REF-1171 · Tanimoto & Ishii, 2006 · Journal of Applied Physiology · Paper · PubMed

Does slow, no-relaxation lifting at ~50% 1RM build muscle?

In 24 untrained men, 12 weeks of 3 s up / 3 s down to failure grew the quads like ~80% 1RM (+5.4% vs +4.3%); normal speed at that load did not. Small trial.

B+moderate evidenceMeasured on non-athletes

Tanimoto and Ishii designed a low-load lifting style that keeps the muscle under constant tension so that intramuscular pressure limits blood flow without a cuff, and called it LST (low-intensity resistance exercise with slow movement and tonic force generation).

Twenty-four untrained young men were randomly assigned to three knee-extension programmes, three sets three times a week for 12 weeks: LST at about 50% 1RM (3 s lifting, 3 s lowering, 1-s pause at full extension, no relaxing phase, sets to failure at 8RM); heavy normal speed (HN) at about 80% 1RM (1 s up, 1 s down, 1 s relaxed, also to failure at 8RM); and light normal speed (LN) at the same load and work as LST (8 repetitions at about 50% 1RM, not taken to failure).

Quadriceps CSA on MRI rose 5.4 ± 3.7% after LST and 4.3 ± 2.1% after HN, with no significant change after LN; LST and HN did not differ. Isometric strength rose in LST and HN, more in HN. 1RM rose in all three groups with no group difference.

In the acute tests LST kept quadriceps EMG almost continuous, caused the deepest muscle deoxygenation and gave blood lactate similar to heavy lifting, while peak systolic pressure was lower than with heavy lifting (HN 227.6 ± 24.9 mmHg).

How this relates to the Russian static-dynamic method: this is the closest tested form of the static-dynamic method with a fully published protocol (the Moscow IBMP no-relaxation trials, REF-1169 REF-1170, are closer in origin but their abstracts give no tempo or set structure): ~50% 1RM, 3 s up / 3 s down, a 1-s pause, no relaxation between reps, sets lasting roughly 56 s (8 reps of 7 s, our calculation) taken to failure, 60 s between sets.

It differs from the usual static-dynamic prescription in using a full knee range (0-90 degrees) with the pause at full extension rather than partial range, and in 60 s rather than about 30 s rests. It measured size, strength, oxygenation and lactate, not mitochondria or capillaries.

Protocol

Participants: 24 right-handed healthy young men without regular exercise training; randomly assigned to three groups of 8, matched for height, weight and knee-extension MVC. Age 19.0-19.8 y, body mass about 59-60 kg.
Exercise: seated knee extension on a nominally isotonic machine, range 0-90 degrees (0 = full extension); metronome-paced.
LST: about 50% 1RM (measured 51.7 ± 3.9% at the start, 52.6-55.1% later), 3 s concentric, 3 s eccentric, 1-s pause at full extension, no relaxing phase; each set to exhaustion with the load set at 8RM, adjusted every session.
HN: about 80% 1RM (81.8 ± 6.4% at the start), 1 s concentric, 1 s eccentric, 1 s relaxed; each set to exhaustion at 8RM.
LN: same load as LST (51.3 ± 1.7% 1RM), HN tempo, 8 repetitions per set, matched to LST for load and mechanical work; not to failure (our inference: 8 fast reps at a load the LST group could lift 8 times slowly). Impulse (force x time) in LST was about three times that in LN.
Volume: 3 sets, 60 s between sets, 3 sessions a week for 12 weeks.
Measurements: MRI quadriceps CSA (mean of three mid-thigh slices); isometric MVC at 80 degrees and isokinetic torque at 90, 200 and 300 deg/s; 1RM; acute VL EMG, VL NIRS oxygenation (occlusion-calibrated), fingertip lactate, continuous radial blood pressure.

Key Results

Outcome (12 wk unless stated)LST, ~50% 1RM slow, no relaxation (n = 8)HN, ~80% 1RM (n = 8)LN, ~50% 1RM normal speed (n = 8)
Quadriceps CSA (MRI)+5.4 ± 3.7% (significant)+4.3 ± 2.1% (significant)no significant change
Between groups, CSALST = HN; both > LN (P < 0.05)
1RM, kg101.0 → 129.4 (+28%, our calculation)104.9 → 138.3 (+32%)99.6 → 115.4 (+16%)
Between groups, 1RMno significant difference
Isometric MVCsignificant risesignificant rise; post value > LSTno significant change; post value < LST
Isokinetic torque 90 / 200 / 300 deg/sno significant changerise at 90 onlysmall rises at 90 and 200
Isometric torque per CSA2.58 → 2.67 (ns)2.63 → 2.90 (ns)2.73 → 2.79 (ns)
Acute: minimum VL oxygenationlowest (P < 0.05 vs HN and LN)intermediateintermediate
Acute: blood lactatesimilar to HNsimilar to LSTlower than LST and HN (P < 0.05)
Acute: peak systolic pressurelower than HN227.6 ± 24.9 mmHglower than HN; LST = LN
Test-retest reliability (9 subjects): CSA ICC 0.99, CV 1.1%; strength ICC 0.97, CV 2.4%.
Typical LST trace: knee torque almost constant at about 50% 1RM, roughly 40% MVC, with continuous VL EMG; HN and LN torque fell to zero in the relaxing phase.

Mechanisms

Measured: LST kept quadriceps activity continuous and produced the deepest deoxygenation and a lactate rise similar to heavy lifting (no significant difference, n = 8 per group), at about 60% of the heavy load.
Measured: isometric torque per unit CSA did not change in any group, so the isometric strength gains tracked hypertrophy.
Authors' interpretation: continuous force above about 40% MVC restricts blood inflow and outflow; the resulting hypoxia and metabolite accumulation, possibly via growth hormone and reactive oxygen species, drive growth. They cite their own separate experiment in which GH after LST matched HN and exceeded LN.
Authors' interpretation: LST does not improve neural function for fast movement; strength gains are velocity-specific, so a fast-velocity exercise should be added for dynamic strength.
Inference: the LST-vs-LN contrast changes tempo, continuous tension and proximity to failure at once (LST sets went to failure, LN sets did not), so it does not isolate tempo. That fits meta-analyses finding tempo itself matters little for hypertrophy when sets are taken near failure (REF-1176, REF-1177) and the base's load and failure entries (REF-677, REF-678).
Inference: NIRS deoxygenation during a set is not a measure of mitochondrial or capillary adaptation; no oxidative outcome was trained or measured.

Practical Application

Does slow, no-relaxation lifting at ~50% 1RM build muscle? In 24 untrained men, 12 weeks of 3 s up / 3 s down to failure grew the quads like ~80% 1RM (+5.4% vs +4.3%); normal speed at that load did not. Small trial.
The set must be hard. LST sets went to failure at 8RM; light normal-speed sets of a fixed 8 reps, not taken to failure and probably far from it (our inference), gave no significant growth. If you use the static-dynamic style, the burn and the near-failure finish are the stimulus, not the slow tempo alone (our inference).
For strength that shows up at speed, add fast work. LST did not raise isokinetic torque at 90-300 deg/s and raised isometric strength less than heavy lifting; the authors suggest pairing it with higher-velocity exercise.
A lower-stress option: peak blood pressure and external load were lower than with heavy lifting, which the authors suggest suits introductory training and rehabilitation.
For endurance athletes, this trial does not test aerobic adaptation; the base's default for strength and economy remains heavy, low-repetition work (SYNTH-43, SYNTH-44) — the base's synthesis, not this paper's.

Limitations

Eight per arm; one exercise (knee extension); untrained young men only; no blinding of assessors reported.
Isometric strength was tested at 80 degrees and isokinetic torque only up to 300 deg/s; no power or sport-performance test.
LN was not taken to failure, so LST vs LN confounds tempo, continuous tension and effort.
Acute oxygenation, lactate and blood pressure were reported mainly in figures; exact group means are not in the text.
Minor inconsistency: 3rd-month SDs for 1RM in Table 2 (HN 18.5, LN 17.5) differ from the text (18.6, 17.6).
Funding and conflicts of interest: no grant or conflict statement appears in the article text read (only the page-charge footnote).
How this entry was checked and changed
Why this level

parallel RCT with 24 participants (rubric B+: 20-39), but only 8 per arm, untrained young men, one exercise, no blinding reported; full text read on the publisher site

Full text (checked 2026-10-16)
Twenty-four right-handed and healthy young men without regular exercise training; randomly assigned into three groups (n = 8 for each group), matched for height, weight and MVC of knee extension (full text).
LST: 3 s for eccentric and concentric actions, 1-s pause, no relaxing phase; HN: 1 s for concentric and eccentric actions and 1 s for relaxing (full text).
Three sets with an interset rest period of 60 s, three times a week for 12 wk; LST and HN sets until exhaustion at 8RM; LN eight repetitions at about 50% 1RM (full text).
Pretraining % of 1RM: LST 51.7±3.9, HN 81.8±6.4, LN 51.3±1.7 (full text).
Impulse in LST about three times as large as in LN at the same intensity and amount of work (full text).
Knee range 0 to 90 degrees; isometric torque at a knee angle of 80 degrees; isokinetic at 90, 200, and 300 deg/s (full text).
CSA +5.4 ± 3.7% after LST and +4.3 ± 2.1% after HN; no significant change after LN (full text).
1RM from 101.0 ± 20.7 to 129.4 ± 11.6 kg in LST, from 104.9 ± 18.6 to 138.3 ± 18.6 kg in HN, from 99.6 ± 20.9 to 115.4 ± 17.6 kg in LN; no significant difference between groups (full text).
Isometric torque per CSA: 2.58 ± 0.29 to 2.67 ± 0.28 (LST), 2.63 ± 0.29 to 2.90 ± 0.39 (HN), 2.73 ± 0.56 to 2.79 ± 0.44 (LN), not significant (full text).
Peak systolic pressure during HN 227.6 ± 24.9 mmHg, significantly higher than LST and LN (full text).
LST knee torque almost constant at about 50% of 1RM, approximately equal to 40% MVC, with almost continuous VL EMG (full text).
Reliability: CSA ICC 0.99 and CV 1.1%; muscular strength ICC 0.97 and CV 2.4% (full text).