Emotions and Gait

Experienced O&P clinicians are skilled at identifying deviations from normal gait and recognizing the underlying neurological, musculoskeletal, and device-related causes of those deviations. When observing gait, students and novice clinicians can be overwhelmed by the number of cues that demand their attention. They tend to assess gait in a less systematic way and fixate on fewer cues for a longer time than experienced clinicians, and often miss the most obvious deviations while focusing on minor deviations that are not as clinically meaningful. They often fail to see the forest for the trees. More experienced clinicians tend to focus on patterns of movement related to pathologies or alignment problems. Over time, this process becomes intuitive, and patterns can be recognized when viewing a patient’s gait for only a short period of time. It is even possible to accurately predict which deviations a particular patient is likely to exhibit based on his or her diagnosis, strength, and range of motion. If asked to replicate the gait of someone diagnosed with hemiparesis or spina bifida, most experienced clinicians could do so in a way that was recognizable to other clinicians. I still recall Tony Herring, MD, demonstrating specific gait patterns and deviations to our class of O&P students at UT Southwestern many years ago and have seen many similar demonstrations since then. Animation and Robotics A video titled “100 Ways to Walk” by stop-motion animator Kevin Parry has amassed over seven million views on YouTube. 1 In the video, Parry demonstrates walking patterns exhibited during a wide variety of activities and emotional states. His ability to replicate each pattern in just a few seconds in a way that is easily recognizable demonstrates his insight into how people move. Each gait pattern is labelled but could be identifiable even without the label. Many of the patterns are exaggerated and stereotypical (as is much of the movement in animation) but demonstrate aspects of gait that are commonly seen in the real world. There appears to be considerable overlap between the skills of an animator wishing to replicate human movements to communicate meaning in digital media and those of clinicians who observe gait to aid in clinical decision-making. One of the many factors that can influence gait is the emotional state of an individual. Like physical impairments, different emotions can result in distinct patterns of walking that are recognizable even by untrained individuals, as Parry’s video demonstrates. Most people are familiar enough with the gait patterns associated with specific emotions that they could recognize (and perhaps even replicate) an angry or happy gait, for example. My interest in the connection between emotions and gait was initially piqued by an article published by Wired magazine in 2020. The article describes research conducted at the University of Maryland to develop “ProxEmo, a novel end-to-end emotion prediction algorithm for socially aware robot navigation among pedestrians.” 2 The goal of the research was to facilitate the navigation of robots in public spaces shared with humans by allowing the detection of human emotions. This ability could be “used for emotion-guided navigation taking into account social and proxemic constraints.” 2 The researchers focused on four general emotions: happy, sad, angry, and neutral. A video describing the research and showing skeletal representations of these different emotions can be viewed at gamma.umd.edu/researchdirections/socrob/proxemo. The researchers concluded: “We can infer from this matrix that our model performs fairly well across all emotion classes with a high accuracy.” 2 The body of robotic research on gait and emotion recognition is extensive and involves advanced scientific methods and analysis that go well beyond most clinicians’ understanding, including my own. Much of the research includes investigating the ability of human subjects to identify emotions from gait, and to identify specific kinematic characteristics of gait that can be used to develop computer systems. This type of information is more relevant to clinical gait assessment. For those interested in a deeper look at this topic, Xu, et al. published an article in 2022 that includes a review and summary of much of this research that clinicians may find helpful. 3 Emotion Recognition by Humans In a 1987 paper, psychologist Joann Montepare, PhD, described research examining “the potential for information provided in a person’s style of walking to reveal certain emotions.” 4 Five people were videorecorded walking while “imagining themselves in each of four emotion situations.” 4 Each walker was read a scenario that described happiness, sadness, anger, and pride and asked to walk in a way consistent with each emotion. Walkers were recorded from the neck down to focus the observation on gait and not facial expression. The scenarios and related emotions were described to ten observers, who then reviewed the videos and identified the emotion they thought was being displayed. Observers then watched the videos again and identified different characteristics of each walking pattern, including stride length, arm swing, light or heavy footedness, and trunk posture. The researchers found that “subjects were able to identify sadness, anger, happiness, and pride from gait information at better-than-chance levels; however, identifications of pride were significantly less accurate than were identifications of sadness and anger…. Results also revealed that gait characteristics such as the amount of arm swing, stride length, heavy footedness, and walking speed differentiated the emotions expressed by walkers.” 4 An angry gait pattern involved more heavy footedness and a sad pattern involved less arm swing than other gait patterns. Both angry and proud gait patterns involve increased stride length. A happy gait pattern was faster than other patterns. Researchers in Australia investigated whether observers of gait “are aware of the gait movements that they used to discriminate between the different displayed emotions” and whether “the gait features that perceivers claim to utilize in the identification of each basic emotion are actually present in the perceived emotion displaying walkers.” 5 The gait patterns of 36 individuals were recorded using a 3D motion analysis system as they simulated happiness, sadness, anger, and fear at three different intensity levels. Thirty-four subjects identified the emotion they observed and described the strategies they used to make that determination. The researchers discussed walking pace (including cadence and stride length), arm swing, bounciness, and stomping. “Happy and angry walkers both displayed long strides with increased arm movement though angry strides had a faster cadence. Fearful walkers walked with fast short strides reminiscent of a scurrying gait. Sad walkers walked with slow short strides consequently creating the slowest walking pace. However, fearful and sad walkers showed less arm movement in their gait in different ways. Sad walkers moved their entire arms whilst fearful walkers primarily moved their lower arms throughout their gait.” 5 These researchers stated “that walking pace, along with the components of walking pace (i.e. cadence and stride length), and the type of arm swing exhibited by the point-light walker provided valid perceptual cues to distinguish between gait displays of happiness, sadness, anger, and fear.” 5 They concluded that “the expression of basic emotions (happiness, sadness, anger, and fear) can be discriminated through specific gait kinematics: walking pace and arm swing.” 5 Venture et al. investigated whether human subjects could correctly identify the gait of four actors exhibiting the emotional states of neutral, joy, anger, sadness, and fear. Subjects were shown actual video footage and animations based on the videos and asked to identify the emotion being demonstrated. They found that “there is little confusion between emotions other than with neutral...for example, joy is not mistaken with sadness….” 6 They concluded that it is “possible to discriminate among the four conveyed emotions with a total average recognition rate of 78 percent....” 6 It seems clear from these examples that emotions can be identified by observing gait in able-bodied individuals, and that the process and cues used to do so are similar to those used in clinical gait assessment. Whether this is relevant to clinical practice is less clear. Clinical Application Artificial intelligence may eventually be used to aid clinicians in identifying pathological gait, even if this technology is not currently available or useful for that purpose. When that occurs, it is likely that it will be used to identify segmental deviations and correlate those patterns with those commonly observed in patients with specific pathologies, rather than to identify emotions. It is interesting to reflect, however, on the relationship between emotions and gait in a clinical setting. Students and trainees struggling to develop gait assessment skills can take encouragement from the fact that they are already skilled at this type of activity, as evidenced by their ability to recognize common emotions by observing gait. Students at Northwestern University Prosthetics-Orthotics Center (NUPOC) are encouraged to watch Parry’s examples and the ProxEmo video to reinforce this point early in their training. An understanding of muscle function, range of motion, and common pathological presentations are all essential for accurate and thorough clinical gait assessment, highlighting the importance of mastering those content areas and skills. Just as with emotion detection, a solid foundation and repetition informs the intuition that follows. Students and trainees must learn to identify both the trees and the forest, and how each impacts the other. An informed understanding of arm swing in pathological gait, for example, requires the knowledge of the variety of factors that can cause it, and an in-depth assessment may be required to accurately assess and address the primary cause. This level of clinical reasoning requires informed experience, a process that can be forgotten or overlooked by clinicians who have achieved a high level of expertise and intuition in this area. Gait pattern recognition involves an element of stereotyping, which requires oversimplification that can lead to misinterpretation. Gait observers in the study by Halovic et al. “partially explained their ability to identify each emotion by citing stereotypic emotional gait movements, some of which were absent from our point-light walker stimuli.” 5 Clinically, what we believe we see may not actually be present, an example of confirmation bias that clinicians must be conscious of to avoid. Here again, the importance of a thorough assessment becomes obvious. As with all stereotyping, intuitive gait assessment is only as useful as it is accurate, and attention to the details of each specific case is required to make meaningful conclusions. Being attuned to the emotional state of patients is at the heart of clinical empathy and benefits clinicians and patients in many ways. It is likely, however, that while the emotional state of our patients does impact their gait, it does so in ways that are much less obvious than musculoskeletal and neurological factors and may mask them. Additionally, emotional states affect aspects of our patients’ lives that influence their experience and our treatment in other ways besides gait, such as engagement with and adherence to the treatment plan. Montepare said of her study that, while it “lends strong support to the claim that important social information is revealed in gait, one may question the extent to which this is true when other sources of information are available such as vocal or facial information.” 4 Sensitivity to what patients verbalize (if they are able) and other observable behaviors that indicate potential barriers to function seems like a more fruitful application of clinicians’ emotional awareness. Clinicians can (and should) develop communication skills that can be used to effectively assess and address the emotional state of our patients, and there are validated tests developed for that purpose. For example, the Patient Health Questionnaire (PHQ-9), used to screen for depression, is included in the Amputee Coalition Well-Being Program. 7 Many outcome measures developed for O&P populations can also create opportunities to discuss psychosocial aspects of patients’ lives with them. The goal is not to diagnose or treat mental health conditions, but to raise awareness, inform O&P treatment decisions, suggest resources that may benefit the patient, and make referrals to healthcare professionals. Emotion detection based on gait may be more of an interesting thought experiment than a practical clinical skill. However, the relationship between clinicians and patients will be strengthened by our attunement to their emotional state. An appropriate and professional emotional connection with our patients can improve our treatment and can be one of the most rewarding aspects of our jobs. John T. Brinkmann, MA, CPO/L, FAAOP(D), is an associate professor at NUPOC. He has over 30 years of experience in patient care and education. References youtube.com/watch?v=HEoUhlesN9E Narayanan, V., B. M. Manoghar, V. S. Dorbala, D. Manocha, and A. Bera. 2020. Proxemo: Gait-based emotion learning and multi-view proxemic fusion for socially-aware robot navigation. 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 8200-07. Xu, S., J. Fang, and X. Hu, et al. 2022. Emotion recognition from gait analyses: Current research and future directions. IEEE Transactions on Computational Social Systems doi: 10.1109/TCSS.2022.3223251. Montepare, J. M., S. B. Goldstein, and A. Clausen. 1987. The identification of emotions from gait information. Journal of Nonverbal Behavior 11: 33-42. Halovic, S., and C. Kroos. 2018. Not all is noticed: Kinematic cues of emotion-specific gait. Human Movement Science 57: 478-88. Venture, G., H. Kadone, T. Zhang, J. Grèzes, A. Berthoz, and H. Hicheur. 2014. Recognizing emotions conveyed by human gait. International Journal of Social Robotics 6: 621-632. https://www.amputee-coalition.org/healthcare-providers/well-being-program/

Emotions and Gait

Every BODY Is an Athlete: Redefining Adaptive Endeavors

Baby Steps: Understanding Infant Crawling

Baby Steps: Understanding Infant Crawling

RECENT NEWS

Get unlimited access!

O&P JOBS

Certified Orthotist & Orthotic Technician

Prosthetic/Orthotic Technician – Boise, ID

Certified Orthotist

Are you sure want to unlock this post?

Are you sure want to cancel subscription?

The most important industry news and events delivered directly to your inbox every week.

Welcome Back!

Retrieve your password

Are you sure want to unlock this post?

Are you sure want to cancel subscription?

Emotions and Gait

Support authors and subscribe to content

Subscribe

Related posts:

Every BODY Is an Athlete: Redefining Adaptive Endeavors

Baby Steps: Understanding Infant Crawling

Baby Steps: Understanding Infant Crawling

RECENT NEWS

Get unlimited access!

O&P JOBS

Certified Orthotist & Orthotic Technician

Prosthetic/Orthotic Technician – Boise, ID

Certified Orthotist

Login to your account

Reset Password

Signup to your Account

Are you sure want to unlock this post?

Are you sure want to cancel subscription?

Account Activation

The most important industry news and events delivered directly to your inbox every week.

Welcome Back!

Retrieve your password

Login to your account

Reset Password

Signup to your Account

Are you sure want to unlock this post?

Are you sure want to cancel subscription?

Account Activation