Cluster analysis demonstrates co-existing sites of fragility fracture and associated comorbidities.

UNLABELLED: We undertook cluster analysis in 11,003 patients who had sustained ≥ 1 fragility fracture, to find associations between fracture sites and comorbidities. We identified three distinct groups of fracture sites and four clusters of fractures and comorbidities. Knowledge of factors associated with fracture sites will aid prophylaxis in at-risk patients.

INTRODUCTION

Fragility fracture (FF) prevalence is increasing. Subsequent fractures lead to greater morbidity and mortality. Few data are available on the association between FF sites and comorbidities.

OBJECTIVES

1. Establish the most common sites of FF and clusters within patients. 2. Identify patterns of co-existing FF and associated comorbidities.

METHODS

We retrospectively reviewed clinical records of patients undergoing bone mineral density estimation at a district hospital in North-West England, 2004-2016, identifying those who had sustained ≥ 1 FF. Demographics, FF site(s), comorbidities, and medications were recorded. Cluster analysis was performed on fracture sites alone, and sites and comorbidities, using Jaccard similarity coefficient. Results were plotted on a dendrogram and divided into clusters.

RESULTS

Of 28,868 patients, 11,003 had sustained ≥ 1 FF, 84.6% female, with overall mean age 67.5 years and median T-score - 1.12 SD. FF of the forearm was more frequent (n = 5045), most commonly co-existing with tibia/fibula fractures. Three FF site clusters were identified: ankle and elbow; forearm, tibia/fibula, ribs and spine; and pelvis, femur and humerus. When including comorbidities, four clusters were identified: forearm, tibia/fibula, spine, associated with family history of FF, smoking, corticosteroids and bisphosphonates; pelvis associated with hyperparathyroidism, PMR, coeliac disease and HRT; femur and humerus associated with IBD and RA; and ribs associated with alcohol and hyperthyroidism.

CONCLUSION

Cluster analysis demonstrated three fracture site clusters, and four subgroups of FF sites and comorbidities. Cluster analysis is a novel method to evaluate comorbidities associated with FF sites. Knowledge of factors associated with FF sites will aid prophylaxis in at-risk patients.