What the science says about training frequency.
How often should you train each muscle group in the gym?
Most gym rats will tell you once or twice a week.
One of the most popular training formats nowadays is the "Bro Split," where you train five days a week, but divide workouts so that each muscle group is hit once weekly.
Other popular formats are four-day-a-week "push/pull" and "upper body/lower body" split routines, where individual muscle groups are trained twice weekly.
While common in years gone by, few people nowadays train muscle groups three times weekly. As for training the same muscle group four or more times per week, most brofessors will vigorously scoff at the idea. Tell them that's what you plan to do, and they'll almost certainly launch into a diatribe about injury and overtraining. "Your muscles won't have enough time to recover, man! You'll end up injured and overtrained!”
Whatever this belief is based on, it's not science (“trust me bro” is not a valid scientific source).
Up until 2018, only one published study that I’m aware of had directly compared a high-frequency routine (i.e. training individual muscle groups four or more times per week) with an equal-volume low-frequency routine.
So why did researchers ignore high-frequency training until recently?
The impetus for change, I suspect, was an unpublished study involving young members of Norway’s national powerlifting squad.
The Norwegian Experience: Train More Often, Get Stronger Quicker
The content below was originally paywalled.
A decade ago, strength-oriented websites were abuzz with news of an experiment dubbed the Norwegian Frequency Project. For many years, most Norwegian powerlifters trained three days a week, hitting each big lift (squat, bench press, deadlift) once or twice a week. That all changed in 1997, when former German Olympic weightlifter and coach Dietmar Wolf was appointed as Norway's national powerlifting coach. He promptly began incorporating higher-frequency training methodologies from his weightlifting background, adjusting them where necessary to match the demands of powerlifting.
The results were positive, and the new high-frequency approach quickly became the favoured training format. However, there was no formal data to prove the new protocol worked better than the traditional three-day program, so the Norwegian School of Sport Sciences decided to compare them in a head-to-head trial.
Sixteen powerlifters, mean age 21.1 years, volunteered for the study (13 men, 3 women). All had competed in national competitions within the previous six months, and all had trained continuously for competitive powerlifting at least 1 year prior to the study.
At the start of the study, mean 1RM squat, bench press and deadlift were similar in the two groups (167 kg, 123 kg and 201 kg, respectively).
So these subjects were quite a bit more advanced than the recreational lifters that usually comprise 'experienced' trainers in most university resistance-training studies.
All subjects were placed on the same fifteen-week program, with one key difference: Training frequency. One group trained three days a week, the other performed 6 shorter training sessions each week. At each session, the lifters in the 6 x a week group performed half as many sets as the 3 x a week group.
After fifteen weeks, a clear pattern was evident:
1RM in the squat increased, on average, 11% in the 6x group versus 5% in the 3x group;
1RM bench press increased 11% in the 6x group versus 6% in the 3x group;
1RM on the deadlift increased 9% in the 6x group versus 4% in the 3x group, although the difference was reportedly not significant.
The researchers also examined thigh hypertrophy. Cross sectional area of the quadriceps increased in the 6x group but did not change in the 3x group (4.2% versus -0.6%, respectively).
So it was a clear win for the 6x group, and high-frequency training continues to be a mainstay for high-level Norwegian powerlifters.
Unlike Olympic lifting, which is highly explosive and employs weights that can be lifted overhead, powerlifting involves grinding heavier weights through much shorter ranges of motion. The Norwegian subjects did not max out every time they hit the gym; the average intensity was 72% to 74% of 1RM for squat, bench, and deadlift. Most folks could knock out around 10 to 12 reps at that intensity, but in this experiment the reps were kept between 3 and 8 for the big three (squat, bench press, deadlift). The only time the subjects ground their way through lifts was when attempting new 1RMs at the end of the study.
In order to maintain a competitive advantage for Norwegian powerlifters, the full methodology and results of the study were never published. There is an abstract of the study presented at a 2012 conference available online (see page 547 of this link), while Dutch researcher Menno Henselmans has obtained some of the study data and diagrams and posted them at his site.
So let’s took a look at the published research, which has examined the effect of high-frequency training in populations ranging from complete newbies to advanced bodybuilders.
This is the only pre-2018 study I’ve found that compared a high-frequency routine with a lower-frequency routine of equal volume. That study, by Gary Hunter, University of Birmingham, Alabama, was conducted in an era when the use of whole-body routines was still common.
Untrained subjects in their 20s were recruited. One group was assigned to a classic 3-day-per-week total-body program comprised of seven exercises. The other group was given the same program with an extra training day. Both groups performed a total of nine sets per exercise, per week.
Hunter does not mention whether the subjects were instructed to train to failure. He does state the subjects aimed for 7-10 reps per set; once more than 10 reps was attained, the weight was increased by 5 lb at the next session.
Each group had eleven females; the 3x and 4x groups had 14 and 10 males, respectively. No mention is made of randomization, so I’m guessing it didn’t happen.
After seven weeks, 1RM bench press increased significantly more in the 4x group compared to the 3x group. The maximum number of reps attained with 50% of baseline 1RM bench also increased significantly more in the 4x group.
In line with the superior bench press results, the 4x group also made greater gains in chest circumference. Marginally greater but non-significant gains in biceps circumference were seen in the 4x group, while little change was seen in waist and hip circumferences.
Unfortunately, lower body strength and circumference changes were not assessed in this study.
The superior 1RM bench press and chest circumference gains were driven largely by the female subjects in the 4x group. After seven weeks, they had increased their 1RM by 9.1 kg, compared to 5.1 kg for the females in the 3x group, a 4 kg difference (the difference between males was 1.7 kg).
Chest circumference increased by 3.1 cm in the 4x females, compared to a -0.7 decrease in their 3x counterparts (the 4x males gained 0.5 cm more than 3x males).
The possibility that female trainees may respond especially well to high-frequency training is something I’ll return to in Part 2.
Fast forward thirty-three years, and we have a sudden burst of published research on the topic. In this University of South Florida trial, healthy young males with a mean age of 22 years and minimum 6 months’ resistance training experience were randomly assigned to either a 3 x or 6 x per week group.
Both groups performed total-body routines, which focused on the three powerlifts (squat, bench, deadlift).
The 3x group performed 4 sets of bench presses and squats on Mondays, Wednesdays and Fridays, and 4 sets of deadlifts on Fridays.
The 6x group performed 2 sets of bench presses and squats every day from Monday through Saturday, and 2 sets of deadlifts on Wednesdays and Saturdays.
The subjects also completed training for the upper back, delts, bis, tris, and abs throughout the training week, but no specific information is provided about their formatting. The researchers report the 6x group’s average training session lasted 1 hour, while the 3x group’s lasted approximately 2 hours, so I can only assume the remaining exercises were also divided up equally on the two routines.
Subjects in both groups were provided approximately 25 grams of whey protein isolate after each workout.
The researchers don't state in the paper whether the subjects trained to failure, but do mention the loading progression was based on that of Bryan et al 2010, where subjects did indeed train to failure on their work sets.
After six weeks of training, the 6x group gained 2.6 kg (3.6%) of fat-free mass, compared to 1.7 kg (2.5%) in the 3x group. 1RM gains on the squat were near identical among groups, while the 6x group experienced marginally greater but non-significant 1RM gains on the bench and deadlift. The 6x group also achieved a marginally higher but non-significant increase in their powerlifting total (the sum total weight of the squat, bench and deadlift).
While the differences between groups were not statistically significant, the researchers commented: “It is interesting to point out, however, that the effect sizes favored the 6x group for increases in all strength-related variables (with the exception of bench press).”
Gomes et al reported on twenty-three resistance-trained subjects (mean age 26.2 years; mean training experience 6.9 years) assigned to one of 2 routines:
A total-body routine, training all muscle groups at every one of the 5 weekly sessions;
A Bro Split, also training 5 days per week but training each specific muscle group only once a week.
On the 5x routine, every muscle group was worked Monday to Friday for 2-3 sets per workout. The same exercises were performed each day, but in varying order.
All work sets were performed to “volitional failure,” which is the point where subjects think they can’t do another rep - without actually trying. This is not the same as “momentary muscular failure,” where no matter how much you grunt, spit and fart, the weight won’t budge and there is no chance of completing the rep in good form.
All volunteers consumed a supplement containing 24 grams of whey protein and 6.4 g of carbohydrate immediately after all workouts.
After eight weeks, mean fat-free mass gains were 0.8 kg and 0.5 kg in the 5x and Bro Split groups, respectively. Muscle mass index (kg/m²) improved by 0.1 in both groups.
1RM squat improved by 12 kg and 8 kg in the 5x and Bro Split groups, respectively. The corresponding improvements in 1RM bench press were 9.7 kg and 5.6 kg.
The only significant difference between groups was in muscle soreness. The Bro Split group showed more delayed-onset muscle soreness (DOMS) than the high-frequency, whole-body group at the beginning, middle and end of the study.
Males in their twenties with a minimum of 6 months’ resistance training experience were randomly assigned to train either 3 or 6 times weekly.
Both groups performed total-body routines. One exercise was employed for each muscle group per workout. In the 3x group, 4 sets of each exercise were performed at every workout. In the 6x group, 2 sets of each exercise were performed at every workout.
All exercises were performed for 6–12 repetitions to muscular failure.
After six weeks, both groups experienced similar increases in squat 1RM, bench 1RM, and bench reps with 60% 1RM.
The subjects were also tested for muscular endurance using lighter weights. The 3x group managed to eek out 3.9 more reps on the squat using 60% of 1RM, but the researchers report large variability among the subjects, so the relevance of this finding is doubtful.
Elbow extensor thickness, rectus femoris thickness and vastus intermedius thickness all increased to a similar degree in both groups.
The researchers reported that elbow flexor (biceps) thickness increased significantly by 7% in the 3x group but decreased by 1% in the 6x group. This makes no sense. There is no tenable reason why the 6x group, having attained similar improvements in strength and triceps and thigh thickness, would suffer a small decrease in biceps size.
The most logical explanation for this anomaly is that the 6x group’s biceps decrease was not real but an artifact of the margin of error and variability inherent even in careful ultrasound testing.
Healthy men, with an average age of 26.4 years and resistance training experience of 6.5 years, were recruited for this study, and assigned to one of the following routines.
Whole-body, training each muscle grouping every day from Monday to Friday;
Bro Split.
Both groups performed a weekly total of 15 sets per muscle group, and 75 sets in total. At each workout, the 5x whole-body group performed 3 sets of a single exercise for each muscle group. A different exercise was used for each muscle group at each of the weekly workouts.
All sets were performed to the point of momentary concentric muscular failure.
After eight weeks of training, 1RM increases on the bench press and seated row machine were near identical between groups, while squat 1RM increased by 18.5% and 8.8% in the 5x and Bro Split groups, respectively. The difference was not statistically significant.
Forearm flexor, triceps and outer thigh muscle thickness increased by 8.5%, 11.2% and 9.7%, respectively, in the 5x group. The corresponding increases for the 1x group were 3.8%, 5.8% and 5.4%. In this case, the differences in forearm and outer thigh thickness between groups were statistically significant.
Perhaps because of the subjects' untrained status, in this trial the researchers gradually increased the volume and frequency of training. Both groups started with 2 weekly workouts, progressing to 5 workouts by week 4.
Between weeks 4-8, the total-body group trained each muscle group every day from Monday to Friday, while the other group followed a Bro Split.
During that period, both groups performed ten sets per week for each of the 7 exercises. The researchers don’t explicitly state whether sets were performed to volitional or momentary muscular failure, but do write that each set was performed for “8–12 maximal repetitions.”
At the end of the study, total lean body mass increased by 1.0 kg in the 5x group, and 1.5 kg in the Bro Split group.
Appendicular lean mass (arms and legs) increased by 0.3 kg and 0.5 kg in 5x and 1x groups, respectively.
Bench press 1RM increased 7.1 kg and 4.5 kg in the 5x and Bro Split groups, respectively.
Right and left leg extension 1RM increased by 21.3 kg and 23.7 kg, respectively, in the 5x group. The corresponding increases for the 1x group were 19.7 kg and 18.1 kg.
None of these differences were statistically significant.
Healthy, resistance-trained young men, with a mean age 20 years and training experience of almost 3 years, were randomly assigned to either a 2x or 4x total-body protocol.
The 2x group trained Tuesdays and Saturdays. Each workout consisted of 4 sets of 1 exercise each for thighs, hamstrings, chest, lats, delts, bis and tris.
The 4x group trained Tuesdays, Wednesdays, Saturdays and Sundays, performing 2 sets of each exercise per workout.
Again, the researchers don’t mention failure, simply stating that sets were performed for 8-10 reps with 70%-80% of 1RM, with weight increases when the prescribed amount of reps were completed.
After eight weeks, both groups experienced similar increases in chest circumference, thigh circumference, arm circumference, 1RM bench press, 1RM leg press, 1RM arm curl, bench press and leg press endurance (max repetitions achieved with 60% of 1RM), countermovement jump, and medicine ball throw.
The only statistically significant differences between groups were the slightly greater increases in 1RM bench press and 1RM leg press seen in the 4x group. Although the differences were not statistically significant, the 4x group also achieved slightly greater gains and corresponding effect sizes in bench press endurance and leg press endurance.
Evangelista et al 2021 (Brazil)
Healthy but untrained subjects in their 20s and 30s took part in this study. The full-body group performed the same routine at each of its four weekly sessions, using 4 sets of 8-12RM per muscle group. The split participants trained each muscle group twice per week, with 8 sets per session.
No mention is made of failure, but the researchers state the subjects executed 8-12RM per set and reported a perceived rating of exertion of 9.5 to 10 for all sets and exercises. Assuming they weren’t exaggerating a la a bunch of flopping soccer players (useless trivia: Brazil faked the most injuries at the 2014 World Cup), this would indicate they were training to failure or at least close to it.
After eight weeks, both groups experienced near-identical increases in bench press and squat 1RM.
Both groups also experienced very similar increases in muscle thickness at the four measured sites (biceps, triceps, vastus lateralis, rectus femoris).
Johnsen and van den Tillaar 2021 (Norway)
In this study, twenty-one resistance-trained male subjects (mean age and resistance training experience of 27.6 and 4.7 years, respectively) were randomly assigned to a full-body or split routine group, both training 4 times per week.
The split group performed an upper/lower body protocol, training muscle groups twice per week.
Sets were performed for 8-12 reps, and the subjects were instructed not to train to failure. Instead, they were told to terminate a set when they felt they still had two repetitions in reserve. If a subject met his rep target but felt he had more than 2 reps left in the tank, he was instructed to increase the weight by 2–10% on the next set.
After eight weeks, both groups significantly increased their squat 1RM (4x group, +14.31 kg; 2x group, +13.25 kg) and bench press (4x group, +8.86 kg; 2x group, +7.75 kg).
No hypertrophy or body composition outcomes were reported in this study.
Hamarsland et al 2022 (Norway)
Twenty-one subjects (11 males, 10 females, mean age 25.9 years) completed this 9-week study. The participants were moderately-trained, having regularly completed at least one workout weekly for the previous 6 months.
They were randomized to a low-frequency or high-frequency group, both using a minimalist whole-body routine comprised only of thigh, chest and upper back exercises.
The low-frequency group trained twice weekly, performing 2-4 sets per exercise for a total of 16 sets per workout. The whole-body group trained four times weekly, performed 1-2 sets per exercise for a total of 8 sets per workout.
The rep ranges were progressively reduced throughout the study, with subjects targeting their 12RM from weeks 1–3, 10RM during weeks 4–6, and 8RM in weeks 7–9.
The authors state that sets were performed to failure, but do not say whether it was of the volitional or momentary muscular type.
Participants were provided 20 grams of whey protein mixed in water after each workout. To balance the supplemental protein intake between groups, the 2x group also ingested the whey drink on the day after workouts.
After nine weeks, both groups made similar 1RM improvements on the squat, hack squat, bench press and chest press.
Lean mass, lean leg mass, lean trunk mass, lean arm mass and vastus lateralis muscle thickness also increased similarly in both groups.
Making Sense Of It All
There’s quite a bit to unpack here, which is why this article requires a sequel. For now, I’ll start with the most obvious take home point:
These studies do not support the widely-held belief that directly training a muscle group four or more times a week is harmful - even when training to momentary muscular failure.
Over periods of 6 to 9 weeks, high-frequency protocols showed themselves to be at least as effective as routines targeting muscle groups 1-3 times weekly.
Not only that, but looking at the results overall, it was the high-frequency routines that displayed a trend for more favourable outcomes.
Three of the 10 published studies above reported statistically significant differences in favour of the high-frequency group, on at least 2 hypertrophy and/or strength endpoints.
Hunter 1985 reported greater strength and endurance gains in the bench press, and a greater increase in chest circumference, for the high-frequency 4x group. The most favourable strength and circumference gains were seen in female 4x subjects.
Zaroni et al reported reported greater increases in forearm and outer thigh muscle thickness for the 5x group. They 5x group also experienced a non-significantly greater gain in squat 1RM.
Arazi 2021 reported slightly greater increases in 1RM bench press and leg press in the 4x group. The 4x group also achieved slightly greater but non-significant gains and corresponding effect sizes in bench press endurance and leg press endurance.
A few more of the studies showed non-significant differences and/or effect sizes in favour of the high-frequency groups: Colquhoun et al 2018 (lean mass, bench/deadlift 1RM, powerlifting total), Gomes et al 2018 (bench/squat 1RM), and Franco 2019 (bench 1RM).
In contrast, Saric et al reported a significant increase in biceps thickness only in the 3x group, but the alleged 1% decrease in the 6x group is untenable, as I detailed above.
Meanwhile, Evangelista et al, Johnsen and van den Tillaar and Hamarsland et al 2022 returned near-identical results between groups.
How Did the Bro Split Fare?
Three of the Brazilian trials (Gomes et al 2018; Zaroni et al 2018; Franco et al 2019) compared 5x a week total-body routines with the ubiquitous Bro Split, which is probably the single most popular weight training format in gyms nowadays.
Zaroni et al showed greater and statistically significant muscle thickness gains in the 5x group, and a non-significantly greater gain in squat 1RM.
In Gomes et al, the 5x group experienced greater but non-significant improvements in bench/squat 1RM.
In Franco 2019 the 5x group experienced a greater but non-significant improvement in bench 1RM.
So while no earth-shattering results emanated from any of the studies, the trend for favourable results is on the side of high-frequency training.
In studies where attrition occurred, dropout rates were similar between the low- and high-frequency groups, further refuting the belief that - at least for the studied durations - the latter leads to increased injury and overtraining.
Limitations of the Above Studies
The studies were small, with initial sample sizes ranging from eighteen subjects in Zaroni et al to 86 in Evangelista et al. Small sample sizes limit the ability to detect statistically significant differences, and this limitation is exacerbated by high dropout rates, which plagued a number of the above studies. The Colquhoun, Franco, Evangelista and Hamarsland studies featured substantial dropout rates of 28-42%.
It may be that the more favourable but non-significant results for high-frequency routines would have achieved statistical significance in larger studies, but until those larger studies are done that remains a matter of speculation. Sadly, sports science is nothing like drug research, where highly profitable pharma companies can spend big on trials with thousands and sometimes tens of thousands of participants.
It should be noted all the above studies involved young, healthy volunteers, most of whom were in their twenties. Caution needs to be exercised when extrapolating the results of these studies to older folks, particularly those who may have lingering muscular or joint issues.
In Part 2, I’ll discuss even more related research, and how you might apply this information to benefit your own training.
I’ll expound upon another limitation of the above studies that, when corrected, may yield the potential to unlock even greater benefits from a high-frequency routine.
I’ll also share with you my experiences with the 5-day-a-week, total-body, one-set-to-failure routine I’ve been following for the last 7-and-a-bit-weeks.
Until then, be well,
Anthony.
Leave a Reply