Prevention of disability by exercise among the elderly: A population-based, randomized, controlled trial

Transcription

1 Scandinavian Journal of Primary Health Care, 2006; 24: ORIGINAL ARTICLE Prevention of disability by exercise among the elderly: A population-based, randomized, controlled trial HEIKKI LUUKINEN 1, SARI LEHTOLA 1, JARI JOKELAINEN 1, RAUNI VÄÄNÄNEN-SAINIO 2, SINIKKA LOTVONEN 3 & PENTTI KOISTINEN 4 1 Department of Public Health Science and General Practice, University of Oulu, Unit of General Practice, Oulu University Hospital, Oulu, 2 Head of Home Care, City of Oulu, 3 Regional Elderly Care, Lassintalo Senior Center, Oulu, Head of Elderly Care, City of Oulu, Finland Abstract Objective. To assess the effectiveness of an intervention planned and implemented by regional geriatric care teams in order to prevent disability in an elderly population. Design. Randomized, controlled trial with intention-to-treat and subgroup analyses. Subjects and setting. All home-dwelling persons aged 85 years or older in the City of Oulu (n/555, 67%), including 486 persons with a history of recurrent falls or at least one risk factor for disability in the activities of daily living (ADL) or mobility. Interventions. Programme consisting of home exercise, walking exercise, group exercise, and self-care exercise. Main outcome measures. Change in ADL and mobility during the intervention. Results. Change in ADL performance did not differ (p/0.462) but a positive change was found in mobility performance (p/0.013) in the intervention subjects (n/243) compared with the controls (n/243). At the end of the intervention period, severe mobility restrictions existed similarly in intervention and control subjects: 48 (34%) and 46 (31%) (p/0.650) respectively. Frequency of admissions into long-term institutional care were similar in intervention and control groups: 15 (7%) and 13 (6%) (p/0.669). Impaired balance existed in fewer intervention subjects (64, 45%) than controls (89, 59%) (p/0.015). Frequency and times of home exercise activity reduced in control subjects (pb/0.001) but did not change in subjects assigned to home exercise intervention (p/ 0.05). Conclusion. A pragmatic intervention slowed down the reduction of movement performance but positive effects could not be found in subjects with severe movement and any ADL dysfunction. Positive effects were also found in balance performance. Key Words: Disability, elderly people, family practice, population-based, prevention, randomized controlled trial Functional disability is a major public health concern, and it places a great burden on healthcare services [1]. Home-based interventions are particularly profitable for older adults at risk of functional decline, because they increase availability of interventions [2]. A multifactorial home-based intervention was effective in reducing the risk of functional decline among older adults with moderate disability but not among those with severe disability [3]. A recent meta-analysis of the home-based interventions also showed that functional decline was reduced in trials using a multidimensional approach. Positive effects were related to younger ages of participants [4]. Interventions may work better among women than men [5]. The question remains Disability among the most elderly home-dwelling subjects can be prevented.. A pragmatic randomized intervention slowed down the reduction of mobility performance.. Positive effects could not be found in subjects with severe movement dysfunction and any dysfunction in activities of daily living.. Positive effects were also found in balance performance. as to whether, in very old populations, functional decline could be reversed. Correspondence: Heikki Luukinen, Department of Public Health Science and General Practice, University of Oulu, Unit of General Practice, Oulu University Hospital, Oulu, Finland. heikki.luukinen@oulu.fi (Received 1 March 2006; accepted 8 August 2006) ISSN print/issn online # 2006 Taylor & Francis DOI: /

2 200 H. Luukinen et al. We conducted a randomized trial in a homedwelling population with the aim of preventing disability. This population was very old, and the interventions consisting of home exercise, walking exercise, group exercise, and self-care exercise were planned and implemented by geriatric teams responsible for the daily care of these people. Material and methods The study protocol was approved by the scientific ethical committee of Oulu University Hospital, and written informed consent was obtained from the subject or her/his guardian before the baseline examinations. Recruitment and randomization The structure of the study population is described in Figure 1. According to the official population register on 23 August 2000 and the geriatric registers of the city of Oulu, a total of 827 persons aged 85 years or older lived at home in Oulu, of whom 555 (67%) were examined at baseline [6]. We chose the following risk factors to recruit the final study population for randomization: recurrent falling during the preceding year, frequent feelings of loneliness, poor self-rated health, depression, low cognitive status, impaired vision, impaired hearing, impaired balance, slow walking speed, and impaired Figure 1. Study progress and follow-up. ability to stand up from a chair. Each of these risk factors was associated with physical and movement disability. Altogether 486 (88%) subjects had at least one risk factor and were selected for randomization into the intervention and control groups. Randomization was done using a random numbers table. Interventions Before the start of the interventions, each participant visited her/his family physician, who in cooperation with the subject assessed the feasibility of the interventions planned by a physiotherapist and an occupational therapist. Their suggestions were based on the risk factors presented itemized on a structured form. Home exercise was prioritized over the other interventions suggested. After the assessment, 136 (63%) of the 217 participants alive at home at the beginning of the intervention period were started on novel intervention(s), of a kind not previously practised by the subject. For a total of 121 (56%) subjects, a home exercise plan with or without other interventions was approved. Correspondingly, walking exercises with or without other interventions were recommended for 56 (26%), group exercises for 46 (21%), and self-care exercises for 21(10%). Home exercise without self-care exercise was prescribed for 103 (47%), while 72 (33%) were with home exercise, but not with self-care exercise and group exercise, and 64 (29%) were with home exercise, but not with self-care exercise and walking exercise. Home exercise was the main or only intervention for 48 (22%). The home exercise interventions included exercises performed in a standing position for those who could manage that: marching on the spot, rising and standing on toes, ankle extension and flexion, hip abduction, hip extension, and transition of weight from one foot to the other. Exercises in a sitting position were suggested for those unable to exercise standing: chair stands, marching in a sitting position, knee extension, hip abduction, ankle flexion and extension, and rotation with extended knees. Exercises in a lying position were suggested if the subject was unable to exercise in a standing or sitting position. The suggested exercises were: raising the pelvis, lifting an extended lower extremity, flexion and extension of the foot without lifting it from the ground, abduction and rotation of the hip, flexion and extension of the ankles. These exercises were recommended to be done three times daily with 515 repetitions. Group exercises consisted mainly of physical exercises in small groups and rehabilitation for war veterans. The occupational therapist planned the self-care exercises.

3 Prevention of disability by exercise among the elderly 201 The control subjects were asked to visit their physicians without a written intervention form. Primary outcome measures The survey lasted from the baseline examinations (median day 4 December 2000), through the beginning of the intervention (1 November 2001) until the follow-up examinations (7 April 2003). Examined in the baseline and follow-up postal questionnaires, physical disability was characterized as a need for personal assistance or difficulty in performing at least one of the following basic activities of daily living (ADL): toileting, washing, getting in and out of bed, eating, dressing, and walking indoors (range of sum scores 6 30). The corresponding items concerning disability of movement were the ability to walk outdoors and up and down stairs (210) [7]. Secondary outcome measures In postal questionnaires, feelings of loneliness (moderately often/more seldom), self-rated health (fairly bad or worse/average or better), problems with vision (yes/no) and hearing (great difficulty/better than that) were assessed according to Leinonen et al. [7]. Depression was assessed according to the short version of the Geriatric Depression Scale [8]. In the clinical examinations undertaken by the geriatric nurses (n /40), cognitive status was assessed according to the Mini Mental State Examination test (MMSE) (300) [9], and poor cognitive status was defined as a test score of5/20. Balance was assessed as an ability to stand with the feet in a tandem position for 10 seconds. Lower extremity function was assessed by an ability to rise from a chair (5 iterations) without using one s arms (yes/ no). Walking speed was measured during a 2.4- metre walk (m/s) [10]. Prescribed medications were examined based on all data the subjects had available at home. A research nurse working at the university department enquired bi-monthly by phone about the physical exercise done by the subject during the preceding two weeks. The frequency (times) and duration (minutes) of home exercise, walking (ordinary walking during the daily activities, including shopping), and group exercise (group gymnastics) were asked about, but the types of exercise were not queried because the research nurse was blinded to the randomization. Mortality data were derived from death certificates filed at the Central Statistical Office of Finland. Data on admissions to long-term institutional care were derived from the statistics of the city of Oulu. Statistical analyses The required sample size was on the basis of data from a similar trial [10]. To show a 40% reduction in the number of subjects with disability 200 subjects per group were needed (a /0.05, power /80%). Primary outcomes were analysed according to the intention-to-treat principle, comparing changes in ADL and mobility scores from baseline until the follow-up examinations in the randomized study population. Those who died or refused during the follow-up were included in the analyses as having no change in the sum-score of functional abilities in question. Secondary outcomes were analysed among the participants who had the data needed for these analyses from intervention baseline until the end of follow-up. Student s t-tests, and corresponding non-parametric tests (Wilcoxon two-sample tests), as appropriate, were used to compare continuous data, and chi-squared statistics were used to compare categorical data. Cox regression analyses were used to analyse mortality. The results are shown summarized as hazard ratios with 95% CIs. When appropriate, continuous variables were categorized through one standard deviation from the mean towards a greater risk. Computation was carried out using commercially available software (BMDP Statistical Software Inc, Los Angeles, California) on a SUN/UNIX mainframe computer (Sun Microsystems, Inc, Palo Alto, California). Results Baseline subject characteristics The similar baseline characteristics of the subjects in the intervention and control groups are presented in Table I. Primary outcome measures Table II shows the numbers of intervention and control subjects according to changes in mobility and ADL scores during the intervention. The ADL score changed similarly in intervention and control groups but change in the mobility score was favourable in intervention subjects compared with controls. Secondary outcome measures Regardless of intervention, positive changes were found in mobility performance in intervention groups compared with controls. The strongest change was found among home exercisers without

4 202 H. Luukinen et al. Table I. Distribution of baseline characteristics in the intervention (n/243) and control (n/243) groups. Characteristic Intervention Control Age, years; mean9/sd 889/3 889/3 Number of used medications; 69/4 79/4 mean9/sd Mini Mental State Examination 239/4 239/4 test (300), pts; mean9/sd Activities of daily living (ADL) 7.89/ /2.7 sum score (630); mean9/sd Mobility sum score (210); 4.79/ /2.6 mean9/sd Severe mobility restriction; n (%) 76 (32) 70 (30) Severe restriction in activities of 49 (20) 48 (20) daily living (ADL); n (%) Male; n (%) 52 (21) 49 (20) Poor self-rated health; n (%) 47 (20) 42 (18) Trouble with vision; n (%) 50 (21) 59 (25) Trouble with hearing; n (%) 90 (38) 102 (43) Feelings of loneliness; n (%) 85 (35) 76 (31) Depression; n (%) 52 (22) 60 (25) Impaired balance with feet in 55 (23) 60 (26) tandem position; n (%) Impaired chair stand; n (%) 89 (37) 95 (40) Slow walking speed; n (%) 63 (28) 48 (22) Recurrent falling during the previous year; n (%) 66 (28) 61 (26) Severe restriction: inability to perform at least one mobility function (moving outdoors or on stairs) or ADL (toileting, eating, washing, getting in and out of bed, dressing, or moving indoors); p/0.05 for all measurements; Student s t-test (continuous) or chi-squared test (dichotomized variables). Numbers may differ because of missing data. any other intervention. All corresponding changes in the ADL scores were non-significant (Table III). Risk factor follow-up data are presented in Table IV. Numbers of subjects with severe ADL and movement disability were similar in intervention and control groups, regardless of the intervention (p/0.05 for all). Impaired balance existed in fewer intervention subjects than controls. The difference was evident among home exercisers with or without other interventions (pb/0.05 for all). Distributions of other risk factors were similar in intervention and control groups. Altogether, 187 (43%) subjects 91 (42%) of the intervention group (n /217), 57(47%) of the home exercise group (n /121), and 96 (44%) of the control group (n /220) reported in some of the phone interviews that they were actively doing home exercise. Frequency and duration of home exercise and corresponding changes during the intervention period in subjects assigned to home exercise intervention and controls are presented in Table V. Frequency and duration did not differ prior to the intervention. During the intervention period the home exercise activity reduced in the control population but did not change in the group assigned to home exercise interventions. The differences in times and duration of walking exercise and group exercise before and during the intervention were similar among those having suggestions for these exercises, home exercisers, and controls (data not shown). Admissions to long-term institutional care occurred as often in intervention as in control groups: 15 (7%) and 13 (6%) (p /0.669), respectively. Cox regression analysis showed that the mortality risk did not differ in the intervention group as compared with the control group: 0.89 ( ). Missing data The randomized intervention subjects without follow-up data (n/99) more often at baseline had low cognitive status (p B/0.01), impaired chair stand (p B/0.01), and had poorer ADL (p B/0.01) and mobility (pb/0.01) scores compared with the subjects with follow-up data (n /144). Correspondingly, the control subjects without follow-up data (n/91) more often had slow walking speed (pb/ 0.05), low cognitive status (p B/0.01), impaired balance (p B/0.05), and had poorer ADL (p B/0.01) and mobility (pb/0.01) scores compared with the subjects with these data (n /152). Table II. Numbers of intervention (n/243) and control (n/243) subjects according to changes in mobility and activities of daily living scores during the intervention. Factor Intervention n (%) Control n (%) p* Mobility score: Improvement 32 (14) 16 (7) No change (56) 124 (53) Impairment 70 (30) 93 (40) Activities of daily living score: Improvement 12 (5) 18 (8) No change (63) 152 (63) Impairment 77 (32) 70 (29) *Chi-squared statistics. 1 Subjects without follow-up data (died or refused during the intervention) included in the group.

5 Prevention of disability by exercise among the elderly 203 Table III. Changes in mobility and activities of daily living (ADL) scores in intervention and control groups from baseline until the end of the intervention. Intervention status Mobility score p* ADL score p* Targeted intervention group 0.5 (02.0) (03.0) (n/217) Home exercise, with or without 1.0 ( ) (02.0) other intervention (n/121) Home exercise, without self-care 1.0 ( ) (02.0) exercise (n/103) Home exercise, without self-care 1.0 ( ) (03.0) and group exercise (n/72) Home exercise, without self-care 0(1.02.0) (02.0) and walking exercise (n/64) Home exercise, without any other 0(1.01.5) (02.0) intervention (n/48) Control group (n/220) 1.0 (03.0) ref. 0 (02.0) ref. Numbers are median (interquartile range), negative values are favourable. * Wilcoxon two-sample test. Discussion The results of the present study showed that pragmatic interventions planned and implemented by regional geriatric care teams slowed down the reduction of mobility performance in a very old, home-dwelling population. We could not find any positive effects among the subjects with severe movement disability. Mortality and long-term institutional admissions did not, either, differ between intervention and control groups. In the highest ages interventions more likely preserve than improve physical performance [3], and they tend to work best at early and reversible stages of the disablement process [12]. Therefore, it is understandable that the present trial was not effective among those with severe functional impairments. The fact that significant improvements in the performance of activities of daily living were not found may be related to the hierarchical pattern of development of disabilities. Mobility is affected before other physical activities of daily living [13]. Lower extremity dysfunction is highly predictive of subsequent disability among the elderly [14]. Also in our study, the positive effects appeared especially strong among subjects with simple home exercise of the hip region, the thigh muscles, and the ankles. The effects of interventions on balance were similar to those found on mobility. Nelson et al. [2] also suggested that improvements in physical performance tend to occur along with those occurring in balance. In agreement with our results, a programme implemented by community nurses reduced the risk of impaired balance performance in primary healthcare practice [15]. In our trial and some others [15 17], home exercise activity was relatively scant at the population level. Older adults are often unwilling to change their lifestyle and habits [18], and the recommended physical activities should be based on habits and Table IV. Distributions of risk factors at follow-up examinations among elderly people in the intervention (n/144) and control (n/150) groups participating in a follow-up examination. Characteristics Intervention Control p Male; n (%) 31 (22) 33 (22) Severe mobility restriction; n (%) 48 (34) 46 (31) Severe restriction in activities of daily living (ADL); n (%) 38 (27) 34 (23) Poor self-rated health; n (%) 28 (20) 24 (17) Mini Mental State Examination test5/20; n (%) 44 (32) 39 (27) Trouble with vision; n (%) 53 (39) 47 (32) Trouble with hearing; n (%) 51 (37) 50 (35) Feelings of loneliness; n (%) 50 (36) 57 (39) Depression; n (%) 45 (33) 41 (28) Impaired balance with feet in tandem position; n (%) 64 (45) 89 (59) Impaired chair stand; n (%) 55 (39) 75 (50) Slow walking speed; n (%) 53 (41) 58 (43) P is from the chi-squared test. Numbers may differ because of missing data. Severe restriction: inability to perform at least one mobility function (moving outdoors or in stairs) or ADL function (toileting, eating, washing, getting in and out of bed, dressing, or moving indoors).

6 204 H. Luukinen et al. Table V. Frequency and duration of home exercise before intervention and corresponding changes during the intervention in home exercisers (n/57) and controls (n/96) expressing home exercise activity in a telephone interview. Before intervention p 1 Change p 2 Frequency, times in two weeks: Home exercise 14 (4.014) (/4.74.0) Control 14 (8.314) /1.7 (/130) B/0.001 Duration, minutes: Home exercise 6.0 (5.0 10) (/5.05.0) Control 7.3 (5.0 10) /3.3 (/7.30) B/0.001 Numbers are median (interquartile range), positive values are favourable; p is from Wilcoxon two-sample test, and denotes 1 contemporary difference between home exercisers and controls; 2 difference during intervention before intervention. skills learned earlier [19]. Patient compliance is the most common barrier to successful interventions [20]. Although the intervention was successful in stopping reduction in the frequency and duration of home exercise, the differences in relation to controls were relatively small. Obviously, changes in the types of interventions also contributed to the differences in outcomes [21]. This pragmatic intervention study has notable strengths. First, the population-based trial diminished problems related to identification of the target population. It also showed availability and compliance related to simple novel home exercise interventions among the most elderly home-dwelling subjects. Second, the study was conducted as part of daily work by the personnel responsible for the care of the study participants. This allows direct recommendation of home-based interventions of this kind in geriatric healthcare. The intervention also has limitations. The design was exceptional in that there was a delay between the baseline examinations and the start of the intervention. This was because the randomization, planning, and education took time. However, this does not affect the main results of the study, because primary outcomes were analysed according to the intentionto-treat principle. Furthermore, missing data did not distort the results derived from subgroup analyses. These experiences can be used by geriatric teams responsible for the care of subjects aged 85 years or older. We suggest the implementation of interventions, including home exercise involving the lower extremities, in order to slow down the reduction in balance performance and mobility. However, preventability of severe disability among the elderly remains a major challenge for further studies [4]. Acknowledgements This study was funded by the Ministry of Health and Social Affairs of Finland. The authors express their gratitude to the personnel of the Regional Geriatric Care of the City of Oulu for carrying out the measurements and implementing the intervention and to program designer Paavo Soini for performing the study s automated data processing. References [1] Fried TR, Bradley EH, Williams CS, Tinetti ME. Functional disability and health care expenditures for older persons. Arch Intern Med 2001;/161:/ [2] Nelson ME, Layne JE, Bernstein MJ, Nuernberger A, Castaneda C, Kaliton D, et al. The effects of multidimensional home-based exercise on functional performance in elderly people. J Gerontol Med Sci 2004;/59A:/ [3] Gill TM, Baker DI, Gottschalk M, Peduzzi PN, Allore H, Byers A. A program to prevent functional decline in physically frail, elderly persons who live at home. N Engl J Med 2002;/347:/ [4] Stuck AE, Egger M, Hammer A, Minder CE, Beck JC. Home visits to prevent nursing home admission and functional decline in elderly people: A systematic review and meta-regression analysis. JAMA 2002;/287:/ [5] Vass M, Avlund K, Kvist K, Hendriksen C, Andersen CK, Keiding N. Structured home visits to older people: Are they only of benefit for women? A randomized controlled trial. Scand J Prim Health Care 2004;/22:/ [6] Lehtola S, Koistinen P, Luukinen H. Falls and injurious falls late in home-dwelling life. Arch Gerontol Geriatr 2006;/42:/ [7] Leinonen P, Heikkinen E, Era P, Heikkinen R-L, Hirvinen H, Kauppinen M, et al. Iäkkäiden henkilöiden terveys- ja toimintakykytarkastusten toteutus perusterveydenhuollossa. Kela [English abstract: Evaluating health and functional capacity in the primary social and health care of elderly people]. Studies in social security and health 12. Helsinki, Finland: The Social Insurance Institution; [8] De Craen AJ, Heeren TJ, Gussekloo J. Accuracy of the 15- item geriatric depression scale (GDS-15) in a community sample of the oldest old. Int J Geriatr Psychiatry 2003;/18:/ 636. [9] Folstein MF, Folstein SE, McHugh PR. Mini-Mental State: A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;/12:/ [10] Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, et al. A short physical performance battery assessing lower extremity function: Association with selfreported disability and prediction of mortality and nursing home admission. J Gerontol Med Sci 1994;/49:/M8594. [11] Stuck AE, Aronow HU, Steiner A, Alessi CA, Bula CJ, Gold MN, et al. A trial of annual in-home comprehensive geriatric

7 Prevention of disability by exercise among the elderly 205 assessments for elderly people living in the community. N Engl J Med 1995;/333:/ [12] Rubenstein LZ, Stuck AE. Preventive home visits for older people: Defining criteria for success. Age Ageing 2001;/30:/ [13] Dunlop DD, Hughes SL, Manheim LM. Disability in activities of daily living: Patterns of change and a hierarchy of disability. Am J Public Health 1997;/87:/ [14] Guralnik JM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med 1995;/332:/ [15] Gardner MM, Robertson MC, McGee R, Campbell AJ. Application of falls prevention program for older people to primary health care practice. Prev Med 2002;/34:/ [16] Day L, Fildes B, Fitzharris M, Lord S. Randomized controlled trial of falls prevention among older people living in their own homes. BMJ 2002;/325:/ [17] Robertson MC, Gardner MM, Devlin N, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls, 2: Controlled trial in multiple centers. BMJ 2001;/322:/7015. [18] Devor M, Wang A, Renvall M, Feigal D, Ramsdell J. Compliance with social and safety recommendations in an outpatient comprehensive geriatric assessment program. J Gerontol Med Sci 1994;/49:/M [19] Shephard RJ. The scientific basis of exercise prescribing for the very old. J Am Geriatr Soc 1990;/38:/ [20] Fortinsky RH, Iannuzzi-Sucich M, Baker DI, Gottschalk M, King MB, Brown CJ, Tinetti ME. Fall-risk assessment and management in clinical practice: Views from health-care providers. J Am Geriatr Soc 2004;/52:/ [21] Kesäniemi YA, Danforth E, Jensen MD, Kopelman PG, Lefebvre P, Reeder BA. Doseresponse issues concerning physical activity and health: An evidence-based symposium. Med Sci Sports Exerc 2001;/33(Suppl):/S3518.

8