Assessing P1NP/β‐CTX as Potential Risk Factors Associated With Different Subtypes of Fragile Hip Fracture in Elderly Patients

ABSTRACT

Objective

Identifying the risk factor for occurrence of trochanteric/femoral neck fractures has always been difficult and crucial for prognosis of the diseases. The goal of this study is to evaluate the difference of P1NP/β‐CTX between trochanteric and femoral neck fractures and establish these two BTMs as independent risk factors for these two subtypes of hip fractures in the elderly.

Methods

Following case–control study design, 83 elderly patients with fragile hip fractures from two tertiary hospitals recruited between February 2023 to August 2024 were assigned into intertrochanteric fracture group (42 cases) and femoral neck fracture group (41 cases). Baseline characteristics including age, sex, height, weight, BMI, and history of diabetes/Parkinson's disease were collected after admission. Besides, P1NP/β‐CTX/PTH/25(OH)D blood tests and dual energy x‐ray absorptiometry (DXA) were performed on all participants before surgical operation. The statistical significances of categorical variables and continuous variables were established through χ ²‐tests and independent samples t‐tests, respectively. Multivariable logistic regression analysis was applied to detect potential associations between the selected variables and hip fracture subtype.

Results

The difference in P1NP between the trochanteric fracture group and femoral neck fracture groups is statistically significant. Specifically, patients with trochanteric fractures showed a markedly higher value of P1NP than those with femoral neck fractures (p = 0.044). However, no significant difference was observed for β‐CTX between the two groups (p = 0.941). Furthermore, BMDs at trochanter and total hip regions in the trochanteric fracture group were significantly lower than those in the femoral neck fracture group (p = 0.028; p = 0.022). After multifactorial logistic regression, P1NP, weight, and female were stronger risk factors for trochanteric fracture (OR 0.969, 95% CI 0.948–0.990, p = 0.004; OR 0.531, 95% CI 0.284–0.992, p = 0.047; OR 0.082, 95% CI 0.008–0.837, p = 0.035), while height was found to be independently associated with femoral neck fracture (OR 1.715, 95% CI 1.023–2.874, p = 0.041).

Conclusions

Compared with femoral neck fractures, P1NP was established as an independent risk factor for trochanteric fractures, suggesting the great potential of this BTM as a predictive indicator and postoperative complication monitor for trochanteric fractures.

By exploring the correlation between P1NP/β‐CTX and the subtypes of hip fractures, this study attempts to establish these two bone turnover markers as predictive indicators and postoperative complication monitors for such common fractures in the elderly.

1. Introduction

Proximal femoral fractures represent a major clinical concern due to their profound impact on both life expectancy and quality of life in the elderly population. These fractures have a notably high incidence, with studies estimating that 6% of men and 18% of women will sustain such an injury during their lifetime. The global burden is expected to rise substantially, with projections indicating that the number of patients affected by proximal femoral fractures would reach 6.3 million by 2050 [1].

This condition is commonly categorized into two main types: femoral neck fractures and intertrochanteric fractures. A number of previous studies have sought to identify the risk factors associated with each of these distinct fracture subtypes. According to studies involving only postmenopausal female patients, those with intertrochanteric fractures were older on average compared to those with femoral neck fractures [2, 3]. Similarly, another study including patients of both sexes found that the association between intertrochanteric fractures and advanced age was stronger in female patients [4]. Besides advanced age, gender has also been established as a significant risk factor influencing the type of hip fracture. Sari Tal et al. found that intertrochanteric fractures are more prevalent among women [5]. Additionally, specific structural parameters of the proximal femur have been implicated in determining hip fracture types. For instance, a smaller neck‐shaft angle is a risk factor for intertrochanteric fractures, while a larger femoral neck diameter is a protective factor against such fractures [6, 7]. The connection between specific chronic diseases and the type of proximal femoral fracture has also been the subject of previous investigations. Among these, diabetes and Parkinson's disease have been found to be associated with intertrochanteric fractures and femoral neck fractures, respectively [5]. Beyond these, physical function‐related indicators such as falls and frailty, as well as biochemical markers like blood calcium and hemoglobin, have also been demonstrated to be associated with the occurrence of different proximal femoral fracture types [5, 8].

Bone turnover markers (BTMs), with C‐terminal telopeptide of type I collagen (CTX) and N‐terminal propeptide of type I procollagen (P1NP) as representatives, are typical metabolites released during bone remodeling [9, 10]. As a C‐terminal telopeptide fragment released into the bloodstream during osteoclast‐mediated bone breakdown [11, 12], β‐CTX is a highly specific marker for the activity of bone resorption. Conversely, P1NP is cleaved from procollagen type I during the synthesis of new type I collagen by osteoblasts, which makes it highly specific for monitoring osteoblastic bone formation activity [11, 12]. These markers aid in osteoporosis diagnosis and treatment monitoring—a decrease in β‐CTX following anti‐osteoporosis therapy indicates a positive anti‐resorptive response, while an increase in P1NP signifies the desired anabolic effect [13]. Many previous studies have shown that bone mineral density is lower in patients with intertrochanteric fractures compared to those with femoral neck fractures, suggesting a potential correlation between osteoporosis and hip fracture subtype [4, 14, 15]. In this study, we aim to: (i) explore the correlation between P1NP/β‐CTX and the subtypes of hip fractures; (ii) establish these two bone turnover markers as effective predictive indicators for such common fractures in the elderly.

2. Methods

2.1. Study Design and Patient Screening

This was a tertiary hospital‐based, multi‐center, case–control study. A total of 322 patients diagnosed with hip fractures who were admitted to Peking Union Medical College Hospital/Peking University People's Hospital between February 2023 and August 2024 were screened based on the following inclusion and exclusion criteria (Figure 1).

FIGURE 1.

Flow chart of patient screening based on inclusion/exclusion criteria.

Inclusion criteria: (1) senior patient of age > 65 years; (2) first‐onset, unilateral hip fracture; (3) low‐trauma fractures caused by spontaneous fall on flat ground; (4) patients targeted for postoperative anti‐osteoporosis treatment who underwent related biochemical test and examinations before the surgery, including dual energy x‐ray absorptiometry (DXA), P1NP/β‐CTX/PTH/25(OH)D blood test.

Exclusion criteria: (1) mental disability or nervous system diseases preventing the patients from cooperating with the required examinations of this study, such as dementia, severe delirium, etc.; (2) diseases that severely damage the external and internal structural integrity of proximal femur, such as malignant bone tumors, congenital hip dysplasia or malformation; (3) history of medications known to significantly affect bone metabolism (anti‐resorptive or anabolic agents for osteoporosis, systemic glucocorticoids, etc.).

According to the criteria above, the patients with qualified conditions were included in this study, which were further divided into intertrochanteric fracture group (case group) and femoral neck fracture group (control group) based on physical examination and medical imaging, respectively. Other than DXA and blooding testing, baseline characteristics including age, sex, height, weight, body mass index (BMI), and bone metabolism‐affecting comorbidities (diabetes and Parkinson's disease) were also collected after admission.

This study was approved by the ethical committees of Peking Union Medical College Hospital/Peking University People's Hospital (I‐25PJ3195/2026PHB058‐001). Informed consent was obtained from all participants.

2.2. BMD Measurement

The healthy hip joint of the recruited patients was scanned by DXA bone densitometer (GE Healthcare, Waukesha, WI, USA) with equal measurement routine in both institutes. Bone density (g/cm²), T‐score, and Z‐score of the hip were determined for the following three regions of the proximal femur: femoral neck, trochanter, and total hip. The examination was acquired within 7 days after admittance before the surgical treatment (the surgery is regularly performed 1 week within admission in both centers), and the measurement was obtained with hip joint in center of the image and femur shaft in 15°–25° internal rotation [16]. Here, the precision error was expressed as the percent coefficient of variation (%CV), with a value of 1% [4].

2.3. Biochemical Testing

Fasting blood samples measuring P1NP (ng/mL), β‐CTX (ng/mL), PTH (pg/mL), and 25(OH)D (ng/mL) were obtained from the patients of both subgroups 3 days within the time of admission and before the surgical operation [17]. The samples were analyzed using an automatic biochemical analyzer (Cobas 8000, Germany).

2.4. Statistical Analyses

In this study, statistical analyses were performed using SPSS 22.0 software (IBM Corporation, Armonk, NY, US). Continuous data presented as mean ± standard deviation were assessed for normality with the Shapiro–Wilk test. The statistical significance of categorical variables was established through χ ²‐tests, and the comparison of continuous variables between the two hip fracture type subgroups were evaluated by independent samples t‐tests. Besides, Odds ratio and P value corrected by multivariable logistic regression analysis were computed to detect potential association between the selected variables and hip fracture subtype. Continuous predictors were standardized (as z‐scores) prior to inclusion in the regression model to facilitate interpretation and comparison of effect sizes. Potential collinearity among independent variables (e.g., BMI is derived from height and weight) was assessed using the variance inflation factor (VIF), with a threshold of VIF < 5 considered acceptable [18]. A p value of < 0.05 was considered statistically significant for all primary analyses.

3. Results

3.1. Baseline Characteristics

After screening, a total of 83 hip fracture patients (41 femoral neck fractures and 42 trochanteric fractures) satisfying the inclusion and exclusion criteria were recruited for subsequent analysis. As for baseline characteristics, no statistical difference was detected between the patients of the two hip fracture subgroups. These include diabetes and Parkinson's disease, which highly correlate to an abnormal state of bone metabolism (Table 1). According to the values of average ± standard deviation, however, patients with trochanteric fractures tend to be shorter and heavier compared to those with femoral neck fractures, resulting in a larger BMI (Table 1).

TABLE 1.

Baseline characteristics, including gender, age, height, weight, body mass index (BMI), and comorbidities (diabetes and Parkinson's disease) for patients of femoral neck and trochanteric fractures (mean ± standard deviation).

Parameters	Femoral neck (n = 41)	Trochanteric (n = 42)	p
Male	7	7
Female	34	35	0.961
Age (years)	80.10 ± 7.79	78.40 ± 9.79	0.387
Height (cm)	159.68 ± 5.73	158.34 ± 7.91	0.379
Weight (kg)	59.32 ± 11.23	60.04 ± 13.56	0.794
BMI (kg/m2)	23.26 ± 4.29	24.00 ± 5.75	0.508
Diabetes (%)	19.51	23.81	0.635
Parkinson's disease (%)	2.44	2.38	0.986

Note: Body mass index (BMI) was calculated by height and weight accordingly. *p < 0.05 for consideration of statistical significance.

3.2. Differences in BMD and Results of Biochemical Testing Between Femoral Neck and Trochanteric Fractures

Significant difference was observed between the femoral neck fracture and trochanteric fracture patients in regards to hip BMD. Specifically, the patient with trochanteric fracture showed a significantly lower BMD compared to those with femoral neck fracture at the regions of trochanter and total hip (p = 0.028; p = 0.022, respectively, Table 2). Among all of the biomechanical indicators examined in this study, only P1NP exhibited a difference between the two groups with statistical significance, whereby the corresponding value was larger in the trochanteric fracture group than that in the femoral neck fracture group (p = 0.044, Table 2).

TABLE 2.

Comparisons of bone mineral densities and biochemical tests between the femoral neck and trochanteric fracture groups (mean ± standard deviation).

Parameters	Femoral neck (n = 41)	Trochanteric (n = 42)	p
NeckBMD (g/cm2)	0.67 ± 0.11	0.64 ± 0.11	0.166
TrBMD (g/cm2)	0.59 ± 0.14	0.52 ± 0.12	0.028*
TotalBMD (g/cm2)	0.73 ± 0.12	0.67 ± 0.12	0.022*
25OH D (ng/mL)	42.48 ± 38.95	39.05 ± 35.68	0.677
PTH (pg/mL)	38.46 ± 18.37	43.19 ± 19.73	0.262
P1NP (ng/mL)	50.06 ± 33.47	68.12 ± 45.81	0.044*
β‐CTX (ng/mL)	0.59 ± 0.30	0.60 ± 0.30	0.941

Note: Bone mineral densities were measured at femoral neck (NeckBMD), trochanter (TrBMD), and total hip (TotalBMD), and biochemical indicators including C‐terminal telopeptide of type I collagen (CTX), N‐terminal propeptide of type I procollagen (P1NP), 25‐Hydroxyvitamin D (25OH D), and Parathyroid Hormone (PTH) were assessed.

^* p < 0.05 for consideration of statistical significance.

3.3. Risk Factors for Femoral Neck and Trochanteric Fractures

Consisted with the significance of difference demonstrated in Table 2, P1NP is identified as risk factor for trochanteric fracture after multi‐factorial regressive analysis (OR 0.969, 95% CI 0.948–0.990, p = 0.004, Table 3). Furthermore, male patients with higher height and lower weight were at lower risk for trochanteric fracture (OR 0.082, 95% CI 0.008–0.837, p = 0.035; OR 1.715, 95% CI 1.023–2.874, p = 0.041; OR 0.531, 95% CI 0.284–0.992, p = 0.047, respectively, Table 3).

TABLE 3.

All predictor variables of the two hip fracture subtypes that were assessed by logistic regression for identification of potential significant risk factors.

Parameter	Odds ratio	95% Confidence interval	p
Gender	0.082	0.008–0.837	0.035*
Age (years)	1.075	0.993–1.163	0.075
Height (cm)	1.715	1.023–2.874	0.041*
Weight (kg)	0.531	0.284–0.992	0.047*
BMI (kg/m2)	4.484	0.988–20.356	0.052
Diabetes (%)	0.673	0.188–2.413	0.543
Parkinson's disease (%)	3.009	0.084–107.867	0.546
NeckBMD (g/cm2)	2.133	0.000–20542.903	0.871
TrBMD (g/cm2)	1.461	0.000–409126.002	0.953
TotalBMD (g/cm2)	18464.574	0.003–5,220,000	0.218
25OH D (ng/mL)	1.003	0.988–1.019	0.663
PTH (pg/mL)	0.989	0.960–1.018	0.443
P1NP (ng/mL)	0.969	0.948–0.990	0.004*
β‐CTX (ng/mL)	12.127	0.970–151.669	0.053

^* p < 0.05 for consideration of statistical significance.

4. Discussion

4.1. High Level of P1NP as Risk Factor for Trochanteric Fracture Occurrence

Numerous studies have been dedicated to investigate the relationship between bone turnover markers and the risk of hip fracture. One controlled clinical trial enrolling 68 elderly osteoporotic patients matched with equivalent healthy individuals showed that higher serum levels of P1NP and β‐CTX independently associate with high fragile hip fracture risk [17]. Another study focusing on the population of diabetic patients concluded that P1NP and CTX associate with higher risk of fracture in nondiabetes but not in T2D [19]. Previous research of applying BTMs for predicting the occurrence of different subtypes of hip fracture, however, has been relatively scarce. According to the biomechanical testing results of the hip fracture patients enrolled in our study, a significant difference in serum P1NP level exists between patients with trochanteric fractures and femoral neck fractures (p = 0.044, Table 2), which persists after multivariable regression analysis (p = 0.004, Table 3). Our data helps to successfully establish the association between high level of P1NP and occurrence of trochanteric fractures, confirming this BTM as a novel predictor for the prognosis of trochanteric fractures.

4.2. Pathophysiological Mechanism for the Association Between P1NP and Trochanteric Fractures

During the synthesis of new bone matrix, osteoblasts secrete procollagen type I as the precursor of type I collagen. In the process of converting procollagen into collagen fibers extracellularly, P1NP is cleaved off as one of the pro‐peptides at both ends [11]. Thus, it is commonly regarded as an effective indicator for monitoring the rate of bone formation [12]. The P1NP level in blood would increase with enhanced activity of osteoblasts. On the other hand, the cross‐linked structures between type I collagen molecules, specifically C‐terminal telopeptide (CTX), are hydrolyzed and released into the bloodstream during the degradation of bone matrix [11, 12]. This suggests that a higher blood beta‐CTX would be observed when the activity of osteoclasts is elevated.

In general, patients with osteoporosis indicate an imbalance of bone formation and degradation, with the activity of osteoclasts exceeding that of osteoblast [20, 21]. In patients with high‐turnover osteoporosis (common in postmenstrual females), both of the markers above would increase due to elevated bone degradation and formation [22]. In contrast, the low‐turnover osteoporosis patients would exhibit relatively lower values of both P1NP and beta‐CTX due to a reduced bone turnover rate [11].

From the anatomical perspective, the intertrochanteric region contains more cancellous bone compared to femoral neck region [23]. Cancellous bone has a porous structure and abundant blood supply, making it the most active site for bone remodeling (i.e., bone resorption and formation) [24]. The significantly lower BMD observed at the region of trochanteric region observed in the trochanteric fracture group compared to the femoral neck fracture group [4, 14, 15] (Table 2) means that the patient with trochanteric fracture suffered from a more active bone loss process at the “high‐bone turnover” region, which makes the osteoporosis suffered by trochanteric fracture patients to be “high‐turnover” type. Therefore, higher rates of both bone formation and resorption are expected in trochanteric fractures. This is corroborated by the relatively higher average levels of P1NP and β‐CTX observed from the patients with trochanteric fractures (68.12 ± 45.81 ng/mL and 0.60 ± 0.30 ng/mL, respectively) in Table 2.

4.3. Prospects of Clinical Application

While previous studies have already demonstrated higher levels of P1NP and β‐CTX in hip fracture patients compared with healthy population [17], the even stronger association between trochanteric fracture and high turnover osteoporosis suggested in our study provides certain insight into the postoperative treatment strategy of hip fracture. Patients in high bone turnover state suggest a more rapid bone metabolism [25], which may affect the fracture healing process. On one hand, high turnover could mean more active callus formation in the initial stages [26]. On the other hand, the unbalance of this state may lead to excessively rapid bone resorption [27], compromising the holding strength of internal fixation and even increasing the risk of internal fixation failure, cut‐out, or loosening. Thus, compared to patients with femoral neck fractures, those with trochanteric fractures may require more frequent postoperative follow‐up of BMTs to monitor the stability of internal fixation and fracture healing [28]. To establish the specific strategies for postoperative monitoring (i.e., the optimal frequency of P1NP testing and the requisite threshold value of P1NP required for clinical intervention), further clinical researches quantifying the relationship between BMTs level and postoperative complication occurrence in intertrochanteric fractures are expected.

4.4. Baseline Characteristics Identified as Risk Factors for Hip Fracture Subtypes

Other than P1NP, baseline characteristics including height and weight were also identified as independent risk factors for occurrence of different hip fracture subtypes (Table 3). Specifically, patients with taller stature were more likely to have femoral neck fracture, which concords with the results from previous publications [29]. A potential biomechanical rationale proposed by MichaeÈlsson suggests that higher height corresponds to longer hip axis length. It provides a longer moment arm magnifying the force transmitted to the subcapital part of the femoral neck upon fall impact, causing femoral neck fracture [29]. In addition, large weight was confirmed to be risky for trochanteric fractures, which is also suggested by other studies [30, 31]. Maffulli et al. proposed a novel metabolism‐related postulation for this observation, claiming that patients with trochanteric fractures are deficient in nutrient affecting bone metabolism, whose deficiency can occur equally in overweight subjects [31].

4.5. Strengths and Limitations

With the study conducted at two tertiary hospitals of distinct geography and community, patients from a variety of socioeconomic statuses and medical practice environments were enrolled. In this way, the representativeness and generalizability of our study results can be greatly strengthened by the multi‐center study design. However, certain limitations exist in our work. As a retrospective study with cross‐sectional design, selection bias could result. While our rigorous inclusion and exclusion criteria effectively minimized bias, restricting the study to patients targeted for postoperative anti‐osteoporosis treatment might introduce subjects with more severe osteoporosis and higher treatment motivation, thereby compromising the sample's representativeness. While the bias might influence the observed severity of osteoporosis and the bone formation/resorption balance in both fracture groups, the higher overall bone metabolism in the trochanteric fracture group would not be affected, nor would the association between high P1NP level and this fracture type. Besides, the relatively small sample size may have been insufficient to detect a statistical significance underlying the true association between specific hip fracture types and certain secondary outcomes (β‐CTX, PTH, and 25(OH)D, etc.), limiting the generalizability of our findings. Specifically, while our study failed to deduced a statistical difference for PTH and 25OH D between the two hip fracture groups, another study enrolling a total of 310 Japanese patients with fragility hip fractures showed significant lower level of vitamin D and higher level of PTH in patients with trochanteric fractures [32]. Hence, future study with a larger sample size and more advanced study design (prospective cohort, etc.) is required.

4.6. Conclusion

Collectively, our experimental data indicates a clear association between certain BTMs and different subtypes of hip fracture. Together with the BMDs at the regions of trochanter and total hip, P1NP was observed with significant statistical difference between the patients of trochanteric and femoral neck fractures. More importantly, high serum PINP was established as an independent risk factor for the occurrence of trochanteric fractures, suggesting its crucial role in the prognosis and postoperative monitoring of hip fracture.

Author Contributions

Lijia Zhang designed the study, collected and analyzed the data, and composed the manuscript. Jiabao Ju, Yichong Zhang and Chen Xiong assisted in data analyzation. Baozhong Zhang, Jia Zhang and Peng Gao assisted in data acquisition. Yanhua Wang and Xiaomeng Zhang revised the manuscript. Jianguo Zhang supervised the study.

Funding

The authors have nothing to report.

Disclosure

This article declare that all authors listed meet the authorship criteria according to the latest guidelines of the International Committee of Medical Journal Editors and that all authors are in agreement with the manuscript.

Ethics Statement

This study was approved by the Ethics Committees of Peking Union Medical College Hospital and Peking University People's Hospital (I‐25PJ3195/2026PHB058‐001).

Conflicts of Interest

The authors declare no conflicts of interest.

Zhang L., Ju J., Zhang Y., et al., “Assessing P1NP/β‐CTX as Potential Risk Factors Associated With Different Subtypes of Fragile Hip Fracture in Elderly Patients,” Orthopaedic Surgery 18, no. 5 (2026): 944–950, 10.1111/os.70292.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.