HUMERAL DIAPHYSEAL NONUNION OCCURS MORE COMMONLY WITH ELEVATED BODY MASS INDEX AND SIMPLE FRACTURE TYPES

Fractures of the humeral diaphysis occur in a bimodal distribution and represent 3–5% of all fractures. Nonunion remains a common complication following humeral diaphyseal fractures, with reports of 4% nonunion rates for surgically treated fractures and up to 33% for non-operative treatment. Prior risk factors from small studies or retrospective reviews have suggested advanced age, current smoking status, obesity, and unstable fracture pattern as possible risk factors for nonunion. This study aimed to systematically evaluate risk factors for humeral diaphyseal nonunion. This is a prespecified secondary analysis from a large randomized controlled trial (RCT) comparing open reduction and internal fixation (ORIF) with non-operative treatment (functional bracing) of humeral diaphyseal fractures. Inclusion criteria were patients 18 years or older with a displaced humeral diaphyseal fracture (AO/OTA 12-A, B, C) amenable to both treatments and presentation within 21 days from injury. Patients were excluded if there was an open fracture, multiple injuries, or nerve injury requiring repair. Eligible patients were followed clinically and radiographically for 1-year post-injury. Non-union was defined as lack of radiographic progression of fracture healing on orthogonal radiographs combined with failure to improve clinically. T-tests and chi-square were used to compare those with and without nonunion, and logistic regression was used to evaluate potential risk factors for nonunion. A total of 168 adult patients were enrolled (n=84 per treatment group), with an 85% 1-year follow-up rate. There were no significant differences between treatment groups for age, sex, smoking status, body mass index (BMI), or fracture classification (Table 1). Time to union was achieved earlier in the ORIF group (p=0.02). There were 13 patients diagnosed with nonunion, with 12 (92.3%) being in the non-operative treatment group. There was no difference between age for those who went on to nonunion (46.7 ± 16.2 years) and those who healed without complication (43.3 ± 17.0 years; p = 0.62). Sex distribution between those with (53.8% female) and without nonunion (37.4%) was similar (p = 0.48). All nonunions occurred in simple AO/OTA A-type fractures (A1 = 38.5%, A2 = 23.1%, and A3 = 38.5%). Regression analysis identified elevated BMI (OR = 1.12; 95% CI = 1.02 to 1.25; p = 0.027) as a risk factor for nonunion. There were 32 current smokers, with three patients experiencing nonunions (9.4%; p = 0.16). Surgical intervention for nonunion management in those initially treated non-operatively occurred at an average of 18.3 (± 10.5) weeks after their initial fracture. Our large, multi-centre RCT reported improved function with ORIF for 4–6 months and inferior outcomes have been reported for non-operatively treated humeral fractures requiring subsequent surgical treatment; therefore, early identification of risk factors for nonunion is important in guiding decision making. This large RCT confirms that elevated BMI is a risk factor for nonunion, that all nonunions in this cohort occurred in simple transverse, short oblique, or spiral fracture patterns, and that non-operative treatment carries a minimum of 15% risk for nonunion. For any figures or tables, please contact the authors directly.