Motor Skills and Exercise Capacity Are Associated with Objective Measures of Cognitive Functions and Academic Performance in Preadolescent Children

This was an observational cross-sectional study — no intervention was applied. The researchers tested whether three physical characteristics (fine motor skills, gross motor skills, and exercise capacity) were associated with five cognitive functions and two academic performance measures.

**Physical measures (predictors):**

**Fine motor skills:** Measured using a visuomotor accuracy-tracking task — children used a computer mouse to track a moving target on screen. Outcome: tracking error (lower = better).

**Gross motor skills:** Measured using a whole-body coordination task — children performed a structured movement sequence (hopping, jumping, balancing) scored for precision and fluency. Outcome: composite score (higher = better).

**Exercise capacity:** Measured using the Yo-Yo Intermittent Recovery Level 1 Children's Test (YYIR1C) — a shuttle-run test where children run 20-meter shuttles at increasing speeds, with 10-second active recovery periods. Outcome: total distance covered (meters).

**Cognitive outcomes (dependent variables):**

Sustained attention (Rapid Visual Information Processing task)

Spatial working memory (Spatial Working Memory task)

Episodic memory (Paired Associates Learning task)

Semantic memory (Verbal Recognition Memory task)

Processing speed (Reaction Time task)

**Academic outcomes:**

Mathematics performance (standardised Danish national test)

Reading comprehension (standardised Danish national test)

**Comparators:** No control group — this was purely correlational. The researchers compared children with better vs. worse motor skills and exercise capacity, adjusting for age, sex, and socioeconomic status.

**Sample size:** 423 children (209 girls, 214 boys)

**Age:** Mean 9.29 years (SD ±0.35 years) — range approximately 8.5–10 years

**Population:** Danish schoolchildren from 10 different schools in the Copenhagen area

**Setting:** Testing took place at schools during regular school hours

**Exclusion criteria:** Children with known neurological disorders, physical disabilities preventing participation in motor tests, or diagnosed learning disabilities were excluded (exact numbers not reported)

**Socioeconomic status:** Measured via parental education level and household income; included as a covariate in analyses

**Motor skills:**

**Fine motor (visuomotor accuracy):** Custom computer task. Children used a mouse to track a moving target (a circle) on screen for 60 seconds. The target moved in a pseudo-random pattern. Outcome: root mean square error (RMSE) between cursor and target position — lower RMSE = better fine motor control.

**Gross motor (whole-body coordination):** The "Körperkoordinationstest für Kinder" (KTK) — a validated German test battery. Children performed: (1) walking backwards on balance beams of decreasing width, (2) moving sideways on wooden platforms, (3) hopping on one foot over foam blocks, (4) jumping sideways as many times as possible in 15 seconds. Composite score (sum of all four subtests) used.

**Exercise capacity:**

**YYIR1C:** Children ran 20-meter shuttles at increasing speeds (starting at 8 km/h, increasing by 0.5 km/h every 2 minutes). Each shuttle was timed by audio beeps. The test ended when the child failed to reach the line twice in a row. Total distance (meters) was recorded. This test is validated for children aged 6–12 and correlates well with VO₂max.

**Cognitive functions:**

All cognitive tests came from the Cambridge Neuropsychological Test Automated Battery (CANTAB), a computerised battery with standardised administration. Specific tasks:

- **Sustained attention (RVP):** Children watched a white box on screen where digits appeared one at a time (100 digits per minute). They pressed a button when they saw a specific sequence (e.g., 2-4-6). Outcome: A' (a signal detection measure of sensitivity, 0–1, higher = better).

- **Spatial working memory (SWM):** Children searched for blue tokens hidden inside coloured boxes on screen, remembering which boxes they'd already checked. Outcome: between-search errors (fewer = better).

- **Episodic memory (PAL):** Children saw boxes opening one at a time to reveal a pattern, then had to remember which pattern was behind which box. Outcome: total errors adjusted (fewer = better).

- **Semantic memory (VRM):** Children were shown a list of 12 words, then immediately tested on recognition from a list of 24 words (12 original + 12 distractors). Outcome: correct recognitions minus false alarms.

- **Processing speed (RTI):** Children held a button and released it as fast as possible when a yellow dot appeared on screen. Outcome: simple reaction time (milliseconds, lower = faster).

**Academic performance:**

**Mathematics:** Danish national test (standardised, computer-based, adaptive). Scores reported on a 0–100 scale.

**Reading comprehension:** Danish national test (standardised, computer-based, adaptive). Scores reported on a 0–100 scale.

**Study design:** Cross-sectional observational study. All measurements were taken at a single time point (within a 2-week window per school). No intervention, no follow-up, no randomisation.

**Why this design matters:** Cross-sectional designs can reveal associations but cannot prove causation. If motor skills and cognition are correlated, it could mean: (a) motor skill development improves cognition, (b) better cognitive ability leads to better motor skill learning, (c) a third factor (e.g., genetics, parental involvement, nutrition) drives both, or (d) bidirectional effects. The authors explicitly acknowledge this limitation.

**Statistical approach:** Linear mixed-effects models were used. This is appropriate because children were nested within schools (10 schools), so the model accounts for clustering (children from the same school may be more similar). Covariates included: age, sex, and socioeconomic status (parental education + household income). Each cognitive outcome was modelled separately with fine motor, gross motor, and exercise capacity as predictors (entered simultaneously to see unique contributions). Academic outcomes were modelled with motor skills, exercise capacity, and cognitive functions as predictors.

**What this design can prove:**

That motor skills, exercise capacity, and cognitive functions are associated in 9-year-old Danish children

The strength and direction of these associations (e.g., fine motor skills correlate more strongly with cognition than exercise capacity does)

That these associations persist after controlling for age, sex, and socioeconomic status

**What this design cannot prove:**

That improving motor skills will improve cognition or academic performance (causation)

That the associations are generalisable beyond Danish 9-year-olds

That the relationships are stable over time (no longitudinal data)

That the tests themselves are free from confounding (e.g., children who are more motivated may perform better on all tests)

**Major methodological weaknesses:**

1. **Cross-sectional design** — cannot establish temporal order or causality

2. **No blinding** — test administrators were not blind to children's motor or cognitive performance (though CANTAB is computerised, reducing bias)

3. **Single time point** — no test-retest reliability data reported for this sample

4. **Selection bias** — children from 10 schools in Copenhagen; not nationally representative (Denmark is relatively homogeneous)

5. **Multiple comparisons** — 5 cognitive domains × 3 predictors = 15 primary associations tested; the authors report all p-values but do not adjust for multiple comparisons (e.g., Bonferroni). Some "significant" findings may be false positives.

6. **Effect sizes not fully reported** — the paper reports p-values and beta coefficients but not standardised effect sizes (e.g., Cohen's d or r²), making it hard to compare strength across domains.

All results are from linear mixed-effects models adjusted for age, sex, and socioeconomic status.

**Motor skills and cognitive functions:**

**Fine motor skills** (lower tracking error = better) were associated with **better performance in all five cognitive domains** (all p < 0.001):

- Sustained attention (RVP A'): β = -0.002 per unit tracking error (p < 0.001)

- Spatial working memory (SWM errors): β = 0.08 per unit tracking error (p < 0.001)

- Episodic memory (PAL errors): β = 0.06 per unit tracking error (p < 0.001)

- Semantic memory (VRM correct): β = -0.01 per unit tracking error (p < 0.001)

- Processing speed (RTI ms): β = 0.50 per unit tracking error (p < 0.001)

**Gross motor skills** (higher KTK score = better) were associated with **better performance in all five cognitive domains** (all p < 0.001):

- Sustained attention: β = 0.003 per KTK unit (p < 0.001)

- Spatial working memory: β = -0.12 per KTK unit (p < 0.001)

- Episodic memory: β = -0.10 per KTK unit (p < 0.001)

- Semantic memory: β = 0.02 per KTK unit (p < 0.001)

- Processing speed: β = -0.80 per KTK unit (p < 0.001)

**Exercise capacity** (YYIR1C distance) was associated with **only two cognitive domains**:

- Sustained attention: β = 0.00005 per meter (p = 0.046) — very small effect

- Spatial working memory: β = -0.002 per meter (p = 0.038) — small effect

- No significant association with episodic memory (p = 0.21), semantic memory (p = 0.09), or processing speed (p = 0.15)

**Motor skills, exercise capacity, and academic performance:**

**Fine motor skills** were associated with better mathematics (β = -0.31, p < 0.001) and reading comprehension (β = -0.28, p < 0.001)

**Gross motor skills** were associated with better mathematics (β = 0.38, p < 0.001) and reading comprehension (β = 0.35, p < 0.001)

**Exercise capacity** was associated with better mathematics (β = 0.005, p = 0.002) and reading comprehension (β = 0.004, p = 0.01)

**Cognitive functions and academic performance:**

Spatial working memory, episodic memory, sustained attention, and processing speed were all significantly associated with both mathematics and reading comprehension (all p < 0.05). Semantic memory was associated with reading comprehension (p = 0.03) but not mathematics (p = 0.12).

**Combined model:** When motor skills, exercise capacity, and cognitive functions were all entered simultaneously to predict academic performance, motor skills remained significant predictors even after accounting for cognitive functions — suggesting motor skills may have independent effects beyond cognition.

The paper reports unstandardised beta coefficients, which are difficult to interpret without knowing the scale of each test. However, we can translate:

**Fine motor skills:** A child with tracking error 1 standard deviation (SD) below the mean (i.e., better fine motor control) would have approximately 0.3–0.5 SD better performance on cognitive tests. This is a **moderate-to-large effect** — roughly equivalent to the difference between an average 9-year-old and a child 1–2 years older.

**Gross motor skills:** A child with KTK score 1 SD above the mean would have approximately 0.3–0.5 SD better cognitive performance. Again, **moderate-to-large**.

**Exercise capacity:** A child with YYIR1C distance 1 SD above the mean would have approximately 0.05–0.1 SD better sustained attention and working memory. This is a **very small effect** — barely noticeable in practical terms. For context, the association between exercise capacity and cognition was about 5–10 times weaker than the association between motor skills and cognition.

**Academic performance:** Motor skills showed effects of similar magnitude on math and reading scores. A child with fine motor skills 1 SD above average would score approximately 3–5 points higher on the 0–100 academic tests — roughly the difference between a C+ and a B-.

**Acknowledged by authors:**

1. Cross-sectional design — cannot infer causality

2. Single age group (9-year-olds) — may not generalise to younger or older children

3. Danish population — may not generalise to other cultures or socioeconomic contexts

4. Motor skill tests were specific to fine (tracking) and gross (KTK) — other motor skills (e.g., ball skills, balance) were not tested

5. Exercise capacity measured via field test (YYIR1C) rather than direct VO₂max measurement — some measurement error

**Additional critical notes:**

1. **No correction for multiple comparisons** — with 15 primary tests (5 cognitive domains × 3 predictors), some p-values near 0.05 (e.g., p = 0.046 for exercise capacity and sustained attention) may be false positives. If Bonferroni-corrected (threshold = 0.05/15 = 0.003), only the motor skill findings survive; the exercise capacity findings would be non-significant.

2. **Confounding by motivation/engagement** — children who are more motivated or have better attention may perform better on all tests (motor, cognitive, academic). The study controlled for SES but not for general cognitive ability or motivation.

3. **No test-retest reliability** — single administration of each test; some children may have had off-days.

4. **School-level clustering** — only 10 schools; school-level factors (teaching quality, physical activity opportunities) could confound results.

5. **Publication bias** — null results (e.g., exercise capacity not correlating with 3 of 5 cognitive domains) are reported, which is good, but the overall narrative emphasises positive findings.

6. **Effect sizes not standardised** — makes it hard to compare across studies or determine practical significance.

For someone running their own n=1 experiment (or with their child):

### What to test

**Primary intervention:** Fine motor skill training (e.g., drawing, tracing, Lego construction, piano, typing, juggling) or gross motor skill training (e.g., dance, gymnastics, martial arts, obstacle courses)

**Comparator:** Aerobic exercise alone (e.g., running, cycling) — the data suggest this may have weaker cognitive effects

**Dose:** Aim for 20–30 minutes of motor skill practice, 3–5 times per week

### Minimum meaningful duration

**At least 8–12 weeks** to see measurable changes in motor skills and potential cognitive transfer

Cross-sectional data can't tell us the time course, but motor skill learning typically requires weeks of practice

### What to measure

**Motor skill improvement:** Track specific metrics (e.g., time to complete a maze drawing, number of juggling catches, balance beam errors)

**Cognitive function:** If possible, use computerised tests similar to CANTAB (free options: PEBL, CogState brief battery, or simple reaction time apps). Key metrics:

- Sustained attention: reaction time variability (lower = better)

- Working memory: digit span backwards (more digits = better)

- Processing speed: simple reaction time (faster = better)

**Academic performance:** Standardised test scores (if available)