| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
a Schools of Exercise and Sport Science, Faculty of Health Sciences, The University of Sydney, Australia.
b Schools of Physiotherapy, Faculty of Health Sciences, The University of Sydney, Australia.
Sandra K. Hunter, Neural Control of Movement Laboratory, Department of Kinesiology and Applied Physiology, University of Colorado at Boulder, Campus Box 354, Boulder, Colorado 80303-0354 E-mail: shunter{at}colorado.edu.
Decision Editor: Jay Roberts, PhD
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionConclusionReferences |
|---|
performed maximal voluntary contractions with the dominant and nondominant knee extensors, plantar flexors, and handgrip. Anthropometric measurements were made and questionnaire responses used to obtain current physical activity levels. Trend analysis within analysis of variance and regression analysis on strength was performed. Limb muscle strength was found to be associated with increased age, muscle group, limb dominance, and activity. Self-reported physical activity levels declined with age but women who were more physically active for their age group were stronger in all muscle groups and had more lean body mass and lean thigh and leg cross-sectional area than relatively inactive women. Slopes of the linear reductions of maximal voluntary strength of the knee extensors, plantar flexors, and handgrip with age were significantly different ( p < .05) at 9.3%, 7.4%, and 6.2% per decade, respectively. The limb muscle strength of healthy Australian independent and urban-dwelling women aged 20 to 89 years was found to be associated with age and three aspects of disuse: muscle group, relative levels of physical activity, and limb dominance. THE physiological changes in the human musculoskeletal system with increasing chronological age have occupied both poets and physiologists over the centuries. Shakespeare's "shrunk shank" depiction of the age-related loss of gastrocnemius muscle volume (1) is in agreement with numerous studies that show that the aging process is associated with loss of skeletal muscle mass and subsequent weakness (2). Lower limb weakness experienced by many older men and women is related to a loss of mobility in activities such as ambulation, rising from a chair, and stair climbing (3)(4)(5)(6)(7). Some older individuals, in particular women, are so weak that they have a limited physiological reserve of strength (8), and tasks such as rising from a chair may require close to or all their maximal leg strength (6)(9).
The onset and the magnitude of muscle weakness in older adults have been described by a large number of cross-sectional studies for various populations, with a range of conclusions reached. Although some studies have concluded that isometric strength declines in upper- and lower-limb muscles around the fifth to sixth decade of life (10)(11)(12)(13)(14), other data have shown that the strength decrement begins as early as 30 years of age in some muscle groups (15)(16). The discrepancy may be a result of the range of muscles tested and the varying degrees of disuse specific to individual muscle groups and/or the populations studied.
Several cross-sectional studies have reported that the greatest decrement in strength is experienced in the lower limb muscle groups of an older individual (17)(18)(19). Bemben and colleagues (16), however, measured the strength of a number of upper- and lower-limb muscle groups in 153 men, aged 20 to 74 years, and they reported the greatest reduction in the elbow extensors compared with the plantar flexors (PFs), dorsiflexors, thumb abductors, and finger flexors. Further, a longitudinal study of 70- to 75-year-old men and women showed that, although reductions in the maximal isometric strength of handgrip (HG), arm flexion, trunk flexion, and trunk extension ranged between 5% and 16% over a 5-year period, knee extensor (KE) strength in both men and women increased significantly by 4% over the same time period in these older individuals (20).
A potential confounding factor in defining the strength decrements across age groups is the physical activity level of the sample group assessed. Older adult volunteers in particular will tend to be more active and healthy (21) and not representative of the broader older population. We revisited this issue of muscle-specific strength reductions that occur with age in a group of Australian independent and urban-dwelling women aged 20 to 89 years. These women were representative in their physical activity levels of a broader sample of Australian women who had been randomly surveyed by the National Heart Foundation (22). Further, we investigated the interaction of relative disuse on limb strength. It is well established that participation in physical activity is reduced in older men and women compared with younger individuals (22). If disuse is a major determinant of muscle strength, then strength changes with increasing chronological age should be associated with the physical activity level of the individual and their preference for using a particular limb (dominance). We aimed to determine whether those women who were relatively more active for their age had a strength advantage that could be functionally significant.
Thus the purpose of this study was to investigate the magnitude and the rate of age-associated strength reductions in Australian independent urban-dwelling women and the relationship among different muscle groups, limb dominance, and physical activity level. We expected that there would be a difference in the rate and the magnitude of strength reduction among muscle groups with increasing age and that strength would be greater in the preferred (dominant) limb and also in women who were relatively more active for their age.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionReferences |
|---|
The women were required to attend a university laboratory for one session of testing. The experimental procedures of the study were as follows.
Anthropometric Measurements
Height and body weight were measured and used to calculate body mass index (BMI). The percentage of body fat was estimated with Siri's equation (23) along with an estimation of body density. Body density was calculated with four skinfold thickness measurements and regression equations according to Durnin and Womersley (24). The four skinfold thicknesses were measured to the nearest millimeter by an experienced anthropometrist who used skinfold calipers (Harpenden Caliper, British Indicators Ltd) at the following sites on the right-hand side of the body: biceps, triceps, suprascapular, and suprailiac. Lean body mass (LBM) was calculated by subtraction of the estimated body fat weight from the gross body weight. Lean mass cross-sectional area (CSA) of the midthigh and thickest portion of the lower leg were calculated by use of girths and subtraction of the estimated subcutaneous fat calculated from skinfolds.
Physical Activity Rating
Current level of physical activity was assessed with the National Heart Foundation Prevalence Risk Study Questionnaire (22). By use of this information, each subject was scored on a 6-point physical scale. The rating was as follows:
Maximal Voluntary Isometric Strength
The maximal voluntary strength of the KEs, PFs, and HG was assessed on both the dominant and the nondominant sides in random order. Each subject performed a minimum of three verbally encouraged maximal voluntary contractions (MVCs) for each muscle group, with one minute rest between contractions. After the third MVC, the contractions were repeated until it was apparent that the subject was unable to improve further. Each MVC lasted 2–5 seconds and the largest recorded MVC was used for analysis.
Force transducers (X Tran 1000 N and 2000 N; Applied Measurement Technology, Australia) set within steel frameworks designed for each muscle group, recorded the isometric force of the KE, PF, and HG. The electrical signal from each force transducer was amplified (Applied Measurement Technology) and digitized by a DT2801A analog-to-digital card (Data Translation, Malboro, USA). Data was sampled at a frequency of 1000 Hz and analyzed on computer.
The isometric force of the KE was measured while the subject was seated in a steel-framed chair with a hip angle of 90 deg and a knee joint angle of 60 deg from full knee extension. Sixty degrees from full knee extension corresponds to the angle for optimal force on the KE length–force curve (25). The PF MVC was measured with the foot positioned at a joint angle of 10 deg dorsiflexion, corresponding to the optimal region on the length–tension curve for the PF (5–15-deg dorsiflexion) (26). The subject performed the HG MVC while in a standing position with arm straightened by their side.
Reliability of MVC
The test–retest reliability of MVC performance was determined on 107 of the subject population ranging in age between 20 and 82 years [mean ± standard error (SE), age 52.0 ± 1.4 years; height 162.7 ± 0.6 cm; weight 64.8 ± 0.9 kg]. The women returned to the laboratory 1–2 weeks after their first visit to perform a retest MVC of the dominant and nondominant KE, PF, and HG. A minimum of 10 women were retested from each decade up to 80 years.
Statistical Analysis
The subjects were grouped into 5-year age brackets for analysis (with the exception of the 80–89-year-old subjects). The main effects of age, physical activity, and limb dominance were analyzed by trend analysis within 13 x 2 x 2 (age group x activity level x dominance) analysis of variance (ANOVA) (27) conducted across the age groups. The women in each of the 5-year age groups were divided into those subjects who were relatively active (ACT: equal to or above the mean activity rating of their age group) or inactive for their age (INACT: activity rating below the mean score of their age group). The main effect of muscle group (KE, PF, and HG) on strength across age groups was analyzed by trend analysis within ANOVA (13 x 3) by use of relative strength scores (strength as a percentage of 20–24-year age group of the dominant and the nondominant sides averaged). Linear and quadratic regression analysis was used to determine the line of best fit when the linear/quadratic trend component with age was significant for a dependent variable. Pearson-product moment correlations were used to assess the association among a range of variables including age, physical characteristics, and strength of the muscle groups. Stepwise linear multiple-regression models were used to determine the predictive nature and contribution of age, physical activity levels, and anthropometric measures to the total variation of muscle strength. For all results, data are expressed as mean ± SE and significance accepted at p < .05.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionReferences |
|---|
|
A comparison of physical activity was based on whether the subject had participated in any vigorous exercise within the two weeks prior to answering the questionnaire (22). The proportion of women aged 20–69 years who had participated in vigorous exercise from the present study sample and the National Heart Foundation study was 26.1% 
and 30.2% 
, respectively (Fig. 1). Of the 10 age group categories from the present study, 9 were within 5.8% of the National Heart Foundation Study (22) whereas the 35–39-year-old group deviated by 8.4%.
|
|
and height decreased linearly 
across age groups. BMI increased across age categories in a linear trend 
. There was no significant difference between the BMI scores of ACT and INACT subjects.
LBM..--
Estimated LBM decreased with age in a linear trend 
F
(Table 1 ). ACT subjects had significantly greater amounts of estimated LBM (43.3 ± 0.5 kg) than INACT subjects (41.3 ± 0.5 kg) F
.
Percentage of body fat..--
Estimated percentage of body fat increased in a linear trend across age groups F
with a significant quadratic trend component F
representing a slowing down in the rate of increase. ACT subjects had significantly less estimated body fat (33.4% ± 0.7%) than INACT subjects (35.7% ± 0.5%) F
. There was no significant interaction between any trend components with age and physical activity, indicating that the shapes of the age trend functions for body fat did not differ between activity groups.
CSA of midthigh (muscle and bone)..--
Estimated midthigh muscle and bone CSA (dominant and nondominant legs averaged) were negatively associated with age 
and positively associated with KE MVC 
and physical activity 
. Midthigh lean mass CSA decreased with age in a linear trend [dominant F
and nondominant F
. ACT women had significantly larger lean mass CSA than INACT women in both the dominant thigh [ACT, 139.6 ± 2.0 cm2; and INACT, 125.8 ± 1.8 cm2; F
and nondominant thigh [ACT, 138.8 ± 2.0 cm2; and INACT, 124.1 ± 1.8 cm2; F
. There was no significant interaction between the trend components with age and physical activity in thigh CSA, indicating that the linear downward slope with age was similar for both ACT and INACT groups.
CSA of lean mass lower leg (muscle and bone).--
Estimated lower-leg lean mass CSAs (dominant and nondominant legs averaged) were significantly associated with age 
, absolute strength of the PF MVC 
, and physical activity 
.
The dominant and the nondominant lower-leg lean mass CSAs decreased with age in a linear trend, dominant 
and nondominant 
. ACT women had significantly larger lean mass CSA than INACT women in both the dominant lower leg [ACT, 76.4 ± 1.0 cm2; and INACT, 69.0 ± 0.9 cm2; 
and nondominant thigh [ACT, 74.8 ± 1.0 cm2; and INACT, 68.0 ± 0.9 cm2; 
. There was no significant interaction between the linear trend components with age for physical activity in either lower-leg CSA, indicating that the difference between the ACT and the INACT lower-leg lean mass CSA remained consistent across age groups in both legs.
Reliability of MVC
The test–retest reliability of MVC performance was determined on 107 of the women subjects who returned to the laboratory after 1–2 weeks for a second session of MVC testing. Intraclass correlations (ICCs) of the test–retest reliability of the MVC of the dominant and nondominant KE, PF, and HG ranged between 
and 0.95. The coefficient of variation (CV) for test–retest of each muscle group MVC was calculated from the largest MVC performed by each woman during each of the two testing sessions 
and then the CVs of the 107 women were averaged. The mean percent CVs for the dominant and the nondominant KE, PF, and HG ranged between 4.4% and 7.1%.
The effect of age on the test–retest reliability of MVC performance was determined by an independent t test to compare the CVs of subjects aged 20–49 years (young, n 
with the CVs of those aged 50–82 years 
. There was no difference in the CV of MVC between the young and the older age group for the dominant KE (5.4% and 6.3%, respectively, p > .05), nondominant KE (8.7% and 6.1%, respectively, p > .05), dominant PF (5.2% and 6.9%, respectively, p > .05), nondominant PF (4.1% and 5.5%, respectively, p > .05), and the dominant HG (4.9% and 4.9%, respectively, p > .05). In the nondominant HG the older age group (50–82 years old) had a significantly higher CV (5.1%) compared with that of the younger age group (3.4%, p < .05).
Strength as a Function of Muscle Group with Age
The strengths of the dominant and the nondominant limbs were averaged for this analysis. The single main effect of muscle strength across age categories was a downward linear trend 
. The linear reduction of the MVC in each muscle group of women aged between 20 and 89 years was KE, 0.93% per year; PF, 0.74% per year; and HG, 0.62% per year. These linear slopes for MVC were significantly different between the KE and the PF 
, KE and HG 
, and PF and HG 
. For the PF and the HG there was a small but significant quadratic component with age (refer to Fig. 2). Fig. 3 shows the range (strongest and weakest woman) and group means of the MVC for each 5-year age group for the KE, PF, and HG. The best-fit regression equations for each of the muscle groups were
 |
 |
 |
|
|
 |
Age, estimated CSA of muscle and bone of the lower leg, height, body fat, and physical activity levels accounted for 51% (adjusted r2) of the total variation in MVC of the PF:
 |
Age, height, physical activity levels, and LBM accounted for 52% (adjusted r2) of the total variation in the absolute strength of the HG:
 |
Strength as a Function of Age and Physical Activity
Physical activity levels and age..--
Older women were significantly less active than younger women. There was a linear trend downward with age from a mean activity rating of 4.1 ± 0.4 at 20–24 years to 1.7 ± 0.4 at 80–89 years 
.
Strength and physical activity.--
The absolute muscle strength of ACT women was significantly greater in all muscle groups (KE, PF, and HG) when compared with the INACT women (refer to Fig. 4). The ACT women were significantly stronger than the INACT women when absolute strength of the leg muscles was normalized for body weight (KE and PF) and normalized for LBM in all three muscle groups (Table 3 ). There was no interaction in linear or quadratic trends between the ACT and the INACT for the KE, PF, or HG. Linear regression analysis resulted in the best line of fit for the ACT and the INACT KE MVCs with increased age. Quadratic regression analysis resulted in the best line of fit for both the ACT and the INACT PF and HG MVCs with age. The regression equations were as follows:
 |
|
|
, PF F
and HG F
(refer to Table 4 ). The dominance advantage remained significant when the MVC of each muscle group was normalized to body weight (KE and PF only) and LBM (all muscle groups). The dominance advantage also remained significant when the KE MVC was normalized for midthigh lean mass CSA (KE/CSA) but was eliminated for the PF/CSA.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionReferences |
|---|
Stepwise regression analysis indicated that the age of the individual was the most potent predictor of limb strength. Even the strongest women in each age group were unable to maintain the strength of the young women (Fig. 4). Although many studies have demonstrated a large decrement of strength in older men and women for the KE (10)(28)(29)(30)(31)(32)(33)(34), PF (11)(16)(18)(35)(36), and HG (12)(14)(15)(32)(37)(38)(39)(40), we were able to report a comparison of the magnitude of weakness for these muscle groups in women aged up to 89 years (Table 5 ).
|
The KE showed the most consistent and rapid decline in strength across the age groups from 25 years of age on, and this was a linear trend. Other cross-sectional studies have reported the decline in KE strength with increased age to be curvilinear for men (10)(34) and women (34)(42), with a rapid reduction in strength post-50 years of age. Given that the activity levels of our sample studied closely matched that of the National Heart Foundation Study (22), which randomly sampled 4750 women, we are confident that the magnitude and the rate of strength decrements across the age groups were representative of the healthy independent-dwelling woman in urban Australia.
The reductions in strength of the PF and HG across age groups was curvilinear in trend, with a relative preservation of strength before a more rapid decline in women aged 50 years and over. In men, Bemben and colleagues (16) also found that the reductions in strength across age groups of PF and the finger flexors were explained by linear and quadratic trends. These results are also consistent with other studies reporting the trend of strength decline for the PF (11) and HG (12). The relative preservation of the HG and PF strength in the middle years of life compared with the KE may be due to the varying degree of muscle group use and/or their fiber-type composition. At all ages in life, it is essential for use of the hands to be maintained because of their important role during daily tasks, independent of whether the individual is mobile or not. The age-associated decline in physical activity levels shown in this study indicated a reduction in mobility and therefore reduction in use of the lower limbs. Further, the soleus, which is one of the PF muscles, is often in a state of low activation even during standing and has a relatively high number of type I (slow) fibers (43)(44). Type I fibers are not subject to the same degree of age-associated atrophy as type II (fast) fibers (45)(46).
Reductions in lower-limb strength and power experienced by many sedentary older women have been associated with a loss of mobility in activities such as ambulation, rising from a chair, and stair climbing (3)(5)(7)(47). There is a threshold of minimum strength that is required for performing daily living tasks for activities such as walking speed (4) and rising from a chair (6)(9). The torque required for rising independently from a knee-height chair has been estimated to be approximately 120 Nm (both legs) for healthy young and older men and women (9)(48). Using this estimation of the KE torque, we calculated that 5% of the women from this present study would have been completely unable to perform the given task successfully because their maximum KE torque was less than that required for rising from a chair. This is a conservative estimate of the number of women who would truly have difficulty in rising from a chair independently because (i) the KE torques required for rising from a chair were measured at chair stand takeoff when the KEs were at an angle outside the optimal range of force development (9)(48), whereas in this present study we measured the maximal strength at 60 deg from full knee extension at which higher torques are able to be developed; (ii) in real life many chairs are lower than knee height and it is well established that the required torque of the KE increases as the chair height lowers (9); and finally (iii) there are secondary factors including feet width, ankle proprioception, and range of movement that may confound the efforts of a woman who is near the torque threshold required for successfully rising from a chair independently (9).
Physically Active Women Have Delayed Impairment of Muscle Strength with Age
There was a progressive decline in the reported physical activity levels in women of increasing age. This is consistent with other studies reporting that older people in Western cultures engage in fewer regular exercise sessions and spontaneous activities and expend less energy than younger individuals (14)(22)(39)(40)(49).
The reduced participation of women in physical activity was associated with different body compositions and strength levels. The ACT women had significantly less body fat (2.3%), more LBM (5%), and a larger muscle and bone CSA of the thigh and lower leg compared with INACT women, which was consistent across all age groups. The different body compositions between the ACT and the INACT women could prove advantageous for ACT women. The ACT women had a larger amount of muscle mass to move a similar body mass. Therefore, during a given physical task, the ACT women would be able to function at a relatively lower work capacity than INACT women of the same age.
The ACT women had a marked strength advantage at all ages when strength was absolute or normalized for body weight, LBM, or for the lean mass CSA of the legs. The mean relative strength advantage of the ACT women was 22.9%, 20.5%, and 15.4% for the KE, PF, and HG MVC, respectively, compared with that of the INACT women. This consistent additive strength in the ACT women across age groups resulted in considerable delay of age-associated muscle weakness compared with that of INACT women of the same age. Interestingly, because the women were divided into the ACT and the INACT group relative to the mean activity score of their age group, the strength advantages in the active women were significant despite the older active women's being less active than many of the young inactive women. This study does not establish a cause–effect relationship between activity levels and strength; however, we postulate that involvement in even minimal increases in habitual daily activity would be a stimulus for relative maintenance of strength levels into old age.
The additive strength in ACT women is functionally significant particularly for weaker older women. For example, if the requirement of a given task of the KE was 100 Nm an ACT woman would theoretically be able to perform that task until she was 73 years of age (calculated from the predictive regression equations of the KE strength in ACT and INACT women). However, an INACT woman would be limited to lifting that load until she was 63 years of age. The ACT woman would have a 10-year advantage over those who were less active for their age. Thus the strength advantage of the physically active women could delay functional impairment of mobility in older persons. The inability to perform tasks of daily living may have even greater significance in frail and dependent older women who were not included in this study. In recent years greater attention has been focused on the functional benefits of strength recovery in very frail and very weak older adults through strength training interventions (5)(50).
Strength with Age and Limb Dominance
The effect of dominance (limb preference) was common to all muscle groups but greatest in the HG (6%) compared with the KE (5%) and the PF (2%). The dominance difference in the HG was similar to that found by Bassey and Harries (40) who reported a 6% difference between the right and the left HG in women aged 65 years and over. Although the dominance effect in all muscle groups was statistically significant (because of the consistent dominance across all age groups), the effect was relatively small, particularly in the PF. The small difference in limb dominance strength for all muscle groups may have minimal functional consequences for an individual, particularly in comparison with the effect of age and physical activity. However, the dominance difference is consistent with the concept of more use being associated with greater strength.
|
Conclusion |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionReferences |
|---|
Received December 15, 1998
Accepted December 7, 1999
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y.-H. Pua Allometric Analysis of Physical Performance Measures in Older Adults Physical Therapy, September 1, 2006; 86(9): 1263 - 1270. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Hunter, A. Critchlow, and R. M. Enoka Influence of aging on sex differences in muscle fatigability J Appl Physiol, November 1, 2004; 97(5): 1723 - 1732. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| All GSA journals | The Gerontologist |
| Journals of Gerontology Series B: Psychological Sciences and Social Sciences | |
and the present study 
who had participated in vigorous activity within the two weeks before the physical activity questionnaire.
. Shown are the mean (±standard error) of age categories.
and inactive 
women in 5-year age categories ( p < .0001).