Background
Asthma is a chronic disorder of the airways characterized by reversible and intermittent airway obstruction, airway inflammation, and hyper-reactivity of the airways in response to a variety of stimuli (e.g., dust, animal hair, smoke, and airborne pollutants). Despite important advances in diagnosis and treatment, asthma remains one of the most prevalent chronic respiratory disorders, affecting 7-10% of the world's population. Rather than decreasing, prevalence rates of asthma over the past three decades are actually rising in all age, sex, and racial groups in North America [
1].
The global burden of asthma appears to be related to poor asthma control, which is associated with more frequent asthma symptomatology and bronchodilator use, worse pulmonary function, greater emergency health service utilization, and greater functional impairment (absenteeism, participation in social activities) [
2,
3]. In Canada, asthma remains poorly controlled in nearly 60% of patients, which places an excess burden on the health care system, and accounts for between 250-300 deaths per year [
4,
5]. Given that asthma can be well controlled for the vast majority of patients [
2,
3], identifying those patients who may be at greater risk for poorly controlled asthma represents an important goal for global asthma prevention.
Socioeconomic status (SES) has been linked to various health outcomes, with lower SES being associated with higher rates of morbidity and mortality from several chronic diseases, including cardiovascular disease, chronic obstructive pulmonary disease, and diabetes [
6‐
8]. However, SES may be particularly relevant to asthma due to pathways by which it could adversely impact asthma outcomes. At the individual level (e.g., education attainment, income), asthmatics of lower SES may have higher exposures to indoor (e.g., cockroaches, tobacco smoke [
9]) and outdoor (e.g., urban pollution [
9]) allergens, and tend to use less inhaled corticosteroids [
10], thus increasing risk for acute asthma exacerbations [
9,
11]. Though the SES-asthma link has been well established in children [
12,
13] and to some degree using area-level measures of SES (e.g., use of zipcodes or postal codes to define deprivation) in adults [
14,
15], less is known about associations between individual-level SES and asthma in adults.
The purpose of the present study was to assess associations between adult individual-level SES, measured according to education level, and several measures of asthma morbidity and health, including levels of asthma control, emergency health service use, asthma self-efficacy, and asthma-related quality of life in a Canadian cohort of asthmatics. It was hypothesized that SES would be significantly and negatively associated with these measures of asthma morbidity and health.
Discussion
The present study assessed associations between individual-level SES (measured according to educational attainment) and multiple measures of asthma morbidity in a Canadian cohort of adult asthmatics. Results showed that patients with lower SES had worse asthma control, worse asthma self-efficacy, and greater emergency health service use relative to patients with higher SES, independent of age, sex, asthma severity, current smoking, BMI, and having a mood and/or anxiety disorder. We also found that patients with less than 12 years of education were 55% more likely to report any emergency health service use, compared to those with 12 or greater years of education, when controlling for age, sex, and severity. When the additional covariates were included in the model, this relationship was no longer statistically significant. However, though statistical significance was lost, there was a minimal change in the point estimate, suggesting that mediation was unlikely. Furthermore, the Poisson regression models indicated that the relationship between education and emergency healthcare usage may be graded and have a dose-response association, even with the inclusion of all covariates.
These findings are consistent with previous studies finding significant associations between lower childhood SES and worse asthma morbidity, including increased prevalence of asthma and severe asthma [
12,
13], and increased risk of emergency department visits and hospitalizations for asthma [
29,
30]. These findings are also in line with previous studies linking lower SES (assessed using area-level and individual-level measures) to worse asthma morbidity in adults, including increased prevalence of asthma [
31], greater asthma symptomatology [
32], and increased asthma related hospitalisations [
33]. However, this study is, to our knowledge, the first to assess the impact of individual-level SES on multiple measures of asthma morbidity in such a large Canadian cohort of adult asthmatics. Although Lynd et al. [
34] examined the link between both individual and area-level measures of SES and asthma in a Canadian sample, their sample size was modest (n = 202), and their analyses focused on links between SES and short-acting bronchodilator use as a proxy measure of asthma control. Their findings are still consistent with those of our study, though we were able to extend their findings by showing that asthmatics of lower SES have worse asthma control according to the ACQ and emergency health service use.
It is noteworthy that patients with lower SES were more likely to exhibit poor health behaviors that may exacerbate asthma, including higher rates of current smoking, total pack-years, and BMI. This is consistent with previous studies linking higher rates of smoking, obesity, reduced consumption of fruits and vegetables, and higher consumption of saturated fats in low SES individuals compared to high SES individuals [
35‐
37]. The higher prevalence of poor health behaviors among socially disadvantaged adults with asthma may partially explain why these patients were more poorly controlled. However, the fact we found lower SES to be related to worse asthma control after adjustment for BMI and smoking suggests these were not the only potential mechanisms linking lower SES to poor control in this study. For example, and as detailed above, nutrition may also play a role. It must also be noted that our assessments of smoking and BMI may be imperfect (e.g., central adiposity may be more important than total body composition). Though the current study was not designed to assess the potential mechanisms linking lower SES to increased asthma morbidity, they can be found in previous studies. For example, lower SES was associated with lower use of inhaled corticosteroids [
10] and lower corticosteroid adherence [
38], though not all studies have reported this [
39]. The current study did not collect data on medication adherence, but the results were independent of asthma severity, which is primarily derived from the prescribed dosage of inhaled corticosteroids. Furthermore, a previous study has shown that SES was related to ACQ scores independent of corticosteroid use [
40]. There is also evidence that the underlying physiological processes seen in asthma are influenced by SES, where heightened inflammatory responses to similar doses of antigen challenge have been shown in patients with low versus high SES [
41,
42], which may be a consequence of low SES individuals overexpressing genes regulating their inflammatory processes [
43]. However, it should be noted that these findings are drawn from data in children and needs to be replicated in adult samples.
One additional finding that warrants discussion is that asthmatics of lower SES were less likely to be atopic (i.e., have allergic asthma) than asthmatics of higher SES. Although this was not the primary aim of the analyses, this finding is consistent with several studies linking lower SES to lower incidence of allergic asthma [
31,
32,
44,
45]. Although controversial, it has been suggested that this relationship may be due to the "hygiene hypothesis," which proposes that the development of atopic asthma and allergy may be prevented via prenatal and-or early childhood exposure to immune system stimulants (e.g., bacteria, viruses and endotoxins) that shift T-helper type 2 cell (Th2) dominance to T-helper type 1 cell (Th1) dominance [
46,
47]. This shift in cytokine balance is thought to contribute to allergic asthma and allergy, and may be induced by a lack of early exposures to microbial environments [
46], which are typical in lower SES settings (e.g., poor housing conditions that may be overcrowded, infested with cockroaches and dust mites, and poorly insulated, leading to greater exposure to infections, allergens, and mould). Our finding of less atopic asthma in patients of lower SES may therefore lend support for the "hygiene hypothesis." However, given the fact that this is a secondary finding, and the controversies surrounding the "hygiene hypothesis," further investigation is clearly needed.
Surprisingly, we did not observe any significant association between SES and asthma-related quality of life, which was contrary to our expectations and to previous findings [
14,
48]. Both lower area-level SES [
14] and composite individual-level SES [
48] have been associated with worse general and asthma-specific quality of life. The reasons for these inconsistencies are not clear. However, they may be related to issues associated with the nature of the populations assessed and to study design. For example, Blanc et al. [
14] recruited patients from multiple clinics via physician referral, as well as using random-digit telephone recruitment; whereas we recruited consecutive patients from a single tertiary-care clinic where asthma is generally more severe and thus may reduce variability in quality of life measures. The Apter et al. [
48] study found that the relationship between SES and quality of life was highly confounded by race/ethnicity, with non-Caucasians having lower SES and poorer quality of life. While the Apter et al. study consisted of nearly 60% of non-Caucasians, the current study has less than 10% non-Caucasions, suggesting that the results reported by Apter et al. may have been driven by race/ethnicity rather than SES [
49]. In addition, the significant association between SES and worse asthma-specific quality of life in Blanc et al.'s study was observed using a different measure of SES (i.e., area-level), and a different quality of life scale (i.e., Marks Asthma Quality of Life Questionnaire) than those used in the present study. As such, the disparate findings between these two studies may be attributable to the specific choice of measures. Further replication studies are needed to shed more light on the association between SES and asthma-related quality of life in adult samples.
The results of this study need be interpreted in consideration of some methodological limitations. First, patients were recruited from the asthma clinic of a single tertiary-care urban hospital, so results may not generalize to rural centers or community samples. Second, we relied upon education level as our measure of individual-level SES, when it may have been more informative to use a composite measure (e.g., education level, income, and-or occupation), or to triangulate analyses using occupation and income as separate measures of SES. Unfortunately, the only additional variable we collected was on employment status (yes-no). In addition, it should be noted that that education is the most common measure of individual-level SES and is stable over time, unlike occupation and income, that can fluctuate over the life course. Furthermore, participant response rates tend to be higher for educational attainment, unlike income which typically has lower response rates and consequently high response bias [
18]. Third, the study was cross-sectional so reverse causality may be possible, though unlikely, and education and asthma morbidity may be linked in a non-causal fashion. As such, further longitudinal studies are needed to confirm the temporal sequence of the results in the current study. Finally, our study was limited by the fact that we were not able to assess other environmental variables that are associated with SES that may have partially accounted for our findings such as actual exposure levels to allergens, irritants, and pollutants, and living conditions (i.e., overcrowding) which may have increased the risk of respiratory infections that confer risk for worse asthma morbidity [
32]. Despite these limitations, the results of the present study complement and strengthen previous reports by including a large cohort of adult asthmatics with objectively confirmed physician-diagnosed asthma and atopy, and the measurement of a range of asthma morbidity and health measures that included self-reported symptoms and objectively measured emergency health service utilization that was verified by chart review. Due to the range and depth of our assessments, we were also able to control for a number of potential confounders, including smoking status, BMI, psychiatric comorbidity, and asthma severity, which attests to the robustness of the findings.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SLB co-wrote the manuscript, conducted all statistical data analyses, and obtained funding for the study. AB collected primary data and helped develop the conceptual idea. EBL helped develop the conceptual framework and provided critical feedback on manuscript drafts. KLL conceived of the study, participated in its design and coordination, obtained funding for the study, and co-wrote the manuscript. All authors read and approved the final manuscript.