Scand J Clin Lab Invest 1990: 50: 649-655
Plasma osteocalcin: biological variations and reference limits
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Chulalongkorn University on 01/10/15 For personal use only.
P . T A R A L L O , J . H E N N Y , B . F O U R N I E R & G . SIEST Centre de Medecine Preventive, U R A CNRS 597, 2 avenue du Doyen Jacques Parisot, 54500 Vandoeuvre-les-Nancy, France
Tarallo P, Henny J, Fournier B, Siest G . Plasma osteocalcin: biological variations and reference limits. Scand J Clin Lab Invest 1990; 50: 649-655. Osteocalcin, the most abundant non-collagenous protein in the bone matrix, is partly released in blood. We have measured its concentration by a radioimmunoassay procedure in 1096 apparently healthy subjects from both sexes who came for a health screening examination. Their ages varied from 4 years to over 65 years. Venous blood was drawn in the morning from fasting subjects. Plasma osteocalcin was higher in men than in women. Its level increased significantly with age, body weight, height and bone age until age 12-13 years in girls and 14-15 years in boys. In women, osteocalcin level increased after the age of 50 years and was higher than in men. It remained constant over age 60 years in both sexes, but was higher in women. There was no effect of menstrual cycle in girls at puberty. Plasma osteocalcin did not vary with follicular and luteal phases or with the use of oral contraceptive drugs in women. The usual nonsteroid anti-inflammatory drugs had no effect on blood osteocalcin level. Reference limits according to age and sex are provided.
Key words: age; healthy subjects; menopause; menstrual cycle; morphometric parameters; plasma bone Gla protein (BGP); puberty; reference limits Pierrette Tarallo, Laboratoire, Centre de Mtdecine Preventive, 2, avenue d u Doyen Jacques Parisot, 54500 Vandoeuvre-les-Nancy, France
Osteocalcin or bone Gla protein (BGP) is a non-collagenous vitamin K-dependent protein bound to calcium. Synthesized by the osteoblasts, it binds to hydroxyapatite and accumulates in the bone matrix. Nano-molar concentrations of B G P are released in the blood [l]. Osteocalcin is strongly correlated with histomorphometric parameters of the bone formation [ l , 21. Its measurement in serum or plasma is, therefore, a convenient method for the study of bone turnover.
However, important variations of plasma or serum B G P levels were observed from one study to another. These variations depend o n the methods used-radio-immunoassay (RIA) or enzyme-immunoassay ( E I A t a n d on the nature of the antibodies (polyclonal [3-51 or monoclonal [6] from bovine, ovine [7], human [8] origin, or antibody raised against the synthetic fragment of human osteocalcin consisting of amino acids 37 to 49 [9]). W e present in this paper, our reference limits of plasma osteocalcin determined by RIA. 649
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Chulalongkorn University on 01/10/15 For personal use only.
650
P. Tarallo et al.
Osteocalcin was measured in a population admitted for health screening in a preventive medicine centre. We tried to evaluate the influence of age, sex, height, weight, puberty, menstruation, menopause, oral contraceptives and nonsteroid anti-inflammatory drugs on its concentration and to determine the relationships between B G P and other markers of bone metabolism such as calcium, phosphate and alkaline phosphatases as rough indices of bone formation. PARTICIPANTS AND METHODS
Participants We studied 1096 apparently healthy subjects (438 men and 658 women) aged from 4 years to more than 65 years. None of them smoked or used any medication, except 31 women taking oral contraceptive drugs and 18 subjects treated with nonsteroid anti-inflammatory drugs.
Method.$ Fasting blood samples were drawn in the morning between 0800 hours and 0900 hours to avoid variations caused by the circadian rhythm of plasma osteocalcin concentrations (10, 111. Plasma or serum were aliquoted and stored at -20 "C until analysed. Osteocalcin was determined in samples by RIA (OSTK-PR; ORIS-Industrie S . A . , Gifsur-Yvette, France). The within- and betweenrun variations were 7% for levels found in children (n=20, meanfSD=29. I f 2 . 0 @I) and 9.1% for the control serum supplied with the kit (n=20; mean k SD=6.S f 0 . 6 pg/l), respectively. To evaluate the preanalytical variations, we compared serum and heparinized plasma osteocalcin levels and studied the influence of storage at -20 "C. Paired Student's t-test showed no statistical difference between serum and plasma from 30 fasting patients (meankSD: serum=8.6f9.3 pgh; plasma=8.6f7.4 &I; t=8.336, p=0.989); thus, for the following studies we used heparinized plasma. There was no statistical difference between plasma levels measured at day 0 and those measured 1 month later after storage at -20 "C (n=20; day O= 14.6k5.0 vgll, 1 month later=13.0+6.8 kgA; t-test= 1.66, p=O. 109). These results agree with those of Melick ef al. [ 3 ] and Price and Nishimoto 1121.
Statistical evaluation
All data in text, tables and figure are presented a s meankSD. The statistical analysis has been done using the nonparametric MannWhitney's test and the Friedman's test for unpaired and paired series, respectively, and the Kruskal-Wallis' test to compare several unpaired series. Pearson's correlation was used to test the linear relationship between osteocalcin and some biological and morphological variables of the health screening examination [ 131. The variables studied were calcium, phosphate, alkaline phosphatases, creatinine, body weight, height, ponderal index (determined from height and weight, with the use of the Lorentz' formula [14, 151) and bone age (the bone maturity, more than the chronological age, is associated with body weight and height, and was determined by Greulich and Pyle's method 1161). The date of the beginning of menopause, obtained from a questionnaire was used to evaluate the influence of this state on plasma osteocalcin. RESULTS Population distribution The normality test for the observed distributions (Chi-square test) of plasma osteocalcin, for children andlor adults in both sexes, rejects the Gaussian model (p
Plasma osteocalcin: biological variations and reference limits
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Chulalongkorn University on 01/10/15 For personal use only.
P . T A R A L L O , J . H E N N Y , B . F O U R N I E R & G . SIEST Centre de Medecine Preventive, U R A CNRS 597, 2 avenue du Doyen Jacques Parisot, 54500 Vandoeuvre-les-Nancy, France
Tarallo P, Henny J, Fournier B, Siest G . Plasma osteocalcin: biological variations and reference limits. Scand J Clin Lab Invest 1990; 50: 649-655. Osteocalcin, the most abundant non-collagenous protein in the bone matrix, is partly released in blood. We have measured its concentration by a radioimmunoassay procedure in 1096 apparently healthy subjects from both sexes who came for a health screening examination. Their ages varied from 4 years to over 65 years. Venous blood was drawn in the morning from fasting subjects. Plasma osteocalcin was higher in men than in women. Its level increased significantly with age, body weight, height and bone age until age 12-13 years in girls and 14-15 years in boys. In women, osteocalcin level increased after the age of 50 years and was higher than in men. It remained constant over age 60 years in both sexes, but was higher in women. There was no effect of menstrual cycle in girls at puberty. Plasma osteocalcin did not vary with follicular and luteal phases or with the use of oral contraceptive drugs in women. The usual nonsteroid anti-inflammatory drugs had no effect on blood osteocalcin level. Reference limits according to age and sex are provided.
Key words: age; healthy subjects; menopause; menstrual cycle; morphometric parameters; plasma bone Gla protein (BGP); puberty; reference limits Pierrette Tarallo, Laboratoire, Centre de Mtdecine Preventive, 2, avenue d u Doyen Jacques Parisot, 54500 Vandoeuvre-les-Nancy, France
Osteocalcin or bone Gla protein (BGP) is a non-collagenous vitamin K-dependent protein bound to calcium. Synthesized by the osteoblasts, it binds to hydroxyapatite and accumulates in the bone matrix. Nano-molar concentrations of B G P are released in the blood [l]. Osteocalcin is strongly correlated with histomorphometric parameters of the bone formation [ l , 21. Its measurement in serum or plasma is, therefore, a convenient method for the study of bone turnover.
However, important variations of plasma or serum B G P levels were observed from one study to another. These variations depend o n the methods used-radio-immunoassay (RIA) or enzyme-immunoassay ( E I A t a n d on the nature of the antibodies (polyclonal [3-51 or monoclonal [6] from bovine, ovine [7], human [8] origin, or antibody raised against the synthetic fragment of human osteocalcin consisting of amino acids 37 to 49 [9]). W e present in this paper, our reference limits of plasma osteocalcin determined by RIA. 649
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Chulalongkorn University on 01/10/15 For personal use only.
650
P. Tarallo et al.
Osteocalcin was measured in a population admitted for health screening in a preventive medicine centre. We tried to evaluate the influence of age, sex, height, weight, puberty, menstruation, menopause, oral contraceptives and nonsteroid anti-inflammatory drugs on its concentration and to determine the relationships between B G P and other markers of bone metabolism such as calcium, phosphate and alkaline phosphatases as rough indices of bone formation. PARTICIPANTS AND METHODS
Participants We studied 1096 apparently healthy subjects (438 men and 658 women) aged from 4 years to more than 65 years. None of them smoked or used any medication, except 31 women taking oral contraceptive drugs and 18 subjects treated with nonsteroid anti-inflammatory drugs.
Method.$ Fasting blood samples were drawn in the morning between 0800 hours and 0900 hours to avoid variations caused by the circadian rhythm of plasma osteocalcin concentrations (10, 111. Plasma or serum were aliquoted and stored at -20 "C until analysed. Osteocalcin was determined in samples by RIA (OSTK-PR; ORIS-Industrie S . A . , Gifsur-Yvette, France). The within- and betweenrun variations were 7% for levels found in children (n=20, meanfSD=29. I f 2 . 0 @I) and 9.1% for the control serum supplied with the kit (n=20; mean k SD=6.S f 0 . 6 pg/l), respectively. To evaluate the preanalytical variations, we compared serum and heparinized plasma osteocalcin levels and studied the influence of storage at -20 "C. Paired Student's t-test showed no statistical difference between serum and plasma from 30 fasting patients (meankSD: serum=8.6f9.3 pgh; plasma=8.6f7.4 &I; t=8.336, p=0.989); thus, for the following studies we used heparinized plasma. There was no statistical difference between plasma levels measured at day 0 and those measured 1 month later after storage at -20 "C (n=20; day O= 14.6k5.0 vgll, 1 month later=13.0+6.8 kgA; t-test= 1.66, p=O. 109). These results agree with those of Melick ef al. [ 3 ] and Price and Nishimoto 1121.
Statistical evaluation
All data in text, tables and figure are presented a s meankSD. The statistical analysis has been done using the nonparametric MannWhitney's test and the Friedman's test for unpaired and paired series, respectively, and the Kruskal-Wallis' test to compare several unpaired series. Pearson's correlation was used to test the linear relationship between osteocalcin and some biological and morphological variables of the health screening examination [ 131. The variables studied were calcium, phosphate, alkaline phosphatases, creatinine, body weight, height, ponderal index (determined from height and weight, with the use of the Lorentz' formula [14, 151) and bone age (the bone maturity, more than the chronological age, is associated with body weight and height, and was determined by Greulich and Pyle's method 1161). The date of the beginning of menopause, obtained from a questionnaire was used to evaluate the influence of this state on plasma osteocalcin. RESULTS Population distribution The normality test for the observed distributions (Chi-square test) of plasma osteocalcin, for children andlor adults in both sexes, rejects the Gaussian model (p
