Methods for the Assessment of Asthma-Related Health Outcomes Workshop 

May 27–28, 2004
Rosen Centre Hotel
Orlando, FL

This meeting was held in conjunction with the National Children’s Study, which is led by a consortium of federal agency partners: the U.S. Department of Health and Human Services (including the National Institute of Child Health and Human Development [NICHD] and the National Institute of Environmental Health Sciences [NIEHS], two parts of the National Institutes of Health, and the Centers for Disease Control and Prevention [CDC]) and the U.S. Environmental Protection Agency (EPA).

Welcome, Greetings, Introductions
Co-Chairs: Hector G. Ortega, M.D., Sc.D., NHLBI, NIH, DHHS; Lucas Neas, Sc.D., EPA; and Erik Svendsen, Ph.D., EPA

Dr. Svendsen welcomed workshop participants and thanked them for their commitment to the National Children’s Study (Study) and the Methods for the Assessment of Asthma-Related Health Outcomes Workshop. He briefly reviewed the Study’s history, noting the following:

• The Study will follow 100,000 American children from before birth through early adulthood.
• Congress mandated the Study and stipulated that the Study be planned and managed by several collaborating federal agencies (see introductory paragraph).
• The Study’s purpose is to answer critical questions on the health and well being of children.
• The Study will be driven by a limited set of core hypotheses, focusing on five priority health and disease outcomes:
  • Asthma
  • Childhood injury
  • Neurodevelopment and behavior
  • Obesity and physical development
  • Pregnancy outcomes.

Dr. Svendsen explained that the Asthma Working Group has developed and submitted its proposal for core hypotheses on asthma. This 19-page document, titled "Initial Proposal for Core Hypotheses/Questions," was distributed to workshop participants; it is also available from the Asthma Working Group co-chairs.

The primary purpose of the workshop is to develop an informal plan to recommend to the Study’s Interagency Coordinating Committee. This plan will describe how the Study will assess asthma outcomes during early childhood years. Dr. Svendsen stated that the specific workshop goals are to:

• Determine the questions needed to assess asthma in early childhood, when they are most effectively queried, and how they should be queried
• Identify minimally invasive techniques of measuring pulmonary function that could be applied (with modification) to young children
• Identify biomarkers to evaluate asthma and immune system function in young children that are feasible on a national scale
• Determine the need for pilot studies on modifications of some of these biomarkers/methods.

In addition to presenting information on instruments and core clinical measure/markers used in asthma epidemiology, workshop participants were asked to discuss and address the specific goals in breakout sessions. The expected products of breakout sessions include comprehensive lists of:

• Recommended questions, both diary and questionnaire, when they should be used, and how they should be queried
• Feasible biomarkers to evaluate asthma and immune function in young children
• Available pulmonary function measures suitable for the Study
• Pilot studies for modifications to pulmonary function measures and/or biomarkers recommended for the Study.

Charge: Asthma Working Group Co-Chairs
Stephanie J. London, M.D., Dr.P.H., NIEHS, NIH, DHHS, and Fernando D. Martinez, M.D., University of Arizona, Tucson

Dr. London thanked Drs. Ortega and Svendsen for organizing the workshop and also thanked participants for their involvement. She then described the current funding situation for the Study: Funding has been allocated for various planning activities, including workshops and pilot studies, but funding to actually implement the Study has not yet been allocated. Dr. London explained that over the past 2½ years, the Study working groups have been developing suggested hypotheses, identifying issues that can be "profitably" studied in 100,000 children and that require a cohort of this size, and exploring the practicalities of implementing such a large-scale longitudinal study. One of the Asthma Working Group’s last tasks is to write a report on the outcome of the current workshop. The Study’s federal advisory committee expects a best assessment of the asthma-related questionnaires; instruments; biomarkers; and minimally invasive measures that are practical, feasible, valuable, valid, and informative for studying infant and child pulmonary function in a large national cohort.

Dr. Martinez reminded the workshop participants of the Study’s main areas of interest and emphasized the Study’s premise of correlating environmental exposures and subsequent health outcomes, including asthma. According to Dr. Martinez, the Study has the potential to be one of the most significant studies in relation to environmental and genetic determinants of the different expressions of recurrent airway obstructions that occur during childhood. The Study will begin before birth, assess factors that occur during pregnancy and the first years of life, and test the hypotheses developed by the Asthma Working Group. Dr. Martinez emphasized that asthma authorities consider the Study to be a crucial research endeavor because of the opportunity to examine the three main elements—the essential determinants of disease development—of asthma:

• Genes
• Environment
• Development.

If the Study is able to craft the appropriate survey instruments and questionnaires, and establish the most suitable clinical outcome measures and biomarkers, then researchers may be able to identify the genetic and environmental factors expressed during development that determine why certain children get asthma. It is hoped that this knowledge will ultimately be used to plan prevention strategies to ameliorate the disease effects of asthma in American children.

Introduction of Core Data Collection Instruments, Measures, Markers, and Studies
Drs. Ortega and Svendsen

Dr. Svendsen stated that everyone who is involved in the designing and planning of the Study is part of a new groundbreaking effort. Smaller asthma studies have been implemented and many of them have been very successful. Several workshop participants were involved in these smaller successful studies, and they can offer insight on these studies beyond what appears in the literature. The Asthma Working Group is trying to draw upon the knowledge, experience, and expertise of these researchers in developing the core asthma data collection instruments, measures, and markers for the Study.

Dr. Svendsen said that numerous studies have used questionnaire materials to assess asthma in early childhood. Workshop materials include examples from:

• Norwegian Mother and Child Study
• CHAMACO Study
• PIAMA Study (Netherlands)
• Tucson studies (Tucson Children’s Respiratory Study; Infant Immune Study)
• Yale Childhood Asthma Study
• Detroit birth cohort studies (Childhood Allergy Study; Wayne Tri-County Health, Environment, Allergy, and Asthma Longitudinal Study)
• Boston’s Birth Cohort Study
• Dunedin Cohort Study (New Zealand).

Dr. Svendsen explained that workshop participants who were involved with these studies would have an opportunity to informally discuss their perspectives on what was done in studies, that is, "what worked and what did not." To begin the discussion, Dr. Ortega presented some information on measures and markers of asthma. He noted that over the past 50 years, there has been tremendous development of potential methods of monitoring asthma in children. Some of the invasive methods are not appropriate for the Study, but several of the minimally invasive methods might be appropriate for studying children. Dr. Ortega listed the following markers of inflammation:

• Several blood markers of inflammation
• Bronchoalveolar lavage and bronchial biopsy
• Induced sputum
• Exhaled nitric oxide (NO)
• Exhaled breath condensates
• Nasal samples (disc, filter, lavage, condensates, NO, and so on)
• Lymphocyte proliferation assays.

Measures of airways physiology to be considered for the Study include:

• AM peak flow
• Force expiratory volume (FEV) ₁
• FEV ₁ after albuterol
• Methacholine challenge (PC ₂₀ )
• Acoustic rhinometry
• Rhinomanometry
• Nasal challenge.

Dr. Ortega listed the following critical considerations for biomarkers:

• Collection procedure—dilution
• Standardization
• Performing the test
• Explaining the procedure (consent form)
• Home versus clinic collection
• Individual versus multiple samples
• Sample storage.

Other critical considerations for biomarkers include:

• Reliability
• Reproducibility
• Repeatability—intraclass correlation coefficient (ICC)
• Validation
• Utility.

Some characteristics of induced sputum as a biomarker are:

• Repeatability—ICC
• Pizzichini et al.’s (Hamilton, Canada) new method to assess induced sputum—store/refrigerate 7 hours
• ICC = 0.95.

Other methods for biomarkers are:

• Saliva used as noninvasive method
• LTB ₄ and LTE ₄ measured before and after treatment with 5-LO inhibitor using enzyme immunoassay (EIA) and confirmed by RP-HPLC.

Dr. Ortega presented an example of measurement of clinical markers at various study visits. Clinical markers included exhaled NO, spirometry, saline challenge, and sputum analysis. Dr. Ortega also presented a chart that compared exhaled NO from subjects with asthma and control subjects. Control levels were clustered in the 3- to 10-parts-per-billion range, where levels for subjects with asthma were scattered across the 10- to 30-parts-per-billion range (from Brent et al., 1997). In concluding, Dr. Ortega listed advantages and disadvantages of the bag method for collecting exhaled breath samples.

Question-and-answer session/comments. In addressing concerns about the overall design of the Study, Scott T. Weiss, M.D., M.S., Harvard University, asked about the entry criteria for women and infants into the Study and asked whether there were any exclusions. He noted that there is a strong relationship between asthma and premature births. Specifically, Dr. Weiss asked whether premature infants would be included in the Study cohort. In response, Grace K. LeMasters, Ph.D., University of Cincinnati College of Medicine, explained that the sample selection for the Study has not been finalized; it may possibly be a national probability sample. Dr. LeMasters—who is a member of the Study’s federal advisory committee—said that because the Study intends to recruit women early in pregnancy (as early as possible in the first trimester), adverse pregnancy outcomes will be included and that there will be no restrictions in terms of conditions. In response to a participant’s question about access to Study data, Sarah A. Keim, M.A., NHLBI, NIH, DHHS, replied that although a data publication policy has not yet been formulated, the intent is to share data in an effort to foster collaboration among Study groups. There will be a centralized data coordinating center and a central statistical system for wide use among researchers. Other discussion topics included:

• Recruitment of various ethnic groups
• Possibility of oversampling
• Location of Study centers
• Data availability for analysis (exclusive use versus public use)
• Input on Study design and protocol
• Review of applications to participate in the Study.

Presentation of Instruments and Core Clinical Measures/Markers Used in Asthma Epidemiology: Focus on Strengths and Limitations for Application in the Study

Cincinnati Childhood Allergy and Air Pollution Study. In this first presentation, Dr. LeMasters described aspects of this recently implemented study, which is designed to assess health effects of exposure to diesel engine exhaust in the Greater Cincinnati area. Specifically, the study will investigate whether allergy and asthma increases are stimulated by exposure to toxic air pollutants or other environmental triggers. Approximately 20,000 diesel trucks travel daily on each of Greater Cincinnati’s interstate highways, and two of these highways pass directly through the city. Approximately 11-12 percent of the area’s population lives within 400 meters of an interstate highway.

The study investigators used birth certificate records and GIS (geographic information system) to identify two groups of families: those living within 400 meters of interstate highways and those living 1,500 meters beyond the highways. To date, the study has recruited a cohort of 800 children, half of whom live close to an interstate highway. Study investigators used an eligibility-screening questionnaire to identify atopic families. The questionnaire asked mothers 14 questions. If a mother responded positively to one of the questions (that is, reports an allergy symptom), then she was given a skin-prick test to assess a battery of air allergens. Dr. LeMasters said that, of the women who responded "yes" on the questionnaire, 80 percent tested positive to an air allergen on the skin-prick test. This result was greater than expected. At around 6 months of age, children of atopic mothers were given skin-prick tests for both air and food allergens. First-year allergic assessments indicate that 26 percent of these children tested positive to either air or food allergens.

Dr. LeMasters explained that skin-prick tests are performed on the back and that images of the results are created. The images are scanned into a digital format and analyzed using an architectural/engineering application called AutoCAD. This application accurately measures wheal dimensions. At the time of testing, children are also given a physical exam, and the mothers complete a questionnaire about children’s allergy symptoms. Mothers maintain monthly diaries of their children’s symptoms. Primary study outcomes include allergic rhinitis and wheezing. The wheezing assessment tool was adapted from various sources, including instruments used by the International Study of Allergy and Asthma in Children (ISAAC) and the Yale Childhood Asthma Study. Besides the phenotypic markers of the skin-prick test, the study will assess primary disease outcomes such as allergic rhinitis, dermatitis, eczema, wheezing, and middle ear and sinus infections.

The study is monitoring air pollution near interstate highways, including primary exposure levels to diesel exhaust particles, volatile organic compounds (VOCs), and metals. Investigators are sampling in-home dust and mold, performing personal sampling of study subjects, and determining children’s exposures to tobacco smoke.

Question-and-answer session/comments. A participant asked whether the study is using an objective dermatitis assessment tool such as the modified SCORAD index. Dr. LeMasters said that the study is using both questionnaires/diaries and objective assessment through physical examination. Another participant asked whether the study is using the Williams protocol from ISAAC to assess atopic dermatitis. Dr. LeMasters replied that the study is using the Eczema Area Severity Index (EASI). The participants discussed several methodological issues, including:

• Parents’ acceptance of skin-prick tests
• Discomfort/tolerance of skin tests
• Skin test standardization
• Use of radioallergosorbent (RAST) tests
• Use of immunoglobulin assays (that is, serum versus skin testing)
• Test–retest reliability.

Yale Childhood Asthma Study. Kathleen Belanger, Ph.D., Yale University School of Medicine, said that this study recruited 1,002 families. Entry criteria specified mothers with a new baby and an older child with physician-diagnosed asthma. Women were screened in maternity hospitals soon after delivery. The study successfully enrolled 1,000 children, who are now, on average, about 6 years old. Enrollment occurred over a 2-year period. Dr. Belanger explained that the primary outcome measure for the study is physician-diagnosed asthma. The study initially relied on mothers reporting the physician’s diagnosis, which was subsequently verified directly with the physician. Although there was a high concordance between mothers’ reports and physicians’ diagnoses, Dr. Belanger reported that physicians were occasionally less willing to commit to an asthma diagnosis, depending primarily on the child’s age.

During the first 3 years of the child’s life, the study collected information on mothers’ observations every 3 months. Questionnaires asked about asthma diagnosis and symptoms such as wheeze, persistent cough, shortness of breath, and chest tightness. Dr. Belanger commented that collecting observational information on the latter two symptoms was not particularly reliable. The study revealed a poor correlation between observed symptoms and diagnoses. Some physicians were reluctant to diagnose asthma in children younger than 6 years of age, whereas others deemed it appropriate to treat asthma symptoms even without a specific diagnosis. Dr. Belanger noted that there appears to be variability among physicians’ willingness to commit to a diagnosis of asthma. Furthermore, there is little correlation between diagnosis and use of medications, which range from daily prescribed medications to seasonal use of over-the-counter (OTC) products.

Despite these problems, Dr. Belanger expressed her belief that any study of asthma will need to collect some type of data on observable symptoms on a regular basis. Dr. Belanger said that, in addition to wheeze and persistent cough, mothers’ observations on children’s retractions may provide valuable information. A participant commented that information on retractions would be "dangerous" (that is, potentially unreliable); the only reliable observable data would be a child’s sleeping respiratory rate. Dr. Belanger commented that it is difficult to know the appropriate questions to ask in a study without sufficiently knowing the study design.

Frequency of parental contact can be an issue in asthma studies, as can the participant’s burden; that is, how much can investigators ask a mother to do regarding observing and reporting? According to Dr. Belanger, there may need to be an inverse relationship between frequency of contact and the number of years a subject will participate in a study. Dr. Belanger cited a 1-year study in which mothers were contacted every 2 weeks and asked about their children’s daily symptoms. The primary complaint of study subjects was not necessarily the intensity of data collection (because the study’s end was in sight) but the repetition of questions. An important issue for studying asthma in a national cohort will be considering the practical realities of collecting data; that is, balancing what investigators want to do and what can be done in a reasonable and feasible manner.

Question-and-answer session/comments. Peter D. Sly, M.D., D.Sc., M.B.B.S., University of Western Australia, noted that study assessments could have different goals for different subject age groups (that is, different questions are asked at different developmental stages). However, core data collection items must be addressed on a regular basis (for example, annually) for a longitudinal study. Dr. Martinez commented that ultimately workshop participants must consider what is essential to collect good data on asthma and not worry about other aspects or components of the Study. Simply stated, participants need to focus their attention on the measures, markers, and questionnaires regarding the most appropriate assessment of asthma symptoms and disease development. The participants discussed the following issues related to data collection and questionnaires:

• Streamlining questionnaires
• Adding/changing questions as children get older (that is, age-appropriate data collection)
• Analyzing/comparing data combined from various major studies (that is, meta-analyses)
• Questionnaire length as related to subject motivation, incentives, and socioeconomic status
• Central diagnosis of eczema lesions using photographs
• Standardization of symptom assessments (for example, redness, itching, dryness)
• Clinical standards versus parental observations.

Tucson Studies. Anne L. Wright, Ph.D., University of Arizona, Tuscon, reported on two studies: the Tucson Children’s Respiratory Study and the Infant Immune Study. The Tucson Children’s Respiratory Study, which began in 1980, enrolled 1,246 children at birth; these subjects are now 20-24 years old. Questionnaire information on this cohort was collected every 2-3 years, with in-depth, detailed information collected every 4-5 years. The advantage of this approach is in collecting both symptom and diagnosis information repeated throughout the subjects’ first 20 years of life. The detailed evaluations include lung function tests, allergy skin-prick tests, and blood markers (such as immunoglobulin E [IgE] and eosinophils). Dr. Wright said that the study response rate was approximately 78 percent. She commented that, although subjects get tired of repeatedly answering the same questions, it is absolutely essential that the language and phrasing of core questions be consistent and unchanged over the course of a study.

Dr. Wright explained that the Infant Immune Study began in 1997 by enrolling mothers during pregnancy, and enrollment was only recently completed. The response rate was 50 percent; that is, half of the eligible women who were asked to participate accepted and enrolled in the study. Because the study’s primary focus is on development of the immune system, subjects are required to provide blood samples. Dr. Wright suspected that this requirement led to the lower response rate. Currently there are 490 "nonselected" study participants.

Dr. Wright said that one of the challenges of the Infant Immune Study is gathering sufficiently detailed information without overwhelming the participants. Symptom questionnaires are completed every 2 months during the first year of life and then once a year after that. Cord blood is collected at birth. Subsequent blood samples are collected at 2 months and then annually up to age 5. Another blood sample is collected at age 8. Blood samples are collected from both parents, and breast milk is collected from the mother. Dust samples will be collected to evaluate exposures, using limulus assays to assess endotoxins and culturing microbes in the house dust. Children’s medical records will be reviewed to determine physician diagnoses, office visits, and number and type of illnesses. Dr. Wright briefly discussed some of the issues regarding the reliability of parental reporting compared with physician diagnoses and medical records.

The primary questions for the Infant Immune Study questionnaire focus on wheeze and cough, occurrence and frequency of symptoms, and doctor diagnosis. Dr. Wright said that one of the areas for improvement in questionnaires is the ability to differentiate daytime and nocturnal symptoms.

Question-and-answer session/comments. Dr. Martinez commented that during the first years of life, among the most rewarding markers for subsequent development of asthma-like symptoms are those obtained from stimulated peripheral blood mononuclear cells. This approach creates a very significant issue because these are very sophisticated techniques that will need to be performed centrally. Such analyses have specific requirements for cryopreservation, handling of cells, specimen storage, and so on. Capturing such important information in a large national study will require significant logistical effort. Dr. Sly described the methodological approach of a study in India in which whole blood is stimulated onsite; supernatant is then collected, shipped, and analyzed at a central location. George T. O’Connor, M.D., M.S., Boston University School of Medicine, discussed some of the methodological issues regarding the Inner City Asthma Consortium’s upcoming birth cohort study called Eureka (Urban Environment and Childhood Asthma). Investigators of this multi-site study decided to cryopreserve blood cells at the sites and have the stimulations preformed at a central location. Participants briefly discussed the utility of assessing gene expression from blood samples gathered for the Study.

Channing Laboratory Birth Cohort Studies. Dr. Weiss stated that the Channing Laboratory—a multidisciplinary research division of Brigham and Women’s Hospital and Harvard Medical School—has been conducting birth cohort studies for approximately 20 years. Dr. Weiss reviewed the following four birth cohort studies, which enrolled more than 5,000 mothers and infants in birth cohorts.

East Boston Study examined the effect of in utero cigarette exposure on lung function and respiratory illness in early life. This general population sample study enrolled 500 mother-and-infant pairs and examined pulmonary function in 250 of the infants at 2 weeks of life, six months and one year. At the time of the study, it was the largest study to examine pulmonary function in normal infants.

The Home Allergens and Asthma Study examined a birth cohort of 500 high-risk infants. The PI is Diane Gold. The study enrolled infants for which either parent had a history of allergy or asthma. The investigation collected home dust samples when the infants were 2 weeks old. Subsequent samples were collected if the family moved or at 2 years of age. The primary hypothesis of the study involved allergy and allergic sensitization. The main outcome variables were questionnaires and infants’ immune system response, using both lymphocyte proliferation and cytokines as well as serum markers of immune function such as total serum IgE.

The Project VIVA Diet and Asthma Study is an ongoing investigation of a birth cohort of 3,000 children. The Pis are Matt Gillman and Diane Gold. This cohort is another general population sample that was recruited from the Harvard Pilgrim HealthCare system. The study administered a well-validated dietary food frequency questionnaire for the mothers during pregnancy and a second questionnaire for the child after birth. Cord blood was collected for cytokine determinations and also for DNA. The primary outcome is questionnaire information but includes immune function parameters collected from 500 of the children.

The final study, Project Access, has a primary hypothesis to assess psychosocial stress in an inner city general population sample by questionnaire. The PI is Roz Wright. Outcomes for this study include asthma-related questionnaire data on symptoms such as wheezing and the effect of stress on immune system development. Project Access is collecting DNA samples. This study is ongoing.

Dr. Weiss said that the Channing Laboratory has the longest and most comprehensive birth cohort experience of any group in the United States. Given this experience, Dr. Weiss listed the following primary outcome measures for asthma data collection:

• Wheezing and Dr. Diagnosis of Asthma
• Nasal symptoms
• Atopic dermatitis.

Dr. Weiss commented that asthma-related studies using questionnaires could sufficiently limit the overall number of questions but that the questions would have to be asked repeatedly during the first 2 years of life to limit the potential of recall bias.

PIAMA Study. H. A. Smit, Ph.D., National Institute of Public Health and the Environment, Netherlands, presented an overview of the Prevention and Incidence of Asthma and Mite Allergy (PIAMA) Study. The study was originally conceived as an investigation of the effectiveness of mite-impermeable mattress covers. Before the study was actually designed and started, it was decided to incorporate an assessment of the natural history of asthma disease development. Thus, the study design included two intervention arms in high risk children (active and placebo matress covers), one natural history arm in high risk children ("control"), and one natural history arm in low-risk children. The main goals of the PIAMA Study were to assess the: 

• Effect of mite allergen avoidance on the incidence of childhood asthma (intervention arm) 
• Natural history of asthma in children from 0 to 8 years age (natural history arm) 
• Risk factors and markers for early detection of asthma (natural history arm).

The study design included:

• Birth cohort study in general population
• 4,146 children enrolled in 1996–1997
• Recruitment of pregnant women from prenatal clinics across the Netherlands
• Gestational age at recruitment: 3 months.

According to Dr. Smit, intervention measures in the PIAMA Study:

• Show a clear difference in (changes in) mite allergen exposure between "active" and "placebo" covers
• Seem to be more effective at high levels than at low levels of exposure
• No or little clinical effectiveness was observed possibly due to the low levels of mite exposure during the measurement period.

The results indicate that children with a parental history of allergy are actually less exposed to putative risk factors for the development of asthma. These children:

• Experience more lower respiratory tract infections with daycare use and older siblings than children without a parental history of allergy
• Are less frequently exposed to lifestyle-associated risk factors for asthma and allergy such as environmental tobacco smoke and cats.

Dr. Smit listed the following strengths of the PIAMA Study data collection:

• High- and low-risk children
• High compliance during follow-up
• Lifestyle and environmental assessment.

Dr. Smit noted that the availability of biological material at very early age and at preschool age was limited in the PIAMAStudy.

Data collection instruments in the PIAMA Study included:

• Self-administered questionnaires
• Dust sampling at homes
• Heel-prick blood (PKU)
• Physical examination:
  • Venous blood for IgE determination
  • Height and weight
  • Presence of eczema
  • Lung function including peak flow, exhaled NO, airway resistance (Rint)
• DNA from parents and children
• Pharmacy registered medication use
• GIS assessment of environmental exposures.

Dr. Smit described the strengths and limitations of the PIAMA questionnaire:

Topic	Strength	Limitation
Diet and physical activity	Included in questionnaire	Not validated
Environmental exposures: home characteristics, pets, combustion products, siblings, daycare, passive smoking	Includes passive smoking Validated Rest: used frequently
Circumstances of birth	Registered during check-up
Respiratory and eczema ISAAC	Validated for 7-12 years of age Standardized Worldwide application	Not validated at early age Longitudinality not validated "Too European?"
Respiratory, allergy, and eczema in parents and siblings ISAAC and European Community Respiratory Health Survey	Validated and standardized
Other: quality of life, health services use,  and so on

Dr. Smit considered what the optimal frequency of follow-up measurements would be, given optimal validity of the data:

• Too frequent follow-up (for example, every month) leads to many missing intermediate data points, inconsistent reports, and may negatively influence the response
• Too infrequent follow-up measurements (for example, once every 3 years) will produce information that can be considered to be rather retrospective
• In PIAMA:
  • Response on individual questionnaires 90-95 percent 
  • Complete data 1-5 years: 80 percent.

Dr. Smit presented some points worth considering in the planning of future birth cohort studies:

• Development of questionnaires: adapt and validate for young age
• Development of validated questionnaires on lifestyle (diet and physical activity) in the very young
• Careful consideration of frequency of follow-up measurements and longitudinality of wording of the questions 
• GIS assessment of environmental exposures
• Dust sampling by respondents and regular mail.

In reviewing the strengths and limitations of early clinical markers in the PIAMA Study, Dr. Smit said that early biomarkers of increased risk for prevention purposes were not available at very early age, only serum in 1-year-olds in the intervention study. Early clinical markers that were used at age 4 years for early detection and early treatment included:

• Interrupter resistance (marker of obstruction)
• Exhaled NO (marker of inflammation).

The general strengths of interrupter resistance and exhaled NO clinical markers are:

• Noninvasive
• Minimal cooperation required
• Relatively cheap.

The limitations of these clinical markers include:

• Validity
•  Practicality 
• MEC regulations.

Measurement of interrupter resistance in 4-year-old PIAMA children had the following practical limitations:

• Unsuccessful measurements due to lack of cooperation (fear, ignorance, reluctance):24 percent of the 4-year-olds could not complete the measurement successfully. Success rate could be increased by offering a second visit. 
• Exclusion by protocol: medication use 12 hours prior to measurement. 4 percent of participants had to be excluded because they forgot or did not wish to skip treatment before the measurement. Opportunity for second visit could prevent this.

Validity of interrupter as early clinical marker for wheeze (Brussee AJRCCM 2004):

• Presence of wheeze in 4-year-olds (cross-sectional) 
• Rint higher in persistent wheezers, but distributions show large overlap 
• Prospective predictive value for 8-year-olds.

Measurement of exhaled NO in 4-year-old PIAMA children had several practical limitations:

• Availability of NO analyser close to field work and technical problems: 25 percent of 1,245 participants were lost for this reason, leaving 935 children to be measured.
• Unsuccessful measurements: lack of cooperation (fear, ignorance, reluctance): leading to exclusion of 30 percent of 935 children with measurements, leaving 659 children.
•  Exclusion by protocol: another 230 children were excluded.
  • Medication use 12 hours prior to measurement
  • Ambient NO greater than 20 parts per billion
  • Reproducibility less than 10 parts per billion.

Thus, of the 935 children who were available for the NO measurements, only 46 percent of the measurements could be used for analysis. This can be increased by offering a second visit.

Dr. Smit described the validity of exhaled NO as an early clinical marker for inflammation:

• Cross-sectional in 4-year-olds: 8 parts per billion (7.6-8.5, 95 percent confidence interval)
  • Exhaled NO higher in children with elevated specific immunoglobulin E (IgE), but distributions show large overlap
  • No assocation with eosinophils
• Prospective predictive value for 8-year-olds is to be investigated.

Question-and-answer session/comments. Dr. Martinez commented that the ISAAC questionnaire is targeted for school-age children. If the ISAAC questionnaire is to be used for the Study, then it must be adapted to gather information on infants and young children. Dr. Belanger emphasized that because questionnaires are used to gather information over time, the same questions must be asked consistently over the data collection period. She noted that initially the parent is the respondent, but eventually the child becomes the respondent, and therefore, the questionnaire must be adapted to gathering information directly from the children themselves. Dr. Weiss commented on the evolution and standardization of "wheeze" questions in American questionnaires and suggested that using ISAAC may not be appropriate for the Study. Dr. Weiss said that the ISAAC questionnaire was developed as a more cross-sectional, survey-type questionnaire and was not designed to gather information on infants and young children. Thomas A. Platts-Mills, M.D., University of Virginia, discussed the reliablility of physical activity questionnaires. He noted that the information parents provide on the physical activity of their children in not well correlated with reality. Dr. Platts-Mills said that adults’ responses about their physical activity may, in fact, be inversely correlated with actual physical activity. Dr. London explained that other Study working groups are addressing issues of assessing physical activity and diet.

Dunedin Multidisciplinary Health and Development Research Study. Malcolm R. Sears, M.B., Ch.B., McMaster University, presented a brief overview of this study. The study had its basis in 1967-1972 as a neonatal study of all live births in Dunedin’s one maternity hospital. The last year of the neonatal study was April 1, 1972-March 31, 1973. A longitudinal follow-up of children born in that year who still resided in Otago province began in April 1, 1975. The follow-up study:

• Identified 1,139 children eligible by continuing residence in the province of Otago at 3 years of age
• Enrolled 1,037 (91 percent).

The cohort was a general population sample of non-high-risk children. Dr. Sears became involved with the study when the children were 7 years old. At age 9, he began to assess longitudinal pulmonary function and gather asthma disease development data through detailed questioning. Blood collections began at age 11, and skin testing began at age 13.The Dunedin longitudinal asthma study, as it was called, performed the following investigations on the cohort from recruitment until age 26 (subject are now being reassessed at 32 years of age):

• Neonatal/perinatal data
• Asthma diagnosis, all ages
• Symptom questionnaire, all ages
• Smoking questionnaire, from age 15
• Occupational history, from age 15
• Family history, at ages 7 and 18
• Spirometry, from age 9
• Methacholine challenge, from age 9 to 21
• Bronchodilator challenge, at ages 18 and 26
• Serum IgE, at ages 11 and 21
• Skin allergy tests, at ages 13 and 21
• Genetics, at age 26
•  Lung volume and Dlco, at age 26.

Dr. Sears explained that the Dunedin study was missing a great deal of information on asthma and allergy during the early lives of the subjects. For example, there were no data on measured environmental exposures and no blood collected before age 11. Although the study’s original design was not to study asthma, the information that was gathered has proved valuable. Dr. Sears listed what has been learned from the Dunedin asthma study (up to age 26):

• Prevalence rates of wheezing and asthma at age 7 
• Prevalence and characteristics of asthma at age 9 
• Prevalence of airway hyperreactivity (AHR) to methacholine including finding AHR in fully asymptomatic 9-year-olds 
• Family, social background, development, and behavioral characteristics of children with asthma 
• Increased risk of development of asthma in children sensitized to house dust mite (fivefold increased risk) and to cat dander (fourfold increased risk) 
• Serum IgE levels predict AHR, even in children without asthma or skin test evidence of atopy 
• Deleterious effects of passive smoking on longitudinal trends in pulmonary function 
• Effects of AHR, wheezing, and atopy on longitudinal trends in pulmonary function 
• Relationships among serum IgE, AHR, lung function, symptoms, and asthma diagnosis 
• Interrelationships among skin test positivity and diagnoses of asthma and hay fever, effect of gender, and relationship of baseline pulmonary function to airway hyperresponsiveness 
• Impact of parental and neonatal risk factors on atopy, AHR, and asthma 
• Relationships among allergy and other childhood neurological and psychological development 
• Role of Chlamydia in the epidemiology of asthma 
• Relationship between socioeconomic status in childhood and adulthood and health outcomes 
• Respiratory effects of cannabis addiction 
• Risk factors for hospitalizations for asthma 
• Predictive relevance of asymptomatic AHR in childhood on the likelihood of developing asthma 
• Long-term outcome of breast-feeding, increasing the risk of atopy and asthma in mid and later childhood 
• Occurrence of airway remodeling in childhood asthma, as defined by persistently abnormal postbronchodilator airway function in young adults 
• Natural history of childhood asthma and risk factors for persistence, remission, and relapse of asthma 
• Impact of cigarette smoking on alveolar function, suggesting early development of emphysema.

Dr. Sears shared some of the reasons why the Dunedin study has been successful:

• The study participants have assumed considerable "ownership" of the study; they strongly believe in the importance of the science and the importance of their continuing participation. 
• The study data is strictly confidential. Such confidentiality allows for highly accurate data. 
• Participants’ ownership and strict confidentiality has fostered a high rate of compliance.

In discussing important issues to address when gathering data for an asthma-related study, Dr. Sears listed the following:

• Exposure data, including breast-feeding and animal exposures, with objective measurements where possible 
• Family data, family history, parental smoking habits 
• Self-reported severity of symptoms (for example, wheezing) 
• Objective physical data such as airway responsiveness.

Australian Birth Cohort Studies. Dr. Sly described three studies and some of the things that he has learned from them. The first study was a general population study of about 3,000 pregnant women, who were recruited antenatally. The hypothesis of this study was that intensive field monitoring improves pregnancy outcomes (which it did not, Dr. Sly reported). The study continues to monitor this cohort. The children have been seen or contacted at ages 1, 2, 3, 6, 8, and 10. The study is currently gathering data on the cohort as 13-year-olds. Of the original 3,000 pregnancies, there were 2,860 live births, and approximately 2,000 of these children remain in the study. Dr. Sly said that several key, core instruments have been used throughout the study, including age-appropriate psychosocial profiles and assessment of family functioning. Formal asthma assessments were performed at age 6 and are now being performed on the 13-year-olds. Although the study is primarily questionnaire-based, physical examinations have been conducted at ages 1, 6, 8, 10, and 13 for different purposes. For example, at age 8, the exam focused on cardiovascular functioning, whereas at age 10, the exam focused on psychosocial and educational outcomes. The current assessment of 13-year-olds includes physical activity, obesity, and education.

The second study was of a high-risk cohort of 250 children recruited on the basis of parental family history. The primary purpose of this study was to examine incidence of respiratory infection during the first 5 years of life. Data collection included mucus samples, daily diaries, and blood (cord blood and samples at 6 months and then annually). The study just completed its 5-year assessment, which included pulmonary function tests.

Dr. Sly explained that the third study is actually a series of birth cohort studies in India. Approximately 2,500 children are to be enrolled in this study, which is using a sequential antenatal recruitment strategy. A stratified approach will focus primarily on ranges of exposures, including such things as socioeconomic status, allergic/infective exposures, nutrition/diet, and environmental exposures (for example, outdoor, indoor, air, water, pesticides, metals). The main study outcomes are asthma and immune system function but also growth, neurodevelopment, and major childhood illnesses. GIS exposure maps will be used to assess environmental exposures.

Dr. Sly noted that the most useful data to collect in asthma-related studies is information on symptoms such as wheeze, wheeze frequency, and severity of wheezing. Asthma should be defined by very strict criteria. For example, at the age of 6 years, a child should have doctor diagnosis, wheeze in the last 12 months, and currently be using asthma medication. However, separate analyses have indicated that wheeze frequency is probably the most important symptom for data collection. Other reliable data include nocturnal symptoms and severity of asthma attack. Study investigators are currently developing a standardize severity scoring system that includes medication use as well as symptoms. In Dr. Sly’s opinion, the variable that is least useful on questionnaire and other assessments is allergic rhinitis. He elaborated that although this symptom is important to monitor, it is probably the most difficult data to capture reliably from questionnaires.

In regard to key clinical measures/markers, Dr. Sly said that blood is an important specimen to collect for asthma-related studies. Key collection times are at birth (cord blood), 1 year, 2 years, 5 years, prepubertal, and 15 years. The specimens would be used to assess not only asthma and immune function but also cardiovascular status and obesity.

Question-and-answer session/comments. Dr. Martinez asked whether any of studies performed any lung function testing early in life. Dr. Sly replied that the large 2,500-subject study examined lung function at 6 years of age. He commented that standard spirometry is not the appropriate measure for such cohorts. Dr. Sly explained that the other two studies used spirometry and forced oscillation in 5-year-olds. Forced oscillation can be used from about 2-3 years of age. Neither spirometry nor forced oscillation has been attempted with infants. The forced oscillations were performed with bronchodilator response. Dr. Platts-Mills asked how well questionnaires predict either acute asthma attack or chronic severity of symptoms. In response, Dr. Sly noted that questionnaires predict chronic severity reasonably well but do not predict the likelihood of acute attack, hospitalizations, or severe exacerbations. These events can only be recorded after the fact. Dr. Sears commented that questionnaires are subject to patient interpretation of wording, reporting issues, and unreliable descriptions (for example, wheezing episodes versus severe attacks). Questionnaires can reliably capture the frequency of events or symptoms but cannot necessarily capture severity. Frequency data can then be merged with pulmonary function and airway responsiveness measurements. Dr. Sears said that symptoms can be broadly categorized into groups such as "bad," "moderate," and "mild." Dr. Ortega asked how the Australian studies’ strict definition for asthma compared with those from other studies. Dr. Sly explained that the strict definition was applied to 6-year-olds to circumvent some the problems to which investigators alluded in this discussion; that is, the problems inherent with questionnaire data in a longitudinal studies. Dr. Martinez mentioned that there are two types of outcomes that are correlated but not necessarily equivalent: catastrophic attacks and severe chronic symptoms.

Detroit Birth Cohort Studies. Christine Cole Johnson, Ph.D., M.P.H., Henry Ford Health System, reported on two studies: the Childhood Allergy Study (CAS) and the ongoing Wayne Tri-County Health, Environment, Allergy, and Asthma Longitudinal Study (WHEALS).

Dr. Johnson summarized CAS as follows:

• Suburban Caucasian children ( n = 835) 
• Born 1987-1989 
• Broad range of middle class families 
• Health maintenance organization (HMO) members 
• Recruited mothers in second trimester while visiting obstetric clinics 
• Obtained cord blood samples 
• Performed annual and semi-annual surveys at 1, 2, 3, 4, 4½, 5, 5½, and 6 years 
• Home visits conducted at 2 and 4 years 
• Collected dust, air, blood, and urine samples 
• Clinical examination performed at age 6, including skin-prick testing, total and allergen-specific IgE, bronchial hyperactivity, methacholine challenge, pulmonary function tests, and urine (for cotinines).

Dr. Johnson summarized WHEALS as follows:

• Urban and suburban children, born 2003 and later 
• African Americans and Caucasians 
• Target enrollment of 2,600; about 100 enrolled to date 
• All levels of socioeconomic status 
• Multiple insurers, including Medicaid and HMOs 
• Recruiting mothers in second trimester while visiting obstetric clinics 
• Obtaining maternal prenatal and postnatal blood samples 
• Obtaining cord blood samples 
• Obtaining paternal blood samples 
• Storing DNA and mRNA 
• Surveys at 1, 3, 6, 12, 18, and 24 months 
• Home visits at 1 and 6 months, including dust (from five within-home sites), breast milk, stool samples, and blood (6 months and 2 years) 
• Measuring allergens (cockroach, dog, cat, dust mite) and endotoxins in dust samples 
• Measuring IL4, IFNγ, IL10 in stimulated T cells in all samples as well as total and specific IgEs. 

Investigators considered the following issues regarding the new questionnaire for WHEALS:

• Including standardized questions (for example, ISAAC and ECRHS) for comparability 
• Gathering more detailed information on allergy and asthma history for subjects, parents, and siblings, including information on onset of symptoms and time of doctor diagnoses 
• Differentiating hay fever from food and drug allergies 
• Directly asking each parent about their personal history, including age at onset of symptoms and remission patterns 
• Including more data on mothers’ and children’s diets, breast-feeding, pet-keeping (size of animal, behavior in home) 
• Adding more detail on smoking history (many parents changed behaviors throughout CAS study) 
• Collecting information on pregnancy history, pregnancy and delivery characteristics, socioeconomic status, antibiotic use (mother and child), bacterial and viral illnesses, fever, farm and livestock exposure.

WHEALS investigators considered the following data collection issues:

• Larger and more diverse population 
• Consideration of recruitment versus retention (for example, age and mobility of mothers) 
• Need for multiple early visits 
• Multiple environmental samples and measures (for example, endotoxins, allergens) 
• Timing of collection of dust samples 
• Household characteristics and assessments 
• "Objective" measurement from parents and siblings (for example, IgE).

WHEALS investigators considered the following sample collection issues:

• Informed consent-a complex process; use of hierarchical consents 
• Permission for storage of DNA (subject, parents, siblings) 
• Use of EMLA ^TM cream (2.5 percent lidocaine, 2.5 percent prilocaine) 
• Incentives 
• Blood volume from baby 
• Laboratory staffed 16 hours a day, 7 days a week, to obtain and process cord blood samples 
• Stem cell storage companies (prevents study from obtaining cord blood) 
• Sample tracking (infrastructure).

To date, about eight families have dropped out of the study. Despite the good rapport that investigators had with these families, Dr. Johnson reported that the families completed the weekly symptom diaries for about 6 months but then simply stopped filling out the forms ("basically they got sick of it") because of the repetitive nature of the questionnaire. Dr. Johnson noted that WHEALS study investigators have made a concerted effort to use neutral wording for the probes and questionnaire, specifically in regard to finding out whether and why parents acquired or got rid of animals so we can assess that bias. In terms of asthma-specific data collection, the questionnaire inquired about doctor diagnoses, recency and frequency of symptoms (wheeze), and use of medications. In addition, the study has the ability to correlate medical records with mothers’ reports; however, Dr. Johnson said that investigators have not yet determined the strength of this correlation.

Question-and-answer session/comments. Dr. Martinez discussed some of the issues and concerns for a tracking system for samples/specimens, particularly for a large national cohort such as the Study. Dr. Sly asked about the number of home visits for WHEALS and emphasized their important role in data collection. Dr. Johnson agreed that home visits add an element of complexity to longitudinal studies but that information on home environments and exposures are invaluable. Home visits may help develop rapport with study subjects and their families. Participants briefly discussed issues about safety during home visits in high-crime areas and about biological versus nonbiological parents.

Dr. Platts-Mills commented that no birth cohort study has ever correctly predicted its most important outcome. For this reason, collecting objective data is very important. Dr. Platts-Mills cited the tolerance effect due to exposure to cats and dogs that was unexpectedly revealed in the Detroit studies. He commented that researchers do not have a good understanding of the interrelationship between exposures to cats and dogs and subsequent immune responses in human subjects. The immune responses to cat and dog allergens can only be detected in serum samples, and air-borne endotoxins from cats and dogs may be significantly different.

WHEALS Samples. Mary J. Maliarik, Ph.D., Henry Ford Health System, reviewed aspects of sample collection and sample processing in WHEALS. WHEALS investigators considered the following sample processing issues:

• Scientific questions: Determines samples collected and processes used 
• Sample logistics 
• Sample stability
  • Need for timely delivery to laboratory
  • How to process
• Numbers: Generally large 
• Cost
  • Reagents
  •  Personnel requirements.

WHEALS samples and sampling processing were characterized as follows:

Sample type	Process	Analysis
Blood-heparin	Lymphocyte stimulation	Intracellular and secreted cytokines
	Plasma	Allergens, cytokines
	RNA isolation	Gene expression
Blood-EDTA	DNA	Gene polymorphisms
Dust	Vacuum method	Endotoxins, allergens
Breast milk	Whole, aqueous, lipid fractions	Immune mediators
Stool	Frozen	Bacteria, virus

Dr. Maliarik presented other details on processing samples such as peripheral blood lymphocytes, plasma, RNA, DNA, breast milk, stool, and dust. She described blood processing as follows:

• Heparin anticoagulant
  • Cell count
  • Cell stimulation for intra- and extracellular cytokine measurement
  • Plasma for IgE, other
  • RNA
•  EDTA anticoagulant
  •  DNA isolation 
  • Plasma for IgE, if needed 
  • Can use banked blood.

Dr. Maliarik provided details on the algorithm for blood processing for cytokines and on whole blood intracellular cytokine staining (Beckton Dickinson). She commented on some issues of cryopreservation, hierarchical sampling, cell preservation, and future/evolving technologies.

Question-and-answer session/comments. Dr. Martinez briefly discussed the importance of both soluble and insoluble proteins; he said that preserving cells to analyze insoluble proteins might become significant. Dr. Martinez cited several ongoing NHLBI projects on protein markers of heart, lung, and blood diseases. Dr. Sly mentioned the importance of storing/preserving red blood cells. Other significant biomarkers include vaccine allergens, tetanus toxin, and lower level mitogenic stimulations such as a low-level PHA. Small blood volumes from infants and small children are a concern with sample processing.

ATS-DLD Respiratory Disease Questionnaire. George T. O’Connor, M.D., M.S., Boston University School of Medicine, reviewed the proceedings of an October 2001 workshop on this questionnaire. Specifically the goal of the workshop was to revise and expand the ATS-DLD-78 Respiratory Disease Questionnaire (RDQ) to meet the needs of U.S. investigators of chronic respiratory diseases. The American Thoracic Society’s Division of Lung Diseases developed the RDQ. Dr. O’Connor explained that the questionnaire was designed as a disease prevalence survey-type questionnaire. It was not designed as a longitudinal questionnaire, it was not designed to be administered every year, and it was not designed to assess incident outcomes in a cohort study. However, the questionnaire has been adapted extensively for birth cohort studies.

Dr. O’Connor briefly described the early history of the RDQ:

• 1960: Began as British Medical Research Council (MRC) questionnaire focusing on "cough and phlegm" 
• 1968: ATS adopted MRC questionnaire 
• 1971: NHLBI published MRC questionnaire modified for use in the United States 
• 1978: Epidemiology Standardization Project
  • Benjamin Ferris, M.D., principal investigator 
  • ATS-DLD-78-A and ATS-DLD-78-C 
  • Greatly influenced by 1960 MRC questionnaire.

The RDQ’s later history included:

• Explosion of interest in asthma, bronchial reactivity, respiratory allergy, sleep apnea, and so on 
• Explosion of new hypotheses regarding environmental risk factors, genetic risk factors, and so on 
• ISAAC, European Community Respiratory Health Survey, U.S. studies of respiratory disease, and so on

According to Dr. O’Connor, the limitations of the RDQ in 2001 were:

• Little information collected on symptoms of AHR 
• Little information collected on allergic or other triggers of episodic symptoms 
• No items on sleep-disordered breathing 
• Inconsistent recall period 
• Limited environmental data (for example, environmental tobacco smoke for children but not for adults).

The specific goals for the October 2001 workshop were to:

• Develop a draft adult core RDQ 
• Incorporate recommendations of the Occupational Working Group into the adult core RDQ 
• Develop a draft pediatric core RDQ 
• For each of these:
  • Document rationale for each item 
  • Document what has been previously demonstrated concerning the validity and reliability of each item 
  • Develop guidelines for evaluating the validity and reliability of the core RDQ
• Establish a process and timeline for obtaining external review and finalizing core RDQs.

The ATS-DLD-78-A focused on:

• Demographic data 
• Symptoms 
• Medical history (diagnoses) 
• Occupational history 
• Smoking history 
• Family history 
• Instructions (including question clarifications) 
• Supplemental questions 
• Rationale and justification of components.

Details of the ATS-DLD-78-A data included:

• Symptoms
  • Cough 
  • Phlegm 
  • Episodes of coughing and phlegm 
  • Wheezing 
  • Breathlessness 
  • Chest colds and chest illnesses
• Medical history (diagnoses)
  • Lung trouble before age 16 
  • Attacks of bronchitis 
  • Pneumonia 
  • Hay fever 
  • Chronic bronchitis 
  • Emphysema 
  • Asthma 
  • Other chest illnesses 
  • Heart trouble 
  • High blood pressure
• Family history
  • Mother and father 
  • Vital status 
  • Age if living or age at death 
  • Cause of death 
  • Chronic bronchitis 
  • Emphysema 
  • Asthma 
  • Lung cancer 
  • Other chest illnesses.

Details of the ATS-DLD-78-C data included:

• Demographic data 
• Residential history 
• School/daycare; siblings; bedroom sharing 
• Home characteristics
• Symptoms 
• Medical history (diagnoses) 
• Occupational history 
• Parental smoking history 
• Family history 
• Instructions (including question clarifications) 
• Rationale and justification of components.

The guiding principles for revising the adult and pediatric RDQs were:

• ATS-DLD-78 was very well crafted. Components that still work well and appear valuable should be retained. 
• Components of ATS-DLD-78 that do not appear valuable for current research should revised or discarded. 
• What will the RDQ be used for?
  •  Minimal set of descriptive data to describe occurrence of respiratory disease in a population 
  • Ascertain health outcomes in epidemiologic studies 
  • Identify informative subgroups.

Dr. O’Connor presented the current wheezing sequence in the ATS-DLD-78-C and then discussed the following proposed 2004 revisions to this sequence (the current draft revision):

• Has this child ever had wheezing or whistling in the chest?
  • Younger than 2 years of age 
  • 2-5 years of age 
  • Older than 5 years
• Has this child ever had episodes of wheezing lasting a day or more?
  • Two or more such episodes 
  • Age at first episode
  • Required treatment
• In last 12 months, has child had wheezing in chest?
  • How often 
  • In absence of cold or flu 
  • Accompanied by shortness of breath 
  • After exercise.

Question-and-answer session/comments. Dr. Belanger inquired whether the questionnaire asks about wheezing in the absence of cold or flu. She noted that because some questionnaires ask simultaneously about all respiratory symptoms, parental responses are often categorized as something such as the "continuous cold." Parents can observe and report on symptoms, but it is difficult for them to make a judgment on whether the symptoms are specifically related to colds, flu, allergies, or other specific conditions. In an effort to overcome recall bias, Dr. Weiss suggested that questionnaires should be administered more frequently early in life, for example, every month during the child’s first year. Questionnaires need to collect reliable information on both severity and frequency of symptoms, and need to provide insight on persistent or chronic symptoms and whether symptoms ever return to normal.

Poole Birth Cohort Study. Dr. Platts-Mills reviewed his experience with biomarkers used in this 1978 birth cohort study of approximately 80 subjects. He said that the "essence" to start the study was derived from the development of an assay to measure mites in houses. The study was designed to look at the role of viruses in asthma but completely failed in this regard. However, the main focus of Dr. Platts-Mills’s presentation was serum biomarkers.

"Noninvasive" markers of immune response used in the Poole Birth Cohort Study included:

• Skin tests
  • Prick tests using single puncture 
  • Intradermal skin tests—48 hours
•  Blood tests
  • Serum antibodies
    • Total IgE, specific IgE 
    • IgG
  • Lymphocyte responses in vitro (volume critical)
    • Nonspecific 
    • Specific stimuli
      • Crude extract 
      • Purified or recombinant 
      • Peptides
  • Complete blood count differentials for eosinophil counts
• Exhaled breath
  • eNO in the gas 
  • Condensate pH, leukotrienes, cytokines, and so on
• Other options include secretions for antibodies, sputum, induced sputum, secretions for viral culture or PCR.

The study was attempting to answer the following questions:

• What is the time course of immune responses, and what are their relationships to lung symptoms and disease, including details IgG/IgG ₄ antibody response in relation to IgE Ab response and symptoms.
• Is there an interaction between different responses?
  • Does "tolerance" to food allergens influence subsequent responses to inhalants? 
  • Is oral tolerance to inhalants possible or relevant?
• What factors influence persistence, progression, and severity of asthma?
  • Different allergens such as Alternaria, mite, cockroach 
  • Titers of IgE ab → total IgE
  • Lifestyle issues such as obesity, activity, and so on 

Dr. Platts-Mills briefly elaborated on the interaction between different responses, noting that researchers are finding different answers in different parts of the world. He noted that studies have revealed that exposure to certain allergens are risk factors for different outcomes. Storage of serum allows investigators to retroactively assess the relationships among risk factors and outcomes. The important issue, however, is understanding what is really responsible for the progression of disease severity. To this end, Dr. Platts-Mills emphasized the importance of measuring a range of airborne exposures, not just dust. High exposures to certain allergens (for example, cat allergen) can actually sensitize children’s immune responses to subsequent airborne exposures. Dr. Platts-Mills reviewed relevant data from a number of related studies, including elevated levels of total serum IgE. He explained that different allergens (for example, dust mite, Alternaria, cockroach) might have different roles in total serum IgE. Therefore, the Study should consider airborne sampling in addition to dust sampling. Dr. Svendsen agreed that the data indicates a strong correlation between serological markers and exposures to airborne allergens. 

Dr. Platts-Mills concluded by summarizing the relevance of total serum IgE measurements:

• Total serum IgE alone is not a good method of excluding or confirming allergic disease in an individual case. 
• Total IgE correlates very strongly with asthma. 
• Among children and young adults admitted to hospitals, mean total IgE is approximately 300-400 international units per milliliter. 
• Does specific IgE antibody explain differences in total IgE?

Question-and-answer session/comments. Dr. Weiss asked whether elevated levels of total IgEs have a predictive value, that is, whether IgE levels are relevant to exacerbations, hospitalizations, or emergency room visits. Dr. Platts-Mills replied that this biomarker has not been shown to be predictive of hospitalizations. Dr. Weiss then asked whether the effect of inhaled steroids is sufficient to decrease IgE levels? Is it an IgE-mediated mechanism? Do high-dose inhaled steroids decrease hospitalizations? Dr. Platts-Mills answered that, to his knowledge, there is no evidence indicating that the effects of inhaled steroids are related to IgE. Dr. Weiss surmised that data at the genetic level might be necessary to predict exacerbations.

Measurement of Lung Function in Preschool Children. Dr. Sly gave a short presentation on measuring lung function in young children. He began by providing background information on some techniques and standardized measures of lung function, including:

• Forced oscillation 
• Interrupter resistance (Rint) 
• Spirometry 
• Plethsymography
• Gas dilution.

Dr. Sly presented details on his experience using a portable forced oscillation technique (FOT) to measure lung function in preschool children. These details were derived from a preliminary study of approximately 70 children, 4-7 years of age. In this study, lung function was correlated with the children’s height and weight. Dr. Sly reported that FOT requires gentle tidal breathing, but that the children can be easily trained to breath into the apparatus. FOT is applicable to children as young as 3 years of age. Application to children 2 years of age and younger may require sedation. The apparatus is portable, and lung function measurement can be completed in 30-60 minutes.

Breakout Group A: Data Gaps in Questionnaires
Moderator: Dr. Svendsen
Discussion leader: Dr. O’Connor
Participants: Dr. Belanger; Terence Dwyer, M.D., M.P.H., NICHD, NIH, DHHS; Dr. Johnson; Ms. Keim; Andrew Liu, M.D., National Jewish Medical and Research Center; Jeanne Moorman, M.S., CDC, DHHS; Dr. Neas; Mark Raizenne, Ph.D., Canadian Institutes of Health; Dr. Smit; and Dr. Wright

The original charge for this breakout group was to:

• Draft a list of recommended questions drawn from each questionnaire presented during the workshop, focusing primarily on those questions that worked well in previous studies, improvements on current questions, and newly recommended questions.
• Compile and prioritize a final list of recommended questions and indicate timing and method of collection, flagging any questions that may be sensitive or of special burden to the respondent.

Dr. Svendsen asked the group to first discuss and develop a preliminary list of questions focused on early life. There was agreement that data collection tools should address the various important developmental stages of early life. A series of questionnaires should attempt to build on information from a "first contact" enrollment questionnaire or interview. Parental histories will identify prenatal exposures and predisposing factors. Because no other Study working group will focus on parents’ or children’s asthma symptoms or related prenatal exposures, parental determinants of child outcomes were deemed important.

Dr. Dwyer said that Study design decisions will be based on recommendations from the workshop and that this was an opportunity for members to influence Study design. Participants were advised to cover the time before birth until a few years following birth. The group considered points in time when it would be important to make new assessments.

Maternal lifetime history and symptoms during pregnancy are important and should entail:

• A first trimester questionnaire—biological samples should attempt to support information in surveys 
• A chronologic approach to pregnancy—aspects of the third trimester will be different than previous periods 
• Questions about maternal activity to determine whether asthma affected lifestyle during pregnancy 
• Maternal weight measures—fetus growth and maternal weight gain may significantly constrict lung movement during later phases of pregnancy and may be more indicative of shortness of breath during the third trimester than asthma.

Discussions on format covered both interviews and questionnaires.

• Interviews
  • Establish personal rapport with subjects 
  • Require high cost for personnel 
  • Present safety issues when done in homes (for example, two people should be sent to in-home interviews in high-crime areas) 
  • Can be conducted during collection of blood and informed consent could be obtained at a personal interview.
• Questionnaires
  • Are cost-effective 
  • Present literacy issues.

Group members viewed maternal symptoms as determinants for child outcomes. Especially during pregnancy, questions should determine severity of maternal asthma, but subject burden and time required for information retrieval should be limited as much as possible. Information on hospitalization, emergency visits, and medications for breathing difficulties during pregnancy will be an essential factor in determining severity. Participants agreed that early pregnancy—preferably first trimester—is the best time to get maternal history and a baseline of symptoms during pregnancy to assess changes until birth.

Drs. Belanger and Wright advised that asthma is underdiagnosed in women—especially those from lower socioeconomic backgrounds. Therefore, diagnosis should not be considered the only significant indicator of asthma. Based on this premise, the group decided that the focus of questionnaires should be on symptoms. Sex differences regarding symptoms were also discussed.

Dr. Svendsen asked the group to discuss factors that should be considered during the first trimester. Because the Study’s focus is on child outcomes and not the mother, questions for a first trimester interview should be appropriately limited. Various asthma indicators were discussed. Because medications including cough medicines and OTC medications for allergies may hide symptoms, questions should be able to assess for allergic-onset asthma. Wording should be precise enough to identify symptoms yet maintain awareness of cultural differences in descriptions of respiratory symptoms. The group was told to assume that only English- and Spanish-speaking subjects would participate. The ATS-DLD-78-C questionnaire is a popular assessment tool that covers various aspects of asthma. This tool has been validated, but there have been some problems in assessing bronchitis-related symptoms. It lacks questions to address severity. Participants decided to select a subset of questions from the ATS-DLD and implement modifications to elicit detailed information for Study purposes. A response-rating system should be developed to determine degree of severity.

Group members agreed that paternal factors of childhood risk for asthma should be identified and that an interview or questionnaire should take place any time before birth. Information should be collected directly from fathers, if possible, unless there are restrictions. Parents will have less time following the birth, and surveys should be done before, or as close to, birth as possible. To reduce subject burden, group members discussed a "first contact" questionnaire containing general questions to identify a subpopulation with potential for asthma diagnosis, after which, a second questionnaire could be sent that would identify specifics of the condition. After discussion, however, the group decided that a better option was a questionnaire format that would direct responders giving "no" responses to skip asthma-specific questions. This would reduce cost for postage, as well as effort for subjects without asthma.

Group members discussed details that should be contained in the maternal surveys versus the more general parental surveys. They assumed that general medical histories and physical examinations for the Study would be done separately and that details of asthma-related information should be the focus of these deliberations. Because child outcomes are seen in terms of parental determinants of risk for asthma, parental history questionnaires should focus on four basic categories: asthma, skin ailments, allergies, and other serious respiratory ailments. Asthma-related questions should cover:

• Wheeze 
• Persistent cough 
• Nocturnal cough 
• Tightness in chest 
• Timing of symptoms, seasonal factors 
• Frequency of symptoms 
• Severity of symptoms 
• Medications, including episodic or routine, as well as prescription and OTC allergy and cough medicines 
• Age at onset of initial symptoms 
• Timing of last occurrence 
• Whether professional diagnosis of asthma made 
• Hospitalizations or emergency care.

Skin ailments, including eczema, should include:

• Whether professional diagnosis made 
• Treatments 
• Frequency 
• Timing of symptoms.

Allergies, rhinitis, and related symptoms should include:

• Seasonal factors 
• Medications 
• Pets in the home 
• Runny nose and watery eyes 
• Pneumonia, bronchitis, and other respiratory disorders.

Unless diagnosed by a physician, parents’ speculations about potential sources of allergies have been shown to be highly inaccurate. The group thought it was best to simply establish that allergies exist rather than have parents speculate about potential cause. The group emphasized that food allergies should not be included.

A second maternal questionnaire should reflect more detailed information on asthma-related symptoms experienced during the entire pregnancy. To optimize recall, timing of the second survey should be as soon after the birth as possible and should be in interview format. This interview should elicit data on:

• Frequency of symptoms 
• Severity of symptoms 
• Fevers and any serious illnesses 
• Perinatal distress 
• Birth stress.

Breast milk samples should be obtained, if possible.

Dr. Svendsen advised the group that other Study working groups would focus on pharmaceuticals and that there is a planned Study contact at birth to collect clinical information.

Group members agreed that during early childhood questionnaires should focus on symptoms that parents can easily observe. One participant remarked that new mothers have more difficulty identifying health problems in infants; thus, symptom descriptions should use easy-to-understand terms but should also be precise and sensitive to cultural differences. The group agreed that formatting should be essentially consistent for the series of questionnaires for ease of use. Different questions should be added at various stages of child development.

The group’s discussion focused on respiratory-related data collection. Important data collection times following birth are:

• 4-6 weeks
  • Home visit is preferred because most mothers will still be at home 
  • New mothers will be better able to recognize problems at 6 weeks
• 3-4 months
  • Good time to assess feeding 
  • Cytokine assessment
• 6 months 
• 1 year 
• Annual assessments in subsequent years.

At 4-6 weeks of age, because parents’ interpretations are often inaccurate, questionnaires should focus on symptoms that can be easily observed. Group members developed a preliminary list and discussed items that questionnaires should address, including:

• Wheeze
  •  Frequency 
  • Severity
• Cough
  • Colds—questions should address cough with and without colds or flu 
  • Persistence—this factor is intended to differentiate asthmatic or allergy-related coughs from cough due to colds 
  • Nocturnal—sleeping patterns are often irregular during early months, so questions should specifically address respiratory symptoms as opposed to just sleep patterns 
  • Seasonal changes—questions should also address changes that come with development.
• Eczema and other skin ailments
  • Questionnaire should emphasize that diaper rash is not part of this assessment 
  • Hives are usually allergy-related and should not be included
• Ear infections and rhinitis 
• Bronchial infections 
• Fevers 
• Croup 
• Pneumonia 
• Vomiting 
• Unscheduled clinical visits
  •  Assumes "well baby" visits 
  • Includes urgent care and hospitalization
• Antibiotics, acetaminophen, and OTC cough medicines
  •  Recall of medications is enhanced in the context of discussion on specific illnesses 
  •  Another Study Working Group will track all other medication data.
• Any respiratory concerns or questions.

At 3-4 months and at 6 months, questionnaires should include additional inquiries about diarrhea and the child’s overall health. Recent research suggests microbes in the intestinal tract may set the stage for asthma. Inexperienced mothers may have difficulty discerning abnormality in breast-feeding infants who may have softer stools.

At 1 year, questionnaires address both symptoms and diagnoses. The format should be static after this time and should include questions on:

• Physician-diagnosed asthma 
• Reduced physical activity due to respiratory symptoms
  • Number of days breathing problems interfered with play to assess severity 
  • May be problem to define "normal activity" in children prior to age 3
• Medications prescribed and OTC medicines for respiratory symptoms 
• Hay fever
• Allergies
  • Food allergies not included because too many diagnoses have similar symptoms 
  • Serology most useful
• Watery eyes and runny, itchy, or stuffy nose 
• Urgent care or hospital visits for respiratory symptoms.

Because socioeconomic factors can limit access to frequent medical care, unscheduled visits, urgent care, and hospital visits will be more important data points. Dr. Wright advised against relying too heavily on physician records for young children. Children do not typically stay with the same physician in a pediatric group, sometimes insurance coverage is lost, or parents move and records do not follow the child. Hospitalization records are likely to provide more complete information.

Because of potential impact on health care, group members recommended that socioeconomic factors such as income, unemployment, and insurance coverage be assessed at several times during the subjects’ years of participation. Dr. Svendsen stated that other groups involved in the Study would assess those factors.

The group was asked to refine its questionnaire lists. Group members revisited various items, discussed decisions in more detail, and confirmed consensus. They readdressed the advantages and disadvantages of using questions from various assessment tools including ISAAC, SCORAD, and the Wilson questionnaire. A modified version of the ATS-DLD core questionnaire was endorsed for data collection on asthma.

To keep subject burden to a minimum, participants were asked to prioritize questions, eliminate some items, and make final recommendations on details of questionnaires. General assumptions affecting the group’s decisions are as follows:

• Although enrollment is desired in the first trimester of pregnancy, group members believed it was best to assume that many participants would be enrolled at later times during pregnancy. 
• A Study contact is already planned at the time of birth, and if needed, information from that contact can be used in the asthma component.
• Another workshop will address issues about data collection on use of medications.

The group recommended three parental-specific questionnaires:

• Maternal history questionnaire 
• Paternal history questionnaire (same as maternal history questionnaire) 
• Postpartum questionnaire.

The maternal history questionnaire should focus on questions about lifelong symptoms and diagnosis. Symptom questions should be sensitive enough to identify undiagnosed women. The questionnaire’s wording should reflect considerations of cultural differences. Questions on phlegm and congestion should be avoided because interpretation of these words is too varied. For ease of use, questionnaires should be formatted so that "yes" responses to general questions direct subjects to more detailed questions placed inside text boxes.

Postpartum questionnaires should be given as soon as possible after the birth to reduce recall error. They should attempt to determine whether asthma was worse during pregnancy through detailed inquiry on severity and frequency of breathing problems. This determination is important because many women avoid medications during pregnancy. Questions about the pregnancy should cover:

• Wheezing and breathing problems 
• Sinusitis 
• Rhinitis 
• Asthma attacks 
• Urgent care and hospitalization 
• Treatment, especially steroidal 
• Pneumonia, bronchitis, and other respiratory ailments.

Questions on congestion, runny nose, and watery eyes were considered lower priority. It is assumed that these symptoms co-occur with many other diagnoses, such as sinusitis, and would be a potential confound. Child data should be collected at the following times:

• 4-6 weeks—home interview 
• 3-4 months (questionnaires from this point forward) 
• 6 months 
• 1 year and annually thereafter.

The group discussed how to address issues of recall and multiple reports of single illnesses. Parents may not be able to remember exact times of illness, and because several assessments will be done in the first year, parents may overreport incidences of illness. To prevent this problem, one group member suggested that questionnaires prior to the 1-year assessment use the phrase "since birth" to elicit general reporting, and the 1-year evaluation should be considered a cumulative assessment. Another participant suggested using the phrase "since our last contact" to elicit information only for specific time periods between assessments; but even using this cue, many parents may still not be able to report accurately. No definitive recommendation was given for this potential problem, but the group acknowledged that using calendars may aid memory.

Childhood questionnaire items were reviewed again. The group made its final recommendations. The 4-6 week questionnaire should include questions on the following items:

• Wheeze 
• Cough 
• Eczema 
• Ear infections and rhinitis 
• Bronchial infections 
• Fevers 
• Croup 
• Pneumonia 
• Vomiting 
• Unscheduled doctor visits 
• Antibiotics, acetaminophen, and OTC cough medicines 
• Any respiratory concerns or questions.

At 3-4 months and again at 6 months, in addition to the previous list, questionnaires should also include questions on diarrhea and the child’s overall health.

At 1 year, questionnaires should address both symptoms and diagnoses. The format is static after this time and should include questions on:

• Physician diagnosis of asthma 
• Reduced physical activity due to respiratory symptoms 
• Prescribed and OTC medicines for respiratory symptoms 
• Hay fever 
• Allergies 
• Watery eyes and runny, itchy, or stuffy nose 
• Urgent care or hospital visits for respiratory symptoms.

Breakout Group B: Data Gaps in Diaries
Moderator: Dr. Neas
Discussion leader: Dr. O’Connor
Participants: Dr. Belanger, Dr. Dwyer, Dr. Johnson, Ms. Keim, Dr. Liu, Ms. Moorman, Dr. Raizenne, Dr. Smit, Dr. Svendsen, and Dr. Wright

The original charge for this breakout group was to:

• Draft a list of recommended questions drawn from each study/diary presented during the workshop, focusing primarily on those questions that worked well in previous studies, improvements on current questions, and newly recommended questions. 
• Compile and prioritize a final list of recommended questions and indicate timing and method of collection, flagging any questions that may be sensitive or of special burden to the respondent.

The group discussed the potential use of subject diaries for data collection. Because large amounts of missing data can lead to false negatives in data analysis, inconsistent use of diaries was viewed as problematic. Dr. Belanger explained that baby calendars have been a successful data collection tool for her studies. The calendars contain photos of the individual babies in the study and provide space for parent’s notes such as "baby’s first steps." The calendars also provide space for information on health concerns and problems. Dr. Belanger said that calendars are useful in prompting parents’ memory and are useful information-gathering tools. Calendars are returned to parents following data retrieval because parents often want to keep the personal information.

The group agreed that calendars may be a useful tool to assist memory, but they noted that the cost might be prohibitive. One participant suggested a slightly more detailed short-term diary to collect information on effects of environmental hazards such as air pollution. The group was informed that another workshop would discuss general environmental data collection. They generally agreed that few subjects are motivated to keep up diaries in any form, and they rejected the use of short-term diaries. Calendars are viewed as the best option to assist parents’ memory, and the group’s consensus was that diaries should not be considered for use in the Study.

Breakout Group C: Minimally Invasive Clinical Outcome Measures Suitable for the Study
Moderator: Joel N. Kline, M.D., University of Iowa
Discussion leader: Robert Tepper, M.D., Ph.D., Indiana University
Rapporteur:; Stephanie London, MD, DrPH, NIEHS
Participants: Ms. Keim; Dr. LeMasters; Dr. Liu; Dr. Maliarik; Dr. Martinez; Dr. Ortega; Dr. Platts-Mills; Isabelle Romieu, M.D., M.P.H., Sc.D., Instituto Nacional de Salud Publica, Mexico; Dr. Sears; Dr. Sly; Dr. Smit; and Dr. Weiss

The original charge for this breakout group was to:

• Draft a list of potential clinical measures suitable for the Study, focusing primarily on those measures that worked well in previous studies, are minimally invasive, and feasible on a national scale. 
• Compile and prioritize a final list of clinical measures and indicate timing and method of collection, flagging any measures that require administration of medications; indicate any pilot work needed to evaluate potential methods.

After discussion and deliberation, this breakout group developed the following list of minimally invasive clinical measures to gather asthma-related data in children:

• Health care utilization, including medication use, unscheduled physician visits, hospitalizations, and emergency room visits
  • Priority: high 
  • Frequency: collect at least once a year but more frequently in first year of life 
  • Validation: needed; HMOs could provide assistance; could use a subset for validation 
  • Comments: Recall bias is a big problem, even for adults; frequent data collection can help overcome recall bias.
• Quality of life (QOL), of both child and family (for example, parental illness/wellness)
  • Priority: high 
  • Frequency: annually 
  • Validation: need a pilot study on parental QOL instrument 
  • Comments: QOL is considered both an outcome and an exposure. Juniper’s data could be valuable. There is little data available for families when child is 2 years of age or younger. QOL data should be captured as early in life as possible. Study should consider both illness-related QOL and overall QOL. An appropriate instrument is needed to capture parental QOL data. There is some data indicating absence of father in the first year of life can be a predictor of asthma.
• Suffering scale
  • Comments: Suffering in the child is associated with symptoms. There is no known index of suffering for young children. Child behavior checklist can be used from age 1 and up. Use with infant temperament scale. Will pick up behavioral problems. Development and validation of a suffering scale recommended.
• Growth and development
  • Priority: high ("of vital importance") 
  • Frequency: not indicated (This is a major aspect of the Study as a whole and is the topic of a report from another group) 
  • Measures: developmental milestones; height; weight; anterior-posterior and lateral thoracic ratio (measured with calipers); routine anthropometry; skin folds
• Allergy
  • Priority: high 
  • Frequency: cord blood, 6 months, 12 months, 2, 5, 8 and 16 years 
  • Validation: none indicated 
  • Comments: Collect serum and analyze for allergen-specific IgEs. The group recommended that skin testing not be performed.
• Physical examination
  • Priority: high 
  • Frequency: first priority at 6 months; second priority at 12 months 
  • Validation: none indicated because instruments are available 
  • Comments: Need to assess for atopic dermatitis. Scoring should use SCORAD or Williams protocol.
• Nasal physiology
  • Priority: low 
  • Frequency: at 6 months and at 2-3 years of age 
  • Validation: none indicated 
  • Comments:
    • Acoustic rhinometry—sends noise up nose and echo indicates size of nasal chamber. Standardize how far into nose (1-2 mm). Devices available are very well standardized. Measures cross-sectional areas at different points in the nose. Sensitive to the changes resulting from decongestant use in normal people. Easy to do in sleeping children. Measurements are well standardized. 
    • Nasal airflow—total waste of time. 
    • Rhinomanometry—helps sort out "snorty" infants from wheezy infants.
  •  Nasal lavage—not well tolerated although not harmful in anyway Issue raised of getting viral measurements on nasal lavage versus serum antibodies.
    •  Priority: medium 
    •  Frequency: 6, 12, 18 months 
    • Comments: group felt that serum antibodies at these time points would capture children with recent infection. Nasal swab recommended below would be easier to do.
  • Nasal scrape—for viral culture
    • Priority: low. Culture is not practical. PCR would be practical.
  • Respiratory physiology
• Sleep/respiratory studies in infants
  • Priority: high 
  • Frequency: at 1 month, 6 months, and 2 years of age 
  • Validation: Need a feasibility to determine whether parents can accurately record children’s sleeping respiratory rate; compare with oxypneumocardiogram (OPCG) in a validation/pilot. 
  • Comments:
    • Sleeping respiratory rate—first choice because it is relatively easy and inexpensive. Collect data during first 2 years of life. 
    • Pilot of parent’s ability to count and report needed. 
    • Sleeping respiratory rate plus oximetry (oxygen saturation)—second choice when combined testing. 
    • OPCG (also called polysomnography)—records heart rate, respiratory rate, sleep-disordered breathing. There is evidence that SLR is the most predictive for childhood respiratory illness. Requires overnight recording. Three chest leads and finger lead.
• Infant lung function tests
  • Priority: low 
  • Comments: All require sedation unless sleep deprived. Sleep-deprived testing is difficult and time-consuming. Although infant PFT would be interesting to have they are not practical in a large study. 
  • The most practical might be Respitrace—elastic bandage on chest and abdomen; traces breathing by recording chest and abdominal motion. Would be feasible. Provides a measurement of strategy of breathing. Compared with respiratory rate, low information content. Low priority.
• Oscillometry
  •  Priority: very high 
  • Frequency: start age 3 then annually to age 7. Having a few years of overlap with spirometry will be useful. 
  • Ideally conduct pre- and post-bronchodilator response challenge. If not feasible, then do without bronchodilator.
• Spirometry
  •  Priority: very high 
  • Frequency: start annually at age 5 
  • Conduct pre- and post-bronchodilator response challenge. Post-bronchodilator should be considered slightly lower (medium) priority. Issue with use of bronchodilator medication. Preferred order of testing: spirometry, oscillometry, bronchodilation, oscillometry, spirometry.
• Provocation challenge
  • Priority: medium 
  • Frequency: start with 7-8-year-olds, then every 2 years. Standard is methacholine challenge. 
  • Comments: This is a very informative test that would greatly enhance the Study, but the group recognized that it adds some logistic difficulty.

Breakout Group D: Minimally Invasive Clinical Outcome Biomarkers Suitable for the Study
Moderator: Dr. Ortega
Discussion leader: Dr. Martinez
Rapporteur: Dr. London
Participants: Ms. Keim, Dr. LeMasters, Dr. Liu,Dr. Maliarik, Dr. Martinez, Dr. Ortega, Dr. Platts-Mills, Dr. Romieu, Dr. Sears, Dr. Sly, Dr. Smit, and Dr. Weiss

The original charge for this breakout group was to:

• Draft a list of potential clinical biomarkers suitable for the Study, focusing primarily on those biomarkers that worked well in previous studies, are minimally invasive, and feasible on a national scale. 
• Compile and prioritize a final list of clinical biomarkers and indicate timing and method of collection, flagging any biomarkers hat require a clinical procedure such as collection of blood; include any pilot work needed to validate potential methods

After discussion and deliberation, this breakout group developed the following list of clinical biomarkers to gather asthma-related data in children:

• Exhaled NO
  • Priority: low 
  • Frequency: annually from age 5 
  • Comments: conduct offline with an NO-free breathing source. Group felt that EBC would be much more valuable.
• Exhaled breath condensate
  •  Priority: high 
  • Frequency: annually from age 5 when spirometry is done. Can be collected at same sitting as spirometry. 
  • Comments: Key reason to collect is future value of frozen sample including proteomics. However there are a few well-validated measurements on EBC that justify its current collection. These are pH, isoprostanes, and leukotrienes. Group was very enthusiastic about future promise of EBC.
• Saliva
  • Priority: medium 
  • Frequency: annually from age 5 
  • Comments: assess for cotinine, cortisol (stress measure), specific IgA, and perhaps IgG
• Nasal swab
  • Priority: high 
  • Frequency: first priority is at 6 months, 1 year, and 2 years; annually after that would be lower priority 
  • Comment: for bacterial culture (relevant to hygiene hypothesis); qualitative assessment for speciation only. Note that this is really an exposure measure not an outcome measure. Could be used also for PCR for respiratory viruses - this is primarily an exposure measure but could also be considered an outcome measure.
• Cord blood
  •  Priority: high 
  • Frequency: at birth 
  • Comments: established protocols for cell cryopreservation; assess DNA, RNA, plasma; vaginal delivery versus caesarian section affects results but can adjust for cell viability.
• Meconium
  • Priority: high 
  • Frequency: at birth 
  • Comments: valuable for time-integrated exposure measures for toxins such as metals such as lead and pesticides. Not an outcome measure.
• Stool
  • Priority: medium
  • Frequency: at 6 months and 1 year 
  • Comment: An exposure measure for the hygiene hypothesis, not as an outcome measure. Culture for fecal bacteria. Nasal swab for bacterial culture could be substituted.
• Blood
  • Priority: high 
  • Frequency: at birth (cord blood), 6 moths, 12 months (if lead screening), 2 year, 5 year, 8 year, and 16 years 
  • Comment: If have to drop one draw, drop the one at 12 months; annual blood samples (in addition to the above) are medium priority. High priority is cryopreservation to enable RNA, lymphocyte stimulation tests, methylation studies. Blood will also be used as a DNA source and source for outcome measurements of allergy —allergen-specific IgEs, total IgE, and eosinophils. Parental blood: high priority; sibling blood: high priority if both parents not available provides information on haplotype and missing genotype. DNA of parents is high priority.
• Breast milk
  •  Priority: medium 
  • Frequency: at about two weeks of life 
  • Comment: some dissent which moved it from high to medium; ideally include some some colostrum; assess for cytokines. More of an exposure measure than an outcome measure.
• Urine
  • Priority: low as an outcome measure. Urine would be high as an exposure measure 
  • Frequency: 6 months, 1 year 
  • Comment: alternate source of cotinine which would spare blood, but this is an exposure measure not an outcome measure.

In concluding, this breakout group offered the following comments/suggestions for the Study:

• Link outcomes to hypotheses 
• Sampling strategies affect what measures can be feasibly performed 
• The coordinating staff need to make sure that the environmental exposure group recommendations include exposures important for asthma. Environmental exposures were not a part of this workshop on outcomes.

Participants

Kathleen Belanger, Ph.D., Yale University School of Medicine
Terence Dwyer, M.D., M.P.H., NICHD, NIH, DHHS
Christine Cole Johnson, Ph.D., M.P.H., Henry Ford Health System
Sarah A. Keim, M.A., NICHD, NIH, DHHS
Joel N. Kline, M.D., University of Iowa
Grace K. LeMasters, Ph.D., University of Cincinnati College of Medicine
Andrew H. Liu, M.D., National Jewish Medical and Research Center
Stephanie J. London, M.D., Dr.P.H., NIEHS, NIH, DHHS
Mary J. Maliarik, Ph.D., Henry Ford Health System
Fernando D. Martinez, M.D., University of Arizona, Tucson
Jeanne Moorman, M.S., CDC, DHHS
Lucas M. Neas, Sc.D., Office of Research and Development, EPA
George T. O’Connor, M.D., M.S., Boston University School of Medicine
Hector G. Ortega, M.D., Sc.D., NHLBI, NIH, DHHS
Thomas A. Platts-Mills, M.D., University of Virginia
Mark Raizenne, Ph.D., Canadian Institutes of Health Research
Isabelle Romieu, M.D., M.P.H., Sc.D., Instituto Nacional de Salud Publica, Mexico
Malcolm R. Sears, M.B., Ch.B., McMaster University, Canada
Peter D. Sly, M.D., D.Sc., M.B.B.S., University of Western Australia
H. A. Smit, Ph.D., National Institute of Public Health and the Environment, Netherlands
Erik Svendsen, Ph.D., Office of Research and Development, EPA
Robert Tepper, M.D., Ph.D., Indiana University
Scott T. Weiss, M.D., M.S., Harvard University
Anne L. Wright, Ph.D., University of Arizona, Tuscon