REF-1486 · Moesgaard et al., 2022 · Sports Medicine · Paper · PubMed

Does periodizing my lifting beat keeping it constant?

Volume equated: slightly for 1RM (ES 0.31, CI 0.04-0.57), no size difference shown. Undulating beat linear overall; by subgroup only in trained, CI touching 0. No endurance athletes.

Main caveat: Abstract only. No heterogeneity, risk of bias, study sizes or durations, so this entry cannot judge whether trials were short or small.
fairly strong evidenceMeasured on non-athletes to highly trained athletes

Moesgaard and colleagues asked whether periodizing volume and intensity changes maximal strength and muscle growth once volume is the same between conditions, and how best to periodize.

They searched PubMed, Scopus and SPORTDiscus; 35 studies met the criteria; they pooled 1RM and hypertrophy for non-periodized versus periodized and for linear versus undulating, with subgroups by training status. Other model comparisons had too few studies. Periodized training raised 1RM more than non-periodized (ES 0.31) without a difference in hypertrophy (ES 0.13).

Undulating beat linear on 1RM (ES 0.31), and the subgroups put that in trained participants only (ES 0.61, lower bound 0.00), not untrained (ES 0.06); hypertrophy did not differ (ES 0.05). The authors propose that the strength effect is neurophysiological, which was not measured.

Our reading: equating volume is the paper's contribution, because it removes one reason for a periodized group to win, namely that it simply did more work. What remains is a small effect on 1RM with intervals close to zero. A subgroup with a lower bound of 0.00 is a hypothesis about trained lifters, not evidence for undulating load in general, and not for endurance athletes.

Protocol

Search: PubMed, Scopus, SPORTDiscus; dates not given in the abstract.
Inclusion: resistance training programmes with volume equated between conditions from pre to post intervention. How volume was equated (sets, repetitions, or sets × repetitions × load) is not stated in the abstract.
Studies and participants: 35 studies; total participants, the split by comparison, ages and sexes are not given in the abstract.
Comparisons pooled: (a) non-periodized vs periodized; (b) linear vs undulating. Others not pooled because of a low number of studies.
Outcomes: 1RM strength; muscle hypertrophy.
Subgroup: training status, on the linear vs undulating comparison in the results reported in the abstract.
Funding and interests: not available from the abstract; Crossref lists Lundbeckfonden as a funder.

Key Results

ComparisonOutcomeES (95% CI)Z, P
Periodized vs non-periodized1RM0.31 (0.04 to 0.57), favours periodizedZ = 2.28, P = 0.02
Periodized vs non-periodizedHypertrophy0.13 (−0.10 to 0.36)Z = 1.10, P = 0.27
Undulating vs linear, all1RM0.31 (0.02 to 0.61), favours undulatingZ = 2.06, P = 0.04
Undulating vs linear, trained1RM0.61 (0.00 to 1.22), favours undulatingZ = 1.97, P = 0.05
Undulating vs linear, untrained1RM0.06 (−0.20 to 0.31)Z = 0.43, P = 0.67
Undulating vs linearHypertrophy0.05 (−0.20 to 0.29)Z = 0.36, P = 0.72
What "volume-equated" did to the set of studies: it was an inclusion rule, so only designs matched on volume entered. The abstract does not say how many otherwise eligible studies this removed.
How this differs from [REF-683] (Williams 2017, CURRENT level A, not regraded): REF-683 pooled 81 effects from 18 studies of periodized vs non-periodized 1RM, 1988-2015, ES 0.43 (95% CI 0.27 to 0.58), and its entry does not record volume equating.Here volume is equated by design; 35 studies entered across all comparisons, and how many fed this one is not in the abstract, so it cannot be called a larger set; the estimate is lower (0.31) and the interval wider, though the two intervals overlap.REF-683 reports larger gains in untrained participants as an across-study moderator of periodized vs non-periodized; this paper reports an undulating-over-linear subgroup in trained participants. They are different contrasts, so they do not directly conflict, but they point to different people.Different study sets mean the drop from 0.43 to 0.31 cannot be read as the share due to volume (our inference). This paper adds hypertrophy, where REF-683 has nothing.
Against [REF-684] (Harries 2015, A-): 17 head-to-head studies, no linear vs undulating difference. This later review finds a small difference favouring undulating overall, with a lower bound of 0.02.
Against [REF-1476] (ACSM overview 2026; sibling draft): that overview cites this review as the one finding periodized programmes slightly favoured under volume-equated conditions, and rates the impact of periodization on strength as not determinable and on hypertrophy as absent.

Mechanisms

Measured: 1RM rose more with periodization without a pooled difference in hypertrophy.
Authors' interpretation: the effect of periodization on maximal strength may be related to neurophysiological adaptations; no neural outcome is reported in the abstract.
Inference: a 1RM test is skilled and load-specific, so programmes that spend some sessions near maximal loads may practise the test itself; this would fit strength without extra muscle, and it is not shown here.

Practical Application

Does periodizing my lifting beat keeping it constant? Volume equated: slightly for 1RM (ES 0.31, CI 0.04-0.57), no size difference shown. Undulating beat linear overall; by subgroup only in trained, CI touching 0. No endurance athletes.
What it supports: a small 1RM advantage for planned changes of load and volume in people whose goal is maximal strength, with no reason to expect more muscle from the structure itself (A-, general lifters).
What it does not support: telling an endurance athlete that undulating load is proven for them; the paper does not test them, and its subgroup rests on an interval starting at 0.00. It tests no rule for when to add load, how much, or when to change a scheme.
For a cyclist or runner who lifts: Not from this paper; the base's own practice: hold the guides' rep band through a block (SYNTH-43 4-6 reps, SYNTH-44 4-8 reps), which also keeps sessions comparable in the log. This paper gives no reason to add daily undulation for them.

Limitations

Abstract only. No heterogeneity, risk of bias, study sizes or durations, so this entry cannot judge whether trials were short or small; number of studies per comparison is unknown.
Effects near zero. The lower bounds are 0.04 and 0.02, and 0.00 in the trained subgroup; the abstract gives no formal test of the difference between trained and untrained subgroups, so "only in trained" rests on one interval that starts at 0.00 beside another that spans zero.
Trained is undefined in the abstract. Nor does it name a sport.
Strength test specificity. 1RM only; no performance outcome outside lifting.
No progression rule. It compares programme shapes, not decisions from a training log.
Interests. Not available from the abstract.
How this entry was checked and changed
Why this level

a meta-analysis of controlled comparisons (rubric A) held a step lower: read from the abstract only (closed access; OpenAlex marks it closed, and its two repository records hold no file), so heterogeneity, the number of studies in each comparison and the risk of bias are not available, and the lower CI bounds of the three effects that favour something lie at 0.04, 0.02 and 0.00