Part of the normal aging process is a loss of muscle mass and muscle strength.1,2 These changes can contribute to a loss of balance and coordination and ultimately to the inability to perform activities of daily living.2,3 Many studies have shown that older individuals experience large strength gains through properly designed strength training regimens,4–7 but few studies have examined the ability to maintain strength following the termination of the training interventions,6,8 especially when the initial training program was longer than 24 weeks.
McCartney et al6 reported about an 8% decline in strength for men and women between the ages of 60 and 80 years after 10 weeks of detraining following a 42-week progressive overload resistance training program and they reported that the declines in strength were about twice as great in the upper body compared to lower body. One year after an 8-week training protocol, Connelly and Vandervoort9 found that elderly women had a 30% decline in isometric strength for knee flexors and a 68% decline in knee extension strength from posttraining values. The 1-year follow up knee flexion strength was close to the original pre- training strength measures; however, knee extension strength was actually 49% lower than pretraining values. A study by Taaffe et al10 demonstrated that retraining (12 weeks) after a detraining period (24 weeks) that was of equal duration as the initial training period (24 weeks) returned strength measures to their post-initial training levels in older individuals (mean age 70.8, SD 1.5 years). Participants lost 19% of their lower and upper body strength over the detraining period but remained stronger than their pre training levels. Porter et al11 examined the effects of detraining following a 1-year high-intensity strength training program in women with a mean age of 60 years. Participants during the second year of the study chose to either continue to train with little or no supervision or to stop training. There was a 7% decline in lat pull down strength and a 9% decline in knee extension strengthfor those who continued to train, while there was a 20% decline and 24% decline in the lat pull down and knee extensor strength, respectively, for those who stopped training.
The purpose of this study was to examine strength changes that occur 6 months after completion of either 40 or 80 weeks of supervised progressive overload resistance training in older men and women. On the basis of the findings from previous studies, it was hypothesized that strength measured 6 months after training would be higher than pretraining levels; however, individuals that trained for the longer duration (80 weeks) and those that continued to exercise on their own during the 6 months would have greater strength maintenance. In addition, it was hypothesized that lower extremity muscular strength would be better maintained than upper extremity strength.
Participants for this study had previously completed either a 40-week (n = 125) or a 80-week (n = 44) supervised progressive overload resistance training regimen at either 40% or 80% of their one repetition maximums (1RM). All were independently living at the time of the study. Ninety percent of the participants were white and all were deemed competent as indicated by their physician-approved medical clearance. The inclusion criteria for the training programs were the following: (1) aged 55 to 75 years, (2) normally active but not engaging in a weight training regimen over the previous year, (3) free of endocrine and metabolic disorders that might have had an effect on muscle mass, and (4) sufficient cognitive level to comprehend and execute instructions of the program supervisors. Exclusion criteria were as follows: (1) physical disability that prevented participation in all of the lifts, including arthritis; (2) formal diagnosis of congestive heart failure, serious arrhythmias, or uncontrolled blood pressure; and (3) abnormal blood test (on a Chem 25 metabolic panel) results that were indicative of a serious health-related problem. The Chem 25 test is the assessment of serum for standard ions, like calcium, potassium, sodium, etc, and for basic lipoproteins (HDL, LDL, VLDL) and glucose levels and kidney function indicators (creatinine). Participants obtained medical clearance from their personal physicians. Once cleared, participants were allowed to choose their training frequency (2 or 3 days/week). This was done to investigate the dose-response effect of the 40- or 80-week training programs and to increase participant compliance with the training program. All participants were also instructed not to begin any new structured physical activity during the study. The institutional review board of the University of Oklahoma approved this study, and written informed consent was obtained from all participants.
The training programs consisted of 12 isotonic resistance exercises performed on Cybex equipment (Ronkonkoma, NY); 5 upper body (forearm flexion/extension, shoulder press, latissimus pull-down, seated row), and 7 lower body (knee flexion/extension, 2 leg press, hip flexion/extension, hip abduction/adduction). At the beginning and completion of the supervised resistance training program, participants were tested using a standard protocol for assessing strength (1RM) on the same Cybex isotonic weight machines that were utilized during the training programs. All loads were recorded in kilograms of mass moved. Progressive overload was applied to the training programs by assessing 1RMs for each exercise every 5 weeks, then adjusting the training loads' new strength values (either 40% or 80% 1RM). Similar training volumes between the different training programs was accomplished by setting the repetition number to 16 for the low-intensity workload and to 8 for the high-intensity workload for each of the 3 sets per exercise. Exercise logs were kept for each participant in which they recorded all the reps and sets for each exercise that was successfully completed during a given exercise session.
Following completion of the supervised programs, participants were encouraged by program staff to attend a resistance exercise program of their choice in the community from a list of possible venues. All participants (66 men, 103 women) were contacted 6 months after the completion of the training programs and 1RM strength was reassessed in a laboratory setting. Not all participants responded or returned for the 6-month follow-up. A total of 69 of the original 169 participants (41%) returned for the strength testing. From the 40-week training group, 39 returned (31%), while 30 participants returned from the 80-week training group (68%). Twenty-five men (46%) and 44 women (55%) from the original sample had pretraining (PRE), posttraining (POST: either 40 or 80 weeks), and 6-month posttraining (6MP) values for analyses. Participants were also asked about their resistance exercise status. Participants who reported that they continued to perform resistance training at least twice per week for the 6-month period were classified in the exercise group (Ex), and the remaining were placed into the no exercise group (NoEx) for statistical analyses. No logs were collected to verify the exercise information; however, self-report physical activity questionnaires were completed that indicated a clear distinction between those place into the exercise (Ex) and no exercise groups (NoEx)
The 1RM protocol consisted of a 5-minute warm-up on a stationary bicycle, 3 minutes of rest, and then 8 to 10 repetitions at approximately 50% 1RM for the given lift, based on the participant's previous 1RM. Single, progressively heavier repetitions were then attempted until failure to complete the full range of motion with 1 minute of rest between attempts. The 1RMs were determined within 5 attempts. Five minutes separated each different exercise and 6 or 7 exercises, alternating between upper and lower body exercises were tested on a given day with at least 48 hours between 2 test days. The values obtained from each individual exercise were divided into upper and lower body strength values and the means for the upper body and lower body strength were calculated as the sum of the individual 1RMs divided by the number of exercises tested for the upper or lower body.
Data were analyzed using SPSS for Windows version 17.0 (SPSS, Inc, Chicago, Illinois). Descriptive data are reported as means ± SE standard error (SE). Three-way repeated-measures analysis of variance (ANOVA) was used to evaluate the effects of time (PRE, POST, 6MP), training duration (40 or 80 weeks), and posttraining exercise status (Ex, NoEx) on the averaged upper body and lower body muscular strengths. As training volume was equalized between the high- and low-intensity groups, the training intensity (40% 1RM or 80% 1RM) was not analyzed as a dependent variable. Likewise, training frequency was assumed to be a nondependent variable, as it was self-selected by the participants. Dependent variables with no significant 3-way interactions then were analyzed by a 2-way, repeated-measures ANOVA (time X training duration). Paired t tests with Bonferroni correction for multiple comparisons were used as a post hoc procedure for significant time effects. Percent changes in strength were computed for PRE versus POST, PRE versus 6MP, and POST versus 6MP. Two-way ANOVA (training duration X exercise status) was used to determine the effects of weeks of training and continuation of exercise for each percent changes strength variable (PRE-POST, PRE-6MP, and POST-6MP).
Table 1 displays the distribution of participant's demographics by gender, training duration, and whether the individuals reported exercising (Ex) or not (NoEx) between the POST and 6MP measurements. Men and women were similar in age, but the men were significantly heavier (P = .001) and taller (P = .000) than the women in the study. There were no significant differences in baseline values of age, height, or weight between the groups when divided by duration of training. When the data were split on the basis of exercise status only weight as significantly different (P = .016) between the groups with the NoEx group weighing significantly more than those that continued to exercise. Overall, there was a 13% dropout rate and compliance was good, with 75% of the training sessions attended (range was 50% to 100% per individual). Participants that returned for 6MP were not different from those that did not return when divided by training duration at PRE and at POST on the basis of upper body mean strength and lower body mean strength (all P > .1).
The PRE, POST, and 6MP upper and lower body mean strength values for the entire sample, for each gender, training duration group (40 or 80 weeks), and exercise status (Ex or NoEx) are shown in Table 2. There was not a significant time X gender interaction based on changes in strength, indicating that men and women responded in a similar manner at each time point. Table 3 provides the percent changes in upper and lower body strength between the 3 time periods (PRE, POST, and 6MP) for the entire sample, for each gender, training duration group (40 or 80 weeks), and exercise status after the training programs (Ex or NoEx). Percentage increases in upper body strength were similar for both men and women following training, and both genders declined in the upper body strength measures 6MP compared to POST values. Men had slightly greater, but not significant, increases in lower body measures of strength following the training period compared to women. Unlike the changes observed at 6MP for the upper body, lower body strength remained unchanged for both men and women 6MP compared to POST values. Because of the similar, nonsignificant differences, the genders were consolidated for subsequent analysis.
Upper Body Mean Strength
As expected, POST values were significantly higher than PRE (P = .01). Figure 1 represents the time course of the changes in upper body mean strength for training duration and post exercise status. 6MP values were also significantly higher than PRE (P = .01); however, POST values were significantly higher than 6MP (P = .01). There were significant 2-way interactions for time X training duration (P = .01) and time X exercise status (P = .01). There was not a significant 3-way interaction (time X training duration X exercise status) or an interaction between training duration and postexercise status on the basis of the percent changes for the groups. The group that trained for 40 weeks had significantly higher POST (mean 42.1 ± SE 2.1 kg) and 6MP (36.9 ± SE 1.8 kg) upper body mean strength compared to PRE values (28.5 ± SE 1.7 kg, P = .01); however, the upper body mean strength at 6MP was significantly lower than the POST value (P = .01). Participants that trained for 80 weeks exhibited significantly higher POST (36.1 ± SE 2.6 kg, P = .01) and 6MP (37.0 ± SE 2.7 kg, P = .01) values than PRE (29.7 ± SE 2.1 kg), but there was no statistical difference between 6MP and POST (P = .37). The group that completed the 40-week intervention had a significantly greater percent increase in upper body strength from PRE to POST (52.2 ± SE 4.4%, P = .01) compared to the 80-week intervention group (22.1 ± SE 2.5%). However, the 40-week group then lost strength (-11.0 ± SE 2.3%) while the 80- week group continued to gain (3.6 ± SE 2.5%, P = .01).
For the Ex group, POST (40.6 ± SE 2.4 kg) and 6MP (40.0 ± SE 2.3 kg) values were significantly higher than PRE values (29.6 ± SE 1.8 kg, P = .01) but there was no difference between POST and 6MP upper body mean strength values. For the NoEx group, POST (38.1 ± SE 2.3 kg) and 6MP (33.2 ± SE 1.8 kg) values were also significantly higher than PRE values (28.4 ± SE 2.0 kg, P = .01); however, 6MP values were significantly lower than POST upper body strength values (P = .01). There were similar percent changes in upper body mean strength from PRE to POST for both the Ex and NoEx groups, regardless of training duration. The percent changes from PRE to 6MP, for those that continued to exercise, had a significantly greater changes (38.0 ± SE 4.0%, P = .01) compared to the NoEx group (22.9 ± SE 4.4%). Finally, in examining the percent changes from POST to 6MP, there was a significantly greater increase for the Ex group (0.4 ± SE 2.6%, P = .02) compared to the NoEx group (-10.9 ± SE 2.4%).
Lower Body Mean Strength
The time courses for each training duration and post exercise group can be seen in Figure 2. There was only a significant main effect for time (P = .01) from the 3-way repeated-measures ANOVA (time X training duration X exercise status) with both POST (P = .01) and 6MP values (P = .01) being significantly higher than PRE. Unlike upper body mean strength, there was no significant difference between 6MP and POST values. There were significant 2-way interactions for time X training duration (P = .01) and time X exercise status (P = .01).
Participants who trained for 40 weeks had significantly higher POST (77.8 ± SE 4.1 kg, P = .01) and 6MP (71.5 ± SE 3.4 kg, P = .01) values compared to PRE (46.8 ± SE 2.6 kg); however, the mean lower body strength at 6MP was significantly lower than POST (P = .04). The participants who completed the 80-week training program had POST (68.0 ± SE 4.1 kg, P = .001) and 6MP (75.3 ± SE 5.0 kg, P = .01) strength values that were significantly higher than PRE (52.1 ± SE 3.1 kg); in addition, the 6MP values were significantly higher than POST values (P = .01). The 40-week cohort had a percentage increase in their lower body mean strength of (71.1 ± SE 5.9%) from PRE to POST, which was significantly more than the 80-week group (33.3 ± SE 5.0%, P = .01). During the unsupervised period (POST to 6MP), the 40-week group lost significantly more strength than the 80-week group (-5.3 ± SE 3.2% vs 10.5 ± SE 2.8%, P = .01). These differences in gains and losses over the entire duration of the study (PRE to 6MP) resulted in a nonsignificant difference between the percentage gains on the basis of training duration: 40 week, 60.5 ± SE 6.3% versus 80 week, 47.1 ± SE 6.3%.
For the Ex group, POST (76.4 ± SE 4.1 kg) and 6MP (80.5 ± SE 4.2 kg) values were significantly higher than PRE values (50.4 ± SE 2.6 kg, P = .01), but there was no difference between POST and 6MP lower body mean strength values. Similar results were observed for the NoEx group, with POST (70.1 ± SE 4.1 kg) and 6MP (64.2 ± SE 3.2 kg) values being significantly higher than PRE values (47.6 ± SE 3.2 kg, P = .01) and there was only a trend for a difference between 6MP values and POST lower body strength values (P = .07). Figure 3 graphically represents the percent changes in upper and lower body mean strength from POST to 6MP for each training duration and exercise grouping. The percent change for the training period was similar for the Ex and NoEx groups (56.5 ± SE 7.0% and 52.3 ± SE 5.6%, respectively). During the unsupervised period, the Ex group continued to gain strength (7.5 ± SE 3.4%), while the NoEx group regressed (−5.6 ± SE 2.7%); this difference was significant (P = .01) and led to total change in strength (PRE to 6MP) also being significantly different on the basis of exercise status: Ex, 64.1 ± SE 6.2% versus NoEx, 43.1 ± SE 6.2%, P = .01.
A unique aspect of our research design is that the individuals who decided not to continue to exercise at the end of the formal training periods provided a pool of participants that allowed us to examine aspects of detraining, whereas the participants who continued to exercise on their own in the community at the end of the formal training programs allowed us to examine the effects of a nonsupervised resistance training on the ability to maintain or improve on the benefits achieved from the supervised programs. While this research approach allows for the gathering of some novel information, it also presents some potential problems that make the interpretation of the findings somewhat difficult and less robust to standard statistical analyses.
Our findings supported the hypothesis that strength values 6 months after supervised training were higher than the pretraining strength measures for all participants, regardless of training duration or posttraining exercise status. Therefore, older participants were able to retain some of their resistance training benefits, whether they continued toexercise on their own or not. This indicates that while regression of strength does occur, it may occur at a pace close to that seen in young adults12 and not be altered because of age-related changes in protein intakes, protein synthesis rates, and age-related muscle loss.
The results from our study confirm that older men and women can significantly improve both upper and lower body strength following a supervised, progressive overload resistance training program and that these participants could easily tolerate a 40- or 80-week training program, because no adverse events or injuries were reported throughout the training programs. It is also important to note that improvements in strength continued over the entire 40- or 80-week periods with the progressive overload model, similar to the finding of McCartney et al.6 Generally, the improvements for the lower body were significantly greater for both men and women than for the upper body (58.2% and 52.6% vs 38% and 39.8%, respectively), regardless of the total training time. Compliance was also similar to McCartney et al,6 who reported a dropout rate of 12% with about 84% of the training sessions attended over two 10-month periods. Somewhat surprising, in each analyses, there was no main effect for exercise status (Ex or NoEx), suggesting that changes observed in strength following the training program were independent of whether or not participants continued to exercise. This result may arise from a loss of statistical power because of multiple comparisons being performed on a limited number of available participants in each subgroup.
We found that the 6MP strength measures were generally lower than the POST values. The one exception occurred for the group of participants that completed the 80-week training program and then continued to exercise on their own: upper body strength values increased through their own exercise programs compared to POST values (8.7%), and lower body strength increased 17.7%. The percent declines in upper and lower body strength over the 6-month period following the training were greatest for those who completed the 40-week supervised program and then decided to stop exercising (— 15.2% and —9.1%, respectively); followed by those that competed the 40-week program and continued to exercise on their own (— 7.0% and —1.7%, respectively); and finally those that completed the 80-week program and decided to stop training (—4.1% and —0.2%, respectively). The self-selecting nature of the participants that chose to extend their supervised training period to 80 weeks may have been the driving force for their continued gains through the unsupervised period.
Generally, the magnitude of strength declines we found 6MP training is in agreement with other more traditional detraining studies. McCartney et al6 reported that overall strength declined by 9% after a 2-month layoff. Porter et al11 reported smaller declines in strength after 1 year of training and a 1 year unsupervised period of 8% for the knee extensors and lat pull-down. Henwood and Taaffe13 reported declines of 16% and 17% for total body strength over a 6-month detraining period for previous high-velocity and strength trained groups, respectively. The results of this study were considerably less than the losses observed by Henwood and Taaffe.13 While the current study generally found that strength did significantly decline over the 6-month lapse, the 6MP levels were still greater than PRE levels; this is contrary to the findings of Kalapotharakos et al,14,15 who reported that strength returned to baseline levels after 6 weeks of detraining. In older adults (61–75 years, men), 10 weeks of moderate training resulted in modest strength gains in knee flexion and extension, which was lost over 6 weeks of detraining.13 The group also tested very old men (80 years and older) and found some mild retention of strength after 8 weeks of training and 6 weeks of detraining: 13% greater than PRE for lower extremity and 11% for the upper extremity.15 The findings of this study are similar to Taaffe and Marcus,16 who reported muscular strength declined during the detraining period by 7% overall. A difference between Taaffe and Marcus's protocol and the present study was the instruction to not exercise duringthe detraining period by the previous study,16 while the present study encouraged participants to exercise independently. No specific training program (frequency, loads, or progression) was suggested.
The amount of decline appeared to be greater for those completing the 40-week program than for those completing the 80-week program. In contrast to the 5-year (2 years of training, 3 years of detraining: 150% of training time) study of Smith et al,8 our study showed greater retention of both upper and lower body strength over the detraining period. The extrapolated timeline for the current study and the data of Smith et al8 mirrors that of Harris et al,17 who found that some strength was lost at 6 weeks of detraining after 18 weeks of training (33% training time), but a greater loss occurred from week 6 to week 20 (111% of training time) of detraining. After the 20-week detraining period, older adults retained greater strength than they began the 18- week training regimen with: total body 14% to 20%, upper body 6% to 17%, and lower body 15% to 27%.17 Similar to Taaffe and Marcus,16 participants were discouraged from exercising during the detraining period. Our findings do not support a relative time-related reduction in strength retention, as there was not a training duration effect on the percent of strength lost during the detraining period for either the lower or the upper extremities. This may be a result of the sample size of this study. On the basis of previous research, one would expect to see statistically significant declines in strength following 40 weeks of training and 24 weeks of detraining (60% of training time) compared to the 80-week group (30% of training time).
The focus on hip musculature during the supervised training period, 5 of the 7 lower extremity lifts directly involved the hips, should have decreased the risk of falling for the participants. Hip abduction strength and power, in particular, has been linked to fall risk in several studies.18,19 This focuses on decreasing falls from perturbations lateral to the direction of motion. Knee extension strength has also been shown to be important in arresting falls as the quadriceps are the main muscle groups that extend the knee.18 Participants in the current study performed 2 exercises with the quadriceps: seated knee extension and leg press on a sled. While the number of falls was not recorded for this study, no falls were reported by the participants either. The improved maintenance of the lower extremity mean strength, compared to the upper extremity strength, may be indicative of a lowered risk of falls for these participants.
A limitation of the current study was the design, because the results of the relative duration of the decline were most likely confounded by the continuation of an exercise regimen by some, but not all participants, in both the 40- and 80-week training groups following the cessation of the supervised training program. Again, the self-selecting nature of those that continued to exercise after the supervised training period likely confounded the overall results of this study. A lack of quantification of the amount of resistance training (both intensity and volume) further confounds the results of this study.
Older adults were able to easily tolerate and demonstrate significant increases in muscular strength following either 40-week or 80-week supervised strength training programs. Both groups were able to maintain upper body and lower body strength at levels that were significantly greater than pretraining levels, regardless whether they continued to exercise or not. However, the longer-duration training program provided a greater, although non-statistically significant, ability to maintain strength at higher levels at 6MP training compared to the strength values obtained at the end of the two different training periods. Strength was better maintained for the lower body compared to the upper body for both groups.
Perhaps an important aspect of these findings is that strength values declined for both groups, regardless of training status. Because these individual may actually represent a more physically active group of individuals compared to the reference population (because they actually volunteered to train for either 40 or 80 weeks), much greater declines in strength for a more sedentary population should be expected.
It should also be mentioned that the results should be viewed with some caution because the research design allowed for obtaining novel information regarding exercise adherence after a supervised program and about detraining, but it also resulted in some difficulties in statistical interpretation.
1. Esposito F, Malgrati D, Veicsteinas D, Orizio C. Time and frequency domain analysis of electromyogram and sound myogram in the elderly. Eur J Appl Physiol. 1996;73:503–510.
2. Hughes V, Frontera W, Wood M, et al. Longitudinal muscle strength changes in older adults: influence of muscle mass, physical activity, and health. J Gerontol A Biol Sci Med Sci. 2001;56(5):B209-B217.
3. Booth F, Weeden S, Tsengm B. Effect of aging on human skeletal muscle and motor function. Med Sci Sports Exerc. 1994;26:556–560.
4. Charette S, McEvoy L, Pyka G, et al. Muscle hypertrophy response to resistance training in older women. J Appl Physiol. 1991;70:1912–1916.
5. Laforest S, St-Pierre M, Cyr J, Gayton G. Effects of age and regular exercise on muscle strength and endurance. Eur J Appl Physiol. 1990;60:104–111.
6. McCartney N, Hicks A, Martin J, Webber C. A longitudinal trial of weight training in the elderly: continued improvements in year 2. J Gerontol A Biol Sci Med Sci. 1996;51(6):B425-B433.
7. Pyka G, Linderberger E, Charette S, Marcus R. Muscle strength and fiber adaptations to a year-long resistance training program in elderly men and women. J Gerontol. 1994;49:M22-M27.
8. Smith K, Winegard K, Hicks AL, McCartney N. Two years of resistance training in older men and women: the effects of three years of detraining on the retention of dynamic strength. Can J Appl Physiol. 2003;28:462–474.
9. Connelly D, Vandervoort A. Effects of detraining on knee extensor strength and functional mobility in a group of elderly women. J Orthoped Sports Phys Ther. 1997;26:340–346.
10. Taaffe DR, Henwood TR, Spry N, Joseph D, Turner D, Newton RU. Alteration in muscle attenuation following detraining and retraining in resistance-trained older adults. Gerontology. 2009;55:217–223.
11. Porter M, Nelson M, Fiatarone-Singh M, et al. Effects of a long term resistance training and detraining on strength and physical activity in older women. J Aging Phys Act. 2002;10:260–270.
12. Lemmer JT, Hurlbut DE, Martel GF, et al. Age and gender responses to strength training
and detraining. Med Sci Sports Exerc. 2000;32:1505–1512.
13. Henwood TR, Taaffe DR. Detraining and retraining in older adults following long-term muscle power or muscle strength specific training. J Gerontol A Biol Sci Med Sci. 2008;63(7):751–758.
14. Kalapotharakos VI, Smilios I, Parlavatzas A, Tokmakidis SP. The effect of moderate resistance strength training
and detraining on muscle strength and power in older men. J Geriatr Phys Ther. 2007;30:109–113.
15. Kalapotharakos VI, Diamantopoulos K, Tokmakidis SP. Effects of resistance training and detraining on muscle strength and functional performance of older adults aged 80 to 88 years. Aging Clin Exp Res. 2010;22:134–140.
16. Taaffe D, Marcus R. Dynamic muscle strength alterations to detraining and retraining in elderly men. Clin Physiol. 1997;17:311–324.
17. Harris C, DeBeliso M, Adams KJ, Irmischer BS, Gibson TA. Detraining in the older adult: effects of prior training intensity on strength retention. J Stren Cond Res. 2007;21:813–818.
18. Lloyd BD, Williamson DA, Sigh NA, et al. Recurrent and injurious falls in the year following hip fracture: a prospective study of the incidence and risk factors from the Sarcopenia and Hip Fracture study. J Gerontol A Biol Sci Med Sci. 2009;64(5):599–609.
19. Hilliard MJ, Martinez KM, Janssen I, et al. Lateral balance factors predict future falls in community-living older adults. Arch Phys Med Rehabil. 2008;89:1708–1713.