GGEM: Gender, Geography, and EMployment differences based on mobility levels among lower limb prosthesis users living in the United States

Background:

For individuals with a disability, an increase in functional mobility may improve their quality of life and well-being. Greater understanding is needed on how factors such as gender, geography, and employment may play a role in mobility levels among individuals with lower limb amputation.

Objectives:

To assess the relationship between gender, geography, and employment status on mobility among lower limb prosthesis users.

Methods:

A cross-sectional analysis of 7,524 patient mobility outcomes completed across the United States was performed. The regression model included the independent variables, such as age, gender, region, employment status, and amputation level. Mobility was entered as the dependent variable.

Results:

Individuals who were employed had 3.6 times the odds of reaching increased mobility (Prosthetic Limb Users’ Survey of Mobility ≥ 50) than those unemployed (odds ratio 3.56, 95% confidence interval 3.10–4.09). Gender and geography were significantly associated with mobility as well.

Conclusions:

Being employed is associated with greater odds of reaching increased mobility. Addressing factors such as returning to employment may aid in improving mobility levels among prosthesis users.

Introduction

Lower limb amputation (LLA) can potentially have a major effect on a patient’s overall health, well-being, and functional mobility. Prediction of functional mobility outcomes after LLA is of concern to healthcare professionals, given that a dramatic reduction in mobility can decrease quality of life,¹ increase healthcare costs or utilization,^2,3 or result in a decrease in occupational classification, including loss of employment.⁴ The influence of gender, geography, or employment on other outcomes (eg, health-related quality and physical activity) has been extensively investigated^5-7; however, few studies have explored the relationship between functional mobility and LLA. Mobility restoration is one of the ultimate aims of rehabilitation among patients with LLA, and the relationship between gender, geography, and employment and functional mobility may be used by clinicians to make informed care decisions geared toward prosthetic selection and design.

Employment affects health while cutting across demographic attributes, including gender and age; individuals who are employed are healthier.⁸ For individuals with a disability, employment is beneficial for improving self-esteem and reducing social isolation.⁹ The importance of employment should not be undervalued among patients with LLA, considering that not only is employment associated with improved psychosocial outcomes and greater levels of financial independence, but being employed may provide access to healthcare services through employer-sponsored insurance plans. However, the reported return to employment rate for patients with LLA is low, ranging from 43% to 70%.^10-12 Most patients with LLA have difficulties sustaining or returning to employment because of reasons outside of their control. Restricted mobility and the lack of prosthetic rehabilitation to restore functional mobility were noted as 2 possible factors influencing the return to employment.¹⁰ Prosthetic rehabilitation is defined as the restoration of function through the utilization of a prosthesis.¹³

Emphasis has been placed on the return to employment rate for young patients with traumatic amputation who mostly fall below 64 years,^14-16 but very few studies exist on the employment rate for people with diabetic/dysvascular amputation regardless of age. Doukas et al¹⁷ reported that in a sample of veterans with LLA attributed to trauma, the employment rate after amputation was 43.4% at the time of the survey. Pezzin et al¹⁵ found that in a sample of mostly young men with traumatic amputation, 97% of individuals were employed before amputation, and after amputation, only 58% were employed at the time of the interview. One notable limitation for some of the above studies was that the authors failed to mention the number of patients with or without a prosthesis. Arguably, a prosthesis would make it easier to travel to work and expand the type of work possible; without reporting, it is difficult to determine the role prosthetic rehabilitation has on employment retention.

Furthermore, other studies conducted within the United Kingdom and The Netherlands have examined the relationship between employment status and physical functioning or mobility.^18,19 Sinha et al¹⁹ and Fisher et al¹⁸ found that patients with LLA who were employed had increased mobility or physical function scores compared with those who were unemployed. Sinha et al¹⁹ used the Short Form-36 instrument to assess physical functioning, while Fisher et al¹⁸ used the Stanmore Harold Wood mobility grade. Despite its high utilization in the literature, the Stanmore Harold Wood mobility grade is not a well-validated mobility instrument.²⁰ Notwithstanding these dated findings, more studies examining the relationship between employment status and functional mobility are needed to influence the clinical practice guidelines in the United States, especially among the diabetic or dysvascular population.

Gender disparity is another factor that may influence functional mobility outcomes. The relationship between gender and mobility among prosthesis users is varied, mixed, and weakly established in the literature. A recent systematic review found 4 studies that reported no significant relationship between gender and walking ability among patients with LLA.^21-24 In contrast to the findings in the review, a previous study found that gender affects mobility (ie, Timed-Up-and-Go (TUG) scores) and walking speed.²⁵

In addition to gender, geography may have a significant influence on the mobility of prosthesis users. Studies have reported that geography or region of care is associated with differences in mobility, functional limitations, and functional status among a nonamputation population.^6,26-28 Adults living in the South reported poorer functional outcomes than adults living in other regions.²⁹ Although regional disparities exist in mobility among the nonamputation population, and previous research reported a disparity in LLA rates across regions,²⁹ little is known about whether differences across regions exist specific to functional mobility among LLA patients.

Therefore, this study aimed to assess the relationship between the independent variables, such as gender, geography, and employment status, with the dependent variable functional mobility among lower limb prosthesis users. Multiple hypotheses were tested: (1) it was hypothesized that individuals who are employed will have higher reported mobility than patients who are unemployed; (2) it was hypothesized that men will have higher functional mobility than women³⁰; and (3) because regional differences exist in mobility for the nonamputation population, it was hypothesized that this difference will also exist among lower limb prosthesis users.

Methods

Participants

The study sample consisted of unilateral prosthesis users who completed clinical outcomes within different prosthetic clinics across the United States. Prosthesis users with a major LLA 18 years and older were grouped by age into 2 groups: 65 years and older and younger than 65 years. Excluded from the analysis were bilateral prosthesis users and individuals for whom mobility had not yet been assessed. The analysis was a convenience sample of the 4 regions (ie, Northeast, South, West, and Midwest) as defined by the 2010 United States Census classification. The regions had 5–7 states from which participating clinics collecting outcomes were located: Northeast (MA, CT, RI, PA, NY, and NJ), South (TX, OK, AR, LA, and MS), West (CA, NM, UT, CO, WY, and UT), and Midwest (WI, IL, MO, ND, SD, NE, and MN).

Design

A cross-sectional analysis of patient mobility outcomes completed between January 1, 2017, and December 31, 2019, was performed. This study was approved by the WCG Investigational Board (protocol #20170059). As a retrospective analysis of deidentified data, the analysis was deemed exempt from requiring individual informed consent. This study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines for observational studies.³¹

Independent variables

The demographic variables included age, gender, region, employment status, and amputation level. Self-reported employment status (ie, employed/student or unemployed), amputation level (ie, above-the-knee or below-the-knee), and age (ie, <65 or ≥65) were evaluated as dichotomous variables. In the 2000 census report from the United States Census Bureau, working-age was reported to be between 16 and 64 years.³² Therefore, working-age adults were grouped as 64 years and younger, and non–working-age as 65 years and older based on the Census Bureau report and also based on healthcare coverage eligibility with US Centers for Medicare and Medicaid Services. Employment status was evaluated using the survey question, “Are you currently employed (or a student)?” Response options for current employment status were reported as “yes” or “no”. Part-time employees and students who selected “yes” were grouped as being employed. The collapsed categories for the amputation level were above-the-knee (transfemoral and knee disarticulation) and below-the-knee (transtibial and ankle disarticulation) amputations.

Dependent variable

Functional mobility was measured through the Prosthetic Limb Users’ Survey of Mobility v1.2 (PLUS-M T-Score).³³ PLUS-M T-Scores were analyzed as both a continuous and a dichotomous outcome variable. Individuals with a PLUS-M T-Score of 50 and above are considered to have a functional mobility level at or above the population mean of all unilateral leg prosthesis users based on the previous development sample.³⁴ The PLUS-M T-Scores were dichotomized into 2 cutoff points based on population mean: higher PLUS-M (≥50) and lower PLUS-M (<50).³⁴ This dichotomous analysis approach implementing a threshold point is consistent with recommendations for easier comprehension of results by the general population.³⁵

Statistical analysis

Descriptive statistics were applied to summarize clinical and demographic characteristics grouped by age group (age <65 years and age ≥65 years). A separate univariate analysis (ie, analysis of variance) was used to examine the differences across gender, employment status, region, and amputation level on mobility as a continuous variable. If significant differences arose, a post hoc test with Bonferroni correction was conducted to examine differences between groups. Next, a multivariate logistic regression was applied to examine the likelihood of an individual of age younger than 65 years to obtain high mobility levels (PLUS-M ≥ 50) while adjusting for gender, employment, region, age (continuous variable), and amputation level. Using the same variables for adjustment, a subsequent multivariate logistic regression was conducted for individuals of age 65 years and older. The level of significance was set at 0.05. All statistical analyses were performed using R software (version 4.0.2).

Results

A sample of 10,682 prosthesis users had clinical outcomes completed in one of the listed regions or states. Of the initial sample, 3,158 were removed because of bilateral amputation or mobility data not collected. The final sample size used for the data analysis was 7,524 patients. Patients’ demographic characteristics were stratified by mobility (ie, higher and lower) and age groups. The average age for prosthesis users who attained increased PLUS-M scores was 48.6 ± 12.1 and 52.1 ± 9.8 for the age younger than 65 years and the age 65 years and older cohort, respectively (Table 1). Approximately 26.3% (1982/7524) of the entire sample were women. Only 23.6% of respondents reported being employed at the time of the survey. Men (52%) had a higher frequency of achieving increased mobility than women (37%). Approximately 70% of employed individuals with age younger than 65 years achieved increased PLUS-M scores compared with 38% of those individuals who were unemployed. Amputation secondary to diabetes/vascular diseases was the most frequent cause of amputation.

Table 1.

Patients' demographics grouped by age and mobility, N = 752.

	Age < 65 y		Age ≥ 65 y
	Higher PLUS-M (N = 2289)	Lower PLUS-M (N = 2473)	Higher PLUS-M (N = 1009)	Lower PLUS-M (N = 1753)
Employment, n (%)
No	1212 (38)	2013 (62)	858 (34)	1664 (66)
Yes	1077 (70)	460 (30)	151 (63)	89 (37)
Gender, n (%)
Female	459 (37)	785 (63)	168 (23)	570 (77)
Male	1830 (52)	1688 (48)	841 (42)	1183 (58)
Region, n (%)
South	504 (45)	624 (55)	191 (31)	430 (69)
Northeast	597 (51)	578 (49)	289 (37)	486 (63)
West	642 (48)	710 (52)	273 (39)	434 (61)
Midwest	546 (49)	561 (51)	256 (39)	403 (61)
Age (y) (mean ± SD)	48.6 ± 12.1	52.1 ± 9.8	71.3 ± 5.4	73.5 ± 6.9
Height (m) (mean ± SD)	1.76±0.1	1.74 ± 0.1	1.76 ± 0.1	1.72 ± 0.1
Weight (kg) (mean ± SD)	91.4 ± 22.5	93.6 ± 25.8	86.8 ± 20.1	84.6 ± 21.8
Amputation etiology, n (%)
Diabetes/vascular disease	680 (36.5)	1181 (63.5)	402 (28.6)	1005 (71.4)
Trauma	635 (54.1)	539 (45.9)	245 (57.9)	178 (42.1)
Cancer	85 (61.6)	53 (38.4)	18 (32.1)	38 (67.9)
Congenital	115 (73.2)	42 (26.8)	19 (73.1)	7 (26.9)
Unreported	774 (54.1)	658 (45.9)	325 (38.2)	525 (61.8)
Amputation level, n (%)
BKA	1890 (51)	1809 (49)	842 (40)	1251 (60)
AKA	399 (38)	664 (62)	167 (25)	502 (75)
BKA by gender, n (%)
Female	376 (39)	585 (61)	139 (25)	411 (75)
Male	1514 (55)	1224 (45)	703 (46)	840 (54)
AKA by gender, n (%)
Female	83 (29)	200 (71)	29 (15)	159 (85)
Male	316 (40)	464 (60)	138 (29)	343 (71)

Abbreviations: AKA, above-the-knee amputation; BKA, below-the-knee amputation.

Bivariate models

Bivariate analysis for mobility as a continuous variable revealed that prosthesis users who were employed had significantly greater mobility T-scores (age < 65 years: 55.2 ± 10.0 and age ≥ 65 years: 52.2 ± 9.9) for both age groups than unemployed prosthesis users (age < 65 years: 45.9 ± 10 and age ≥65 years: 44.7 ± 11.5, P < 0.001) (Table 2). Men with age 65 years and older had significantly greater mean mobility scores (50.3 ± 11.1) than women (46.1 ± 11.2, P < 0.001). Region and amputation levels were significantly associated with mobility (Table 2).

Table 2.

Bivariate analysis for mobility as a continuous variable.

	Age < 65 y		Age ≥65 y
	Mobility	P	Mobility	P
Employmenta (mean ± SD)
No	46.3 ± 10.6	<0.001	44.7 ± 11.5	<0.001
Yes	55.2 ± 10.0		52.2 ± 9.9
Gendera (mean ± SD)
Female	46.1 ± 10.9		41.0 ± 11.2
Male	50.3 ± 11.1	<0.001	46.9 ± 11.4	<0.001
Regionb (mean ± SD)
South	48.2 ± 11.4	<0.001	43.8 ± 11.2	0.003
Northeast	49.8 ± 11.2		45.7 ± 11.9
West	48.9 ± 11.3		46.0 ± 11.4
Midwest	49.7 ± 10.9		45.6 ± 11.7
Amputation levela (mean ± SD)
BKA	49.9 ± 11.3	<0.001	46.6 ± 11.5	<0.001
AKA	46.5 ± 10.6		41.4 ± 11.1

Abbreviations: AKA, above-the-knee amputation; BKA, below-the-knee amputation; PLUS-M, Prosthetic Limb Users’ Survey of Mobility.

^aIndependent t test variables.

^bVariable used in ANOVA. Mobility calculated from the PLUS-M T-Score.

Multivariate model with mobility as a dichotomous variable

Given that all independent variables used in the bivariate models were significant and none exhibited multicollinearity, all were included in the final logistic regression model. After controlling for age (ie, years continuous), gender, region, and amputation level, among the age younger than 65 years cohort, individuals who were employed had 3.6 times the odds of reaching a higher PLUS-M functional mobility (≥50) than those who were unemployed (odds ratio [OR] 3.56, 95% confidence interval [CI] 3.10–4.09) (Table 3). Men were nearly 2 times more likely to have greater mobility scores than women (OR 1.86, 95% CI 1.61–2.14). Prosthesis users who lived in the Northeast had a 30% increased odds of reaching increased mobility scores compared with those who lived in the South (OR 1.30, 95% CI 1.09–1.55). Age (OR 0.98, 95% CI 0.97–0.99) and above-the-knee amputation (OR 0.49, 95% CI 0.42–0.57) were negatively associated with increased PLUS-M scores.

Table 3.

Adjusted odds of attaining higher mobility.

	Age < 65 y			Age ≥ 65 y
	OR	CI	P	OR	CI	P
Employment
No	*			*
Yes	3.56	3.10–4.09	<0.001	2.80	2.11–3.73	<0.001
Gender
Female	*			*
Male	1.86	1.61–2.14	<0.001	2.28	1.87–2.79	<0.001
Region
South	*			*
Northeast	1.30	1.09–1.55	0.003	1.43	1.13–1.82	0.003
West	1.07	0.90–1.27	0.431	1.36	1.08–1.73	0.010
Midwest	1.11	0.93–1.33	0.238	1.34	1.05–1.71	0.181
Age (y)	0.98	0.97–0.99	<0.001	0.95	0.94–0.96	<0.001
Amputation level
BKA	*			*
AKA	0.49	0.42–0.57	<0.001	0.5	0.40–0.61	<0.001

Abbreviations: AKA, above-the-knee amputation; BKA, below-the-knee amputation; CI, confidence interval; OR, odds ratio; PLUS-M, Prosthetic Limb Users’ Survey of Mobility.

^*Comparison group.

The adjusted odds of reaching mobility greater than the population mean based on age groups. Men, being employed, and living in the Northeast have the highest odds of reaching mobility greater than the population mean (ie, PLUS-M T-Score of 50.0).

Discussion

Prosthetic rehabilitation aims to restore functional independence and provide patients with a mobility level similar to or possibly greater than their preamputation levels. After amputation, a prosthesis can maximize mobility and may allow individuals with employment requiring physical activities such as technical, skilled, and semiskilled jobs to retain their occupation and, most importantly, their financial independence. All factors that may lend themselves to reduce mortality.³⁶

This study aimed to establish the relationship between gender, geography, and employment status on mobility among prosthesis users. The results revealed a significant relationship between employment and mobility. More specifically, an employed prosthesis user (ie, age younger than 65 years) had a 3.5 increased odds of being in the higher mobility category compared with those who were unemployed. Similarly, this relationship was also noted in prosthesis users, age 65 years and older, where employed individuals had a 2.8-fold increased odds of being in the higher mobility category when compared with their unemployed peers within the same age bracket. This may suggest that being employed has greater association with higher mobility. The finding of employment related to higher mobility is consistent with previous work, noting that employment requires the greatest use of a prosthesis.³⁷ Regarding prosthesis utilization or adjustment to amputation, the same study also found that being unemployed was related to being more functionally and socially restricted.³⁷ It could be that unemployed individuals may not be equipped with the appropriate prosthesis device to retain and maintain their occupation. Certain types of prosthesis components are associated with a reduction in injurious falls³⁸ and pain.⁴ These factors, for example, may play a role in encouraging return to employment.

Caution should be taken with interpretation of employment as success of mobility among prosthesis users, especially among individuals 64 years and older given the high likelihood of choosing to retire at 65 years. This in itself creates a limitation to this work—the definition of non–working-age is based on the fact that the general population becomes eligible for Medicare in the United States at age 65 years. Eligibility for such health benefits, in addition to Social Security, is a primary factor for retirement age centered at 65 years. Many individuals with LLA qualify for Medicare and Social Security benefits after amputation. Subsequently, by the very definition of non–working-age being based on eligibility for benefits, arguably, the amputee population may entirely be considered non–working-age. As a result, in particular for retired patients with declining mobility levels or sedentary lifestyle, the ability to engage in other physical activities that offer the benefits of exercise and mental acumen similar to that of employment should be given a greater value. Previous studies have emphasized that functional mobility is correlated with improvements in quality of life, satisfaction, and overall well-being.¹ Moreover, the average mobility score of our 65 years and older, unemployed cohort was 44.7 ± 11.5, which is practically within one-half of a standard deviation of the population mean, suggesting that although unemployed, their functional mobility is close to average.

Although this study sample had more men than women, the findings revealed a functional deficit in mobility after amputation among women. Recent analyses have used different instruments to assess mobility differences across genders among patients with LLA, making the comparison across findings difficult.^18,19 However, the current finding of reduced functional mobility among women is consistent with these studies. Since fewer women tended to achieve higher mobility scores than men, targeted rehabilitation may reduce the gender gap in mobility. In addition to the functional deficit in mobility experienced by women, above-the-knee amputations were less frequent in women than in men. These findings are in contrast to studies conducted in Sweden³⁹ and the United States.⁴⁰ This difference may be explained by the relatively younger patient population at the time of amputation in the current sample compared with the study in Sweden, which reported that nearly 50% of patients were 80 years and older at first amputation. They concluded that having an amputation at a later age suggests that women may have had healthier lives than men before amputation but experience increased complications in adjusting to a proximal amputation. Concerning to the current analysis, however, is that typically below-the-knee amputation is associated with increased functional mobility, and yet despite greater frequency of below-the-knee amputation, there was reduced functional mobility.

Regional differences in mobility and functional limitations are not unique to the nonamputation population. Such differences were noted among the individuals who underwent amputation. The current findings found that prosthesis users in the North had 30% increased odds of being in the higher mobility category (ie, PLUS-M T-Score above the population average of 50.0) than individuals living in the South after accounting for potential confounding factors. Previous findings have reported poorer health, such as higher levels of obesity and diabetes, increased functional limitation, and lower life expectancy among individuals in the South compared with other regions.⁶ This unhealthier lifestyle may be affecting mobility among prosthesis users.

This study has notable advancements to the literature. First, it provides updated findings on the relationship between current employment and mobility using a large sample size. In addition, the sample consisted of patients representing the 4 regions of the United States, and unlike previous studies, these findings are not restricted to veterans and younger patients. A clear understanding of the impact of employment, gender, and geography may help in clinical decision-making regarding prosthetic rehabilitation. In light of these strengths, there are limitations to consider. This study was cross-sectional; therefore, causal relationships must not be assumed. The time of employment was restricted to 1 reference point, which was the time of the survey assessment. We only evaluated patients’ current employment status; therefore, past employment history for legacy prosthesis users before completing the survey was not reported because of the absence of patients’ data. Future work should investigate whether patients who are unemployed had returned to work at least once after amputation to provide a more granular understanding of the unemployment rate among lower limb prostheses and possible extrapolation of economic costs. Programs and facilities offering vocational rehabilitation services were not included as a variable in our study design, but this is possibly a confounding variable, which would affect the relationship between employment and mobility. It is worth noting that vocational rehabilitation in the United States is not often employer initiated; it is more commonly a service individuals are referred to by the rehabilitation team.

This study endpoints (PLUS-M) are also limited to prosthesis users, thus not allowing examination of those who are limited to wheelchair use. However, it is reasonable to speculate that wheelchair users would have lower employment, given the increased difficulty of mobility exacerbated within environments lacking physical accessibility accommodations. Caution must be applied should these findings be generalized to patients who have never used a prosthesis or patients with bilateral amputation. Other factors interacting with geography that could potentially influence mobility differences should be investigated. For example, Resnik and Borgia⁴¹ showed that prosthetic prescriptions were less likely to be filled in the South than in other regions. If there is less likelihood of individuals in a region filling a prescription for a prosthesis, it would seem reasonable that the area would also suffer from lower patient engagement from those who do fill the prosthesis prescription. In addition, future work should aim at looking at the incidence of return to employment and return to employment in relation to regional differences in the United States. Finally, although early work demonstrated that certain comorbid conditions and prosthesis type influence prosthetic mobility,^38,42 these variables were not available for inclusion in the current analyses. However, given the magnitude of the OR, it is believed that similar results would be noted.

Conclusion

Gender, geography, and employment are all factors that are highly associated with prosthetic users’ mobility. In comparison with men, women are less likely to ambulate, with functional mobility exceeding the population mean of prosthesis users. In addition, individuals living in the South in the United States struggle to achieve functional mobility above the population mean for prosthesis users. Finally, being unemployed is associated with low mobility levels, a finding for those above and below the age of 65 years. These findings should be considered when developing targeted prosthetic rehabilitation plans, noting the potential increased challenges of, for example, a female living in the South who may struggle to return to employment.

Funding

The authors disclosed that they received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interest

The authors disclosed no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental material

There is no supplemental material in this article.