Secondary Logo

Journal Logo

Perspectives by Discipline

KAFOs for Ambulation: An Orthotist’s Perspective

Michael, John W. CPO/L, MEd, FISPO, FAAOP

Author Information
JPO Journal of Prosthetics and Orthotics: June 2006 - Volume 18 - Issue 7 - p P187-P191
  • Free


Knee-ankle-foot-orthoses (KAFOs) are frequently considered “orthoses of last resort” by experienced rehabilitation experts, based in large measure on the clinical experience that the initial and long-term acceptance rates for this level of orthotic control are both significantly lower than for devices that do not cross the knee joint.1 The long-term use rate of bilateral KAFOs by adults with complete spinal cord injuries (SCI) at the thoracic level is particularly discouraging.2 However, there are a number of factors that may contribute to this observation that should be considered in more detail.3

Part of the reluctance to recommend KAFOs comes from the widely accepted aphorism that “the best orthosis is the least orthosis.” The author’s experience suggests that the orthosis the patient will most readily accept is the “good WIFE:” the weightless, invisible, free, and effortless-to-use device. In practical terms, this means that lower-profile orthoses come closer to meeting these idealized goals than more extensive designs simply because they are generally lighter, less conspicuous, less costly to create and maintain, and easier for the patient to master. Therefore, orthotists have developed a large array of ankle-foot orthoses (AFOs) to address many biomechanical deficits that affect gait.

An effective orthosis can be designed to support or control the impaired limb using a ground-up strategy. Application of well-established principles from the use of foot orthoses permits the skilled practitioner to ensure that the affected leg has a stable foundation. When it is necessary to cross the ankle, the additional biomechanical control that an intimately fitted, custom-made shank segment provides can build on the skeletal alignment the foot segment of the orthosis provides.4

Most orthoses have an effect on the adjacent joints just external to the device itself and, particularly for lower limb applications, this biomechanical principle can be used therapeutically. Perhaps the best-known example of the application of indirect control to the next proximal joint is the floor reaction AFO originally described by Saltiel,5 who was an orthotist with post-polio paralysis of one leg.

In the hands of a skilled orthotist, many lower limb gait deficiencies can be significantly improved by provision of a custom AFO, so KAFOs are unnecessary in many instances. Even when the effectiveness is uncertain, trial with an AFO is sometimes recommended before consideration of a KAFO. Because rehabilitation experts tend to reserve the application of higher profile orthoses to only those cases for which an AFO will not suffice, KAFOs are commonly restricted to only the most complex cases with the greatest biomechanical deficits. Some of the reduction in short- and long-term acceptance may be the result, in part, of the complexity of the impairments that are managed with KAFOs.

It is interesting to note that there is a similar pattern showing decreased long-term use of a prosthesis when comparing amputations above the knee joint versus those cases who retain a functional biological knee.6,7 This consideration is particularly significant for individuals with bilateral limb loss,8 where having at least one functional biological knee joint is believed to significantly increase the odds for long-term use of an artificial limb.9

The reasons for the apparent value of an intact knee have not been objectively verified, but common explanations include the increased difficulty that loss of knee function causes when sitting down or returning to a standing position, kneeling, walking briskly, turning, and recovering balance. It may also be that increased body weight and height make overcoming these challenges more difficult, as it has been shown that children often accept and use even complex HKAFO designs until they approach adult stature.10 It then appears that those individuals who retain at least one functional biological knee continue to use orthoses into adulthood while the balance of these patients prefer wheeled mobility for traversing both short and long distances.

It is believed that retention of at least some voluntary hip flexion or extension control increases acceptance of KAFOs and HKAFOs.11 Particularly for pediatric patients, various means have been developed to mechanically link bilateral KAFOs so that motion at one hip joint induces the opposite motion at the contralateral hip joint. Orthotist Wally Motloch is credited with promulgating the concept of using mechanical links to create a reciprocating gait, and his isocentric reciprocating gait orthosis (IRGO) has been shown to result in a more energy efficient gait than similar alternatives.12,13

The use of RGOs for adults is less well established than applications for children with spina bifida. The general belief within North America seems to be that although motivated and physically fit patients may well use such extensive orthoses during the rehabilitation process and after discharge, most adults gradually come to prefer wheeled mobility and the orthosis is used primarily for standing tasks such as working at a counter or stovetop, for transfers, or for ambulating very short distances on familiar and level surfaces.14

It is interesting to note that Stallard and others within the United Kingdom report much higher long-term acceptance and use rates for the Orthotic Research and Locomotion Assessment Unit (ORLAU) ParaWalker, which is one type of hip guidance orthosis (HGO).15 HGOs, which are exoskeletal HKAFOs customarily worn over street clothing, have not been widely used in the United States despite studies showing that certain designs provide mechanical advantages compared to other linked KAFO options presently available.16,17

Some evidence exists to support the notion that walking speed and gait efficiency influence long-term use of devices to assist in ambulation. Researchers from Rancho Los Amigos have suggested that, unless users achieve a comfortable walking pace that permits crossing an urban street before the traffic light changes, long-term use of orthoses for ambulation will not be sustained.18 Merritt and Yoshida,19 in their excellent review article on KAFOs, define functional ambulation as the ability to cover 75 meters within 1 minute and to traverse a distance of one city block or 250 meters “without undue stress”19; Not only does being restricted to a slow walking pace make ambulation less practical, but it may also make such locomotion less efficient.

The inverse relationship between level of physical impairment and net energy cost of ambulation is well established for both amputees and persons with neuromuscular impairments. Recently, Gard and Fatone20 proposed that the slower pace resulting from higher levels of impairment may, in and of itself, be one reason why energy consumption to cover a given distance increases with such disabilities. There is evidence that when normal subjects walk at significantly slower speeds they must expend more energy to do so. It may be that, like riding a bicycle, there is a certain critical velocity that must be achieved to generate sufficient momentum to traverse efficiently from one point to another without excessive effort.

Such velocity hypotheses raise an interesting question: Would the long-term use of walking devices improve markedly if designers could enable patients with greater levels of impairment to increase their self-selected walking speed to a more normal pace without expending more than normal effort? If so, then gait velocity rather than gait symmetry may be the more important characteristic to optimize in the design of the orthosis.

The energy required to ambulate with various neuromuscular disorders has been studied in some depth. These data are very well summarized in the 1999 review article by Mulroy and Waters.21 It is clear that immobilizing lower limb joints in normal subjects decreases the efficiency of gait and that blocking knee motion throughout the gait cycle is more harmful than immobilizing the ankle, whereas locking both joints results in the most laborious gait. These data may have important implications for KAFO prescription and design.

Historically, the great majority of KAFOs have locked the impaired knee in full extension throughout the gait cycle. Parisian therapist Eric Viel22 summarized the situation in the 1960s: “Lower extremity ortheses [sic] are nothing but glorified knee locks, slightly modified to suit individual tastes.”

Unfortunately, this statement aptly summarized the state of the art worldwide until very recently. Although it is true that KAFOs that prevent hyperextension while permitting free sagittal plane motion of the knee have been available for many decades,23 these designs are used very selectively for adults with knee control impairments. Even with a knee joint that is offset posteriorly, such “free knee” KAFOs are stable only when there is an external extension moment at the knee. Such ground reaction stability is reliable only on level surfaces. Declines or irregularities in the walking surface can cause the force vector to abruptly change to a flexion moment, resulting in immediate collapse of the knee without warning. The inevitable risk of falls and injury combined with patient trepidation about a knee joint that can collapse without warning at any time makes the use of such KAFOs tenuous at best for most adults.

Within the past 6 years, a number of novel knee joints became available permitting the orthotist to create a KAFO that stabilized the knee during weight bearing but released automatically to permit knee flexion during swing phase.24 Preliminary studies of these stance control orthoses (SCOs) have demonstrated a strong patient preference and significantly less abnormal kinematics than when ambulating with the knee locked.25 Subjective perception of the difficulty ambulating appears to decrease with the use of stance control knee joints but the objective data regarding energy efficiency are limited and inconsistent at this time.26 The heterogeneity of the populations studied, with numerous diagnoses and widely varying patterns of biomechanical deficits, suggests that it may be useful to stratify subjects in future studies into more homogeneous groupings.


From the orthotist’s perspective, KAFO designs are driven primarily by the specific biomechanical deficits of each individual patient. Although most KAFOs look superficially similar, the biomechanical control they provide at the ankle, knee and—for HKAFOs—at the hip are often widely divergent. This makes it difficult to interpret research data unless the design of the orthosis and the biomechanical control of each joint and limb segment have been completely described.

As a general rule, orthoses should interfere with joint motion as little as possible while being consistent with the overall goals of patient safety and effective ambulation. Thus, the patient with advanced osteoarthritis whose primary complaint is painful genu varum and associated ankle varum may do well with a KAFO that effectively stabilizes the knee and ankle in the coronal plane but does not otherwise interfere with normal hip, knee, or ankle motion. A different patient of the same general age and stature with painful arthritis at the ankle that is aggravated by motion plus chronic knee pain secondary to the loss of almost all intra-articular cartilage might do better with a similar-looking KAFO that completely immobilizes both knee and ankle during weight-bearing. Although these KAFOs are radically different biomechanically, the hypothetical patients share a common medical diagnosis. This illustrates that the patient’s diagnosis is a poor predictor of the need for an orthosis or the type of orthosis that will be most effective. The best criteria for prescription and provision of a KAFO are based on the biomechanical function to be provided to address each patient’s unique pattern of neuromuscular deficits. Medical diagnosis can give clues about the biomechanical losses to be expected, but there is no substitution for a detailed physical examination of each potential KAFO wearer before developing a detailed prescription recommendation.

Harris27 has described a detailed and systematic method to specify the biomechanical performance desired at each joint and in each plane. Although the form itself is rarely used in clinical practice, it is important for the clinic team to clearly communicate all such expectations to the orthotist to ensure that design of the KAFO addresses as much of the patient’s deficit as is practical.

After the desired biomechanical controls at each joint and limb segment have been determined, selection of the components and materials to accomplish the desired outcome can commence. When multiple options exist, the orthotist considers the cost, weight, safety, and functional reliability of similar components to select the best alternative.

Some aspects of KAFO design are based on technical considerations or material science principles. For example, small children can often safely use KAFOs with only a lateral side bar, whereas most adult devices have both medial and lateral side bars to increase structural stability. Greater body weight or higher activity levels may dictate the use of specific materials as will efforts to minimize the weight of the final device.

Thermoplastics are tough but somewhat flexible, whereas carbon fiber composite structures are thin, stiff, and lightweight. Aluminum alloys are lighter than steel but subject to fatigue failures, whereas titanium alloys provide strength that is similar to steel elements but at a weight that is nearly as light as aluminums. Orthotists commonly create a custom KAFO that combines several of these materials to optimize the overall performance of the device while reducing the total cost and weight.

Other technical considerations could include the commercial availability of a specific component or material in a particular setting, whether these items can be serviced and repaired locally, and long-term reliability under local conditions. For example, resistance to corrosion is of major importance in oceanfront climates and also for pork farmers, due to the corrosive humidity associated with both settings, but this KAFO characteristic is of much less concern in most landlocked locales. To ensure an optimal outcome, the orthotist must determine such details of KAFO design based on a thorough understanding of each patient’s living and working situation.

The structural configuration of KAFOs varies, with numerous variants described in the literature. Orthotists try to create an orthosis that is not only structurally safe and functionally reliable but also easy for the user to don and doff. There is some evidence that certain designs are easier for patients to manage than others,28 but these design considerations are typically determined by the orthotist, based on individual patient preference.

Finally, it must be noted that whether or not to prescribe KAFOs and associated HKAFO variants to aid in ambulation varies from center to center and from country to country, suggesting that local custom is one factor in this decision. The absence of any strong objective evidence about the use of custom KAFOs to assist in walking after cerebral vascular accident (CVA) is one reason such devices are rarely provided in the United States at this time29 but are advocated in Japan to treat the same condition.30–32 Successful applications of KAFOs with electronic stance control knee joints to hasten post-CVA rehabilitation have recently been reported, whereas other centers are investigating the utility of “walking robots,” such as the Lokomat device.33 As we move toward evidence-based practice, clinicians are eager for more objective information to help them decide with confidence between such competing treatment philosophies.


There is no universally agreed-on method to classify the myriad KAFO designs into logical groupings. At present, these devices are commonly referred by eponymous acronyms such as RGO, ARGO, IRGO, and HGO, but this is more a matter of convenience than a formal classification system per se. Although there is general agreement that orthoses are not diagnosis specific, at present no comprehensive alternative method to group KAFOs into functional categories has been promulgated.

It may be useful to consider development of a matrix that is based, at least in large part, on the primary biomechanical characteristics of the KAFO as well as key clinical considerations, such as the presence of significant spasticity or the level of neuromuscular impairment. Factors to develop such a clinical classification system may include:

  • Whether the knee joint is locked in extension during swing phase
  • The type of hip joint control provided, if any
  • Whether the KAFOs are worn unilaterally or bilaterally
  • The degree of spasticity present at the hip, knee, and ankle joints
  • The strength and range of motion at the ipsilateral hip joint
  • Residual sensation or proprioception in the affected limb

Despite the somewhat discouraging reports in the literature of the long-term rejection rates for KAFOs and HKAFOs to assist in ambulation,34 the recent development of orthotic joints that stabilize the knee during stance phase without interfering with swing phase knee flexion requires rethinking of the old paradigms that were based on locked-knee technology. Although it is unlikely that these biomechanically improved components will overcome all of the shortcomings of prior KAFOs, they clearly result in a less abnormal gait pattern for many patients and seem likely to improve the acceptance rates for at least certain individuals. Investigations are currently underway to examine the effect of augmenting the biomechanical control of microprocessor-controlled stance control KAFOs with muscle activation by surface functional electrical stimulation.35 Further experience and research are required to determine the clinical indications and limitations of this latest addition to our KAFO armamentarium.

Development of a clinical classification system might help busy practitioners to more clearly identify the utility of specific KAFO designs for particular patients, as this choice is likely to remain a largely empirical decision for the foreseeable future.


1.Hoffer MM, Feiwell E, Perry R. Functional ambulation in patients with myelomeningocele. J Bone Joint Surg [Am] 1973;55:137–148.
2.Mikelberg R, Reid S. Spinal cord lesions and lower extremity bracing: an overview and follow-up study. Paraplegia 1981;19:379–385.
3.Subbarao JV. Walking after spinal cord injury: goal or wish? West J Med 1991;154:612–614.
4.Fish DJ, Nielsen J-P. Clinical assessment of human gait. J Prosthet Orthot 1993;5:39–48.
5.Saltiel J. A one-piece laminated knee locking short leg brace. Orthot Prosthet 1969;23:68–75.
6.Mueller MJ, Delitto A. Selective criteria for successful long-term prosthetic use. Phys Ther 1985;65:1037–1040.
7.Cutson TM, Bongiorni DR. Rehabilitation of the older lower limb amputee: a brief review. J Am Geriatr Soc 1996;44:1388–1393.
8.Dougherty PJ. Long-term follow-up study of bilateral above-knee amputees from the Vietnam war. J Bone Joint Surg [Am] 1999;81:1384–1390.
9.Brodska WK, Thornhill HL, Sarapkar SE, et al. Long-term function of persons with atherosclerotic bilateral below-knee amputation living in the inner city. Arch Phys Med Rehabil 1990;71:895–900.
10.Roussos N, Patrick JH, Hodnett C. A long-term review of severely disabled spina bifida patients using a reciprocal walking system. Disabil Rehabil 2001;23:239–244.
11.Hussey RW, Stauffer ES. Spinal cord injury: requirements for ambulation. Arch Phys Med Rehabil 1973;54:544–547.
12.Winchester PK, Carollo JJ, Parekh RN. A comparison of paraplegic gait performance using two types of reciprocating gait orthoses. Prosthet Orthot Int 1993;17:101–106.
13.Harvey LA, Davis GM, Smith MB. Energy expenditure during gait using the Walkabout and Isocentric Reciprocal Gait Orthoses in persons with paraplegia. Arch Phys Med Rehabil 1998;79:945–949.
14.Franceschini M, Baratta S, Zampolini M. Reciprocating gait orthoses: a multicenter study of their use by spinal cord injured patients. Arch Phys Med Rehabil 1997;78:582–586.
15.Major RE, Stallard J, Farmer SE. A review of 42 patients of 16 years and over using the ORLAU ParaWalker. Prosthet Orthot Int 1997;21:147–152.
16.Stallard J, Major RE. The case for lateral stiffness in walking orthoses for paraplegic patients. J Eng Med 1993;2007H:1–6.
17.Banta JV, Bell KJ, Muik EA. ParaWalker: energy cost of walking. Eur J Pediatr Surg 1991;1:7–10.
18.Waters RL, Miller L. A physiologic rationale for orthotic prescription in paraplegia. Clin Prosthet Orthot 1987;11:66–73.
19.Merritt JL, Yoshida MK. Knee-ankle-foot orthoses: indications and practical applications of long leg braces. Phys Med Rehab State of the Art Rev 2000;14:395–422.
20.Gard SA, Fatone S. Biomechanics of lower limb function and gait. In: Condie E, Campbell J, Martina J. Report of a Consensus Conference on the Orthotic Management of Stroke Patients. Copenhagen: International Society for Prosthetics and Orthotics; 2003:55–63.
21.Waters RL, Mulroy S. The energy expenditure of normal and pathological gait. Gait Posture 1999;9:207–231.
22.Viel E. Critique of lower extremity bracing. Orthot Prosthet 1968;12:28–33.
23.Lehmann JF, Warran CG, DeLateur BJ, et al. Biomechanical evaluation of axial loading in ischial weight-bearing braces of various designs. Arch Phys Med Res 1970;51:331–337.
24.Michael JW, McMillan AG, Kendrick K. Stance control orthoses: history, overview, and case example of improved KAFO function. Alignment 2003:60–70.
25.McMillan AG, Kendrick K, Michael JW, Horton GW. Preliminary evidence for effectiveness of stance control orthoses. J Prosthet Orthot 2004;16:6–13.
26.Hebert JS, Liggins AB. Gait evaluation of an automatic stance-control knee orthosis in a patient with postpoliomyelitis. Arch Phys Med Rehabil 2005;86:1676–1680.
27.Harris EE. A new orthotics terminology: a guide to its use for prescription and fee schedules. Orthot Prosthet 1973;27:6–19.
28.Peethambaran A. The relationship between performance, satisfaction, and well being for patients using anterior and posterior design knee-ankle-foot-orthosis. J Prosthet Orthot 2000;12:33–45.
29.Waters RL, Garland DE, Montgomery J. Orthotic prescription for stroke and head injury. In: Atlas of Orthotics, 2nd ed. St. Louis: CV Mosby;1985:270–286.
30.Kakurai S, Akai M. Clinical experiences with a convertible thermoplastic knee-ankle-foot orthosis for post-stroke hemiplegic patients. Prosthet Orthot Int 1996;20:191–194.
31.Fujimoto M, Abe K, Makajima H. Usefulness of KAFOs as training orthoses for severe flaccid stroke patients (abstract). Arch Phys Med Rehabil 1999;80:1202.–
32.Bellantoni L, Heffernan C, Oyedijo O. The effect of knee-ankle-foot orthosis on ambulation outcome after stroke (abstract). Clin Res 1992;40:119A.
33.Greshoffer P. The Lokomat, a new possibility in the rehabilitation of stroke patients. J Neurol 2004;251(Suppl 3):28–31.
34.Oliver M. The misuse of technology: walking appliances for paraplegics. J Med Eng Technol 1977;2:69–73.
35.Stein RB, Hayday F, Chong S, et al. Speed and efficiency in walking and wheeling with novel stimulation and bracing systems after spinal cord injury: a case study. Neuromodulation 2005;8:264–271.

ambulation; hip-knee-ankle-foot-orthosis; HKAFO; KAFO; knee-ankle-foot-orthosis; lower limb orthoses

© 2006 American Academy of Orthotists & Prosthetists