|Home | About | Journals | Submit | Contact Us | Français|
As the prevalence of diabetes is increasing worldwide, research on some of the lesser-known effects, including impaired bone health, are gaining a lot of attention. The two most common forms of diabetes are type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM). These two differ in their physiology, with T1DM stemming from an inability to produce insulin, and T2DM involving an insufficient response to the insulin that is produced. This review aims to highlight the most current information regarding diabetes as it relates to bone health. It looks at biochemical changes that characterize diabetic bone; notably increased adiposity, altered bone metabolism, and variations in bone mineral density (BMD). Then several hypotheses are analyzed, concerning how these changes may be detrimental to the highly orchestrated processes that are involved in bone formation and turnover, and ultimately result in the distinguishing features of diabetic bone. The review proceeds by explaining the effects of antidiabetes medications on bone health, then highlighting several ways that diabetes can play a part in other clinical treatment outcomes. With diabetes negatively affecting bone health and creating other clinical problems, and its treatment options potentiating these effects, physicians should consider the use of anti-osteoporotic drugs to supplement standard anti-diabetes medications in patients suffering with diabetic bone loss.
Diabetes mellitus (DM) is a disease with extreme and increasing prevalence. The World Health Organization estimates that in 2014, 9% of adults (>18 years old) had diabetes, with the number of new cases of diabetes for adults in the United States (US) having tripled since 1980. Due to the high impact of diabetes on mortality and morbidity, it has a significant economic burden, costing the US hundreds of billions of dollars each year as reported by the Centers for Disease Control and Prevention. The treatment of the disease itself is a large part of this burden, but there are several comorbidities that also contribute. These comorbidities include both macrovascular and microvascular complications like stroke, coronary artery disease, neuropathy, and peripheral vascular disease. Interestingly, diabetes also seems to have an effect on bone health. There are several changes in type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) diabetic bone regarding bone strength, bone turnover, and stem cell differentiation that result in altered bone mineral density (BMD) and bone structure. One outcome that is similar to both types of diabetes is that mesenchymal stem cells in the marrow show preferential differentiation towards adipocytes compared with osteoblasts [Botolin et al. 2005; Botolin and Mccabe, 2007]. As such, increased marrow adipogenesis has been observed with long standing diabetes in both type 1 and type 2 diabetic models. In regards to other changes in bone health, the two types of diabetes differ in several areas with many mechanisms yet to be worked out. There is research on several different aspects of the relationship between diabetes and bone health, but there are not many pieces that compile this information for appropriate review. This perspective review attempts to address this disparity in the hope of shedding light on ways to better treat diabetic patients in the clinical setting.
Hyperglycemia in patients with T1DM is the root of many of the disease’s complications. Patients with diabetes are shown to have accumulations of harmful advanced glycation end products (AGEs) [Santana et al. 2003]. These can form when sugars react nonenzymatically with proteins or lipids, and becomes more likely to happen in patients with poor glucose utilization. Interestingly, these AGEs were shown to cause apoptosis of mesenchymal stem cells in humans, preventing the differentiation of osteoblasts, adipocytes, and cartilage [Yamagishi et al. 2005]. Osteoblastic synthesis of the protein osteocalcin was shown to be inhibited by high serum glucose concentrations, resulting in poor bone formation after sustained exposure to this condition [Inaba et al. 1995]. Osteoblastic expression of osteocalcin, matrix metallopeptidase-13 (MMP-13), vascular endothelial growth factor and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was also shown to be down regulated in T1D mice, concluding that chronic hyperglycemia affects all stages of osteoblast maturation [Botolin and Mccabe, 2006]. Another study linked AGEs to increased osteoclastic activity, which would perpetuate bone loss [Zhou et al. 2006]. Based on this information it is not surprising then that nearly 20% of T1DM patients between the ages of 20–56 years meet the criteria for being osteoporotic.
T1DM appears to affect the microarchitecture of bone by decreasing BMD. Changes in the mesenchymal progenitor cells found in the marrow of patients with T1DM favor adipogenesis. Additionally, several studies, including a meta-analysis from Vestergaard, show that BMD is reduced in patients with T1DM [Vestergaard, 2007]. It is hypothesized that this is related to differential expression of growth factors like insulin-like growth factor 1 (IGF-1) and transforming growth factor β1 (TGF-β1). For example, streptozocin-induced diabetic mice, mimicking T1DM, displayed reduced expression of IGF-1, IGF-1 receptors and insulin receptors that were accompanied by decreased bone growth [Lu et al. 2003]. This corroborates other studies linking patients with T1DM to lowered BMD.
The bone of patients with T2DM has an increased BMD, while those with T1DM show lowered BMD. A meta-analysis by Ma and colleagues including 15 observational studies (3437 patients with T1DM and T2DM and 19,139 controls), concluded that BMD was significantly higher in patients with T2DM than in patients without T1DM or T2DM when considering the femoral neck, hip and spine [Ma et al. 2012]. While this may seem like a positive outcome, the higher BMD is reported in conjunction with decreased overall bone turnover. Results from several studies showed decreased expression of bone turnover markers osteocalcin, type 1 cross-linked C-telopeptide (CTX-1), and type 1 cross-linked N-telopeptide (NTX) [Akin et al. 2003; Starup-Linde, 2013]. This lack of bone turnover likely has effects on bone health and may be related to increased BMD.
Hyperinsulinemia as seen in patients with T2DM could also be responsible for increased BMD. Insulin can interact with the IGF-1 receptor present on osteoblasts due to its structural homology. IGF-1 has been shown to have a strong positive correlation with BMD, in human and mouse models [Langlois et al. 1998; Zhao et al. 2000]. Based on this, it is plausible that hyperinsulinemia could create the effects similar to IGF-1 by promoting osteoblast division and differentiation (Figure 1 following this section). Hyperinsulinemia may also be responsible for the increase in adipogenesis seen for T2DM patients within the bone marrow stem cell niche via a similar mechanism noted for hyperlipidemia that involves a signaling cascade via protein kinase B (PKB) mammalian target of rapamycin (mTOR) and subsequent activation of CCAAT/enhancer-binding protein α (C/EBPα) and peroxisome proliferator-activated receptor (PPAR)-γ [Piccinin and Khan, 2014].
Obesity, defined as a body mass index (BMI) > 30, has been shown to be closely associated with T2DM and in a follow-up study of 2204 women, BMI was the number one predictor of T2DM [Colditz et al. 1995]. Additionally, obesity has been shown to have a positive correlation with BMD, and there are several hypotheses to explain this phenomenon. Adipose tissue releases several adipokines such as leptin, which has been found in higher amounts within the plasma of diabetic men than healthy controls [Fischer et al. 2002]. Plasma leptin has been shown to both inhibit osteoclastogenesis, via reduced production of RANK/RANK-ligand (RANK/L), while stimulating osteoblastogenesis thereby inducing bone growth [Holloway et al. 2002, Figure 1]. Additionally, it has been reported that higher adiponectin levels are correlated with lower BMD, and diabetic patients have reduced levels of adiponectin [Weyer et al. 2001; Gomez-Ambrosi and Fruhbeck, 2005; Richards et al. 2007].
Treatment strategies for diabetes include exogenous insulin, insulin-sensitizing drugs, drugs that increase secretion of insulin (secretagogues), and incretin mimetics, which mimic the action of incretin hormones found in the gut [Hinnen et al. 2006]. The absolute insulin deficiency in T1DM limits the effective treatment methods. The main means of management for patients with T1DM include exogenous insulin, blood glucose monitoring, and meal timing and regulation [Daneman, 2006]. In regards to potential side effects, patients with T1DM and T2DM should preferably be treated with medications that have positive or neutral effects on bone.
The relative insulin deficiency in T2DM allows for more target treatment methods. These treatments produce certain beneficial outcomes but have some undesirable side effects like weight gain and further diminished bone health. Most often the first-line treatment method for patients with T2DM is metformin. Metformin’s primary role is increasing insulin sensitivity by decreasing the production of hepatic glucose via inhibition of the mitochondrial respiratory-chain complex 1, while stimulating glucose uptake in muscles [Viollet et al. 2012]. Metformin has also been shown, in vitro and in vivo to promote osteogenesis which could also account for increased BMD in these patients [Molinuevo et al. 2010]. Therefore, metformin proves to be an effective method in regards to both glucose utilization and bone health.
If metformin administration does not allow patients to reach desired blood glucose levels, several other treatment options may be considered. Thiazolidinediones (TZDs), like rosiglitazone and pioglitazone, work by increasing the body’s sensitivity to insulin [Chen et al. 2015]. TZDs accomplish this through agonistic effects on the nuclear receptor PPAR-γ, which is predominantly found in adipose tissue and implicated in the regulation of insulin sensitivity [Hoffmann et al. 2012; Chen et al. 2015]. Activated PPAR-γ facilitates insulin sensitivity and adipocyte differentiation, which increases fatty acid storage and decreases fatty acids in circulation. This potently induces a desired anti-diabetic effect shown to be more effective than other treatments, but with the cost of other serious negative side effects [Meier et al. 2016]. A majority of studies evaluating TZD use in diabetes points to its negative effects on bone. A study by Chen and colleagues including a cohort of 32,466 patients with T2DM from 2000 to 2010, concluded that there was an elevated risk of fracture in diabetic patients using TZD for an extended period, particularly women younger than 64 years old [Chen et al. 2015]. A longitudinal observational cohort study, using participants from the ACCORD study, concluded that TZD use exclusively in women, showed an increase in nonspine fractures, and fracture risk was mitigated by discontinued use of TZDs [Schwartz et al. 2015]. Another serious complication is risk for cardiovascular events, and thus rosiglitazone has been recommended for suspension by the European Medicines Agency and severely limited by the US FDA [Consoli and Formoso, 2013]. Therefore, despite the large potential benefits of TZD use in the treatment of T2DM, the complications they cause have removed them from predominance.
Incretin-based therapies are another popular second-line treatment option for patients with T2DM. Glucagon-like peptide 1 (GLP-1), an incretin mimetic, is a hormone derived from the gut that stimulates the production of insulin while inhibiting glucagon secretion. Additionally, it inhibits gastric emptying and food intake by reducing appetite [Drucker and Nauck, 2006]. GLP-1 works on B-islet cells, as well as those in the kidney, heart, and bone [Campbell and Drucker, 2013]. A study of ovariectomized (OVX) mice determined that the use of GLP-1 receptor agonists (GLP-1RAs) improved trabecular bone mass and architecture concluding that GLP-1RAs could be a good treatment therapy for patients with T1DM and T2DM, particularly post-menopausal women [Pereira et al. 2015]. However, these findings could be contradicted if results similar to those seen in Burghardt and colleagues’ study occurred, which found increased trabecular BMD, but increased pore volume and altered bone distribution, which resulted in a compromised bone structure [Burghardt et al. 2010]. Dipeptidyl peptidase-4 (DDP-4) is a compound that normally degrades incretin hormones. DDP-4 inhibitors have been developed for diabetes patients to impede the actions of this peptidase, slowing the degradation of GLP-1 and therefore increasing its action [Hinnen et al. 2006]. These DDP-4 inhibitors have shown little-to-no effect on bone health and fracture risk in patients using them [Hinnen et al. 2006].
Sodium–glucose co-transporter 2 (SGLT2) inhibitors are a recently introduced class of T2DM medication that act to block reabsorption of glucose in the kidney. These have been shown to be effective in enhancing glycemic control, as well as reducing blood pressure and body weight [Haas et al. 2014]. Through a similar mechanism in the kidney, these inhibitors likely increase serum phosphate levels, stimulating parathyroid hormone (PTH) secretion [Haas et al. 2014]. PTH in turn facilitates bone resorption by affecting osteocyte signaling [Haas et al. 2014]. Data from several sources have shown that SGLT2 inhibitors potentiate overall bone turnover, shown by increased markers for both resorption and formation [Haas et al. 2014]. More studies are ultimately needed to work out negative effects of these inhibitors on bone health.
Sulfonylureas are another effective drug treatment strategy, and are insulin secretagogues. They bind to the sulfonylurea receptor subunits of potassium channels, causing these voltage-gated channels to close [Nelson and Cox, 2008]. In the pancreas, this results in membrane depolarization, which causes an influx of calcium and subsequent insulin release. Although there is only a small amount of data relating sulfonylureas to bone health, some studies have linked it to decreased fracture risk, especially compared with other treatments like TZDs [Meier et al. 2016].
One way to offset negative effects of diabetes and its treatments on bone health could be osteoporotic drugs that improve bone strength. Antiresorptive drugs such as bisphosphonates (specifically alendronate) for osteoporosis have been shown to be effective in some cases of diabetic bone loss [Ikeda et al. 2004; Coe et al. 2015]. Additionally, anabolic bone treatment should be considered. For example, PTH 1–34 (teriparatide), PTHrp, and IGF-1 have all been shown to stimulate bone formation when administered appropriately [Mosekilde et al. 2011]. A study by Poon and colleagues determined that a combination of vitamin K2 and 1,25(OH)2D3 was effective in the reappearance of surface microvilli and ruffles on osteoblasts in mice with T2DM [Poon et al. 2015]. Additionally, a study of type 1 diabetic mice by Motyl and colleagues determined that intermittent treatment of PTH reversed pre-existing bone loss through its anabolic effects on osteoblasts [Motyl et al. 2012].
Clinically, both diabetic populations incur an increase in fracture rate compared with the nondiabetic population. In order to investigate why patients with T2DM have an increased incidence of fracture despite having increased bone density, Burghart and colleagues used high-resolution peripheral quantitative computed tomographic imaging of cortical and trabecular bone microarchitecture of the distal tibia and radius in patients with T2DM [Burghardt et al. 2010]. As expected, results showed that patients with T2DM had higher tibial and radial BMD. They also showed a 10% increase in trabecular volumetric BMD bordering cortical bone and a greater trabecular thickness of the distal tibia in the T2DM cohort as compared with controls [Burghardt et al. 2010]. Interestingly, these changes were accompanied by significant increases in the porosity of the cortical bone in the radius as well as pore volume in the tibia [Burghardt et al. 2010]. These findings could account for the increased fracture rate associated with T2DM, despite these patients having greater BMD. This study postulates that improper redistribution of bone mass, notably, increased trabecular bone density with decreased intracortical bone, results in compromised bending load in patients with T2DM [Burghardt et al. 2010].
With the data showing BMD being associated with high fracture risk at both increased and decreased levels, it seems to suggest that altered BMD alone may not properly account for the increased fracture risk in patients with T1DM and T2DM. One possibility is that increased BMI, through an increased fall risk and altered testosterone levels, is the cause for increased fractures in T2DM while decreased BMD could be directly linked to increased fracture risk in T1DM [Nielson et al. 2011]. Another way of more properly analyzing increased fracture risk in patients with DM is accounting for its effects on microvasculature. Problems with microvasculature, including retinopathy and diabetic neuropathy, are a very common complication of diabetes. This has negative outcomes for bone, as shown by an increased risk of hip fracture being demonstrated in patients with ophthalmic, nephropathic, and neurological microvascular complications [Miao et al. 2005].
Patients with DM are not only at risk for fracture, but also at risk for complications during osseous healing. These patients typically show higher rates of delayed union and nonunion than their nondiabetic counterparts. In looking at patients undergoing arthrodesis, one study found a higher rate of nonunion in patients with non-neuropathic DM as compared with the overall study sample [Perlman and Thordarson, 1999]. A cohort study showed similar results for fracture healing in patients with non-neuropathic DM, with these patients experiencing significant delays in healing as compared with patients without DM [Loder, 1988]. Explanations for poor fracture healing have pointed to deficiencies in the production and localization of collagen and growth factors, resorption of cartilage in the callus of the healing fracture, and damaging glycated protein products [Kayal et al. 2007]. These have negative effects on health of the bone and hinder proper bone healing.
Diabetes has adverse effects on joint health, and predisposes patients to eventually needing surgical interventions on them. Surgical intervention can interrupt the proper management of blood glucose levels that patients with DM rely on, putting them at risk for postoperative complications. One of the complications is that patients are at higher risk for peri-prosthetic infection following joint replacement surgery. Jämsen and colleagues found that having DM more than doubled a patient’s risk of peri-prosthetic infection, and even nondiabetic patients with high preoperative blood glucose levels trended towards a higher infection rate [Jamsen et al. 2012]. Another study found that patients with DM were around four times as likely to experience postoperative infection following a total joint arthroplasty than those without it, with little correlation with hemoglobin A1c levels [Iorio et al. 2012]. In patients undergoing foot and ankle surgery, postoperative infections occurred in 13.2% of those with DM and only 2.8% of those without it [Wukich et al. 2010]. This high rate of postoperative infection for patients with DM puts them at risk for further adverse effects, potentially compounding the health problems they are already experiencing.
As evidenced earlier, there seems to be an intricate interplay between bone density and diabetes. In patients with T1DM this can often lead to osteopenia, with extreme cases being linked to osteoporosis [Takamoto and Kadowaki, 2004]. Conversely there is little-to-no evidence linking T2DM to these outcomes. Investigations into the molecular link between diabetes and osteoporosis show that on the cellular level T1DM has been shown to hinder bone formation, and even favor bone resorption [Hamann et al. 2012]. This facilitates bone loss in these patients and puts them at an increased risk for negative outcomes in relation to bone health.
Diabetes is a major health burden in the US, and has recently been shown to have complications relating to bone health. Studies have shown that several changes occur in diabetic bone. The bone in patients with both types of diabetes seems to show increased adipogenesis. Additionally, patients with T1DM present with decreased BMD while patients with T2DM present with increased BMD. This discrepancy is related to several type-specific factors including hyperglycemia, altered insulin levels, and increased obesity rates, all of which could impact bone turnover markers. Unfortunately, many treatment options that are useful in treating insulin resistance have negative side effects that compound the health issues already seen with diabetic bone.
These complications have an effect on other clinical outcomes, which makes it important to find a way to avoid them. Patients with both types of diabetes are prone to much higher fracture risk than healthy individuals. Diabetic patients that do experience fracture, frequently experience delayed union and nonunion. Postoperative infection rates are high in patients requiring bone-related surgical intervention, which further complicates the healing process. Many of these poor clinical outcomes could be combatted by improving bone health for these patients. Therefore, the benefits of treatment regiments including anti-bone resorptive drugs, such as alendronate, should be explored for patients experiencing diabetic bone loss. Bone anabolic therapies such as a combinations of vitamin K2 and 1,25(OH)2D3 for T2DM, and intermittent PTH treatment for T1DM have shown promise and should also be considered. Additionally, certain antidiabetic drugs, such as TZDs appears to have a worse impact on bone than other strategies like metformin, so these should generally be avoided until better alternatives are discovered. There is still much to learn regarding definitive mechanisms that create diabetic bone and its structure, so further investigations should be completed to evaluate the cause of increased fractures in T1DM and T2DM.
Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement: The authors declare that there is no conflict of interest.
Vikram Sundararaghavan, Department of Orthopaedic Surgery, University of Toledo Medical Center, Toledo, OH, USA.
Matthew M. Mazur, Department of Orthopaedic Surgery, University of Toledo Medical Center, Toledo, OH, USA.
Brad Evans, Department of Orthopaedic Surgery, University of Toledo Medical Center, Toledo, OH, USA.
Jiayong Liu, Department of Orthopaedic Surgery, University of Toledo Medical Center, 3065 Arlington Avenue, Toledo, OH 43614, USA.
Nabil A. Ebraheim, Department of Orthopaedic Surgery, University of Toledo Medical Center, Toledo, OH, USA.