Relationship Between α-Glucosidase Inhibitors Application and Bone Mineral Density in Chinese Patients with Type 2 Diabetes Mellitus: A Cross-Sectional Study

Introduction

The prevalence of diabetes is increasing globally. According to IDF Diabetes Atlas (11th edition), the number of adults aged 20–79 with diabetes worldwide reached 589 million in 2024 and is expected to rise to 853 million by 2050, with 90% cases being type 2 diabetes mellitus (T2DM).¹ Compared with healthy individuals, patients with T2DM have an increased risk of developing osteoporosis and related fractures (such as hip, spinal, or even any site),^2,3 which may be related to insulin resistance and poor glycemic control,⁴ inflammation,⁵ hyperosmotic diuresis triggering the loss of calcium and phosphorus, and the impact of hypoglycemic drugs,⁶ and other factors that lead to the deterioration of the bone micro-environment. The influence of hypoglycemic drugs on bone mineral density (BMD) is intricately linked to the particular drug in question.^6–11 Several hypoglycemic drugs, such as rosiglitazone^12–15 and canagliflozin,¹⁶ have been shown to adversely affect bone health. Since T2DM patients require the use of hypoglycemic drugs for most of the disease process, the influences of different classes of hypoglycemic drugs on BMD and fracture risk need to be carefully evaluated.

China is the country with the most extensive use of α-glucosidase inhibitors (AGIs) worldwide.¹⁷ The results of previous researches on the of AGIs on bone mass are extremely inconsistent, and the approaches used to detect and evaluate bone metabolism are often one-sided, so the effects of AGIs on bone health need to be further investigated.

In this study, BMD was measured by dual-energy X-ray absorptiometry (DXA), and the levels of bone metabolism biomarker and FRAX scores for fracture risk were combined to comprehensively analyze the impacts of AGIs application on the BMD and fracture risk of T2DM patients.

Materials and Methods

Study Population

A total of 251 patients with T2DM admitted to the Zhongshan Hospital, Fudan University (Xiamen Branch) from September 2018 to September 2020 were enrolled. All patients met the 2018 American Diabetes Association’s (ADA’s) criteria for the diagnosis and classification of diabetes and were negative for serum glutamic acid decarboxylase antibodies, including 149 males and 102 females, and all patients underwent DXA scan. Medical records of the patients were reviewed to confirm their medication status. According to whether AGIs (including acarbose and voglibose) were used at admission, the patients were divided into the AGIs group and the Non-AGIs group. In addition, the AGIs group was required to have received the drug for at least 3 months prior to enrollment, while the non-AGIs group had no history of relevant drug use within 3 months prior to enrollment.

Exclusion criteria include: (1) previously and currently diagnosed with type 1 diabetes, secondary diabetes, and special type diabetes; (2) age ≤ 18 years; (3) estimated glomerular filtration rate (eGFR) < 60 mL/(min ·1.73m2); (4) use of estrogen, progestogen, glucocorticoids, bisphosphonates, desmuximab and other agents that may affect BMD within the past six months; (5) presence of any disease prone to secondary osteoporosis, such as hyperparathyroidism, multiple myeloma, autoimmune diseases; (6) long-term bed rest leading to bone loss; (7) metal implants affecting BMD detection; (8) presence of psychiatric disease, infection or stress state, severe organ dysfunction, and any other conditions considering unsuitable for inclusion.

Information and Data Collection

Demographic and clinical information for all patients were obtained from the medical records and interviews, including age, gender, smoking history, drinking history, duration of T2DM, diabetic complications and hypoglycemic drug use. Height, weight, waist circumference, hip circumference, and blood pressure data were measured, and body mass index (BMI) and waist-to-hip ratio were calculated.

Laboratory indicators were based on early-morning venous blood after an overnight fast of 8 hours, and the detection items included: (1) Bone turnover markers: carboxy-terminal β-crosslinking telopeptide of type 1 collagen (β-CTX), type I procollagen amino-terminal peptide (P1NP), osteocalcin (OC) and 25-hydroxy vitamin D3 (25OHD). (2) Blood biochemical indicators: triacylglycerol (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL), high-density lipoprotein cholesterol (HDL), calcium (Ca), phosphorus (P), alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), uric acid (UA), blood creatinine (CREA) and eGFR. (3) Glycometabolic indicators: glycated hemoglobin (HbA1c), fasting blood glucose (FBG), fasting insulin, and the HOMA-IR (Homeostatic Model Assessment for Insulin Resistance) was calculated according to the formula: fasting blood glucose (mmol·L⁻¹) × fasting insulin (μU·mL⁻¹)/22.5.¹⁸ (4) Hormone indicators: thyroid stimulating hormone (TSH), free triiodothyronine (FT3), free thyroxine (FT4), parathyroid hormone (PTH), estradiol (E2). Above indicators were determined by using the same methods as previously described.¹⁹

Definition of Osteoporosis and Assessment of Fracture Risk

Based on the Chinese guidelines for osteoporosis (2022 version),²⁰ the definition of osteoporosis should include any one of the following three items: ① Minimum T-score of BMD ≤−2.5. ② Hip or vertebral fragility fractures, regardless of BMD measurement. ③ Fragility fractures of the proximal humerus, pelvis, or distal forearm with −2.5 < T-score < −1.0. BMD was measured by DXA (QDR4500A, Hologic Inc., Waltham MA, USA) at the femoral neck and lumbar spine (L1-L4), with a dedicated technician from the Department Medical Imaging responsible for the assessments. Fracture Risk Assessment Tool (FRAX) score²¹ was used to assess the 10-year probability of a major osteoporotic fracture and hip fracture.

Statistical Analysis

Data conforming to the normal distribution were expressed as mean ± standard deviation (SD), and T-test is used for statistical difference analysis. Non-normally distributed data were represented in terms of the median (P25, P75), and T-test was applied after the data was transformed into a normal distribution. Non-parametric test was conducted for data that could not be converted to normal. Group differences between the BMD tertiles were analyzed by ANOVA. Count data were expressed as n (%) and χ² test was used. Multiple linear regression analysis was applied to explore the influence of AGIs on bone mass, with BMD as the dependent variable and related factors as covariates. Data were analyzed by SPSS 27.0 software, and statistical graphs were plotted by GraphPad Prism 9.0. P < 0.5 was considered statistically significant.

Results

Demographic and Clinical Characteristics of the Population Stratified by BMD

The patients were divided into groups by tertiles of femoral neck BMD and lumbar spine total BMD, respectively. The demographic and clinical characteristics of different BMD patients were compared in Table 1. In all groups, patients with lower BMI and E2 levels tended to have lower BMD. In the femoral neck BMD subgroups, the decrease of BMD was accompanied by the increase of patient’s age, the prolongation of diabetic course and the decline of LDL and FT4 levels (P < 0.05), and there were no significant differences in the use of hypoglycemic agents. However, in the lumbar spine BMD subgroups, the proportion of patients with AGIs application in the lowest BMD group was obviously higher and the uric acid level was notably lower than that in the other two groups (P < 0.05), while there were no marked deviations in age, diabetes course, LDL and FT4 levels among the groups. No striking differences between groups were found for any of the other analyzed indicators in both femoral neck BMD and lumbar spine BMD groups.

Table 1 Demographic and Clinical Characteristics of the Population Stratified by BMD

Comparison of Basic Data Between T2DM Patients with and without AGIs Application

Of the 251 T2DM patients included, 58 received AGIs therapy and 193 did not. The AGIs group consisted of 26 males and 32 females, and the non-AGIs group consisted of 123 males and 70 females. Table 2 showed that compared with the Non-AGIs group, the AGIs group had obvious higher proportion of female, longer diabetes course, more patients with concomitant diabetic peripheral vascular lesions, and sulfonylurea application, and had significant lower values of HbA1c, TC, TG, LDL, FT3, and FT4 (all P < 0.05). The remaining parameters between the two groups were not statistically significant.

Table 2 Comparison of Basic Data Between Non-AGIs and AGIs Groups

Comparison of the Risk of Osteoporosis and Fracture Between the AGIs Group and the Non-AGIs Group

The comparison of BMD, FRAX score and prevalence of osteoporosis between AGIs group and Non-AGIs group were presented in Table 3. The results showed that the BMD of both femoral neck and lumbar spine were significantly lower and the FRAX score was notably higher in the AGIs group than those in the Non-AGIs group (P < 0.05), indicating a higher risk of severe osteoporotic fracture and hip fracture within 10 years. Meanwhile, the proportion of patients with osteoporosis was obviously higher in the AGIs group than in the Non-AGIs group (P < 0.05). Additionally, statistical analysis of bone turnover markers revealed that the β-CTX and P1NP levels in the AGIs group were significantly lower than those in the Non-AGIs group (P < 0.01), while no significant difference was observed in 25OHD and OC levels between the two groups (Figure 1), suggesting that the bone transformation rate in AGIs group was lower.

Table 3 The Comparison of BMD, FRAX Score and Prevalence of Osteoporosis Between AGIs Group and Non-AGIs Group

Figure 1 Bone turnover markers between AGIs group and Non-AGIs group. Notes: (A) type I procollagen amino-terminal peptide (P1NP); (B) carboxy-terminal β-crosslinking telopeptide of type 1 collagen (β-CTX); (C) osteocalcin (OC); (D) 25-hydroxy vitamin D3 (25OHD). **P < 0.01, ****P < 0.0001.

Correlation Between AGIs Application and BMD Level in T2DM Patients

In order to further explore the correlation between AGIs application and BMD in T2DM patients, a multiple linear regression analysis was conducted with BMD value as the dependent variable. Table 4 indicated a distinguish negative correlation between AGIs application and lumbar spine total BMD after adjusting for potential confounding variables (β = −0.053, 95% CI −0.100~-0.006, P = 0.029). However, there was no significant correlation between AGIs application and femoral neck BMD.

Table 4 Multiple Linear Regression Analysis Evaluated the Association Between AGIs Application and BMD in T2DM Patients

Discussion

With the increasingly serious aging of China’s population, the prevalence of osteoporosis, as an age-related disease, is also ascending. It is worth noting that T2DM patients are more prone to fractures than non-T2DM patients, even at the same BMD levels.^2,3,22,23 Therefore, in the whole process of diabetes management, it is necessary to carefully evaluate the potential factors that promote the decline of bone mass in T2DM patients, especially those that can be avoided or changed, such as medication choice. This study used DXA to measure BMD levels in patients and the FRAX score to assess bone strength and the risk of fractures within 10 years. The findings revealed that BMD levels in T2DM patients were negatively correlated with the application of AGIs. Compared with the patients without AGIs use, patients with AGIs use had lower BMD values, higher incidence of osteoporosis, and higher FRAX scores, accompanied by the reduction of both bone turnover markers P1NP and β-CTX levels. Furthermore, AGIs application was identified as an independent risk factor for a decrease in lumbar BMD.

AGIs delay the digestion of starch and sucrose in the intestines by reversibly and competitively inhibiting pancreatic α-amylase and membrane-bound intestinal α-glucosidase, flattening the postprandial blood glucose excursions,²⁴ with high safety and no drug interaction. Current research on oral AGIs and bone health is limited. In 2021, a meta-analysis compared the fracture risks associated with various hypoglycemic drugs and found that voglibose reduced the risk of fractures in diabetic patients and was significantly beneficial than insulin, sodium-glucose cotransporter 2 inhibitors (SGLT2i), dipeptidyl peptidase 4 inhibitors (DPP4i), and some glucagon-like peptide 1 receptor agonists (GLP-1RAs) in terms of fracture.²⁵ In the same year, a double-blind, randomized clinical trial indicated that attenuating postprandial glycemic indices, with an AGI, markedly decreased postprandial bone resorption and could be explained by the rise in serum GLP-1,²⁶ suggesting that AGI may have bone-protective effects. However, a nationwide observational study in Japanese elderly with diabetes found AGI was related to increased hip fracture risk compared with metformin.²⁷ Additionally, a small sample clinical control study in 2023 demonstrated that treatment with acarbose reduced postprandial suppression of bone resorption in patients with T2DM.²⁸ Overall, the current evidence is inconclusive and conflicting regarding the effects of AGIs on bone metabolism and risk of fractures.

Based on the statistical results of 251 T2DM patients included in this study, we found that the use of AGIs was negatively correlated with the level of lumbar spine BMD and may increase the risk of osteoporosis and fracture. This result may be partially due to the longer duration of diabetes and more diabetic peripheral vascular lesions in the patients in AGIs-group, since previous studies had shown that T2DM complications were associated with a higher risk of fracture.²⁹ In a Danish case-control study, diabetic retinopathy (OR 2.1), kidney disease (OR 2.0), neuropathy (OR 1.9), and macrovascular complications (OR 1.9) were all associated with an increased risk of fracture30. Fracture events were also more common when T2DM persisted over 10 years.³⁰ In addition, the levels of HbA1c, FT3 and FT4 in AGIs group were significantly lower than those in Non-AGIs group, hinting that the reduction of BMD was not caused by worse blood glucose control,³¹ hyperthyroidism,³² and other adverse factors for bone. Although the proportion of female in AGIs group was higher than that in the Non-AGIs group, there was no significant difference between the two groups in age and E2 level. It should be noted that in terms of the types of hypoglycemic drugs, only one of the 251 patients used GLP-1RA because it was not widespread at that time. The use of sulfonylureas was higher in the AGIs group than in the Non-AGIs group, but the impact of sulfonylureas on bone mass remains unclarified. Previous studies showed a reduced incidence of fractures in patients treated with sulfonylureas,^33,34 while paradoxically, other studies suggested that sulfonylureas may increase the risk of falls and fractures by increasing hypoglycemia.^8,35

To further clarify the bone conversion after AGIs use, we analyzed the levels of bone turnover markers and found that the levels of P1NP and β-CTX were significantly declined in the AGIs group, suggesting a lower bone transformation rate.

Existing researches have shown that in T2DM patients, AGIs regulate postprandial glucose and bone resorption by stimulating GLP-1 secretion and inhibiting DPP-4 to improve circulating GLP-1 levels,^36–38 and acarbose was found to have a positive effect on the intestinal microbiota.³⁹ Therefore, it is speculated that the alteration of intestinal microbiota and intestinal insulin may be related to the mechanism by which AGIs affect bone metabolism, but the above conclusions have not been fully verified and recognized.

More interestingly, we found that AGIs application was an independent risk factor for lumbar spine BMD reduction in T2DM patients, but there was no significant correlation with femoral neck BMD. It was speculated that this may be related to the different metabolic and bone composition characteristics of different skeletal sites. The lumbar spine is cancellous bone, which may be more prone to bone loss.

So far, there is insufficient research on the impacts of commonly used hypoglycemic drugs AGIs on bone metabolism. Our study provides more evidence for the effects of AGIs on bone mass in T2DM. We found that in clinically severe diabetic patients, the use of AGIs was associated with a decrease in BMD and potentially an increased the risk of osteoporosis and fractures, which may suggest caution in the use of AGIs among patients with T2DM patients and poor BMD. However, as a cross-sectional study, this study has several limitations: 1. The sample size is small; 2. As a single-center study, there is a selection bias; 3. There is a lack of data on the duration and dosage of AGIs used by patients, which limits further analysis, thus reducing its value. The effects and mechanisms of AGIs on bone metabolism require further large-sample and scientifically designed studies for validation.

Conclusion

This study indicates that the use of AGIs in patients with T2DM is significantly associated with an increased risk of reduced BMD, osteoporosis, and fracture.