Abstract
Background Mexican Americans comprise the most rapidly growing segment of the older US population and are reported to have poorer functional health than European Americans, but few studies have examined factors contributing to ethnic differences in walking speed between Mexican Americans and European Americans.
Objective The purpose of this study was to examine factors that contribute to walking speed and observed ethnic differences in walking speed in older Mexican Americans and European Americans using the disablement process model (DPM) as a guide.
Design This was an observational, cross-sectional study.
Methods Participants were 703 Mexican American and European American older adults (aged 65 years and older) who completed the baseline examination of the San Antonio Longitudinal Study of Aging (SALSA). Hierarchical regression models were performed to identify the contribution of contextual, lifestyle/anthropometric, disease, and impairment variables to walking speed and to ethnic differences in walking speed.
Results The ethic difference in unadjusted mean walking speed (Mexican Americans=1.17 m/s, European Americans=1.29 m/s) was fully explained by adjustment for contextual (ie, age, sex, education, income) and lifestyle/anthropometric (ie, body mass index, height, physical activity) variables; adjusted mean walking speed in both ethnic groups was 1.23 m/s. Contextual variables explained 20.3% of the variance in walking speed, and lifestyle/anthropometric variables explained an additional 8.4%. Diseases (ie, diabetes, stroke, chronic obstructive pulmonary disease) explained an additional 1.9% of the variance in walking speed; impairments (ie, FEV1, upper leg pain, and lower extremity strength and range of motion) contributed an additional 5.5%. Thus, both nonmodifiable (ie, contextual, height) and modifiable (ie, impairments, body mass index, physical activity) factors contributed to walking speed in older Mexican Americans and European Americans.
Limitations The study was conducted in a single geographic area and included only Mexican American Hispanic individuals.
Conclusions Walking speed in older Mexican Americans and European Americans is influenced by modifiable and nonmodifiable factors, underscoring the importance of the DPM framework, which incorporates both factors into the physical therapist patient/client management process.
Ethnic minorities are the most rapidly growing segment of the older US population.1 Projections estimate that the percentage of non-Hispanic white, or European American, older adults (≥65 years of age) will decrease from 80% in 2010 to 58% by 2050.1,2 In 2010, ethnic minorities comprised only 20% of older adults: 8.4% were African Americans, 7.0% were Hispanic, 3.5% were Asian or Pacific Islanders, fewer than 1% were American Indian or Native American, and 0.8% self-identified with 2 or more races.3 Although all ethnic groups will contribute substantially to the increasing diversity among older Americans, the older Hispanic population is projected to grow the fastest,2 to become the largest ethnic minority group in the older population by 2019,4–6 and to comprise 20% of older Americans by 2050.2,4,7 Among Hispanics, Mexican Americans are the largest subgroup, comprising about 70% of that population.8
Ethnic disparities in health and functional independence among older adults are a major public health concern, particularly among the growing Hispanic population.9 Compared with non-Hispanic whites, older Hispanics have been reported to be more functionally impaired, have greater dependencies (ie, need for assistance) in both activities of daily living (ADLs) and instrumental activities of daily living (IADLs), and greater cognitive difficulty compared with European Americans.9–11 Higher rates of disability (ie, difficulty performing ADLs and IADLs), more frequent use of hospital services, and greater prevalence of frailty also have been reported in older Hispanics compared with non-Hispanic whites.12–14
Walking speed is a powerful indicator of health and disease15,16 and among older adults, in particular, has been shown to be a valid and reliable predictor of changes in function,17–20 future health status,16,21,22 and survival.23 As a consequence, walking speed has been suggested as a “vital sign”21,24 that should be included as part of a comprehensive clinical assessment to evaluate the health needs of older adults. Data from the Hispanic Established Populations for the Epidemiological Study of the Elderly (H-EPESE), based on an exclusively Hispanic cohort, has shown that walking speed is a good predictor of disability25 and mortality26,27 in Mexican Americans; however, relatively few studies have directly examined ethnic differences in walking speed between older Hispanics and non-Hispanic whites.28,29 Although these studies have consistently shown slower walking speed in older Hispanics compared with non-Hispanic whites, most have reported these ethnic differences without examining factors that may explain the differences. A notable exception is the study by Haas et al,28 which showed that a large proportion of the ethnic disparity in walking speed results from health and socioeconomic status (SES) disadvantages in both early life and adulthood. No studies that we are aware of have used the disablement process model (DPM)30–32 to examine factors that explain ethnic differences in walking speed between older Hispanics and non-Hispanic whites or to comprehensively examine factors that contribute to walking speed in older Hispanics and non-Hispanic whites.
Thus, the purpose of the present study was to use data from the San Antonio Longitudinal Study of Aging (SALSA) survey, a community-based study of the disablement process in older Mexican Americans and European Americans, to examine not only factors that may explain ethnic differences in walking speed but also factors that contribute to walking speed regardless of their contribution to observed ethnic differences. The study was guided by the SALSA DPM,33 which was based on Verbrugge and Jette's30 conceptualization of the process as having 2 major components: (1) the main disease-disability pathway (disease to impairment to functional limitation to disability) proposed by Nagi31 and (2) modifying factors outside that pathway that can slow, speed, prevent, or even reverse progression toward disability. This approach is of particular relevance, as the scope of physical therapist practice—examination, evaluation, diagnosis, prognosis, and intervention—lies within the context of the DPM.34
It is important to note that the SALSA DPM, described more fully in the Method section, closely parallels the International Classification of Functioning, Disability and Health (ICF)35 model adopted by the American Physical Therapy Association (APTA) as its guiding framework in 2008.36 A comparison of the terminology and components of the SALSA DPM and ICF models is provided in Figure 1. Our study contributes to the evidence base in support of the utility of the DPM approach encompassed in the ICF model as well as the SALSA DPM in physical therapist practice.
Comparison of components of the San Antonio Longitudinal Study of Aging (SALSA) disablement process model (DPM) and the International Classification of Functioning, Disability and Health (ICF) model.35
Method
Participants
The study sample comprised Mexican American and European American men and women, aged 65 to 80 years, who participated in the baseline examination (1992–1996) of the SALSA survey.14 The SALSA cohort comprised the oldest people who had participated earlier in the San Antonio Heart Study (SAHS), the first large epidemiological study of diabetes and cardiovascular risk factors in Mexican Americans and European Americans, carried out in 2 phases between 1979 and 1988.37 Details of the SAHS design, sampling approach, recruitment, and field procedures have been reported previously.37 Participants were randomly sampled from low-, middle-, and high-income neighborhoods to provide a cohort with comparable numbers of Mexican Americans and European Americans within neighborhoods and to maximize sociocultural variation among Mexican Americans in the study. The response rate to the SALSA baseline examination among age-eligible SAHS cohort members was 70.5% (749 of 1,063), with no evidence of major attrition bias between the initial SAHS survey and the SALSA survey.
The SALSA baseline examination consisted of a comprehensive home-based assessment, conducted in the participant's home, and a performance-based assessment, conducted at a clinical research center. Trained, bilingual staff administered assessments in English or Spanish, according to the participant's preference, using standardized protocols. The SALSA survey was approved by the University of Texas Health Science Center at San Antonio's Institutional Review Board, and all participants gave informed consent.
Participants for the current study were the 703 participants who had complete data for variables included in the analyses (Fig. 2.)
Flow chart of study participants. SAHS=San Antonio Heart Study, SALSA=San Antonio Longitudinal Study of Aging, BMI=body mass index, HBP=high blood pressure, COPD=chronic obstructive pulmonary disease, FEV1=forced expiratory volume in 1 second, LVH=left ventricular hypertrophy, ROM=range of motion.
Measures
As operationalized in the SALSA DPM (Fig. 3), modifying factors outside the main disease disability pathway included contextual and lifestyle factors. Age, sex, ethnic group, and SES (education and income) were treated as contextual factors that provide the context of daily living for individuals as they age and that may have a substantial independent effect on discrete stages in the disablement process. Body mass index (BMI), a function of height and weight, was categorized as a lifestyle factor because it is, to a large extent, the product of behavioral choices related to food intake and physical activity. Because of the potential independent contribution of height to walking speed, we have included height in the analyses as a separate variable and hereafter refer to lifestyle/anthropometric factors rather than lifestyle factors.
San Antonio Longitudinal Study of Aging (SALSA) disablement process model: main disease-disability pathway and modifiers.33 MI=myocardial infarction.
The SALSA DPM main disease-disability pathway incorporated a number of chronic diseases, with a primary focus on diabetes, cardiovascular diseases (ie, myocardial infarction, high blood pressure, angina, stroke), and arthritis. Following Verbrugge and Jette,30 impairments were defined as significant structural abnormalities and dysfunctions in specific body systems30; cardiopulmonary, neurosensory, and musculoskeletal impairments were included in the model. Walking speed is a basic physical action used in many different situations in daily life by an individual's sex-age group and involves the whole person.30 Thus, poorer performance in walking speed was treated as a functional limitation. Measures of variables included in the present analyses are described below, organized according to components of the SALSA DPM.
Contextual factors.
Age and sex were assessed by participant self-report. Ethnic classification as Mexican American or European American was based on a validated, standardized algorithm, which considers concordance of parental surnames (maiden name of mother), birthplace of both parents, self-declared ethnic identity when it indicates a distinct national origin, and ethnic background of grandparents.38 Years of formal education and annual household income were assessed by self-report.
Lifestyle/anthropometric factors.
Height and weight were measured after participants had removed their shoes and upper garments and donned an examination gown. Body mass index was calculated as weight (in kilograms) divided by height (in meters) squared. Physical activity also was measured using the Minnesota Leisure Time Questionnaire, which allows calculation of kilocalories of energy expenditure per week.39 Weekly kilocalories were collapsed into 3 categories: <500 kcal/week, 500–999 kcal/week, and 1,000+ kcal/week.
Chronic diseases.
Diabetes mellitus was classified, according to American Diabetes Association criteria,40 as fasting blood glucose ≥126 mg/dL or currently taking glucose-lowering medication. Hypertension was defined, according to Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure 6 (JNC6) guidelines, as systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg.41 Myocardial infarction was assessed by the presence of ischemic electrocardiograph abnormalities using a 12-lead electrocardiograph.42 Angina pectoris was measured using the Rose questionnaire.43 Stroke and arthritis were assessed based on participant self-report of physician-diagnosed disease.
Chronic obstructive pulmonary disease (COPD) was assessed with the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria, based on forced expiratory volume in 1 second (FEV1) and the ratio of FEV1 to forced vital capacity as previously described.44
Impairments.
Pulmonary impairment was measured by FEV1 using the Welch Allyn Pneumocheck spirometer (Welch Allyn Inc, Skaneateles Falls, New York)45,46; the average of 3 measures was used for analysis. When compared with clinical pulmonary function measures, the Pneumocheck FEV1 is directly comparable (r=.98) to no systematic differences.47 Left ventricular hypertrophy (LVH) was classified based on readings from a 12-lead electrocardiograph.42
Near and far visual acuity was measured under standardized conditions using the charts developed for the Early Treatment of Diabetic Retinopathy Study.48,49 Visual acuity was measured as a log of the minimal angle of resolution (logMAR). A single measure for both eyes was calculated as a weighted logMAR, with higher values indicating poorer vision.50
Lower extremity strength and joint range of motion (ROM) were measured using the Physical Disability Index (PDI).51 Three measurements were taken at each of 4 muscle groups (ie, hip flexors, knee flexors and extensors, and ankle dorsiflexors) using a hand-held dynamometer (Nicholas Manual Muscle Tester, Nicholas Institute of Sports Medicine and Athletic Trauma, New York, New York). The score for each group was calculated by computer at data entry as the average of the 3 measurements. A summary measure of lower extremity strength was calculated according to the PDI scoring algorithm, with scores representing percent of maximum performance in the tested sample.
Lower extremity ROM measures included knee flexion and extension, ankle dorsiflexion and plantar flexion, and hip flexion. Hip flexion was measured using a Cybex EDI Inclinometer (Cybex, Ronkonkoma, New York), and a standard goniometer was used for all other joints. A summary measure for lower extremity joint ROM was calculated according to the PDI scoring algorithm, with scores representing percent of maximum performance in the tested sample.
Interrater reliability among 4 observers who measured muscle strength and ROM was evaluated in a sample of 24 ambulatory adults aged 65 to 80 years using a Latin square design. Agreement among the 4 observers was excellent (intraclass correlation coefficient [ICC] for lower extremity strength=.71, ICC for lower-extremity ROM=.80).
Pain located in the upper and lower leg was assessed with the McGill Pain Map (MPM) using a standardized scoring procedure.52
Functional limitation.
Walking speed was measured as part of the performance-based assessment and calculated as distance (in meters) divided by time (in seconds). Participants were instructed to walk at their usual and comfortable pace over a 50-ft (15.24-m) distance beginning from a standing start. The examiner walked behind the participant to ensure safety while not influencing pace and measured walking time with a stopwatch. The mean of 2 trials was used for analysis.
Data Analysis
Bivariate associations of ethnic group with contextual and lifestyle/anthropometric variables, diseases, impairments, and 50-ft walking speed were examined using the chi-square statistic for categorical variables and t tests for continuous variables.
A series of multivariable regression models were performed to examine the contribution of these variables to walking speed and ethnic differences in walking speed and to assess the amount of variance explained by contextual and lifestyle/anthropometric factors, chronic diseases, and impairments as operationalized in the SALSA DPM. Four hierarchical regression models were tested: model 1 analyzed contextual factors (ie, age, sex, ethnicity, SES); model 2 analyzed contextual and lifestyle/anthropometric factors (ie, BMI, height, and physical activity); model 3 analyzed contextual and lifestyle/anthropometric factors and chronic diseases (ie, diabetes, myocardial infarction, angina, hypertension, stroke, arthritis, and COPD); and model 4 analyzed contextual and lifestyle/anthropometric factors, chronic diseases, and cardiopulmonary, neurosensory, and musculoskeletal impairments (ie, FEV1, LVH, near and far vision, upper and lower leg pain, and lower extremity strength and joint ROM). There was no imputation for missing data. Means and standard errors for ethnic differences were obtained by analysis of covariance. The adjusted R2 for each model, for covariates, and for ethnic group net of the covariates was generated by regression analyses using a hierarchical approach. Statistical significance was set at P<.05. All analyses were performed with SPSS version 21 (IBM Corp, Armonk, New York).
Role of the Funding Source
This research was supported by National Institute on Aging (NIA) R01-AG10444 and NCRR Grant M01-RR01346 (Fredrick C. Bartler General Clinical Research Center). Dr Quiben was supported by the Castella Geriatric Research Fellowship from the School of Health Professions, University of Texas Health Science Center at San Antonio.
Results
Sample characteristics, grouped according to components of the SALSA DPM,33 are shown by ethnic group in Table 1. Mexican Americans were significantly younger and had lower SES, shorter height, and higher BMI values compared with European Americans. Among the chronic diseases examined, the ethnic difference was statistically significant only for diabetes, with Mexican Americans having an almost threefold higher prevalence of this disease compared with European Americans. Near and far vision, FEV1, lower-extremity ROM, and 50-ft walking speed were significantly lower in Mexican Americans than in European Americans.
Sample Characteristicsa
Table 2 shows results for the 4 models that regressed walking speed hierarchically on contextual factors, lifestyle/anthropometric factors, chronic disease, and impairments. Adjusted variances (adjusted R2), mean walking speed for Mexican Americans and European Americans, and mean ethnic difference in walking speed (m/s) are included with each model.
Regression Models Examining Contributors to 50-ft Walking Speed in a Biethnic Cohort of Community-Dwelling Older Adultsa
Prior to examining multivariable models, the unadjusted mean walking speed for Mexican Americans and European Americans was 1.17 and 1.29 m/s, respectively (Tab. 1); the ethnic difference was 0.12 m/s (P<.001). In model 1, after adjustment for other contextual factors (ie, age, sex, SES), the ethnic group difference remained statistically significant (P=.04), but the mean difference in walking speed was substantially reduced (Mexican Americans: 1.21 m/s; European Americans: 1.25 m/s; ethnic difference=0.04 m/s), and ethnicity explained only 0.4% of the variance in walking speed. Age and sex were significantly associated with walking speed and together explained 5.9% of the variance; age was negatively associated with walking speed, whereas male sex was positively associated with walking speed. Socioeconomic status was significantly and positively associated with walking speed and explained an additional 14% of the variance. The full contextual model, including ethnicity, explained 20.3% of the variance in walking speed.
Additional adjustment for lifestyle/anthropometric variables (ie, BMI, height, and physical activity) in model 2 eliminated the association of ethnicity with walking speed. Adjusted mean walking speed was identical for Mexican Americans and European Americans (1.23 m/s). Lifestyle/anthropometric factors were all highly statistically significant and explained 8.4% of the variance in walking speed net of contextual factors. Body mass index was negatively associated with walking speed and explained the largest amount of variance (4.0%); both height and physical activity were positively associated with walking speed, with height explaining 2.5% of the variance and physical activity explaining 1.9% of the variance. The full contextual-lifestyle/anthropometric model explained 28.7% of the variance in walking speed.
Model 3 further adjusted for diseases (ie, diabetes, angina, myocardial infarction, stroke, arthritis, hypertension, and COPD). Only diabetes, stroke, and COPD were significantly associated with walking speed net of contextual and lifestyle/anthropometric factors; all 3 diseases were associated with slower walking speed and explained an additional 1.9% of the variance. The full contextual-lifestyle/anthropometric-disease model explained 30.6% of the variance in walking speed.
Model 4 included additional covariate adjustment for impairments in the following systems: cardiovascular (ie, FEV1 and LVH), neurosensory (ie, near and far vision), and musculoskeletal (ie, upper and lower leg pain and lower extremity strength and ROM). Among these impairments, FEV1, upper leg pain, and lower extremity strength and ROM were significantly associated with walking speed; FEV1 and lower extremity strength and ROM were positively associated, whereas upper leg pain was negatively associated with walking speed. These impairments explained an additional 5.5% of the variance net of contextual, lifestyle/anthropometric, and disease factors. Musculoskeletal impairments accounted for 2.7% of the variance, whereas cardiopulmonary and neurosensory impairments together accounted for 2.8% of the variance. The full contextual-lifestyle/anthropometric-disease-impairment model explained 36.1% of the variance in walking speed.
Discussion
Guided by the DPM developed in the SALSA survey, the present study was designed to extend current knowledge not only about factors that explain reported ethnic differences in walking speed but also about factors that influence walking speed more generally in older Mexican Americans and European Americans. Using a hierarchical analytic approach, we found that the ethnic difference in walking speed could be fully explained by contextual and lifestyle/anthropometric factors (adjusted mean walking speed in both Mexican Americans and European Americans was 1.23 m/s) without considering diseases and impairments. In addition, contextual factors, including ethnicity, explained 20.3% of the variance in walking speed, with SES alone accounting for 14% of the variance. Lifestyle/anthropometric factors accounted for an additional 8.4% of the variance, with BMI explaining 4%, height 2.5%, and physical activity 1.9%. Together, contextual and lifestyle/anthropometric factors explained 28.7% of the variance in walking speed and, thus, explained not only the ethnic difference in walking speed but also a substantial proportion of the variance in overall walking speed in older Mexican Americans and European Americans. With the exception of BMI and physical activity, these factors are essentially nonmodifiable.
Although contextual and lifestyle/anthropometric factors were sufficient to explain the ethnic difference in walking speed, diseases and impairments explained an additional 7.4% of the variance in walking speed. Diabetes, stroke, and COPD explained 1.9% of the variance over and above that explained by contextual and lifestyle/anthropometric factors. Forced expiratory volume in 1 second, upper leg pain, and lower extremity strength and ROM explained 5.5% of the variance over and above that explained by contextual, lifestyle/anthropometric, and disease factors. Thus impairments, primarily in the musculoskeletal system, explained almost 3 times more of the variance in walking speed than that explained by disease variables. This finding would be expected based on the DPM, which posits that diseases lead to impairments in various body systems and that impairments, in turn, lead to functional limitations, including slower walking speed. Consistent with the DPM, these impairments, which are key modifiable targets for physical therapist intervention to improve walking speed, result from antecedent contextual, lifestyle/anthropometric, and disease characteristics.
Methods used to measure walking speed across studies have varied in terms of starting protocol (standing versus moving), pace (usual versus fast), and distance timed (2–15 m). Recent reviews examining the influence of these methodological differences on reported walking speed in older populations showed there was no significant difference between unadjusted mean walking speed and walking speed adjusted for starting protocol53,54 and that distances walked (up to 50 ft) during the gait test did not significantly influence the gait speed recorded.54–56
Both the Health and Retirement Study28 and the Whitehall II Study57 measured walking speed from a standing start and at participants' usual pace (as we did in the SALSA survey) but measured walking speed over 8 ft (2.5 m) rather than 50 ft. As in our study, which showed that SES accounted for 14% of the variance in walking speed, these studies demonstrated a substantial effect of SES on walking speed. The Health and Retirement Study28 showed that both childhood and adult SES are major determinants of race/ethnic differences in walking speed, and the Whitehall II Study57 showed that age- and ethnicity-adjusted mean walking speed was 13% higher in the highest employment grade (defined by salary and pension entitlement) compared with the lowest employment grade. Higher SES contributes not only to better access to health care resources and overall health care use but also to improved knowledge of health behaviors58,59 that may play a role in physical functioning, including walking speed. Lower SES, on the other hand, has been related to sedentary lifestyle, obesity, chronic disease, and physiological impairments,57,58 all of which negatively influence walking speed. Adult height is also a reflection of childhood nutrition and other early life circumstances shaped by SES.57
Although our findings indicate that ethnicity is not an independent risk factor for slower walking speed (ie, the observed ethnic differences are explained by contextual and lifestyle/anthropometric factors), the unadjusted ethnic difference in mean walking speed was statistically significant (P<.001), although relatively small (0.12 m/s). Thus, it is important to consider whether the observed ethnic difference is clinically meaningful. Perera et al60 reported that a change of 0.05 m/s in walking speed over time represents a small meaningful change and a change of 0.10 m/s represents a substantial change and further indicated that, for clinical use, even a small meaningful change in walking speed is detectable. Cutpoints for clinically meaningful values of walking speed performance among community-dwelling older adults also have been established: superior=≥1.4 m/s, normal=1.0 to 1.4 m/s, mildly abnormal=0.6 to 1.0 m/s, and seriously abnormal=<0.6 m/s.15,61,62 The proportion of Mexican American and European American participants of the SALSA survey whose walking speed fell within each performance level was: superior: Mexican Americans 18.8%, European Americans 30.6%; normal: Mexican Americans 59.02%, European Americans 60.5%; mildly abnormal: Mexican Americans 18.8%, European Americans 7.1%; and seriously abnormal: Mexican Americans 3.3%, European Americans 1.8% (P<.001 for ethnic difference).
Compared with European Americans in our cohort, the proportion of Mexican Americans whose walking speed would be classified as seriously abnormal was about 2 times greater and mildly abnormal was 2.6 times greater, whereas the proportion of Mexican Americans whose walking speed would be classified as superior was almost 40% less. The magnitude of the ethnic difference in unadjusted mean walking speed and the proportions classified as having seriously abnormal, mildly abnormal, and superior walking speed support the conclusion that the difference in unadjusted walking speed between older Mexican Americans and European Americans in our cohort is clinically meaningful.
The walking speed that older Mexican American and European American patients seeking physical therapist services present with is, in effect, unadjusted walking speed; therefore, some clinicians may observe ethnic disparities in walking speed. In the event that this occurs, clinicians should be aware of the modifiable and nonmodifiable factors that contribute to these disparities and consider them in the examination and development of potential targets for the components of intervention34: patient-related instruction, coordination and communication, and direct intervention.
Limitations and Strengths
The present study has several important strengths. The SALSA survey focused on the disablement process and used standardized measures. The sets of variables analyzed in this study were guided by the SALSA DPM. A published, validated algorithm38 was used to classify participants as Mexican American or European American. Participants were community-dwelling older adults randomly sampled from socioeconomically diverse neighborhoods so that they represented a broad range of SES. The SALSA cohort comprised both Mexican Americans and European Americans, permitting us to directly compare walking speed in the 2 ethnic groups.
The study also had some limitations. It was an observational study conducted in a single geographic area and included only Mexican Americans and European Americans 65 years of age and older. The findings, therefore, may have limited generalizability to older Mexican Americans and European Americans in other geographic areas, to other Hispanic subgroups, or to individuals younger than 65 years of age. In addition, the summative model was based on 79.4% (558/703) of the sample due to missing data.
Clinical Implications of Findings for Physical Therapy
Our findings about the relative contribution to walking speed of contextual and lifestyle/anthropometric factors outside the DPM main disease-disability pathway and diseases and impairments within that pathway highlight the utility of the ICF model as a guide to evidence-based physical therapist practice and the importance of considering all major components of the DPM.
Socioeconomic status is a fundamental, nonmodifiable factor that has a major influence on walking speed. Even though clinicians taking a holistic approach to rehabilitation cannot directly change a patient's SES, clinicians need to be aware of the impact of patients' life circumstances on walking speed and related physical function, particularly those influenced by SES, and take them into account when developing a plan of care.
Much has been written about SES inequalities and their contribution to disparities in health outcomes. Knowledge of the impact of SES63 on economic opportunities, environmental risks, housing, and nutrition can directly influence the delivery of physical therapist interventions. In physical therapy, consideration of the individual's SES involves attention to the environment, available support systems, and health behaviors that may influence participation and adherence to patient-related instructions (eg, home exercise programs, education on physical activity) or direct interventions that involve the patient's or client's environment/community (eg, self-care, home management, work integration). Physical therapists also should take into account education level when choosing appropriate written materials and patient-related instructions. Consideration of SES, therefore, influences multiple aspects of physical therapist intervention: communication and coordination, content of the instructions provided to patients and their families, and procedural interventions that are influenced by existing resources (eg, walking as an intervention when the patient has limited access to safe environments). Consideration of other nonmodifiable factors such as age, sex, and ethnicity will similarly influence multiple aspects of physical therapy intervention.
The DPM classifies slower walking speed as a functional limitation and posits that functional limitations drive disability.64 The Mexican Americans in our study were about 2 to 2.6 times more likely than European Americans to have walking speed in the abnormal ranges, reflective of functional limitations in walking speed, and, therefore, may be at greater risk than European Americans for developing incident disability, depending on their specific contextual and lifestyle/anthropometric characteristics. However, given its posited role as a driver of disability, slower walking speed may be an optimal target for clinical intervention in both ethnic groups to prevent, slow, or reverse progression toward disability.
Finally, our findings support the recommendation that walking speed be considered a vital sign in physical therapist practice21,24 and that it be routinely incorporated into a standard physical examination for older adults.65 Because walking requires the coordination of organ systems, even if referral for physical therapy was due to a specific diagnosis-related reason, slow walking speed in the severely or mildly abnormal ranges should trigger a referral to the appropriate medical profession for further evaluation of disease-related factors that may affect walking speed.23,65,66
Footnotes
Both authors provided concept/idea/research design, writing, and data analysis. Dr Hazuda provided data collection, fund procurement, and participants.
The University of Texas Health Science Center at San Antonio Institutional Review Board approved the study.
This research was supported by National Institute on Aging (NIA) R01-AG10444 and NCRR Grant M01-RR01346 (Fredrick C. Bartler General Clinical Research Center). Dr Quiben was supported by the Castella Geriatric Research Fellowship from the School of Health Professions, University of Texas Health Science Center at San Antonio.
- Received April 6, 2014.
- Accepted January 2, 2015.
- © 2015 American Physical Therapy Association
References
Issue highlights
