National physical activity recommendations for older Australians: Discussion Document

Appendix 5 - Randomised controlled trial and review articles

The National Ageing Research Institute was commissioned by The Department of Health and Ageing to review the evidence and develop physical activity recommendations for older people.

Page last updated: 01 February 2011

Table A5.1 Randomised controlled trials: residential care setting

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Hauer 2002 GermanyTo assess the feasibility, safety and efficacy of intensive, progressive physical training in rehabilitation after hip surgery.3 month intensive progressive resistance and functional physical training 6-8 weeks after hip surgery.28 (15 intervention, 13 control) elderly patients with history of injurious falls in acute care or inpatient rehab due to fall-related hip fracture or elective hip replacement.RCTBaseline, 3 months (end of training) and 6 months
  • muscle strength
  • functional performance
  • training events
  • training adherence
  • physical activity
  • emotional status
Training increased strength, functional motor performance and balance and reduced fall related behavioral and emotional problems. Training improvements lost after stopping training.Patients need at least partly preserved functional ability to take part in training. No mention of patient allocation, including whether blind.
Becker et al. 2003 GermanyTo determine the effectiveness of a multifaceted in effectiveness of the complete multi-faceted intervention on falls in a nursing home residents.Multifactorial intervention involving staff and resident education, advice on environmental adaptations, exercise, and hip protectors.981 residents from 6 long term care facilities (nursing homes) aged 60 and older (mean age 85 years).RCTFalls rate, Fracture ratesThe intervention group demonstrated a significantly fallers (p=0.038) and frequent fallers (p=0.015) than the control group. There were no significant difference found between groups for rate of hip fractures.
Gillies et al. 1999 UKTo investigate the effect of a 12 week exercise program on the functional status of institutionalised elderly people.Circuit of functional exercises twice/week for 12 weeks. Compared to reminiscence plus seated range of movement exercises for trunk and upper limbs.Residents from 2 nursing homes (n= 15, mean age 88 years).Cluster, RCTDistance walked in 15 seconds, Timed up and go, Timed stair ascent and descentSignificant improvements on all outcomes for intervention group. Controls improved stair ascent and single chair rise. Significant group difference in walking.Small sample size: underpowered given attrition rate. Good adherence (92%). No exercise related injuries reported.
Top of page

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Jensen et al. 2002 SwedenTo determine whether a multifactorial intervention program would reduce falls and fall-related injuries.An 11 week multidisciplinary multifactorial program including staff education, environmental modification, exercise, supply or repair of aids, medication review, hip protectors, post-fall problem-solving conference and staff guidance. Compared with usual care control group.439 residents from 9 residential care facilities, aged 65 and over.Cluster, RCTFalls rate, Time to first fall, FracturesThe intervention group had significantly fewer fallers and the time to first fall was longer than the control group. When adjusted for baseline factors the intervention group had significantly fewer multiple fallers and falls per 1000 person days. Significant difference between the intervention and control groups for having a femoral fracture (adjusted OR=0.23, 95% CI=0.06-0.94).
Lazowski et al. 1999 CanadaTo improve strength, balance, mobility, flexibility and function using a tailored exercise programWalking, strengthening and balance exercises 45 minutes 3 times/week for 4 months. Compared with seated exercises to improve range of movement.Residents of a long term care institution (n= 68, mean age 80 years).RCTFunctional independence (FIM), Timed up and go, Stair climbing, Upper and lower body flexibility, Upper and lower limb strength, Gait speed, BalanceIntervention group had significant improvements in mobility, balance, flexibility, knee and hip strength. Range of motion group had improved shoulder strength.Interventions required minimal training of staff and volunteers.
Morris et al. 1999 USTo evaluate how weight training or nurse-based rehabilitative care programs in nursing homes impact on residence performance of activities of daily living and objective tests of physical performancePRT (2 sets of 8 repetitions) 3 times per week; walking on alternate days. Compared with tailored nursing rehabilitation intervention designed to maintain function/prevent decline. Control group received usual care.Nursing home residents (n = 468, mean age 84.7 years, 79% female) from six nursing homes.Cluster, RCTFunction: ADL and Minimum Data Set assessment summary Strength: Timed sit to stand Endurance: 6 minute walk Balance: Time standing in five positions Mood: Geriatric Depression ScaleIntervention groups had significantly reduced declines in function, in particular early and late loss ADL and locomotionAble to compare exercise types. No evidence of cluster analysis.
McMurdo and Rennie 1993 ScotlandUpper and lower limb strengthening exercises 45minutes twice/week. Compared to same frequency program of music and reminiscence therapy designed to promote social interaction. 7 months.Residents of 4 nursing homes (n = 49, mean age 80 years).Cluster, RCTBarthel ADL Knee flexion/extension, Spinal flexion, Sit to stand time, Grip strength, Balance, Life Satisfaction Index, Geriatric Depression Scale, MiniMentalSignificant improvements in exercise group for grip strength, spinal flexion, sit to stand ADL and depressive status.Average attendance was 91% for exercise program and 68% for the reminiscence sessions. Eight lost to follow up due to death (5 in exercise group) and 3 of control group due to lack of interest.
Meuleman et al. 2000 USTo assess the impact of exercise training on strength, endurance and functional status in debilitated older peopleResistance training 3 times/week, with twice weekly cycle ergometry for 4-8 weeks. Compared to usual care.Residents of 2 nursing homes (n= 78, mean age 75 years).RCTStrength in dominant arm and leg, Mobility (self selected walk speed) Cardiovascular fitness (timed endurance test) IADL, ADL, Geriatric Depression ScaleTrends for greater improvement in strength and function in intervention group. Those with least function at baseline showed the greatest improvements.Only one person withdrew due to injury.
Top of page

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Mulrow et al. 1994 USTo assess whether a tailored physical therapy program improved physical function and self perceived healthOne on one tailored program 3 times/week for 4 months. Compared to visit to read to resident 3 times/week for 4 months.Residents of 9 nursing homes (n = 194, mean age 80 years).RCTADL, Physical Disability Index (PDI) Range of movement, Mobility, Upper and lower limb strength, Sickness Impact Profile (SIP) Geriatric Depression Scale, Falls, MiniMentalNo significant group differences on PDI, ADL or SIP. Intervention group used assistive devices less. No difference in falls rates (79 vs 60).Program facilitated by an English or Spanish physiotherapist, as appropriate. Good adherence (89% and 92% sessions for intervention and controls respectively). Cost $1220 on average, compared with control $189.
Rubenstein et al. 2000 USTo assess the effectiveness of an exercise program in reducing falls in at risk older menUsual care vs. 12 week low to moderate intensity program including strength, endurance and balance exercisesAmbulatory men with at least one falls risk factor (n= 59)RCTFalls rate Endurance Gait Balance Strength (hip, ankle) Physical functionIntervention group had significantly lower three month fall rate (adjusted for activity level)- 6 falls/1000 hours activity vs 16.2 falls/1000 hours of activity Intervention group had significant improvements on gait and endurance measures.
Sauvage et al. 1992 USTo assess the impact of a resistance and aerobic exercise program on mobility, strength and endurance in frail older menLower limb PRT and cycle ergometry 3 times/week for 12 weeks. Compared with usual careResidents from one nursing home (n = 14)RCTTinetti mobility test, Gait, Strength testing, Balance, Maximum oxygen consumptionIntervention group had significantly improved mobility scores, endurance, left stride length and gait velocity.Small sample size. Clinical significance of results to be questioned as only 5-10% above baseline
Simons 2006 USTo assess the effects of two types of exercise on multiple measures of functional fitnessWalking (self paced, progressive pace and duration) or PRT (One set 10 reps at 75% 1 RM for 6 exercises) versus control. Twice/week for 16 weeks.64 older people (45 women, 19 men) from supported residential care. Aged 66-96 mean 83.5 years.RCTUpper and lower body strength Flexibility Agility BalancePre-post within exercise group improvements. Walking and resistance training significantly improved functional health compared to controls, particularly upper body strength (20% more improved than controls).Walking achieved similar improvements to PRT when compared to controls (no between group analyses). No follow-up. Attrition 10%. Participants encouraged to attend healthy ageing classes? intervention?
Shaw 2003 UKTo assess the effectiveness of a multifactorial assessment and interventions for older people with cognitive impairment presenting at ED after a fallUsual care vs. medical, physiotherapy and occupational therapy assessment, referral to interventions based on assessment.People presenting to A&E with a fall who were not hospitalized. 80 % from residential care units.RCTOne year follow up Falls rate Falls injuriesTrend for fewer falls and falls related injuries in intervention group.Specific impact of exercise not ascertained
Resnick 2002 USTo assess the feasibility of a multifaceted walking and behaviour change program and determine its effects on self efficacy, outcome expectations, activity, health, falls and falls-related injuries6 month program. Goal to walk for 20 minutes three times a week alone or in a group. Supported by nurse who monitored progress weekly in first month, then monthly20 sedentary older women in retirement community (assisted living). WALC vs. routine care. Mean age 88 (+/- 3.7 years)RCTBaseline, 2 and 6 month post
  • program initiation
  • self efficacy
  • outcome expectations
  • exercise and free living activity
  • health status
  • falls and falls-related injuries
Intervention group had increased self-efficacy, more exercise and overall activity and showed a trend for stronger outcome expectations90% of intervention group initiated and continued to exercise regularly. 15% attrition due to illness. 2/10 used existing walking group, indoor walking Calendar cue used to measure exercise behaviour (intervention not controls) Falls self reported Small sample size-underpowered for some outcomes
Top of page
Table A5.2 Randomised controlled trials: Community setting

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Barnett et al. 2003 AustraliaTo determine whether participation in a weekly group exercise program with ancillary home exercises improves balance, muscle strength, reaction time, physical functioning, health status and prevents falls in ‘at risk’ community dwelling older people.Intervention: Weekly structured exercise group run in a community setting conducted by an accredited instructor. I hour classes over 4 terms for 1 year (37 classes) with progressive work, included provision of home program with diaries to record participation, and received information repractical falls prevention strategies. Control: Received written information about falls prevention strategies. 1 yearCommunity dwelling older people (n=163); aged > 65 years and identified as ‘at risk’ of falling.RCTFalls rate Balance Strength Reaction time Walking speed SF36 PASE Fear of FallingIntervention Group: - performed significantly better than the control group in three of six balance measures (but not in measures of strength, reaction time and walking speed or on Short-Form 36, Physical Activity Scale for the Elderly or fear of falling scales). - the rate of falls was significantly (40%) lower than the control group (IRR=0.60, 95% CI 0.36-0.99). - had non-significantly (34%) fewer injurious falls and a lower proportion of fallers than controls.
Binder et al 2002 USTo evaluate whether a multidimensional exercise training program can significantly reduce frailty in community dwelling older men and womenIntervention: group exercise 3 times per week Control: home based low intensity (flexibility) program. 9 months115 sedentary frail, community dwelling adults (mean age 83 years)RCTPhysical performance, peak oxygen uptake, ADL, IADL, functional statusIntervention group had significant improvements on physical performance, peak oxygen uptake, functional status compared to controlsIntensive pre-screening produced a selective sample. Unclear whether findings transferable into routine practice. Participants given free transport to training.
Blumenthal 1999 US Babyak 2000 USTo explore the adjuvant effects of aerobic exercise in older depressed people16 wks aerobic exercise. Compared with antidepressants and aerobic exercise or antidepressants alone.Older people (n =156, with 133 at follow up, aged 50 years or over) with major depression.RCTDepressionExercise equivalent efficacy to medications. Treatment adherence better among exercisers.Short term follow up. No control group.
Day et al 2002 AustraliaTo determine the effectiveness of three interventions to prevent falls among older people and explore the interactions between them.Interventions: Exercise, home hazard modifications, vision improvement (combined into 8 groups defined by the presence/absenc e of each intervention). 18 months follow-up.1090 people aged 70 and over living at home in the City of Whitehorse. Most were Australian born, aged between 70-84 and rated their health as good.RCTFalls rateThe strongest effect that was observed was for all three interventions combined (rate ratio 0.67 (0.51 to 0.88, P=0.004). Significant benefit for exercise alone (rate ratio for exercise was 0.82 (95% confidence interval 0.70 to 0.97, P=0.02), and a significant effect (P<0.05) for the combinations of interventions that involved exercise. Home hazard management and vision improvement showed no significant effect.
Top of page

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Dubbert 2002 USSTEPS program. All received baseline brief advice from nurse about increasing walking (aim: 3x20 minutes per week). Two intervention arms received follow-up support a) 20 nurse-initiated calls or b) 10 nurse-initiated calls plus 10 motivational calls using automated phone system.Primary care patients (n=212, 60-80 years old, mostly male), independent livingRCTWalking: activity diaries, interviews, 7-day PAR questionnaire, endurance , mobility, SF36, falls, injuries, healthcare utilisation.All groups improved walking and fitness from baseline. Nurse plus automated calls walked significantly more frequently than those with no phone contacts. No change in quality of life.85% (181/212) of randomised patients completed 12 month follow-up. No evidence of increased adverse events
Engel & Lindner 2006 AustraliaTo explore whether using a pedometer would increase time spent walking in older people with type 2 diabetesInterventions: Coaching (education, goal setting, motivational and support strategies) without or without pedometer.57 older (mean age 62 years) overweight people with type 2 diabetes.RCTTime spent walking, HbA1c, blood pressure, 10 m shuttle, BMITime spent walking was greater in the coaching only group. There were improvements in weight, waist circumference and cardiovascular fitness for both groups.Confirms physiological benefits of physical activity. Small sample.
Kriska et al. 1986 USTo explore compliance with a brisk walking program designed to reduce bone lossSupervised brisk walking, 3 miles 3 times/week. Compared to usual care group.229 women aged 50-65 yearsRCTPaffenbarger physical activity survey, Caltrac monitorIntervention group significantly more active at 1 and 2 year follow up. Compliers tended to be more active, non-smokers, lower weight at baseline. Compliers had significantly less illness.Cause and effect of health status not clear. No analysis of unique impact of prompts and incentives used.
Duncan et al. 1998 USTo develop a home based balance, strength and endurance program, evaluate the ability to recruit and retain stroke subjects and to assess the effects of the interventions used8 week, 3 times per week supervised home based balance, strength and endurance program.Minimally and moderately impaired stroke patients 30-90 days post strokeRCTUpper and lower extremity motor function, 6 minute walk, Berg balance score, Jebsen test of hand function, Barthel Index, Lawton IADL scaleImprovements in neurological impairment and lower extremity function. Found that many didn’t have sufficient space in their home to do walking component so introduced cycle ergometers- feasibility implicationsSmall study. Underpowered.
Li et al. 2005 USTo determine the effectiveness of 24 form Yang style tai chi in reducing falls among older people.Intervention: 24 form Yang style tai chi, 3 times weekly for 6 months. Control group undertook a stretching /relaxation program 3 times weekly for 6 months256 physically inactive community dwelling adults (aged > 70 years), 70% female.RCTFalls, multipleTwo groups comparable at baseline. At six months, significantly fewer falls occurred in the tai chi group (38 vs 73), with greatest effect achieved after 3 months of training. Also significantly fewer injurious falls in the tai chi group (7 vs 17). Significantly longer time to first fall in tai chi group. Significant intervention effect for the tai chi group on all dynamic balance measures, and significantly lower fear of falling. Benefits appeared sustained six months after end of formal program.Median compliance of both programs was 61 of a total of 78 sessions (78%). Prospective recording of falls using falls calendars. Limitations included some assessors being aware of some participant’s intervention status, as well as the high proportion of people screened not being eligible for the project (62%).

Top of page

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Lord et al. 2003 AustraliaTo determine whether falls could be prevented in frail older people living in retirement villages using group exercise designed to improve activities of daily living.Intervention: group exercise (most weight-bearing exercises), consisting of one hour/twice weekly classes over a 12 month period. Control: Two groups, one attending a one hour/twice weekly flexibility and relaxation classes and one not taking part in any group activity.551 frail older people living in retirement villages (self-care and intermediate-care) aged 62-95.RCTFalls rate Reaction time Walking test Knee extension strength Standing balanceAt six month retest the exercise group performed significantly better in simple and choice reaction time and 6 minute walk tests, though not in knee extension strength and standing balance. At one year follow-up, adjusting for age and gender, there were 22% fewer falls in the exercise group than the combined control groups and 31% fewer falls in the 173 subjects with a history of previous falls.
Mather 2002 UKTo assess the impact of exercise on depression in older people not responding to medication.Group exercise 2/week for 10 weeks; chiefly weight bearing exercise. Compared with health information sessions86 people, median age 63 and 65 years respectively for exercise and control groupRCTDepressionGreater proportion (55% vs. 33%) less depressed post intervention in exercise group (OR=2.51, P=0.05, 95% CI 1.00-6.38), defined as a reduction in Hamilton score from baseline of 30% or more. Difference persisted to 24 week follow upGeneralisable to those poorly responsive to antidepressants. All participants were also on antidepressants. 100% compliance with programs. No adverse events.
McLafferty et al. 2004 USTo explore the effects of resistance training on measures of mood in healthy older men and women.24 weeks of PRT. 3 times/week at high intensity (80% 1RM) or variable intensity (50-80% 1RM).28 healthy sedentary older people ( mean age 67 years).Pre-postProfile of Mood StatesBoth groups had significant improvements on confusion, tension, anger and total mood scores.Study underpowered. Sample healthy, not clinically depressed.
Manson et al. 2002 USTo compare the roles of walking and vigorous exercise in the prevention of coronary and cardiovascular events in a large ethnically diverse cohort of postmenopausal women.6 year follow up of physical activity and coronary and cardiovascular events in women free from CHD and cancer at baselineWomens Health Initiative study participants aged 50-79 yearsProspective CohortWalking Vigorous activity Coronary events CVD risk factorsSignificant inverse association between physical activity and risk of a) coronary disease and b) cardiovascular events, with trends across quintiles of METS, walking and vigorous activity respectively. Brisk walking and vigorous activity reduce the incidence of coronary disease by up to 30% across ethnic groups of differing ages and body mass.Ethically diverse cohort but analyses simplified to white vs. black.
McNeil 1991 USTo assess interventions to improve depressive status in older peopleAccompanied walking. Compared with social visits and usual care control group.Older people with moderate levels of depressive symptoms (n = 30)RCTDepressionSocial visits and walking reduced depressive symptoms. Walking also reduced somatic symptoms.Small sample.
Penninx 2002 USTo evaluate the benefits of aerobic exercise in older people with knee osteoarthritis.Group based aerobic exercise. Compared with health education.Older people with knee osteoarthritis (n = 439, 60 years or over). Not clinically depressed.RCTDepressionAerobic exercise reduced depressive symptoms in both high and low level depression. No effect for resistance training.18 month follow up.
Robertson et al. 2002 New ZealandTo evaluate the effect of a home-based exercise program and identify subgroups most likely to benefit.Individually prescribed muscle strengthening, balance retraining program.Community dwelling adultsMeta-analysis of RCTsFalls rate Falls injuriesFalls and falls related injuries significantly lower in exercise group. Those over 80 years most likely to benefit
Top of page

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Singh et al. 1997, 2001 USATo evaluate the long term impact of a progressive resistance training program in older people with clinical depressionHigh intensity progressive resistance training 3 times/week for 10 weeks supervised then 10 weeks unsupervised. Compared with health education lectures.Community dwelling older people with clinical depression (n = 32, mean age 71.3 years).RCTDepression73% of intervention group no longer depressed at 20 weeks, compared to 36% of control group. The depression score was significantly improved at both 20 weeks and 26 months follow-up. Improved self efficacy and morale.Contacted weekly for 20 weeks. Choice of continuing exercise venue. Shows benefits of ongoing unsupervised activity. Small sample.
Song et al. 2003 South KoreaTo evaluate the effectives of a special version of tai chi (Tai chi for Arthritis) in improving function and reducing symptoms for people with arthritis.The intervention consisted of a 12 form version of Sun style tai chi once weekly for 12 weeks.72 older women with osteo-arthritis.RCTPhysical symptoms, body mass index, cardiovascular functioning, self perceived difficulties in physical function.The tai chi group achieved significantly less pain and stiffness in their joints, reported fewer difficulties in physical functioning, relative to the control group who either had no change or deterioration in these measures. Program was well tolerated by patients with osteo-arthritis.Moderate (41%) drop-out rate. Analyses limited to univariate analyses. Functional improvement based on self report rather than observation.
Toshima et al 1990 USATo evaluate a comprehensive rehabilitation programIntervention: 12 four hour sessions over 8 weeks. Education, psychosocial support and individualised walking program. Controls: Four bi-weekly meetings education only. 6 month follow up.119 COPD patients (32 female, mean age 62.6 years)RCTExercise endurance, quality of wellbeing, depression and self efficacyIncreases in exercise endurance significantly greater in intervention group. Self efficacy also tended to be greater in the intervention group.Limited attrition-89% underwent the 6 month follow-up. Patients were required to make up any sessions they missed.
Wolf et al. 1996 USATo evaluate the effect of Tai Chi and computerised balance training on measures of frailty and falls.Intervention (Tai Chi): Tai Chi Quan in group classes, with encouragement to practice at home Intervention (Balance Training): Individual sessions using computerised balance machine Control Group discussion (non-specific): with no change to exercise levelsCommunity - dwelling adults (n = 200), aged >70 yearsRCTFalls rate Blood pressure Fear of fallingTai Chi reduced the risk of multiple falls by 47.5% Lower Blood Pressure in Tai Chi group before and after 12 minute walk Fear of falling responses were reduced after Tai Chi
Wadell et al. 2005 SwedenInvestigate effects of decreased training frequency in patients with COPDHigh intensity training program 3 times a week for 3 months and once a week for 6 months43 (30 intervention, 13 control) COPD patientsControlled trial – no mention of randomisationBaseline, 3- month and 9 months -walking tests -cycle ergometer tests -St George’s Respiratory Questionnaire -SF-36 -spirometry -bone mineral density.Training once a week not enough to maintain improvements from initial high frequency training.Low frequency training for 6 months with a preceding 3-month period of high frequency training might prevent decline in physical capacity and health related quality of life.

Top of page

StudyAimInterventionSampleDesignOutcome measuresMain findingsStrengths/ weaknesses
Fatouros et al. 2005 GreeceTo determine the effect of intensity on strength, anaerobic power and mobility of older men24 week high or low intensity strength training, 48 week detraining52 older men (mean 71.2 +/- 4.1 years)RCTBaseline, post training and 4, 8 and 12 months follow-up upper and lower body strength anaerobic power timed up and go walking, climbing stairs.High intensity produced greater gains in strength, anaerobic power and mobility, maintained through detraining. Anaerobic power returned to baseline after 4 months in both groups.Small sample size. No adherence figures given. The sample were very motivated (attended 4 baseline preparation sessions).
Singh et al. 2005 AustraliaTo test whether: PRT is an effective antidepressant in older adults with clinical depression; high intensity PRT is superior to low intensity PRT and standard GP care8 week PRT 3 times per weekOlder community dwelling people (60) aged > 60 years with major or minor depression (DSM-IV psychiatric, assessment, GDS > = 14RCTBaseline, 8 weeks Depression score Outcome expectation Eysenck personality questionnaire Social support network Self efficacy Locus of control SF-36 Pittsburgh Sleep quality Index Muscle strength (1 RM).Significantly more of high intensity group had a 50% reduction in symptoms (61% vs. 29% low intensity and 21% GP care, p = 0.03). All reported significantly improved sleep quality. Vitality quality of life improved more in high intensity group.High intensity PRT is more effective than low intensity PRT or usual care in reducing depressive symptomatology in older depressed people.
Tessier et al. 2000 CanadaTo examine the impact of an aerobic exercise program on metabolic control, physical performance, QoL and attitudes in elderly ambulant patients with type 2 diabetesSupervised rapid walking, strength and endurance program, 3 times a week for 16 weeks39 people aged 65 or over being treated for type 2 diabetesRCTGlucose tolerance test, Quality of life Attitudes towards type 2 diabetes.Intervention group had significantly improved area under curve for glucose tolerance, extended treadmill exercising, improved attitudes towards type 2 diabetes.Exercise can be helpful for older people with type 2 diabetes where other treatments do not manage the condition.
Wilbur et al. 2003 USTo assess the effectiveness of a home based moderate intensity walking intervention in sedentary African American and Caucasian middle aged women24 week home based moderate intensity walking intervention with behaviour strategies to promote adherence153 sedentary employed women recruited via 9 worksites, newspaper and TV advertising. Mean age 49.8 years. 33% African American.One year cross over trial with controls doing intervention after 24 weeksBaseline, 24 weeks and 1 year Aerobic fitness BP, Lipoproteins Depressive status.Adherence significantly higher in Caucasians (71.5% vs. 56%, those with higher self efficacy and those with less previous exercise experience. Average 64% of expected walks during 24-48 week maintenance phase.Mainly professional women. Volunteer sample. Low attrition rate 20%.

Top of page
Table A5.3 Review articles for the promotion of physical activity to prevent or treat depression

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Brosse 2002 US Critical review To determine whether exercise is effective in treating depression in adultsNo details given for general review. 12 English language papers of RCTs where participants met diagnostic criteria for depression/reported elevated depressive symptomatology and did not require a specific comorbid medical conditionInconclusive evidence for added value of exercise in those receiving antidepressants, standard psychotherapy or psychiatric care. Overall benefit compared to no treatment or social contact alone. The longer the intervention, the greater the benefit. Methodological weaknesses limit findingsNeed for focus on older people
Paluska and Schwenk 2000 US Critical Review To review the current literature in the effects, mechanisms and potential benefits of using physical activity as a component in the treatment of depression and anxietyMedline and selected bibliographies. No dates recorded. All types of designIncreased aerobic or strength training exercise significantly reduces depressive symptoms for adults and older adults. Habitual physical activity has not been shown to prevent the onset of depression. Although the evidence base is smaller, gains similar to those from meditation and relaxation have been found for anxiety and panic disorder, with acute anxiety responding better to physical activity therapy.Better understanding of the underlying mechanisms is required.

Table A5.4 Review articles for the promotion of physical activity in older people

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Adams and White 2003 UK Critical review To discern if there is evidence of an additional effect of transtheoretical model (TTM) based activity promotion interventions over non-staged interventions.1982-2001, 26 papers. Controlled and uncontrolled trials.In general, the variety of TTM based activity promotion interventions reviewed reported short term gains in stage of change and/or activity behaviour. Positive changes were only seen in 29% of the long term ( > 6 months) studies.The findings come from studies largely employing white, middle class females. The complete range of TTM stages were not always represented nor were outcomes attained for all. A stage change without concomitant effect on activity behaviour has limited value.No specific focus on older adults.
Atienza 2001 US Integrative, commentary Summarise the research literature on home-based physical activity programs for middle-aged and older adults (ages 40+ years).1979-2000 29 studies Randomised clinical trials – some with controls.Home-based training effective for improving cardiorespiratory health, physical and overall functioning, muscle strength. Can be more effective than group-based training.Home-based programs effective and perhaps more convenient for some groups.Little information on review methodology
Boule et al. 2001 Canada Meta-analysis To systematically review the effect of exercise interventions on glycaemic control in adults with type 2 diabetes.To December 2000 14 trials (11 RCT, 3 non-RCT).Weighted mean postintervention HbA1c lower in exercise than control groups –0.66, p < 0.001) No significant difference in postintervention body mass.Exercise has benefits that are not influenced by exercise volume or intensity, nor by weight loss. Both aerobic and resistance training are effective.Results generalize to middle aged people across many ethnic backgrounds. Need more evidence for older people.
Top of page
Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Conn et al. 2003 US Integrative Effectiveness of interventions to increase physical activity or aerobic exercise in older adults.1960-2000 17 studies (42 retrieved) RCT English language.Wide range of intervention types. Few used tailored, mediated or theory driven techniques.Overall, physical activity increased significantly in response to the interventions. Amount of physical activity usually failed to reach the current guidelines Need to obtain more details on strategies to reach a wider audience, sex and ethnic differences, achieving greater increases in physical activity.Common weaknesses: Small sample sizes, time limited, non-validated outcome measures.
Dugdill et al. 2005 UK (Dugdill, Graham et al. 2005) Review of development, impact and evaluation of UK exercise referral schemes.Commentary review Participatory action research with two schemes (case/cohort design).For 342 adults tracked for 1 year at 3 monthly intervals adherence was 35-45%, with older adults more likely to complete. Physiological changes of population health rather than clinical significance. Adherence greater in those referred by cardiac and practice nurses than by GPs.Further evidence of benefits for some population groups. Need to engage younger adults. Need for evaluation designs that give better understand of the context surrounding the effectiveness of a scheme at an individual and systems level.
Eakin 2001 Integrative, commentary Review interventions designed to increase physical activity in middle-aged and older adults initiated in health care settings.1995-1999 Four studies No mention of design.No short-term increase in short-term physical activity, one intervention achieved long term change.Limited literature in the area, need for assessment of more intensive interventions.No information on review methodology. No mention of search and inclusion methods. No mention of criteria used to assess validity of studies.
Ellingson and Conn 2000 US Integrative commentary To review the empirical evidence that older people who exercise regularly have improved quality of life.7 descriptive, 11 trials.Largely positive findings: no negative findings.Limited by
  • tools used to measure quality of life
  • small sample sizes
  • lack of follow up - limited conceptual frameworks.
Need more common quality of life framework and tools to compare between studies. Too many studies used only indirect measures of QoL , or interventions where the proposed mechanism was not outlined.
Fiatrone Singh 2002 Australia US Commentary review To discuss non-pharmacological modulation of body composition through appropriate dietary intake and physical activity patterns.No strategy outlined.Discusses relationship of body composition to chronic disease, adipose tissue mass and distribution, muscle mass and quality, bone mass. Provides exercise and dietary recommendations for body composition.Body composition changes can take a while to materialize. Studies need to use behaviour change measures as the proximal outcomes.
Gregg et al 2000 USA Integrative, commentary To assess the relationship between physical activity and risk for falls and osteoporotic fractures among older adults.1975-1998 Observational (case control and prospective cohort) studies and RCTs (12 papers, 5 studies).Level III-2 and III-3 evidence for 20-30% risk reduction for hip fracture from physically active compared to sedentary people. No evidence that physical activity protective for other osteoporotic fracture types.Observational studies equivocal. Small sample sizes and underpowered studies limit findings. Recommend for future studies:
  • Greater specification of types and quantity of physical activity
  • Identify those who benefit most.
Top of page

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Hillsdon 2005 UK Cochrane review To assess the effects of interventions for promoting physical activity in adults aged 16 years and older, not living in an institution.1979-2001 17 studies All RCT.Effect of interventions on self reported physical activity and cardio-respiratory fitness was positive and moderate (SMD 0.31, 95% CI 0.12-0.51 and 0.4, 95% CI 0.09-0.70 respectively). No age differential effects (above or below 75 years) reported (1 study). Interventions incorporating professional guidance and self direction, plus ongoing support had more consistent effect estimates.Interventions can produce moderate short to mid-term increases in physical activity, at least in middle-aged adults (mean 52 years SD 12.1). Clinical and statistical heterogeneity, limited interpretation of results. Majority of studies stopped after 12 months: no clear information on long term effectiveness. No examination of effectiveness in minority groups. Need for more cost-effectiveness data, details of person delivering intervention, the underlying theory and the translation into routine practice.Studies failed to examine interaction between baseline levels of physical activity and exposure to intervention. Weaknesses in: allocation concealment, describing randomisation methods, blinding of assessors to group membership.
Holland 2002 Integrative, commentary Reviews literature on influence of physiological aging processes on connective tissue, joint integrity, flexibility and physical function of older adults.32 studies 19 cross-sectional studies, 13 trials (8 general exercise and 5 specific stretch programs).Loss of flexibility occurs between young adulthood and middle to old age, although no causal relationship established. Age related changes in musculoskeletal tissue and increased incidence of joint pathology, negatively affect joint extensibility and stiffness. Decreased flexibility involved in etiology of physical impairments.Research on improving flexibility using stretch protocols has produced encouraging results.No information on review methodology.
Houde 2002 US Integrative, commentary To review the literature on physical activity and its relationship to cardiovascular risk factors and mortality in older adults in order to clarify the specific benefits and optimal level of physical activity for cardiovascular health in the older adult population.1990-August 2000 44 research articles Trials and observational studies.No clear evidence regarding quantity or type of physical activity to recommend. Evidence that physical activity decreases mortality rates. Equivocal evidence for physical activity impact on cardiovascular risk factors.For future studies recommend: Larger sample sizes Better controlling for confounder variables, e.g. dietary factors, medication, comorbidity Use of psychometrically tested, population group appropriate, physical activity measurement instruments.
Keysor and Jette 2001 US Systematic review.1985-2000 , 2 databases 31 studies, 29 RCT, 2 quasi-experimental.Positive effects for strength, balance, flexibility, aerobic capacity and walking outcomes. Few improvements where physical, social, emotional and/or overall disability assessed.Exercise in older people improves physical function and health. The current evidence does not clearly support a reduction in disability risk arising from physical activity adoption.
Koltyn 2002 US Summarizes results from studies that have examined whether improvements in pain occur after an exercise intervention in older adults.1977-2000 17 research articles Experimental and quasi-experimental.Most investigations show exercise-training programs appear to improve pain in community-dwelling older adults.Further research needed to know optimal conditions for pain reduction, and underlying mechanisms of the pain-relieving effects of exercise. More research needed on older adults in care.
Linnan 2001 US Reviews the present status of worksite-based physical activity programs.Experimental, quasi-experimental and observational.Worksite physical activity programs yield health behaviour changes. However, evaluation methodology used is flawed.Attention needed on work-based physical activity for older workers. No generisability of positive findings from past studies as flawed methodology.
Top of page

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Australia Sytematic review to determine whether progressive resistance training reduces impairments, activity limitations and participation restrictions after stroke.8 studies reviewed, experimental or quasi-experimental.Improved muscle strength ( d= 1.2-4.5). Improvements in walking and stair climbing (3/8 studies).Muscle strength consitently improved; some improvements in function, but insufficient evidence for social participation.
Rhodes et al. 1999 US Reviews factors at the individual level associated with regular exercise among older adults.27 cross-sectional and 14 longitudinal studies Participants mean age 65 years or over Four databases searched Dates not given.Poor health and perceptions of frailty are major barriers to exercise adoption and adherence Education and exercise history positively associated with regular exercise behaviour.Social cognitive theory guides understanding factors that predict exercise adherence, but there are relatively few studies that have tested the associations. There can be gender differences on some factors e.g. self efficacy and differences between younger and older adults.
Rydwik et al. 2004 Sweden Systematic review To describe the effect of physical training on physical activity performance in institutionalized elderly patients with multiple diagnoses.1980-2002; original RCTs in peer reviewed journals (16).Trials with samples of people aged 70+ years, with multiple diagnoses and institutionalised. Strong evidence for muscle strength and mobility benefits. Moderate evidence for range of motion gains Equivocal evidence for gait, balance, endurance and ADL.Heterogeneity highlighted. Future research needs:
  • Larger samples
  • More focus on clinically relevant outcomes
  • Greater specification/definition of interventions and their outcomes.
Seefeldt 2002 USA Commentary review To summarise the literature dealing with the determinants of and barriers to an active lifestyle in adults.No search strategy or methodology included. Descriptive studies and evaluation of the effectiveness of interventions designed to promote active living. 246 references.Invariable and modifiable influences on physical activity adoption and maintenance discussed. Socio- environmental factors systematically appear as determinants across ethnic groups. Social support common factor across cultures, generations and genders.Methodological inconsistencies have produced equivocal findings. Successful programs are tailored to individual needs and incorporate personal control and social support. Recommend for future studies: - Greater specification/definition of interventions and their outcomes.
Spirduso and Cronin 2001 US Systematic review To determine if exercise operates in a dose-response fashion to influence wellbeing and to postpone dependency.To X Search strategy outlined Cross sectional and longitudinal studies, RCTs.Level III evidence indicates that older people who are physically active have higher levels of physical function and well being. Level II evidence is equivocal in supporting this premise. Where improvement in function and wellbeing reported, no evidence that intensity acted in a dose-response manner to produce these effects.Interventions were not always well adhered to, in part due to their inappropriateness for participants. Outcome measures not always appropriate for participants.
Taylor et al. 2004 International Meta-analysis To update the systematic review of the effects of exercise-based cardiac rehabilitation in patients with CHD and to address previous concerns regarding the applicability of the evidence to routine practiceTo March 2003 48 trials (19 exercise only) 8940 patientsReduced all cause mortality (OR 0.8, 95% CI 0.68-0.93), cardiac mortality (OR 0.74, 95% CI 0.61-0.96), total cholesterol (wmd –0.37 mmol/l, 95% CI - 0.63, -0.11), triglyceride wmd –0.23mmol/l, 95% CI –0.39, -0.07 mmol/l) and systolic blood pressure (wmd –3.2 mmHg, 95% CI –5.4, -0.9 mmHg). Lower self report rates of smoking.Improvements are still seen even in newer studies where broader range of patients, more on improved drug therapy. Benefits independent of diagnosis, type of rehabilitation, exercise intervention dose, length of follow-up, trial quality and publication date. No conclusions about quality of life. All groups report improvements and analyses hampered by variable measures used.
Top of page

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Taylor et al. 2004 International Narrative review To critically review evidence on the effects of sedentary behaviour on health, for an ageing society, and whether trends towards increasing sedentary behaviour can be reversed and at what economic cost.Not stated.Few older people participate in physical activity to a level that may benefit their health. Healthy ageing is accompanied by physiological changes that are exacerbated by illness. There is some evidence that these can be reversed by physical activity uptake. Exercise has a role in improving biopsychosocial aspects of health.The UK needs national trend data for physical activity in older people. The influence of the social and physical environment on physical activity uptake requires further examination. Strategies to encourage adoption and maintenance of behaviour need further evaluation. Service delivery targets and multidisciplinary education will assist the support of physical activity uptake amongst older people.
Van der Bij et al. 2002 Netherlands Systematic review To evaluate the effectiveness of physical activity interventions among older adults.1985-August 2000 7 databases Search strategy listed 38 RCTs.Health based interventions (9), group based (38) and educational (10) interventions can promote increased physical activity. Changes are small and short-term.Home and group based programs have good participation rates (from 55- 100%). Participation appeared unrelated to type or frequency of physical activity. Behavioural reinforcement did not add to the improvements. Longer-term outcomes need to be explored, using appropriate measurement tools.
Witham et al. 2003 UK To review the benefits of exercise training for older people with chronic heart failure.Not given. 18 RCTS tabled.Efficacy studies indicate that physiological improvements occur that impact on clinical symptoms and health status.Most people with chronic heart failure are 70 years or over and studies have tended to include younger adults. More studies are needed with older adults with a range of disability and comorbidities. The long term effects of maintained training remain to be assessed.
Top of page
Table A5.5 Review articles for types of physical activity used to promote health in older people

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Verhagen et al. 2004 Netherlands Systematic review To assess the effect of Tai Chi Chuan on fall prevention, balance and cardiorespiratory function in older people.To 7/2001; 9 papers from 7 studies; Experimental design.Focused on samples aged 50 years and over. 478/505 were healthy seniors 53-96 years. Programs largely modified Yang forms, of varying frequency and duration. One trial (FICSIT) found significant reduced risk of falls. Where reported, compliance rates high (80% or greater).Benefits based largely on pre-post analyses. Limited evidence for Tai Chi in reducing falls and BP in older people.Limited blinded assessment, no follow-up assessment.
Roddy et al. 2005 UK Systematic review and consensus process To produce evidence-based recommendations on exercise in the management of hip and knee OA to guide health care practitioners.Guideline development group and literature search of reviews, trials and lower evidence literature.Both strengthening and aerobic exercise can reduce pain and improve function and health status in those with knee OA (level 1B) but limited evidence for hip OA Group exercise and home exercise are equally effective (level 1A) Exercise programs that include advice and education can promote a positive lifestyle change with an increase in physical activity (level 1B) Adherence is a principal predictor of long-term outcome from exercise in those with knee or hip OA (level 1).Number of contraindications is small (level 4 evidence) Based on the general literature, strategies to improve and maintain adherence should be adopted e.g. long-term monitoring/review and inclusion of spouse/family in exercise (only level 4 specifically for knee or hip OA) The current evidence does NOT support exercise effectiveness independent of the presence or severity of radiographic findings.

Table A5.6 Review articles of benefits of physical activity for adults

Author, review type and aimStudies reviewed – timeframe, numbers, designFindingsConclusions
Lee and Skerrett 2001 US Systematic review To assess the dose-response relationship between physical activity and all-cause mortality.1966-7/2000; 44 level III studies. Excluded where only 2 physical activity levels or specific mortality rates examined.34/44 papers provide evidence of a linear inverse dose-relationship. Five report a threshold effect (L shaped curve), 5 no significant association. Applicable to men and women, younger and older adults.Adherence to guidelines will reduce risk of all-cause mortality by 20-30%.10 studies give specific results for older adults (> 65 years).
McTiernan 2000 US Commentary To review data on the associations between BMI, physical activity and breast carcinoma risk across populations.Methodology not stated.Discusses possible physiological mechanisms for the influence of physical activity on breast cancer development. Similar association seen in Japanese and White US women. Sedentary behaviour is prevalent in CLDB groups and, with obesity, may be a key risk factor.There is a need for more research, focusing on CLDB populations, not only from an epidemiological perspective, but to gain further information about sub-groups perceptions of physical activity and appropriate means to measure it across population groups.
Eldar and Marincek 2000 Slovenia Commentary review To review the evidence for physical activity in elderly people who have identifiable neurological impairment.Methodology not stated.There is some level IV evidence for the amelioration of fatigue and functional improvement in those with multiple sclerosis and post-polio syndrome. There is level II and IV evidence for health improvements for stroke patients.The existing evidence indicates that exercise does not have any adverse effects when used to manage neurological conditions. More robust studies that focus on the older population are needed.
Sherrington et al 2004 Australia Reviews systematic reviews and RCTs which investigated the effect of physical activity / exercise on falls in older people.Search strategy defined. Studies published between Jan 2001 and Feb 2004 were included. RCTs were included if they were not included in the identified systematic reviews, and had more than 100 participants. Studies excluded if they didn’t have falls as an outcome, non-English papers, and those focusing primarily on bone health. Identified 6 systematic reviews and 3 additional RCTs meeting inclusion criteria.Strong evidence of effectiveness of individualized home exercise programs prescribed by trained health professionals (Level I). Level II evidence that group exercise programs, including tai chi, are effective in reducing falls. There is emerging evidence that individual prescription of physical activity is of greater value for frailer older people.Considerable growth in recent randomized trial research and reviews. Some methodological difficulties encountered, including inconsistent definitions of falls, differing populations used, and differences in type, intensity and dosage of physical activity programs. Also a number of successful falls prevention studies have included physical activity as part of a multi-factorial program, and while the program may have been successful in reducing falls, the individual contribution of the physical activity program to the observed effect is often unable to be delineated.
Taylor 1998 US Systematic review To evaluate physical activity interventions in low income and ethnic minority groups and people with disability.1983-1997 papers focused on healthy adults. 14 studies identified. Mainly level II-2 or III- 3 design.Ethnic minority group interventions commonly used community advisory panels, community needs assessments and community members delivering intervention. Mostly process articles: two reported increased physical activity.The four disabilities papers reported changes to physical activity behaviour.More work is needed to test the effectiveness of processes before program outcomes can be assessed. Increased use of theoretically driven approaches and the use of validated tools will be useful.Specifically excluded studies focused on older people. However, the findings about disability may be generalisable to this group.
Top of page

Document download

This publication is available as a downloadable document.

National physical activity recommendations for older Australians: Discussion Document(PDF 1160 KB)