D. Q. Ma, G. Jones. Menzies Research Institute, Hobart, Australia.
Fractures in younger life are very common especially those involving the upper limb. However, the causes of these fractures are poorly understood. The aim of this population based case-control study was to evaluate the association of both bone dependent and independent factors with upper limb fracture risk in children 9-16 years of age. Areal (aBMD) and apparent (BMAD) bone mineral density were measured by dual energy X-ray absorptiometry (DXA). Skeletal age (SA) and metacarpal index (MI) was determined by hand radiograph. Types and patterns of physical activity, risk taking and diet were assessed by interview-administered questionnaires. Coordination was measured using the 8-point movement ABC. A total of 321 fracture cases and 321 randomly selected age- and gender- matched controls were recruited. Fracture sites were as follows: hand: n=91, wrist and forearm: n=190, upper arm: n=40. Conditional logistic regression analysis showed wrist and forearm fracture risk was significantly associated with lumbar spine BMAD (OR: 1.52/SD reduction, 95% CI 1.18-1.96), MI (OR: 1.43/SD reduction, 95% CI 1.12-1.79), television, computer and video viewing (OR: 1.58/category, 95% CI 1.14-2.20), light physical activity (OR: 0.81/unit, 95% CI 0.65-1.00), cola consumption (OR: 1.43/category, 95% CI 1.03-1.97) and walking backwards score (OR: 1.17/unit, 95% CI 1.02-1.34). Hand fracture was significantly associated with the difference between skeletal age and chronological age (OR 1.45/year reduction, 95%CI 1.05-1.96), high-risk sport participation (1.43/sport, 95% CI 1.08-1.88) and total risk taking score (OR: 1.94/unit, 95% CI 1.01- 3.72). Upper arm fracture risk was only significantly associated with high-risk sport participation (OR: 2.73/sport, 95% CI 1.27-5.84) but conclusions may be limited by sample size considerations. Obesity or overweight was not associated with any fracture type. For total fractures, there was a gender discordant effect of sports participation with an increased fracture risk in boys and decreased fracture risk in girls which reached statistical significance for total, contact, non-contact and high-risk sports participation as well as four individual sports (soccer, cricket, surfing and swimming). In conclusion, both bone dependent and independent factors are important determinants of upper limb fracture risk in children. There is heterogeneity of cause for both gender and different fracture sites which will necessitate different approaches to prevention.