Previous studies about pediatric pain have consistently found that the lower limb is the most commonly reported musculoskeletal location with pain and has been ranked among the leading general pain symptoms in childhood and adolescence [1–3]. The lower limb is also the most commonly injured site in sports [4] and accounts for a substantial percentage of pediatric rheumatology and orthopedic consultations [5].

Two studies have reported consequences and disabilities related to lower limb pain in schoolchildren, both limited to non-traumatic knee pain. Fairbank et al. found that 18% of children aged 13–17 with knee pain had to stop playing during the past year because of their symptoms. In addition, 22% of girls and 40% of boys needed to visit a doctor [6]. In a Finnish study [7], 45% of schoolchildren aged 9–15 reported moderate to severe disability attributed to lower limb pain during past three months. Despite these high figures, less attention has been given to determinants of lower limb pain in children compared to other pediatric pain symptoms. This might be due to a belief that lower extremity pain in children is a transient, self-limiting condition [8]. Further, the substantial burden of other pain symptoms in adults [e.g. lower back pain [9] and headache [10]] has directed research interests in understanding the origin of these symptoms in childhood. However, lower limb pain may limit children's activities and lead to adoption of an inactive lifestyle persisting to adulthood [11]. Hence, more studies are needed to provide further insight into factors that influence the occurrence of this condition in schoolchildren.

Some potential determinants of lower limb pain have been studied in children and adolescents, but the results are inconsistent. A positive relationship has been reported between frequency of physical exercise and lower limb pain in most studies [6, 7], but no correlation was observed in a recent follow-up study [11]. Similarly, conflicting results have been reported for the predictive role of physical fitness [12, 13] hypermobility/hyperflexibility [11, 14, 15] and psychosocial factors [16, 17]. We hypothesized that distinction of traumatic from non-traumatic pain would provide a clearer picture of the risk factors for lower limb pain.

The aims of this study were (a) to determine the prevalence of pain symptoms in different lower limb parts in a population-based sample of preadolescents, (b) to examine the associations of traumatic and non-traumatic lower limb pain with age, sex, hypermobility, frequency of exercise, physical fitness, and psychosomatic factors, (c) to assess school absence and subjective disabilities attributed to traumatic and non-traumatic lower limb pain.

The initial study took place in Lahti, a town of 94 827 (1995) inhabitants in southern Finland. All the 21 primary schools in town were asked to participate in the survey. Two schools refused and 19 took part in the study. The Steiner school, the hospital school, and the schools for hearing and physically disabled and the mentally handicapped were excluded because some of the methods used in this study were not suitable for their pupils. All pupils from the third and fifth grades participated in the survey except those who were not at school on the day of the study [18, 19].

The study was approved by the Ethics Committee of the Health Care Center of the City of Lahti. The parents were informed about the study by letters in advance. Both the children and their parents had the possibility to refuse enrollment in the survey. The final sample consisted of 1756 preadolescents, of whom 867 were third-grade (mean age 9.8 [SD = 0.4] years) and 889 were fifth-grade (mean age 11.8 [SD = 0.4] years) school children. This population represents 82.9% of all schoolchildren in these grades attending normal or special schools in Lahti (Figure 1). In March 1995, these children completed a pain questionnaire, during a class lesson, to assess lower limb pain during the preceding 3 months and were tested for hypermobility and cardio-respiratory fitness.

In our study, the knee was the most common site of pain in the lower limb, followed by the ankle and foot. This is in accordance with previous studies in children and adolescents [24, 25]. The knee was also the most common location of pain due to a direct trauma, as has been reported previously [26]. About half the children who had lower limb pain reported pain also in other musculoskeletal location. This figure is substantially higher than the 19% prevalence of musculoskeletal pain in those without lower limb pain. This finding is consistent with previous surveys which have demonstrated frequent co-occurrence of different musculoskeletal complaints [12, 27].

On the basis of our cross-sectional study, lower limb pain – both traumatic and non-traumatic- was more common among the younger age group (less than 11 years old) than among the older age group (from 11 to 13 years old). This contradicts the previous study by Vähäsarja, showing a clear increase in prevalence of knee pain from 4% in the 9–10 olds to 19% in the 14–15 olds schoolchildren [7]. However, this might be due to the different age distribution of our study population; which is composed of two age groups (9–10 and 11–12 years); and the wide case definition of lower limb pain used in our study (including pain in all lower limb area). More frequent non-traumatic lower limb pain among younger children (aged 9–10 years) is in accordance with the literature suggesting that "growing pains" mainly affects children aged 4 to 10 with occasional relapses in teenage years [8]. In this study, traumatic lower limb pain was more common among boys. Likewise, Smedbraten and co-workers found knee to be the only site in which pain was more common among boys [24].

Difficulty in walking long distances was the most common disability, reported by approximately 40% of children with lower limb pain. This is much higher than the 11% reported in a previous study about knee pain in schoolchildren [7]. However, the minimum distance was not specified in the later study. Disturbances of sleeping was the only disability associated to some extent more closely with non-traumatic than traumatic pain. This suggests that "growing pains" constituted a substantial proportion of non-traumatic pain.

Practicing vigorous exercise was associated with occurrence of non-specific lower limb pain. As mentioned earlier, conflicting results have also been reported about the role of physical exercise in occurrence of lower limb pain in schoolchildren [6, 7, 11]. we hypothesized that psychosomatic symptoms are associated only with non-traumatic lower limb pain, and that physical exercise is only related to traumatic lower limb pain, and this might be the reason for this conflicting results. As we expected, our stratified analysis showed that practicing vigorous exercise was associated with pain in the traumatic group but not in the non-traumatic group. After adjusting for physical activity and all other baseline predictors in our multivariate analysis, day tiredness was the only significant psychosomatic factor in the traumatic group. Day tiredness might have been a consequence of pain experience rather than had psychological origin. We have used the term "psychosomatic" to refer to symptoms that are chiefly considered to be expressions of psychological stress rather than manifestations of organic disorders based on previous research in children [28, 29]. The same term was used in our previous reports [19, 30] as well as by other researchers [31, 32]. However it might be argued that some of these symptoms were manifestations of lower limb pain rather than related to emotional problems in childhood.

High level of physical fitness was independently associated with the occurrence of traumatic lower limb pain and practicing vigorous exercise was a significant effect modifier for this relationship. These results seem to contradict previous findings by Knapik et al. who reported that musculoskeletal injuries in young soldiers were associated with lower aerobic fitness [13]. However, several factors may prevent a valid comparison between our results and those of the latter study. Firstly, the age groups and the study settings in these studies were different; secondly, Knapik et al. were studying the relationship between physical fitness and all kinds of musculoskeletal injuries, including overuse injuries. Overtraining injury is not caused by a direct trauma but is related to repetitive loading on musculoskeletal structures without sufficient recovery. It must be noted that in the current study we have not collected information about overuse injuries, as we assumed that children would find it hard to understand the exact meaning of such type of injury. We have categorized our children into two groups according to whether or not they have directly traumatized their lower limb area with pain. Children who might have had overuse injuries were not included in the traumatic group, but rather in the non-traumatic group, as they have not indicated a direct trauma to the lower limb in the pain questionnaire. It might be possible that the predictive roles of physical fitness on the development of overuse and traumatic musculoskeletal injuries are different. However, this possible explanation cannot be examined with our data, and hence further research is needed to find out if such a difference really exists.

Occurrence of traumatic and non-traumatic lower limb pain did not differ between children with and without hypermobility. This is similar to our previous finding [19], regarding occurrence of musculoskeletal pain in general. However, children with regional knee hypermobility reported significantly less traumatic pain than children with normal joint laxity. This inverse association was not explained by differences in physical exercise frequency between these two groups. These results go well with the previous observations that hypermobility is not associated with increased athletic injuries [33] and that stretching plays an important role in decreasing the rate of injury [34]. These studies as well as ours, challenge the recommendation that children with hypermobility should avoid strenuous physical activities.

The most important strengths of this study are its population based setting and classifying lower limb pain according to location and association with injuries. In addition, pain interference with the children's daily activities was assessed. Yet, we did not have information about anatomic-structural or psychosocial factors, which might have been linked to occurrence of lower extremity pain in children.

Our study indicates that risk factors and consequences of non-traumatic and traumatic lower limb pain are not similar. Non-traumatic pain is correlated with headache, recurrent abdominal pain and sleeping problems, while the traumatic pain is associated with practicing vigorous exercise and high level of physical fitness. This dissimilarity is consistent with our previous finding that traumatic lower extremity pain in preadolescents has a more favorable long-term natural course compared to non-traumatic pain [35]. Including or excluding traumatic injuries in the case definition of lower limb pain in research work investigating its determinants in children will significantly affect the results, and this might be the reason for the discrepancy of previous research conclusions. These two conditions need to be treated as different disorders in future studies.

The author(s) declare that they have no competing interests.