Humans walk an average of 10 000 steps per day [1]. Functional gait forms an integral part of life, allowing individuals to function within their environment and participate in activities of daily living. The importance of locomotion from a psychosocial point is often overlooked. It facilitates normal social interaction and participation in recreational activities [2, 3]. The ability to walk is one of the critical elements in measuring and improving quality of life and reflects the individual’s health status [2, 4].

Gait assessment is part of the physical examination and can help screen for a range of physical impairments and abnormalities [5]. Similarly, analysis of gait at an early age can help predict motor outcome in cerebral palsy [6]. Evidence shows that a better understanding of normal development may be useful in interpreting abnormal findings [7, 8]. Gait analysis can also be used as an outcome measure to evaluate the effect of an intervention such as the single event multi-level surgery for children with cerebral palsy [9].

Although gait analysis has been conducted in children since the 1980’s, surprisingly little is known about age related gait patterns in children with typical development [10]. The gait data of 85 healthy children (4–16 years) at self-selected walking speed were examined using the VICON Plug-In-Gait (PIG) model [11]. Gait cycles of thorax, spine and pelvis kinematics in the sagittal, frontal and transverse planes were recorded, and stratified by age and normalized speed. The sagittal thorax and spine movements were found to be gradually and significantly associated with age, but less so with speed, so that with increasing age, children tended to lean their trunk forward relative to the pelvis. In contrast, the frontal and transverse parameters of spine and pelvic movements were found to be mainly dependent on speed, not age [11]. A 3D motion analysis study of fifty children between 7 and 11 years old using the ZEBRIS CMS 70 P system measured flexion (F), extension (E), abduction (abd) and adduction (add) angles of the hip joint, the F and E of the knee and ankle joints and foot rotations for each age group [12]. Their findings were consistent with other published literature reporting on joint kinematics and suggested that children 7–11 years old presented with adult-like gait patterns [12–14].

Speed strongly influences other spatiotemporal parameters, joint kinematics and kinetics of walking gait in children aged 4–17 years [14–16]. Van der Linden et al. [15] and Schwartz et al. [16] found that kinematics, kinetics and EMG readings corresponded strongly with speed. The kinetic values of peak propulsive forces were found in most of the joints of the lower limb during increased walking speeds, as well as significant differences in the kinematics of ankle dorsiflexion (DF), knee E and hip F and E ranges in the sagittal plane [15, 16]. EMG readings showed greater muscle activity at increased speeds for the hamstrings, rectus femoris and tibialis anterior muscles [15, 16]. In a South African based study of 200 children between the ages of 1 and 13 years, the study reported an increased speed as age increased. The data for children from 4–13 years of age were centred on the data for adults confirming that neuromaturation of gait patterns occurs from four years onwards and the authors concluded that speed is a reliable measure of gait maturation [17].

To the researchers’ knowledge, no studies describing the three-dimensional (3D) gait analysis of South Africa children have been conducted. Currently there exists no normative dataset for the gait parameters of typically developed children in South Africa. A normative database of typically developed South African children will provide a valid reference dataset to determine how gait is affected in children with gait abnormalities due to e.g. cerebral palsy which is highly prevalent in Africa [18]. Furthermore, a South African database of normative values is required to demonstrate that a South African gait analysis laboratory, protocols and procedures compares to international standards. It is known that gait laboratories from different countries have reported variability in gait patterns particularly hip rotation and foot progression angles. These differences could be due to the different marker placement or data processing protocols between laboratories [19]. Ferrari et al. [20] is one of the first studies that compared five different gait analysis protocols to assess inter-protocol variability. Prior to their study there had not been an emphasis on the standardization of gait analysis protocols between different laboratories, thus no gold standard for evaluation of gait. However, recently there has been an increase in studies measuring the reproducibility of data within and between gait laboratories. The primary aim of this study is to describe joint kinematics and spatiotemporal parameters of gait in South African children to constitute a normative database. Secondly we assessed if there are age related differences in the aforementioned gait parameters. We hypothesized that there will be no differences in gait parameters in children 6–8 years old compared to 9–10-year-old children.

This is the first report on normative gait patterns of typically developed South African children. The findings of this study suggest that the kinematic patterns and spatiotemporal parameters of gait in typically developed children 6–10 years old are consistent with the published international literature which reported on the gait patterns of children in developed countries such as Australia, Norway, Germany, and China [11, 14, 34–36]. The results are also in agreement with recent studies indicating that there are no significant differences in spatiotemporal parameters or kinematics between genders [8, 14, 37]. Moreno-Hernández et al. [37] suggests that it is not until the adolescent years when neurological and musculoskeletal maturity is reached, that gender differences may be notable. Children reach adult-like sensory integration at the age of 12 years and may be gender specific [38]. Other studies have concluded that a child’s gait will continue to evolve in terms of spatiotemporal parameters (step and stride length, speed and cadence, balance and percentage of support) until a child is fully grown due to the changes in anthropometric measurements [39–41].

Chagas et al. [8] and Moreno-Hernández et al. [37] studied children between the ages of 6–13 years and reported a non-normalized mean cadence of 122.48 ± 13.83 steps/min and 117.9 ± 11.4 steps/min respectively. This compares well with our study. Our study has shown that non-normalized cadence was significantly lower (p = 0.02), the speed faster and the step and stride length longer for the older children (9–10 years) compared to the younger children (6–8 years). This concurs with Dusing and Thorpe [7] and Holm et al. [14] who also reported reduced cadence in older children compared to younger children. Consistent with the findings of the present study, Chagas et al. [8] and Moreno-Hernández et al. [37] reported no significant differences in normalized cadence when comparing age sub-groups of children. Comparisons between children and adults also revealed an on-going decrease in cadence as age increased [35, 42]. Cadence decreased with age when children, aged 5–13 years, were compared with young adults (mean age 19.7 years) [35]. Bovi et al. [42] compared children (6–17 years) with 20 adults and found no significant difference in cadence between the two groups. This could be since the younger group included adolescents, who already showed matured gait patterns and adult-like sensory-motor integration [13, 38]. Step and stride length increased with age, but was not significantly different between the two age sub-groups of our study. Both findings agree with published studies [7, 14, 34, 35, 42]. Non-normalized step and stride length increased with age, but normalized values remained unchanged [7, 14]. Although speed affects cadence, step length, stride length and other spatiotemporal parameters, as well as kinematics during gait, our study did not show a significant difference in walking speed between younger and older children [11, 15, 16, 43]. The median speed for the group in the current study compares well with international research based on Mexican children (6–13 years), Australian children (5–13 years) and American children (9–11 years) who walked at a self-selected speed of 1.13(±0.19) ms^-1, 1.37(±0.17) ms^-1 and 1.22 (±0.04) ms^-1 respectively [35, 37, 44]. Thus, 6–10-year-old South African children’s spatiotemporal parameters of gait fall within the international norms when compared with those of other countries.

The kinematic gait patterns were similar between the younger and older children which could be attributed to negligible differences in walking speed, exposure to similar levels and type of physical activities and the absence of gross developmental or structural abnormalities in our participants [43]. We also noted similar peak hip, knee and ankle angles between the older and younger age groups. These differences in peak angles were from 0.1° up to 5.8° between the two age groups (see Table 4 and 5). Our finding compares to published reports [11, 13, 36].

Cigali et al. [12] and Shih et al. [36] reported similar mean peak hip abd/add angles in 50 children aged 7–11 years (−3.30 ± 2.32 – 6.33 ± 5.54) and 10 children aged 9.7 ± 0.9 years (0.42 ± 3.52 – 8.93 ± 4.39) respectively compared to our reporting of median values. Cigali et al. [12] also reported mean peak values for knee E/F (−7.06 ± 6.76 – 55.56 ± 3.11), ankle PF/DF (−21.85 ± 6.03 – 12.08 ± 12.27) and foot progression (−18.50 ± 11.80 – 11.00 ± 11.80) which falls within the standard deviation band width of our study (see Fig. 2). Shih et al. [36] reported comparable mean peak knee add/abd and knee external/internal rotation values of−2.21 ± 4.42 – 3.42 ± 4.89 and−10.18 ± 6.54 – 3.08 ± 5.07 respectively. The studies by Nikolajsen et al [44] and Kung et al. [45] reported only on the joint kinematics during the stance phase of children aged 10 years old and reported similar hip, knee and ankle mean peak values as see in our study for the stance phase. For example, Kung et al [45] reported mean peak knee F, ankle DF and PF of 41.74 ± 3.72; 10.18 ± 3.15 and−11.78 ± 5.14 respectively. This could indicate that the kinematic gait patterns of the pelvis and lower limb of 6–10-year-old children are established, comparable to the joint kinematics of children from other countries and mimic more the adult-like patterns observed by Sutherland et al. [13]. They evaluated the gait of 309 children ranging from the onset of walking to seven years of age. They found that between the age of 3.5–4 years, children achieve maturation of gait. In a later study, they concluded that growth alone can explain most changes throughout the rest of the growing years [41]. As children mature and grow, their leg length and body height increase, which directly affect the time-distance parameters of gait [41].

When comparing joint kinematics within the two age subgroups, hip rotation was significantly different between the groups (p = 0.036). Older children (Group B) presented with more external rotation at the hip joint than the younger children (Group A). Femoral anteversion and hip internal rotation are highly correlated and both reduce significantly with advancing age. Thus, our study supports the fact that as a child develops, the degree of anteversion of the femoral head decreases and causes the older child to walk with more relative external rotation of the hip than a younger child [46, 47]. The degree of hip internal rotation may indicate surgical intervention in children with pathological gait. Hip rotation kinematic patterns might be age specific and should be considered accordingly when interpreting gait analysis data.

The study was limited by small numbers in certain sub-groups such as the number of 6-year-old boys. Kinetics were also not included in this study, but we recommend that future studies include kinetics as it could add valuable information to the understanding and interpretation of the gait patterns in typically developed 6–10-year-old children in South Africa.

This study evaluated the 3D kinematics and spatiotemporal parameters of gait in 28 typically developed 6–10-year-old South African children. It provides normative values for gait parameters that show that this South African gait analysis laboratory compares well with international gait laboratories and values can be used for comparison during gait analysis. The study’s findings concluded that normalized spatiotemporal parameters were similar between the two age groups and are consistent with the values of children from other countries. The joint kinematic values showed significant differences for hip rotation, indicating that older children had more external rotation than younger children.