A low physical activity level is associated with increased cardiometabolic risk factors (Biswas et al., 2015; Leskinen et al., 2018), and time spent in a sedentary posture (i.e., sitting) is associated with cardiovascular risk (Tigbe et al., 2017). Many individuals with a lower limb amputation are not physically active enough to meet health recommendations (Langford et al., 2019), increasing the risk of developing metabolic syndrome (Bhatnagar et al., 2019). Among amputees, positive relationships have been reported between prosthetic mobility and quality of life and general satisfaction (Wurdeman et al., 2018) as well as between physical activity and body image (Wetterhahn et al., 2002). If the amputation includes loss of the knee joint, as in the case of a transfemoral amputation (TFA) or knee disarticulation (KD), prosthesis mobility is generally more challenging, and walking demands greater energy than for those with lower amputation levels (Ettema et al., 2021; van Schaik et al., 2019).
Most research investigating mobility in individuals with lower limb amputations has focused on gait analysis and performance tests assessed in a laboratory or rehabilitation environment or by capturing patient-reported outcomes on questionnaires. However, these assessments do not fully describe real-life daily activity. Recently, step monitors and more advanced accelerometers, that is, activity monitors, have allowed researchers to assess daily activity. Two recent reviews have focused on studies involving different sensors (Chadwell et al., 2020) and outcomes (Mellema & Gjovaag, 2022) when monitoring daily activity in individuals with a lower limb amputation. Both reviews concluded that the most reported outcome was the number of daily steps, which is commonly assessed using a step monitor attached to the prosthetic ankle. Other available devices and outcomes include phone-based accelerometers, pedometers attached to the waist, sensors embedded in the prosthetic socket, or accelerometers attached to the nonamputated limb or the prosthetic limb. Depending on the study aim, real-life activity among amputees may be best monitored on the prosthesis, capturing activity only while wearing the prosthesis, or by attaching the sensor elsewhere on the body, reflecting the overall activity of the individual. Hypothetically, by simultaneously monitoring the activity on the prosthesis and at another location of the body, researchers could assess both the overall activity and, more specifically, activity while using the prosthetic limb. To the best of our knowledge, no studies have reported daily activity of amputees using this method.
One commonly used activity meter is activPAL™. This small sensor is placed on the frontal part of the thigh and collects data on the number of steps, body position, and the time spent on different types of activities over several days (Aguilar-Farias et al., 2019; Edwardson et al., 2017). Using this device, a few studies on amputees, mainly including patients with transtibial amputation (TTA), have reported low levels of physical activity and a majority of time spent sitting (Buis et al., 2014; Miller et al., 2019; Pepin et al., 2019). Furthermore, a few validation studies performed in laboratory settings are available. Salih et al. (2016) assessed 21 individuals (n = 17 with TTA, n = 4 with TFA) performing activities with simultaneous activPAL recording (one unit attached to each thigh). The output was compared to observed performance, revealing acceptable validity (Salih et al., 2016). More recently, Deans et al. (2020) assessed 15 individuals (n = 11 TTA, n = 4 TFA) and concluded that recordings of incidental steps (e.g., steps taken in a confined space) were not reliably captured from the prosthesis (Deans et al., 2020). They recommended that future research focus on real-world conditions.
This study aimed to expand our understanding of daily activity among individuals with lower limb prostheses, especially among those with a prosthetic knee, and to assess the use of activPAL monitors in this population. Our specific aims were to assess overall daily activity in individuals with an established TFA or KD and to explore the amount of activity performed with the prosthesis as compared to overall activity.
Methods
Study Participants
This study included a convenience sample of individuals with unilateral TFA or KD who were established prosthetic users. The inclusion criteria were as follows: living in Sweden, above 18 years of age, used a prosthesis in daily life for at least 1 year, able to walk 100-m indoors without any gait aid or supported by a single cane/crutch, and able to understand instructions and answer questionnaires in Swedish or English. Participants were recruited through an advertisement distributed to clinicians at a university hospital, five selected prosthetic workshops and an amputee social media group. Interested participants were asked to contact the first author (Hagberg) to obtain the full study information. If inclusion criteria were fulfilled and consent to participate was obtained, an appointment was scheduled. Participants had two participation options: an appointment at the university hospital or an appointment at their local prosthetic workshop or other suitable location closer to their home. In each, case the appointment lasted for less than 2 hr and included collecting demographic information, administering self-report questionnaires on prosthetic use and mobility, performing the 2-min walk test (2MWT), and finally, preparing the participant for the activity data measures. The activity data were captured by two activity monitors, one located on the nonamputated thigh and the other on the prosthesis, during a 7-day period. A detailed description of the measures is provided below. The study was approved by the Swedish Ethical Review Authority (Dnr 2020-03762), and all participants provided written informed consent to participate in the study.
Study recordings were collected between September 2020 and December 2021. Due to the COVID-19 pandemic, data collection had to be stopped for 4 months within this timeframe. Before the study started, two pilot assessments were performed; these two individuals were later contacted and provided their written informed consent to be included in the study after ethical approval was obtained. In total, 42 individuals were included, but simultaneous activity measures failed for three individuals; thus, data for 39 participants are reported.
Measures
The demographic information included sex; age; level and cause of amputation; number of years using a prosthesis; other current medical disorders or condition(s) that affected overall activity; employment status; and general information on the current kind of prosthetic suspension, prosthetic knee, and foot. The waist circumference (in centimeters), height (in meters), and weight without prosthesis (in kilograms) were recorded, and body mass index was calculated. In individuals with a TFA, the residual femur length was categorized as short, medium, or long (Persson & Liedberg, 1983).
Self-reported prosthetic mobility included two validated amputee-specific questionnaires: the Questionnaire for Persons with a Transfemoral Amputation (Q-TFA), from which the prosthetic use score and the prosthetic mobility score were recorded (Hagberg et al., 2004), and the Prosthetic Limb Users Survey of Mobility (PLUS-M; Hafner et al., 2017) and a nonvalidated prosthetic activity grade (Supplementary Material [available online]). The two Q-TFA scores range from 0 to 100, with a higher score indicating more time wearing the prosthesis during a week and better prosthetic mobility. The PLUS-M provides a T score ranging from 17.5 to 76.6, with a higher score representing a higher level of prosthetic mobility; a T score of 50 is equivalent to the mean score reported by more than 1,000 users of lower limb prostheses included in the PLUS-M development study (Sions et al., 2020). The 12-item short version of the PLUS-M was used. The prosthetic activity grade (0–4) gives a gross view of the normal amount of prosthetic use and activity, ranging from 0 (no prosthesis use) to 4 (all-day prosthesis use, lots of walking without the use of an walking aid and/or performing other loading or demanding activities while using prosthesis, i.e., gym training or biking) and was answered at the end of the activity recordings as described in next section. The prosthetic activity grade has previously been reported by patients using a bone-anchored TFA prosthesis (Hagberg & Brodtkorb, 2021; Hagberg et al., 2020).
The 2MWT is a generic gait test (Bohannon et al., 2015) frequently used to assess patients with lower limb amputations (Gaunaurd et al., 2020; Miller et al., 2019; Sions et al., 2020). The test includes walking as far as possible on an even indoor surface for 2 min; if a cane or crutch is normally used for support, participants are encouraged to do so. In the present study, the length of the walking track varied between 18 and 30 m (depending on the location of data collection), with most participants (68%) performing the test on a 30-m track and 5% on the shortest track (18 m). The test was performed once, and the distance traveled within 2 min was recorded to the closest meter.
Activity Measurements and Procedure
\ActivPAL4™ (PAL Technologies Ltd.) activity monitors are based on a triaxial accelerometer and have a size of 53 × 35 × 7 mm, a weight of 15 g, and a sampling frequency of 20 Hz (Edwardson et al., 2017). The device is valid and reliable for recording data on number of steps, for distinguishing among various body postures (i.e., laying, reclining, sitting and standing), and for estimating energy expenditure in metabolic equivalents (METs; Alothman et al., 2020; An et al., 2017; Edwardson et al., 2017; O’Brien et al., 2022).
After completing the questionnaires and performing the 2MWT, each participant was supplied with two activity monitors, one of which was placed on the frontal-mid position on the thigh of the nonamputated limb as recommended (Edwardson et al., 2017), and the other on the frontal side of the prosthesis at approximately the same level (Figure 1). To protect from water, both devices were wrapped in plastic before being secured. Then, a 10 × 10-cm water-resistant skin-friendly surgical dressing (FoamLite®, ConvaTec) was used to adhere the device to the skin on the nonamputated limb. Various tapes were used to attach the device to the prosthesis, depending on the prosthesis surface. Both devices were programmed to record activity for 7 days, starting at midnight (i.e., after the participant had returned home). Participants were instructed not to remove the devices during the 7-day recording period. However, in the case of discomfort, pain, or loosening of the dressing, they were shown how to properly replace the device with a new dressing. Participants were also instructed to make short notes in a diary sheet but otherwise to go about their days as they normally would. After the recording period ended, participants were asked to remove the devices and send them back in a prepaid envelope.
—Illustration of the attachment of activity monitors (arrows) to the nonamputated limb and the prosthetic limb, shown with a socket-suspended prosthesis (left) and a bone-anchored prosthesis (right).
Citation: Journal for the Measurement of Physical Behaviour 6, 3; 10.1123/jmpb.2022-0053
The diary sheet included dates and time intervals in which the participant was asked to make short notes on prosthesis usage and to note if a specific activity had been performed (e.g., gym training, spinning, longer walk or longer time without activity/without wearing the prosthesis). The sheet also included summary questions to be answered at the end of the 7 days. These included questions of whether the recording period represented a normal week of daily life (yes/no); if not, participants were asked to report whether more or less activity than usual had occurred. Participants were also asked whether they needed to change the dressing/tape used to fasten the devices (yes/no) and to rate their current prosthetic activity grade (Supplementary Material [available online]). The diary sheet was returned to the researchers along with the accelerometers.
Data and Statistical Analysis
Activity data were analyzed if at least 4 days of simultaneously recorded data from both devices were available. For each participant, the PDF files provided from the analysis software (PALanalysis v.8.11.6.70, CREA algorithm v1.3) were visually compared between the nonamputated side and the prosthesis side, and grossly compared to the information given in the diary sheet. Furthermore, each participant was provided with their own PDF files from the software.
The detailed activity analysis included the number of sit-to-stand transitions, number of steps, time spent standing, walking, bicycling, sitting, seated while commuting, and laying per day. In addition, METs per hour as well as the time upright (i.e., standing, walking, and cycling); sedentary (i.e., sitting, reclining, and seated while commuting); and laying per day were analyzed for the nonamputated limb.
Depending on the data, the descriptive statistics are provided as count, percent, mean, SD, median, and minimum–maximum values. The mean amount of daily activity measured from the prosthesis compared to that from the nonamputated limb was analyzed for each participant in terms of the number of sit-to-stand transitions, number of steps, and walking time (in hours). To detect differences between the recordings, the paired sample t test was used. All tests were two-sided, and a p value <.05 was considered statistically significant. To further explore differences between recordings of steps taken between the nonamputated limb and the prosthesis, a descriptive subanalysis of selected participants was performed. This subanalysis included six participants for whom the diary notes on prosthetic use and activity performed gave information when the participant had taken a longer duration walk (>20 min) or had taken steps on a confined space (defined as incidental steps). The analysis included a comparison of the number of steps simultaneously recorded from each device for seven “walk sessions” and seven “incidental steps sessions” separately. The results are presented as the percentage of steps recorded from the prosthesis as compared to the recordings from the nonamputated limb.
Data analysis were performed using SPSS (version 28).
Results
The study group consisted of 39 participants (17 women and 22 men) between 21 and 79 years of age, with 1.5–57 years of experience with a prosthesis. The most common cause of amputation was trauma (59%), and the majority of individuals in the study group (87%) had a TFA. Forty-nine percent of the participants reported some other condition or mobility restriction in addition to amputation. Participant demographic information is provided in Table 1, and prosthetic information is provided in Table 2. Descriptive information on self-reported mobility and the 2MWT are provided in Table 3. The mean prosthetic use score was 86, and the mean PLUS-M T-score was 52.1.
Characteristics of Study Participants
| Demographic characteristics | Total (N = 39) | Women (n = 17) | Men (n = 22) |
|---|---|---|---|
| Age at inclusion (years) | 54 (14.5) | 56 (14.4) | 52 (14.6) |
| Years using prosthesis | 24 (17.3) | 32 (18.9) | 18 (13.9) |
| Reason for amputation | |||
| Trauma | 23 (59%) | 6 (35%) | 17 (77%) |
| Tumor | 11 (28%) | 7 (41%) | 4 (18%) |
| Infection | 2 (5%) | 2 (12%) | — |
| Other | 3 (8%) | 2 (12%) | 1 (5%) |
| Amputation side | |||
| Right | 21 (54%) | 7 (41%) | 14 (64%) |
| Left | 18 (46%) | 10 (59%) | 8 (36%) |
| Amputation level | |||
| KD | 5 (13%) | 1 (6%) | 4 (18%) |
| TFA | 34 (87%) | 16 (94%) | 18 (82%) |
| TFA length classification | |||
| Short | 6 (15%) | 3 (18%) | 3 (17%) |
| Medium | 17 (44% | 10 (59%) | 7 (32%) |
| Long | 11 (28%) | 3 (18%) | 8 (36%) |
| Other conditiona | |||
| No | 20 (51%) | 8 (47%) | 11 (50%) |
| Yes | 19 (49%) | 9 (53%) | 11 (50%) |
| Occupation | |||
| Work and/or study | 26 (67%) | 8 (47%) | 18 (82%) |
| Retired (age) | 12 (31%) | 8 (47%) | 4 (18%) |
| Retired (disability) | 1 (2%) | 1 (6%) | — |
| Living arrangement | |||
| Single | 9 (23%) | 6 (35%) | 3 (14%) |
| Cohabitation/married | 30 (77%) | 11 (65%) | 19 (86%) |
| Height (m) | 1.76 (0.94) | 1.68 (0.06) | 1.82 (0.06) |
| Weight (kg) without prosthesis | 77.6 (17.0) | 69.2 (9.6) | 84.1 (18.8) |
| BMIb | 28.0 (5.1) | 27.6 (3.5) | 28.3 (6.2) |
| Waist (cm), n = 38 | 98 (15.3) | 94 (12.1) | 102 (16.8) |
Note. Values shown are the n (%) or mean ± SD. KD = knee disarticulation; TFA = transfemoral amputation; BMI = body mass index.
aOther condition included mobility restrictions in the lower extremity (n = 9; e.g., hip and/or knee arthrosis, replacement of the hip or knee joint and drop foot), impaired function of the upper extremity affecting the use of gait aids (n = 5), Type 2 diabetes mellitus (n = 4), and impaired function of the heart or lung (n = 2). bBMI was calculated by adding 12% to the weight to compensate for the loss of the limb.
Prosthetic Information
| Prosthetic details | Total (N = 39) | Women (n = 17) | Men (n = 22) |
|---|---|---|---|
| Current prosthesis | |||
| Bone-anchored prosthesis | 15 (38.5%) | 7 (41%) | 8 (36%) |
| Socket-suspended prosthesis | 24 (61.5%) | 10 (59%) | 14 (64%) |
| Socket suspension | |||
| Vacuum without liner | 6 (25%) | 4 (40%) | 2 (14%) |
| Liner and vacuum | 9 (37.5%) | 2 (20%) | 7 (50%) |
| Liner with other suspension (e.g., pin, seal-in) | 9 (37.5%) | 4 (40%) | 5 (36%) |
| Prosthetic knee | |||
| Microprocessor controlled | 26 (67%) | 11 (65%) | 15 (68%) |
| Nonmicroprocessor controlled | 13 (33%) | 6 (35%) | 7 (32%) |
| Prosthetic foot | |||
| Energy saving | 34 (87%) | 13 (76%) | 21 (95.5%) |
| Energy saving, adjustable | 3 (8%) | 3 (18%) | — |
| Energy saving, intelligent | 2 (5%) | 1 (6%) | 1 (4.5%) |
Note. Values shown are the n (%).
Description of Participants Self-Reported Prosthetic Mobility and Performance
| PROM and performance test | Total (N = 39) | Women (n = 17) | Men (n = 22) |
|---|---|---|---|
| Use of gait aid outdoors | |||
| No aid | 12 (31%) | 2 (12%) | 10 (45.5%) |
| One cane/crutch | 24 (61%) | 13 (76%) | 11 (50%) |
| Two crutches | 3 (8%) | 2 (12%) | 1 (4.5%) |
| Q-TFA prosthetic use score 0–100 | 86 (15) | 85 (13) | 86 (16) |
| 90 (37–100) | 90 (52–100) | 90 (37–100) | |
| Q-TFA prosthetic mobility score 0–100 | 74 (13) | 72 (13) | 76 (14) |
| 74 (50–98) | 68 (50–98) | 75 (51–97) | |
| PLUS-M T score (21.8–71.4) | 52.1 (6.81) | 51.4 (5.84) | 52.5 (7.57) |
| 50.5 (38.6–67.1) | 50.5 (39.7–67.1) | 50.5 (38.6–67.1) | |
| Prosthetic activity gradea | |||
| 2 | 10 (26%) | 6 (35%) | 4 (18%) |
| 3 | 18 (46%) | 8 (47%) | 10 (46%) |
| 4 | 11 (28%) | 3 (18%) | 8 (36%) |
| 2-min walk test (m)b | 135 (36.5) | 135 (39.1) | 135 (35.4) |
| 127.5 (76–241) n = 38 | 127 (76–241) n = 16 | 130 (84–212) |
Note. Values shown are the n (%), mean ± SD, or median (minimum–maximum). PROM = patient-reported outcomes; PLUS-M = Prosthetic Limb Users Survey of Mobility; Q-TFA = Questionnaire for Persons with a Transfemoral Amputation.
aThe Prosthetic activity grades answered by participants represent: 2 = I use the prosthesis most of the day. Walking is performed with or without the support of a walking aid at home, but I always use a walking aid outdoors. 3 = I use the prosthesis for the entire day. Walking is often performed without the support of a walking aid, but a walking aid might be used for longer distances. I walk a lot, but I rarely perform other demanding or high-load activities with the prosthesis. 4 = I use the prosthesis for the entire day. Walking is performed without a walking aid, I walk a lot and/or routinely perform other demanding or loading activities involving the prosthesis (e.g., cycling or gym exercises). bThe 2-min walk test was performed with the support of one cane or crutch by n = 17 participants.
Seven days of simultaneously recorded activity data were collected in 92% (n = 36) of the study group. Six days were analyzed from one individual, and 4 days were analyzed from two individuals, all including both weekdays and holidays. Fewer than 7 days of data were collected for a variety of reasons as follows: phantom limb pain leading to medication and bed rest for 1 day (n = 1), dead battery in one monitor on Day 5 (n = 1), and trying out a new prosthesis (which interrupted the use of the current prosthesis) for 3 days (n = 1). Sixty-nine percent (n = 27) of the participants reported the measurement period represented a normal week in their daily life. Twenty-eight percent (n = 11) reported the measurement period differed from a normal week; of these, two participants had been more active than normal, and nine had been less active than normal. The question regarding the need to change the dressing and/or tape used to attach the devices was answered by 34 participants (data missing from n = 5). Among these participants, 47% (n = 16) had not changed the dressing, and 53% (n = 18) had changed the dressing one or several times (on the prosthesis: n = 2, on the nonamputated limb: n = 14, on both sides: n = 2).
The overall activity recorded from the nonamputated limb showed that the participants spent 5.0 hr/day in an upright position, 10.2 hr/day sedentary, and 8.7 hr/day laying (Table 4). Table 5 displays the activity from both devices regarding the daily number of sit-to-stand transitions, number of steps, and time spent walking. On average, 6,124 overall steps and 4,479 prosthesis steps were recorded. The results showed that, on average, 85% of all sit-to-stand transitions, 73% of all steps, and 68% of all time spent walking were recorded from the prosthesis (Table 5, Figure 2). There were no statistically significant differences between women and men in any outcome.
Mean Overall Daily Activity
| Activity nonamputated limb | Total (N = 39) | Women (n = 17) | Men (n = 22) |
|---|---|---|---|
| Upright time (hr)a | 5.0 (1.7) | 4.9 (1.7) | 5.0 (1.8) |
| 4.7 (1.4–8.5) | 4.2 (2.3–7.3) | 4.8 (1.4–8.5) | |
| Sedentary time (hr)b | 10.2 (1.8) | 10.2 (1.9) | 10.1 (1.7) |
| 10.0 (6.3–13.7) | 10.5 (6.3–13.3) | 10.0 (7.2–13.7) | |
| Lying time (hr) | 8.7 (1.1) | 8.7 (0.9) | 8.8 (1.2) |
| 8.7 (6.5–12.1) | 8.7 (7.1–10.8) | 8.8 (6.5–12.1) | |
| MET/hour | 33.0 (1.3) | 32.9 (1.4) | 33.1 (1.2) |
| 33.0 (30.8–36.1) | 32.4 (30.9–36.0) | 33.0 (30.8–36.1) |
Note. Values shown are the mean ± SD, median (min–max), MET/hour = estimated metabolic equivalents per hour.
aUpright time includes standing, walking, and cycling. bSedentary time includes sitting and seated while commuting.
Mean Daily Activity Simultaneously Recorded From the Nonamputated Limb and the Prosthesis
| Activity | Nonamputated limb | Prosthetic limb | Diff p value |
|---|---|---|---|
| Number of sit-to-stand transitions | 38 (10.5) | 32 (12.1) | <.001 |
| 37 (13–66) | 31 (9–62) | ||
| Amount of sit-to-stand transitionsa | 85% (19.0) | NA | |
| 88% (25–116) | |||
| Number of steps | 6,124 (3,056) | 4,479 (2,465) | <.001 |
| 5,955 (1,568–13,271) | 3,781 (1,061–11,178) | ||
| Amount of stepsa | 73% (14.3) | NA | |
| 75% (40.5–101) | |||
| Walking time in hours | 1.5 (0.67) | 1.0 (0.48) | <.001 |
| 1.5 (0.4–3.0) | 0.9 (0.3–2.2) | ||
| Amount of walking timea | 68% (15.0) | NA | |
| 69% (32–98) |
Note. Values shown are the mean ± SD and median (min–max). NA = not applicable.
aActivity recorded from the prosthesis as compared to the nonamputated limb, %.
—Box plot of the percentage of recorded sit-to-stand transitions, number of steps, and walking time with the prosthesis compared to the nonamputated limb.
Citation: Journal for the Measurement of Physical Behaviour 6, 3; 10.1123/jmpb.2022-0053
The subanalysis of the number of steps recorded from the prosthesis, as compared to the nonamputated limb while taking a longer walk (walking session) and while moving on a confined space (incidental step session), showed that in mean 95% of the prosthetic steps were recorded during walking sessions and 60% during incidental step sessions (Table 1 in the Supplementary Material [available online]).
Discussion
This study explored physical activity and use of a prosthetic limb in individuals living with loss of the knee joint via simultaneous activity metrics from the nonamputated limb (i.e., overall activity) and the prosthesis (i.e., activity using the prosthesis). The main results showed that sedentary time accounted for the majority of the day (on average, more than 10 hr) and that the recordings from the prosthesis, on average, accounted for 85% of all sit-to-stand transitions, 73% of all steps, and 68% of the time spent walking as compared to the overall recordings.
Sedentary behavior has been defined as “any waking behavior characterized by an energy expenditure of 1.5 MET or lower while in sitting, reclining, or lying posture” (Tremblay et al., 2017). Accumulating evidence suggests a relationship between a more sedentary lifestyle and a higher risk of welfare diseases such as diabetes, metabolic syndrome, and cardiovascular diseases (Tigbe et al., 2017; van der Berg et al., 2016). Recently, the World Health Organization launched a global status report on physical activity and recommended that all countries prioritize the promotion of physical activity to promote health. The same report advocated the use of digital and wearable devices to record physical activity and sedentary behaviors (WHO, Global status report on physical activity, 2022). The data recorded from the nonamputated limb in the present study reflected the participants’ overall activity. On average, our participants performed 38 sit-to-stand transitions, took 6,124 steps, spent 10.2 hr sedentary, and spent 5.0 hr in an upright position (e.g., standing or walking) per day. In a recent Swedish epidemiology study, including >27,000 participants aged 50–65 years, accelerometer-based daily physical activity was assessed. In this large group, sedentary time was reported to be just below 8 hr/day (Ekblom-Bak et al., 2022). The authors argued that more than 9.5 hr sedentary per day could be associated with increased risk for mortality. Our recordings from the nonamputated limb can also be compared to results reported for healthy Scottish employees (n = 111) using the same activity monitor (Tigbe et al., 2017): 62 sit-to-stand transitions, 14,708 steps, 7.1 hr sedentary time, and 7.1 hr standing/stepping per day. Of note is that these healthy participants were younger than our study group. The lower overall activity in our study group is not surprising given the inevitable mobility difficulties following the loss of a limb. Nevertheless, the finding of more than 10 hr a day spent sedentary clearly demonstrates the need to encourage a much higher long-term physical activity in this group of individuals.
Wearable devices have previously been used to investigate physical activity in amputees (Chadwell et al., 2020; Mellema & Gjovaag, 2022). However, many of these studies only report activity performed while wearing the prosthesis. It is well known that people with limb loss, especially among those with higher levels of amputations, do not choose to wear their prosthesis in all waking hours (Hagberg & Branemark, 2001; Sørensen et al., 2022). Chadwell et al. (2020) strongly recommended that future research with prosthetic users combine the monitoring of daily activity in the community with the assessment of prosthetic wear and mobility. The participants in the present study reported using their prosthesis to a high degree (prosthetic use score, median: 90), and the mobility measures were on par with studies on individuals with TFA or KD and similar demographic backgrounds (Carse et al., 2021; Hagberg et al., 2004; Sions et al., 2020).
Prosthetic steps have been reported in several studies; in general, fewer prosthetic steps are taken by those with a TFA compared to those with a TTA (Wong et al., 2021). For example, Sions and coworkers reported an average of 5,491 prosthetic steps/day in 47 individuals with a TTA (Sions et al., 2018); they concluded that this number of steps represented about 77% of the 6,500–7,000 steps/day recommended for older adults or people with limited mobility (Tudor-Locke et al., 2011). In another study that included 17 participants with TFA, with similar demographic characteristics as the present study, prosthetic steps were recorded over a year. In this study, on average, 1,540 (SD: 726) strides/day were recorded; this number can be multiplied by two to provide the number of steps/day, a number comparable to our results, that is, 3,080 (SD: 1,452) steps/day (Halsne et al., 2013). In contrast, Lin et al. (2014) used a pedometer worn at the waist (thus not only recording steps taken with the prosthesis) and reported, on average, 3,985 (SD: 1,246) steps/day in eight individuals with TFA (Lin et al., 2014). In our study, which consisted only of individuals with TFA/KD, the average number of steps was somewhat higher and thus closer to the report by Tudor-Locke et al. (2011). Notably, the high individual variation reflects the large spread of daily activity among the study group.
We found fewer prosthetic steps than overall steps, which might reflect times when the participants moved with a gait aid while not wearing a prosthetic limb, but it might also stem from the manner of moving around, as very short or slow prosthetic steps might not be detected as a step by the activity monitor. The latter explanation was reported by Deans et al. (2020) in a study comparing observed prosthetic steps to recorded steps among 11 TTA and four TFA participants. Their results showed a lower criterion-related validity for incidental prosthetic steps than purposive steps (intraclass correlation coefficient .56 and .93, respectively; Deans et al., 2020). Our subanalysis of selected walking sessions and incidental step sessions point in the same direction as the relationship between steps recorded from the prosthesis compared to the nonamputated limb was 95% while walking but only 60% while moving in a confined space. Recently, DeGrasse et al. (2023) tested a new kind of activity monitor placed inside the prosthetic socket, in 14 individuals with TTA, to explore which type of active motion dominates the total time of prosthetic use in daily life: walking or weight shifts. These authors defined weight shifts as active movements within a defined space and concluded that “weight shifts are meaningful parts of prosthesis use activity” (DeGrasse et al., 2023), as the results showed that time in walking and in weight shifts were similar.
We also observed a few other pitfalls in the activity data collection. For instance, on a few nights, the monitor on the prosthesis recorded standing time while the monitor on the nonamputated limb recorded laying time, and the participant noted not wearing the prosthesis in their diary. Another example is that while performing wheelchair sports without wearing the prosthesis, the overall activity of participants was recorded as sitting (i.e., sedentary time) while the prosthesis was recorded as standing or sitting, but again noted as not worn in the diary. A table illustrating such pitfalls is provided in Table 2 of Supplementary Material (available online). These pitfalls led to our decision not to report upright time and sedentary time recorded from the prosthesis, as we could not fully trust these data. The estimated energy expenditure reported should also be interpreted with caution since an increase in energy cost during prosthetic walking (Ettema et al., 2021; van Schaik et al., 2019) is not accounted for in the MET/hr estimation by activPAL. In a recent review, O’Brien et al. (2022) described generally high levels of validity of activPAL to distinguish among postures; however, they also noted pitfalls. For example, differences in the recorded and actual posture occurred while sitting on a bar stool, which involves a smaller hip angle than sitting on a normal height chair. This sitting position is common among TFA socket-prosthesis users to avoid discomfort from the socket. Thus, there is a need for further research to appropriately identify activities, postures, and movements in people with mobility restrictions, including those with lower limb amputations (Deans et al., 2020; Piazza et al., 2017; van Rooij et al., 2020).
In trauma-related amputations, it is not uncommon for other body parts to sustain life-long impairments, and the presence of secondary conditions in individuals after limb loss is well known (Gailey et al., 2008). Indeed, in our study group, almost half of participants had other such conditions. The individual variation in our results might thus be explained by the presence of additional impairments, such as hip or knee arthrosis, as well as the wide age span of the participants (29–79 years). On average, the prosthesis was used in 83% (25%–116%) of all sit-to-stand transitions (Figure 2). Upon deeper examination, we found that seven participants used the prosthesis in >100% of sit-to-stand transitions; most of these individuals were highly dependent on their prosthesis due to other conditions affecting mobility. In fact, four individuals needed to use the prosthesis for all standing or walking activities, meaning that they did not have the option of using crutches without wearing a prosthetic limb. The high number of sit-to-stand transitions recorded from the prosthesis could also hypothetically stem from handling of the prosthesis during donning.
Strengths and Limitations
In Sweden, the majority of individuals undergoing lower limb amputation are old and fragile, and the 1-year mortality is high (Kamrad et al., 2020). The purpose of the current study was to describe physical activity in a substantially smaller group of individuals long after amputation and who used a prosthesis in daily life. One strength of the current study is that the study group had, on average, more than 20 years of experience using a prosthesis. Another strength is that 44% of the participants were women, as there is a general dominance of males in the literature involving prosthetic users (Mellema & Gjovaag, 2022). Recently, Wong et al. (2021) noted a lack of studies that explored sex differences in stepping. Thus, our reporting of results for women may start to fill this gap in knowledge.
The present study had some limitations. The study was not designed to investigate differences between different kinds of prosthetic solutions or to validate the use of the activity monitor. Therefore, the results are not differentiated based on prosthetic solutions, and the pitfalls noted for the activity measures are briefly described. The subanalysis to explore whether the activity monitor detected fewer prosthetic steps while moving in a confined space than while taking a longer walk was limited to a low sample and a low number of sessions. To further investigate the detection of prosthetic steps in real life, a more detailed method to control prosthetic use, movement patters, and type of activity is warranted. Moreover, the position of the device on the prosthetic limb could differ from that placed on the nonamputated thigh (i.e., either more proximal or more distal), depending on the individual residual limb length and prosthetic solution. Previous research has indicated that such variability in monitor positions might affect recorded data (Stanton et al., 2016). Another limitation is that the current study does not report detailed data regarding the length or intensity of each activity bout. Previous studies have reported short activity bouts and long durations of time spent sitting among amputees (Klute et al., 2006; Miller et al., 2019). Relations between daily activity and demographics (e.g., age, residual limb length, body mass index, other conditions, and type of prosthesis), patient-reported prosthetic mobility and gait performance were not analyzed. An additional study, involving more participants, would be of great interest for the analysis of such relationships. Finally, the 2MWT was performed on different tracks which might have influenced the length participants walked.
In summary, we encourage future research to discover methods to better detect physical activity performed while sitting, for example, during wheelchair sports, and to further validate activity monitoring, specifically among prosthetic users. We also encourage future research to include outcomes from activity monitors in interventions that aim to improve the ability of people with limb loss to live a less sedentary life.
Conclusions
In conclusion, the results of the present study revealed that established prosthetic users with a unilateral TFA or KD spend most of their time sedentary, indicating a low overall physical activity level. Therefore, actions should be taken to encourage higher physical activity to promote long-term health in this group. By simultaneous recordings of activity data, we found that 85% of all sit-to-stand transitions, 73% of all steps, and 68% of all time spent walking during a week were recorded from the prosthesis as compared to the nonamputated limb. These observations indicate a major importance of the prosthesis for daily activity and emphasize the need for efficient access to appropriate prosthetic services throughout life. We also noted pitfalls in the activity data, especially for data recorded from the prosthetic limb, which highlight the need for further research to more accurately obtain real-life activity data in individuals using a prosthetic limb.
References
Aguilar-Farias, N., Martino-Fuentealba, P., Salom-Diaz, N., & Brown, W.J. (2019). How many days are enough for measuring weekly activity behaviours with the ActivPAL in adults? Journal of Science and Medicine in Sport, 22(6), 684–688. https://doi.org/10.1016/j.jsams.2018.12.004
Alothman, S., Hoover, J.C., Alshehri, M.M., Alenazi, A.M., Wick, J., LeMaster, J., ... Kluding, P.M. (2020). Test-retest reliability of ActivPAL in measuring sedentary behavior and physical activity in people with type 2 diabetes. Journal of Physical Activity and Health, 17(11), 1134–1139. https://doi.org/10.1123/jpah.2019-0506
An, H.S., Kim, Y., & Lee, J.M. (2017). Accuracy of inclinometer functions of the activPAL and ActiGraph GT3X+: A focus on physical activity. Gait & Posture, 51, 174–180. https://doi.org/10.1016/j.gaitpost.2016.10.014
Bhatnagar, V., Richard, E., Melcer, T., Walker, J., & Galarneau, M. (2019). Retrospective study of cardiovascular disease risk factors among a cohort of combat veterans with lower limb amputation. Vascular Health and Risk Management, 15, 409–418. https://doi.org/10.2147/VHRM.S212729
Biswas, A., Oh, P.I., Faulkner, G.E., Bajaj, R.R., Silver, M.A., Mitchell, M.S., & Alter, D.A. (2015). Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: A systematic review and meta-analysis. Annals of Internal Medicine, 162(2), 123–132. https://doi.org/10.7326/M14-1651
Bohannon, R.W., Wang, Y.C., & Gershon, R.C. (2015). Two-minute walk test performance by adults 18 to 85 years: Normative values, reliability, and responsiveness. Archives of Physical Medicine and Rehabilitation, 96(3), 472–477. https://doi.org/10.1016/j.apmr.2014.10.006
Buis, A., Dumpleton, T., Murray, K., McHugh, B., & McKay, G. (2014). Measuring the daily stepping activity of people with transtibial amputation using the ActivPAL™ activity Monitor. Journal of Prosthetics and Orthotics, 26(1), 43–47. https://doi.org/10.1097/JPO.0000000000000016
Carse, B., Scott, H., Davie-Smith, F., Brady, L., & Colvin, J. (2021). Minimal clinically important difference in walking velocity, gait profile score and two minute walk test for individuals with lower limb amputation. Gait & Posture, 88, 221–224. https://doi.org/10.1016/j.gaitpost.2021.06.001
Chadwell, A., Diment, L., Mico-Amigo, M., Morgado Ramirez, D.Z., Dickinson, A., Granat, M., ... Worsley, P. (2020). Technology for monitoring everyday prosthesis use: A systematic review. Journal of NeuroEngineering and Rehabilitation, 17(1), Article 93. https://doi.org/10.1186/s12984-020-00711-4
Deans, S., Kirk, A., McGarry, A., & Rowe, D. (2020). Reliability and criterion-related validity of the activPAL™ accelerometer when measuring physical activity and sedentary behavior in adults with lower limb absence. Journal for the Measurement of Physical Behaviour, 3(3), 244–252. https://doi.org/10.1123/jmpb.2019-0045
DeGrasse, N.S., Mertens, J.C., Brzostowski, J.T., Allyn, K.J., Vamos, A.C., Krout, A.J., ... Sanders, J.E. (2023). Beyond step counts: Including wear time in prosthesis use assessment for lower-limb amputation. Journal of Rehabilitation and Assistive Technologies Engineering, 10, Article 20556683231163337. https://doi.org/10.1177/20556683231163337
Edwardson, C.L., Winkler, E.A.H., Bodicoat, D.H., Yates, T., Davies, M.J., Dunstan, D.W., & Healy, G.N. (2017). Considerations when using the activPAL monitor in field-based research with adult populations. Journal of Sport and Health Science, 6(2), 162–178. https://doi.org/10.1016/j.jshs.2016.02.002
Ekblom-Bak, E., Borjesson, M., Bergman, F., Bergstrom, G., Dahlin-Almevall, A., Drake, I., ... Ekblom, O. (2022). Accelerometer derived physical activity patterns in 27.890 middle-aged adults: The SCAPIS cohort study. Scandinavian Journal of Medicine & Science in Sports, 32(5), 866–880. https://doi.org/10.1111/sms.14131
Ettema, S., Kal, E., & Houdijk, H. (2021). General estimates of the energy cost of walking in people with different levels and causes of lower-limb amputation: A systematic review and meta-analysis. Prosthetics and Orthotics International, 45(5), 417–427. https://doi.org/10.1097/PXR.0000000000000035
Gailey, R., Allen, K., Castles, J., Kucharik, J., & Roeder, M. (2008). Review of secondary physical conditions associated with lower-limb amputation and long-term prosthesis use. Journal of Rehabilitation Research and Development, 45(1), 15–30. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=18566923
Gaunaurd, I., Kristal, A., Horn, A., Krueger, C., Muro, O., Rosenberg, A., ... Gailey, R. (2020). The utility of the 2-minute walk test as a measure of mobility in people with lower limb amputation. Archives of Physical Medicine and Rehabilitation, 101(7), 1183–1189. https://doi.org/10.1016/j.apmr.2020.03.007
Hafner, B.J., Gaunaurd, I.A., Morgan, S.J., Amtmann, D., Salem, R., & Gailey, R.S. (2017). Construct validity of the prosthetic limb users survey of mobility (Plus-M) in adults with lower limb amputation. Archives of Physical Medicine and Rehabilitation, 98(2), 277–285. https://doi.org/10.1016/j.apmr.2016.07.026
Hagberg, K., & Branemark, R. (2001). Consequences of non-vascular trans-femoral amputation: A survey of quality of life, prosthetic use and problems. Prosthetics and Orthotics International, 25(3), 186–194. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=11860092
Hagberg, K., Branemark, R., & Hagg, O. (2004). Questionnaire for Persons with a Transfemoral Amputation (Q-TFA): Initial validity and reliability of a new outcome measure. Journal of Rehabilitation Research and Development, 41(5), 695–706. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15558399
Hagberg, K., & Brodtkorb, T.H. (2021). Patient-reported benefits of bone-anchored transfemoral prostheses as assessed by MedTech20: A general outcome measure for medical products. Prosthetics and Orthotics International, 45(4), 355–361. https://doi.org/10.1097/PXR.0000000000000008
Hagberg, K., Ghassemi Jahani, S.A., Kulbacka-Ortiz, K., Thomsen, P., Malchau, H., & Reinholdt, C. (2020). A 15-year follow-up of transfemoral amputees with bone-anchored transcutaneous prostheses. Bone & Joint Journal, 102-B(1), 55–63. https://doi.org/10.1302/0301-620X.102B1.BJJ-2019-0611.R1
Halsne, E.G., Waddingham, M.G., & Hafner, B.J. (2013). Long-term activity in and among persons with transfemoral amputation. Journal of Rehabilitation Research and Development, 50(4), 515–530. http://www.ncbi.nlm.nih.gov/pubmed/23934872
Kamrad, I., Soderberg, B., Orneholm, H., & Hagberg, K. (2020). SwedeAmp—The Swedish Amputation and Prosthetics Registry: 8-year data on 5762 patients with lower limb amputation show sex differences in amputation level and in patient-reported outcome. Acta Orthopaedica, 91(4), 464–470. https://doi.org/10.1080/17453674.2020.1756101
Klute, G.K., Berge, J.S., Orendurff, M.S., Williams, R.M., & Czerniecki, J.M. (2006). Prosthetic intervention effects on activity of lower-extremity amputees. Archives of Physical Medicine and Rehabilitation, 87(5), 717–722. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=16635636
Langford, J., Dillon, M.P., Granger, C.L., & Barr, C. (2019). Physical activity participation amongst individuals with lower limb amputation. Disability and Rehabilitation, 41(9), 1063–1070. https://doi.org/10.1080/09638288.2017.1422031
Leskinen, T., Stenholm, S., Aalto, V., Head, J., Kivimaki, M., & Vahtera, J. (2018). Physical activity level as a predictor of healthy and chronic disease-free life expectancy between ages 50 and 75. Age and Ageing, 47(3), 423–429. https://doi.org/10.1093/ageing/afy016
Lin, S.J., Winston, K.D., Mitchell, J., Girlinghouse, J., & Crochet, K. (2014). Physical activity, functional capacity, and step variability during walking in people with lower-limb amputation. Gait & Posture, 40(1), 140–144. https://doi.org/10.1016/j.gaitpost.2014.03.012
Mellema, M., & Gjovaag, T. (2022). Reported outcome measures in studies of real-world ambulation in people with a lower limb amputation: A scoping review. Sensors, 22(6), Article 2243. https://doi.org/10.3390/s22062243
Miller, M.J., Cook, P.F., Kline, P.W., Anderson, C.B., Stevens-Lapsley, J.E., & Christiansen, C.L. (2019). Physical function and pre-amputation characteristics explain daily step count after dysvascular amputation. PM&R, 11(10), 1050–1058. https://doi.org/10.1002/pmrj.12121
O’Brien, M.W., Wu, Y., Petterson, J.L., Bray, N.W., & Kimmerly, D.S. (2022). Validity of the ActivPAL monitor to distinguish postures: A systematic review. Gait & Posture, 94, 107–113. https://doi.org/10.1016/j.gaitpost.2022.03.002
Pepin, M.-E., Devour, A., Coolsaet, R., & Galen, S. (2019). Correlation between functional ability and physical activity in individuals with transtibial amputations: A cross-sectional study. Cardiopulmonary Physical Therapy Journal, 30(2), 70–78. https://doi.org/10.1097/cpt.0000000000000091
Persson, B.M., & Liedberg, E. (1983). A clinical standard of stump measurement and classification in lower limb amputees. Prosthetics and Orthotics International, 7(1), 17–24. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=6856447
Piazza, L., Ferreira, E.G., Minsky, R.C., Pires, G.K.W., & Silva, R. (2017). Assessment of physical activity in amputees: A systematic review of the literature. Science & Sports, 32(4), 191–202. https://doi.org/10.1016/j.scispo.2017.07.011
Salih, S.A., Peel, N.M., & Burgess, K. (2016). Monitoring activity of inpatient lower limb prosthetic users in rehabilitation using accelerometry: Validation study. Journal of Rehabilitation and Assistive Technologies Engineering, 3, Article 2055668316642387. https://doi.org/10.1177/2055668316642387
Sions, J.M., Arch, E.S., & Horne, J.R. (2018). Self-reported functional mobility, balance confidence, and prosthetic use are associated with daily step counts among individuals with a unilateral transtibial amputation. Journal of Physical Activity and Health, 15(6), 423–429. https://doi.org/10.1123/jpah.2017-0196
Sions, J.M., Beisheim, E.H., & Seth, M. (2020). Selecting, administering, and interpreting outcome measures among adults with lower-limb loss: An update for clinicians. Current Physical Medicine and Rehabilitation Reports, 8(3), 92–109. https://doi.org/10.1007/s40141-020-00274-4
Sørensen, I.L., Berg, G.V., & Tramontano, M. (2022). Facilitators and barriers that transfemoral amputees experience in their everyday life: A Norwegian qualitative study. Rehabilitation Research and Practice, 2022, 1–8. https://doi.org/10.1155/2022/2256621
Stanton, R., Guertler, D., Duncan, M.J., & Vandelanotte, C. (2016). Agreement between activPAL3c accelerometers placed at different thigh positions. Gait & Posture, 48, 230–236. https://doi.org/10.1016/j.gaitpost.2016.06.013
Tigbe, W.W., Granat, M.H., Sattar, N., & Lean, M.E.J. (2017). Time spent in sedentary posture is associated with waist circumference and cardiovascular risk. International Journal of Obesity, 41(5), 689–696. https://doi.org/10.1038/ijo.2017.30
Tremblay, M.S., Aubert, S., Barnes, J.D., Saunders, T.J., Carson, V., Latimer-Cheung, A.E., ... Participants, S.T.C.P. (2017). Sedentary Behavior Research Network (SBRN)—Terminology consensus project process and outcome. The International Journal of Behavioral Nutrition and Physical Activity, 14(1), Article 75. https://doi.org/10.1186/s12966-017-0525-8
Tudor-Locke, C., Craig, C.L., Aoyagi, Y., Bell, R.C., Croteau, K.A., De Bourdeaudhuij, I., ... Blair, S.N. (2011). How many steps/day are enough? For older adults and special populations. The International Journal of Behavioral Nutrition and Physical Activity, 8, Article 80. https://doi.org/10.1186/1479-5868-8-80
van der Berg, J.D., Stehouwer, C.D., Bosma, H., van der Velde, J.H., Willems, P.J., Savelberg, H.H., ... Koster, A. (2016). Associations of total amount and patterns of sedentary behaviour with type 2 diabetes and the metabolic syndrome: The Maastricht Study. Diabetologia, 59(4), 709–718. https://doi.org/10.1007/s00125-015-3861-8
van Rooij, W.M.J., van den Berg-Emons, H.J.G., Horemans, H.L.D., Fanchamps, M.H.J., de Laat, F.A., & Bussmann, J.B.J. (2020). Validation of a clinically feasible activity monitor which measures body postures and movements in adults with lower-limb amputation who wear a prosthesis. Journal for the Measurement of Physical Behaviour, 3(2), 135–146. https://doi.org/10.1123/jmpb.2019-0038
van Schaik, L., Geertzen, J.H.B., Dijkstra, P.U., & Dekker, R. (2019). Metabolic costs of activities of daily living in persons with a lower limb amputation: A systematic review and meta-analysis. PLoS One, 14(3), Article e0213256. https://doi.org/10.1371/journal.pone.0213256
Wetterhahn, K.A., Hanson, C., & Levy, C.E. (2002). Effect of participation in physical activity on body image of amputees. American Journal of Physical Medicine & Rehabilitation, 81(3), 194–201. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=11989516
WHO, Global status report on physical activity. (2022). www.who.int/teams/health-promotion/physical-activity/global-status-report-on-physical-activity-2022
Wong, C.K., Rissland, M.S., Madagan, D.M., & Jones, K.N. (2021). A scoping review of physical activity in people with lower-limb loss: 10,000 steps per day? Physical Therapy, 101(8), Article pzab115. https://doi.org/10.1093/ptj/pzab115
Wurdeman, S.R., Stevens, P.M., & Campbell, J.H. (2018). Mobility analysis of AmpuTees (MAAT I): Quality of life and satisfaction are strongly related to mobility for patients with a lower limb prosthesis. Prosthetics and Orthotics International, 42(5), 498–503. https://doi.org/10.1177/0309364617736089
