Effects of Incretin Therapy on Skeletal Health in Type 2 Diabetes—A Systematic Review

ABSTRACT Diabetes poses a significant risk to bone health, with Type 1 diabetes (T1D) having a more detrimental impact than Type 2 diabetes (T2D). The group of hormones known as incretins, which includes gastric inhibitory peptide (GIP) and glucagon‐like peptide 1 (GLP‐1), play a role in regulating bowel function and insulin secretion during feeding. GLP‐1 receptor agonists (GLP‐1 RAs) are emerging as the primary treatment choice in T2D, particularly when atherosclerotic cardiovascular disease is present. Dipeptidyl peptidase 4 inhibitors (DPP‐4is), although less potent than GLP‐1 RAs, can also be used. Additionally, GLP‐1 RAs, either alone or in combination with GIP, may be employed to address overweight and obesity. Since feeding influences bone turnover, a relationship has been established between incretins and bone health. To explore this relationship, we conducted a systematic literature review following the PRISMA guidelines. While some studies on cells and animals have suggested positive effects of incretins on bone cells, turnover, and bone density, human studies have yielded either no or limited and conflicting results regarding their impact on bone mineral density (BMD) and fracture risk. The effect on fracture risk may vary depending on the choice of comparison drug and the duration of follow‐up, which was often limited in several studies. Nevertheless, GLP‐1 RAs may hold promise for people with T2D who have multiple fracture risk factors and poor metabolic control. Furthermore, a potential new area of interest is the use of GLP‐1 RAs in fracture prevention among overweight and obese people. Based on this systematic review, existing evidence remains insufficient to support a positive or a superior effect on bone health to reduce fracture risk in people with T2D. © 2023 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.


Understanding the incretin system: implications for bone health
To comprehensively assess the impact of incretin therapy on bone health, it is crucial to grasp the functioning of incretins and their effects on the skeleton.
The incretins are a group of hormones released by the gastrointestinal tract in response to nutrient intake.Their primary role is to regulate insulin release in response to feeding, thereby modulating blood glucose levels.
The incretins comprise 1. Glucose-dependent insulinotropic polypeptide, formerly known as gastric inhibitory peptide (GIP): GIP is secreted by the enteroendocrine K cells in the small intestine and has local inhibitory effects on gastric acid secretion.Moreover, GIP stimulates insulin secretion in a glucose-dependent manner, contributing to the lowering of blood glucose. (1). Glucagon-like peptide 1 (GLP-1) is secreted by the enteroendocrine L cells in the small and large intestines and is rapidly degraded by dipeptidyl peptidase 4 (DPP-4).GLP-1 stimulates insulin secretion in a glucose-dependent manner and inhibits gastric emptying. (2)spite belonging to the glucagon superfamily, GLP hormones suppress endogenous glucagon secretion, further contributing to the glucose-lowering effect. (3)iabetes is associated with compromised bone health, although recent studies have shown improvements in fracture risk. (4,5)The interplay between blood glucose, insulin, feeding, and bone health is complex. (6)(9) A somewhat smaller effect of feeding is observed on formative markers, among others procollagen I N terminal propeptide (P1NP).The effect of feeding on bone turnover markers can be negated by the somatostatin analogue octreotide. (10)n healthy control people, an oral glucose tolerance test (OGTT) suppresses CTX by approximately 50% compared to fasting levels.However, intravenous infusion of glucose keeping the same glucose levels as during the OGTT (isoglycemic intravenous glucose infusion [IIGI]) only marginally suppressed CTX (slightlymore than 30%) compared to the fasting state (a decline of around 30%).For P1NP no difference was present between IIGI and the fasting state, whereas the OGTT decreased P1NP by around 5%. (11) GLP-1 levels were similar during fasting and the IGII, whereas they-along with GIP-increased during the OGTT compared to the IGII and fasting. (11)As a significant correlation was present between GIP and nadir CTX, these observations demonstrate an effect of incretins on bone turnover.GIP was higher during the IIGI than during fasting and lower during the OGTT. (11)As mentioned earlier, octreotide, which inhibits incretins, negates the effects of the OGTT on bone turnover markers. (10)A mixed meal test suppresses CTX by more than that observed during the OGTT. (9)The suppression of P1NP is inversely dependent on insulin resistance, though this does not seem to be the case for CTX. (9)Likewise, fasting P1NP seems to be inversely dependent on insulin resistance. (9)oreover, weight loss is a keystone in the prevention and treatment of Type 2 diabetes (T2D). (12)(15) In people with T2D, a modest (and recommended) weight loss of approximately 7% over 1 year is found associated with a significant loss of bone mass that persists even if weight is maintained for the next 3 years. (16)The loss of bone loss might also result in a further increased risk of fractures in these patients. (17)The use of GLP-1 RAs as treatment for T2D has emerged and shows beneficial aspects in both glycemic controls, as a weight loss facilitator, and in the prevention of other diabetes-related comorbidities.Thus, it seems reasonable to investigate the potential impact on bone metabolism in people with T2D.
Collectively, incretin receptors are widely expressed, suggesting effects of incretin beyond the regulation of glucose homeostasis.Associations between incretin hormones and bone metabolism have emerged and opened up an interesting possibility of the interplay between feeding, obesity, T2D, and bone health.Frequent and excessive feeding may lead to prolonged suppression of bone turnover and increased insulin resistance.Likewise, significant weight loss might also induce a concomitant loss of bone mass.Consequently, incretin therapy may have the potential to impact bone health.

Incretin therapy
In clinical practice, incretin therapy primarily involves the use of GLP-1 RAs or DPP-4is, commonly known as gliptins.There are several GLP-1 RAs available, each with varying degrees of similarity to native GLP-1 and different potencies for reducing HbA1c levels.Likewise, there are multiple DPP-4is available on the market.Generally, DPP-4is have a less potent glucose-lowering effect compared to GLP-1 RAs.However, unlike GLP-1 RAs, which are typically administered via subcutaneous injections, DPP-4is can be taken orally.
GLP-1 RAs may be employed for the treatment of T2D, particularly in people with severe insulin resistance (SIRD) often associated with significant overweight. (18)Notably, GLP-1 RAs have also received therapeutic approval for managing obesity in the absence of diabetes. (19)) ) Currently, the following GLP1-RAs are approved by the US Food and Drug Administration and the European Medicines Agency for the treatment of T2D: dulaglutide, exenatide, liraglutide, lixisenatide, and semaglutide.Approved DPP-4is are sitagliptin, saxagliptin, linagliptin, and alogliptin. (23)Vildagliptin is only EMA approved.Tirzepatide is the first approved GIP and GLP1-RA combination therapy for T2D.Only liraglutide and semaglutide are approved for treatment of obesity without existing diabetes.

Diabetes and bone health
Type 1 diabetes (T1D) and T2D are associated with impaired bone health; in particular, a discrepancy between BMD and fracture risk has been reported. (5)In general, people with T1D are more severely affected by lower BMD and higher fracture risk than people with T2D. (5)In people with T2D, only some fracture types may be increased, whereas others may be lower than in the general population, perhaps due to a protective effect of being overweight. (24)ncretins are mainly used to treat people with T2D and were recently upgraded from second-line therapy in addition to or substituting metformin to a more prominent role in case of intended weight loss or presence of atherosclerotic cardiovascular disease. (25)Incretins have also been used in T1D, although to a much lesser degree, and are not standard of care at present.
It is important to acknowledge that T2D encompasses various subtypes, ranging from the classic phenotype characterized by severe insulin resistance often associated with significant overweight and multiple diabetes-related complications to milder forms such as obesity-related (MOD) or age-related (MARD) subtypes, as well as severe insulin-deficient subtypes (SIDD). (18)hese different T2D phenotypes can have implications for bone health, considering that body weight itself is linked to fracture risk and BMD, as mentioned earlier.Therefore, when evaluating the choice of drug and comparator drugs, such as GLP-1 RAs or DPP-4is, these factors should be taken into consideration. (26)

Inclusion and exclusion
See Figure 1A,B (PRISMA flow) for flow of inclusion and exclusion of studies.Systematic reviews and meta-analyses were given priority.If original studies were not included in the meta-analyses, either due to oversight or being more recent than the meta-analyses themselves, they were still cited.Older metaanalyses were cited alongside more recent ones if there were discrepancies in their findings.If older meta-analyses aligned with the newer ones, they were not cited.Additionally, other relevant articles referenced within the included records were reviewed for eligibility.While studies involving people with diabetes were prioritized, studies on nondiabetic people were also considered for comprehensive evaluation.Among studies involving humans, randomized controlled trials (RCTs) took precedence over observational studies.Regarding fractures, only human studies were considered eligible.It is worth noting that studies were often limited and frequently did not focus specifically on people with T2D.Finally, GLP-1 RAs were given priority.Additionally, an overview of included studies regarding fracture risk, bone turnover, and bone structure for GLP-1 RAs and DPP-4is are presented in Tables 1-4.Furthermore, some overlap was seen for eligibility between GLP-1 RAs and DPP-4is as several studies included both exposures.Hence, 44 studies were reviewed for GLP-1RAs and 41 for DPP-4is.

Fractures
In general, a cohort study comparing GLP-1 RAs and DDP-4is showed no significant difference in incident major osteoporotic fractures. (27)Using the same cohort, the authors reported no difference in major osteoporotic fracture risk between GLP-1 RAs or SGLT2is (both in combination with metformin). (28)Another cohort study using people aged ≥65 years from Medicare reported that initiation of SGLT-2 inhibitors was not associated with an increased risk of fracture in older T2D adults compared with initiating a GLP-1 RA. (29) Compared with insulin, a network meta-analysis using data from RCTs did not report any increase in fracture risk with GLP-1 RAs, (30) although a meta-analysis using observational studies concluded that current GLP-1 RA use was associated with a decreased risk of fracture. (31)This finding was consistent with the results from two additional cohort studies conducted by the same group of authors. (32,33)n older meta-analysis found no reduction in fracture risk by GLP-1 RAs in T2D compared to other antidiabetic medications. (34)Interestingly, another older meta-analysis of RCTs reported that, compared with placebo and other antidiabetic drugs, liraglutide and lixisenatide were associated with a significant reduction in the risk of bone fractures, and the beneficial effects were dependent on the treatment duration, with effects only observed after more than 1 year. (35)Yet another metaanalysis also looked at subtypes of drugs from RCTs and reported that albiglutide (GLP-1-RA) decreased the risk of fracture (relative risk [RR] of 0.29 [95% CI: 0.04 to 0.93]) compared to placebo. (36)one of the other subtypes of GLP-1 RAs were significantly associated with fracture risk.Additionally, a newer meta-analysis of RCTs reported that GLP-1 RAs were associated with a decreased fracture risk compared to placebo or other antihyperglycemic drugs, and the reduction was highest for exenatide. (37)Yet an older meta-analysis of RCTs reported that liraglutide was associated with a reduced risk of incident fractures (odds ratio [OR] of 0.38 [95% CI: 0.17 to 0.87]), whereas exenatide was associated with an elevated risk (OR of 2.09 [95% CI: 1.03 to 4.21]). (38)his section presents a collection of cohort studies, metaanalyses, and network meta-analyses that investigate the association between GLP-1 RAs with the risk of fractures in people with T2D. Bone turnover and bone structure

Human studies
The section presents various clinical trials and studies that investigate the effects of GLP-1 RAs in human studies, particularly liraglutide and exenatide, on bone metabolism, BMD, and bone turnover markers in people with diabetes and in overweight.
The effects of liraglutide on bone resorption and turnover was exploited in two trials.One RCT that included T2D people found that liraglutide treatment for 6 months did not affect bone resorption despite a significant weight loss compared to the control group without weight loss. (39)These observations may suggest that liraglutide expresses antiresorptive effects.Moreover, a trial on healthy obese people exposed to liraglutide for 52 weeks after a 12% weight loss intervention reported increased P1NP and no change in CTX in the liraglutide group compared to the control group, which indicates a higher bone formation. (40)he exenatide effects on bone metabolism in healthy T1D people were discussed in three studies.One human trial included healthy normal-weight controls and showed suppression of CTX after exenatide injection. (41)Another study in T1D           people concluded that, despite an exenatide-induced body weight reduction, no changes in bone metabolism were observed with exenatide added to insulin therapy after 26 weeks. (42)A double-blinded RCT investigated the effects of exenatide 2 mg once weekly (n = 23) or placebo (n = 22) on bone turnover markers in chronic, obese, antipsychotic-treated people. (43)They reported, in general, no significant effects on bone turnover markers (P1NP and CTX) after 3 months of treatment.However, they did observe a numerical reduction of both PINP and CTX in the exenatide group and a numerical increase in the placebo group. (43)These findings suggest that weightinduced bone turnover is mitigated by GLP-1.
The effects of liraglutide on bone metabolism in healthy and obese.A clinical trial included people (without diabetes) with common obesity and people with one of the most common monogenic causes of obesity in humans, a mutation in the melanocortin-4 receptor (MC4R). (44)They observed a 6% weight loss in both groups but no differences in markers of bone turnover (CTX, osteocalcin and P1NP) after 16 weeks of liraglutide treatment (neither in groups of between groups).Although not directly related to diabetes, the use of liraglutide in people with prediabetes due to olanzapine-or clozapine-treated schizophrenia did not change bone turnover markers. (45)Atrial that involved obese people with and without an MC4R mutation, it was observed that the control group exhibited an increase in bone mass (measured by mineral apparent density) in response to liraglutide treatment.However, no such increase was observed in the MC4R group. (44)o elucidate the effects of exenatide and dulaglutide on BMD in people with diabetes, a single-blind study in overweight T2D people randomized to exenatide, dulaglutide, insulin glargine, or placebo reported a BMD increase of the total hip in the exenatide group.In the dulaglutide-treated group, only the femoral neck BMD decreased, but the magnitude of the decrease was less than that observed in the placebo group; the BMD of the first to fourth lumbar vertebrae (L1-L4), femoral neck, and total hip decreased significantly in the placebo group, while in the insulin glargine group, the BMD of L2, L4, and L1-4 increased. (46)Compared with the placebo group, the BMD of the femoral neck and total hip increased significantly in the exenatide group and the insulin glargine group.Compared with the exenatide group, the BMD of L4 in the insulin glargine group increased as well. (47)n the RCT on T2D people randomized to liraglutide or placebo, mentioned earlier, liraglutide treatment prevented a hip BMD decrease (despite a weight loss), as was otherwise observed in the control group. (39)The effect was observed as early as 13 weeks after initiation.No differences were observed for lumbar spine BMD or in the evaluation of bone microstructure by HR-pQCT. (39)LP-1 RA treatment on BMD in T2D people was first evaluated in a retrospective cohort study that included people with diabetes and compromised bone quality (T-score < À1), the effects of switching from a DPP-4i to a GLP-1 RA were examined.Although an improvement of glycemic control and a larger weight loss was observed, the GLP-1 RA group experienced a more significant decrease in the lumbar spine BMD than the group continuing DPP-4i treatment (À0.028 g/cm 2 versus À0.019 g/cm 2 , p = 0.041, adjusted for body mass index [BMI]).(48) However, in the study examining bone turnover markers, a 6-month exposure to exenatide in people with T1D did not result in any changes in BMD, as assessed through measurements of the whole body, hip, lumbar spine, and forearm.(42) Furthermore, in a trial involving obese people without diabetes who were on antipsychotic treatment, a 3-month exposure to exenatide did not lead to any changes in BMD in the lumbar spine, femoral neck, or total hip.(43) Lastly, the effect of switching from DPP-4i to GLP-1 RA on BMD was seen in a trial conducted on healthy obese people, the effects of liraglutide over a period of 52 weeks were investigated.It was found that during weight maintenance, the control group experienced a decrease in total, pelvic, and arm-leg BMC.However, no changes were observed in these measurements in the liraglutide treatment group.(40) Overall, the studies suggest that GLP-1 RAs, particularly liraglutide (39,40,44,45) and exenatide, (41)(42)(43)46,47) may have varying effects on bone metabolism and BMD in different populations, including T2D people and obese people, although the effect of the drugs is generally neutral.The impact on bone health appears to be influenced by factors such as weight loss, genetic mutations, and the presence of diabetes.

Animal studies
This section provides an overview of several animal studies investigating the effects of GLP-1 RAs, particularly liraglutide and exenatide, on bone health, BMD, bone turnover markers, and bone strength.
First, the effects of liraglutide on bone remodeling and bone loss was evaluated.An animal study on rats showed that liraglutide improved bone remodeling parameters by increasing the bone formation marker osteocalcin and the activity of alkaline phosphatase (ALP) while decreasing CTX (a bone resorption marker) with no change in P1NP (another bone formation marker). (49)In ovariectomized mice, both liraglutide and a quinoxaline-based compound prevented bone loss, as evidenced by increased P1NP and decreased CTX levels, indicating enhanced bone formation and reduced bone resorption.However, these effects were observed only with exenatide treatment, not with liraglutide. (50,51)A study evaluated the effects on bone mediated by the SGLT2 inhibitor dapagliflozin versus liraglutide and reported a different but positive impact on bone microarchitecture and material properties by both drugs that was not explained by the lowering of blood glucose. (52)They reported that liraglutide administration resulted in higher mature collagen crosslinks and lower collagen glycation as well as restoring a normal postprocessing of collagen molecules.These findings may indicate the ability of GLP-1 RA to restore the immature divalent collagen crosslinks seen in diabetes.
Then the effects of exenatide on bone resorption and bone healing was assessed.A study using exenatide on mice reported a lowering of bone resorption measured by CTX and the receptor activator of nuclear factor-κB ligand (RANKL); the latter acts as an essential part of osteoclast formation and activation. (53)The study further investigated any potential mechanism and found that GLP-1 RAs may affect the inhibition of the lipopolysaccharide (LPS)-induced TNF-α expression of macrophages more than a direct effect on osteoclast precursors or RANKL expression.LPS promotes inflammation and inflammatory bone loss and induces the production of proinflammatory cytokines, e.g., TNF-α, and TNF-α enhances osteoclast formation and RANKL expression on stromal cells. (54)57) The effects of GLP-1 RAs on bone quality in various mice models was assessed in several studies.One study used a leptin receptor-deficient mice model with obesity and severe T2D to JBMR Plus (WOA) assess changes in bone quality after 4 weeks of exenatide administration. (58)They reported an increased bone formation rate in leptin-deficient mice but no effect on lean mice also exposed to exenatide treatment.They did not observe an effect on bone resorption.Then, a study on mice with a femoral defect treated with a GLP-1 polymer found an increased number of osteoblasts and decreased number of osteoclasts with corresponding effects on the expression of formation and resorption markers, suggesting that the GLP-1 polymer enhances bone formation. (59)A study from 2008 investigated GLP-1 RA knockout mice and reported lower total and cortical BMD evaluated by a computed tomography (CT)-based analysis with no difference in the trabecular evaluation. (60)The study on mice with a femoral defect also analyzed by micro-CT (μCT) and observed higher trabecular number, trabecular thickness, and bone volume to total volume (BV/TV) ratio in the mice treated with a polymer of GLP-1. (59)These findings suggest a promoted bone formation and a potential increasing rate of bone healing facilitated by GLP-1.
GLP-1 RAs' effects on BMD and bone strength in several mice models.In a short-term animal study lasting 5 weeks, it was found that the antidiabetic treatment using exenatide as a GLP-1 RA and sitagliptin as a DPP-4i restored bone elasticity in comparison to untreated diabetic rats.However, the treatment resulted in a deterioration of bone strength. (61)Both the untreated diabetic rats and the treated rats showed significantly higher bone bending stress, but this was improved by the treatment, with exenatide demonstrating a slightly more pronounced effect than sitagliptin. (61)Another study involving diabetic rats reported improvements in trabecular and cortical bone measures after 4 weeks of liraglutide treatment. (62)In a study conducted on ovariectomized mice, both liraglutide and exenatide were found to enhance bone mass, as indicated by increased trabecular bone indices measured using μCT. (50)Notably, no changes were observed in cortical bone measurements.Another animal study demonstrated that liraglutide prevented the deterioration of trabecular microarchitecture and improved bone strength. (49)Additionally, a different animal study showed that a GLP-1 RA improved bone loss primarily through the induction of bone formation. (51)Studies on nondiabetic mice also revealed that exenatide increased the transcription of osteogenic differentiation-related genes, promoted osteogenic differentiation and bone repair, and improved BMC. (63,64) In a short-duration study conducted on animals with T1D lasting 8 weeks, liraglutide treatment, either alone or in combination with insulin, was found to restore decreased BMD and partially correct compromised trabecular microarchitectures compared to control animals without T1D without any changes in bone turnover markers. (65)A trial involving ovariectomized rats with streptozotocin-induced diabetes reported a reduction in femoral BMD and the destruction of bone microarchitecture, which was alleviated by liraglutide treatment along with decreased RANKL/OPG ratio. (46)Evaluation based on counted numbers of osteoblasts and osteoclasts using H&E-and TRACP-stained sections was used. (46)Another trial conducted on ovariectomized rats without diabetes found that liraglutide improved trabecular volume, thickness, and number, increased BMD, and reduced trabecular spacing in the femurs, as determined by μCT analysis.Similar results were observed in the lumbar vertebrae. (66)In a study involving leptin receptor-deficient mice exposed to exenatide, improved trabecular bone structure was observed.Consistent with the aforementioned studies, no differences were found in the cortical bone parameters following exenatide treatment. (58)Lastly, biomechanical bone strength was evaluated in six studies.Two studies found an improvement of bone strength by GLP-1 RAs (51,57) one study found that bone elasticity was restored but bone strength deteriorated, (61) and two studies found no effect on bone strength. (52,58)Though results were inconsistent, most animal studies suggested an improved or no effect on bone strength after GLP-1 RA therapy.

Cell studies
The section discusses several cell culture studies conducted on rodents, including bone marrow mesenchymal stem cells, bone marrow-derived macrophages, fibroblasts, preosteoclasts, and osteoblasts, to investigate the presence and effects of GLP-1 RAs on bone remodeling processes.
As for the presence and effects of GLP-1RAs on bone cells, several studies on cell cultures from rodents, including bone marrow mesenchymal stem cells, bone marrow-derived macrophages, fibroblasts, and preosteoclasts, demonstrated that the presence of GLP-1 receptors and GLP-1 RAs (mainly exenatide) could potentially inhibit bone resorption and promote osteogenesis and osteoblast proliferation. (50,57,67)These findings suggest that GLP-1 binds to and modulates cells involved in bone remodeling in favor of formation by increasing osteoclast number and serum CTX.However, mixed evidence exists on the effect of GLP-1 receptor presence on osteoblasts, with one study reporting that GLP-1 receptors were not identified on primary osteoblasts cultured from rat bone marrow stem cells but another study reporting the presence of GLP-1 receptors on osteoblasts cultured from murine preosteoblasts. (68)However, yet another study reported that exenatide only influenced bone nodule formation of cultured primary murine osteoblasts during high glucose concentrations. (58)nother study exploited human dental pulp-derived stem cells and observed increased osteoblast differentiation during liraglutide treatment, high activity of bone ALP, and increased expression of osteoblast marker genes, e.g., Runx2, type 1 collagen, osteonectin, and osteocalcin. (69)verall, the cell culture studies suggest that GLP-1 RAs are present on various bone cells, and GLP-1 RAs, especially exenatide and liraglutide, have the potential to promote bone formation and inhibit bone resorption.However, the presence of GLP-1 RAs on osteoblasts remains a subject of debate due to some contradictory findings.These studies provide valuable insights into the potential mechanisms through which GLP-1 RAs might impact bone metabolism, supporting the notion that these drugs could have beneficial effects on bone health.

Fractures
The section presents a collection of cohort studies, meta-analyses, and network meta-analyses that investigate the association between the use of DPP-4is and the risk of fractures in people with diabetes.
Regarding specific subtypes of incretins, a meta-analysis suggested that the risk of fracture for alogliptin was decreased compared to placebo (OR of 0.51 [95% CI: 0.29 to 0.88]). (72)Aloglitpin also reportedly reduces the fracture risk compared with linagliptin (OR of 0.45 [95% CI: 0.20 to 0.99]) and saxagliptin (OR of 0.46 [95% CI: 0.25 to 0.84]); the risk was higher with saxagliptin versus sitagliptin (OR of 1.90 [95% CI: 1.04 to 3.47]) and sulfonylureas (OR of 1.98 [95% CI: 1.06 to 3.71]).In a direct pairwise metaanalysis, alogliptin was associated with a nonsignificant tendency to fracture risk reduction compared with placebo (OR of 0.54 [95% CI: 0.29 to 1.01]). (72)In addition, from the first metaanalysis, a subgroup analysis revealed that sitagliptin 100 mg daily (OR of 0.495 [95% CI: 0.30 to 0.80]) was associated with a reduced risk of fractures. (70)Interestingly, among people with T2D and long-term usage of DPP-4is as a second-line antidiabetic drug, a decreased fracture risk, including a decreased risk of upper arm fractures, was observed for a period of up to 5 years compared to those without DPP-4i use. (73)omparison of various antidiabetic medications on fracture risk reveals divergent and intercontinental differences.A cohort study on T2D people observed no statistically significant increase in fracture risk with SGLT2i use compared with DPP-4i use after 1 year. (74)dditionally, another study found that treatment with metformin (hazard ratio [HR] of 0.88 [95% CI: 0.85 to 0.92]) and DPP-4is (HR of 0.93 [95% CI: 0.88 to 0.98]) was associated with a reduced fracture risk, while insulin (HR of 1.26 [95% CI: 1.21 to 1.32]), thiazolidinediones (HR of 1.23 [95% CI: 1.18 to 1.29]), and meglitinides (HR of 1.12 [95% CI: 1.00 to 1.26]) were associated with an increased risk (p value <0.05). (75)However, a Japanese study reported that both insulins, alpha-glucosidase inhibitors, and DPP-4i use were related to increased hip fracture risk compared to metformin.The risk of vertebral fractures was higher in people prescribed insulin, thiazolidine, and DPP-4is compared with metformin. (76)In contrast, another Japanese study reported a reduced fracture risk among those exposed to DPP-4is. (77)A Danish retrospective cohort study did not observe a significantly different risk of osteoporosis-related fractures between people exposed to DPP-4i, (27) while a Taiwanese cohort study reported that the risk of osteoporosis was significantly lower among people exposed to DPP-4is compared with those without DPP-4i treatment (HR of 0.616 [95% CI: 0.358 to 0.961; p = 0.011]).Kaplan-Meier analysis showed that the preventive effect on osteoporosis was positively correlated with the cumulative dose of DPP-4i (log-rank, p = 0.039) (78) ; although no fracture outcomes were reported, these findings may be relevant as the fracture risk seems to increase when T2D and osteoporosis coexist. (79)A Korean study reported that initiating an SGLT2i was not linked with increased fracture risk compared to DPP-4is. (80)This was backed by another Korean study reporting no difference in fractures between DPP-4i users and nonusers. (81)Similarly, another study found that initiation of a SGLT2i versus a DPP-4i was not associated with a higher risk of fractures regardless of estimated glomerular filtration rate (eGFR) among people with renal impairment. (82)In agreement with the two aforementioned studies, an American study did not report differences in fracture risk between users of SGLT2is versus DPP-4is. (83)Lastly, a cohort study reported that the SGLT2i empagliflozin was associated with a similar risk of fractures compared to the use of DPP-4is. (84)No discrepancies in fracture risk were reported between users of DPP-4is and SGLT2is in a large cohort study, and similar findings were reported by another study. (85,86)stly, it has been reported that significantly fewer falls and fractures occurred with linagliptin treatment compared to the sulphonylurea (SU) glimepiride. (87)However, glimepiride may per se be associated with hypoglycemia and thus increase the risk of falling.In contrast to these findings, another study reported that DPP-4is were not associated with an increased risk of fragility fractures compared with SU or insulin; however, they were associated with a lower risk versus thiazolidinediones. (88)verall, several meta-analyses (70,71) and cohort studies (27,74- 78,81,84,85) demonstrated that DPP-4is were either decreased or not associated with an increased risk of fractures. (70,71)Only one study reported an increase risk. (76)Additionally, subgroup analyses showed that certain specific DPP-4is, such as alogleptin (72) and sitagliptin, (70,72) were associated with a reduced risk of fractures.Furthermore, intercontinental differences were observed in the association between antidiabetic medications and fracture risk.However, it is important to note that factors such as study design, population characteristics, and medication dosage may contribute to the variations in findings.

Human studies
This section includes three studies examining the effects of DPP-4is on bone health in people with T2D.
Effects of vildagliptin on bone markers.A RCT investigated the effects of vildagliptin, a DPP-4i, on bone markers. (89)The study found that 1 year of vildagliptin exposure did not result in any significant changes in postprandial serum CTX concentrations (a marker of bone resorption) compared to pretreatment levels.Similarly, fasting serum ALP, calcium, and phosphate were also unaffected by 1 year of vildagliptin treatment.These results suggest that vildagliptin did not have a substantial impact on bone resorption markers or bone metabolism in the studied population.
Association of plasma DPP activities with bone health.A human study reported that elevated plasma DPP activities were associated with a higher proportion of osteoporosis/osteopenia in people with newly diagnosed T2D. (90)While not directly linked to DPP-4i use, this finding suggests a potential relationship between DPP activity and bone health in people with T2D.Elevated DPP activity may be an indicator of underlying mechanisms that could influence bone health, warranting further investigation.
Trabecular Bone Score (TBS) and BMD in DPP-4i users was estimated by a retrospective study with T2D. (91)They found that exposed to DPP-4is had a higher TBS which is an index of bone microarchitecture.However, the BMD increased in both DPP-4 exposed and unexposed groups, with no significant differences between the groups.The study suggests that DPP-4i use might be associated with improved trabecular bone microarchitecture, but it did not show a significant difference in BMD between DPP-4i users and non-users.
Overall, the findings from these studies suggest that DPP-4i use might have some effects on bone health in people with T2D, but the evidence is still limited and inconclusive.

Animal studies
This section includes several animal studies investigating the effects of various DPP-4is on bone health in different diabetic animal models.
The effects of linagliptin on bone health were observed in diabetic mice.In a mouse study using a high-fat diet to induce diabetes, impaired bone microarchitecture, reduced BMD, and altered bone turnover biomarkers were observed. (92)Treatment with linagliptin alone and in combination with metformin significantly improved bone architecture, BMD, and bone turnover biomarkers.However, metformin alone did not show significant improvement in bone health.The results suggest that linagliptin, especially when used in combination with metformin, positively modulates bone health in diabetic mice.
The effects of vildagliptin, pioglitazone, and sitagliptin on bone health were studied in diabetic rats.A 5-week study in male Zucker Diabetic Fatty (ZDF) rats used different treatment groups, including vildagliptin, pioglitazone, and their combination. (92)he pioglitazone group showed decreased osteocalcin levels and increased TRACP 5b, indicating impaired bone metabolism.The results suggest that pioglitazone adversely affects bone health in diabetic rats.
A study using Wistar rats reported that sitagliptin-treated diabetic animals had significantly lower serum levels of CTX-I (a bone resorption marker) compared to untreated diabetic animals. (93)μCT analysis showed that sitagliptin prevented cortical bone growth stagnation in diabetic rats, resulting in stronger femora during three-point bending.However, another study in T1D rats reported that sitagliptin administration did not reverse the negative effects of T1D on bone indices, such as trabecular number and thickness. (94)Additionally, the study on ZDF rats reported that vildagliptin treatment significantly increased BMD and trabecular bone volume.The combination therapy restored BMD, trabecular bone volume, and trabecular bone thickness, which were otherwise decreased by pioglitazone alone. (95)Furthermore, the study using Wistar rats reported sitagliptin prevented cortical bone growth stagnation in diabetic rats assessed by micro-CT, resulting in stronger femora during three-point bending. (93)astly, the effects of DPP-4i co-administration with LPS in mice was studied and resulted in a lower osteoclast number and decreased bone resorption compared to LPS administration alone. (96)This suggests that DPP-4i may have a protective effect on bone resorption during inflammatory conditions.
Overall, animal studies provide valuable insights into the effects of various DPP-4is on bone health in different diabetic animal models.Linagliptin, (92) especially when used in combination with metformin, showed positive effects on bone health, improving bone architecture and turnover.4) Cell studies A combined in vitro and in vivo rat study in diabetes induced by a high-fat diet and streptozotocin reported that cell growth was disturbed while both the abnormal macrophage polarization and the endothelial impairment in diabetes were significantly alleviated by sitagliptin. (97)Additionally, DM animals showed angiogenesis inhibition and poor bone formation on the boneimplant interface, which were significantly ameliorated by sitagliptin treatment.
To summarize, the consensus from most studies suggests that the use of DPP-4is does not increase the risk of fractures.Some studies observed slight decreases in fracture risk with certain subtypes of DPP-4is, although differences exist.However, it's important to consider that the choice of comparator drugs, the duration of follow-up, and the specific indications for prescribing the drug being compared to the comparator might influence the findings.

GLP-1-RAs
In general, the studies included in this review indicate that GLP-1 RAs have a neutral to positive effect on bone health in people with and without T2D, whether for blood glucose regulation or weight loss therapy.Most of the presented studies describe a potential positive impact on bone health, as evidenced by various correlations observed in cell and animal studies.However, it should be noted that a definitive causal relationship demonstrating positive effects on bone health in humans has yet to be established.
GLP-1-RAs might prevent bone loss during weight loss assessed by evaluation of bone turnover markers, bone mass, and bone microstructure.The mechanism behind the positive effects of GLP-1 on bone has been discussed widely in several systematic and narrative reviews.There have been suggestions of a direct effect on osteoclasts and osteoblasts through intracellular signaling pathways (98,99) or indirect effect on bone cells through thyroid c cells and calcitonin-dependent inhibition of bone resorption. (100)Moreover, a potential incretin-mediated decrease of the chronic inflammation in T2D, which is believed to negatively modify bone tissue by nonenzymatic glycosylation, mineralization imbalance, and bone microdamage, has been suggested. (99)See Figure 2 for an overview of the cellular mechanisms.However, human studies including T2D people are scarce, and this includes diverse endpoints, making it difficult to compare findings and draw reasonable conclusions.In T2D, BMD measurements do not sufficiently predict low bone quality.The clinical presentation of bone fragility is fractures, which are therefore, the most favorable outcome measure in people with T2D.No detrimental effects with GLP-1 RAs were reported on the risk of fractures, although results were inconsistent, with heterogeneity between studies regarding people, study time, and comparison drugs.However, fracture assessment requires longterm follow-up, making it challenging to conduct intervention studies for assessment.Consequently, more sophisticated methods have emerged, i.e., HR-pQCT and microindentation.Several studies reported a positive effect on trabecular bone microstructure but no effect on cortical bone.However, these studies were primarily on rodents without T2D.It has been suggested that people with T2D have a compromised cortical bone structure expressed as a higher cortical porosity, making these results even more challenging to interpret. (101,102)Human trials that include people exposed to GLP-1 RAs have suggested that GLP-1 may prevent an otherwise expected decrease in hip BMD during significant weight loss.However, only one of the included studies evaluated microarchitecture by HR-pQCT, and no differences were observed.None of the identified human studies evaluated bone strength by microindentation.Although results were inconsistent, most animal studies suggested an improved or no effect on bone strength after GLP-1 RA therapy.Collectively, human studies of bone health in response to GLP-1 RA treatment are limited and trials including T2D people are warranted.As the validity of BMD and the cut-off value for the osteoporosis diagnosis with a T-score of À2,5 SD has been questioned in people with T2D, we propose that state-of-theart future research be conducted to focus on the assessment of bone indices using more advanced techniques, e.g., HR-pQCT and microindentation.However, in addition, studies that differentiate between different phenotypes of T2D with a focus on hard endpoints like fracture risk, fracture localization, and potential risk reduction during longer-term follow-up.Additionally, it is worth noting that not everyone may experience the beneficial effects of GLP-RA therapy (approximately 10% to 15%), while others may experience significant and rapid weight loss and improved blood glucose control (in the case of T2D only), which raises questions about whether similar effects on bone health are maintained. (103)rthermore, most studies on GLP-1 RAs and weight loss are conducted in conjunction with exercise and dietary advice.Exercise has a positive effect on weight maintenance by preserving muscle mass and overall benefits bone tissue. (104)Therefore, robust evidence is needed regarding the impact on body composition.Moreover, the specific doses or dose-response effects of GLP-1 RAs on bone health have not been thoroughly investigated, especially when considering the higher dosages used for weight loss compared to the treatment of T2D.
(107) Consequently, the modern-day treatment approach utilizing incretin hormones for both weight loss and T2D is still in its early stages.Therefore, this topic is complex, and the current evidence supports a neutral or even a positive effect of incretin hormones on bone health.The crucial question that remains is whether GLP-1 RAs represent a novel treatment for osteoporosis or merely a supplementary option for future strategies.

DPP-4is
DDP4is exhibit effects similar to those of GLP-1 RA therapy but are considered less effective at regulating blood sugar levels and have fewer side effects due to their indirect mechanism of action.It is important to note that DPP-4is are only prescribed to people withT2D and are not approved for weight loss.They are typically used in combination with other antidiabetic drugs and are rarely prescribed as monotherapy.Consequently, GLP-1 RA drugs are generally regarded as superior to DPP-4is.
Existing evidence indicates that DPP-4is are neutral for bone health, as they do not increase the risk of fractures and may even have some positive effects on bone structure.However, further human studies are necessary to gain a clearer understanding of the impact of DPP-4is on bone turnover, bone structure, and fracture risk.Future research should include longer-term follow-up and comparisons with other antidiabetic medications to better comprehend the role of DPP-4is in promoting bone health.
Although DPP-4is have their place in the treatment of T2D, their potential as a weight loss drug or their overall efficacy should be considered limited when compared to the numerous beneficial effects of GLP-1 RAs and potential future combinations.

Conclusion
Based on this systematic review, existing evidence is yet insufficient to support a positive or superior effect on bone health to reduce fracture risk in people with T2D.Collectively, incretin treatment is a potential therapeutic strategy for fracture prevention in people with T2D, but further evaluation is needed.Acknowledgement of T2D as an independent risk factor for osteoporosis-related fractures, and the potential beneficial effects of incretins could encourage future T2D guidelines to include fracture risk assessment in the treatment algorithm for the choice of glucose-lowering drug.However, further, larger, and more thorough clinical trials are necessary to investigate changes in bone indices and fracture risk after incretin treatment to confirm any potential advantageous effects on bone health in people with T2D.

Perspectives
Ideally, large-scale RCTs are needed, focusing on fractures as primary outcomes or using BMD as a proxy measure.Many existing RCTs did not prioritize fractures as the main outcome and had limitations in terms of duration, which may not capture changes in fracture risk adequately.Additionally, considering the diverse patterns of fracture risk in T2D and its different phenotypes, it becomes necessary to study subtypes of fractures rather than overall fracture risk.The interaction between T2D type and fracture risk also needs exploration, as factors such as body weight, insulin, insulin resistance, and metabolic control could potentially modify fracture risk.Moreover, reports on BMD from existing RCTs are scarce.However, considering that the excess fracture risk in T2D is limited and declining, conducting largescale RCTs with fractures as primary outcomes might not be feasible.Instead, studies should preferably focus on high-risk groups, including people with prior fractures, low BMD, and severe insulin resistance, for example, as well as those with poor metabolic control.
Furthermore, there is a need for research investigating the impact of incretin-mediated effects on bone health and loss in obese and overweight people.This area of study holds particular interest because bariatric surgery has been associated with an increased risk of fractures. (108,109)

2023
Investigate role of dental-derived stem cells (hDPSCs) and their association with GLP-1 R In vitro GLP-1 RA expression was found to be upregulated during osteoblast differentiation; GLP-1 RA liraglutide peptide treatment increased osteoblast differentiation in hDPSCs by increasing calcium deposition, ALP activity, and osteoblast marker genes, Runx2, type 1 col, osteonectin, and osteocalcin DPP-4is and cell studies Xiang et al 2020 Examine role of angiogenesis with sitagliptin treatment in diabetes-induced poor osteointegration of titanium implants and underlying mechanisms In vitro, human umbilical vein endothelial cells (HUVECs) incubated on titanium (Ti) surface were subjected to (1) normal milieu (NM); (2) diabetic milieu (DM); (3) DM + sitagliptin; (4) NM + macrophage; (5) DM + macrophage; or (6) DM + macrophage + sitagliptin In vitro, when cells were incubated alone, DM caused M1 polarization of macrophage, evidenced by increased iNOS and decreased CD206 expressions, and obvious dysfunctions of HUVECs; DM-induced injury of endothelial cells was significantly worsened when the two cells were co-cultured; addition of sitagliptin markedly reversed changes of macrophage but not of HUVECs in DM when cells were cultured alone; when cells were co-cultured, however, both abnormal macrophage polarization and endothelial impairment in DM was significantly alleviated by sitagliptin Note: The MAPK/ERK pathway (also known as the Ras-Raf-MEK-ERK pathway) is a chain of proteins in the cell that communicates a signal from a receptor on the surface of the cell to the DNA in the nucleus of the cell.cAMP = cyclic adenosinmonophosfat; iNOS = Nitric oxide synthase.