Advertisement

Calcified Tissue International

, Volume 100, Issue 2, pp 109–121 | Cite as

Epidemiology of Fractures in Diabetes

  • Jakob Starup-Linde
  • Morten Frost
  • Peter Vestergaard
  • Bo AbrahamsenEmail author
Review

Abstract

Diabetes mellitus is associated with an increased risk of fracture. The risk of a hip fracture is up to sevenfold increased in patients with type 1 diabetes and about 1.3-fold increased in patients with type 2 diabetes. However, these relative risk estimates may depend on the age and gender distribution of the population in question. Bone mineral density and the fracture risk assessment tool do not explain the increased fracture risk in patients with diabetes. Shared risk factors as pancreatitis, alcohol use, smoking and oral glucocorticoids may influence the observed fracture risk in patients with diabetes. This review examines the association between diabetes and fracture and attempts to disentangle the tight connection between diabetes per se, diabetes-related complications, comorbidities and shared risk factors. This is of great importance as the number of diabetes patients’ increases with growing and aging populations and putting even more at risk of fracture.

Keywords

Diabetes mellitus Fracture Risk factors Epidemiology 

Introduction

Though the prevalence of diabetes varies between geographical areas, the global burden of diabetes has been increasing substantially over the past decades, chiefly triggered by increasing obesity and adverse lifestyle risk factors. Type 1 diabetes accounts for 5–10 % of all cases of diabetes [1]. While the incidence rates of childhood type 1 diabetes (T1D) are increasing [2, 3], the incidence rates of adult autoimmune diabetes have not been evaluated extensively. In a European study, autoimmune diabetes was observed in 9.7 % of adult-onset cases of primary diabetes, and Latent Autoimmune Diabetes of Adults (LADA) was three times more common than classical T1D [4]. Whether the incidence of autoimmune diabetes in adults has changed is unknown [5]. Compared to other forms of diabetes, monogenic types of diabetes are rare, accounting for only about 1 % of children with diabetes [6].

Globally, 9 out of 10 cases of all patients with diabetes have type 2 diabetes (T2D) [1]. Reports generally show increasing incidence rates of T2D [7, 8], which emphasizes the importance of introducing measures that effectively prevent not only progression of T2D but also the complications caused by inadequately managed T2D.

The apparent incidence of any of these types of diabetes is sensitive to misclassification. In underdeveloped countries, patients with T1D may die before diagnosis or incident cases of ketoacidosis caused by T1D may wrongly be interpreted as infection [9]. Distinction between T1D and LADA relies on both medical history and measurement of antibodies [10], a test that is not always available to health professionals, and monogenic forms of diabetes are often misclassified as type 1 or 2 diabetes [11]. In addition, many patients with T2D remain undiagnosed even in areas of the world with highly developed universal healthcare systems. It has proven challenging to distinguish reliably between T1D and T2D in most national health databases due to inaccuracies in coding practices. Though a young age of onset and insulin as the first therapy used may suggest T1D, ambiguous coding can be difficult to disentangle due to partly overlapping treatment modalities. Dedicated national diabetes databases can have issues with accurately separating patients with diabetes from patients with General Practice billing for repeated glucose measurements even in highly register-driven countries [12]. Methodological issues aside, the latest estimates for T1D and T2D combined from the International Diabetes Foundation cite a figure of 328 million adults (age 20–89) with diabetes in 2013 with an expected increase to 592 million in 2035, of which 143 million will be in China alone.

The incidence rates of T1D vary between regions with the highest levels being observed in the Nordic countries, Saudi Arabia, Canada and the UK [9]. Unlike osteoporosis that has the highest prevalence in the Nordic countries and the USA [13], the highest proportion of adults with T2D is found in Latin America, the Middle East, India and China. The gender distribution is different in diabetes and osteoporosis. Hence, T1D is an exception among autoimmune diseases because it does not show a strong female bias. The male/female ratio is close to one before puberty [14] and 1.5 in adults [5]. For T2D, an increasing male preponderance seems to have followed increasing urbanization in the twentieth century and men appear more sensitive to abdominal adiposity as Swedish men are being diagnosed with T2D earlier than women and at a lower body mass index (BMI) [15].

The global fracture burden reflects both true traumatic fractures in persons without skeletal fragility and fractures related to compromised bone structure and reduced bone strength as observed in osteoporosis. In practice, this constitutes a continuum as patients with impaired bone strength suffer high-energy trauma fractures at a higher rate than skeletally healthy subjects. In epidemiological practice, fractures occurring at major osteoporotic sites including forearm, humerus, spine and hip in subjects aged 50 or over often form the basis for fracture burden calculations. Current global estimates attribute some 9 million fractures annually to osteoporosis, with 22 million women and 5.6 million adult men in the European Union fulfilling the criteria for osteoporosis [16]. In addition to BMD, several conditions are known to influence the integrity of bone and increase the risk of fractures. In recent years, awareness of adverse effects of T1D and T2D on the skeletal integrity and fracture risk has advanced [17]. It is important to identify and modify risk factors for fracture, including non-skeletal factors such as balance and falls risk. As will be discussed below, while diabetes is a non-modifiable risk factor in the sense that diabetes irrespective of type is a chronic disease, much of the ensuing fracture risk could be due to essentially modifiable risk factors.

Type 1 Diabetes and Fractures

The incidence of fractures in adults with T1D has been investigated in a number of studies. The vast majority of these investigations show an increased risk of fractures including hip fractures. However, published data on fractures in T1D vary substantially with regard to the composition and characterization of the study populations including age ranges, gender and study size as well as fracture sites reported (Supplemental Table 1). This is a potential obstacle to a clear synthesis of the results.

Meta-analyses published in 2007 by Janghorbani et al. [18] and Vestergaard [19] both showed an increased risk of hip fracture in T1D of 6.3 (95 % CI 2.6–15.1) and 6.94 (95 % CI 3.25–14.78), respectively. More recent meta-analyses have suggested that the risk estimates may be lower, most likely due to more studies being available for the analyses. In 2014, Fan et al. [20] reported that the relative risk of a hip fracture in individuals with T1D was 5.76 (95 % CI 3.66–9.07), whereas a meta-analysis from Shah et al. [21] from 2015 showed that the relative risk of a hip fracture was 3.78 (95 % CI 2.05–6.98). Stratifying the latter analysis according to gender revealed minor differences in relative risks of hip fracture between women [5.19 (95 %CI 2.22–12.11)] and men [4.05 (95 %CI 0.99–16.47)] [21]. However, data from the meta-analysis suggested that the relative risks of any fracture differ between women [4.10 (95 % CI 1.79–9.38)] and men [1.79 (95 % CI 1.38–2.33)] with T1D. Further, Shah et al. [21] observed substantial effects of study design on the risk estimates. Thus, the risk of a fracture in cross-sectional studies was lower than in cohort studies {[2.13 (95 %CI 1.24–3.67)] and 4.45 (95 %CI 1.33–14.89), respectively}, possibly explained by varying sizes of study populations and number of fractures as confidence intervals differ remarkably.

In most cases, T1D is diagnosed in childhood [1]. Therefore, T1D has a huge theoretical potential for affecting bone mineral accrual, which would reduce peak bone mass and increase the risk of fractures even in early life. Nevertheless, the majority of studies on skeletal events in T1D have been conducted in individuals aged 40 years or over. Importantly, the incidence of fractures in children with T1D was recently examined in a large study from the UK. Using information from electronic medical records from general practitioners, Weber et al. [22] reported incident fractures among more than 30,000 men and women aged 0–89 years with T1D, comprising a total more than 2500 fractures. Interestingly, the investigators observed an increased risk of any fracture in individuals with T1D aged 0–19 years in both men [HR 1.14 (95 % CI 1.01–1.29)] and women [1.35 (95 % CI 1.12–1.62)]. Furthermore, a higher risk of a hip fracture among women but not men aged 0–29 years [HR 4.71 (95 % CI 1.45–15.28) and 2.01 (95 % CI 0.99–4.10), respectively] was reported. By contrast, based on data from Scottish national diabetes and hospital admission registries including more than 20,000 persons with T1D, Hothersall et al. [23] observed higher risk of a hip fracture in both women and men aged 20–39 years with T1D [incidence risk ratio 8.92 (95 % CI 2.22–35.81) and 6.35 (95 % CI 3.44–11.72), respectively]. Correspondingly, compared to the general population, higher risks of being admitted to hospital due to a hip fracture were observed among Swedish women and men aged <40 years with T1D [standardized hospitalization ratio (SHR) 7.6 (95 % CI 4.5–12.0) and 3.4 (95 % CI 2.1–5.3), respectively] [24]. Although limited in number, these investigations show that fracture risk is increased in individuals with T1D even in adolescents and young adults.

Current knowledge of fracture risk in adults aged more than 40 years is much more extensive. Thus, the risk of any fracture and hip fracture was higher in both women and men aged more than 40 years with T1D than in age- and sex-matched controls in the study by Weber et al. [22], who reported hazard ratios ranging between 1.64 (95 % CI 1.35–1.99) and 2.07 (95 % CI 1.78–2.41) for any fracture and 1.99 (95 % CI 1.43–2.78) and 5.06 (95 % CI 2.80–9.14) for hip fractures in women and men aged more than 40 years (Supplemental Table 1). Hothersall et al. [23] reported similar risks of hip fractures in women and men aged 40–84 years, although for some of the age ranges, the estimates were significant in men (Supplemental Table 1). A similar risk of an incident hip fracture was observed among the 292 participants in the Nurse’s Health Study with T1D [RR 6.4 (95 % CI 3.9–10.3)] [25]. By comparison, substantially higher risks of being admitted to hospital due to a hip fracture in women and men with T1D were observed in Sweden [SHR 11.7 (95 % CI 8.0–16.4) and 15.0 (95 % CI 11.1–19.8), respectively] [24]. The investigators of the latter study speculated that rehospitalization of particularly severe cases of T1D after the same fracture could have inflated risk estimates. Other factors such as osteoporosis being particularly prevalent in Sweden and the fairly low number of incident fractures (n = 112) in the study could have influenced the findings. To that end, compared to age-matched controls, Nicodemus et al. [26] observed an equivalent increase in the risk of a hip fracture among postmenopausal women with T1D that were on average 60.9 years [RR 12.25 (95 % CI 5.05–29.7)]. However, the study comprised only five incident fracture cases (Supplemental Table 1). By contrast, Vestergaard et al. [27] reported risk estimates for any and hip fractures in a Danish population- and register-based case–control study that included more than 4000 patients with T1D and almost half a million controls. The relative risk of both any and a hip fracture was increased in patients with T1D by 1.30 (95 % CI 1.16–1.46) and 1.70 (95 % CI 1.31–2.21) (Supplemental Table 1), respectively. Using a Taiwanese health insurance database, Liao et al. [28] observed a modest increase in fracture risk among 2992 individuals with T1D that sustained 382 fractures [HR 1.22 (95 % CI 1.1–1.36)]. Compared to most of the previously mentioned investigations, risk estimates were substantially lower in the latter two investigations. Several factors may account for the discrepancies including inadequacies of registries to correctly identify those with T1D and differences in the composition of study populations. For example, Miao et al. [24] included cases that had been diagnosed with T1D before the age of 30, which most probably increased the likelihood of the cases being correctly classified as having T1D. However, this method of case selection may have inflated the risk estimates as cases developed T1D prior to or at time of peak bone mass and had endured diabetes for a longer duration.

Importantly, not all previous investigations have shown increased fracture risk in T1D. Melchior et al. [29] observed similar distal forearm and hip fracture risks in insulin-treated patients with diabetes, and Hanley et al. [30] reported comparable prevalence of vertebral deformities in men and women with T1D and controls. Notably, their investigations comprised only a modest number of fractures (n = 30) or deformities (n = 19), which may have influenced the results.

Thus, the risk estimates reported in the previously mentioned investigations on fracture risk in T1D vary noticeably. The estimates are however bound to diverge considerably due to differences in the design of the investigations, variation in age of the participants, inclusion of diverse ethnic groups, uneven availability of important health-related information such as body mass index, concomitant medication, lifestyle factors such as smoking and previous medical history including past fractures.

In conclusion, fracture risk including in particular that of the hip is significantly increased in individuals with T1D. Future studies that provide information on fracture risks at other fracture sites in different ethnic groups and in young adults are very much needed.

Type 2 Diabetes and Fractures

Although lower than in patients with T1D, the relative risk of fractures—including that of the hip—is also increased in patients with T2D. Biologically, these two types of diabetes differ markedly as T2D is characterized by reduced insulin sensitivity in addition to an inadequate secretion of insulin but also a tendency toward co-existing conditions related to metabolic syndrome such as obesity.

Although the incidence of T2D in children and adolescents is increasing [2], the disease is most often diagnosed in adulthood. Therefore, T2D is of course much less likely than T1D to influence peak bone mass. Several other factors may influence bone mass in T2D including bone anabolic effects of insulin and common disorders such as osteoarthritis.

In developing countries, the highest prevalence of T2D is observed in individuals aged 45–64 years. By contrast, T2D is most common in adults aged more than 64 years in developed countries [7], which corresponds with the incidence of osteoporotic fractures. The risk of fractures including fragility fractures in individuals with T2D has been addressed in several investigations that cover these age ranges in men as in women. However, a minority of investigations includes elderly patients or different ethnic groups. Further, most of the studies report risks of hip but not other types of fractures.

Numerous studies have reported an increased risk of hip fracture in individuals with T2D [23, 25, 26, 27, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41]. Risk estimates reported in these studies vary significantly (Supplemental Table 2), with the incidence rate ratio reported to be as low as 1.05 (95 % CI 1.01–1.10) in the oldest group (women aged more than 84 years) [23], compared with a relative risk of 3.1 (95 % CI 2.3–4.0) in women known to have had T2D for at least 11 years [25]. Furthermore, a number of studies have not shown an increased risk of hip fractures in T2D [42, 43, 44], or the risk estimate was insignificant until data were adjusted for confounders [45, 46]. Similar fracture rates in individuals with T2D and controls despite a tendency toward higher BMI and BMD in the former suggest that fracture risk is increased in T2D after adjustment for differences in body composition or bone mass. In some investigations, the risk of hip fractures was increased in women but not men [23, 40, 46] or in men but not women [34]. A limited number of studies stratified data according to age of the participants, and the results are not consistent across strata. Chen et al. [37] observed higher fracture risk among the youngest but not among the most senior participants, whereas Hothersall et al. [23] showed lower or no risks in the youngest participants and increased risk among elderly women. Ambiguities aside, the risk of a hip fracture appears modestly increased in T2D. By contrast, data on risks of fractures at other sites are limited.

Investigating the risk of any fractures, most studies have reported higher risk of a fracture in T2D [27, 28, 34, 35, 36, 40, 41, 47]. Conversely, the risk was not increased in a few investigations [48, 49]. Furthermore, although higher risk was observed in men, Ahmed et al. [46] observed similar risk of non-vertebral fractures in women with T2D as in controls. In general, investigations with the largest study populations and highest number of fractures observed an increased risk of any or non-vertebral fractures. Though not uniformly, increased risk of fractures at specific sites such as proximal humerus fractures [32, 35, 43], wrist or distal forearm [27, 41], lower leg [35, 41], ribs [34] and foot have been reported [34, 35]. Furthermore, vertebral fractures seem more prevalent in some [34, 35, 50] but not all studies [30, 48]. Importantly, some studies have not observed higher risks of fractures at any of these sites [48], and van Daele et al. [49] even reported lower risk of non-vertebral fracture in Dutch women with T2D. The conclusions of the latter investigation were based on self-reported fractures among a modest 38 women with non-insulin-dependent diabetes mellitus (NIDDM) and 449 women without NIDDM. Based on the presently available information on the distribution of fractures in individuals with T2D, peripheral fractures as well as those of the hip appear most prevalent. These observations suggest that T2D mainly has detrimental effects on cortical and not on trabecular bone as generally observed in primary osteoporosis. However, the specific effects of T2D on fractures are difficult to disentangle as the disease develops in later life and may well co-exist with osteoporosis.

Of course, T2D does not only differ from T1D in terms of pathophysiology and average debut age but also in the panel of medications used to control it.

Ivers et al. [43] observed higher risk of any fracture only in individuals with T2D for at least 10 years [RR 2.9 (95 % CI 1.2–7.0)], whereas Leslie et al. [33] reported higher risks of a composite fracture endpoint consisting of hip, wrist and spine fractures in men and women who had sustained diabetes for at least 5 years. Potentially to some extent mirroring the effects of disease duration, studies have shown that the risks of non-vertebral but not hip fracture in elderly women as well as any fracture in senior men are higher in those on insulin treatment [32, 51]. In addition, proper glycemic control may influence fracture risk. Using data from the Rotterdam Study, Oie et al. [44] observed higher risk of any fracture in those with unsatisfactorily controlled diabetes [RR 1.31 (95 % CI 1.00–1.71)] but not in patients who were adequately treated [RR 0.85 (95 % CI 0.63–1.15)]. Analogous differences were however not observed at the hip [RR 1.15 (95 % CI 0.68–1.94)]. Antidiabetic medication and treatment effect as well as duration of the disease may have deleterious effects on bone. The influence of antidiabetic medication is not within the scope of this review and is further examined in the review by Palermo et al. [52].

The majority of the published investigations of fracture risk in T2D are based on study populations recruited in Europe or the USA; therefore, most of the current information on fracture risk is based on effects observed in white adults even though the incidence of T2D is significant in other ethnic groups. A minority of mainly US investigations such as the Women’s Health Initiative Study [35] and The Health, Aging, and Body Composition Study [47] include black adults in the study populations, and although Japanese, Taiwanese and Singapore Chinese study populations have been investigated, future studies need to diversify in order to provide much needed data on bone status in all ethnic groups since the burden of T2D appears to expanding.

Acknowledging the limitations, currently available information on the risk of fractures including hip fractures in individuals with T2D has been assessed in meta-analyses that were published in 2007 and 2015. Janghorbani et al. [18] and Vestergaard [19] reported relative risks of hip fractures of 1.7 (95 % CI 1.3–2.2) and 1.38 (95 % CI 1.25–1.53), respectively. The meta-analysis by the latter included assessments of the risks of spine, wrist and any fracture in patients with T2D, whereas the former only assessed the risk of a hip fracture. In Vestergaard’s [19] meta-analysis, the risk of a fracture in the spine was not different from that observed in controls [RR 0.93 (95 % CI 0.63–1.37)], whereas the risks of a wrist and any fracture were increased [RR 1.19 (95 % CI 1.01–1.41) and 1.19 (95 % CI 1.11–1.27), respectively], although the risk of any fracture was increased only after exclusion of two studies that contributed significant heterogeneity to the meta-analysis [19]. Incorporating studies available after 2007 in their meta-analysis, Fan et al. [20] observed marginally lower risks of hip fractures in patients with T2D [RR 1.34 (95 % CI 1.19–1.51)]. Contrary to the smaller meta-analysis by Janghorbani et al. [18], their meta-analysis showed equal risks of hip fracture in men and women.

Different ethnicities, duration of disease and use of antidiabetic drugs, variable methods used to recruit participants as well as the shortcomings that are inherent in register-based studies may all contribute to the fairly substantial variation in the results of presently published investigations of fracture risk in T2D. Future studies of the effects of T2D on bone in children and adolescents as well as different ethnic groups are necessary. Further, evaluations of the impact of glucose control on fracture risk in T2D are also much needed.

Risk Factors for Fracture in Diabetes

Several Risk Factors may Influence the Fracture Risk in Patients with Diabetes

First, the prevalence of diabetes and of fractures increases with age [53, 54]. Age is a strong predictor of fracture in non-diabetics [54], and age is also related to fractures in diabetes patients [55]. Longevity remains substantially decreased in patients with diabetes [56]; therefore, the increasing fracture burden in diabetes patients compared to non-diabetics cannot be explained by patients reaching high age. Patients with T1D are at increased risk of incident fractures throughout life [22]. As in non-diabetics the incidence of fracture increases with age in T1D [22]. Higher age increased the risk of a fracture in a study pooling patients with T1D and T2D [55]. When applying the fracture risk assessment tool (FRAX) model that predicts 10-year fracture, the gradient of risk for age as regards major osteoporotic fracture was the same in persons with and without diabetes [55]. However, for hip fractures, age had a somewhat lower impact on the risk of fracture. The hazard ratios of fracture for patients with diabetes are higher in younger age groups compared to those aged 80 or more. Fracture risk in diabetes patients is reported to range from no increase in relative risk (RR = 1) for T2D [23] to a ninefold increased relative risk of T1D [57] (Supplemental Tables 1, 2). The difference in risk estimates between studies may depend on the age of cohort analyzed (low absolute risk of fractures among young persons) as well as different fracture rates between patients with T1D and T2D [19].

Second, female gender is a strong fracture predictor in non-diabetics [58]. Fracture rates are higher in women with diabetes than in men with diabetes [22, 23]. In particular, young women with T1D have an increased fracture risk compared to non-diabetes women, whereas men are at an increased risk at older age [22]. When applying the FRAX model, gender impacts fracture risk similar between diabetes and non-diabetics based on the current knowledge [55]. Based on the existing evidence, gender influences the fracture risk in patients with diabetes in the same manner as in non-diabetics.

Third, a previous fracture impacts on the risk of sustaining a new fracture in both patients with diabetes and in non-diabetic individuals [55, 59].

Smoking and high alcohol consumption are associated with the incidence of diabetes [60, 61], but they are also strong risk factors for fracture [62, 63]. Information on smoking and alcohol use is often inaccurate or missing in epidemiological studies [64] or may be assessed by proxy variables as chronic obstructive pulmonary disease for smoking and diagnoses of alcohol and substance abuse as high alcohol intake [65] and in some studies based on self-report of previous use [32]. The inadequate collection of information on smoking and alcohol limits the current knowledge on the impact on the risk of fracture. However, both smoking and alcohol use are associated with fracture risk in patients with diabetes and with a similar gradient of risk when comparing with non-diabetics [55]. Further evaluation on the effect of smoking and alcohol on fracture risk is needed in patients with diabetes.

Fourth, diabetes may not only induce a need for drug prescriptions that is different from that of the general population but some therapeutics may also of course lead to diabetes. Oral glucocorticoids may induce diabetes [66] and bone loss [67]. Diabetes is induced due to insulin resistance and beta-cell dysfunction [66, 68]. The bone loss develops through a direct inhibition of the osteoblast and by inhibition of estrogens and growth hormone that indirectly increase osteoclast activity and bone resorption [67].

Previous register-based studies have not excluded oral glucocorticoid users in their assessments of diabetes on fracture risk [55, 64]. Therefore, it is impossible to determine whether the effect of diabetes on fracture risk is explained solely by diabetes or, at least in part, glucocorticoids use inducing diabetes. One study reports that the effect of glucocorticoids on fractures is not different between individuals with and without diabetes [55], whereas another study finds no association between glucocorticoids use and fracture within diabetes patients [59]. Thus, glucocorticoids seem to have the same impact on fracture risk in non-diabetics and patients with diabetes.

Patients with diabetes complications are more likely to receive antiepileptics than is the case for the background population as it is used to control neuropathic pain [69]. However, antiepileptic treatment or epilepsy per se is a risk factor for fracture [70] and has also been associated with fracture risk in diabetes [59]. Cytochrome P450 enzyme-inducing antiepileptics are mainly associated with negative effects on bone and decrease bone mineral density (BMD) possibly by inactivating vitamin D into inactive metabolites causing vitamin D deficiency [71]. Patients with diabetes are already associated with a decreased vitamin D level compared to non-diabetics [72]; thus, the effect of antiepileptics may be enhanced in patients with diabetes. In addition, antiepileptics may also be used as seizure prevention and thus represent individuals at risk of fracture. Further research is needed to determine the effect of antiepileptic treatment on the fracture burden in patients with diabetes. This is however somewhat challenging due to the clustering of diabetes complications and antiepileptic use, both of which may signal increased fracture risk compared with patients with milder, better controlled or earlier-stage disease.

Fifth, some comorbid conditions are more prevalent among patients with diabetes. Pancreatitis causes pancreas destruction with exocrine and endocrine dysfunction, potentially resulting in malabsorption and diabetes [73]. Malabsorption and malnutrition may also increase the risk of a fracture as well as alcohol abuse that may cause pancreatitis [73]. The type 3c diabetes, which is diabetes caused by exocrine pancreatic insufficiency, is likely to be underdiagnosed within the diabetes population and registered as either T1D or T2D [74]. Diabetes type 3c may make up 9 % of the diabetes population; however, this estimate is based on hospital-admitted patients [75]. It is unknown whether some patients with diabetes suffer from undiagnosed pancreatic insufficiency which may influence fracture rates. Thus, registry-based studies may analyze patients with diabetes that is caused by pancreatic disease. Many of these registry-based studies do not exclude individuals with a diagnosis of pancreatitis [22, 55, 64, 65]. Pancreatitis is strongly associated with fracture (3.5-fold increased risk of a hip fracture) [76] that may explain some and perhaps most of the increased fracture risk in patients with diabetes. Future studies should determine to what extent pancreatitis influences fractures risk in diabetes. This may be accomplished by either excluding patients with pancreatitis or conduct sensitivity analyses with and without patients with pancreatitis.

Furthermore, T1D and T2D are frequently accompanied by hyperglycemia that may cause a loss of minerals due to polyuria. However, the effects of inadequate nutrient utilization and insulin deficiency on bone metabolism have not been thoroughly investigated. T2D is associated with obesity [1]. The effect of obesity on fracture risk is ambiguous. BMI was protective against osteoporotic fracture and hip fracture before BMD adjustment, and hereafter a slight significant increase in osteoporotic fractures was observed [77]. According to a meta-analysis, obesity may in fact decrease the risk of hip fracture [78], possibly due to protective effects of fat deposits at the hip that theoretically would decrease fracture risk in relation to falls [79, 80]. Obesity per se does not seem to increase fracture risk, although humeral and elbow fractures may be increased [81]. Obesity may increase fracture risk by causing insulin resistance and inflammation [82]. While little is known about the effects of insulin resistance at tissue level in humans, inflammation may theoretically increase bone resorption through activation of NF-kB signaling. Bone marrow adipose tissue (BMAT) expands with age, obesity and diabetes [83]. While the physiological function of BMAT remains unknown, human studies have shown an inverse correlation between bone and BMAT formation, possibly due to a shift toward adipogenesis rather than osteogenesis [84]. Future studies that address the effects of BMAT on bone metabolism and fracture risk in patients with diabetes are needed.

T1D is associated with other autoimmune diseases, and among these are autoimmune thyroid disease and celiac disease [85]. Both hypothyroidism and hyperthyroidism are associated with an increased fracture risk [86], whereas it is controversial whether celiac disease relates to fracture [87]. These diseases may increase the number of fractures among patients with T1D. Studies of fracture risk have controlled for celiac disease but not consistently as the diagnosis is susceptible to underdiagnosis and misclassification. Further research is needed to investigate the extent of the fracture burden caused by other autoimmune diseases in patients with T1D.

Assessment of Risk

BMD measured by dual-energy X-ray absorptiometry (DXA) scan is both diagnostic for osteoporosis and predicts fracture risk [88, 89]. In contrast to T1D, bone mineral density (BMD) tends to be at the same level or even higher in patients with T2D [19, 90] and in patients with T1D the BMD is not decreased to a level that explains the observed fracture risk.

It is however uncertain whether the BMD difference is due to diabetes per se or disease-related characteristics. Furthermore, the apparently increased BMD in patients with T2D may be partly attributable by measurement error due to aortic calcification, osteoarthritis and diffuse idiopathic skeletal hyperostosis that all increase the BMD estimation. However, it is unlikely that these factors explain the entire increase in BMD among patients with T2D.

BMD measures the mineralized tissue and will not detect deterioration of any non-mineralized bone. It has been hypothesized that the bone is weakened in diabetes due to hyperglycemia and reactive oxygen species that create non-enzymatic collagen cross-links [91]. Changes in the collagen strength and structure are not measurable by the DXA which therefore may underestimate the fracture risk. Although BMD does not explain the difference in fracture risk, bone mass may still be a predictor but with another optimal threshold. In older patients with T2D, it has been suggested that a higher T-score threshold is needed to predict fracture in diabetes [92]. At any given T-score, the risk of a hip fracture or non-vertebral fracture is increased in 75-year-old patients with T2D compared to non-diabetics [93]. However, the relative fracture risk for patients with T2D compared to controls increases at lower BMD levels. To apply another T-score threshold in patients with diabetes, studies of younger diabetes individuals and investigation of T1D and T2D separately are needed because the fracture burden mainly is increased younger age groups and T1D.

FRAX uses common risk factors including BMD and provides 10-year fracture risk prediction estimates [94]. However, irrespective of the type of diabetes, fracture risk is higher in individuals with diabetes at any given level of T-score or FRAX score than in non-diabetics [65, 92]. The ability of T-score to predict fracture is similar in persons with and without diabetes, and as for T-scores another threshold for diagnosing osteoporosis may be needed [92]. It is tempting to apply another threshold to the FRAX model or T-scores. Caution should be used because a change in the threshold would provide guidelines that suggest medical treatment of possible bone healthy individuals that would not develop a major fracture. A much more potent and reliable outlook would be to determine the pathogenesis of the increased fracture risk and thereby find a reliable fracture predictor in patients with diabetes.

Falls may besides a decrease in bone quality also explain the increased fracture risk in patients with diabetes. Diabetes patients may be more prone to falls caused by complications as neuropathy (decreased peripheral sensation) and retinopathy (decreased eyesight). In addition, hypoglycemic events [95] and orthostatic hypotension due to the use of antihypertensives or cardiovascular autonomic neuropathy may lead to falls and fractures. Patients with T1D and T2D without any complications had an almost twofold increased risk of a hip fracture, and furthermore, no single complication explained the increased fracture risk [64]. The risk of a fracture is still increased in diabetes patients after adjustment for hypoglycemic events [27] and after adjustment by self-reported falls [32, 35, 51]. Furthermore, intensive glycemic control did not increase fracture risk in the ACCORD trial [96]. Therefore, diabetic complications, hypoglycemic events and falls cannot fully explain the fracture risk in patients with diabetes. This suggests that additional underlying mechanisms as a decreased bone quality may influence fracture risk or that the registration of falls and hypoglycemic events in observational studies underestimates the true fracture risk. Currently, no studies have addressed the potential importance of orthostatic hypotension on fracture risk in patients with diabetes. Multidisciplinary intervention studies aiming at reducing falls and fracture risk in elderly patients with diabetes are needed.

Although many fracture predictors seem to be reliable in diabetes: age, gender, previous fracture, glucocorticoids use, smoking and alcohol use, other fracture predictors such as BMD and the FRAX model underestimate the fracture risk. This suggests that diabetes per se may increase fracture risk. However, it is unknown how factors such as falls, pancreatitis and autoimmune diseases impact the fracture risk in patients with diabetes and whether these factors may explain the observed fracture burden.

Population Attributable Risk and Global Forecast

The population attributable risk of fractures depends on the prevalence of diabetes and the relative risk of a fracture associated with diabetes. If the fracture risk and the prevalence of diabetes are high, a very large number of fractures may potentially be attributable to diabetes. The risk of a fracture attributable to diabetes may be high among older individuals where the absolute risk of fractures is high (it may be as high as 4–6 % per year in some Scandinavian countries) and the prevalence of diabetes (predominantly T2D) is high. With a prevalence of diabetes of 9 % [97] and an estimated RR of hip fracture for patients with T1D of 6.9 and a RR of hip fracture for patients with T2D of 1.4 [19], the population attributable risk of hip fractures for T1D is 5.1 % and for T2D it is 3.0 % assuming that 10 % of the prevalence is T1D and 90 % is T2D. Supplemental Table 3 further explains the calculation.

Figure 1 shows a theoretical relationship between prevalence of diabetes and population attributable risk of hip fractures. This model assumes that 10 % of the diabetes population has T1D and 90 % of the diabetes population has T2D.
Fig. 1

Population attributable risk of hip fractures with varying prevalence of diabetes assuming that 10 % of the prevalence is T1D and 90 % is T2D. Diamonds depict the total diabetes population, squares depict the type 1 diabetes population, and triangles depict the type 2 diabetes population. The prevalence of diabetes in adults is estimated to be 6.8 % in Europe, 9.6 % in North America and 10.9 % in the Middle East and North Africa [8]

Under the assumption that the relative risk is 1.19 for patients with T2D [19] and the relative risk is 3.16 for patients with T1D [21], the population attributable risk of any fracture is 1.9 and 1.5 % for type 1 and T2D, respectively. Figure 2 shows a theoretical relationship between prevalence of diabetes and population attributable risk of any fracture with the same assumptions as in Fig. 1. Changing the assumptions and subgroups (say age strata) may change the risk attributed to diabetes.
Fig. 2

Population attributable risk of any fracture with varying prevalence of diabetes assuming that 10 % of the prevalence is T1D and 90 % is T2D. Diamonds depict the total diabetes population, squares depict the type 1 diabetes population, and triangles depict the type 2 diabetes population

Thus, a significant proportion of all fractures and especially hip fractures may potentially be attributable to diabetes. Although T1D is associated with the highest risk of fracture, the number of patients with T2D is so large that it makes a significant contribution to the overall fracture burden. With an expected increase in the diabetes population from 328 to 592 million 2035 the number of hip fractures also may increase [8].

The WHO forecast a significant increase in the number of subjects with diabetes. In certain regions such as North America and the Indian subcontinent, the prevalence of diabetes is very high and the mean population age is increasing worldwide. In such populations the increase in the prevalence of diabetes and the aging of the population may lay the foundations for a significant increase in the number of fractures.

Limitations on Fracture Epidemiology in Diabetes

The epidemiological knowledge of fracture rates relies on the capture of fracture events. Clinically significant fractures including a hip fracture are unlikely to be undetected, and thus, the outcome is reliable. The outcome becomes more unreliable in clinically less conspicuous fractures such as vertebral fractures. Studies have reported prevalence of 37–50 % for a vertebral fracture in diabetes populations based on the Genant classification [98, 99]. This prevalence is higher than what is reported in population-based studies [100] and suggests that diabetes patients in addition to an increased risk of hip fracture also may be at an increased risk of vertebral fracture. These findings suggest an underestimation of vertebral fracture in patients with diabetes; however, injudicious use of the Genant classification may group patients with degenerative spine diseases such as Scheurmann’s disease as having sustained vertebral fracture and thereby overestimate fracture prevalence. Direct comparisons between patients with and without diabetes are needed to disentangle whether increased vertebral fracture prevalence in patients with diabetes can be confirmed. Furthermore, the increased risk of hip fracture may be caused by especially diaphyseal and subtrochanteric fractures and thereby represent less common types of fracture as the femoral neck [101]. Hip fractures may present differently in patients with diabetes than in primary osteoporosis; however, further research is needed. As discussed above, epidemiological studies have difficulty distinguishing between patients with T1D and T2D, and in many cases, they are therefore analyzed as one patient group despite the difference in risk [55, 65]. Although the FRAX model underestimates the fracture risk in patients with diabetes, we speculate that this could be caused by the subpopulation of T1D patients in the total diabetes population [65]. Future studies should clearly distinguish between patients with T1D and T2D because pooled population studies are difficult to interpret. Furthermore, most investigations of fracture risk in T1D and T2D have been conducted in European or North American populations. Although reports on fracture risk in Asian populations are available, future studies that incorporate different ethnic groups are needed.

Studies evaluating the effect of pancreatitis on fracture risk in patients with diabetes are needed. Pancreatitis is related both to diabetes and to fracture, and it may increase the observed fracture risk in the diabetes population. It is unknown to what extent pancreatitis may drive the increased risk of fracture in current studies [19]. Also, oral glucocorticoids use may induce both diabetes and fracture and inflate the association between diabetes and fracture. This needs further investigation.

Smoking and high alcohol use may also affect both the risk of diabetes and fracture. These risk factors are difficult to collect in larger populations, and proxy variables of uncertain value have been used. It is largely unknown whether the increased fracture risk in diabetes partly may be caused by modifiable lifestyle choices including physical inactivity, alcohol intake and smoking. Similarly, the registration and validity of falls and hypoglycemia are uncertain in population-based studies, as many events may not be reported. Again, further research is needed to investigate the association between falls and fracture in patients with diabetes. We speculate that the observed risk of fracture in diabetes may be partly caused by falls and the lack of controlling for shared risk factors such as smoking, alcohol, pancreatitis and oral glucocorticoids. This may overestimate the association of diabetes per se with increased fracture risk and reduced bone quality.

Clinical Recommendations

Physicians should be aware that patients with diabetes are at an increased risk of a fracture. Moreover, current risk factors may not adequately predict these fractures so particular attention is warranted.

Imaging of the spine by vertebral fracture assessment or spine X-ray may help predict fractures in diabetes patients as a previous fracture is a predictor of new fractures and that the prevalence of a vertebral fracture may be high enough in patients with diabetes to justify an increased diagnostic vigilance. Fall prevention may also decrease the fracture risk in patients with diabetes, but randomized studies on the effect of falls prevention in diabetes are missing. In patients with diabetes specifically, this may be achieved by detecting and preventing hypoglycemic events and orthostatic hypotension due to antihypertensive treatment and of course by optimum long-term management to reduce the risk of neuropathy and retinopathy.

Directions for Future Research

The global burden of both diabetes and fractures is estimated to increase substantially in the following years. Although patients with diabetes contribute significantly to the number of fractures and especially hip fractures, current fracture predictors do not explain this increase though. Furthermore, the effect of falls and hypoglycemia—reversible risk factors—on the risk of fracture in patients with diabetes should be further investigated. Future epidemiological studies should aim at disentangling the effects of shared risk factors for diabetes and fracture as pancreatitis and oral glucocorticoids. Furthermore, the incidence of vertebral fracture in patients with diabetes needs to be investigated in order to develop preventive strategies better suited for reducing the fracture burden in conjunction with an increasing societal challenge posed by diabetes.

Notes

Acknowledgments

Morten Frost was supported by a grant from the Danish Council for Independent Research.

Conflict of interest

Bo Abrahamsen reports current institutional research grants and contracts with Novartis and UCB, past institutional research contracts with Amgen and NPS Pharmaceuticals. Jakob Starup-Linde, Morten Frost, Peter Vestergaard declare that they have no conflict of interest.

Supplementary material

223_2016_175_MOESM1_ESM.docx (43 kb)
Supplementary material 1 (DOCX 42 kb)

References

  1. 1.
    American Diabetes Association (2012) Diagnosis and classification of diabetes mellitus. Diabetes Care 35(Suppl 1):S64–S71CrossRefGoogle Scholar
  2. 2.
    Hamman RF, Bell RA, Dabelea D, D’Agostino RB Jr, Dolan L, Imperatore G, Lawrence JM, Linder B, Marcovina SM, Mayer-Davis EJ, Pihoker C, Rodriguez BL, Saydah S, SEARCH for Diabetes in Youth Study Group (2014) The SEARCH for Diabetes in Youth study: rationale, findings, and future directions. Diabetes Care 37:3336–3344PubMedPubMedCentralCrossRefGoogle Scholar
  3. 3.
    DIAMOND Project Group (2006) Incidence and trends of childhood type 1 diabetes worldwide 1990–1999. Diabet Med 23:857–866CrossRefGoogle Scholar
  4. 4.
    Hawa MI, Kolb H, Schloot N, Beyan H, Paschou SA, Buzzetti R, Mauricio D, De Leiva A, Yderstraede K, Beck-Neilsen H, Tuomilehto J, Sarti C, Thivolet C, Hadden D, Hunter S, Schernthaner G, Scherbaum WA, Williams R, Brophy S, Pozzilli P, Leslie RD, Action LADA Consortium (2013) Adult-onset autoimmune diabetes in Europe is prevalent with a broad clinical phenotype: action LADA 7. Diabetes Care 36:908–913PubMedPubMedCentralCrossRefGoogle Scholar
  5. 5.
    Diaz-Valencia PA, Bougneres P, Valleron AJ (2015) Global epidemiology of type 1 diabetes in young adults and adults: a systematic review. BMC Public Health 15:255-015-1591-yCrossRefGoogle Scholar
  6. 6.
    Irgens HU, Molnes J, Johansson BB, Ringdal M, Skrivarhaug T, Undlien DE, Sovik O, Joner G, Molven A, Njolstad PR (2013) Prevalence of monogenic diabetes in the population-based Norwegian Childhood Diabetes Registry. Diabetologia 56:1512–1519PubMedCrossRefGoogle Scholar
  7. 7.
    Wild S, Roglic G, Green A, Sicree R, King H (2004) Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 27:1047–1053PubMedCrossRefGoogle Scholar
  8. 8.
    IDF Diabetes Atlas G (2015) Update of mortality attributable to diabetes for the IDF Diabetes Atlas: Estimates for the year 2013. Diabetes Res Clin Pract 109:461–465CrossRefGoogle Scholar
  9. 9.
    Patterson C, Guariguata L, Dahlquist G, Soltesz G, Ogle G, Silink M (2014) Diabetes in the young—a global view and worldwide estimates of numbers of children with type 1 diabetes. Diabetes Res Clin Pract 103:161–175PubMedCrossRefGoogle Scholar
  10. 10.
    Fourlanos S, Dotta F, Greenbaum CJ, Palmer JP, Rolandsson O, Colman PG, Harrison LC (2005) Latent autoimmune diabetes in adults (LADA) should be less latent. Diabetologia 48:2206–2212PubMedCrossRefGoogle Scholar
  11. 11.
    Shields BM, Hicks S, Shepherd MH, Colclough K, Hattersley AT, Ellard S (2010) Maturity-onset diabetes of the young (MODY): how many cases are we missing? Diabetologia 53:2504–2508PubMedCrossRefGoogle Scholar
  12. 12.
    Green A, Sortso C, Jensen PB, Emneus M (2014) Validation of the Danish national diabetes register. Clin Epidemiol 7:5–15PubMedPubMedCentralCrossRefGoogle Scholar
  13. 13.
    Johnell O, Kanis JA (2006) An estimate of the worldwide prevalence and disability associated with osteoporotic fractures. Osteoporos Int 17:1726–1733PubMedCrossRefGoogle Scholar
  14. 14.
    Gale EA, Gillespie KM (2001) Diabetes and gender. Diabetologia 44:3–15PubMedCrossRefGoogle Scholar
  15. 15.
    Wandell PE, Carlsson AC (2014) Gender differences and time trends in incidence and prevalence of type 2 diabetes in Sweden—a model explaining the diabetes epidemic worldwide today? Diabetes Res Clin Pract 106:e90–e92PubMedCrossRefGoogle Scholar
  16. 16.
    Hernlund E, Svedbom A, Ivergard M, Compston J, Cooper C, Stenmark J, McCloskey EV, Jonsson B, Kanis JA (2013) Osteoporosis in the European Union: medical management, epidemiology and economic burden: a report prepared in collaboration with the International Osteoporosis Foundation (IOF) and the European Federation of Pharmaceutical Industry Associations (EFPIA). Arch Osteoporos 8:136-013-0136-1 (Epub 2013 Oct 11) CrossRefGoogle Scholar
  17. 17.
    Hamann C, Kirschner S, Gunther KP, Hofbauer LC (2012) Bone, sweet bone—osteoporotic fractures in diabetes mellitus. Nat Rev Endocrinol 8:297–305PubMedCrossRefGoogle Scholar
  18. 18.
    Janghorbani M, Van Dam RM, Willett WC, Hu FB (2007) Systematic review of type 1 and type 2 diabetes mellitus and risk of fracture. Am J Epidemiol 166:495–505PubMedCrossRefGoogle Scholar
  19. 19.
    Vestergaard P (2007) Discrepancies in bone mineral density and fracture risk in patients with type 1 and type 2 diabetes—a meta-analysis. Osteoporos Int 18:427–444PubMedCrossRefGoogle Scholar
  20. 20.
    Fan Y, Wei F, Lang Y, Liu Y (2016) Diabetes mellitus and risk of hip fractures: a meta-analysis. Osteoporos Int 27(1):219–228PubMedCrossRefGoogle Scholar
  21. 21.
    Shah VN, Shah CS, Snell-Bergeon JK (2015) Type 1 diabetes and risk of fracture: meta-analysis and review of the literature. Diabet Med 32:1134–1142PubMedPubMedCentralCrossRefGoogle Scholar
  22. 22.
    Weber DR, Haynes K, Leonard MB, Willi SM, Denburg MR (2015) Type 1 diabetes is associated with an increased risk of fracture across the life span: a population-based cohort study using the health improvement network (THIN). Diabetes Care 38(10):1913–1920PubMedPubMedCentralCrossRefGoogle Scholar
  23. 23.
    Hothersall EJ, Livingstone SJ, Looker HC, Ahmed SF, Cleland S, Leese GP, Lindsay RS, McKnight J, Pearson D, Philip S, Wild SH, Colhoun HM (2014) Contemporary risk of hip fracture in type 1 and type 2 diabetes: a national registry study from Scotland. J Bone Miner Res 29:1054–1060PubMedCrossRefGoogle Scholar
  24. 24.
    Miao J, Brismar K, Nyren O, Ugarph-Morawski A, Ye W (2005) Elevated hip fracture risk in type 1 diabetic patients: a population-based cohort study in Sweden. Diabetes Care 28:2850–2855PubMedCrossRefGoogle Scholar
  25. 25.
    Janghorbani M, Feskanich D, Willett WC, Hu F (2006) Prospective study of diabetes and risk of hip fracture: the Nurses’ Health Study. Diabetes Care 29:1573–1578PubMedCrossRefGoogle Scholar
  26. 26.
    Nicodemus KK, Folsom AR, Iowa Women’s Health Study (2001) Type 1 and type 2 diabetes and incident hip fractures in postmenopausal women. Diabetes Care 24:1192–1197PubMedCrossRefGoogle Scholar
  27. 27.
    Vestergaard P, Rejnmark L, Mosekilde L (2005) Relative fracture risk in patients with diabetes mellitus, and the impact of insulin and oral antidiabetic medication on relative fracture risk. Diabetologia 48:1292–1299PubMedCrossRefGoogle Scholar
  28. 28.
    Liao CC, Lin CS, Shih CC, Yeh CC, Chang YC, Lee YW, Chen TL (2014) Increased risk of fracture and postfracture adverse events in patients with diabetes: two nationwide population-based retrospective cohort studies. Diabetes Care 37:2246–2252PubMedCrossRefGoogle Scholar
  29. 29.
    Melchior TM, Sorensen H, Torp-Pedersen C (1994) Hip and distal arm fracture rates in peri- and postmenopausal insulin-treated diabetic females. J Intern Med 236:203–208PubMedCrossRefGoogle Scholar
  30. 30.
    Hanley DA, Brown JP, Tenenhouse A, Olszynski WP, Ioannidis G, Berger C, Prior JC, Pickard L, Murray TM, Anastassiades T, Kirkland S, Joyce C, Joseph L, Papaioannou A, Jackson SA, Poliquin S, Adachi JD, Canadian Multicentre Osteoporosis Study Research Group (2003) Associations among disease conditions, bone mineral density, and prevalent vertebral deformities in men and women 50 years of age and older: cross-sectional results from the Canadian Multicentre Osteoporosis Study. J Bone Miner Res 18:784–790PubMedCrossRefGoogle Scholar
  31. 31.
    Cortes-Sancho R, Perez-Castrillon JL, Martin-Escudero JC, Iglesias S, Alvarez-Manzanares P, Ramos R (2004) Type 2 diabetes mellitus as a risk factor for hip fracture. J Am Geriatr Soc 52:1778–1779PubMedCrossRefGoogle Scholar
  32. 32.
    Schwartz AV, Sellmeyer DE, Ensrud KE, Cauley JA, Tabor HK, Schreiner PJ, Jamal SA, Black DM, Cummings SR, Study of Osteoporotic Features Research Group (2001) Older women with diabetes have an increased risk of fracture: a prospective study. J Clin Endocrinol Metab 86:32–38PubMedCrossRefGoogle Scholar
  33. 33.
    Leslie WD, Lix LM, Prior HJ, Derksen S, Metge C, O’Neil J (2007) Biphasic fracture risk in diabetes: a population-based study. Bone 40:1595–1601PubMedCrossRefGoogle Scholar
  34. 34.
    Melton LJ III, Leibson CL, Achenbach SJ, Therneau TM, Khosla S (2008) Fracture risk in type 2 diabetes: update of a population-based study. J Bone Miner Res 23:1334–1342PubMedPubMedCentralCrossRefGoogle Scholar
  35. 35.
    Bonds DE, Larson JC, Schwartz AV, Strotmeyer ES, Robbins J, Rodriguez BL, Johnson KC, Margolis KL (2006) Risk of fracture in women with type 2 diabetes: the Women’s Health Initiative Observational Study. J Clin Endocrinol Metab 91:3404–3410PubMedCrossRefGoogle Scholar
  36. 36.
    Lipscombe LL, Jamal SA, Booth GL, Hawker GA (2007) The risk of hip fractures in older individuals with diabetes: a population-based study. Diabetes Care 30:835–841PubMedCrossRefGoogle Scholar
  37. 37.
    Chen HF, Ho CA, Li CY (2008) Increased risks of hip fracture in diabetic patients of Taiwan: a population-based study. Diabetes Care 31:75–80PubMedCrossRefGoogle Scholar
  38. 38.
    Koh WP, Wang R, Ang LW, Heng D, Yuan JM, Yu MC (2010) Diabetes and risk of hip fracture in the Singapore Chinese Health Study. Diabetes Care 33:1766–1770PubMedPubMedCentralCrossRefGoogle Scholar
  39. 39.
    Forsen L, Meyer HE, Midthjell K, Edna TH (1999) Diabetes mellitus and the incidence of hip fracture: results from the Nord-Trondelag Health Survey. Diabetologia 42:920–925PubMedCrossRefGoogle Scholar
  40. 40.
    de Liefde II, van der Klift M, de Laet CE, van Daele PL, Hofman A, Pols HA (2005) Bone mineral density and fracture risk in type-2 diabetes mellitus: the Rotterdam Study. Osteoporos Int 16:1713–1720PubMedCrossRefGoogle Scholar
  41. 41.
    Holmberg AH, Johnell O, Nilsson PM, Nilsson J, Berglund G, Akesson K (2006) Risk factors for fragility fracture in middle age: a prospective population-based study of 33,000 men and women. Osteoporos Int 17:1065–1077PubMedCrossRefGoogle Scholar
  42. 42.
    Dobnig H, Piswanger-Solkner JC, Roth M, Obermayer-Pietsch B, Tiran A, Strele A, Maier E, Maritschnegg P, Sieberer C, Fahrleitner-Pammer A (2006) Type 2 diabetes mellitus in nursing home patients: effects on bone turnover, bone mass, and fracture risk. J Clin Endocrinol Metab 91:3355–3363PubMedCrossRefGoogle Scholar
  43. 43.
    Ivers RQ, Cumming RG, Mitchell P, Peduto AJ, Blue Mountains Eye Study (2001) Diabetes and risk of fracture: The Blue Mountains Eye Study. Diabetes Care 24:1198–1203PubMedCrossRefGoogle Scholar
  44. 44.
    Oei L, Zillikens MC, Dehghan A, Buitendijk GH, Castano-Betancourt MC, Estrada K, Stolk L, Oei EH, van Meurs JB, Janssen JA, Hofman A, van Leeuwen JP, Witteman JC, Pols HA, Uitterlinden AG, Klaver CC, Franco OH, Rivadeneira F (2013) High bone mineral density and fracture risk in type 2 diabetes as skeletal complications of inadequate glucose control: the Rotterdam Study. Diabetes Care 36:1619–1628PubMedPubMedCentralCrossRefGoogle Scholar
  45. 45.
    Martinez-Laguna D, Tebe C, Javaid MK, Nogues X, Arden NK, Cooper C, Diez-Perez A, Prieto-Alhambra D (2015) Incident type 2 diabetes and hip fracture risk: a population-based matched cohort study. Osteoporos Int 26:827–833PubMedCrossRefGoogle Scholar
  46. 46.
    Ahmed LA, Joakimsen RM, Berntsen GK, Fonnebo V, Schirmer H (2006) Diabetes mellitus and the risk of non-vertebral fractures: the Tromso study. Osteoporos Int 17:495–500PubMedCrossRefGoogle Scholar
  47. 47.
    Strotmeyer ES, Cauley JA, Schwartz AV, Nevitt MC, Resnick HE, Bauer DC, Tylavsky FA, de Rekeneire N, Harris TB, Newman AB (2005) Nontraumatic fracture risk with diabetes mellitus and impaired fasting glucose in older white and black adults: the health, aging, and body composition study. Arch Intern Med 165:1612–1617PubMedCrossRefGoogle Scholar
  48. 48.
    Gerdhem P, Isaksson A, Akesson K, Obrant KJ (2005) Increased bone density and decreased bone turnover, but no evident alteration of fracture susceptibility in elderly women with diabetes mellitus. Osteoporos Int 16:1506–1512PubMedCrossRefGoogle Scholar
  49. 49.
    van Daele PL, Stolk RP, Burger H, Algra D, Grobbee DE, Hofman A, Birkenhager JC, Pols HA (1995) Bone density in non-insulin-dependent diabetes mellitus: the Rotterdam Study. Ann Intern Med 122:409–414PubMedCrossRefGoogle Scholar
  50. 50.
    Yamamoto M, Yamaguchi T, Yamauchi M, Kaji H, Sugimoto T (2009) Diabetic patients have an increased risk of vertebral fractures independent of BMD or diabetic complications. J Bone Miner Res 24:702–709PubMedCrossRefGoogle Scholar
  51. 51.
    Napoli N, Strotmeyer ES, Ensrud KE, Sellmeyer DE, Bauer DC, Hoffman AR, Dam TT, Barrett-Connor E, Palermo L, Orwoll ES, Cummings SR, Black DM, Schwartz AV (2014) Fracture risk in diabetic elderly men: the MrOS study. Diabetologia 57:2057–2065PubMedPubMedCentralCrossRefGoogle Scholar
  52. 52.
    Palermo A, D'Onofrio L, Eastell R, Schwartz AV, Pozzilli P, Napoli N (2015) Oral anti-diabetic drugs and fracture risk, cut to the bone: safe or dangerous? A narrative review. Osteoporos Int 26(8):2073–2089PubMedCrossRefGoogle Scholar
  53. 53.
    Center for Disease Control and Prevention (2015) Incidence of diagnosed diabetes per 1,000 population aged 18–79 years, by sex and age, United States. CDC, Atlanta, GA, pp 1997–2014Google Scholar
  54. 54.
    Sanders KM, Seeman E, Ugoni AM, Pasco JA, Martin TJ, Skoric B, Nicholson GC, Kotowicz MA (1999) Age- and gender-specific rate of fractures in Australia: a population-based study. Osteoporos Int 10:240–247PubMedCrossRefGoogle Scholar
  55. 55.
    Leslie WD, Morin SN, Lix LM, Majumdar SR (2014) Does diabetes modify the effect of FRAX risk factors for predicting major osteoporotic and hip fracture? Osteoporos Int 25:2817–2824PubMedCrossRefGoogle Scholar
  56. 56.
    Tancredi M, Rosengren A, Svensson AM, Kosiborod M, Pivodic A, Gudbjornsdottir S, Wedel H, Clements M, Dahlqvist S, Lind M (2015) Excess mortality among persons with type 2 diabetes. N Engl J Med 373:1720–1732PubMedCrossRefGoogle Scholar
  57. 57.
    Meyer HE, Tverdal A, Falch JA (1993) Risk factors for hip fracture in middle-aged Norwegian women and men. Am J Epidemiol 137:1203–1211PubMedCrossRefGoogle Scholar
  58. 58.
    Khosla S, Riggs BL (2005) Pathophysiology of age-related bone loss and osteoporosis. Endocrinol Metab Clin North Am 34:1015–1030PubMedCrossRefGoogle Scholar
  59. 59.
    Starup-Linde J, Gregersen S, Vestergaard P (2016) Associations with fracture in patients with diabetes: a nested case–control study. BMJ Open 6:e009686-2015-009686CrossRefGoogle Scholar
  60. 60.
    Foy CG, Bell RA, Farmer DF, Goff DC Jr, Wagenknecht LE (2005) Smoking and incidence of diabetes among U.S. adults: findings from the Insulin Resistance Atherosclerosis Study. Diabetes Care 28:2501–2507PubMedCrossRefGoogle Scholar
  61. 61.
    Carlsson S, Hammar N, Grill V, Kaprio J (2003) Alcohol consumption and the incidence of type 2 diabetes: a 20-year follow-up of the Finnish Twin Cohort Study. Diabetes Care 26:2785–2790PubMedCrossRefGoogle Scholar
  62. 62.
    Kanis JA, Johansson H, Johnell O, Oden A, De Laet C, Eisman JA, Pols H, Tenenhouse A (2005) Alcohol intake as a risk factor for fracture. Osteoporos Int 16:737–742PubMedCrossRefGoogle Scholar
  63. 63.
    Kanis JA, Johnell O, Oden A, Johansson H, De Laet C, Eisman JA, Fujiwara S, Kroger H, McCloskey EV, Mellstrom D, Melton LJ, Pols H, Reeve J, Silman A, Tenenhouse A (2005) Smoking and fracture risk: a meta-analysis. Osteoporos Int 16:155–162PubMedCrossRefGoogle Scholar
  64. 64.
    Vestergaard P, Rejnmark L, Mosekilde L (2009) Diabetes and its complications and their relationship with risk of fractures in type 1 and 2 diabetes. Calcif Tissue Int 84:45–55PubMedCrossRefGoogle Scholar
  65. 65.
    Giangregorio LM, Leslie WD, Lix LM, Johansson H, Oden A, McCloskey E, Kanis JA (2012) FRAX underestimates fracture risk in patients with diabetes. J Bone Miner Res 27:301–308PubMedCrossRefGoogle Scholar
  66. 66.
    Hwang JL, Weiss RE (2014) Steroid-induced diabetes: a clinical and molecular approach to understanding and treatment. Diabetes Metab Res Rev 30:96–102PubMedPubMedCentralCrossRefGoogle Scholar
  67. 67.
    Canalis E, Mazziotti G, Giustina A, Bilezikian JP (2007) Glucocorticoid-induced osteoporosis: pathophysiology and therapy. Osteoporos Int 18:1319–1328PubMedCrossRefGoogle Scholar
  68. 68.
    van Raalte DH, Nofrate V, Bunck MC, van Iersel T, Schaap JE, Nassander UK, Heine RJ, Mari A, Dokter WH, Diamant M (2010) Acute and 2-week exposure to prednisolone impair different aspects of beta-cell function in healthy men. Eur J Endocrinol 162:729–735PubMedCrossRefGoogle Scholar
  69. 69.
    Collins SL, Moore RA, McQuay hj, Wiffen P (2000) Antidepressants and anticonvulsants for diabetic neuropathy and postherpetic neuralgia: a quantitative systematic review. J Pain Symptom Manage 20:449–458PubMedCrossRefGoogle Scholar
  70. 70.
    Carbone LD, Johnson KC, Robbins J, Larson JC, Curb JD, Watson K, Gass M, Lacroix AZ (2010) Antiepileptic drug use, falls, fractures, and BMD in postmenopausal women: findings from the women’s health initiative (WHI). J Bone Miner Res 25:873–881PubMedGoogle Scholar
  71. 71.
    Meier C, Kraenzlin ME (2011) Antiepileptics and bone health. Ther Adv Musculoskelet Dis 3:235–243PubMedPubMedCentralCrossRefGoogle Scholar
  72. 72.
    Starup-Linde J, Eriksen SA, Lykkeboe S, Handberg A, Vestergaard P (2014) Biochemical markers of bone turnover in diabetes patients—a meta-analysis, and a methodological study on the effects of glucose on bone markers. Osteoporos Int 25:1697–1708PubMedCrossRefGoogle Scholar
  73. 73.
    Sarles H (1992) Chronic pancreatitis and diabetes. Baillieres Clin Endocrinol Metab 6:745–775PubMedCrossRefGoogle Scholar
  74. 74.
    Hardt PD, Brendel MD, Kloer HU, Bretzel RG (2008) Is pancreatic diabetes (type 3c diabetes) underdiagnosed and misdiagnosed? Diabetes Care 31(Suppl 2):S165–S169PubMedCrossRefGoogle Scholar
  75. 75.
    Ewald N, Kaufmann C, Raspe A, Kloer HU, Bretzel RG, Hardt PD (2012) Prevalence of diabetes mellitus secondary to pancreatic diseases (type 3c). Diabetes Metab Res Rev 28:338–342PubMedCrossRefGoogle Scholar
  76. 76.
    Munigala S, Agarwal B, Gelrud A, Conwell DL (2015) Chronic pancreatitis and fracture: a retrospective, population-based Veterans Administration Study. Pancreas 45(3):355–361CrossRefGoogle Scholar
  77. 77.
    De Laet C, Kanis JA, Oden A, Johanson H, Johnell O, Delmas P, Eisman JA, Kroger H, Fujiwara S, Garnero P, McCloskey EV, Mellstrom D, Melton LJ III, Meunier PJ, Pols HA, Reeve J, Silman A, Tenenhouse A (2005) Body mass index as a predictor of fracture risk: a meta-analysis. Osteoporos Int 16:1330–1338PubMedCrossRefGoogle Scholar
  78. 78.
    Tang X, Liu G, Kang J, Hou Y, Jiang F, Yuan W, Shi J (2013) Obesity and risk of hip fracture in adults: a meta-analysis of prospective cohort studies. PLoS One 8:e55077PubMedPubMedCentralCrossRefGoogle Scholar
  79. 79.
    Santesso N, Carrasco-Labra A, Brignardello-Petersen R (2014) Hip protectors for preventing hip fractures in older people. Cochrane Database Syst Rev 3:CD001255Google Scholar
  80. 80.
    Kannus P, Parkkari J, Niemi S, Pasanen M, Palvanen M, Jarvinen M, Vuori I (2000) Prevention of hip fracture in elderly people with use of a hip protector. N Engl J Med 343:1506–1513PubMedCrossRefGoogle Scholar
  81. 81.
    Johansson H, Kanis JA, Oden A, McCloskey E, Chapurlat RD, Christiansen C, Cummings SR, Diez-Perez A, Eisman JA, Fujiwara S, Gluer CC, Goltzman D, Hans D, Khaw KT, Krieg MA, Kroger H, LaCroix AZ, Lau E, Leslie WD, Mellstrom D, Melton LJ III, O’Neill TW, Pasco JA, Prior JC, Reid DM, Rivadeneira F, van Staa T, Yoshimura N, Zillikens MC (2014) A meta-analysis of the association of fracture risk and body mass index in women. J Bone Miner Res 29:223–233PubMedCrossRefGoogle Scholar
  82. 82.
    Esser N, Legrand-Poels S, Piette J, Scheen AJ, Paquot N (2014) Inflammation as a link between obesity, metabolic syndrome and type 2 diabetes. Diabetes Res Clin Pract 105:141–150PubMedCrossRefGoogle Scholar
  83. 83.
    Lecka-Czernik B, Rosen CJ, Kawai M (2010) Skeletal aging and the adipocyte program: new insights from an “old” molecule. Cell Cycle 9:3648–3654PubMedPubMedCentralCrossRefGoogle Scholar
  84. 84.
    Gimble JM, Zvonic S, Floyd ZE, Kassem M, Nuttall ME (2006) Playing with bone and fat. J Cell Biochem 98:251–266PubMedCrossRefGoogle Scholar
  85. 85.
    Witek PR, Witek J, Pankowska E (2012) Type 1 diabetes-associated autoimmune diseases: screening, diagnostic principles and management. Med Wieku Rozwoj 16:23–34PubMedGoogle Scholar
  86. 86.
    Vestergaard P, Mosekilde L (2002) Fractures in patients with hyperthyroidism and hypothyroidism: a nationwide follow-up study in 16,249 patients. Thyroid 12:411–419PubMedCrossRefGoogle Scholar
  87. 87.
    Compston J (2003) Is fracture risk increased in patients with coeliac disease? Gut 52:459–460PubMedPubMedCentralCrossRefGoogle Scholar
  88. 88.
    Marshall D, Johnell O, Wedel H (1996) Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 312:1254–1259PubMedPubMedCentralCrossRefGoogle Scholar
  89. 89.
    Blake GM, Fogelman I (2007) The role of DXA bone density scans in the diagnosis and treatment of osteoporosis. Postgrad Med J 83:509–517PubMedPubMedCentralCrossRefGoogle Scholar
  90. 90.
    Ma L, Oei L, Jiang L, Estrada K, Chen H, Wang Z, Yu Q, Zillikens MC, Gao X, Rivadeneira F (2012) Association between bone mineral density and type 2 diabetes mellitus: a meta-analysis of observational studies. Eur J Epidemiol 27:319–332PubMedPubMedCentralCrossRefGoogle Scholar
  91. 91.
    Saito M, Kida Y, Kato S, Marumo K (2014) Diabetes, collagen, and bone quality. Curr Osteoporosis Rep 12:181–188CrossRefGoogle Scholar
  92. 92.
    Schwartz AV, Vittinghoff E, Bauer DC, Hillier TA, Strotmeyer ES, Ensrud KE, Donaldson MG, Cauley JA, Harris TB, Koster A, Womack CR, Palermo L, Black DM, Study of Osteoporotic Fractures (SOF) Research Group, Osteoporotic Fractures in Men (MrOS) Research Group, Health, Aging, and Body Composition (Health ABC) Research Group (2011) Association of BMD and FRAX score with risk of fracture in older adults with type 2 diabetes. JAMA 305:2184–2192PubMedPubMedCentralCrossRefGoogle Scholar
  93. 93.
    Schwartz AV (2016) Epidemiology of fractures in type 2 diabetes. Bone 82:2–8PubMedCrossRefGoogle Scholar
  94. 94.
    Kanis JA, McCloskey EV, Johansson H, Oden A, Strom O, Borgstrom F (2010) Development and use of FRAX in osteoporosis. Osteoporos Int 21(Suppl 2):S407–S413PubMedCrossRefGoogle Scholar
  95. 95.
    Johnston SS, Conner C, Aagren M, Ruiz K, Bouchard J (2012) Association between hypoglycaemic events and fall-related fractures in Medicare-covered patients with type 2 diabetes. Diabetes Obes Metab 14:634–643PubMedCrossRefGoogle Scholar
  96. 96.
    Schwartz AV, Margolis KL, Sellmeyer DE, Vittinghoff E, Ambrosius WT, Bonds DE, Josse RG, Schnall AM, Simmons DL, Hue TF, Palermo L, Hamilton BP, Green JB, Atkinson HH, O’Connor PJ, Force RW, Bauer DC (2012) Intensive glycemic control is not associated with fractures or falls in the ACCORD randomized trial. Diabetes Care 35:1525–1531PubMedPubMedCentralCrossRefGoogle Scholar
  97. 97.
    World Health Organisation (2014) Global status report on noncommunicable diseases. World Health Organisation, GenevaGoogle Scholar
  98. 98.
    Herrera A, Mateo J, Gil-Albarova J, Lobo-Escolar A, Artigas JM, Lopez-Prats F, Mesa M, Ibarz E, Gracia L (2015) Prevalence of osteoporotic vertebral fracture in Spanish women over age 45. Maturitas 80:288–295PubMedCrossRefGoogle Scholar
  99. 99.
    Kiyohara N, Yamamoto M, Sugimoto T (2015) Discordance between prevalent vertebral fracture and vertebral strength estimated by the finite element method based on quantitative computed tomography in patients with type 2 diabetes mellitus. PLoS One 10:e0144496PubMedPubMedCentralCrossRefGoogle Scholar
  100. 100.
    Waterloo S, Ahmed LA, Center JR, Eisman JA, Morseth B, Nguyen ND, Nguyen T, Sogaard AJ, Emaus N (2012) Prevalence of vertebral fractures in women and men in the population-based Tromso Study. BMC Musculoskelet Disord 13:3-2474-13-3Google Scholar
  101. 101.
    Napoli N, Schwartz AV, Palermo L, Jin JJ, Wustrack R, Cauley JA, Ensrud KE, Kelly M, Black DM (2013) Risk factors for subtrochanteric and diaphyseal fractures: the study of osteoporotic fractures. J Clin Endocrinol Metab 98:659–667PubMedPubMedCentralCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  • Jakob Starup-Linde
    • 1
    • 2
  • Morten Frost
    • 3
  • Peter Vestergaard
    • 2
    • 4
  • Bo Abrahamsen
    • 5
    • 6
    Email author
  1. 1.Department of Endocrinology and Internal MedicineAarhus University HospitalAarhus CDenmark
  2. 2.Department of Clinical MedicineAalborg University HospitalAalborgDenmark
  3. 3.Department of EndocrinologyOdense University HospitalOdense CDenmark
  4. 4.Department of EndocrinologyAalborg University HospitalAalborgDenmark
  5. 5.Department of MedicineHolbæk HospitalHolbækDenmark
  6. 6.Odense Patient Data Explorative Network (OPEN), Institute of Clinical ResearchUniversity of Southern DenmarkOdense CDenmark

Personalised recommendations