REF-1488 · Halonen et al., 2024 · Scandinavian Journal of Medicine & Science in Sports · Paper · PubMed
Does a long break from lifting cost me for good?
In 42 untrained adults, 10 on, 10 off, 10 on ended with no clear difference from 20 straight weeks; the loss was back within 5 weeks. Not tested in trained athletes.
Main caveat: No clear difference is not equality. Power came from training vs control figures, not an equivalence margin, and the between-group CIs overlap widely.
Halonen and colleagues randomised 55 healthy, untrained adults to periodic training (PRT: 10 weeks lifting, 10 weeks detraining, 10 weeks lifting) or continuous training (CRT: 10 weeks without lifting, then 20 weeks lifting); 20 and 22 completed the comparison. Both did the same supervised programme twice a week, with the same 40 planned sessions and similar volume load.
Over those 20 weeks of lifting both groups improved 1RM, muscle area and jump height with no significant group-by-time differences, against the authors' hypothesis that continuous training would win. The detraining cost measurable strength and size: 1RM by 5.4% and 3.6%, muscle area by 9.9% and 7.3%.
After five weeks of retraining no measure differed significantly from the group's own week-10 value, and over those five weeks gains in leg press and muscle area were larger than the continuous group's over their training weeks 10-15 (calendar weeks 20-25).
Our reading: "regained rapidly" means 1RM and ultrasound muscle area no longer differing significantly from the earlier peak after five weeks of two sessions a week, in beginners. This is a reassurance about one voluntary break in early training, and no clear difference is what was found, not proof of equality.
The authors themselves note the continuous group slowed after 10 weeks, perhaps partly because the programme was monotonous. The study also prints a repetition-based weekly load table, which was used in both arms and not compared with anything.
Protocol
Design: randomized, parallel-group, 1:1, matched pairs by combined Z score of BMI and age, within each sex; the first 10 weeks also served as an RCT of training vs no training. Measures were taken twice at the beginning and every fifth week, not during the control and detraining periods. Registered NCT05553769; ethics University of Jyväskylä.
Who: n = 42 completed the comparison (PRT n = 20, CRT n = 22) of 55 eligible volunteers; 45% women in each group. Inclusion: age 18-40 y, fewer than 10 resistance sessions a year, not doing systematic endurance training (fewer than 2 endurance sessions a week lasting under 30 min, for the last 6 months), BMI 18.5-30 kg/m².PRT age 32.9 ± 5.8 y, body mass 78.2 ± 14.9 kg; CRT age 31.5 ± 4.0 y, 73.2 ± 14.0 kg. Dropouts: 4 from the first training period, 4 PRT and 5 CRT later.
Training: two supervised whole-body sessions a week, at least 48 h apart, with 2 min rest between sets: leg press 4 × 8-10, knee extension 4 × 8-10, Smith bench press 3 × 8-10, barbell biceps curl 4 × 8-10, chest-supported row 4 × 8-10. Concentric as fast as possible, eccentric about 2 s.Sets stopped at about 1-2 repetitions in reserve, except the last set of each exercise in the second session of each week, taken to volitional failure. Starting loads 70% 1RM (leg press, Smith bench, row) and 50% 1RM (curl, knee extension). Every fifth week the second session was replaced by the tests, which also acted as a small volume deload.Compliance at least 92.5% of sessions.
Detraining: habitual lifestyle, with no resistance or endurance training or other unaccustomed exercise for 10 weeks; checked by a survey and a mid-period email.
Outcomes: leg press and barbell biceps curl 1RM (test-retest CV 3.1% and 4.2%); ultrasound cross-sectional area of vastus lateralis and biceps brachii (CV 1.6% and 2.3%); countermovement jump height (CV 4.0%).
Analysis: two-way repeated-measures ANOVA with sex and baseline as covariates; Holm-Bonferroni post hoc. Sample size from earlier studies with 10-20 per group for training vs control.
Funding and interests: the paper prints funding from the Rehabilitation Foundation Peurunka and Suomen Urheilututkimussäätiö, and that J.P.A. was funded by the Research Council of Finland (grant 357185). The authors declare no conflicts.
Key Results
| Outcome | PRT week 0 → 10 → 20 (after the break) → 25 → 30 | CRT week 10 → 20 → 25 → 30 | Week 30 change from pretraining, 95% CI: PRT vs CRT |
|---|---|---|---|
| Leg press 1RM (kg) | 165.5 → 197.6 → 187.5 → 201.5 → 209.3 | 165.1 → 195.7 → 202.4 → 207.0 | 43.8 (33.8-53.7) vs 41.9 (36.4-47.4) |
| Biceps curl 1RM (kg) | 28.0 → 34.2 → 32.8 → 35.1 → 36.7 | 28.9 → 33.5 → 35.3 → 36.2 | 8.8 (7.1-10.4) vs 7.7 (6.3-9.1) |
| Vastus lateralis area (cm²) | 26.2 → 30.3 → 27.2 → 30.2 → 31.3 | 27.2 → 32.6 → 32.9 → 33.6 | 5.2 (3.8-6.5) vs 6.4 (5.4-7.4) |
| Biceps brachii area (cm²) | 9.1 → 10.6 → 9.8 → 10.7 → 11.3 | 9.0 → 10.7 → 11.1 → 11.4 | 2.2 (1.8-2.7) vs 2.4 (2.0-2.8) |
| Jump height (cm) | 25.7 → 30.4 → 28.4 → 29.8 → 30.5 | 27.6 → 30.2 → 30.5 → 31.2 | 4.8 (3.4-6.1) vs 3.7 (2.2-5.1) |
Between groups: no significant Group × Time difference over the 20 weeks of lifting, and also none when analysed over 30 calendar weeks. No difference in total volume load (PRT 862 275 ± 187 295 kg vs CRT 891 115 ± 183 946 kg, p = 0.618). Week 20 vastus lateralis area differed between groups, as expected, because one group had just stopped.
Loss during the 10 weeks off (PRT): leg press 1RM −5.4% ± 4.4%, biceps curl 1RM −3.6% ± 6.8%, vastus lateralis area −9.9% ± 4.1%, biceps brachii area −7.3% ± 3.9%, jump height −6.9% ± 5.3%. All stayed above pretraining values, though vastus lateralis area only just (p = 0.03). Muscle area fell more than 1RM in the legs (p = 0.01).
Second 10 weeks of lifting, PRT vs CRT: leg press 12.2% ± 5.4% vs 5.9% ± 3.5%; vastus lateralis area 15.1% ± 4.8% vs 3.3% ± 4.0%; biceps brachii area 15.1% ± 6.1% vs 6.8% ± 4.1% (all p < 0.001); biceps curl 12.2% ± 6.2% vs 9.3% ± 4.6% (p = 0.11). The larger gains arose in the first five weeks of it. Those five weeks were not larger than the same group's first five weeks of lifting ever.
10 weeks of training vs no training (first phase, n = 24/23 vs n = 27): leg press 1RM difference between groups 26.4 kg (19.2-33.7), p < 0.001.
The load rule used in both arms, not tested (Table 1). The repetitions reached in the failure set set next week's load, in kg:
| Reps in the failure set | Leg press, knee extension | Smith bench | Biceps curl | Row |
|---|---|---|---|---|
| <5 | −7.5 | −5 | −2.5 | −5 |
| 6-7 | −5 | −2.5 | −1 | −2.5 |
| 8-10 | 0 | 0 | 0 | 0 |
| 11-12 | +2.5 | +2.5 | +1 | +2.5 |
| 13-15 | +5 | +5 | +2.5 | +5 |
| 16-20 | +7.5 | +7.5 | +3.5 | +10 |
| >20 | +10 | +10 | +5 | +15 |
Mechanisms
Measured: muscle size fell more than maximal strength during the break, and regain was fastest in the first five weeks back.
Authors' interpretation: the fast regain may support muscle memory, or it may be resensitization of muscle after a break; because the first five weeks of retraining were not faster than the first five weeks of lifting ever, they call the mechanism unclear. Neural adaptations may explain retained strength, and were not measured.
Inference: this does not contradict REF-1360 (a previously trained leg did not regrow faster than an untrained one after 20 weeks); both fit a comeback that is fast against a plateau, not faster than a beginner's start.
Inference: the continuous group's slowing after 10 weeks is part of why the schedules came out level; a less monotonous programme might have separated them (the authors raise this).
Practical Application
Does a long break from lifting cost me for good? In 42 untrained adults, 10 on, 10 off, 10 on ended with no clear difference from 20 straight weeks; the loss was back within 5 weeks. Not tested in trained athletes.
After a voluntary break: expect some loss and expect it back. Muscle area and 1RM were back at their previous level within five weeks of two sessions a week, after 10 weeks without lifting (B+, untrained adults). Our reading: a reason not to panic, not a schedule; the study tests no restart load.
Which restart load: the study does not test one; both blocks began from a new 3-5RM test or the pretest 1RM. Not from this paper; the base's own practice: restart below the last load and rebuild; the reason for the break decides the return (SYNTH-36).
The Table 1 repetition rule is a worked example of a kg-and-reps rule: add load when the last set overshoots the rep band, cut when it falls short. It served in a trial that worked, but was not compared with another rule, was built for 8-10 repetitions in untrained adults, and needs a set to failure each week, which SYNTH-43 and SYNTH-44 avoid for endurance athletes.
Do not extend it: not to illness or injury (the authors exclude these), to trained or older people, or to the claim that one should plan breaks.
Limitations
No clear difference is not equality. Power came from training vs control figures, not an equivalence margin, and the between-group CIs overlap widely; the week 30 vastus lateralis gain was numerically smaller in PRT.
Who. Untrained adults, 18-40 y, without endurance training. The authors call for research in trained people. Early lifting may respond unusually fast.
Attrition. 13 of 55 did not complete; analysis was of completers.
Measures. Only two exercises and two muscles; leg press is a machine and curl a single joint. The authors note that free-weight strength may fall further. No nutrition control. Activity during the break was self-reported.
Reporting. The Results say no significant differences "(p ≤ 0.150)", which reads as a typo for p ≥ 0.150 (our reading), and the abstract writes "CTR" for CRT; Table 1 has no row for exactly 5 repetitions.
Interests. None declared; no device or commercial funder.
How this entry was checked and changed
Why this level
randomised, 42 completers in the comparison; held below A- because the sample-size calculation addressed training vs control, not equivalence of the two schedules, and 13 of 55 did not complete. Read from the full text (publisher's version, CC BY, self-archived at the University of Jyväskylä; the publisher's own link returned a bot check, which was not bypassed)
Full text (checked 2026-10-30)
Age 18–40 years, fewer than 10 RT sessions a year; endurance training was an exclusion (full text)
Two whole-body sessions a week, at least 48 h apart, 2-min rests, 8–10 reps at about 1–2 repetitions in reserve, with a failure set in the second session of each week that set the following week's load (full text)
Total volume load over the 20 weeks of RT: PRT 862 275 ± 187 295 kg vs CRT 891 115 ± 183 946 kg, p = 0.618 (full text)
During detraining, leg press 1RM −5.4% ± 4.4%, biceps curl 1RM −3.6% ± 6.8%, VL CSA −9.9% ± 4.1%, BB CSA −7.3% ± 3.9%, CMJ −6.9% ± 5.3% (full text)
Second block PRT vs CRT: leg press 12.2% ± 5.4% vs 5.9% ± 3.5%, VL 15.1% ± 4.8% vs 3.3% ± 4.0%, BB 15.1% ± 6.1% vs 6.8% ± 4.1%, biceps curl 12.2% ± 6.2% vs 9.3% ± 4.6%, p = 0.11 (full text)
Leg press 1RM at week 30: PRT 209.3 ± 50.7 kg, change 43.8 (33.8–53.7); CRT 207.0 ± 44.5 kg, change 41.9 (36.4–47.4) (full text)
The authors say results cannot be directly applied to breaks caused by illnesses or injuries and that trained populations need study (full text)
Conflicts: none declared; thanked for financial support: Rehabilitation Foundation Peurunka and Suomen Urheilututkimussäätiö (full text)