This study investigated the effects of heparin-induced osteoporosis on the alignment of the developing spine in a bipedal rat model. Results demonstrated that the incidence of spinal malalignment at week 40 is not significantly different between the control group, which did not receive any medication, and the low bone mass group, which received daily subcutaneous heparin injections. DEXA scan results and histomorphometric analyses of spinal segments clearly show that daily heparin injections prove useful to decrease the trabecular bone density.
The death of experimental rats (six from heparin, two from control groups) decreased the number of specimens which may have decreased the statistical power of the findings. This may have contributed to the insignificance of the curve magnitudes between the groups.
Heparin, in addition to its anticoagulant effects, is of interest to orthopaedic surgeons because of its effects on bone healing and density. Since the 1960s, multiple clinical reports linked long term heparin use to osteoporosis [24–27]. There has been extensive research on topic and this relationship has been proven with clinical and laboratory studies on rats [19–22, 28, 29]. These studies showed an increase in osteoclast surface area and a decrease in osteoblast surface area thereby suggesting that Heparin causes osteoporosis by both increasing bone resorption and decreasing bone formation. Daily subcutaneous injections of Heparin causes 30% bone loss in rats at 28 days and continues even after heparin injections have been withdrawn. This effect of continued bone loss been linked to the sequestration of heparin molecules in bone tissue . Daily heparin injections have minimal morbidity and proven efficacy in creating decreased bone mineral density. Therefore, in order to avoid the morbidity of additional surgery, we opted to administer daily heparin injections instead of ooferectomy. Our results are concordant with the available literature utilizing subcutaneous heparin injections.
Osteoporosis has previously been reported as a causative or a coexisting factor in idiopathic scoliosis patients in several studies [8, 9, 11, 15, 16, 31, 32]. These studies used DEXA scans and were able to show differences between scoliotic and age matched control groups. A recent study by Park et al.  looked into the osteogenic and adipogenic abilities of mesenchymal stem cells in 19 girls with AIS and 16 age- and sex-matched healthy control subjects. In children with AIS, the osteogenic differentiation ability of mesenchymal stem cell were lower when compared with controls whereas the adipogenic ability was not significantly different between groups. On the other hand, Szalay et al.  argued that the decreased Z scores in DEXA scans were related to the low body mass index of children with AIS and may not be related to scoliosis. They suggested that the decreased BMD in scoliosis patients is due to the lower BMI of these children and they mentioned that in their series scoliotic patients of normal and heavy weight demonstrated Z scores only slightly lower than the controls. However, their argument does not refute the theory that low BMD may be a contributing factor in development of scoliosis in children. However, currently available data strongly suggests a relationship between osteoporosis and AIS .
One major limitation of the current study is about the definition of 'osteoporosis' in rats. According to WHO criteria [35, 36] osteoporosis is defined as a BMD value more than 2.5 standard deviations below the mean BMD of young adult women (BMD T-score < -2.5), and osteopenia (low bone mass) is defined as a BMD value between 1 and 2.5 standard deviations below the mean BMD of young adult women (-2.5 < BMD T-score < -1). The only study that defines osteoporosis in rats is by Sristava et al.  but the definition was only for categorical purposes and was not able to evaluate fracture risk. Therefore we chose not to use the term 'osteoporosis' but instead 'low bone mass', 'decreased bone mineral density' or 'osteopenia' to denote statistically significant differences. Heparin injections did cause a statistically significant decrease in BMD of the rats in all body regions. It is not known, however, whether this decrease is enough to cause clinically relevant problems, i.e. 'osteoporosis'. Another shortcoming of our study is that we did not use a placebo or sham injection group. We do not believe that subcutaneous injections of a placebo would be any different than no intervention at all, however, technically this is still a shortcoming of the study design. Also, it is of significance that the curve magnitudes are small and yet standard deviations are large which, makes the interpretation of the results difficult.
With the assumption that there is a relationship between osteoporosis and idiopathic scoliosis, one logical explanation could be that there is a treshold level of BMD needed for development of deformity and the BMD of our rats may have stayed below this threshold within the osteopenia region. Another argument could be that 40 weeks is not enough for the negative effects of osteoporosis to cause a significant change in vertebral bodies. Also, administration of heparin may have effects currently unknown to us. As an example, during the review process of this paper we have been informed by one of the reviewers that heparin has been shown to interfere with melatonin binding and signal transduction in chick brain . This interaction has not yet been shown in rats, however the same effect maybe confounding our data. For this reason, we believe that another method to induce osteopenia, such as ooferectomy should be used to induce a more profound osteopenia which would also avoid the potential effects of a medication on the curve formation.
Our data implies that bipedality may be the major factor for the development of scoliosis in bipedal animal models and low bone mass may not significantly contribute to the degree of deformity in the developing spine. Yet, bipedality alone as a factor for the development of scoliosis is not supported by the literature. There are at least two other studies showing that rats that were rendered bipedal with a protocol very similar to the one used in this study did not develop scoliosis in the absence of pinealectomy [17, 18]. It should be noted that these studies did not specify how they defined scoliosis. The disconcordance between these studies and the current study may be related to the differences in reporting threshold of the spinal deformity or simply due to differences in the protocols, such as the size of the cages, the amount of time animals spent on two hindlimbs, et cetera. The relationship of bipedality alone as an etiologic factor for spinal deformity should be further scrutinized with future studies.
Our study did not show a statistically significant relationship between low mineral density and development of scoliosis in a bipedal rat model. Another interesting finding of this study was that in heparin treated animals, a relative thoracic hypokyphosis was present. Considering that idiopathic scoliosis is a three dimensional deformity presenting with hypokyphosis in the sagittal plane, we believe that the relative hypokyphosis in the heparin treated animals maybe of some value. With these findings it is not possible to prove or refute the long questioned relationship between osteoporosis and idiopathic scoliosis. Further studies are required to clarify the relationship between osteoporosis and development of spinal deformity in the growing spine.