TY - JOUR
T1 - Birth weight and adult cardiovascular risk factors using multiple birth status as an instrumental variable in the 1958 British Birth Cohort
AU - Kwok, Man Ki
AU - Yeung, Shiu Lun Au
AU - Leung, Gabriel M.
AU - Mary Schooling, C.
N1 - Publisher Copyright:
© 2016 Elsevier Inc.
PY - 2016/3/1
Y1 - 2016/3/1
N2 - Background: Birth weight is classified as a risk factor for cardiovascular disease by the World Health Organization, but appropriate preventive interventions remain unclear because the observations have not been confirmed in experiments and appear to be contextually specific. Methods: Using 9452 participants of the 1958 British Birth Cohort at age 42 years in 2000 (58% follow-up), we examined the credibility of multiple birth status as an instrumental variable (IV) for birth weight and, if appropriate, use it to obtain less confounded estimates of the associations of birth weight with cardiovascular disease risk factors including self-reported height, body mass index and hypertension than conventional regression in 2014. Results: Multiple birth (203 twins and 6 triplets) was associated with older maternal age, but not with paternal occupation or maternal smoking. Multiple births had lower birth weight-for-gestational age z-score. Multiple birth status was not directly associated with height, BMI or hypertension. Using IV estimates birth weight-for-gestational age z-score was not clearly associated with height (0.99cm, 95% confidence interval (CI) -0.27, 2.25), body mass index (BMI) (0.42kg/m2, 95% CI -0.17, 1.01) or hypertension (risk ratio 0.82, 95% CI 0.54, 1.23) adjusted for maternal age, with a first-stage F statistic of 145.3 from IV analysis. Conclusions: Multiple birth status is a credible IV for obtaining a less confounded estimate of the association of birth weight with height, BMI and blood pressure. Such analysis suggests that birth weight may be spuriously related to height, BMI and blood pressure, and thus not an effective target for intervention.
AB - Background: Birth weight is classified as a risk factor for cardiovascular disease by the World Health Organization, but appropriate preventive interventions remain unclear because the observations have not been confirmed in experiments and appear to be contextually specific. Methods: Using 9452 participants of the 1958 British Birth Cohort at age 42 years in 2000 (58% follow-up), we examined the credibility of multiple birth status as an instrumental variable (IV) for birth weight and, if appropriate, use it to obtain less confounded estimates of the associations of birth weight with cardiovascular disease risk factors including self-reported height, body mass index and hypertension than conventional regression in 2014. Results: Multiple birth (203 twins and 6 triplets) was associated with older maternal age, but not with paternal occupation or maternal smoking. Multiple births had lower birth weight-for-gestational age z-score. Multiple birth status was not directly associated with height, BMI or hypertension. Using IV estimates birth weight-for-gestational age z-score was not clearly associated with height (0.99cm, 95% confidence interval (CI) -0.27, 2.25), body mass index (BMI) (0.42kg/m2, 95% CI -0.17, 1.01) or hypertension (risk ratio 0.82, 95% CI 0.54, 1.23) adjusted for maternal age, with a first-stage F statistic of 145.3 from IV analysis. Conclusions: Multiple birth status is a credible IV for obtaining a less confounded estimate of the association of birth weight with height, BMI and blood pressure. Such analysis suggests that birth weight may be spuriously related to height, BMI and blood pressure, and thus not an effective target for intervention.
KW - Adult
KW - Birth weight
KW - Cardiovascular diseases
KW - Cohort studies
KW - Multiple births
UR - http://www.scopus.com/inward/record.url?scp=84954414998&partnerID=8YFLogxK
U2 - 10.1016/j.ypmed.2015.12.016
DO - 10.1016/j.ypmed.2015.12.016
M3 - Article
C2 - 26748345
AN - SCOPUS:84954414998
SN - 0091-7435
VL - 84
SP - 69
EP - 75
JO - Preventive Medicine
JF - Preventive Medicine
ER -