Muscle engagement and focus of attention: the evidence

ME-A-01The default focus is external

PrescriptionUnless a set meets the conditions in ME-A-03, the cue a member is given directs attention to the outcome of the movement (the bar, the floor, the handle) rather than to the body part producing it.

Evidence gradeA

EffectMeta-analyses of 73 studies (1,824 participants) for performance and 40 studies (1,274 participants) for learning found external focus more effective than internal focus for performance (Hedges' g = 0.264, 95% CI [0.217, 0.310]), retention learning (g = 0.583, 95% CI [0.425, 0.741]) and transfer learning (g = 0.584, 95% CI [0.325, 0.842]). Metaregression found neither age group, health status, skill level, nor their interactions moderated the difference (all p > .100) [1]. Fifteen years of research show the advantage extends across task types, skill levels and age groups, in movement effectiveness (accuracy, consistency, balance) and efficiency (muscular activity, force production, cardiovascular response) [2].

Population73 and 40 studies respectively across motor tasks and populations [1]; comprehensive review [2].

Sources[1][2]

CaveatsThe learning effects (g ≈ 0.58) are roughly double the acute performance effect (g = 0.26), so the strongest claim is about skill retained rather than about a single set. Most constituent studies are motor tasks such as balance, throwing and jumping rather than loaded barbell work. The unmoderated result across skill level is the finding that makes this safe to apply to a novice and an experienced lifter alike.

ME-A-02The load above which internal focus stops working

PrescriptionAn internal cue is given only below roughly 60% of one-repetition maximum. At and above that, the member gets an external cue regardless of goal. Band: 60% to 80% of 1RM, standard 60%.

Evidence gradeC

EffectEighteen resistance-trained men performed bench press at 20, 40, 50, 60 and 80% of 1RM under three conditions: regular, focusing on the pectoralis major, and focusing on the triceps brachii. Focusing on a muscle increased that muscle's activity at loads between 20 and 60% of 1RM, but not at 80%. The authors describe a threshold between 60 and 80% rather than a linear decline. The increase did not come at the expense of the other muscle's activity [4]. In a separate study, eleven college football players given verbal instruction to use only the chest or only the triceps at 50% and 80% of 1RM showed the effect at the lower load [7].

Population18 resistance-trained men [4]; 11 male Division III football players [7].

Sources[4][7]

CaveatsThis is the single most load-bearing number in the document and it rests on one study in eighteen people, in one exercise. The threshold is a range, not a line, and MVIII programs the conservative end of it. Whether the 60 to 80% boundary found in the bench press holds for a squat, a row or a curl has not been tested. The effect is on EMG amplitude, which ME-E-02 explains is not growth.

ME-A-03When an internal cue is used at all

PrescriptionAn internal cue is given only when all three hold: the goal is hypertrophy-oriented, the set's role is not a top set, and the load is below the ME-A-02 threshold. Otherwise the cue is external.

Evidence gradeC

EffectThirty untrained college-aged men trained 3 sessions per week for 8 weeks, 4 sets of 8 to 12 repetitions, randomly assigned to focus on contracting the target muscle or on the outcome of the lift. Elbow flexor thickness increased more in the internal group (12.4% vs 6.9%), with similar direction in quadriceps thickness. Isometric elbow flexion strength was greater for internal and isometric knee extension strength greater for external, though neither strength difference reached statistical significance [3].

Population30 untrained college-aged men, 8 weeks.

Sources[3][4]

CaveatsOne study, untrained participants, eight weeks, and a sample of fifteen per group. Untrained people grow from nearly anything, which is exactly the population in which a between-group difference is hardest to attribute to the variable of interest. The strength findings split in opposite directions and neither was significant, which is the honest summary of the strength question: unknown. MVIII gates this rule on three conditions rather than one specifically because the evidence for it is thin enough that its boundaries should be tight rather than generous.

ME-A-04Role outranks goal

PrescriptionWhere the goal and the set's role disagree, the role decides. A hypertrophy-focused member's heavy top set receives an external cue.

Evidence gradeC · derived

EffectNo study located tested this ordering directly. It is derived from two rows that were measured: the external-focus advantage on force production and movement efficiency is the better-evidenced of the two halves [1][2], and the selective-activation effect disappears above the ME-A-02 threshold [4], which is where a top set sits by definition.

PopulationNone. This row is an inference from [1][2][4].

Sources[1][2][4]

CaveatsDerived, and labeled as such. No trial has compared cueing strategies assigned by set role. The inference is that a set whose job is to move a heavy load should be cued the way the evidence says heavy loads are best cued, and that a member's overall goal does not change what a given set is doing. # PART B — Engagement

ME-E-01Attention changes measured activation

FindingDirecting attention at a specific muscle raises that muscle's surface EMG amplitude at low to moderate loads, without reducing the other prime mover's activity [4], and verbal technique instruction produces the same shift [7]. In the opposite direction, adopting an external focus reduced integrated EMG during biceps curls while movements were performed faster, and the reduction persisted when movement time was controlled with a metronome [8].

Evidence gradeC · contested

Population18 resistance-trained men [4]; 11 football players [7]; 11 and 12 participants across two experiments [8].

Sources[4][7][8]

CaveatsNote that [8] points the opposite way from [4] and both are correct: an internal focus raises the target muscle's signal, and an external focus lowers total signal while producing better movement. Less electrical activity for the same or better output is efficiency, not weakness. Reading these two rows as a contradiction is the mistake ME-E-02 exists to prevent.

ME-E-02Surface EMG amplitude is not motor unit recruitment, and it is not hypertrophy

FindingConclusions drawn by comparing sEMG amplitudes across muscles, exercises, techniques and loads, and used to hypothesize about longitudinal adaptations such as strength and hypertrophy, are frequently unsubstantiated and unwarranted [9]. Using "motor unit recruitment" synonymously with EMG amplitude is an incorrect assumption and a common mistake in sports and exercise science, particularly under fatiguing and dynamic conditions, and greater electromyographic responses do not imply greater motor unit recruitment, nor can hypertrophic potential be inferred from them [10].

Evidence gradeC

PopulationMethodological reviews, no primary sample.

Sources[9][10]

CaveatsThis is the most important entry in the document, and it undercuts the evidential basis of Part A's exception and most of Part C. Every acute EMG row here is a measurement of a signal under a condition, not a measurement of growth. MVIII cites activation research to decide which cue to say, which is a low-stakes use, and never to decide which exercise builds more muscle, which is the use these two papers refuse. Graded C as methodological argument rather than trial evidence, and load-bearing anyway. # PART C — Position

ME-R-01Training at long muscle lengths

PrescriptionWhere two variations of a movement differ in the length of the target muscle under load, MVIII selects the variation that loads it long.

Evidence gradeB

EffectForty-five untrained women trained knee extension in full, initial-partial (lengthened), final-partial (shortened) or daily-varied ranges. The lengthened-partial group showed greater relative increases than all other groups at 70% of femur length, and greater than shortened-partial, full-ROM and control at 50% and 60%; the shortened-partial group was similar to non-training control at 60% and 70%. Strength gains were specific to the range trained [11]. Fourteen untrained adults trained one leg standing (gastrocnemius lengthened) and the other seated (shortened) for 12 weeks; muscle volume increased in the whole triceps surae for both legs except the gastrocnemius of the seated leg, with greater changes for the standing condition [12]. A systematic review and meta-analysis of whether muscle length influences regional hypertrophy is cited in agreement [13].

Population45 untrained women [11]; 14 untrained adults, within-subject [12].

Sources[11][12][13]

CaveatsBoth primary trials are in untrained participants, and both are single-joint. The regional measurement in [11] is the point: an effect at 70% of femur length and no effect at another site means "the muscle grew" is the wrong resolution for this question. The within-subject design in [12] is a strength, since each participant is their own control.

ME-R-02Range of motion

PrescriptionFull range unless a stated reason exists to shorten it. Partial work is programmed at long muscle lengths rather than short ones, per ME-R-01.

Evidence gradeA

EffectA systematic review and meta-analysis of range of motion and resistance training adaptations is cited [14], in agreement with the regional findings at [11] and [13].

PopulationSystematic review and meta-analysis of range of motion and resistance training adaptations; the prescription itself is specified in the bodybuilding document. [14]

Sources[11][13][14]

CaveatsThis row exists so that the cueing document does not contradict the training documents. The primary specification of range of motion lives in bodybuilding-programming.md; this entry cites the same works through this document's numbering rather than restating the prescription.

ME-R-03Grip width in the lat pulldown

PrescriptionMVIII does not prescribe a wide grip for latissimus development. Grip width is a comfort and strength variable, not an activation lever.

Evidence gradeC

EffectFifteen men performed 6RM lat pulldowns at narrow, medium and wide grips (1, 1.5 and 2 times biacromial distance). 6RM load was higher with narrow (80.3 ± 7.2 kg) and medium (80 ± 7.1 kg) grips than wide (77.3 ± 6.3 kg; p = 0.02). There was similar EMG activation between grip widths for latissimus, trapezius and infraspinatus across the whole movement, with greater latissimus and infraspinatus activation for wide versus narrow in the eccentric phase only (p ≤ 0.04) [6]. A separate study separating grip width from forearm orientation found the two variables had been conflated in earlier recommendations [15].

Population15 men [6]; 12 healthy men [15].

Sources[6][15]

CaveatsThe authors of [6] state plainly that the belief a wider grip activates latissimus more is a general belief "without any broad scientific support". Small samples, acute EMG, and subject to ME-E-02. The practical finding is the load one: you are weaker with a wide grip and no better activated across the full movement.

ME-R-04Squat variation and gluteal activation

PrescriptionMVIII does not select between front, full and parallel squats on the basis of gluteal activation. Selection is on load tolerance, equipment and the range-of-motion rule at ME-R-02.

Evidence gradeC

EffectThirteen trained women performed 10 repetitions at estimated 10RM of front, full and parallel squats. There were no statistical differences between variations in any of the tested muscles: upper gluteus maximus, lower gluteus maximus, biceps femoris or vastus lateralis. The authors conclude the variations can be performed for similar EMG amplitudes, recommend full range on the basis of other research, and note the front squat may provide a similar stimulus to the back squat despite requiring lighter loads [5].

Population13 healthy trained women.

Sources[5]

CaveatsA negative finding in a sample of thirteen, which is exactly the design least able to detect a small difference. It is cited as a refusal to differentiate rather than as proof of equivalence, and that distinction matters. # PART D — Findings

ME-F-01External focus produces more output for less signal

FindingParticipants performing biceps curls under an external focus moved faster than under an internal focus, and integrated EMG activity was reduced. When movement time was equalized with a metronome, iEMG remained lower under external focus. The authors interpret this through the constrained-action hypothesis, that an external focus promotes more automatic control processes [8].

Evidence gradeC

Population11 and 12 participants across two experiments.

Sources[8]

CaveatsSmall samples. Included because it is the mechanism sentence for the whole of Part A: external focus is not a motivational trick, it is a measurable change in how movement is controlled.

ME-F-02Push-up variants load measurably differently

FindingKinetic and electromyographic analysis of push-up variants establishes that hand and foot position change the proportion of bodyweight supported [16].

Evidence gradeC

PopulationKinetic and electromyographic analysis of push-up variants; see the progression specification, which is where the ladder built on it is specified. [16]

Sources[16]

CaveatsCited here through this document's numbering; the progression ladder built on it is specified in progression-programming.md, not here.

ME-C-01How many cues to give in a set

FindingNo systematic review or controlled trial establishing an optimal number of verbal cues per set, or the point at which additional cues degrade performance, was located. MVIII caps the cue list at four bullets. That cap is a design decision.

Evidence grade—

PopulationNone.

SourcesNone.

CaveatsSearched for specifically. The attentional focus literature is about where attention goes and is silent on how much instruction to give at once.

ME-C-02Whether an app's cue works like a coach's cue

FindingNo trial was located comparing a cue delivered by software, on a screen or through audio, against the same cue spoken by a present coach. Every study in Part A used verbal instruction from an experimenter.

Evidence grade—

PopulationNone.

SourcesNone.

CaveatsThis is the gap under MVIII's entire cueing feature. The instructions in [1] through [8] were given by a person in a room. Whether a member reads a line on a phone mid-set and attends to it the same way is unknown.