Resistance Training Programme Variables and  Strength Recommendations

Image by Anastase Maragos

ACSM 2002 Set Volume

ACSM 2009


ACSM 2002 Frequency 

ACSM 2009 Frequency

ACSM 2002


ACSM 2009 


ACSM 2002 Inter-set Recovery

ACSM 2009 

Exercise Order 

Health and Fitness Recommendations  and Muscular Strength Development

Over the last few decades, scientific training principles have prevailed over pseudoscience and these principles are now fundamental in the development of valid RT prescription for different population ranges (Bird et al., 2005). RT is an exercise modality that is frequently recommended for different population groups (clinical, fitness and athlete) and it has been reported to improve physical health and well-being (ACSM, 2002; Kraemer, Ratamess and French, 2002). The associated health benefits for individuals performing RT cannot be underestimated. These benefits go beyond that of muscular strength and athletic performance, with the application towards muscular tension provided by RT alters and enhances an individual's well-being and longevity. 


The health benefits of RT are predominantly used as a preventative training modality or as a countermeasure to conditions where muscle weakness compromise an individual's functional capacity (muscle atrophy, sarcopenia, musculoskeletal disorders or injury). It is necessary to recognise that without the body's ability to contract muscles, we could not walk, lift, breathe, digest, or perform essential biological functions. A well-conditioned muscular system enables a physically active lifestyle and enhances health factors that may reduce the risk of degenerative diseases and medical complications (ACSM, 2009). Recent evidence has suggested that muscle mass index is a stronger predictor of life expectancy than body mass index and offers additional support to the importance of muscular strength and muscle mass (Srikanthan and Karlamangla, 2014). However, during the early 20th century, misconceptions surrounded RT with the belief that it was detrimental to physical health.

Formation of Health and Fitness Organisations

The development of scientific evidence on RT since the 1940s has progressively increased with a further upsurge in magnitude throughout the 1960s and 1970s. Researchers in the 1960s refined the seminal research performed by Delorme (1945) with systematic manipulation of different RT variables from longitudinal studies (Berger, 1962; Hellebrant and Houtz 1956; Berger, 1962; Berger 1965; Capen, 1950; O'Shea, 1966). This scientific evidence and promotion of RT helped transform the landscape with increased participation and understanding of the value of RT towards physical health and well-being. This upsurge in popularity since the early 1970s has now led to national health and fitness organisations recommending that this mode of exercise be incorporated into exercise programs. These organisations included the National Strength and Conditioning Association (NSCA), the American Heart Association (AHA), the American Association for Cardiovascular and Pulmonary Rehabilitation (AACPR), Cardiovascular and Pulmonary Rehabilitation and the American College of Chest Physicians (ACCP), and the American College of Sports Medicine (ACSM) all advocated the use of RT to increase physical health, performance and longevity.

The National Strength and Conditioning Association


Traditionally, strength athletes only performed resistance training (RT) to improve muscle strength, power, hypertrophy, and sports-specific fitness. Unfortunately, the strength programs were established specifically from the applied experiences of the trainer or coach. This generated an explosion of unsupported RT programs that created confusion for those that were prescribed it. Indeed, RT program design was initially more of an art than science, with trainers and coaches modifying the training dose based on restricted knowledge of the physiological responses to exercise. Fortunately, in the 1960s and 1970s, American football recognised the value in RT and hired strength coaches to improve performance. By the 1970s, most strength and power athletes were performing RT. A critical development during this time was the NSCA that spread RT recommendations and attempted to disperse misconceptions and started to distribute practical and scientific evidence throughout the different sports. By the 1980s, most athletes and their coaches embedded RT for almost all sports. From the conception of this professional body, there are now over 30,000 members in 72 countries.


The NSCA have released several books and training recommendations intended for athletic development rather than universal recommendations on RT. For example, Pearson et al., (2000) produced a series of basic guidelines for RT of athletes that allowed coaches to develop athletic performance safely. These recommendations were broader ranging, unlike other organisations with programming designed specifically to match the profile of the athlete and the sport. Unlike other generic training suggestions, the approach to loading and volume is periodised to avoid the potential of over-training syndrome. The NSCA recommendations are positioned around the stress-adaptational responses and sports performance.

The NSCA specified that RT programs for athletes should be developed based on the individual needs of the athlete and specific sport.  The NSCA, unlike other organisations, stated that RT programs should be developed from integrating of scientific knowledge while addressing the practical requirements of the sport. This is unlike other recommendations, as ultimately RT programming [in part] seeks to develop long-term adherence to the individualised program. This allows athletes to have programs that best meet their needs and the sport. Therefore, because of the sport, training approaches and programming draw on a broader range of knowledge and skills typically discussed in other recommendations. Strength trainers and coaches need to be creative while applying scientific knowledge to athletes’ specific needs, which is unlike the general population who have very linear goals (Kraemer, 2006).

Barbell and Kettlebell Weights

The American Heart Association


In 1995, the Centres for Disease Control and Prevention and ACSM published a joint public health recommendation on Physical Activity and Public Health. These recommendations were provided to send a concise and clear public health message concerning sedentary lifestyles and the role physical activity can have on health improvements. These recommendations were supported and endorsed by the AHA, with healthy adults aged between 18-to-65 years encouraged as part of a physical activity routine to perform moderate-intensity aerobic (endurance) physical activity for a minimum of 30-minutes on five days each week or vigorous-intensity aerobic physical activity for a minimum of 20-minutes for three-days.


However, no recommendations were provided for individuals to engage in RT to improve health. In 2007, the updated recommendations included clarifications to the 1995 recommendations with muscle strengthen activities incorporated into the physical activity recommendations (Haskell et al., 2007). Healthy adults were recommended to perform eight-to-ten exercises (weight training, weight-bearing callisthenics) performed on two or more non-consecutive days each week using major muscle groups. For improvements in muscle strength, resistance loading of eight-to-12 repetitions of each exercise be performed until volitional fatigue was encouraged (Pollock et al., 2000).

The American Association for Cardiovascular and Pulmonary Rehabilitation


The AACVPR and the ACCP released a previous set of evidence-based guidelines in 1997 with the inclusion of physical activity to increase recovery. Since then, the ACCP and AACVPR released further recommendations with the addition of a strength training component to a program of pulmonary rehabilitation increases muscle strength and mass (Ries et al., 2007). However, no specific recommendations concerning the required RT dose were provided even though the evidence they produced supported inclusion within a structured exercise program.

The American College of Sports Medicine


The scientific community throughout the 1970s had a limited appreciation towards the benefits of RT towards improving population-level health, except for increasing muscular strength and endurance. Some researchers believed that these functional variables played a limited role in the general population’s health.  Indeed, ACSM (1978) initially produced exercise guidelines for population-level physical activity, stating that individuals need only perform aerobic exercise. This positional statement reflected the health and fitness trends and limited research on RT being performed during the 1970s. This evident lack of scientific research led to the exclusion of RT within exercise guidelines with greater importance placed on cardiorespiratory fitness and body composition. For example, epidemiological research during this period reported a strong relationship between aerobic endurance exercise and the prevention of cardiovascular disease (Fox and Skinner 1964, Kannel, 1970). This led to the significant promotion of aerobic activities to increase individuals V̇𝑂2max and was interpreted to improve physical health (Blair, LaMonte and Nichaman, 2004). Unfortunately, due to RT omission and increased promotion of aerobic exercise, some interpreted that RT was insignificant for improving physical health (Feigenbaum and Pollock, 1999).


In the early 1980s, RT was recognised as a method that could positively affect athletic performance. By the mid-1980s the medical community began to acknowledge the therapeutic value of RT on health-related aspects including low back health, weight management, bone health, and basal metabolic regulation (Feigenbaum and Pollock, 1999). The ACSM in 1990 included within its recommendations that an RT component should be incorporated within the exercise prescription for physical health and development. This acknowledgement of the therapeutic benefits of RT leads to other prominent organisations, including the AACPR, ACCP and AHA, to integrate RT within exercise programmes. This endorsement by other leading health and fitness organisations has helped to establish the ACSM as the main authoritative body of exercise prescription.


Currently, there are large volumes of published information and scientific data on RT with numerous recommendations on how to improve muscular strength across all age classifications that are founded on pre-existing literature (Table 1 and Figure 1). ACSM has published several position statements that provide recommendations for enhancing physical conditioning for specific population groups (novice to athletes). These recommendations include guidance on the resistance loading, training frequency, volume (sets x repetitions), exercise order, and exercise selection. Most novice trainees and newly qualified trainers are typically directed towards these ACSM position statements. These statements are made available via open access, and due to marketing, most individuals accept what has been published as scientifically correct and assume that the evidence has sufficiently filtered through the peer-review process. ACSM states that their guidance is the most authoritative, evidence-based statement issued by ACSM and often considered by some as definitive with suggestions that evidence is extrapolated from a large body of scientific data that provides a substantial burden of proof.

ACSM Position Statements and Training Recommendations 

The ACSM and other health and fitness organisations have provided statements on strength development from as early as 1990 through to the present recommendations. The physical activity recommendations for population-level health now acknowledge the benefits of RT to improve health-related factors including, improved bone density, functional capacity, basal metabolism and back health (Feigenbaum and Pollock 1999). Consequently, health-related recommendations began to include RT within an integrated physical activity program that also included aerobic and other forms of exercise (ACSM, 1990).

The ACSM from the early 1990s emerged as the only plausible organisation that attempted to provide RT recommendations for population-level health. For example, in the ACSM (1998) position statement, “The recommended quantity and quality for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults,” provided recommendations with RT included within the program design. These suggestions encouraged individuals to perform RT with one-set of eight-to-12 repetitions for eight-to-ten exercises, including one exercise for all major muscle groups, and 10-to-15 repetitions for older adults and the infirm (ACSM, 1998). This then leads to a series of recommendations that are diversified by population group and training objectives (ACSM, 2002 and 2009).

In 2002, ACSM produced a position statement that evolved from the previous 1998 statement. These new recommendations now included guidance for those healthy adults that wish to progress their muscular fitness. The purpose of this 2002 position statement was to increase guidance from beginner RT programs to progression models that can apply to novice, intermediate, and advanced trainees. The ACSM (2009) position statement further advanced training recommendations on strength development from previous recommendations. This was generated from further studies including reviews, epidemiological studies, clinical studies, and meta-analyses on various acute RT programme variables. These new recommendations stated that novice trainees perform one-to-three-sets for eight-to-12 repetitions with 60-to-70% 1RM loading and two-to-three-minute recovery. For intermediate trainees, multiple-sets for six-to-12 repetitions at 60-to-70% 1RM and two-to-three-minute recovery. Athletes were directed towards performing multiple-sets with varying repetitions at a loading of 80-to-100% 1RM with four-to-six-minute recovery, in which to avoid over-training.

Table 1. Development of standards, guidelines and position statements regarding strength training for adults

Screenshot 2020-09-11 at 13.55.23.png

Concerns Regarding ACSM Position Statements and Strength Development

The development of effective RT programming for strength development has significantly evolved in terms of scientific understanding compared to the pioneering work of Roux (1895), DeLorme (1945) and Berger (1962). There is now a comprehensive range of scientific literature that can guide the layperson with RT recommendations that are concise and easy to implement. However, this does not imply that current training recommendations for healthy adults are without criticism. For example, the ACSM (2002 and 2009), have been debated for the apparent lack of scientific rigour when considering the acute RT variable recommendations (i.e. load, frequency, volume, type, intensity of effort, exercise order, etc.) (Carpinelli, Otto, and Winett, 2004; Carpinelli, 2009).

In fact, Carpinelli (2009) suggested that if one were to follow the ACSM (2009) guidelines with intentions to attain the desired components of muscular fitness (e.g. strength, hypertrophy, power and endurance) then trainees would need to devote a minimum of 20 hours per week (5 hours per day x 4 days per week) performing resistance exercise. This is a far greater volume, and frequency than organisations recommend (NSCA, 2008) and is potentially over-complicating the suggested requirements for resistance exercise. This is of concern since previous research has identified that perceived difficulty and lack of time are barriers to resistance exercise (Winett, Williams and Davy, 2009; Owen and Bauman, 1992; Ainsworth et al., 2000; Grubbs and Carter, 2002).

The ACSM is the most recognised health and fitness organisation that produces guidance via evidence-based recommendations. This has led to other organisations producing recommendations from scientific evidence derived or in association with ACSM. These position statements circulated by ACSM are freely available via open access, with most trainees, personal trainers and coaches accepting the current evidence as accurate.  The ACSM states that there is a large body of scientific evidence that supports their recommendations for strength development. However, it is debatable whether all included evidence provides enough scientific credence to support the contentions made by the ACSM position statements.


Regrettably, the ACSM is the only plausible health and fitness organisation that has endeavoured to provide RT recommendations for population-level health. Due to the significance of ACSM position statements within the field of exercise prescription, there is a need to examine whether the cited evidence is substantial enough to provide guidance for muscular strength development. This section seeks to appraise and summarise a selection of evidence used to produce recommendations on several acute RT variables (set-volume, loading, frequency, and inter-set recovery). This review of evidence follows on from the excellent work of Carpinelli, Otto and Winett (2004) but focuses only on strength development recommendations. It should be noted that this literature review aims to appraise the body of evidence and not directly condemn the ACSM.

Screenshot 2020-09-11 at 14.03.08.png

Figure 1. Cited studies used to generate ACSM (2002) position statement and recommendations.

Overview of ACSM (2002) recommendations

The ACSM (2002) position statement placed greater emphasis upon considerations made to key programme variables than the previous recommendations. A total of > 85 cited studies were used to provide recommendations on nine key programme variables (Figure 3.0). These recommendations advised that novice trainees perform one-to-three-sets for eight-to-12 repetitions with 60-to-70% 1RM loading with a two-to-three-minute recovery between sets. For intermediate trainees, multiple-sets for six-to-12 repetitions at 70-to-80% 1RM with a two-to-three-minute recovery between sets were recommended. Athletes were instructed to perform multiple-sets for one-to-12 repetitions at a loading of 80-to-100% 1RM with a two-to-three-minute recovery.




The ACSM (2002) position statement reports that multiple-set programming is superior to one-set programmes for strength development. However, closer inspection reveals that several cited studies on muscular strength development do not support these recommendations. Studies cited by ACSM (2002) on set-volume (Table 2 and Table 3) did not provide a credible body of evidence that supports the use of multiple-sets. The recommendations suggested that novice trainees perform one-to-three-sets with both intermediate and advanced athletes advised to complete multiple-sets for strength gains. However, from the 23 studies that are cited on strength development, only four studies (17%) support the use of multiple-sets. A further three studies (13%) cited that support the use of multiple-sets but contained errors within the methodology or data. Intriguingly, 13 studies (57%) that were used to support the recommendations did not support the use of multiple-sets.

Table 2. Summary Findings on Set-volume for Strength Development

Screenshot 2020-09-11 at 14.09.28.png

Table 3. Cited Studies by ACSM (2002) on Set-volume for Strength Development

Screenshot 2020-09-11 at 14.12.29.png
Screenshot 2020-09-11 at 14.14.16.png
Screenshot 2020-09-11 at 14.18.06.png

Training Frequency


The ACSM (2002) position statement cites 18 studies that recommend multiple training sessions are required per week for strength development (Table 4 and Table 5). ACSM (2002) recommended that novice trainees train two-to-three days per week (d.wk-1), with increased frequency for intermediate (two-to-three d.wk-1) and advanced trainees (four-to-six d.wk-1). These recommendations seem conceivable, however, from the 18 cited studies on strength development, only four (22%) cited studies supported the use of multiple training days, with two studies (11%) supporting multiple training days but had errors with the methodology or data. Regrettably, eight studies (44%) did not support the recommendations for weekly training frequency.

Table 4. Summary Findings on Training Frequency for Strength Development

Screenshot 2020-09-11 at 14.27.41.png

Table 5. Cited studies by ACSM (2002) on training frequency for strength development

Screenshot 2020-09-11 at 14.31.03.png
Screenshot 2020-09-11 at 14.32.30.png

Resistance Loading


The ACSM (2002) recommendations on repetition ranges and resistance loading on specific strength outcomes have limited supporting body of evidence (Table 6 and Table 7). The ACSM (2002) position statement suggests that for the untrained or novice trainees perform between eight-to-12 repetitions with a resistance loading of 60-to-70% 1RM for strength development. For intermediate trainees, six-to-12 repetitions at a loading of 70-to-80% 1RM with advanced trainees and competitive athletes advised performing one-to-12 repetitions at a loading of 80-to-100% 1RM. Seventeen studies were cited that support the ACSM (2002) position statement that increased resistance loading with a reduction in repetitions is necessary to increase muscular strength. However, three studies (18%) support the repetition range of eight-to-12, with ten studies (59%) unable to back these recommendations. Furthermore, one study had issues regarding the methodology.

Table 6. Summary Findings on Resistance Loading and Repetition Range

Screenshot 2020-09-11 at 14.39.39.png

Table 7. Cited Studies by ACSM (2002) on Resistance Loading

Screenshot 2020-09-11 at 14.42.25.png
Screenshot 2020-09-11 at 14.43.59.png

Inter-Set Recovery


The ACSM (2002) recommendations claim that the amount of inter-set recovery and rest between resistance exercises has a significant effect on training adaptations (Table 8 and Table 9). The position statement founded from eight studies recommends that a two-to-three-minute recovery period is required between sets and resistance exercises. However, these recommendations do not have a large body of cited evidence to support these recommendations. From the eight studies, seven did not support the inter-set recovery period with one study supporting these recommendations but contains errors within the methodology and data reporting.


Table 8. Summary findings on inter-set recovery for strength development

Screenshot 2020-09-11 at 14.48.00.png

Table 9. Cited Studies by ACSM (2002) on Inter-set Recovery for Strength Development

Screenshot 2020-09-11 at 14.52.31.png
Screenshot 2020-09-11 at 14.53.59.png

Conclusion on ACSM (2002) Position Statement and Cited Evidence


The ACSM (2002) position statement regrettably contains limited supporting evidence for important RT variables (set-volume, loading, frequency, and inter-set recovery) and the application towards strength development. The recommendations provided by ACSM did not provide a substantial body of cited evidence to support the position statement. There appear to be several errors within the 2002 recommendations that create an unreliable position statement. The ACSM argue that the recommendations provided are “the most authoritative, evidence-based…” with critics stating that this position statement may have eluded the peer review process (Carpinelli, Otto, and Winett, 2004). Moreover, ACSM states that ST practices should vary for the novice, intermediate, and advanced trainees; however, these recommendations are unsubstantiated within the 2002 position statement. It is difficult based on the supporting cited evidence to conclude that one population group (untrained vs trained) responds better than others.


Significantly, assertions made on RT programme designs must be established upon the evidence presented from appropriately designed RT studies, with the entire burden of proof, therefore, resting with the authors. Therefore, these recommendations do not meet the standards required for a scientifically established, methodologically robust consensus statement. Readers should be vigilant and are encouraged to examine all cited studies in this or any position or consensus statement. That is read the entire paper and decide if the evidence presented supports the original resistance studies cited as reported data. There is unfortunately limited evidence within this position statement to suggest that the manipulation of set-volume, loading, frequency, or inter-set recovery would elicit superior strength adaptations.


Concerns and Issues Regarding ACSM (2009 and 2011) Recommendations

The most recent ACSM position statement (2011) is an updated version from ACSM 2009. This new statement has over 400 cited scientific publications ranging from scientific reviews, epidemiological studies, clinical studies, meta-analyses, consensus statements, and evidence-based guidelines. ACSM previously published an updated quantity and quality position statement (2009) that endeavoured to advance previous ACSM (2002) recommendations. The claimed purpose of the existing ACSM position statement (2011) was to provide evidence-based direction for health and fitness professionals concerning the framework and modification of exercise programme training variables. However, signposting was given towards the ACSM’s (2009) standpoint if trainees wanted to develop maximal muscular strength. Intriguingly, the authors of this position statement, also referred towards ACSM (2002) stating that a superior framework was provided towards exercise prescription recommendations for novice, intermediate, and advanced trainees physical fitness. Unfortunately, the claimed framework is founded on conjecture and opinions rather than quantifiable scientific evidence.


The ACSM (2009) position statement on strength development is claimed to be established from 117 cited studies that were sourced from scientific reviews, epidemiological studies, clinical studies, and meta-analyses on ten programme variables (Figure 10 and Table 10). Recommendations include that novice trainees perform one-to-three-sets for eight-to-12 repetitions with 60-to-70% 1RM loading and two-to-three-minute recovery. For intermediate trainees, multiple-sets for six-to-12 repetitions at 60-to-70% 1RM and two-to-three-minute recovery. Athletes were advised to perform multiple-sets for varying repetition ranges at a loading of 80-to-100% 1RM with four-to-six-minute recovery. Ironically, the ACSM (2011) position statement differs in the set-volume and recovery period, with no justification for these changes. On the contrary, signposting was given towards the ACSM (2009) recommendations for strength development.


Carpinelli (2009) stated that the ACSM (2009) position statement is similar to the 2002 recommendations, which had inaccurate reporting. The ACSM (2009) and the new position statement (2011) may be unfortunately founded upon the misinterpretation and representation of selective RT studies. Therefore, this section of the literature review seeks to appraise and summarise evidence used to produce the ACSM (2009) position statement in terms of muscular strength recommendations. This follows on from the work of Carpinelli (2009) who questioned the validity of the previous position statements identifying failings in the reported evidence. However, this section focuses entirely on strength development rather than pooling studies to generate general conclusions on muscular functioning. Several specific areas are reviewed from five subsections (set-volume, inter-set recovery, training frequency, the order of exercise, and resistance loading). The rationale for selecting these five training variables is that they are the most frequently modified by trainees to increase maximal strength. It should be stated again that this literature review aims to appraise the body of evidence and not directly condemn the ACSM positional statement.

Screenshot 2020-09-11 at 15.21.18.png

Figure 10. Cited Studies Used to Generate ACSM (2009 and 2011) Position Statement Recommendations

ACSM (2009) Recommendations on Single and Multiple-Sets


The current opinion on the number of sets for developing muscular strength has primarily been indefinable and controversial. This conflict of opinion regarding the optimum number of sets (volume) is derived from the ‘overload principle’. With the belief that to develop muscular strength, trainees must train the muscles beyond their existing capacity in which to yield physiological adaptations. Baechle and Earle (2008) have described training volume as the summation of the total number of repetitions achieved during a single RT session multiplied by the resistance used (kg) in which the muscles are being physiologically overloaded. The ongoing debate concerning whether increased RT volume (multiple-sets) results in more significant strength gains compared to one-set. Research has indicated that multiple-sets elicit greater muscular strength gains than one-set training (Berger, 1962; Borst et al., 2000; Kraemer, 1997; Kraemer et al., 2000). However, the daily RT set-volume needed to increase strength maximally remains equivocal.


Many studies concerning training volume compared the effects of performing one- or three-sets of each exercise per training session on strength increases on trained and untrained subjects (Pollock et al., 1993; Starkey et al., 1996; Hass et al., 2000; Rhea et al., 2003; Paulsen et al., 2003; Kelly et al., 2007; Bottaro et al., 2009; Baker et al., 2013; Sooneste et al., 2013). Studies comparing one- or three-sets in this area have reported inconsistent findings with some favouring multiple-sets for producing increases in strength compared to one-set. This contrasts with several meta-analyses (Rhea et al., 2003; Peterson, Rhea and Alvar, 2004; Wolfe, Lemura and Cole, 2004; Peterson, Rhea and Alvar, 2005; Krieger, 2009; Fröhlich, Emrich and Schmidtbleicher, 2010; Krieger, 2010) that reported support for increased session volume.


The ACSM (2009) cited 13 original studies and two meta-analyses within the evidence statement and recommendations that support the contention of multiple-sets over single-set training (Table 3.1.8). The ACSM position statement claims that multiple-sets are superior to single-set programmes in intermediate and advanced athletes, with novice trainees recommended to perform between one-set and three-sets per exercise. The evidence that supports the recommendations regarding set-volume and muscular strength development is sparse. These recommendations are founded from the previous standpoint with only one additional reference (Wernbom, Augustsson and Thomee, 2007) included giving additional weighing to that of the recommendations provided by ACSM (2009). It is essential to note that the previous standpoints did not have a substantial body of evidence to support such recommendations and with a closer examination revealing that most studies cited do not support the claims in either the 2009 or the 2011 position statement.


Untrained Subjects and Single vs Multiple-Sets


The ACSM (2009) cited the seminal study by Berger (1962) to support of multiple-sets. Nine groups of college-age males (n = 177) completed training, three days per week (d.wk-1) for 12-weeks. Subjects performed the bench press exercise using one of nine combinations of sets and repetitions (one-set x two repetition maximum [RM], one-set x 6RM, one-set x 10RM, two-sets x RM, two-sets x 6RM, two-sets x 10RM, three-sets x 2RM, three-sets x 6RM, or three-sets x 10RM). When Berger (1962) pooled, the nine groups results according to the set number (i.e. one, two or three-sets) it was reported that significant pre-to-post 1RM bench press strength difference occurred in all treatment groups (three-sets = 25.5%, vs two-sets = 22.0%, and one-set = 22.3%). However, there was no reported significance between the one and two set groups and only a negligible difference of 3.2% between three-sets and one-set. Berger reported that seven out of the nine groups had no significant difference in the magnitude of strength, with most outcomes not supporting multiple-set programming.


Intriguingly, the study by Ostrowski et al., (1997) were cited within the evidence statement for novice trainees even though the subjects (n = 27) had between one- and four-years weight training experience and could squat and bench press at least 130% and 100% of their body mass. Subjects were randomly assigned to one of three groups: low volume (three-sets per muscle group per week [one-set per session]); moderate volume (six-sets per muscle group per week [two-sets per session]); or high-volume (12-sets per muscle group per week [four-sets per session]) and trained four d.wk-1. Significant pre-to-post-test 1RM bench press and 1RM squat was reported but did not differ significantly between groups. The increases in 1RM strength in low, moderate and high-volume groups were 10kg, 8kg, and 14kg for the 1RM squat, and 3.6kg, 4.5kg, and 1.6kg for the 1RM bench press. Ostrowski et al., (1997) reported that low volume (three-sets per muscle group per week) is as effective as six or 12-sets for increasing strength when performed over ten-weeks at a frequency of one-day d.wk-1. This study does not, therefore, support the belief that multiple-set training is superior to single-set training, as recommended by ACSM (2009).

Table 9. Cited Studies by ACSM (2002) on Inter-set Recovery for Strength Development

Screenshot 2020-09-11 at 15.27.44.png
Screenshot 2020-09-11 at 15.29.29.png
Screenshot 2020-09-11 at 15.32.21.png

Meta-analytical Evidence on Single vs Multiple-Sets


Regrettably, RT set-volume has historically been an often-debated subject based on subjective recommendations that support the use of multiple-set programming. Evidence cited from meta-analytical studies (Rhea et al., 2003; Peterson, Rhea and Alvar, 2004; Wolfe, Lemura and Cole, 2004; Peterson, Rhea and Alvar, 2005) included within the ACSM (2009) recommendations suggest when ES is aggregated that strength outcomes are more significant with higher volume (multiple-sets) compared to lower volume (one-set) for producing strength gains. However, Winnett (2004) states that most meta-analyses on RT fail to provide clear evidence of the importance of any specific parameter of training or protocol. The ACSM (2009) cited two meta-analyses within the evidence statement and recommendation section (Peterson, Rhea and Alvar, 2004; Peterson, Rhea and Alvar, 2005). Both meta-analyses suggested that multiple-sets increased strength in both untrained and trained subjects.

A meta-analysis by Rhea et al., (2003) on Muscular Strength 


A meta-analysis by Rhea et al., (2003) reported a significant magnitude of strength increases between trained and untrained population groups (Table 10). Rhea and colleagues reported strength increases in the bench press of 20% for one-set compared with a 33% increase with three-sets. Pre- to post-leg strength increased by 25.4% for one-set compared with an increase of 52.1% with three-sets. Substantial ES differences were reported between one and three-sets (ES = 2.3 bench press and ES = 6.5 leg press) for men and women irrespective of training status.

Table 10. Summary Findings on Set-volume Performed by Rhea et al., (2003)

Screenshot 2020-09-11 at 15.41.02.png

In their meta-analysis, Rhea and colleagues (2003) calculated the ES using the difference between the post-training and pre-training means divided by the pre-training standard deviation. They did not include any mean or standard deviation until after the pre-training strength test.  Furthermore, Rhea and colleagues did not describe how they determined the specific muscle groups that accounted for strength improvements reported in each study or specify how they coded sets per muscle group. The authors noted that potential sources of bias existed in published research and indicated that both published and unpublished studies would be included in the meta-analysis.  They claimed that they performed a literature search of published and unpublished RT invention studies. However, there is no clear evidence of any unpublished studies being omitted or retrieved in their narrative or reference section of, their inclusion and exclusion criteria for these studies, or how many unpublished studies were found in their search. 


The results within their meta-analysis are problematic with Rhea and colleagues (2003) reporting ES of 1.75, 1.94, and 2.28 for two, three, and four-sets per muscle group in previously untrained populations (Table 11). However, they did not specify if there were any significant differences in ES between set volume groups.  For example, the five-sets per muscle group elicited an ES of 1.34, almost one standard deviation below four-sets (ES = 2.28).  The inference made by Rhea and colleagues (2003) is that the previously untrained population elicits the most significant strength gains in the triceps by performing one-set of the bench press, military press, triceps press, and dipping exercises.  However, by additionally performing only one more set of any one of these exercises or one more triceps exercise would reduce the training effect by almost one standard deviation.


Rhea et al., (2003) also stated that further strength increases accompany RT beyond one-set protocols for advanced trainees.  Their data (Table 11) reported an ES of 0.92, 1.0, and, 1.17 for two, three and four-sets per muscle group.  However, there is no reported statistical analysis of differences between these ES; hence, any conclusions from their data are debatable. These conclusions for the untrained and trained population are inconsistent and have limited practical application to RT.

Meta-analyses by Peterson, Rhea and Alvar on Muscular Strength Gain


In the Peterson, Rhea and Alvar (2004) meta-analysis a total of 37 studies were selected for analysis with the following variables coded: sex and age, the frequency of training, RT loading (intensity), the number of sets performed, training to failure, periodisation of RT programme, and the use of creatine. Peterson and colleagues (2004) stated that superior pre-to-maximal strength gains were produced when performing eight-sets per muscle group during each RT session. They further suggested that there were additional strength benefits of performing higher training volumes (Table 11).


Nevertheless, the evidence contained within the Peterson, Rhea and Alvar (2004) meta-analysis cannot be used to validate such a conclusion. Inferences were derived supporting the contention that athletes should perform eight-sets per muscle group to develop muscular strength. Unfortunately, Peterson, Rhea and Alvar (2004) did not provide a rationale for study inclusion within the meta-analysis or specify how they coded the number of sets per muscle group or indicated which muscles they were coding. Additionally, one of the inclusion criteria of studies was that subjects must have been competitive collegiate or professional athletes.  Unfortunately, several included studies (Aagaard et al., 1996; Hoff, Helgerud and Wisloff, 1999; Kraemer et al., 2000; Newton and McEnvoy, 1994) by Peterson and colleagues involved subjects who had no prior experience with RT. They also did not explain how these competitive male and female athletes from varied sports would respond differently from other RT novices. This should have been considered and specified within the meta-analysis as by freely allocating varied population groups impedes any conclusion about different responses to a specific set-volume of RT.


Peterson, Rhea and Alvar (2004) also stated that a one-way analysis of variance was performed to compare differences in ES by variable and training protocol, with the level of significance set at P < 0.05. However, no statistical differences were reported between ES. Moreover, the mean ES was 1.22 for eight-sets per muscle group (Table 11) but was generated from only six ES. Peterson and colleagues did not state the source of those ES or how many studies produced this data. Such an assumption is unreliable due to the small number of ES included for eight-sets.  Although not stated by Peterson and colleagues, the ES reported, could be from only one study with four-sets accumulated from 119 ES.  No data supporting strength changes in kilograms were specified to allow informed decisions to be made on the presented evidence. Furthermore, Peterson, Rhea and Alvar (2004) provided restricted descriptions and only included the mean ES and standard deviations that confirmed maximal strength gains.

Screenshot 2020-09-11 at 15.49.30.png

Table 11. Summary of Meta-analytical Outcomes Performed by Peterson, Rhea and Alvar, (2004 and 2005) on Set-volume

According to Otto and Carpinelli (2006), the meta-analysis performed by Peterson and colleagues (2004) incorrectly computed the ES as they used the post-training mean minus the pre-training mean divided by the pre-training. According to Thomas, Salazar and Landers (1991), a pooled standard deviation is recommended when there is no control group. However, several studies included within Peterson et al., (2004) meta-analysis (i.e. Berger, 1962 and Kraemer, 1997), had no control group.  This absence of a control group should have led to a pooled standard deviation rather than the pre-training standard deviation used in this meta-analysis (Otto and Carpinelli, 2006). 


In the second cited meta-analysis, Peterson, Rhea and Alvar (2005) examined two meta-analyses (Rhea et al., 2003; Peterson, Rhea and Alvar, 2004) that comprised of 177 studies and a total of 1803 ES that investigated the dose-response trends for intensity, frequency, set-volume on strength gains. The authors reported no new relevant data on the athletic population, and information derived was merely a repeat of their previous meta-analysis (Peterson, Rhea and Alvar, 2004). Furthermore, Peterson and colleagues concluded that recreationally trained (non-athletes) should perform RT with a four-sets volume if they are required to increase maximal strength.

A meta-analysis by Wolfe, Lemura and Cole (2004) on muscular strength gain


A meta-analysis by Wolfe, Lemura and Cole (2004) sought to determine if muscular strength outcomes were affected by the number of sets performed and the duration of the RT programme (Table 12). Wolfe et al., (2004) indicated that the ES for trained subjects was significantly greater because of training with multiple sets (ES = 0.70) compared with a single set (ES = 0.29), but not significantly different in untrained subjects (ES = 1.73 and 1.69, multiple sets and one-set, respectively). Analysis related to subjects set-end point was also performed. Set-end point denoted as subjects encouraged to perform resistance exercises until muscular failure vs subject self-determining their own end-point. Analysis of the main effects revealed no significant difference for the set-end point and the number of sets performed (P < 0.052). Examination by training status revealed that the highest mean ES (1.86) was produced by untrained subjects who did not train to failure. Wolfe, Lemura and Cole (2004) concluded that one-set programmes were more suitable for untrained individuals, as comparable strength gains were observed with one-set and multiple-set programmes. These observations led Wolfe and colleagues to suggest that as the subject’s strength advances, there should be an associated change in programming from one to multiple-sets to stimulate continuous strength gains.


Unlike previous meta-analyses on set-volume (Rhea et al., 2003; Peterson et al., 2004), Wolfe and colleagues met minimal methodological criteria from which ES could be extracted. They controlled for several essential variables within the sampled studies, including publication bias, research design quality, the number of subjects included in each study, the number of ES generated and possible outliers. However, Wolfe et al., (2004) specified within their narrative that after examination, no studies were excluded as outliers. Unfortunately, only 16 studies met acceptable inclusion criteria, with six studies included subjects with previous training experience. Lipsey and Wilson (2001) note that having a limited number of studies makes it problematic to conclude and may result in ‘second order sampling error’ (i.e. sampling error at the meta-analysis level). There are additional issues with this meta-analysis, as Wolfe et al., (2004) did not clearly define trained, and incorrectly included a study by Kraemer et al., (2000) as trained were if fact the subjects were previously untrained collegiate female tennis players.  However, the results from this meta-analysis are debatable, as only a limited number of studies involved in the meta-analysis were experienced trainees. Additionally, Wolfe et al., did not critically evaluate each study, and multiple non-independent ES were applied. 

Table 12. Summary Findings on Set-volume Performed by Wolfe, Lemura and Cole (2004)

Screenshot 2020-09-11 at 15.56.29.png

Post ACSM 2009 Meta-analysis on Set-volume and Strength gain


A meta-regression by Krieger (2009) compared the effects of one-set vs multiple-sets of resistance exercises on muscular strength. Krieger’s analysis comprised of 14 studies (440 subjects), with 30 treatment groups and a total of 92 ES. The results suggested that multiple-sets were associated with a larger ES than one-set (difference = 0.26 ± 0.05; 95% confidence interval [95% CI]: 0.15 – 0.37; P < 0.0001). When Krieger’s meta-regression was further assessed there was a drift towards two-to-three-sets per exercise compared to one-set (difference = 0.25 ± 0.06; 95% CI: 0.14 – 0.37; P = 0.0001). No significant difference was observed between one-set per exercise and four-to-six sets per exercise (difference = 0.35 ± 0.25; 95% CI: -0.05­ – 0.74; P = 0.17) or between two-to-three-sets and four-to-six sets per exercise (difference = 0.09 ± 0.20; 95% CI: -0.31 – 0.50; P = 0.64). Krieger reported that the sensitivity analysis revealed no highly influential studies and no evidence of publication bias. It was concluded that two-to-three-sets per resistance exercise produced 46% greater strength gains compared to one-set in both trained and untrained subjects.


However, Carpinelli (2012), questioned the findings of a meta-regression performed by Krieger (2009). Carpinelli identified a series of fundamental errors within the reporting of critical outcomes. The methods used to select the studies, extract the data or assess their validity were not clearly described and it was not known whether efforts were made to reduce reviewer errors or bias. The assessment of study quality and validity was scored on two published ten-point scales, one described by Bågenhammar and Hansson (2007) and the other by Durall et al., (2006). These scores were pooled together to form an overall score, but the criteria were not reported, and only the aggregate score was presented. This, therefore, makes it difficult to critique the quality of the studies and raises questions about the reliability of the evidence. There are issues with scoring consistency with the included studies (i.e. Rhea et al., 2002; Kremmler et al., 2004; Kraemer, 1997) that question the credibility of the conclusions made by Krieger.  For example, Carpinelli (2012) identifies several reasons for the three highest quality scored studies to be excluded from the meta-regression (Table 13).


Table 13. Confounding studies in Krieger’s (2009) meta-regression

Screenshot 2020-09-11 at 16.00.30.png

Conclusion on Set-volume


Although meta-analytical evidence on the effects of single-set-vs three-sets have been published (Rhea et al., 2002; Peterson, Rhea and Alvar, 2004; Wolfe, Lemura and Cole, 2004; Krieger, 2009) disagreement remains regarding the appropriate dose-response relationship (additional sets of resistance exercises) and strength gain. Reviews by Otto and Carpinelli (2006), Carpinelli (2012), and Fisher et al., (2011) have examined the validity of published meta-analyses on increased set-volume, concluding that the reported data do not support a dose-response relationship between the increased number of sets and strength gains (Table 14). These critical reviews report issues concerning the inclusion of low-quality studies that create spurious inferences regarding strength increases. Due to reported imprecisions and variations in the included studies (variation in subject characteristics, and variances in experimental programme design) have been reported to affect meta-analyses reliability and accuracies. These limitations, therefore, do not adequately provide necessary evidence from previous meta-analyses that partially contribute towards strength development recommendations.

Table 14. Adherence to meta-analysis recommendations 

Screenshot 2020-09-11 at 16.05.48.png

Previous observations on set-volume have been reported to lack well-controlled selection and screening procedures that create difficulties for those endeavouring to summarise the available data. Findings on previous meta-analyses suggest that researchers should be cautious when performing mixed model meta-analyses (mixed-sex subject groups), as this may produce unfounded conclusions. The body of scientific knowledge would be significantly enhanced if more volume equated random controlled studies were performed on comparable subjects (i.e. similar age, same sex and training status) to help clarify the optimal set-volume required for muscular strength gains. This would help establish the optimum set dose-response and provide larger sample sizes for meta-analyses, thus reducing the need to include low power studies.

To date, researchers have oversimplified their RT study designs and may have inadvertently produced data that provides unreliable guidance concerning the appropriate number of sets and strength development. Intriguingly, there seems to be limited research that focuses specifically upon the variation of the number of sets performed per exercises and on sets per muscle group. There is a moderate-to-large body of evidence that has led to the current recommendations on set volume. Unfortunately, these training recommendations have been established from mixed-sex groups, broad population age ranges, use of multiple and different strength measurements that does not fully answer the question at hand individually or collectively. However, there is undoubtedly scope for well-controlled meta-analytical studies examining this area


ACSM (2009) Recommendations on Training Frequency

Traditionally, weekly RT frequency has not been adequately considered nor debated; however, there have been different recommendations that support a variety of RT frequencies for developing muscular strength. Limited scientific attention has been given to RT frequency and the potential effect it has on increasing muscular strength. Physical activity guidelines from organisations including the ACSM that recommend an RT frequency of two-to-three d.wk-1 for healthy adults (ACSM, 2009 and 2011). These training frequency recommendations are, however, notional, derived from inference rather than evidence.


Unfortunately, existing conjecture and limited evidence suggest that the RT frequency should differ for untrained, intermediate and experienced trainees (Campos et al., 2002). Several studies have stated that an RT frequency of two-to-three d.wk-1 for previously untrained individuals produces optimum strength gains (Carroll et al., 1998; Braith et al., 1994; Campos et al., 2002; Paulsen et al., 2003) whereas Feigenbaum and Pollock (1999) suggest that a single-set programme of 15 repetitions performed at a frequency of two-to-three d.wk-1 allows for sufficient regeneration, while still providing 80-to-90% strength gains of more frequent RT programmes. Nonetheless, these recommendations are derived from samples that include both healthy and chronic disease patients.


The current evidence on the effect of weekly RT frequency on strength development has mostly been undefined with inconsistent results and unsubstantiated recommendations. The ACSM (2009 and 2011) position statement cites 16 RT studies (Table 3.2.5) that support the frequency recommendations for untrained (Candow and Burke, 2007; Coyle et al., 1981; Dudley et al., 1991; Graves et al., 1989; Hickson, Hidaka, and Foster, 1994; McLester, Bishop and Guilliams, 2000; Rhea et al., 2003); intermediate (Candow and Burke, 2007; McKenzie-Gilliam , 1981; Graves et al., 1989; McLester, Bishop and Guilliams, 2000; Rhea et al., 2003) and well-trained subjects (Häkkinen et al., 1988; Hoffman et al., 1990; Rhea et al., 2003). Although these recommendations appear plausible, the cited studies may not apply to the suggested RT frequency. In addition, to the ACSM (2009) citing supporting evidence is the category rating scale that attempts to quantify the scientific evidence-based on the National Heart, Lung, and Blood Institute evidence category (NHLBI, 1998). ASCM (2009) awarded the untrained population group the highest category rating (Cat A) with the sources derived from randomised control trials (RCT). This has been defined as evidence derived “from well-designed RCT that provide a consistent pattern of findings in the population for which the recommendation is made. Requires a substantial number of studies involving a substantial number of participants.” (ACSM, 2009). However, not all cited studies were RCTs and were derived from six studies and one meta-analysis (Table 15). The intermediate population group had a category rating of B with cited evidence from three randomised studies (RAN), one RCT and the same meta-analysis used for the untrained population group (Rhea et al., 2003). The trained population group had a category rating of C from two studies and two meta-analyses. The recommendations provided by ACSM (2009 and 2011) are invalid and call into question the authors, editors, and the peer-reviewers of the publication and the methods used to classify and cite the evidence.


The studies cited by ACSM (2009) for untrained and intermediate subjects use the same studies to support recommendations. Furthermore, two meta-analyses were also included to support RT frequency, and if excluded, then only nine studies generated the recommendations on RT frequency. Thus, the ACSM (2009) recommendations have been generated from limited heterogeneous training and population groups with differing RT variables. These differing variables cast doubt on the ACSM (2009 and 2011) position statement as studies cited may not support the recommended RT frequency as it is difficult and often inappropriate to extrapolate the findings of their included cited studies across different population groups.

Table 15. Weekly Resistance Training Frequency

Screenshot 2020-09-11 at 20.40.19.png
Screenshot 2020-09-11 at 20.43.48.png
Screenshot 2020-09-11 at 20.46.45.png

Untrained Subjects and Weekly Training Frequency


The ACSM (2009 and 2011) position statement recommends that novices (those with no RT experience or have not trained for several years) train the entire body two-to-three d.wk-1. A study by Candow and Burke (2007) was cited as supporting evidence for this recommendation. Here, the authors investigated the effects of short-term equal-volume RT with different training frequencies on lean tissue and muscular strength. Twenty-nine untrained male (n = 6) and female (n = 23) subjects trained two or three d.wk-1. The pre-to-post results reported that both groups increased squat strength (28%), bench press strength (22-to-30%) (P < 0.05) and increased lean tissue mass. Candow and Burke suggested that the training volume might be more important than the frequency in developing strength in untrained men and women but did not suggest a frequency specific response.


Another study cited in the ACSM (2009 and 2011) position statement was Dudley et al., (1991) who examined the effects of eccentric actions on muscular adaptations to RT. Thirty-five middle-aged males performed four-to-five sets of six-to-12 repetitions per set of leg press and leg extension exercises two d.wk-1 for 19-weeks. Subjects were randomly separated into three exercising groups (concentric / eccentric action; concentric only; concentric / concentric) and a control group that did not train. The resistance exercises per set induced muscular failure within the specified number of repetitions. The results reported pre-to-post significant increases in the subjects’ 3RM leg press when performing concentric and eccentric actions (concentric / eccentric actions = 26%; concentric / concentric = 15%; concentric only = 14%). Subjects demonstrated significant pre-to-post strength increases when performing leg press 3RM with only concentric actions (concentric / eccentric actions = 22%; concentric / concentric = 18%; concentric only = 14%). For the 3RM leg extension exercise when performing both concentric and eccentric actions, pre-to-post significant strength increases were observed (concentric / eccentric actions = 29%; concentric only = 16%). This study focused on muscular actions (concentric vs eccentric) rather than the specific training frequency that would induce muscular strength development. It would be difficult and inappropriate to infer these findings to support of any RT frequency recommendations, as Dudley et al., (1991) investigated muscle actions.


The ACSM (2009) cited a study by Hickson, Hidaka and Foster (1994) that examined the effects of three submaximal exercise intensities on strength and muscle fibre type. A small sample group of eight subjects (four male and four female) completed a three d.wk-1 RT programme that continued for 16-weeks. Each subject performed the same relative work rates at 40,60 and 80% 1RM. After training, the bench press and parallel squat strength performance increased by 23% and 37% respectively. Subjects parallel squat repetitions at 40% were compared with the percentage of slow twitch fibres in the vastus lateralis muscle. Post-training analysis observed a similar relationship at 40% and 60%. Hickson, Hidaka and Foster (1994) study did not need to consider training frequency and the impact it has on strength development, as the objective was to access the role RT has with muscle fibre development.


 Intermediate Subjects and Weekly Training Frequency


For intermediate trainees, the ACSM position statement suggests a similar training frequency of two-three d.wk-1 for total-body workouts or split routines (upper body/lower body) to provide a higher volume of exercise. There is one additional study (McKenzie-Gilliam, 1981) cited in the ACSM document to support intermediate trainees’ recommendations. McKenzie-Gilliam (1981) examined subjects’ bench press muscular strength responses to five different weekly training frequencies (one, two, three, four or five d.wk-1). The subjects were high school male volunteers (n = 75) that trained either one, two, three, four or five d.wk-1 for nine weeks. The results from McKenzie-Gilliam inferred that there was a sequential dose-response curve towards higher weekly training frequency. McKenzie-Gilliam reported that the group that trained five d.wk-1 demonstrated superior strength gains than the other groups leading to suggest that more frequent stress leads to greater muscular adaptations. However, it must be emphasised that all groups performed identical RT programmes of 18-sets of 1RM and the total weekly volume between groups were not equalised. Consequently, the greater weekly frequency groups had significantly increased training volume at the end of the nine-weeks. It is highly unlikely that most healthy adults would apply the training procedures of McKenzie-Gilliam (1981) (18-sets of 1RM) due to the time-consuming and unfeasible training methodology.


These short-term strength adaptations, however, may be because the subjects performed the 1RM bench press throughout the nine-week period. This increase in the subject’s strength could be attributed to the principle of specificity as strength developments may not be increased by additional weekly frequency. This was determined by Dankel et al., (2016) who performed 1RM and maximal voluntary isometric contraction (MVC) testing on elbow flexion exercises on five trained subjects. Subjects performed a 1RM and MVC on one arm while the other arm performed a 1RM test and MVC, in addition to three-sets of exercises (70% 1RM) for 21 days. Dankel and colleagues stated that the increase in the subjects’ 1RM might not have been completely related to exercise volume but was driven by the specificity of the exercise. Therefore, the results from the study by McKenzie-Gilliam (1981) may not have been solely related to the subjects training frequency as performing 18-sets of 1RM may have been due to a ‘learning effect’ caused by performing repeated RT sessions.


Trained Subjects and Weekly Training Frequency


The training frequency of four-to-five d.wk-1 for advanced weightlifters, powerlifters, and bodybuilders has been suggested for strength development for trained individuals (Table 16). Cited by the ACSM (2009) position statement is the study by Häkkinen et al., (1988). The authors concluded that high-intensity resistance exercise on the same day results in acute adaptive responses on the neuromuscular and endocrine system. However, no suggestion was given regarding training frequency; instead, it was reported that diurnal variations might have concealed endocrinological adaptations in the morning session.


The study by Hoffman et al., (1990) was also cited as evidence that supports higher weekly frequency for well-trained subjects. Hoffman and colleagues examined the effects of self-selection of RT frequency on muscular strength. Sixty-one American football players participated in a ten-week winter training programme. Each subject self-selected a weekly RT frequency from three d.wk-1 (n = 12), four d.wk-1 (n = 15), five d.wk-1 (n = 23), or six d.wk-1 (n = 11) and participated in sport specific conditioning twice per week. Sports specific field tests were conducted before and after the ten-week RT programme, including strength measurements (1RM squat and bench press). The post-test analysis revealed significant increases in the 1RM squat for the four d.wk-1 and six d.wk-1 training frequency, with the five d.wk-1 having significant increases in both 1RM squat and bench press. However, Carpinelli, Otto and Winett (2004) were highly critical of the Hoffman et al., (1990) study as they identified several methodological issues. These included concerns with subjects not being randomly assigned to different training programmes, subjects self-selecting training frequency, and expansive resistance loading ranges with subjects performing four-to-five-sets of between two-to-ten repetitions.


Furthermore, this study was not matched for total weekly set-volume or repetitions. Moreover, the four and five d.wk-1 groups performed less total weekly training than that of the three and six d.wk-1 groups which could suggest that it was the reduction in training volume and not the frequency that produced physical changes. Unfortunately, these factors were not fully considered by the authors when interpreting their findings.


Meta-analytical Recommendations for Training Frequency

The two meta-analyses that are included within the ACSM (2009) recommendations (Rhea et al., 2003; Peterson, Rhea and Alvar, 2004) have reported methodological constraints, and the published evidence provided is disputed (Carpinelli, 2009). The two meta-analyses (in part) sought to determine the optimum weekly training frequency for strength development. Rhea et al., (2003) analysis comprised of 140 studies with a total of 1433 ES and reported that the ES for training frequency was different by training experience (Table 3.2.6). Rhea et al., (2003) reported that the untrained subject’s ES increased as the frequency increased to three d.wk-1 with the trained subject’s ES eliciting the most significant strength change with two d.wk-1 training. However, the programme design for the trained group had an increased training volume that may have been too aggressive for the untrained populations. The data produced by Rhea et al., (2003) supported the contention that increased weekly training frequency is superior to that of a single training session per muscle group. However, as observed (Table 3.2.6), potential issues with the ES data sample size might warrant caution when interpreting the findings on training frequency. The untrained group had an unbalanced sample with 17 ES in the one d.wk-1 treatment compared to 158 ES for two d.wk-1 and 965 ES for three days. The standard deviation in both the two and three d.wk-1 represents a significant deviation from the ES mean (two d.wk-1 = 3.1; three d.wk-1 = 2.3 respectively).

Furthermore, when observing the trained subject’s ES, no inclusion of one d.wk-1 treatment was provided within the meta-analysis. The reported optimum frequency per muscle group is founded upon 69 ES compared to 133 ES for three d.wk-1. This variance could therefore decrease the reliability regarding the optimum exercise weekly frequency. The Rhea et al., (2003) study reported the pooled effects for training frequency up to three d.wk-1, however, no additional information is available concerning if increased weekly training frequency would increase muscular strength. No accurate conclusions can be fully extracted from this meta-analysis, as several confounding factors decreased the strength and power of the evidence. These confounding factors included the reporting of training each muscle group twice per week which had an ES of 1.4 which was two times greater than training three times a week (ES = 0.70) for trained individuals. The pooled ES expose substantial differences between training frequencies with two d.wk-1 (n = 69) and three d.wk-1 (n = 133) influenced the reliability of the meta-analysis.

Table 16. Summary of Meta-analytical Outcomes on Training Frequency

Screenshot 2020-09-11 at 20.56.55.png

The Peterson, Rhea and Alvar, (2004) meta-analysis attempted to identify whether a specific dose-response relationship existed for loading (intensity), training frequency, and exercise volume on strength gains in an athletic population. Peterson and colleague’s analysis comprised of 37 studies with a total of 370 ES, with training frequency determined by the number of days’ subjects trained a specific muscle group per week. The authors proposed that maximal strength development is elicited when subjects train two d.wk-1. However, such a conclusion is erroneous and inconsistent with the presented evidence in their results (Table 17). Specifically, when comparing two vs three d.wk-1 as the ES variance between treatments is 0.01. Therefore, any conclusions drawn about the direct impact of training frequency would be unreliable.

Considerations Towards Weekly Training Frequency


The weekly training frequency is dependent on several training variables that can impact upon muscular strength adaptations. These variables include training volume, intensity, exercise selection, training status, and speed of muscle actions (eccentric, concentric movements). Logan and Abernethy (1992) reported that a time course after a single bout of heavy resistance exercise leads to a decrement in strength among inexperienced trainees. Logan and Abernethy recommended that recovery of approximately three-days may aid in the complete restoration of maximum voluntary contraction capacity and one repetition maximum (1RM) performance.


However, there are studies contrary to this supposition and propose that increases in strength occur with reduced intersession recovery. For example, Clarkson et al., (1992) reported that the upper body responds better to a higher training frequency of five d.wk1 and the lower body responds best to a training frequency of between three- and four-days d.wk1. A review by Tan (1999) suggested that untrained individuals may require longer intersession recovery periods to stimulate optimum muscular strength. Research by Hoffman et al., (1990) and Stowers et al., (1983) suggest that smaller muscles produce smaller observed strength gains which may require trainees to have more stimulus or more extended observations before reporting statistically significant differences. Stowers et al., (1983) further stated that athletes are possibly closer to their strength potential and that higher training frequencies may evoke more significant strength gains. Binkley (2002) suggested a higher training frequency of four-to-six dwk-1 in the off-season period to improve an athlete’s strength. Fleck and Kraemer (2014) have specified that to increase strength, training frequency should be performed daily with trainees performing each muscular action maximally.


Conclusion on Weekly Training Frequency


Attention must be given regarding the time-consuming nature of trainees performing higher frequency training. There should at least be a substantial body of evidence that supports that higher frequency training produces significantly superior results than lower frequency training. Unfortunately, there is limited evidence to support such recommendations. As reported in this section, the current literature does not provide a significant body of support for higher weekly training frequency. For both novice and advanced trainees, there is limited evidence that training each muscle group at higher frequencies (two-to-three d.wk-1) provide additional strength benefits. Smith and Low (2004) examined the scientific evidence that supported the ACSM’s (2002) position statement on recommendations that trainees perform higher frequency training to lower frequency training.


These recommendations for both novice and experienced trainees had limited support for training the muscle group more than once per week. Smith and Low (2004) provided supporting evidence from six studies (McLester, Bishop and Guilliams, 2000; Taaffe et al., 1999; Carpenter et al., 1991; Graves et al., 1990; Pollock et al., 1993; DeMichele et al., 1997) that demonstrated limited strength increases by performing additional weekly training sessions. For example, a study by Graves et al., (1990) investigated the effects of 12-weeks of training on the lumbar strength of untrained subjects. Subjects performed one-set of lumbar extensions either one d.wk-1, two d.wk-1, three d.wk-1 or once every 14-days. Graves and colleagues reported that all groups increased significantly in peak isometric torque across seven joint angles examined, with no differences between groups in isometric strength. Intriguingly, one of the subjects in the three dwk-1 reported strength losses and muscular atrophy from overuse. Smith and Low (2004) stated that there might be significant inter-individual responses due to the subject’s specific exercise tolerance. Furthermore, limited considerations were made by the ACSM (2009) position statement on the issues surrounding the importance of individualising resistance programmes based on the subject’s RT tolerance.


Order of Resistance Exercise

Exercise selection involves choosing exercises for an RT programme (Baechle and Earle, 2008). Several terms have been proposed for resistance exercise classifications, including primary or assistance exercises, structural or body part exercises, and multi-joint or single-joint exercises (Fleck and Kraemer, 1997). All those classifications are based upon the size of the muscle involved. For example, single-joint exercises (leg curl, tricep extension, and chest flye) are often prescribed to isolate specific muscle groups with a reported reduction of injury due to the minimum level of skill or technique required to complete the movement actions (ACSM, 2009). That said, multi-joint exercises (squats, deadlifts and shoulder press) have been suggested to be more neutrally demanding and have been regarded as the most effective means for increasing muscular strength (ACSM, 2009). According to ACSM (2002 and 2009), position statement healthy adults should perform movements that activate large muscle groups first in a training session. Equally, several resistance studies have specified that repetition performance was significantly higher for resistance exercises that involved either large or small muscle groups when performed at the beginning of a session. With relevance towards chronic adaptations, the few studies that analysed maximal strength in response to exercise orders suggest more significant increases in maximal strength occur when exercises are performed at the commencement of the training session (Dias et al., 2010; Simão et al., 2010; Spineti et al., 2010).


Current ACSM (2009) Recommendations for the Order of Resistance Exercise


Recommendations concerning exercise order in a training session suggest that trainees should perform large muscle group exercises before exercises involving small muscle groups (e.g. bench press before chest flye). The suggested rationale is centred on the premise that smaller muscle groups become pre-fatigued (via single-joint exercises) and this would then place less overload during multi-joint exercises and therefore less training stimulus on the larger muscle groups. For several years, multi-joint (structural) exercises have been suggested that they should be performed before single-joint exercises (Stone et al., 1983). However, relatively few studies have carefully examined these recommendations, especially the acute responses and chronic adaptations under controlled conditions. Currently, ACSM (2009) has cited two studies (Simão et al., 2005; Simão et al., 2007) within their evidence statement and recommendations that claim to support exercise order. However, the classification and sources of evidence are from non-randomised trials (Table 17), thus rendering these recommendations weak.

Table 17. ACSM (2009) Supporting evidence for Exercise Order 

Screenshot 2020-09-11 at 21.17.47.png

ACSM (2009) Supporting Evidence for Exercise Order for Strength Development


The Simão et al., (2005) study cited by ACSM (2009) observed the performance effects of exercise order during the upper body only RT exercises. A total of 28 trained subjects (male = 14; female = 4) with a minimum of six-month RT experience performed five upper body exercises. Subjects completed two training sessions separated by 48 hours’ recovery in a counterbalanced cross-over design. The first treatment group comprised of the large muscle groups trained before smaller muscles (L-SMG). The exercise order for the first session was bench press, lat pull-down, seated machine shoulder press, biceps curl, and triceps extension. For the second treatment group (S-LMG) subjects performed exercises that engaged the small to large muscle groups (triceps extension, biceps curl, seated machine shoulder press, lat pull-down and bench press). During both sessions, subjects performed three-sets of each exercise to concentric failure, with two-minutes recovery between sets and exercises. The findings of the study stated that during the performance of three-sets, the mean number of repetitions for each exercise varied significantly between groupings (except shoulder press). When comparing between groupings (L-SMG vs S-LMG), no significant differences were reported in the number of repetitions performed between the first and second sets for all exercises.


Furthermore, no significant differences between groups in the third set except for the bicep curl. However, there were significant differences when sets within each group were examined. In the L-SMG group, subjects completed fewer repetitions in the third set than the first or second sets for all exercises (except tricep extension). In S-LMG grouping, fewer repetitions were achieved for the shoulder press in set three compared to the first set. Subjects also performed fewer repetitions on the lat pull-down in the third set compared to set one or two. Simão et al., (2005) concluded that whenever an exercise is performed last in an exercise sequence, the trainee’s performance of that exercise will be negatively affected. There was no reference or endorsement made from this study that supported the contention of performing L-SMG to promote strength development.


The second study (Simão et al., 2007) cited to support of ACSM (2009) position statement investigated the influence of exercise order and the number of repetitions achieved, and trained female subjects perceived exertion rates during RT. Simão et al., (2007) collected data from 23 young women (24.2 ± 4.5 years) with a minimum of two years of RT experience. The exercise treatment order for L-SMG was bench press, shoulder press, tricep extension, leg press, leg extension, and leg curl. For the S-LMG treatment, the reverse exercise order was performed by subjects. During both training sessions, three-sets of each exercise was completed with subjects performing resistance exercise actions until concentric failure using resistance loading at 80% 1RM, with two-minute recovery between sets and resistance exercises. Simão and co-workers reported that the mean number of repetitions achieved was consistently less than when resistance exercises were performed later in the exercise sequence with both L-SMG and S-LMG affected by the exercise order.


The ACSM (2009) recommendations for sequencing exercises include performing L-SMG and multiple-joint before single-joint exercises. However, only two studies were cited by ACSM (2009) to support this, with these studies not directly referring that large to small muscle group training was the most effective. On the contrary, both studies stated that regardless of the training sequence, the exercises performed last would produce the fewest number of repetitions. There is an apparent lack of evidence to support such a claim on exercise order; unfortunately, the ACSM (2009 and 2011) position statement contains errors with cited references that failed to support their opinions and recommendations. This, therefore, renders the recommendations void when considering exercise order as no published studies on the chronic effects relative to the order of RT prior to 2010, the ACSM (2009) assertion was established only on opinion.


Current Evidence on Resistance Exercise Order and Chronic Adaptations


The current training recommendations concerning exercise order are unsubstantiated and are without any strong scientific foundation. Many trainees consider that the exercise sequence is an essential training variable. Regrettably, there is limited quantifiable evidence to confirm or deny whether specific exercise sequences have any significant effect on strength development (Carpinelli, 2013). There are, at present, an insufficient quantity of evidence that has investigated the effects of exercise sequence on strength gain (Table 18).


A study by Dias et al., (2010) examined the effect exercise order has on muscular strength in 48 untrained young men (age = 19.01 ± 0.5 years) after eight-weeks of training. Subjects were randomly assigned into three groups: (1) group one (L-SMG) performed bench press, lat pull-down, seated military press, biceps curl, and elbow extension; (2) group two (S-LMG) performed the same resistance exercises, but in the opposite order (elbow extension, biceps curl, seated military press, lat pull-down, and bench press); (3) group three served as a control group and did not train. Subjects in the L-SMG and S-LMG trained three times a week at a prescribed resistance loading of 8-to-12RM with a minimum recovery period of 48-hours given between training sessions. All treatment subjects were encouraged to perform all sets to concentric failure with a two-minute rest period given between sets and exercises. The subjects 1RM was assessed for all exercises at baseline, and after eight-weeks of training, however, Dias et al., (2010) did not state whether the 1RM assessors were blinded to the different training treatments. Furthermore, Dias and colleagues made no attempt to control for repetition duration during training or the subjects’ 1RM assessment. Results demonstrated that both L-SMG and S-LMG treatment groups significantly increased 1RM in the five exercises compared to baseline. Additionally, both groups exhibited significant pre-to-post strength increases of 18.6-to-76.4% for all exercises. However, the only significant difference between exercise sequences was that S-LMG reported a significantly greater increase in bicep curl and tricep extension strength.


A similar study by performed Simão et al., (2010) investigated the influence of exercise order has on strength and muscle thickness in untrained males following 12-weeks of linear periodised training. Thirty-one physically active males (age = 29.5 ± 0.6 years) but who had less than six-months’ experience were randomly assigned to one of two treatment groups or a control group. Both treatment groups followed a linear programme design; week one-to-four each group performed four-sets of 12-to-15RM with one-minute recovery between sets; weeks five-to-eight subjects completed three-sets of eight-to-ten RM with two-minutes recovery, and in weeks nine-to-12 subjects performing two-sets of 3-to-5RM with three-minute recovery periods. Subjects in the L-SMG group performed four exercises (bench press, later pull-down, tricep extension, and bicep curl). The other treatment group (S-LMG) performed the same exercises but in reverse order (bicep curl, tricep extension, lat pull-down, and bench press). The 1RM was assessed for all exercises at baseline, and after 12-weeks of training, however, like Dias et al., (2010) it was not stated whether the 1RM assessors were blinded to the different treatment protocols. Results by Simão et al., (2010) revealed that both L-SMG and S-LMG treatment groups had significant increases in pre-to-post 1RM strength for all four exercises (Table 3.2.8). This was regardless of the exercise sequence with only a two per cent variance between treatment groups. Interestingly, the tricep extension exercise was the only exercise that had a significantly larger ES in S-LMG (ES = 2.07) compared to L-SMG (ES = 0.75). However, when the other three exercises were combined per treatment group, similar pre-to-post strength changes were observed regardless of the exercise sequence (S-LMG ES = 0.75 vs L-SMG = 0.74) with a negligible ES difference between treatments (ES difference = 0.01).


In a further study, Spineti et al., (2010) randomly assigned 30 untrained males to one of two treatment groups (L-SMG and S-LMG) or a control group in which to assess the influence exercise order has on strength. In the L-SMG group, the following exercises were performed in this sequence; bench press, lat pull-down, tricep extension and bicep curl with the order reversed for the S-LMG group (bicep curl, tricep extension, lat pull-down, and bench press). The treatment groups followed a nonlinear programme design, during week one-to-four each training group completed four-sets of 12-to-15RM with one-minute rest. In weeks five-to-eight subjects completed three-sets of 8-to-10RM with two-minutes rest and in weeks nine-to-12 subjects performed two-sets of three-to-five RM with three-minute recovery. Subjects in both treatment groups trained twice per week and were encouraged to perform all sets to concentric failure while being supervised by a qualified strength and conditioning professional. Like the other two studies (Simão et al., 2010; Dias et al., 2010), it was not clearly stated whether the 1RM assessors were blinded to the different training procedures. Results suggested that both training groups significantly increased pre-to-post 1RM strength when compared to the control group. When resistance exercises were individually evaluated, the ES of the 1RM bench press was significantly greater w