Bone 131 (2020) 115158

Contents lists available at ScienceDirect

Bone
journal homepage: www.elsevier.com/locate/bone

Full Length Article

The effect of bisphosphosphonates on bone turnover and bone balance in T

postmenopausal women with osteoporosis: The T-score bone marker
approach in the TRIO study
⁎
F. Gossiela, , M.A. Paggiosia, K.E. Naylora, E.V. McCloskeya, J. Walsha, N. Peelb, R. Eastella
a
Academic Unit of Bone Metabolism, The Mellanby Centre for Bone Research, University of Sheffield, Sheffield, United Kingdom
b
Metabolic Bone Centre, Sheffield Teaching Hospitals NHS Foundation Trust, Northern General Hospital Sheffield, United Kingdom

A R T I C LE I N FO A B S T R A C T

Keywords: Postmenopausal osteoporosis is characterised by increased bone turnover and an imbalance between bone re-
Bone turnover markers sorption and formation.
Osteoporosis Bisphosphonate treatment reduces bone turnover but their effect on bone balance is yet to be fully in-
Bone turnover vestigated.
Bone balance
Using the T-score approach our aims were to: i) investigate the effects of oral nitrogen-containing bispho-
sphonates on bone balance and turnover in postmenopausal women with osteoporosis and ii) determine the
relationship of the change in bone balance and turnover with the change in BMD at the lumbar spine and total
hip.
Women were recruited, mean age 67 years, and randomised to receive: ibandronate (n = 55, 150 mg/
month), alendronate (n = 54, 70 mg/week) or risedronate (n = 56, 35 mg/week). They also received calcium
and vitamin D daily. A fasting serum sample was collected at baseline and weeks 1, 2, 4, 12, 13, 48 and 96. The
control group were 226 healthy premenopausal women receiving no treatments. PINP and CTX were measured
using the iSYSIDS analyser and BMD (in g/cm2) of the lumbar spine and total hip were measured by DXA
(Hologic Inc). PINP and CTX values were log10-transformed and normalised. T-scores were calculated using the
mean and standard deviation from the premenopausal group. Bone turnover and bone balance were calculated
from the T-scores.
Mean levels (95% CI) of balance and turnover are shown in the table. The change in turnover at weeks 4, 12
and 48 was inversely correlated with the change in lumbar spine and total hip BMD at weeks 48 and 96,
(p < .01 to p < .001). The change in balance at week 4 positively correlated with the change in total hip BMD
at weeks 48, (p < .01).
Bisphosphonates resulted in an initial positive balance and a reduction in turnover. Some of these changes
were associated with increases in BMD. Bone turnover is a better predictor of BMD than bone balance.

1. Introduction
Postmenopausal osteoporosis is commonly treated with bispho-
sphonates. These drugs work by decreasing the rate of bone remodel-
ling and this is reflected in a decrease in the bone turnover markers
(BTMs) [1]. There is a decrease in bone resorption and then a few weeks
later there is a decrease in bone formation [1]. The decrease in bone
formation is a consequence of the coupling between bone resorption
and bone formation [2]. It has been reported that early treatment
changes in BTMs are associated with long-term changes in bone mineral
density (BMD) [3]. It has not yet been shown whether the balance
between bone formation and resorption relates to the increase in BMD.
BTMs provide the opportunity to study both overall bone turnover
and bone balance and hence allow us to investigate the mechanism of
action of antiresorptives. Bone turnover and bone balance have been
evaluated by bone histomorphometry [4]. There appears to be little
effect on the balance between resorption and formation at the in-
dividual bone remodelling unit (no change in remodelling balance).
However, there were only two studies of bone balance [5,6].
BTMs provide the opportunity of studying balance and turnover at
the whole-body level, rather than at the bone remodelling unit. Bone
balance has been estimated previously using several approaches by: i)

⁎
Corresponding author.
E-mail address: f.gossiel@sheffield.ac.uk (F. Gossiel).

https://doi.org/10.1016/j.bone.2019.115158
Received 11 July 2019; Received in revised form 25 October 2019; Accepted 13 November 2019
Available online 15 November 2019
8756-3282/ Crown Copyright © 2019 Published by Elsevier Inc. All rights reserved.
F. Gossiel, et al. Bone 131 (2020) 115158

calculating the Z-score for postmenopausal women relative to an age-

matched control group [7,8], ii) using the multiple of the medians of
the BTM to calculate bone turnover and bone balance [9], and iii)
calculating a bone balance index based on the relationship between
urinary NTX and serum osteocalcin in women less than two years from
their final menstrual period compared with > 5 years after their final
menstrual period [10].
Here, we describe a T-score approach for assessing bone turnover
and bone balance in postmenopausal women with osteoporosis re-
ceiving bisphosphonate treatment. It is a statistical approach that first
expresses bone turnover markers levels in standard deviation units from
the mean of young women (T-score), then using the averages the T-
scores for resorption and formation markers to make a global estimate
of bone turnover, and subtracts the resorption T-score from the for-
mation T-score to obtain a whole-body balance.
Using the T-score approach our aims were: i) to investigate and
describe the effect of oral nitrogen-containing bisphosphonates on bone
balance and turnover in postmenopausal women with osteoporosis and
ii) determine the relationship of the change in bone balance and turn-
over with the change in BMD at the lumbar spine and total hip. The
hypotheses were; i) bisphosphonates reduce bone turnover and make
bone balance more positive during treatment and ii) a positive bone
balance and a decrease in turnover are related to an increase in BMD.
Fig. 1. The T-score bone marker plot for postmenopausal women receiving
2. Methods
bisphosphonates. The filled circles represent the mean T-score at each time
point. The large circle represents the 95% confidence ellipse for premenopausal
2.1. Study design and population women with the mean in the centre.

The TRIO study was a 2-year, open-label, parallel randomised

control intervention trial of three orally administered bisphosphonates,
at their licensed dose. The overall study design and population has been
fully described previously [1,11].
Women with postmenopausal osteoporosis were recruited through a
hospital metabolic bone clinic and from general practice registers. They
had either a (i) a BMD T score ≤ −2.5 at the lumbar spine or proximal
femur or (ii) a BMD T score ≤ −1.0 at the lumbar spine or proximal
femur plus a previous fracture sustained during a fall from standing
height or less. 165 subjects were used for this analysis. The study
medications used were (i) ibandronate (Bonviva, Roche, 150 mg once a
month) (n = 55), (ii) alendronate (Fosamax, Merck, 70 mg once a
week) (n = 54), and (iii) risedronate (Actonel, Warner Chilcott, 35 mg
once a week) (n = 56). The women also received calcium carbonate 3 g
(1200 mg elemental calcium) and cholecalciferol 20 micrograms
(800 IU) per day (Adcal D3, two tablets daily, ProStrakan) which was
initiated at the baseline 1 visit, one week before the bisphosphonate
was started.
226 pre-menopausal women (ages 35 to 40 years) were recruited to
act as the comparator group. They were not prescribed any bone-active
medications throughout the study. All had regular menstrual cycles and
did not take any hormonal contraception. The study was approved by
the Sheffield Research Ethics Committee and the Medicines and
Healthcare Products Regulatory Agency (MHRA) and was carried out in
accordance with the Declaration of Helsinki and the International
Conference on Harmonisation Good Clinical Practice (ICH GCP)
guidelines. Written informed consent was obtained for all individual
participants included in the study.
for each subject were measured in one analytical batch.
The C-telopeptide of type I collagen (CTX) and intact pro-collagen I
N-propeptide (PINP) were measured using IDS-iSYS automated im-
munoassays (Immunodiagnostic Systems, Boldon, UK). The inter-assay
coefficients of variation (CVs) were 6.5% and 7.2% respectively.
2.3. Bone mineral density
For this analysis BMD (in g/cm2) of the lumbar spine (L1 to 4) and
total hip were measured by DXA using a Discovery A densitometer
(Hologic Inc., Waltham, MA), [11]. The measurements used were per-
formed at baseline (duplicate measurements at baseline 1 and 2), week
12 (duplicate measures at treatment weeks 12 and 13), and weeks 48
and 96. Quality assurance was implemented and assessed. Daily mea-
surements the manufacturer's device-specific test objects were per-
formed in accordance with manufacturer guidelines to monitor the
stability of the densitometer, the coefficient of variation was 0.376%.
2.4. Statistical analyses
Data for CTX and PINP were skewed so these were log10 trans-
formed. To show PINP and CTX levels over 96 weeks, the mean and
95% confidence intervals (CI) were calculated and plotted at each time
point. The BMD T-scores at the lumbar spine and total hip were also
plotted at each time point.
2.5. The T-score bone marker approach

2.2. Biochemistry
Blood was collected following an overnight fast from the post-
menopausal women receiving treatment at baseline 1 (treatment week
−1) and baseline 2 (treatment week 0) then at treatment weeks 1, 2, 4,
12, 13, 48 and 96; samples were collected at baseline from the healthy
premenopausal women. The samples were left to clot for 30 min at
room temperature then centrifuged at 2500g for 10 min. The serum was
extracted and stored at −80 °C until the time of measurement. All visits
The T-score for PINP and CTX were used to calculate bone turnover
and bone balance.
The T-scores for postmenopausal women were then calculated from
the mean and standard deviation for the premenopausal women:
• T-score = (lg10 BTM − mean lg10 BTM)/lg10 standard deviation).
Average bone turnover and bone balance were calculated in each
person from the T-scores for bone resorption and bone formation:

2
F. Gossiel, et al. Bone 131 (2020) 115158

Table 1
Baseline characteristics.
Variable Ibandronate Alendronate Risedronate Premenopausal

n 57 57 58 226
Age (years) 66.9 (7.2) 67.8 (7.8) 66.8 (6.7) 37.9 (1.7)
Height (cm) 159.8 (6.9) 160.1 (5.3) 160.7 (6.0) 164.6 (6.5)
Weight (kg) 67.4 (10.8) 66.3 (10.2) 69.1 (9.4) 67.1 (11.0)
Lumbar spine BMD T-score −2.21 (1.06) −2.30 (0.96) −2.18 (0.842) 0.25 (1.10)
Total hip BMD T-score −1.29 (0.79) −1.59 (0.75) −1.15 (0.82) 0.26 (0.90)
Bone balance T-score 0.21 (0.01, 0.41) −0.01 (−0.23, 0.14) 0.00 (−0.20, 0.19) 0.00
Bone turnover T-score 1.56 (1.34, 1.77) 1.34 (1.12, 1.56) 1.22 (1.00, 1.43) 0.00

Data are shown as mean and standard deviation and mean and 95% confidence intervals for bone balance and bone turnover T-scores.

Fig. 2. The geometric mean (95% CI) levels of CTX (top) and PINP (bottom) for postmenopausal women receiving ibandronate, alendronate and risedronate.

3
F. Gossiel, et al.
Fig. 3. Bone turnover (top) and bone balance (bottom) for postmenopausal women receiving ibandronate, alendronate and risedronate. The dotted line at y = 0 is
the mean for premenopausal women.
• Bone turnover = (T-score bone formation + T-score bone resorp-
tion) / 2
• Bone balance = (T-score bone formation − T-score bone resorp-
tion)
Presentation of the data using the T-score bone marker plot:
Bone balance and bone turnover were plotted against each other
and this produced a 4-quadrant graph (Fig. 1). A bivariate 95% con-
fidence ellipse was calculated, surrounding the data for postmenopausal
women at each time point and for premenopausal women at baseline.
The Mahalanobis distance was calculated from the mean of the pre-
menopausal ellipse to the mean of the postmenopausal ellipse at each
4
Bone 131 (2020) 115158
time point [12]. The distance is zero if the means are equal and it in-
creases if they are not equal. The two-sample Hotelling T2 test was used
to test the significance in the differences between means.
These analyses were performed using programme R, [13]. The 95%
confidence ellipses were plotted using the R library, [14].
Pearson correlations and 95% confidence intervals were calculated
to assess the relationship between the change in lumbar spine BMD T-
scores at weeks 48 and 96 from baseline and the change in bone
turnover and bone balance at weeks 4, 12 (mean of weeks 12 and 13)
and 48. A P-value of < 0.05 was considered significant.
The mean and 95% calculated from the T-scores for bone balance
and bone turnover were plotted separately, to show the differences
F. Gossiel, et al. Bone 131 (2020) 115158

Fig. 4. The T-score bone marker plot for postmenopausal women receiving
Alendronate (black), Ibandronate (blue) and Risedronate (green). The filled
circles represent the mean T-score at each time point. The red circle represents
the 95% confidence ellipse for premenopausal women with the mean in the
centre. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)
between bisphosphonates. The significance from baseline and the pre-
menopausal mean were calculated using Dunn's post hoc test.
3. Results
Baseline characteristics for the postmenopausal receiving bispho-
sphonates, and the premenopausal women, are shown in Table 1.
Fig. 1 shows the T-score bone marker plot for postmenopausal
women receiving bisphosphonate relative to the range of values ob-
served in premenopausal women. It is divided in to 4 quadrants
representing high and low turnover and high and low balance. The
centre of the plot is the mean bone turnover and bone balance for
premenopausal women. At baseline 1, postmenopausal women with
osteoporosis were in a state of high bone turnover and a normal bal-
ance, relative to premenopausal women. At week 2 of treatment their
bone turnover was normal and bone balance was positive; by weeks 48
and 96 bone turnover was low and bone balance was slightly positive.
At each time point the Mahalanobis distance was significantly different
from the mean T-score for premenopausal women, (P < .001).
Fig. 2 shows the mean (95% CI) levels for CTX and PINP for each
bisphosphonate. At week 1, there was a significant reduction from
baseline in CTX with ibandronate, alendronate and risedronate treat-
ment, p < .001. This was sustained for the remainder of the study
period. At week 12, there was a significant reduction from baseline in
PINP with ibandronate, alendronate and risedronate treatment,
p < .001. This was sustained for the remainder of the study period.
Figs. 3 and 4 shows the mean T-scores for bone balance and bone
turnover separately and together, repectively. At baseline 1, the mean
bone turnover T-scores for ibandronate, alendronate and risedronate
were 1.6 (1.3, 1.8), 1.3 (1.1, 1.6) and 1.2 (1.0, 1.4) respectively, in-
dicating high bone turnover at baseline. These were significantly higher
than the mean for premenopausal women, (P < .001). At week 4 there
was a significant decrease in bone turnover with each bisphosphonate
from baseline 1. The mean T-scores were −0.1 (−0.4, 0.2), −0.3
(−0.6, 0.0) and −0.1 (−0.4, 0.2), respectively, (P < .001). The T-
scores were not significantly different from the mean for pre-
menopausal women. After week 4 bone turnover continued to decrease
until week 96 and this was significantly different from baseline and the
premenopausal mean, (P < .001).
At baseline 1, the mean bone balance T-scores for ibandronate,
alendronate and risedronate were 0.2 (0.0, 0.4), −0.01 (−0.3, 0.1) and
0.0 (−0.2, 0.2), respectively. At week 2, there was a marked increase in
bone balance with ibandronate treatment, from baseline 1. The mean T-
score was 4.0 (3.5, 4.4). There were increases with alendronate and
risedronate of 2.3 (1.8, 2.8) and 2.0 (1.6, 2.5) respectively, (P < .001).
After week 4 there was a decrease in bone balance until week 96 where
the mean T-scores were 0.1 (−0.4, 0.5), 0.7 (0.2, 1.1) and −0.1 (−0.6,
0.4), respectively. At weeks 48 and 96 bone balance remained sig-
nificantly higher than the premenopausal mean with alendronate
treatment only.

Table 2
Pearson correlations between the changes in balance and turnover with the changes in lumbar spine and total hip BMD with bisphosphonate treatment.
Correlation (95% CI) P-Value

Lumbar spine T-score – 48 week change Balance – 4 week change 0.09 (−0.08, 0.25) 0.306
Balance – 12 week change −0.07 (−0.23, 0.10) 0.441
Balance – 48 week change −0.08 (−0.24, 0.90) 0.356
Turnover – 4 week change −0.22 (−0.38, −0.06) 0.009
Turnover – 12 week change −0.38 (−0.52, −0.23) < 0.001
Turnover – 48 week change −0.42 (−0.54, −0.27) < 0.001
Total hip t-score – 48 week change Balance – 4 week change 0.25 (0.09, 0.40) 0.003
Balance – 12 week change 0.07 (−0.10, 0.24) 0.395
Balance – 48 week change 0.13 (−0.04, 0.29) 0.122
Turnover – 4 week change −0.34 (−0.48, −0.18) < 0.001
Turnover – 12 week change −0.40 (−0.54, −0.25) < 0.001
Turnover – 48 week change −0.42 (−0.55, −0.27) < 0.001
Lumbar spine T-score – 96 week change Balance – 4 week change 0.18 (−0.03, 0.37) 0.094
Balance – 12 week change −0.01 (−0.22, 0.20) 0.921
Balance – 48 week change −0.01 (−0.22, 0.20) 0.934
Turnover – 4 week change −0.39 (−0.55, −0.20) 0.001
Turnover – 12 week change −0.52 (−0.66, −0.35) < 0.001
Turnover – 48 week change −0.40 (−0.56, −0.21) < 0.001
Total hip T-score – 96 week change Balance – 4 week change 0.18 (−0.03, 0.37) 0.095
Balance – 12 week change 0.19 (−0.16, 0.39) 0.099
Balance – 48 week change 0.10 (−0.12, 0.30) 0.370
Turnover – 4 week change −0.35 (−0.49, −0.12) 0.003
Turnover – 12 week change −0.32 (−0.50, −0.12) 0.002
Turnover – 48 week change −0.39 (−0.56, −0.20) < 0.001

5
F. Gossiel, et al.
To assess the relationship between changes in bone balance and
bone turnover with changes in BMD from baseline, Pearson correlations
were performed, (Table 2). The change in bone turnover at weeks 4, 12
and 48 was inversely correlated with the change in lumbar spine and
total hip BMD at weeks 48 and 96, (p < .01 to p < .001). The change
in bone balance at week 4 positively correlated with the change in total
hip BMD at weeks 48, this was significant, (p < .01). The correlation
of change in bone balance and change in BMD was not significant at the
other time points or at the lumbar spine.
4. Discussions
The T-score bone marker approach was designed to quantify bone
turnover and bone balance between formation and resorption of the
whole skeleton. Using this approach, untreated postmenopausal women
with osteoporosis have a high bone turnover and a normal balance.
Treatment results in a positive bone balance for at least 12 weeks and a
reduction in bone turnover. Some of these changes are associated with
increases in BMD at the spine and hip.
Bone balance based on bone turnover markers has been studied
before, but not to our knowledge in the evaluation of anti-resorptive
therapy. This is the first report of any treatment effect on bone balance
using the T-score approach. A Z-score approach has been used to study
the effect of the anabolic therapy, teriparatide, on bone balance.
Teriparatide therapy of osteoporosis results in an early, large increase
in bone formation markers followed by increases in bone resorption
markers [15]. These sequences of events shown by changes in BTMs
suggested there is a period of time where the anabolic action of PTH is
at its maximum, i.e. the concept of ‘the anabolic window’ [16]. Lane
et al., 2000 reported an ‘uncoupling index’ based on the Z-score ap-
proach and showed that balance was positive for the first 9 months in
glucocorticoid-induced osteoporosis treated with teriparatide for
24 months [17].
The Z-score approach was first developed to compare women with
vertebral fracture with age-matched controls [18]. This Z-score was
calculated from the bone resorption marker urinary deoxypyridinoline
and the formation marker osteocalcin and the mean index was −0.45
as deoxypyridinoline was increased more than osteocalcin in osteo-
porosis. This difference in bone balance is consistent with studies of
bone histomorphometry after tetracycline labelling and reconstruction
of the remodelling cycle, [19]. However, it is not consistent with the
current work. This may be explained by the less severe osteoporosis in
our study – the presence of vertebral fracture was not an inclusion
criterion and the comparison to healthy young rather than older
women.
Bone balance using this same Z-score approach correlated with the
rate of spine bone loss in early postmenopausal women, with
r = −0.35, [8]. Bone balance using regression approach (based on
urinary N-telopeptide of type I collagen and osteocalcin) related to bone
loss from the spine and femoral neck in early postmenopausal women,
whereas N-telopeptide alone did not relate to the rate of bone loss [10].
The Z-score approach has also been used to predict fracture risk. In
the Rochester Epidemiology study, fractures were strongly predicted by
the uncoupling index based on urinary pyridinoline and osteocalcin
(P < .01) [20]. This approach was developed further to include an
overall estimate of bone turnover as well as an estimate of balance, this
time using the multiple of medians rather than using standard devia-
tions [9]. Postmenopausal women had higher turnover and more ne-
gative balance than premenopausal women, but there was no statistical
comparison. There was no separation of women with and without os-
teopenia.
The present study takes the bone marker plot approach [9], and
modifies it by first using standard deviation units rather than multiple
of medians and second using a healthy control group in order to esti-
mate T-scores. We have clearly shown that bisphosphonates result in
short term positive balance and short and long-term reductions in bone
6
Bone 131 (2020) 115158
turnover. It was notable that ibandronate had a pronounced short-term
effect on bone balance as it more rapidly reduces bone resorption and
alendronate may still result in positive balance after 2 years. We plan to
use this approach to study other treatments, such as anabolic therapy
and to test whether this approach allows us to improve the ability of
BTMs to predict fractures.
Secondly, we used the T-score bone marker approach to relate the
changes in bone turnover and in bone balance with changes in BMD.
We have shown that short and long-term reductions in bone turnover
were related with the biggest long-term increases in BMD at the lumbar
spine and total hip. This is consistent with findings from others, [3]. It
was demonstrated that the greatest reductions in BTMs at 6 months
predicted changes in BMD at 3 years in women receiving alendronate or
HRT or combination [3]. Changes in bone balance were weakly related
with BMD, with the exception at the total hip and bone balance at week
4. The latter showed the greatest increases in balance were related to
biggest increases in BMD. The weak relationship between bone balance
and BMD may be explained by the increase noise inherent in the esti-
mate of balance as it combines two measures of bone turnover.
T-score approach may be a useful way of using bone turnover
markers prove in predicting fractures and in monitoring the effects of
anabolic and other anti-resorptive therapy for osteoporosis.
5. Conclusion
Using the T-score bone marker method we can conclude that oral
nitrogen-containing bisphosphonates reduce bone turnover and result
in a transient positive bone balance. In a clinical setting, if BMD change
is considered as a gold standard for assessing osteoporosis, then bone
turnover is a better predictor of BMD than bone balance.
Acknowledgments
This study was funded by Warner Chilcott, the bone turnover
marker measurements were funded by Immunodiagnostics Systems.
Professor Richard Eastell (Academic Unit of Bone Metabolism, The
University of Sheffield) is a National Institute for Health Research
(NIHR) Senior Investigator.
We are grateful to the data safety monitoring board, the Clinical
Trials Research Unit, School of Health and Related Research, for data
management and statistical support and the staff of the Academic Unit
of Bone Metabolism for conducting the study. We would also like to
acknowledge the Lay Advisory Panel for Bone Research and the parti-
cipants of the TRIO study. We acknowledge the support of the NIHR
Clinical Research Facility. The views expressed in this publication are
those of the author(s) and not necessarily those of the National Institute
for Health Research.
References
[1] K.E. Naylor, R.M. Jacques, M. Paggiosi, F. Gossiel, N.F. Peel, E.V. McCloskey, et al.,
Response of bone turnover markers to three oral bisphosphonate therapies in
postmenopausal osteoporosis: the TRIO study, Osteoporosis International: A
Journal Established as Result of Cooperation between the European Foundation for
Osteoporosis and the National Osteoporosis Foundation of the USA 27 (1) (2015)
21–31.
[2] R. Eastell, P. Szulc, Use of bone turnover markers in postmenopausal osteoporosis,
The Lancet Diabetes & Endocrinology 5 (11) (2017) 908–923.
[3] S.L. Greenspan, N.M. Resnick, R.A. Parker, Early changes in biochemical markers of
bone turnover are associated with long-term changes in bone mineral density in
elderly women on alendronate, hormone replacement therapy, or combination
therapy: a three-year, double-blind, placebo-controlled, randomized clinical trial, J.
Clin. Endocrinol. Metab. 90 (5) (2005) 2762–2767.
[4] R. Eastell, J.S. Walsh, N.B. Watts, E. Siris, Bisphosphonates for postmenopausal
osteoporosis, Bone 49 (1) (2011) 82–88.
[5] P.M. Chavassieux, M.E. Arlot, C. Reda, L. Wei, A.J. Yates, P.J. Meunier,
Histomorphometric assessment of the long-term effects of alendronate on bone
quality and remodeling in patients with osteoporosis, J. Clin. Invest. 100 (6) (1997)
1475–1480.
[6] E.F. Eriksen, F. Melsen, E. Sod, I. Barton, A. Chines, Effects of long-term risedronate
F. Gossiel, et al.
on bone quality and bone turnover in women with postmenopausal osteoporosis,
Bone 31 (5) (2002) 620–625.
[7] C.Y. Guo, W.E. Thomas, A.W. al-Dehaimi, A.M. Assiri, R. Eastell, Longitudinal
changes in bone mineral density and bone turnover in postmenopausal women with
primary hyperparathyroidism, J. Clin. Endocrinol. Metab. 81 (10) (1996)
3487–3491.
[8] A. Rogers, R.A. Hannon, R. Eastell, Biochemical markers as predictors of rates of
bone loss after menopause, Journal of Bone and Mineral Research: The Official
Journal of the American Society for Bone and Mineral Research 15 (7) (2000)
1398–1404.
[9] C. Bieglmayer, S. Kudlacek, The bone marker plot: an innovative method to assess
bone turnover in women, Eur. J. Clin. Investig. 39 (3) (2009) 230–238.
[10] A. Shieh, W. Han, S. Ishii, G.A. Greendale, C.J. Crandall, A.S. Karlamangla,
Quantifying the balance between total bone formation and total bone resorption: an
index of net bone formation, J. Clin. Endocrinol. Metab. 101 (7) (2016) 2802–2809.
[11] M.A. Paggiosi, N. Peel, E. McCloskey, J.S. Walsh, R. Eastell, Comparison of the
effects of three oral bisphosphonate therapies on the peripheral skeleton in post-
menopausal osteoporosis: the TRIO study, Osteoporosis International: A Journal
Established as Result of Cooperation between the European Foundation for
Osteoporosis and the National Osteoporosis Foundation of the USA 25 (12) (2014)
2729–2741.
[12] P.C. Mahalanobis (Ed.), On the Generalised Distance in Statistics, Proceedings
National Institute of Science, India, 1936.
[13] R: A Language and Environment for Statistical Computing. R Foundation for
Statistical Computing V, Austria. URL: http://www.R-project.org/.
[14] R.W. Hayden, A review of: “an R companion to applied regression, second edition,
by J. Fox and S. Weisberg”, J. Biopharm. Stat. 22 (2) (2012) 418–419.
Bone 131 (2020) 115158
[15] H. Dobnig, A. Sipos, Y. Jiang, A. Fahrleitner-Pammer, L.G. Ste-Marie,
J.C. Gallagher, et al., Early changes in biochemical markers of bone formation
correlate with improvements in bone structure during teriparatide therapy, J. Clin.
Endocrinol. Metab. 90 (7) (2005) 3970–3977.
[16] J.P. Bilezikian, Combination anabolic and antiresorptive therapy for osteoporosis:
opening the anabolic window, Current Osteoporosis Reports 6 (1) (2008) 24–30.
[17] N.E. Lane, S. Sanchez, H.K. Genant, D.K. Jenkins, C.D. Arnaud, Short-term increases
in bone turnover markers predict parathyroid hormone-induced spinal bone mi-
neral density gains in postmenopausal women with glucocorticoid-induced osteo-
porosis, Osteoporosis International: A Journal Established as Result of Cooperation
between the European Foundation for Osteoporosis and the National Osteoporosis
Foundation of the USA 11 (5) (2000) 434–442.
[18] R. Eastell, S.P. Robins, T. Colwell, A.M. Assiri, B.L. Riggs, R.G. Russell, Evaluation of
bone turnover in type I osteoporosis using biochemical markers specific for both
bone formation and bone resorption, Osteoporosis International: A Journal
Established as Result of Cooperation between the European Foundation for
Osteoporosis and the National Osteoporosis Foundation of the USA 3 (5) (1993)
255–260.
[19] E.F. Eriksen, S.F. Hodgson, R. Eastell, S.L. Cedel, W.M. O’Fallon, B.L. Riggs,
Cancellous bone remodeling in type I (postmenopausal) osteoporosis: quantitative
assessment of rates of formation, resorption, and bone loss at tissue and cellular
levels, Journal of Bone and Mineral Research: The Official Journal of the American
Society for Bone and Mineral Research 5 (4) (1990) 311–319.
[20] L.J. Melton 3rd, S. Khosla, E.J. Atkinson, W.M. O’Fallon, B.L. Riggs, Relationship of
bone turnover to bone density and fractures, Journal of Bone and Mineral Research
the Official Journal of the American Society for Bone and Mineral Research 12 (7)
(1997) 1083–1091.