Fractures in children : a population-based study from northern Sweden

Abstract: Fractures and other injuries are notoriously common in childhood and adolescence and a major cause of morbidity. They place a heavy burden on individuals, families, health systems and society. In a population-based study using data from the Injury Database at Umeå University Hospital we analysed injuries, and especially fractures, in Umeå and it´s five surrounding municipalities to increase knowledge on the epidemiology of injuries in children. We found an injury rate requiring visits to the hospital’s emergency department of 110/1000 person-years among those 0-19 years of age. For the same age group, the incidence of injuries resulting in admittance to hospital was 132/104 person-years. The incidence of fractures was 201/104 person-years. The accumulated risk of sustaining a fracture before 17 years of age was 34%, and the peak sex-specific incidence of fractures coincided with the pubertal growth spurt in both sexes. Fractures were more common from May-September when temperatures were warmer and days longer. The most common activity at injury in toddlers and preschool children was play, whereas teenagers were more often injured in sports and traffic-related activities. In an analysis of fracture incidence between municipalities we noted that children and adolescents growing up in rural communities appeared to sustain fewer fractures than their peers in an urban municipality, risk ratio 0.81 (95 % confidence interval 0.76-0.86). We speculated that this observation might be due to differences in behaviour and activities among teenagers in rural and urban communities. In a further analysis we combined data from the Injury Database with socioeconomic microdata. By linking data between individual children and their parents/families we could analyse the relevance of socioeconomic variables and the number of siblings on fracture rate. We observed that children in households with higher incomes sustained significantly more fractures. Comparing the highest and lowest quintiles for income, the rate ratio was 1.40 (1.28-1.54). We also found that having siblings correlated with an increased risk of fractures, rate ratio 1.28 (1.19-1.38). Parents’ education level and family type did not influence the prevalence of fractures significantly. It appeared that the association between variables at the individual level and fractures was stronger than the association between municipality and fractures.  Conclusions: Our results show that the incidence of fractures varies with sex, age, developmental stages, skeletal growth, activities of daily life, and seasons. However, it is also influenced by place of residence, family’s socioeconomic status and having siblings. Refined methods and data allowing an analysis of subsets of fractures and other injuries on the individual/family- and area-level simultaneously should lead to further insights into the possible causal mechanisms underlying observed correlations.