Validation of the Nelson Formula for Estimating the Height of Children in Ogbomoso, Nigeria

When direct height measurement using a stadiometer or infantometer is problematic or impossible, an age-based formula is needed, but the commonly used Nelson formula has not been adequately validated for use in Nigeria. This study investigated the Nelson height formula in healthy children. In a cross-sectional study, we recruited 1361healthy children aged 2 to 12 years using a multi-stage random sampling technique. Heights of the children was measured according to the World Health Organisation standard and estimated using the Nelson formula. We compared the two methods of height determination using the Bland-Altman analysis approach with clinically acceptable limit set at 5 cm a priori. Linear regression analysis was also used to examine the relationship between height and age at p = 0.05. There were 708 (52.0%) males and 653 (48.0%) females, mean age 7.13.0 years. Male and female mean heights at ages 2 to 12 did not differ statistically .On the average, the Nelson formula was 1.3 cm higher than the actual stadiometer height, with 95% limits of agreement ranging from 13.2 to 15.8.The linear regression model that depicts the relationship between height and age as: 𝐇𝐞𝐢𝐠𝐡𝐭 (𝐜𝐦) = 𝟓. 𝟐 𝐱 𝐚𝐠𝐞 (𝐲𝐞𝐚𝐫𝐬) + 𝟖𝟏. 𝟔 𝒄𝒎. A reasonable agreement between heights calculated using the Nelson formula and height measurements using a stadiometer, but systematic and random biases could be more than plus or minus 5 cm, which is outside acceptable clinical limits. A modified Nelson formula is proposed.