Knee alignment in adolescents is correlated with participation in weight-bearing sports

One thousand eight young healthy children between 12 and 19 years of age were recruited for this cross-sectional correlation study. Only healthy volunteers with no orthopaedic or trauma history could participate. We excluded subjects who had been treated or seen for a musculoskeletal condition or trauma and/or had been seen by a specialist. All participants were recruited as volunteers at three different schools. All volunteers consented to participate in the study, which was approved by the ethical commission of our institution before the first inclusion.

Included were 564 male and 444 female volunteers. The volunteers were first divided into two groups by gender. Then these groups were further split into three age groups: group 1 (12–13 years of age), group 2 (14–15 years of age) and group 3 (16–18 years of age). Participant’s demographic data are presented in Table 1.

All participants were weighted in kilogrammes (kg), and their height was measured in centimetres (cm). From these measurements, the body mass index (BMI) was calculated.

The alignment of the knee was determined by measuring the intercondylar (IC) and intermalleolar (IM) distance. The subject was asked to stand in an anatomic position with the hips and knees in maximal extension with the medial condyles or malleoli touching as it has been described before [1].

To have the subjects stand in a normal relaxed position, we used a specially designed instrument, first described by Witvrouw et al. [3]. The subjects were asked to assume a relaxed standing position, with their feet at shoulder width, each foot on a different platform. They were asked to position their feet as such that the medial border of the feet was above the inner border of the platform. Then a motor was used to bring the two platforms together at a constant low speed until the medial condyles or malleoli touched [3] (Fig. 1). The position of the patella facing forward was critically checked, as was done by authors who have studied lower leg alignment [5, 12]. This was done to control the rotational position of the lower extremity, since this might influence the outcome of the measurement [12, 13]. Eventually, the intercondylar distance (mm) in subjects with genu varum was measured with an inside calliper with the medial malleoli touching using the method described by Heath et al. [14]. The intermalleolar distance (mm) in subjects with genu valgum was measured with an inside calliper with the medial condyles touching. Intercondylar distances were expressed as positive values and intermalleolar distances were expressed as negative values. Both measurements were combined into one parameter: the IC/IM distance [2]. When a subject was positioned in a correct way, but neither the two malleoli nor the condyles touched, the bony intercondylar and intermalleolar distance was measured. When they had genu varum, half of the IM measurement was subtracted from the IC measurement. This value was then considered as the absolute IC/IM distance. When they had genu valgum, twice the IC measurement was subtracted from the IM measurement. This value was then considered as the absolute IC/IM distance. All IC/IM distances from all subjects were measured by the same examiner. The inter-tester reliability (ICC = 0.95) and the intra-tester reliability (ICC = 0.96) for the used technique of measuring IC/IM had a high interclass correlation coefficient [3].

All participants were asked to fill in a questionnaire. They were asked about the date of birth, their sports participation, for how many years they had done so and for how many hours a week on average. They also answered the question whether they had already consulted a doctor in musculoskeletal injuries. They were not informed about the purpose of the study before the measurements were done and the questionnaire was filled in.

For every sport the subjects participated in, we calculated the product of the number of years and the average number of hours a week. We then added the products for all impact/weight-bearing sports (e.g. tennis, basketball, volleyball, athletics). We defined the result as ‘the cumulative sports factor’ and calculated it for every participant in the study in order to describe the cumulative load of impact sports. A similar approach has been described before by Asadi et al. [15].

Properties of the different groups are presented as mean and standard deviation. Graphs present the relation between height, IC/IM and age.

Within the group of boys and the group of girls, a one-way analysis of variance was performed to analyse the differences among the three age groups separately. Post hoc tests were performed using the least significant difference test. For each age group, the differences between the data of boys and girls were analysed using analyses of variance. Spearman’s rank correlation coefficients between IC/IM, cumulative sports factor and BMI were calculated. Scatter plots were used to present the IC/IM in function of the cumulative sports factor on both the male and female subjects, separately for each group.

For this statistical analysis, SPSS statistics 18 (SPSS Inc. Released 2009. PASW Statistics for Windows, Version 18.0. Chicago: SPSS Inc.) was used. Statistical significance was accepted at the level of P < 0.05.

Linear models are used for data analysis with IC/IM score as the response variable. A possible nonlinear shape of the association between two continuous measures (e.g., IC/IM score and cumulative sports factor) is explored using both quadratic effects and restricted cubic splines with 4 and 5 knots. Likelihood ratio tests are used for the comparison linear and nonlinear trends. The R² measure is used to compare the relative importance of different sports/cumulative sports factors in the explanation of the association between the cumulative sports factor and IC/IM. All these analyses have been performed using SAS software version 9.2 (SAS Institute Inc. 2002. SAS System for Windows, Version 9.2. Cary: SAS Institute Inc.).