How strength adapts

Start lifting and you'll get stronger within a few weeks, well before anything looks different in the mirror. Then, some months in, the progress slows. This lesson explains what's changing inside you in those first weeks, why the fast start doesn't last, and why "stronger" always means stronger at something. That last point gives this course the question it will ask of every training claim from here on, because it's the question that decides whether a study's result applies to you.

Starting is the big effect

Two terms first, because the whole lesson uses them. Resistance training is any training where muscles work against a load: a barbell, a machine, a band, your own body. The one-repetition maximum, or 1RM, is the most weight you can lift once in a given exercise, and it's the standard test of strength in this literature. It's a research test. You don't need to find yours, and lesson 7 shows how to estimate it from a lighter set.

In 2026 the American College of Sports Medicine replaced its 2009 advice on resistance training with a new position stand. It is an overview of reviews: its authors pooled 137 systematic reviews of randomised trials, covering more than 30,000 participants, all healthy adults training for at least six weeks.¹ This course read the full text, which is free, and it is the anchor for this lesson, the next one and lesson 7.

Lesson 1 quoted its abstract's headline result: compared with no exercise, resistance training improved strength, muscle size and a long list of physical functions. The sentence straight after it says the same thing Currier's team found: "Few RT prescription (RTx) variables affected primary adaptations."¹ RT is resistance training; a prescription is the recipe, meaning the load, the sets, the days and so on. The stand's summary table lists circuit training, elastic bands and home-based training among the forms that improve strength compared with no exercise.¹

So training beats not training, on almost everything measured, and arguing about the recipe buys you much less than starting does, at least over the weeks these trials ran, in mostly untrained people.

Mental Fitness lesson 2 taught you to ask what the comparison group got, and this is that lesson arriving in a gym. Against a group that did nothing, every programme wins. Against another programme, almost nothing can be told apart. When somebody tells you a programme "works", ask which of those two comparisons they mean, because the answer changes the size of the claim by a lot.

One more thing belongs here, because this lesson leans on both papers. Two of the 2026 stand's authors, Currier and Phillips, also led the 2023 analysis, which the stand cites, so the stand is partly a summary of its own authors' work. Lesson 1 mentioned this alongside the stand's one declared interest, an author's former advisory role with a fitness-equipment maker. Neither is a reason to distrust the findings, and both are things you'd want to know.

The first weeks: the nervous system learns first

A beginner lives through a puzzle. In the first month the weights go up almost every week, and nothing has visibly grown. If strength were only a matter of muscle size, that couldn't happen.

The standard explanation is that the nervous system changes first. A muscle is controlled by motor units, each a single nerve cell in the spinal cord together with the fibres it switches on. That definition is standard physiology this course supplies, not a finding from the papers below, but it makes their findings readable.

The idea is old. In 1979 Moritani and deVries trained seven young men and eight women for eight weeks, and this course read the abstract only. It reads: "The results indicated that neural factors accounted for the larger proportion of the initial strength increment and thereafter both neural factors and hypertrophy took part in the further increase in strength, with hypertrophy becoming the dominant factor after the first 3 to 5 weeks."⁴ Hypertrophy is muscle growth. Their participants trained one arm, and the untrained arm got stronger too, which they read as "cross education". The abstract does not say how they estimated muscle growth, so this lesson cannot tell you.

In 1988 Sale put the mechanism in one sentence, again read at abstract level:

"Strength training may cause adaptive changes within the nervous system that allow a trainee to more fully activate prime movers in specific movements and to better coordinate the activation of all relevant muscles, thereby effecting a greater net force in the intended direction of movement."⁴

Notice three phrases in it. "More fully activate prime movers" means switching on more of the main muscles doing a movement than you could before. "Better coordinate" means all the muscles involved working together towards one force. And "in specific movements" is the seed of the specificity section below.

The modern account is a 2021 review by Škarabot and colleagues, which this course read in full.⁴ It keeps the hedge: "the initial (< 2–4 weeks) increases in force production are thought to be primarily underpinned by neural adaptations". It lists the evidence for that, including strength gains that are specific to the trained task without much change in the muscle, force rising faster than muscle size, cross-education, and strength gains after weeks of only imagined contractions. And it reports what recordings of single motor units show: "increased force production following resistance training is accompanied by decreased motor unit recruitment threshold and increased discharge rate".⁴ In plain terms, after training, motor units switch on at lower forces and fire faster.

Then it says what nobody knows yet: "the precise site of putative neural adaptations remains unclear."⁴ Brain, spinal cord or the connections in between, the review can't say. It also warns that most of the studies it reviews rely on "simplified models of resistance exercise", usually tested with "single-joint/single-muscle isometric contractions".⁴ Isometric means pushing against something that doesn't move, which is a long way from a squat.

Two timings, then: Moritani's hypertrophy becoming "the dominant factor" after three to five weeks, and Škarabot's neural dominance in the first two to four. They do not quite agree, and they are not fully independent either, since the review cites Moritani's study for its own figure. The honest summary is the first few weeks rather than a number.

Why the fast start doesn't last

Put the two phases together and the slowdown makes sense. Learning to use the muscle you already have is quick, and it runs out, because once you can switch on most of a muscle there is less left to learn. Growing new muscle is slow. That reading is this course's inference from the timing in the sources above, not a result any one of them measured, and its second half assumes growth adds to strength, which is the dispute below. Lesson 5 shows that each extra set in a week buys less, and lesson 7 comes back to why progress slows over months and what to do when it stalls.

There's one more thing, and it's a live dispute rather than a settled point. You'd think bigger muscles obviously cause more strength. One research group, led by Loenneke, argues the experiments don't show it: "at present, there is no experimental evidence that would support any paradigm in which muscle hypertrophy is a mechanism for increasing strength with exercise." Their reason is that the two come apart too often. They report "dissociations between changes in muscle size and strength" after exercise, drugs and ageing, "suggesting that these changes may be completely separate phenomena." They are careful about what they aren't saying: "We are not dismissing the potential importance of maintaining muscle mass, particularly in clinical populations. What we are suggesting, however, is that muscle function may not necessarily be improved by these exercise or pharmacological induced increases in muscle size."⁵ Loenneke is also the senior author of the specificity meta-analysis in the next section.

In a published exchange with that group, Taber, Vigotsky, Nuckols and Haun argue that growth is one contributing cause among several. Nuckols is also a practitioner whose work lesson 7 uses. They grant part of the other side's case: "We concede that strength can increase without increases in muscle size (not necessary)", and that "an increase in muscle size may not be accompanied by an increase in strength (not sufficient)". They think the better question is not whether but how much: "We, therefore, find the question of “Does hypertrophy contribute to strength gain?” less interesting than “To what extent and under what circumstances does hypertrophy contribute to strength gain?”" They conclude "that in the long-term muscular hypertrophy contributes to strength", and they add that "short-term studies utilizing untrained subjects may not fully capture the influence of hypertrophy on strength."⁵

So the two sides agree on more than it first looks. Bigger and stronger tend to go together, and either can change without the other. They disagree about what those mismatches mean: the Loenneke group reads them as a sign that size and strength may be separate, and the Taber group reads them as other causes hiding a real contribution from growth. This course doesn't take a side. Where this lesson mentions growth taking over after the first weeks, that is Moritani's reading, and whether the growth adds to strength is this dispute.

Stronger at what? Specificity

Say you train the bench press, lying on a bench and pressing a barbell up from your chest, for eight weeks, and then test your bench press. You have spent eight weeks practising the exact thing you are being tested on.

A 2023 meta-analysis by Spitz and colleagues asked what that does to the result. It pooled twelve studies of healthy adults that had a non-exercising control group and tested strength twice: once in the movement people trained, and once on a different device, a machine that measures force in a way they had not practised, using a muscle they had trained. This course read the abstract only.⁶

The dots are the estimates and the lines through them are the 95 per cent confidence intervals. They overlap a little, between 1.23 and 1.4, so read the gap as large but imprecise, not as an exact ratio.

The same pattern turns up again in a comparison of equipment. A 2023 meta-analysis of free weights against machines, which this course read in full, found that "strength changes are specific to the training modality": free-weight training improved free-weight tests more, machine training tended to improve machine tests more, and in direct comparison neither won.⁷ So the choice between them, in the authors' words, is "down to individual preferences and goals".⁷

The 2026 stand puts the other half of this in its own words: "there is a considerable carryover of training effects on general muscular performance in various domains in nonadvanced trainees".¹ So the practised lift overstates the gain; it doesn't invent it.

Now the part that matters for reading any study in this field. The stand says what its strength numbers mean: "Unless otherwise specified, strength was voluntary isotonic one-repetition maximum (1RM) in the same mode as that in which the training was performed."¹ Isotonic means lifting a load through a movement. In other words, most of the strength evidence in the most important document in this lesson is the practised lift, tested. That does not make the evidence wrong, but it's worth knowing before you read "strength rose" anywhere in it.

What 24 per cent is a measurement of

This trial is constructed for teaching, and no source read for this course reports it.⁸

Suppose a trial takes untrained young men, has them bench press twice a week for eight weeks, and reports that their bench press 1RM rose from 50 kg (110 lb) to 62 kg (137 lb), a gain of 24 per cent. A fitness article turns that into "eight weeks of bench pressing makes you 24 per cent stronger". Take it apart one question at a time.

What was measured? The bench press, the lift they practised. Spitz's meta-analysis found the effect on a trained movement more than twice the size of the effect on an unpractised test, so a good share of that 24 per cent is likely to be skill at benching. How big a share, for this trial, you cannot say: effect sizes aren't percentages, and one meta-analysis can't be converted into a correction for one study.⁶

On whom? Untrained young men, close to the typical population in this literature, which is young, untrained and more often male, as the table below shows. The trial says nothing direct about a trained lifter, an older woman, or anyone after week eight.

What survives? Something true and worth knowing: untrained young men got much better at benching in eight weeks. What doesn't survive is "24 per cent stronger", as if the whole 24 per cent would show up on a test they'd never done.

The course's question: trained whom, for how long, measured on what?

You've just watched it work, and lesson 3 used it on the one-minute interval study. From here it's this course's standing question, its addition to the institute's claim sort, and every evidence lesson after this one uses it.

Trained whom, for how long, measured on what?

It overlaps with Memory's sample question, and the difference fits in one sentence: Memory's question asks how many people and who they were, while this one asks who, for how long, and scored on which test, because in training the duration and the test change the answer as much as the people do. It also overlaps Focus and Deep Work's instrument question, and the difference is what makes this one a training question: in a training study the test is often the very movement people practised, so the instrument improves along with them. That's specificity, and it's why "measured on what?" has to be asked separately here.

So who is in this literature, and for how long? The table is assembled from the meta-analyses' own descriptions of the studies they pooled, because no single paper describes the whole field.⁹

Read down the columns. Apart from the one analysis limited to older adults, the typical study is two to three months long, in untrained people in their twenties, more often men, and, from the section above, tested on the lift they practised. When somebody says "the research shows" about sets or reps, that is usually the population and the duration they mean.⁹ Lesson 5 takes those sets and reps one at a time, with this table beside it.

The stand's authors would add a caution of their own. In their 2023 analysis, they write, "training experience had a minimal impact on strength and hypertrophy outcomes", and they call their recommendations "evidence-based across all ages".¹ So the question doesn't say the findings are wrong for you. It says what they were measured on.

This matters most for questions a short study can't see. The Taber group's point was that "short-term studies utilizing untrained subjects may not fully capture the influence of hypertrophy on strength."⁵ If they're right, a literature of eight-week studies in beginners is poorly placed to see what growth adds. The Loenneke group, who find size and strength coming apart "throughout the literature", see no experimental evidence yet of a contribution waiting to be found.⁵ Either way, most of what the table describes is short.

There is one long record, and it is worth knowing what kind of thing it is. Steele and colleagues analysed the training logs of "14,690 participants (60% female; aged 48 ± 11 years)" at a private exercise company, doing "1x/week, single sets to momentary failure of six exercises" for up to 352 weeks, about 6.8 years.⁹ They report "approximately ~30-50% gains over the first year reaching ~50-60% of baseline 6 years later", with strength close to a plateau by about one to two years. This course read the abstract only. Apply the question: older and more female than the trials, far longer, measured on whatever the company recorded, which the abstract as read doesn't describe, with no control group. It is a picture of what happened to those people, not a trial of what caused it.

Not everyone is the average

Every finding so far is an average, and people vary around it more than you might guess. The best known demonstration is Hubal and colleagues' 2005 study of 585 people, 342 women and 243 men, who trained the elbow flexors of one arm (the muscles that bend the elbow, mainly the biceps) for twelve weeks, read here at abstract level.¹⁰ Muscle size changes ranged from −2 to +59 per cent, and "1RM strength gains ranged from 0 to +250% (0 to +10.2 kg)".¹⁰ Same programme, same twelve weeks, and some people barely changed while others more than doubled.

Two studies, one word

That result gave rise to the idea of the non-responder, somebody whose body simply does not answer training, and the idea is contested. Lesson 3 met the same argument over endurance. Take two studies and read them with the course's question.

The first. A 2007 study by Bamman and colleagues, read here at abstract level, grouped 66 previously untrained people after sixteen weeks into "extreme responders (Xtr, n=17), modest responders (Mod, n=32), and nonresponders (Non, n=17) based on mean myofiber hypertrophy", which is fibre growth.¹⁰ As its title says, it then used those groups to test "the importance of myogenic gene expression during myofiber hypertrophy", that is, whether the three groups differed in the genes that build muscle. Whom: untrained people. How long: sixteen weeks. Measured on: one outcome, fibre size.

The second. A 2015 study of adults over 65, also read at abstract level, trained them for twelve and twenty-four weeks and concluded: "Nonresponsiveness was not apparent in any subject, as a positive adaptive response on at least one training outcome was apparent in every subject."¹⁰ Its authors added that responses grew with the length of training. Whom: older adults. How long: up to six months. Measured on: several outcomes, with a response on any one of them counting.

What this means for you: your result may land far from the average in either direction, and nothing in this course can predict it. One slow measure over a few weeks doesn't make you a non-responder.

The same care applies to one popular belief, that women gain strength poorly. A 2020 meta-analysis by Roberts, Nuckols and Krieger, read at abstract level, found no significant difference between men and women in muscle growth, and a larger effect for women in what it calls "relative upper-body strength".¹¹ Nuckols, a co-author, is the practitioner from the growth dispute above. The authors hedge it themselves: "it is possible that untrained females display a higher capacity to increase upper-body strength than males", and the difference might be "an artifact of the short duration of studies included".¹¹ Those results are effect sizes, the standardised scale from the specificity section, not per cents. For per cents, Hubal's study, above, is the one this course read: "Despite greater absolute gains in men, relative increases in strength measures were greater in women versus men", and its authors sum up that men "had only a slight advantage in relative size gains".¹⁰ So in kilograms men gained more, and as a share of where they started, women gained more strength and nearly as much size. Lesson 9 takes women and older adults properly.

What people get wrong

"My early gains mean my muscles grew." The first few weeks are mostly the nervous system learning to use the muscle, and some of any gain on a practised lift is skill at that lift.

"A bigger lift means I'm that much stronger at everything." Gains on the practised lift ran at more than twice the effect size of gains on an unpractised test. Real strength came on the unpractised test too, just less of it.

"The research shows" means it applies to me. Usually it means eight to twelve weeks in untrained people in their twenties, tested on the lift they practised.

"One slow result means I'm a non-responder." Lesson 3 found the same argument for endurance. Responses vary enormously, but the label depends on what was measured and for how long, and in the longer study read here everyone improved on something.

"Women respond poorly to training." In short studies women's gains had effect sizes as large as men's or larger, and in the one large trial read, women gained more strength as a share of where they started and only slightly less size. In kilograms, men usually gain more.

Practice

Connections

Back. Lesson 1 said the steepest gain is from none to some, and the resistance training evidence says the same thing in its own terms. Lesson 3 met non-response in endurance training, where dose and measurement explained much of it; here the outcome and the duration do similar work. From earlier courses, Memory's sample question is inside this lesson's "whom", Focus and Deep Work's instrument question is inside its "measured on what?", and Mental Fitness's comparison group is why every programme looks good against doing nothing and hard to tell apart from another programme.

Forward. Lesson 5 asks the question of the arguments people have about load, sets, effort and frequency. Specificity is the reason lesson 6 treats technique as a skill: you get better at the movement you practise, in the way you practise it. Lesson 10 joins this lesson's question to the institute's claim sort.

Go deeper

• Škarabot and colleagues 2021, "The knowns and unknowns of neural adaptations to resistance training", European Journal of Applied Physiology. Read here in full. Worth it for the "unknowns" half: a review that is exact about what it can't say.
• Currier and colleagues 2023, the network meta-analysis, British Journal of Sports Medicine. Read here in full, apart from the supplements. The discussion's argument about participation against optimisation is the clearest statement of the view the 2026 stand adopted.
• Both sides of the growth dispute. Loenneke and colleagues 2019, "Is muscle growth a mechanism for increasing strength?", Medical Hypotheses, read here at abstract level only; and Taber and colleagues 2019, "Exercise-induced myofibrillar hypertrophy is a contributory cause of gains in muscle strength", Sports Medicine, read here in full. The Loenneke group's own paper in the Sports Medicine exchange wasn't opened here, so the Taber paper is the fuller statement only because it's the one this course could read.

Sources

1. B. S. Currier, A. C. D'Souza, M. A. F. Singh and others, S. M. Phillips (last author), "Resistance training prescription for muscle function, hypertrophy, and physical performance in healthy adults: an overview of reviews", American College of Sports Medicine position stand, Medicine & Science in Sports & Exercise 58(4), 2026, pp. 851 to 872, doi 10.1249/MSS.0000000000003897, PMC12965823. Read level: full text of the article (Europe PMC XML: introduction, methods, all results sections, Table 1 to Table 6 as rendered, discussion, limitations, disclosures). Supplemental appendices (per-review data, AMSTAR scores, evidence grades) not opened. Press release read in full. Supports: the design and numbers, "Few RT prescription (RTx) variables affected primary adaptations", the forms of training in its summary table, the carryover sentence, the strength definition, the population quotation in the table, and the authors' sentences on training experience and age. Two of its authors also led source 2, as the body says.
2. B. S. Currier, J. C. Mcleod, L. Banfield and others, "Resistance training prescription for muscle strength and hypertrophy in healthy adults: a systematic review and Bayesian network meta-analysis", British Journal of Sports Medicine 57(18), 2023, pp. 1211 to 1220, doi 10.1136/bjsports-2023-106807. Read level: full text (Europe PMC), including results, discussion and limitations; supplements not opened. Supports: the network size, "All RTxs were superior to CTRL", the 101 of 111 comparisons, the authors' hedge on equivalence, and the table row.
3. The muscle-structure image: US National Cancer Institute, SEER Training Modules, "Structure of a Skeletal Muscle", via Wikimedia Commons, public domain. The alt text describes the drawing's own labels (bone, tendon, fascicle, muscle fiber). The caption's note on how growth is measured comes from Bamman's fibre sizes and Hubal's muscle cross-section, source 10.
4. Neural adaptation: T. Moritani and H. A. deVries, "Neural factors versus hypertrophy in the time course of muscle strength gain", American Journal of Physical Medicine 58(3), 1979, pp. 115 to 130, PMID 453338, abstract only; D. G. Sale, "Neural adaptation to resistance training", Medicine & Science in Sports & Exercise 20(5 Suppl), 1988, pp. S135 to 145, doi 10.1249/00005768-198810001-00009, abstract only; J. Škarabot, C. G. Brownstein, A. Casolo, A. Del Vecchio and P. Ansdell, "The knowns and unknowns of neural adaptations to resistance training", European Journal of Applied Physiology 121(3), 2021, pp. 675 to 685, doi 10.1007/s00421-020-04567-3, full text. Supports: the Moritani sample, finding and "cross education"; Sale's sentence; Škarabot's timing, lines of evidence, motor-unit finding, unknown site, its caution about simplified models, and its citing of Moritani. How Moritani estimated growth is not known to this course, and the body says so.
5. The growth-and-strength dispute: J. P. Loenneke, S. J. Dankel, Z. W. Bell and others, "Is muscle growth a mechanism for increasing strength?", Medical Hypotheses 125, 2019, pp. 51 to 56, doi 10.1016/j.mehy.2019.02.030, abstract only; C. B. Taber, A. Vigotsky, G. Nuckols and C. T. Haun, "Exercise-induced myofibrillar hypertrophy is a contributory cause of gains in muscle strength", Sports Medicine 49(7), 2019, pp. 993 to 997, doi 10.1007/s40279-019-01107-8, full text. The Loenneke group's Sports Medicine paper in that exchange, which the Taber paper replies to, was not opened. Supports: both sides' quotations, including the Loenneke group's reason and the Taber group's concession. Loenneke's place as last author of the Spitz meta-analysis is from its citation at source 6.
6. R. W. Spitz, R. Kataoka, S. J. Dankel and others, J. P. Loenneke, "Quantifying the generality of strength adaptation: a meta-analysis", Sports Medicine 53(3), 2023, pp. 637 to 648, doi 10.1007/s40279-022-01790-0. Read level: abstract only (PubMed). Supports: the design, including that the unpractised test had to use a trained muscle, the two effect sizes and intervals, the conclusion and the chart.
7. M. E. Haugen, F. T. Vårvik, S. Larsen and others, "Effect of free-weight vs. machine-based strength training on maximal strength, hypertrophy and jump performance: a systematic review and meta-analysis", BMC Sports Science, Medicine and Rehabilitation 15(1), 2023, 103, doi 10.1186/s13102-023-00713-4. Read level: full text. Supports: the modality-specific findings, the conclusion, and the table row.
8. The course's own constructions, each labelled where it appears in the body. The eight-week bench press trial is invented, and the body says so where it begins. The motor-unit definition is standard background this course supplies, as the body says. The reading of the slowdown as neural learning running out while growth continues slowly is this course's inference, said as such in the body, which also says it assumes one side of the growth dispute.
9. Who these studies trained: assembled by this course from the meta-analyses' own descriptions, because no paper characterising the whole literature was found. The rows come from Pelland and colleagues 2026, Sports Medicine 56(2), pp. 481 to 505, doi 10.1007/s40279-025-02344-w (published version, abstract only; preprint, full text, not peer reviewed, and the duration is the preprint's figure); Lopez and colleagues 2021, Medicine & Science in Sports & Exercise 53(6), pp. 1206 to 1216, doi 10.1249/MSS.0000000000002585 (full text); Grgic and colleagues 2022, Journal of Sport and Health Science 11(2), pp. 202 to 211, doi 10.1016/j.jshs.2021.01.007 (full text); Borde and colleagues 2015 (source 11); and sources 7, 2 and 1 above. Plus J. Steele and others, "Long-term time-course of strength adaptation to minimal dose resistance training through retrospective longitudinal growth modeling", Research Quarterly for Exercise and Sport 94(4), 2023, pp. 913 to 930, PMID 35591809, abstract only. Supports: the table, and the Steele sample, routine, results and near-plateau.
10. Individual variation: M. J. Hubal, H. Gordish-Dressman, P. D. Thompson and others, "Variability in muscle size and strength gain after unilateral resistance training", Medicine & Science in Sports & Exercise 37(6), 2005, pp. 964 to 972, PMID 15947721; J. P. Ahtiainen, S. Walker, H. Peltonen and others, "Heterogeneity in resistance training-induced muscle strength and mass responses in men and women of different ages", Age 38(1), 2016, 10, doi 10.1007/s11357-015-9870-1; T. A. Churchward-Venne, M. Tieland, L. B. Verdijk and others, "There are no nonresponders to resistance-type exercise training in older men and women", Journal of the American Medical Directors Association 16(5), 2015, pp. 400 to 411, doi 10.1016/j.jamda.2015.01.071; M. M. Bamman, J. K. Petrella, J. S. Kim and others, "Cluster analysis tests the importance of myogenic gene expression during myofiber hypertrophy in humans", Journal of Applied Physiology 102(6), 2007, pp. 2232 to 2239; J. C. Pelland and others, bioRxiv preprint, doi 10.1101/2025.07.24.666533. Read level: abstract only, all five; the Pelland preprint is not peer reviewed. Supports: Hubal's sample, ranges, absolute and relative gains by sex and its conclusion on relative size, Bamman's clusters and title, the Churchward-Venne conclusion and duration point, the preprint's design and conclusion, and the Ahtiainen figures in the exercise.
11. Sex and age: B. M. Roberts, G. Nuckols and J. W. Krieger, "Sex differences in resistance training: a systematic review and meta-analysis", Journal of Strength and Conditioning Research 34(5), 2020, pp. 1448 to 1460, doi 10.1519/JSC.0000000000003521, abstract only; R. Borde, T. Hortobágyi and U. Granacher, "Dose-response relationships of resistance training in healthy old adults", Sports Medicine 45(12), 2015, pp. 1693 to 1720, doi 10.1007/s40279-015-0385-9, full text; J. Grgic, A. Garofolini, J. Orazem and others, "Effects of resistance training on muscle size and strength in very elderly adults", Sports Medicine 50(11), 2020, pp. 1983 to 1999, abstract only. Supports: the sex findings and hedges, which the abstract reports as effect sizes, not per cents; the Borde and Grgic descriptions in the exercise.