Postoperative pressure ulcers (PU) represent a prevalent and costly complication in the management of hip fractures, with reported prevalence rates ranging from 8.8 to 55% (1, 2). These ulcers significantly prolong the length of hospitalization (3) and serve as a critical indicator of poor health outcomes in older adults with hip fractures (4, 5). In older adults undergoing hip fracture surgery, the development of PU is not solely attributed to mechanical pressure, but is also influenced by underlying vulnerabilities such as diminished physiological reserve, impaired tissue repair capacity, and difficulty mobilizing postoperatively (6). Clinical guidelines emphasize the need for early mobilization within 24 h (7); however, delayed mobilization often occurs in older adults due to various factors, which can initiate a cascade of ischemic tissue damage (7).

While prolonged mechanical pressure over bony prominences remains a primary causative factor, with external pressure >32 mm Hg impairing perfusion (8–10), this explanation is insufficient in the context of geriatric hip fracture patients. In this population, the development of PU results from a combination of factors, including systemic vulnerability, chronic inflammation, malnutrition, and comorbidities (5, 6). Immobility-induced ischemia further prolongs exposure to pressure (4), while microvascular dysfunction, arising from age- and disease-related impairments in the hypoxic response, contributes to delayed tissue repair mechanisms (10, 11).

Current research primarily focuses on alleviating mechanical pressure at bony prominence sites to prevent the onset of PU (12, 13). Additionally, a study by Aline et al. has investigated the potential involvement of inflammatory responses in PU development (11). More recently, nutritional status has emerged as a significant factor influencing PU formation, garnering increasing attention in related research (14, 15). However, many of these studies exhibit methodological limitations and often lack generalizability and clinical applicability.

Considering these factors, the present study aims to investigate four well-established hematological markers that reflect systemic inflammation and nutritional status. By integrating these markers into a composite predictive index, this study seeks to develop a reliable, clinically applicable tool for early prediction of PU development and prolonged LOS, ultimately guiding early intervention strategies and improving patient management in hip fracture cases.

In this study, 2,573 electronic medical records were collected between January 2017 and November 2024. After applying the inclusion and exclusion criteria, 1,279 patients were enrolled. The results indicated that 156 patients (12.2%) developed PU postoperatively. These patients were subsequently randomized into a training cohort (896 patients) and a validation cohort (383 patients), as shown in Figure 1. The mean length of hospital stay (LOS) for the entire cohort was 10.69 days. The median age of the participants was 76 years (interquartile range [IQR]: 66.00–82.00), with 39.72% male and 60.28% female patients. Among the four hematologic markers used to construct the predictors, lymphocyte count, blood glucose levels, and albumin levels showed significant differences between the PU and non-PU groups (p < 0.05). Table 1 summarizes the baseline characteristics of all participants.

After identifying covariates significantly associated with PU through univariate logistic regression, these variables were included in the multivariate logistic regression analysis. The detailed information of the adjustment variables is provided in Supplementary Table 1. No multicollinearity was detected between the covariates (VIF < 10). The baseline characteristics of the patients before and after PSM based on GAR thresholds are presented in Supplementary Table 2. Table 2 summarizes the results of both multivariate logistic regression and PSM analyses. The final analysis revealed a complex association between GAR levels and the occurrence of postoperative PU. After adjusting for confounders, GAR was included as a continuous variable in the analysis, showing a significant positive association with the risk of postoperative PU (OR = 1.84, 95% CI: 1.44–2.35). Specifically, each 0.1-unit increase in GAR was associated with an 8.4% increase in PU risk. This indicates that higher GAR values, reflecting elevated glucose and lower albumin levels, strongly predict PU development.

Among all predictors, preoperative glucose levels, GAR, and GNR were significantly associated with PU development (p < 0.001). Among these closely related predictors, GAR exhibited the strongest predictive ability for PU in hip fracture patients (see Figure 2). ROC curve analysis revealed that GAR, as a predictor, had an AUC of 0.720, with a sensitivity of 79.5% and specificity of 55.1% (see Figure 3a). For further characterization parameters of each predictor, refer to Table 3.

In both the overall study population and the validation set, GAR demonstrated moderate predictive ability for PU (Figures 3d,g, AUC of 0.720 [95% CI: 0.674–0.766] and 0.660 [95% CI: 0.615–0.726], respectively). Analysis using restricted cubic spline (RCS) curves confirmed a positive correlation between preoperative GAR levels and PU risk: as the preoperative GAR level increased, the risk of postoperative PU also increased significantly (Figures 3b,e,h). The model was adjusted for all covariates to control for potential confounding factors.

Figures 3c,f,i display the predicted probability of postoperative PU in hip fracture patients at different levels of preoperative GAR. To further explore the relationship between preoperative GAR and PU, a threshold effect analysis was performed. The results showed a significant nonlinear relationship between preoperative GAR and PU (see Table 4 for details). The analysis identified GAR = 2.2 as a key inflection point: when the GAR level was below 2.2, the probability of PU increased significantly with rising GAR levels.

To further examine the relationship between preoperative GAR and LOS, we conducted a GLM regression analysis. After adjusting for confounders, the results showed that each 0.1-unit increase in preoperative GAR was associated with a significant prolongation of LOS by 0.17 days (95% CI: 0.11–0.22) (see Table 5 for details).

Additionally, subgroup analyses were performed to assess the impact of other covariates on the relationship between preoperative GAR levels and PU (see Figure 4). The results revealed a significant interaction between fracture type, preoperative GAR level, and PU incidence (all interaction p-values < 0.05). Specifically, patients with proximal femur fractures exhibited a higher incidence of PU at the same preoperative GAR level. Therefore, clinicians should give special attention to elevated GAR levels when managing these patients, as this factor is strongly associated with an increased risk of postoperative PU.

PU remain a significant clinical challenge, particularly in frail older adults undergoing hip fracture surgery, where the incidence is high and outcomes are often poor (20). This persistent issue is largely attributed to the multifactorial nature of PU development, driven by a complex interplay of metabolic stress, nutritional depletion, impaired tissue resilience, and frailty (6, 11–15). The glucose-albumin ratio, which integrates metabolic stress (elevated blood glucose) and nutritional depletion (hypoalbuminemia), provides a novel and valuable tool for assessing PU risk in this vulnerable population, as both factors significantly impair tissue healing capacity.

This retrospective study systematically investigates hematological indicators of inflammation and nutritional status, both of which are closely linked to PU development. We identified key predictors most strongly associated with the occurrence of postoperative PU in older adults undergoing hip fracture surgery, thereby providing clinicians with a reliable and easily monitored predictive tool for PU risk. Our findings demonstrate an independent correlation between preoperative GAR levels and PU occurrence in hip fracture patients, with higher GAR levels significantly increasing the risk of PU development. Notably, the association between preoperative GAR levels and PU occurrence was even more pronounced in the proximal femur fracture subgroup, which is typically associated with frailty due to osteoporosis and other geriatric factors. ROC curve analysis revealed that the optimal GAR threshold was 1.65, with an AUC of 0.72, proving effective in predicting the risk of postoperative PU in hip fracture patients. However, the specificity of 55.1% suggests a relatively high false-positive rate, which may limit the clinical applicability of GAR as a stand-alone predictor. A higher false-positive rate could result in unnecessary interventions or monitoring for patients who may not develop PU, leading to increased healthcare costs and resource utilization. To address this limitation, we suggest that GAR could be used in combination with other clinical screening tools, such as the Braden Scale or Norton Scale, which evaluate additional factors like sensory perception, mobility, and friction/shear. By combining GAR with these tools, clinicians could improve specificity and reduce the number of false positives, enhancing the overall accuracy of PU risk prediction. Moreover, incorporating other clinical assessments, such as patient frailty or nutritional assessments, could further refine the predictive model and help guide targeted interventions. These findings underscore the potential of preoperative GAR levels as a predictor of PU risk. Additionally, threshold analysis provides clinicians with a critical inflection point of 2.2 for preoperative GAR levels, which can help enhance clinical vigilance.

Numerous studies have demonstrated that elevated blood glucose levels significantly increase the risk of developing surgery-related PU in patients (21, 22). Hip fracture patients are typically mobilized within 24 h of surgery, in accordance with current best practices, which significantly reduces the risk of immobility-related complications, including PU. However, those who are not mobilized early due to severe frailty or other medical conditions remain at a heightened risk of PU development (7). The stress associated with hip fractures and subsequent surgeries often leads to elevated blood glucose, which, in turn, contributes to delayed wound healing and an increased risk of PU development (23, 24). Hyperglycemia impairs leukocyte function, particularly during the inflammatory phase of wound healing, by inhibiting neutrophil migration, adhesion, and cytokine release (25). This compromises the immune response and prolongs the inflammatory phase, further delaying PU healing.

Moreover, hyperglycemia interferes with collagen deposition by impairing collagen synthesis and fibroblast function (26). These alterations result in delayed wound healing and increased susceptibility to ischemic necrosis, particularly at sites of sustained pressure, such as bony prominences (27, 28). Hyperglycemia also induces oxidative stress and the formation of advanced glycosylation end products (AGEs), which exacerbate tissue damage, impair microcirculation, and worsen ischemia in pressure ulcer-prone areas (29–32). This impaired tissue repair process, further exacerbated by hyperglycemia, highlights the critical need for better metabolic management in frail patients to mitigate PU development.

In addition to metabolic stress, nutritional status plays a critical role in PU development (33, 34). Malnutrition, on one hand, results in a reduction of connective tissue, thereby weakening the skin’s resistance to external pressure and lowering tissue tolerance, which in turn increases the risk of PU formation (35). In malnourished individuals who develop severe PU, the wound healing process is often significantly delayed. Among patients with hip fractures, malnutrition is particularly prevalent, rendering this population at higher risk for PU development (36). On the other hand, undernutrition is typically accompanied by a reduction in adipose tissue, diminishing the cushioning and protective functions over bony prominences. In addition, conditions such as generalized weakness, dehydration, and edema are frequently observed in malnourished patients, collectively impairing the skin’s barrier function, reducing mobility, and compromising immune defenses, thereby further elevating the risk of PU occurrence (37–40).

Serum albumin is widely used as a biomarker of nutritional status and plays a vital role in maintaining tissue integrity during mechanical loading (41–43). Low serum albumin levels are strongly associated with PU risk, as they reflect both reduced tissue repair capacity and impaired resistance to pressure-induced tissue damage (44, 45).

Frailty is a multidimensional syndrome characterized by reduced physiological reserve and increased vulnerability (6), often seen in older adults, particularly those undergoing hip fracture surgery (46). It encompasses a combination of factors, including impaired mobility, poor nutritional status, and weakened immune responses (47–49). Although frailty was not directly assessed in this study, it has been proposed as a potential unmeasured confounder influencing the relationship between GAR and postoperative PU risk. Frailty is prevalent among older adults with hip fractures and has been linked to poorer surgical outcomes and an increased susceptibility to complications, including PU. For instance, a recent meta-analysis demonstrated that frailty was associated with poor postoperative outcomes following hip fractures (50), and studies by Kistler et al. (51) and Pizzonia et al. (52) showed associations between frailty and short- and long-term hip fracture outcomes, respectively. However, due to the absence of formal frailty measures in our dataset, we were unable to draw definitive conclusions about its role in PU development. Future studies should explicitly assess frailty using validated scales, such as the Frailty Phenotype or the Clinical Frailty Scale, to gain a deeper understanding of how frailty interacts with metabolic and nutritional factors in predicting PU. This approach would enable more targeted interventions and enhance the accuracy of predictive models for PU development in frail patients.

In conclusion, this study comprehensively analyzed the role of GAR in predicting the development of PU in older adults undergoing hip fracture surgery. By combining hyperglycemia and hypoalbuminemia, the GAR provides a new, clinically relevant predictor that integrates metabolic and nutritional factors that are important for tissue healing. GAR not only facilitates the identification of at-risk patients but also lays the foundation for tailored interventions, which should be based not solely on objective indicators but also on individualized care. Future research should further investigate the direct role of frailty in PU risk and explore how GAR can be incorporated into clinical practice to improve the management and outcomes of older adults undergoing hip fracture surgery.

Preoperative GAR levels are a reliable predictor of postoperative PU development in older adults undergoing hip fracture surgery, exhibiting a significant nonlinear dose–response relationship, particularly in those with proximal femur fractures. Specifically, each 0.1-unit increase in the GAR index was associated with an 8.4% increase in the risk of PU development. This relationship is reflective of the combined effects of elevated glucose levels and lower albumin levels. Furthermore, preoperative GAR levels were also correlated with patients’ length of stay, with a notable increase of 0.17 days in LOS for every 0.1-unit rise in preoperative GAR. In conclusion, timely medical intervention and precautionary measures are essential when preoperative GAR levels fall within the range of 1.65 to 2.2.