Year: 2007
Type: review
Study design: Observational Study
Abstract: Asthma and other wheezing disorders are the most common chronic health problems in childhood and place a large burden on children, their families and society.1 Prevalence is highest in infancy or early childhood and the aetiology is complex, with a strong influence of intrauterine and early life exposures. Clinical presentation, response to treatment and prognosis differ by age and the natural history is highly variable. It has been questioned whether asthma, especially in young children, should be regarded as one disease with a single underlying aetiology but a wide range of severity,2 or as a syndrome comprising several separate conditions.3,4 Furthermore, respiratory illness in early life is associated with adult respiratory disease and diminished lung function.5,6 Despite this, most epidemiological studies of asthma before the 1990s had focused on schoolchildren and adults, and there were sparse population-based data on infants and preschool children. The Leicester respiratory cohort studies were set up to fill this gap. The first Leicester cohort was established in 1990 by Hamish Simpson (with David Luyt and Adrian Brooke as principal fellows) as a community-based sample of 1650 children born between 1985 and 1990. This study provided the first data on prevalence7–9 and subsequently on natural history of preschool wheeze in Europe.10–12 The large age-spread of the Leicester 1990 cohort permitted an overview of the prevalence and characteristics of wheeze in preschool children, but complicated studying the natural history. This cohort was relatively small and did not include ethnic minority groups. Therefore, Michael Silverman and Claudia Kuehni recruited a second cohort in 1998, consisting of 8700 children born between 1993 and 1997. A stratified sampling design was used to recruit 5400 children aged 1, and 3300 children aged 2–4 years. Of these, 6100 were white and 2600 were south Asians. The direct costs of the cohort studies (salaries, consumables and equipment) came from short-term project funding from bodies such as Asthma UK, The Wellcome Trust, the Swiss National Science Foundation, R&D Directorates of NHS Trusts, Children's Research Fund, Medisearch (Leicester), BUPA Foundation and many others. Infrastructure is provided by the Universities of Leicester, UK and Berne, Switzerland, where the data centre is located. Leicester City Council is a partner in research into air pollution. Studies are approved by the local Research Ethics Committee. The broad purpose of the Leicester cohorts is to study the childhood epidemiology of wheezing disorders and other common respiratory problems such as chronic cough, chronic rhinitis and habitual snoring. Four major aims of the study were: (i) to estimate the prevalence of these respiratory disorders in young children and to determine whether secular trends in asthma and wheeze described for adults and schoolchildren were also evident for preschool children; (ii) to study the contribution of potential risk factors on incidence and persistence of wheeze and other respiratory disorders taking into account different age-exposure windows; (iii) to study wheezing disorders in children of south Asian origin, the predominant British ethnic minority and (iv) to study the natural history of wheeze and chronic cough and to determine how many disease phenotypes exist within the ‘asthma’ spectrum in childhood. The results of these studies have implications for public health, giving detailed population-based information on respiratory disorders in children, on health care utilization and on treatment, and should lead to preventive interventions. For clinical purposes, research from the Leicester cohorts should provide diagnostic algorithms to predict prognosis and facilitate the management of children suffering from wheeze or chronic cough. Both Leicester cohorts are stratified random samples from the Leicestershire Health Authority Child Health Database, a constantly updated database with perinatal demographic and health-related information for all children resident in the Leicestershire Health area. This region in the East Midlands of England covers the government local authority areas of Leicester City Council, Leicestershire County Council and Rutland County Council (Figure 1) with about 924 000 inhabitants (Census 2001). Over 106 000 of these were of south Asian ethnicity, accounting for 3.7% of the population of Leicestershire county and 29.9% of Leicester city.13 The region has a mix of people from urban and rural neighbourhoods, long-established and immigrant, wealthy and poor, typical of England as a whole. The only hospital with a paediatric department in the region is located in Leicester city. Annual means of daily maximum and minimum temperatures in the English Midlands are 13 and 5.5°C, respectively and annual precipitation is about 800 mm. Geographical situation of the counties of Leicestershire and Rutland Eligibility criteria for children taking part in the cohort studies were: registered in the Leicestershire Health Authority Child Health Database, complete birth records, born and living in Leicestershire and aged between 0 and 4 years at recruitment (1–4 years for 1998 cohort). In total 46 160 children fulfilled these criteria in 1998. Of these 30 771 (66.7%) were white (British or European) and 6194 (13.4%) were south Asian by self-assigned maternal ethnicity in the birth record. The remaining children were of other or unknown ethnicity. The Leicester cohorts consist of stratified random samples of fixed sizes drawn from among all eligible children in 1990 and 1998, respectively. In total 1650 children were sampled in 1990 and 8700 in 1998. Table 1 summarizes the sample sizes for each age and ethnic group in both cohorts. In 1998 a large proportion of all preschool children in Leicestershire was recruited for the study (55.7% and 81.0%, respectively of white and south Asian 1-year olds, and 8.4% and 29.1% of white and south Asian 2–4-year olds, respectively). Amongst south Asian mothers (Table 2), 609 (23.4%) were born in the UK, 666 (25.6%) in East Africa and 1146 (44.1%) in the Indian subcontinent (902 from India, 137 from Bangladesh and 107 from Pakistan). The largest ethnic subgroup (874) were Gujarati-speakers born in India (398), Africa (289) or the UK (164). The second important subgroup were Punjabi-speaking families (207), born mainly in the UK (94) or in India (105). Summary of children recruited in the two cohorts, by age and ethnicity 1998(a): These 4400 children were primarily recruited to assess secular trend since 1990 and perform cross-sectional analyses in different age groups. Respiratory questions were identical to those used in 1990. 1998(b): Additional 4300 children aged 1 year, to study natural history in more detail. Summary of children recruited in the two cohorts, by age and ethnicity 1998(a): These 4400 children were primarily recruited to assess secular trend since 1990 and perform cross-sectional analyses in different age groups. Respiratory questions were identical to those used in 1990. 1998(b): Additional 4300 children aged 1 year, to study natural history in more detail. Characteristics of children of south Asian ethnicity recruited in 1998, by maternal country of birth, main household language and religion Characteristics of children of south Asian ethnicity recruited in 1998, by maternal country of birth, main household language and religion Baseline routine data and recurrent updates were available for all children (100%) of the cohorts from the Leicestershire Health Authority Child Health Database. A first detailed postal questionnaire was sent to parents at the time of recruitment, in 1990 and 1998, respectively. Since then, the families received similar questionnaires every few years, adapted to the age of the children. Questionnaire surveys were carried out in 1990, 1992–94, 1998 and 2003 (addressed to the young people themselves) for the 1990 cohort, and in 1998, 1999, 2001, 2003 and 2006 for the 1998 cohort (Table 3). Summary of surveys conducted with the Leicester cohorts Postal questionnaire Interviews Measurements Postal questionnaire Interviews Measurements Postal questionnaire Interviews Measurements Postal questionnaire Interviews Measurements Summary of surveys conducted with the Leicester cohorts Postal questionnaire Interviews Measurements Postal questionnaire Interviews Measurements Postal questionnaire Interviews Measurements Postal questionnaire Interviews Measurements Both cohorts underwent physiological measurements at early school age, when they were old enough to comply (Table 3). A nested sample of the 1990 cohort was studied in 1992–94 targeting all children for whom parents had reported wheeze ever (n = 222) or chronic cough (n = 226) in the first survey, and a nested sample of asymptomatic children (n = 347). The lab measurements for the 1998 cohort are ongoing. All children whose parents had completed the first survey in 1998 and at least one follow-up survey (n = 5180) are being invited for lab measurements, beginning with the oldest children. Currently about 500 children have finished their tests. Response rates to questionnaire surveys in 1990 cohort were good, with 1422/1650 (86.2%) responding to the initial survey, 79.1% replying in 1998 and 57.3% in 2003. Figure 2a shows that a significant proportion of those who had not replied in the first survey answered one of the follow-up questionnaires, so that in total, 94.4% had replied at least once. Participation in laboratory measurements in 1992–94 was 65.3%, 57.5% and 61.4% for children with wheeze, chronic cough and no symptoms, respectively. Patterns of response to the first three surveys in (a) the 1990 cohort and (b) the 1998 cohort. The black bar in the lowest layer represents response to the initial survey as a percentage of the total cohort. The two black bars of the next layer show response to this follow-up as a percentage of respondents and non-respondents to the initial survey, respectively. In the top layer, response to the next follow-up is given for each of the four patterns of response to the first two surveys. Figures in the black bars represent numbers of respondents. Overall response rates (RR) are also reported for each survey Follow-up rates in the 1998 cohort were lower, with 78.3%, 59.6% and 48.6% responding in 1998, 2001 and 2003, respectively, and with 86.2% replying at least once (Figure 2b). The 2006 survey is still ongoing. The main reason for attrition was the mobility of these families. Within 4 years (April 1997 to April 2001), the Leicestershire Health Authority Child Health Database had been notified of address changes in 1933/4400 [cohort 1998(a)] (44%) families, with 1265 moving once, 438 moving twice and 230 moving more than twice in this period.14 With increasing age, notification of address changes in the database became increasingly incomplete. Subjects participating in more than one survey tended to be of higher social class and of white ethnicity, but had a lower frequency and severity of symptoms than children participating only once. The entire 1990 and 1998 cohorts of 1650 and 8700 children respectively, are representative, for the population of preschool children in Leicestershire and Rutland, being randomly sampled. Baseline routine data are available for all of them. However, there were some systematic differences between respondents and non-respondents to specific surveys. For instance, respondents to the first questionnaire in 1998 had a higher socioeconomic status than non-respondents (in whites 44.5% of respondents and 24.8% of non-respondents were in the least deprived tertile of the Townsend score; in south Asians these proportions were 17.5% and 12.2%, respectively) and a higher maternal age at birth (mean age in whites: 28.9 vs 25.9 years; south Asians: 28.1 vs 27.6 years). Information was available for both cohorts from routine databases, questionnaires and physiological examinations (Table 4). Summary of what has been measured Birth notification and perinatal details: gender; date of birth; number of older siblings; gestational age; birth weight Demographic and socio-economic information: mother's occupation, ethnicity, religion, country of birth and language; father's occupation at time of child's birth Health visitor records: complete information on growth (height, weight, head circumference at 1, 8, 18, 38 months); complete records of childhood immunizations (type of vaccine, dose, date); routine development tests Townsend score (1991 census data) Proximity to roads and modelled exposure to air pollutants based on postcode (Indic–Airviro dispersion model Version 2.2, Swedish Meteorological and Hydrological Institute, Norrköping, Sweden) All home addresses since the child was born Birth notification and perinatal details: gender; date of birth; number of older siblings; gestational age; birth weight Demographic and socio-economic information: mother's occupation, ethnicity, religion, country of birth and language; father's occupation at time of child's birth Health visitor records: complete information on growth (height, weight, head circumference at 1, 8, 18, 38 months); complete records of childhood immunizations (type of vaccine, dose, date); routine development tests Townsend score (1991 census data) Proximity to roads and modelled exposure to air pollutants based on postcode (Indic–Airviro dispersion model Version 2.2, Swedish Meteorological and Hydrological Institute, Norrköping, Sweden) All home addresses since the child was born aMeasurements of 1990 cohort were described elsewhere 10,11. bUsing a 20% decrease in transcutaneous oxygen tension (tc-PO2) to assess response. cUsing a 20% decrease in FEV1 to assess response. Summary of what has been measured Birth notification and perinatal details: gender; date of birth; number of older siblings; gestational age; birth weight Demographic and socio-economic information: mother's occupation, ethnicity, religion, country of birth and language; father's occupation at time of child's birth Health visitor records: complete information on growth (height, weight, head circumference at 1, 8, 18, 38 months); complete records of childhood immunizations (type of vaccine, dose, date); routine development tests Townsend score (1991 census data) Proximity to roads and modelled exposure to air pollutants based on postcode (Indic–Airviro dispersion model Version 2.2, Swedish Meteorological and Hydrological Institute, Norrköping, Sweden) All home addresses since the child was born Birth notification and perinatal details: gender; date of birth; number of older siblings; gestational age; birth weight Demographic and socio-economic information: mother's occupation, ethnicity, religion, country of birth and language; father's occupation at time of child's birth Health visitor records: complete information on growth (height, weight, head circumference at 1, 8, 18, 38 months); complete records of childhood immunizations (type of vaccine, dose, date); routine development tests Townsend score (1991 census data) Proximity to roads and modelled exposure to air pollutants based on postcode (Indic–Airviro dispersion model Version 2.2, Swedish Meteorological and Hydrological Institute, Norrköping, Sweden) All home addresses since the child was born aMeasurements of 1990 cohort were described elsewhere 10,11. bUsing a 20% decrease in transcutaneous oxygen tension (tc-PO2) to assess response. cUsing a 20% decrease in FEV1 to assess response. Routine data were available from the Leicestershire Health Authority Child Health Database for all children, whether or not they had responded to the questionnaires. This included demographic and ethnic information from maternity records and birth notification, perinatal data and data on growth, immunizations and other health-related events during childhood. The complete postcodes (7-digits) for all addresses at which the child had ever lived allowed area-based deprivation measures (Townsend score) and exposure to air pollutants and road traffic to be calculated.14 The baseline questionnaire for the 1990 survey was developed de novo, including questions derived from the American Thoracic Society (ATS) childhood questionnaire15 and other validated questionnaires adapted for preschool children.16,17 Exactly the same questions were applied in 1998 to 2600 white and 1800 south Asian children [cohort 1998(a)] to assess secular trends. The questionnaire for the 4300 additional 1-year olds recruited in 1998 [cohort 1998(b)] was slightly modified. Some questions were adapted or added to allow a better phenotype description of children with wheeze and cough. We also included those questions from the International Study of Asthma and Allergies in Childhood (ISAAC) which could be applied to infants and young children.18 The initial questionnaires for cohorts 1998(a) and (b) were used with little variation in all subsequent surveys. They define wheezing as ‘breathing that makes a high-pitched whistling or squeaking sound from the chest, not the throat.’ Additional questions on symptoms and exposures were introduced at follow-up surveys according to the age of the children. Repeatability of both the 1990 and the 1998 questionnaires was good.7,19,20 A major aim of the Leicester cohorts was to provide data on prevalence, risk factors and secular trends of wheeze in preschool children. In 1990, 11% of 1–5-year olds had doctor-diagnosed asthma, 16% had wheezed and 13% reported attacks in the past 12 months.7,8 These were the first population-based prevalence data for preschool wheeze in the UK and Europe. When we repeated the survey 8 years later (1998) the prevalence of wheeze and asthma had doubled.21 The fact that all wheezing disorders increased (including viral wheeze, a non-atopic condition occurring only with respiratory infections) suggested that factors unrelated to atopy were probably in the epidemiology of that in 1–5-year olds with wheeze the health a total of of total NHS This is a large for one single in a and the to at the population to a air was associated with increased prevalence of chronic cough and with increased incidence of wheeze and cough in young This was of many and incidence data on wheeze with records from the a conducted that the risk of childhood asthma, especially of the asthma phenotype and suggested that be is to have the The secular trends for preschool wheeze were not by main aim of the 1998 cohort was to asthma in children of south Asian ethnicity. Asthma is common in children living in India with the from south Asian children living in the UK are and a lower prevalence of wheeze but a higher with We in mothers of the 1998 cohort, that south Asian who to the UK at the age of years or older had a lower risk of asthma than those born in the UK or who before age This the that early life factors influence the risk of adult In and prevalence of wheeze was lower in south Asian with in south Asians were at risk than whites of wheeze asthma for and with this, wheeze, a significant of disease and was more common in south with the of south Asian children with these the in The ethnic studies are being by physiological characteristics and in south Asian and white in lab The Leicester studies reported on management of wheeze in preschool children in the of and A large proportion of preschool children used asthma but that treatment was to asthma phenotype and with of wheeze, especially in and south Asian children, but of wheeze and chronic cough. This the for development and of treatment The main aim of the Leicester cohorts to natural history and phenotypes in childhood Follow-up of the 1990 cohort that than of the children who wheezed during preschool years to so at early school age, but that those who did many with asthma in some children with chronic cough many typical of asthma, few to wheeze 2–4 years studies the natural history of preschool wheeze and chronic cough with are We are a for disease phenotypes These and and data and are an for disease phenotypes data from epidemiological or clinical cohorts. this in a study on the 1990 cohort, we for the of three phenotypes of wheeze, and two phenotypes of chronic cough, and The be and on the 1998 cohort. of were higher in more childhood by treatment a of or a in asthma The short-term of preschool questionnaire was and of preschool symptoms at was In we that of the in questionnaire surveys of prevalence of wheeze, to an in children from south Asian and deprived Questionnaire surveys should both a of wheeze and measures of of the of air an exposure which had received in the of respiratory according to the respiratory health of the children. We that could some or all the between reported exposure to road traffic and exposure measures should be used in surveys. The Leicester respiratory cohorts are the largest community-based cohort studies on childhood respiratory disease in the In to other UK cohorts which aim to on many health the Leicester cohorts include a range of questions to and lower respiratory symptoms at each survey, to different phenotypes within the syndrome of or asthma with and The 1998 cohort an data for studying ethnic differences in wheezing and is in fact the largest population-based database on respiratory disorders in British south Asian children. We care to the most of the main by an of and including questions on of the in the We also to other of the questionnaires and the influence of different on We information on many potential to study A wide range of data from the Leicestershire Health Authority Child Health Database data from questionnaires. the study two separate cohorts from the same population that with to years of birth and age-spread of their This to study cohort and the to perform both cross-sectional and analyses in different age groups. Within age and ethnicity the cohorts are they were sampled randomly from the entire local Routine data is available for the entire cohort and be used for to account for systematic differences between respondents and non-respondents to postal questionnaires. include the attrition in later surveys especially of the 1998 cohort of in Respiratory were mainly by a of the families received the first questionnaire when the child was aged 1 year, a proportion of children in both cohorts were aged 2–4 at the time of the first These children are being in analyses to natural history of respiratory to we have measurements only on one at early school Some exposures as or were not but only by data available on growth, immunizations and exposure to air pollution. For data to be with those from other cohort to The Leicester cohort studies are conducted at the of Child Health at Leicester in with the Leicester City Trust, Child Health for Leicester, Leicestershire and Rutland, and the of and at the of Berne, The Leicester detailed information on and in are invited to the principal or Claudia We all families in Leicestershire for participating in the are and were by from the Swiss National Science Foundation no and
DOI: 10.1093/ije/dym090
Authors: C. E Kuehni, A. M Brooke, M.-P. F Strippoli, B. D Spycher, A. Davis, M. Silverman
Institutions: University of Bern, University of Leicester
Language: English
Subfield: Physiology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2096580836