Intended for healthcare professionals

Endgames Statistical Question

Standard deviation or the standard error of the mean

BMJ 2015; 350 doi: (Published 17 February 2015) Cite this as: BMJ 2015;350:h831
  1. Philip Sedgwick, reader in medical statistics and medical education1
  1. 1Institute for Medical and Biomedical Education, St George’s, University of London, London, UK
  1. p.sedgwick{at}

The effects of a diet with a low glycaemic index during pregnancy on maternal and neonatal morbidity for women at risk of fetal macrosomia (large for gestational age infants) were investigated. A randomised controlled trial was performed. The intervention consisted of a low glycaemic index diet from early pregnancy. The control treatment was no dietary intervention. Participants were women without diabetes, all in their second pregnancy, who had previously delivered an infant weighing greater than 4000 g. In total, 800 women were recruited and randomised to the intervention (n=394) and control treatment (n=406) groups.1

The baseline characteristics for the treatment groups were presented; these included body mass index (BMI) (intervention: mean 26.8 (standard deviation 5.1); control 26.8 (4.8)). The outcome measures included birth weight and gestational weight gain. Of those women allocated to the intervention, 372 provided data at follow-up, compared with 387 of those allocated to the control. A per protocol analysis was performed. Mean birth weight was greater in the intervention group than in the control group, although the difference was not significant (mean 4034 (standard error 26.4) v 4006 (25.3) g; mean difference 28.6 g, 95% confidence interval −45.6 to 102.8; P=0.449). Mean gestational weight gain was significantly less for the intervention arm (12.2 (standard error 0.23) v 13.7 (0.25) kg; mean difference −1.35 kg, −2.45 to −0.24; P=0.01). The researchers concluded that a low glycaemic index diet in pregnancy did not significantly reduce birth weight for large for gestational age infants but it did have a significant effect on reducing gestational weight gain for women at risk of fetal macrosomia.

Which of the following statements, if any, are true?

  • a) The standard deviation of the BMI quantified the variation in measurements at baseline for the sample members allocated to a treatment group

  • b) The standard …

View Full Text

Log in

Log in through your institution


* For online subscription