Although physical fitness is relevant for well-being and health, knowledge on the feasibility of instruments to measure physical fitness in older adults with intellectual disability (ID) is lacking. As part of the study Healthy Ageing with Intellectual Disabilities with 1,050 older clients with ID in three Dutch care services, the feasibility of 8 physical fitness tests was expressed in completion rates: box and block test, response time test, Berg balance scale, walking speed, grip strength, 30-s chair stand, 10-m incremental shuttle walking test, and the extended modified back saver sit and reach test. All tests had moderate to good feasibility in all subgroups, except for the participants with profound ID (all tests), severe ID (response time test and Berg balance scale), and wheelchair users (all tests that involve the legs). We conclude that the 8 tests are feasible to measure physical fitness in most older adults with ID.
Physical fitness is important for the health and well-being of older adults (Mazzeo & Tanaka, 2001; U.S. Department of Health and Human Services, 2008; World Health Organization, 2009, 2010). Furthermore, physical fitness is a necessary attribute to independent functioning or to prevent disability (American College of Sports Medicine, 1998; Cooper et al., 2011; van Heuvelen, Kempen, Brouwer, & de Greef, 2000).
The aging population with intellectual disability (ID) is rapidly increasing because of longer life expectancy as a result of improved health care and an increase in absolute numbers of the total population (Patja, Iivanainen, Vesala, Oksanen, & Ruoppila, 2000). Childhood mobility impairments, lifelong low physical activity levels (Hilgenkamp, Reis, van Wijck, & Evenhuis, 2012; Temple, Frey, & Stanish, 2006), as well as multiple chronic health conditions, increasing with age, may be barriers to the maintenance of fitness in this group. Indeed, lower fitness than in the general population has already been demonstrated for younger adults with ID (Carmeli, Ayalon, Barchad, Sheklow, & Reznick, 2002; Fernhall, 1993; Lahtinen, Rintala, & Malin, 2007).
Low physical fitness is preventable or reversible in older adults through physical activity and structured exercise (Chodzko-Zajko et al., 2009), which opens up possibilities to maintain or positively influence health and independence into old age. But to identify which groups would benefit most from intervention programs or treatments and to evaluate the effect of any intervention or treatment, instruments to measure physical fitness that are applicable to this group, are required. So far, research has focused mainly on younger adults or adolescents, and mostly on those with a mild or moderate level of ID (Carmeli et al., 2002; Carmeli, Barchad, Lenger, & Coleman, 2002; Fernhall, 1993; Lahtinen et al., 2007; Pitetti, Climstein, Mays, & Barrett, 1992). Like in the general population, tests used for younger adults can be difficult for older adults because of the physical limitations caused by aging. On the other hand, tests used in the general older population often rely, at least partly, on specific cognitive abilities, which cannot be assumed to be at the same level in the cognitively heterogeneous population with ID: Some instruments consist of familiar daily tasks that are not so common for older adults with intellectual disability, such as writing in the Jebsen hand function test (Jebsen, Taylor, Trieschmann, Trotter, & Howard, 1969), and some of the instruments used in older adults have complicated verbal instructions, such as the Moberg pick-up test (Moberg, 1958) or the arm curl test (American Alliance for Health, Physical Education, Recreation and Dance, 1980). It was therefore necessary to investigate how to measure physical fitness in this group.
To propose a fitness test battery for older adults with ID, a multidimensional concept of physical fitness needed to be described for the group first. Based on the descriptions of physical fitness used by the American College of Sports Medicine (2005) and Bouchard and Shephard (1994) and the notion that some fitness components are more relevant to daily functioning than others (Rikli & Jones, 2001), we proposed a combination of seven health-related and performance-related fitness components to describe physical fitness in older adults with ID: manual dexterity, response time, balance (static and dynamic), muscle strength, muscle endurance, and aerobic endurance and flexibility (Hilgenkamp, van Wijck, & Evenhuis, 2010). In an extensive literature review, all available instruments to measure physical fitness were identified, including those already used in older adults with ID or mild cognitive impairments. They were then evaluated for their functionality, validity, and reliability in other populations. Finally, expected feasibility in older adults with ID was evaluated by an expert meeting of physiotherapists, experienced in working with this group. Feasibility encompassed several aspects, among others, regarding the demands of the test to participants: inclusion criteria, level of difficulty of the instructions to the participant, level of difficulty of the execution of the task itself, and duration of tests, all influencing completion rates of a test in a specific population. A thorough description of the selection process has been published elsewhere (Hilgenkamp, van Wijck, & Evenhuis, 2010). With this procedure, we assembled a battery consisting of eight tests, which proved to have sufficient test–retest reliability (intraclass correlation coefficient > 0.6) in a heterogeneous sample of 36 older people with ID (Hilgenkamp, van Wijck, & Evenhuis, 2012). Feasibility of these tests was expressed in completion rates, and sufficient feasibility was confirmed for use of these instruments in large-scale epidemiological research in the latter study as well. For use in clinical practice, however, detailed information about selective dropout is required; this would enable practitioners and researchers to estimate in advance which test would be feasible for a specific individual or group of individuals. To assess completion rates of these fitness tests in subgroups of this population, a large sample is required.
In previous research, differences in performance on physical fitness tests have been shown between men and women, age groups, different levels of ID, different levels of mobility, having Down syndrome or not, and being physically active (Bouchard, Shephard, & Stephens, 1994; Carmeli, Ayalon, et al., 2002; Carmeli, Barchad, et al., 2002; Fernhall, 1993; Jones, Rikli, & Beam, 1999; Lahtinen et al., 2007; Pitetti et al., 1992). These characteristics may also be used to evaluate completion rates, and therefore the study question of this article is: What is the completion rate of the eight physical fitness tests in a large sample of older adults with ID and in subgroups based on gender, age, level of ID, presence of Down syndrome, level of mobility, and level of physical activity?
This study was part of the large-scale Dutch cross-sectional study Healthy Ageing and Intellectual Disability (HA-ID), executed by a Dutch consort of three ID care services, Abrona at Huis ter Heide, Amarant at Tilburg, and Ipse de Bruggen at Zwammerdam, in collaboration with two university institutes, Intellectual Disability Medicine, Erasmus Medical Center at Rotterdam, and the Center for Human Movement Sciences, University Medical Center at Groningen. All 2,150 clients with intellectual disability, aged 50 years and over, of the three care providers were invited to participate, resulting in a sample of 1,050 clients, which was nearly representative for the total older client population. There was a slight overrepresentation of women and an underrepresentation of the most independent group with only ambulant support while the group of residents with intensive support and care was overrepresented in the study population. Since there is no national registry of people with ID in the Netherlands, only representativeness for the client population of care providers could be calculated. Details about design, recruitment, and representativeness of the sample have been presented elsewhere (Hilgenkamp et al., 2011). Data collection took place between February 2009 and July 2010.
Ethical approval was provided by the Medical Ethical Committee of the Erasmus Medical Center (MEC 2008-234) and by the ethical committees of the participating ID care services. Informed consent was obtained from all participants; however, unusual resistance was a reason for aborting measurements at all times.
Gender and age were collected from the administration systems of the ID care services. Professional caregivers provided information about mobility (independent, with walking aid, or wheelchair user). Level of ID was categorized by psychologists or behavioral therapists as: borderline (IQ = 70–84), mild (IQ = 50–69), moderate (IQ = 35–49), severe (IQ = 20–34), or profound (IQ < 20) based on criteria of the International Statistical Classification of Diseases and Related Health Problems (World Health Organization, 1992). The presence of Down syndrome was collected through the medical files. Level of physical activity was measured with a pedometer (NL-1000), of which detailed methods have been described elsewhere (Hilgenkamp, Reis, et al., 2012) and classified in an active group (7,500 steps/day or more) and a less active group (less than 7,500 steps/day) (Tudor-Locke, Hatano, Pangrazi, & Kang, 2008). This cutoff point is in accordance with recent research on sufficient steps per day (Tudor-Locke, Craig, Aoyagi, et al., 2011; Tudor-Locke, Craig, Brown, et al., 2011). Table 1 provides an overview of the battery of eight physical fitness tests (Hilgenkamp et al., 2010).
For the box and block test (BBT), which measures manual dexterity, a participant had to move as many colored blocks of 2.5 cm3 from one side of a wooden box to the other side in 1 min (Mathiowetz, Volland, Kashman, & Weber, 1985). The test instructor was instructed to be aware of the level of understanding of participants of the “as fast as possible” part of the instruction and whether the participant was picking up only one color or color sequences, or was placing the blocks very neatly (and slowly) at the other side of the box. If these faulty executions would not change after attempts of the test instructor to correct them, the test was not completed successfully. In the general population, reliability and validity of this test were excellent, with intraclass correlation coefficients (ICC) from 0.87 to 0.97 (Desrosiers, Bravo, Hebert, Dutil, & Mercier, 1994). In patients with stroke, multiple sclerosis, or brain injury, interrater reliability and test–retest reliability were excellent (ICC > 0.95), as was construct validity, measured with correlation coefficients between different motor scales (rho > 0.92) (Platz et al., 2005). Test–retest reliability in 36 older adults with ID was good—ICC was 0.90 for both the same-day interval and 2-week interval—(Hilgenkamp, van Wijck, et al., 2012).
In the response time test, participants were asked to respond as quickly as possible to an auditive (doorbell) or visual stimulus (white dot on the screen) by pressing a button on a laptop's keyboard. Each condition—auditive signal (RTA) first, then visual signal (RTV)—was repeated 15 times, with random presentation of the signal (between 1.0 and 9.0 s). Time was recorded between the presentation of the stimulus and pushing the button, and the median of 15 scores was used as the participant's result for that condition. If the test instructor was not convinced the participant understood the action–reaction part of the instruction and the “as fast as possible” part of the instruction, the test was not completed successfully. Reliability and validity in the general older population were good, test–retest reliability, r = 0.63, p < .01 (Deary & Der, 2005; Lord, Clark, & Webster, 1991). Test–retest reliability in older adults with ID was good: visual, ICC was 0.75 (same-day interval) and 0.72 (2-week interval); auditive, 0.87 (same-day interval) and 0.74 (2-week interval) (Hilgenkamp, van Wijck, et al., 2012).
The Berg balance scale (BBS) consists of 14 balance tasks with varying difficulty, ranging from unsupported sitting in a chair to tandem stance and standing on one leg (Berg, Wood-Dauphinee, Williams, & Maki, 1992). For this article, a successful completion of the BBS meant that a participant was able to execute all 14 items of the test. The freely available original test instructions were followed, but some aids were used to enhance understanding of the tasks, such as two carpet feet and a carpet circle on the floor, to point out where the participant had to stand or turn around. Use of walking aids was not allowed. Validity and reliability in the general older population have been demonstrated previously: test–retest reliability of ICC > 0.90 (Steffen & Seney, 2008) and 0.97 (Conradsson et al., 2007), and construct validity: moderate correlation with the timed up and go and usual gait speed (Spearman's rho = −0.53 and 0.46, respectively; Wang et al., 2006), as well as in the population with ID: substantial to almost perfect interrater agreement (kappa = 0.74–1.00) and low to almost perfect test–retest agreement (kappa = 0.37–1.00; Sackley et al., 2005), and in the older population with ID: test–retest reliability of ICC = 0.96 (de Jonge, Tonino, & Hobbelen, 2010).
Walking speed was measured three times for comfortable (WSC) and fast (WSF) speed over a distance of 5 m (after 3 m for acceleration). The three attempts of WSC were averaged to get the participant's result; for the WSF, the best score out of three attempts was used as the participant's result. The participants had to walk the distance without someone walking alongside or physically supporting them to avoid influencing the comfortable speed and the balance of the participant. Walking aids were allowed but recorded by the test instructor. Reliability and validity in the general population were good (test–retest reliability, ICC > 0.90; Steffen, Hacker, & Mollinger, 2002), construct validity: risk ratio for falls: 1.4, risk ratio (Abellan van Kan et al., 2009). Test–retest reliability in older adults with ID was good: comfortable: ICC was 0.96 for same-day interval and 0.93 for 2-week interval, and fast: 0.96 for same-day interval and 0.90 for 2-week interval (Hilgenkamp, van Wijck, et al., 2012).
Muscle endurance was measured with the 30-s chair stand test (30sCS; Rikli & Jones, 2001). The participant was asked to sit down and stand upright as often as possible for 30 s without using his or her hands. Test instructors did not record the result if participants could not stand up from a chair without supporting themselves with their arms on arm rests, knees, or walking aid (after attempts of the test instructor to motivate participants to do so). In the general older population, test–retest reliability was good (ICCs of .84 for men and .92 for women), and criterion-related validity as a measure of lower body strength was confirmed (r = .78 for men and .71 for women; Jones et al., 1999). Test–retest reliability in older adults with ID was moderate (ICC was 0.72 for same-day interval and 0.65 for 2-week interval; Hilgenkamp, van Wijck, et al., 2012).
Grip strength (GS; Mathiowetz, Kashman, et al., 1985) was measured to determine muscle strength. The participant squeezed a Jamar Hand Dynamometer (#5030J1, Sammons Preston Rolyan, USA) to his or her maximum ability in a seated position, according to the recommendations of the American Society of Hand Therapists (Fess & Moran, 1981). The best result of three attempts for both the left and the right hand (with a 1-min pause between attempts) was recorded. A visual example was provided by squeezing a rubber ball by the test instructor. The test instructor had to be convinced the participant squeezed with maximal effort; otherwise he did not record the result. Reliability and validity in the general population were good (test–retest reliability: ICC = 0.99, 95% CI [0.98–0.99]); construct validity with correlation coefficients (p < .01): Barthel Index = 0.197, Lawton IADL = −0.451, Timed-Up-and-Go = −0.330, Tinetti's POMA = 0.239, and 6-min walk test = 0.316 (Abizanda et al., 2012; Stark, Walker, Phillips, Fejer, & Beck, 2011). Test–retest reliability in older adults with ID was good (ICC was 0.94 for same-day interval and 0.90 for 2-week interval; Hilgenkamp, van Wijck, et al., 2012).
For the 10-m incremental shuttle walking test (ISWT; Singh, Morgan, Scott, Walters, & Hardman, 1992), the participant starts walking a 10-m section at .50 m/s with the test instructor. Every minute, the test instructor increased walking speed by .17 m/s; the participant continued walking until he or she could no longer keep up with the pace. The number of completed minutes in which the participant is able to walk at the right speed was recorded, which resulted in a number of meters walked in this test, which can be used to calculate maximal oxygen uptake (VO2max; Singh, Morgan, Hardman, Rowe, & Bardsley, 1994). The test was stopped if the participant was too breathless to maintain the required speed, failed to complete a 10-m shuttle in the time allowed, or reached 85% of the predicted maximal heart rate (Singh et al., 1992). The participants wore heart-rate monitors to assess their effort during the tests and to check whether this was sufficient for a valid test result. Validity of this test was confirmed (comparison with VO2max during a treadmill walking test: correlation coefficient: r = 0.88; Singh et al., 1994). Test–retest reliability has been investigated in 10 patients with chronic airway obstruction and who had Pearson correlations of 0.98 and higher, but a significant mean difference was observed between the first and the second session, suggesting that a practice session is necessary to obtain valid results (Singh et al., 1992). In 353 patients attending cardiac rehabilitation, the ICC was 0.94, with, again, a significant mean difference between the first and second session, confirming the need for a practice walk (Jolly, Taylor, Lip, Singh, & Committee, 2008). The pilot study could not confirm this need definitively in this group, partly due to small numbers. To avoid the risk of underestimating cardiorespiratory endurance, the ISWT was executed twice for every participant. Test–retest reliability in older adults with ID was good: ICC = 0.90 (same-day interval) and 0.76 (2-week interval) (Hilgenkamp, van Wijck, et al., 2012).
Flexibility was measured with an extended version of the modified back saver sit and reach test (EMBSSR; Hilgenkamp et al., 2010; Hui & Yuen, 2000), in which the participant sits on a chair, stretches one leg on a second chair, and bends forward. The distance from the participant's distal point of the phalanx distalis of the digitus medius to the malleolus lateralis was measured while he or she was bending over. This test was executed for both legs. Hui and Yuen measured from finger to heel, but because of high prevalences of deviations in foot position in people with ID, the ankle was considered to be a more stable, although a more proximal, anatomical point across this population than the heel. For comparison reasons with the measurement of Hui and Yuen from finger to heel, a standard correction of subtracting 6 cm will be applied to the measured results. Test–retest reliability and validity were good in the general population: test–retest reliability: ICCs were 0.96 (men) and 0.97 (women); construct validity: comparison low-back and hamstring criterion measure for men (r = 0.47–0.67) and women (r = 0.23–0.54) (Hui & Yuen, 2000); and good to moderate in older adults with ID: left leg: ICC was 0.96 (same-day interval) and 0.63 (2-week interval); right leg: 0.95 (same-day interval) and 0.71 (2-week interval) (Hilgenkamp, van Wijck, et al., 2012).
The Revised Physical Activity Readiness Questionnaire (rPARQ) was administered by the professional caregivers in advance of participation in the physical fitness tests, to determine if these tests could be performed safely by the participant (Cardinal, Esters, & Cardinal, 1996; Thomas, Reading, & Shephard, 1992). This instrument consists of seven questions regarding the presence of health conditions or complaints that can be answered with yes or no. If any of the questions was answered with yes or left blank, the general practitioner or intellectual disability physician was consulted to determine whether exercise was safe for this particular participant, considering his/her medical history. If not, the tests involving physical effort were excluded from the assessment: fast walking speed, 30-s chair stand, and the 10-m incremental shuttle walking test.
The physical fitness assessments were executed on locations familiar or close to participants: a large room within their home, a familiar day care center, or a gym. Participants conducted the box and block test and the response time test in a separate room from the other tests to avoid visual distraction or distraction by noise.
All assessments with the physical fitness instruments were guided by test instructors, who all were physiotherapists, occupational therapists, or physical activity instructors with experience in working with people with ID. They all followed a 2-day course for the execution of the eight tests and received an instruction manual specifically designed for this study, with specific emphasis on scoring either a successful measurement with a result, or an unsuccessful measurement, without any results, to avoid ambiguity about the validity of the results.
To avoid undesirable influences of consecutive tests, both 10-m incremental shuttle walking tests (ISWT) were executed on the same day at the beginning and at the end of the assessment, with at least an hour in between. Furthermore, a recovery period of at least 5 min was included after the execution of the ISWT. Tests that were more cognitively challenging (box and block test, response time tests, Berg balance scale) were executed in the first part of the test session after the first ISWT to avoid the influence of fatigue. Flexibility was measured directly after the last ISWT, to use the exercise of this test as a warm-up to measure flexibility (Williford, East, Smith, & Burry, 1986). In all sessions one or more breaks were allowed if the participants showed signs of exhaustion or distraction. No formal time limit was placed on the breaks, but they took no longer than half an hour.
Test instructions developed for people with normal intellectual capabilities give standardized descriptions on how to motivate the participant and how much encouragement is allowed during a test (i.e., “The tester should encourage participants a few times by saying ‘you're doing well’ and ‘keep up the good work’”; instructions for the 6-min walk test, Rikli & Jones, 2001). In the above-mentioned expert meeting of physiotherapists with years of experience of working with people with ID, consensus was reached that such standardized encouragement is ineffective in people with intellectual disability because of the large variation in behavior and responses. To keep the motivational aspect as equal as possible, we prescribed “maximal motivation” for all tests to the test instructors. In some cases, this meant that participants were motivated to engage in the assessments by constant verbal encouragement and verbal rewarding, in other cases the test instructor had to remain very calm and quiet to motivate the client as much as possible and to prevent stress or anxiety. The specific background, knowledge, and experience of the test instructors were important conditions to ensure the most suitable maximal motivation for every participant while considering safety as well.
Fine adjustments have been made to the tests, such as squeezing a stress ball as an example for the grip strength test and using carpet feet to indicate the position of the feet in the Berg balance scale. The test instructors found these adjustments helpful when giving instructions to the participant, as well as in the execution itself. They have applied these adjustments for a large part of the study population.
For each physical fitness test, numbers of participants are provided for the total sample and for all categories of baseline characteristics (gender, 10-year age-groups, level of ID, presence of Down syndrome, mobility impairment, and level of physical activity [PA]). Because only three participants fell into the age category of 90 years and over, they were merged with the preceding category (80-plus years).
Completion rates were divided in quartiles, and characterized as low ( ≤ 25%), moderate ( > 25% and ≤ 50%), good ( > 50% and ≤ 75%), and excellent ( > 75%).
Descriptive characteristics of the study population are presented in the third column of Table 2. For level of physical activity, only the most functionally able part of the participants could be measured successfully with the pedometer (257 successful out of 1,050 participants). Of the 257 participants with a successful measurement with the pedometer, three did not participate in the physical fitness assessment because of behavioral problems or noncooperation and were left out in further analyses.
Numbers and percentages of participants successfully performing each physical fitness test are presented in Table 2, specified for gender, age category, level of ID, presence of Down syndrome, level of mobility, and level of physical activity.
The chance of completing a test for an individual can be derived by looking at the lowest completion rate of the separate categories involved, for example, a participant with a moderate level of ID in a wheelchair has a 38% chance of being able to complete the test successfully.
In Table 3 a summary of the results is given, with the labels low feasibility (≤ 25%), moderate feasibility (> 25% and ≤ 50%), good feasibility (> 50% and ≤ 75%), and excellent feasibility (> 75%).
This large-sample study provides new and relevant information for clinical practice about the feasibility of eight physical fitness tests in older adults with ID. All eight tests had moderate to good feasibility in the total sample of 1,050 older adults with ID. The Berg balance scale and the 30-s chair stand had the lowest completion rates across all subgroups, with good feasibility only among those with borderline, mild, and moderate ID and the independent walkers. The box and block test, comfortable walking speed, and grip strength had the highest completion rates for the total group as well as for the subgroups.
Subgroups that had trouble with some or all tests were participants with profound ID (low feasibility on all tests), with severe ID (low feasibility on the response time tests and the Berg balance scale), and wheelchair users (low feasibility on all tests that involved the legs). Other studies have also identified the need for specific instruments to measure physical fitness in adults with severe and profound intellectual and/or motor disabilities in some studies combined with visual impairments (Gagnon, Decary, & Charbonneau, 2011; Waninge, Evenhuis, van Wijck, & Van der Schans, 2011; Waninge, van Wijck, Steenbergen, & van der Schans, 2011). On the other hand, the oldest age groups, participants with Down syndrome, and participants with low levels of physical activity all had moderate to excellent completion rates for all tests.
In the epidemiological study of older people with ID (HA-ID) that this study was a part of, this dropout in the physical fitness tests will lead to an underrepresentation of participants with severe or profound ID and people in a wheelchair and an overrepresentation of younger, independently walking older adults with borderline to mild ID.
Because of a lack of large-scale research into the health of adults with ID with objective measures (Hilgenkamp et al., 2011), it is hard to compare this study's results with previous research in similar populations. Measures used in health screenings in the general population often are not applicable to the population of adults with ID because self-report questionnaires, neuropsychological tests, and physical tests often require a level of cognitive and physical capabilities that is too demanding for adults with ID (Eastwood, Nobbs, Lindsay, & McDowell, 1992; Fried et al., 1991). Most published epidemiological research in adults with ID is based on existing medical records or observations of professional caregivers (Cooper, 1998; Minihan & Dean, 1990; Perry et al., 2010; van Schrojenstein Lantman-de Valk et al., 1997). In the small number of previous studies with objective measures in younger adults with ID, dropout or nonresponders are not always described or studied (Baynard, Pitetti, Guerra, Unnithan, & Fernhall, 2008; Fernhall et al., 1996; Skowronski, Horvat, Nocera, Roswal, & Croce, 2009), or feasibility is used to describe an optimal protocol for fitness testing (Fernhall & Tymeson, 1987).
It is possible to compare these completion rates with those of feasibility studies in the general older population, and although the used tests are different, some differences stand out. Suni et al. (1998) deployed a health-related fitness assessment in 500 middle-aged men and women and found that exclusion rates (excluded due to refusing to participate or interrupting) increased with age, in contradiction with the results of the current study. For balance, hand grip, and flexibility tests, exclusion was less than 5%; for lower-extremity muscle tests, exclusion was less than 10%. Up to 27% was excluded from the muscle endurance tests in the highest age range (modified push-up test and isometric back extension muscle endurance test; Suni et al., 1998). This is comparable to other results with submaximal tests (Laukkanen, Oja, Pasanen, & Vuori, 1992).
Malmberg et al. (2002) measured exclusion rates in a sample of 1,133 middle-aged and older adults and found exclusion rates of less than 5% for the 6.1-m walk, chair stand, stair climb and descent. Less than 15% was excluded from the one-leg standing balance test, the 1-km walk, the trunk-side bending test, and the knee extension range of motion test, but for women older than 70 years, the exclusion rates increased for the one-leg standing balance (25%) and the 1-km walk (35%). Men and women in the oldest age groups had high exclusion rates for the dynamic back extension test (20%–40%), and the one-leg squat had the highest exclusion rates, varying from 15% to 63% in different age groups for men and varying from 40% to 80% in different age groups for women (Malmberg et al., 2002).
This age effect is seen in our study population, too, but only in the three most strenuous tests: maximal walking speed, 30-s chair stand, and the 10-m incremental shuttle walking test, whereas in the Berg balance scale and the comfortable walking speed only the 80-plus group with ID had considerably lower participation rates. An explanation could be that the age-related decline in muscle mass and cardiorespiratory endurance (American College of Sports Medicine, 1998) is expressed most clearly in the most strenuous tests.
Comparing the results of Suni et al. (1998) and Malmberg et al. (2002) with the present study, the large absolute differences of feasibility on all tests stand out. One explanation regarding the balance tests might be the use of a multi-items test in our study (Berg balance scale) versus the use of single-item tests in both the other studies. The absolute participation rates in the general population might also suggest that the older population with ID is at best comparable to the oldest age groups of the general population, suggesting an early aging process, possibly because of an inactive lifestyle. The actual results on the physical fitness test will provide more insight in this hypothesis.
A third explanation may be offered by the heterogeneity of the population. The study population does not only vary by age, but is very diverse in level of intellectual disability and level of mobility. These characteristics showed to be of great influence on participation rates in all tests, with low feasibility for the most disabled groups, resulting in lower average participation rates.
Differences in feasibility among tests are most likely to be caused by the specific demands of the tests. The tests vary in physical and cognitive demands, thus resulting in different completion rates for each test. The influence of test instructors on differences in feasibility across the tests is considered minimal, because all test instructors performed all tests, there were no specific test instructors for specific tests. The influence of procedures, for instance by the sequence of the tests or the duration of the total fitness battery, was kept as low as possible. As described in the methods, the sequence was specifically designed to avoid influence of one test on another; sufficient time was planned for the execution of tests; and breaks in-between tests were allowed.
Both a strength and a limitation of most of the tests (excluding comfortable walking speed) is that they are performance-based, and they provide objective and more direct information about physical fitness than using questionnaires. At the same time it is a limitation because the tests assume that the participant will perform at his or her best, and qualification of results (e.g., norm values) is based on this assumption. Although the test instructors had been instructed to motivate the participants to maximally exert themselves, relying on their experience with this population, this remained difficult. This is indicative in itself of the limited experience of people with ID with physical exertion throughout their lives, causing fear of, or resistance against, unfamiliar bodily responses to exercise.
More practical limitations include the space required for both the 10-m incremental shuttle walking test (at least 14 m long) and the walking speed test (at least 14 m long), specifically for the maximal walking speed, and quiet areas with little distraction from sounds or visual stimuli that were required for the reaction time tests and the box and block test.
A limitation of this study is that the HA-ID population does not include older adults with ID who do not receive any form of registered professional support or care for people with ID. Therefore, these results are not generalizable to this group, which consists mainly of older persons with borderline or mild ID living independently or in general geriatric settings. Within the HA-ID population, a bias during the informed consent procedure led to an underrepresentation of older adults with ID that receive only ambulatory support or only go to a day care center, which is also the most independent group within the client population with ID. Since the trend in the presented results is more positive toward the more functionally able subgroups and the younger subgroups, it is to be expected that the independently living group with ID and younger adults with ID can be measured with these same instruments as well.
This study provides information regarding in which groups eight physical fitness tests can be used with a sufficient chance of completing the test and generating useful results. Outcomes of this study are of particular importance to professionals who want to measure physical fitness in older adults with ID. Various groups can benefit from the results. Clinicians can use these instruments to assess participants at the start of a treatment, to decide what kind of intervention to choose, and to evaluate the effect of interventions aimed at improving specific components of physical fitness (subject to further research on sensitivity to detect small changes). These instruments can be used in program evaluation as well, because they proved to be feasible in the majority of older adults with ID. Researchers can use the complete battery or separate tests, because no specific lab equipment is required.
The tests are useful as well for epidemiological research. Although researchers need to be aware that some groups will be excluded by the tests, the tests can measure a wide range of individuals. Use of the tests in epidemiological research will enable comparison with other populations, resulting in increasing knowledge about the physical fitness of (older) adults with ID and the relevance of this information for practice and policy.
Editor-in-Charge: Georgia Frey
Thessa I. M. Hilgenkamp (e-mail: email@example.com), Erasmus Medical Center, Department of General Practice, Intellectual Disability Medicine, P.O. Box 2040, Rotterdam, Rotterdam 3000 CA, Netherlands; Ruud van Wijck, University of Groningen, Groningen, The Netherlands; and Heleen M. Evenhuis, Erasmus Medical Center, Rotterdam, The Netherlands.