Summary of published studies examining the association between birth weight and lung function
| First author (year of publication) | Study type | N | Country | Age at lung function tests | Source of birth weight data | Main results (as published) | β (SE) coefficient(l/kg) used in meta-analysis* | Factors controlled for |
|---|---|---|---|---|---|---|---|---|
| SE, standard error. | ||||||||
| *These were either calculated from the information provided or from information obtained from the authors. | ||||||||
| Barker (1991)7 | Retrospective cohort | 825 | England | 59–70 | Obstetric records | Change per 1 lb birth weight: | 0.03 (0.014) | Age, height, smoking, social class at birth |
| FEV: 0.06 l(0.02 to 0.09) | ||||||||
| Strachan (1994)27 | Prospective cohort | 1186 | England & Wales | 34–35 | Obtained at birth | Positive linear trend of FEV1 across quarters of birth weight(p = 0.011) | Not available | Sex, height, smoking, father’s social class |
| Frankel (1996)26 | Prospective cohort | 1258 | Wales | 45–59 | Self-report | Positive linear trend of FEV1 across quarters (p = 0.06) | 0.047 (0.023) | Age, height squared (male only study) |
| Stein (1997)9 | Birth cohort follow-up | 286 | India | 38–59 | Obstetric records | Change per 1 lb birth weight: | 0.06 (0.03) | Age, height, body mass index, social class, smoking (results stratified by sex) |
| FEV1 in men: 0.16 l (−0.02 to 0.33) | ||||||||
| FEV1 in women: 0.14 l (0.00 to 0.30) | ||||||||
| Shaheen (1998)15 | Retrospective cohort | 239 | Scotland | Mean (SD)57.6 (4.3) | Obstetric record | Change per kg birth weight: | 0.036 (0.057) | Age, sex, height, smoking and type of spyrometer |
| FEV1: 0.036 l (−0.074 to 0.147) | ||||||||
| FVC: 0.010 l(−0.119 to 0.139) | ||||||||
| Lopuhaa (2000)14 | Retrospective cohort | 726 | The Netherlands | 50 | Birth records | p values for linear trends across four categories of birth weight: | 0.03 (0.03) | Age, sex, height |
| FEV1: p = 0.6 | ||||||||
| FVC: p = 0.6 | ||||||||
| FEV1/FVC: p = 0.8 | ||||||||
| Cheung (2001)13 | Prospective birth cohort | 120 | China | 30 | Birth records | p values for linear trends across seven categories of birth weight: | Not available | Age, sex, height |
| FEV1: p = 0.36 | ||||||||
| FVC: p = 0.87 | ||||||||
| FEV1/FVC: p = 0.30 | ||||||||
| Boezen (2002)36 | Prospective cohort | 590 | The Netherlands | 18–22 | Obstetric records | Change per 100 g decrease in birth weight | 0.13 (0.04) | Age, height, smoking, maternal smoking in pregnancy, sex, birth order, gestational age |
| FEV1: −0.013(SE 0.004) | ||||||||
| Edwards (2003)10 | Record linkage cohort study | 323 | Scotland | 45–50 | Obstetric records | p values for linear trends across five categories of birth weight: | Not available | Age, height, weight, smoking, deprivation category, sex, parity, gestational age |
| FEV1 men: p = 0.04 | ||||||||
| FEV1 women: p = 0.01 | ||||||||
| FVC men: p<0.01 | ||||||||
| FVC women: p<0.01 | ||||||||
| Laerum (2004)16 | Retrospective cohort | 1514 | Denmark, Estonia, Iceland, Norway, Sweden | 20–44 | Birth or obstetric records | Change per 500 g birth weight | 0.09 (0.024) | Age, sex, height, prematurity, study centre (country), BMI, allergic rhinitis, parental and adult smoking (the results in the published study were also adjusted for birth length; the results included in our meta-analysis were not adjusted for birth length) |
| FEV1: 0.45 (−0.52 to 1.41) % predicted | ||||||||
| FVC: 0.47 (−0.52 to 1.46) % predicted | ||||||||