Author (year, country)	Number of participants and participant characteristics	Motor measures and ages at measurement	Factor measures and ages at measurement		Confounders (*in the final model)	Statistical analysis and results
Child factors
Yaari (2018, Israel)	n = 149 Groups: FT n = 39 M GA = 39.8 weeks (SD = 1.0) (range = 37.7–41.2) M BW = 3,373 g (SD = 346) MPT n = 57 M GA = 33.2 weeks (SD = 0.6) (range = 32.1–34) M BW = 1865 g (SD = 320) VPT n = 34 (NI) EPT n = 19 (NI)	Motor: MSEL Age:1, 4, 8, 12 and 18 months	Factor: GA Measurement: medical status FT (GA 37–41 weeks, BW > 2,500 g) MPT (GA 33–34 weeks, BW < 2,500 g) Age: at birth		Sex*	Regression with pairwise comparisons show that average level of gross motor outcome across time (from 1 to 18 months) is lower for MPT than for FT infants (b* = −2.19, SD = 1.09, p = 0.045). Pairwise comparisons between MPT and FT on differences between gross motor outcomes between 18 months and 1 month, shows that MPT are more delayed in GMD than FT infants (b* = − 6.60, SD = 2.23, p = 0.0036).
Espel et al. (2014, USA)	n = 232 Groups: Early term: 27% Full term: 56% Late term: 6.6% M GA = 39.46 weeks (SD 1.06) (range = 37 0/7–41 6/7) M BW = 3,418 g (SD = 420)	Motor: BSID‐II Age: 3, 6 12 months	Factor: GA Measurement: ultrasound <20 weeks of gestation Early FT (37–38 weeks) FT (39–40 weeks) Late FT (41–42 weeks) Age: at birth		BW GA Sex Birth order Ethnicity*	ANCOVA reveals group differences in psychomotor development at each assessment age. At 3 months, PDI is lower for early FT infants than for late FT (F(2, 179) 54.01, p < 0.05). Early FT infants exhibit lower psychomotor development scores than FT and late FT infants at 6 months (F(2, 168) 56.69, p < 0.01) and 12 months (F(2,155) 55.32, p < 0.01). FT infants had lower psychomotor development scores than late FT infants at 12 months.
Field et al. (1978, USA)	n = 151 Groups: Post term n = 46 M GA = 42 weeks M BW = 3,600 g FT n = 59 M GA = 40 weeks M BW: 3,300 g Post‐term RDS n = 46 (NI)	Motor: DDST, BSID Age: 4, 8 and 12 months	Factor: GA Measurement: N/A Age: at birth		No confounders considered	MANOVA showed that, at 4 months, post FT infants had inferior ratings on the DDST in comparison with the normal FT infants (p < 0.001). At 8 and 12 months, there were no significant differences between post FT and FT infants on the PDI scores.
Datar & Jacknowitz (2009, USA)	n = 7,425 Groups: Singletons n = 6,750 Twin pairs n = 625 Twins and other higher births whose siblings not included n = 50 M GA = 38.3 weeks M BW not reported	Motor: BSID‐II SF Age: 9 and 24 months	Factor: BW VLBW < 1,500 g MLBW 1,500–2,499 g NBW ≥ 2,500 g Measurement: weight Age: at birth		BW GA* Sex Birth order* Height* Ethnicity* education* Income Marital status* Pregnancy/delivery risk factors*	At 9 months, multiple linear regression revealed large and significant effects of VLBW (b* = −8,764; p < 0.001) and MLBW (b* = −2,901; p < 0.001) on GMD. At 2 years, the cross‐sectional estimates of VLBW (b* = −4.123; p < 0.001) and MLBW (b* = −1.383; p < 0.001) were considerably smaller, these changes being statistically significant at α = 0.01. This suggests some catch‐up is taking place between LBW and NBW children by age of 2 years.
Grantham‐McGregor et al. (1998, Brazil)	n = 262 Groups: ABW n = 131 LBW n = 131 GA > 37 weeks M GA and M BW not reported	Motor: BSID Age: 6, 12 months	Factor: BW ABW 3,000–3,499 g LBW 1,500–2,499 g Measurement: weight Age: at birth		SES	At 6 months, multiple linear regression showed that LBW‐FT infants have significantly lower scores than ABW infants on the PDI (−7.3 points, p < 0.001). This difference increased by 12 months of age (PDI −9.9 points, p < 0.001).
Lung et al. (2009a, Taiwan)	n = 20,112 Groups: FT n = GA ≥ 37 weeks BW ≥ 2,500 g PT n = GA < 37 weeks BW < 2,500 g M GA and M BW not reported	Motor: TBCS Age: 6, 18 months	Factor: twin, BW Measurement: N/A Age: at birth		BW* GA* Sex* Twin* Maternal education* Parental income*	Using structural equation modelling at 6 months, infants of parents with a higher income and infants born FT or with normal BW showed advanced GMD (b* = 0.03, p < 0.001; b* = −0.11, p < 0.001; b* = −0.10, p < 0.001). At 18 months, infants of mothers with a higher education, and of parents with higher income, who were male, twin, born FT of normal BW, had better GMD (b* = 0.03, p < 0.001; b* = 0.06, p < 0.001; b* = 0.02, p = 0.019; b* = −0.02, p = 0.026; b* = −0.02, p = 0.036; b* = −0.05, p < 0.001). (model with p value = 0.227 and AGFI = 0.999).
Nan (2013, UK)	n = 152 Twins M GA = 37 weeks (range = 26–39) M BW = 2,300 g (range = 940–3,500)	Motor: ASQ Age: 3, 6, 9, 12, 18 and 24 months	Factor: twin, BW Measurement: birth chart Age: at birth		GA* BW Sex*	Cross‐sectional multilevel linear regression analysis adjusted for sex and GA showed that twins scored lower on GMD than singletons (p < 0.001) during the first year of life. After the age of 12 months, twins catch up on GMD. BW was not a significant predictor of GMD at any age of measurement.
Brouwer et al. (2006, The Netherlands)	n = 3,490 Groups: Monozygotic twins n = 786 Dizygotic twins n = 1,645 Singletons n = 1,059 GA > 36.5 weeks BW > 2,500 g M GA and M BW not reported	Motor: MM 4 milestones: turn, sit, crawl and walk Age: 0–24 months	Factor: twin Measurement: questionnaires, blood typing Age: ≥ 3 years		GA* BW (highly correlated with GA)	ANOVA shows that no remarkable differences are seen between healthy singletons and healthy twins in the achievement of gross motor milestones within the normal range. Dizygotic twins were faster than monozygotic twins in reaching the moment for sit (p < 0.001), crawl (p = 0.013), stand (p < 0.001) and walk (p < 0.001).
Goetghebuer et al. (2003, UK)	n = 408 Groups: Twin pairs n = 168 M GA twins = 38.9 weeks (range = 38.7–39.2) M BW twins = 2,790 g (range = 2,700–2,800) Singletons n = 72 BW > 2,500 g M GA singletons = 38.3 weeks (range = 38.1–38.6) M BW = 3,240 g (range = 3,100–3,300)	Motor: MM 8 milestones adapted from DDST Age: 1, 2, 3, 4, 5, 9, 12, 18 months	Factor: twin Measurement: twin delivery Age: at birth		BW* Number of siblings* Non‐independence within twin pairs* Length*	Age of milestone achievement was higher in twins for each milestone and significant for: Maintaining head (p = 0.003), sitting without support (p = 0.03), walking (holding on) (p = 0.03). Age of milestone achievement was highly concordant within twins. The concordance was significantly higher (p < 0.05) in monozygotic than in dizygotic twins for crawling, sitting, standing holding on, and taking two steps. At 12 months, after adjustment for BW, length and sex, twin status and number of siblings were significantly associated with ‘parental report infant shows slower development than siblings’ (p = 0.05) and ‘maintaining head’ (p = 0.05).
Wilson and Harpring(1972, USA)	n = 261 M GA and M BW not reported	Motor: BSID Age: 3, 6, 9, 12 and 18 months	Factor: Twin Measurement: blood typing Age: at birth		BW*	Correlations show that twins have significantly lower scores on the motor scale at 6 and 18 months. Low GA in twins has a major effect on developmental status in the first half year of life (correlations at 3, 6, 9 and 12 months r = 0.30, r = 0.40, r = 0.20, r = 0.20, by 18 and 24 months p < 0.001).
Scharf et al. (2016, USA)	n = 950 GA: ≥37 weeks = 3% 32–37 weeks = 18% 28 < 32 weeks = 46% 22 < 28 weeks = 34% VLBW: <1,500 g Groups: Anthropometric scores < −2 SD Anthropometric scores > 2 SD	Motor: BSID SF Age: 9 and 24 months	Factor: weight, length and head circumference Measurement: weight Age: at birth, 9, 24 months		BW* Sex GA	Linear regression analysis adjusted for BW, sex and SES show that length and weight z‐scores at 9 months were correlated with (1) children's Bayley motor scores at 2 years and (2) the change in Bayley motor scores from 9 to 24 months. Children who scored more than 2 SDs below the mean in weight at 9 months showed a significant odds ratio (OR 2.64, p < 0.01) for Bayley motor scores of 2 SDs below the mean at 2 years.
Slining et al. (2010, USA)	n = 217 GA > 35 weeks M GA = 39.48 weeks (SD = 1.47) M BW = 3.23 g (SD = 0.48)	Motor: BSID‐II Age: 3, 6, 9, 12 and 18 months	Factor: weight Measurement: weight and subcutaneous fat Age: at birth, 3, 6, 9, 12 and 18 months		Age* Age squared* Sex* Weight status	Multivariate models showed that motor delay is 1.80 times more likely in overweight infants compared with nonoverweight infants (i.e., weight‐for‐length z‐score > 90th percentile) (95% CI [1.09, 2.97]) and 2.32 times as likely in infants with high subcutaneous fat compared with infants with lower subcutaneous fat (95% CI [1.26, 4.29]). High subcutaneous fat was also associated with delay in motor development (OR 2.27, 95% CI [1.08, 4.76]).
Bartlett (1998, Canada)	n = 132 BW > 2,500 g M GA and M BW not reported	Motor: AIMS and PDMS Age: 6 weeks, 3, 5, 7, 10 and 15 months	Factor: head proportion, BMI and body length Measurement: standard anthropometric measurements Age: 6 weeks, 3, 5, 7, 10 and 15 months		No confounders considered	Pearson correlations between head proportion and AIMS total, and subscale scores, revealed that infants with proportionately larger heads had significant lower scores on the AIMS total (r = −0.38, p = 0.001), this outcome being fully explained by the prone motor scores at 6 weeks of age. There was no correlation between BMI and body length and motor outcome scores.
Capute et al. (1985, USA)	n = 381 M GA and M BW not reported	Motor: MM 12 motor milestones Age: N/A	Factor: ethnicity Measurement: N/A Age: time of recruitment		Sex SES	Analysis of variances show that infants with an Afro‐American background achieve motor milestones, on average, at an earlier age, except ‘roll prone to supine’. Between 4 and 5 months of age, the milestones ‘roll supine to prone’ was reached 0.5 month earlier by infants with Afro‐American background. This advantage increases to 1.1 months for the milestone ‘walk’ (10.9 months vs. 12 months). Association of ethnicity with motor gradient without adjustment is F(16.88 p < 0.01)). After adjusting for SES and sex, the association of ethnicity still exceeds p < 0.01‐level.
Atun‐Einy et al. (2013, Israel)	n = 27 M GA and M BW not reported	Motor: AIMS (video) Age: 7–12 months/every 3 weeks	Factor: MTM Measurement: MTM scale Age: 7–12 months/every 3 weeks. Seven measurements		No confounders considered	A repeated‐measures ANOVA on the MTM score over the course of the seven observations reveals a main effect for the group: F(1, 18) = 0.25, p = 0.11. A significant interaction effect (F(6, 108) = 2.96, p < 0.01) showed an increase in motivation scores by the lower scoring group across time and a decrease in motivation scores by the higher scoring group. No significant effect of time was found. Infants with higher AIMS scores had higher motivation to move scores than infants who scored lower on the AIMS. The t test shows that strongly motivated infants had earlier onset for all motor milestones (sitting, pulling‐to‐stand, hands‐and‐knees, crawling and cruising) than weakly motivated infants (t(13) = 2.39, 2.98, 2.25, 2.50 p < 0.05). Infants' MTM score was positively correlated with the AIMS percentile at the same and subsequent sessions (Pearson correlations ranging from r = 0.36 to 0.69; with only r = 0.36 ns (p = 0.06).
Environmental factors
Majnemer and Barr (2006, Canada)	n = 155 GA > 38 weeks M GA and M BW not reported	Motor: AIMS, PDMS, Battelle Developmental Inventory Age equivalent (mon) Age: 4 or 6 and 15 months	Factor: sleep position Measurement: parental diary 3 consecutive days/24 h every 5 min Age: 4 or 6 months		Sex* Parental education* Parental age* Parity* Weight at assessment* Age at testing*	Linear regression showed there were no significant differences between sleep position on AIMS total score and PDMS score at 4 months. At 6 months of age, infants sleeping prone had significantly better motor scores on the AIMS total raw scores (p = 0.02) and PDMS (p = 0.03). At 15 months, no significant differences in PDMS score and Battelle developmental inventory age equivalent scores. Linear regression models at 4 months shows that the AIMS prone raw score (r 2 = 0.27, p = 0.0001) and the total raw score were predicted by sleep position (prone versus supine) (r 2 = 0.17, p = 0.0001), when adjusting for confounders. When adjusting for confounders on linear regression models, sleep position consistently predicted AIMS motor scores and Peabody gross motor quotient, accounting for 22% to 31% of the variance. Univariate analyses indicated that the Battelle gross motor subscale score was significantly associated (p = 0.05) with sleep position, which was further demonstrated on simple linear regression analysis (r 2 = 0.8, p = 0.048). At 15 months of age, prone sleepers attained motor milestones significantly earlier: walking upstairs (p = 0.04) and walking (p = 0.05).
Davis et al. (1998, USA)	n = 351 M BW = 3,490 g (SD = 41) M GA not reported	Motor: MM 9 motor milestones Age: 0–18 months	Factor: sleep position Measurement: position recorded by parents: prone and supine Age measurement: 2–6 months		BW* Sex* Maternal education* Ethnicity* Number of siblings*	Linear regression analysis shows that prone sleepers acquire motor milestones significantly earlier for: rolling prone to supine (p < 0.002), sitting unsupported (p = 0.003), creeping (p = 0.0002), crawling (p = 0.003) and pulling to stand (p = 0.001). Walking alone was not associated with prone sleeping (p = 0.4). Increased prone playtime was associated with tripod sitting, sitting alone, crawling and pulling to stand (p < 0.05). After controlling for maternal education, ethnicity, sex, BW and number of siblings, only pulling to stand remained significant (p < 0.01).
Lung and Shu (2011, Taiwan)	n = 1,630 Birth cohort with 7.1% infants with chronic illness included M GA and M BW not reported	Motor: TBSC Age: 6, 18 and 36 months	Factor: sleep position Measurement: interview at home Age: 6 months		Maternal education* Paternal education* Acute hospital admissions* Chronic illness*	At 6 months, structural equation model shows that infants sleeping supine had slower GMD (b* = −0.11, p < 0.001). Supine sleeping position did not affect infant development at 18 and 36 months. Other factors were associated with infant GMD at 6 months: acute hospital admission (b* = −0.07), chronic illness (b* = −0.05) and paternal education (b* = 0.06). At 18 and 36 months, maternal education (b* = 0.11 and b* = 0.07) and chronic illness (b* = −0.13 and b* = −0.05) were also associated.
Ratliff‐Schaub et al. (2001, USA)	n = 205 GA < 34 weeks (range = 29.33–29.65) BW < 1,750 g (range = 174–1,257) M GA and M BW not reported	Motor: BSID second edition Age: 4 and 13 months corrected age	Factor: Sleep position Measurement: Question on infants' usual sleeping position Age: 4 and 13 months corrected age		Maternal education* Ethnicity*, Days hospitalized*, Methyxanthine use*, Marital status* head circumference* Other maternal and infant characteristics were potential confounders, but were excluded from analysis due to p > 0.2	Multiple linear regression analyses show that the PDI scores of PT infants at 4 and 13 months corrected age did not differ significantly between prone sleepers and supine or side sleepers in both adjusted and unadjusted analyses (4 months: p = 0.7371; 13 months p = 0.1454). Individual items of the BSID show that supine sleepers were less likely than prone sleepers to receive credit for: maintaining head at 45° and 90° (p = 0.021) and lowering the head with control (p = 0.001).
Jardí et al. (2018, Spain)	n = 154 GA ≥ 37 weeks BW ≥ 2,500 g	Motor: BSID second Edition Age: 6 and 12 months	Factor: BF (exclusive BF, mixed feeding and total time BF) Measurement: 24‐h food diary and questionnaires Age: at birth, 1, 4, 6 and 12 months		BW* GA* Sex* Maternal education Maternal age* Maternal SES* Head circumference at birth, 6 and at 12 months* Height at birth, 6 and at 12 months* Iron status at 6 and 12 months* Infant haemoglobin at 6 and 12 months* BMI at 6 and 12 months*	Multiple linear regression showed in the adjusted model, that exclusive BF during the first 4 months increased the PDI by 7.712 points (p = 0.019), while mixed feeding increased it by 6.393 points (p = 0.039) at 6 months. Higher GA and higher BMI increased the PDI scores (p = 0.005 and p = 0.024 respectively) At 12 months, the adjusted model showed that exclusive BF during the first 4 months increased the PDI by 7.223 points (p = 0.033), while mixed feeding did not significantly increase the PDI (b* = 4.620; p = 0.160). Higher iron status at 6 months increased the PDI scores (p = 0.015).
Michels et al. (2017, USA)	n = 4,270 Groups: PT = 17%) FT = 83% M GA and M BW not reported	Motor: MM Age: 4, 8, 12, 18 and 24 months	Factor: BF (exclusive BF, mixed feeding) Measurement: Parent report Age: 4 months		Maternal factors Ethnicity* Education* Age* BMI* PPD* Paternal factors Education* Age* Infant characteristics Sex* Plurality* Rapid weight gain until 4 months postpartum* ASQ pass/failure at 4 months* Postpartum* Conception via fertility treatment*	Accelerated failure time models reveal that feeding differences at 4 months do not greatly affect the timing of gross motor milestone achievement. After adjustment for confounders, infants who were fed solids in addition to breastfeeding achieved standing faster than infants exclusively breastfed at 4 months (AF: 0.93; 95% CI [0.87, 0.99]). After controlling for multiple testing, these differences were no longer significant. No differences were found for PT and FT infants.
Morris et al. (1999, Brazil)	n = 262 Groups: LBW (1,500–2,499 g) n = 131 GA ≥ 37 weeks M LBW = 2,338 g (SD = 152) HBW (3,000–3,499 g) n = 131 GA ≥ 37 weeks M HBW = 3,210 g (SD = 142)	Motor: BSID Age: 6, 12 months	Factor: BF intensity in first 4 weeks or 5–26 weeks Measurement: frequency of BF Age: at birth, 6 and 12 months		BW* SES* Diarrhoea morbidity*	Weak and non‐significant correlations were observed between BF intensity in weeks 1–4 and 6 months. There was no association between BF intensity over weeks 5–26 and PDI scores at 6 and 12 months. Multiple linear regression models, adjusted for confounders, showed that BF frequency over the first 4 weeks of life was significantly associated with motor development at 6 months in both LBW and HBW infants (b* = 0.23; 95% CI [0.00–0.45]; p = 0.047).
Oddy et al. (2011, Australia)	n = 2,868 All infants eligible M GA = 38.8 weeks (SD = 2.13) M BW not reported	Motor: IMQ Age: 24, 26 and 36 months	Factor: BF duration Measurement: parental questionnaire Age: 0–12 months		GA* Sex* Maternal education* Maternal age* Maternal smoking in pregnancy* Biological father living with family* Total family income* Total amount of stressful life events during pregnancy* Apgar score infant at 5 min*	Overall, t tests show no significant differences in GMD of infants who were breastfed <4 months and >4 months. In subsequent analysis separated by sex, boys receiving BF < 4 months did have an increased risk for one atypical score on GMD at one time point between 0 and 3 years (OR 2.03; 95% CI [1.17, 3.50]; p = 0.011).
Smith‐Nielsen et al. (2016, Denmark)	n = 83 Groups: PPD‐group n = 53: M GA 40.2 weeks (SD = 1.3) M BW 3,466 g (SD = 450) Control group n = 83 M GA = 40.6 weeks (SD = 1.2) M BW = 3,583 g (SD = 526)	Motor: BSID‐III Age: 4, 13 months	Factor: maternal PPD Measurement: EPDS Age of measurement: 4, 13 months		Sex* Maternal co‐morbid personality disorder*	Multivariate analyses of variance (MANOVA) showed no significant effects of PPD on motor scales at 4 and 13 months. Also, after adjustment for confounders, the effect remained non‐significant (at 4 months p = 0.187; at 13 months p = 0.562).
Sutter‐Dallay et al. (2011, France)	n = 515 BW < 2,500 g = <1% M GA and M BW not reported	Motor: BSID‐II Age:3, 6, 12, 18 and 24 months	Factor: maternal depression Measurement: EPDS Age: 6 weeks, 3, 6, 12, 18 and 24 months		GA* Maternal education level* Maternal age* Mean income* Parity* EPDS score*	Multivariate regression models revealed no concurrent association between EPDS scores and infant motor scores over the follow up (b* = 0.60; 95% CI [−0.40, 1.60]; p = 0.24). This association remained non‐significant after adjustment for EPDS score at the time of infant assessment.
Lung, Shu, Chiang et al. (2011, Taiwan)	n = 1,693 All infants eligible M GA and M BW not reported	Motor: BSID Age: 6, 18 and 36 months	Factor: maternal mental health Measurement: Interview, SF‐36 Age: 6 months		Maternal education* Parental income* Family support*	Structural equation analysis showed that maternal mental health at 6 months was not significantly associated with GMD of infants at 6, 18 and 36 months. The study revealed the association of GMD with several other factors like family support, prenatal income and maternal education.
de Borba and Valentini (2015, Brazil)	n = 40 Groups: Infants with adolescent mothers M GA = 37.3 (SD = 2.7) M BW = 2,914 (SD = 734) Infants with adult mothers M GA = 38.7 (SD 2.4) M BW = 3,194 (SD 539)	Motor: AIMS Age: three assessments with an interval of 2 months between 0 and 18 months	Factor: maternal age Adolescent: 15–19 years Adult: 25–39 years Measurement: questionnaire Age: maternal age at infant birth		No confounders considered	Generalized estimated equations showed that AIMS percentile (F(938.2) = 0.003, p = 0.874) and total AIMS score (F(38.2) = 0.085; p = 0.755) did not differ between infants of adolescent mothers and adult mothers. Infants of adolescent mothers had lower scores in the third evaluation in supine position (p = 0.046).
Lung et al. (2009b, Taiwan)	n = 17,595 M GA and M BW not reported	Motor: TBCS Age: 6,18 months	Factor: Parental mental health Measurement: SF‐36 Age: 6 months		Parental education* Parental age*	Multiple linear regression showed that parental mental health (paternal and maternal) was not significantly associated with children's 6‐month development (paternal b* = −0.01, t = 1.04, p = 0.298; maternal b* = 0.01, t = 0.74, p = 0.458). At 18 months, only maternal mental health was predictive of infants' GMD (maternal b* = 0.017, p = 0.01). When the covariates of parental education and age of childbirth were added, the effect of maternal mental health decreased (b* = 0.02, t = 2.12, p = 0.034).
Hernández‐Martínez et al. (2011, Spain)	n = 72 M GA = 39.8 weeks (SD = 1.32) M BW = 3,277.7 g (SD = 456.23)	Motor: BSID Age: birth, 12 months	Factor: parental neonatal perceptions Measurement: NPI Age: 3 days, 3 months		GA* BW SES Father and mother neonatal perception scores* NBAS (endurance item)*	Using stepwise multiple regression models, more negative maternal neonatal perceptions (b* = −0.325, p = 0.024) and a higher GA (b* = 0.340, p = 0.018) predicted psychomotor development at 4 months and accounted for 21.8% of the variance. At 12 months, paternal neonatal perceptions (b* = 0.383, p = 0.010), together with the NBAS endurance item (b* = 0.339, p = 0.021) were significant in accounting for 17.2% variance of the psychomotor development.
Siegel and Burton, (1999, USA)	n = 109 M GA and M BW not reported	Motor: BSID, MM Age: 6 and 9 months (n = 34) 9 and 12 months (n = 35) 12 and 15 months (n = 40)	Factor: use of a baby walker Measurement: exposure baby walker from parent interview Age: 6 and 9, 9 and 12, and 12 and 15 months		Parental education	A three‐by‐three between‐subjects MANCOVA showed a significant effect of walker experience on infants' motor milestones in general (multivariate F(6,154) = 4.81 p = <0.0005). The univariate test showed that the use of a baby walker significantly affects the developmental onset of sitting, crawling and walking (F(2,79) = 11.07, 4.97 and 4.25, p = 0.0005, p = 0.01 and p = 0.02), with a later onset of the motor milestones. A significant main effect of the use of a baby walker was observed for motor and mental scores considered together (multivariate F(4,196) = 6.16 p < 0.0005). The univariate tests showed significant effects for motor development (F(2,99) = 6.06, p < 0.03). Parental education was added as a covariate in the analyses.
Souza et al. (2010, Brazil)	n = 30 Groups: FT = 86.2% PT = 13.8% M GA, M BW not reported	Motor: BSID‐III Age: 12, 17 months	Factor: daycare attendance Measurement: full time daycare attendance Age: 0–17 months		No confounders considered	Descriptive statistics showed that 13% (n = 4) of the infants attending daycare full‐time had suspected delays in GMD at 12 and 17 months, according to the reference means of the BSID. Of these four infants, one infant was PT with LBW.
Tsuchiya et al. (2012, Japan)	n = 742 GA = 39.0–39.2 weeks BW = 2,948–2,985 g Infants with pathology affecting motor function n = 5 M GA and M BW not reported	Motor: MSEL Age: 6, 10, 14 months	Factor: seasonal variation Measurement: month of birth Age: at birth		GA BW Sex SES Parental ages Parity	In a linear regression model (month of birth is transformed to a trigonometric form), the season of birth was significantly associated with GMD at 6 months (F(2,736) = 21.71, p < 0.001 and 10 months (F(2,736) = 12.36, p < 0.001. At 14 months, the season of birth was not significantly associated with gross motor score (F(2,736) = 1.21, p = 0.30). Infants born in Mar to Apr show a peak in GMD and those born in autumn (Sep to Oct) show the lowest GMD scores. The cyclic fluctuation of motor development according to month of birth disappears at 14 months of age.
Vierhaus et al. (2011, Cameroon/Germany)	n = 345 Groups: Cameroonian infants n = 73 German infants n = 272 M GA and M BW not reported	Motor: BSID III Age: 3, 6 and 9 months	Factor: cultural context Measurement: N/A Age: time of inclusion		No confounders considered	Univariate analysis of variance of the BSID outcomes, depending on cultural background (Cameroonian Nso versus Germans, between subjects factor) and cultural background by age (3, 6 and 9 months, within subjects factor), showed large differences between the two cultural backgrounds (F = 65.58; df 1/251; p < 0.001; η 2 = 0.207) in favour of the Cameroonian infants at 3 months. These differences decrease over time and are almost non‐existent at 9 months (F = 23.63; df 2/502; p = <0.001; η 2 = 0.086). The largest deviance is related to GMD at 6 months due to items as sitting and standing being reached by Cameroonian infants much earlier than German ones. The sequence of BSID items differ between the groups.
Multiple factors
Bjarnadóttir et al. (2019, Denmark)	n = 650 GA > 37 weeks BW > 2,500 g	Motor: MM 13 predefined milestones Age: N/A	Factor: BF, predictors pregnancy and birth, home environment Measurement: interviews/questionnaires Age: ongoing parental interviews from 1 week to 24 months		GA* BW Sex* SES Maternal age* Maternal education Paternity leave*	Principal components analysis was used to analyse motor milestone outcomes, grouping the milestones into ‘late’ and ‘early with late in opposite directions’. Multivariate analysis showed that sex, GA and maternal age (M = 0.32, p = 0.05. b* = −0.23, p = < 0.001 and b* = 0.05, p = 0.02, respectively) were significant predictors for the achievement of later milestones (crawling, walking and standing). Boys achieved these late milestones at an earlier age. For the early milestones, GA (b* = −0.11, p = 0.01) and paternity leave (M = −0.28, p = 0.01) were significant predictors. Linear and logistic regression analysis revealed that motor milestone achievement from 1 to 24 months was not significantly related to BF duration (exclusive or total).
Flensborg‐Madsen & Mortensen (2017, Denmark)	n = 5,601 M GA = 39.1 weeks (range = 27–46.5) M BW = 3,250 g (range = 850–5,450)	Motor: MM Age: N/A	Factor: GA, BW and other predictors Measurement: questionnaire, measurements Age: at birth and 12 months		Sex	Multiple linear regression analysis showed that most of explained variance (14.5%) in motor milestone attainment is due to GA (b* = −0.15; p < 0.001) and BW (b* = −0.16; p ≤ 0.001), after adjustment for confounders. Other predictors (p values ≤ 0.10 were considered significant) in the final model were: BF, paternal age, higher birth order, weight increase (all negative associations) and larger head (positive association).
Pereira et al. (2016, Brazil)	n = 49 Groups: Preterm n = 12 (24.5%) Term n = 37 (75.5%) M GA = 38.20 weeks (range = 32–42 weeks M BW = 3,156 g (range = 2,200–3,995 g	Motor: AIMS Age: three assessments from 2 to 12 months	Factor: home environment, maternal practices, cognition Measurement: DAIS, A‐HEMD‐IS and KIDI Age: between 2 and 12 months		GA BW Sex* Cognition* DAIS score* Family income* Mechanical ventilation*	Generalized estimating equations used for longitudinal analysis showed that the scores on motor development increased over time and strongly significant correlations were found between the motor outcomes at the three time points. Multivariate analysis revealed at assessment 1 that: family income (p = 0.011), score on cognition (p > 0.001), days of mechanical ventilation (p = 0.099) and being put in more stimulating and independent positions (p = 0.037) explained motor performance significantly (Adj R 2 = 0.876). At assessment 2, the multivariate model included cognition (p > 0.001) and family income (p = 0.003, Adj R 2 = 0.860); at the third assessment, only cognition (p > 0.001) remained in the model (Adj R 2 = 0.751). Variability in motor development is better explained by environment and parental knowledge and practice.

Abbreviations: ABW, adequate birthweight; AHEMD‐IS, Affordances of the home environment—Infant‐Scale; AIMS, Alberta Infant Motor Scale; ASQ, Ages and Stages Questionnaire; BSID, Bayley Scales of Infant Development; BW, birthweight; DAIS, Daily Activities of Infants Scale; DDST, Denver Developmental Screening Test; EPDS, Edinburgh Postnatal Depression Scale; EPT, extremely preterm; FT, full term; HBW, high birthweight; GA, gestational age; IMQ, Infant Motor Quotient (now ASQ); KIDI, Knowledge Infant Development Inventory; LBW, low birthweight; M, mean; M‐ABC, Movement‐ABC; MM, motor milestones; MLBW, medium low birthweight; MPT, moderately preterm; MSEL, Mullen Scale of Early Learning; MTM scale, motivation to move scale; N/A, not applicable; NBAS, Neonatal Behavioural Assessment Scale; NBW, normal birthweight; NI, not included in the results of this review due to pathology; NPI, Neonatal Perception Inventory; PDI, Psychomotor Developmental Index; PDMS, Peabody Developmental Motor Scales; PPD, postpartum depression; PT, preterm; RDS, respiratory distress syndrome; SES, socio‐economic status; SF‐36, 36‐item Short Form Health Survey; TBCS, Taiwanese Birth Cohort Study developmental instrument; VLBW, very low birthweight; VPT, very preterm.