Incidence of Childhood Distal Forearm Fractures Over 30 Years: Title and subTitle BreakA Population-Based Study FREE

Previous studies have shown that the incidence of distal forearm fractures in children peaks during early adolescence around the time of the pubertal growth spurt.^{1 - 3} This observation has been explained on the basis of a transient increase in cortical porosity that results from enhanced bone turnover in response to greater calcium demand at the time of maximal longitudinal bone growth.⁴ Such a process might be especially problematic given the thin cortex of the metaphyseal region of the distal radius,⁵ and growth-related structural changes in the metaphysis further compromise bone strength in the distal forearm.⁶

Adolescence is also characterized by the increase in physical activity necessary to maximize skeletal mass.^{7 - 9} Thus, Parfitt⁴ has suggested that forearm fractures in adolescence are "an inescapable consequence of an appropriate level of physical activity, and [are] the price that has to be paid in order to maximize bone accumulation during growth and minimize fracture risk in old age." To some extent, then, forearm fractures in adolescence may represent a transient imbalance between physical activity and acquisition of bone mass during puberty. This relationship may have changed over the years, due either to changing patterns of physical activity or to alterations in acquisition of bone mass, related perhaps to changing dietary habits.

The incidence of distal forearm fractures among Rochester, Minn, residents younger than 35 years in 1969-1971 was reported previously.¹⁰ The purpose of the present investigation was to estimate the incidence of such fractures among comparable Rochester residents 10, 20, and 30 years later in 1979-1981, 1989-1991, and 1999-2001. In addition, we extended the newest study in Rochester to include the remainder of Olmsted County, which is largely rural. Thus, in addition to providing updated incidence rates for distal forearm fractures in adolescence, and a comparison of urban with rural rates, we also tested for possible changes in incidence over time.

Population-based epidemiologic research can be conducted in Rochester because medical care is virtually self-contained within the community and there are relatively few provider organizations. Through the medical records linkage system provided by the Rochester Epidemiology Project, the details of almost all of the medical care delivered to the residents are available for study.¹¹ We used this unique database to identify all distal forearm fractures that occurred among community residents younger than 35 years during the 3-year periods 1979-1981, 1989-1991, and 1999-2001, and we rereviewed previously published data for 1969-1971.¹⁰ Only a quarter of patients with forearm fracture are hospitalized,¹² but it was possible in our data system to identify those treated solely on an outpatient basis. Moreover, the indexing system is very redundant.

The complete (inpatient and outpatient) medical records were reviewed for all local residents with any diagnosis attributable to rubrics 813.4 through 813.9 in the International Classification of Diseases, Ninth Revision, Clinical Modification.¹³ In accordance with Minnesota privacy law that took effect in 1997,¹⁴ we had authorization to review the 265 cases reported in 1969-1971 and all 944 potential cases in 1979-1981, 1595 of 1610 (99%) in 1989-1991, and 2362 of 2383 (99%) in 1999-2001. The study was approved by the Mayo institutional review board.

Patients were excluded if they were not residents at the time of fracture, if their fracture occurred outside the time frame of the study, or if it was a fracture of the shaft rather than the distal forearm. Except for 36 patients who declined to provide the research authorization, fracture ascertainment is believed to be complete,¹⁵ and ascertainment methods were consistent over time.

All fractures, including torus fractures, were radiographically confirmed at the time of injury. Fractures were classified according to etiology using information about each event that was recorded in the medical record: those caused by a specific pathological process (eg, osteogenesis imperfecta), those due to serious trauma (a category combining fractures clearly resulting from severe trauma, eg, motor vehicle crashes or falls from a height, with an intermediate category made up of recreational and occupational injuries), and those due to moderate trauma (by convention, the equivalent of a simple fall from standing height or less).

Olmsted County is a Metropolitan Statistical Area and contains 1 centrally located city, Rochester, and 13 small towns or villages. We determined that each child was a resident of either Rochester or the remainder of Olmsted County on the date of fracture. Rochester had about 86 000 residents in 2000 and was considered "urban" for the purposes of this study. The remainder of the county is mostly farming country with a 2000 population of 38 000. Although no one in Olmsted County is more than 23 miles from sophisticated medical care available at Mayo Clinic, the population density of the area outside of Rochester is only 61.7 people per square mile. Thus, the balance of Olmsted County was considered "rural" for the purposes of this study, even though 4% of the overall county population lived on the outskirts of Rochester in 2000 and another 7% lived in one small town.

In calculating incidence rates, the entire population of Rochester (or Olmsted County) younger than 35 years was considered to be at risk. Denominator age- and sex-specific person-years were estimated from decennial census data.¹⁶ In order to estimate the underlying variability, it was assumed that, given a fixed number of person-years, the number of fracture cases follows a Poisson distribution. This allowed for the estimation of SEs and the calculation of 95% confidence intervals (CIs) for the incidence rates. Overall rates were directly age- or age-sex–adjusted to the population distribution of US whites younger than 35 years in 2000. The SEs and CIs for the adjusted rates were based on the same assumption as above.

The relationships of crude incidence rates to age, sex, and time period (1969-1971 vs 1979-1981 vs 1989-1991 vs 1999-2001) or age, sex, and residence (urban vs rural) were assessed using generalized linear models assuming a Poisson error structure.¹⁷ Such models fit the natural logarithms of the crude incidence rates as linear combinations of sex, age group, residence, and time period. In our modeling, we first fit sex and age group, testing for an interaction between them. Differences between the 4 time periods were assessed using the subset of Rochester data alone. Using the 1990-2001 data for all of Olmsted County, the difference between residence (urban vs rural) was then assessed. Model fit was evaluated using the model deviance, which is a measure of how well the observed and predicted incidence rates agree. The model fits the data reasonably well if the expected value of the deviance is approximately equal to its degrees of freedom. The distributions of moderate and serious trauma fractures between Rochester residents and residents of rural Olmsted County in 1999-2001 and between Rochester residents in the various time periods were compared using the Pearson χ² statistic. All analyses were carried out in SAS version 8.02 (SAS Institute Inc, Cary, NC), with P<.05 used to determine statistical significance.

There were 1458 total fractures in these time periods. The number of fractures increased from 265 in 1969-1971 to 469 in 1999-2001 (Table 1). The overall annual age- and sex-adjusted incidence of distal forearm fractures per 100 000 Rochester residents younger than 35 years increased from 263.3 (95% CI, 231.1-295.4) in 1969-1971 to 322.3 (95% CI, 285.3-359.4) in 1979-1981 and to 399.8 (95% CI, 361.0-438.6) in 1989-1991 before leveling off at 372.9 (95% CI, 339.1-406.7) in 1999-2001. Overall, comparably age- and sex-adjusted annual incidence rates were 42% greater (P<.001) in 1999-2001 compared with 1969-1971. The rates per 100 000 were 32% greater among male residents in 1999-2001 compared with 1969-1971 (409.4 [95% CI, 359.9-459.0] vs 309.4 [95% CI, 259.3-359.5]) and 56% greater among female Rochester residents in 1999-2001 (334.3 [95% CI, 288.6-380.1] vs 214.6 [95% CI, 174.9-254.4]). The increase in incidence was statistically significant for both male (P = .01) and female residents (P<.001), but the rate of increase did not differ significantly between the 2 groups (P = .31). As is evident from Table 1, the discrepancy was almost entirely due to differences in fracture rates for Rochester residents younger than 20 years. However, the overall age-specific pattern was relatively unchanged from one time frame to the next (Table 1).

To obtain additional insight into the secular change in incidence, we compared causes of the distal forearm fractures among Rochester residents in the 4 time periods (Table 2). There was a slight decline in the relative proportion of fractures due to simple falls, but this was related to an absolute increase in fractures due to serious trauma. In fact, the age- and sex-adjusted annual incidence of distal forearm fractures due to moderate trauma in 1999-2001 was very similar to the rate in 1969-1971 (72.7 per 100 000 [95% CI, 56.2-89.2] vs 69.6 per 100 000 [95% CI, 55.1-84.1]). The increase in fractures due to serious trauma, eg, from 179.0 per 100 000 (95% CI, 152.2-205.8) in 1969-1971 to 293.8 per 100 000 (95% CI, 263.7-323.9) in 1999-2001, completely accounted for the overall change in fracture rates between the 2 time periods. This increase, in turn, was explained by an increase in the fractures associated with recreational activities (Table 2). The annual incidence of fractures due to recreational injuries doubled, from 114.7 per 100 000 (95% CI, 93.5-135.9) in 1969-1971 to 228.6 per 100 000 (95% CI, 202.0-255.2) in 1999-2001 (P<.001). However, the distribution of recreational injuries by type was fairly comparable across the time periods (Table 3), although there was some increase associated with skiing and skating (especially skateboarding and rollerblading) among boys and with basketball, soccer, and skating among the girls. Injuries on playground equipment, included in the category of other recreational causes, accounted for 10% of all distal forearm fractures in 1999-2001 and for a similar 6% in 1969-1971.

In the most recent 3-year period, 1999-2001, we had the opportunity to compare urban with rural distal forearm fractures and to estimate the overall incidence for Olmsted County as a whole. All together, 661 county residents younger than 35 years experienced a total of 690 fractures, giving an overall age- and sex-adjusted annual incidence rate of 361.6 per 100 000 (95% CI, 334.6-388.6) for this age group. There were 385 fractures among male and 305 among female residents of Olmsted County. The annual incidence among male residents was 394.1 per 100 000 (95% CI, 354.7-433.6) compared with 327.2 per 100 000 (95% CI, 290.4-363.9) among female residents, for a male to female ratio of age-adjusted incidence rates of 1.2:1 (P = .01). The incidence of fractures varied strikingly with age in both sexes (Figure 1), with the highest rate among girls between the ages of 8 and 11 years and boys between 11 and 14 years. Peak annual rates were 1277 per 100 000 for girls aged 11 years and 1536 per 100 000 for boys aged 12 years. All together, 632 individuals (95% of the female and 96% of the male residents) experienced a single distal forearm fracture during the 3-year study period, but 8 male and 6 female residents had 2 separate fractures while 7 male and 8 female residents experienced bilateral distal forearm fractures. For 332 male (90%) and 259 female residents (89%), the initial distal forearm fracture observed during the study period was the first such fracture that the individual had ever experienced.

The overall age- and sex-adjusted annual incidence rate in 1999-2001 was 10% greater among the residents of the central city of Rochester (372.9 per 100 000; 95% CI, 339.1-406.7) compared with those living in the rural portion of Olmsted County (340.2 per 100 000; 95% CI, 294.7-385.8) as shown in Table 4 (P = .58). Age-adjusted rates were 13% greater among urban than among rural male residents (P = .81) and 5% greater comparing female residents of Rochester with female residents of rural Olmsted County (P = .58), but neither difference was statistically significant. There was also no significant difference in the distribution of causes when comparing Rochester residents with the residents of rural Olmsted County (P = .45). All together, 79% of the distal forearm fractures among young Olmsted County residents in 1999-2001 resulted from serious trauma. Again, this was mostly in conjunction with recreational activities, which accounted for 420 fractures. An additional 72 fractures were caused by falling from greater than standing height, while 18 were due to motor vehicle crashes and 38 to occupational and other injuries. One hundred twenty-six fractures (18%) were due to moderate trauma. The final 16 forearm fractures could not be confidently ascribed to a specific injury.

These results demonstrate a significant increase in the incidence of distal forearm fractures among children in Olmsted County, Minnesota, between 1969-1971 and 1989-1991, with a leveling off in 1999-2001. Although the increase could be partly due to more aggressive treatment of these injuries in recent years, this is unlikely to account for an overall increase in distal forearm fractures of 56% in female and 32% in male residents. Moreover, investigators in other settings have made similar observations. Landin¹ documented an overall increase in distal forearm fractures of roughly 60% in girls and 35% in boys in Malmö, Sweden, between 1950 and 1979, and a subsequent study revealed a further 31% increase in girls but no change in boys between 1975-1979 and 1993-1994.¹⁸ Likewise, Danish investigators estimated that fracture rates had increased 33% in girls and 5% in boys between 1975-1979 and 1985.² Overall, Jónsson and colleagues¹⁹ reported a 60% increase in distal forearm fractures among Swedish boys aged 10 to 19 years between 1953-1957 and 1991-1992, although they saw no increase among girls in Malmö. In the comparable time frame, 1989-1991, annual Rochester rates for 10- to 19-year-old boys and girls were similar to those reported for Malmö (for boys, 85 in Rochester vs 57 per 10 000 in Malmö and, for girls, 34 vs 29 per 10 000, respectively). By contrast, there was no change in rates of distal forearm fracture among Rochester residents 35 years and older between 1969-1971 and 1989-1991.¹⁵ This has been confirmed by other investigators who found no change in incidence even among children.²⁰

As have others,^{1 ,3 ,19 ,21 - 25} we documented a greater incidence of distal forearm fractures in boys than in girls. There was little difference between the sexes before age 5 years, but fracture rates rose rapidly thereafter to peak among girls aged 8 to 11 years and boys aged 11 to 14 years. These results are consistent with those of other researchers^{2 - 3} who have found that the peak incidence of distal forearm fractures spans the pubertal growth spurt, which reaches its maximum around age 12 years in girls and age 14 years in boys.⁸ Moreover, the greater height velocity in boys than in girls (4.9 vs 2.9 cm per year) may help explain the excess of distal forearm fractures in boys,³ although different injury patterns may also contribute. Generalized deficits in bone density have been reported for girls with distal forearm fractures.^{26 - 27} However, the subsequent rapid decline in fracture rates, which also was observed by Bailey and colleagues,³ points to a more transient process. According to Parfitt,⁴ the pubertal growth spurt entails a dramatic increase in the requirement for calcium that is met by an increase in intracortical bone turnover; when growth velocity slows, the cortical porosity is corrected with a concomitant increase in bone mass. Thus, the cessation of longitudinal growth in the distal radius after age 12 years in girls and age 14 years in boys is accompanied by a rapid increase in cortical thickness and bone strength.⁶

Since cortical porosity⁵ and increased bone turnover²⁸ could both be expected to impair bone strength at an age when physical activity is high,³ these fractures may well represent an imbalance between the demands placed on the skeleton by the increase in physical activity and the ambient bone mass.⁴ However, the increase in incidence that we observed between 1969-1971 and 1999-2001 also suggests either that the pattern of physical activity has changed, that acquisition of bone mass is being impaired, or both. Previous investigators linked both the increasing incidence of distal forearm fractures and the convergence of female with male rates to greater participation in sports and recreational activities.^{1 - 2 ,18} Similarly, we found an increase in fractures due to recreational activities, and the fact that forearm fractures were more frequent in the summer months reinforces this notion. This indicates that at least part of the explanation for our findings may lie with changes in physical activity. Indeed, national data indicate a substantial number of distal forearm fractures attributable to rollerskates or in-line skates in recent years.²⁹ Others have suggested instead that decreasing physical activity may be more important by inducing overweight; thus, the increase in distal forearm fractures could relate to skeletal overloading by obesity,²⁷ which is also increasing among children.³⁰ In other countries, where obesity is less prevalent, reduced physical activity in recent years has been associated with fewer forearm fractures.¹⁸

Although we were unable to directly address the issue in a retrospective study such as this, these data also raise concerns about whether acquisition of bone mass may be impaired in the later time periods, perhaps related to changing dietary habits. Over the past 20 years, there has been a dramatic increase in the consumption of carbonated soft drinks, with a corresponding decline in milk consumption.^{31 - 32} Among girls aged 12 to 19 years, for example, consumption of soft drinks increased from 207 to 396 g per day between 1977-1978 and 1994-1996, while milk consumption fell from 303 to 189 g daily; overall, daily calcium intake declined slightly from 784 to 771 mg/d for girls and remained unchanged at 1145 mg/d for boys.^{33 - 34} Both values are below the recommended dietary allowance for calcium of 1200 mg/d for adolescents,³⁵ the 1300 mg/d recommended for maximizing peak bone mass in this age group,³⁶ or the 1300 mg/d recommended as an adequate intake in the new dietary reference intake values.³⁷ In addition to changing dietary habits, other lifestyle factors, such as smoking or other drug use, may also impair acquisition of bone mass; however, our study was not designed to study these problems. We also were not able to study the severity of fractures because different classifications were used over the period.

In summary, our study demonstrates a significant increase in distal forearm fractures in children and adolescents between 1969-1971 and 1999-2001, although the results are based on estimates made at specific points in time. The comparison is a fair one because we used comparable methods throughout and rereviewed the 1969-1971 fractures previously reported,¹⁰ reassigning diagnoses and etiologies as needed for consistency with the new data. Moreover, these data are population-based and complete. However, Rochester is a small midwestern community that is predominantly white and better educated than the white population of the country as a whole.¹¹ To the extent that increasing rates of forearm fracture in childhood are a more generalized phenomenon, there may be cause for concern. This is because distal forearm fractures are a harbinger of additional osteoporotic fractures, at least among adults,^{38 - 40} and because impairments in childhood growth rates are associated with a dramatic increase in risk of hip fracture later in life.⁴¹ Indeed, a quarter of the bone mass in adult women and men is accumulated during the adolescent growth spurt⁴² and it is essential that adequate nutrition in childhood be ensured.⁴³ Suggestions for accomplishing this in practice have been published by the American Academy of Pediatrics.⁴⁴ In addition, further studies are needed to determine whether the changes in the incidence of distal forearm fractures are due primarily to increased physical activity, decreased acquisition of bone mass, or both.