SEARCH

SEARCH BY CITATION

Keywords:

  • TYPE 2 DIABETES;
  • MICROINDENTATION TESTING;
  • BONE MATERIAL STRENGTH;
  • HRPQCT;
  • BONE TURNOVER

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

Although patients with type 2 diabetes (T2D) are at significant risk for well-recognized diabetic complications, including macrovascular disease, retinopathy, nephropathy, and neuropathy, it is also clear that T2D patients are at increased risk for fragility fractures. Furthermore, fragility fractures in patients with T2D occur at higher bone mineral density (BMD) values compared to nondiabetic controls, suggesting abnormalities in bone material strength (BMS) and/or bone microarchitecture (bone “quality”). Thus, we performed in vivo microindentation testing of the tibia to directly measure BMS in 60 postmenopausal women (age range, 50–80 years) including 30 patients diagnosed with T2D for >10 years and 30 age-matched, nondiabetic controls. Regional BMD was measured by dual-energy X-ray absorptiometry (DXA); cortical and trabecular bone microarchitecture was assessed from high-resolution peripheral quantitative computed tomography (HRpQCT) images of the distal radius and tibia. Compared to controls, T2D patients had significantly lower BMS: unadjusted (−11.7%; p < 0.001); following adjustment for body mass index (BMI) (−10.5%; p < 0.001); and following additional adjustment for age, hypertension, nephropathy, neuropathy, retinopathy, and vascular disease (−9.2%; p = 0.022). By contrast, after adjustment for confounding by BMI, T2D patients had bone microarchitecture and BMD that were not significantly different than controls; however, radial cortical porosity tended to be higher in the T2D patients. In addition, patients with T2D had significantly reduced serum markers of bone turnover (all p < 0.001) compared to controls. Of note, in patients with T2D, the average glycated hemoglobin level over the previous 10 years was negatively correlated with BMS (r = −0.41; p = 0.026). In conclusion, these findings represent the first demonstration of compromised BMS in patients with T2D. Furthermore, our results confirm previous studies demonstrating low bone turnover in patients with T2D and highlight the potential detrimental effects of prolonged hyperglycemia on bone quality. Thus, the skeleton needs to be recognized as another important target tissue subject to diabetic complications. © 2014 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

Type 2 diabetes (T2D) is one of the most common chronic diseases worldwide.[1, 2] As changing lifestyles lead to increased obesity,[3] the prevalence of T2D will continue to grow and the economic public health burden will worsen significantly.[1, 2] Indeed, the direct medical costs of T2D are estimated at over $116 billion annually in the United States alone.[1] Globally, 285 million people have T2D, and this number is predicted to increase to 439 million by 2030.[2] Although patients with T2D are at significant risk for premature mortality and morbidity from macrovascular disease, retinopathy, nephropathy, and neuropathy,[2] there is growing evidence that T2D is also an independent risk factor for fragility fractures at skeletal sites such as the hip, spine, and distal forearm.[4-8] These findings are perhaps surprising given that patients with T2D often have normal or increased dual-energy X-ray absorptiometry (DXA)-derived bone mineral density (BMD),[9] even when normalized for body mass index (BMI).[10]

Recently, however, using data from three large prospective observational studies in the United States (Study of Osteoporotic Fractures [SOF]; Osteoporotic Fractures in Men Study [MrOS]; and Health, Aging, and Body Composition [ABC] Study), Schwartz and colleagues[11] demonstrated that patients with T2D have a higher fracture risk for a given femoral neck BMD T-score and age or for a given FRAX probability (defined by the World Health Organization's Fracture Risk Algorithm [FRAX] score[12]). These results have since been corroborated by data from a large clinical database in Canada.[13] Collectively, these findings indicate that BMD and FRAX underestimate fracture risk in T2D patients and suggest that other factors are responsible for this increased fracture risk.[14]

Although DXA is commonly used in fracture risk assessment, it provides little information about bone microarchitecture and material composition—the primary determinants of bone “quality.”[15, 16] Fortunately, quantification of bone microarchitecture has become possible with the advent of high-resolution peripheral quantitative computed tomography (HRpQCT), although this technique cannot measure bone material strength (BMS), an often neglected, but nonetheless important, component of bone quality.[15, 16] Progress in understanding how BMS might be changed in T2D has been hampered by the invasive nature of the direct measures previously needed to quantify this property in vivo, which has rendered routine clinical assessments impractical. However, recent advances in microindentation technology have now made it safe to quantify BMS in humans with minimal discomfort.[17, 18] Indeed, previous studies have used bone microindentation testing to show that bone material properties are worse in hip fracture[19] and atypical femoral fracture[20] patients. However, whether patients with T2D have deficits in BMS as compared to nondiabetic controls is not known.

Therefore, because the underlying skeletal abnormalities in T2D have not been identified, we performed in vivo microindentation testing of the tibia to directly measure BMS in 60 postmenopausal women (age range, 50–80 years) including 30 patients diagnosed with T2D for >10 years and 30 age-matched, nondiabetic controls. In addition, we measured regional BMD by DXA and bone microarchitecture of the distal radius and tibia by HRpQCT. Thus, the aims of this study were to: (1) determine whether BMS, bone imaging parameters (derived from DXA and HRpQCT), and/or bone turnover are altered in patients with T2D compared to age-matched, nondiabetic controls; and (2) examine the associations of BMS with duration of T2D and circulating levels of glycated hemoglobin.

Subjects and Methods

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

Subjects

This cross-sectional study was approved by the Mayo Clinic Institutional Review Board, and informed written consent was obtained from all participants. We recruited 60 normal postmenopausal women (serum follicle stimulating hormone [FSH] >20 IU/L), including 30 patients diagnosed with T2D for >10 years and 30 age-matched, nondiabetic controls, by letters from an age-stratified (50–80 years) random sample of Olmsted County, MN, USA, residents, augmented by newspaper and website advertisements. All potential subjects were rigorously screened for coexisting disease and excluded if they had low body stores of vitamin D (serum 25-hydroxyvitamin D of <20 ng/mL), any fractures within the previous 6 months (because fracture may alter bone turnover in this time period[21]), or any disorders associated with altered skeletal structure or function, including presence of chronic renal impairment (chronic kidney disease [CKD] stage IV or V), chronic liver disease, severe neuropathic disease, unstable cardiovascular disease, malignancy, chronic gastrointestinal disease, hypoparathyroidism or hyperparathyroidism, acromegaly, Cushing's syndrome, hypopituitarism, severe chronic obstructive pulmonary disease, alcoholism, or type 1 diabetes. In addition, subjects with a history of traumatic fractures (eg, falls >3 m, motor vehicle accidents, etc.) or pathological fractures (eg, due to Paget's disease, myeloma, metastatic malignancy) were excluded. Further, subjects were excluded if undergoing treatment for blood clots or coagulation defects, or treatment with any of the following drugs: corticosteroids (>3 months at any time or >10 days within the previous year), anticonvulsant therapy (within the previous year), pharmacological doses of thyroid hormone (causing thyroid stimulating hormone [TSH] to decline below normal), adrenal or anabolic steroids, aromatase inhibitors, calcitonin, calcium supplementation >1200 mg/d (within the preceding 3 months), bisphosphonates, estrogen or selective estrogen receptor modulator (SERM) (within the past year), parathyroid hormone (PTH), sodium fluoride, teriparatide, or thiazolidinediones (TZDs).

We also ascertained a complete list of current and past medications, including the preparation, duration, and dose of diabetic medications (eg, biguanides, insulin, dipeptidyl peptidase-4 [DPP-4] inhibitors, sulfonylureas, α-glucosidase inhibitors, glucagon-like peptide analogs and agonists, meglitinides, and sodium glucose cotransporter-2 [SGLT-2] inhibitors), as well as β-blockers, angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers (ARBs), calcium channel blockers, statins, thiazides, proton pump inhibitors, and selective serotonin reuptake inhibitors (SSRIs), because these drugs may affect bone turnover. A summary of these medications is provided in Supplementary Table 1. All patients with T2D were on at least one oral diabetic medication; thus, none were being treated with diet alone. Clinical details in the medical records were reviewed to determine if subjects met study criteria.

Between February 2012 and March 2013, fasting morning blood was drawn from potential candidates in the outpatient Clinical Research Unit (CRU) at the Mayo Clinic (Rochester, MN, USA) and serum was analyzed for determination of vitamin D (total 25-hydroxyvitamin D), calcium, phosphorus, alkaline phosphatase, aspartate transaminase, creatinine, TSH, and FSH. Inclusion criteria required that all subjects were within the normal reference range for each of these parameters, except for FSH; menopause was defined as a serum FSH level >20 IU/L and no menses for >1 year. In addition, we measured serum glycated hemoglobin; inclusion criteria required that controls had a normal glycated hemoglobin level (4.0% to 5.9%), whereas patients were required to be diagnosed with T2D for >10 years according to the American Diabetes Association criteria[22] (ie, glycated hemoglobin level ≥6.5%). In the patients with T2D, we also ascertained the average glycated hemoglobin level over the previous 10 years from clinical records for use as an indicator of past glycemic control. All subjects were residents of Olmsted County; this community is highly characteristic of the U.S. white population but underrepresented with respect to persons of African or Asian ancestry.[23] Reflecting the ethnic composition of Olmsted County, 98% of the sample was white.

Study protocol

All subjects were interviewed for fracture, medical, and medication history; these data were supplemented by medical records and cross-checked with redundant sources (radiographic reports, primary care or other health care provider notes, etc.) for accuracy. In addition, smoking and alcohol habits were ascertained during the interview. Anthropometric data were collected on all subjects wearing light-weight clothing and no shoes. Weight was obtained using an electronic scale (Model 5002; Tronic, Inc., White Plains, NY, USA) and height was measured using a customized stadiometer (Mayo Section of Engineering). We assessed BMS (primary end point) from in vivo microindentation testing at the midshaft of the nondominant anterior tibia as detailed below (power calculations are provided in the Supplementary Materials). Secondary endpoints included the DXA-derived and HRpQCT-derived bone parameters, as well as the serum biochemical markers of bone turnover. BMD of the hip, radius, lumbar spine (L1–L4), and total body regions was measured by DXA; cortical and trabecular bone microarchitecture was assessed from HRpQCT images of the nondominant distal radius and tibia, but data from two radius scans (0 T2D; 2 control) were excluded because of motion artifact. Our research team has extensive experience using DXA and HRpQCT;[24-26] these methods are detailed in the Supplementary Materials.

Bone microindentation testing

The OsteoProbe Reference Point Indenter (Active Life Scientific Inc., Santa Barbara, CA, USA) is a handheld microindentation instrument designed for in vivo BMS measurements.[17, 18] Supplementary Fig. 1 shows the device positioned over the testing site. Primary components of the device include an impact mechanism, a displacement transducer, and a sterilized stainless steel disposable probe with a 90-degree conical tip (375 µm diameter; <10 µm tip sharpness radius). Additional details regarding the device specifications have been published previously[17, 18] and are summarized in the Supplementary Materials.

The testing site (midshaft of the nondominant anterior tibia) was determined by calculating the midpoint from the proximal end of the medial border of the tibial plateau to the distal edge of the medial malleolus. After administrating local anesthesia (1% lidocaine), the probe was inserted through the soft tissue and periosteum until residing on the bone surface. While keeping the device perpendicular to the bone surface (within 10 degrees), the measurement was actuated by slowly compressing the device's outer housing unit, compressing the internal primary spring until the trigger mechanism initiated an impact (Supplementary Fig. 2). The impact mechanism creates a force to drive the probe into bone, while the displacement transducer measures indentation distance increase (IDI, µm) from impact. The IDI from the impact was converted by a computer to BMS, defined as 100 times the ratio of the harmonic mean IDI from five separate impacts into a polymethylmethacrylate plastic calibration phantom relative to the IDI from the impact into the bone.[17] For each subject, BMS was calculated as the average of 10 measurements at different midshaft tibial sites (separated by >2 mm).

The indentations are small (mean IDI from impact ± SD, 187 ± 24.0 µm; range, 154–323 µm), yet large enough to create cracks (ie, microfractures) in the bone.[17, 18] Thus, BMS is a direct measure of fracture resistance insofar as the farther the probe indents the bone (higher the IDI from impact), the more easily the bone is fractured (lower the BMS). This technology is well validated[17, 18] and a similar device has been used in previous studies involving animals[27-29] and humans.[19, 20] For example, in a study of human cadaveric bone specimens, IDI from impact was inversely correlated with crack growth toughness (r = −0.90), and scanning electron microscope images of cracks induced by microindentation and by experimental fractures were virtually identical.[19] In our laboratory, the within-subject (n = 10 controls; mean age ± SD: 63.1 ± 8.9 years; range, 50–76 years) precision error (coefficient of variation [CV]) of the OsteoProbe was 1.65% for BMS. It should be noted that patients who have a significant skin disorder, bruising, local edema, or infection, as well as those undergoing treatment for blood clots or severe coagulation defects should not be exposed to this procedure. Based on our experience with 60 human subjects (30 patients with T2D and 30 nondiabetic controls), the procedure causes minimal discomfort (only during the local anesthesia injection) and no complications have been observed. Thus, this procedure poses nonsignificant risk to humans.

Serum bone turnover measurements

Fasting morning blood was drawn and stored at −80°C. Bone formation was assessed by serum amino-terminal propeptide of type I collagen (P1NP; µg/L) as measured by RIA (interassay CV, <10%; Immunodiagnostic Systems, Fountain Hills, AZ, USA). Bone resorption was evaluated by serum cross-linked C-telopeptide of type I collagen (CTx; ng/mL) as measured by ELISA (interassay CV, <8%; Roche Diagnostics, Indianapolis, IN, USA).

Statistical analysis

Data are expressed as mean ± SEM unless otherwise specified. All variables were tested for skewness and kurtosis; plots and regression models were used to check the data for normality, linearity, outliers, and potential influential observations. Because all variables satisfied the requirements for parametric statistics, transformations were not performed. Demographic and clinical characteristics as well as anthropometric, biochemical, and bone parameters were compared between the T2D and control groups using two-sample t tests and χ2 tests as appropriate. Further comparisons of bone parameters were made using an analysis of variance model adjusted for BMI, and then subsequently for additional covariates (ie, age, hypertension, nephropathy, neuropathy, retinopathy, and vascular disease). Associations of duration of T2D and glycated hemoglobin level with BMS were examined using Spearman correlations, which are appropriate for smaller sample sizes and more robust to potential outliers. Testing was performed at a significance level of p < 0.05 (two-tailed).

Results

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

In the patients, duration since diagnosis of T2D ranged from 10 to 34 years (mean age at diagnosis ± SD: 48.4 ± 7.9 years). Given age-matching, the T2D and control groups were virtually identical in age (Table 1). Further, as evident in Table 1, the groups were similar in smoking and alcohol habits, and had comparable fracture histories (3 subjects per group with a prior fracture). In addition, height did not differ between the groups, but the patients with T2D were heavier and had significantly higher BMIs than controls (Table 1). As expected, glycated hemoglobin levels were significantly (p < 0.001) higher in patients with T2D compared to controls. Further, relative to controls, bone formation (P1NP; p < 0.001) and bone resorption (CTx; p < 0.001) were significantly lower in patients with T2D (Table 1). By contrast, 25-hydroxyvitamin D levels did not differ between the groups (Table 1). Last, use of proton pump inhibitors and SSRIs did not differ between the T2D and control groups, whereas use of drugs commonly prescribed to treat diabetic complications (eg, β-blockers, ACE inhibitors, ARBs, calcium channel blockers, statins, and thiazides) tended to be higher in patients with T2D (Supplementary Table 1).

Table 1. Characteristics of the Patients With T2D and Age-Matched, Nondiabetic Controls
CharacteristicT2D (n = 30)Control (n = 30)pa
  1. Values are shown as means ± SEM unless otherwise noted. Statistically significant p values are shown in bold.

  2. T2D = type 2 diabetes; NA = not applicable; P1NP = amino-terminal propeptide of type I collagen; CTx = cross-linked C-telopeptide of type I collagen.

  3. a

    p values are shown for comparisons between the T2D and control groups using two-sample t tests and χ2 tests as appropriate.

  4. b

    The presence of hypertension, nephropathy, neuropathy, retinopathy, or vascular disease was determined by clinical records.

Demographic
Age (years)65.4 ± 1.465.7 ± 1.60.906
Duration of T2D (years)17.0 ± 1.1NANA
Current smoker, n (%)3 (10.0)2 (6.7)0.640
Past smoker, n (%)9 (30.0)10 (33.3)0.781
Alcohol servings, units/d0.78 ± 0.090.77 ± 0.100.988
Previous fracture, n (%)3 (10.0)3 (10.0)1.000
Coexisting conditionsb
Hypertension, n (%)30 (100.0)10 (33.3)<0.001
Nephropathy, n (%)5 (16.7)0 (0.0)NA
Neuropathy, n (%)9 (30.0)0 (0.0)NA
Retinopathy, n (%)6 (20.0)0 (0.0)NA
Vascular disease, n (%)5 (16.7)0 (0.0)NA
Anthropometry
Height (cm)162 ± 1.2164 ± 1.50.188
Weight (kg)85.3 ± 2.476.0 ± 3.00.021
Body mass index (kg/m2)32.6 ± 0.828.1 ± 1.00.001
Biochemical
Glycated hemoglobin level, screen visit (%)7.7 ± 0.25.4 ± 0.1<0.001
Glycated hemoglobin level, 10-year average (%)7.4 ± 0.1NANA
P1NP (µg/L)34.3 ± 2.748.6 ± 2.7<0.001
CTx (ng/mL)0.305 ± 0.0280.463 ± 0.028<0.001
Total 25-hydroxyvitamin D (ng/mL)34.0 ± 1.833.1 ± 1.50.719

Compared to controls, T2D patients had significantly lower BMS: unadjusted (−11.7%; p < 0.001; Fig. 1A) and following adjustment for BMI (−10.5%; p < 0.001; Fig. 1B). Furthermore, in unadjusted analyses, T2D patients tended to have higher regional BMD, had thicker radial and tibial cortices, and tended to have more advantageous trabecular microarchitecture compared to controls, although these differences were not statistically significant following adjustment for confounding by BMI (Table 2). Nevertheless, after adjustment for BMI, radial cortical porosity tended to be higher in T2D patients compared to controls. Adjustment for additional covariates (age, hypertension, nephropathy, neuropathy, retinopathy, vascular disease) resulted in similar findings; of note, the differences in BMS between the T2D patients and controls remained significant even following these additional adjustments (−9.2%, p = 0.022; Table 3).

image

Figure 1. Unadjusted (A) and BMI-adjusted (B) comparisons of BMS between patients with T2D and age-matched, nondiabetic controls. Values are shown as mean ± SEM. ‡p < 0.001. BMI = body mass index; BMS = bone material strength; T2D = type 2 diabetes.

Download figure to PowerPoint

Table 2. BMS, Regional BMD by DXA, and HRpQCT-Derived Bone Parameters at the Distal Radius and Tibia in Patients With T2D and Age-Matched, Nondiabetic Controls, Unadjusted and Adjusted for BMI
Bone parameterT2D (n = 30)Control (n = 30)Difference (%)papb
  1. Values are presented as mean ± SEM or percent difference adjusted for BMI. Statistically significant p values are shown in bold.

  2. BMS = bone material strength; BMD = bone mineral density; DXA = dual-energy X-ray absorptiometry; HRpQCT = high-resolution peripheral quantitative computed tomography; T2D = type 2 diabetes; BMI = body mass index.

  3. a

    Unadjusted.

  4. b

    Adjusted for body mass index.

In vivo microindentation testing
BMS77.2 ± 1.685.7 ± 1.6−10.5%<0.001<0.001
Distal radius parameters (HRpQCT)
Cortical porosity (%)2.95 ± 0.372.15 ± 0.3831.5%0.3440.157
Cortical pore volume (mm3)14.7 ± 1.910.0 ± 1.937.8%0.1670.108
Cortical volumetric BMD (mg/cm3)919 ± 10.3926 ± 10.7−0.8%0.8260.637
Cortical thickness (mm)0.95 ± 0.030.90 ± 0.035.9%0.0240.237
Endocortical circumference (mm)49.7 ± 1.149.6 ± 1.10.0%0.4150.988
Periosteal circumference (mm)68.6 ± 1.168.2 ± 1.10.7%0.6380.792
Trabecular bone volume fraction0.134 ± 0.0060.129 ± 0.0063.8%0.1800.619
Trabecular number (1/mm)1.89 ± 0.051.80 ± 0.055.0%0.0290.240
Trabecular thickness (mm)0.070 ± 0.0020.072 ± 0.002−2.8%0.9600.695
Trabecular separation (mm)0.47 ± 0.020.50 ± 0.02−5.6%0.0270.296
Distal tibia parameters (HRpQCT)
Cortical porosity (%)6.9 ± 0.67.1 ± 0.6−3.8%0.9820.750
Cortical pore volume (mm3)66.5 ± 5.065.1 ± 5.02.1%0.1970.850
Cortical volumetric BMD (mg/cm3)845 ± 13.5832 ± 13.51.6%0.3980.506
Cortical thickness (mm)1.21 ± 0.041.13 ± 0.046.6%0.0080.177
Endocortical circumference (mm)84.2 ± 1.786.3 ± 1.7−2.4%0.2680.424
Periosteal circumference (mm)104 ± 1.7106 ± 1.7−1.4%0.6580.568
Trabecular bone volume fraction0.144 ± 0.0050.138 ± 0.0054.3%0.0500.398
Trabecular number (1/mm)1.80 ± 0.051.84 ± 0.05−2.0%0.1870.644
Trabecular thickness (mm)0.080 ± 0.0020.076 ± 0.0025.1%0.4450.265
Trabecular separation (mm)0.49 ± 0.020.49 ± 0.02−1.0%0.0620.864
Regional BMD (DXA)
Femoral neck hip (g/cm2)0.93 ± 0.030.94 ± 0.03−0.5%0.1770.900
Trochanter hip (g/cm2)0.85 ± 0.020.81 ± 0.024.6%0.0070.264
Total hip (g/cm2)1.03 ± 0.020.98 ± 0.024.2%0.0070.247
Ultradistal radius (g/cm2)0.43 ± 0.010.41 ± 0.014.1%0.0100.295
Total radius (g/cm2)0.64 ± 0.010.62 ± 0.013.2%0.0050.286
Lumbar spine (L1–L4) (g/cm2)1.27 ± 0.041.15 ± 0.049.4%0.0030.057
Total body (g/cm2)1.16 ± 0.021.14 ± 0.021.4%0.0250.633
Table 3. BMS, Regional BMD by DXA, and HRpQCT-Derived Bone Parameters at the Distal Radius and Tibia in Patients With T2D and Age-Matched, Nondiabetic Controls, Adjusted for Potential Confounding Variables
Bone parametersT2D (n = 30)Control (n = 30)Difference (%)papb
  1. Values are presented as mean ± SEM or percent difference adjusted for BMI, age, hypertension, nephropathy, neuropathy, retinopathy, and vascular disease. Statistically significant p values are shown in bold.

  2. BMS = bone material strength; BMD = bone mineral density; DXA = dual-energy X-ray absorptiometry; HRpQCT = high-resolution peripheral quantitative computed tomography; T2D = type 2 diabetes; BMI = body mass index.

  3. a

    Adjusted for BMI, age, and hypertension.

  4. b

    Adjusted for BMI, age, hypertension, nephropathy, neuropathy, retinopathy, and vascular disease.

In vivo microindentation testing
BMS77.7 ± 1.985.2 ± 1.9−9.2%0.0050.022
Distal radius parameters (HRpQCT)
Cortical porosity (%)2.84 ± 0.432.27 ± 0.4522.3%0.2460.458
Cortical pore volume (mm3)14.6 ± 2.310.0 ± 2.437.3%0.1420.256
Cortical volumetric BMD (mg/cm3)921 ± 13.0923 ± 13.7−0.2%0.7890.929
Cortical thickness (mm)0.95 ± 0.040.90 ± 0.045.1%0.4910.492
Endocortical circumference (mm)50.9 ± 1.448.4 ± 1.55.1%0.2970.324
Periosteal circumference (mm)69.7 ± 1.467.0 ± 1.54.0%0.2680.295
Trabecular bone volume fraction0.129 ± 0.0070.135 ± 0.008−4.5%0.6140.641
Trabecular number (1/mm)1.84 ± 0.061.85 ± 0.07−0.4%0.7570.953
Trabecular thickness (mm)0.069 ± 0.0030.073 ± 0.003−5.6%0.5440.434
Trabecular separation (mm)0.49 ± 0.020.48 ± 0.022.5%0.6060.747
Distal tibia parameters (HRpQCT)
Cortical porosity (%)6.5 ± 0.77.5 ± 0.7−13.2%0.7340.425
Cortical pore volume (mm3)63.5 ± 6.368.1 ± 6.3−7.0%0.9870.669
Cortical volumetric BMD (mg/cm3)849 ± 15.8827 ± 15.82.6%0.8930.416
Cortical thickness (mm)1.17 ± 0.051.16 ± 0.051.0%0.7700.880
Endocortical circumference (mm)86.2 ± 2.184.3 ± 2.12.2%0.2320.598
Periosteal circumference (mm)106 ± 2.1104 ± 2.11.8%0.2320.594
Trabecular bone volume fraction0.140 ± 0.0060.142 ± 0.006−1.4%0.9710.849
Trabecular number (1/mm)1.77 ± 0.071.86 ± 0.07−5.0%0.6650.439
Trabecular thickness (mm)0.079 ± 0.0030.078 ± 0.0031.3%0.9780.868
Trabecular separation (mm)0.50 ± 0.020.48 ± 0.024.3%0.8540.606
Regional BMD (DXA)
Femoral neck hip (g/cm2)0.91 ± 0.030.96 ± 0.03−6.1%0.2110.315
Trochanter hip (g/cm2)0.83 ± 0.030.83 ± 0.03−0.5%0.6980.937
Total hip (g/cm2)1.00 ± 0.031.01 ± 0.03−0.9%0.5980.877
Ultradistal radius (g/cm2)0.43 ± 0.010.41 ± 0.016.4%0.5350.254
Total radius (g/cm2)0.64 ± 0.010.62 ± 0.014.1%0.5790.295
Lumbar spine (L1–L4; g/cm2)1.25 ± 0.051.16 ± 0.057.5%0.3200.304
Total body (g/cm2)1.16 ± 0.031.13 ± 0.032.8%0.9530.520

We next used Spearman correlations to examine associations of BMS with duration of T2D and glycated hemoglobin level. Duration of T2D was not significantly correlated with BMS (r = −0.09, p = 0.732). In patients with T2D, the average glycated hemoglobin level over the previous 10 years was negatively correlated with BMS (r = −0.41; p = 0.026). However, the single screen visit glycated hemoglobin level was not related to BMS in either the T2D patients (r = −0.19; p = 0.306) or in the control subjects (r = −0.09; p = 0.630).

Discussion

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

Despite their increased risk of fragility fractures (eg, hip, spine, and distal forearm),[4-8] we found that postmenopausal women with T2D have bone density (derived by DXA) that is similar to age-matched, nondiabetic women after adjustment for confounding by BMI. Further, although radial cortical porosity tended to be higher in patients with T2D, we found that bone microarchitectural parameters at the distal radius and tibia (derived by HRpQCT) did not differ significantly between women with and without T2D. By contrast, patients with T2D did have significantly reduced BMS compared to nondiabetic controls. Our findings therefore raise the possibility that fragility fractures in patients with T2D are due principally to compromised BMS, and point to the need for further studies comparing BMS in T2D patients with and without fragility fractures.

The recent evidence suggesting that fracture risk is higher for a given femoral neck BMD T-score and age or for a given FRAX probability in patients with T2D compared to nondiabetic controls[11, 13] has led to the suggestion that fragility fractures in T2D may result from underlying skeletal abnormalities (ie, altered microarchitectural and/or bone material properties).[14] To test this hypothesis, several previous studies have used HRpQCT to assess bone microarchitecture in patients with T2D compared to nondiabetic subjects. For example, Burghardt and colleagues[30] found that 19 postmenopausal women with T2D had more advantageous trabecular microarchitecture, but significantly compromised cortical microarchitecture (ie, higher cortical porosity), at the distal radius, and similar changes at the distal tibia, compared to 19 nondiabetic controls. Shu and colleagues,[31] however, found no significant differences in trabecular or cortical bone parameters at the radius or tibia in postmenopausal women with and without T2D, although the subset of subjects who underwent HRpQCT scanning was relatively small (14 subjects per group) and cortical porosity was not reported. More recently, Patsch and colleagues[32] found significantly higher cortical porosity at the distal radius in T2D patients with fragility fractures compared to T2D patients without fracture, suggesting that cortical bone may be preferentially compromised in T2D patients who fracture. Unfortunately, quantification of cortical porosity in vivo remains challenging both because median pore size is ∼50 µm[33] and because cortical remnants in the endosteal “transition zone” can be interpreted as either porous cortical bone or trabeculae.[34] However, no previous study has measured BMS, an important component of bone quality.[15, 16] Indeed, our data show that T2D patients have significantly reduced BMS compared to age-matched, nondiabetic controls, even after adjustment for potential confounders. By contrast, after adjustment for BMI, we found that T2D patients have BMD (by DXA) and bone microarchitectural parameters (by HRpQCT) that did not differ significantly from nondiabetic controls; however, as noted earlier, radial cortical porosity tended to be higher in the T2D patients, which is consistent with the findings of Burghardt and colleagues.[30] Nevertheless, the contribution of the observed increase in cortical porosity to fracture risk in T2D patients is currently unknown.

T2D is more common with advancing age, and therefore frequently coexists with age-related bone loss.[14] Thus, established risk factors for fragility fractures with normal aging contribute to fracture risk in T2D patients. However, considerable evidence exists that subsets of fracture risk factors are either exacerbated in T2D (eg, poor balance, reduced muscle quality, and falls) or are T2D-specific (eg, poor glycemic control, T2D duration, and diabetic medications and complications).[35] For example, evidence exists that poor glycemic control (ie, a glycated hemoglobin level ≥7.5%) is associated with higher fracture risk in T2D patients,[36] although this relationship has not been established unequivocally.[37] In the present study, the average glycated hemoglobin level over the previous 10 years was negatively associated with BMS in patients with T2D, suggesting that prolonged hyperglycemia could have detrimental effects on bone quality, and raising the testable hypothesis that intensive glucose control may not only reduce microvascular and neuropathic complications,[38] but also fragility fractures. By contrast, duration of T2D was not related to BMS, although selecting patients on the basis of diagnosis of T2D for >10 years may have limited our ability to detect a possible relationship between duration of disease and BMS.

A number of diabetic medications have also been implicated in fracture risk in T2D. For example, long-term TZD use doubles female fracture risk;[39] thus, we excluded T2D patients on TZDs. Also, in a population-based study of Olmsted County, MN, residents with T2D,[7] we found that use of insulin, but not biguanide (eg, metformin), was associated with increased fracture risk. However, associations with fracture risk may reflect the fact that diabetic medications are also markers of poor glycemic control rather than having direct negative effects on bone quality. This premise is supported by in vitro data demonstrating stimulatory effects of insulin and metformin on osteoblast proliferation and differentiation.[40, 41] In addition, as expected, we found that use of drugs commonly prescribed to treat diabetic complications (eg, β-blockers, ACE inhibitors, ARBs, calcium channel blockers, statins, and thiazides) tended to be higher in patients with T2D. Although some of these drugs may affect bone metabolism in humans, most evidence suggests that these medications are beneficial, rather than harmful, to bone.[42] In addition, the presence of multiple diabetic complications may be linked to increased fracture risk in T2D patients, although findings to date have been inconsistent.[4-8] Indeed, some evidence suggests that diabetic complications do not sufficiently explain the increased fracture risk in patients with T2D.[43] In support of this premise, our data show that adjustment for diabetic complications does not nullify the significant difference in BMS between patients with T2D and nondiabetic controls.

Consistent with previous studies,[44-47] we found that bone turnover was reduced in the T2D patients as compared to the controls. Whereas the underlying causes for low bone turnover and, in particular, reduced bone formation in T2D are not well understood, studies in rodents have demonstrated that metabolic dysfunction in T2D is related to an accelerated onset of aging.[48, 49] Like aging, T2D may interfere with skeletal homeostasis by suppressing osteoblast differentiation and enhancing osteoblast apoptosis—potentially leading to a defect in bone formation.[48] Reduced bone formation and an imbalance between bone formation and resorption are also suggested by the few available histomorphometric studies in patients with T2D.[44, 47] Therefore, current collective evidence indicates that T2D is associated with low bone turnover, with perhaps a disproportionate reduction in bone formation relative to bone resorption.

The cellular mechanisms responsible for reduced BMS in patients with T2D are incompletely understood, although the proximal culprits likely include advanced glycation end-products (AGEs) and oxidative stress.[14, 48] AGEs are intermediate protein products that become glycated and undergo undesired chemical modifications after excessive glucose exposure. Thus, prolonged high circulating glucose concentrations in T2D may lead not only to higher glycated hemoglobin levels, but also to the accumulation of AGEs in both circulation and bone tissue.[50] In T2D, both non-enzymatic glycation via AGEs and enzymatic biochemical processes may create collagen crosslinks that lead to biomechanically more brittle bone with impaired toughening mechanisms (ie, resistance to fracture) at both microstructural and nanostructural levels that negatively impact bone quality.[50] Moreover, analogous to high-dose bisphosphonate therapy,[51] prolonged low bone turnover in T2D may result in further increases in bone AGEs, defective microdamage repair, and increased bone microcrack accumulation—creating a “vicious cycle” that contributes to increased fracture risk.

Our study had a number of limitations. For example, our sample size is relatively small and our findings are cross-sectional; thus, they need to be confirmed prospectively. In addition, our study included women diagnosed with T2D who were otherwise healthy; BMS might be much worse in patients with multiple comorbidities or in those with fragility fracture. Additional concerns are the lack of histomorphometric indices of bone turnover and measures of AGEs in bone biopsies, as well as the generalizability of these data from a sample that was predominantly white. Nevertheless, our results can be reasonably extrapolated to a large part of the general population. Further, a concern with HRpQCT imaging is that despite permitting much higher resolution, measures of bone microarchitecture (eg, cortical porosity) are, in fact, only estimates of the true values. Nonetheless, HRpQCT is perhaps at the current limit of feasibility for assessing bone microarchitecture in vivo in humans.

Another valid concern is safety of the OsteoProbe instrument. Notably, our group and others[19, 20] have experience using this device in human clinical research, and based on follow-up we have found no complications. However, as noted earlier, patients who have a significant skin disorder, bruising, local edema, or infection, as well as those undergoing treatment for blood clots or severe coagulation defects should not be exposed to this procedure. In addition, because the OsteoProbe is a relatively new technology, data relating BMS to fracture risk in humans are somewhat limited. However, as noted earlier, previous work has shown alterations in bone material properties in hip fracture[19] and atypical femoral fracture[20] patients. Furthermore, the in vivo OsteoProbe instrument is similar, in principle, to the in vitro BioDent instrument, which has been validated extensively in animals.[27-29] Indeed, in two different animal models with compromised bone quality, including a model of T2D, strong correlations were shown between bone material properties derived from the BioDent and traditional “gold standard” bone mechanical testing techniques (three-point bending and axial compression) at both appendicular and axial skeletal sites.[29] Furthermore, based on data in diabetic and nondiabetic rats,[29] a 23.4% difference in femoral IDI (by the Biodent) corresponds to a 66.7% difference in femoral toughness (by three-point bending testing), a measure of fracture resistance. Based on these estimates, the 11.7% deficit in BMS (by the OsteoProbe) we observed in T2D patients relative to nondiabetic controls would result in a 33.4% deficit in toughness, which would be expected to be clinically significant. Therefore, the reductions in BMS we observed in postmenopausal women with T2D are concerning. However, further work is needed to more precisely define the relations between in vivo bone microindentation testing and fracture resistance in humans.

In conclusion, our findings represent the first demonstration, using a direct in vivo measure of BMS, of compromised bone material properties in patients with T2D. Further, our results confirm previous studies demonstrating low bone turnover in patients with T2D and highlight the potential detrimental effects of prolonged hyperglycemia on bone quality. Thus, the skeleton needs to be recognized as another important target tissue subject to diabetic complications.

Acknowledgments

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

This work was supported by NIH grants AG004875, AR027065, GM065354, and UL1 TR000135 (Mayo Clinical and Translational Science Award). JNF is supported by T32 DK007352: Diabetes and Metabolism. Active Life Scientific provided the microindentation instrument and probes used in this study but had no control over the outcomes or the content of the manuscript. We thank Active Life Scientific for their technical support during the study, in particular Drs. Paul Hansma and Hal Kopeikin, as well as Mr. Davis Brimer. SK had full access to all data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.

Authors' roles: Study design: JNF, MTD, SA, LJM, and SK. Study contact: SK. Study conduct: JNF and LKM. Data collection: JNF and LKM. Data analysis: JNF. Data interpretation: JNF, MTD, SA, LJM, and SK. Drafting manuscript: JNF and SK. Revising manuscript content: JNF, MTD, SA, LJM, LKM, and SK. Approving final manuscript: JNF, MTD, SA, LJM, LKM, and SK.

References

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgments
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
jbmr2016-sm-0001-SupplementaryData.pdf150K

Supplementary Fig 1. Image of the OsteoProbe® Reference Point Indenter (Active Life Scientific Inc., Santa Barbara, CA), a hand-held microindentation instrument designed for in vivo measurements of bone material strength (BMS) in humans at the midshaft of the non-dominant anterior tibia.

Supplementary Fig 2. Time (s; seconds) versus force (N; newtons) plot showing a single OsteoProbe® measurement cycle including the four basic operations of the device: (1) pre-load, (2) trigger, (3) impact, (4) unloading. Reproduced from Bridges et al. (Rev Sci Instrum 2012;83:044301) with permission.

Supplementary Table 1. Medications of the patients with type 2 diabetes and age-matched, non-diabetic controls.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.