Sex Differences Among Adults Presenting to the Emergency Department With Acute Asthma FREE

ASTHMA IS a common disease, and its prevalence and natural history vary by sex.¹ From 1982 to 1992, the prevalence of asthma among US women increased 82% (from 2.9% to 5.4%) compared with an increase of only 29% among US men (from 4.0% to 5.1%).² During these same years, asthma mortality increased among women 59% (from 15 to 25 deaths per million population) vs a 34% increase among men (from 12 to 16 deaths per million population).² Although some investigators^3- 4 attribute these trends to sex differences in biological susceptibility or environmental exposures, they may also be caused by differences in asthma diagnosis and management.

There is preliminary evidence that women comprise almost two thirds of emergency department (ED) visits for acute asthma⁵ but only 52% of ED visits overall.⁶ Furthermore, because acute asthma accounts for almost 2 million ED visits and 460,000 hospitalizations in the United States annually¹—together costing at least $2 billion per year⁷—any sex differences in acute asthma would have important medical and economic implications. To date, however, there are few publications on this topic.⁵ Thus, we examined sex differences in a prospective multicenter study of more than 1200 adults presenting to the ED with acute asthma.

This study combines data from 2 prospective cohort studies performed from October to December 1996 and from April to June 1997 as part of the Multicenter Asthma Research Collaboration.⁸ Using a standardized protocol, investigators at 36 EDs in 18 states provided 24-hour coverage for a median of 2 weeks and enrolled 1281 patients with acute asthma. All patients were cared for at the discretion of the treating physician. Inclusion criteria were physician diagnosis of acute asthma, age 18 to 54 years, and ability to give informed consent. Multiple visits by individual patients were excluded. The institutional review board at each of the 36 participating hospitals approved the study.

For the present analysis, pregnant women (n=53) were excluded because their ED course might differ significantly from that of other women. This left 1228 patients for analysis.

The ED interview assessed patients' demographic characteristics, asthma history, and details of current asthma exacerbation. Data on ED management and disposition were obtained by medical chart review. Follow-up data were collected by telephone interview 2 weeks later and included details of any urgent asthma visits, changes in medical management, compliance with prescribed corticosteroid therapy, and current asthma symptoms. All forms were reviewed by site investigators before submission to the Data Coordinating Center in Boston, Mass, where they underwent further review by trained personnel and double data entry.

Primary care provider (PCP) status was assigned on the basis of the following question: "Do you have a primary care provider (such as a family doctor, internist, or nurse practitioner)?" If yes, patients were asked to provide the name and address of their PCP. Median family income was estimated using patients' home ZIP codes.⁹ Triggers of the patient's asthma, in general, were assessed using a standardized list of 9 potential triggers. Peak expiratory flow rate (PEFR) was expressed as a percentage of the patient's predicted value based on age, sex, and height.¹⁰ Changes in PEFR were expressed as the absolute change in percent predicted (eg, an improvement from 40% predicted to 70% predicted would be expressed as a change of 30%). Relapse was defined as any urgent visit to an ED or clinic for worsening of asthma during 2-week follow-up. Treatment failure after the index ED visit was assigned to patients who reported "severe symptoms" during the preceding 24 hours on any of 3 questions (ie, asthma symptoms "most of the time," "severe" discomfort and distress because of asthma, or "severe" activity limitations because of asthma) or who stated that their asthma was "about the same" or worse than at ED presentation. The severe symptoms classification was also used to characterize the 24 hours before ED presentation.

All analyses were performed using a statistical software program (STATA 5.0; StataCorp, College Station, Tex). Data are presented as proportions, means±SDs, or medians (with interquartile range). The association between sex and other factors was examined using χ², Student t, and Wilcoxon rank sum tests, as appropriate. Variables associated with sex (or with the outcome of interest) at P<.1 were evaluated for inclusion in multivariate logistic regression models. The possibility of a period effect was examined by adjusting for period of enrollment, but this did not materially affect any of the results that follow (data not shown). To optimally control for probable confounding by PCP and hospital admission status, we reran the final multivariate models in these a priori subgroups. Interaction was formally tested by multiplying the 2 factors of interest and then including this interaction term in the multivariate model. All odds ratios (ORs) are presented with 95% confidence intervals (CIs). All P values are 2-sided, with P<.05 considered statistically significant.

Study enrollment did not differ between women and men (68% of consecutive patients for both; P=.97). Of the 1228 patients, 790 (64.3%) were women and 438 (35.7%) were men, which yielded a female-male ratio of 1.8. Patients' demographic characteristics, according to sex, are shown in Table 1. There was no significant age difference between women and men. Men were more likely to be black. There were no significant sex differences in education level or estimated household income, but women were more likely to have insurance and a PCP. Among patients with a PCP, a similarly high percentage of women and men had seen their PCP within the past year (92% vs 89%; P=.19).

Clinical characteristics related to chronic asthma are also shown in Table 1. Age of asthma diagnosis differed by sex, with men more likely to report asthma diagnosis during the first decade of life and women more likely to report asthma diagnosis thereafter. Patients did not differ by sex in previous use of systemic corticosteroids, history of asthma hospitalization, or history of intubation for asthma. Women more often reported a history of hay fever and more known asthma triggers. Although environmental allergens were a common trigger among women and men (77% and 74%; P=.20), psychosocial factors were more common among women (62% vs 49%; P<.001). Smoking status did not differ by sex.

β-agonist use in the 4 weeks before ED presentation was equally common among women and men, although women received this medication more often from a home nebulizer machine (33% vs 22%; P<.001). Inhaled corticosteroid use in the past 4 weeks was more common among women. Among patients using inhaled corticosteroids, however, a similar percentage of women and men reported use at least 4 times per week (74% vs 78%; P=.41). Although women reported more urgent asthma visits to their PCP and slightly more asthma hospitalizations in the past year, they did not differ from men in the number of ED visits for acute asthma. Women and men also reported the same number of days their activity was limited by asthma.

Because several of these results may have been mediated by the patient's PCP status, we repeated selected analyses among patients with a PCP (n=797). Inhaled corticosteroid use in the past 4 weeks was identical among women and men with a PCP (52%; P=.99). Likewise, there were no sex differences in the median number of urgent clinic visits for asthma in the past year (1 visit; P=.56) or in history of asthma hospitalization during the past year (38%; P=.96).

Acute asthma presentation according to sex is shown in Table 2. Women and men differed by time of ED presentation, with men presenting more often during the early morning hours and women in the evening. There were no statistically or clinically significant sex differences in duration of asthma symptoms or treatment before the ED visit, but women more often described their symptoms as severe. Women and men did not differ significantly according to the presence of any major, relevant, concomitant disorders (7% vs 8%; P=.23), including chronic obstructive pulmonary disease (2% vs 3%; P=.13). Women presented to the ED with slightly higher respiratory rates and oxygen saturation levels, although these differences were of questionable clinical significance. By contrast, sex differences in initial PEFR values, either absolute or as percent predicted, seemed to be of statistical and clinical significance (Figure 1). Expressed another way,¹¹ acute asthma severity markedly differed by sex, with mild flares (initial PEFR ≥80%) and moderate flares (initial PEFR of 50%-79%) being more common among women (11% vs 3% and 42% vs 18%, respectively), but severe flares (initial PEFR <50%) being more common among men (46% vs 78%; overall P<.001).

Overall, the ED course for asthmatics did not differ according to sex (Table 2). Despite initial differences in PEFR, women and men received comparable numbers of inhaled β-agonist treatments in the first hour and during the entire ED stay. Patients also did not differ by corticosteroid treatment, with only two-thirds receiving systemic corticosteroids in the ED. During a comparable ED length of stay, women and men experienced similar increases in PEFR (Figure 1); this led women to have a significantly better final percent predicted PEFR. Despite this, women were significantly more likely to be admitted to the hospital than men.

Restricting the analysis to patients admitted to the hospital (n=274), women had a lower final PEFR than men (229±72 vs 285±83 L/min; P<.001). However, when PEFR was standardized (ie, expressed as percent predicted for patient's age, sex, and height),¹⁰ women also had significantly higher preadmission values than men (59%±19% vs 48%±14%; P<.001). A similar pattern was found for initial PEFR (179±72 vs 206±68 L/min [P=.02] and 46%±19% vs 34%±12% [P<.001]). Women and men who were admitted to the hospital experienced comparable increases in PEFR during the ED stay (12%±14% vs 14%±15%; P=.49). Although details of the hospital course are not available, median inpatient length of stay tended to be longer for women than for men (3 vs 2 days; P=.07).

To further examine the association between sex and risk of hospital admission, a multivariate logistic regression was performed (Table 3). This analysis excluded the few patients who left the ED against medical advice, an ED disposition that differed between women and men (1% vs 5%; P<.001). The final model, controlling for 17 potential confounders, showed that women were 2.2 times more likely than men to be admitted to the hospital for acute asthma. Even if one assumes that all 30 against-medical-advice visits would have led to hospital admission, women remained at significantly higher risk than men (model 4: OR, 1.9; 95% CI, 1.1-3.2). Furthermore, restricting the cohort to patients with a PCP (n=797) did not materially change the overall result (model 4: OR, 2.3; 95% CI, 1.2-4.4).

Women were more likely than men to be reached for the 2-week follow-up interview (75% vs 58%; P<.001). Among the 595 women and 256 men with available data, risk of relapse was slightly higher among women, but the difference was not statistically significant (17% vs 13%; P=.12). On multivariate analysis adjusting for 10 factors, any possible difference essentially disappeared (OR, 1.1; 95% CI, 0.7-1.8; P=.65).

By contrast, women were more likely than men to have an ongoing exacerbation at 2-week follow-up (25% vs 18%; P=.02). This difference was not caused by differences in reported compliance with corticosteroid therapy, which was high in women and men (84% vs 87%; P=.34). To further examine the association between sex and treatment failure, a multivariate logistic regression was performed (Table 4). The final model, controlling for 12 potential confounders, found that women remained 1.5 times more likely than men to report an ongoing exacerbation. Somewhat higher ORs were obtained in 2 subgroups: the 587 patients with a PCP (model 3: OR, 2.1; 95% CI, 1.2-3.8; P=.01) and the 614 patients who were discharged from the ED (model 3: OR, 1.8; 95% CI, 1.1-3.0; P=.03). Formal tests of interaction between sex and these 2 factors were not statistically significant (P>.1 for both).

In this prospective multicenter study, we found significant sex differences among adults who presented to the ED with acute asthma. These patients were 1.8 times more likely to be women than men, a female-male ratio that exceeds that of patients in the ED overall (1.1)⁶ and that of asthma prevalence among adults (1.0-1.5).^2,12- 13 Furthermore, although men received less outpatient care and had worse pulmonary function in the ED, women were more likely to be admitted to the hospital and to report an ongoing exacerbation at 2-week follow-up.

Sex differences regarding asthma begin in early childhood, with evidence of greater asthma incidence among younger boys than among girls; reversal of this relationship occurs sometime during adolescence.^14- 15 Men in the present study were more likely to report asthma diagnosis during the first decade of life, whereas women were more likely to report asthma diagnosis thereafter. Male predominance in childhood asthma has been attributed to boys having smaller airway diameters relative to lung volume (dysanapsis)¹⁶ and more allergen sensitivities.¹⁷ Another possible explanation involves the "Yentl syndrome," ie, the undertreatment of women compared with men unless women develop severe manifestation of disease or a "typical" disease presentation.¹⁸ In recent cross-sectional surveys,^19- 21 there was substantial underdiagnosis and undertreatment of asthma among girls.

Pubertal changes in sex hormones coincide with a late adolescent increase of asthma in girls, which suggests that estrogen or progesterone may cause asthma. The literature^22- 27 is contradictory, however, and a coherent hypothesis is lacking. Other investigators invoke women's increased likelihood of indoor environmental exposures³ and greater atopy prevalence²⁸; the latter observation is supported by our hay fever data. Results of cross-sectional population studies²⁹ also show more bronchial hyperreactivity among women than among men. In terms of health care use, results of a recent cross-sectional study³⁰ show that asthmatic women aged 35 to 55 years are more likely to report previous ED treatment for asthma than their male counterparts; the authors did not find this sex difference among younger age groups. More research is needed on the biological or social basis for sex differences in the epidemiology of asthma.

With respect to outpatient management of adults who presented to the ED with acute asthma, our study revealed important sex differences. Women were more likely than men to be insured and to have a PCP, despite a lack of significant sex differences in age, level of education, or estimated household income. Use of preventive medicine is paramount in primary care and, accordingly, women's increased contact with PCPs explained why a larger proportion of women took inhaled corticosteroids, as recommended by National Asthma Education and Prevention Program guidelines.¹¹ As long as ED personnel do not assume some responsibility for the long-term care of their asthmatic population, PCP status will remain the strongest predictor of whether an asthmatic patient in the ED is taking inhaled corticosteroids.³¹

We also noted large sex differences in the acute asthma visit; women presented to the ED with less severe exacerbations than men. Women's initial and final percents of predicted PEFR were significantly higher than those of men, which is consistent with the results of a recent study by Awadh and colleagues⁵ in Vancouver. In this relatively small retrospective study (n=137), initial PEFR did not differ significantly between women and men (46±20 vs 40±20; P=.11), but initial forced expiratory volume in 1 second did (49±20 vs 33±15; P=.001). At the time of ED disposition, either discharge or admission to the hospital, forced expiratory volume in 1 second remained higher among women compared with men (68±20 vs 58±19; P=.03).

Despite such objective data showing that men had more severe asthma exacerbations than women, we found that women were admitted to the hospital twice as often. Awadh et al⁵ found insufficient statistical power to address sex differences in admission rates, but several other groups^32- 37 report increased admission rates among women. For example, billing data from southeastern Pennsylvania show nearly 3 times as many asthma hospitalizations among women than men.³³ Some^32- 33 but not all³⁵ of these studies also report that women had slightly longer hospital stays, which was taken as evidence that women with asthma have more severe asthma flares. In our prospective multicenter study, we also found that women tended to have longer hospital stays (P=.07), despite their better pulmonary function before admission.

Whether women are "overtreated" or men are "undertreated" is uncertain. If women are overtreated, this would be the opposite of female undertreatment for childhood asthma.^19- 21 Overtreatment would resemble, however, patterns seen for several other adult diseases for which women tend to get more care for the same type of illness or complaint unless care involves a major diagnostic or therapeutic intervention.³⁸ Alternatively, perhaps men are being undertreated because they tend to ignore symptoms and may be more likely to decline a recommended hospital admission. If all patients who left the ED against medical advice were admitted to the hospital, the sex difference would be attenuated but remain significant. However, we are unable to account for "borderline" hospital admission decisions that, if declined, may not have been classified as having left the ED against medical advice.

Another possible explanation for our paradoxical PEFR admission finding might be that emergency physicians are not familiar with sex differences in absolute PEFR. In other words, emergency physicians may not be aware of large differences in absolute PEFR between healthy women and men.¹⁰ For example, a PEFR of 300 L/min represents approximately 70% to 80% of the number predicted for a typical 35-year-old woman but only 50% to 55% of the number predicted for a typical 35-year-old man. Likewise, 70% of predicted PEFR is approximately 280 L/min for a typical woman vs 400 L/min for a typical man. If male norms were applied to all patients—without regard to sex—women would seem more ill than they really were and, consequently, would get admitted more often to the hospital.

At 2-week follow-up, it seemed that women and men were at similar risk of acute asthma relapse but women were 50% more likely to report an ongoing asthma exacerbation. In light of women's better outpatient care, less severe obstruction at ED presentation, and increased likelihood of hospitalization (and its presumed benefits), we were surprised by the latter finding. Women may be more likely to use an inadequate inhaler technique,³⁹ but we found the same self-reported compliance with taking systemic corticosteroids, and we doubt that medication-related differences can explain our finding. Instead, we believe that sex differences in the way mild symptoms are perceived, evaluated, and acted on⁴⁰ provide a more likely explanation. Women may be more likely than men to perceive airway obstruction as dyspnea⁴¹ and to link psychological issues with bodily well-being.⁴² Indeed, women in our study reported a greater number of known asthma triggers, with psychosocial factors especially common among women compared with men. In another setting, Osborne et al³⁰ recently found that asthmatic women reported more symptoms and poorer quality of life than men, although measures of airway obstruction were comparable. These observations have led some physicians to erroneously attribute women's health complaints to emotional rather than physical causes.³⁸ Thus, it may be preferable to view men as underanxious patients who are less aware of their bodies or more likely to deny their symptoms rather than viewing women as "different" from a male-based norm.

The present study has several potential limitations. First, PEFR measurement was not standardized across sites. There is, however, no reason to believe that there were sex-related differences, particularly of this magnitude, in the accuracy of PEFR measurement. Second, ED management and the decision on hospital admission were left to the discretion of the emergency physician; there undoubtedly were different approaches to acute asthma management across EDs. To address this variability, we collected information about all asthma therapies recommended in the National Asthma Education and Prevention Program guidelines (eg, β-agonists, corticosteroids, and anticholinergic agents)¹¹ and found that adjusting for differences in treatment during the ED visit did not materially change our findings. Third, we had limited information about the inpatient course. A longer inpatient length of stay by women seemed to contradict the PEFR findings, but length of stay is a crude measure of disease severity because it is dependent on many other factors. Fourth, we were unable to examine the role of interviewer factors, such as interviewer sex and training, on patient response. A variety of women and men performed the structured interviews, and the effect of interviewer factors on subjective outcomes (like treatment failure) is unknown. Last, we were unable to confirm self-reports of an ongoing asthma exacerbation (eg, by comparing the patient's PEFR with that at time of ED discharge). This shortcoming does not allow us to comment on the biological aspects of patients' disease, but quality-of-life issues, such as "severe" discomfort and "severe" activity limitations, are worthy of our concern.

In summary, we found large sex differences among adults presenting to the ED with acute asthma. The female majority has important pathophysiological implications, but it may also represent sex differences in health care use. Men and women differed on a variety of chronic and acute asthma factors related to biological aspects of asthma and its process of care. The paradoxical PEFR admission findings support further investigation of how PEFR values are used in the ED and greater caution in the interpretation of asthma hospitalization data; reported sex differences in asthma hospitalization probably overstate the true differences in acute asthma severity. Moreover, our data suggest that we should pay greater attention after the ED visit to quality-of-life issues as opposed to more traditional end points such as relapse. For unclear reasons, increased access to outpatient care, higher PEFR values in the ED, and higher admission rates did not produce a female advantage at 2-week follow-up. Further studies are needed to better understand the complex relationship between sex and acute asthma.

Accepted for publication October 8, 1998.

This study was supported by grants HL-07427 (Dr Woodruff) and HL-03533 (Dr Camargo) from the National Institutes of Health, Bethesda, Md, and by unrestricted grants from Glaxo Wellcome Inc, Research Triangle Park, NC, and Monaghan Medical Corp, Syracuse, NY.

We thank Frank Speizer, MD, for his helpful suggestions and all of the Multicenter Asthma Research Collaboration investigators for their ongoing dedication to emergency asthma research.

Multicenter Asthma Research Collaboration Steering Committee members: Carlos A. Camargo, Jr, MD (chair); Rita K. Cydulka, MD; Michael A. Gibbs, MD; Robert A. Silverman, MD; and Janice L. Zimmerman, MD.

Operations Committee and Data Coordinating Center members: Carlos A. Camargo, Jr, MD (chair); George T. Mathew; Leo T. Mayer; Anita K. Singh; and Prescott G. Woodruff, MD, Massachusetts General Hospital, Boston.

Principal investigators at the 36 participating sites: Courtney A. Bethel, MD, MPH (Mercy Hospital, Philadelphia, Pa); Leonard Bielory, MD (University Hospital, Newark, NJ); Michelle P. Blanda, MD (Summa Health System, Akron, Ohio); Barry E. Brenner, MD, PhD (Brooklyn Hospital Center, Brooklyn, NY); Carlos A. Camargo, Jr, MD, DrPH (Massachusetts General Hospital, Boston); Rita K. Cydulka, MD (MetroHealth Medical Center, Cleveland, Ohio); Daniel J. Dire, MD (University of Oklahoma Medical Center, Oklahoma City); Nabil El Sanadi, MD (Broward General Hospital, Fort Lauderdale, Fla); Stephen D. Emond, MD (St. Luke's/Roosevelt Hospital Center, New York, NY); Ted J. Gaeta, DO (St. Barnabas Hospital, Bronx, NY); Michael A. Gibbs, MD (Carolinas Medical Center, Charlotte, NC); CPT Theodore E. Glynn, MD (Brooke Army Medical Center, Fort Sam Houston, Tex); Louis G. Graff IV, MD (New Britain General Hospital, New Britain, Conn); Richard O. Gray, MD (Hennepin County Medical Center, Minneapolis, Minn); John P. Hanrahan, MD (Beth Israel Hospital, Boston); Fred Harchelroad, MD, DABMT (Allegheny General Hospital, Pittsburgh, Pa); Ahamed H. Idris, MD (University of Florida Health Center, Gainesville); Marian E. Johnson, MD (Jackson Memorial Hospital, Miami, Fla); Luis F. Lobon, MD (Beth Israel Medical Center, New York); Michael F. McDermott, MD (Cook County Hospital, Chicago, Ill); Eric S. Nadel, MD (Brigham and Women's Hospital, Boston); Richard M. Nowak, MD (Henry Ford Hospital, Detroit, Mich); Edward Paul, MD (Charity Hospital, New Orleans, La); Charles V. Pollack, Jr, MD, MA (Maricopa Medical Center, Phoenix, Ariz); David J. Robinson, MD (University of Maryland Medical Center, Baltimore); Robert M. Rodriguez, MD (Southwestern Medical Center, Dallas, Tex); Gail Rudnitsky, MD (Allegheny University—MCP Division, Philadelphia); Robert E. Sapien, MD (University of New Mexico Health Sciences Center, Albuquerque, NM); Donald Schreiber, MD (Stanford University Medical Center, Stanford, Calif); Robert A. Silverman, MD (Long Island Jewish Medical Center, New Hyde Park, NY); Sarah A. Stahmer, MD (Hospital of the University of Pennsylvania, Philadelphia); Andrew Sucov, MD (Strong Memorial Hospital, Rochester, NY); David M. Taylor, MD (University of Pittsburgh Medical Center, Pittsburgh); Carol A. Terregino, MD (Cooper Hospital/University Medical Center, Camden, NJ); Debbie Travers, RN (University of North Carolina Hospitals, Chapel Hill); and Janice L. Zimmerman, MD (Ben Taub General Hospital, Houston, Tex).

Reprints: Carlos Camargo, MD, Department of Emergency Medicine, Clinics 116, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114 (e-mail: marc@mgh.harvard.edu).