Keywords

Key Facts

Key Facts of Diabetes Mellitus and Related Bone Disease

  • Diabetes is an extremely common disease throughout the world, with an estimated 592 million cases in 2035.

  • Type 1 diabetes is characterized by a near complete loss of insulin production.

  • Type 2 diabetes is characterized by a decrease in insulin production relative to insulin sensitivity.

  • Diagnosis of diabetes mellitus can be made by measuring fasting plasma glucose above 7 mmol/l, a 2 h oral glucose tolerance test plasma glucose value above 11 mmol/l or an HbA1c of ≥ 48 mmol/mol.

  • Apart from the well-known microvascular and macrovascular complications, diabetes is also related to poor bone health.

  • The risk of hip fracture is suggested to be sevenfold increased in patients with type 1 diabetes and twofold increased in patients with type 2 diabetes.

  • Bone mineral density is increased in type 2 diabetes and slightly decreased in type 1 diabetes, but the lower bone mineral density in type 1 diabetes cannot explain the increased risk of fracture.

  • Patients with diabetes display lower C-terminal cross-linked telopeptide of type-I collagen and osteocalcin levels representing lower bone resorption and bone formation.

  • The mechanisms behind the increased risk of fractures in diabetes are still unclear but could be related to lack of insulin, disturbed glucose metabolism, medication use, renal impairment, falls or other factors.

Key Facts of Bone Remodeling

  • Bone consists of a mineralized matrix, mainly hydroxyapatite, a non-mineralized matrix, mainly collagen, and a cellular compartment.

  • The cellular component of bone consist of osteoclasts that resorb bone tissue, osteoblasts that form new bone tissue, and osteocytes that are thought to regulate the bone turnover process.

  • Bone remodeling is a highly coordinated process of degradation of old bone and creation of new bone.

  • Bone remodeling consists of three phases: A bone resorption phase maintained by the osteoclasts, a reversal phase where the bone is prepared for the osteoblasts, and a bone formation phase where matrix is produced by osteoblasts and subsequently matured and mineralized.

  • When the bone remodeling is out of balance, typically with degradation exceeding creation of bone, osteoporosis can arise.

  • Bone turnover markers are biomarkers that reflect bone remodeling.

  • Bone turnover markers can easily be measured in blood and are a useful tool in assessing bone remodeling. See Table 1.

    Table 1 Overview of commonly used bone turnover markers
    Full size table

Key Facts of Glucose and Bone Turnover

  • C-terminal cross-linked telopeptide of type-I collagen decreases within 20 min of an oral glucose tolerance test.

  • When glucose is given intravenously, a decrease in C-terminal cross-linked telopeptide of type-I collagen is seen; only it is delayed by 1 h compared to the oral glucose tolerance test.

  • The decrease in C-terminal cross-linked telopeptide of type-I collagen seen in an intravenous glucose tolerance test is significantly lower than that of the oral glucose tolerance test.

  • In healthy males, a hyperglycemic clamp has been shown to induce a decrease in osteoprotegerin, whereas no change was seen during euglycemia.

  • Procollagen type 1 N-terminal propeptide has both been reported to be stable and to decrease during an oral glucose tolerance test.

  • Hyperglycemia may decrease osteoblast differentiation and bone formation and impair bone resorption by increasing osteoprotegerin.

  • Hyperglycemia decreases the number of osteoclasts, inhibits osteoclastogenesis and osteoclast differentiation, and impairs the ability of osteoclasts to resorb mineralized matrix.

  • Osteocytes react to hyperglycemia by increasing sclerostin production, which in turn inhibits the Wnt pathway and thereby bone formation.

Definitions of Words and Terms

Bone formation:

The process of producing new bone tissue.

Bone resorption:

The process of degrading old bone tissue.

Bone turnover:

The process of old and damaged bone tissue being degraded and replaced by new bone tissue.

Euglycemia:

The state with a normal concentration of glucose in the blood.

Hyperglycemia:

The condition of having a higher than normal concentration of glucose in the blood.

Hyperglycemic clamp:

A technique used in experiments where a constant but varying amount of glucose is infused intravenously in accordance to insulin secretion and glucose metabolism to keep blood glucose at a constant high level. Same technique can be used for achieving euglycemia (euglycemic clamp).

Hypermineralization:

A state where bone is over-mineralized relative to its collagenous matrix.

Mineralization:

The process of impregnate mineral in the matrix of the bone. The mineral content of bone is primarily hydroxyapatite, which primarily consists of calcium and phosphate.

Osteoblasts:

The cell type responsible for bone formation.

Osteoclastogenesis:

The development of osteoclasts.

Osteoclasts:

The cell type responsible for bone resorption.

Osteocytes:

The most common cell type in bone tissue, thought to be encased osteoblasts that control the activity of osteoblasts and osteoclasts through mechanosensory mechanisms.

Introduction

Diabetes Mellitus

Diabetes Mellitus is a highly prevalent condition throughout the world with an estimate of 592 million suffering from it in 2035 (International Diabetes Federation). Diabetes is characterized by a relatively decreased insulin production, with a complete lack of insulin production in type 1 diabetes and a decreased insulin production relatively to the insulin resistance in patients with type 2 diabetes. The decreased insulin production causes unstable fasting conditions and patients with diabetes may be diagnosed by increased fasting plasma (p-) glucose of >7 mmol/l, an 2 h value of >11 mmol/l at an oral glucose tolerance test (OGTT) (American Diabetes Association 2012) or an elevated glycated hemoglobin A1c (HbA1c) level of ≥ 48 mmol/mol. The glycemic regulation in patients with diabetes is disturbed and fasting p-glucose levels do not follow the same pattern as in non-diabetes subjects.

Bone Remodeling

Bone remodeling is the process of degradation of old bone and creation of new bone (Hadjidakis and Androulakis 2006; Khosla and Riggs 2005). Under optimal circumstances the degradation (bone resorption) and creation (bone formation) of bone are balanced. In osteoporotic individuals, the rate of resorption is higher than the rate of formation (Khosla and Riggs 2005) which diminish bone mass and bone mineral density (BMD). Bone is constructed by a mineralized matrix consisting of mainly hydroxyapatite, a non-mineralized matrix consisting of primarily collagen, and a cellular compartment consisting of the bone cells osteoclasts, osteoblasts , and osteocytes . Mechanical resistance is provided by the hydroxyapatite crystals whereas stability and elasticity are provided by the network of type I collagen (Boskey 2013).

Osteoclasts are the bone resorping cells and osteoblasts are the bone forming cells. The osteocytes may serve as main regulators of the bone remodeling as they react to mechanical stress. The osteocytes produce and secrete sclerostin, a Wnt pathway inhibitor that decreases bone formation, and fibroblast growth factor-23 (FGF-23) that stimulate phosphate excretion in the kidneys. Furthermore, Receptor Activator of Nuclear factor Kappa beta Ligand (RANKL), a promoter of bone resorption, and its antagonist osteoprotegerin (OPG) are secreted by osteocytes. However, the main production of OPG and RANKL are from the osteoblasts (Hadjidakis and Androulakis 2006).

The bone remodeling is divided into three phases; a bone resorption phase performed by the osteoclasts, a reversal phase where the bone is prepared for the osteoblasts, and finally the bone formation phase that consists of the production, maturation, and mineralization of the matrix. The resorption phase may take 2 weeks, whereas bone formation may continue for as long as 4 months. In healthy subjects, the production of matrix and mineralization of matrix are at the same rate (Hadjidakis and Androulakis 2006). Figure 1 depicts the systems involved in bone remodeling.

Fig. 1
figure 1

Systems regulating bone remodeling. The osteocyte secretes sclerostin, which decrease bone formation and FGF-23 which increase phosphate excretion in the kidneys. The osteoblasts maintain bone formation and regulate the osteoclast through both OPG and RANKL. Red arrows are inhibitory actions, green arrows stimulatory actions, and blue arrows secreted products (With permission from the author (Starup-Linde 2015))

Full size image

Bone Turnover Markers

Bone turnover markers are biomarkers of the bone remodeling and a specific marker represents a specific phase of the bone remodeling. Bone turnover markers are released to the blood during the bone remodeling and are thus easily measured providing information on the bone turnover. Table 1 describes the most commonly used bone turnover markers. C-terminal cross-linked telopeptide of type-I collagen (CTX) and N-terminal cross-linked telopeptide of type-I collagen (NTX) are commonly used resorption markers that reflect collagen degradation. Tartrate resistant acid phosphatase (TRAP) reflects the activity of osteoclasts and is also a resorptive marker. Procollagen type 1 N-terminal propeptide (P1NP) is a formative marker, which is released during the cleavage of immature collagen. Osteocalcin, another formative marker, is an unmineralized matrix component and reflects the bone formation. Bone-specific alkaline phosphatase (BAP) is an enzyme produced during the mineralization phase of the bone formation (Starup-Linde 2013; Starup-Linde and Vestergaard 2015).

Analytical Factors

Bone turnover marker assays are offered by a large number of immunodiagnostic kit companies. Most are research-grade assays that are not intended for diagnostics use. Assays intended for diagnostic use are regulated by national and international bodies in terms of a range of validation parameters, particularly assay standardization (White 2011). This is in contrast to research-grade assays, which are unregulated and frequently missing assay characterization and standardization (Bowsher and Sailstad 2008). The above leads to considerable measurement differences between commercial assays and may result in conflicting research findings and slowing the implementation of bone turnover markers into routine clinical practice (Seibel et al. 2001; Whitham and Milford-Ward 2000). In this regard a recent publication stands out. The joint International Osteoporosis Foundation has recommended the use of CTX and P1NP as reference bone turnover markers in clinical trials and proposed strategies for standardization aiming for future inclusion in routine clinical practice and comparable values across assays (Vasikaran et al. 2011). Until such standardization has been attained, results and reference intervals from different assays should not be used interchangeably for clinical use, and care must be taken to address this important issue in research studies (Meier et al. 2009).

In addition to analytical issues, preanalytical factors are considered problematic with significant influence on measurements (Hannon and Eastell 2000). Preanalytical factors are factors such as sample handling, circadian, age, gender, menopausal status, and fractures. Of these, sample handling and circadian changes can be controlled by standardized sampling, sample handling, and collecting samples at the same time of day. Most other preanalytical factors cannot be controlled and their recognition is important in the interpretation of bone turnover marker results. Therefore, clinicians and researchers should be familiar with conditions where bone turnover levels are expected to be altered, for example in children, menopause, and after recent fracture (Seibel 2005). Another preanalytical factor commonly viewed as an obscuring factor but also of scientific interest and the focus of this review is the observed postprandial suppression of bone turnover levels (Clowes et al. 2003). This subject is further described in section “The Effect of Glucose Intake on Bone Turnover Markers in Humans.”

Diabetes and Bone

Diabetes is related to microvascular and macrovascular complications (American Diabetes Association 2012). Until recently the increased risk of fracture was an overseen complication. Thus, diabetes and bone may be related. The risk of hip fracture has been suggested to be sevenfold increased in patients with type 1 diabetes and twofold increased in patients with type 2 diabetes compared to non-diabetes individuals (Vestergaard 2007; Janghorbani et al. 2007). One would expect a similar lowering of BMD as it is the primary fracture predictive tool. However, BMD is increased in patients with type 2 diabetes and only slightly decreased in patients with type 1 diabetes and does not explain the increased fracture risk (Vestergaard 2007). Patients with diabetes have hypermineralized bone relative to their decreased bone material competence. When adding further fracture predictors to the model in The Fracture Risk Assessment Tool (FRAX), which also includes BMD, it underestimates both hip fracture risk and major osteoporotic fracture risk in patients with type 2 diabetes (Giangregorio et al. 2012). Furthermore, the increased fracture risk is not explained by either hypoglycemic events or the number of falls in patients with diabetes (Bonds et al. 2006; Vestergaard et al. 2005), or why the increased risk of fracture in diabetes patients seems to be bone related. The mechanism of the decreased bone quality and lack of fracture predictors is not well understood in diabetes; however, it may relate to the relative lack of insulin and disturbed glucose metabolism, but also factors as obesity, medication use, and renal impairment may affect the bone metabolism in patients with diabetes.

Glucose and Bone Turnover Markers

Clinical Studies

The Effect of Glucose Intake on Bone Turnover Markers in Humans

Patients with diabetes are at increased risk of fracture, thus their bone turnover may be altered. Clinical studies have investigated the effect of glucose on bone turnover markers. Table 2 presents the studies that have examined the effect of glucose ingestion on bone turnover markers. In men and women subjected to an OGTT, the bone resorption marker s-CTX and u-CTX decreased (Clowes et al. 2003; Henriksen et al. 2003; Bjarnason et al. 2002; Nissen et al. 2014; Chailurkit et al. 2008; Viljakainen et al. 2014; Paldanius et al. 2012; Schwetz et al. 2014; Karatzoglou et al. 2014). The decrease in s-CTX has been reported as early as 20 min after glucose ingestion (Clowes et al. 2003), whereas the decrease was delayed by one hour during the intravenous glucose tolerance test (IVGTT) (Bjarnason et al. 2002). The decrease in s-CTX was apparent in patients with type 2 diabetes but lower than in healthy controls (Chailurkit et al. 2008). Furthermore, the osteoclast specific marker TRAP decreased in both healthy obese and healthy non-obese individuals during OGTT (Viljakainen et al. 2014). A convincing effect of glucose intake on bone resorption markers was observed in these OGTT and IVGTT studies. The effect on bone formation markers was more unsettled, although both P1NP and osteocalcin have been reported to decrease. A decrease in s-osteocalcin has been shown two hours after an OGTT (Clowes et al. 2003; Viljakainen et al. 2014; Paldanius et al. 2012; Schwetz et al. 2014); however, other studies showed stable osteocalcin levels when comparing to fasting conditions (Henriksen et al. 2003; Bjarnason et al. 2002; Holst et al. 2007). P1NP has both been reported to be stable (Karatzoglou et al. 2014) and to decrease (Clowes et al. 2003; Viljakainen et al. 2014; Paldanius et al. 2012; Schwetz et al. 2014) during an OGTT. The mineralization marker BAP decreased during OGTT in both obese and non-obese subjects (Viljakainen et al. 2014).

Table 2 Studies that examine the effects of glucose ingestion and bone turnover marker response
Full size table

As the effects of OGTT and IVGTT are related to time, it was important to show whether time itself affected the bone turnover markers. Maintenance of a euglycemic p-glucose level of 5 mmol/l did not change the levels of s-CTX, osteocalcin, or P1NP (Clowes et al. 2002) when followed for two hours. Furthermore, when comparing fasting condition with OGTT, CTX decreased significantly more during the OGTT than during the fasting state (Henriksen et al. 2003). Thus, glucose intake has a time independent effect on bone turnover markers.

Glucose intake decreased the bone turnover markers, but a decrease was also observed during hypoglycemia where parathyroid hormone (PTH), P1NP, s-CTX, and osteocalcin decreased (Clowes et al. 2002). Bone turnover markers may decrease with p-glucose values both lower and higher than 5 mmol/l. Therefore the effect of glucose on bone turnover markers may be u-shaped with an optimal state in the normal healthy fasting condition. Although both IVGTT and OGTT decreased s-CTX, the decrease was significantly smaller during IVGTT (Bjarnason et al. 2002), which suggests that an additional component from the gastrointestinal tract affects bone turnover. Glucagon-like peptide-2 (GLP-2) increased while CTX decreased in gastrectomized patients (Holst et al. 2007), and this supports that the gastrointestinal absorption may affect bone turnover. Furthermore, the decrease in CTX during hyperglycemia was enhanced in combination with infusion of gastric inhibitory peptide (GIP) (Nissen et al. 2014), which suggests that the gastrointestinal hormones potentiate the effect of glucose on bone turnover. An intravenous injection of GIP and a subcutaneous injection of glucagon-like peptide-1 (GLP-1) did not affect s-CTX, whereas subcutaneous injection of GLP-2 decreased s-CTX (Henriksen et al. 2003).

The clinical studies show a strong relation between glucose intake and bone turnover, which may be either mediated or enhanced by gastrointestinal hormones. However, no direct pathway was established. The effect may be from an alteration of the OPG/RANKL pathway. During a hyperglycemic clamp OPG decreased in healthy males, while no change was observed during euglycemia (Knudsen et al. 2007). In type 2 diabetes women, OPG remained stable during an OGTT whereas it decreased in healthy women (Chailurkit et al. 2008). Thus, the OPG system may be altered in patients with diabetes compared to healthy subjects.

Bone Turnover Markers in Diabetes

Bone turnover markers in patients with diabetes have been examined in a meta-analysis (Starup-Linde et al. 2014). Both osteocalcin and CTX were decreased in patients with diabetes. compared to non-diabetes controls, whereas NTX was borderline significantly increased in diabetes patients. 25 hydroxy vitamin D levels were lower in diabetes patients, and phosphate levels were increased in patients with diabetes. PTH, calcium, and BAP were not different from controls in patients with diabetes. When stratifying by diabetes type, patients with type 1 diabetes had lower 25 hydroxy vitamin D and osteocalcin, whereas patients with type 2 diabetes had lower phosphate levels and borderline decreased osteocalcin compared to non-diabetes subjects (Starup-Linde et al. 2014). Further studies add to a decreased bone turnover in both patients with type 1 and type 2 diabetes and report that BAP is not decreased, when other bone markers were decreased (Starup-Linde and Vestergaard 2015). Thus, bone turnover in diabetes is altered in comparison to non-diabetes individuals with lower CTX and osteocalcin levels representing lower bone resorption and bone formation. BAP, which represents mineralization, was not different, thus the bone matrix mineralization seems not to be impaired.

All bone turnover markers displayed heterogeneity between studies (Starup-Linde et al. 2014). The heterogeneity may be due to differences in patient characteristics, due to analytical and preanalytical factors, due to using different assays (no marker was evaluated with same method through all studies), or due to differences in p-glucose levels in the patients with diabetes. An in vitro study revealed that the decrease in bone turnover markers is not due to an immunochemical masking effect by bone marker glycation, as addition of glucose to serum samples with increasing dose and incubation time did not change P1NP, osteocalcin, and CTX (Starup-Linde et al. 2014). The heterogeneity among bone turnover markers also makes them unreliable fracture predictors as they may change depending on p-glucose. However, decreased osteocalcin levels and increased P1NP/CTX ratio have been associated with fractures in patients with type 2 diabetes (Starup-Linde and Vestergaard 2015).

In Vitro Studies

The Effect of Glucose on Osteoblasts

Osteoblast-like cells have been exposed to different hyperglycemic conditions, and indices of bone turnover have been assessed. Table 3 presents the studies that have evaluated the addition of glucose to osteoblast-like cells. In human osteoblast-like cells, hyperglycemia of both 12 mmol/l and 24 mmol/l for 7 and 14 days, respectively, increased the matrix calcification. The quality of the mineral was reduced with low calcium phosphate ratios (Garcia-Hernandez et al. 2012). Alkaline phosphatase activity increased at a glucose level of 12 mmol/l but decreased at a glucose level of 24 mmol/l (Garcia-Hernandez et al. 2012). Both bone formation markers osteocalcin and runt-related protein 2 (Runx2), and the bone resorptive marker RANKL increased while OPG decreased; this suggests an overall increased bone turnover (Garcia-Hernandez et al. 2012). Two other studies using a different human cell line showed decreased proliferation, alkaline phosphatase activity, and expression of OPG but with glucose concentrations from 16.7 mmol/l to 49.5 mmol/l (Terada et al. 1998; Shao et al. 2014). Continuous glucose levels of 49 mmol/l are life threatening in vivo and even sustained levels above 20 mmol/l are unphysiological and lead to ketoacidosis or hyperglycemic hyperosmolar nonketotic coma. Studies investigating murine osteoblast-like cells exposed to glucose have reported varying results. Increased proliferation and increased matrix mineralization have been reported (Balint et al. 2001; Liu et al. 2015; Wu et al. 2012; Zhen et al. 2010); however, both decreased matrix mineralization (Zhen et al. 2010, Bartolome et al. 2013; Cunha et al. 2014; Ma et al. 2014) and unchanged matrix mineralization (Botolin and McCabe 2006) have also been reported. Alkaline phosphatase activity reflects the mineralization process and has been reported to be both decreased (Balint et al. 2001; Liu et al. 2015; Wu et al. 2012; Zhen et al. 2010; Cunha et al. 2014) and increased (Liu et al. 2015; Zhen et al. 2010; Botolin and McCabe 2006) during hyperglycemia. Runx2 is an important transcription factor in osteoblast differentiation and osteocalcin is a marker of osteoblasts activity. RANKL and osteocalcin have been reported decreased (Wu et al. 2012; Bartolome et al. 2013; Botolin and McCabe 2006; Zayzafoon et al. 2000) and increased (Liu et al. 2015; Zhen et al. 2010) in hyperglycemic circumstances. Furthermore, the OPG/RANKL pathway may be disturbed as OPG increased 30-fold during hyperglycemia and RANKL only increased two- to threefold, suggesting an inhibitory effect on bone resorption (Cunha et al. 2014). The Wnt pathway was also downregulated during hyperglycemia by decreasing β-catenin accumulation (Lopez-Herradon et al. 2013). Thus, hyperglycemia may decrease osteoblast differentiation and bone formation and also impair the bone resorption by increasing OPG.

Table 3 Studies that examined the effects of in vitro added glucose on osteoblasts
Full size table

Very different doses of glucose have been used in the studies to induce hyperglycemia ranging from 11 mmol/l to 49.5 mmol/l. The studies by Liu et al. and Zhen et al. (Liu et al. 2015; Zhen et al. 2010; Li et al. 2007) use different levels of hyperglycemia and present the importance of the glucose levels as both studies reported increased alkaline phosphatase activity and mineralization at the lowest level of hyperglycemia (15.5 mmol/l and 11 mmol/l) whereas higher glucose levels (22 mmol/l, 25.5 mmol/l, 35.5 mmol/l, and 44 mmol/l) decreased alkaline phosphatase activity and did not increase mineralization at the lowest levels of hyperglycemia. The glucose levels of 22 mmol/l or more are very high and is life threatening if sustained for longer periods, whereas levels of 11 mmol/l or 15 mmol/l may be tolerated for a longer period. The studies reporting decreased mineralization have used glucose levels higher than 16 mmol/l. The effect of glucose on bone markers may thus depend on the glucose levels; small increases of glucose may increase alkaline phosphatase activity and increase mineralization, whereas supraphysiological levels of glucose may decrease mineralization and alkaline phosphatase activity.

The Effect of Glucose on Osteoclasts

CTX is a marker of bone resorption and as CTX decreases during glucose intake, glucose may directly affect the osteoclasts. Table 4 presents the studies that have examined the effect of hyperglycemia on osteoclasts. Only three studies have examined this relationship and all on murine cells. Hyperglycemia was shown to have a detrimental effect when directly added to osteoclast-like cells. Hyperglycemia decreased the number of osteoclasts, TRAP expression, osteoclastogenesis, cell to cell fusion (which is an important step in creation of the multinucleated osteoclasts), and osteoclast differentiation (Wittrant et al. 2008; Xu et al. 2013, 2015). Furthermore, a decreased pit resorption area was observed during hyperglycemia. This reflects an impairment in the ability of the osteoclast to resorp mineralized matrix at elevated glucose concentrations (Xu et al. 2013). These in vitro studies clearly show a direct detrimental effect of hyperglycemia on osteoclasts. Thus, the hyperglycemia inhibits osteoclasts, this is in line with the clinical human studies where glucose ingestion decreased CTX (Bjarnason et al. 2002).

Table 4 Studies examining the effects of in vitro added glucose on osteoclasts and osteocytes
Full size table

The Effect of Glucose on Osteocytes

Only a single study has assessed the effect of hyperglycemia on osteocytes. It is presented in Table 4. This study showed an increased expression of sclerostin protein, whereas RANKL was unchanged during hyperglycemia (Tanaka et al. 2015). The regulatory activity of osteocytes may thus also be affected by glucose, with an increased sclerostin production and thereby an inhibitory effect on the osteoblasts by blocking the Wnt pathway. Further research is needed to confirm the effect of glucose on osteocytes.

The Effect of Glucose on Mesenchymal Stem Cells

Both osteoblasts and osteoclasts are derived from the mesoderm, and mesenchymal stem cells may thus differentiate to both types. Only immortalized human mesenchymal stem cells proliferated during hyperglycemia. Primary human mesenchymal stem cells had lower proliferation rate but differentiated towards osteogenic cells during hyperglycemia over 4 weeks and had enhanced mineralization compared to cells exposed to lower glucose levels (Li et al. 2007). Murine mesenchymal stem cells decreased mineralization, TRAP, and alkaline phosphatase activity but increased collagen production when exposed to hyperglycemia (Dienelt and zur Nieden 2011). Glucose may thus affect the differentiation of mesenchymal stem cells to osteoblasts and osteoclasts (Table 5).

Table 5 Studies examining the effects of in vitro added glucose on mesenchymal stem cells
Full size table

Glucose and Diabetic Bone Disease

Ingestion of glucose decreased both bone resorption markers and bone formation markers in healthy individuals and a link between glucose and bone turnover markers was established. In patients with diabetes the increased fracture risk was not explained by the apparent normal or increased BMD. This paradox suggests that BMD is not equal to bone quality in patients with diabetes. The mineralization marker BAP remained unchanged while CTX and osteocalcin were decreased in diabetes patients compared to non-diabetes subjects. Thus, a dissociative bone remodeling may be present in patients with diabetes with a decreased bone resorption and bone matrix formation whereas the mineralization is relatively increased. In this state, BMD is increased in comparison to the quality of the bone, which also is the case in some osteopetrotic patients (Starup-Linde and Vestergaard 2015). A study examining bone biopsies from patients with type 2 diabetes reports reduced bone formation rate and mineralizing surface but normal adjusted apposition rate in comparison to controls (Manavalan et al. 2012). Therefore the decreased mineralizing surface is due to a decreased bone formation. The in vitro studies on both human and murine cells report direct effects of glucose on bone cells. A physiological hyperglycemia (as in many patients with diabetes) with glucose levels of 11–15 mmol/l increased the mineralization and alkaline phosphatase activity by osteoblasts, whereas higher glucose levels decreased mineralization and alkaline phosphatase activity. Furthermore, hyperglycemia was mainly reported to decrease the Runx2 and osteocalcin expression and thus leading to decreased bone formation, although some in vitro studies also report increased Runx2 and osteocalcin. Besides a direct effect on osteoblasts, hyperglycemia decreased osteoclastogenesis and osteoclast activity and thus impaired bone resorption and increased the osteocytes expression of sclerostin, which decrease bone formation by antagonizing the Wnt pathway.

In patients with diabetes the p-glucose level is cyclical depending on medication use and food intake. As the p-glucose level may rapidly increase, the bone formation and bone resorption may drop. However, the mineralization process seems to be spared, which may be caused by an increased mineralization in the p-glucose levels of 11–15 mmol/l. In patients with diabetes, the level of 11–15 mmol/l is a physiological level and may be sustained for longer periods. Furthermore, HbA1c was a positive effect modificator of BAP in a meta-regression analysis (Starup-Linde et al. 2014) and mineralization reflected by BAP may increase due to long-term hyperglycemia. The bone of patients with diabetes may be hypermineralized due to long-term hyperglycemia. Figure 2 presents the hypermineralization hypothesis. Hyperglycemia may decrease both bone formation and bone resorption by a direct inhibitory effect on osteoblasts and osteoclasts. During hyperglycemia, osteocytes produce more sclerostin, and osteoblasts may increase OPG production which will decrease bone formation and bone resorption, respectively. At p-glucose levels of 11–15 mmol/l, the hyperglycemia may increase the bone mineralization as reflected by both calcification and alkaline phosphatase activity but with low quality mineralized material. Therefore, a state where the mineralization does not correspond to the bone turnover may exist in patients with diabetes, and the bone is hypermineralized relatively to the bone material strength.

Fig. 2
figure 2

The hypermineralization hypothesis. Hyperglycemia decreases the osteoclasts’ ability to resorp bone directly and indirectly by increasing the OPG/RANKL ratio. Bone formation by the osteoblasts is directly inhibited by hyperglycemia reflected by decreased Runx2 and osteocalcin expression and indirectly by an increased production of sclerostin by the osteocytes. Both bone formation and resorption is decreased; however at hyperglycemic levels of 11–15 mmol/l the mineralization is increased by the osteoblast, whereas the quality of the mineralized matrix is decreased due to decreased calcium/phosphate ratio

Full size image

Potential Applications to Prognosis, Other Diseases or Conditions

The interaction between glucose and bone turnover markers is a potential predictive and prognostic marker of fracture risk in patients with diabetes. The hyperglycemia, which characterizes patients with diabetes, may have detrimental effects on bone turnover, bone composition, and bone strength. The combination of bone turnover markers and continuous glucose monitoring may be of use to determine the bone turnover in patients with diabetes and evaluate whether this may be prognostic of fracture. No therapy of diabetic bone disease is available. Strict glucose control may be beneficial for bone health in patients with diabetes however, it is unknown whether antiresorptive therapies that decrease bone resorption are beneficial in a state already characterized by decreased bone resorption.

Future investigations may apply the advanced techniques of continuous glucose monitoring to determine the circadian rhythm of bone turnover in patients with diabetes and its relation to p-glucose levels, which may be very different from what is seen in non-diabetes subjects. Furthermore, additional studies investigating the effect of glucose on osteoblasts, osteoclasts, and osteocytes are needed. Translation of these results into animal models is also important. Randomized controlled trials are needed to determine whether a specific antidiabetic treatment is beneficial for bone health and if antiresorptive treatment may be of use.

Summary Points

  • Diabetes mellitus patients have higher risk of bone fracture which cannot be explained by their bone mineral density.

  • Bone turnover in patients with diabetes is altered in comparison to non-diabetes individuals as they display lower C-terminal cross-linked telopeptide of type-I collagen and osteocalcin levels representing lower bone resorption and bone formation.

  • An oral glucose tolerance test induces a decrease in C-terminal cross-linked telopeptide of type-I collagen in healthy individuals, but the decrease is attenuated in patients with diabetes.

  • An intravenous glucose tolerance test does not induce the same reduction in C-terminal cross-linked telopeptide of type-I collagen as an oral glucose tolerance test, indicating that gastrointestinal hormone release influence the glucose-bone interaction.

  • Hyperglycemia may decrease osteoblast differentiation and bone formation and may also impair the bone resorption by increasing osteoprotegerin and inhibiting osteoclast activity directly.

  • The effect of glucose on bone markers may be dependent on the glucose level, where small increases in blood glucose may increase alkaline phosphatase activity and mineralization, whereas supraphysiological levels decrease mineralization and alkaline phosphatase activity.

  • The bone of patients with diabetes may be hypermineralized due to hyperglycemia that increase bone mineralization relatively to the bone turnover.