Stretch grows muscle longer, not thicker — and mTORC1 was the switch in rodent diaphragm

Start with what this study is not. It is not humans training in a stretched position. It is unilateral diaphragm denervation (UDD) in rats and mice, an in vivo model that holds muscle under mechanical stretch, used to ask what stretch itself switches on.

But the question that model answers overlaps precisely with the one lifters argue about. What does loading a muscle in a lengthened position actually build? The answer here is specific: not thickness, but length.

What longitudinal hypertrophy means

Muscle does not grow in only one direction. Fibres thickening and sarcomeres being added in series so the muscle gets longer are different adaptations. This study is about the second — longitudinal hypertrophy — and the UDD model is stretch-specific in that the response it triggers depends on titin stiffness.

To separate stretch signalling from the effects of denervation itself, the researchers performed unilateral (UDD) and bilateral (BDD) diaphragm denervation in rats and ran global transcriptomic and proteomic analyses on both. Without that contrast, ‘changes caused by stretch’ and ‘changes caused by losing the nerve’ stay mixed together.

Titin senses it, MARPs hold it back

What came up in that comparison was upregulation of titin-associated muscle ankyrin repeat proteins (MARPs). Phosphorylation enrichment mass spectrometry in mouse diaphragm then pointed at titin's N2A element. Titin is the giant protein that senses stretch inside the sarcomere, and N2A is the sensing site.

Then the interesting result. In MARP knockout mice, the same UDD stimulus did not reduce longitudinal hypertrophy — it produced more of it, with Western blot confirming mTOR pathway activation. MARPs are therefore not what drives the growth; they are what constrains it. The growth response to stretch comes with a brake fitted as standard.

Block mTORC1 and the length growth disappears

The final test was pharmacological. Inhibiting mTORC1 with rapamycin suppressed longitudinal hypertrophy. That closes the loop: stretch sensed by titin runs, in a MARP-dependent way, through mTOR signalling to produce growth in length. The authors' conclusion is that mTORC1 is the central regulator of longitudinal muscle hypertrophy.

Treat mTORC1 as a dial that volume turns up and this result reads badly. A separate experiment in which mTORC1 was kept switched on did not grow muscle evenly at all — that one is covered in mTORC1 is not a volume knob. Read together, the picture is consistent: mTORC1 is a junction that helps determine which stimulus produces which kind of growth, not a pedal that grows muscle the harder you press it.

So what changes in training?

The honest answer is nothing, for now. This study prescribes no set count, no range of motion, no rep scheme. What it does change is the shape of the argument. If lengthened-position work is special, the specialness may not be more growth but a different kind of growth — and a tape measure around the limb would never detect it.

That makes the logging advice modest too. Longitudinal adaptation is not visible in a tape measure, a mirror, or a Big 3 1RM. What is trackable is performance at the end of the range — how a full-depth squat or a controlled fully-lengthened position feels and moves — and those are comparable over time only if you write them down. Thickness and length being separate axes is the practical perspective this study leaves behind.