Reviewing the association between motor competence and physical activity from a behavioral genetic perspective

Children 0-18; 106 systematic reviews/meta-analyses summarizing 1,344 primary studies (426 of them on the MC-PA pathway), plus a separate systematic search of twin studies

Systematic review of 106 meta-analyses/systematic reviews covering the five bidirectional pathways of the Stodden model in 0-18-year-olds, re-read through twin/family heritability evidence. Argues the entire MC-PA literature omits genetic confounding. Design mix of the reviewed reviews: 25% cross-sectional, 12% longitudinal, 32% intervention, 31% (cluster) RCT. The twin search excluded studies with fewer than 50 complete twin pairs or mean age above 18, and turned up only FIVE twin studies of motor competence and NONE of perceived motor competence, against very many for fitness and physical activity. All heritability figures below are sample-size-weighted averages reported separately by sex, not ranges. Note the authors' own twin work (Zi et al. 2023b, 2024) is two of the five MC twin studies.

The direct hereditarian read on this cluster: motor competence (h2 55% boys / 58% girls), fitness (65% / 67%) and physical activity (37% boys / 29% girls) are all heritable, share genetic variance, and "essentially none of the past systematic reviews or meta-analyses of the cross-sectional or longitudinal observational studies on the Stodden model had considered, let alone ruled out, genetic confounding". The observed MC->PA link is therefore plausibly familial confounding (pleiotropy + rGE: abler children move more and get selected into sport) rather than a trainable cascade, and the evidence gradient fits: cross-sectional associations are strong (unanimously strong in all 11 middle/late-childhood analyses, r ~ 0.25), longitudinal prediction is weak (strong in only 3 of 7 analyses), and the two reviews of MC interventions find indeterminate evidence at best that raising MC raises PA. The authors conclude "reverse causality with familial confounding are the main sources of the observed association". THE DIRECTION THAT DOES WORK, and which this record previously omitted entirely: physical-activity interventions raise motor competence, with strong evidence in 12 of 16 analyses in early childhood (13 of 15 reviews) and 13 of 17 analyses in middle childhood, mostly from RCTs, with larger effects when the PA includes deliberate motor-skill training. Note also PA in childhood has a LARGE shared-environment component (c2 33% boys / 49% girls) — family/school environment moves participation even if it does not move capacity — whereas fitness has essentially none (c2 3% / 2%), so any confounding of fitness-dependent paths must be purely genetic.

This is the source's entire thesis: MC, PA and fitness share genes; observed MC-PA correlations are uncontrolled for genetic and shared-family confounding. The authors name this "a huge knowledge gap in the fields of motor development and youth public health" and note the needed genetically informative study has not been done. All heritability rows here are DESCRIPTIVE trait estimates, not intervention effects - hence baseline `none` and horizon `not-applicable` throughout.

Narrative/methodological review, not a new genetically informative test; the needed within-twin-pair MC->PA study has not been done. Independent of the Barnett/Hill consortium, which is what makes its corroboration of their MC->PA reading worth something.