Time trends in pediatric fracture incidence in Sweden during the period 1950-2006

Abstract

Background and purpose - Pediatric fracture incidence may not be stable. We describe recent pediatric fracture epidemiology and etiology and compare this to earlier data. Patients and methods - The city of Malmö (population 271,271 in 2005) in Sweden is served by 1 hospital. Using the hospital diagnosis registry, medical charts, and the radiographic archive, we identified fractures in individuals <16 years that had occurred during 2005 and 2006. We also retrieved previously collected fracture data from between 1950 and 1994, from the hospital's pediatric fracture database. We used official population data to estimate period-specific fracture incidence (the number of fractures per 10⁵ person-years) and also age- and sex-adjusted incidence. Differences are reported as rate ratios (RRs) with 95% confidence intervals. Results - The pediatric fracture incidence during the period 2005-2006 was 1,832 per 10⁵ person-years (2,359 in boys and 1,276 in girls), with an age-adjusted boy-to-girl ratio of 1.8 (1.6-2.1). Compared to the period 1993-1994, age-adjusted rates were unchanged (RR =0.9, 95% CI: 0.8-1.03) in 2005-2006, with lower rates in girls (RR =0.8, 95% CI: 0.7-0.99) but not in boys (RR =1.0, 95% CI: 0.9-1.1). We also found that the previously reported decrease in unadjusted incidence in Malmö from 1976-1979 to 1993-1994 was based on changes in demography, as the age-adjusted incidences were similar in the 2 periods (RR =1.0, 95% CI: 0.9-1.1). Interpretation - In Malmö, pediatric fracture incidence decreased from 1993-1994 to 2005-2006 in girls but not in boys. Changes in demography, and also other factors, influence the recent time trends.

Figure 1.

Sex- and age-specifi c fracture incidence in 1-year classes during the period 2005–2006. Data are expressed as number of fractures per 105 person-years, with 95% CI.

Figure 2.

Age-specifi c boy-to-girl rate ratio (RR) with 95% CI per 1-year age class during the period 2005–2006.

Figure 3.

Sex-specifi c unadjusted (A) and age-adjusted (B) fracture incidence during different periods, from 1950 to 2006. Data are expressed as number of fractures per 10⁵ person-years per period of examination, with thick line markers representing the number of years measured in the period. Rate ratios (RRs) between periods of interest are presented with 95% CI on arrows with pointer between the time periods compared.

Figure 4.

Age-specifi c fracture incidence in boys (A) and girls (B) during the periods 1950–1955, 1976–1979, and 2005–2006 in Malmö, Sweden. These 3 periods were selected as to show incidences at study start, during the period with highest incidence, and at study end. Data are incidences in 1-year age classes.