Bone mass and biochemical markers of bone turnover increase significantly during puberty. We studied the possible relationships between markers of bone formation and bone resorption and increases in skeletal size, bone volume, and bone density in healthy children at different stages of sexual development. Serum concentrations of bone specific alkaline phosphatase (BALP) and osteocalcin (bone Gla protein, BGP), urinary levels of pyridinoline (Pyr) and deoxypyridinoline (Dpyr) and computed tomography (CT) measurements of the cross-sectional areas of the vertebrae and the femurs, the apparent density of cancellous bone in the vertebrae, and the volume and the material density of cortical bone in the femurs were determined in 126 boys and 143 girls, ages 7–18 years. Serum levels of BALP and BGP and urinary concentrations of Pyr and Dpyr peaked in early puberty and were lowest in the later stages of puberty. CT measurements for the cross-sectional areas of the vertebrae and the femurs, the femoral cortical bone areas, and the apparent density of cancellous bone increased in all children during puberty, while values for material bone density did not change significantly with the stage of sexual development. BALP and BGP showed significant inverse correlations with the material density of bone (r = –0.23 and –0.24, respectively), but no association with bone volume in the appendicular or axial skeleton. In contrast, Pyr and Dpyr correlated with femoral cross-sectional area (r = –0.24 and –0.33, respectively) and cortical bone area (r = –0.29 and –0.33, respectively), and with the apparent density of vertebral cancellous bone (r = –0.26 and –0.19, respectively), but not with the material density of bone. We conclude that, during puberty, there is a differential association between the two components of bone mass and the markers of bone formation and bone resorption; while markers of bone formation are related to the material density of bone, markers of bone resorption are related to the volume of bone.
Osteoporosis, a disease of reduced bone mass and fragility fractures in the elderly, has recently been recognized as having its antecedents during childhood. Because of the lack of definitive treatment, much emphasis is being placed on the prevention of this disease. To this end, reliable and convenient tests for quantifying bone turnover during growth may ultimately be of value in the identification of children who may be at risk for osteoporosis later in life.
Bone metabolism rates are higher in children than in adults. In recent years, bone cell activity has been assessed by measurements of specific markers of bone formation and bone resorption. Biochemical assays for monitoring bone turnover rely on the measurement, in serum or urine, of enzymes or matrix proteins synthesized by osteoblasts or osteoclasts that spill over into body fluids, or of osteoclast-generated degradation products of the bone matrix itself. Serum levels of bone-specific alkaline phosphatase (BALP) and osteocalcin (bone Gla protein, BGP) are currently the most convincing formation markers. The most commonly used resorption markers are pyridinoline (Pyr) and deoxypyridinoline (Dpyr), which are products of collagen degradation that for now are best measured in urine.
Several studies have shown that bone mass and markers of bone turnover change during adolescence.(1–5) Regardless of gender, levels of bone turnover markers appear to reach peak values during midpuberty (Tanner stages II and III) and decrease markedly in late puberty (Tanner stages IV and V).(6–8) Similarly, in both males and females, bone mass increases markedly during puberty and reaches its peak soon after sexual maturity.(9–11) Previous studies have identified significant inverse relationships between levels of bone turnover indices and changes in bone mass during adolescence.(12,13) However, whether bone turnover markers reflect the changes in bone volume and/or bone density that occur during puberty is unknown. Such knowledge may ultimately be of value in identifying the mechanisms by which bone mass is regulated during growth.
In this study, we assessed the possible relations between values for markers of bone formation and bone resorption and measurements of bone volume and bone density at the various stages of sexual development in children.
MATERIALS AND METHODS
The study subjects were healthy white children and adolescents who were recruited from schools of Los Angeles County. The investigational protocol was approved by the Institutional Review Board for clinical investigations at this facility, and informed consent was obtained from all subjects and/or their parents.
Candidates for this study underwent a physical examination by a pediatric endocrinologist to determine their general healthiness and stage of sexual development. Subjects were excluded if they had any chronic illness, if they had been ill for longer than 2 weeks during the previous 6 months, or if they had taken any medications, vitamin preparations, or calcium supplements within the previous 6 months. For determinations of sexual development, the grading system of Tanner was used and subjects were classified as Tanner stages I, II, III, IV, or V.(14) The Tanner system utilizes assessments of the pattern of development of pubic hair in all children and of breast development in girls and penile and testicular size in boys.(14) If discrepancies existed among criteria, greater emphasis was placed on the degree of breast development or testicular and penile size for designation of Tanner stage. All subjects were appropriately physically active for their age.
Measurements of height and weight were obtained, and children in whom either height or weight were not within the 5th and 95th percentiles for the mean age-adjusted normal values were excluded from further evaluation.(15) Thereafter, body surface area (SA) and body mass index (BMI) were calculated as previously described.(16) Skeletal maturation was assessed on the basis of roentgenograms of the left hand and wrist according to the method of Greulich and Pyle.(17) Those in whom chronologic and bone age differed by more than 1 year were also excluded.
A total of 269 healthy white children and adolescents (126 boys and 143 girls) between the ages of 7.2 and 18.2 years were enrolled in this study.
After an overnight fast, blood was drawn for determinations of markers of bone turnover. Timed 2-h urine samples were collected the same day. Serum BALP was measured using a specific two-site immunoradiometric assay with an interassay coefficient of variation (CV) of 8.5%. Serum BGP was measured using a specific two-site immunometric assay with an interassay CV of 11.3%. Pyr and Dpyr were measured in urine following acid hydrolysis and chromatographic purification using high performance liquid chromatography with fluorometric detection at 395 nm. Interassay CVs for Pyr and Dpyr were 5.3% and 9.9%, respectively. Creatinine (Cr) was measured in urine using a kinetic alkaline picrate method and an automated chemistry analyzer (Hitachi, Tokyo, Japan) with 5.3% interassay CV. These assays were performed at Quest Diagnostics, Nichols Institute (San Juan Capistrano, CA, U.S.A.).
Computed tomography bone measurements
All computed tomography (CT) measurements were obtained with the same scanner (CT-T 9800; General Electric Co., Milwaukee, WI, U.S.A.) and mineral reference phantom (CT-T bone densitometry package; General Electric). For determinations in the axial skeleton, the apparent density of cancellous bone and the cross-sectional areas were measured at the lumbar vertebrae, and, in the appendicular skeleton, the cross-sectional area, the cortical bone area, and the material density of cortical bone were measured at the midshaft of the femurs, as previously described.(11,18)
For this study, the apparent density of cancellous bone was defined as the mean value of the CT unit of measurement (mg/cm3) at the midportion of the first three lumbar vertebral bodies. Because of the relatively small size of the trabeculae when compared with the pixel, CT values for apparent cancellous bone density reflect not only the amount of mineralized bone and osteoid, but also the amount of marrow per pixel.(19) The material density of cortical bone was defined as the amount of bone per pixel (mg/cm3) at the midshaft of the femur. Because of the thickness and the relative lack of porosity of cortical bone in the femur, CT values at this skeletal site reflect the material or true density of the bone (the amount of collagen and mineral in a given volume of bone).(18)
The CVs for repeated CT measurements of vertebral cross-sectional area, cancellous bone density, femoral cross-sectional area, cortical bone area, and cortical bone density were calculated to be between 0.6% and 2.5%.(11,18) The time required for the procedure was ∼10 minutes and the radiation exposure was ∼100–150 mrem (1–1.5 mSv) localized to the first three lumbar vertebrae and the midshaft of the femurs; the effective radiation dose was ∼8 mrem.(20,21)
The results are expressed as the mean ± SEM. Data were analyzed using one-way analysis of variance and linear regression analysis. All tests were conducted at the α = 0.05 level and were two-tailed. The computer software program STATVIEW SE+ (Abacus Concepts, Inc., Berkeley, CA, U.S.A.) was used for the analyses.
The anthropometric characteristics of the 269 children and adolescents, as grouped by Tanner stages of sexual development, are summarized in Table 1. The heights of the boys and girls at each stage of sexual development are also depicted in Fig. 1.
Table Table 1. Age and Anthropometric Measurements in Healthy Boys and Girls Grouped by Tanner Stage of Sexual Development
Biochemical markers of bone turnover at each stage of puberty in girls and boys are shown in Table 2 and in Figs. 2 and 3, respectively. Bone formation markers, assessed by measurements of serum levels of BALP and BGP were highest at midpuberty and were considerably lower in late puberty. These differences were significant in boys (BALP: F = 3.0, p = 0.022; BGP: F = 4.7, p = 0.001), but not in girls (BALP: F = 1.1009, p = 0.36; BGP: F = 1.2, p = 0.31). To determine bone resorption markers, urinary measurements of Pyr and Dpyr were corrected by urinary Cr concentration to account for urine dilution. Urinary concentrations of bone resorption markers showed significant differences in both genders at all stages of sexual development with peak values at Tanner stage II (Pyr: F = 9.4, p = 0.0001 for girls, F = 10.9, p = 0.0001 for boys; and Dpyr: F = 7.56, p = 0.0001 for girls, F = 6.2, p = 0.0002 for boys).
Table Table 2. Biochemical Markers of Bone Formation and Bone Resorption in Healthy Boys and Girls Grouped by Tanner Stage of Sexual Development
The apparent density of cancellous bone in the vertebrae increased significantly in all children (F = 8.89, p = 0.0001 in boys, and F = 10.1, p = 0.0001 in girls) throughout puberty (Fig. 4). In contrast, CT values for material bone density in the femur did not change significantly in either boys or girls at any stage of sexual development (Fig. 4). Vertebral cross-sectional area, femoral cross-sectional area, and femoral cortical bone area also increased in all children during puberty (p < 0.0001) (Figs. 5 and 6).
Table 3 shows the intercorrelations between the biochemical markers of bone turnover and the correlations between these markers and the anthropometric measurements. Markers of bone turnover were significantly intercorrelated; correlations between BGP and BALP were, however, relatively weaker (r = 0.35) than those for Pyr and Dpyr (r = 0.89). Indexes of bone formation showed a significant correlation with indexes of bone resorption (r between 0.22 and 0.44; all p < 0.002). While bone resorption markers correlated weakly with anthropometric variables, there was no association between values for BALP and BGP and age, height, weight, SA, or BMI of the subjects.
Table Table 3. Correlation Coefficients for Age, Anthropometric Measurements, and Biochemical Markers of Bone Turnover in Healthy Boys and Girls
As expected, age and anthropometric measurements showed strong and highly significant correlations with vertebral cross-sectional area and with femoral cortical bone area and cross-sectional area (Table 4). While the apparent density of cancellous bone correlated weakly with age and anthropometric variables, the material density of cortical bone at the midshaft of the femur did not correlate with these variables (Table 4).
Table Table 4. Correlation Coefficients for CT Bone Measurements, Age, and Anthropometric Measurements in Healthy Boys and Girls
Table 5 shows the correlations between CT bone measurements in the appendicular and axial skeletons and markers of bone turnover. In the appendicular skeleton, significant inverse correlations were seen between femoral cortical bone density and bone formation markers, and between femoral cross-sectional area and femoral cortical bone area and bone resorption markers. In the axial skeleton, values for vertebral cancellous bone density did not correlate with bone formation markers, but significant inverse correlations were observed with Pyr and Dpyr.
Table Table 5. Correlation Coefficients for CT Bone Measurements and Biochemical Markers of Bone Turnover in Healthy Boys and Girls
Adolescence is the period of life during which the growth spurt and the greatest accumulation of bone occur. The objective of this study was to examine the possible relationships between biochemical markers of bone turnover and skeletal size, bone volume and bone density at each stage of puberty. In the present study, we measured Pyr and Dpyr as bone resorption markers and BGP and BALP as bone formation markers in children at different levels of sexual development. Our results showed a differential relationship between bone resorption and bone formation markers and the density and the volume of bone in children and adolescents. While Pyr and Dpyr showed significant inverse correlations with the size and the volume of bone in the axial and appendicular skeletons, negative associations were observed between BGP and BALP and measurements of the material density of bone.
Our results corroborate previous studies showing the biochemical markers of bone resorption are strongly interrelated with similar patterns of change during puberty. Both Pyr and Dpyr were highest at early puberty (Tanner stage II) and were lowest in late puberty (Tanner stage V). At Tanner stage V, values for both Pyr and Dpyr were similar to adult levels. These markers of bone resorption showed significant inverse correlations with measurements of the cross-sectional areas of the femurs and the vertebrae and CT measurements reflecting the volume of bone in the axial and appendicular skeletons, cortical bone area in the femur, and apparent density of cancellous bone in the vertebra. Due to the small size of the trabeculae when compared with the pixel, CT measurements of the apparent density of cancellous bone in the vertebrae are mainly a reflection of the number and the thickness of the trabeculae, not of the true material density of bone.(22) Pyr and Dpyr were not, however, associated with the material density of the bone.
In contrast to the markers of bone resorption, the markers of bone formation were not associated with the size or the volume of the bones in the axial or appendicular skeletons. There were, however, significant inverse correlations between BGP and BALP and the material density of bone. We found that, in normal children, these indices of bone formation accounted for 5–6% of the variance in the material density of cortical bone. In this study, we used CT to quantitate bone mineralization in vivo at the midshaft of the femur.(18) At this level, the thickness and the relative lack of porosity of cortical bone circumvents partial volume averaging errors, and CT values reflect the densities and concentrations defined by the osteoid and mineral. While the nonmineral fraction may contribute to minor fluctuations in measurements of cortical bone density, the CT numbers are primarily based on the calcified bone fractions which have a high attenuation coefficient.(18) These measurements are analogous to in vitro determinations of the intrinsic mineral density of bone, which are commonly expressed as the ash weight per unit volume of bone.(23)
It should be noted that although BGP and BALP are both indicators of bone formation they did not show as parallel a response to puberty as bone resorption markers. The relatively weak correlation between these two markers may reflect the expression of these proteins at different stages of osteoblast development and synthetic activity.
Our results corroborate previous studies, suggesting that osteocalcin and alkaline phosphatase play a role in the mineralization process. The best evidence for the involvement of BALP in bone mineralization is the disease hypophosphatasia, an inborn error of metabolism characterized by a reduction of activity of BALP. Hypophosphatasia is characterized clinically by defective skeletal mineralization that manifests as rickets in infants and children and osteomalacia in adults.(24) Several hypotheses have been used to explain the action of BALP in skeletal mineralization. It has been suggested that BALP could act as a plasma membrane transporter for inorganic phosphate, or an extracellular calcium-binding protein that stimulates calcium phosphate precipitation and orients mineral deposition into osteoid.(25) Alternatively, it could be based on the increase of inorganic pyrophosphate seen in patients with hypophosphatasia, which, at high concentrations, has been reported to impair the growth of hydroxyapatite crystals.(26) Support for the participation of BGP in the mineralization process is more limited, because there is no human genetic disease that lacks the function of this macromolecule of the osteoblast. Nevertheless, several studies have suggested that BGP has an effect on bone mineral deposition.(27–30)
In interpreting biochemical indices of bone turnover during childhood, the contribution from both remodeling and modeling must be considered. Whereas, in adults, bone formation and bone resorption are part of the turnover mechanism, called remodeling, by which old bone is replaced by new bone, during growth the activities of the skeletal cells are also devoted to the development of the skeleton, a process called modeling. In children, modeling and remodeling of bone tissue cannot be independently assessed by measurements of biochemical markers. During the pubertal growth spurt, modeling is responsible for most of the bone formation and resorption. Thereafter, remodeling prevails and modeling activity gradually decreases, as longitudinal growth ceases. Thus, the lower values for serum and urine concentrations of all markers observed in this cross-sectional study during late puberty could reflect the decrease of bone modeling.
In summary, the results of this study shows that there is a differential relationship between the various structural parameters of the growing skeleton and markers of bone formation and bone resorption. While markers of bone formation showed a significant inverse correlation with the material density of bone, markers of bone resorption were negatively associated with the volume of bone and with skeletal size.
The authors would like to thank Ms. Cara L. Beck for her technical assistance and comments on this manuscript. This work was supported in part by a grant (R01-AR4–1853–01A1) from the National Institute of Arthritis Musculoskeletal and Skin Diseases and a grant (1RO1 LM06270–01) from the National Library of Medicine.