Bone Mass from Childhood to Adulthood

Attaining high peak bone mass (PBM), the highest bone mass value in life which is
reached in young adulthood, is important as it reduces the risk of having low bone mass
in old age69, 80. Low bone mass is associated with high fracture risk3, 60. Osteoporosis is
the result of bone loss, a physiological process related to aging and/or low PBM. It
would therefore be of great value to identify children at risk of reaching low PBM for
possible interventions. But the level of correlation, in the thesis referred to as “tracking”,
in bone mass from childhood to adulthood is unclear. Making predictions about adult
bone mineral density (BMD) from childhood measurements is difficult as bone
properties change rapidly during growth59. Most studies that have evaluated the
question are either cross-sectional, have a short follow-up time or end close to the final
growth spurt, making reliable predictions difficult. There are some reports suggesting
that a childhood excess62, 76 or deficit77, 116 in BMD remains in adulthood, and the few
prospective studies that have addressed the question infer that there is a partial
“tracking” in BMD during growth. Longitudinal studies with serial measurements that
cover both the pre- and post-pubertal phases and that follow the participants until peak
bone mass (PBM) would provide data with a higher level of evidence and thereby
increase our knowledge.
In this thesis, with a long-term prospective study design, we have evaluated the
“tracking” of bone mass from childhood to adulthood, and specifically evaluated two
risk factors linked to low BMD. The first is a fracture in childhood which has been an
event identified as associated with low BMD both in childhood31 and in adulthood54.
The second is premature birth in relation to low birth weight, since both traits have
been associated with low PBM67, 84.
We invited subjects from three previous studies63, 86-87 published during 1981–1985 to
be re-measured almost three decades after the initial measurement. The study subjects
with a mean age of 10 years (range 3–17) at the first measurement were re-measured a
mean 27 (range 25–29) years later. Bone traits were prospectively evaluated with singlephoton
absorptiometry (SPA) in 214 individuals consisting of three cohorts: healthy
control subjects, children with fracture during childhood and children born preterm,
either small for gestational age (SGA) or appropriate for gestational (AGA). In the
second cohort we evaluated bone traits prospectively by dual-energy X-ray
absorptiometry (DXA) in 121 children from the Pediatric Osteoporosis Prevention
(POP) study, an exercise intervention study that is primarily designed to assess
10
musculoskeletal development and fracture risk in response to increased physical
education in school children. The study subjects with a mean age of 8 years (range 7–
9) at the first measurement were re-measured a mean 11 (range 10–12) years later.
Our aim was to evaluate (i) whether a bone mass scan in childhood can be used to
predict bone mass in adulthood, (ii) whether children who sustain a fracture are at
increased risk of reaching low adult BMD and (iii) whether prematurely born children,
either AGA or SGA, are at increased risk of reaching low adult BMD.
The correlation coefficients (r) between pre-pubertal and young adulthood
measurements for distal radius BMC and BMD varied between 0.35 and 0.64 and for
femoral neck BMC, BMD and bone area it varied between 0.37 and 0.65. A childhood
fracture in men was associated with a low BMC Z-score (–0.4 (95% CI –0.6, –0.1))
and low BMD Z-score (–0.4 (95% CI –0.7, –0.1)) at baseline and with a low BMC Zscore
(–0.5 (95% CI –0.8, –0.2)) and low BMD Z-score (–0.4 (95% CI –0.7, –0.1))
at follow-up. Preterm-born children were still shorter in adulthood (p=0.03), they also
had lower femoral neck (FN) BMC, FN BMD, tibial cortical BMD, tibial crosssectional
area and SSI than controls (all p-values 0.001 to <0.05). The deficits were
driven by lower bone traits in preterm SGA individuals, while no differences were seen
in preterm AGA individuals compared to controls.
This thesis shows that an individual pediatric bone mass scan, regardless of whether it
is evaluated with SPA or DXA and independent of the measured skeletal region, has
poor ability to predict an adult bone mass value. We also show that a childhood fracture
in men was associated with low BMD and smaller bone size in young adulthood and
that prematurity and being born SGA is another risk factor for low bone mass in young
adulthood.