20 Swedish National Fod Administration. Swedish Food ReguLlationis. Food
additives. Uppsala: Swedish Nationali Food Administration, 19X5.
21 Fine DH, Ross R, Rooinhehler DlP, Silvcrgleid A, Song L. Formation in vivo
of volatile N-nitrosamines in man aftcr ingestion of cookcd hacon and
spinach. Nature 1977;265:753-5.
22 Osterdahl BG. Volatile nitrosamines in foods on the Swedish market and
estimation of their dails intake. FtoodAddit Contam 1988;5:587-95.
23 Wagner DA, Shuker D)EG, Bilmazes C, et al. Effect of v%itamin C and E.
on endogenotis synthesis of N-nitrosamino-acids in humans: precursor-
product studies with (' N) nitrate. Cancer Res 1985;45:6519-22.
24 Jenkins D)JA, Wolever 'rMs, Jenkins AL,
carbohydrate metabolism and diabetes. Mo/lAspectsMed 1987;9:97-1,12.
25 AMiettinen 0. Components of the crtude risk ratio. Am J Epidemiol 1972;96:
26 Klecka WR. Discriminant analysis. In: Nie NH, Hull CH, Jenkins JG,
laylor RH. Dietary fibre,
Steinhrenncr K, Bent DH, cds.Starisnical packageforih.social sclent(cs 2nid
ed. Ncw York: McGraw-Hill, 1975:434-47.
27 Miettinen 0. Estimahility anid cstimation in cast-referent studies. Am 7
28 Rakicten N, Rakicten ML, Nadkarni MV. Studics on the dia!sctogenic action
of streptozotocin. Cancer Chemoiheraps' Reports 1963;29:91 -8.
29 Scott FW, Daneman 1), Martin JM. Evidence for a critical rolc of diet in tlc
development of insulin-dependent diahetes incillitis. Diabeties Ris 1988;7:
30 Nerup J, Mandrup-PoulsenT, Molvig J, Helqvist 8, S
Mechanisms of pancreatic B-cell destructioin in type I diahetes. Dlsahts
31 'Iibblin G. A pipulation studv oii 50-year-old-men. An analsis ot the nuoi-
participating grotip. Acia Med St-and 1965;178:453-9.
'ogcnscn L, Egebergi.
(Accepted 6 March 1990)
Has the prevalence of asthma increased in children? Evidence
from the national study of health and growth 1973-86
P GJ Burney, S Chinn, RJ Rona
Objectives-To estimate changes in the preva-
lence of reported symptoms of respiratory disease
and reported diagnoses of asthma and bronchitis in
primary school children in England between -1973
Design-Mixed longitudinal survey.
primary schools in 22 areas.
Participants-15 000 Boys and 14156 girls each
studied at least once between 1973 and 1986.
Data collected-Whether, according to the parent
or guardian, the child had wheezed, wheezed on
most days or nights, or had attacks of bronchitis or
asthma in the past year.
Results-Within age groups trends in successive
annual cohorts showed an increasing prevalence of
asthma for each annual birth cohort (boys, 6-9%,
p<0001; girls, 12-8%, p<OOOl) and of wheeze
on most days or nights (boys, 4-3% per cohort,
p<0-001; girls, 6-1% per cohort, p<0001) and a
falling prevalence of bronchitis (boys, -4-7% per
cohort, p<0001; girls, -5-8% per cohort, p<0001).
There was a smaller increase in the prevalence of
wheeze whether or not it occurred on most days or
nights, and this increase was significant only among
the girls (boys, 1-0% per cohort, p>005; girls, 1-7%
per cohort, p<O0O5). Although the rate ofincrease of
"asthma" was greater than the rate of decrease in
"bronchitis," the baseline prevalence ofasthma was
much lower than that of bronchitis, and the total
proportion ofchildren with either diagnosis declined
slightly over the whole period. The main change was
an increase in the proportion of children whose
parents stated that they had persistent wheeze and
yet did not have a report of either "asthma" or
Conclusions-These results suggest that there has
been a true increase in morbidity that is not simply
due to changes in diagnostic fashion. The increase is
large enough to explain much ifnot all ofthe increase
in admission to hospital and mortality, and
underlines the importance ofan understanding ofthe
aetiology of asthma in tackling the causes of the
Mortality from asthma among 5-34 year olds
increased between the mid-1970s and the mid-1980s.'
This increase was the more surprising as mortality had
been falling in most conditions for which there was
effective prevention or treatment.2 At the same time
more children with asthma had been admitted to
hospital34 and consultations with general practitioners
for asthma virtually doubled between 1970-1 and
Each of these changes could be due to a change in
medical practice, including a change in the use of
diagnostic labels, or to a change in the prevalence ofthe
disease. There is some indirect evidence that the
prevalence of asthma may have been increasing, but
other evidence suggests that there has been no change.
Smith found an increase in the prevalence ofasthma in
Birmingham in -the 1960s,6 but this evidence was
collected at a time of great demographic change
when the city was being largely rebuilt. Wadsworth
compared those who had been studied in the 1946
national perinatal survey with their firstborn offspring
and showed that. the children had a threefold greater
chance of having been treated for asthma before their
fifth birthday,7 but this could have been explained by
differences in management or differences in diagnostic
practice. Hill- et al estimated that the prevalence of
.1-28% between 1985 and 1988 and suggested that
the large increase in estimated '.'asthma" was due
to a change in diagnostic labelling.8 More recently
Burr et al showed increases between 1973 and.1988 in
both symptoms and bronchial response to exercise in
12 year old children attending schools in Caerphilly.9
other hand, Hay and Higginbottam,
Anderson, and Hill et alhave all pointed to the lack of
any trend over time in the results from published
surveys reporting the prevalence of asthma.'0'2 As
Anderson points out, however, the interpretation of
this evidence is difficult because the methods used
in each survey are different and estimates of the
prevalence ofasthma are likely to be sensitive to these
differences.'3 Geographical variation in the prevalence
of asthma would also confound any estimate of trend
from these data.
that asthma may be an increasing
problem are not confined to the United Kingdom.
Upward trends in mortality have also been noted
-in New Zealand, France, Germany; Denmark, and
possibly the United States.4'-6 Upward trends in
hospital admissions have also been noted in New
Zealand and the United States.'7
This paper reports trends in the prevalence of
respiratory conditions, including both diagnoses and
symptoms, reported by the parents and guardians of
children in the-national study of health and growth
between 1973 and 1986. Estimates of the trends in
prevalence have been made for cohorts of children
living in England and born between 1961 and 1981.
Health Medicine, United
Medical and Dental
Schools ofGuy's and
St Thomas's Hospitals,
St Thomas's Hospital,
London SEI 7EH
P G J Burney, FFPHM, senior
lecturer in public health
S Chinn, MA, senior lecturer in
R J Rona, MFPHM, senior
lecturer in public health
BrMedJ 1990;300: 1306-10