The prevalence of obesity is increasing in the United States and throughout the world [1, 2]. In 2011–2012, 34.9% of adults in the United States were obese [3]. Musculoskeletal disorders are commonly seen in obese individuals and one of the most common and disabling of these is knee osteoarthritis (OA) [4].

Individuals who are obese and have knee pain may adopt different gait patterns to compensate both for the extra weight and joint pain. While no prior studies have focused specifically on obese persons with knee pain, studies of obese adults showed that they walked with decreased velocity [5, 6]. As body mass index (BMI) increases, gait speed decreases [7]. Past studies of knee joint kinematics mainly focused on sagittal and coronal plane motions, i.e., knee joint flexion/extension [5, 8–10] and varus/valgus rotation [9, 10]. Even so, there is no clear consensus on knee joint kinematics in obese individuals during walking. For example, Haight et al.[8] reported that obese individuals walked with a less flexed knee during the stance phase compared to non-obese individuals. Vismara et al. [11] concluded that the range of knee flexion excursion during gait was not significantly different than a healthy group. The inconsistency and variation in the literature may be due to differences in measurement methods or the presence of different lower extremity joint pathology such as pain which is extremely common in obese adults and may cause gait modifications. For instance, knee pain is a major symptom in individuals with knee OA and reduced range of knee flexion during gait has been frequently reported [12–14].

Most previous investigations of obese gait used skin marker motion analysis systems [5, 6, 8, 15]. The kinematic data derived from a skin marker-based motion capture system are vulnerable to soft tissue artifacts [16, 17]. According to Peters et al.[16] the magnitudes of soft tissue artifacts were greater than 30 mm on the femur and up to 15 mm on the tibia. Cappozo et al. [18] quantified rotation errors of 6–20° on the femur and 4–10° on the tibia [18]. Although some biomechanical researchers have tried to reduce soft tissue artifact by using an obesity-specific marker set to investigate gait patterns [15], accurate detailed kinematics in 6 degree-of-freedom (6DOF) among obese individuals have not been elucidated. As past studies have shown that standing posture and gait pattern were affected by body weight [5, 8–10, 19], a better understanding of the standing posture and gait pattern in 6DOF is critical. Such information would provide a foundation on which to build better walking programs, which are commonly suggested as a means to increase the energy expenditure of obese individuals [20, 21] and especially to design a treatment program for obese persons with knee pain.

Recent advancement of imaging technology, such as the dual fluoroscopy imaging system (DFIS), makes it possible to track in-vivo bone motion without soft tissue artifacts [22–24]. This methodology has been validated[25] with submillimeter and subdegree accuracy in translation and rotation. In this study, we evaluated knee joint kinematics in standing and during gait in obese individuals with knee pain using DFIS[25] and compared them to a healthy non-obese control group without knee pain. We hypothesized that there would be distinct motion patterns in obese individuals with knee pain which would differ from typical patterns in a healthy population.

This study investigated individuals with obesity and knee pain in standing and during gait under DFIS surveillance. While standing, the knee was in about 7.4° hyperextension, slight abduction (valgus), and about 5.6° of external rotation. During treadmill gait, the largest rotational excursion was in flexion-extension and the largest translational excursion was in the anterior-posterior direction, while motions in the other planes were smaller.

Few studies have evaluated standing posture in individuals with obesity. One study using a radiographic hip-knee-ankle measurement, found that obese individuals stood in a slightly knee flexed position [19]. However our obese individuals with knee pain tended to use a hyperextension strategy in standing. The adaptation in standing posture may be due to the high BMI in our sample (≥ 35) and knee pain. This posture is thought to reduce the demand on the quadriceps and could potentially prevent fatigue [30]. Our study also provided information in the other degrees of freedom. The analysis of the relative position between the femur and tibia indicated that the center of the femur was near the center in anterior-posterior direction, but located at the medial portion of the tibial plateau. These data provide additional data for more complete understanding of standing posture correction.

During walking, spatiotemporal parameters may be affected by gait speed. The speed we tested was controlled at 0.67m/s (1.5 mph), which is slightly slower than the reported self-selected walking speed of 0.73 to 1.08 m/s for an obese population [7, 31]. At this gait speed, we found that obese individuals with knee pain had slightly increased cadence and decreased stride length compared to individuals walking at a similar gait speed [7, 32], and had increased stride duration and stance phase compared to individuals walking at a faster gait speed [9, 31]. The speed preference and participant characteristics, such as knee pain and presence of osteoarthritis, may contribute to differences between studies.

Several studies have reported knee kinematics during gait using skin marker motion analysis in obese individuals [5, 6, 8]. In these studies, obese individuals were found to walk with a similar pattern or with more knee extension and reduced range of motion in the sagittal plane compared to non-obese individuals, and this strategy was assumed to decrease the exertion of the knee extensors and prevent fatigue [10, 33]. Our findings were not in full agreement with past studies [5, 8]. Similar to past studies, our obese individuals walked with a smaller range of flexion-extension motion during the stance phase compared to a healthy population (Fig 3A) [22]. However, our obese individuals walked in a more flexed knee position. While it is commonly thought that a more knee extended position will reduce demand on muscle, our participants did not display this adaptation. The combined smaller range of flexion-extension motion with a more knee flexed position during walking may increase the demands on the quadriceps and other extensors, and this gait pattern may result in early muscle fatigue. Our findings also showed a larger variation in knee flexion angles at initial contact, indicating different gait adaptation strategies could be used among individuals. The initial contact flexion angles were weakly correlated with their WOMAC pain score during walking (r = -0.34), suggesting that pain status may be a major contributor to the gait pattern.

Knee pain is a commonly reported symptom by patients with knee OA, and kinematic changes in the sagittal plane during gait were frequently reported [12–14]. Individuals with knee OA walked in a more extended position and reduced range of flexion-extension motion. This gait pattern had been further replicated by studies of experimental knee pain [34]. Obese individuals have also been found to walk with a similar gait pattern. Our study including individuals with combined obesity and knee pain walked with a smaller range of flexion-extension motion similar to patients with knee OA, but in a more knee flexed position during most stance phase when compared to a healthy non-obese group. This finding suggests that our obese individuals with knee pain may adopt a unique gait pattern, which is not typically found in patients with either obesity or knee OA.

A previous study found no difference in axial rotation between obese and non-obese groups [6], and our findings were in agreement with this; however, the pattern of axial rotation in our participants seemed to differ from that of healthy knees [22]. Healthy knees internally rotate after initial contact to a peak rotation at the end of loading response (Fig 3C) [22]. Instead, the tibia in our obese individuals was externally rotating to a peak of about 3.9° around the end of mid-stance and reversed after that point, implying the pivoting mechanism observed in healthy participants [22] may be changed in obese individuals. The large variation in the internal-external rotation angles in our obese individuals suggests that obese individuals may adopt diverse strategies to avoid pain during walking.

In the coronal plane, past studies found similar or more adduction in obese individuals compared to non-obese group during gait [5, 6]. In our study, we found adduction-abduction in coronal plane motion in our obese group was not significantly different compared to our control group (Fig 3B) [22]. We note that the majority of our subjects did not have medial osteoarthritis which might have been present in previous studies.

The smaller range of motion of the obese patients indicated that these individuals use a stiffening knee strategy despite maintaining the knee in more flexion during functional activities [35]. This strategy for reduced range of motion is also found in knee OA patients [12–14, 36] and obese individuals without knee pain [8], meaning both extra body weight and pain contribute to the gait pattern change. As walking has been routinely suggested for obese patients as a safe activity to increase energy expenditure, increasing the knee joint range of motion while walking should be addressed, and this could potentially better distribute contact stress to prevent the local stress concentration.

Several limitations need to be mentioned in interpreting the results. First, our participants were severely obese (BMI ≥ 35.0) and with knee pain; therefore, the results may not generalize to the less obese population (BMI: 30.0–34.9) with or without knee pain. Our study sample was 80% female, so the results may not generalize to all obese individuals with knee pain, although, like our sample, most obese individuals with knee pain are women. The control group was predominantly male, and this may also have contributed to the differences in kinematics found. Treadmill walking was used in this study to assess the kinematics, so the results may not generalize to walking overground. The definition of the coordinate system in this study was based on local bone geometry; therefore, the values found in this study may be slightly different from studies using other coordinate system definitions. Lastly the loading from the lead protection gowns (9.1kg, 7.0–10.6% of body mass of the obese participants) could have contributed to the gait pattern changes; however, it is unlikely that this contributed significantly to the differences found.

In conclusion, obese individuals with knee pain used hyper-extension knee posture while standing, but maintained the knee in more flexion during gait with reduced overall range of motion in the 6DOF analysis compared to a healthy group. In addition to the facilitation of greater total range of sagittal plane motion with a knee extension strategy during gait, increasing the knee joint range of motion in other directions should be addressed in weight management exercise programs.