Deficient Bone Formation in Idiopathic Juvenile Osteoporosis: A Histomorphometric Study of Cancellous Iliac Bone


  • Presented in part at the 15th Annual Meeting of the ASBMR (J Bone Miner Res 1993;8:S158) and at the 4th International Meeting on Osteoporosis, Hong Kong, China, 1993 (Proceedings of the Fourth International Symposium on Osteoporosis and Consensus Development Conference, pp. 200–202)


Idiopathic juvenile osteoporosis (IJO), a rare cause of osteoporosis in children, is characterized by the occurrence of vertebral and metaphyseal fractures. Little is known about the histopathogenesis of IJO. We analyzed by quantitative histomorphometry iliac crest biopsies from 9 IJO patients (age, 10.0–12.3 years; 7 girls) after tetracycline labeling. Results were compared with identically processed samples from 12 age-matched children without metabolic bone disease and 11 patients with osteogenesis imperfecta type I. Compared with healthy controls, cancellous bone volume (BV) was markedly decreased in IJO patients (mean [SD]: 10.0% [3.1%] vs. 24.4% [3.8%]), because of a 34% reduction in trabecular thickness (Tb.Th) and a 37% lower trabecular number (Tb.N; p < 0.0001 each; unpaired t-test). Bone formation rate (BFR) per bone surface was decreased to 38% of the level in controls (p = 0.0006). This was partly caused by decreased recruitment of remodeling units, as shown by a trend toward lower activation frequency (54% of the control value; p = 0.08). Importantly, osteoblast team performance also was impaired, as evidenced by a decreased wall thickness (W.Th; 70% of the control value; p < 0.0001). Reconstruction of the formative sites revealed that osteoblast team performance was abnormally low even before mineralization started at a given site. No evidence was found for increased bone resorption. Compared with children with osteogenesis imperfecta (OI), IJO patients had a similarly decreased cancellous BV but a much lower bone turnover. These results suggest a pathogenetic model for IJO, in which impaired osteoblast team performance decreases the ability of cancellous bone to adapt to the increasing mechanical needs during growth. This will finally result in load failure at sites where cancellous bone is essential for stability. (J Bone Miner Res 2000;15:957–963)


Osteoporosis in childhood is uncommon and is mostly secondary to a spectrum of diverse conditions. These include prolonged immobilization, osteogenesis imperfecta (OI), malabsorption syndromes, corticoid excess, and homocystinuria. Serious diseases such as leukemia may temporarily present as osteoporosis. Apart from OI, the exclusion of these primary disorders usually is not difficult. If no underlying cause can be detected, idiopathic juvenile osteoporosis (IJO) is said to be present.(1)

This entity was first recognized by Dent et al.(2) IJO typically develops in a prepubertal, previously healthy child of either sex. There is an insidious onset of diffuse pain (back, hips, and feet) and difficulty in ambulation. Long bone fractures, mostly at metaphyseal sites, may occur. Vertebral compression fractures are frequent and cause the upper/lower segment ratio to decrease below unity. On X-rays, the new bone formed in metaphyseal areas appears as a radiolucent band for which Dent coined the term “neo-osseous osteoporosis.”(3) Spontaneous recovery is the rule after 3–5 years of evolution, although spine deformities and severe functional impairment may persist.(4) The first reports on IJO (reviewed in Ref. 5) described patients with a negative family history who were between 8 and 13 years of age at disease onset. More recently, some familial cases of increased bone fragility with collagen abnormalities have been classified under the heading of “juvenile osteoporosis.”(6,7)

Table Table 1.. Study Groupsand Structural Histomorphometric Data
 IJOControlsOIp (C)p (OI)
  1. Values are mean (SD).

  2. p(C), significance of difference between IJO patients and controls; p(OI), significance of difference between IJO patients and OI patients.

n (m/f)9 (2/7)12 (7/5)11 (7/4)  
Age (years)10.9 (0.9)11.2 (1.2)11.7 (1.3)0.520.15
BV/TV (%)10.0 (3.1)24.4 (3.8)11.8 (5.2)<0.00010.40
Tb.Th (μm)96 (13)145 (16)116 (22)<0.00010.03
Tb.N (/mm)1.03 (0.28)1.69 (0.23)1.00 (0.39)<0.00010.85
BS/BV (mm2/mm3)22.2 (3.1)14.8 (1.5)17.3 (3.3)<0.00010.003

The mechanisms leading to IJO are unclear. There are no defined biochemical characteristics. Alterations of the markers of bone and mineral metabolism are nonspecific and inconsistently noted.(4,8–10) In some cases a negative calcium balance has been reported.(2,11) Little is known about the changes in the bone tissue itself, and no study has compared static and dynamic histomorphometric parameters of IJO patients and controls of the same age. Also, the histomorphometric differences between IJO and OI have not been characterized. To address these questions, we obtained iliac crest biopsy specimens from 9 IJO patients after tetracycline labeling and analyzed them by quantitative histomorphometry. Only nonfamilial IJO cases corresponding to the classical description of the disease were included in this study to avoid possible admixture of other disease entities. Results in these patients were compared with identically processed samples from 12 age-matched subjects without metabolic bone disease and 11 patients with OI type I.



Nine children, 10.0–12.3 years old (Table 1), were referred to us for evaluation of spine compression fractures and/or nontraumatic long bone fractures. The presentation was typical of IJO in each case. Other causes of osteoporosis were excluded. Basic biochemical workup did not show significant disturbances of mineral homeostasis.

The control population consisted of 12 age-matched children (age, 9.1–12.9 years; Table 1). These were part of a larger group of children and adolescents without metabolic bone disease, who were biopsied with the aim of establishing reference data.(12) Iliac bone biopsy specimens were obtained during various orthopedic procedures for conditions such as lower limb deformities, scoliosis, clubfoot, and other problems, which require corrective surgery. None of these children had evidence of metabolic bone disease. All subjects were ambulatory and had normal renal function as assessed by serum creatinine. None were immobilized before biopsy or received medications known to affect bone metabolism.

The third study group comprised 11 age-matched children (age, 9.1–13.2 years) with typical clinical features of OI type I. A detailed analysis of the histomorphometric findings in children with various forms of OI will be published elsewhere.

Informed consent was obtained in each instance from the subject and/or legal guardian. The Ethics Committees of the participating institutions (Shriners Hospital and DuPont Institute) approved the study protocol.

Bone biopsy and histomorphometry

Transiliac bone biopsy specimens were collected on the fourth or fifth day after dual labeling with demeclocycline (Declomycin, Wyeth-Ayerst Montréal, Canada; 15–20 mg/kg per day, taken orally during two 2-day periods separated by a 10-day free interval). Labeling was successful in 7 IJO patients, 11 control subjects, and all 11 OI patients. Biopsy preparation and histomorphometric analyses with the exception of wall thickness (W.Th) measurements were performed as described previously.(12)

W.Th was determined at currently active formation sites as described by Steiniche et al.(13) This method uses the basic assumption that matrix apposition is completed at locations where very thin osteoid seams overlie very thick walls, which are not yet completed. In addition, reconstruction of the formation site is possible with this technique.(13,14) These analyses were performed on Goldner stained sections. The abrupt change in collagen fiber orientation, as seen under polarized light, was taken as the basis of the wall.

Three additional parameters were derived from the reconstruction of the formative site (see Fig. 1): (1) “preminer-alization osteoid thickness” (O.Th), which is defined here as the mean O.Th at sites where mineralization had not yet begun; (2) “initial mineralization lag time,” which indicates the interval between the occurrence of the first signs of osteoid and the appearance of the first mineralized bone (initial mineralization lag time was calculated as described)(13); and (3) “initial osteoid apposition rate,” defined here as the mean rate of matrix deposition during the initial mineralization lag time. This was calculated as the ratio between premineralization O.Th and the initial mineralization lag time divided by two.

Figure Fig. 1..

Reconstruction of the formative site in controls (n = 11) and IJO patients (n = 7). Values on the time axis represent geometric means; values on the ordinate are arithmetic means. The y value of the first data point on the upper line corresponds to premineralization O.Th. The slope of the line from the origin to this first point represents the initial osteoid apposition rate. The × value of the first data point on the lower line represents the initial mineralization lag time. The y value of the last point of both lines represents W.Th. uW.Th = uncompleted wall thickness

Statistical analysis

Results are given as arithmetic mean and SD. Geometric means and geometric SD were calculated for skewed parameters. Differences between IJO patients and controls or OI patients were tested for significance using Student's unpaired t-test. Skewed parameters were compared by t-tests after logarithmic transformation. All tests were two-tailed, and a 5% significance level was maintained. These calculations were performed using the SPSS software (Statistical Package for Social Sciences by SSCP Inc., Chicago, IL, U.S.A.), version 6.0 for Windows.


Comparison with healthy controls

Figure 2 shows typical histological sections from an IJO patient, a child with OI, and a healthy control subject. The decreased amount of cancellous bone and the scarcity of osteoid seams in IJO are readily apparent.

Structural parameters

As expected, all parameters of cancellous bone structure differed significantly between IJO patients and controls (Table 1). Bone volume per tissue volume (BV/TV) was decreased markedly, because of a 34% reduction in trabecular thickness (Tb.Th) and a 37% lower trabecular number (Tb.N). Therefore, the bone surface to bone volume ratio (BS/BV) was much higher in IJO.

Bone formation

Parameters reflecting cancellous bone formation are given in Tables 2 and 3. O.Th was decreased in IJO patients compared with controls. All BS-based formation parameters tended to be lower in IJO patients, but the difference achieved statistical significance only for mineralizing surface (MS/BS; Table 2) and bone formation rate (BFR/BS; Table 3). BFR/BS was decreased to 38% of the value in controls.

The osteoid was covered adequately by osteoblasts (Ob.S/OS). Nevertheless, mineral apposition rate (MAR) and adjusted apposition rate (Aj.AR) were decreased (Table 2). Accordingly, BFR relative to osteoblast surface (BFR/BS) was diminished in IJO patients (Table 3).

Activation frequency (Ac.f) in IJO was 54% of the control result (Table 2). Because of high variability this difference was only of borderline significance (p = 0.08). Formation period (FP) was prolonged by about 50%, but again the difference to controls did not reach statistical significance. W.Th was clearly decreased in IJO.

Reconstruction of the formative site in IJO patients and controls allows a visual appreciation of the differences in W.Th and FP (Fig. 1). As described in the methods section, these data also were used to calculate premineralization O.Th and initial osteoid apposition rate. Premineralization O.Th was lower in IJO patients than in controls (mean [SD]: IJO, 5.75 μm [0.64 μm]; controls, 8.00 μm [1.44 μm]; p = 0.001). However, premineralization O.Th constituted a similar proportion of reconstructed W.Th in both groups (mean [SD]: IJO, 18.0% [4.0%]; controls, 17.6% [3.4%]; p = 0.81). Initial osteoid apposition rate in IJO patients was decreased to 38% of the control value (geometric means [geometric SD]: IJO, 0.73 μm/day [1.80 μm/day]; controls, 1.92 μm/day [1.90 μm/day]; p = 0.005).

Figure Fig. 2..

Typical histological sections of biopsy specimens from a (A and B) 10-year-old boy without bone disease, (C and D) an 11-year-old boy with OI type I, and (E and F) a 10-year-old girl with IJO. Original magnifications: A, C, and E, ×40; B, D, and F, ×250.

Bone resorption

Osteoclast number [N.Oc/Bone Perimeter (B.Pm)] and the percentage of BS covered by osteoclasts surface (Oc.S/BS) or displaying an eroded surface aspect (ES/BS) were not significantly different between IJO patients and controls (Table 4). However, mean values for N.Oc and Oc.S were about 30% lower in IJO patients, whereas erosion surface was slightly higher. Therefore, the ratio between osteoclast and eroded surfaces was significantly lower in IJO patients (Table 4).

Comparison with OI type I

Comparative inspection of histological sections of patients with IJO and OI type I suggests that trabeculae are similarly scarce in both conditions but that bone turnover is more active in OI (Fig. 2). Bone marrow cellularity appeared to be higher in a number of sections from OI patients, but this was not quantified in the present study. Histomorphometric evaluation revealed that IJO and OI patients had similarly low BV and Tb.N (Table 1). However, trabeculae were thinner in IJO, which also is reflected in the higher BS/BV ratio. Ac.f- and BS-based parameters of bone formation were markedly higher in OI (Tables 2 and 3). In particular, mean BFR/BS was three times higher in OI patients (Table 3). The difference in the yearly turnover of cancellous bone tissue (BFR/BV) was somewhat smaller, because of the higher BS/BV ratio in IJO (Tables 1 and 3). MAR was similar in both patient groups, whereas Aj.AR tended to be higher in OI (Table 2). This was a result of a similar trend in the percentage of osteoid displaying mineralization (MS/OS). Walls were significantly thicker in OI. All osteoclast indices were significantly higher in OI, while ES was similar in both groups (Table 4).


This is the first report comparing static and dynamic histomorphometric findings in IJO patients and age-matched controls. Several histomorphometric reports on IJO have been published but were limited to static methods to quantify bone metabolism, described single cases, or did not have adequate control groups.(15–19) No conclusive picture has emerged from these reports. Schematically, either “increased bone resorption” or “decreased bone formation” were incriminated as causing IJO.

Biochemical studies of bone metabolism have not detected any consistent abnormality in children with IJO.(4,8,9) One study found a normal rise in serum osteocalcin in 6 IJO patients after calcitriol was administered orally, which was postulated to indicate “normal osteoblast function.”(10) However, the fact that osteoblasts in this test released normal amounts of osteocalcin into the circulation does not necessarily mean that they also deposited matrix on the bone surface in a normal fashion.

Table Table 2.. Bone Formation Parameters
 IJOControlsOIp (C)p (OI)
  1. Values are mean (SD).

  2. p(C), significance of difference between IJO patients and controls; p(OI), significance of difference between IJO patients and OI patients.

  3. a Values are geometric mean (geometric SD).

O.Th (μm)4.2 (1.0)6.8 (1.5)5.5 (1.5)<0.00010.05
OS/BS (%)23 (6)30 (11)45 (15)0.070.0006
OV/BV (%)2.1 (0.8)2.9 (1.0)4.4 (2.4)0.060.01
MS/BS (%)7.9 (4.2)15.1 (4.2)22.1 (8.5)0.0030.0009
Ob.S/BS (%)6.4 (3.6)9.2 (4.1)21.0 (7.3)0.11<0.0001
Ob.S/OS (%)28 (15)31 (14)47 (15)0.670.02
MS/OS (%)a29 (1.81)49 (1.41)48 (1.34)0.030.07
MAR (μm/d)0.76 (0.13)0.92 (0.07)0.77 (0.18)0.0030.86
Aj.AR (μm/d)a0.22 (2.08)0.45 (1.37)0.36 (1.31)0.030.10
Mlt (days)22.2 (10.2)14.9 (3.7)15.2 (4.2)0.110.13
Ac.f (per year)a0.58 (2.15)1.08 (1.37)1.55 (1.43)0.080.01
FP (days)a151 (2.2)101 (1.5)107 (1.38)0.230.29
W.Th (μm)32.4 (3.6)46.2 (5.5)39.2 (4.4)<0.00010.002
Table Table 3.. Bone Formation Rate Using Different Referents
 IJOControlsOIp (C)p (OI)
  1. Values are mean (SD).

  2. p(C), significance of difference between IJO patients and controls; p(OI), significance of difference between IJO patients and OI patients.

  3. a Values are geometric mean (geometric SD).

BFR/BS (μm3/μm2 per year)a18.8 (2.02)49.1 (1.31)56.5 (1.50)0.010.0006
BFR/Ob.S (μm3/μm2 per year)a338 (1.24)624 (1.82)297 (1.16)0.0080.15
BFR/BV (%/year)48 (34)72 (22)101 (42)0.080.01
BFR/TV (%/year)4.9 (3.4)17.6 (6.9)12.2 (7.3)0.00040.03
Table Table 4.. Bone Resorption Parameters
 IJOControlsOIp (C)p (OI)
  1. Values are mean (SD).

  2. p(C), significance of difference between IJO patients and controls; p(OI), significance of difference between IJO patients and OI patients.

  3. a Values are geometric mean (geometric SD).

N.Oc/BPm (per mm)a0.20 (2.04)0.30 (1.68)0.45 (1.69)0.160.01
Oc.S/BS (%)a0.71 (1.93)0.99 (1.78)1.33 (1.90)0.230.05
ES/BS (%)17.3 (7.1)16.3 (3.5)18.5 (8.1)0.700.72
Oc.S/ES (%)4.8 (1.7)6.8 (2.4)8.3 (3.6)0.050.02

Defects in bone formation

The present study shows that IJO is characterized by defective bone formation, as BFR/BS was decreased to 38% of the value found in controls. The total rate of bone formation depends on the number of new osteoblast teams recruited per unit time(20) (Fig. 3). The best estimate of this quantity is activation frequency, which in IJO patients was decreased to 54% of the control value. The second determinant of BFR is the average volume of bone matrix made by individual teams of osteoblasts, which is reflected by mean W.Th.(21) Mean W.Th in IJO patients was decreased to 70% of control values. Thus, in IJO fewer osteoblast teams are recruited and the individual team performs less than is normal.

Figure Fig. 3..

Summary of bone formation defects in IJO. In bold print physiological quantities and their direction of change in IJO. In parenthesis, the histomorphometric parameter that was used to estimate each physiological quantity. Each mean value is given as a percentage of the mean in controls.

Our data additionally provide some insight into what is wrong with the individual osteoblast team. The performance of the team depends on the number of team members as well as the amount of work done by each member.(20) The number of osteoblasts recruited per team is reflected by the fraction of osteoid covered by osteoblasts.(22) In IJO patients this fraction was 91% of the control value, no significant difference. This suggests that the insufficient osteoblast team performance in IJO is not a consequence of deficient osteoblast recruitment, as it is in postmenopausal osteoporosis.(22)

The volume of matrix deposited by osteoblasts is the product of their active life span and the average rate of matrix deposition during this period (Fig. 3). The average rate of matrix production is reflected by the bone formation rate per osteoblast surface.(22) In IJO patients this was decreased to 54% of the control value. This low work rate was partially compensated by a trend toward a longer active life span of the osteoblasts. As mentioned above, the integrated effect of these three factors was a 30% reduction in osteoblast team performance (Fig. 3).

Reconstruction of the formative site showed that the osteoid apposition rate of IJO patients is already very much decreased during the first few days of osteoblast activity at a new formation site. This suggests that the overall performance of the osteoblast team is not weak because the cells “are running out of steam.” Rather, the osteoblast team appears to be headed for a lower target right from the start, with accordingly scaled down intermediate steps. This observation also is supported by the finding that premineralization O.Th was lower in IJO patients than in controls but constituted the same percentage of final team performance (W.Th).

The role of bone resorption

Bone resorption was studied in less detail than formation. Given the low Ac.f of bone remodeling in IJO, a decrease in all of the resorption indices measured in this study is to be expected. Indeed, N.Oc and Oc.S were about 30% lower than in controls. ES was unique among the resorption parameters in that it was relatively high, given the low bone turnover. The “erosive appearance” of many sections may have misled some of the earlier investigators to propose increased bone resorption as the primary cause of IJO.(15,17) However, it is unlikely that this is indicative of higher bone resorption activity, because a smaller than normal percentage of the ES was covered by osteoclasts. Therefore, the relatively high ES probably reflects a prolongation of the reversal phase.

If bone resorption was not increased in IJO, why then were Tb.Th and Tb.N so markedly decreased? A decreased performance of the individual osteoblast team will result in a negative bone balance per remodeling unit, if osteoclast erosion depth remains unchanged. No attempt was made to estimate erosion depth in this study. However, an imbalance between osteoclast erosion and osteoblast refilling must be postulated to explain the markedly decreased Tb.Th in IJO patients.

The role of deficient trabecular production

A negative balance within individual remodeling sites would have even more devastating effects if it also occurred in the metaphysis, because the turnover of metaphyseal bone is much faster than that of mature cancellous bone. Low Tb.N in children with IJO might not only be caused by removal of mature trabeculae, but may also be caused by insufficient production of secondary trabeculae in the metaphysis. This could explain the radiological observation that the decrease in metaphyseal cancellous bone mass is most severe in newly formed bone (“neo-osseous osteoporosis”).

Comparison of IJO and OI

IJO may be difficult to distinguish from OI type I by noninvasive means. Bone biopsy may clarify the diagnosis, because there are typical histological and histomorphometric differences between the two entities. Microscopically, a “lack of activity” usually is noted in IJO, while there is “hypercellularity” in OI. In histomorphometric terms, this translates into low Ac.f- and BS-based remodeling parameters in IJO and an increase in these values in OI. This suggests that there is a fundamental difference in the pathophysiological basis of “typical” IJO and OI type I.


IJO is characterized by a 2-fold dysfunction of cancellous bone formation. First, fewer remodeling cycles are initiated. This will not be detrimental to bone integrity in the short run, as the recruitment of osteoblast and osteoclast teams is similarly decreased. Second, the amount of bone formed in each remodeling cycle is decreased markedly. The consequences are thinning of mature trabeculae and possibly decreased production of secondary trabeculae in the metaphysis.

During growth the mechanical load on bones increases continuously, creating a constant need for adaptive changes.(23,24) Insufficient production of cancellous bone can be expected to first create weaknesses at locations where trabeculae are most needed to maintain bone stability, that is, the metaphyses of long bones and vertebrae. Mechanical strain eventually exceeds the fracture threshold, and fractures at these sites ensue.

This study focused on the analysis of cancellous bone structure and remodeling in IJO. Whether this disease also affects cortical bone modeling and remodeling remains to be investigated. Furthermore, it is becoming increasingly recognized that an intricate relationship exists between marrow stromal cells, blood vessels, and the cells that are active on the bone surface.(25,26) Further studies of the bone marrow composition in patients with IJO might therefore help to elucidate the etiology of the bone cell abnormalities described in the present report.


We thank Guy Charette for technical assistance with sample processing and Mark Lepik for art work. This study was supported by the Shriners of North America and by the Deutsche Forschungsgemeinschaft grant Ra 803/1–1 (F.R.).