Sex differences in the clinical presentation and management of airflow obstruction

The most common causes of airflow obstruction in primary-care practices are chronic obstructive lung disease (COPD), characterised by progressive, partially reversible airway obstruction, and asthma, characterised by variable airflow limitation 1, 2. Previous studies have shown that the clinical diagnosis of chronic bronchitis is made more commonly in females than males, and emphysema is more commonly diagnosed in males than females 3. Hypothetical case presentations to primary-care physicians reveal that, for the same clinical history, males were more likely to be diagnosed with COPD and females with asthma 4. Whether or not females are more susceptible to cigarette smoke than males is controversial. A recent study in a pulmonary clinic matched females to males on forced expiratory volume in one second (FEV₁) per cent predicted 5. Compared with males, females were younger, smoked less, were more breathless and reported poorer quality of life scores. The present study explores differences in clinical expression of airway disease, diagnoses and management in a primary-care setting. The study group was not selected based on the presence or absence of respiratory symptoms or known lung disease.

METHODS

Statistical analysis

Two different criteria were used for determining the prevalence of airflow obstruction. The first was an FEV₁/FVC <70% pred, the clinical threshold criteria popularised by the Global Initiative for Chronic Obstructive Lung Disease guidelines for the diagnosis of airflow obstruction 8. The second definition was a pre-bronchodilator FEV₁/FVC1 less than the lower limits of normal (LLN), based on reference equations determined by the third National Health and Nutrition Examination Survey spirometric data 9. Results were stratified by sex and differences were tested either with an independent unpaired t-test or Chi-squared statistics.

RESULTS

The study included eight communities with populations between 1,206 and 4,406, of which 1,046 subjects were enrolled. A total of 1,034 subjects had acceptable and reproducible spirometry, of whom 550 (53%) were males and 484 (47%) were females. There was no significant difference in prevalence of airflow obstruction between the eight study sites (Chi-squared, p = 0.17). The mean±sd age of the study patients was 59±12.7 yrs, with patients having a smoking history mean±sd of 35.5±28.5 pack-yrs.

To determine the degree to which the study group was representative of all eligible subjects, the entire clinical population that visited the eight primary-care practices was surveyed for several days. Of the 1,800 subjects who were 100% sampled, 561 were ≥35 yrs of age and had ever smoked. Compared with all of those eligible (aged ≥35 yrs and had ever smoked), the group studied using spirometry differed as follows: 1) aged 2 yrs younger; 2) 1% more males; and 3) 1 yr extra smoking.

In the study group of 1,034, males were on average 5.5 yrs older and had smoked for 12 pack-yrs more than females (p<0.0001; table 1⇓). Females reported dyspnoea (p = 0.008) and wheeze (p = 0.031) more frequently. Based on responses to the ATS Questionnaire 6, females were twice as likely to have been diagnosed with asthma, 20 versus 10% (p<0.0001), respectively, and two-thirds more likely to have been diagnosed with chronic bronchitis, 19 versus 11% (p = 0.001), respectively. The prevalence of airflow obstruction was higher in males than females. This was statistically significant when defined by FEV₁/FVC <70% pred, but not significant when defined by the LLN criteria. In total, 123 (22%) males and 57 (12%) females had an FEV₁/FVC <70% pre-bronchodilator (p<0.0001). FEV₁/FVC less than LLN was present in 76 (14%) males and 54 (11%) females (p = 0.20). Females were also almost twice as likely to report using respiratory medications, 23 versus 15% (p = 0.0005), respectively.

When the analysis was restricted to only those with obstruction defined by the LLN criteria, females had smoked less than males, 39 versus 53 pack-yrs (p = 0.0008), respectively. They were also much less likely to report sputum, 17 versus 41%, respectively, (table 2⇓). There were no significant sex differences in reported respiratory diseases and medications.

Respiratory medication use was stratified by severity of obstruction from none to severe (table 3⇓). Females were twice as likely as males to be prescribed medications unless airway obstruction was severe. When severe, 70% of both males and females reported taking at least one respiratory medication.

To determine if airflow obstruction was more likely to be reversible in females than males, bronchodilator response was stratified by sex and severity of obstruction (table 4⇓). No significant differences were found between the sexes.

DISCUSSION

Females were more likely than males to report respiratory symptoms, which is consistent with the recent findings of de Torres et al. 5. In addition, the current authors found that females were more likely to have received a diagnosis of a chronic airway disease, and were more likely to be taking respiratory medications. One possible explanation is a sex difference in the severity or expression of the disease. More females than males reported breathlessness, which was not explained by age, smoking history or FEV₁. This relative increase in symptoms may explain the increased probability of being both diagnosed with lung disease and treated.

There also appears to be a sex-bias in the diagnosis of chronic bronchitis, which may influence population health statistics for respiratory disease. The defining symptom of bronchitis is mucus hypersecretion from the chest and does not require airflow obstruction to be present 10. Although males generally reported more sputum production, consistent with this diagnosis, females in the present study were more likely to be labelled as having chronic bronchitis. American national population database studies have also reported a greater prevalence of chronic bronchitis in females compared with males 3. The current results suggest that this observed difference may not reflect differences in mucus hypersecretion, but rather a physician diagnostic bias.

It is not possible to determine if the prevalence of asthma was truly higher in the female study population or whether there was a reporting bias. Self-reported asthma and bronchial hyperresponsiveness are known to be more common in adult females than adult males 11, 12. Chapman et al. 13 reported that clinicians presented with hypothetical cases were more likely to diagnose asthma in females than males despite similar age, smoking histories and symptoms. The present results also suggest a sex-bias in diagnosing asthma. Females were twice as likely as males to be diagnosed with asthma although bronchodilator responsiveness did not differ between the two groups. Females have been reported to use primary-care services with greater frequency, but to receive fewer specialist referrals than males and to be less likely than males to be referred for invasive cardiac procedures 4, 14, 15. These findings raise the expectation that females may also be relatively under-treated for respiratory disease. The current authors found the opposite to be true. Females were more likely to be prescribed respiratory medications until severe obstruction was present (table 3⇑). Males may be relatively under-treated at milder stages of airflow obstruction.

The two different criteria used to define airflow obstruction affected males and females differently in the present study. The FEV₁/FVC ratio decreases with age among healthy adults. Elderly people without respiratory disease may have a FEV₁/FVC within the LLN defined by a healthy reference population, yet have a FEV₁/FVC <70% pred. This scenario occurred more often in males than females in the present study because the mean age of males was 62 versus 56 yrs for females.

In conclusion, the current authors found that sex differences in symptom reporting, diagnoses and management of respiratory illness exist in primary-care practices. Physician awareness of this issue may help reduce this presumably unintentional bias. It may stem from a difference in the prevalence of breathlessness, a subjective indicator. Perhaps increased use of spirometry, an objective measure, would reduce the sex-related biases and improve diagnosis and management of airway diseases in the primary-care practice setting.