Odds ratios and adjusting for confoundingBMJ 2011; 343 doi: https://doi.org/10.1136/bmj.d6833 (Published 26 October 2011) Cite this as: BMJ 2011;343:d6833
- Philip Sedgwick, senior lecturer in medical statistics
- 1Centre for Medical and Healthcare Education, St George’s, University of London, Tooting, London, UK
Researchers investigated the association of cerebral palsy with birth weight and Apgar score five minutes after birth. A population based cohort study design was used. The cohort consisted of 543 064 singletons born without malformations between 1986 and 1995 in Norway who had survived the first year of life. Data were extracted from the Medical Birth Registry of Norway and the Norwegian Registry of Cerebral Palsy for the 10 year period.1
The prevalence of cerebral palsy was reported to be 18 in 10 000 children. The table⇓ shows the adjusted and unadjusted odds ratios for categorised Apgar score and birth weight for each of the subdiagnoses of cerebral palsy—quadriplegia, diplegia, and hemiplegia.
Which of the following statements, if any, are true?
a) Reference categories are indicated by (1) in the table
b) The association between cerebral palsy and Apgar score was confounded by birth weight
c) The adjusted odds ratios were derived using Cox proportional hazard regression
d) The odds ratios are estimates of the population relative risks
Answers a, b, and d are true, …