Table 2

 Power calculations: required sample size for cross sectional studies (80% confidence for 0.05 two sided significance)

Duration of treatmentAnticipated height deficitAnticipated change in height z-scoreRequired number of individuals per group (n)
*Calculated using the formula: n = 16/f2; where n, number in each group, f, expected change expressed as a fraction of the standard deviation. The standard deviation for height was taken as 6.0 cm and the cohort standard deviation in z-scores was estimated as 1.0 to 1.2.
1 year1.0 cm−0.17554–797*
2 years2.0 cm−0.33147–212*
3 years or more3.0 cm−0.564–92*