What are the actions of antidiabetic drugs?

5.1 Mechanism of action

The sulfonylureas remain one of the most widely utilized anti-diabetic medication classes since 1942, when the hypoglycemic effects of sulfonylureas were first discovered (Sola et al., 2015). They bind to the sulfonylurea receptor (SUR1) which is most commonly expressed on pancreatic beta islet cells, though they are also found in the brain. Some of the sulfonylureas have an affinity for the SUR2A receptor, an isoform of the SUR1 which is more present in heart and skeletal muscle tissue (Sola et al., 2015).

Their binding leads to closure of ATP-sensitive potassium channels (Fig. 3). This causes the cell membrane to depolarize and open voltage-gated calcium channels. The subsequent influx of calcium ions leads to insulin exocytosis from within the beta-cells (Hivelin et al., 2016).

5.3.2 Topiramate

Topiramate is a sulfamate-substituted derivative of d-fructose (White, 2005) whose antiepileptic activity was discovered by serendipity while researching for new anti-diabetic medication (Phabphal and Udomratn, 2010). It is rapidly absorbed, and has a good bioavailability, however, reaching peak plasma concentration could be slightly slowered when administered with food. It does not bound to proteins (Bialer, 1993; Khalil et al., 2019). The half-life of topiramate ranges from 19 to 23 h and independently of the dose. Its pharmacokinetics are linear and serum concentration reaches its peak within 1.8–4.3 h. It is mainly excreted in the urine (Bialer, 1993), and partially by oxidation and hydrolysis (Khalil et al., 2019). In humans, six metabolites have been identified (Khalil et al., 2019).

Several mechanisms of action have been described. Topiramate has been demonstrated to modify several receptor-gated and voltage-sensitive ion channels, including negative modulation of the voltage-activated Na+ and L-type Ca2+ channels. Topiramate negatively modulates the excitatory neurotransmitter glutamate through the α-amino-3-hydroxy-5-methyl-4 isoxazole propionic acid (AMPA) receptors and reducing kainate currents (Andreou and Goadsby, 2011). Unlike valproate, topiramate does not modulate NMDA (Cutrer, 2001; White, 2005; Andreou and Goadsby, 2011). It also reduces carbonic anhydrase (Dodgson et al., 2000) and enhances GABA evoked chloride single-channel currents (White, 2005). Its mechanism of action is consistent with an effect on trigeminovascular activation, by ways of inhibiting the trigeminocervical complex (Storer and Goadsby, 2004) attenuating neurogenic dural vasodilation and nitric oxide-induced vasodilation (Akerman and Goadsby, 2005b), possibly by inhibiting CGRP secretion from trigeminal neurons (Durham et al., 2006) and it could also inhibit cortical-spreading depression in a dose-dependent manner (Akerman and Goadsby, 2005a). Based on the last effect, topiramate could be helpful in the treatment of migraine aura (Fernandez-Diaz et al., 2008), although this has not been proven clinically in an underpowered open trial (Lampl et al., 2004).

Its efficacy at a dose of 50–200 mg per day has been proven in numerous randomized, placebo-controlled studies in patients with episodic (Brandes et al., 2004; Storey et al., 2001; Silberstein et al., 2004b; Mei et al., 2004; Gupta et al., 2007) and chronic migraine (Storey et al., 2001; Mei et al., 2006; Silberstein et al., 2007; Diener et al., 2007) and also in those with medication overuse (Silvestrini et al., 2003; Mei et al., 2006; Diener et al., 2007) and prominent associated vertiginous symptoms (Carmona and Settecase, 2005; Beh, 2018). It may be less efficacious in patients with daily headache (Rothrock et al., 2005). Its efficacy as a migraine preventive resulted similar (Diener et al., 2004b) or slightly better to that of propranolol (Ashtari et al., 2008) and efficacy and tolerability were equivalent to those of valproate (Shaygannejad et al., 2006), amitriptyline (Dodick et al., 2009) and onabotulinumtoxinA (Mathew and Jaffri, 2009; Cady et al., 2011). Topiramate has also been approved as an anti-obesity drug.

Adverse events include paresthesia (Krymchantowski and Tavares, 2004; Silberstein et al., 2004b; Diener et al., 2007), close angle glaucoma (Di Legge et al., 2002), weight loss (Krymchantowski and Tavares, 2004; Silberstein et al., 2004b), mood disturbances (Krymchantowski and Tavares, 2004; Phabphal and Udomratn, 2010) and cognitive detriment (Privitera et al., 1996; Krymchantowski and Tavares, 2004; Silberstein et al., 2004b; Martin et al., 1999; Gordon and Logan, 2006). Treatment with topiramate causes increased urinary pH, which in addition to lower urinary citrate excretion and systemic metabolic acidosis increase the possibility to form kidney calcium phosphate stones (Welch et al., 2006). Infrequent undesirable manifestations such as myoclonus, fasciculations (Alonso-Navarro and Jimenez-Jimenez, 2006), palinopsia (Sierra-Hidalgo and de Pablo-Fernandez, 2013; Evans, 2006; Yun et al., 2015) and Alice in Wonderland syndrome (Evans, 2006; Jurgens et al., 2011) and anorgasmia (Sun et al., 2006) have also been reported as reversible side effects that had complete resolution after discontinuation of the drug. Topiramate is FDA pregnancy category D and has been associated to orofacial cleft in exposed offspring (Hunt et al., 2008; Alsaad et al., 2015).

Topiramate could interact with other antiepileptic drugs, especially those enzyme inducers such as phenytoin or carbamazepine could increase its clearance (Khalil et al., 2019), and decrease levels of digoxin (Bourgeois, 1996). Dose adjustments may be required in those patients with chronic renal or hepatic impairment (Khalil et al., 2019) and high doses of topiramate could affect oral contraception (Johnston and Crawford, 2014), for what contraception counselling is advisable in patients taking combined oral contraceptive medication (Viana et al., 2014).

Treatment

An important pillar of T2D prevention and treatment is lifestyle change, with increased exercise, decreased caloric intake, nutritional modifications and weight loss. The Diabetes Prevention Program (DPP) showed that progression to T2D for patients with IGT was reduced 58% with intensive lifestyle modification during the 2.8 year follow-up (Knowler et al., 2002). Weight loss of 5% can improve glycemia and reduce the need for anti-diabetic medications in patients with T2D. A 500–700 kcal deficit per day is recommended to achieve weight loss (American Diabetes Association, 2018b). Patients with T2D should be counseled on reduction in refined carbohydrates and limitation of dietary fat to 20%–30% of calories with an emphasis on monounsaturated fats. Exercise recommendations from the ADA are 150 min of moderate to vigorous intensity exercise spread over the week, similar to the DPP design. Motivation for continued lifestyle changes can wane and effective programs, such as with the DPP, incorporate “coaches” and continued counseling and encouragement to achieve success.

There is a plethora of non-insulin therapies approved for the treatment of T2D with diverse mechanisms of action (Table 3). Often, a combination of medications is required to attain glycemic control initially or with time, as a patient progresses toward complete beta cell failure. In considering a combination, the clinician should choose therapies that will target different defects present in T2D for an additive or synergistic effect on glycemic control. Patient specific factors also need to be weighed—age, weight, hypoglycemia risk, cardiovascular disease, etc.—to derive the maximum benefit and minimize the risks of each medication or combination.

Table 3. Available pharmacologic therapies for type 2 diabetes

When glycemic control cannot be attained with non-insulin therapies, insulin can be initiated (Table 4), as a basal injection of a long-acting or intermediate-acting insulin. Further addition of prandial (bolus) insulin for one to three meals per day may be necessary. The commencement of insulin therapy does not necessitate the discontinuation of all other T2D therapies as there can be continued benefit. For example, the combination of metformin or a GLP-1 receptor agonist with insulin can be insulin sparing.

Table 4. Available insulins and insulin analogues with estimated pharmacokinetics

1.2.2 Type 2 diabetes

Type 2 diabetes (T2D) results from insulin resistance and inadequate insulin secretion to compensate for the reduced insulin sensitivity, and it accounts for approximately 90% of all cases of diabetes. During the 1920s, an antihyperglycaemic extract from the flowering plant French lilac was isolated and used to generate the biguanide synthalin (Frank and Nothmann, 1926), but its hepatotoxicity meant that it was not a clinically viable therapy (White, 2014). However, another biguanide, metformin, was introduced in the 1950s and it is now the most widely prescribed anti-diabetic medication, with over 100 million users worldwide. Metformin's primary mechanism of action is to reduce hepatic glucose production and it also improves insulin sensitivity through increasing glucose uptake into muscle (Rena et al., 2017). Another important group of therapies for diabetes are sulphonylureas, which were developed after the discovery in 1942 that anti-bacterial sulphonamides reduced blood glucose levels (Quianzon and Cheikh, 2012b). Sulphonylureas have been used clinically since the early 1950s, initially with first-generation drugs such as tolbutamide and chlorpropamide, then with analogs with improved safety profiles such as glipizide, glyburide and glimepiride (White, 2014). However, the mode of action of these drugs, which is to stimulate insulin secretion by directly closing β-cell ATP-sensitive K+ (KATP) channels, means that they are associated with unwanted hypoglycaemia since they can promote insulin release even when blood glucose levels are not raised (Rendell, 2004). This potentially life-threating risk led the American Diabetes Association to provide guidelines in 2019 that sulphonylureas should be used as a last-line therapy for T2D, only if all other classes of anti-diabetic medications fail or if glycated hemoglobin levels are above the target goal, unless cost is a major issue.

While metformin and sulphonylureas were the only available therapies for T2D for nearly 40 years, since the 1990s a wide family of anti-diabetic agents has been introduced. Thus, α-glucosidase inhibitors such as acarbose and miglitol delay post-prandial increases in blood glucose levels by inhibiting α-glucosidase-mediated digestion of oligosaccharides in intestinal enterocytes (Derosa and Maffioli, 2012) and meglitinides such as repaglinide have sulphonylurea-like effects to close β-cell KATP channels, and their shorter half-lives mean that they are not commonly associated with the hypoglycaemia that is seen with sulphonylurea use (Guardado-Mendoza et al., 2013). Thiazolidinediones (TZDs) were also first used clinically in the late 1990s and they improve glucose homeostasis by activating nuclear PPARγ receptors, which leads to increased expression of genes responsible for transducing the anabolic effects of insulin (Soccio et al., 2014). The first clinically used TZD, troglitazone, was withdrawn in 2000 due to its hepatotoxicity and rosiglitazone was also withdrawn in Europe in 2010 because of an increased risk of cardiovascular complications. Pioglitazone is the only TZD currently marketed in the UK, but it was withdrawn in France and Germany in 2011 because of an increased risk of bladder cancer. It has been known for nearly 60 years that gut-derived incretin peptides are responsible for enhanced insulin secretion when glucose is delivered orally (McIntyre et al., 1964) and, more recently, levels of one of these incretins, glucagon-like peptide-1 (GLP-1) were found to be reduced in people with T2D (Toft-Nielsen et al., 2001). GLP-1 replacement in T2D is not an option since it has a half-life of < 5 min, but GLP-1 analogs, with considerably longer half-lives than native GLP-1, act at β-cell GLP-1 receptors to potentiate glucose-stimulated insulin secretion (Holst, 2019). GLP-1 analogs such as exenatide and liraglutide must be delivered by subcutaneous injection, a potential disadvantage compared to the other T2D therapies, which are all taken orally. However, an oral formulation of the GLP-1 analog semaglutide (Rybelsus®) has proven to be very effective in clinical trials (Aroda et al., 2019) and it was approved for use in the United States in 2019 (Yamada et al., 2020; Rodbard et al., 2019). DPP-IV inhibitors such as sitagliptin, were first used clinically in 2006 to prolong the circulating half-life of endogenous incretins and thus allow enhanced insulin secretion (Gupta and Kalra, 2011). Sodium-glucose co-transporter 2 (SGLT2) inhibitors, introduced in the last decade, are the most recent class of compounds used to treat T2D. These agents, known as “flozins,” inhibit the low-affinity SGLT2 glucose transporters in the kidney proximal tubules so that excess glucose is excreted in the urine (Rizos and Elisaf, 2014).

Conundrums

The relationship between MS, obesity, and sleep apnea is as complex as it is intriguing. Attempts to establish unequivocal links between sleep apnea and MS should take into account the common risk factors for these entities (obesity and age being the chief of these). As Dempsey puts it, “solving the algebraic equation Z–X and extracting the Z component from syndrome X is proving extremely difficult” (Dempsey, Veasey, Morgan, & O’Donnell, 2010).

Obesity, of course, predisposes to OSA; on the other hand, OSA (through the fatigue, sleepiness, and physical inactivity it engenders) promotes obesity. It has been argued that common pleiotropic effects such as hypertension, central obesity, dyslipidemia—as well as sleep apnea—are determined by common genetic factors; syndrome Z may just be an extended form of MS.

Excess fat deposits in the parapharyngeal area doubtless play a part in the genesis of OSA by compromising the airway (Davies & Stradling, 1990). On the other hand, many persons with OSA do not manifest any of the structural abnormalities of the upper airway that are known to predispose to airway obstruction during sleep (Carmelli, Swan, & Bliwise, 2000). Indeed, there seems to be a stronger relationship between OSA and central obesity than between OSA and parapharyngeal adiposity (Larsson et al., 1984), which suggests mere mechanical loading may not be the only contributory factor involved.

The relationship between OSA and diabetes is unresolved. The Wisconsin study found that OSA subjects ran twice the risk of developing diabetes, and this risk was independent of known confounders, but somewhat confusingly, the increase in the incident risk for diabetes was not perceived to be high after 4 years of follow-up (Reichmuth, Austin, Skatrud, & Young, 2005). Not all authors were able to establish an unequivocal relationship between OSA and glycemic control (Lam et al. 2010). Several studies have demonstrated improvements in glycemic control in the short (Harsch et al. 2004) and long- term (Schahin et al. 2008) with CPAP therapy. On the other hand, other investigators found no evidence that untreated OSA adversely impacts glycemic control (Coughlin, Mawdsley, Mugarza, Wilding, & Calverley, 2007; West, Nicoll, Wallace, Matthews, & Stradling, 2007). As in the case of hypertensive patients, CPAP therapy appears to benefit most those patients who report excessive daytime somnolence (Barceló et al., 2008), raising more questions. It is also possible that at least some of the improvements in glycemic control could have been achieved by lifestyle modification and anti-diabetic medication, and not by CPAP therapy alone. By the same token, it is likely (Dorkova, Petrasova, Molcanyiova, Popovnakova, & Tkacova, 2008; Lam et al., 2010), but not undisputable (Comondore et al. 2009; Robinson, Pepperell, Segal, Davies, & Stradling, 2004), that CPAP has a beneficial effect on lipid metabolism.

As discussed, obesity itself is strongly linked to cardiovascular disease. Unfortunately, most of the earlier studies that addressed the risks of obesity in cardiovascular disease failed to examine the role of OSA (Romero-Corral et al., 2006). Case series and case-control studies that comprise the bulk of the literature on this subject cannot tease out the independent risks of OSA from those of obesity. Most studies differ considerably with regard to the degree of obesity and gender distribution: most authors have typically studied moderate to severe OSA in male subjects.

The vast majority of studies utilize AHI to diagnose and quantify the severity of sleep apnea—unfortunately, to the exclusion of other important variables. For instance, the Respiratory Disturbance Index measures perturbances other than overt apneas and hypopneas—such as periodic limb movements and snore arousals—the sleep disturbances that occur as a consequence of which can impact on metabolic disturbances (Novoa et al., 2011). In addition, most studies have traditionally relied on surrogate indices such as measurements of blood pressure, cardiac function, and serum catecholamine levels rather than more relevant indicators of outcome such as quality of life, functional class, frequency and duration of hospitalization, and mortality. Only a handful of relatively small trials have examined the efficacy of interventions considered the gold standard (such as CPAP) against these outcomes. The possibility of a placebo effect with CPAP confounds issues further. For example, with respect to hypertension, despite data from several randomized, placebo-controlled trials, it is still unclear how much of the improvement in blood pressure can be ascribable to CPAP. The fact that sub-therapeutic CPAP does not improve blood pressure despite cutting the AHI by half (Becker et al., 2003) argues in favor of pathogenetic pathways other than intermittent hypoxia—such as repetitive arousals—that CPAP might also counteract.

Most patients with untreated OSA are at a survival disadvantage. OSA morbidity is especially high in those with excessive daytime somnolence, in males (Punjabi & Polotsky, 2005), and in younger individuals (Lavie, Lavie, & Herer, 2005). It is likely that these individuals benefit the most from early diagnosis and treatment. Why older individuals are relatively protected from the adverse cardiovascular consequences of sleep apnea is unclear, but it is feasible that a certain amount of ischemic preconditioning occurs with advancing age (Lavie & Lavie, 2006). Although obese individuals are at greater risk for cardiovascular disease and congestive heart failure (Kenchaiah et al., 2002), obesity may paradoxically confer a survival advantage in chronic, stable heart failure (Curtis et al., 2005). This advantage appears to taper off when BMI rises to 35 kg/m2 (Parameswaran, Todd, & Soth, 2006).

The propensity to develop sleep-disordered breathing depends on the complex interaction between multiple compensatory processes that vary substantially between person to person, and also within a given individual. Several phenotypes clearly exist within OSA patients. Genome-wide linkage studies of OSA phenotypes (Palmer et al., 2003) suggest that fully half of the genetic determinants of AHI are independent of obesity (Patel, 2005). Notably, there is a much higher prevalence of OSA in males, and in certain ethnic groups (Africans, East Asians) as compared to Western Caucasians (Ip et al., 2004; Young et al., 1993). Indeed, the relationship between obesity and OSA appears to be influenced by ethnicity. In Asian Indians, sleep apnea of comparable severity has been associated with a relatively low BMI (Udwadia, Doshi, Lonkar, & Singh, 2004). Asian Indians have a relatively high proportion of body fat and a commensurately low muscle mass, which also predisposes them to MS (Misra et al., 2001; Ramachandran, Snehalatha, Satyavani, Sivasankari, & Vijay, 2003).

The critical importance of defining those phenotypes with the highest risk of morbidity underscores the importance of any current research that focuses on the elaboration of candidate genes linking the genetic mechanisms of obesity with sleep apnea.

Could anti-diabetic medications be important?

Whether anti-diabetic medications impact on OA outcomes has been investigated in an analysis of longitudinal data from the Osteoarthritis Initiative study, finding that medication-treated diabetes has no effect on knee OA incidence (OR = 0.53; 95% CI 0.23, 1.5), but reduces knee OA progression, measured as JSN or knee replacement therapy (OR = 0.66; 95% CI 0.44–0.98) [30].

Metformin is the first recommended antidiabetic drug for the management of T2DM. In a UK cohort study set within the Consultations in Primary Care Archive, of 3217 patients with T2DM, there was no association between prescribed metformin treatment at baseline and OA outcome during follow up (adjusted HR = 1.02; 95% CI: 0.91, 1.15) [31]. However, in a case-control study performed in Taiwan, patients who have OA and T2DM receiving combination NSAIDs and metformin therapy had lower joint replacement surgery rates than those without metformin (adjusted HR = 0.742; 95% CI 0.601, 0.915; p = 0.005) [32]. It has been suggested that this effect may be attributable to a reduction in pro-inflammatory factors associated with combined therapy with metformin. Indeed, metformin and even more thiazolidinediones are antidiabetic agents that have been shown to exert some anti-inflammatory activity [33]. However, a large population based case-control study was performed in UK using the Clinical Practice Research Datalink; it did not find any evidence for a disease modifying osteoarthritic effect of thiazolidinediones [34], despite promising results from animal in vivo studies [34]. Finally, there are no clinical data available yet with new antidiabetic agents such as incretin-based therapies - dipeptidyl peptidase-4 (DPP-4) inhibitors and glucagon-like peptide-1 (GLP-1) receptor agonists or sodium-glucose cotransporter type 2 (SGLT2) inhibitors.

5 Effects of anti-diabetic agents on bone

The effect of anti-diabetic medications on skeletal health has lately received great interest. More data concern the use of thiazolidinediones (TZDs), an insulin sensitizing class of drugs that are peroxisome proliferator-activated receptor gamma (PPARγ) agonists. Activation of PPARγ results in stimulation of adipocyte differentiation and in inhibition of osteoblastogenesis [62]. Furthermore, TZDs induce osteoblast and osteocyte apoptosis [63,64] and upregulate the expression of sclerostin [63], a potent inhibitor of osteoblastogenesis. TZDs use has been associated with BMD reductions at both the lumbar spine and the hip, mainly in postmenopausal women [65], although there are reports for a similar effect in premenopausal women [66] and men [67]. There is increasing evidence that the use of TZDs increases the risk for any fracture in postmenopausal women, while no definite effect could be demonstrated in men [68,69]. It is postulated that longer medication use [69], as well as higher doses [68] might both increase the risk for fracture. It is estimated that treatment of 86 patients with a thiazolidinedione for 3 years would produce 1 additional peripheral fracture [69]. However, the overall increased fracture risk in T2DM could not be attributed to the use of TZDs alone, since the percentage of patients on TZDs in the large epidemiological studies demonstrating increased fracture risk was rather small and an increase in fracture risk was demonstrated in cohorts even before the introduction of these drugs.

Apart from TZDs, the rest of anti-diabetic drugs probably have either a neutral or a positive effect on bone metabolism, as suggested by studies so far, even though it is perhaps too early to draw any definite conclusions about the newest drugs such as the glucagon-like peptide-1 (GLP-1) analogues, dipeptidyl peptidase-4 (DPP-4) inhibitors and sodium-glucose cotransporter-2 (SGLT-2) inhibitors.

Sulphonylureas may have a beneficial impact on the risk for any fracture [7], as well as the risk for vertebral fractures [70], although a neutral role has also been suggested [6]. However, the pathogenetic mechanisms are quite difficult to explain and their effects are probably explained by their efficiency in improving glycemic control. Nevertheless, a possible anabolic effect through the increase in endogenous insulin release induced by sulphonylureas and a direct anabolic effect by stimulating proliferation and differentiation of osteoblasts, as demonstrated by in vitro studies [71], cannot be excluded.

Metformin use has also been associated with a decrease in fracture risk [6,7], and it is likely that it exerts direct effects on bone metabolism, apart from those mediated through improved glycemic control. Metformin has been shown to produce direct osteogenic effects in vitro [72], to reduce receptor activator of nuclear factor κB ligand (RANKL) expression and thus inhibit osteoclast differentiation [73] and to protect from ovariectomy induced bone loss in vivo [73]. Moreover, in animal models, metformin has demonstrated an ability to protect from the deleterious effects of rosiglitazone on bone when administered in combination [74]. However, the latter finding was not confirmed in humans; the combination of metformin and rosiglitazone for 80 weeks in diabetic men and women resulted in significant decreases of 2.2% in lumbar spine and of 1.5% in total hip BMD, as compared with metformin monotherapy [75]. Of note, there was no difference in BMD reduction between men, premenopausal and postmenopausal women. Finally, there is evidence that metformin might protect from the adverse effects of AGEs on osteoblastic cells [76].

GLP-1 receptor knockout mice are shown to exhibit enhanced bone resorption accompanied by elevated osteoclasts numbers and cortical osteopenia, leading to increased bone fragility [77]. It is postulated that the effects of GLP-1 on bone homeostasis are mediated through its action in stimulating calcitonin secretion from the parafollicular cells of the thyroid. In animal studies, administration of GLP-1 or exendin-4 exerted osteogenic effects [77], however, administration of exenatide in diabetic patients for 44 weeks did not significantly affect BMD or bone turnover markers, comparing to the administration of insulin glargine [78]. Of note, patients treated with exenatide displayed a significant reduction in body weight which, however, had no negative effect on bone metabolism. Mabilleau et al, in a recent meta-analysis, demonstrated that the use of GLP-1 analogues does not have any impact on fracture risk, nevertheless, most studies were of short duration and not adequately powered to establish an association between GLP-1 analogues use and fractures [79].

DPP-4 inhibitors inhibit the degradation of endogenous incretins, namely GLP-1 and gastric inhibitory polypeptide (GIP), as well as other gastrointestinal peptides such as glucagon-like peptide-2 (GLP-2), increasing their plasma levels. GIP enhances osteoblasts’ function [80]. GIP receptor knockout mice (GIPR −/−) exhibit decreased bone size, mainly due to a reduction in bone formation, while GIP administration prevents the bone loss associated with ovariectomy [81]. Data about the effect of GLP-2 at the tissue level are limited. However, its administration in postmenopausal women resulted in suppression of bone resorption without affecting bone formation [82]. Prolonged administration, for four months, was associated with a significant increase in total hip BMD [83]. DPP-4 knockout male and female mice display normal skeletal phenotypes, however ovariectomized DPP-4 knockout models exhibited significantly reduced femoral size. Sitagliptin treatment resulted in significant improvements in trabecular architecture in female, but not in male, or ovariectomized mice [84]. Data regarding the effects of DPP-4 inhibitors in human skeletal health are quite scarce. In a retrospective population based cohort study, the use of DPP-4 inhibitors was not associated with increased risk for fractures, as compared with the use of other anti-diabetic medications, nevertheless, the duration of use was short [85]. A meta-analysis of randomized controlled trials with DPP-4 inhibitors suggested that the use of this class of medication could be associated with a reduction in fracture risk [86]. However, the duration of the studies included was quite short and fractures were not the primary end point, so larger studies concerning the effects of DPP-4 inhibitors on skeletal physiology are necessary in order to draw any definite conclusions.

Dapagliflozin, a selective SGLT2 inhibitor, is a relatively new medication for diabetes and data concerning its effects on bone metabolism are therefore limited. In a 50 week trial, the addition of dapagliflozin on metformin produced neither changes on markers of bone turnover, namely amino-terminal propeptide of type 1 collagen (P1NP) and carboxy-terminal collagen crosslinks (CTX), nor any alterations in BMD [87]. More data on SGLT2 inhibitors on skeletal homeostasis are warranted.

Pramlintide is an amylin analogue, with an indication for T2DM in patients who use mealtime insulin. Amylin enhances proliferation of osteoblasts, inhibits osteoclastic activity and osteoclastogenesis, while systemic administration in adult mice increases skeletal mass [81]. However, in a study regarding patients with T1DM, pramlintide administration for 12 months had no effect on BMD and bone turnover markers [88]. There are currently no available data concerning its effects on bone metabolism in T2DM patients.

2.8.4 Do anti-diabetic agents influence the risk of pancreatic cancer?

Recent studies suggest that anti-diabetic medications may modulate the risk of pancreatic cancer in type 2 diabetes. Numerous studies report the protective effect of metformin on overall cancer mortality including a recent meta-analysis by Ben et al. Li et al.25 in a case–control study of 973 patients with pancreatic cancer (259 of whom had diabetes) and 863 controls (109 diabetic patients) observed that diabetic patients who had taken metformin had a significantly lower risk of pancreatic cancer compared with those who had not (odds ratio—0.38). Diabetic patients who had taken insulin or insulin secretagogues were found to have a significantly higher risk of pancreatic cancer. However, a subsequent study by Bodmer et al.26 did not find any association between metformin use and pancreatic cancer. At this time, a definitive conclusion on a potential protective role for metformin cannot be drawn Table 4.

Table 4. Anti-diabetic agents and pancreatic cancer risk.

4.6 Other drug targets

Other targets for developing anti-diabetic medications may also be of benefit in the prevention or treatment of AD. One of these is amylin (otherwise known as islet amyloid polypeptide- IAPP), a peptide hormone that is secreted from pancreatic β-cells. Amylin lowers blood glucose levels through delayed gastric emptying, decreased glycagon secretion and increased satiety. However, amylin does have the propensity to aggregate and form amylin oligomers and fibrils, particularly in a chronic diabetic state, resulting in reduced insulin production, hyperglycaemia and β-cell dysfunction or death. Therefore, a soluble, non-aggregating synthetic analogue of amylin, pramlintide, has been developed and approved as adjunctive therapy to insulin for the treatment of type 1 and type 2 diabetes (Aathira and Jain, 2014; Grunberger, 2013).

Interestingly amylin can readily cross the BBB and its receptors are distributed throughout the brain (Christopoulos et al., 1995; Sexton et al., 1994; Gotz et al., 2013), Jackson et al., 2013). Increasing evidence suggests that amylin may play a role in a variety of CNS functions including mood, memory, anxiety and satiety (Kovacs and Telegdy, 1996; Morley et al., 1995; Srodulski et al., 2014; Zhu et al., 2015). The amylin receptors have roles in Aβ-mediated neurotoxicity (Kimura et al., 2012) and amylin deposition has been reported in grey matter, cerebral blood vessels and in the peri-vascular space of diabetic and AD brain as well as in association with β-amyloid plaques (Jackson et al., 2013). Very recent studies have shown that administration of pramlintide to AD transgenic mice reduces cerebral amyloid burden and improves learning and memory in the mice (Zhu et al., 2015). It also increases synaptogenesis and reduces oxidative stress and inflammation in the senescence accelerated mouse model (SAMP-8), which shows AD-like pathology (Adler et al., 2014).

Drug development for diabetes is ongoing, with novel agents being tested in pre-clinical studies as well as clinical trials; these may also be of value in the prevention or treatment of neurodegenerative diseases. These include inhibitors of tyrosine-protein phosphatase (PTP1B, a negative regulator of insulin signalling), which have shown to reduce inflammation and restore insulin signalling in the brain (Picardi et al., 2008). Inhibitors of the oxidoreductase 11-β hydroxysteroid dehydrogenase type 1 (11β-HSD1), which converts inactive keto-forms of cortisone to active glucocorticoids (cortisol and corticosterone), are in early stage of clinical trials for T2D. Elevated levels of glucocorticoids are characteristic of the metabolic syndrome and are pathogenic in T2D, and are associated with neuronal damage, memory impairments and hippocampal atrophy (Yau and Seckl, 2001; Holmes et al., 2010). Therefore, reducing glucocorticoid levels may provide significant benefits in the prevention of neurodegeneration.

9 Conclusion

Metformin, although a time tested anti-diabetic medication, has probably much more to offer. With time scientists are deciphering the mysterious actions of this drug and the potential benefits are coming up day by day, to date it is an unexplored territory for the entire medical community. It is therefore necessary to put efforts in order to research the advantages of metformin. Although much studies have been done on rodents, but still they provide us with the scientific basis on which further human trials can be conducted. Its role in the treatment of PCOS, obesity and NAFLD has almost been established based on scientific evidence Keeping in view the potential therapeutic advantages of metformin, this drug should be given chance in the arsenal of therapeutics apart from its traditional usage in diabetic dilemma. The current limelight should focus as its role being a potential anti-cancer, anti-inflammatory and rejuvenation compound so that the focus of future research on metformin should have a direction.