From the perspective of feasibility, our results indicate that the classical and hypothetical 1-RM testing procedure can be well applied in breast and prostate CS following cancer treatment. However, strength testing results vary between the three investigated methods with the occurrence of over- as well as underestimation of patients’ strength performance depending on training machines. Further, the NOR prediction accuracy of all three strength testing procedures seems to be very poor for all tested strength training machines and intensity regimens.
RT has become a central pillar in the supportive care of cancer patients in the last two decades. Studies show impressive effectivity regarding relevant clinical outcomes like chemotherapy compliance rates, onset of lymphedema as well as quality of life, fatigue, distress and functional well-being and RT is therefore recommended and prescribed in various exercise oncology guidelines [5–7]. Consequently, a variety of research groups has performed projects to enhance the application range and quality of RT regimens as well as to test different RT intensity protocols in CS. With regard to the application range, the field of RT research in CS has moved from the “classical” breast CS treated with curative intent [22] to studies that were enrolling cachectic head and neck [23], pancreatic [24], advanced renal cell [25] or lung CS [26] as well as CS with unstable bone metastasis [27]. Other studies have successfully tested RT regimens not following the “classical” progressive hypertrophy RT approach (2–3 sets with 12 reps), but a maximal strength training protocol with 4x4 repetitions of dynamic leg press at approximately 90% of the 1-RM twice a week for 12 weeks in early stage breast CS [28]. On the other end, also gentle strength training approaches with 50% of the 1-RM just once a week for about six months were investigated [29]. Other studies also used intensity-varying approaches, like daily undulating training protocols (e.g., high intensity on Mondays, moderate on Wednesdays and again high intensity on Fridays) shown to be feasible and effective with regard to various clinically relevant endpoints [30, 31].
However, as already mentioned in the study descriptions before, the different studies used various 1-RM/h1-RM testing procedures to determine training intensities as well as to evaluate the efficacy of the invested RT intervention. In light of our results, it has to be questioned whether it is acceptable to incorporate RT studies with different 1-RM/h1-RM testing procedures in the same meta-analysis approach that might lead to invalid or inaccurate conclusions. Obviously, these questions are relevant to all RT studies and are not a unique problem of cancer patient populations. Furthermore, the question arises whether the participants in the above-mentioned studies really trained at the intended intensities. This is obviously a key question since a higher stress than normal must occur for fitness to improve [4]. Since the deviations between the h1-RM and the 1-RM results occurred in both directions in the present study, training weights could also turn out too high which bears the risk of overstraining the patients.
Generally, 1-RM test procedures were evaluated with regard to test–retest reliability and show mostly good to excellent results, regardless of e.g., previous RT experience, sex, and age of the participants [32]. However, patient cohorts are underrepresented in this type of research with only 4 out 32 studies having patients with disease conditions involved and CS are completely missing. There are studies showing in multiple following testing sessions (incorporating sufficient recovery time between sessions) that 1-RM performance significantly increased from the first to the last (4th) testing session by about 10–17% [33, 34]. These findings support our observation of patients being able to perform more repetitions in the first training session after 1-RM testing than expected by the prediction formulas which has been already shown for healthy older adults [35]. However, studies investigating such research questions in CS are currently missing. To the best of our knowledge, there is only one study that investigated the test-retest reliability of 10-RM tests for the leg press and bench press in breast CS [36]. The authors report a high to very high rate of reliability between the tests for both strength machines (ICC of 0.94 and 0.98, respectively). Even though these results contrast with the above-mentioned results from other studies, one has to keep in mind that the two test procedures (1-RM and 10-RM) are only comparable with each other to a limited extent.
Bringing the existing knowledge with our observation together, it has to be suggested that (I) h1-RM tests lead to errant maximal values with the occurrence of over- as well as underestimation of maximal values, (II) the NOR at given percentages of maximal values derived from 1-RM and h1-RM tests varies extremely between individuals. This implies that training intensities of the till to date published RT studies in oncology might be inaccurate regarding the intended training intensities when training intensities were prescribed as %1-RM/%h1-RM. Studies in the RT research area in non-cancer populations have partly addressed these questions with non-satisfying results mainly in accordance with our findings. For example, a study published in the 1990s tested 91 participants to determine the NOR they could perform at 40, 60, and 80%1-RM on various RT machines [37]. They observed large variations in the NOR that the participants were actually able to perform at the different intensities. Based on their findings the authors concluded that a given %1-RM will not always result in the same NOR. Interestingly, this finding was neither influced by gender nor by training status [37, 38] and also age does not seem to play an important role [39]. Another study undelines these findings, further showing that this phenomenon is true for all existing testing methods and equation procedures, and is not just limited to one of them [40]. According to our findings, the variations in the NOR tend to be greater for larger muscle groups of the lower extremity than those of the upper extremity [38].
Given the findings of our study as well as the current discussion in the field about how to optimize RT in general and in particular for CS, one aspect remains crucial: Reporting. It is well known from exercise oncology research that description and in particular reporting of exercise regimens need to be improved [41, 42]. Therefore, a variety of researchers have focused on comprehensive methods to describe and report RT adequately with a special focus on the cancer domain [43–45] as well as in general [46]. All approaches mentioned, move the field forward by suggesting relevant reporting parameters like objective volume load or velocity documentation or subjectively perceived intensity reporting. In addition, the cancer specific recommendations should incorporate reporting variables which are similar to medical/drug research outcomes in oncology, like relative exercise dose intensity, total cumulative planned and completed dose or dose modification, comparable to Schluter et al. [31]. Two implications can be drawn from the mentioned paper: (I) To be able to draw solid conclusions from RT trials concerning the effectivity and efficacy, a comprehensive approach for reporting RT regimens is mandatory; (II) No matter which additional method or recommendation will be used, it is of great importance to find a consensus regarding how these different reporting strategies can be best implemented and integrated to complement each other.
Our study is to our knowledge the first that addresses questions about strength test prediction accuracy and comparability for intensity prescription of RT in CS. However, there are limitations that have to be mentioned. Due to the study design with multiple 1-RM tests as well as multiple training sets with different intensities, there is a probability of biased findings due to the effect of increasing muscular fatigue with increasing number of sets and tests. We accounted for that by randomizing the order of the 1-RM testing procedures as well as the order of the intensity specifications the patients were asked to follow within one set. Due to the relatively low number of participants and to the nature of our research questions, we do have partly skewed data. However, we accounted for that by only applying non-parametric testing procedures within the analysis.
In conclusion, our study shows that from a feasibility perspective various 1-RM testing procedures can be applied in breast and prostate CS after acute cancer treatment and are well tolerated. Nevertheless, the finding that strength testing results vary largely between 1-RM procedures and over- as well as underestimate patients’ strength depending on which method is used and which muscle group is tested, limits the comparability of studies using different 1-RM testing procedures. Findings from other studies suggest that the extent of misestimation of h1-RM procedures largely depends on the NOR achieved in the test [47–52]. Therefore, it can be assumed that even intraindividual values reported in studies are inaccurate. Consequently, it has to be critically asked whether 1-RM/h1-RM testing procedures are adequate methods to be used to evaluate resistance-training efficacy in CS, and whether data from different 1-RM procedures are acceptable to be used for review and meta-analysis purposes. Future studies should invest this cross-sectional phenomenon on a longitudinal perspective to elucidate the potential problem of different non-comparable 1-RM procedures through the RT intervention period. Furthermore, given that the h1-RM methods showed the largest accuracy of repetitions for TM12, it can be concluded that these methods may be most suitable for intensity prescription of exercise sessions with similar numbers of repetitions to the h1-RM test. More simply put, if you intend to exercise at e.g., 10-RM a 10-RM test might be more suitable for determining the training weights. This would also be a safer alternative to the 1-RM test, which given the occurrence of SAE in the past, should better be replaced by a safer testing procedure for clinical populations.
Having discussed accuracy and methodological aspects of strength testing and intensity prescription of RT before, it has to be noted that there is quite a lot of discussion in the area of RT research about whether RT intensity is the central key variable when it comes to RT efficacy. A recent meta-analysis from studies performed with healthy participants shows that muscle hypertrophy can be reached by applying low- as well as high-load RT protocols [53]. However, maximal strength benefits were significantly greater in favor of high- vs. low-load training. This is mostly in line with recent findings from Lopez et al. in the cancer area. They found that the prescribed volume was inversely associated with gains in muscle strength, although there was no relationship between RT intensity and strength gains [54]. The authors therefore conclude that low volume RT might be a viable approach for breast CS to gain benefits from RT regardless of the training intensity. The possible opportunities again, stress the importance of accurate intensity assessment and prediction to set up an adequate resistance training regimen.
Regarding intensity prescription, given the fact that the accuracy of %1-RM/h1-RM (prediction of repetitions a patient can perform with a certain weight) was mostly inadequate, raises the question whether this method is adequate for intensity prescription in CS. From a practical point of view, exercise trainers should not rely on 1-RM/h1-RM testing procedure results and closely monitor their patients predominately through initial training weeks to be sure that the intended exercise intensity is reached. A more practical approach would be to directly approach the desired weight (gradually increase or decrease the weight until the patient can lift the weight the intended number of times). Prescribing the actual NOR to dictate the intensity and not vice versa is an approach which has already been advocated by other authors [55] and is also part of the Australian recommendations for CS [56]. Our results together with those from previous studies suggest that alternative methods than the ones commonly used for strength testing and RT prescription may be more suitable for CS, which should be investigated in future studies