Does lifelong endurance training keep fast-twitch muscle?
In 31 men over 65, lifelong lifters had young-like type II fibres, 1RM and RFD; lifelong runners resembled active peers. Cross-sectional, 8-13 per group: selection not excluded.
Tøien, Wang and colleagues (Molde University College in Norway, with the University of Southern Denmark and Norwegian hospitals) asked whether lifelong strength and lifelong endurance training leave different marks on ageing muscle.
They studied 42 men: 10 strength-trained master athletes (powerlifting and weightlifting clubs, 73 ± 4 y), 8 endurance-trained master athletes (track-and-field distance runners in events of 3,000 m or longer, 72 ± 6 y), 13 recreationally active older men (hiking and skiing, 75 ± 6 y) and 11 moderately active young men (25 ± 4 y).
Vastus lateralis biopsies were stained for fibre types and for markers of denervation and reinnervation (fibre-type grouping, atrophic fibres, nuclear clumps, NCAM-positive fibres); leg-press 1RM, leg-press rate of force development (RFD) and treadmill VO2max were measured.
The strength athletes had a type II fibre share like the young (52.0% vs 51.1%), higher than active controls (35.0%, P = 0.006) and, as a tendency, than the endurance athletes (39.3%, P = 0.063); they had less type I grouping and fewer atrophic fibres than both other older groups, and the highest 1RM and RFD.
The endurance athletes did not differ from active controls in fibre composition, grouping or atrophic fibres, had a higher RFD than controls (P = 0.049) and a 1RM that only tended to be higher (P = 0.083), and had by far the highest VO2max per kilogram among older men (47.5 ml/kg/min), though about 25% below the young (63.1, our calculation) and not different from the other older groups in L/min.
The authors conclude that strength training "appears to counteract age-related denervation processes" and should be used to preserve fast-twitch muscle, strength and RFD.
Our reading: a coherent cross-sectional pattern that fits the trial evidence that heavy lifting builds type II muscle in older adults, but it is not proof that lifting preserved these men's muscle: strength athletes may have started with more fast-twitch muscle, and the groups are small.
Protocol
Participants: 42 men; older groups recruited at > 65 y on training status (part of another study by the same group).Strength-trained master athletes (OS, n = 10, 73 ± 4 y) from local powerlifting and weightlifting clubs: mostly heavy training of 5 reps or fewer at > 80% 1RM in squat, deadlift, snatch, clean and jerk and bench press, plus sporadic swimming, cycling or walking.Endurance-trained master athletes (OE, n = 8, 72 ± 6 y) from local track-and-field clubs, events of 3,000 m or longer; half did some general strength work, under one session a week, mostly circuit training for the core. "Athlete" = training and competing over the past 2 years; master athletes had trained systematically most of their lives with some sporadic breaks.Recreationally active older controls (OC, n = 13, 75 ± 6 y) from senior societies; moderately active young references (YC, n = 11, 25 ± 4 y) from local university students; activity of OC and YC confirmed by the IPAQ questionnaire. No women: no female strength master athletes could be found.
Exclusions: neurological, musculoskeletal, cardiovascular or pulmonary disease. None reported use of growth hormone, testosterone or related anabolic substances (questionnaire).
Testing (one day): leg-press 1RM on an instrumented horizontal leg press (slow lowering, about 1-s pause at about 90 degrees knee flexion, then a maximal push); dynamic leg-press RFD on a force plate at a load at or just above body weight, 10-90% of peak concentric force, best of three; treadmill VO2max (highest 30-s average; plateau and RER ≥ 1.05 required). Thigh muscle volume from skinfolds and circumferences (an equation validated against MRI).
Biopsies: vastus lateralis, 10-14 days after testing, after 48 h without alcohol or strenuous exercise. Immunofluorescence for MHC I, IIa and IIx and laminin (fibre type, area, grouping), NCAM; H&E for nuclear clumps. Atrophic fibre = at or below 1,494 µm2 (the lowest 1st percentile of the young group). On average 867 fibres per biopsy for fibre type; analyses blinded to subject and group.
Statistics: one-way ANOVA with Fisher's LSD for normally distributed outcomes; Kruskal-Wallis with Mann-Whitney U for nuclear clumps, NCAM-positive and atrophic fibres; outliers outside 1.5 × IQR removed (several per variable); P ≤ 0.05, "borderline" reported at P ≤ 0.10. No correction across the many pairwise comparisons; no sample-size calculation reported.
Funding: Molde University College; imaging at the Danish Molecular Biomedical Imaging Center, supported by the Novo Nordisk Foundation. No conflicts of interest declared.
Key Results
| Outcome (mean ± SD) | Young (n = 11) | Old strength (n = 10) | Old endurance (n = 8) | Old control (n = 13) |
|---|---|---|---|---|
| Age (y) | 25 ± 4 | 73 ± 4 | 72 ± 6 | 75 ± 6 |
| Body mass (kg) | 76.0 ± 8.9 | 85.9 ± 12.1 | 69.0 ± 8.1 | 83.5 ± 9.6 |
| Thigh muscle volume (cm3) | 5,259 ± 950 | 5,199 ± 1,418 | 4,657 ± 860 | 4,863 ± 942 |
| VO2max (ml/kg/min) | 63.1 ± 6.6 | 33.2 ± 6.8 | 47.5 ± 8.1 | 35.2 ± 3.8 |
| VO2max (L/min) | 4.79 ± 0.65 | 2.84 ± 0.61 | 3.24 ± 0.48 | 2.94 ± 0.46 |
| Leg-press 1RM (kg) | 151.0 ± 24.4 | 170.0 ± 18.9 | in figure only | in figure only |
| Leg-press RFD (N/s) | 3,470 ± 1,394 | 3,993 ± 894 | in figure only | in figure only |
| Type II fibres (% of fibres) | 51.1 ± 14.4 | 52.0 ± 16.4 | 39.3 ± 11.9 | 35.0 ± 12.4 |
| Type IIa (% of fibres) | 41.9 ± 9.8 | 41.4 ± 9.2 | 33.5 ± 6.5 | 29.7 ± 11.0 |
| Type IIx (% of fibres) | 9.3 ± 8.9 | 10.5 ± 12.9 | 5.9 ± 5.8 | 5.3 ± 7.1 |
| Type II fibre area (µm2) | 4,644 ± 1,204 | 5,353 ± 1,233 | 4,176 ± 1,148 | 4,146 ± 1,485 |
| Grouped fibres (% of all fibres) | 4.5 ± 6.3 | 12.3 ± 9.3 | 31.4 ± 21.9 | 35.9 ± 23.4 |
| Atrophic fibres (%) | 0.1 ± 0.4 | 0.2 ± 0.7 | 1.2 ± 1.0 | 1.1 ± 1.4 |
| Nuclear clumps (per 1,000 fibres) | 0.0 ± 0.0 | 1.9 ± 3.7 | 0.0 ± 0.0 | 3.6 ± 6.3 |
| NCAM-positive (per 1,000 fibres) | 2.2 ± 2.3 | 7.3 ± 9.8 | 1.0 ± 1.4 | 7.4 ± 8.5 |
1RM: strength athletes above endurance athletes and controls (both P < 0.001); vs young only a tendency (P = 0.080). Endurance athletes vs controls: a tendency only (P = 0.083). Young above controls (P < 0.001).
RFD: strength athletes above controls (P < 0.001), vs endurance athletes a tendency (P = 0.072), not different from young (P = 0.329). Endurance athletes above controls (P = 0.049), not different from young (P = 0.334). Controls below young (P = 0.002). The Discussion summarises strength athletes as having about 30-294% higher maximal strength and RFD than the other older men.
VO2max: endurance athletes above strength athletes and controls per kilogram (both P < 0.001); young above all older groups (P < 0.001); strength athletes vs controls not different (P = 0.469); no difference in L/min between any of the older groups (P = 0.135-0.664). The endurance athletes' value per kg is about 25% below the young (our calculation).
Fibre type: type II share in strength athletes above controls (P = 0.006), vs endurance athletes a tendency (P = 0.063), equal to young (P = 0.895); type IIa above controls (P = 0.006), vs endurance athletes P = 0.088; type IIx no group effect (P = 0.480). Area-weighted type II share in strength athletes above endurance athletes (P = 0.016) and controls (P = 0.006). Mean fibre areas did not differ between groups (type I P = 0.590, type II P = 0.140).
Denervation-reinnervation markers: grouped fibres in strength athletes below controls (P = 0.004) and endurance athletes (P = 0.031); endurance athletes vs controls not different (P = 0.566).Atrophic fibres: endurance athletes (P = 0.009) and controls (P = 0.034) above strength athletes; endurance athletes vs controls not different (P = 0.606). Nuclear clumps, a marker of long-term neurogenic atrophy, were found in 3 strength athletes and 5 controls and in no endurance athlete; strength athletes above young (P = 0.044).NCAM-positive fibres: no group effect (P = 0.097).
Thigh muscle volume: Table 1 marks no group difference (no P value is given), and the strength athletes' mean (5,199 cm3) was no larger than the young men's (5,259 cm3); the Discussion's "enlarged thigh muscle volume" for the strength athletes is credited to an earlier study ("as also previously documented").
Mechanisms
Measured: lifelong strength athletes had a type II share and a fibre-grouping profile like young men, and fewer atrophic fibres than the other older men; lifelong distance runners did not differ significantly from recreationally active older men on any of the three (8 vs 13 men, so modest differences cannot be excluded).
Authors' interpretation: recurrent high-force contractions keep the large high-threshold motor units in use, preserving their innervation and the type II fibres they supply; in endurance training type II fibres are rarely recruited, so their motor neurons may degrade and the fibres are either reinnervated by slow motor units (type I grouping) or waste away. This is consistent with motor-unit data in ageing muscle (REF-596) but was not measured as motor-unit loss here.
Authors' interpretation: "the specific training modality and its pattern of muscle loading are of greater importance than training volume per se for type II fiber maintenance", since the runners' higher activity volume did not change their fibre profile compared with controls.
Measured, against the narrative: nuclear clumps, read by the authors as severe neurogenic atrophy, occurred in strength athletes and controls but in no endurance athlete; the authors open the passage with "Somewhat surprisingly" and say it "may in turn suggest superior reinnervation in the endurance-trained", with too few affected fibres (8 of 42 men) to conclude.
Inference: fibre areas did not differ between groups, so the strength athletes' higher area-weighted type II share comes mainly from more type II fibres rather than significantly larger ones; the Discussion's "larger type II fiber area" is a non-significant difference (P = 0.140).
Inference: a higher type II share in men who chose and stayed in power sports for decades is what selection would also produce; trials in older adults show that resistance training raises strength and muscle mass (REF-597), and the authors cite trials in which high-load training enlarged type II fibres, but none shows that it prevents fibre-type shift over 40 years.
Practical Application
Does lifelong endurance training keep fast-twitch muscle? In 31 men over 65, lifelong lifters had young-like type II fibres, 1RM and RFD; lifelong runners resembled active peers. Cross-sectional, 8-13 per group: selection not excluded.
For a masters endurance athlete (the base's practice, B): running alone, even for life, did not go with preserved fast-twitch muscle here, while heavy lifting did; add heavy, low-rep strength work alongside the endurance training. The trial evidence that heavy lifting works at 50+ is in REF-1086 and REF-597; this study adds the long-term cross-sectional picture.
Endurance still helps the aerobic side, within limits: lifelong runners had 47.5 ml/kg/min vs about 33-35 in the other older men, but about 25% below young men and not significantly higher in L/min (3.24 vs 2.84 and 2.94; P = 0.135-0.664, 8-13 per group); much of the per-kg gap is lower body mass. Training slows the decline but does not stop it (REF-595).
Endurance athletes' force was in between: their RFD was above controls (P = 0.049) and their 1RM only tended to be (P = 0.083); do not read the study as "running keeps you strong".
TRANSFER for cyclists (inference): only distance runners were studied, and the authors warn the endurance findings may not extend to cycling, rowing or skiing.
Limitations
Cross-sectional: genetics, self-selection into power or endurance sports, nutrition and other exposures cannot be separated from training; the authors state this.
Small groups (8-13 per group); several key contrasts only tendencies (strength vs endurance athletes for type II share P = 0.063, for RFD P = 0.072; endurance athletes vs controls for 1RM P = 0.083); many pairwise tests without correction; outliers removed per variable.
Men only, from local clubs; competition level not reported; only distance runners represent endurance training.
Thigh muscle volume estimated from skinfolds and circumferences, not imaged; one biopsy site.
The endurance and strength athletes' training histories ("most of their lives", "some sporadic breaks") are self-reported and not quantified; the young group was moderately active students, not athletes.
The Discussion goes beyond Table 1 in places (the strength athletes' "enlarged thigh muscle volume" is not shown in Table 1, and their "larger type II fiber area" was not significant here, P = 0.140; the endurance athletes' "higher levels of maximal lower limb muscle strength" than controls was P = 0.083).
1RM and RFD for the endurance and control groups are given only in a figure, not in the text.
Funding: Molde University College; no conflicts of interest declared.
How this entry was checked and changed
Why this level
cross-sectional, n = 42 (10, 8, 13 and 11 per group), read from the full text (open access, PMC). A lifelong-athlete vs control comparison cannot separate training from selection, genetics, nutrition or other exposures, which the authors state; the group sizes are small and several key contrasts were only tendencies
Full text (checked 2026-10-22)
A total of 42 men; OS, 73 ± 4 yr, n = 10; OE, 72 ± 6 yr, n = 8, endurance events of long duration (≥3,000 m); OC, 75 ± 6 yr, n = 13; YC, 25 ± 4 yr, n = 11 (full text).
Strength athletes typically 5 reps or less at >80% of one repetition maximum; half of the endurance athletes did general strength exercises <1 session/wk (full text).
Table 1 VO2max 63.1 ± 6.6, 33.2 ± 6.8, 47.5 ± 8.1, 35.2 ± 3.8 mL·kg−1·min−1; 4.79 ± 0.65, 2.84 ± 0.61, 3.24 ± 0.48, 2.94 ± 0.46 L·min−1; body mass 85.9 ± 12.1 and 69.0 ± 8.1 kg in OS and OE (full text).
No difference in absolute VO2max between any of the older groups (P = 0.135–0.664); no difference between OS and OC (P = 0.469) (full text).
1RM OS: 170.0 ± 18.9 kg, YC: 151.0 ± 24.4 kg; RFD OS: 3,993 ± 894, YC: 3,470 ± 1,394 N·s−1 (full text).
OS 1RM above OC and OE (both P < 0.001), tending above YC (P = 0.080); OE tended above OC (P = 0.083); OE RFD above OC (P = 0.049); OS vs OE RFD P = 0.072 (full text).
Type II fibres OS 52.0 ± 16.4%, YC 51.1 ± 14.4%, OE 39.3 ± 11.9%, OC 35.0 ± 12.4%; OS vs OC P = 0.006, vs OE P = 0.063, vs YC P = 0.895 (full text).
No main effect for mean cross-sectional area of type I fibers (P = 0.590) or type II fibers (P = 0.140) (full text).
Grouped fibres OS 12.3 ± 9.3%, OC 35.9 ± 23.4% (P = 0.004), OE 31.4 ± 21.9% (P = 0.031); YC 4.5 ± 6.3% (full text).
Atrophic fibres OS 0.2 ± 0.7%, YC 0.1 ± 0.4%, OE 1.2 ± 1.0%, OC 1.1 ± 1.4%; OC (P = 0.034) and OE (P = 0.009) above OS (full text).
Nuclear clumps in eight subjects, three in OS and five in OC; OS above YC (P = 0.044); NCAM-positive no main effect (P = 0.097) (full text).
Results "may not extend to other forms of endurance training, such as cycling, rowing, and skiing"; funding from Molde University College; no conflicts declared (full text).