Ethnic and gender related longitudinal changes in metacarpal dimensions through late childhood and adolescence, and their association with bone growth in a South African cohort
Date
2018
Authors
Magan, Ansuyah
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Introduction
Attainment of peak bone mass in childhood influences skeletal fragility in late adulthood.
Longitudinal studies on bone growth in South African (SA) children are sparse; however, ethnic
differences in fracture rates have been reported in these children. This thesis presents unique
longitudinal data on metacarpal bone development in two SA ethnic groups, spanning 12 years
through puberty and adolescence, a period during which significant physiological changes are
experienced.
Aims
The main aims of this thesis were to assess metacarpal bone growth longitudinally during
adolescence in SA children of different ethnic and sex groups, and, to compare its growth and
strength patterns (measured by radiogrammetry) to those of the non-weightbearing radius
and the weight-bearing tibia (measured by pQCT). I further examined the association between
radiogrammetric metacarpal bone measures and indices and fracture risk.
Hypothesis
There are ethnic and sex differences in appendicular bone growth and strength in South
African children.
Objectives
1. To monitor the drift of the periosteal and endocortical bone surfaces during
metacarpal growth longitudinally in black and white South African children and
adolescents.
2. To compare the growth and strength of weight-bearing and non-weight bearing bones
longitudinally in South African adolescents.
3. To determine whether metacarpal indices predict fracture risk in South African
children and adolescents.
Design
Radiogrammetry measures of the second metacarpal bone form the basis of the three studies
reported in this thesis. Radiogrammetry was performed at the midshaft of the second
metacarpal bone on hand-wrist radiographs obtained from 572 children from the Birth to
Twenty Bone Health Cohort (BHC) annually from ages 9 until 21 years. These radiographs
represent the largest longitudinal collection in African children. Bone length, bone width and
medullary width were measured in millimetres using digital vernier callipers on a total of 4730
radiographs. Superimposition by Translation and Rotation (SITAR) was used to obtain age at
peak metacarpal length velocity (aPLV). Bone width and medullary width were modelled using
SITAR against both chronological age and age from PLV.
In the second study of the thesis, radiogrammetry measures of the second metacarpal bone
were compared to measurements obtained from peripheral quantitative computed
tomography scans of the radius (at 65% of the length of the radius) and tibia (at both 38% and
65% of tibial length) taken annually on 372 black and 152 white participants from the BHC
between the ages of 12 and 20 years. Measurements for individuals who had images of all 3
bones were aligned by age from PLV. Metacarpal, radial and tibial bone width (mm), cortical thickness (mm), medullary width (mm), the stress-strain index (SSIp, mm3) and muscle crosssectional area (MCSA, mm2) were measured. Bivariate regression analysis was performed to
determine the association between metacarpal bone width, cortical thickness and SSIp, and
MCSA of the radius and tibia. Mixed-effects models were used to compare the growth
patterns of bone width, cortical thickness, medullary width and SSIp of the metacarpal, radius
and tibia, for each sex and ethnic group when aligned by age from PLV.
In the third study of the thesis, metacarpal bone length, bone width and medullary width were
measured in 359 BHC children between 10 and 17 years of age, and the following proxies for
bone strength calculated: Metacarpal Index (MCI), Bone Mineral Density (BMD), Section
Modulus (SM), Stress-strain Index (SSIp) and Slenderness Index (SLI). Height and weight were
measured annually. Self-reported physical activity was obtained at age 15 years for the
preceding 12 months. Fracture history was obtained retrospectively at age 17 years.
Multivariate logistic regression analyses, adjusting for height and formal physical activity,
were performed to examine the associations between anthropometric measures, bone
measures and bone indices, and fracture history at 17 years separately for each sex and ethnic
group.
Results
In the first study I showed that in black and white females, tempo and velocity of metacarpal
length growth was similar. However, in males, black children attained PLV seven months later
than whites (p < 0.0001), and had a longer second metacarpal (p < 0.05), greater bone width
size (p < 0.02), later growth tempo (p < 0.0009) and reduced velocity (p < 0.0001) than white
males. Medullary width growth velocity in black females and males peaked 2 years prior to
attainment of PLV and exceeded that of their white peers (p < 0.0001) in whom it peaked 6
12 months after PLV attainment. Black adolescents therefore had wider bones with relatively
thinner cortices and wider medullary cavities than their white peers. Ethnic and sex
differences also occurred in the timing of medullary width contraction that accompanied
expansion in bone width and cortical thickness. In black males, medullary width contraction
commenced approximately 3 years later than in black females, whereas in white males this
occurred a year later than in white females.
The second study showed that, from 12 until 20 years, the associations between metacarpal measures (bone width, cortical thickness and SSIp) and MCSA at the radius (males R2: 0.33 - 0.45; females R2: 0.12 - 0.20) were stronger than the tibia (males R2: 0.01 - 0.11; females R2:
0.007 - 0.04). In all groups, radial bone width, cortical thickness and medullary width accrual
rates were similar to those of the metacarpal, except in white females who had lower radial
cortical thickness (0.04 mm/year) and greater radial medullary width (0.06 mm/year) accrual.
In all groups, except in white males for cortical thickness, tibial bone width and cortical
thickness accrual rates were greater than at the metacarpal. Tibial medullary width (0.29 -
0.35 mm/year) increased significantly compared to metacarpal medullary width (-0.07 - 0.06
mm/year) in males only. In all groups, the SSIp increased in each bone throughout
adolescence, except tibial SSIp in white females.
In the third study I found that, at age 17 years, 82 (23%) participants (black, 16%; white, 42%;
p < 0.001) reported a previous fracture. None of the metacarpal bone measures or indices
were associated with fracture in black participants. In white females, after adjusting for
physical activity, a 1 SD increase in SLI was associated with doubling the fracture risk (95% CI:
1.08; 3.98). In white males, indices usually associated with bone strength were associated with
an increase in fracture risk; a 1 SD increase in BMD was associated with a 2.62-fold increase in
fracture risk (95% CI: 1.22; 10.75), whilst a 1 SD greater SM (95% CI: 1.07; 4.89) and SSIp (95%
CI: 1.11; 4.47) were associated with a more than twofold increase in fracture risk, after
adjustment for height and physical activity.
Conclusions
This longitudinal African study on bone growth during adolescence revealed distinct ethnic
and sex differences in bone acquisition, measured using radiogrammetry. Irrespective of sex
or ethnicity, metacarpal growth was similar to the growth of the non-weight bearing radius
but differed from that of the weight-bearing tibia, when aligned on a common biological
maturity indicator. The local and systemic factors influencing site-specific differences require
further investigation. No single metacarpal index consistently predicted fracture across the
four groups which may be due to ethnic and sex differences in bone micro-structure, muscle
mass and skeletal loading. Thus, metacarpal radio grammetry does not reliably predict fracture
in SA children.
Description
A thesis submitted to the Faculty of Health Sciences, University of the Witwatersrand
for the degree of
Doctor of Philosophy
Johannesburg, 2018
Keywords
Citation
Magan, Ansuyah (2018) Ethnic and gender related longitudinal changes in metacarpal dimensions through late childhood and adolescence, and their associations with bone growth in a South African cohort, University of the Witwatersrand, Johannesburg, <http://hdl.handle.net/10539/27898>