Background and Purpose. This study looked at adherence, and factors affecting adherence, to a prescribed home exercise program (HEP) in older adults with impaired balance following discharge from physical therapy.

Subjects. The subjects were 556 older adults (≥65 years of age) who were discharged from physical therapy during the period 2000 to 2003.

Methods. A survey was developed to determine participation in a HEP. Univariate logistic regressions identified specific barriers and motivators that were associated with exercise participation following discharge from physical therapy.

Results. Ninety percent of respondents reported receiving a HEP; 37% no longer performed it. Change in health status was the primary reason for poor adherence to a HEP. Eight barriers (no interest, poor health, weather, depression, weakness, fear of falling, shortness of breath, and low outcomes expectation) were associated with a lack of postdischarge participation in exercise.

Discussion and Conclusions. Exercise adherence following discharge from a physical therapy program is poor among older adults. Barriers, not motivators, appear to predict adherence. [Forkan R, Pumper B, Smyth N, et al. Exercise adherence following physical therapy intervention in older adults with impaired balance. Phys Ther. 2006;86:401–410.]

The benefits of exercise in older adults are well established. Research has shown that for older adults (≥65 years of age) exercise can reduce frailty,1 increase walking speed,2, 3 improve the ability to live independently,4 and increase life expectancy.5, 6 Exercise significantly reduces the risk for cardiovascular disease,6 adult-onset diabetes, and osteoporosis,4, 5, 7 and it is associated with a reduction in health care costs.8, 9 In addition, participation in regular exercise can significantly improve balance and reduce the risk of falls in older adults.35, 1017

Researchers4, 1823 have identified a variety of factors that determine adherence to exercise in older adults. Some of these factors increase adherence (motivators), whereas others decrease adherence (barriers). For example, one of the strongest motivators affecting exercise adherence in older adults is self-efficacy (the concept that a person is capable of controlling his or her own behavior).1820 A second motivator is outcome expectation, which is the belief that specific consequences will result from specific personal actions.18, 20 Resnick and Spellbring21 examined factors that facilitated adherence to a walking program in a group of 23 older adults (≥65 years of age) in a continued care retirement community. They concluded that older adults who adhered to exercise were characterized by an inner motivation to exercise, a belief that they were able to exercise safely (self-efficacy), a recognition of the benefits of exercise (outcome expectation), the ability to set specific activity-related goals, and an enjoyment of walking.

Barriers shown to decrease adherence to exercise in older adults include insufficient time, lack of social support, no place to exercise, no transportation to an exercise site, and insufficient money to either buy exercise equipment or join an exercise facility.4, 22, 23 Fear of falling and fear of injury while exercising also are significant barriers.4, 22, 23 Finally, researchers1820, 23 have found that increased stress and depression levels, increased age, decreased health status, and lack of enjoyment while exercising are associated with poor exercise adherence.

According to the American Physical Therapy Association’s Guide to Physical Therapy Practice (Guide), physical therapists should strive to “restore, maintain, and promote not only optimal physical function but optimal wellness and fitness and optimal quality of life as it relates to movement and health.”24(p21) Prescribing and promoting exercise, and educating patients on the importance and value of exercise are key responsibilities of the physical therapist. Several studies13, 14, 25, 26 have shown the benefits of participation in a multidimensional physical therapist–supervised exercise program on balance, walking, and fall risk among older adults. However, there is limited information on the degree to which benefits are sustained after participating in a physical therapy program. Some studies2, 3, 26 have shown that benefits of exercise gained during physical therapy often are not maintained after discharge. Lack of sustained benefits from physical therapy may be the result of poor adherence to a prescribed home exercise program (HEP) that is designed to promote the maintenance of improved function following discharge.

Little is known about adherence to a HEP prescribed by a physical therapist in older adults following discharge or about the factors that affect postdischarge adherence. Thus, the purpose of this study was to develop an instrument to survey older adults following discharge from a physical therapy balance-training program in order to: (1) determine postdischarge adherence and factors that limit adherence to a prescribed HEP and (2) to characterize overall participation in exercise, including both prescribed and nonprescribed exercise, as well as factors influencing exercise participation.



Survey questionnaires were mailed to a convenience sample of 630 adults aged 65 years and older who had completed a physical therapist–supervised balance-training program at either the University of Washington Medical Center (Strong and Steady program) or the Northwest Hospital (Safety and Gait Enhancement [SAGE] program) between the years of 2000 and 2003. The SAGE and Strong and Steady programs are hospital-based outpatient physical therapy programs and were developed by the same individual using a model for improving balance and reducing fall risk through the use of multidimensional exercises, including cardio-vascular fitness, strength (muscle force-generating capacity) and flexibility exercises, and balance and mobility training.14

Older adults are referred to the Strong and Steady or SAGE program by a physician because of balance and gait impairments with a history of falls or near falls, an increased risk of falls, or restricted activity levels due to fear of falling. The majority of patients have impaired balance and gait associated with chronic health conditions such as arthritis, diabetes, and hypertension; however, a limited number are referred with neurological or degenerative pathology. Once admitted into the program, patients are examined by a licensed physical therapist who is trained in the special examination and treatment procedures for these programs, and an individualized exercise program is developed following standardized treatment algorithm. Patients are treated 1 to 3 times per week for 4 to 6 weeks depending on the severity of the balance problems.

An integral part of both programs is the establishment of an individualized, home-based exercise program that includes resistance strength training, flexibility exercises, cardiovascular fitness (usually a progressive walking program), and balance exercises, which are to be performed on most days of the week. Patients are provided with an exercise log to track adherence. Patients are discharged with an individualized HEP, with the instructions to continue this lifelong exercise program on a regular basis. Records of the duration of each individualized session were not obtained for this study.

Patients from these 2 programs were selected to participate in the survey because both programs emphasize lifelong exercise as an integral part of the discharge plan. In addition, both programs keep computerized demographic records of patients completing the program. Demographic information was used to conduct the survey research.

Development of the Survey

A survey questionnaire was the selected method of data collection in order to ensure subject anonymity, to provide a standard form to decrease potential bias, and to reach a large population inexpensively in a short period of time.27 The MEDLINE database was used to review research examining factors affecting adherence to exercise in community-dwelling older adults. In addition, research focused on developing instruments that measure factors affecting adherence to exercise in older adults were identified. Based on the outcomes of this literature search, an initial survey instrument was developed that included 43 questions divided into 7 sections to determine adherence to a prescribed HEP, participation in additional forms of physical activity excluding a HEP, factors affecting adherence to a prescribed HEP, barriers to and motivators for physical activity, fall history, general health status, and demographics.

A pilot survey was administered to 5 older adults and 3 physical therapists to reveal any limitations in the design and sequence of questions and to clarify operational definitions. The 5 older adults were selected from among those volunteering to participate in a geriatric physical therapy class. Of the 3 physical therapists selected, one had extensive experience in survey research, one was the coordinator of the Strong and Steady program, and one was a therapist in the SAGE program. A feedback questionnaire containing questions regarding time to complete the survey, clarity of the survey questions, and difficulty in completing the survey accompanied the pilot survey instrument. Revisions to the survey instrument included changes to the sequence and layout of the survey, clarifications of questions, and corrections to punctuation or spelling errors. Because minimal revisions were made, a second pilot survey instrument was not necessary.27

The final survey instrument was a self-administered, 43-item questionnaire composed predominantly of closed-ended questions. Closed-ended questions were developed to provide uniform answers and to simplify coding for data analysis.27 One open-ended question was included to give respondents an opportunity to describe any new medical problems that had been diagnosed since discharge from either the SAGE or Strong and Steady program. Three questions included an “other” category with prompts and lines provided to further describe their response. Neutral and interesting questions were included in the beginning to establish curiosity and subject trust, whereas demographics and more sensitive questions (eg, fall history, barriers and motivators, depression level) were placed at the end.27 Questions were grouped according to topic, resulting in 7 sections. The subjects’ answers were numerically coded and analyzed using SPSS version 11.5.28,* Nonresponses were coded as missing.

Survey Instrument

Section 1 began with a question to determine whether the subject received a HEP upon discharge; if so, subjects were directed to answer questions that queried participation in the prescribed HEP over the previous 4 weeks. Continuous scale questions were used to quantify behavior in terms of frequency of participation and duration of each exercise bout. A closed-ended multiple-choice question regarding mode of exercise was included in this section. The modes of exercise included in the list of options were consistent with those that were an integral part of the multidimensional HEP incorporated in the SAGE and Strong and Steady programs. The final question in this section queried barriers associated with the subject’s HEP.

Due to lack of literature regarding barriers specific to HEP adherence, literature on barriers shown to decrease adherence to general exercise in older adults was reviewed.4, 1820, 22, 23 Those barriers that appeared most relevant to a HEP were selected and included in a “check all that apply,” multiple-choice question to determine the reason or reasons for nonadherence to the HEP. An open-ended opportunity for providing other reasons that were not included in the multiple-choice answer was provided.

Section 2 determined participation in exercise other than the prescribed HEP, including frequency of participation in muscle strengthening, flexibility, and aerobic activity of at least 30 minutes in duration. These modes of exercise and the duration for aerobic activity are consistent with those prescribed for older adults by American College of Sports Medicine and the Centers for Disease Control and Prevention.4, 29 Examples of modes were provided to help the subject determine participation accurately. Section 3 consisted of questions regarding availability of social support associated with any physical activity and the location at which physical activity was performed. These questions were included because previous literature has shown these 2 items can be associated with exercise participation.4, 22, 23 Section 4 included 2 subscales from a published survey instrument examining factors affecting adherence to exercise in community-dwelling older adults.30 The barrier subscale contained 13 questions, and the motivator subscale contained 11 questions.

In addition, a published single-question instrument31 was used to examine the subject’s perception regarding the expected outcome for participating in a home-based exercise program. Outcomes expectation has been identified as an important factor in determining long-term adherence to an exercise program among older adults.30

Section 5 addressed fall frequency in the previous 3 months, fall-related injuries requiring medical attention, and reasons for falls. Questions on fall history were included because previous research has shown that among older adults, falls can affect participation in exercise.4, 22, 23 Section 6 included questions related to current health status. Previous research3234 has shown a relationship between the presence of adverse health conditions and exercise participation. A checklist of common comorbidities was included to determine number and type of comorbidities. An opportunity for subjects to report other comorbidities was provided. Participants listed current prescription medications. Finally a 2-question instrument was used to screen for depression.35 Section 7 was composed of dichotomous and “fill in the blank” questions to determine subject demographics, including the following: date of discharge from physical therapy, age, weight, height, sex, and current living status, to compile a subject profile.


A cover letter explaining the purpose of the survey and a self-addressed stamped envelope were included with the survey questionnaire. Survey questionnaires were returned anonymously; however, they were color-coded for each program. Reminder postcards were sent out 4 weeks after the initial mailing. Completion of the questionnaire implied informed consent to participate in the survey.

Data Collection and Analysis

Descriptive statistics coded and analyzed in SPSS28 were used to summarize demographic and health status data in the 2 groups (SAGE and Strong and Steady). Differences in baseline characteristics were analyzed using the t test and chi-square test as appropriate to assess the statistical difference between the 2 groups.

The number of people who reported receiving a prescribed HEP on discharge was analyzed. Mode, frequency, and duration of participation in the prescribed HEP was described, and the most frequently reported combinations of exercises were determined. For nonprescribed physical activity, percentages were calculated to determine the frequency of participation in strengthening, flexibility, and aerobic activity.

Survey respondents reported on motivators and barriers using a 4-point ordinal scale. Barriers were scored as follows: 1=strongly agree, 2=agree, 3=disagree, and 4=strongly disagree; motivators were scored as follows: 4=strongly agree, 3=agree, 2=disagree, and 1=strongly disagree. A barrier subscale score was calculated by summing all of the barrier scores for an individual (0–52). Motivator subscale scores (0–44) were calculated in the same manner. The 4-point ordinal scale was dichotomized with “strongly agree” or “agree” responses combined and coded as “item endorsement,” while the “disagree” or “strongly disagree” responses were combined and coded as “no endorsement.” Using the dichotomous scoring system, the total number of endorsed motivators (0–11) and barriers (0–13) was calculated.

A t test was used to test for differences among the 2 groups (adults who reported participating in some exercise versus those who did not participate) in total number of barriers or motivators. Univariate logistic regressions were used to study the specific barriers and motivators that predicted continued participation in exercise following discharge from physical therapy. Participants were grouped into 1 of 3 categories based on time since discharge: <12 months, 12–48 months, and >48 months. Chi-square analysis was used to determine whether adherence to exercise (either to a prescribed HEP or nonprescribed physical activity) was associated with time since discharge. All statistical tests were performed at the .05 significance level.


Characteristics of Survey Respondents

From January 2004 to March 2004, 179 completed survey questionnaires were returned. Seventy questionnaires were returned as undeliverable and 4 questionnaires were excluded from the study (2 survey participants were <65 years of age and 2 survey participants were discharged prior to 2000). The final response rate for the total survey was 31.5% (175/556); however, because not all respondents answered all questions, sample size varied by question.

There were no statistically significant differences in age, sex, or health status between the SAGE and Strong and Steady respondents. As shown in Table 1, the average age was 81.9 years (SD=6.0) for the SAGE respondents and 81.0 (SD=7.1) for the Strong and Steady respondents; the majority were female (70.8% of the SAGE respondents, 69.2% of the Strong and Steady respondents); and the average number of comorbid health conditions reported was 3.1 (SD=1.9) for the SAGE respondents and 2.9 (SD=1.9) for the Strong and Steady respondents. Finally, the percentage of respondents who reported receiving a prescribed HEP was similar for both programs (SAGE=91.7%, Strong and Steady=80.8%). Because demographic characteristics were comparable between both groups they were combined for subsequent data analysis.

Table 1.

Demographic and Health-Related Characteristics of Survey Respondentsa

Characterizing Participation in a Prescribed HEP

Ninety percent (153/170) of survey respondents reported receiving a HEP as a part of their physical therapy program. Data on characterization and adherence to a prescribed HEP is shown in Table 2. Frequency of participation in the prescribed HEP in the past 4 weeks was low, 36.6% (56/153) of the respondents reported no exercise, and only 9.2% (14/153) of the respondents reported adherence on 5 or more days per week. Duration of HEP was most often reported as 30 minutes or less. The 2 most frequently reported modes of exercise were strength training (70.9%, 73/103) and balance training (69.9%, 72/103). Flexibility exercises were preformed by 52.4% (54/103) of the respondents, and 45.6% (47/103) of the respondents reported some form of aerobic exercise. As shown in Table 2, most respondents participating in their HEP reported performing more than one mode of exercise.

Table 2.

Reported Postdischarge Participation in a Prescribed Home Exercise Program (HEP)

Adherence to exercise was independent of time since discharge (chi-square test for independence of the 2 variables, P=.553). This finding is shown in Table 3, which compares the distribution of subjects who reported some participation in a prescribed HEP versus those who reported no participation in a prescribed HEP as a function of time since discharge (categorized as <12 months, 12–48 months, and >48 months).

Table 3.

Effect of Time Since Discharge on Adherence to a Prescribed Home Exercise Program (HEP)

The most frequently reported reason given for poor adherence to a prescribed HEP was a change in health status (30.7%, 31/101). Table 4 lists other reasons reported for lack of adherence among survey respondents.

Table 4.

Reported Reasons for Nonadherence to Prescribed Home Exercise Program (HEP)

Participation in Nonprescribed Physical Activity

Participation in nonprescribed physical activity was examined to determine whether this was a factor in low adherence to a prescribed HEP. Table 5 summarizes the frequency of participation by mode of exercise. For all modes, including strengthening, aerobic, and flexibility exercises, the highest percentage of respondents reported participating less than one time per week.

Table 5.

Reported Frequency of Participation in Exercise (by Mode), Not Prescribed in the Home Exercise Program (HEP)

Older adults were grouped according to participation in exercise. Categorization was determined either through participation in a prescribed HEP or through participation in other forms of physical activity. Respondents who reported exercising at least one time per week in the past 4 weeks were classified as exercisers (83.3%), whereas respondents reporting no exercise in the past 4 weeks were categorized as nonexercisers (16.7%). Participation in nonprescribed physical activity was not associated with time since discharge (chi-square test for independence of the 2 variables, P=.185).

Factors Affecting Participation in Exercise

The exercise and the nonexercise groups differed (P<.001) in the number of barriers to exercise, with the nonexercise group reporting more barriers (X̄=7.2, SD=1.8) compared with the exercise group (X̄=4.6, SD=2.6). The number of motivators did not differ between the 2 groups (P=.65). A series of univariate logistic regressions was performed on individual barriers and motivators to determine which were most associated with adherence to exercise following discharge from a physical therapy program.

Table 6 shows the results for the logistic regressions. This table summarizes the proportion of participants from each group (exercisers versus nonexercisers) who endorsed the item and the number who did not answer the question (missing data), which is followed by the P for a specific motivator or barrier in the model and its odds ratio and 95% confidence interval. For example, among respondents, 68.4% of the nonexercisers and 59.7% of the exercisers endorsed the motivator “prefer to be in scheduled exercise program” (second motivator in the table), while 37 participants did not answer that item. Additionally, the P of .469 implies that this motivator was not statistically significant (ie, this motivator does not explain why a person becomes an exerciser or why he or she does not). Odds ratios and confidence intervals were calculated for all items for which the calculation was possible. The odds ratio is the odds that a person who endorses an item is an exerciser when compared to a person who did not endorse the item. For example, for the first barrier in the table, a person who is not interested in exercise is 0.21 times less likely to be an exerciser than a person who did not endorse that item.

Table 6.

Univariate Logistic Regressionsa

In the models, none of the motivators were associated with exercise participation. Eight barriers were associated (P<.05) with decreased participation in exercise following discharge from a physical therapy program and are shown in Table 6 in boldface. The barrier most associated with reduced adherence was “I am not interested in exercise,” followed in order by “In the past 4 weeks my health status affected my ability to exercise regularly,” “Bad weather prevents me from exercising,” “I do not have the strength to exercise,” “It is difficult to exercise when I feel depressed,” “Fear of falling prevents me from exercising,” “I get short of breath when I exercise,” and “I feel the same whether I am active or not.” Their associated odds ratios were all less than 1, indicating that for all those items, endorsement would decrease likelihood of exercise participation.


A major responsibility of physical therapists is to prescribe, promote, and educate patients on the importance and value of exercise as it relates to optimal physical function, wellness, and quality of life.24 Ideally, if a patient was given a physical therapy HEP, he or she would adhere to the exercise program over the long term. We found that following discharge from a balance training physical therapy program, 90% of survey participants reported receiving a HEP; however, 36% of older adults were no longer participating in their prescribed HEP, and less than 10% participated 5 or more days per week.

Surprisingly, time since discharge did not affect adherence to a prescribed HEP. Adherence was not greater among older adults discharged from physical therapy within 12 months compared to those had been discharged for more than 48 months. Change in health status was the most commonly reported reason for lack of participation in a HEP. Burton et al18 reported that current state of health was a predominant factor in initiating or maintaining physical activity. This finding suggests that physical therapists need to educate patients and physicians that a return to physical therapy for modification of a HEP may be necessary following a change in health status.

An additional reason reported for nonadherence to a prescribed HEP was participation in other forms of physical activity, which may explain poor postdischarge HEP adherence. However, in this study, older adults most frequently reported performing additional physical activity less than one time per week, suggesting that participation in additional physical activity was not a major contributor to nonadherence to a prescribed HEP.

The respondents reported that aerobic activity was the least common mode of exercise in their HEP. This was an unexpected finding, considering previous research31 has shown that older adults report walking as a preferred form of activity. There are several possible explanations for this finding. First, physical therapists within these 2 balance-training programs may emphasize strength more than aerobics when prescribing a HEP. Alternately, in our study, one of the major barriers reported by respondents to maintaining participation in physical activity was weather. Strength training exercises may be less affected by adverse weather because older adults can perform strength training activities in their home.

Results from our study also suggested that barriers play a greater role in determining postdischarge exercise participation than motivators. No motivator was significantly associated with exercise participation in our analysis. Eight barriers were significantly associated with a decreased adherence to exercise following discharge from physical therapy. Barriers found to decrease adherence to exercise following discharge from physical therapy included the following: lack of interest, poor health, bad weather, depression, lack of strength, fear of falling, shortness of breath, and low outcomes expectation. These findings are consistent with those of other studies demonstrating a relationship between barriers and exercise participation.18, 20, 21

Limitations of the Study

There are several limitations associated with this study. Survey respondents were drawn from 2 physical therapy programs targeting balance training in older adults; thus, the degree to which findings could be generalized to a broader population of adults receiving physical therapy is unknown. Were both groups of subjects located in a cool climate? In addition, because we were only able to use information gathered from the returned surveys, we do not know to what extent these results are representative of the entire survey population. Descriptions of HEP were self-reported by survey respondents, and the accuracy of this information was not verified. In addition, respondents did not report the prescribed HEP as defined by their physical therapist, rather only what they were currently doing. This limited our ability to determine adherence to the prescribed HEP. We also did not inquire about the length of the entire balance training program, the time it took to complete the HEP, or additional physical therapy received since discharge, all of which could have affected adherence to the HEP.

The survey questionnaires were mailed after the winter holidays, which may have affected participation in physical activity in the 4 weeks prior to receiving the survey instruments. Missing values may reflect an individual’s perception that not answering the item was equivalent to nonendorsement of that item. We chose to use available data rather than assuming that missing data reflected nonendorsement. Worthy of mention is the fact that the question most frequently not answered was “It is difficult to exercise when I am depressed,” which may mean that depression was not a condition that applied to the person or that the respondent refused to answer because this was a sensitive question. Finally, we did not collect data on race or ethnicity; thus, the effect of these variables on exercise adherence and factors affecting adherence cannot be determined from this study.

Clinical Implications and Conclusions

This study has a number of implications for the practice of physical therapy. In this study, barriers rather than motivators were more likely to predict postdischarge exercise participation. This finding suggests that physical therapists may need to place a greater emphasis on addressing patients’ barriers when establishing postdischarge exercise programs. In addition, physical therapists need to address current as well as potential changes in health status as they relate to long-term adherence and participation in physical activity. Because of the broad and significant benefits of exercise, increasing physical activity in older adults to the level necessary to achieve health benefits needs to be a key focus in physical therapy. Further research should examine what strategies help patients overcome their barriers and increase adherence to a HEP or participation in additional physical activity.


  • All authors provided concept/idea/research design, writing, and data analysis. Ms Forkan, Ms Pumper, Ms Smyth, Ms Wirkkala, and Dr Shumway-Cook provided data collection. Dr Shumway-Cook provided project management, subjects, facilities/equipment, and institutional liaisons.

  • The University of Washington Institutional Review Board reviewed and approved this study.

  • The results of this study were presented as a platform presentation the Combined Sections Meeting of the American Physical Therapy Association; February 23–27, 2005; New Orleans, La.

  • * SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

  • Received May 25, 2005.
  • Accepted September 26, 2005.


View Abstract