Serum CTRP3 and CysC correlate with glucose, lipid, and bone metabolism in elderly patients with type 2 diabetes mellitus complicated with osteoporosis

ABSTRACT

Objective

This study analyzed the correlation between serum C1q/TNF‐related protein 3 (CTRP3) and cystatin C (CysC) and glucose‐lipid metabolism, bone mineral density (BMD), and bone metabolism markers in elderly patients with type 2 diabetes mellitus (T2DM) complicated with osteoporosis (OP).

Methods

A retrospective analysis was conducted on 235 patients with T2DM admitted to Taizhou Municipal Hospital between January 2021 and June 2024. They were categorized into the T2DM group (n = 131) and the T2DM + OP group (n = 104). A healthy control group (n = 95) was established, comprising elderly individuals without diabetes or OP. Serum CTRP3 and CysC were measured. The correlation between serum CTRP3 and CysC levels and relevant metabolism markers and BMD was analyzed. The diagnostic values of serum CTRP3 and CysC were assessed, as well as their role in T2DM + OP.

Results

Serum CTRP3 was lower and CysC was higher in the T2DM + OP group. Serum CTRP3 and CysC levels in the T2DM + OP group were correlated with some indicators of glucose‐lipid metabolism and bone metabolism and BMD. The area under the curve (AUC) values of serum CTRP3 and CysC levels in distinguishing T2DM + OP were 0.719 and 0.702 (sensitivity: 60.58%; 61.54%, specificity: 74.05%; 74.05%), respectively, with their combination showing better performance (AUC: 0.773, sensitivity: 76.92%, specificity: 64.89%). Elevated serum CysC (OR: 1.422, 95%CI: 1.161–1.742) and CTRP3 (OR: 0.960, 95%CI: 0.943–0.976) were independent risk and protective factors for T2DM + OP in elderly patients, respectively (all P < 0.05).

Conclusion

CTRP3 and CysC combined detection helps distinguish OP in elderly patients with T2DM.

Serum levels of CTRP3 and CysC in elderly patients with T2DM and OP are closely associated with glucose‐lipid metabolism, bone mineral density, and bone metabolism markers. Combined detection of these two markers can assist in distinguishing OP in elderly patients with T2DM.

INTRODUCTION

The chronic metabolic disease known as type 2 diabetes mellitus (T2DM) is characterized by hyperglycemia due to insufficient insulin and insulin resistance ¹ . Osteoporosis (OP) is a systemic skeletal disorder characterized by reduced bone mineral density (BMD), microarchitectural degeneration, impaired bone strength, increased fragility, and heightened fracture risk. It represents one of the major complications of type 2 diabetes mellitus (T2DM) ² . The incidence of T2DM‐associated OP continues to rise. Research indicates that the risk of this condition gradually increases with age among individuals with T2DM, leading to a significant rise in fracture risk ³ . Dual‐energy X‐ray absorptiometry (DXA) remains a crucial tool for diagnosing OP and assessing fracture risk, but it has limitations, such as delayed reflection of bone metabolism and limited detection of specific skeletal regions ⁴ . By contrast, evaluating relevant biomarkers in blood or urine can provide real‐time insights into bone metabolism, bone quality, and fracture risk, offering valuable guidance for clinical diagnosis and treatment.

C1q/tumor necrosis factor‐related protein 3 (CTRP3) is a newly identified adipokine, a member of the CTRP superfamily with high homology to adiponectin. It enhances insulin sensitivity, glucose regulation, and lipid modulation, as well as has anti‐atherosclerosis effects ⁵ . Plasma CTRP3 levels are reduced in high‐fat diet‐induced obese mice ⁶ , and there is a negative correlation between circulating CTRP3 levels and T2DM status ⁷ . Cystatin C (CysC), a small molecular protein, undergoes glomerular filtration followed by complete reabsorption and degradation, resulting in low plasma concentrations under normal renal function. As a biomarker of glomerular filtration rate (GFR), cystatin C (CysC) exhibits a negative correlation with GFR and is widely used for the early diagnosis of kidney disease and the assessment of changes in renal function ⁸ , ⁹ . It has been reported that elevated serum CysC levels may increase the risk of osteoporotic fractures, predominantly hip fractures ¹⁰ .

Research indicates that CTRP3 not only ameliorates renal fibrosis ¹¹ but also affects lipid metabolism in renal tubular cells, suggesting that CTRP3 is a potential therapeutic target for diabetic kidney injury ¹² . Furthermore, a recent study has detected elevated serum CysC levels in patients with T2DM with increased visceral fat area ¹³ . Nevertheless, the potential interplay between CTRP3 and CysC remains unclear, particularly in the context of T2DM complicated with OP. This study aimed to evaluate serum CTRP3 and CysC in elderly patients with T2DM complicated with OP and their correlations with glucose‐lipid metabolism, BMD, and bone metabolism. This study aimed to offer new insights into the clinical treatment of elderly patients with T2DM complicated with OP.

MATERIALS AND METHODS

Participants

Three hundred and twenty‐three patients with T2DM treated at Taizhou Municipal Hospital between January 2021 and June 2024 were analyzed for our retrospective study. On the basis of the inclusion and exclusion criteria, 235 patients were ultimately selected for the study. These patients were divided into a T2DM group (n = 131) and a T2DM + OP group (n = 104). A healthy control group (n = 95) was established, comprising elderly individuals who underwent health examinations at Taizhou Municipal Hospital during the same period and had no history of diabetes or OP. This study was authorized by Taizhou Municipal Hospital ethics committee and complied with the Declaration of Helsinki.

Inclusion and exclusion criteria

The inclusion criteria were as follows: (1) Patients with T2DM met the 1999 World Health Organization (WHO) diagnostic criteria for T2DM, defined as typical symptoms of diabetes (polydipsia, polyphagia, and polyuria) plus one of the following: random venous plasma glucose level ≥11.1 mmol/L, fasting blood glucose level ≥7.0 mmol/L, 2‐h blood glucose level ≥11.1 mmol/L during an oral glucose tolerance test, or glycated hemoglobin (HbA1c) ≥6.5%. (2) Patients with T2DM and OP met the 1994 WHO diagnostic criteria for OP, based on DXA BMD measurements and a history of fragility fractures, defined as a DXA‐measured BMD T‐score ≤ −2.5. Alternatively, regardless of the T‐score, patients may be diagnosed with OP if they have a history of fragility fractures involving the hip or vertebrae (Osteoporotic fractures, also known as fragility fractures, refer to pathological fractures that occur in individuals with OP. Due to reduced bone density and quality, bone strength diminishes, leading to fractures that can result from minor trauma or even routine activities. These represent a severe consequence of OP) ¹⁴ , ¹⁵ , ¹⁶ . (3) Complete clinical data. (4) Age >60 years.

Exclusion criteria: (1) Acute or chronic infections, autoimmune illnesses, gastrointestinal problems, skeletal disorders, connective tissue diseases, hematologic disorders, or other endocrine abnormalities. (2) Presence of malignancy. (3) The use of medications affecting bone metabolism and glucose‐lipid metabolism, such as hormone replacement therapy, bisphosphonates, proton pump inhibitors, or glucocorticoids in the past 6 months. (4) Physical disability, long‐term bedridden status, or inability to perform daily activities for other reasons. (5) T1DM or specific types of diabetes.

Sample size estimation

During the sample size design phase of this study, we referenced preliminary pilot study data and existing literature on expression differences of serum CTRP3 and CysC ¹⁷ , ¹⁸ . Based on pre‐experimental findings, CTRP3 and CysC exhibited moderate effect sizes (Cohen's f ≈ 0.25) across the three groups (healthy controls, T2DM group, and T2DM + OP group). Using G*Power 3.1.9.7 software for sample size estimation, with α = 0.05 (two‐tailed) and statistical power 1 − β = 0.90, the minimum total sample size required for three‐group comparisons via one‐way anova was calculated to be 207. This study estimated 10 potential variables affecting the co‐occurrence of OP in patients with T2DM based on prior research and literature reports for multivariate logistic regression analysis. Following the event‐per‐variable (EPV) principle, the target outcome was the presence of OP in patients with T2DM, with an anticipated event rate of approximately 44%. To ensure model robustness, at least 100 events (10 variables × 10 EPV) were required, resulting in an estimated total sample size of 100/0.44 = 227 cases. Accounting for potential data loss and group imbalance, the study ultimately enrolled 235 patients with T2DM and 95 healthy controls, yielding a total sample size of 330 cases that met the statistical requirements. For ROC curve analysis, using the sample size estimation tool provided by MedCalc software, with an expected AUC of 0.80, α = 0.05, and 90% power, the sample sizes for each group in this study exceeded the minimum required per group, demonstrating strong statistical stability.

Data collection

Data collected included age, gender, body mass index (BMI), blood pressure, course of T2DM, history of smoking and alcohol use, underlying diseases, and indicators of lipid metabolism (triglycerides [TG], total cholesterol [TC], high‐density lipoprotein cholesterol [HDL‐C], low‐density lipoprotein cholesterol [LDL‐C]) and glucose metabolism (fasting plasma glucose [FPG], HbA1c, fasting insulin [FINS], and homeostasis model assessment of insulin resistance [HOMA‐IR]). All blood samples were collected as 5 mL of fasting antecubital venous blood upon patient admission. Serum was separated and sent for testing. A fully automated biochemical analyzer (Model AU5800, Beckman Coulter, Brea, CA, USA) was used to assess lipid metabolism markers (TG, TC, HDL‐C, and LDL‐C). FINS was measured using an electrochemiluminescence immunoassay analyzer (Model Cobas8000, Roche, Basel, Switzerland), and HbA1c was measured using high‐performance liquid chromatography (D‐10 System, Bio‐Rad, Hercules, CA, USA). FPG was detected using the glucose oxidase method. HOMA‐IR was calculated as FINS × FPG/22.5.

Enzyme‐linked immunosorbent assay (ELISA)

The levels of bone metabolism markers, including serum C‐terminal telopeptide of type I collagen (β‐CTX) (ml038555), osteocalcin (OC) (ml058528), procollagen type I N‐terminal propeptide (PINP) (ml038554), tartrate‐resistant acid phosphatase‐5b (TRACP‐5b) (ml060039), CysC (ml058113), and CTRP3 (ml063269), were measured using ELISA kits (MLBio, Shanghai, China) with an FK‐SY96S multifunctional microplate reader (Fangke, Shandong, China). The intra‐assay coefficient of variation (CV) was <8%, and the inter‐assay CV was <10%, meeting clinical testing standards. The entire process was conducted using a blinded method. Testing operators were unaware of the sample group assignments or clinical data, performing experiments and recording data solely based on coded identifiers to prevent subjective bias.

BMD measurement

BMD measurements were performed using a DXA scanner (Hologic, Inc., Discovery‐WI model) to assess BMD (g/cm²) at the hip, left femoral neck, and L1–L4 spine.

Statistical analysis

GraphPad Prism 9.5 (GraphPad Software Inc., San Diego, CA, USA) and SPSS 27.0 (IBM Corp., Armonk, NY, USA) software programs were used for data analysis and graphing. The Kolmogorov–Smirnov test was used to assess normality. Quantitative data with a normal distribution were represented as mean ± SD. Comparisons between two groups were performed using the independent‐samples t‐test, while comparisons among multiple groups were analyzed using one‐way anova. Post hoc analysis employed Tukey's multiple comparisons test. For correlation analysis, Pearson correlation coefficients (r) were employed. Non‐normally distributed quantitative data were represented by the median (Q1, Q3). Comparisons between two groups were performed using the Mann–Whitney U test, while comparisons among multiple groups were conducted using the Kruskal–Wallis test. Post hoc analysis employed Dunn's multiple comparisons test. Count data were expressed as case numbers and percentages. Intergroup comparisons were performed using the chi‐square test. The ability of serum CTRP3 and CysC to distinguish OP in elderly patients with T2DM was examined using ROC curves. AUCs were compared using MedCalc software, and Delong's test was utilized. The risk factors for T2DM complicated with OP were identified using the multivariate logistic regression analysis. VIF ≥ 5 and tolerance ≤0.1 indicate severe collinearity between respective variables. For two‐tailed tests, the threshold for statistical significance was established at P < 0.05.

RESULTS

Comparisons of baseline data

Differences in gender, BMI, blood pressure, smoking and alcohol use history, or underlying disease between the three patient groups were not statistically significant (all P > 0.05). However, statistically significant differences were observed among the three groups in age, course of T2DM, eGFR, markers of glucose‐lipid metabolism, markers of bone metabolism, and BMD (all P < 0.05), as shown in Table 1.

Table 1.

Baseline characteristics comparison among elder patients with T2DM

Parameters	Control group (n = 95)	T2DM group (n = 131)	T2DM + OP group (n = 104)	P‐value
Gender (Male/Female)	55/40	76/55	64/40	0.828
Age (years)	67 (64, 72)	67 (64, 69)	69 (65, 71)	0.010
BMI (kg/m2)	22.97 ± 1.80	23.09 ± 2.06	23.55 ± 2.49	0.118
Systolic blood pressure (mmHg)	125.46 ± 6.39	125.29 ± 7.58	126.44 ± 8.33	0.459
Diastolic blood pressure (mmHg)	81.28 ± 5.29	81.41 ± 5.19	82.56 ± 6.34	0.186
Duration of T2DM (years)	–	6 (2, 13)	7 (2, 12)	0.006
Smoking history (case, %)	35 (36.84)	50 (38.17)	41 (39.42)	0.932
Alcohol history (case, %)	51 (53.68)	72 (54.96)	58 (55.77)	0.957
Hypertension (case, %)	–	45 (34.35)	38 (36.54)	0.728
Hyperlipidemia (case, %)	–	65 (49.62)	46 (44.23)	0.411
Stroke (case, %)	–	15 (11.45)	12 (11.54)	0.983
Coronary heart disease (case, %)	–	14 (10.69)	10 (9.62)	0.788
eGFR (mL/min/1.73m2)	72.37 ± 5.72	71.52 ± 6.62	69.39 ± 4.43	<0.001
Lipid metabolism index
TG (mmol/L)	1.51 ± 0.24	1.67 ± 0.26	1.72 ± 0.35	<0.001
TC (mmol/L)	4.61 ± 0.31	4.83 ± 0.37	5.02 ± 0.32	<0.001
HDL‐C (mmol/L)	1.39 ± 0.22	1.31 ± 0.31	1.29 ± 0.25	0.015
LDL‐C (mmol/L)	2.10 ± 0.45	2.25 ± 0.52	2.46 ± 0.60	<0.001
Glucose metabolism index
FPG (mmol/L)	4.92 ± 0.60	8.67 ± 1.07	8.45 ± 1.62	<0.001
HbA1c (%)	5.2 ± 0.37	7.62 ± 0.65	8.10 ± 1.12	<0.001
FINS (mU/L)	8.88 ± 1.83	11.74 ± 2.92	14.35 ± 4.30	<0.001
HOMA‐IR	1.95 ± 0.38	4.63 ± 1.64	5.38 ± 1.84	<0.001
Bone metabolism marker
β‐CTX (ng/mL)	0.42 ± 0.11	0.51 ± 0.12	0.66 ± 0.15	<0.001
OC (ng/mL)	9.56 ± 2.07	12.25 ± 2.09	14.97 ± 3.22	<0.001
PINP (ng/mL)	38.51 ± 4.12	44.15 ± 5.88	51.83 ± 6.47	<0.001
TRACP‐5b (U/L)	2.94 ± 0.57	4.05 ± 0.84	6.17 ± 1.49	<0.001
Lumbar spine BMD (g/cm2)	0.99 ± 0.08	0.98 ± 0.12	0.84 ± 0.16	<0.001
Femoral neck BMD (g/cm2)	0.86 ± 0.10	0.84 ± 0.12	0.71 ± 0.12	<0.001
Hip BMD (g/cm2)	0.92 ± 0.09	0.90 ± 0.13	0.78 ± 0.13	<0.001

Measurement data that conformed to normal distribution were expressed as mean ± SD, and the independent‐sample t‐test was used to compare two groups, while one‐way anova was used to compare multi‐groups, followed by the Tukey's multiple comparison test. Measurements that were not normally distributed were expressed as median values (minimum, maximum), and comparisons between the two groups were made using the Mann–Whitney U test, while comparisons among multiple groups were made using the Kruskal–Wallis test, followed by the Dunn's multiple comparison test. BMD, body mass density; BMI, body mass index; FINS, fasting insulin; FPG, fasting plasma glucose; HbA1c, glycosylated hemoglobin; HDL‐C, high‐density lipoprotein cholesterol; HOMA–IR, homa insulin resistance; LDL‐C, low‐density lipoprotein cholesterol; OC, osteocalcin; OP, osteoporosis; PINP, amino‐terminal propeptide of type I procollagen; T2DM, type 2 diabetes mellitus; TC, total cholesterol; TG, triglyceride; TRACP‐5b, Tartrate‐resistant acid phosphatase 5b; β‐CTX, beta C‐terminal cross‐linked telopeptides of type I collagen.

Serum CTRP3 is reduced while CysC is increased in elderly patients with T2DM complicated with OP, and CTRP3 is negatively correlated with CysC

Serum CTRP3 and CysC levels were measured in elderly patients with T2DM and OP using ELISA. Serum CTRP3 level was lower, whereas serum CysC level was higher (both P < 0.001, Figure 1a,b) in the T2DM + OP group than in the T2DM and Control groups. Pearson correlation analysis revealed a negative correlation between serum CTRP3 and CysC in elderly patients with T2DM with OP (r = −0.219, P = 0.025; Figure 1c).

Figure 1.

Comparisons of serum CTRP3 and CysC levels and their correlation analyses. (a, b) Normally distributed measurement data are expressed as mean ± SD, and comparisons between the two groups were performed using the independent‐sample t‐test. ***P < 0.001. (c) Pearson correlation analysis of the serum CTRP3 and CysC levels in elderly patients with T2DM with OP. r represents the correlation coefficient.

Serum CTRP3 and CysC correlate with glucose metabolism in elderly patients with T2DM with OP

Serum CTRP3 level in elderly patients with T2DM with OP showed a moderate negative correlation with FPG (r = −0.485) and HOMA‐IR (r = −0.435; all P < 0.05, Figure 2a,c). Serum CysC was weakly positively correlated with FPG (r = 0.218) and HOMA‐IR (r = 0.215; all P < 0.05, Figure 2e,g). No significant correlations were found between serum CTRP3 or CysC level and FINS or HbA1c level (all P > 0.05, Figure 2b,d,f,h).

Figure 2.

Correlation analyses of serum CTRP3, CysC, and glucose metabolism indicators in patients with T2DM with OP. Serum CTRP3 and CysC levels were measured using ELISA. (a–d) Pearson correlation analysis of CTRP3 and glucose metabolism indicators. (e–h) Pearson correlation analysis of CysC and glucose metabolism indicators. r represents the correlation coefficient.

Serum CTRP3 and CysC correlate with lipid metabolism in elderly patients with T2DM with OP

Serum CTRP3 level in elderly patients with T2DM with OP was positively correlated with HDL‐C (r = 0.292) and negatively correlated with TG (r = −0.346) and LDL‐C (r = −0.378) levels (all P < 0.001; Figure 3a,c,d). Serum Cys and TG were positively correlated (r = 0.323; P < 0.001, Figure 3e). Serum CysC level did not correlate with TC, HDL‐C, or LDL‐C (all P > 0.05, Figure 3f–h). There was no correlation between serum CTRP3 and TC levels in elderly patients with T2DM with OP (P > 0.05, Figure 3b).

Figure 3.

Correlation analyses of serum CTRP3, CysC, and lipid metabolism indicators in patients with T2DM with OP. Serum CTRP3 and CysC levels were measured using ELISA. (a–d) Pearson correlation analysis of CTRP3 and lipid metabolism indicators. (e–h) Pearson correlation analysis of CysC and lipid metabolism indicators. r represents the correlation coefficient.

Serum CTRP3 and CysC correlate with bone metabolism markers and BMD in elderly patients with T2DM with OP

Pearson correlation analysis revealed that the serum CTRP3 level in elderly patients with T2DM with OP was moderately negatively correlated with β‐CTX (r = −0.501), OC (r = −0.515), PINP (r = −0.432), and TRACP‐5b (r = −0.470; all P < 0.001, Figure 4a–d) and moderately positively correlated with the lumbar spine BMD (r = 0.575), femoral neck BMD (r = 0.521), and hip BMD (r = 0.496; all P < 0.001, Figure 4e–g). Serum CysC was weakly positively correlated with β‐CTX (r = 0.318), OC (r = 0.329), PINP (r = 0.220), and TRACP‐5b (r = 0.357; all P < 0.001, Figure 4h–k) and negatively correlated with lumbar spine BMD (r = −0.408), femoral neck BMD (r = −0.226), and hip BMD (r = −0.313; all P < 0.001, Figure 4l–n) in elderly patients with T2DM with OP.

Figure 4.

Correlation analyses of serum CTRP3, CysC, and bone metabolism markers and BMD in patients with T2DM with OP. Serum CTRP3 and CysC levels were measured using ELISA. (a–g) Pearson correlation analysis of CTRP3 and bone metabolism markers and BMD. (h–n) Pearson correlation analysis of CysC and bone metabolism markers and BMD. r represents the correlation coefficient.

Serum CTRP3 and CysC assist in distinguishing OP in elderly patients with T2DM

Further investigation was conducted on the discriminatory ability of serum CTRP3 and CysC, both individually and in combination, for identifying OP in elderly patients with T2DM. With a cutoff value of 83.05, sensitivity of 60.58%, and specificity of 74.05%, the AUC of serum CTRP3 was 0.719 (95%CI: 0.657–0.776; Figure 5a). With a cutoff value of 0.97, sensitivity of 61.54%, and specificity of 74.05%, the AUC of serum CysC was 0.702 (95%CI: 0.639–0.760; Figure 5b). The sensitivity of the combined prediction model was 76.92%, the specificity was 64.89%, and the AUC was 0.773 (95%CI: 0.714–0.825; Figure 5c). Additionally, further comparison using MedCalc software revealed that the combined testing of serum CTRP3 and CysC had a significantly better distinguishing ability than either biomarker tested alone (P = 0.020 for CTRP3 and P = 0.012 for CysC; Table 2 and Figure 5d).

Figure 5.

ROC curve analysis of serum CTRP3 and CysC levels and their combined detection for distinguishing OP in elderly patients with T2DM. (a) ROC curve of CTRP3; (b) ROC curve of Cys C; (c) ROC curve of combined CTRP3 and Cys C; (d) ROC curves for predicting OP in elderly patients with T2DM using various indicators.

Table 2.

Comparison of the distinguishing ability of serum CTRP3, CysC alone, and combined tests for the presence of OP complication in elderly T2DM patients

Indicators	AUC	95%CI	Sensitivity (%)	Specificity (%)
Serum CTRP3	0.719	0.657–0.776	60.58	74.05
Serum CysC	0.702	0.639–0.760	61.54	74.05
Combination	0.773	0.714–0.825	76.92	64.89
Serum CTRP3~combination	P = 0.020
Serum CysC~combination	P = 0.012

Multiple AUC comparisons were made using the Delong test in MedCalc software.

Analysis of risk factors for OP in elderly patients with T2DM

Subsequently, we conducted an in‐depth analysis of the risk factors for OP in patients with T2DM. With the presence of OP (0 = No, 1 = Yes) as the dependent variable, gender, age, BMI, blood pressure, smoking and alcohol use history, underlying diseases, course of T2DM, eGFR, glucose‐lipid metabolism indicators, and serum CTRP3 and CysC were included in a univariate logistic regression model. Then, indicators with P < 0.05, as well as gender, were included in the multivariate logistic regression analysis. The composite index HOMA‐IR (HOMA‐IR = fasting insulin × fasting glucose/22.5) was used to represent the combined information conveyed by FINS and FPG. Therefore, in the multivariate logistic regression model, HOMA‐IR was included as a composite indicator to assess its independent association with OP, while its constituent components, FINS and FPG, were not included separately. This approach also avoided multicollinearity within the model. Collinearity diagnosis found that none of the independent variables mentioned above showed significant collinearity (Table S1). Elevated serum CysC (OR: 1.422, 95%CI: 1.161–1.742) was an independent risk factor for elderly patients with T2DM complicated with OP, while elevated serum CTRP3 (OR: 0.960, 95%CI: 0.943–0.976) was an independent protective factor (all P < 0.05, Table 3). In addition, the Hosmer–Lemeshow test showed a P‐value of 0.065, indicating that the fitting effect met the requirements and did not show statistical evidence of overfitting.

Table 3.

Analysis of risk factors for elderly T2DM complicated with OP

Items	Univariate analysis			Multivariate analysis
	P	OR	95%CI	P	OR	95%CI
Gender (Male = 1, Female = 0)	0.585	1.158	0.684–1.959	0.561	1.214	0.632–2.334
Age (years)	0.002	1.117	1.040–1.200	0.129	1.069	0.981–1.164
BMI (kg/m2)	0.119	1.096	0.977–1.229	–	–	–
Systolic blood pressure (mmHg)	0.268	1.019	0.986–1.053	–	–	–
Diastolic blood pressure (mmHg)	0.130	1.036	0.990–1.084	–	–	–
Duration of T2DM (years)	0.016	1.195	1.034–1.381	0.092	1.154	0.977–1.365
Smoking history (Yes = 1, No = 0)	0.844	1.054	0.622–1.788	–	–	–
Alcohol history (Yes = 1, No = 0)	0.902	1.033	0.616–1.734	–	–	–
Hypertension (Yes = 1, No = 0)	0.728	1.100	0.643–1.884	–	–	–
Hyperlipidemia (Yes = 1, No = 0)	0.411	0.805	0.480–1.350	–	–	–
Stroke (Yes = 1, No = 0)	0.983	1.009	0.450–2.260	–	–	–
Coronary heart disease (Yes = 1, No = 0)	0.788	0.889	0.378–2.092	‐	‐	‐
eGFR (mL/min/1.73m2)	0.007	0.936	0.892–0.982	0.060	0.947	0.895–1.002
TG (mmol/L)	0.187	1.780	0.755–4.193	–	–	–
TC (mmol/L)	<0.001	4.886	2.233–10.691	0.052	2.696	0.993–7.321
HDL‐C (mmol/L)	0.492	0.727	0.293–1.806	–	–	–
LDL‐C (mmol/L)	0.005	1.992	1.235–3.212	0.877	1.067	0.469–2.429
FPG (mmol/L)	0.220	0.886	0.729–1.075	–	–	–
HbA1c (%)	<0.001	1.814	1.338–2.459	0.055	1.434	0.993–2.071
FINS (mU/L)	<0.001	1.224	1.130–1.326	–	–	–
HOMA‐IR	0.001	1.283	1.101–1.496	0.821	0.969	0.738–1.272
Serum CTRP3 (ng/mL)	<0.001	0.959	0.945–0.973	<0.001	0.960	0.943–0.976
Serum CysC (10−1 mg/L)	<0.001	1.552	1.301–1.850	<0.001	1.422	1.161–1.742

Given the relatively small absolute differences between groups for CysC, its OR value may still exhibit statistically high values. To validate the robustness of this result, the original CysC values were multiplied by 10 and re‐entered into the multivariate logistic regression. Then, the OR value for CysC indicates that for every 0.1 mg/L increase in CysC, the risk of combined OP increases by approximately 1.422 times.

DISCUSSION

Individuals with T2DM face a heightened risk of developing OP and are also more likely to exhibit calcium and phosphorus metabolism abnormalities, lower osteocalcin levels, and impaired bone remodeling, all of which further increase the likelihood of OP ¹⁹ . In this study, we evaluated serum CTRP3 and CysC levels in elderly patients with T2DM with OP.

Persistent blood glucose fluctuations in patients with T2DM often result in metabolic disturbances affecting three primary nutrients, such as proteins, fats, and carbohydrates, which adversely impact the bone matrix ²⁰ . Moreover, lumbar spine, femoral neck, and total hip BMD measurements are commonly employed to assess BMD in patients with T2DM ²¹ . In our analysis, significant differences were observed among three groups in terms of age, course of T2DM, eGFR, glucose‐lipid metabolism indicators, bone turnover markers, and BMD measurements. These findings align with earlier studies, such as one showing significantly higher LDL‐C levels in the OP group and higher HDL‐C levels in the non‐OP group (P < 0.05) ²² . Other studies have indicated that FPG, postprandial glucose, and HbA1c levels are higher in patients with T2DM with OP (T‐score ≤ −2.5) ²³ . Elderly patients with T2DM‐OP have a longer course of disease, increased BMI, HOMA‐IR, β‐CTX, and TRACP‐5b, alongside reduced PINP levels ²⁴ .

Our subsequent experiments revealed reduced serum CTRP3 levels and elevated CysC levels in elderly patients with T2DM with OP. Although this study observed a negative correlation between serum CTRP3 and CysC in elderly patients with T2DM‐OP (r = −0.219, P = 0.025), its biological mechanism remains unexplained. Based on existing literature, we propose the following potential hypothesis to explain this observation: First, dysregulation of the metabolic‐inflammatory axis may serve as the core underlying mechanism linking the two. As an adipokine with insulin‐sensitizing and anti‐inflammatory properties, reduced CTRP3 levels signify weakened metabolic protective mechanisms, thereby exacerbating insulin resistance and chronic low‐grade inflammation ¹² , ²⁵ . Meanwhile, mounting evidence indicates that CysC is not merely a passive marker of renal function but actively participates in inflammatory processes. Elevated levels of CysC reflect systemic inflammatory burden ¹³ , ¹⁸ . Therefore, decreased CTRP3 and elevated CysC may represent either the same pathological condition (i.e., worsening metabolic inflammation) or a pair of biomarkers that change inversely within that condition. They synergistically disrupt bone metabolism balance by creating an inflammatory environment. Second, the kidneys may serve as a key organ linking CTRP3 and CysC interactions. CTRP3 possesses anti‐renal fibrosis properties and protective effects on renal function ¹¹ . Therefore, we hypothesize that lower CTRP3 levels may predispose the kidneys to early injury, leading to elevated accumulation of CysC, a sensitive marker of glomerular filtration rate, thus constituting a potential indirect link. Finally, the two may be indirectly associated through their synergistic negative impact on bone metabolism. As previously mentioned, low CTRP3 and high CysC levels collectively signal a pro‐inflammatory and metabolically disordered internal environment. This environment promotes osteoclast activation and osteoblast suppression, jointly promoting the onset and progression of OP by diminishing bone protective factors and increasing the risk of bone destruction, respectively. Of course, these hypotheses require further experimental validation such as observing the effects of regulating CTRP3 levels on CysC levels and bone metabolism in animal models. Elucidating the mechanism between CTRP3 and CysC will open new perspectives for understanding the pathogenesis of T2DM complicated by OP.

Furthermore, the changes in serum CTRP3 and CysC levels in elderly patients with T2DM with OP were correlated with some glucose‐lipid metabolism and bone metabolism markers, as well as BMD. First, studies have confirmed that CTRP3 improves insulin sensitivity, has anti‐inflammatory effects, and regulates lipid metabolism ²⁶ , ²⁷ . In the context of T2DM and OP, insulin resistance and chronic low‐grade inflammation are two key common pathological bases ¹⁷ . Insulin resistance can impair the function of osteoblasts, while inflammatory factors, such as TNF‐α and IL‐6, can strongly promote osteoclastogenesis and activation. Therefore, the observed decrease in serum CTRP3 levels might indirectly disrupt the balance of bone remodeling by exacerbating insulin resistance and inflammation, leading to reduced bone formation and enhanced bone resorption, ultimately contributing to OP. CTRP3 levels are positively correlated with BMD ²⁸ and HDL‐C and negatively correlated with TG ²⁹ . This study consistently found that CTRP3 was negatively correlated with HOMA‐IR, negatively correlated with bone resorption markers, and positively correlated with BMD. CysC has traditionally been regarded as a biomarker of renal function. Renal dysfunction directly affects calcium and phosphorus metabolism and activation of vitamin D, thereby interfering with bone metabolism. However, increasing evidence suggests that CysC itself may directly participate in bone regulation ³⁰ , ³¹ . Under pathological conditions, such as chronic kidney disease or metabolic disorders, an abnormal increase in CysC levels may reflect a disruption in its physiological inhibitory function, or it may itself become a pathological signal that promotes bone resorption ³² . Increasing evidence directly indicates that elevated Cre/CysC ratios are independently associated with BMD in elderly patients with T2DM ³³ . Since this ratio increases as CysC levels decrease, the biological observation aligns with current and previous data showing a negative correlation between absolute CysC concentrations and BMD ¹⁰ , ³² . In our study, CysC was positively correlated with bone resorption markers (β‐CTX and TRACP‐5b), suggesting that high levels of CysC might be associated with enhanced osteoclast activity in elderly patients with T2DM, and its specific role deserves further investigation.

Further analyses suggest that decreases in serum CTRP3 levels may provide value as a diagnostic indicator for postmenopausal OP ²⁸ . Moreover, CysC can serve as a potential biomarker for predicting OP in middle‐aged and elderly adults ¹⁸ . In addition, elevated CysC levels have been linked to an increased risk of OP and fractures, with renal function and physical activity as contributing factors ³² . CysC is a sensitive renal marker ³⁴ and participates in the regulation of bone metabolism. Even mild renal impairment in elderly adults may disturb calcium, phosphate, and vitamin D metabolism, reducing 1,25(OH)₂D production and increasing bone resorption ³⁵ . Although few studies have explored the correlation between CTRP3 and CysC, these pathways may explain the opposing effects of CTRP3 and CysC to serve as independent risk and protective factors in T2DM‐OP in the elderly, respectively, and support their potential value for distinguishing OP in elderly patients with T2DM.

This study demonstrated that, in elderly patients with T2DM, a decrease in serum CTRP3 levels and an increase in CysC levels were significantly correlated with adverse glucose‐lipid metabolism conditions and deterioration of bone health. The combined detection of the two showed potential clinical application value in assisting in the identification of patients with concomitant OP. These associations provide important clinical evidence for further exploration of their role in the pathogenesis of T2DM complicated by OP. Serum CysC testing is already widely used in clinical practice with standardized, low‐cost kits (e.g., immunoturbidimetric assays). Although CTRP3 detection remains in the research phase, commercial ELISA kits (e.g., R&D Systems, Cloud‐Clone Corp.) make testing technically feasible. Compared with traditional OP screening tools such as DXA, blood‐based CTRP3/CysC detection is more cost‐effective and reflects metabolic risk. CysC is routinely tested in clinical labs, and CTRP3 can be measured via third‐party services without the need for specialized equipment, making combined testing feasible even in primary care, especially in resource‐limited areas. Future studies should assess its clinical cost‐effectiveness using health economic models. This study revealed links between CTRP3/CysC and metabolic‐skeletal markers in elderly patients with T2DM, suggesting their combined use for supporting timely identification and individualized interventions and promoting integrated management of metabolic and bone health.

Multivariate logistic regression analysis revealed that after adjusting for age, gender, course of disease, glucose‐lipid metabolism, elevated serum CysC (OR: 1.422, 95%CI: 1.161–1.742) was an independent risk factor for OP in elderly patients with T2DM, while elevated serum CTRP3 (OR: 0.960, 95%CI: 0.943–0.976) was an independent protective factor. This indicates that for every 0.1 mg/L increase in serum CysC, the risk of OP increases by approximately 1.422‐fold, and for every 1 ng/mL increase in serum CTRP3, the risk of OP decreases by approximately 0.960‐fold. CysC, as a sensitive marker of renal function, exhibits elevated levels that may indicate early renal impairment. This impairment exerts a potent negative effect on bone metabolism through classic pathways such as calcium‐phosphorus homeostasis and vitamin D metabolism ¹⁸ , ³² , ³⁶ . Therefore, CysC may serve as a potent risk indicator reflecting dysfunction within the kidney‐bone axis. Despite controlling for multiple known confounding factors, residual confounding effects from unmeasured variables, such as vitamin D levels, physical activity, and gender hormone levels, cannot be entirely ruled out. This finding underscores the importance for clinicians to prioritize monitoring CysC levels in elderly patients with T2DM. Even mild elevations within the normal range may indicate significant OP risk, warranting early intervention in conjunction with bone density testing. Additionally, in univariate analysis, significant differences in age were observed among the three patient groups (P = 0.011). However, after adjusting for multiple factors including CysC, blood glucose, and blood lipids in the multivariate model, age did not emerge as an independent risk factor for OP (P > 0.05). This seemingly contradictory result aligns with clinical logic and can be understood through the concept of mediation in statistics. Age serves as a fundamental risk factor for multiple chronic diseases, with its effects often mediated through more specific pathological alterations. In this study, advancing age may lead to renal decline (as indicated by elevated CysC) and dysregulation of glucose‐lipid metabolism, which are key metabolic indicators directly influencing bone metabolism. Multivariate analysis revealed that after adjusting for intermediate variables such as CysC, FBG, and HbA1c, the residual independent effect of age itself became non‐significant. This suggests that in elderly patients with T2DM, quantifiable metabolic abnormalities driven by age, such as early renal decline, hold stronger and more direct predictive value for assessing OP risk compared with the unmodifiable factor of age itself. This finding shifts clinical focus from general demographic characteristics toward more specific, potentially interventionable pathological indicators.

Despite its important findings, this study has several limitations. First, only the baseline serum levels of CTRP3 and CysC were measured, without longitudinal monitoring. This limited our ability to evaluate dynamic changes in these biomarkers in relation to disease progression or treatment response. Second, and most importantly, as a cross‐sectional clinical correlation analysis, this study mainly revealed correlations between CTRP3 and CysC with bone metabolism indicators but did not delve into the specific molecular mechanisms underlying them, which needed to be elucidated through future basic experiments and prospective cohort studies. The direct causal chain and signaling pathways (such as Wnt/β‐catenin, RANKL/RANK/OPG, etc.) connecting these two biomarkers with bone metabolism imbalance (such as osteoblast differentiation inhibition or osteoclast activity enhancement) have not been elucidated in this study either. Therefore, our findings are more suggestive than conclusive, providing hypotheses and directions for future mechanism research. Third, the diagnosis of OP was based solely on BMD, which does not capture other key components of bone quality, such as turnover rates, microarchitecture, or geometry. Fourth, although body fat plays a critical role in metabolic syndrome and may affect CTRP3 and CysC levels through mechanisms such as inflammation, insulin resistance, and adipokine secretion ³⁷ , we did not directly assess body fat percentage or visceral fat area in the elderly population. Instead, we used BMI as a surrogate indicator; however, BMI cannot distinguish between fat and muscle mass and may underestimate the true influence of adiposity. This limitation may affect the precision of our conclusions regarding the metabolic‐bone axis. Future studies should aim to include direct measures of body composition, employ larger and more diverse samples, and adopt longitudinal and mechanistic approaches. These efforts will help clarify the roles of CTRP3 and CysC in the pathophysiology of T2DM complicated by OP and ultimately support more effective prevention and treatment strategies for this vulnerable population. Finally, this study noted that gender was a significant factor influencing BMD and metabolic phenotypes. Although there were no statistically significant differences in gender distribution among study groups at baseline, and gender was adjusted as a covariate in multivariate analyses, we were unable to conduct systematic gender‐stratified analyses for all indicators. This limitation primarily stems from the potential insufficiency of sample sizes in stratified subgroups (e.g., n = 40 for the T2DM + OP female subgroup) to support robust statistical inference. Forcing such analyses could instead yield misleading conclusions due to insufficient test power. The primary objective of this study was to preliminarily elucidate the overall association patterns between CTRP3 and CysC in elderly individuals with T2DM and OP. Our preliminary analysis revealed that after controlling for gender, the primary correlation between CTRP3, CysC, BMD, and metabolic indicators remained consistent. Future research should include recruiting larger cohort samples specifically designed to explore potential gender‐specific differences in the role of CTRP3 and CysC in the development of OP among male and female patients with T2DM. Additionally, investigations should examine whether these associations are influenced by factors such as menopausal status, thereby providing more precise evidence for personalized diagnosis and treatment.

AUTHOR CONTRIBUTIONS

YPW is the guarantor of the integrity of the entire study and contributed to the study design; YW contributed to the literature research and statistical analysis; TC contributed to the study concepts, data acquisition, and data analysis; GC contributed to the definition of intellectual content, clinical studies, and manuscript editing; ZYY contributed to the manuscript preparation and review; all authors read and approved the final manuscript.

DISCLOSURE

The authors declare no conflict of interest.

Approval of the research protocol: The study was approved by the academic ethics committee of Taizhou Municipal Hospital and abided by the Declaration of Helsinki.

Informed consent: All patients were fully aware of the study objective, with signatures on the informed consent forms obtained from them.

Registry and the registration no. of the study/trial: N/A.

Animal studies: N/A.

FUNDING

No funding was received for this study.

CONSENT FOR PUBLICATION

Not applicable.

Supporting information

Table S1. Collinearity diagnostics.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.