|Home | About | Journals | Submit | Contact Us | Français|
SIR—Bone loss, the major determinant of hip fracture risk in the elderly , results from an increased rate of bone turnover  that will amplify the imbalance at the level of the individual remodelling unit. However, while the increase in bone turnover in elderly women results in part from secondary hyperparathyroidism , the mechanisms responsible for the age-related decline in bone formation have not yet been identified.
Numerous studies have compiled evidence that growth regulatory factors–including the insulin-like growth factors (IGFs, IGF-I and IGF-II)–play a critical role in the control of the cellular events involved in bone formation . IGFs in serum are bound to binding proteins (IGFBPs), including IGFBP-3, the main carrier of IGFs in the circulation , IGFBP-5, which modulates IGF actions on bone positively , and the inhibitory IGFBP-4 .
We previously observed a down-regulation of the serum stimulatory components of the IGF system in elderly women with a femoral neck fracture . To address further the potential impact of deficiency of the IGF system on skeletal integrity, we compared IGF system components in patients with different types of osteoporotic hip fracture. Compared with femoral neck fractures, trochanteric fractures are associated with a more severe degree of osteoporosis . The aim of the current study was to test the hypothesis that differences in skeletal fragility between both hip fracture types might be related to differences in exposure to stimulatory and/or inhibitory IGF system components.
The study population consisted of 69 patients with hip fracture, including 50 femoral neck fractures and 19 intertrochanteric fractures. Details of the recruitment of the femoral neck (cervical) fracture patients have been reported previously . To be eligible for participation, women had to be over 60 years of age and to have suffered a fall resulting in a first femoral neck fracture treated by prosthetic replacement with a hemiarthroplasty or trochanteric fracture treated by internal fixation with a sliding screw. Patients were excluded in the case of haemodynamic instability and/or need for intravenous fluids or blood products prior to surgery; non-osteoporotic metabolic bone disease; diabetes or thyroid disease, whether controlled or uncontrolled; current use of calcium, vitamin D glucocorticoids or osteoporosis treatments.
An age-matched sample of 20 control subjects was randomly selected from a previously reported study population of 245 healthy elderly women over 70 years of age . Control subjects were non-institutionalised, functionally independent, and free from diseases known to affect the musculoskeletal system or the somatotrophic axis.
Fasting blood samples were obtained within 18 hours after trauma and before surgical treatment. Circulating 25(OH)D, intact PTH, osteocalcin, skeletal alkaline phosphatase and pyridinium crosslinks (pyridinoline and deoxypyridinoline) were measured as previously described . IGF-I and IGF-II were measured by radioimmunoassays (RIAs) after separation of IGFBPs by a rapid acid gel filtration protocol . Serum IGFBP-3, IGFBP-4 and IGFBP-5 were measured by RIAs [12, 13].
Within 10 days after fracture, patients were scanned by dual-energy X-ray absorptiometry (DXA) (Lunar DPX-L, Lunar Radiation Inc., Madison, WI) .
All P values are based on Student’s t-test for differences between groups and on age-adjusted (partial) Pearson correlation coefficients (rp) for associations between parameters among fracture patients, following logarithmic transformations when appropriate, to obtain a normal Gaussian distribution. Data were re-analysed using non-parametric tests (Wilcoxon matched pairs signed rank sum test and Spearman rank correlations), which yielded the same results (not shown).
Data from the femoral neck fracture patients have been published , including a comparison with a control group. This paper extends these observations with data from trochanteric fracture patients, sampled and analysed within the same batches as the original patients and controls.
A comparison of both patient groups is shown in Table 1. Trochanteric fracture patients were older (P =0.01). No differences were observed for height, weight, and body mass index between the groups. Trochanteric bone mineral density (BMD) was 14% lower (P =0.01) in women with trochanteric fracture. Femoral neck BMD, vitamin D, PTH, serum IGF-I and IGFBP-3 were similar in both groups, but trochanteric fracture patients had lower levels of IGF-II and IGFBP-5 and increased concentrations of IGFBP-4 (P < 0.001). Compared with (younger) controls , trochanteric patients showed a deficit in IGF-I, IGF-II, IGFBP-3 and IGFBP-5 and an increase in IGFBP-4 (P < 0.001 for all comparisons, not shown).
To further address the significance of the differences in IGF regulatory system components between both fracture types, the 19 women with trochanteric fracture were further compared with age-matched subsets of 20 individuals from the published [8, 10] control and cervical fracture populations (Table 3), again showing a deficit in serum stimulatory components and increased levels of the inhibitory IGFBP-4 (Tables 2 and and3).3). Similar results were obtained when using a different statistical approach, i.e. age-adjusted analyses. After adjustment for age, serum IGF-II, IGFBP-4, and IGFBP-5 in trochanteric fracture patients remained statistically different from those in cervical fracture patients (n = 50) or controls (n = 245), with a P value <0.01 for all comparisons (data not shown).
Considering the hip fracture population as a whole (n = 69), significant age-adjusted correlations were observed between trochanteric BMD and the binding proteins IGFBP-5 and IGFBP-4 (rp =0.39, P =0.007 and rp = −0.29, P =0.02, respectively). Similar relationships were statistically not significant when femoral neck and trochanteric fracture patients were analysed separately (data not shown), emphasising the exploratory character of the analyses on pooled data. Femoral neck BMD was statistically unrelated to any of the IGF system components (not shown).
Trochanteric fractures have been associated with up to double the short-term mortality of cervical fractures . While fall characteristics have been shown not to be associated with fracture type , previous studies have reported differences in the characteristics of patients who sustain both fracture types with respect to age, presence of other skeletal fractures, and bone density, suggesting that femoral neck and trochanteric fractures may be caused by different aetiologies . In particular, trochanteric fracture may involve a greater degree of trabecular osteopaenia. In this regard, our bone density measurements are in agreement with previous studies reporting lower densities in trochanteric fractures compared with cervical fractures [9, 16]. The greatest reduction has been observed in the trochanteric region , an area of predominantly trabecular bone. In line with these densitometric findings, vertebral fractures are twice as common in women with trochanteric fractures than in those who have sustained cervical fractures .
The findings of the present study–a significant deficit in systemic IGF-II and IGFBP-5 and increased circulating levels of IGFBP-4–suggest that differences in the degree of trabecular osteopaenia between both fracture types could partly be related to differences in exposure to stimulatory and inhibitory IGF components.
IGFBPs are important in presenting IGF to its receptor and modulating the activity of IGFs by binding the biologically active free IGF . While IGFBP-5 enhances the mitogenic potential of IGFs added to osteoblasts , IGFBP-4 inhibits IGF actions on bone cells , suggesting that the balance between the stimulatory and inhibitory classes of IGFBPs will determine the degree and extent of IGF-induced cellular responses in target tissues such as bone. In addition, IGFs are fixed in bone by means of IGFBP-5, which binds with high affinity to both hydroxyapatite and IGFs . The fact that IGF-II is the most abundant growth factor in human bone (and the deficit in IGF-II observed in this study) may be related to the higher affinity of IGFBP-5 for IGF-II than IGF-I.
Cross-sectional studies have shown a positive association with circulating IGF-I and BMD of the hip in elderly individuals [10, 19]. More recently, longitudinal data in postmenopausal women have confirmed that low serum concentrations of IGF-I are associated with femoral bone loss  and an increased risk of osteoporotic fractures . Overall, these studies and our findings support the concept that the IGF system contributes to the preservation of bone mass with ageing.
The strengths of our study include the recruitment of well-defined patient populations, the use of well-validated assays to measure IGF(BP)s, and the timing of the sampling. We excluded patients with haemodynamic instability and/or need for intravenous fluids or blood products prior to surgery, and all patients were sampled before surgical treatment, within 18 hours after fracture. Within this time frame, there is no fracture-induced down-regulation of the stimulatory IGF components .
There are several potential limitations to our analysis which may affect the inferences derived from these data, however. In particular, the extent to which the serum levels of the IGF components reflect their local skeletal activity cannot be determined from our data. Furthermore, although the hip fracture patients were frail, a number of exclusion criteria were applied and those included in this study represented a relatively healthy subset of all persons sustaining hip fracture; thus, our results may not be generalisable to all hip fracture patients. Finally, and most importantly, the data are observational and descriptive: cause and effect associations cannot be inferred and will require confirmation from future studies.
We conclude that the exposure to stimulatory and inhibitory components of the IGF system is different between femoral neck and trochanteric fractures. The extent to which these differences reflect a partially different patho-physiology between the two types of hip fractures remains to be further clarified.
This study was supported by grant G.0171.03 from the Fund for Scientific Research-Flanders, Belgium (F.W.O.-Vlaanderen). S.B. and D.V. are senior clinical investigators of the Fund for Scientific Research-Flanders, Belgium (F.W.O.-Vlaanderen). S.B. is holder of the uncommitted Leuven University Chair in Metabolic Bone Diseases.