Splinting the juvenile arthritic wrist: A clinical observation



In children as well as in adults, the wrist is frequently involved in rheumatic arthritic processes. With persistent activity of rheumatoid arthritis (RA), the adult wrist joint will become affected bilaterally in 95% of the patients, while in children the wrist is affected in approximately 60% of the cases (1–5). Only the knee joint is more frequently involved in children. One of the earliest clinical signs in children is the loss of full extension of the wrist, even before the palpable synovitis can be detected. To diagnose and treat wrist involvement in an early phase of the disease, and to prevent the typical rheumatoid changes of the wrist and hand, the underlying process leading to this deformity must be understood. This understanding is far from complete, as our knowledge is based mainly on the pathogenesis and radiologic changes that occur during the course of the disease. Radiographic assessments particularly focus on erosive changes and the existence of malalignment, which have been shown to correlate well with clinical measures of joint deformity and impaired range of motion (6).

Hand function in children is very important in the development of skills needed for daily childhood activities; therefore, preserving this function warrants special attention. One of the interventions aiming to preserve function is splinting of the affected wrist joint. The rationale for the use of splinting is to balance rest and activity, thereby preserving wrist function. However, the evidence-based knowledge to support this rationale is scarce. To determine the validity of “balancing rest and activity of the affected wrists to preserve function,” we reviewed our own data, data in the literature regarding rheumatic wrist problems in children, and data from a 15-year followup study of rheumatic wrist problems in an adult cohort of our adult rheumatology counterpart within the medical center.

Type of adult wrist involvement in rheumatoid arthritis

Van Vugt et al found at least 4 types of early wrist involvement (7) in a 15-year followup of 20 adult patients (40 wrists). Starting from baseline radiographs in the initial phase of the disease, they measured angular and linear displacement to quantify malalignment of the wrist and fingers. They used linear displacement of the center of rotation, while calculating carpal–ulnar distance and the ratio of each distance to the length of the third metacarpal. A value for linear malalignment was obtained by comparing this ratio of a malaligned wrist to that of the same wrist at baseline. Angular measurements were obtained using Shapiro's coordinates (8) to determine the amount of radial deviation of the wrist and ulnar drift of the fingers.

Based on the first localization of radiologic damage in the initial stage they discriminated 1) the radial type, 2) the ulnar type, 3) the central type, and 4) a diffuse type involvement (Figures 1 and 2). Radial deviation of the wrist was increased in the centrally affected wrist compared to the diffuse involvement. Furthermore, radial deviation of the wrist was positively correlated with ulnar drift of the fingers. Radiologic abnormalities were progressive and correlated with disease duration, but showed individual differences in their course. Almost all patients showed linear malalignment.

Figure 1.

Three different types of early wrist involvement. First location of radiologic damage in the A, radial region (radial type), B, central region (central type, with decreased carpal height due to arthritic disruption of the scapholunate ligament), and C, ulnar region (ulnar type).

Figure 2.

Diffuse type wrist involvement.

Because of its complicated structure, the wrist is prone to deformity, subsequently leading to disability of the hand. An anatomic as well as functional characteristic plays an important role in the development of wrist problems. Other factors include loss of bone structure and ligament stability secondary to synovitis, tendon (sub)luxation, and intrinsic muscles contractures (9, 10).

However, the underlying mechanism of this deformative course remains unknown. The central type of wrist involvement, as described by van Vugt and colleagues (7), induced more deviation of the wrist than the other types. Also, based on radiographic scoring at 5 years, van Vugt et al found the wrist to be more severely affected than the fingers, possibly indicating the role of malalignment of the wrist in the development of hand deformities. Shapiro (8) also indicated that the “key” to rheumatoid deformities may be found in the wrist. To our knowledge there are no publications describing the existence of different types of wrist involvement in juvenile rheumatology.

The juvenile wrist

We found abnormalities (i.e., periarticular osteoporosis, osteophytosis, narrowing of the jointspace, and growth disturbances of the epiphysis) at both radiocarpal junctions (11) in 21 of the 23 patients included in a cross-sectional study of children with polyarticular juvenile arthritis visiting our rheumatology outpatient department in a 6-month period. A local growth disturbance almost always results in a relative shortening of the ulna with respect to the radius. Whether this relative shortening is caused by overgrowth of the radius or undergrowth of the ulna due to premature closure of the epiphysis seems to be determined by age (i.e., the developmental stage in which the wrist is affected by the disease). This growth problem is characteristic for juvenile arthritis, and marks a significant difference between adult RA and juvenile RA (12). From our clinical experience erosive changes in the capital bone most frequently occur at the central carpometacarpal joint of metacarpus 3 and at the intercarpal joint space between os hamatum and os capitatum. This could be recognized as the juvenile representation of the central type of wrist involvement seen in adults. Multiple erosions and fusions, occurring almost simultaneously, between the distal carpal row and metacarpus 1–5 are also often seen. This could be viewed as a combined radial–central–ulnar type of involvement, and may best be described as a transsectional type of juvenile wrist involvement.

Malalignment in juvenile wrists

Malalignment is an important feature of wrist deformity in adult and juvenile RA patients (Figure 3). In juvenile wrists, it is commonly associated with ulnar deviation and/or subluxation (12, 13). However, as we have seen in the adult wrist, patterns vary individually, and the mechanisms underlying this pathokinesiologic phenomenon remain unclear.

Figure 3.

Radiograph of the left hand of a 10-year-old girl, showing malalignment of the carpus, shortening of the ulna, and slanting of the radius.

Because treatment and the understanding of the pathophysiology of malalignment rely on assessment of the malalignment, valid assessment is crucial. Therefore, we studied the methods for analyzing and assessing malalignment in 40 juvenile wrists (14) and determined the assessment of malalignment of the wrist should include measurements representing both modes of displacement with planar motion: rotation and translation.

Two posteroanterior radiographs were taken for each child, using a standardized technique. The first radiograph was taken with the wrist resting in neutral position (the third metacarpal was placed parallel to the longitudinal axis of the forearm), and the second radiograph was taken with the wrist in the “position of comfort” (the child was asked to place his or her hand/wrist in the position that felt most comfortable). After the radiographs were digitized, osseous landmarks were identified, angles measured, and lengths calculated from the x and y coordinates. Measurements based on the approximate to the wrist's center of rotation (Acr), according to the method of Youm et al (15), were used to assess the translation mode of displacement of the wrist. The perpendicular distance between the Acr and the longitudinal axis of the forearm and the radius, the carpal–forearm distance (CFD) (16), and carpal–radial distance (CRD) (17) respectively, were used as measurements. Ratios were calculated using the intermetacarpal width as defined by Poznanski et al (18), because this measurement is less influenced by growth disturbances affecting metacarpal length in children (19).

If Acr functions as a true or reasonable approximate to the center of rotation in children, by principle CFD and CRD ratios would remain the same between two different positions of a wrist. We found however, that the location of this approximate to the center changes slightly with changes in wrist position. These changes were statistically significant, and therefore this line of assessment is probably not clinically valid in children.

Regarding rotation we found that what is considered to be the “normal” position of the wrist, and thus the angle that is used to assess rotation, has profound implications on the conclusions to be drawn from that assessment. Furthermore, the problem of positioning of the wrist while taking radiographs is probably underrated or underestimated. Radiologic protocols are needed to avoid artifacts and misinterpretations of affected wrists. It has been suggested that the patient be positioned lying down, with the shoulder slightly flexed, the elbow at 90 degrees, and the palm up (20), but due to contractures this may not always be possible.

In a different study, we investigated the displacement response of the wrist during grasp as seen on radiographs (21). We hypothesized that due to ligamentous instability, the increased muscular force during grasp would induce ulnar displacement, as is theorized in adult patients. Radiographs of the wrists were taken in 30 children. Standardized techniques were used, with the wrist in a resting position, and the child grasping an inflated cuff of a sphygmomanometer and exerting maximum possible strength. With grasp, we found an increase in carpal narrowing, an ulnar lunate displacement, and a decrease in ulnar variance. These changes are similar to those found in the healthy wrist as described by Schuind et al (22). Furthermore, these changes suggest that the wrists of juvenile RA patients act in accordance with the generally accepted explanation for the development of malalignment of the wrist in adult RA patients.

However, radial displacement of the lunate, though slight, was found in 2 wrists, and the amount of ulnar displacement varied substantially (3.1% to 22.5%). Not all wrists responded to the increased muscular compressive force to the same extent, indicating individual differences, also noted before in the adult cohort. However, one child with Down syndrome and juvenile rheumatoid arthritis, showed a substantial displacement. Because hypermobility is known to exist in Down syndrome, this could indicate that during grasp the ligamentous loss of constraint leads to a greater displacement response, indicating that grasp could be an indirect measure of ligamentous instability. We concluded that measuring displacement response during grasp in juvenile arthritic wrists is not easy, as it requires strict radiologic protocols and consensus about the interpretation of previous and current radiologic findings. We are currently performing a long-term followup on radiographs made with and without grasp, and based on a strict radiologic protocol, to collect information on the course of the radiologic changes over time.


There is a current tendency to treat rheumatic problems in general, and wrist joint problems specifically, rather aggressively with inflammation suppressive medication, intraarticular steroid injections, and synovectomy. However, the problem of how to best prevent deformity of the juvenile wrist and hand remains. Physical and occupational therapy are the most frequently used nonmedical and nonsurgical interventions. The goals of these interventions are to prevent further damage to the affected joints, to preserve function, and to strengthen the periarticular muscles. The preservation of function usually means splinting of the affected wrist. Commonly, a standardized form of splinting is used in which the wrist is placed in pronation, with the device extending to the midpalm of the hand. Fingers are free to move, but the wrist is “laid to rest” in a “neutral position.” Neutral often means a zero-degree position, and a functional position of the wrist usually means a 15 to 20 degree extension position with a slight ulnar deviation. However, the intraarticular pressure, which is increased in the acute phase of the disease as a consequence of inflammation, is the lowest in a flexed position, thereby protecting the wrist joint from (further) inflammatory damage. The best method to balance movement and rest of the wrist and fingers is unclear, because we do not yet know the impact of muscular force on wrist malalignment.

Some clinicians suggest that splinting is important to prevent daily use of the hands in a malposition, because permanent misloading results in fixed joint deformity. Therefore, splints should be worn most of the day. However, there is no evidence from controlled trials to substantiate that rationale.

Hypotheses emphasizing over- and/or underpower of the different flexors and extensors of the forearm have been described by a great number of authors (4, 21, 23–27). We therefore tried to gain insight into the role of forearm muscles in wrist malalignment by using interactive graphics based musculoskeletal modeling (28). With this modeling, the effect of musculoskeletal geometry, joint kinematics, and muscle–tendon parameters on muscle tendon lengths, momentums, muscle force, and joint movements can be quantified (29, 30). Individual models of the wrist were constructed of one healthy and 3 children with juvenile RA with different expressions and stages of malalignment. Joint movements were calculated and shear forces were estimated by applying rigid body spring modeling to the modified wrists, based on the method described by Schuind et al (31). In this study we confined the analysis to the 2-dimensional, frontal plane. Because more details regarding the modeling of the wrists are beyond the scope of this contribution, the reader is referred to the actual study (28) and subsequent findings (32) for more information.

Radial joint movement for all patients with juvenile RA was lower than found in the healthy subject, while for the JRA patients with ulnar–carpal translation, shear forces were large compared to the shear force found in the healthy subject. This would suggest that translation might still increase.

To our knowledge, this was the first study that attempted to determine the role of the forearm muscles in the course of wrist malalignment in juvenile RA. The availability of juvenile RA wrist models allow further experimental research into the pathokinesiologic mechanism underlying wrist malalignment, and can give additional information on how best to intervene (with or without splinting).

Meanwhile, we should question whether adult and juvenile arthritic wrists need the same orthotic approach: static and in a neutral position. In children, local growth disruptions and dysmorphic growth can aggravate malalignment. Malalignment in itself hampers wrist function and may aggravate drift of the fingers. So it seems of utmost importance to create the opportunity for normal osseous configuration and maintain normal alignment in a timely manner. This might mean that some form of correction should be built into the orthotic device, or perhaps it means the wrist and fingers should be allowed to freely move in a functional way but under “corrected anatomic and kinesiologic circumstances.”

Because wrist and hand deformity place a heavy burden on the child with arthritis and his or her family, more research into basic pathokinesiologic mechanisms and prevention is urgently needed.


We would like to thank Richard van Vugt, MD, PhD, for providing the radiographs of his patients, and Sonja Raaff for her secretarial assistance.