Intended for healthcare professionals


Asthma in children: epidemiology

BMJ 1994; 308 doi: (Published 18 June 1994) Cite this as: BMJ 1994;308:1584
  1. P D Phelan

    Asthma is one of the commonest childhood illnesses, and its prevalence and severity may be increasing. Determining whether this is so has been difficult, however, because of the lack of a widely accepted epidemiological definition of asthma and an objective measure of asthma that is applicable to children and is reasonably sensitive and specific. Furthermore, many studies purporting to show an increase in prevalence have used different measures and studied far from comparable populations over the period of the suggested increase.

    Several valid studies have now reported a substantial increase in the prevalence of asthma in some countries. In 1991 Robertson and colleagues in Melbourne showed a doubling over 26 years, to 46%, in the prevalence of children with wheeze on one or more occasions before the age of 7; 21- 23% had wheezed in the previous 12 months.1 These findings have been confirmed by Peat and colleagues, who studied children aged 8-10 in two country towns in New South Wales (p 1591).2 They found about a doubling over 10 years, to 23-27%, in the prevalence of children reporting wheeze in the previous 12 months. This is not solely an Australian phenomenon - Anderson et al report a more modest increase over 13 years, to 13%, in the prevalence of children aged 7.5 to 8.5 who reported wheeze in the previous 12 months (p 1600).3

    The study by Peat and colleagues is particularly important as it reports objective measurements whereas the other studies were based on responses to a questionnaire. It showed that airway hyperresponsiveness increased 1.4-fold to twofold. Although airway hyperresponsiveness is not equivalent to asthma (8-15% of children who have never wheezed will show bronchial hyperresponsiveness and up to 30% with typical asthma will fail to show it on several occasions), its increase supports the validity of the questionnaire completed by parents. Interestingly, this study found that airway hyperresponsiveness increased mainly in children with other evidence of atopy, although the prevalence of atopy was unchanged.

    Even when differences in methodology are allowed for, there is considerable international variation in the prevalence of childhood asthma, defined as wheeze, within the past 12 months. Similar high rates to those reported in Australia have been noted when almost identical questionnaires were used in Fiji4 and in La Serena, a non-polluted town in northern Chile.5 In Switzerland the prevalence is much lower at about 7%5 and Germany has rates of 4% to 6%.6

    The reasons for this increase are not known. As genetic change seems highly unlikely some environmental factors must be responsible. The increase has occurred during a time when general atmospheric pollution has fallen, and high prevalences were seen in Fiji and northern Chile - where pollution is not a problem. The prevalence was lower in Leipzig (formerly in East Germany) than in Munich.6 Bronchodilating drugs have been relatively little used in Fiji so attributing the high prevalence to some effect from them is difficult. Peat and colleagues suggest, with some supporting data, that the increased responsiveness in atopic children may be due to a much greater allergen load in a child's environment.2 compatible with such a mechanism is the lower frequency of wheeze in the Swiss mountains.

    The studies of Peat et al and Anderson et al indicated that the number of episodes of wheeze in affected children had increased,2,3 but Anderson et al's study suggested, however, that the morbidity from asthma as indicated by days lost from school and interference with normal activities had fallen substantially over the 13 years of study. Hyndman and colleagues in a study from East Anglia found that the increase in admission to hospital of children with asthma, which has been reported from many countries, may be reaching a plateau or even falling (p 1596).7 These are reassuring findings and could well be due to more effective treatment in those with recognised asthma. The greater willingness of doctors to diagnose asthma in children with recurrent wheeze is probably the single most important factor in improving treatment.8

    The increase in the prevalence of asthma in children may well have serious implications for adults as 40% of children with infrequent trivial wheeze and 70-90% of those with more troublesome asthma continue to have symptoms in mid-adult life.9 As yet studies have not shown an increase in asthma in elderly people, although it may be increasing in young adults.10 If the increase in the prevalence in children continues into adult life this will create substantial health problems and may well have important economic consequences through time off work and the cost of treatment.

    Prophylactic drugs, particularly inhaled corticosteroids, might be expected to increase the likelihood of childhood asthma resolving, yet there is no evidence to support this. The long term use of inhaled corticosteroids in children with more troublesome asthma has failed to alter permanently the long term course of the disease.11 When inhaled steroids were stopped after 24-36 months of treatment the pattern of asthma rapidly returned to the pattern that had been present before they were started. Furthermore, increasing evidence exists that even low dose inhaled corticosteroids may have a prolonged growth suppressant effect in children with mild asthma.12 The more widespread use of the currently available inhaled corticosteroids in the hope of modifying the natural course of asthma cannot be recommended. Their use should be limited to those children in whom asthma substantially interferes with the enjoyment of a normal lifestyle or in whom there is persisting airways obstruction. Whether the newer inhaled corticosteroids will have fewer side effects, particularly effects on growth, remains to be determined.

    Ways to limit the increase in the prevalence of asthma and if possible to reduce it are clearly vitally important. Until the reasons for the observed increase are determined, however, it is not possible to suggest intervention strategies. The hypothesis of Peat and colleagues that the likely explanation is a change in the degree of exposure to allergens is interesting and is further discussed by Cullinan and Newman Taylor below (p 1585).13 Reducing exposure to allergens, particularly to house dust mites, is expensive and substantially alters lifestyles. It is too early to recommend interventions in the absence of proof of efficacy and benefit.