Repetitive Strain Injury in Upper Limb Amputees

W. Hill, S. LeBlanc, L. Manor, E. Biden


Atlantic Clinic for Upper Limb Prosthetics is a specialty upper limb prosthetics clinic based in Fredericton, New Brunswick. The Clinic has been in operation since 1981 and typically has a caseload of approximately 140 upper limb amputees drawn primarily from Atlantic Canada (New Brunswick, Nova Scotia, Prince Edward Island, and Newfoundland and Labrador). The overall population of our catchment area is approximately 2.3 million. For some time we have been curious as to whether our clinic population reported similar levels of repetitive strain issues as reported in the literature.

To address this question, we present data from a survey of our working age clinic population aimed at assessing risk factors for repetitive strain injuries and a similar survey for a control group of normally limbed individuals. The surveys were conducted in the summer and fall of 2012 following a protocol approved by the University Research Ethics Board.


Overuse injuries, also referred to as repetitive strain injuries (RSI's) or cumulative trauma disorders in unilateral upper limb amputees have had relatively few reports in the literature. Such injuries can be caused by relatively low loads where there are many repetitions or by higher loads with fewer repetitions. RSIs can also be associated with physical exertions causing high loads in the joints, vibration, mechanical compression, and awkward positions. RSI's fall in many categories including carpal tunnel syndrome, tendonitis, trigger finger or tennis elbow. In many cases the repeating loads are much less than would be considered a risk for a single loading cycle.

Three papers of note are Jones and Davidson(1999) who reported on a survey of upper limb acquired amputations in Australia, Gambrell (2008) who built on Jones and Davidson with a literature review of related studies and Østlie et al (2011) who reported on a large group of acquired amputees along with a control population drawn from Norway.

All three papers conclude that RSI's are a substantial risk for the amputee population with the two survey based studies both showing wrist and hand pain in over half of their populations. By contrast the control study of Østlie et al found only about 10% of the normally limbed controls reported such pain.


The project developed two surveys, one directed at working age individuals from our clinic population and the other at normally limbed individuals of similar age. In addition some other data were gleaned from our clinic database.

The survey of the clinic population established the level and type (congenital, traumatic, disease related) and the time since amputation. Respondents were asked what type of prosthesis they used, as well as questions about how frequently they used a prosthesis. Questions were designed to elicit information on activities associated with RSI risk including school or work activities, computer use, sport and recreational activities, and hobbies.

Respondents were asked about employment or school, how much they worked, and whether and how they used their prosthesis at work or in their educational pursuits. Questions on posture, lifting, and sitting were posed. Participants were asked about whether they had pain, the location of the pain and were polled as to pain intensity using a visual analog scale from 0 to 10 where 0 was no pain and 10 was the worst possible excruciating pain. Finally, participants were asked for their age, height and weight, which allowed calculation of their Body Mass Index.

For the control survey of normally limbed subjects, participants were asked the same series of questions regarding upper limb function and discomfort. The structure of this survey matched closely the original survey of the clinic population. The data generated informed the question of whether the prevalence of repetitive or overuse injuries is higher in the amputee population than in the general public.

Reviews of charts identified 75 individuals who were considered to be of working age, basically 16 and older, and were at the transradial or transhumeral level without other medical complications. These people were contacted either during a visit to the clinic, by phone or by mail and a survey of their activities and related pain levels was conducted.

Of the 75 amputees identified, 13 could not be found. Of the remaining 62, 33 respondents completed the survey. Of these, 14 were female and 19 were male. Seventeen were congenital and most of these had been followed by our clinic since birth.

We specifically asked questions in such a way as to acknowledge that people might have phantom limb pain, but at the same time focused on pain that seemed to occur along with activities of daily life. To this end about 13% of the congenital amputees and 33% of the acquired amputees reported phantom pain.

Results & Discussion

Figure 1 shows the age composition of our clinic in 2011 when the clinic had been in operation for 30 years. Each dot is a person. Square dots represent people with congenital limb deficiency and the median age is 23 years. The round dots represent acquired limb loss and the median age is 43. Having been fitting prostheses for over 30 years and having a mandate for all ages from infants to adults, we have a wide age range.

For the clinic population, the individuals with congenital limb loss were 88% transradial. For the acquired group 62% were transradial and 31% were transhumeral. Time since amputation varied from new amputees to 60+ years since limb loss.

Figure 2 shows the reported intensity of pain for both the affected and unaffected sides. For box plots, such as are shown here, the box represents the 25th to the 50th percentiles and the whiskers represent the limits of the data except for significant outliers which are indicated by a *. The line in the mid portion of the box shows the median value and the bull's eye shows the mean.

Median pain levels reported were in the 3 to 4 range which was described as "Nagging uncomfortable or troublesome". Isolated individuals had levels up to 8 on the scale taking them to the "intense" level. Individuals with limb loss indicate pain on both sides.

When we compare reported pain for the clinic population and control group (Table 1 ), the significant differences (indicated by a *) are in the hand and thumb, and wrist and arm, which are prime locations for RSI's to develop. This comparison is done using the sound side for those with limb loss.

When we compare pain intensity at the sound wrist for those with limb loss versus normally limbed individuals, the limb deficient individuals report substantially higher pain levels. Most of these people rate their pain at the "nagging and troublesome" level. Perhaps not surprisingly, those with congenital limb deficiency report lower median pain levels than the acquired group for the affected side and somewhat lower median pain levels on their sound limb.

The group reporting pain in their residual limb generally have a higher BMI than those with no pain.

Figure 3 shows that heavier, higher BMI, individuals indicate more frequently that they make limited use of their prosthesis.

Texting emerged from our analysis as a potential risk factor. Of the clinic population, 79% report texting at least occasionally which is actually higher than the 73% of normally limbed who did (Table 2 ). Of those in the control group who reported being texters, only 18% reported any pain at all and less than 3% reported more than minimal pain levels. By contrast 57% of the clinic population who were texters reported pain and the mean level was in the "nagging and troublesome" category.

Discussion & Conclusions

Pain incidence and pain levels reported for the clinic population in this study are much more prevalent than in a control group of normally limbed individuals and are generally at a higher level. We are concerned that the risks of RSI's are higher in our clinic participants and that these people should be encouraged to avoid activities which could harm them in the long term.

Widespread texting is a relatively new activity but may have long term consequences.

We are currently developing a study protocol to examine force levels during texting and also the amount of texting which can occur before people report symptoms. If you are interested in contributing to the study reported here or to our upcoming studies, please contact Wendy Hill OT at .


The authors acknowledge the assistance of Greg Bush, Research Prosthetist and David Smith Master's Candidate and currently prosthetics student in the development and execution of this work.

W. Hill, S. LeBlanc, L. Manor, E. Biden are affiliated with Atlantic Clinic for Upper Limb Prosthetics, Institute of Biomedical Engineering, University of New Brunswick, Canada