REF-1616 · García-Hermoso et al., 2022 · Translational Pediatrics · Paper · PubMed · Free full text

Fitness carries over partly from youth into adulthood: in 21 prospective studies (6197 people at follow-up, mean 20.8 years) the correlation between youth and adult values was 0.38 for cardiorespiratory fitness, about 0.5 for muscular strength and endurance and 0.69 for sit-and-reach flexibility, slightly stronger in women and from adolescence, with very high heterogeneity. From the abstract

Main caveat: Abstract only; risk-of-bias results not seen.
moderate evidenceMeasured on non-athletes and recreational peopleRead from the abstract only

García-Hermoso and colleagues pooled prospective studies that measured fitness in healthy youth aged 6-18 and again in adulthood, using correlations as the effect size and reading below 0.30 as low, 0.30-0.60 as moderate and above 0.60 as high stability. Twenty-one studies (6197 participants at follow-up, 47.4% women, mean follow-up 20.8 years) were included.

Cardiorespiratory fitness, muscular strength and muscular endurance showed moderate tracking regardless of the test used or the length of follow-up, slightly stronger in women than men and from adolescence than from childhood; sit-and-reach flexibility showed high stability, rising with longer follow-up.

The authors note the inconsistency between studies and argue that high fitness should be targeted from childhood because low adult fitness is linked to chronic disease and death.

Participants

21 prospective studies, 6197 participants at follow-up (47.4% women), apparently healthy at 6-18 years, mean follow-up 20.8 years.

Protocol

Search: MEDLINE and Web of Science, two authors; PRISMA.
Effect size: correlation coefficients, random-effects pooling.
Risk of bias: Quality Assessment Tool for Observational Cohort and Cross-sectional Studies (results not in the abstract).
Meta-regression: length of follow-up.

Key Results

Componentr (95% CI)I²
Cardiorespiratory fitness0.38 (0.29 to 0.48)92.7%
Muscular strength0.51 (0.43 to 0.59)87.9%
Muscular endurance0.50 (0.36 to 0.86)94.5%
Flexibility (sit and reach)0.69 (0.58 to 0.81)92.9%
Moderators: moderate tracking held regardless of test and follow-up length; slightly stronger in women than in men and from adolescence than from childhood (no figures in the abstract).
Flexibility: the correlation rose with follow-up length (β = 0.017, 95% CI 0.012 to 0.021).
Interval as printed: the muscular-endurance interval is lopsided around its estimate; it is reported here as in the abstract.
Size (our arithmetic): r = 0.38 means youth aerobic fitness accounts for about 14% of the variance in adult aerobic fitness; r = 0.51 for strength, about 26%.

Practical Application

In 21 cohorts of young people followed about 21 years, those fitter as children and teenagers tended to stay somewhat fitter as adults (r = 0.38 for aerobic fitness, about 0.5 for strength), a moderate carry-over with very mixed results between studies. A fit youth goes with a partial head start in rank, not an aerobic base that lasts on its own: youth aerobic fitness accounts for about 14% of the spread in adult fitness (our arithmetic), and the rest is open to what the adult does (inference, not tested here).
Strength carried over more than aerobic fitness, and flexibility most.
A single long cohort found a weaker link by mid-adulthood (r = 0.16, REF-1617); this review reports no moderation by follow-up length for these components, so the two point the same way but differ in size.
Inference, not tested here: aerobic adaptations fade within weeks once training stops (PHYS-41), which fits a modest carry-over of aerobic fitness across decades.

Limitations

Abstract only; risk-of-bias results not seen.
Very high heterogeneity in every component.
Tests, adjustment and attrition not described in the abstract.
Correlations show stability of rank, not retained fitness.
Women and adolescence effects described as slight, without figures.
How this entry was checked and changed
Why this level

a meta-analysis of cohorts, capped by rule 1: risk of bias not read (abstract only); heterogeneity very high (I² 87.9-94.5%)