Diabetes, diabetic medications, and risk of fracture


Diabetes mellitus is characterized by high levels of blood glucose; the main forms are type 1 diabetes (T1D), caused by a lack of insulin production due to an autoimmune process toward pancreas beta cell, and type 2 diabetes (T2D), mainly caused by insulin resistance. In the last Diabetes Atlas, published by the International Diabetes Foundation in 2017, it was estimated that almost 425 million people have diabetes worldwide and this number is going to increase up to 693 million by 2045 . Diabetes-related complications are the real burden of the disease, accounting for up to 60%–70% of health-care costs related to diabetes leading to a reduction of life expectancy and relevant decrease in the quality of life for patients affected by diabetes. Besides the well-known micro- and macrovascular complications of diabetes, the increased risk for fragility fractures has recently been recognized as an important complication of both T1D and T2D .

Fractures, bone turnover, bone morphology, and bone strength in diabetes mellitus

Type 1 diabetes and fracture risk

In the second decade of 1900, it was first suggested that T1D causes an increased prevalence of fragility fractures . Many years later, the Iowa Women’s Health Study, an 11-year follow-up of 32,089 postmenopausal women, was one of the first studies proving diabetes-induced bone fragility, with hip fractures being 12 times more common in T1D women compared to controls . Thereafter, a metaanalysis by Vestergaard confirmed the higher risk of hip fractures in T1D and T2D (although mostly among T1D) . The Nurses’ Health Study reported a sixfold increase in hip fracture risk and a twofold increase in vertebral fracture risk compared to control subjects . Similar results have also been reported for men. In a cohort of Norwegian subjects, men with T1D had a 17.8-fold increased risk of hip fractures in a 6-year follow-up period . Similar evidence was also obtained in the Tromsø study, a Swedish cohort of more than 24,000 patients with T1D, reporting an 8- to 12-fold increase in hip fracture risk in men with T1D . In 2007 two large metaanalyses conducted by Vestergaard and Janghorbani et al. reported 6.9- and 6.3-fold increase in hip fracture risk, respectively, in patients with T1D compared to subjects without diabetes.

Furthermore, in recent years, the increased fracture risk among T1D has been confirmed again. In one study conducted among 30-year-old patients with T1D, it was found that the prevalence of morphometric vertebral fractures was higher in patients with T1D (24%) compared to controls (6%) . In 2015 a study, based on a population from The Health Improvement Network (THIN) database in United Kingdom, analyzed fracture incidence among more than 30,000 T1D patients aged 0–89 years compared with more than 303,000 age- and sex-matched controls without diabetes, with an average follow-up of 4.7 years. Patients with T1D showed increased fracture risk through all ages compared to controls, starting from childhood. In both sexes the lowest hazard ratio (HR) for fracture in T1D was in those <20 years [men: 1.14; 95% confidence interval (CI) 1.01–1.29, women: 1.35; 95% CI 1.12–1.63]. The risk was highest in men 60–69 years of age (HR 2.18; 95% CI 1.79–2.65) and in women of 40–49 years (HR 2.03; 95% CI 1.73–2.39). In this study the greatest fracture incidence in men was aged between 10 and 20 years and in women between 80 and 90 years of age. Finally, the authors found that fractures at the lower extremity sites (hip/femur, lower leg/ankle, foot) were more common in patients with T1D compared to controls . Another population study, based on a national diabetes database in Scotland, further confirmed these findings. In particular, the incidence risk ratio (IRR) of hip fractures in men was 3.28 (95% CI; 2.52–4.26) and in women was 3.54 (95% CI; 2.75–4.57) compared with the general population. Moreover, increased IRR was observed at younger ages, but absolute risk difference was greatest at older ages . Shah et al. analyzed the results of 14 studies, reporting 2066 fracture events among 27,300 people with T1D (7.6%) and 136,579 fracture events among more than 4 million of people without diabetes (3.1%). They found that the relative risk (RR) of any fractures in T1D was 3.16, the RR of hip fractures was 3.78, and the RR for spinal fractures was 2.88 . Finally, a recent metaanalysis reported an increased RR of 1.88 for all fractures in young and middle-aged patients with T1D, aged 18–50 years .

In summary, patients affected with T1D showed an increased fracture risk. Differences in the age of study participants, ethnicity, duration of disease, level of glucose control, and diabetic complications may explain the variability in fracture risk. Several factors may be contribute to an increased fracture risk in T1D, including reduced bone mineral density (BMD), in particular at the femur . However, decreased BMD only partially explains the increased fracture risk in T1D.

Type 2 diabetes and fracture risk

Fracture risk is also increased in T2D but to a lesser extent compared to T1D . Furthermore, unlike T1D, T2D is characterized by increased BMD . One of the first reports of the increased fracture risk in T2D was observed in the “Study of Osteoporotic Fractures,” a prospective study of women aged 65 years or older. In this study, despite elevated BMD, the risk of all nonspine fractures was higher in women with T2D in age-adjusted models (RR 1.22; 1.06–1.41). The authors divided T2D patients based on treatment and they found that women not treated with insulin had higher risks of hip and proximal humerus fractures and those treated with insulin had a higher risk of foot fracture, compared with healthy controls . Later studies confirmed this observation. “The Health, Ageing, and Body Composition Study” reported that T2D was associated with a 64% increase in fractures compared with participants without diabetes, even after adjusting for BMD . Similar results were observed in the Women’s Health Initiative Observational Study, which studied a group of more than 90,000 postmenopausal women, reporting that those with T2D had a 20% increase risk of fractures compared to women without T2D, despite a higher baseline BMD . Two large metaanalyses summarized the evidence published over the past two decades, showing a threefold increased risk of fractures in T2D , depending on the skeletal site.

Of the nonvertebral fractures, hip fractures have been the site most extensively studied. In both men and women, T2D was associated with a threefold higher risk of hip fracture in a systematic review of 16 observational studies from Europe and the United States . This observation was further confirmed in other studies conducted in western population in women and men and also in Asian population . A recent Spanish study, conducted on “Sistema de Información para el Desarrollo de la Investigación en Atención Primaria” (SIDIAP) database among more than 50,000 patients with T2D, showed an increased incidence of hip fractures compared to control subjects without diabetes even if death for any cause was considered as a competitive event .

The impact of T2D on nonvertebral and nonhip fractures is less consistent. In particular, fractures of the wrist , the upper arm , and the foot seem to be increased in patients with T2D, although a metaanalysis reported either no change in fracture risk or a small increase for nonvertebral sites other than hip.

Finally, there are contrasting reports regarding the impact of T2D on vertebral fractures. The “Women’s Health Initiative Observational Study” reported a 20% increased risk of vertebral fracture in women with T2D (adjusted RR 1.2; 95% CI, 1.1–1.3) . The Malmö Preventive Project observed similar findings, T2D was the main risk factor for vertebral fractures, increasing the risk more than threefold (RR 3.56, 1.75–7.23; P =.001) . These results were also confirmed in Japanese men with T2D, with an odds ratio (OR) of 4.73 (95% CI, 2.19–10.2) for vertebral fractures after adjusting for age, BMI, and lumbar spine BMD . However, other studies failed to find an association between T2D and vertebral fracture risk, in particular in older women in the Study of Osteoporotic Fractures , in elderly men enrolled in the Osteoporotic Fractures in Men (MrOS) study and also in both men and women over 50 years of age and older in the population-based Canadian Multicentre Osteoporosis Study .

In conclusion, there are numerous reports with strong evidence for an increased risk of nonvertebral and hip fractures in T2D patients, irrespective of gender and ethnicity. However, there are conflicting findings on vertebral fractures, which may be attributed to differences in study design, the population enrolled, and method of fracture definition.

Bone turnover in diabetes

A large body of evidence has been published investigating bone biomarkers in patients with diabetes. Both formation markers, osteocalcin (OC), tartrate-resistant acid phosphatase (TRAP), and procollagen type 1 amino-terminal propeptide (P1NP), and resorption markers, C-terminal cross-linked telopeptide (CTX), seem to be lower at onset of T1D , but they showed a trend toward normalization over time.

Bone turnover markers have also been widely investigated in adult patients with both T1D and T2D, with two large metaanalyses being published recently. In 2014 Starup-Linde et al. found that both OC and CTX were decreased in patients with diabetes compared to healthy controls . Then, Hygum et al. published another metaanalysis on 66 studies, confirming the previous results . Furthermore, they found that not only CTX and OC but also P1NP were reduced in patients with diabetes, whereas osteoprotegerin (OPG) was increased . They estimated also that Sclerostin was borderlined significantly higher for patients with diabetes compared to controls . There were no differences in the markers TRAP, N-terminal telopeptide of type I collagen, bone alkaline phosphatase, and receptor activator of nuclear factor kappa-B (NF-kB) ligand (RANKL) .

Dividing patients based on diabetes type, authors found that TRAP was significantly decreased in T2D patients, but not in T1D, compared to controls . CTX and OC were significantly lower in both T1D and T2D patients compared to controls but P1NP was significantly decreased only in T2D. Of note, few of the studies considered in the analysis investigated P1NP in T1D. Serum Sclerostin levels have been extensively evaluated in several studies showing increased levels in T2D compared to T1D, LADA, or controls .

All these results may suggest a state of low bone turnover, confirmed by further evidence from bone histomorphometric analysis. Patients with T2D have a decreased number and a more immature form of circulating osteogenic precursor cells . The osteogenic precursor cells of T2D patients also showed a reduction in the expression of Runx2, the major regulator of osteoblast formation . Furthermore, histomorphometric indices of bone formation, as mineralizing surface, bone formation rate, and osteoblast surface, were reduced in subjects with T2D . Those findings were not confirmed in young patients with T1D by Armas et al. , although patients with previous bone fractures showed impaired structural and dynamic trends compared to nonfracturing subjects .

In conclusion, current evidence suggests that diabetes may be characterized by low bone turnover , as suggested by the reduction of both bone formation and bone resorption markers and by histomorphometric findings, as summarized in Table 50.1 .

Table 50.1

Bone formation and bone resorption in patients with diabetes mellitus.

Bone formation
Bone biomarkers Histomorphometry

  • ↓ Children with T1D

  • ↓ adults with T1D and T2D

↓ Mineralizing surface, bone formation rate and osteoblast surface in T2D impaired structural and dynamic trends in adults with T1D and bone fractures

  • ↓ Children with T1D

  • ↓ adults with T2D

Sclerostin ↑ Adults with T1D and T2D
Bone resorption
Bone biomarkers

  • ↓ Children with T1D

  • ↓ adults with diabetes

  • ↓ adults with T1D and T2D

Osteoprotegerin ↑ Patients with diabetes

  • ↓ In children with T1D

  • ↓ patients with T2D

OC , osteocalcin; P1NP , procollagen type 1 amino-terminal propeptide; T1D , type 1 diabetes; T2D , type 2 diabetes; TRAP , tartrate-resistant acid phosphatase.

Bone mineral density and microarchitecture in diabetes

As shown in previous sections, it is well established that fracture risk is increased in patients affected by both T1D and T2D, but this observation is not sustained by the same trend in BMD values, as one might expect. Vestergaard et al. reported in metaanalysis increased fracture risk among patients with both T1D and T2D. They found reduced BMD Z -score at the spine and hip in patients affected by T1D, but increased BMD at the spine and the hip in T2D patients . Most studies have reported a significant decrease in BMD at either the spine, hip, or total femur in patients with T1D. Long duration of disease , poor glycemic control, and chronic complications were associated with lower values of BMD. In contrast, T2D is characterized by an increased risk of fractures despite normal-to-high BMD at both the hip and spine , with 4%–5% higher BMD in T2D patients compared to controls . This feature is similar for both male and females and across ethnic groups including Mexican American, White, Black, and Asian .

It is important to highlight the relationship between BMD with BMI. Some authors showed that BMI is strongly related to BMD and, for this reason, they stated that BMI could explain higher BMD in T2D . However in some studies, increased BMD in T2D persisted after BMI adjustment . Hence, in patients with T2D, BMD underestimates their fracture risk. Furthermore, areal BMD does not account for possible alterations in bone geometry and/or bone microstructure that may be present in T2D.

Indeed, alteration in bone microarchitecture might contribute to bone fragility in patients with diabetes. Using the quantitative computed tomography (QCT) in men with T2D, it was shown that higher trabecular bone density compensated for the lower bone area and that higher volumetric BMD (vBMD) was associated with lower risk of vertebral fractures .

Using high-resolution peripheral QCT (HR-pQCT), it has been found that T1D is characterized by lower trabecular vBMD at the ultradistal radius and tibia and lower cortical thickness at the tibia compared to controls. Furthermore, impairments in bone microarchitecture were more severe in patients with microvascular complications .

In some, but not all studies, patients with T2D have deficits in cortical bone, represented by cortical porosity and lower cortical density. Burghardt et al. reported a deficit in cortical bone among postmenopausal women with T2D, whereas trabecular bone microarchitecture was similar to control subjects . Similar findings of lower cortical density, worse cortical microarchitecture, and greater cortical porosity were reported in postmenopausal African-American women with T2D and in postmenopausal women affected by T2D with prevalent fragility fractures .

Lower cortical vBMD and higher cortical porosity were also confirmed in a large cohort of T2D patients from the Framingham Study after adjusting for gender and obesity status, and these parameters were more severe in patients with a history of prior fracture . Recently, a large prospective study in the Bone Microarchitecture International Consortium showed that HR-pQCT parameters improved prediction of bone fracture beyond DXA-BMD and Fracture Risk Assessment Tool (FRAX) . Further studies are needed to validate HR-pQCT in clinical practice, establishing reference values and cutoff for fracture risk.

Pritchard et al., using magnetic resonance imaging, observed greater cortical porosity and larger holes within the trabecular bone network in patients with T2D compared to subjects without diabetes .

In summary, altered bone microarchitecture and increased bone porosity might contribute to increased fracture risk in patients with diabetes.

Bone mechanical properties in diabetes

Alterations in microarchitecture and bone porosity may contribute to bone strength impairment in patients with diabetes. In particular, patients with T2D have shown lower bone bending strength indices at the midshaft of the radius and tibia . Recently, studying bone specimens from the proximal femur, Karim et al. used cyclic reference point indentation to identify altered cortical bone biomechanical properties in patients with T2D, consistent with lower bone material strength in T2D .

Studying tissue composition and mechanical performance, increased advanced glycation end products (AGEs) concentration and mineralization were also observed in the cancellous bone at the femoral neck in men with T2D . Statistical modeling suggested that altered tissue composition in T2D would decrease postyield strain and postyield toughness after accounting for the influence of bone volume .

In vivo assessment of “bone material strength index” (BMSi) via impact microindentation at the tibial diaphysis revealed lower BMSi in 30 postmenopausal women with T2D duration of more than 10 years compared with matched controls . Furthermore, HbA1c levels were negatively correlated with BMSi . These findings were also confirmed in older women, aged 57–80, in a population-based study conducted in Gothenburg, reporting a lower BMSi in women with T2D compared to controls .

In conclusion, this evidence shows that poor bone quality decreases bone resistance and could contribute to increased bone fragility in patients with diabetes.

Bone metabolism in diabetes: from obesity and insulin resistance to insulin deficiency

Insulin deficiency

Insulin receptors (IRs) are present on osteoblast and osteoclast surfaces . Both in vitro and in vivo studies have shown a role of insulin in bone formation, regulating osteoblast proliferation and differentiation.

By stimulating the substrate of the IR substrate (IRS) complex, insulin activates the intracellular MAPK and PI3-K/Akt pathways leading to growth, differentiation, and survival of osteoblasts. Therefore dysfunction of osteoblasts may result from impaired insulin signaling.

In particular, IRS-1 in osteoblasts is required to promote osteoclast differentiation factors (ODFs) such as the RANKL/ODF .

Forkhead box (FOX) protein is a member of a family of transcription factors, with a relevant role in bone metabolism, activated by insulin signaling. FOX-1 is involved in several functions, it regulates RANK-induced osteoclast formation and cytokine production and it also protects osteoblast from oxidative stress . Insulin deficiency could lead to an imbalance of FOX production and regulation.

Some in vivo studies have pointed out that damage of osteoclast and osteoblast function caused by diabetes could be restored by insulin. In streptozotocin (STZ)–induced diabetic rats, insulin administration reverses the upregulation of osteoclastogenesis and downregulation of osteoblastogenesis caused by high blood glucose levels . This finding was confirmed in a clinical study conducted on 62 patients with new onset T1D, in whom stabilization of BMD at all sites was associated with insulin therapy .

C-peptide is a short 31-amino acid polypeptide release from proinsulin during the insulin synthesis. It has been previously observed that C-peptide can exert effects beyond the pancreas on multiple metabolic pathways . Serum C-peptide is negatively associated with BMD in more than 6000 subjects without diabetes, also after stratification based on insulin levels . Also Montalcini et al. showed that lumbar spine BMD was positively associated with C-peptide levels, unrelated to insulin levels . The same group of authors also found that, in a population of nondiabetic postmenopausal women, serum C-peptide was inversely correlated with bone fractures . This evidence seems to suggest that C-peptide itself, and not only insulin, could have a role in bone metabolism.


Obesity is one of the most important risk factors for diabetes developing and the role of obesity in bone metabolism has been deeply studied, but with contrasting findings. Until a few years ago, it was widely accepted that obesity was a protective factor in relation to bone health ; however, recent evidence has questioned the protective role of obesity on bone metabolism . The chapter goes beyond the systematic revision of studies conducted on obesity, considering that the possible relationship between bone and fat is complex and not yet entirely understood. In particular, obesity has shown a protective role mainly against hip fractures; however, ankle and humerus fractures are more common in obese subjects . Furthermore, obesity is related to higher BMD and increased bone strength, but bone strength does not sufficiently increase enough to protect against all fractures . Higher mechanical loading (positive for BMD) influences this complex relationship; moreover, increased production of proinflammatory cytokines and adipokines in obesity may cause skeletal fragility.

Osteoblasts, osteocytes, and adipocytes share the same cell precursors, deriving from a common mesenchymal stem cell (MSC), but follow different evolution based on the activation of different signaling pathways. The Wnt/β-catenin signaling pathway stimulates the differentiation of osteoblast, while peroxisome proliferator–activated receptor gamma (PPARγ) induces adipogenesis. PPARγ mRNA expression in adipose tissue is increased during obesity and the increased level of PPARγ may be involved in reduced bone formation .

Role of adiponectin and ghrelin on bone metabolism and risk of fracture has been investigated in several studies, although data have not been consistent .

Leptin is produced by adipose cells and its main role is to suppress appetite. Obesity leads to a reduction of sensitivity to leptin, resulting in higher levels in obese subjects compared to normal-weight people. Leptin exerts its function on both bone formation and bone resorption process. Leptin shows a direct anabolic effect on osteoblast and seems to prevent osteoclast differentiation increasing OPG levels . In most studies, leptin is positively correlated with BMD although this association is sometimes not independent of fat mass . In addition, reduced levels of leptin are found in subjects with fractures . Furthermore, the administration of leptin in obese subjects, aside from weight reduction, leads also to preservation of bone density or increment of bone mass, suggesting an anabolic effect to bone .

Increased aromatase expression in the obese may also explain the protective effect of obesity on bone metabolism.

Hyperinsulinemia and insulin resistance

T2D is characterized by hyperinsulinemia and insulin resistance, which are considered the hallmarks of the disease. As previously stated, the increase in BMD in T2D patients is correlated with increased BMI, this correlation might be explained by high levels of insulin in those patients.

Cross-sectional studies have demonstrated that there is a positive correlation between insulin and BMD , this evidence is confirmed by the association of high BMD levels in most of the clinical conditions characterized by high insulin levels, such as metabolic syndrome .

In contrast, animal studies show that insulin resistance may have a negative effect on bone metabolism, playing a detrimental role in insulin signaling in osteoblasts . Epidemiological data have not been consistent, showing either a negative, a neutral, or a protective effect of insulin resistance on BMD or fracture risk . Furthermore, a recent study conducted in nondiabetic older adults in the Health, Aging and Body Composition prospective cohort study found that greater insulin resistance is associated with higher BMD and there was no association between insulin resistance and fracture risk after adjustment for BMI and BMD .

Insulin-like growth factor 1 (IGF-1), or somatomedin C, binds not only to the IGF-1 but also to the IR. Although with a lower binding affinity with the latter, IGF-1 can trigger the insulin signaling. Furthermore, the IGF receptor has been found on the surface of osteoblast , and IRS-1 and IRS-2 are main mediators in the IGF signaling cascade .

Adolescent T1D patients have low levels of IGF-1 compared to subjects without diabetes ; earlier age at diagnosis and poor metabolic control are predictive of lower IGF-1 . Low values of IGF-1 are associated with low femoral and lumbar spine BMD in T1D patients and with vertebral fractures in postmenopausal women with T2D . These data were recently confirmed by a cross-sectional study, conducted among almost 1000 postmenopausal women and men with T2D, in which patients with lower IGF-1 had significantly increased risk of vertebral fractures compared to the subjects with higher IGF-1 .

Diabetes mellitus and bone health: pathways to skeletal fragility

Glucose toxicity

Bone formation: osteoblast and osteocyte function


Hyperglycemia results in a detrimental effect on osteoblastogenesis. As discussed, the precursor of osteoblasts is the MSC and high glucose concentrations reduce MSC viability and clonogenicity . Hyperglycemia also has a negative effect on bone marrow stromal cells (BMSC). In particular, a reduction in proliferation of BMSC and alkaline phosphatase (ALP) activity has been observed in several in vitro studies after using a high glucose concentration in addition to a downregulation of bone mineralization rate in BMSC from STZ-induced diabetic rats . BMSC exposed to hyperglycemia exhibit increased activity of adipogenic pathways. Indeed, when BMSC are exposed to a high glucose concentration, an increase of adipogenic, rather than osteogenic, pathways was observed. Activation of PPARγ and enhanced expression of cyclin D3 were identified as mechanisms involved in this process . The dysregulation of PI3K/Akt pathway, responsible for bone formation, also stood out as one of the mechanisms that could explain the stimulation of adipogenesis over osteoblastogenesis, due to the negative impact of reactive oxygen species, caused by hyperglycemia . A growing body of evidence derived from animal studies has also shown other mechanisms in osteoblast function affected by diabetes. In particular, in these studies a low rate of mineralization and decreased trabecular bone volume were observed and linked to the reduction of RUNX2 gene expression and ALP activity , downregulation of OC, OPG , and bone morphogenetic protein-2 expression .

Another important pathway in bone formation is the WNT/β-catenin signaling, which results in the transcription of genes involved in the proliferation and differentiation of osteoblast. High levels of blood glucose exert an adverse effect on this pathway. Notably, STZ diabetic rats have decreased in β-catenin and increased expression of WNT signaling inhibitors SOST and Dickkopf-related protein 1 (Dkk1) .


In recent years, several authors highlighted the increasing importance of osteocyte’s function and contribution on bone fragility in diabetes. In STZ diabetic mice, increased osteocyte apoptosis may lead to reduced osteocyte density and numbers. Furthermore, osteocytes seem to play a central role in osteoblast function due to their production of Sclerostin, one of the major inhibitors of WNT pathway, resulting in a reduction of bone formation. Sclerostin may also play a role in bone–vascular axis in diabetes, being involved in vascular atherosclerosis and calcification .

Bone resorption: osteoclast function

The impairment of bone resorption is another mechanism through which diabetes and hyperglycemia could affect bone metabolism. This detrimental effect is linked to osteoclastogenesis; in particular, it is well known that differentiation of embryonic stem cells into osteoclasts is promoted by physiological glucose levels, and hyperglycemia could negatively impact this process. In support of this hypothesis, STZ-induced diabetic mice have impaired bone resorption, due to the reduced levels of dendritic cell-specific transmembrane protein , which are involved in osteoclast differentiation .

In addition, studies report a reduction of RANKL concentration , NF-kB activity , cathepsin K, and TRAP activity in vitro when osteoclasts are exposed to a high concentration of glucose . Recently, Cai et al. found that high glucose levels, suppressing AMPK/mTOR/ULK1 signaling, reduced the autophagy in osteoclast cells, affecting osteoclast formation and function .

Advanced glycation end products and “diabetic collagenopathy”

Many authors have hypothesized and partially demonstrated that the impairment of tissue material properties in diabetes seems to be related to the alterations in bone collagen. Indeed, bone elasticity, toughness, and strength depend, in part, on the type of cross-links between adjacent collagen molecules . Newly formed collagen fibers need to be stabilized by covalent cross-links that are represented by both enzymatic and nonenzymatic cross-links. Diabetes can negatively affect both ways. In particular, diabetes decreases the lysyl oxidase activity that is responsible for the prevention of excessive accumulation of enzymatic cross-links in the physiological mineralization process . AGEs are nonenzymatic chemical modifications of proteins by aldose sugars that are elevated in individuals with hyperglycemia and, in particular, are associated with macro- and microvascular complications of diabetes. Several studies indicate that formation of AGEs within collagen fibers collagen can adversely affect mechanical properties, which may ultimately contribute to increased skeletal fragility. Moreover, as diabetes is characterized by a low bone turnover, there is an accumulation of a large quantity of altered type 1 collagen, which may determine biomechanical changes both in the cortical and trabecular bone. Recently, the additional role of pentosidine in “diabetic collagenopathy” has been found by Hunt et al.; they observed that in cancellous bone of T2D men, increased concentrations of pentosidine were associated with a reduction of postyield strain and toughness .

Karim et al. have demonstrated that cortical, but not trabecular, bone AGEs tend to be higher in T2D subjects and serum AGEs were positively correlated with cortical and trabecular bone AGEs . Findings from in vitro studies have suggested that AGEs themselves, and the interaction of AGEs with their receptor RAGE (expressed in osteoblast), may negatively affect osteoblast activity . Clinical studies have reported variable associations between diabetes, AGEs, bone turnover, and fracture risk in humans. Bone turnover markers have been reported as unrelated to, or positively associated with, serum pentosidine , and negatively associated with carboxymethyllysine (CML); instead, a positive association between serum pentosidine and fracture risk in cohorts of people with T1D and T2D has been shown . Endogenous secretory RAGE (esRage) has also been tested in diabetes in order to evaluate the potential relationship with the risk of fragility fractures. In particular, low esRAGE levels have been shown to be a risk factor for vertebral fractures in patients with T2D . Recently, Lamb et al. confirmed that, in more than 3000 community-dwelling men aged 70–89 years, circulating esRAGE modulates bone turnover in older men while CML level predicts incidence of hip fracture .

However, these results should be carefully evaluated as the immunoassay for serum AGEs (mainly pentosidine) has a low grade of sensitivity and specificity due to numerous factors in blood and urine that interfere with immunoassay standardization.

Incretin system

The incretin system encompasses a group of hormones that regulates the cross talk between stomach, bowel, and pancreas. Glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1) are the most important players in this system, accounting for 90% of this physiological role. It is well known that the incretin system is impaired in diabetes and that food-stimulated insulin secretion is reduced in T2D. While GIP seems to be normal or even elevated in T2D, meal-stimulated GLP-1 is reduced in subjects with T2D. GIP and GLP-1 have a short half-live because they are rapidly degraded by dipeptidyl peptidase-4 (DPP-4).

GIP is a 42-amino acid hormone, secreted by K-cells of proximal regions of the small intestine (duodenum and proximal jejunum) . The action of GIP starts with the activation of a specific G protein–coupled receptor (GIPR) expressed by different cells, not only pancreatic beta cells and adipocytes, but also osteoclasts , osteoblasts , osteocytes, and chondrocytes . Some evidence confirmed the action of GIP on both osteoblast and osteoclast function, sustaining bone formation. GIP promotes osteoblast proliferation, raising the expression of collagen type I and the activity of ALP . Contrasting evidence has been reported on the impact of GIP on osteoclast activity, . However, reduced cortical mass and bone strength were found in mice lacking the GIP receptor , whereas GIP receptor-knockout mice have impaired mechanical properties of bone matrix, BMD, and the ratio of mature/immature collagen cross-links .

Few studies were conducted in humans, suggesting a positive association between GIP and bone resorption. Nissen found that during euglycemic or hyperglycemic glucose clamps GIP decreased C-terminal telopeptide (CTX) levels , these findings was also recently confirmed by Bergmann that evaluated the CTX reduction during glucose and GIP infusions in overweight and obese subjects .

GLP-1 interacts with osteoblasts through a GPI/IPG-coupled receptor (GLP-1R) but the mechanism of action is not fully elucidated yet. GLP-1 seems to induce osteogenic differentiation of adipose-derived stem cells and may reduce osteoclasts through a calcitonin-dependent pathway .

Acute and chronic diabetic complications


Large studies have found that hypoglycemia is closely correlated with risk of falls, the primary risk factor for bone fracture, in T2D patients . Among several evidences, in the Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus-Thrombolysis in Myocardial Infarction 53 (SAVOR-TIMI 53) trial, over 16,000 subjects enrolled fracture risk were associated with major hypoglycemic events . Recently, also data from THIN database show that hypoglycemia is a significant risk factor for fragility fractures in patients with T2D .

Chronic diabetes complications

Growing evidence supports the link between chronic diabetes complications and risk of fracture in both T1D and T2D by several mechanisms . Diabetic polyneuropathy and retinopathy cause impaired postural stability and visus deficits, respectively, which increase risk of falls . Cardiovascular diseases, such as stroke, peripheral vasculopathy, and arrhythmia, may also increase the hazard of collapse and falls. In addition, fractures are increased in patients with T1D and diabetic kidney disease, maybe due to altered vitamin D and parathyroid hormone (PTH) levels . In a cross-sectional study, only a subgroup of T2D subjects with microvascular complications showed increased cortical porosity compared to patients without microvascular complications or healthy controls, but the mechanism of this finding remains unclear . Further, poor glycemic control, presence of neuropathy, and longer disease duration are associated with increased fracture risk in T1D patients .

In the Joslin 50-Year Medalist Study, following patients with T1D with more of 50 years of disease duration, a low rate of nonvertebral fractures and normal Z -scores were reported, suggesting that long-term T1D survivors, presenting low cardiovascular complications, are also protected from bone fragility .

Calcium metabolism

Low levels of vitamin D are observed in both T1D and T2D, and low levels of vitamin D at onset of T1D seem to also suggest a pathogenic role in the development of T1D . However, calcitriol supplementation did not result in any change in bone turnover markers in T1D subjects .

Moreover, urinary loss of calcium, phosphorus, and magnesium could be caused by polyuria due to hyperglycemia leading to an imbalance in the PTH/vitamin D axis. In contrast, it has been found that both men and women with T2D have decreased levels of PTH, leading to low bone formation and reduced bone resorption . A Japanese study has shown that low levels of PTH in T2D might increase the risk of vertebral fracture . Finally, an impairment in parathyroid function was also observed in T1D; Thalassinos et al. described several years ago that a “functional hypoparathyroidism” in subjects affected with T1D is likely associated with poor glycemic control, while improvement of metabolic control positively affects serum calcium levels .


Inflammation, despite its pathway, plays an important role in both T1D and T2D; hence, it has been suggested to consider diabetes as an inflammatory disease. T1D is a result of a complex interplay between genetic susceptibility and environmental factors, although a possible role is also played by natural inflammation. In particular, interleukin-1 (IL-1), a proinflammatory cytokine with a central role in innate immunity, and tumor necrosis factor-α (TNF-α) have been involved in T1D onset and beta cell damage, leading to a negative effect on insulin production . Both IL-1 and TNF-α promote bone resorption increasing osteoclast activity, expressing RANK and inhibiting apoptosis . In addition, it has been observed that TNF-α may decrease RUNX2 and osterix, and as a consequence, the expression of Sclerostin and Dkk1 is increased . Reducing TNF-α levels or suppressing TNF-α by TNF soluble binding protein or an anti-TNF-α, a reduction of RANKL-induced osteoclast formation is achieved , resulting in a reduction of bone resorption.

The pathogenic process leading to T2D is associated with low-grade chronic inflammation . IL-6 is one of the cytokines most strictly related with obesity and T2D, as it is released from the adipose tissue . It has been shown that the increased level of IL-6 is associated with both insulin resistance and hyperglycemia. Recently, new findings support the relationship between IL-6 and bone health, even if the specific role is not clear yet due to the contrasting results. IL-6 promotes bone resorption, increasing osteoclastogenesis, and both serum IL-6 and gene polymorphisms in IL-6 seem to play a role in the reduction of BMD and muscle strength . In contrast, IL-6 stimulates mesenchymal progenitor differentiation toward osteoblastic lineage with a beneficial effect on bone formation in conditions of higher bone turnover . Despite the presence of inflammation, usually leading to high bone turnover, diabetes is characterized by a condition of low bone turnover; therefore the role of inflammation on bone health in T2D still needs further clarification. .

Marrow adiposity

Age and environmental and health conditions contribute to the constitutional transformation of bone marrow . In particular, development of marrow adiposity is a physiological age-related phenomenon that includes the transformation of an active marrow (hematopoietic/red marrow) into a less active one (fat/yellow) . Bone marrow fat may act as an insulin-sensitive tissue, related to systemic energy metabolism, because it exerts genetic and metabolic functions similar to brown adipose tissue . In an animal model of T2D mice, bone loss has been associated with increased bone marrow adiposity . Many studies have suggested that in patients with bone fractures and diabetes, the composition of bone marrow fat, especially increased fatty acid saturation, could play a role ; as a matter of fact a negative relationship between bone marrow adiposity and BMD has been observed in humans . Furthermore, an association between circulating level of Sclerostin with higher vertebral marrow fat in men has been found, suggesting an involvement of osteocyte function in marrow adipogenesis . These observations propose that bone marrow microenvironment disruption could increase adipogenesis at the expense of osteoblastogenesis .

Muscle: role of sarcopenia and new biomarkers

As discussed earlier, diabetes through hyperglycemia exerts its effect on several tissues and also on the musculoskeletal system, including muscle and connective tissue, aside from bone. T1D patients sometimes present with “diabetic myopathy,” a condition characterized by reduced muscle growth and strength and decreased stem cell differentiation into the myogenic lineage . In addition, diabetes may also be the cause of sarcopenia, defined as a disease of the muscle characterized by a decline in skeletal muscle mass, muscle strength, and physical performance by The European Working Group on Sarcopenia in Older People 2 . Risk of developing sarcopenia is threefold higher in T2D patients compared to nondiabetic individuals after adjusting for multiple factors . Then, sarcopenia leads to an increased risk of falls and fractures . Although the mechanism implied in the relationship between diabetes and sarcopenia has not been fully elucidated yet, insulin resistance is responsible for decrease in the synthesis of muscle protein and increase in protein degradation and intramuscular fat.

Irisin could be counted as a further mechanism to explain the connection among diabetes, increased risk of fracture, and skeletal muscle. Irisin is a myokine, regulated by physical exercise, which can trigger “browning” of white adipose tissue. Both in vitro and in vivo findings have shown that irisin could have a positive effect on bone health ; recently it has been described that irisin binds to αV integrin receptor on the osteocytes and fat tissue cells and activates osteocyte to produce Sclerostin , inhibiting osteoblasts.

Effect of diabetes therapies on bone metabolism


International guidelines on the management of diabetes indicate metformin as the first line therapy in the treatment of T2D and suggest also a role in the prevention of development of T2D from a condition of prediabetes. Both preclinical and clinical data available seem to point out a good safety profile of metformin regarding bone health. Regarding clinical data, it is possible to exclude a harmful effect of this drug on bone, but contrasting findings are available on a positive or neutral effect on fracture risk.

Preclinical data

Preclinical findings showed that metformin was able to promote cellular proliferation, ALP activity, calcium deposition, and the number of nodules formed in rat primary osteoblasts , counteracting the detrimental effects of hyperglycemia on osteoblast function , suggesting a potential anabolic role on bone metabolism. metformin acts also on bone resorption by modulating the RANK/RANKL/OPG pathway and inhibiting osteoclast function , resulting in the reduction of bone resorption.

Clinical data

In large cohort studies conducted in Denmark and Rochester, treatment with metformin in patients with T2D has a protective effect toward hip fracture risk, with a lower risk of 20%–30%, compared to insulin and sulfonylureas or thiazolidinediones (TZDs) users . Similarly, in “A Diabetes Outcome Progression Trial” (ADOPT), subjects enrolled in the metformin group had a lower cumulative incidence of fracture compared to glyburide or rosiglitazone groups . Other cohort studies have reported a neutral effect of metformin on hip fracture risk in elderly T2D subjects . The Osteoporotic Fractures in Men (MrOS) study showed that the use of metformin in men with T2D (aged ≥65 years old) did not affect the risk of bone fractures ( Table 50.2 ).

Table 50.2

Effects of hypoglycemic treatment on bone mineral density (BMD) and fracture risk.

BMD Fracture
Metformin =/↑ ↓/=
Sulfonylureas =/↑
Thiazolidinediones ↓↓/= ↑↑/= a
Incretin GLP-1 analog =/↑ =/↓
DPP-4 inhibitor =/↓
SGLT2-I =/↓ =/↑

↑ Increase, ↓ decrease, =unchanged, – data not available.

Only gold members can continue reading. Log In or Register to continue

Oct 27, 2020 | Posted by in ENDOCRINOLOGY | Comments Off on Diabetes, diabetic medications, and risk of fracture

Full access? Get Clinical Tree

Get Clinical Tree app for offline access