Nephrol Dial Transplant (2007) 22: 3202–3207
doi:10.1093/ndt/gfm347
Advance Access publication 13 June 2007
Original Article
Regulation of fibroblast growth factor-23 in chronic kidney disease
Per-Anton Westerberg1, Torbjörn Linde1, Björn Wikström1, Östen Ljunggren1, Mats Stridsberg2
and Tobias E. Larsson1
1
Department of Medical Sciences and 2Department of Clinical Chemistry, Uppsala University Hospital, Uppsala, Sweden
Abstract
Background. Fibroblast growth factor-23 (FGF23)
is a circulating factor that regulates the renal
reabsorption of inorganic phosphate (Pi) and is
increased in chronic kidney disease (CKD). The aim
of the current investigation was to study the regulation
of FGF23 in CKD subjects with various degree of
renal function. As such, we analysed the relationship
between FGF23, Pi, calcium, parathyriod hormone
(PTH), 25(OH) vitamin D3(25(OH)D3), 1,25(OH)2
vitamin D3(1,25(OH)2D3) and estimated glomerular
filtration rate (eGFR).
Methods. Intact FGF23 and other biochemical
variables were analysed in 72 consecutive adult outpatients with various stages of CKD (eGFR ranging
from 4–96 ml/min.) Association studies were performed
using linear univariate and multivariate analysis.
Results. FGF23 was significantly elevated at CKD
stage 4 (266 315 pg/ml, P < 0.001) and 5 (702 489
pg/ml, P < 0.001) compared with CKD 1–2 (46 43
pg/ml). In CKD 4–5 an independent association
between log FGF23 and Pi (P < 0.001), 25(OH)D3
(P < 0.05) as well as eGFR (P < 0.01) was observed.
In contrast, in CKD 1–3 log PTH (P < 0.05) was
the only independent predictor of log FGF23 in
multivariate analysis. In CKD 1–5, Pi (P < 0.00001)
and log PTH (P < 0.01) were explanatory variables
for log FGF23 in multivariate analysis.
Conclusions: We conclude that serum FGF23 increases
in CKD 4–5, in parallel with the emerging hyperphosphataemia. Serum Pi is the most important predictor
of FGF23 when GFR is less than 30 ml/min.
In contrast, our data suggest that Pi may not be an
important determinant of FGF23 in normophosphataemic CKD subjects. Finally, the association between
FGF23 and PTH in CKD may suggest a co-regulation
that remains to be further elucidated.
Correspondence and offprint requests to: Tobias E. Larsson,
Department of Medical Sciences, Uppsala University Hospital,
Ing. 70, 3 tr., UAS, 75185 Uppsala, Sweden.
Email: [email protected]
Keywords: chronic kidney disease; fibroblast growth
factor-23(FGF23); hyperphosphataemia; parathyroid
hormone; phosphate
Introduction
Fibroblast growth factor-23 (FGF23) is a novel
circulating phosphaturic factor that plays a critical
role in renal inorganic phosphate (Pi) reabsorption
[1,2]. Importantly, two genetic disorders of disturbed
Pi homeostasis, autosomal dominant hypophosphataemic rickets (ADHR, OMIM#193100) [2] and hyperphosphataemic familial tumoral calcinosis (HFTC,
OMIM#211900) [3,4] are caused by activating and
inactivating mutations in the human FGF23 gene.
Unlike most other members of the FGF family,
FGF23 contains a signal peptide and is secreted into
the circulation. Bone is the major physiological site of
FGF23 expression [5].
FGF23 decreases renal Pi reabsorption, and subsequently serum Pi levels, by reducing the apical
expression of the sodium-dependent Pi co-transporter
IIa (NPT2a) in the brush-border membrane of the
renal proximal tubules [1]. Consequently, hypophosphataemia is a hallmark of a variety of disorders
associated with increased FGF23 levels, such
as tumour-induced osteomalacia [6] and X-linked
hypophosphataemic rickets [5].
The role of FGF23 in chronic kidney disease (CKD)
and secondary hyperparathyroidism has been subject
to investigation by several independent research
groups. We and others have previously demonstrated
that serum FGF23 increases as renal function declines
[7–12]. The aetiology of increased FGF23 levels in
renal failure is likely a compensatory mechanism to the
hyperphosphataemia, although retention of FGF23 in
the circulation [10,11] and calcitriol treatment may also
contribute [12,13]. Recent data suggest that FGF23
directly signals in the parathyroid glands [14] and
Nakanishi et al. [7] demonstrated that elevated FGF23
levels are an important predictor of future secondary
hyperparathyroidism in dialysis patients.
ß The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please email: [email protected]
FGF23 regulation in CKD
3203
Although over-expression of FGF23 in vivo unequivocally causes phosphaturia and hypophosphataemia,
it is still not entirely clear to what extent FGF23
regulates Pi levels in normal physiology. In this regard,
several groups have reported a small increase of
FGF23 levels after a dietary Pi load [11,15–17]. This
effect does not occur rapidly, thus it is unlikely that
acute changes or adaptations in serum Pi regulate
FGF23 expression.
In this cross-sectional study, we wanted to elucidate
the regulation of intact, biologically active, FGF23 in
CKD patients with a wide range of renal function.
Additionally, we wanted to establish at which degree
of renal function elevated FGF23 levels are manifest.
calculated
by
using
the
following
estimate:
eGFR ¼ [79.901(cystatin C)1.4389] ml/min/1.73 m2. Parathyriod hormone (PTH) levels were analysed for biointact
PTH (1–84) measured on an automated immunoassay system
(Nichols Advantage, Nichols Institute Diagnostics, San Juan
Capistrano, CA, USA). Measurement of 25(OH)D3 was
performed on the Nichols Advantage automated assay
system (Nichols Institute Diagnostics, San Juan Capistrano,
CA, USA) and 1,25(OH)2D3 using an ELISA kit from
Immundiagnostik Inc., (Bensheim, Germany). Serum intact
FGF23 concentrations were measured using an ELISA
according to the manufacturer’s protocol (Kainos
Laboratories International; Tokyo, Japan) [18]. This
second-generation, two-site, monoclonal antibody ELISA
has previously been shown to recognize the biologically
active, intact FGF23 [18].
Subjects and methods
Subjects
Statistical analysis
We recruited 72 consecutive individuals with CKD and
varying degrees of residual renal function at the out-patient
clinic at the Department of Nephrology, Uppsala University
Hospital. Blood samples were drawn and stored at 708C
until analysis. Written consent was obtained from all patients
and samples were drawn and stored according to ethical
guidelines (ethical approval number 2006:094). The aetiology
of CKD was 15% diabetes mellitus, 30% glomerulonephritis,
6% polycystic kidney disease, 24% other renal disease and
25% unknown renal disease. 25% (n ¼ 18) were treated with
low dose of per oral active vitamin D, 25% (n ¼ 18) used
calcium carbonate as Pi binder and 7% (n ¼ 5) were on
sevelamer treatment. Mean age SD was 59.6 18 years;
gender distribution was 53% males/47% females; mean BMI
(kg/m2) SD was 26.1 4.9; mean systolic blood pressure
(mmHg) SD was 131 19 and mean diastolic blood
pressure (mmHg) SD was 78 10.
All statistical analyses were performed using STATISTICA
software (StatSoft Inc., Tulsa, USA). Univariate correlation
analyses were used where the Pearson correlation coefficient,
r, measured the degree of linear relationship between two
normally distributed continuous variables. Multiple linear
regression analysis was used to investigate the relationship
among the target and several independent predictor variables
simultaneously and to define the relative contributions of the
independent variables to the variation of the target variable.
For all analyses, standardized b-values are presented and
a P-value <0.05 was considered as statistically significant.
A comparison of biochemical changes between various stages
of CKD was performed using ANOVA followed by Tukey’s
HSD post-hoc test.
Biochemical analysis
Routine serum biochemistries for Pi, calcium, albumin
and cystatin C were assessed by standard methods at the
Department of Clinical Chemistry at Uppsala University
Hospital. All calcium levels presented within this paper
were corrected for albumin concentration as follows:
[calcium
(corrected)] ¼ (calcium) [0.018(albumin)] 42].
Estimated glomeral filtration rate (eGFR) was indirectly
Results
A biochemical summary of the CKD cohort is
presented in Table 1. Serum FGF23 levels in the
entire cohort were 235 367 pg/ml, which is higher
than the previously determined mean values
(28.9 1.1 pg/ml, ranging from 8.2 to 54.3 pg/ml)
among 104 healthy control individuals [18]. To further
analyse serum biochemistries, we stratified the cohort
into four parts, depending on degree of renal function:
CKD stage 1–2 (eGFR>60 ml/min), CKD stage
Table 1. Summary of serum biochemistries of the entire CKD cohort as well as stratified into CKD stages 1–5
Number of subjects
FGF23 (pg/ml)
Phosphate (mmol/l)
Calcium (mmol/l)
PTH (pg/ml)
25(OH) vitamin D (ng/ml)
1,25(OH)2 vitamin D (pg/ml)
Creatinine (mmol/l)
eGFR (ml/min/1.73 m2)
Ref range
0.75–1.40
2.15–2.50
6.6–38.4
10–68
15–50
60–110
>100
CKD 1–5
n ¼ 72
CKD 1–2
n ¼ 21
CKD 3
n ¼ 20
CKD 4
n ¼ 16
CKD 5
n ¼ 15
235 367
1.18 0.35
2.27 0.16
75 86
22 11
46 27
222 144
41.0 27
46 43
0.97 0.21
2.25 0.08
28 11
24 10
62 29
93 13
77.3 9.1
56 31
1.04 0.26
2.34 0.12
41 22
23 12
39 19
166 51
42.9 8.9
266 315
1.29 0.24
2.27 0.16
105 78
22 13
45 32
269 116
22.5 5.2
702 489
1.54 0.41
2.23 0.26
158 126
19 11
33 18
395 141
7.7 2.8
Significantly changed variables compared with CKD 1 2 are in bold.
P < 0.05; P < 0.01.
3204
P.-A. Westerberg et al.
3 (eGFR 30–60 ml/min), CKD stage 4 (eGFR
15–30 ml/min) and CKD stage 5 (eGFR<15 ml/min).
Since FGF23 and PTH did not follow a Gaussian
distribution, the log values of FGF23 and PTH were
used for subsequent association analysis.
Mean FGF23 levels were 46 43 pg/ml in CKD 1–2
and did not change in CKD 3 (56 31 pg/ml) as
determined by ANOVA and Tukey’s HSD posthoc test. FGF23 was significantly increased in
CKD 4 (266 315 pg/ml, P < 0.001) and CKD 5
(702 489 pg/ml, P < 0.001) (Table 1) compared with
CKD 1–2. Since significant changes in serum FGF23
occurred between CKD 3 and 4, we further subdivided
CKD 3 subjects into CKD 3a (eGFR 45–60 ml/min)
and CKD 3b (eGFR 30–45 ml/min). No difference in
FGF23 levels were observed between the two groups
(CKD 3a; 54 31 pg/ml and CKD 3b; 60 33 pg/ml).
Thus, our results indicate that FGF23 significantly
rises when eGFR is below 30 ml/min. Similarly, serum
Pi was in the normal range for subjects with CKD 1–3,
although increased in CKD 4 (P < 0.01) and CKD 5
(P < 0.001) (Table 1). PTH levels were normal in CKD
1–3 but also increased in CKD 4 (P < 0.001) and CKD
5 (P < 0.001) (Table 1). No difference in serum
calcium, 25(OH)D3 or 1,25(OH)2D3 between the
different groups was observed (Table 1). In summary,
significant biochemical abnormalities appeared in
CKD 4–5, concomitant with changes in serum FGF23.
We then performed linear univariate analysis with
log FGF23 as dependent variable. Importantly,
A
Pi (r ¼ 0.69, P < 0.001), log PTH (r ¼ 0.56, P < 0.001),
1,25(OH)2D3 (r ¼ 0.24, P < 0.05) and eGFR
(r ¼ 0.68, P < 0.001) all correlated to log FGF23
levels when all subjects were included (Table 2).
In contrast, in a multivariate analysis, only
Pi (b ¼ 0.46, P < 0.00001) and log PTH (b ¼ 0.56,
P < 0.01) remained as independent predictors of log
FGF23 levels (Table 2).
Based on the finding that FGF23 were significantly
changed at eGFR below 30 ml/min, we split the cohort
into two parts, namely CKD 1–3 and CKD 4–5. There
was no significant correlation between log FGF23 and
eGFR in CKD 1–3 in a univariate model (Figure 1B).
In contrast, a strong association was found between
log FGF23 and eGFR in CKD 4–5 (r ¼ 0.62,
P < 0.001) (Figure 1A). No correlation was found
Table 2. Linear univariate and multivariate analysis of biochemical
variables in the CKD cohort with log FGF23 as dependent variable
n ¼ 72
r-value
Phosphate
0.69
Calcium
0.00
Log PTH
0.56
25(OH) vitamin D
0.0009
1,25(OH)2 vitamin D3 0.24
eGFR
0.68
P-value
b-value P-value
<0.001
0.46 <0.00001
0.97 N.S.
0.16
0.075 N.S.
<0.001
0.36 <0.01
0.99 N.S.
0.13
0.12 N.S.
<0.05
0.054
0.52 N.S.
<0.001
0.17
0.16 N.S.
N.S. means not significant, significant P-values are in bold.
B
3.2
log FGF23 (pg/ml)
2.6
2.0
1.4
0.8
0
30
60
90
0
30
60
90
eGFR(ml/min/1.73m2)
Fig. 1. Correlation between log FGF23 and eGFR; (A) CKD stage 4–5 (r ¼ 0.62, P < 0.001); (B) CKD 1–3 (r ¼ 0.24, P ¼ 0.13).
FGF23 regulation in CKD
3205
between log FGF23 and Pi in CKD 1–3 (Figure 2B),
however, there was a linear correlation between
FGF23 and Pi in CKD 4–5 (r ¼ 0.75, P < 0.001)
(Figure 2A). Importantly, we also found a significant
A
3.2
correlation between log FGF23 and log
1–3 (Figure 3B) (r ¼ 0.44, P < 0.01),
correlation was found in CKD 4–5
Finally, there was no association
PTH in CKD
but no such
(Figure 3A).
between log
B
log FGF23 (pg/ml)
2.6
2.0
1.4
0.8
0.4
1.0
1.6
2.2
0.4
1.0
1.6
2.2
phosphate (mmol/l)
Fig. 2. Correlation between log FGF23 and Pi; (A) CKD stage 4–5 (r ¼ 0.75, P < 0.001); (B) CKD stage 1–3 (r ¼ 0.03, P ¼ 0.83).
A
B
3.2
log FGF23 (pg/ml)
2.6
2.0
1.4
0.8
0.6
1.2
1.8
2.4
0.6
1.2
1.8
2.4
log PTH (pg/ml)
Fig. 3. Correlation between log FGF23 and log PTH; (A) CKD stage 4–5 (r ¼ 0.17, P ¼ 0.35); (B) CKD stage 1–3 (r ¼ 0.44, P < 0.01).
3206
P.-A. Westerberg et al.
FGF23, calcium, 25(OH)D3 or 1,25(OH)2D3 in neither
of the two groups (data not shown).
In a multivariate analysis with log FGF23 as
dependent variable, log PTH was the only significant
predictor of log FGF23 levels in CKD 1–3 (b ¼ 0.39,
P < 0.05) (Table 3). In contrast, in CKD 4–5,
Pi (b ¼ 0.60, P < 0.001), eGFR (b ¼ 0.37, P < 0.01)
and 25(OH)D3 were also independent predictors of
log FGF23 (Table 4).
Finally, we sought to further elucidate the association between FGF23 and PTH at earlier stages of
CKD. Notably, median values of log PTH were
proportionally higher in CKD 4 compared with
Table 3. Linear univariate and multivariate analysis of biochemical
variables in CKD 1–3 with log FGF23 as dependent variable
CKD 1-3. n ¼ 41
r-value P-value
Phosphate
0.03
Calcium
0.28
Log PTH
0.44
25(OH) vitamin D
0.25
0.31
1,25(OH)2 vitamin D3
eGFR
0.24
b -value P-value
0.84 N.S.
0.034
0.079 N.S.
0.22
<0.01
0.40
0.12 N.S.
0.24
0.052 N.S. 0.28
0.13 N.S.
0.073
0.83 N.S.
0.18 N.S.
<0.05
0.13 N.S.
0.11 N.S.
0.69
N.S. means not significant. Significant P-values are in bold.
Table 4. Linear univariate and multivariate analysis of biochemical
variables in CKD 4-5 with log FGF23 as dependent variable
CKD 4–5 n ¼ 31
r-value P-value
b -value P-value
Phosphate
0.75 <0.001
0.60
Calcium
0.05
0.77 N.S.
0.29
Log PTH
0.17
0.35 N.S
0.23
25(OH) vitamin D
0.060
0.75 N.S.
0.26
0.041
0.83 N.S.
0.035
1,25(OH)2 vitamin D3
eGFR
0.62 <0.001
0.37
<0.001
0.050 N.S.
0.12 N.S.
<0.05
0.77 N.S.
<0.01
N.S. means not significant, significant P-values are in bold.
CKD 3, than that of median FGF23 levels
(Figure 4). Moreover, only 18 subjects revealed PTH
within the normal reference range in our cohort,
whereas 30 individuals displayed FGF23 levels
<50 pg/ml (data not shown). These findings may
imply that a gradual increase in PTH occurs before
a reciprocal rise in FGF23 in CKD.
Discussion
In the current study, we sought to determine the
regulation of serum FGF23 in relation to other
biochemical parameters in CKD subjects with various
degrees of renal function. In agreement with previous
reports, we demonstrate that FGF23 are increased at
later stages of CKD [9–12]. Importantly, we did not
detect a substantial rise in FGF23 until CKD 4, which
coincided with significant changes in serum Pi and
PTH. In parallel, multivariate analysis revealed that
eGFR was negatively associated to log FGF23 in CKD
4–5, but not in CKD 1–3, supporting that GFR is an
independent determinant of serum FGF23 at later
stages of CKD. Further studies are needed to
determine whether minor changes in GFR at earlier
stages of CKD are associated to variations in serum
FGF23. We conclude that intact FGF23 significantly
rises at GFR below 30 ml/min., in the presence of
other biochemical abnormalities including hyperphosphataemia.
The association between FGF23 and Pi in normal
physiology has been subject to extensive investigation.
Dietary Pi load appears to increase serum FGF23,
although the effect is small [11,15–17]. Other studies
support the notion that supraphysiological concentrations of Pi may be required to induce FGF23
expression [19]. In the current study, we did not
find any correlation between log FGF23 and Pi in
CKD 1–3, whereas Pi was the most significant
predictor of log FGF23 in CKD 4–5. Our data favours
the idea that a rise in FGF23 is detected at a critical
A
B
3.0
log FGF23 (pg/ml)
log PTH (pg/ml)
2.6
2.0
1.4
2.4
1.8
1.2
CKD1-2
CKD3
CKD4
CKD5
CKD1-2
CKD3
CKD4
CKD5
Fig. 4. Medians and quartiles for log PTH and log FGF23 categorized for various CKD stages; (A) log PTH; (B) log FGF23.
FGF23 regulation in CKD
point where manifest hyperphosphataemia occurs and
that hyperphosphataemia is one major stimulus to
elevated FGF23 levels in CKD.
Interestingly, log PTH was positively and independently associated to log FGF23 in CKD 1–3. This may
imply a possible co-regulation of FGF23 and PTH in
normophosphataemic CKD subjects, although further
studies are needed to clarify such a regulation.
FGF23 and PTH are both phosphaturic factors,
decreasing renal Pi reabsorption in the kidney proximal tubules. In CKD, FGF23 and PTH gradually
increase with declining renal function, which mainly
occurred in CKD 4–5 in our study. Importantly, we
observed small increments in PTH in many patients
before a reciprocal rise in FGF23 was detected.
The physiological relevance of this is unclear but may
be explained by several facts. A mild hyperphosphataemia may be a stronger stimulus for PTH
expression than for FGF23. The phosphaturic potency
of PTH in vivo may also be stronger than that of
FGF23. Furthermore, an early rise in PTH may be
more beneficial in CKD because it promotes renal
Pi secretion and protects systemic calcium levels by
increasing circulating serum calcitriol.
In summary, we conclude that serum FGF23 is
increased in CKD 4–5, mainly due to manifest
hyperphosphataemia and decreased renal clearance of
FGF23. Our data suggest that FGF23 is not a suitable
marker for early Pi retention, but rather a sign that
renal Pi excretion no longer can be sufficiently
increased by a further rise in PTH. A positive
association between FGF23 and PTH in CKD 1–3
supports a possible co-regulation of FGF23 and PTH
that needs to be further explored.
Acknowledgement. This work was supported by the Novo Nordisk
Foundation, the Swedish Kidney Foundation and the Swedish
Society of Medicine. We would also like to thank Anna-Lena
Johansson for technical assistance.
Conflict of interest statement: T.E.L. and Ö.L. have received lecture
fees. All other authors have nothing to declare.
References
1. Shimada T, Mizutani S, Muto T et al. Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl Acad Sci USA 2001; 98: 6500–6505
2. ADHR-Consortium.Autosomal dominant hypophosphataemic
rickets is associated with mutations in FGF23. Nat Genet 2000;
26: 345–348
3207
3. Larsson T, Yu X, Davis SI et al. A novel recessive mutation in
fibroblast growth factor-23 causes familial tumoral calcinosis.
J Clin Endocrinol Metab 2005; 90: 2424–2427
4. Benet-Pages A, Orlik P, Strom TM, Lorenz-Depiereux B.
An FGF23 missense mutation causes familial tumoral calcinosis
with hyperphosphatemia. Hum Mol Genet 2005; 14: 385–390
5. Liu S, Guo R, Simpson LG, Xiao ZS, Burnham CE,
Quarles LD. Regulation of fibroblastic growth factor 23
expression but not degradation by PHEX. J Biol Chem 2003;
278: 37419–37426
6. Jonsson KB, Zahradnik R, Larsson T et al. Fibroblast growth
factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia. N Engl J Med 2003; 348: 1656–1663
7. Nakanishi S, Kazama JJ, Nii-Kono T et al. Serum fibroblast
growth factor-23 levels predict the future refractory hyperparathyroidism in dialysis patients. Kidney Int 2005; 67: 1171–1178
8. Shigematsu T, Kazama JJ, Yamashita T et al. Possible
involvement of circulating fibroblast growth factor 23 in the
development of secondary hyperparathyroidism associated with
renal insufficiency. Am J Kidney Dis 2004; 44: 250–256
9. Imanishi Y, Inaba M, Nakatsuka K et al. FGF-23 in patients
with end-stage renal disease on hemodialysis. Kidney Int 2004;
65: 1943–1946
10. Weber TJ, Liu S, Indridason OS, Quarles LD. Serum FGF23
levels in normal and disordered phosphorus homeostasis. J Bone
Miner Res 2003; 18: 1227–1234
11. Larsson T, Nisbeth U, Ljunggren O, Juppner H, Jonsson KB.
Circulating concentration of FGF-23 increases as renal function
declines in patients with chronic kidney disease, but does not
change in response to variation in phosphate intake in healthy
volunteers. Kidney Int 2003; 64: 2272–2279
12. Gutierrez O, Isakova T, Rhee E et al. Fibroblast growth factor23 mitigates hyperphosphatemia but accentuates calcitriol
deficiency in chronic kidney disease. J Am Soc Nephrol 2005;
16: 2205–2215
13. Nishi H, Nii-Kono T, Nakanishi S et al. Intravenous calcitriol
therapy increases serum concentrations of fibroblast growth
factor-23 in dialysis patients with secondary hyperparathyroidism. Nephron Clin Pract 2005; 101: C94–C99
14. Urakawa I, Yamazaki Y, Shimada T et al. Klotho converts
canonical FGF receptor into a specific receptor for FGF23.
Nature 2006; 444: 770–774
15. Antoniucci DM, Yamashita T, Portale AA. Dietary phosphorus
regulates serum fibroblast growth factor-23 concentrations in
healthy men. J Clin Endocrinol Metab 2006; 91: 3144–3149
16. Burnett SA, Gunawardene SC, Bringhurst FR, Juppner H,
Lee H, Finkelstein JS. Regulation of C-terminal and intact
FGF-23 by dietary phosphate in men and women. J Bone Miner
Res 2006; 21: 1187–1196
17. Ferrari SL, Bonjour JP, Rizzoli R. Fibroblast growth factor-23
relationship to dietary phosphate and renal phosphate handling
in healthy young men. J Clin Endocrinol Metab 2005; 90:
1519–1524
18. Yamazaki Y, Okazaki R, Shibata M et al. Increased circulatory
level of biologically active full-length FGF-23 in patients with
hypophosphatemic rickets/osteomalacia. J Clin Endocrinol
Metab 2002; 87: 4957–4960
19. Burnett-Bowie SA, Mendoza N, Leder BZ. Effects of gonadal
steroid withdrawal on serum phosphate and FGF-23 levels in
men. Bone 2006; 40: 913–918
Received for publication: 7.2.07
Accepted in revised form: 7.5.07