Answers are grounded in this paper and cite their sources.
OBJECTIVES: To evaluate the association between preexisting anemia and proximal humerus fracture (PHF) nonunion and to determine whether risk varies by anemia type and associated co-occurring conditions. METHODS: Design: Retrospective cohort study. SETTING: Administrative claims database study using commercial and Medicare fee-for-service data from 2016 to 2023. PATIENT SELECTION CRITERIA: Adults aged ≥19 years who sustained a PHF (OTA/AO 11) and had continuous insurance enrollment for ≥6 months after PHF were included, allowing entry into the at-risk period for nonunion. Anemia diagnosed during the 1-year period before the fracture was the primary exposure and was evaluated in three ways: (1) presence or absence of any anemia, (2) number of different co-occurring anemia conditions (range, 0-14), and (3) clinically relevant anemia categories, including nutritional, hemolytic, and aplastic/other anemias. OUTCOME MEASURES: The incidence rate (IR), IR ratio (IRR), and hazard ratio (HR) was estimated for the primary outcome of nonunion of the PHF from 6 to 18 months post-fracture, before and after accounting for potential confounders. Additionally, sub-group estimates were generated for patients 40, 50, 60, 70, and 80 years of age. RESULTS: There were 25,109 adults with a PHF and any type of anemia (mean [SD] age=76.0 [12.8] years; 78.2% women; 6.8% commercially-insured) and 64,389 with a PHF without anemia (68.3 [17.0] years; 74.6% women; 26.7% commercially-insured). The crude IR and IRR of PHF nonunion was significantly higher for the full anemia cohort (IR=6.0 [95%CI=5.1-6.8]; IRR=1.37 [95%CI=1.15-1.62) and for each anemia sub-cohort (IR [IRR] for 1, 2, and ≥3 co-occurring anemic conditions was 5.5 [1.25], 6.2 [1.42], and 8.1 [1.85], respectively) than the non-anemia cohort. Compared to the non-anemia cohort, the adjusted HR was not significantly different for the full anemia cohort (HR=0.97; 95%CI=0.80-1.17) or anemia sub-cohorts (HR range, 0.85-1.00). However, age strongly modified all associations (P for interaction, all ≤0.001). For all three types of anemia exposure (yes/no, count of co-occurring, clinically distinct groups), the HR of nonunion was elevated at the younger ages (age 40: HR=2.06 [95% CI=1.33-3.18]; age 50: HR=1.63 [95% CI=1.18-2.27]; age 60: HR=1.30 [95% CI=1.02-1.65]) and declined with older age, close to 1.00 by 70 years old and below 1.00 by 80 years old. CONCLUSIONS: Anemia was associated with a higher rate of nonunion of the proximal humerus fracture among younger and middle-aged adults (HR=2.06, 1.63, and 1.30 for 40, 50, and 60 year old patients, respectively), but not for older adults. LEVEL OF EVIDENCE: Prognostic Level III.
No takes yet. Share an insight, caveat, or question.
Nakken et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: