Table 1 Birth cohort studies
Study | Study population | Age | Exposure assessment | Agent | Range of exposure | Outcome | Relative risk | Comments |
|---|---|---|---|---|---|---|---|---|
Gehring et al 2002 | Birth cohort (GINI and LISA), 1756 children in the city of Munich | 1–2 yrs | Individual exposure estimated from regression models Annual mean at birth | NO2, PM2.5 | 20–67 μg/m3 12–22 μg/m3 | Questionnaire-reported symptoms | Slightly increased OR of non-specific respiratory symptoms, significant only in males | Adjustment for important confounding variables |
Morgenstern et al 2007 | Birth cohort (GINA and LISA), 3577 children from the city of Munich and surrounding area | 1–2 yrs | Individual exposure estimated from regression models and buffer zones variables. | NO2, PM2.5 | 19–72 μg/m3 7–15 μg/m3 Annual mean at birth | Questionnaire-reported symptoms | Distance to nearest main road less than 50 m, OR 1.23 (1.00–1.51) for asthmatic bronchitis Very few children with doctor-diagnosed asthma at this age | Adjustment for important confounding variables |
Morgenstern et al 2008 | Birth cohort (GINA and LISA), 3066 children from the city of Munich and surrounding area | 6 yrs | Individual exposure estimated from regression models and buffer zones variables | NO2, PM2.5 | 6–74 μg/m3 19–13 μg/m3 Average exposure up to 6 years of age. | Questionnaire-reported symptoms Circulating IgE | Distance to nearest main road less than 50 m: OR 1.66 (1.01–2.59) for doctor-diagnosed obstructive bronchitis or asthma OR 1.30 (1.02–1.66) sensitization to pollen | Adjustment for important confounding variables. Blood samples were obtained from 1353 children (an unspecified subset) – the loss in retention rate is not commented |
Brauer et al 2002 | Birth cohort (PIAMA) from the Netherlands, 4,146 children at start, 3,745 at one year and 3,730 at 2 yrs. | 2 yrs | Individual exposure estimated from regression models | NO2, PM2.5 | 13–58 μg/m3 13–25 μg/m3 Annual mean at birth | Questionnaire-reported symptoms | Slightly but significant increased risk of upper respiratory infections | Adjustment for important confounders. |
Brauer et al 2007 | Birth cohort (PIAMA) 3,538 children (retention 85%) A subgroup of 713 children | 4 yrs | Individual exposure estimated from regression models | NO2, PM2.5 | 13–58 μg/m3 13–25 μg/m3 Annual mean at birth | Questionnaire-reported symptoms Circulating IgE | OR for IQR of PM2.5 1.32 (1.04–1.69) for doctor-diagnosed asthma ever and 1.75 (1.23–2.47) for any sensitization to food allergens | Adjustment for important confounders. High rate of retention High risk children were overrepresented in the IgE screening subgroup Low rate of sensitization to outdoor allergen |
Nordling et al 2008 | Birth cohort (BAMSE) of 4,089 children in Stockholm, Sweden 3,515 replied to questionnaires at 4 yrs and 2,543 delivered blood samples | 4 yrs | Individual exposure based on atmospheric dispersion model, high resolution | NOx, Traffic PM | 5–49 μg/m3 1–7 μg/m3 (P5–P95) Annual mean at birth | Questionnaire-reported symptoms Circulating IgE | OR for 95th % range of NOx 1.60 (1.09–2.36) for persistent wheeze and 1.67 (1.10–2.53) for any sensitization to pollen | Adjustment for important confounders. Analyses based on exposures during 1st year of life Significant difference between extreme percentiles of exposure. Dose-response relations not presented. |
Melén et al 2008 | Case-cohort within the BAMSE birth cohort in Stockholm (a randomly sampled subcohort of 542 nonwheezers and 167 wheezers. In addition 375 wheezers from the original cohort) | 4 yrs | Individual exposure based on atmospheric dispersion model, high resolution | NOx, | Questionnaire-reported symptoms Circulating IgE | Variants in the GSTP1 and TNF genes modify the association between sensitization and NOx. | ||
Oftedal et al 2008 | Birth cohort study in Oslo, Norway 2,244 children who lived in Oslo since birth | 10–11 yrs | Individual exposure based on atmospheric dispersion model with contributions from busy roads | NO2, PM2.5 PM10 | Mean (IQR) life time estimate 29.0 (19.5) μg/m3 NO2 and 12.3 (3.6) μg/m3 PM2.5 Smaller ranges compared to Nordling et al | Skin prick test | No association between long-term exposure and sensitization to any allergen (except for D. farinae) | Very few children were sensitized to D farinae and the association with traffic exhaust was likely to be caused by confounders |
Ryan et al 2005 | Birth cohort study (the Cincinnati Childhood Allergy and Air Pollution Study, CCAAPS) – 622 children with at least one allergic parent were enrolled at 6 months | 1 year | Individual exposure (distance to various traffic conditions) based on GIS model | Not recorded | Questionnaire-reported wheeze without a cold | Distance to stop-and-go traffic less than 100 m: OR 2.5 (1.15–5.42) for wheezing without a cold No effect from smoking | A small study with limitations in the control of confounding | |
Ryan et al 2007 | CCAAPS See above! | 1 year | Individual exposure (distance to various traffic conditions) based on GIS model and regression model estimating elemental carbon attributable to traffic | ECAT | 0.30 – 0.90 μg/m3 | Questionnaire-reported wheeze without a cold | Significant exposure-response association between ECAT level and risk of wheeze | The strength of this study is the improved exposure assessment |
Clougherty et al 2007 | Birth cohort – 888 pregnant women were enrolled and the caregivers of 417 children responded to questionnaires after 6–10 yrs | ~7 yrs | Individual exposure based on a regression model | NO2 | 38–85 μg/m3 | Frequent telephone or face-to-face-interviews | OR for 8 μg/m3 increase in NO2 exposure 1.63 (1.14–2.33) for diagnosed asthma but only in children exposed to violence. | NO2 was included as a continuous variable. Concentration at the year of diagnosis showed the closest association. Limitations: Low retention rate, reporting bias and potential confounding |