Bone, 6,63-67 (1985) Printed in the USA. All rights reserved
8756-3282185 $3.00 + .oo 0 1965 Pergamon Press Ltd.
The Effect of Trauma on Serum Concentrations of Vitamin D Metabolites in Patients with Hip Fracture P. LIPS,’ R. BOUILLON,2 M.J.M.JONGEN,’ NETELENBOS’
F.C. VAN GINKEL,’ W.J.F. VAN DER VIJGH,’ and J.C.
Ziekenhuis der Vrije Universifeit, Amsterdam,
The Netherlands. ‘Rega Instituut, Leuven, Belgium.
Address for correspondence and reprints: P. Lips, M.D., Department of Internal Medicine, Academlsch Zlekenhurs der Vrije Untversiteit, P.B. 7057, 1007 MB Amsterdam, The Netherlands.
et al., 1982). Serum concentrations of the vitamin D metabolites, 25(OH)D, 24,25(OH),D, and 1,25(OH),D, were significantly lower in the patients than in aged control subjects. Serum 25(OH)D levels were lower than 15 nmol/L in 38% of the patients. However, blood samples of patients can only be obtained after the fracture, and the effect of trauma on vitamin D metabolism is unknown. Serum calcium concentrations fall after a hip fracture or after a major surgical procedure. This fall is caused by the decrease in serum albumin, which occurs within 24 hr after trauma (Lal et al., 1976). We observed large differences in serum protein concentrations between patients with hip fracture and aged controls (Lips et al., 1982). Low serum protein levels may be pre-existent in some patients, but probably they are a consequence of trauma in most. Serum concentrations of vitamin D metabolites may also be influenced by trauma, i.e., the hip fracture, as they are greater than 99% protein bound (Haddad and Walgate, 1976; Bouillon and van Baelen, 1981). We have measured the serum concentration of vitamin D binding protein (DBP) in our patients with hip fracture and aged control subjects in order to assess the effect of trauma on the vitamin D metabolite levels.
In a previous study we observed lower serum concentrations of 25(OH)D, 24,25(OH),D, and 1,25(OH),D in patients with hip fracture than in aged control subjects. In order to evaluate the effect of trauma on vitamin D metabolite levels, we measured serum concentrations of vitamin D binding protein (DBP) in 118 patients with hip fracture and 71 aged control subjects. Serum DBP was lower in the patients than in the controls (mean f SD 315 +- 80 vs 371 f 44 mgll, P < 0.001). Serum DBP correlated positively with serum total protein, albumin, q-globulin, and thevitamin D metabolite levels in the patients. When correcting for differences in serum DBP, serum 25(OH)D and 24,25(OH)*D still were significantly lower in patients than in controls, whereas serum 1,25(OH),D was not. The free 1,25(OH),D index (1O5X molar ratio 1,25(OH),DIDBP) was lower in patients than in controls, but the level of significance was marginal. This difference was not significant when patients and controls with impairment of renal function were excluded. It is concluded that the difference in serum 25(OH)D and 24,25(OH),D between patients and controls is largely preexistent. However, the lower serum 1,25(OH),D in the patients is mainly caused by the trauma. The free 1,25(OH)*D concentrations are almost similar in the two groups when renal function is normal.
Subjects and Methods
Key Words: Vitamin D Metabolites-Vitamin D Binding Protein-Hip Fractures-Free Hormone Concentration-Change of Serum Proteins after Trauma.
The study was conducted in 125 patients with hip fracture and 74 control subiects (mean aae rt SD 75.9 f 11 .O vears and 75.6 + 4.2 years respectively), and Gas described in detail previously (Lips et al., 1982). Blood samples were drawn from the patients usually within 1 or 2 days of the fracture. The patients were sampled before or immediately after the operation. Control subjects were sampled throughout the year in order to account for seasonal vanation (Lips et al., 1983). Patients and control subjects had similar renal function, that IS, there was no significant difference in serum creatinine concentrations (mean f SD: 86 f 26 vs 85 f 16 PmollL. NS). lmparrment of renal function (serum creatinine > 110 PmollL maximum value 184) was more common in the patient group (n = 22) than in the control subjects (n = 4). Medication that might influence vitamin D metabolism included diphenylhydantoin in two patients and phenobarbital in one patlent. Estrogens
Introduction Osteomalacia and vitamin D deficiency have been observed with varying incidence in patients with hip fracture. In a large study in 125 patients with hip fracture, we did not find serious osteomalacia, whereas many patients were vitamin D deficient 63
P. Lips et al.: Trauma, DBP, and vitamin D metabolites
or progestogens were not used by any patient or control subject. One patient was treated with vitamin D,50,000 IU twice weekly, in addition to sodium fluoride and calcium carbonate. This patient was excluded from the analysis. Serum protein fractions were derived by densitometry from an electrophoretogram using cellulose acetate as support material and Ponceau S as dye. Vitamin D binding protein (DBP) was measured by single radial immunodiffusion with an interassay coefficient of variation of 2.5% (Bouillon et al., 1977). The vitamin D metabolrtes [25(OH)D, 24,25(OH),D and 1,25(OH),D] were extracted from serum samples and purified by gradient HPLC. Concentrations of 25(OH)D and 24,25(OH),D were measured by a competitive protein binding assay using rat serum as binding protein, Interassay coefficients of variation are 3.9 and 8 O%, respectively. Concentrations of 1,25(OH),D were measured by a competitive protein binding assay using chicken duodenal mucosa receptor as bindrng protein. The interassaycoefficient of variation is 4.8% (Jongen et al., 1981a,b). Differences in parameters between patients and controls were statistically evaluated using analysis of variance (ANOVA), with age and sex as covariates. Correlations were calculated using Pearson’s product moment correlation coefficient. A correction for the effect of trauma was made in two different ways: (1) When evaluating the difference in serum vitamin D metabolite levels between patients and controls, serum DBP was included as covariate in the analysis of variance. (2) Every vitamin D metabolite value was divided by the corresponding DBP value. In fact, the free 25(OH)D index was calculated as 1O3 X molar ratio 25(OH)D/DBP. The free 24,25(0H),D index was calculated in the same way. The free 1,25(OH),D index was calculated as 1O5 X molar ratio 1,25(OH),D/DBP. These molar ratios probably are a good estimation of the true free concentrations (Bouillon et al., 1981). Differences were considered significant if the P value was less than 0.05.
Results The serum concentration of DBP was lower in patients with hip fracture than in control subjects, as was previously observed for serum albumin and total protein (Table I). Correlations of serum DBP with age, serum proteins, and the vitamin D metabolites are presented in Table II. The latter correlations are not an effect of a mutual relationship with age, as is demonstrated by the partiat correlations. The correlations of serum DBP with serum 25(OH)D and serum 1,25(OH),D in the patients are shown in Figure 1. Serum concentrations of the vitamin D metabolites were lower in the patients than in the control subjects. However, when correcting for the trauma by including serum DBP as covariate in ANOVA, the difference in serum 1,25(OH),D was not significant anymore (Table III). The free indexes of the vitamin D metabolites are presented in Table IV. Individual values of the free 25(OH)D index and the free 1,25(OH),D Index are shown in Figure 2. The free 1,25(OH),D index was lower in patients with impaired renal function (serum creatinine > 110 PmollL) than in patients with normal renal function (mean +- SD: 0.92 f 0.59 vs 1.35 + 0.65, P < 0.02). When excluding all subjects with renal functional impairment,
the difference in the free 1,25(OH),D index between patients and controls was not significant (Table IV).
Discussion The vitamin D status in patients with hip fracture is usually studied after the fracture, and the results are compared with those of healthy aged control subjects(Lund et al., 1975; Baker et al., 1979; Lips et al., 1982). The results are modified by the fracture trauma and sometimes by the operation, depending on the time of sampling. The problem of comparability may be bypassed by choosing a control group that has also suffered trauma, such as patients having undergone nonacute elective orthopedic surgery (Wootton et al., 1979). We made a correction for the effects of trauma by means of DBP. The present survey shows lower values of serum DBP in the patients than in the control subjects, corresponding to lower levels of total protein and albumin. The correlations between these proteins in the patients presumably are due to parallel changes after the trauma, whereas the correlation between DBP and a,-globulin in patients and controls follows from their parentage. The correlations between OBP and the vitamin D metabolites in the patients probably originate from a decrease in both after the fracture. This result was not unexpected, as we previously observed in the patients positive correlations between serum concentrations of the vitamin D metabolites and albumin (Lips et al., 1982). These findings correspond to the parallel decrease of serum calcium and albumin after hip fracture, discussed in the same paper. The serum concentration of DBP may rapidly change during acute illness or after operation (Brown et al., 1980). The decrease of serum DBP and the vitamin D metabolites after hip fracture may be due to hemorrhage, which often is considerable. Alternatively, a synthesis problem might exist, as has been observed in liver disease, where similar relationships among serum albumin, DBP, and 25(OH)D are found (Imawari et al., 1979). The quantitative effect of the trauma on the vitamin D metabolites may be inferred from the correction. When correcting by including serum DBP as covariate in ANOVA, the difference between patients and controls in serum 25(OH)D and 24,25(OH),D remans significant, which Indicates pre-existing differences. On the other hand, the difference in serum 1,25(OH),D is not significant anymore. Thus, lower levels of 1,25(OH),D in the patients may be an effect of the fracture itself. The other correction method, by means of calculating free indexes, is a usual method for estimating free or effective hormone concentrations of protein-bound hormones, such as thyroxine or cortisol (Bouillon et al., 1981). The results of both correction methods point to the same direction. The free 1,25(OH),D index is lower in patients than in controls, but the difference is small and marginally significant (Table IV and Figure 2). It is not significant when patients and control subjects with impaired renal function are excluded. Thus, the free serum
Table I. Serum concentrations of DBP and protein fractions rn patients with femoral neck fracture and control subjects. Patients
Control subjects n
mean 2 SD
mean t SD
8.2 -r- 1.1
+ 60 mg/L
P. Lips et al: Trauma, DBP, and vitamin D metabolites
Table II. Correlations of serum DBP with age, serum protein fractions, and serum concentrations of the vitamin D metabolites. Patients
Control subjects n DBP with age
DBP wrth total protein
P < 0.001
DBP with Albumin
% the patients, pantal correlationscontrolhng for age are. r = 0.49, 0.28, and 0.57 respectively, ‘In the patients, partial correlationscontrolltng for age are: r = 0.27 (P < 0.005).
P < 0.001. lo), and 0 39 (P < 0 OOl), respectively.
1,25(OH),D concentration probably is almost similar in patients with hip fracture and aged control subjects with normal renal function. However, the free serum 25(OH)D and 24,25(OH),D concentrations are substantially lower in patients than in controls. The degree of substrate [ie., 25(OH)D] deficiency in the patients might be insufficient to result in a manifest deficiency of free 1,25(OH),D. Free serum 1,25(OH),D concentration is maintained at a constant level by an increase in serum parathyroid hormone (PTH) concentration, as is observed in winter (Lips et al., 1983). A high bone turnover, compatible with secondary hyperparathyroidism, was observed in more than 20% of the bone biopsies of our patients with hip fracture (Lips et al., 1982). However, one should expect a significantly higher mean serum PTH concentration in the patients than in the control subjects, which was not observed (Lips et al., 1982). This may be due to the fact that the patients were less active and less mobile than the control subjects (Lips, 1982). Immobilization suppresses serum PTH levels (Stewart et al., 1982). The effect of trauma on serum PTH concentration is not well known. Our data are cross-sectional, and low serum DBP levels may be pre-existent in some patients, corresponding to low levels of other serum proteins. A sequential study should confirm if serum DBP falls after major bone trauma, for example, orthopedic operations. In conclusion, our results indicate that nearly normal serum 1,25(OH),D concentrations are preserved in the presence of a low 25(OH)D concentration. Our study again demonstrates the tight feedback control of serum 1,25(OH),D in contrast with the more loosely regulated serum concentrations of 25(OH)D and 24,25(OH),D.
Acknowledgment: This study was partly supported by the Praeventiefonds, The Hague. We are grateful to Dr. H. Kruijswijk and Dr. C. L. van Schaik from the Onze Lieve Vrouwe Gasthuis for their cooperation in the study. Prof. C. van der Meer(A.Z.V.U.) and Prof. P. Meunier (Lyon) gave invaluable advice. Mr. I. Jans performed the DBP measurements, and Mrs. A. Wiggers gave secretarial assistance.
Fig. 1. Relationship between serum DBP and the serum concentrations of 25(OH)D and 1,25(OH),D in patients with hip fracture.
P. LIPS et al.: Trauma, DBP, and vttamin D metabolites
Table III. Serum concentrations of the vitamin D metabolites in patients with femoral neck fracture and control subjects. Control subjects n
Mean + SD
Mean + SD
18.3 + 10.0 nmol/L
2.3 + 105
NS (P = 0.07)
aANOVA.covarlates age and sex bANOVA.covanates age, sex, and DBP.
Table IV. Free indexes of the vitamin D metabolites in patients with femoral neck fracture and control subjects Control subjects
mean + SD
mean + SD
Free 25(OH)D index
4.63 + 1.98
3.02 + 1.56
0.32 + 0.27
0.14 + 0.21
Free 1,25(OH),D index
1.48 ” 0.45
1.29 + 0.66
<0.05 (P = 0.04)
Free 1,25(OH),D Index”
1.47 + 0.43
1.35 f 0.65
NS (P = 0.18)
aPatlents and controls with normal renal function (serum creatlnlne s 110 pmol/L)
. free 25(OH)D
. . i
. : a
: controls n = 71
hip fractures n=117
Fig. 2. lndrvrdual values of the free 25(OH)D Index and the free 1,25(OH),D Index In patients wrth hrp fracture and aged control subjects
P. Lips et al.: Trauma, DBP, and vitamin D metabolites
Lal SK., JacobK.C Nag1O.N.. Annamalar A.L. and Narr C.R.: Variation of some plasma components after closed fractures. J TIiluma 16:206-211, 1976.
Baker M.R., McDonnell H , Peacock M. and Nordtn B.E.C.: Plasma 25.hydroxyvitamrn D concentrations rn patrents wrth fractures of the femoral neck. Br. Med. J. 1.589, 1979. BoutlIon R. and van Baelen H.: Transport of vrtamin D: Srgnrficance of free and total concentrations of the vrtamtn D metabolites. Calcif. Tissue Int. 33.451-453, 1981. Bourllon R , vanBaelen H. and de Moor P: The measurement of the vrtamtn D brndrng protean In human serum. J. C/in. Endocrinol. Metab. 45225-231, 1977 Bouillon R., van Assche F.A., van Baelen H., Heyns W. and de Moor P.: Influence of the vitamin D btnding protein on the serum concentratton of 1.25 drhydroxyvttamrn D J. C/m Invest. 67589.596, 1981, Brown I.R.F., Sood A. and Carter N.D.: Vrtamin D brndrng globulin levels and afknity in various clintcal conditrons. J. C/in. Patho/. 33:966-970, 1980. Haddad J.B. and Walgate J.: 25.hydroxyvitamin D transport tn human plasma. J. ho/. Chem. 251:4803-4809, 1976. ImawanM.,AkahumaY.. ItakuraH.. MutoY.. KosakaK. andGoodmanD.S:The effects of drseases of the liver on serum 25hydroxyvrtamrn D and on the serum binding protern for vitamrn D and its metabolites. J. Lab. C/m Med. 93.171-180, 1979.
Lips P.: Metabolrc causes and preventron of femoral neck fractures. Ph.D. Theses, Vnte Universitert. Amsterdam, 1982. Lips P.. Hackeng W.H.L., Jongen M.J.M , van Grnkel F.C and Netelenbos J.C.: Seasonal varrakon tn serum concentrattons of parathyroid hormone in elderly people. J. C/in. Endocnnol Mefab. 57.204-206, 1983.
Jongen M.J.M., van der Vifgh WJ.F., Willems H.J J. and Netelenbos J.C.: Analysts for 1.25.dihydroxyvttamrn D in human plasma, after a liqutdchromatographrc purification procedure with modified competitive protean btndrng assay. C/in Chem. 27.444-450, 1981 a.
Lips P.. Netelenbos J.C.. Jongen M.J M , van Gtnkel F.C.. Althuts A.L.. van Schatk C.L., van der Vrtgh W.J.F, Vermeiden J P.W. and van der Meer C.: Hrstomorphometrrc profile and vitamrn D status rn patrents wtth femoral neck fracture. Metab. Bone Dis. Rel Res. 4:85-93, 1982. Lund B.. Sorensen O.H. and Chrrstensen A.B.: 25.hydroxycholecalctfeml and fractures of the proxrmal femur. Lancet 2:300-302. 1975. Stewart A.F.. Adler M., Byers C.M., Segre G.V. and Broadus AE.: Calcium homeostasrs in rmmobrlrzatton: an example of resorptive hypercalctuna. N Engl. J. Med. 306:1136-l 140. 1982. Wootton R., Brereton P.J., Clark M B., Hesp R Hodkinson H.M., Klenerman L.. Reeve J.. Slavrn G. and Tellez-Yudilevrch M.: Fractured neck of femur in the elderly: an attempt to Identify patients at rusk C//n So. 57:93-101, 1979.
Received: December 12, 1983 Revised: July 23, 1984 Accepted: August 20, 1984
Jongen M J M.. van der Vttgh W.J.F., Wtllems H.J.J.. Netelenbos J C. and Lips P Srmultaneous determrnation of 25hydroxyvitamin D, 24.25.dthydroxyvrtamin D, and 1,25dihydroxyvitamin D In plasma or serum Clan. Chem. 27.1757-1760, 1981b.
Nous avons observe lors dune etude antbrieure que les taux &riques de 250HD, de 24,25(OH),D et de 1,25(OH),D etalent plus bas chez ks maladea presentant une fracture du col femoral que chez les temolns de m6me Ige. Afin d’evaiuer i’effet du traumatisme sur ies taux des metaboiites de la vitamine D, nous avons mesure ies concentrations s6riques de la proteine porteuse de la vftamine D (PPD) chez 118 malades avant une fracture du col et 71 temoins iiges. Le taux &ique de la PPD Btait corn% positivement chaz ies malades avec celui des protines totales, ceux de I’albumine, des o 2 globuiines et des m&aboiites de la vitamine D. Apres correction en fonction de ces differences du taux de la PPD, les concentrations Jrtques de la 250HD et de la 24,25(0H),D etaient encore signlficatlvement pius bas chez Ies maiades qua cher ies tamoins, alors que feS taux de ia 1,25(OH),D n’etalent plus dlfferents. Cindlce de 1,25(OH),D libre (IO5 x rapport molaire 1,25(OH),D/PPD) dtait plus bas chez les malades que chsz fes t6moins, mals la difference n’existait plus apres exclusion des malades et des timolns avant une fonctlon r&ala aiteree. On peut en conciure que la difference des taux seriques de la 25(OH)D et de la 24,26(OH),D entre fes malades et ies tlmolns pr6exiate nettement a la fracture. L’abaissement du taux skique de la 1,26(OH),D chez ies maladesest cats&e principalement par le traumatisme. Les concentrations de la 1,25(OH),D sont presque simitairesdans les deux groupes lorsque la fonctlon r&tale est normale.