2) following adjustment for covariates. Black Latinx and AA/B participants had similar numbers of hospitalizations for asthma and similar levels of asthma control and asthma-related quality of life.

Compared with AA/B, CAO/B [Caribbean, Continental African, and Other Black participants; n=80] were more likely to be younger and residing in the Northeast US, and to have fewer medical comorbidities (Supplementary Table 1). However, there were no significant differences between CAO/B and AA/B participants in terms of ED/UC visits, SC courses, hospitalizations for asthma and asthma control or asthma-related quality of life after adjustment for covariates listed in the statistical methods (Figure 2).

Compared with CAO/B, Black Latinx adults were significantly more likely to be residing in Puerto Rico and to prefer Spanish as their spoken language, to have lower BMI and more medical comorbidities, and to be depressed (Supplementary Table 3). However, there were no significant differences between Black Latinx and CAO/B participants in terms of ED/UC visits, SC courses, hospitalizations for asthma or in asthma control or asthma-related quality of life after adjustment for covariates listed in the statistical methods (Figure 2).

We then set out to determine whether a specific subgroup of Black Latinx adults (i.e., Black Puerto Rican participants (n=120) or non-Puerto Rican Black Latinx participants (i.e., Mexican, Central and South American, Dominicans and Cuban Black participants; n=26) explained the greater rates of ED/UC visits for asthma in Black Latinx adults relative to AA/B. Compared with AA/B, Puerto Rican Black participants were more likely to be younger, be from the Northeast US or Puerto Rico, prefer Spanish as their spoken language, have lower BMI and fewer medical comorbidities, use more baseline controller medications, be depressed, have lower health literacy and higher blood eosinophil counts (Supplementary Table 1). Puerto Rican Black participants had significantly greater rates of ED/UC visits (IRR=1.64, 95%CI=1.22-2.21) and SC use (IRR=1.43, 95%CI=1.06-1.92) for asthma than AA/B participants (Figure 3) following adjustment for covariates listed in the statistical methods. There were no significant differences in the risk of asthma hospitalizations, baseline asthma control, or baseline asthma-related quality of life between these two subgroups.

Figure 3:

Panel E: Forest plot comparing Puerto Rican (PR) Black Latinx with AA/B using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for Puerto Rican Black Latinx. Panel F: Forest plot comparing Non-Puerto Rican (Non-PR) Black Latinx with AA/B using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for Non-Puerto Rican Black Latinx. Panel G: Forest plot comparing Puerto Rican (PR) Black Latinx with Non-Puerto Rican (Non-PR) Black Latinx using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher rates of outcomes for Puerto Rican Black Latinx. Panel H: Forest plot comparing Puerto Rican (PR) Black adults from Puerto Rico with Puerto Rican Black adults from continental United States (US) using negative binomial regression and controlling for age, BMI, comorbidities, education, depression, health literacy, perceived stress, smoking status, and blood eosinophil count. Values to the right of the vertical line mean higher incidence of outcomes for Puerto Rican Black adults from Puerto Rico.

Compared with AA/B participants, Non-Puerto Rican Black Latinx participants (n=26), were more likely to be younger and from the Northeast US, to prefer Spanish as their spoken language and to have lower BMI, and to have lower educational attainment and health literacy (Supplementary Table 1). However, there were no significant differences in rates of ED/UC visits for asthma, number of SC courses for asthma, hospitalizations for asthma, asthma control or asthma-related quality of life between these two groups after adjustment for covariates listed in the statistical methods (Figure 3).

We questioned whether the greater rates of ED/UC visits among Black Latinx adults is due to a specific subgroup. We found that Puerto Rican Black Latinx participants (n=120) were more likely to be more educated compared with non-Puerto Rican Black Latinx participants (n=26) (Supplementary Table 1). We found that Puerto Rican Black Latinx participants had significantly greater rates of ED/UC visits (IRR=1.91, 95%CI=1.01-3.60) for asthma compared with non-Puerto Rican Black Latinx participants following adjustment for covariates listed in the statistical methods (Figure 3). There were no significant differences in SC courses or hospitalizations for asthma, asthma control or asthma-related quality of life between these two groups.

Finally, we determined whether the greater rates of ED/UC visits among Black Puerto Rican adults is due to residing in Puerto Rico as opposed to some other characteristic of the Black Puerto Rican community. We found no significant differences in rates of ED/UC visits, SC courses, or hospitalizations for asthma even when taking covariates into account (Figure 3).

DISCUSSION

In a cohort of adults with moderate-severe asthma from various Black ethnic subgroups residing in the continental US and Puerto Rico, ME/B participants, specifically Black Puerto Ricans, had greater risk of annualized asthma-related ED/UC visits and SC use than AA/B. To our knowledge this is the first study to compare clinical features and morbidity measures among Black adults with asthma in several ethnic subgroups.

While Puerto Rican adults have higher prevalence and mortality from asthma than adults in other Latinx groups (28, 29, 30), this has not been previously investigated among Black Latinx adults. Similarly, Puerto Rican children with asthma and a higher percentage of African ancestry were linked with lower lung function(11), yet this correlation was not explored in Puerto Rican adults.

Making the distinction among Black ethnic subgroups is important from a public health perspective. For instance, Black Caribbean adults differ from African American adults in terms of ethnic identity (31) and beliefs (32). Another study demonstrated that African-born Black adults living in the United States exhibited better overall health outcomes than Caribbean-born Black adults living in the US (33). A US National Health Interview Survey that identified 896 Black Latinx adults, found that their health status was influenced by both race and ethnicity. For example, Black Latinx adults were similar to both non-Latinx Black adults and White Latinx adults in regard to physical activity and smoking status (34). Taken together, these data suggest that grouping all Black adults in the US under the umbrella “African American” overlooks the significant variations existing among this population.

In our study, self-identified Caribbean Black, Continental African, Black Latinx, and Other Black participants comprised the ME/B group because they self-identified with a culture rather than being solely American. Although all participants resided in the US or Puerto Rico, the ME/B subgroups identified their native culture rather than being American.

The findings are contrary to our original hypothesis that AA/B adults would have worse asthma morbidity compared to ME/B subgroups. A study showed that Caribbean-born Blacks in New York City had lower all-cause and cardiovascular mortality rates compared to their US-born counterparts (35). This contrasts with our asthma-related finding that CAO/B, which encompass Black Caribbeans, have no significant differences in asthma morbidity measures when compared to AA/B. As described in previous studies, psychological distress was a possible factor contributing to worse health outcomes seen in US-born Black adults compared to foreign-born Black adults (13). We expected to see similar findings in regard to asthma morbidity in our study population, however this was not the case. It is possible that there are other sociocultural factors contributing to greater asthma morbidity in ME/B adults, specifically Black Puerto Rican adults. Studies that take into consideration the existence of subgroups/cultures within the Black population are lacking, particularly in regard to health status.

Previous studies provide clues as to why Puerto Ricans have higher asthma morbidity. These included obesity, psychosocial stress, low health literacy, inadequate access to health care, and air pollution (36, 37). Our findings that Black Puerto Rican Latinx adults experience more ED/UC visits and SC use than adults in other Black subgroups remained significant after adjustment for several of these covariates. This suggests that such differences are explained by other risk factors not measured in our study, including increased perception of severe asthma symptoms, cultural behaviors leading to increased healthcare utilization, diet, exposure to allergens, access to health care, and prematurity. It should also be noted that ME/B participants, particularly Black Puerto Rican Latinx adults, had higher blood eosinophil counts compared to AA/B participants. Although adjusted for in the multivariate analyses, high blood eosinophil counts have been considered a risk factor for asthma exacerbations in adults with persistent asthma (38). However, this finding has not been specifically explored in Black adults with asthma. Possible explanations to the higher blood eosinophil counts seen in ME/B adults in our study population include air pollution, genetic variation, and allergen exposure.

In many cases, asthma care in the ED is more accessible than at primary care physicians’ offices(39); barriers to access to care may partly explain our findings. Indeed, access to the ED/UC is less restricted than access to a primary care or specialist office in terms of hours and availability (40). Thus, ED/UC has become a first-line option for some Black adults with asthma. This may account for the higher incidence of ED/UC visits in Black Puerto Rican Latinx adults without an increase in hospitalizations. Since this study was not designed to determine the reasons for the differences noted, these factors remain speculative.

Another significant finding from our study was that Black Puerto Rican Latinx adults experienced higher rates of ED/UC visits for asthma than Non-Puerto Rican Black Latinx adults. This finding further suggests that factors other than language, race, or ethnicity impact higher asthma morbidity.

Our study has several limitations. First, subgroups were created based on participants’ self-identification of race and ethnicity and not genetic ancestry, as genetic data were not captured for the PREPARE trial. Regardless of the impact of genetics, our study clearly illustrates the importance of cultural differences. Second, asthma exacerbations were based on self-reported retrospective data and could be subject to recall bias. Third, there are relatively few participants in the study which limits statistical power to show differences across subgroups. In addition, some subgroups, such as Caribbean Black, Continental African and Other Black were small, which led us to combine these participants. This could result in loss of representation of these subgroups. Fourth, we could not adjust by geography or language preference because of their collinearity with Black subgroups (e.g., zero AA adults resided in Puerto Rico or preferred Spanish as their spoken language). We acknowledge that geographic differences in terms of climate (41), health care delivery (42), quality of nutrients (43), and pollutants (44) may account for some of our findings. Similarly, preference for Spanish often restricts patient-clinician communication (45) leading to reduced healthcare access(46), further aggravating asthma morbidity(47). However, we found no significant differences in any asthma exacerbation outcome between Puerto Rican participants who reside in the mainland US vs. those who reside in Puerto Rico (of whom 21% vs. 100% preferred Spanish, respectively), suggesting that characteristics other than geography and language underlie the greater morbidity experienced by Black Puerto Rican adults. Fifth, most participants in this cohort were female. While this finding is consistent with Black women being disproportionately affected by asthma (48), our findings may not be generalizable to men. Sixth, we did not capture place of birth in the PREPARE study; therefore our results cannot discern whether Black subgroup categories denote exposures related to being foreign-born versus having a foreign culture. Future studies may wish to address this point. Lastly, our cohort was comprised of Black adults with moderate-severe asthma and our findings cannot be generalized to all adults with asthma. Participants volunteering for a clinical trial may not be representative of the general population, as those with more exacerbations or worse asthma control may be more inclined to volunteer. There may be additional cultural factors varying among subgroups that influence participation in a clinical study. This type of recruitment bias could potentially distort findings.

We find that ME/B adults, specifically Puerto Rican Black adults, with moderate- severe asthma experience worse asthma morbidity than other Black adult subgroups. This study highlights the importance of distinguishing ethnic subgroups within the Black population since cultural factors impact asthma morbidity. Clinicians should tailor asthma education in a culturally sensitive manner. Policy makers and researchers should realize the unique barriers to healthcare access among adults with asthma that vary not only by country of origin, race, and ethnicity, but also by culture, and design interventions that break down such barriers.

Clinical Implications.

Disparities in asthma morbidity exist for Black adults. Black Latinx adults, especially Puerto Ricans, are at higher risk for ED visits for asthma than other Black adult ethnic subgroups.

Abbreviations

ACT^®

Asthma Control Test

AA/B

African American/Black

APGAR

Activities, Persistent, triGGers, Asthma medications, Response to therapy

ASUI

Asthma Symptom Utility Index

BHLS

Brief Health Literacy Screen

BMI

Body Mass Index

CAO/B

Caribbean, African, or Other/Black

ED

Emergency department

FeNO

Fractional Exhaled nitric oxide

IRR

Incident rate ratio

ME/B

Multiethnic/Black

OR

Odds ratio

PARTICS

Patient-Activated Reliever-Triggered Inhaled CorticoSteroid

PCORI

Patient Centered Outcomes Research Institute

PHQ-2

Patient Health Questionnaire-2

PR

Puerto Ricans

PREPARE

PeRson EmPowered Asthma RElief

PSS

perceived stress scale

SABA

Short-acting beta₂-agonist

SC

Systemic corticosteroids

UC

Urgent care

US

United States