The present prospective study, approved by the Ethics Review Board of the Ottawa Hospital, is derived from an ongoing birth-cohort study of the influence of indoor environmental factors on respiratory illness during the first 2 years of life. The study began in 1997 and is being conducted in the province of Prince Edward Island, Canada, which has a population of approximately 150,000. All physicians who practice obstetrics in the province participated in the study. Women in the third trimester of pregnancy received letters from their physicians' offices describing the study and requesting participation. Those who expressed interest to their doctor were contacted by telephone by a member of the research team to obtain informed consent. Excluded from the study are babies born more than 4 weeks prematurely, those with neonatal respiratory difficulties requiring prolonged hospitalization at birth, and those whose families expect to change residence within 2 years of the child's birth. Because of resource constraints, we recruited approximately 40 consecutive newborns each year. Baseline socio-demographic information and family histories were obtained.
The participating parents maintained a daily symptom diary on large multipurpose calendars on which they recorded the presence or absence of symptoms of respiratory illness in the index child. For this study, four symptoms defining a respiratory-illness event were tracked daily: stuffy nose, coughing, wheezing, and shortness of breath. "Wheezing" referred to audible high-pitched musical sounds during breathing, and shortness of breath was defined by the parents' perception of rapid or laboured breathing. Each study family was phoned every 2 weeks to document information from the diary. If the parents had omitted to record symptoms on a daily basis, they provided information for that 2-week period based on recall.
We used the method of Samet to define a respiratory illness event . A respiratory illness event began when 2 consecutive days with at least one of the above four symptoms occurred and ended when there were 2 consecutive symptom-free days. This was done to identify discrete acute illness events as opposed to persistent ongoing symptoms such as a chronically runny nose. There were no doctors' visits, physical examinations, cultures, or blood tests to confirm an infectious event, nor did we collect information on nonrespiratory illness events. The primary outcome of interest was the number of respiratory illness events per year averaged over the 2 years of follow-up.
Lymphocyte phenotyping was performed in participating children at the age of 2 years. Peripheral blood samples were collected in a 5 mL Vacutainer containing ethylenediaminetetraacetic acid. One hundred microlitres of whole blood were incubated with 5 μL of fluorescein isothiocyanate (FITC), phycoerythrin (PE) or phycoerythrincyanin 5.1 (PC5) anti-CD3 (UCHT1), CD4 (13B8.2), CD8 (B9.11), HLA-DR (Immu-35), CD38 (T16), CD25 (B1.49.9), CD45RA(ALB11), CD45RO (UCHL1), CD62L (DREG56) (all from Beckman Coulter, Fullerton, CA, USA), anti- CD69 (L78), and CD29 (MAR4) (Pharmingen, Mississauga, ON) mABs for 25 minutes. The antibody combinations were as follows: CD3-FITC/CD4-PE/CD8-PC5, CD4-PE/HLA-DR-PC5/CD38-FITC, CD8-PE/HLA-DR-PC5/CD38-FITC, CD4-PE/CD25-PC5/CD69-FITC, CD8-PE/CD25-PC5/CD69-FITC, CD4-PC5/CD45RA-P/CD62L-FITC, CD8-PC5/CD45RA-P/CD62L-FITC, CD4-PC5/CD45RO-FITC/CD29-PE, and CD8-PC5/CD45RO-FITC/CD29-PE. Autofluorescence and colour compensation tubes were also included.
The samples were then lysed with the Immuno-Prep reagent system on a Coulter Multi-Q-Prep workstation. Ten thousand lymphocytes were acquired on a Beckman Coulter EPICS ALTRA flow cytometer, and the data was analyzed with EXPO2 analysis software. Lymphocyte subsets were distinguished from nonlymphoid cells by light scattering profiles , and the expression of the various cell surface molecules was determined on CD4+ and CD8+ T cells.
The illness rate per year was correlated with clinical data, and the percentage of lymphocytes expressing one or more cell surface antigens was measured at age 2 years with Pearson's correlation coefficient. Results were then adjusted for any characteristics that were either associated with the incidence of illness or the lymphocyte subpopulation distributions. Multivariate analysis was not performed, and because this was an exploratory study, p values were not corrected for the multiple comparisons.