REF-311 · Trappe, Costill, Fink & Pearson, 1995 · Journal of Applied Physiology, · Paper · PubMed

Skeletal Muscle Characteristics Among Distance Runners: A 20-Year Follow-up Study

28 men who were highly trained distance runners in 1966-74, re-biopsied ~20 years later: the 11 who had stayed highly trained (>70% type I at baseline) were unchanged on average, while the 10 fitness-trained and 7 untrained men showed a mean INCREASE in type I fiber proportion

Blimited evidenceMeasured on highly trained to elite athletesWritten before October 2026, being re-checked

Participants

n=28 men first studied between 1966 and 1974, when all were highly trained distance runners; classified by their training in the interim as highly trained (HI, n=11), fitness trained (FIT, n=10) or untrained (UT, n=7)

Protocol

Gastrocnemius biopsies at the original study and at follow-up (1993)
Fiber type proportion, fiber areas, and enzyme activities (succinate dehydrogenase, citrate synthase, phosphorylase)

Key Results

FIT and UT groups: mean INCREASE (P < 0.05) in the proportion of type I fibers over 20 years
HI group (initially > 70% type I): mean proportion unchanged — but 6 of the 11 individuals did show an increase in type I percentage
Subgroup of elite runners who had kept training for competition: approximately 25% reduction in SDH activity and decreases in type I and type II fiber areas — both NOT statistically significant (P > 0.05)
In 1993 the HI group had higher (P < 0.05) SDH and citrate synthase activities than FIT and UT; phosphorylase did not differ among groups
Authors' summary: these middle-aged men had a significantly greater proportion of type I fibers than 20 years earlier

Mechanism

Whichever way the training went over two decades — continued heavy training, fitness training or stopping — the type I share did not fall, and in the less-trained groups it rose. Inference, not tested (the abstract is truncated before any mechanism): ageing-related preferential loss or atrophy of type II fibers, rather than training, is the most common explanation for a rising type I proportion in men who trained less; the study cannot separate ageing from detraining.

Practical Application

In 28 former highly trained distance runners re-biopsied after ~20 years, the type I proportion rose on average in those who had dropped to fitness training (n=10) or stopped (n=7); those who stayed highly trained (n=11) were already > 70% type I and unchanged as a group, though 6 of 11 still rose
Continued hard training preserved oxidative enzyme capacity: in 1993 the HI group had higher SDH and citrate synthase than the FIT and UT groups; glycogenolytic capacity (phosphorylase) did not differ
Inference, not tested: a rising type I share in the men who trained LESS argues against reading "more type I" as a training achievement — in this cohort it is at least as much an ageing signature
Not from this paper; the base's own practice: for masters athletes the actionable part is the enzyme result — staying in hard training kept oxidative capacity above that of ex-runners; the fiber-type shift is not something to train for

Limitations

Self-selected groups (not randomised to training conditions); n=7-11 per group
20-year gap — many confounders (ageing itself, diet, other activity)
Only men, only runners; abstract truncated at 250 words
How this entry was checked and changed
Why this level

regraded 2026-09-27 from A-: small observational cohort, n=28, groups self-selected

Content correction 2026-09-27
"Type II fiber atrophy more pronounced in untrained group" -> not in the abstract; removed. The abstract's fiber-area finding is a non-significant decrease in type I and II areas in the elite subgroup that kept competing.
"Enzyme activities maintained in HI group, declined in UT" -> the abstract: HI had higher SDH and CS than FIT and UT in 1993; the competing elite subgroup itself showed a ~25% (non-significant) SDH reduction.
"Decades of endurance training can progressively shift fiber composition toward Type I" and "Confirms Bakken's observation" -> the increase occurred in the FIT and UT groups, not in the highly trained group; relabelled accordingly, Bakken line dropped.
Level A- -> B (small self-selected cohort); Tier "mixed→Tier 4-5 (review...)" -> baseline Tier 3-4, follow-up groups Tier 3 / 1 / 0.