Study identifies fragility fracture risk factors in peritoneal dialysis patients

Patients on peritoneal dialysis face a substantial burden of fragility fractures, most of which turn out to be silent vertebral fractures that would go undetected without targeted screening, according to a retrospective cohort study published in Archives of Osteoporosis. The study, led by Mariana Diz-Lopes and Bernardo Fernandes of the Unidade Local de Saúde de São João in Porto, Portugal, also found that the widely used FRAX® fracture risk assessment tool performs poorly in this population, substantially underestimating the likelihood that these patients will break a bone.

Peritoneal dialysis is a form of renal replacement therapy in which the lining of the patient’s abdomen filters waste products from the blood, offering an alternative to hemodialysis for many people with end-stage kidney disease. Although fractures are a recognized and serious complication of chronic kidney disease in general, contributing to pain, disability, institutionalization, and death, patients treated with peritoneal dialysis have been comparatively understudied. Most fracture research in dialysis populations has centered on hemodialysis, leaving clinicians with limited evidence to guide fracture prevention decisions for the peritoneal dialysis cohort. The new study was designed to fill that gap by quantifying how often fragility fractures occur during peritoneal dialysis follow-up, identifying the factors that predict them, and testing whether a standard risk calculator works in this setting.

The research team conducted a single-center retrospective cohort study at their Portuguese institution, including adult patients who had been receiving peritoneal dialysis for at least twelve months. In total, 325 patients were analyzed, with a mean age of 53 years and a majority, 57 percent, being male. Fragility fractures occurring during peritoneal dialysis follow-up were identified through two complementary routes: systematic review of clinical records and review of spine radiographs. This dual approach mattered, because it allowed the investigators to capture vertebral fractures that had never caused symptoms and therefore never prompted clinical attention. Vascular calcification, a hallmark of the disordered mineral metabolism that accompanies kidney failure, was quantified using the Adragão vascular calcification score, a simple radiographic tool originally developed for hemodialysis patients that evaluates calcification in the pelvic and hand arteries on plain films.

To evaluate whether existing risk tools could be applied to these patients, the researchers calculated FRAX® probabilities for each individual without incorporating bone mineral density measurements. FRAX®, the Fracture Risk Assessment Tool developed by the World Health Organization collaboration led by John Kanis and colleagues, integrates clinical risk factors such as age, sex, prior fracture, parental hip fracture history, smoking, glucocorticoid use, and secondary causes of osteoporosis to estimate ten-year fracture probability. The investigators then used logistic regression to determine which variables independently predicted fracture occurrence, and constructed receiver operating characteristic curves to judge how well FRAX® discriminated between patients who fractured and those who did not. Statistical analyses were performed by the joint first authors, Diz-Lopes and Fernandes, who contributed equally to the work.

The headline finding was that fragility fractures were far from rare. Thirty-two of the 325 patients, or 9.8 percent, sustained at least one fragility fracture during peritoneal dialysis follow-up. Strikingly, the majority of these were asymptomatic vertebral fractures, accounting for 23 of the 32 fracture events, or 7.1 percent of the entire cohort. In other words, roughly seven in every hundred peritoneal dialysis patients in the study had vertebral fractures that were detectable on radiographs but had produced no clinical symptoms. Only a minority of fractures were the clinically obvious events, such as hip or other peripheral fractures, that typically bring a dialysis patient to medical attention.

When the researchers ran multivariable logistic regression to isolate independent predictors of fracture, two factors emerged. A history of previous fracture was the strongest, conferring nearly a five-fold increase in odds of sustaining a new fragility fracture during follow-up, with an odds ratio of 4.85 and a 95 percent confidence interval of 1.22 to 19.27. This aligns with a large body of literature, including a 2023 meta-analysis used to update FRAX® itself, establishing prior fracture as one of the most potent predictors of subsequent fracture in the general population. The second independent predictor was a higher Adragão vascular calcification score, with an odds ratio of 2.10 and a 95 percent confidence interval of 1.10 to 4.03, meaning each increment in the score roughly doubled the odds of fracture.

The link between vascular calcification and skeletal fragility is a recurring theme in dialysis research. Prior studies in hemodialysis populations, including work by Fusaro and colleagues using quantitative morphometry and by Rodriguez-Garcia and colleagues in the Asturias cohort, have documented strong associations between vascular calcifications, vertebral fractures, and mortality. The prevailing explanation, articulated by Cannata-Andia and colleagues, is that the connections between vascular and bone health run deep: the same disturbances of mineral metabolism, inflammation, and bone turnover that drive calcium deposition in arteries can also impair bone quality. Some investigators, including Adragao’s group, have shown that higher mineralized bone volume is associated with lower plain radiographic vascular calcification scores, suggesting that the two processes may be two faces of a shared underlying pathology. The new peritoneal dialysis data extend this vascular-bone axis to a population where it had not been well characterized.

The performance of FRAX® in this cohort was the study’s most sobering result. The tool showed only modest discriminatory ability for any fragility fracture, with an area under the receiver operating characteristic curve of 0.69. An AUC of 0.69 falls well short of what is generally considered good discrimination and implies limited practical utility for individual decision-making. More concerning still, sensitivity at the standard intervention thresholds was very low, meaning that the great majority of patients who actually fractured would not have been flagged for treatment under existing guidelines. Notably, this remained true even when the researchers classified chronic kidney disease as a secondary cause of osteoporosis within the FRAX® algorithm, the most favorable adjustment available to them. The conclusion is that FRAX® systematically underestimates fracture risk in peritoneal dialysis patients.

This finding has context in a growing literature on FRAX® and kidney disease. Studies by Whitlock and colleagues in 2019 showed that FRAX® can predict fracture risk in chronic kidney disease populations, while other work by Przedlacki and colleagues in hemodialysis and by Hayashi and colleagues in Japanese hemodialysis patients has produced more mixed results. A recent overview of systematic reviews by Cruz-Priego and colleagues in 2025 examined the predictive capacity of fracture risk assessment tools broadly and highlighted persistent questions about their transportability across populations. The FRAX® algorithm was derived from population-based cohorts that excluded or underrepresented patients with advanced kidney disease, whose bone disease differs fundamentally from postmenopausal or age-related osteoporosis. In dialysis patients, the spectrum of skeletal pathology, encompassing high-turnover bone disease, adynamic bone disease, and mixed uremic osteodystrophy under the umbrella of chronic kidney disease-mineral and bone disorder, is far more heterogeneous than the bone loss captured by standard osteoporosis models.

The implications for clinical practice are significant. The European consensus statement on the diagnosis and management of osteoporosis in chronic kidney disease stages G4 to G5D, published by Evenepoel and colleagues in 2021, acknowledged the difficulty of distinguishing the causes of bone fragility in advanced kidney disease, and recent KDIGO Controversies Conference conclusions from Ketteler and colleagues in 2025 similarly emphasized unresolved questions in chronic kidney disease-mineral and bone disorder. Given that bone mineral density by dual-energy X-ray absorptiometry, the cornerstone of conventional osteoporosis assessment, may not fully capture bone quality in uremic patients, and given that DXA studies in peritoneal dialysis patients have shown discordance between lumbar spine and femoral neck measurements, the study’s authors argue that improved fracture risk assessment strategies tailored specifically to peritoneal dialysis patients are needed. The identification of vascular calcification as an independent risk factor is particularly appealing from a practical standpoint, because the Adragão score relies on simple plain radiographs already obtained for other reasons in many dialysis patients, making it a feasible and inexpensive addition to risk stratification.

The authors are careful to frame their findings within the limits of the study design. As a retrospective, single-center cohort of 325 patients at one Portuguese institution, the study cannot establish causation, is subject to the ascertainment biases inherent in clinical record review, and may not generalize to peritoneal dialysis populations with different demographic or clinical characteristics, or to healthcare systems with different screening practices. The confidence interval around the odds ratio for prior fracture is wide, from 1.22 to 19.27, reflecting the relatively small number of fracture events and the uncertainty that accompanies estimates derived from them. Verbal informed consent was obtained from subjects given the retrospective design, and ethical approval was granted by the local ethics committee of Hospital de São João under approval number CE 90/2024. The authors declared no conflicts of interest and no external funding. Data supporting the findings are available from the corresponding author upon reasonable request.

Even with these caveats, the study delivers a clear message to nephrologists and rheumatologists who care for peritoneal dialysis patients. Fragility fractures are common in this group, driven disproportionately by vertebral fractures that remain invisible unless deliberately sought, and the standard risk calculator that guides osteoporosis treatment decisions in the general population cannot be trusted here. Patients with a prior fracture or with vascular calcification on plain radiographs deserve heightened vigilance. Until tools validated specifically for the peritoneal dialysis population are developed, clinicians may need to rely on clinical judgment, attention to prior fracture history, and simple radiographic markers of vascular calcification to identify those at greatest skeletal risk. As the population of people living long-term on dialysis grows, and as reports of increasing hip fractures in both hemodialysis and peritoneal dialysis patients accumulate, closing the evidence gap in bone health for peritoneal dialysis patients becomes an increasingly urgent priority for the kidney and bone research communities alike.