Introduction: Metaphyseal inwaisting is a process that occurs during long bone growthand remodelling of epiphyses and results in a proportional increase in growth plate width(GPW) and a decrease in metaphyseal width (MW). The Metaphyseal Index (MI)compares GPW to MW, usually in the distal femur. However, due to bone-ageassessments, the most commonly performed radiograph in children is that of thehand/forearm. Previous work showed that gymnasts have a more widened growth plateat the distal radius than normal children, but these studies did not quantify themorphological changes using the MI and pQCT measures. Previous studies have shownthat the use of DXA hand/forearm images for the purposes of bone-age assessment wereunreliable for children aged 11 and under.Aims: Examine distal radius morphology of 378 Normal subjects (155 male), 36Gymnast subjects (15 male), 17 NF1 subjects (7 male) and 108 CDGP subjects (83male) to calculate the precision of MI, MW and GPW measurements, to determine anormal reference range for the MI in Normal subjects and use this to compare to theother 3 groups, and to compare longitudinal measurements. Also, to investigate whetherDXA software upgrades have improved the ability to make TW3 bone-age assessments,to investigate how closely DXA compares with standard radiographs using 98 (38 male)DXA hand images and radiographs, and calculate the precision (CV%) of the GP andTW3 bone-age assessment methods.Methods: Anthropometric data, Tanner stage, posterior-anterior hand radiographs andpQCT scans of the non-dominant hand/forearm were obtained. MI was measured usinga semi-automated computer-assisted method. Statistical analyses were used to comparemales and females, and compare the Normal group to other groups. Also, DXA imagesand radiographs were assessed by the same assessor and the TW3 and GP bone-ageingmethods were compared. A CV% was calculated for both comparisons.Results: The CV% of MI, MW and GPW = 1.05%, 0.92% and 1.28% respectively. MIof males and females was not statistically different in any group. The MI of Gymnastswas significantly lower than the Normal group (p = 0.008). The NF1 and CDGP groupswere not significantly different from the Normal group. Longitudinal measurementsindicated those with a low/high MI at the first visit were likely to have a low/high MI atthe second visit, though occasionally the MI would decrease between visits. DXA boneageassessments proved to be reliable in subjects of all ages assessed in this study andshowed a CV% only slightly higher than standard radiographs (CV = 2.95% DXA vs2.68% radiograph). The CV% of GP and TW3 methods = 2.68% & 1.61% respectively.Discussion: The CV% of MI, MW and GPW shows these methods to be very precise.The mean MI of gymnasts is significantly lower than in normal children due to awidening of the growth plate and not due to a reduction in metaphyseal width.Insufficient subject numbers and smaller age ranges, particularly in the Gymnast andNF1 groups may play a part in the non-significant differences between them and theNormal group. DXA CV% shows that DXA is almost interchangeable with standardradiographs. The TW3 and GP CVs% show that TW3 bone-age assessment is moreprecise than the GP method. This confirms the tight control that the MW and GPW havein proportion to each other. This is the first study to quantify changes in distal radiusmorphology in normal, athlete and disease groups, and create a range of normalreference values, which could be useful for future work in this area.