Global Asthma Network Logo Global Asthma Network Global Asthma Network Logo
  • The Global Asthma Network strives for a world where no-one suffers from asthma

  • The Global Asthma Network is the asthma surveillance hub for the world

  • The Global Asthma Network researches ways of reducing the burden of asthma

  • The Global Asthma Network promotes access to appropriate asthma management

  • The Global Asthma Network stimulates and encourages capacity building in LMICs

  • The Global Asthma Network strives to ensure access to quality-assured essential asthma medications

  • The Global Asthma Network raises the profile of asthma as a major NCD

10. Validation of instruments

Each question in each questionnaire is justified, its source specified, validation described where available and analyses outlined.

10.1 Adolescent age group questions (self completed)

10.1.1  Height and weight measurements of adolescent

Overweight hypothesis
ISAAC Phase Three hypothesized that if a child’s weight was excessive in comparison with their height, this could be associated with an increased risk of symptoms of asthma, rhinoconjunctivitis and eczema53. For ISAAC Phase Three no sample questions were identified from the literature. There were some papers that commented on the inaccuracies of self-reporting, however another paper found insignificant differences between self-reported and measured height and weight in bank employees54.

  • Source of ISAAC questions: ISAAC Steering Committee

Mitchell et al43 in ISAAC Phase Three reported associations between obesity and symptoms of asthma and eczema and clear evidence of dose-effect relationships with the magnitudes of the risks of symptoms of asthma and eczema greater with obesity than with overweight.
Analysis of variables. Body Mass Index (BMI), weight ÷ height2, is calculated for each individual and the BMI used as the variable for analyses. Weight data is converted to kilograms and height data converted to Metres. Although the majority of centres in ISAAC Phase Three self-reported height and weight (72% by parents of the children and 76% by the adolescents themselves), no major differences in the associations were seen in the analyses with measured and self-reported values43. However recent evidence exists that half of parents underestimate their children’s overweight/obese status and a significant minority underestimate children’s normal weight55. Therefore in the Global Asthma Network height and weight will be measured by fieldworkers and/or school staff using a standardised protocol56 (see section 7 and section 20.4).

10.1.2  Asthma questions

These questions are built on those used in ISAAC. The ISAAC Phases One, Two and Three core questions (questions 1-6 below) were piloted before Phase One57 and published49, 58, 59 . Most were based on questions used in previous respiratory epidemiological studies (prior to ISAAC) and include both sensitive and specific indicators of asthma60. Many of these questions have been used in ISAAC in 306 centres in 105 countries. They were translated from English into 52 languages in ways which were understood locally50. They have been used in key ISAAC asthma symptoms publications; Phase One61, 62 , Phase Two33 and Phase Three12, 16, 18 .
The core questions have been validated against bronchial hyper-responsiveness3, 4, 22, 23, 63-70 .
Due to experience in the presentation and analyses of the ISAAC asthma data12, 16, 18, 33, 61, 62 and due to the identification of new information from the ISAAC data further asthma questions have been added to gain information on asthma.

Question 1.
This is based on the IUATLD questionnaire (pre 1993)60. ‘Attacks’ of wheezing are not mentioned in order to identify adolescents with persistent symptoms which are not obviously characterised as episodes or attacks. This is seen as a very sensitive question, was successfully used in all ISAAC Phases and it will be used unchanged for the Global Asthma Network.

Question 2.
Limitation to a 12 month period reduces errors of recall71 and is believed to be independent of month of completion72. This is considered to be the most useful question for assessing the prevalence of wheezing illness. This question was used successfully in ISAAC and it will be used unchanged for the Global Asthma Network.

Questions 3, 4 & 5.
These questions used in ISAAC offer three alternative quantitative measures of the frequency and severity of wheezing. Problems with the concept of ‘attacks’ and difficulty in quantifying the frequency of recurrent asthma, led to the inclusion of question 5 to identify and quantify persistent wheeze. Question 5 was created by ISAAC to identify acute severe asthma which had previously little study. These three questions were used to identify asthma severity in ISAAC, in a comparison of asthma symptom prevalence, mortality and hospital admissions which demonstrated correlations with all three parameters73. In that paper, the 12-month prevalence of moderate to severe wheezing comprised one or more of: (i) four or more attacks of wheeze; (ii) woken by wheeze on one or more nights per week or; (iii) wheezing severe enough to limit speech to only one or two words at a time, between breaths. These questions were used to define severe wheezing in ISAAC Phase Three12 and will be used unchanged for the Global Asthma Network.

Question 6.
This is the first time in the questionnaire that ‘asthma’ is mentioned. It is deliberately asked after the questions on asthma symptoms. The asthma label is affected by many factors such as awareness of asthma, medical training and experience, cultural and societal factors74. Occasionally asthma may be suggested in the absence of wheeze (on the basis of recurrent nocturnal cough etc). This question was used in all phases of ISAAC. It has not been clear whether the answer represents the opinion of the adolescent, or was a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (question 7).

Question 7.
As mentioned above in Question 6, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had asthma.

Question 8.
This question, adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 3)75, has been added to the Global Asthma Network questionnaire. Written asthma management plans are part of most guidelines. Thus this question will provide new information on these plans and will be analysed with the prevalence and severity questions.
Questions 9 & 9a.
Questions 9 and 9a on asthma management are essential to enable the Global Asthma Network to assess its target to decrease the proportion of symptomatic people with asthma not on inhaled corticosteroids. These questions are adapted from ISAAC Phase Two questions (Module 2.2 Asthma management. Questions 1 and 2)75. The wording ‘(when you didn’t have a cold)’ has been added to the end of the question to reinforce that question 9 is asking about breathing problems without the complication of cold symptoms.

Question 10 & 10a.
Questions 10 and 10a ask about swallowed medicines as opposed to inhaled medicines in questions 9 and 9a and are adapted from ISAAC Phase Two questions (Module 2.2 Asthma Management, questions 1 and 2)75. The wording ‘(when you didn’t have a cold)’ has been added to the end of the question to reinforce that question 10 is asking about breathing problems without the complication of cold symptoms.

Question 11 & 12.
These questions about urgent visits for asthma in the past 12 months are essential to enable the Global Asthma Network to assess its target to decrease unplanned visits for asthma. The questions are adapted from ISAAC Phase Two (Module 2.2 Asthma management. Question 5)75. Different patterns of medical care may contribute to variations in the severity of asthma between countries or over time. These questions along with questions 9 and 10 will allow the relationship (cross-sectional) to be explored between treatment and morbidity.

Question 13.
This question, about hospital admissions for asthma in the past 12 months, is essential to enable the Global Asthma Network to assess its target to decrease hospital admissions for asthma. This question is adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 6)75.

Question 14.
This question is adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 9)75. It has been added to the Global Asthma Network questionnaire to provide information on school absenteeism, in the past 12 months, which is an additional indicator of asthma morbidity. It will enable the Global Asthma Network to assess its target to decrease time off school due to asthma.

Question 15.
When piloting questions for ISAAC Phase One57, this question was a stem question under question 2, however it was found in some Australasian surveys to identify some adolescents who deny wheezing or whistling in questions 1 or 2, yet report wheeze with exercise, and so was moved to question 7 for the full ISAAC study. It was successful in ISAAC placed as question 7 and it will now follow the new questions on asthma diagnosed asthma and management for the Global Asthma Network questionnaire.

Question 16.
Nocturnal cough is widely accepted as an alternative presentation of asthma, and this question was included in ISAAC to increase the overall sensitivity of the questionnaire, although its specificity in population surveys remained unclear. In ISAAC Phase One dry night cough in the past 12 months was reported more frequently than 12 month wheeze in all countries except for Australia, New Zealand and Sweden62. Phase Three saw a global increase of night cough of 0.51% per year18. This question has been retained for the Global Asthma Network so that trends over time can be monitored.

10.1.3  Rhinitis questions

ISAAC studied not only asthma but also the related diseases of rhinitis10, 13, 19, 30 and eczema11, 14, 20, 24, 76 . While the Global Asthma Network is focussed on asthma, the surveillance undertaken by the Global Asthma Network gives an opportunity to continue surveillance of rhinitis and eczema too. This would enable the monitoring of time trends in these two related diseases16, 19, 20, 49, 58, 59 where to date there have been only two time points in worldwide studies19, 20 . In addition, the responses to the questions will enable the Global Asthma Network to continue to explore the relationship of asthma with rhinitis and eczema.

The ISAAC Phases One, Two and Three rhinitis core questions (questions 17-22 below) were based on questions used in previous respiratory epidemiological studies and include both sensitive and specific indicators for respiratory symptoms77, 78 . Due to the experience in the presentation and analyses of the ISAAC rhinitis data10, 13, 16, 19, 61 and due to the identification of new information from the ISAAC data one further question (question 23) on doctor diagnosis has been added to enhance the Global Asthma Network rhinitis questions.

Questions 17 & 18.
These questions (prior to ISAAC) were found to have a positive predictive value of 80% in detecting rhinitis in a community sample of adults (aged 16-65 years) in south west London78. These questions were used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 19.
This question (new for the Global Asthma Network) was piloted successfully in the EISL study79 to determine the presence of allergic rhinitis in children. This question was not used in ISAAC worldwide surveys.

Question 20.
This question had the highest positive predictive value (78%) in detecting atopy among participants with rhinitis78. In ISAAC question 20 was used in conjunction with question 18 to determine the prevalence of rhinoconjunctivitis by calculating the number of children responding positively to both questions divided by the total number of completed questionnaires10. Rhinoconjunctivitis was justified to be considered in ISAAC to be the variable to be analysed in detail because this symptom combination most closely related to objective indicators of allergic sensitisation in European populations78, 80, 81 . This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 21.
This question was considered a crude qualitative measure of severity when considering questions to use in ISAAC Phase One52 but was considered to correlate well with other indicators of morbidity from rhinitis including reported symptom severity, interference with specific activities of daily living and medical service use. This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 22.
This question has been used to investigate the labelling of rhinitis in relation to the prevalence of rhinitis symptoms. The label ‘Hay fever’ had a positive predictive value of 71% in detecting atopy among participants with rhinitis78. This question was used in all ISAAC phases. It has not been clear whether the answer represents the parent’s own opinion, or was a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (question 23)

Question 23.
As mentioned in Question 22, this is a new question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had rhinitis.

10.1.4  Eczema questions

ISAAC studied not only asthma but also the related diseases rhinitis10, 13, 19, 30 and eczema11, 14, 20, 24, 76 . While the Global Asthma Network is focussed on asthma, the surveillance undertaken by the Global Asthma Network also gives an opportunity to continue surveillance of rhinitis and eczema. This would enable the monitoring of time trends in these two related diseases16, 19, 20, 49, 58, 59 whereas to date there have been only two time points in worldwide studies19, 20 . In addition the responses to the questions will enable the Global Asthma Network to continue to explore the relationship of asthma with rhinitis and eczema.

The ISAAC Phases One, Two and Three eczema core questions (questions 24-29 below) were based on questions used in previous eczema studies82-84 prior to ISAAC and included both sensitive and specific indicators for eczema symptoms. Due to experience in the presentation and analyses of the ISAAC eczema data11, 14, 16, 20, 61 and due to the identification of new information from the ISAAC data one further question (question 30) on diagnosis has been added to enhance the Global Asthma Network eczema questions.

Question 24.
Prior to ISAAC, this screening question was evaluated in a UK pilot study of factors which discriminated ‘typical’ mild-moderate atopic dermatitis from non-atopic eczema and other inflammatory dermatoses presenting for the first time in British hospital outpatient clinics85. This screening question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 25.
Following the form of the ISAAC core questionnaires for wheezing and rhinitis, further enquiry focused on those children with recent rashes (in the past 12 months) to minimise problems of incomplete and selective recall. This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.
Question 26.
This question used in ISAAC worldwide surveys, was found to have high sensitivity (94%) and specificity (96%). This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Questions 27.
Question 27 was included in all ISAAC phases in the adolescent questionnaire as a measure of eczema morbidity83. In the Global Asthma Network, this question will be used unchanged.

Question 28.
Respondents with sleep loss of one or more nights per week were considered to have severe eczema based on previous studies86, 87 . This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 29.
This question permitted investigation of the labelling of eczema in relation to the prevalence of eczema symptoms. This question was used in all ISAAC phases. It has not been clear whether the answer represents the participant’s own opinion, or was a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (question 30)

Question 30.
As mentioned in question 29, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had eczema.

Analyses: asthma, rhinitis and eczema
The analysis of the asthma, rhinoconjunctivitis and eczema core questions will enable the Global Asthma Network to obtain internationally comparable estimates of the direction and magnitude of change in prevalence of symptoms. Comparisons of prevalence rates between different centres will be made using appropriate statistical methods. Crude rates can be compared by using contingency tables or logistic regression. Comparison of standardised rates or data that needs controlling for confounding will involve multivariate logistic regression.

In addition, calculations will be made of the proportion of participants that have asthma who are not on inhaled corticosteroids, time off work/school because of asthma, unplanned visits for asthma, and hospital admissions. The direction and magnitude of symptoms and severity of asthma will be derived for each centre.

10.1.5  Environmental questions

The ISAAC Phase Three EQ (questions 31,32,35-43 below) were either sourced from existing questions used in previous epidemiological studies (prior to ISAAC), or where these were not available, developed by the ISAAC Steering Committee. Due to experience in the presentation and analyses of the ISAAC EQ and due to the identification of new information from the ISAAC data, further questions have been added to enhance the Global Asthma Network EQ (questions 33, 34, 37a, and 44-48).
Inclusion of standardised questions on past and present living and exposure conditions will permit:

  • Between-centre correlations of disease prevalence and risk factor distribution.
  • A pooled evaluation of within-centre analyses of the association between disease and risk factors at the individual level.

Analyses:
For each EQ, prevalence odds ratios will be calculated using generalised linear mixed models with a binomial distribution and a logit link, with centres being modelled as a random effect. Analyses will be adjusted for gender, region of the world, language and per capita gross national income. Regression models will allow for the sampling of schools by scaling the size of the sample by the design effect. Further multiple regression analyses will be conducted to investigate whether the association between symptoms and a particular EQ is confounded by other risk factors for the information collected from an EQ and which had shown an association with wheezing, rhinoconjunctivitis or eczema in the univariate analyses.

Questions 31, 32, 33.
Exercise
ISAAC Phase Three hypothesized that regular exercise and physical fitness are protective against asthma. Possible aggravating factors: Being sedentary and lack of physical fitness. An assessment of physical activity in adolescents was undertaken by Aaron et al88 but not undertaken in relationship to asthma and allergies.

  • Source of questions 31 & 32 for ISAAC: Aaron et al88 and Kohl et al89

Mitchell et al in ISAAC Phase Three43 found a positive association between physical activity and symptoms of asthma, rhinoconjunctivitis and eczema in adolescents but not in children43, 90 and an increased risk of asthma symptoms with television viewing for five or more hours/day in children and adolescents. Question 32 has added ‘of 7 days’ following ‘a week’ and ‘(include DVD’s, films)’ as well as ‘videos’ following comments made by participants during the ISAAC fieldwork.

  • Source of question 33, the Global Asthma Network Steering Group. Question 33 is a new question that takes into account time spent on the computer (PlayStation, smartphone, tablet) and/or internet (Chat, Facebook, Games, Twitter YouTube) and more

Analysis of variables. Combinations of these questions will allow classification of participants into groups based on their level of exercise and whether they are sedentary.

Question 34.
Twin sibship
A new question on whether the participant is a twin has been added following the Global Asthma Network surveillance pilot study 2015. This will avoid confusion when twins complete questions 35 and 36 on older/younger siblings.

  • Source of question 34: the Global Asthma Network Steering Group

Questions 35 & 36.
Parity
ISAAC hypothesized that increased household size was associated with a decreased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema91. A decrease in the cumulative incidence and prevalence of allergic rhinitis and the cumulative incidence of asthma with increasing number of siblings was reported by Hesselmar et al91.
Related factors could be small family size and early birth order.

  • Source of ISAAC questions 35 & 36: ISAAC Steering Committee

Strachan et al in ISAAC Phase Three48 found new observations of greater symptom severity among children from larger families which seems to be a more consistent phenomenon worldwide, and deserves further exploration and explanation.
These questions will remain unchanged for the Global Asthma Network, however the instruction ‘(Please put 0 if there are no older [younger] siblings)’ will be added to these questions following comments made by participants when undertaking the ISAAC fieldwork.
Analysis of variables. Associations with symptoms will be identified. The questions allow identification of separate effects of birth order and number of siblings and for the Global Asthma Network will build on the information from ISAAC.

Questions 37, 37a & 38.
Migration
ISAAC hypothesized that migrants to a new country will adopt the prevalence of symptoms of asthma, allergic rhinoconjuncitivitis and atopic eczema of their new country over time92-97.
There was conflicting evidence in the literature regarding the health of immigrants. Robertson et al92 reported that symptoms of asthma, rhinoconjunctivitis and eczema were more common in children born in Australia compared with children born in other countries but resident in Australia. Leung et al95 reported that prevalence of hay fever and asthma increased with length of stay among Asian immigrants to Australia.

  • Source of ISAAC questions 37 & 38: ISAAC Steering Committee

García-Marcos et al in ISAAC Phase Three47 found that recent migration to high prevalence/affluent countries is associated with a lower prevalence of allergic diseases. The pre-migration environment might be protective, but this effect quickly decreases with increasing time in the host country.
These questions will be used unchanged in the Global Asthma Network. For the country of birth question, if the respondent answered ‘NO’ to question 37, the Global Asthma Network will include a question about what country the respondent was born in (question 37a).

  • Source of question 37a: The Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be identified. The questions allow assessment of the effect of time spent in the country and country of birth.
Question 39.
Traffic
ISAAC hypothesized that respiratory irritants such as sulphur dioxide (SO2) nitrogen oxides (NOx) and particulates from diesel combustion cause local respiratory inflammation, increasing tissue contact with inhaled allergens and the likelihood of an allergic response98, 99 .

  • Source of ISAAC question: Weiland et al98

Brunekreef et al in ISAAC Phase Three37 found that higher exposure to self-reported truck traffic on the street of residence is associated with increased reports of symptoms of asthma, rhinitis and eczema in many locations in the world.
This question will remain unchanged for the Global Asthma Network, however ‘(not often)’ will be added next to ‘seldom’ as an option as the Global Asthma Network pilot study found some participants did not understand the word ‘Seldom’.
Analysis of variables. Associations with symptoms will be further explored in the Global Asthma Network.

Question 40.
Diet
ISAAC Phase Three hypothesized that a plant-based diet protected against asthma and allergies and a ‘Western’ diet was positively associated with asthma and allergies.
Protective and aggravating factors found in the ISAAC Diet ecological analysis100 included starch, cereals, rice, vegetables, fish, other seafood, fibre, fruit, nuts, olive oil (protective); trans fatty acids, fast foods (aggravating). Other foods considered included, eggs, animal fats, milk, polyunsaturated fatty acids. Dietary surveys in the literature were lengthy questionnaires and either interviewer administered or adult self-completed, food frequency questionnaires either prospective using a diary or retrospective by recall (7 day, 3 months, 12 months) or ‘dietary history’ which is a recalled food frequency questionnaire. No short diet questionnaires suitable for inclusion in the ISAAC EQ were identified.

  • Source of question: ISAAC Steering Committee

Ellwood et al, in ISAAC Phase Three44 found a negative association between the intake of fast food and asthma, rhinitis and eczema symptoms and a positive association between these symptoms and the intake of fresh fruit and vegetables. These associations were also found by Nagel et al29 and Wickens et al101 in ISAAC Phase Two. The ISAAC Phase Three EQ questions will be used in the Global Asthma Network with some food items being separated (such as cooked and raw vegetables) and additional food items such as fizzy or soft drinks. Since we found the association with fast food we have been questioned about its precise nature, so we are adding a food type to separate out the types of fast food. The fast food outlets that are common in most countries are McDonald’s and Burger King which enables burgers to be studied as a separate group. Food types that have been separated out are: Cooked vegetables; Raw vegetables; Cereals (excluding bread); Bread; Olive oil; Other dairy (include cheese and yoghurt); Sugar (includes lollies, candies, sweets); Fast food/ burgers; Fast food excluding burgers. ‘Fizzy or soft drinks (include local terminology)’ has been added due to experience in ISAAC.
Analysis of variables. Components will be analysed individually, some dietary items may be combined, such as a plant based diet, or a Mediterranean diet. Details of these combinations are yet to be developed.

Question 41.
Paracetamol
It has been hypothesized that frequent paracetamol use is associated with an increased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema102.
Newson et al102reported, in arecent ecological analysis, a positive association between paracetamol consumption and the prevalence of asthma, allergic rhinoconjunctivitis and atopic eczema symptoms in children. They speculate that paracetamol may influence atopic disease by depleting glutathione in the airways and in immune cells.

  • Source of question: ISAAC Steering Committee

Beasley et al35 in ISAAC Phase Three found that the use of paracetamol may represent an important risk factor for the development and/or maintenance of asthma, rhinoconjunctivitis and eczema in adolescent children. This question will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified.

Questions 42 & 43.
Allergens
ISAAC hypothesized that exposure to allergens is associated with increased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema.
Roost et al103concluded that current cat ownership represented a significant risk for sensitization to cat if cats were allowed indoors. They do, however suggest that childhood exposure to pets, including cats, might modulate immunologic mechanisms and reduce sensitization to cat in adulthood.

  • Source of ISAAC questions: Roost et al103

Brunekreef et al, in ISAAC Phase Three40 found that early life exposure to cats is a risk factor for symptoms of asthma, rhinoconjunctivitis and eczema in 6 and 7 year olds especially in less affluent countries.
These questions will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified. The questions allow identification of separate trigger and sensitisation effects and with data from ISAAC Phase Three.

Questions 44 – 47.
Tobacco smoke
It is hypothesized that exposure to tobacco smoke in early life is associated with increased risk of symptoms of asthma, rhinoconjunctivitis and eczema. That exposure to tobacco smoke is a trigger for asthma attacks for asthmatics104. In the ISAAC Phase Three EQ, the questions from Jarvis104 (1999) were used. However in the Global Asthma Network, updated questions (2012) from a WHO survey, the ‘Global Adult Tobacco survey’, will be used. The option ‘don’t know’ has been removed for the Global Asthma Network

  • Source of questions: The Global Adult Tobacco survey 2011105

Analysis of variables. Associations with symptoms will be identified.

Question 48.
Water pipe smoking
ISAAC Phase Three in Syria found a stronger association between Mother smoking water pipe (narghile) than with cigarette smoking106. They concluded that international studies investigating environmental tobacco smoke should include questions on narghile smoking. The Global Asthma Network adolescent questionnaire and the Global Asthma Network adult questionnaire has used an adapted question from the ISAAC Phase Three Syrian study106. Narghile is also known as bong, crack pipe, hookah, hubble bubble, shisha, vapourizer, water vapour and water pipe. This terminology has been added to question 48.

  • Source of question: ISAAC Phase Three study in Syria106

Analysis of variables. Associations with symptoms will be identified.

10.1.6  VIDEO questionnaire

Questions 49-53.
In response to possible translation problems with written questionnaires, for ISAAC Phase One a video questionnaire was developed and validated in Wellington, New Zealand1, 2, 6 .

In particular the video questionnaire was developed to avoid problems of translation and comprehension of terms such as ‘wheeze’ or ‘whistling’ and their use in culturally heterogeneous populations.

In Phase One of ISAAC, it was the first occasion in which the video questionnaire had been used in an international comparison. The similarities and differences found between countries were generally consistent with previously published work, and the video and written questionnaires showed a similar pattern of results1-6, 107 .

The video is non-verbal and shows 5 different scenes of breathing. The students are then asked if their breathing has ever been like that of the person in the video; if YES, they are asked whether this has occurred in the past year; and if YES, they are asked whether this occurs more often than once a week. This video takes 6 minutes to play. The video has the advantage of obtaining data from a large number of students quickly and efficiently.

The international version of the video questionnaire will remain unchanged for the Global Asthma Network; it will remain a highly recommended questionnaire for centres to use.

10.2  Children’s questions (parent completed)

10.2.1  Height and weight measurements of child

Overweight hypothesis
ISAAC Phase Three hypothesized that if a child’s weight was excessive in comparison with their height, this could be associated with an increased risk of symptoms of asthma, rhinoconjunctivitis and eczema53. For ISAAC Phase Three, no sample questions were identified from the literature. There were some papers that commented on the inaccuracies of self-reporting, however another paper found insignificant differences between self-reported and measured height and weight in bank employees54.

  • Source of ISAAC questions: ISAAC Steering Committee

Mitchell et al43 in ISAAC Phase Three reported an association between obesity and symptoms of asthma and eczema and clear evidence of a dose-effect relationship with the magnitude of the risk of symptoms of asthma and eczema greater with obesity than with overweight.
Analysis of variables. Body Mass Index (BMI), weight ÷ height2 is calculated for each individual and the BMI used as the variable for analyses. Weight data is converted to Kg and height data converted to metres. Although the majority of centres in ISAAC Phase Three self-reported height and weight (72% by parents of the children and 76% by the adolescents themselves), no major differences in the associations were seen in the analyses with measured and self-reported values43. However recent evidence exists that half of parents underestimate their children’s overweight/obese status and a significant minority underestimate children’s normal weight55. Therefore in the Global Asthma Network height and weight will be measured by fieldworkers and/or school staff using a standardised protocol56 (see section 8 and section 20.4).

10.2.2  Asthma questions

These questions are built on those used in the International Study of Asthma and Allergies in Childhood (ISAAC). The ISAAC Phases One, Two and Three core questions (questions 1,3-7,16,17 below) were piloted before Phase One57 and published49, 58, 59, 108 . Most were based on questions used in previous respiratory epidemiological studies (prior to ISAAC) and include both sensitive and specific indicators of asthma60. Many of these questions have been used in ISAAC in 306 centres in 105 countries. They were translated from English into 44 languages in ways which were understood locally50. They have been used in key ISAAC asthma symptoms publications; Phase One61, 62 , Phase Two33 and Phase Three12, 16, 18 .
The core questions have been validated with bronchial hyper-responsiveness3, 4, 22, 23, 63-70 .
Due to experience in the presentation and analyses of the ISAAC asthma data12, 16, 18, 33, 61, 62 and due to the identification of new information from the ISAAC data further asthma questions have been added to gain information on asthma.

Question 1.
This question is based on the IUATLD questionnaire (pre 1993)60. ‘Attacks’ of wheezing are not mentioned in order to identify children with persistent symptoms which are not obviously characterised as episodes or attacks. This is seen as a very sensitive question, was successfully used in all ISAAC Phases and it will be used unchanged for the Global Asthma Network.
Question 2.
This question was not included in the ISAAC worldwide surveys. This question determines the age of onset of wheeze which is pertinent to defining phenotypes109.

Question 3.
Limitation to a 12 month period reduces errors of recall71 and is independent of month of completion72. This is considered to be the most useful question for assessing the prevalence of wheezing illness. This question was used successfully in ISAAC and it will be used unchanged for the Global Asthma Network.

Questions 4, 5 and 6.
These questions used in ISAAC offer three alternative quantitative measures of the frequency and severity of wheezing in the past 12 months. Problems with the concept of ‘attacks’ and difficulty in quantifying the frequency of recurrent asthma, led to the inclusion of question 5 to identify and quantify persistent wheeze. Question 6 was created by ISAAC to identify acute severe asthma which had previously little study. These three questions were used to identify asthma severity in ISAAC, in a comparison of asthma symptom prevalence, mortality and hospital admissions which demonstrated correlations with all three parameters73. In that paper, the 12-month prevalence of moderate to severe wheezing comprised one or more of: (i) four or more attacks of wheeze; (ii) woken by wheeze on one or more nights per week or; (iii) wheezing severe enough to limit speech to only one or two words at a time, between breaths. These questions were used to define severe wheezing in ISAAC Phase Three12 and will be used unchanged for the Global Asthma Network.

Question 7.
This is the first time in the questionnaire that ‘asthma’ is mentioned. The asthma label is affected by many factors such as awareness of asthma, medical training and experience, cultural and societal factors74. Occasionally asthma may be suggested in the absence of wheeze (on the basis of recurrent nocturnal cough etc). This question was used in all phases of ISAAC. It has not been clear whether the answer represents the parent’s own opinion, or was a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (question 8).

Question 8.
As mentioned in Question 7, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had asthma.

Question 9.
This question, adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 3)75, has been added to the Global Asthma Network surveillance questionnaire. Written asthma management plans are part of most asthma guidelines. Thus this question will provide new information on these plans and will be analysed with the prevalence and severity questions.

Questions 10 & 10a.
Questions 10 and 10a on asthma management are essential to enable the Global Asthma Network to assess its target to decrease the proportion of symptomatic people with asthma not on inhaled corticosteroids. These questions are adapted from ISAAC Phase Two questions (Module 2.2 Asthma management. Questions 1 and 2)75. The wording ‘(when he/she did not have a cold)’ has been added to the end of question 10 to reinforce that the question is asking about breathing problems without the complication of cold symptoms.

Question 11 & 11a.
Questions 11 and 11a ask about swallowed medicines as opposed to inhaled medicines in questions 10 and 10a and adapted from ISAAC Phase Two questions (Module 2.2 Asthma Management. Questions 1 and 2)75. The wording ‘(when he/she did not have a cold)’ has been added to the end of question 11 to reinforce that question 11 is asking about breathing problems without the complication of cold symptoms.

Question 12.
This question about urgent visits for asthma is essential to enable the Global Asthma Network to assess its target to decrease unplanned visits for asthma. This question is adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 5)75.

Question 13.
Different patterns of medical care may contribute to variations in the severity of asthma, between countries or over time. This question and question 12 will explore the relationship between treatment and morbidity and has been adapted from ISAAC Phase Two (Module 2.2 Asthma management. Question 5)75.

Question 14.
This question is about hospital admissions for asthma is essential to enable the Global Asthma Network to assess its target to decrease hospital admissions for asthma. This question is adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 6)75.

Question 15.
This question is adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 9)75. It has been added to the Global Asthma Network questionnaire to provide information about school absenteeism which is an additional indicator of asthma morbidity. It will enable the Global Asthma Network to assess its target to decrease time off school due to asthma.

Question 16.
When piloting questions for ISAAC Phase One, this question was a stem question under question 2. However it was found in some Australasian surveys to identify some children who deny wheezing or whistling at question 1 or 2, yet report wheeze with exercise, and so was moved to question 7 for the full ISAAC study. It was successful in ISAAC placed as question 7108 and it will now follow the new questions on asthma diagnosis and management for the Global Asthma Network questionnaire.

Question 17.
Nocturnal cough is widely accepted as an alternative presentation of asthma, and this question was included in ISAAC to increase the overall sensitivity of the questionnaire, although its specificity in population surveys remained unclear. In ISAAC Phase One, dry night cough in the past 12 months was reported more frequently than 12 month wheeze in all countries except for Australia, New Zealand and Sweden62. Phase Three saw a global increase of night cough of 0.51% per year18. This question has been retained for the Global Asthma Network so that trends over time can be monitored.

10.2.3  Rhinitis questions

ISAAC studied not only asthma but also the related diseases of rhinitis10, 13, 19, 30 and eczema11, 14, 20, 24, 76 . While the Global Asthma Network is focussed on asthma, the surveillance undertaken by the Global Asthma Network gives an opportunity to continue surveillance of rhinitis and eczema too. This would enable the monitoring of time trends in these two related diseases16, 19, 20, 49, 58, 59 where to date there have been only two time points in worldwide studies19, 20 . In addition, the responses to the questions will enable the Global Asthma Network to continue to explore the relationship of asthma with rhinitis and eczema.

The ISAAC Phases One, Two and Three rhinitis core questions (questions 18,20,22-24 below) were based on questions used in previous respiratory epidemiological studies and include both sensitive and specific indicators for respiratory symptoms77, 78 . Due to the experience in the presentation and analyses of the ISAAC rhinitis data10, 13, 16, 19, 61 and due to the identification of new information from the ISAAC data three questions (questions 19, 21 and 25) have been added to enhance the Global Asthma Network rhinitis questions.

Questions 18 & 20.
These questions (prior to ISAAC) were found to have a positive predictive value of 80% in detecting rhinitis in a community sample of adults (aged 16-65 years) in south west London78. These questions were used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 19.
This question determines the age of onset of rhinitis. This question was not included in the ISAAC worldwide surveys.

Question 21.
This question (new for the Global Asthma Network) was piloted successfully in the EISL study79 to determine the presence of allergic rhinitis in children. This question was not used in ISAAC worldwide surveys.

Question 22.
This question had the highest positive predictive value (78%) in detecting atopy among participants with rhinitis78. In ISAAC, question 22 was used in conjunction with question 20 to determine the prevalence of rhinoconjunctivitis by calculating the number of children responding positively to both questions divided by the total number of completed questionnaires10. Rhinoconjunctivitis was justified to be considered in ISAAC to be the variable to be analysed in detail because this symptom combination most closely related to objective indicators of allergic sensitisation in European populations78, 80, 81 . This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 23.
This question was considered a crude qualitative measure of severity when considering questions to use in ISAAC Phase One52 but was considered to correlate well with other indicators of morbidity from rhinitis including reported symptom severity, interference with specific activities of daily living and medical service use. This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 24.
This question permitted investigation of the labelling of rhinitis in relation to the prevalence of rhinitis symptoms. The label “hay fever” had a positive predictive value of 71% in detecting atopy among participants with rhinitis78. This question was used in all ISAAC phases. It has not been clear whether the answer represents the parent’s own opinion, or was a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (question 25)

Question 25.
As mentioned in question 24, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had rhinitis.

10.2.4  Eczema questions

ISAAC studied not only asthma but also the related diseases rhinitis10, 13, 19, 30 and eczema11, 14, 20, 24, 76 . While the Global Asthma Network is focussed on asthma, the surveillance undertaken by the Global Asthma Network gives an opportunity to also continue surveillance of rhinitis and eczema. This would enable the monitoring of time trends in these two related diseases16, 19, 20, 49, 58, 59 where to date there have been only two time points in worldwide studies19, 20 . In addition the responses to the questions will enable the Global Asthma Network to continue to explore the relationship of asthma with rhinitis and eczema.

The ISAAC Phases One, Two and Three eczema core questions (questions 26-32 below) were based on questions used in previous eczema studies82-84 prior to ISAAC and included both sensitive and specific indicators for eczema symptoms. Due to experience in the presentation and analyses of the ISAAC eczema data11, 14, 16, 20, 61 and due to the identification of new information from the ISAAC data one further question (question 33) on diagnosis has been added to enhance the Global Asthma Network eczema questions.

Question 26.
Prior to ISAAC, this screening question was evaluated in a UK pilot study of factors which discriminated “typical” mild-moderate atopic dermatitis from non-atopic eczema and other inflammatory dermatoses presenting for the first time in British hospital outpatient clinics85. This screening question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 27.
Following the form of the ISAAC core questionnaires for wheezing and rhinitis, further enquiry focused on those children with recent rashes to minimise problems of incomplete and selective recall. This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.
Question 28.
This question used in ISAAC worldwide surveys was found to have high sensitivity (94%) and specificity (96%) if case-definition was based on both flexural involvement and age of onset (question 29)84 . This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Questions 29 & 30.
Questions 29 and 30 were included in all ISAAC phases as measures of the severity of eczema, one assessed chronicity, the other morbidity83. In the Global Asthma Network, these questions will be used unchanged.

Question 31.
Respondents with sleep loss of one or more nights per week were considered to have severe eczema based on previous studies86, 87 . This question was used in all ISAAC phases and will be used unchanged for the Global Asthma Network.

Question 32.
This question permitted investigation of the labelling of eczema in relation to the prevalence of eczema symptoms. This question was used in all ISAAC phases. It has not been clear whether the answer represents the parent’s own opinion, or was a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (question 33).

Question 33.
As mentioned in Question 32, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had eczema.

Analyses: asthma, rhinitis and eczema
The analysis of the asthma, rhinoconjunctivitis and eczema core questions will enable the Global Asthma Network to obtain internationally comparable estimates of the direction and magnitude of change in prevalence of symptoms of each condition. Comparisons of prevalence rates between different centres will be made using appropriate statistical methods. Crude rates can be compared by using contingency tables or logistic regression. Comparison of standardised rates or data that needs controlling for confounding will involve multivariate logistic regression.

In addition, calculations will be made of the proportion of participants that have asthma who are not on inhaled corticosteroids, time off work/school because of asthma, unplanned visits for asthma, and hospital admissions. The direction and magnitude of symptoms and severity of asthma will be derived for each centre.

10.2.5  Environment questions

The ISAAC Phase Three EQ (questions 35, 39, 40, 42, 44, 46-48, 53, 54, 58-68 below) were either sourced from existing questions used in previous epidemiological studies (prior to ISAAC), or where these were not available, developed by the ISAAC Steering Committee. Due to experience in the presentation and analyses of the ISAAC EQ and due to the identification of new information from the ISAAC EQ data, further questions have been added to enhance the EQ to be used in the Global Asthma Network.

Inclusion of standardised questions on past and present living and exposure conditions will permit:

  • Between-centre correlations of disease prevalence and risk factor distribution.
  • A pooled evaluation of with-centre analyses of the association between disease and risk factors at the individual level.
  • Monitoring time trends in exposure to risk factors and asthma.

Analyses:
For each EQ, prevalence odds ratios will be calculated using generalised linear mixed models with a binomial distribution and a logit link, with centres being modelled as a random effect. Analyses will be adjusted for gender, region of the world, language and per capita gross national income. Regression models will allow for the sampling of schools by scaling the size of the sample by the design effect. Further multiple regression analyses will be conducted to investigate whether the association between symptoms and a particular EQ is confounded by other risk factors for the information collected from an EQ and which had shown an association with wheezing, rhinoconjunctivitis or eczema in the univariate analyses.

Question 34.
Paracetamol
Beasley et al, in ISAAC Phase Three34 found that use of paracetamol in the first year of life and in later childhood is associated with risk of asthma, rhinoconjunctivitis and eczema at age 6 to 7 years, suggesting to the authors that exposure to paracetamol might be a risk factor for the development of asthma in childhood. There are suggestions that prenatal use of paracetamol might be a risk factor for the development of asthma.

  • Source of question: The Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be undertaken on paracetamol use in pregnancy.

Question 35.
Farm animals
That exposure to farm animals in pregnancy is protective against the development of asthma symptoms in children.
Douwes et al110 in ISAAC Phase Two found a combination of prenatal and current exposure was most strongly associated with wheeze, asthma medication, asthma ever, hay fever and eczema. Although ISAAC has not used these questions they will be used in the Global Asthma Network to test out this hypothesis.

  • Source of question: The Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be undertaken.

Question 36.
Smoking in pregnancy
There is evidence that low birth weight is associated with maternal smoking in pregnancy111-114. ISAAC Phase Three assessed the child’s birth weight and found an association between low birth weight and an increased risk of reported asthma ever and symptoms of asthma (Mitchell 2014). However the authors commented that a limitation of that study was the absence of data on maternal smoking during pregnancy.

  • Source of question: the Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be undertaken.

Question 37.
Prenatal carpet exposure
Douwes et al110 found an association between prenatal exposure to farm animals and prevalence of asthma symptoms. This association further supports the importance of early environmental exposure affecting the risk of subsequent development of asthma115. Carpets are large reservoirs for dust and associated microbial contaminants116. In particular, exposure to dust and microbial agents would be considerably higher in homes with carpets versus those that only have smooth floor covering. This question could therefore be used as a crude proxy of dust/microbial exposure.

  • Source of question: Douwes J and the Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be undertaken.

Question 38.
Premature birth
There is some evidence that premature birth may be associated with bronchial hyperresponsiveness117 which can be explored with this question.

  • Source of question: The Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be undertaken.

Question 39.
Birth weight
It has been hypothesized that low birth weight is associated with increased risk of symptoms of asthma, allergic rhinoconjuncitivitis and atopic eczema. A potential association between low birth weight and an increased risk of chronic obstructive lung disease has been identified118 and is worth further exploration.

  • Source of ISAAC question: ISAAC Steering Committee

Mitchell et al, in ISAAC Phase Three46 confirmed that low birth weight is a risk factor for symptoms of asthma, but not for rhinoconjunctivitis and that the findings for eczema are equivocal.
This question will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified

Questions 40, 40a, 40b, 41.
Breast feeding
It has been hypothesized that breast feeding in the first year of life is protective against the development of asthma, rhinoconjunctivitis and eczema. Introduction of other milk at less than 4 months of age was found by Oddy et al119 to increase the risk of current wheeze 1.31 times (95% CI 1.05 - 1.64) when controlled for gestational age, sex, smoking in the household and childcare attendance in prospective birth cohort study of 2187 children in Western Australia. Rusconi et al22, found that breast feeding for six or more months was associated with a decreased risk of transient early wheeze (OR 0.82 [95% CI 0.68 - 0.97]). Breast feeding was also associated with an increased risk of late-onset wheezing (OR 1.22 [95% CI 0.99 - 1.5]).

  • Source of question 40 for ISAAC: Rusconi et al120. “Was this child ever breastfed?”

To extend the information from question 40 for the Global Asthma Network, the ISAAC Phase Two questions (questions 40a and b) have been included which define how long the child was breast fed and the timing of adding other foods or juices.

  • Source of question 40a, 40b: ISAAC Phase Two75

Bjorksten et al in ISAAC Phase Three38 did not find a consistent association between breast feeding in the first year of life and either a history or current symptoms of wheezing, rhinoconjunctivitis or eczema in 6/7 year old children, but a possible effect on severe symptoms of the latter two conditions.
A further question in the first year of life has been added to the Global Asthma Network questionnaire to define what type of milk did the child most often drink in the first year of life.

  • Source of question 41: Douwes et al110

Analysis of variables. Associations with symptoms will be identified and a more detailed analysis will be undertaken.

Question 42.
Paracetamol use in the first 12 months of life
It has been hypothesized that frequent paracetamol use in the first 12 months of life is associated with an increased risk of symptoms of asthma, rhinoconjunctivitis and eczema. Newson et al102reported, in anecological analyses, a positive association between paracetamol consumption and prevalence of asthma, rhinoconjunctivitis and eczema symptoms in children. They speculated that paracetamol use in the first 12 months of life may influence atopic disease by depleting glutathione in the airways and in immune cells.

  • Source of question: ISAAC Steering Committee

Beasley et al, in ISAAC Phase Three34 found that use of paracetamol in the first year of life and in later childhood, is associated with risk of asthma, rhinoconjunctivitis and eczema at age 6 to 7 years, suggesting to the authors that exposure to paracetamol might be a risk factor for the development of asthma in childhood.
This question will be used unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified and data on paracetamol use in pregnancy will gather prenatal information.

Questions 43, 44, 44a, 44b.
Antibiotics
It has been hypothesized that antibiotic use in the first year of life is associated with an increased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema.
Wickens et al demonstrated an associated risk between symptoms of asthma, allergic rhinoconjunctivitis and atopic ezcema and the use of antibiotics in early childhood121.
Beasley et al, in ISAAC Phase Three36 found an association between antibiotic use in the first year of life and current symptoms of asthma, rhinoconjunctivitis and eczema in children 6 and 7 years old. This question will remain unchanged for the Global Asthma Network.

  • Source of questions 43; 44a & 44b: The Global Asthma Network Steering Group
  • Source of question 44: ISAAC Steering Committee

Analysis of variables. Associations with symptoms will be identified and allow comparisons to be made with data from ISAAC Phase Three. Other questions on antibiotic use will allow further information to be gathered on frequency of chest infections and use of antibiotics.

Question 45.
Sheepskin
This question has the same hypothesis as question 37 and has been added to the Global Asthma Network questionnaire to determine if sleeping on a sheepskin as an infant is associated with symptom prevalence of asthma, rhinoconjunctivitis and/or eczema.

  • Source of question 45: Douwes et al110

Analysis of variables. Associations with symptoms will be identified.

Questions 46 & 47.
Animal allergens
ISAAC hypothesized that exposure to animal allergens is associated with increased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema.
Roost et al103concluded that current cat ownership represented a significant risk for sensitization to cat if cats were allowed indoors. They did, however suggest that childhood exposure to pets, including cats, might modulate immunologic mechanisms and reduce sensitization to cat in adulthood. Hesslemar et al91found that children exposed to cat or dog during the first year of life was associated with a lower symptom prevalence of allergic rhinitis and asthma in school children.

  • Source of ISAAC questions: Hesselmar et al

Brunekreef et al, in ISAAC Phase Three40 found that early life exposure to cats is a risk factor for symptoms of asthma, rhinoconjunctivitis and eczema in 6 and 7 year olds especially in less affluent countries.
These questions will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified. The questions allow identification of separate trigger and sensitisation effects and with data from ISAAC Phase Three.

Question 48.
Farm animals
That exposure to farm animals in pregnancy (question 35) or in the child’s first year of life is protective against the development of asthma symptoms in children.
Douwes et al110 in ISAAC Phase Two found that a combination of prenatal and early life exposure was most strongly associated with wheeze, asthma medication, asthma ever, hay fever and eczema. ISAAC used question 35 but not this current question. Question 48 has been included in the Global Asthma Network to test out this hypothesis.

Questions 49, 50 & 50a.
Wheezing in infancy
The multicentre International Study of Wheezing in Infants (EISL)122-125 was developed to study the prevalence of recurrent wheezing and related risk factors in infants during the first year of life. The questions have been validated and found to be reliable and reproducible for obtaining data on wheezing in children below 36 months of age and for identifying those with probable asthma126. We have taken two questions on the first year of life from the EISL questions for the Global Asthma Network. The wording ‘(when he/she did not have a cold)’ has been added to the end of question 50 to reinforce that question 50 is asking about breathing problems without the complication of cold symptoms.

  • Source of questions: EISL125

Analysis of variables. Associations with symptoms will be identified.

Questions 51, 51a & 52, 52a.
Additional risk factors
These questions have been added to the Global Asthma Network questionnaire to test out other risk factors such as attendance at a child care facility, nursery school, or kindergarten or play centre. These questions have been derived from the German ISAAC Phase Two studies.

  • Source of questions: ISAAC Phase Two modules75

Analysis of variables. Associations with symptoms will be identified.

Questions 53, 54, 55.
Exercise
ISAAC Phase Three hypothesized that regular exercise and physical fitness are protective against asthma. Possible aggravating factors: Being sedentary and lack of physical fitness. An assessment of physical activity in adolescents was undertaken by Aaron et al88 but not undertaken in relationship to asthma and allergies.

  • Source of questions 53 & 54 for ISAAC: Aaron et al88 and Kohl et al89

Mitchell et al in Phase Three43 found a positive association between physical activity and symptoms of asthma, rhinoconjunctivitis and eczema in adolescents but not in children43, 90 and an increased risk of asthma symptoms with television viewing for five or more hours/day in children and adolescents.

  • Source of question 55: The Global Asthma Network Steering Group

This question takes into account time spent on the computer and internet such as games, Facebook, Twitter and YouTube.
Analysis of variables. Combinations of these questions will allow classification of participants into groups based on their level of exercise and whether they are sedentary.

Question 56.
Pneumonia
This question has been adapted from the EISL study125 and the wording ‘had’ changed to ‘diagnosed’ for the Global Asthma Network as previous experience in ISAAC shows that diagnosis is a more robust data source.

  • Source of question: EISL125

Analysis of variables. Associations with a diagnosis of pneumonia or bronchopneumonia and symptoms will be identified in the first year of life.

Question 57.
Twin sibship
A new question on whether the participant is a twin has been added following the Global Asthma Network pilot study. This will avoid confusion when twins complete questions 58 and 59 on older/younger siblings.

  • Source of question 57: The Global Asthma Network Steering Group

Questions 58 & 59.
Parity
ISAAC hypothesized that increased household size was associated with a decreased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema91. A decrease in the cumulative incidence and prevalence of allergic rhinitis and the cumulative incidence of asthma with increasing number of siblings was reported by Hesselmar et al91.
Aggravating factors could be small family size and early birth order.

  • Source of ISAAC question: ISAAC Steering Committee

Strachan et al in ISAAC Phase Three48 found new observations of greater symptom severity among children from larger families which seems to be a more consistent phenomenon worldwide, and deserves further exploration and explanation.
These questions will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified. The questions allow identification of separate effects of birth order and number of siblings and for the Global Asthma Network analysis will build on the information from ISAAC.

Questions 60, 60a & 61.
Migration
ISAAC hypothesized that migrants to a new country will adopt the prevalence of symptoms of asthma, allergic rhinoconjuncitivitis and atopic eczema of their new country over time92-97.
There was conflicting evidence in the literature regarding the health of immigrants. Robertson et al92 reported that symptoms of asthma, rhinoconjunctivitis and eczema were more common in children born in Australia compared with children born in other countries but resident in Australia. Leung et al95 reported that prevalence of hay fever and asthma increased with length of stay among Asian immigrants to Australia.

  • Source of ISAAC questions 60 & 61: ISAAC Steering Committee

García-Marcos et al in ISAAC Phase Three47 found that recent migration to high prevalence/affluent countries is associated with a lower prevalence of allergic diseases. The pre-migration environment might be protective, but this effect quickly decreases with increasing time in the host country.
These questions will be used unchanged in the Global Asthma Network. Use of these questions in the Global Asthma Network will allow an up to date observation of these findings. For the country of birth question, if the respondent answered ‘NO’ to question 60, the Global Asthma Network will include a question about what country the respondent was born in (question 60a).

  • Source of question 60a: The Global Asthma Network Steering Group

Analysis of variables. Associations with symptoms will be identified. The questions allow assessment of the effect of time spent in the country and country of birth.

Questions 62, 63.
Additional risk factors
These questions have been added to the Global Asthma Network questionnaire to test out other risk factors such as floor covering in the child’s bedroom or changes made to the home because of allergies. These questions have been derived from the German ISAAC Phase Two studies.

  • Source of questions: ISAAC Phase Two modules75

Analysis of variables. Associations with symptoms will be identified.

Question 64.
Traffic
ISAAC hypothesized that respiratory irritants such as sulphur dioxide (SO2) nitrogen oxides (NOx) and particulates from diesel combustion cause local respiratory inflammation, increasing tissue contact with inhaled allergens and the likelihood of an allergic response98, 99 .

  • Source of ISAAC question: Weiland et al98

Brunekreef et al in ISAAC Phase Three37 found that higher exposure to self-reported truck traffic on the street of residence is associated with increased reports of symptoms of asthma, rhinitis and eczema in many locations in the world.
This question will remain unchanged for the Global Asthma Network, however ‘(not often)’ will be added next to ‘seldom’ as an option as the Global Asthma Network pilot study found some participants did not understand the word ‘Seldom’.
Analysis of variables. Associations with symptoms will be further explored in the Global Asthma Network.

Question 65.
Diet
ISAAC Phase Three hypothesized that a plant based diet protected against asthma and allergies and a “Western” diet was positively associated with asthma and allergies.
Protective and aggravating factors found in the ISAAC Diet ecological analysis100 included starch, cereals, rice, vegetables, fish, other seafood, fibre, fruit, nuts, olive oil (protective); trans fatty acids, fast foods (aggravating). Other foods considered included eggs, animal fats, milk, polyunsaturated fatty acids. Dietary surveys in the literature were lengthy questionnaires and either interviewer administered or adult self-completed, food frequency questionnaires either prospective using a diary or retrospective by recall (7 day, 3 months, 12 months) or “dietary history” which is a recalled food frequency questionnaire. No short diet questions suitable for inclusion in the ISAAC EQ were identified.

  • Source of question: ISAAC Steering Committee

Ellwood et al, in ISAAC Phase Three44 found a negative association between the intake of fast food and asthma, rhinitis and eczema symptoms and a positive association between the intake of fresh fruit and vegetables. These associations were also found by Nagel et al29 and Wickens et al101 in ISAAC Phase Two. The ISAAC Phase Three EQ questions will be used in the Global Asthma Network with some food items being separated (such as cooked and raw vegetables) and additional food items such as fizzy or soft drinks. Since we found the association with fast food we have been questioned about putting, for example, McDonald’s with fish and chips, so we are adding a food type to separate this out. The fast food outlets that are common in most countries are McDonald’s and Burger King which enables burgers to be studied as a separate group. Food types that have been separated out are: Cooked Vegetables; Raw Vegetables; Cereals (excluding bread); Bread; Olive Oil; Other dairy (include cheese and yoghurt); Sugar (include lollies/candies/sweets); Fast food/ burgers; Fast food excluding burgers; Fizzy or soft drinks (include local terminology).
Analysis of variables. Components will be analysed individually, some dietary items may be combined, such as a plant based diet, or a Mediterranean diet. Details of these combinations are yet to be developed.

Questions 66 & 67.
Allergens
ISAAC hypothesized that exposure to allergens is associated with increased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema.
Roost et al103concluded that current cat ownership represented a significant risk for sensitization to cat if cats were allowed indoors. They do, however suggest that childhood exposure to pets, including cats, might modulate immunologic mechanisms and reduce sensitization to cat in adulthood.
Source of ISAAC questions: Roost et al103
Brunekreef et al, in ISAAC Phase Three40 found that early life exposure to cats is a risk factor for symptoms of asthma, rhinoconjunctivitis and eczema in 6 and 7 year olds especially in less affluent countries.
These questions will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified. The questions allow identification of separate trigger and sensitisation effects and with data from ISAAC Phase Three.

Question 68.
Paracetamol
It has been hypothesized that frequent paracetamol use is associated with an increased risk of symptoms of asthma, allergic rhinoconjunctivitis and atopic eczema102.
Newson et al102reported, in an ecological analyses, a positive association between paracetamol consumption and prevalence of asthma, allergic rhinoconjunctivitis and atopic eczema symptoms in children. They speculate that paracetamol may influence atopic disease by depleting glutathione in the airways and in immune cells.

  • Source of questions: ISAAC Steering Committee

Beasley et al35 in ISAAC Phase Three found that the use of paracetamol may represent an important risk factor for the development and/or maintenance of asthma, rhinoconjunctivitis and eczema in adolescent children. This was not explored in the 6/7 year old children. This question will remain unchanged for the Global Asthma Network.
Analysis of variables. Associations with symptoms will be identified.

10.3  Parent/guardian questions (about their own health)

The analysis of the asthma, rhinoconjunctivitis and eczema core and management questions will enable the Global Asthma Network to obtain internationally comparable estimates of the direction and magnitude of change in prevalence of symptoms. Comparisons of prevalence rates between different centres will be made using appropriate statistical methods. Crude rates can be compared by using contingency tables or logistic regression. Comparison of standardised rates or data that needs controlling for confounding will involve multivariate logistic regression. These analyses will also allow confirmation (or otherwise) of previously hypothesized associations and from the analysis of the ISAAC data.

10.3.1  Asthma questions

These questions are built on those used in the International Study of Asthma and Allergies in Childhood (ISAAC)52 and in the European Community Respiratory Health Survey (ECRHS)60, 127 . The ISAAC Phases One, Two and Three core questions (questions 2, 3, 4, 8 and 9 below) were piloted before Phase One57 and published49, 58, 59 . Most were based on questions used in previous respiratory epidemiological studies (prior to ISAAC) and include both sensitive and specific indicators of asthma60. Many of these questions have been used in ISAAC in 306 centres in 105 countries. They were translated from English into 52 languages in ways which were understood locally50. They have been used in key ISAAC asthma symptoms publications; Phase One61, 62 , Phase Two33 and Phase Three12, 16, 18 .
The core questions used have been validated with bronchial hyper-responsiveness3, 4, 22, 23, 63-70 .
Due to experience in the presentation and analyses of the ISAAC asthma data12, 16, 18, 33, 61, 62 and the identification of new information from the ISAAC data, this questionnaire was designed for the adults of the 6/7 and 13/14 year old participants. New questions have been added to this questionnaire about adult health.

Question 1.
This question is based on the ECRHS survey, prior to ISAAC, which includes sensitive and specific indicators of asthma60, 127

Question 2.
Limitation to a 12 month period reduces errors of recall71 and is believed to be independent of month of completion72. This is considered to be the most useful question for assessing the prevalence of wheezing illness. This question was used successfully in ISAAC and it will be used unchanged for the Global Asthma Network.

Question 3 & 4.
These questions used in ISAAC offer alternative quantitative measures of the frequency and severity of wheezing. Problems with the concept of “attacks” and difficulty in quantifying the frequency of recurrent asthma lead to the inclusion of question 4 to identify and quantify persistent wheeze. These two questions were used to identify asthma severity in ISAAC, in a comparison of asthma symptom prevalence, mortality and hospital admissions which demonstrated correlations with all three parameters73. In that paper, the 12-month prevalence of moderate to severe wheezing comprised one or more of: (i) four or more attacks of wheeze; (ii) woken by wheeze on one or more nights per week or; (iii) wheezing severe enough to limit speech to only one or two words at a time, between breaths. These questions were used to define severe wheezing in ISAAC Phase Three12 and will be used unchanged for the Global Asthma Network adult group.

Questions 5, 6 & 7.
Question 5 is used in the ECRHS60, 127 screening questionnaire (Q1.1) preceded by Question 2 on this questionnaire (Have you had wheezing or whistling in the chest at any time in the past 12 months?). Question 6 is a modified question of Question 5 of the ECRHS II main questionnaire (shortness of breath). Question 7 is a modified question of Question 6 of ECRHS II main questionnaire (woken by an attack of coughing).

Question 8. This question was used in ISAAC Phases One Two and Three, as there was a dearth of epidemiological information relating to acute, severe asthma, which is of direct relevance for international comparisons of hospital admissions and mortality statistics. This question filled this gap and has provided good data on asthma severity12, 22, 61 . This question will used unchanged for the Global Asthma Network adults.

Question 9.
This is the first time in the questionnaire that ‘asthma’ is mentioned. The asthma label is affected by many factors such as awareness of asthma, medical training and experience, cultural and societal factors74. Occasionally asthma may be suggested in the absence of wheeze (on the basis of recurrent nocturnal cough etc). This question was used in all phases of ISAAC. It has not been clear whether the answer will represent the parent’s own opinion, or is a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (Question 10).

Question 10.
As mentioned, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had asthma for question 9.

Question 11.
This question, adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 3)75, has been added to the Global Asthma Network surveillance questionnaire. Written asthma management plans are part of most asthma guidelines. Thus this question will provide new information on these plans and will be analysed with the prevalence and severity questions.

Question 12.
This is a new question set by the Global Asthma Network Steering Group for the Global Asthma Network to determine age of onset of asthma.

Question 13.
This question, ‘Have you had an attack of asthma in the past 12 months?’ was Question 5 in the ECHRS screening questionnaire.

Question 14, 14a, 15 & 15a.
Questions 14 and 15 on asthma management are essential to enable the Global Asthma Network to assess its target to decrease the proportion of symptomatic people with asthma not on inhaled corticosteroids. Question 14 is adapted from ISAAC Phase Two questions (Module 2.2 Asthma management. Questions 1 and 2)75. Question 15 is adapted from the ECRHS main questionnaire (Question 61). Questions 14 and 15 have had the wording ‘(when you did not have a cold)’ added to the end of the question to reinforce that the questions are asking about breathing problems without the complication of cold symptoms.

Question 16 & 17.
These questions about urgent visits for asthma in the past 12 months are essential to enable the Global Asthma Network to assess its target to decrease unplanned visits for asthma. These questions have been adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 5)75.

Question 18.
This question is about hospital admissions for asthma is essential to enable the Global Asthma Network to assess its target to decrease hospital admissions for asthma. This question is adapted from an ISAAC Phase Two question (Module 2.2 Asthma management. Question 6)75

Question 19 & 20.
Questions 19 and 20 are adapted from the ECRHS II main questionnaire. Question 19 is adapted from ECHRS questions 90 and 91 and question 20 is adapted from ECHRS question 29.

10.3.2  Rhinitis questions

Question 21.
This is the first time in the questionnaire that ‘hay fever’ is mentioned. This question was used in all phases of ISAAC. It has not been clear whether the answer will represent the adult’s own opinion, or is a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (Question 22).

Question 22.
As mentioned, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had hay fever for question 21.

10.3.3  Eczema questions

Question 23.
This is the first time in the questionnaire that ‘eczema’ is mentioned. This question was used in all phases of ISAAC. It has not been clear whether the answer will represent the adult’s own opinion, or is a label given by a doctor. In the Global Asthma Network the same question will be used, followed by a clarification question (Question 24).

Question 24.
As mentioned, this is an additional question for the Global Asthma Network to clarify that a doctor had confirmed that the participant had eczema for question 23.

10.3.4  Environmental questions

(see sections 10.1.5 and 10.2.5 for analyses)
Question 25.
Socio economic status
It has been hypothesized that increased socio-economic status (SES) is associated with increased risk of symptoms of asthma.
Studies of the relationship between SES and health have shown that SES is multidimensional, incorporating elements of occupational characteristics, education, income, wealth and residential characteristics.

  • Source of ISAAC question: Durkin et al 128

This question was included in the ISAAC 6/7 year EQ and has been removed from that age group questionnaire for the Global Asthma Network and transferred to the Global Asthma Network adult questionnaire
Analysis of variables. Associations with symptoms will be identified.

Questions 26, 27 & 28.
Moisture in the home
It has been hypothesized that moisture/mould in the home environment is associated with increased risk of asthma.
These questions, taken from ISAAC Phase Two75, and have been adapted and expanded for the Global Asthma Network adult questionnaire.
Weinmayr et al31 found in ISAAC Phase Two a consistent association of dampness with respiratory and other symptoms was found in both affluent and non-affluent countries, among both atopic and non-atopic children.

  • Source of question: ISAAC Phase Two

Analysis of variables. Associations with symptoms will be identified.

Questions 29-33.
Cooking
It has been hypothesized that gas cooking is associated with increased risk of symptoms of asthma129.
Gas cooking has shown mixed effect in epidemiological studies. Kerkhof et al129 showed increased bronchial responsiveness among persons with high total IgE levels who use gas for cooking suggesting that atopic subjects are sensitive to adverse effects of gas cooking on respiratory health. However, Moran et al130in their study concluded that the use of gas for cooking was unlikely to be a major influence on respiratory morbidity in young adults.

  • Source of ISAAC Phase Three questions: ISAAC Steering Committee
  • Source of ISAAC Phase Two questions: ISAAC Phase Two Steering Group
  • Source of the Global Asthma Network Adult questions: Adapted from ISAAC Phases One and Two
  • Source of the Global Asthma Network question 30: WHO. Tuberculosis prevalence surveys: a handbook. 2011131

Wong et al132 in ISAAC Phase Three found the use of open fires for cooking is associated with an increased risk of symptoms of asthma in children. Because a large percentage of the world’s population uses open fires for cooking, this method of cooking might be an important risk factor if the association is proven to be causal. Worldwide, respiratory health effects account for nearly a half of the overall deaths and disabilities from household air pollution133.
Analysis of variables. Associations in the symptoms will be identified. Combinations of the variables may be used to identify severe exposure to gas combustion products.

Questions 34-36.
Heating

  • Source of question 34: ISAAC Phase Three EQ
  • Source of the Global Asthma Network Question 35 & 36: WHO. Tuberculosis prevalence surveys: a handbook. 2011131

Question 37.
Diet
ISAAC Phase Three hypothesized that a plant based diet protected against asthma and allergies and a “Western” diet was positively associated with asthma and allergies. Protective and aggravating factors found in the ISAAC Diet ecological analysis100 included starch cereals, rice, vegetables, fish, other seafood, fibre, fruit, nuts, olive oil (protective); trans fatty acids, fast foods (aggravating). Other foods considered included, eggs, animal fats, milk, polyunsaturated fatty acids. Dietary surveys in the literature were lengthy questionnaires and either interviewer administered or adult self-completed, food frequency questionnaires either prospective using a diary or retrospective by recall (7 day, 3 months, 12 months) or “dietary history” which is a recalled food frequency questionnaire. No short diet questions suitable for inclusion in the ISAAC EQ were identified.

  • Source of question: ISAAC Steering Committee

Ellwood et al44, in ISAAC Phase Three found a negative association between the intake of fast food and asthma, rhinitis and eczema symptoms and a positive association between the intake of fresh fruit and vegetables. These associations were also found by Nagel et al29 and Wickens et al101 in ISAAC Phase Two. The ISAAC Phase Three EQ questions will be used in the Global Asthma Network with some food items being separated (such as cooked and raw vegetables) and additional food items such as fizzy or soft drinks. Since we found the association with fast food we have been questioned about putting, for example, McDonald’s with fish and chips, so we are adding a food type to separate this out. The fast food outlets that are common in most countries are McDonald’s and Burger King which enables burgers to be studied as a separate group. Food types that have been separated out are: Cooked Vegetables; Raw Vegetables; Cereals (excluding bread); Bread; Olive Oil; Other dairy (include cheese and yoghurt); Sugar (include lollies/candies/sweets); Fast food/ burgers; Fast food excluding burgers; Fizzy or soft drinks (include local terminology).
Analysis of variables. Components will be analysed individually, some dietary items may be combined, such as a plant based diet, or a Mediterranean diet. Details of these combinations are yet to be developed.

Questions 38-41.
Smoking
It has been hypothesized that exposure to tobacco smoke in early life is associated with increased risk of symptoms of asthma, rhinoconjunctivitis and eczema. That exposure to tobacco smoke is a trigger for asthma attacks for asthmatics104. In the ISAAC Phase Three EQ, the questions from Jarvis104 (1999) were used. However in the Global Asthma Network, updated questions (2012) from a WHO survey, the Global Adult Tobacco survey, will be used.

  • Source of questions: The Global Adult Tobacco survey 2012105

Analysis of variables. Associations with symptoms will be identified.

Question 42.
Water pipe smoking
It has been hypothesized that water pipe smoking is associated with increased risk of symptoms of asthma.

  • Source of question: Adapted from Mohammad Y et al. Respiratory effects in children from passive smoke of cigarettes and narghile: ISAAC Phase Three in Syria106