Evidence Table 3

Studies Examining Exercise-Related Interventions in the Community (listed chronologically)

SourceSafety Issue Related to Clinical PracticeDesign TypeStudy Design, Study Outcome Measure(s)Study Setting & Study PopulationStudy InterventionKey Finding(s)
Reinsch 1992124Exercise-related interventions for fall prevention in the communityRCTDesign: Four-arm RCT

Outcomes: Fall rates, time to first fall, injury rates
Setting: 16 senior centers in Orange County, California
Population: 230 older adults who were participants at senior centers
Exercise in conjunction with cognitive behavioral therapy for safety self-awarenessFalls: No effect on fall rates, falls efficacy, or fear of falling.
Time to first fall: Longer time to first fall.
Injuries: Decreased injuries. Even though a relatively high percentage (38.6 percent) suffered at least one fall, only 7.8 percent of these community-residing elderly required medical attention.
Province 1995115Exercise-related interventions for fall prevention in the communityMeta-analysisDesign: Preplanned meta-analysis of 7 RCTs

Outcomes: Time to each fall (fall-related injury) by self-report and/or medical records
Setting: Two nursing homes and five community sites

Population: Patients ages 60–75, ambulatory, cognitively intact
Exercise training one area or more of endurance, flexibility, balance platform, Tai Chi (dynamic balance), and resistanceFalls:
  • Fall rates decreased in group with general exercise (odds ratio = 0.90, 95% CI = 0.81–0.99).
  • Fall rates decreased for those with exercise plus balance training (odds ratio = 0.83, 95% CI = 0.70–0.98).


Injuries: Patients who did not exercise had an increase in injurious falls, but power was low to detect this outcome.
Wolf 1997120Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Frailty indicators, occurrence of falls
Setting: Community

Population: 200 men and women ≥ 70 years
Tai Chi, computerized balance training, or educationMultiple falls: Risk of multiple falls decreased by 47.5 percent
Steinberg 2000143Exercise-related interventions for fall prevention in the communityRCTDesign: RCT with four arms

Outcomes: Self-reported slips, trips, or falls
Setting: Community, Australia

Population: 252 active, community-dwelling Australians ≥ 50 yrs.
Education re: fall risk factors, strength/balance exercises, home safety advice, medical evaluationFalls: 30 percent reduction in falls; hazard ratio 0.70 (95% CI = 0.48–1.01).
Slips: 58 percent reduction in slips; hazard ratio 0.42 (95% CI = 0.29–0.69).
Trips: 64 percent reduction in trips; hazard ratio 0.36 (95% CI = 0.26–0.66).
Rubenstein 2000116Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Muscle strength, endurance, mobility, balance, fall rates
Setting: Community-living men

Population: 59 men ≥ 65 years with specific fall risk factors
90 min. exercise sessions 3x/week Focus on increased strength and endurance, improving mobility and balanceFalls:
  • Exercise group had lower fall rates than nonexercisers when adjusted for baseline activity level (6 falls/1,000 hours of activity vs 16.2 falls/1,000 hours, P < 0.05).
  • Total number of falls not decreased.

Strength: Exercise achieved no significant effect on hip or ankle strength, balance, self-reported physical functioning.
Robertson 2002123Exercise-related interventions for fall prevention in the communityMeta-analysisDesign: Meta-analysis of four studies

Outcomes: Fall rates, injury rates
Setting: Community setting: nine cities and towns in New Zealand

Population: 1,016 women and men ages 65 to 97
Muscle strengthening and balance retraining exercises designed specifically to prevent fallsFalls and injuries: Fall and injury rates decreased by 35 percent; no difference between genders.
  • Fall rate incidence rate ratio (IRR) = 0.65, 95% CI = 0.57–0.75
  • Participants reporting a fall in the previous year had a higher fall rate (IRR = 2.34, 95% CI = 1.64–3.34).
  • Injury rate IRR = 0.65, 95% CI = 0.53–0.81.
Barnett 2003 113Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Fall rates, balance, muscle strength, fear of falling
Setting: Community, South Western Sydney, Australia.

Population: 163 subjects ≥ 65 years identified as at risk of falling using a standardized assessment screen by general practitioner or physical therapist
Weekly group exercise program with ancillary home exercises over 1 yearFalls: Fall rates decreased by 40 percent in the exercise group (IRR = 0.60, 95% CI = 0.36– 0.99).

Balance measures: Improved in exercise group.

Other measures: No difference between groups in strength, reaction time, and walking speed or on Short-Form 36, Physical Activity Scale for the Elderly or fear of falling.
Wolf 2003122Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Time to first fall, fall rates, balance
Setting: 20 congregate living facilities in the greater Atlanta area

Population: 291 women and 20 men ages 70 to 97 who were transitioning to frailty
Intense Tai Chi exercise program or wellness education programFalls: Fall rates decreased in Tai Chi group, but no statistical difference between groups (relative risk = 0.75, 95% CI = 0.52–1.08).
Clemson 2004114Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Fall rates
Setting: Community

Population: 310 men and women ≥ 70 years who had had a fall in the previous 12 months or were concerned about falling
Occupational therapy home visits, lower-limb balance and strength training, environmental safety educationFalls:
  • 31 percent reduction in falls for both genders (relative risk = 0.69, 95% CI = 0.50–0.96; P = 0.025).
  • For men alone, 68 percent reduction in falls (relative risk = 0.32, 95% CI = 0.17–0.59).
Morgan 2004140Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Falls
Setting: Community

Population: 294 men and women ≥ 60 years who had either a hospital admission or bed rest for 2 days or more within the previous month
Exercise sessions lasting 45 minutes, including warm-up and cool-down, 3 times a week for 8 weeks (24 sessions)Falls:
  • 49 percent reduction in falls for patients with low baseline physical functioning.
  • 3.5 times increase in falls for patients with high baseline physical functioning.
Suzuki 2004118Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Fall rates
Setting: Community, Japan

Population: 52 elderly Japanese women
Exercise intervention— home and community centerFalls: Fall rates decreased in intervention group (13.6 percent v. 54.5 percent; P = 0.0097).
Li 2005112Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcome: Fall rates, functional balance, physical performance, fear of falling
Setting: Community in Portland, Oregon

Population: 256 physically inactive elders ages 70 to 92
Tai Chi or stretching 3x/week for 6 monthsFalls: 55 percent reduction in falls in Tai Chi group (relative risk = 0.45, 95% CI = 0.30–0.70). Fewer falls in the Tai Chi group (Tai Chi = 38 vs. stretch = 73; P = 0.007), (Tai Chi = 28 percent vs. stretching = 46 percent; P = 0.01).
Injuries: Fewer injurious falls (Tai chi = 7 percent vs. stretching = 18 percent; P = 0.03).
Lord 200588Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcome: Fall rates
Setting: Community in Australia

Population: 620 people ≥ 75 years
Interventions to maximize vision and sensation or brief advice or usual careFalls: The rate of falls during the trial period were similar in the three groups.
Injuries: The rate of injurious falls during the trial period were similar in the three groups.
Faber 2006125Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Falls, mobility, physical performance, and self-reported disability
Setting: 15 homes for the elderly in Amsterdam, The Netherlands

Population: 287 elderly men and women (mean age +/− standard deviation, 85+/− 6yrs)
20-week exercise program of balance training inspired by Tai Chi or daily mobility activities or controlFalls: Fall incidence rate lower in balance training group (2.4 falls/yr) compared to the mobility activities group (3.3 falls/yr) and control (2.5 falls/yr), but not statistically significant.
For frail subjects: Risk of becoming a faller in the exercise groups increased almost 3 times (hazard ratio = 2.95; 95% CI = 1.64–5.32).
For pre-frail subjects: Risk of becoming a faller decreased by 61 percent (hazard ratio = 0.39; 95% CI = 0.18–0.88).
Lin 2006126Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Falls, fall-related injuries, related functional outcomes
Setting: 6 rural villages in Taiwan: 2 villages received intervention, 4 villages acted as controls
Population: 1,200 men and women ≥ 65 years screened; 88 participants
Tai Chi training plus fall-prevention education or fall-prevention education aloneFalls: 50 percent greater decrease in fall rates among the Tai Chi practitioners (relative risk = 0.5; 95% CI = 0.11–2.17), but not statistically significant.
Tinetti Balance Scale: Tai Chi practitioners increased by 1.8 points (95% CI = 0.2–3.4).
Tinetti Gait Scale: Tai Chi practitioners increased by 0.9 point (95% CI = 0.1–1.8).
Fear of Falling: No significant changes in the fear of falling.
Luukinen 2007127Exercise-related interventions for fall prevention in the communityRCTDesign: RCT

Outcomes: Fall rates, time to first fall
Setting: Community, home-dwelling Finnish

Population: 555 older men and women (67 percent ≥ 85 years), most with history of recurrent falls or at least one mobility risk factor
Suggestions for a program consisting of home exercise, walking exercise, group activities, self-care exercise, or routine careFor all subjects:
Falls:
  • 12 percent decrease in falls from baseline for intervention group (hazard ratio = 0.88, 95% CI = 0.74–1.04).
  • 7 percent decrease in all falls, but not statistically significantly (hazard ratio = 0.93, 95% CI = 0.80–1.09).

For subjects not homebound:
Falls:
  • 22 percent decrease in falls (hazard ratio = 0.78, 95% CI = 0.64–0.94).
  • 12 percent decrease in first four falls (hazard ratio = 0.88, 95% CI = 0.74–1.05).

Study Design Type: (1) Meta-analysis, (2) Randomized controlled trials, (3) Nonrandomized trials, (4) Cross-sectional studies, (5) Case control studies, (6) Pretest and post-test (before and after) studies, (7) Time series studies, (8) Noncomparative studies, (9) Retrospective cohort studies, (10) Prospective cohort studies, (11) Systematic literature reviews, (12) Literature reviews, nonsystematic/narrative, (13) Quality-improvement projects/research, (14) Changing-practice projects/research, (15) Case series, (16) Consensus reports, (17) Published guidelines, (18) Unpublished research, reviews, etc.

Study Design Type: (1) Meta-analysis, (2) Randomized controlled trials, (3) Nonrandomized trials, (4) Cross-sectional studies, (5) Case control studies, (6) Pretest and post-test (before and after) studies, (7) Time series studies, (8) Noncomparative studies, (9) Retrospective cohort studies, (10) Prospective cohort studies, (11) Systematic literature reviews, (12) Literature reviews, nonsystematic/narrative, (13) Quality-improvement projects/research, (14) Changing-practice projects/research, (15) Case series, (16) Consensus reports, (17) Published guidelines, (18) Unpublished research, reviews, etc.

From: Chapter 10, Fall and Injury Prevention

Cover of Patient Safety and Quality
Patient Safety and Quality: An Evidence-Based Handbook for Nurses.
Hughes RG, editor.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.