Our program of resistance + impact exercise produced statistically significant improvements in upper and lower body maximal strength compared to a control program of low-intensity stretching in BCS who were older than 50 years of age and postmenopausal at diagnosis. However, the POWIR program did not appear to improve either objective or subjective measures of physical function more than the control condition. In fact, both groups improved significantly over time in a measure of mobility, e.g., gait speed. Women were able to perform either study program as prescribed. Within the POWIR program, higher attendance was associated with greater improvements in maximal upper and lower body strength.
Our study is the first to report on a resistance training program specifically for women who were older when they were diagnosed and treated for breast cancer. Approximately 85% of women who receive a first diagnosis of breast cancer are aged 50 and over, thus older women constitute the largest group of BCS [58
]. While other resistance training trials have included older women as part of a broader age range of BCS [25
], these studies cannot evaluate the specific capacity of older BCS to tolerate and respond to rehabilitative interventions. We found that older BCS could tolerate a moderate-vigorous resistance training program as evidenced by compliance to the prescribed training program and the absence of program-related injuries. We are also among the first to examine physical function using objective measures rather than self-report [59
]. Objective measures of physical function, particularly those that can capture different domains of function such as balance and mobility, overcome limitations of self-reported physical function because objective tests can indicate which body systems underlie limitations and detect declines before a woman recognizes a change in her abilities [8
The notable limitations of our study were the modest size of the sample and generalizability of the program. Our study may not generalize to the broader population of BCS, since women who dropped out of the program had some different characteristics than women who did not. Women who withdrew from the program were closer to diagnosis and reported more difficulties with activities of daily living and lower mental health scores. It is unlikely that these women dropped out because they could not tolerate the POWIR program since a greater proportion dropped out of the control program due to poor health than the POWIR program (); however, dropouts from our study may represent a subgroup of older BCS that have different exercise preferences and warrant further attention. Though our choice against a non-exercise control group may also be perceived as a limitation, we view it as a strength because an exercise placebo group reduces the potential for unequal attrition among participants allocated to a non-exercise group and also keeps the level of attention from instructors and peers similar across groups. We also felt that withholding exercise completely from BCS would be unethical given the recommendations that cancer survivors avoid inactivity [34
]. Stretching exercise has been used as a control condition for resistance training studies in older adults and did not improve strength or bone density [60
], but has been shown to provide other benefits such as increased range of motion [64
] that could improve mobility [65
]. In fact, the main effects of time on gait speed in our study () indicate that even low intensity stretching might have had a small effect on function outcomes in BCS, though neither group could be compared to a no exercise condition.
Women in the POWIR program improved upper and lower body muscle strength as evidenced by 12% and 20% increases in 1-RM bench and leg press performance, respectively. Our findings agree with other resistance training trials in BCS but the magnitude of strength increases was slightly lower among our older sample [25
]. Differences between those studies and ours may be attributed in part to the different age ranges of participants, but could also reflect differences in strength testing protocols and training programs. Resistance training trials in older adults without cancer consistently report significant strength gains, but those studies comparing the degree of change between young and old participants report similar improvements or slightly less improvement in old versus young adults [66
]. Within our sample of BCS ranging in age from 53–83 years old, age did not moderate significant group differences over time suggesting that resistance exercise benefits are similar across this older age span.
In contrast to Twiss et al who reported improvements in dynamic balance after strength training in BCS [24
], we did not find differences between POWIR and FLEX on simple static balance tests. We expected that the POWIR program would improve balance based on our prior reports of stability improvements with this program in women without cancer [27
] and because muscle strength contributes to balance control [13
]. We used a simple clinical test of static balance, the one-leg stance test, whereas other studies used laboratory balance tests [27
] or more complex clinical tests of dynamic balance, e.g., timed backward tandem walk [24
]. Our simple clinical test may have lacked the specificity and sensitivity to detect balance changes from strength training, particularly attributes most relevant to falls [47
], and future trials should select the most appropriate measurement tool for their intervention and population. Another reason that we may have failed to find group differences on either objective or perceived self-report physical function is that we did not select women for low baseline function nor did we exclude aerobically active women (). We also excluded women who could not engage in moderate-vigorous intensity resistance training sufficient to improve bone health and thus may have inadvertently excluded women who might have been more likely to improve function because of low baseline fitness [67
]. Average baseline values for usual gait speed, handgrip strength and the PPB among our participants () were at or above values reported in cohort studies of well functioning, community-dwelling older adults [68
Campbell et al. have urged investigators to thoroughly report compliance and adherence to prescribed exercise interventions in cancer survivors so that specific evidence-based exercise prescriptions, administered by healthcare professionals, could be available for BCS [67
]. In addition to reporting participant compliance to our prescribed protocol, we evaluated the influence of adherence on study outcomes. Women who came to class an average of 1 or more times per week had significantly greater improvements in maximal upper and lower body strength measures than women attending less often (). These findings are consistent with those of Taafe et al and Di Franco et al who reported significant maximal strength gains among older adults who participated in moderate-vigorous intensity resistance training 1 day per week compared to inactive controls [71
]. However, while Taafe and Di Franco reported that strength gains were no better among older adults who resistance trained 2 or 3 days per week compared to 1, our results indicate that adherence rates above an average of 1 time per week promoted greater upper body strength gains. Our study was in women only and the others included men, thus it is possible that in women the upper body may be more responsive to further increases in training than the lower body. This body region may be more sensitive to increased frequency of training in women because faster declines in upper extremity strength than lower extremity strength across the 6th to 9th decades have been reported for community-dwelling women [72
]. For these older BCS, an average participation rate equal to one day per week was sufficient to cause gains in muscle strength and this frequency may meet the preferences of older BCS for gradual increases in the level of exercise over time [73
]. Our data suggest, though, that even greater gains in upper extremity strength could be achieved by progressing toward ACSM recommendations for twice-weekly resistance training for older adults [36
Given the known roles of muscle weakness in the etiology of falls and functional decline, improvements in muscle strength could translate to lower fall and disability risk in older BCS. When considering the multi-factorial nature of falls, muscle weakness ranks as the lead risk factor in older adults [11
]. In addition to fall risk, muscle weakness is also a precursor to disability onset because it precedes functional limitations that can lead to loss of independence [8
]. Muscle weakness is a consistent predictor of ADL disability among community-dwelling older adults [75
]. Since we did not measure falls nor disability in our study, we cannot assume that our older BCS who gained strength will fall less or remain independent longer; however, future controlled exercise trials that track falls and disability in adequately powered samples are warranted. In 2011, the first generation of baby boomers will reach 65, and the aging of this generation will contribute to a projected doubling of cancer survivors by 2050 [76
]. Developing safe, effective and translatable interventions that optimize function and quality of life specific to older cancer survivors is an important area for future work.