Article Text


Wheeze, asthma diagnosis and medication use in developing countries
  1. M R Becklake
  1. Correspondence to:
    Professor M R Becklake
    Respiratory Epidemiology and Clinical Research Unit, Montréal Chest Institute, 3650 St Urbain Street, Montréal, Québec, Canada H2X 2P4;

Statistics from

Sociodemographic, environmental, and medical predictors in a national adult survey conducted in South Africa

Standardised population based health surveys have produced much useful information on the prevalence and determinants of asthma in children (International Study of Asthma and Allergy in Children (ISAAC) study)1 and adults (European Community Respiratory Health Study (ECRHS) study).2 Most sites participating in these studies have been in high income countries and, because some centres took part at their own expense, the populations studied were probably not representative of their countries as a whole. In addition, a number of national health surveys such the US National Health and Nutrition Survey (NHANES),3 the Canadian National Health Survey,4 and National Birth Cohort Studies (UK, New Zealand)5,6 have been conducted, all in high income countries.

In this issue of Thorax Ehrlich et al,7 in what is one of the first such studies in a developing (in this case middle income) country, describe the results of a national health survey conducted among 13 826 South Africans aged 15 years and over to determine the national prevalence and predictors of three outcomes: (1) wheezing symptoms, (2) an asthma diagnosis, and (3) use of asthma medication. This was an ambitious undertaking (and a remarkably successful one) given the racial, linguistic, and socioeconomic heterogeneity of the population and its quadruple burden of HIV, other infections (including tuberculosis), chronic non-communicable disease, and injury. What is surprising is not that the findings in this paper7 and in its companion paper on chronic bronchitis8 were different from those published for developed (high income) countries, but that in how many respects they were similar.


Great care was taken by Ehrlich and colleagues to maximise the quality of the data gathered, as evidenced by (1) the use of a national sampling frame (the South African Demographic and Health Survey) designed to give cross sectional data on a nationally representative sample of the non-institutionalised population; (2) validation of reported recent wheeze against peak expiratory flow (PEF) measurements: in men and women with recent wheeze % predicted PEF was 4–11% lower than in those without recent wheeze (see table 1 in their paper7); (3) although only values for recent wheeze were reported in the paper, its validity was supported by similar findings for three other asthma related symptoms—wheezing with breathlessness, wheezing not only with a cold, and wheezing at night; (4) the absence of proxy reporting while still achieving a high response rate (92% in women, 86% in men); (5) visual verification of current medications by interviewers and their subsequent classification by a health professional; and (6) internal standardisation of PEF values based on sex, age, height and weight (table 1 in their paper7).


The sex differences in the prevalence of the three main outcomes studied by Ehrlich and colleagues are of interest. For recent wheeze there was an excess in women compared with men (OR 1.6, 95% CI 1.3 to 1.8); for an asthma diagnosis there was no significant difference (OR 1.2, 95% CI 0.9 to 1.6); while for current asthma medication use the rates were lower in women than in men (OR 0.6, 95% CI 0.4 to 0.9) (table 3 in their paper7). These rates refer to adults aged 15 years and over.

For the purposes of comparison with countries using the ECRHS questions and protocol which covered patients aged 20–44 years, Ehrlich and colleagues7 reported separately on subjects within this age range. The prevalence rates of an asthma diagnosis were 3.1% in women and 2.7% in men, below the ECRHS median (4.5%) and close to its lower range (2.0%).9 However, the question used in the study by Ehrlich et al to establish a diagnosis of asthma (“Have you ever been told by a doctor, nurse or other health professional that you have asthma?”7) is not the same as the question used in the ECRHS study (“Have you had an attack of asthma in the last 12 months?”10). Thus, recent wheeze (in the last 12 months) may be a more useful outcome than an asthma diagnosis for comparing the results of the study of Ehrlich and colleagues7 with those from high income countries using the ECRHS questions and protocol. Additional justification for use of current wheeze as the outcome is that it has been validated against PEF measurements. This also allowed the authors to avoid use of the term “asthma” in their questionnaire, for which differences in validity were found for some of the ECRHS sites.2,9

The prevalence of recent wheeze in the 20–44 year old age range (that is, the ECRHS age range) of 13.4% in men and women combined was below the ECRHS median (20.7%). However, when stratified by sex, the prevalence of recent wheeze was higher in women than in men (17.6% v 14.4%). Rates were also higher in women than in men in some low income ECRHS sites—for example, in Mombai (Bombay) in urban India (4.7% in women, 4.4% in men in a study of 2313 adults11) and in Tirana, Albania (7.3% in women, 6.0% in men in a study of 2653 adults12). Although recent wheeze was not reported in an analysis of the combined data pertaining to 18 659 subjects from 16 high income ECRHS sites in Europe, the US and Australia,13 the prevalence of reported asthma was significantly higher in women than in men (RR 1.3–1.8). This pattern was consistent in all 16 countries.

Ehrlich and colleagues speculated that the sex differences in the prevalence of wheeze and reported asthma in their study were artefactual, due to the oversampling of women and/or the undersampling of men away from home at the time of the interviewer’s visits.7 However, this age range also corresponds to the reproductive period of a woman’s life. The fact that the female excess in the prevalence of recent wheeze was strengthened from an OR of 1.3 (95% CI 1.1 to 1.4) in the univariate analysis to an OR of 1.6 (95% CI 1.3 to 1.8) in the logistic regression model, thus adjusting for all the other risk factors studied, suggests that a biological explanation is more likely.

There are recognised sex differences in the morphology and physiology of human airways which affect airway behaviour during a woman’s reproductive years.14,15 These include: (1) dimensional factors: girls and women have larger airways in relation to their airspaces than boys and men throughout the human life span; (2) immunological factors: atopy, measured as skin prick test positivity to common allergens, is also higher in women in their reproductive years than in men; and (3) hormonal factors: these are probably the most important determinants of airway behaviour in women throughout their reproductive years, imposing variations related to the menstrual cycle, oral contraceptives, pregnancy, and postmenopausal hormone replacement.


In addition to sex, the sociodemographic predictors of recent wheeze examined in the study by Ehrlich et al7 included age, household asset index (based on 10 household characteristics such as electricity, water and flush toilets, and reporting that nobody in the household ever went hungry), education, racial group, urban versus rural residence, and medical insurance coverage (table 3 in their paper). For recent wheeze the important predictors were female sex (discussed above), education (a monotonic decrease with increasing years of education), and race (Africans and coloureds exhibited lower rates than whites and Asians).

The protective effect of education for recent wheeze in the study of Ehrlich and colleagues7 is a provocative finding. The authors suggest that this effect may be partly explained by cultural and linguistic differences in responses to symptom questions, amplified by a wide gap in education within the study population. However, there is also a plausible alternative explanation.

Although recent wheeze showed a strong inverse relationship with years of education, there was no relationship with wealth as reflected in the asset index, an index carefully constructed to reflect living conditions and household assets. A similar protective effect of years of education, together with wealth (being in the highest quintile of asset index) was also shown for chronic bronchitis in the authors’ companion paper to their present study carried out in the same study population,8 and in two other studies cited in their paper—one for chronic bronchitis in Brazil16 and one for obstructive lung disease and spirometric airflow limitation in Norway.17 This “protective” effect for different respiratory outcomes in countries with such different economies (the gross domestic product per capita in South Africa, Brazil and Norway was $8466, $6477 and $27 557, respectively18), different life expectancies (48.0, 63.2 and 78.9 years, respectively18), and different literacy rates (82%, 83% and 100%, respectively18) suggests that the “protective” effect of education is a more universal phenomenon than is generally thought.


In their study, Ehrlich et al7 identified occupational exposure and, to their surprise, a past history of tuberculosis as predictors for recent wheeze (table 3 of their paper). The estimated contributions of occupational exposure, expressed as population attributable fraction) for recent wheeze and for asthma were 12.2% and 13.6%, respectively, close to the median estimate of 15% for asthma published in a recent American Thoracic Society statement based on a review of European and US data.19

Although Ehrlich and colleagues note that the PAF of 12.2% for occupational exposure in respect of recent wheeze approximates that of light and heavy smoking combined (11.5%), they do not comment specifically on the role of smoking. Others have called cigarette smoking and asthma “a dangerous mix”.20 These authors draw attention to the fact that, in Canada for instance, smoking is as common in those suffering from asthma as it is in the general population, including the 20–35 year age group. Active smoking is also associated with increased asthma morbidity and impairment of the response to steroids, leading the authors to call for a reassessment of guidelines for asthma treatment.20

In their national survey, Ehrlich and colleagues7 captured what had already been described in black gold miners21 (that tuberculosis can cause chronic impairment of lung function) and what was subsequently described in black coal miners22 in whom a past history of tuberculosis accounted for reductions in the level of FEV1 and FVC % predicted of 20% and 14%, respectively. They also advocate for greater recognition of post-tuberculous obstructive lung disease. This implies that the clinical history of the adult South African who wheezes should always include detailed questioning on past exposure to tuberculosis, tuberculin testing, and BCG administration as well as having contracted the disease. Their message may be relevant for other countries with high rates of tuberculosis, HIV/AIDS, or both. Their recommendation for increased future research into the mechanisms certainly deserves widespread support.


In addition to establishing the national prevalence of wheeze, an asthma diagnosis and asthma medication use in South Africa, the study by Ehrlich and colleagues has led to four interesting observations:

  • For the outcome “current wheeze”, the most important determinants in adults aged 15 years and older (expressed as population attributable fraction %) were occupational exposure (12.2%), light (7.8%) and heavy smoking (3.7%), tuberculosis ever (5.4%), and obesity (3.8%). All were significant except obesity.

  • Sex differences in the prevalence of asthma (higher rates in women in their reproductive years than in men) were confirmed in the South African population. They are most unlikely to be artefactual.

  • Control of hazardous occupational exposure merits equal attention to the more visible tobacco control activities in the public health institutes in South Africa. The authors also make the case for greater clinical recognition of post-tuberculous obstructive lung disease. These conclusions may be applicable to other populations in high as well as in low and middle income countries.

  • In South Africa (which Erlich et al characterise as a medium income country but with large scale urban poverty and rural underdevelopment), education was protective for the outcome “recent wheeze”, a relationship not accounted for by a marker of “worldly wealth” in the form of the carefully constructed asset index developed for the national health survey. This was a consistent finding for which the reasons and/or mechanisms are not obvious and one which invites further investigation.


I am grateful for comments and advice from Professors Maurice McGregor and Richard Menzies, McGill University.

Sociodemographic, environmental, and medical predictors in a national adult survey conducted in South Africa


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Airwaves
    Wisia Wedzicha