Muscles Worked by Exercise 3D Explorer

What muscles are activated during compound strength exercises?

Compound exercises move more than one joint, so several muscles share the work. The prime movers of the back squat are the quadriceps, gluteus maximus and adductor magnus; of the bench press, pectoralis major, the anterior deltoid and triceps brachii; and of the deadlift, gluteus maximus, the hamstrings, adductor magnus, the quadriceps and the erector spinae, which hold the spine still.

Prime mover, synergist, stabiliser: how a role is decided here

Every lift here is described joint by joint. First come the main joint actions the lift is made of, such as knee extension and hip extension in a squat. Then each muscle gets the one role that fits what it does in that technique:

• Prime mover. It produces the lift’s main joint action against the load.
• Synergist. It assists that action or adds a secondary one.
• Stabiliser. It contracts mostly isometrically to hold a joint or segment still.

A prime mover always sits under one of the lift’s joint actions, and no muscle has two roles in one lift. Where a prime mover works without changing length, the card says so: in the deadlift the erector spinae are prime movers that hold the spine still rather than move it. The roles follow kinesiology texts (Neumann; Kendall and colleagues; the NSCA’s Essentials of Strength Training and Conditioning) and the electromyography studies named on each lift, which settle how two variations differ: that a chin-up uses more biceps and a pull-up more lower trapezius (Youdas and colleagues, 2010), or that a sumo deadlift uses more of the vasti (Escamilla and colleagues, 2002).

Worked example: the back squat joint by joint

The back squat opens the page, done this way: barbell high on the upper back, shoulder-width stance, lowered until the thighs are parallel to the floor.

• Knee extension. The quadriceps, all four heads, straighten the knee out of the bottom against the load, and that joint action is what makes them prime movers. They lengthen under load on the way down and shorten on the way up.
• Hip extension. Gluteus maximus and adductor magnus drive the hips forward and up. The hamstrings help, but they also cross the knee, which straightens as the hip does; they lengthen at the knee as they shorten at the hip, so their length changes much less than that of gluteus maximus: synergists.
• The ankle. Soleus controls how far the shin tips forward, a synergist.
• Everything held still. Erector spinae, multifidus, rectus abdominis, external oblique, internal oblique, transversus abdominis, quadratus lumborum, gluteus medius, gluteus minimus, upper trapezius, middle trapezius and rhomboids: the trunk, the pelvis and the shoulder blades the bar rests on, which do their work without moving.

That is 3 prime movers, 2 synergists and 12 stabilisers, the counts beside the map. Going deeper or widening the stance moves more of the work to the hip extensors (Escamilla, 2001; Schoenfeld, 2010), which a role cannot show and the card notes where it matters.

Why muscle charts disagree

Two charts of the same lift often light different muscles, and both can be right, because each assumes a technique. Depth, stance width, grip, bar position, tempo and load all shift the work between muscles, and so do the lifter’s proportions. That is why every lift here states the technique its roles are for, under the map.

Charts are also often read as rankings of growth, which they are not. Surface electromyography measures how strongly a muscle is activated, and the signal depends on where the electrodes sit, how it was normalised and the person; a larger amplitude does not necessarily mean more force or more growth (Vigotsky et al. (2018), Front Physiol 8:985). So the studies cited here are used for which muscles a lift involves and how two variations differ, never to put a percentage on a muscle. How much a muscle grows depends mostly on its weekly volume and how close sets go to failure, and is much the same across a wide range of loads (Schoenfeld and colleagues, 2017), which no chart of a single lift can say.

Deep muscles the flat map cannot show

The maps are traced from the meshes as an orthographic camera sees them, so a muscle under another is hidden, as it is on a body. In both views the map cannot show any of the 10 muscles or muscle groups that lie wholly under others: levator scapulae, rhomboids, multifidus, quadratus lumborum, pectoralis minor, supraspinatus, subscapularis, supinator, internal oblique and gluteus minimus. Parts of others are hidden while the rest shows, such as the vastus intermedius under the other three heads of the quadriceps, and abductor digiti minimi of the hand is under the skin the map draws over the hands. Tick Deep muscles as outlines and every muscle the lift uses that is mostly hidden is drawn as a dashed outline of its whole shape, in its role’s colour, and the 3D model below shows each in place.

Three thin sheets are drawn as if transparent, because the meshes join each to its muscle. The external oblique carries its aponeurosis, which forms the front of the rectus sheath, so rectus abdominis would otherwise be hidden behind it, and is seen through to 15 mm. Latissimus dorsi carries the thoracolumbar fascia over the erector spinae, seen through to 20 mm, and platysma, a skin muscle, lies over sternocleidomastoid, seen through to 10 mm. Deeper than that, a muscle stays hidden.

Every lift at a glance

All 25 lifts, in the 9 movement patterns the picker uses, with the muscles in each role, from the same records the map is coloured from.

About the model

The muscles are 168 meshes, each muscle on both sides, taken from five of the six BodyParts3D muscle models: one adult male body, segmented from imaging. They are grouped into 55 muscles and muscle groups, 47 of them named by the lifts and the rest there for the root mode or for context. The muscles and the roles apply to everyone; the shapes are this one person’s.

At full resolution these meshes come to 10.39 million triangles, more than the 3D viewer’s 8 million prompt, because one lift reaches across several regional models: a squat alone uses muscles from the leg, back and abdominal models. So each mesh is simplified at the largest step, from 0.7 mm down, that keeps every original vertex within 0.7 mm of the simplified surface; the right external oblique, which no step kept that close, is shipped whole. That leaves 4.89 million triangles, the farthest original vertex 0.667 mm from the surface, and the download is about 18 MB. Nothing is fetched until you press the button, and the model is kept on your device afterwards with a button to remove it.

The flat maps do not use the simplified meshes: they are traced from the full-resolution meshes of all six muscle models and the skeleton, with the skin over the head, hands and feet, and shaded from the meshes’ own surface normals. More of the body’s muscles are in the 3D models of each region: the lower limb, upper limb, back and abdominal muscle explorers.

What makes a muscle pull, and how much it turns a joint

A role says which muscles work; how they make force is the subject of the muscle twitch and tetanus simulator, and why the same muscle is stronger at one joint angle than another is moment arms, which the torque calculator works through. The spinal roots that supply each of these muscles are on the dermatome and myotome map, and the root mode of this page colours the map by them.