The effects of environmental enrichment in infants with or at high risk of cerebral palsy: an updated systematic review and meta-analysis

Abstract

Background

This systematic review and meta-analysis aimed to evaluate the effects of environmental enrichment (EE) in infants with or at high risk of cerebral palsy (CP), as well as to identify the most effective age window for intervention.

Methods

PubMed, Embase, the Cochrane Library, the Cumulative Index to Nursing and Allied Health Literature, the Web of Science Core Collection, Psychological Information Database, and Sociological Index were searched from database inception to 27 February 2025. All data analysis was performed using Stata 17.0. Differences were expressed using standard mean difference (SMD) with 95% confidence interval (CI). Outcomes, including motor development, gross motor function, fine motor function, and cognitive development, were investigated.

Results

Fourteen randomized controlled trials with 592 participants were included. Of the 14 included articles, 50% were assessed as low risk, 36% were assessed as some concerns, and 14% were assessed as high risk. EE interventions significantly improved motor development (SMD = 0.35; 95% CI = 0.11 to 0.60; p = 0.004), gross motor function (SMD = 0.25; 95% CI = 0.06 to 0.44; p = 0.011), and cognitive development (SMD = 0.32; 95% CI = 0.10 to 0.54; p = 0.004) in infants with or at high risk of CP. No significant effect was observed on fine motor function. Subgroup analyses indicated that the optimal age window for EE is 6-18 months for motor development and 6-12 months for cognitive development. The overall quality of evidence ranged from high to low.

Conclusions

EE interventions significantly improve motor development, gross motor function, and cognitive development in infants with or at high risk of CP.

Trial registration

The protocol of this study was registered in PROSPERO (CRD42024523400).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12887-025-05954-5.

Introduction

Cerebral palsy (CP), the most common physical disability in childhood [1], is a lifelong neurodevelopmental disorder caused by non-progressive brain maldevelopment or injury occurring prenatally or during early infancy [2]. It encompasses heterogeneous neuromotor impairments, primarily characterized by movement and posture limitations, and commonly co-occurs with sensory, cognitive, or communication deficits [3]. Because it originates during a critical phase of brain development, CP occurs at a time when neural circuits exhibit a high degree of plasticity [4, 5]. Numerous studies have identified critical periods in early postnatal life—particularly between birth and two years of age—during which the brain is especially responsive to environmental inputs, offering a unique window of opportunity for targeted interventions [5–7]. Accordingly, early identification and intervention in infants with or at high risk of CP have been recognized as crucial for optimizing neurodevelopmental outcomes [8, 9]. Experimental and clinical research further supports that enriched environments and sensory-motor experiences during these periods can enhance neuroplasticity and positively influence developmental trajectories [7, 10, 11].

Previous studies often lacked a clear definition of infants at high risk for CP, leading to the misclassification of cases such as prematurity, motor delays, or brain injuries without confirmed CP. However, prematurity does not guarantee CP [12], motor delays are nonspecific [13], and brain injuries vary in outcome [14]. The 2017 International Guidelines for Diagnosis standardized the definition of high-risk CP, requiring motor dysfunction plus at least one additional criterion (e.g., abnormal neuroimaging or clinical history) [9]. To improve accuracy, future studies should adopt these criteria to ensure proper identification of high-risk CP infants.

Environmental enrichment (EE), originally developed in animal studies, has been shown to play a beneficial role in neural development. Specifically, EE interventions can effectively promote dendritic branching and synaptic density, increase cortical thickness, enhance neuronal differentiation, exploratory behavior, and learning and memory abilities, stimulate hippocampal neurogenesis, and improve the balance between hippocampal inhibition and excitation [15, 16]. Given that EE not only enhances normal neural development through neuroplasticity but also supports neural repair after central nervous system injury via morphological, cellular, and molecular adaptations, it has been increasingly recognized as a promising early intervention in both experimental and clinical neuroscience [17–19]. In infants, EE typically targets motor, sensory, cognitive, or social stimulation through play and caregiver interaction [20]. Various EE protocols—such as COPCA (coping with and caring for infants with special needs), GAME (goals - activity - motor - enrichment), and SPEEDI (supporting play exploration and early development intervention)—are tailored to different developmental goals [21–23]. Despite these variations, researchers consistently emphasize that effective EE interventions for infants should ideally integrate a stimulating, play-based environment with active social interactions involving caregivers or healthcare professionals [24–26]. In 2013, Morgan et al. investigated the effects of EE on motor development in infants diagnosed with CP or considered at high risk based on abnormal general movements or atypical neuroimaging findings. However, the included population had a wide age range, from birth to 96 months, which may introduce developmental variability and affect the interpretation of intervention effects across different age groups. A meta-analysis of data collected immediately after the intervention demonstrated that EE effectively improved motor development in both infants with and at high-risk of CP [20]. Motor development refers to the dynamic, progressive changes in an individual’s motor skills over time, with particular focus on the sequential acquisition of abilities during childhood—from basic movements such as grasping and crawling to more complex skills like running, jumping, and writing [27]. Despite this, the study by Morgan et al. conceptualized motor development as a single, unified outcome, without distinguishing between its major subcomponents [20]. Given that motor development encompasses both gross and fine motor functions [28], further investigation is warranted to examine the differential effects of EE on these specific domains. Such an approach may yield more nuanced insights into which aspects of motor function development are most responsive to EE, thereby informing the development of more targeted early intervention strategies. Moreover, the potential impact of EE on cognitive development in this population has not been well established.

As research on EE continues to evolve, there is a growing need to update our understanding of its therapeutic potential in infants with or at high risk of CP. While previous reviews have provided valuable insights, further investigation is warranted to address areas such as more specific population definitions, distinctions between different domains of motor development, and the inclusion of cognitive outcomes. This systematic review and meta-analysis aimed to address these gaps by evaluating the effectiveness of EE interventions in improving motor development—specifically gross and fine motor functions—as well as cognitive development. Moreover, by focusing on infants aged 0–2 years, a critical period for neuroplasticity, we sought to identify the optimal age window for intervention and to provide more targeted guidance for early clinical applications.

Methods

Search strategy and study selection

This systematic review and meta-analysis follow the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [29]. The PRISMA checklist is shown in Supplementary material 1. The protocol was registered in PROSPERO (CRD42024523400). We conducted a systematic literature search of published studies in PubMed, Embase, the Cochrane Library, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), the Web of Science Core Collection, Psychological Information Database (PsycINFO), and Sociological Index (SocINDEX) from their inception to 27 February 2025. Searches were conducted using keywords such as Cerebral Palsy, Randomized controlled trials, high risk of Cerebral Palsy, and early intervention. The search strategy is shown in Supplementary material 2. All retrieved literature was imported into EndNote X9 software, and the Systematic Review Accelerator Helper was used to assist with deduplication, initial screening, and classification. Two reviewers (XYZ and MDJ) independently screened titles and abstracts, then assessed full texts for eligibility. Disagreements were resolved through discussion, with a third reviewer (XL) consulted if needed.

Eligibility criteria

The inclusion criteria were as follows: (1) participants aged ≤ 24 months, diagnosed with or at high risk of CP according to the 2017 International Guidelines, requiring motor dysfunction plus at least one additional factor (e.g., abnormal neuroimaging or clinical history) [9]; (2) intervention involved EE-based therapies; control groups received conventional treatments without EE; (3) studies reported at least one predefined outcome, including motor development (e.g., assessed by Bayley Scales of Infant and Toddler Development (BSID)), gross and fine motor function (e.g., Gross Motor Function Measure (GMFM), Peabody Developmental Motor Scales (PDMS)), or cognitive development (e.g., BSID); (4) outcomes had to be clearly defined and quantitatively measured; and (5) randomized controlled trials (RCTs) published in English from database inception to 27 February 2025.

Exclusion criteria were as follows: (1) studies with diagnoses limited to prematurity, motor developmental delay, or brain injury; (2) studies where the control group received environmental enrichment interventions; (3) studies primarily focusing on caregiver outcomes; and (4) studies lacking sufficient statistical data to calculate effect sizes.

Data extraction

Two reviewers (JPL and XZ) independently extracted data from the included studies. If there was a disagreement, the article was re-evaluated by the third reviewer (XL). The third reviewer was also responsible for verifying the accuracy of data extraction. The WebPlotDigitizer (by Ankit Rohatgi, version 5.2) was used to extract data from the figures. To collect the general characteristics of enrolled studies, a table was designed and filled by us (Table 1). For the intervention provided in the studies, we extracted the EE elements used in the intervention group and the provider, sites, and duration of the intervention, as well as the details of other treatments. The mean, standard deviation, and sample size of the outcome measures in each group were extracted for meta-analysis. Authors were contacted for additional data when data were not directly provided in the article.

Table 1.

Characteristics of included studies

Author, year	Country	Sample size (female)	Diagnosis	Age, Median (range)/Mean ± SD		Intervention	Interest outcome measures	Assessments timepoint
				Baseline (mon)	Gestational age (week)
Nelson (2001) [30]	America	EG: 21(14) CG: 16(5)	High risk of CP	EG: 7.72 ± 0.19 of CA CG: 7.74 ± 0.21 of CA	EG: 26.28 ± 2.24 CG: 26.75 ± 1.35	EG: EE + Standard care CG: Standard care	Motor: BSID-II Cognitive: BSID-II	12 mon of CA
Ohgi (2004) [31]	Japan	EG: 12(3) CG: 11(3)	High risk of CP	EG: 0 CG: 0	EG: 30.3 ± 3.3 CG: 30.3 ± 2.7	EG: EE CG: Standard care	Motor: BSID-II Cognitive: BSID-II	6 mon of CA
Blauw-Hospers (2011) [32]	America	EG: 21(12) CG: 25(14)	High risk of CP	3–6 of CA	EG: 29 (27–40) CG: 30 (25–39)	EG: EE CG: Standard care	Motor: AIMS Cognitive: BSID-II	6 mon of CA, 18 mon of CA
Morgan (2015) [33]	Australia	EG: 6(1) CG: 7(1)	High risk of CP	EG: 4.10 ± 0.96 of CA CG: 4.01 ± 0.91 of CA	EG: 35.50 ± 5.21 CG: 33.57 ± 7.76	EG: EE CG: Standard care	Motor: PDMS-II	after 12-week intervention
Morgan (2016) [34]	Australia	EG: 15(7) CG: 15(6)	High risk of CP	EG: 3.62 ± 1.10 of CA CG: 4.61 ± 1.17 of CA	EG: 34.27 ± 5.27 CG: 35.27 ± 5.09	EG: EE CG: Standard care	Motor: PDMS-II, GMFM-66 Cognitive: BSID-III	12 mon of CA
Dusing (2018) [35]	America	EG: 7(2) CG: 7(4)	High risk of CP	EG: 0 CG: 0	EG: 25 (24–27) CG: 26 (25–28)	EG: EE + Standard care CG: Standard care	Motor: TIMP, BSID-III Cognitive: BSID-III, EPSI	after 15-week intervention, 12 mon of CA
Holmström (2019) [36]	Sweden	EG: 19(7) CG: 20(6)	High risk of CP, motor disfunction and preterm	EG: 6.3 ± 1.62 of CA CG: 6.7 ± 1.96 of CA	EG: 33 ± 6.5 CG: 33 ± 6.95	EG: EE CG: Standard care	Motor: PDMS-II, GMFM-66, BSID-III Cognitive: BSID-III	after 35-week intervention, 24 mon of CA
Harbourne (2019) [37]	America	EG:11(N/A) CG:9(N/A)	High risk of CP and CP	15 ± 6.9	N/A	EG: EE CG: Optimal Pattern Intervention	Motor: GMFM-88 Cognitive: EPSI	after 12-week intervention
Hielkema (2020a) [38]	Netherlands	EG: 23(8) CG: 20(9)	High risk of CP	EG: 1.4 (0.1–8.6) CG: 2.5 (0.9–9.0)	EG: 32 (26–41) CG: 29 (26–41)	EG: EE CG: PT	Motor: IMP, AIMS, Bayley-II, GMFM-66, GMFM-88 Cognitive: Bayley-II	after 48-week intervention, 21 mon of CA
Hielkema (2020b) [39]	Netherlands	EG: 23(8) CG: 20(9)	High risk of CP	EG: 1.4 (0.7–2.8) CG: 2.5 (1.8–4.7)	EG: 32 (29–40) CG: 29 (27–38)	EG: EE CG: PT	NA	after 48-week intervention, 21 mon of CA
Harbourne (2021) [40]	America	EG: 55(26) CG: 57(22)	High risk of CP	EG: 10.67 ± 2.57 of CA CG: 10.93 ± 2.63 of CA	N/A	EG: EE + Standard care CG: Standard care	Motor: GMFM-88, BSID-III Cognitive: APSP, BSID-III	after 36-week intervention,
Ziegler (2021) [41]	Switzerland	EG: 8(2) CG: 8(1)	High risk of CP	EG: 1.04 (0-3.45) CG: 3.34 (0-5.06)	EG: 27 (25–30) CG: 29.5 (26–31)	EG: EE CG: Standard care	Motor: IMP, BSID-III Cognitive: BSID-III	18 mon of CA, 24 mon of CA
Cemali (2022) [42]	Turkey	EG: 17(5) CG: 17(11)	CP	EG: 14.47 ± 1.28 CG: 13.82 ± 1.55	N/A	EG: EE + PT CG: PT	Motor: AIMS	after 8-week intervention
Alberge (2023) [43]	France	EG: 63(28) CG: 59(22)	High risk of CP	EG: 8.5 (8.28–8.97) of CA CG: 8.5 (8.28–8.97) of CA	EG: 27.6 ± 1.3 CG: 27.4 ± 1.4	EG: EE CG: Standard care	Motor: BSID-III Cognitive: BSID-III	9 mon of age, 24 mon of age

EG experiment group, CG control group, CP cerebral palsy, N/A not applicable, w weeks, mon months, CA corrected age, EE environmental enrichment, PT physiotherapy, BSID-II Bayley Scales of Infant Development-Second Edition, AIMS Alberta Infant Motor Scale, PDMS-II Peabody Developmental Motor Scales- Second Edition, GMFM-66 Gross Motor Function Measure-66, BSID-III Bayley scales of infant development-Third edition, TIMP Test of Infant Motor Performance, GMFM-88 Gross Motor Function Measure-88, EPSI Early Problem Solving Indicator, IMP Infant Motor Profile, APSP Assessment of Problem Solving in Play

Quality assessment

Two reviewers (MDJ and JYZ) used the Cochrane Risk of Bias Tool 2.0 (RoB 2.0) to assess the methodological quality of RCTs. In the RoB 2.0 tool, the risk of bias for each outcome is categorized as low risk of bias, some concerns, or high risk of bias. A rating of “low risk” indicates that the study demonstrates sound methodological practices in the assessed domain, with no identifiable sources of bias, and its results are considered trustworthy. “Some concerns” suggests that there is limited information or some uncertainty about the study’s methods in a specific domain, raising the possibility of bias that is not severe enough to be classified as high risk. A judgment of “high risk of bias” reflects substantial methodological flaws in the study, such as inadequate randomization, improper handling of missing data, or selective reporting, which could significantly compromise the credibility of the outcome [44].

The quality of evidence was assessed by two reviewers (MDJ and JYZ) using the GRADE (Grading of Recommendations Assessment, Development and Evaluation). The quality of the evidence was classified by GRADE as high, moderate, low, and very low [45]. High quality indicates a high level of certainty, meaning the results accurately reflect the true effect and are likely to be reproducible in other studies. Moderate quality suggests a high degree of certainty but with some limitations or uncertainty that may affect the interpretation of the results. Low quality implies that there is considerable uncertainty, and the results may be influenced by bias or other factors, making them less reliable. Very low quality indicates extremely poor evidence, with high uncertainty and a lack of reliability in the conclusions [46].

Statistical analysis

All data analysis was performed using Stata 17.0 (StataCorp, College Station, TX, USA). Standard mean difference (SMD) with 95% confidence interval (CI) was used to estimate continuous variables. The I-squared (I²) test was used to estimate heterogeneity, and the values of 25%, 25–50%, 50–75%, and > 75% indicated no heterogeneity, low heterogeneity, moderate heterogeneity, and high heterogeneity, respectively. If the I² value was less than 50%, the fixed effects model was used; otherwise, a random-effects model was used. Subgroup analyses for motor and cognitive development outcomes were conducted based on the infants’ age at assessment. The Egger test evaluated publication bias. The leave-one-out sensitivity analysis was used to asssess the robustness of the meta-analysis results. The statistically significant p value was set at 0.05.

Results

Search results

A total of 12,242 records were identified from the initial database search. After removing 390 duplicate records and excluding 5030 studies using automation tools, 6822 potentially relevant references were identified for further screening. A total of 6784 articles were excluded during the initial screening process, including 6702 articles based on titles and abstracts that did not meet the inclusion criteria, 62 non-RCTs, and 20 non-human studies. Finally, we identified 38 studies that were subjected to full-text review. Fourteen RCTs met the eligibility criteria and were included [30–43], the PRISMA flow diagram is shown in Fig. 1.

Fig. 1.

The flow diagram. CINAHL: the Cochran Library, the Cumulative Index to Nursing and Allied Health; WOS: the Web of Science; PsycINFO: Psychological Information; SocINDEX: Sociological Index; RCTs: Randomized controlled trials; EE: environmental enrichment; CP: cerebral palsy

Study characteristics

Five hundred ninety-two participants were included for systematic review and meta-analysis. Most of the included studies recruited infants identified as being at high-risk of CP. However, two studies included participants who had already been diagnosed with CP [37, 42]. Since CP is typically diagnosed around the age of two, the age of participants at the time of enrollment (i.e., baseline) in these two studies was notably higher than in the others, contributing to considerable variation in baseline ages across studies. The youngest participants began receiving EE interventions immediately after birth [21, 31], while the oldest began around 15 months of age in the study by Cemali et al. [42]. The characteristics of the RCTs are summarized in Table 1.

Due to the multifaceted nature of EE as a stimulus-based intervention, different studies have employed diverse EE protocols. As a result, many have assigned specific program names to their interventions, such as “COPCA” [32, 38, 39, 41], “GAME” [23, 47], “SPEEDI” [21], “START-Play”(Sitting Together and Reaching to Play) [22], and Small Step [48]. Most studies incorporated enrichment strategies targeting motor, cognitive, and social development. Only two studies included sensory enrichment in their programs [30, 42]. Motor enrichment was primarily implemented through educating parents to recognize their infant’s abilities and developmental needs, promoting self-initiated motor behaviors, practicing motor tasks, and setting up motor-enriched play environments. Cognitive enrichment similarly included parental education on infant’s abilities and developmental needs, as well as task-oriented practice to facilitate experiential learning and the promotion of exploratory behaviors. Social enrichment was mainly delivered through educating parents on parent-infant interaction strategies. The EE in four studies was provided by caregivers [31, 35, 36, 41], and six studies by healthcare workers [30, 37–39, 42, 43], and four studies by both caregivers and healthcare workers [32–34, 40]. The sites where EE was conducted can be categorized as hospital-based and home-based. Three of the included studies were hospital-based (e.g., special care nursery, neonatal intensive care unit, and therapy room) [30, 31, 42], and eleven were home-based [32–41, 43]. More details about the intervention are summarized in Table 2.

Table 2.

Environmental enrichment intervention and other treatments in the included trials

Author, year	Details of EE intervention			Other treatments	Duration
	Program and Enrichment used	Providers	Sites
Nelson (2001) [30]	Program: multisensory intervention Sensory enrichment: auditory stimuli via a female human voice, tactile stimuli administered as light stroking, visual stimuli involving eye-to-eye contact, and vestibular stimulation consisting of rhythmic rocking. Social enrichment: educating parents on parent-infant interaction was conducted.	Healthcare workers	Hospital (special care nursery)	Standard care: a special care nursery environment was designed to reduce environmental stress (reduce sound and light) and promote sleep cycle and motor development (stress reduction program)	until 2 mon of CA
Ohgi (2004) [31]	Program: NBAS-based intervention combined with developmental support Motor and cognitive enrichment: based on the infant’s capabilities and developmental needs, parents were taught how to care for the baby. Social enrichment: optimize caregiving interactions by enhancing the mothers’ adjustment to their infants’ behavior.	Caregivers	Hospital (NICU)	Standard care: general guidance was imparted to parents, and upon discharge, developmental services were provided to all infants and parents as per their individual needs.	until 6 mon of CA
Blauw-Hospers (2011) [32]	Program: COPCA Motor and cognitive enrichment: caregivers identified infants’ signals and responded appropriately to their actual needs while promoting self-generated motor behaviors, variations, and trial-and-error experiences through play Social enrichment: educating parents on parent-infant interaction was conducted	Caregivers and Healthcare workers	Home	Standard care: traditional infant physical therapy based on neurodevelopmental treatment was employed	3–6 mon of CA
Morgan (2015) [33]	Program: GAME Motor and cognitive enrichment: motor task practice and set up motor enriched play environments; instructed parents to recognize and understand infants’ motor behaviors, and subsequently provided them with opportunities for motor development	Caregivers and Healthcare workers	Home	Standard care: the content of child guidance encompassed physical instruction, along with advice given to parents regarding posture and manipulation; the selection of therapeutic approaches was determined by the therapist and could include NDT, motor learning, developmental skills approaches, or a combination of methodologies	12w
Morgan (2016) [34]	Program: GAME Motor and cognitive enrichment: motor task practice and set up motor enriched play environments; instructed parents to recognize and understand infants’ motor behaviors, and subsequently provided them with opportunities for motor development	Caregivers and Healthcare workers	Home	Standard care: intervention methods included NDT, sensory integration training, etc.	16w
Dusing (2018) [35]	Program: SPEEDI Motor and cognitive enrichment: parents provided daily opportunities designed to support the infants’ emerging motor control and exploratory behaviors Social enrichment: educating parents on parent-infant interaction was conducted	Caregivers	Phase 1: Hospital (NICU) Phase 2: Home	NA	15w
Holmström (2019) [36]	Program: Small Step Motor enrichment: learn new gross motor activities Cognitive enrichment: object exploration triggered different hand actions	Caregivers	Home	Standard care: based on family-centered intervention and functional training	35w
Harbourne (2019) [37]	Program: Motor-Based Problem-Solving Intervention Motor and cognitive enrichment: set up the environment to facilitate small increments of movement, which infants could utilize to solve movement problems.	Healthcare workers	Home	Optimal Pattern Approach: helped babies initiate movement within normal or optimal movement patterns and directly blocked movement patterns that could lead to errors or misalignment	12w
Hielkema (2020a) [38]	Program: COPCA Motor and cognitive enrichment: caregivers identified infants’ signals and responded appropriately to their actual needs, while promoting self-generated motor behaviors, variations, and trial-and-error experiences through play Social enrichment: educating parents on parent-infant interaction was conducted	Healthcare workers	Home	PT: based on NDT principles	12 mon
Hielkema (2020b) [39]	Program: COPCA Motor and cognitive enrichment: caregivers identified infants’ signals and responded appropriately to their actual needs, while promoting self-generated motor behaviors, variations, and trial-and-error experiences through play Social enrichment: educating parents on parent-infant interaction was conducted	Healthcare workers	Home	PT: based on NDT principles	12 mon
Harbourne (2021) [40]	Program: START-play Motor and cognitive enrichment: Embed motor learning and problem-solving skills into games	Caregivers and Healthcare workers	Home	Usual care: most of the infants in the control group received only natural environment Part C services, some of them received only outpatient therapy, and some received both types of services	3 mon
Ziegler (2021) [41]	Program: COPCA Motor and cognitive enrichment: caregivers identified infants’ signals and responded appropriately to their actual needs, while promoting self-generated motor behaviors, variations, and trial-and-error experiences through play Social enrichment: educating parents on parent-infant interaction was conducted	Healthcare workers	Home	Standard care: Standard care was heterogeneous and eclectic, using parent training, and often included components of neurodevelopmental treatment with hands-on techniques	6 mon
Cemali (2022) [42]	Program: Sensory integration training Motor, cognitive and sensory enrichment: provided sensory opportunities, posed just right challenges, avoided negative experiences, cooperated in activity choices, helped with self-organization, supported with the optimum stimuli, created a play context, maximized the child’s success, ensured physical safety, arranged the child’s play environment, and provided an alliance during treatment	Healthcare workers	Hospital (therapy rooms)	PT: based on what the infant can and cannot do in the AIMS	8w
Alberge (2023) [43]	Program: Early post-hospital psychomotor therapy Motor and cognitive enrichment: infants are encouraged to engage and maintain attention, visuospatial environmental exploration, physical exploration, and interactive behaviors	Healthcare workers	Home	N/A	8 mon

EE environmental enrichment, NBAS Neonatal Behavioral Assessment scale, COPCA COPing with and CAring for infants with special needs, GAME Goals-Activity-Motor-Enrichment, SPEEDI Supporting Play Exploration and Early Development Intervention, START Play Sitting Together and Reaching to Play, NICU Neonatal Intensive Care Unit, PT physiotherapy, NDT Neuro-Developmental Treatment, N/A not applicable, AIMS Alberta Infant Motor Scale, mon months, CA corrected age, w weeks

Risk of bias

The risk of bias in the included RCTs was assessed using the Cochrane Risk of Bias Tool 2.0 and is presented in Fig. 2. Of the 14 included articles, 50% were assessed as low risk [22, 23, 38, 39, 41, 47, 48], 36% were assessed as some concerns [21, 30, 31, 37, 42], and 14% were assessed as high risk [32, 43]. The most common cause of the risk of bias was the selection of the reported result.

Fig. 2.

The risk of bias assessment of included studies

Meta-analysis results

Motor development

Six studies, including 269 participants, were included, and a meta-analysis was performed using a fixed-effects model. We extracted the data of BSID-II/III, PDMS-II, and Infant Motor Profile. The pooled meta-analysis revealed a significant difference between the two groups (SMD = 0.35; 95% CI = 0.11 to 0.60; p = 0.004; I² = 0.0%; Fig. 3a), indicating that EE had a significantly greater effect on motor development in infants with or at high risk of CP compared to the control group. The effect size of 0.35 represents a small to moderate beneficial effect of EE on motor outcomes. Moreover, the I² value of 0.0% suggests negligible heterogeneity among the included studies, indicating that the results were highly consistent across trials. The Egger found no publication bias for motor development outcomes (p = 0.886).

Fig. 3.

Forest plot of motor development. a Pooled meta-analysis results of motor development. b Subgroup analysis was conducted by age in months at assessment. I²: I-squared; SMD: standard mean difference; CI: confidence interval.

Endpoint and follow-up data assessed using BSID-II/III were extracted. Subgroup analyses were performed according to infants’ age at assessment: 6-12, 12-18, and 18-24 months. The results of the subgroup analyses revealed significant differences within the 6-12 months subgroup (SMD = 0.47; 95% CI = 0.15 to 0.79; p = 0.004; I² = 0.0%; Fig. 3b) and the 12-18 months subgroup (SMD = 0.57; 95% CI = 0.10 to 1.03; p = 0.018; I² = 0.0%; Fig. 3b), indicating small to moderate beneficial effects of EE on motor development in these age ranges. Specifically, the SMD values of 0.47 and 0.57 suggest that early environmental enrichment interventions have a positive impact on motor outcomes for infants between 6 and 18 months of age. In contrast, no significant effect was observed in the 18-24 months subgroup. Furthermore, the I² values of 0.0% in both significant subgroups indicate no observed heterogeneity, suggesting that the effects reported across the included studies within each subgroup were highly consistent and reliable. This lack of heterogeneity strengthens the confidence in the pooled effect estimates for these age groups. The leave-one-out sensitivity analyses showed that no studies showed a meaningful effect of excluding them from the analysis.

Gross motor function

Seven studies, including 421 participants, were included, and a meta-analysis was performed using the fixed-effects model. We extracted the data of the Alberta Infant Motor Scale, GMFM-66, and the BSID-III. The meta-analysis results found significant differences between the two groups (SMD = 0.25; 95% CI = 0.06 to 0.44; p = 0.011; I² = 0.0%; Fig. 4a), indicating that EE had a significantly greater effect on gross motor function in infants with or at high risk of CP compared to the control group. The effect size of 0.25 represents a small but meaningful improvement in gross motor function due to the EE intervention. Additionally, the I² value of 0.0% indicates no observed heterogeneity among the included studies, suggesting that the effect of EE on gross motor function was consistent across studies. The leave-one-out sensitivity analyses showed that no studies showed a meaningful effect of excluding them from the analysis. The Egger test found no publication bias for gross motor function outcomes (p = 0.856).

Fig. 4.

Forest plot of gross motor function and fine motor function. a Meta-analysis results of gross motor function. b Meta-analysis results of fine motor function. I²: I-squared; SMD: standard mean difference; CI: confidence interval

Fine motor function

Three studies, including 272 participants, were included, and a meta-analysis was performed using a fixed-effects model. We extracted the data of the BSID-II/III and PDMS-II. No significant differences were found between the two groups (SMD = 0.20; 95% CI= −0.04 to 0.43; p = 0.109; I² = 0.0%; Fig. 4b). The leave-one-out sensitivity analyses showed that no studies showed a meaningful effect of excluding them from the analysis. The Egger test found no publication bias for fine motor function outcomes (p = 0.486).

Cognitive development

Six studies, including 372 participants, were included, and a meta-analysis was performed using a fixed-effects model. We extracted the data from the BSID-II/III. The leave-one-out sensitivity analysis revealed that after excluding the study by Blauw-Hospers et al. [32], there was a significant difference between the two groups (SMD = 0.32; 95% CI = 0.10 to 0.54; p = 0.004; I² = 0.0%; Fig. 5a). This effect size of 0.32 indicates a small to moderate beneficial effect of the EE intervention on cognitive developmental outcomes as measured by BSID-II/III. Additionally, the I² value of 0.0% suggests no observed heterogeneity among the remaining studies, indicating that the results were consistent and robust after the exclusion of this study.

Fig. 5.

Forest plot of cognitive function. a Pooled meta-analysis results of cognitive development. b Subgroup analysis was conducted by age in months at assessment; I²: I-squared; SMD: standard mean difference; CI: confidence interval

Similarly, endpoint and follow-up data assessed using the BSID-II/III were extracted, and subgroup analyses were conducted based on infants’ age at assessment. Subgroup analysis revealed significant differences within the 6-12 months subgroup (SMD = 0.33; 95% CI = 0.01 to 0.66; p = 0.046; I² = 0.0%; Fig. 5b). This effect size of 0.33 indicates a small but meaningful improvement in cognitive developmental outcomes for infants receiving the EE intervention within this age range. Furthermore, the I² value of 0.0% suggests no observed heterogeneity among the studies in this subgroup, indicating consistent effects across the included trials. The Egger test found no publication bias for cognitive development outcomes (p = 0.105).

The quality of the evidence

The quality of the evidence was rated using GRADE. The results of the GRADE were as follows: the motor development outcome was moderate, the gross motor function outcome was low, the fine motor function outcome was low, and the cognitive development outcome was high. The main reasons for the downgrading of evidence were high risk of bias, high statistical heterogeneity, and small sample size for many of the outcomes.

Discussion

The meta-analysis conducted by Morgan et al., which included 7 RCTs published between 1988 and 2011, demonstrated the superiority of EE in improving motor development in infants with or at high risk of CP [49]. Given the emergence of updated diagnostic guidelines for high-risk CP and the growing number of high-quality studies in recent years, we were motivated to build upon previous work by establishing refined inclusion and exclusion criteria to evaluate the effectiveness of EE interventions. In this study, we included 14 RCTs involving 592 participants, enabling a more comprehensive and targeted analysis. Beyond confirming the beneficial effects of EE on motor development, our meta-analysis also provides new evidence regarding its positive impact on gross motor function and cognitive development in this population. Moreover, our findings suggest that the optimal time window for EE interventions is between 6 and 18 months of age for improving motor development, and between 6 and 12 months of age for enhancing cognitive development in infants with or at high-risk of CP.

EE interventions appear to have a beneficial effect on motor development, although the magnitude of improvement may vary across studies. This positive impact is likely attributable to the way enriched environments support key physiological and behavioral processes. Motor development is closely linked to both the maturation of the nervous system and the development of the musculoskeletal system. An enriched environment offers infants increased opportunities for environmental exploration, balance and coordination activities, social interaction, voluntary movements, and sensory integration [17]. These experiences collectively stimulate neuromuscular activity and promote the practice and enhancement of muscle strength and coordination [17, 41]. Notably, our findings suggest that infants between 6 and 18 months of age exhibited the most pronounced improvements in response to EE interventions, indicating that this developmental window may represent a particularly sensitive period for environmental stimulation. This effect may be because motor development is inherently age-related. The period from 6 to 18 months of age is critical for the development of key motor skills, including reaching and grasping, sitting, crawling, standing, and walking independently [50, 51]. This finding also aligns well with the observed benefits of EE interventions on gross motor function, as many of these skills fall under the category of gross motor development. However, no significant improvement in fine motor function was observed following EE interventions. Several potential explanations may account for this finding. First, fine motor function often matures later than gross motor function and may require more targeted, repetitive, and task-specific training to yield measurable improvements. Second, some of the assessment tools used may not have been sensitive enough to detect subtle changes in fine motor performance within the limited intervention periods. Finally, the developmental trajectory of fine motor skills is closely intertwined with cognitive and perceptual development [52], which may require longer interventions to show clear effects.

EE interventions provide a stimulating learning environment that encourages infants to develop adaptability and problem-solving skills through trial and error [48]. Our study found that EE interventions can promote cognitive development in infants with or at high risk of CP. This finding is consistent with the results reported by Villouta-Gutiérrez et al., who demonstrated that EE has a positive impact on various neurocognitive dimensions in children and adolescents with intellectual disabilities, including self-determination, intellectual abilities, and motor skills [53]. The horizontal improvement across different neurocognitive variables through EE protocols not only contributes to better cognitive functioning but also equips individuals with practical tools for daily living and social participation. This improvement, in turn, facilitates more effective inclusion in educational, occupational, and social settings. Findings from basic research further support the cognitive benefits of EE by providing plausible neurobiological mechanisms. Animal studies have shown that EE leads to a range of adaptive neural changes, including (1) increases in synaptogenesis and synaptic strength, (2) increased cortical thickness and overall brain weight, and (3) enhanced cell proliferation and neuronal survival [11, 54–56]. These neurobiological adaptations help sustain and enhance cognitive abilities and behavioral functioning.

Since infants are unable to choose their environments voluntarily, caregivers play an irreplaceable role in the implementation of EE interventions. A qualitative study on parental experiences with the Small Step program highlighted the benefits of integrating interventions into daily routines, providing guidance to parents, and fostering collaboration with therapists [25]. These strategies may enhance the effectiveness of interventions and improve adherence. Educating caregivers to recognize infants’ capabilities and engage in meaningful interactions not only contributes to better infant neurodevelopmental outcomes but also supports caregivers’ mental health. The prevalence of anxiety is higher among parents of children with CP compared to those of typically developing children [57]. Among the included studies, three studies examined caregiver-related outcomes in addition to infant-related outcomes [23, 47, 48]. However, due to substantial methodological heterogeneity and the limited number of studies, a meta-analysis was not conducted. Nevertheless, existing evidence suggests that caregivers’ mental health can influence infant development. Balikci found that the “HEP” (Homeostasis–Enrichment–Plasticity), a family-centered EE program, can improve functional impairments in infants with CP and reduce anxiety levels in their caregivers [58]. In summary, we believe that future EE protocols should include at least three key components: (1) educating caregivers to recognize the abilities of infants; (2) teaching caregivers how to interact more effectively and frequently with infants; and (3) providing infants with more learning opportunities.

This study has several strengths. First, this study is the first meta-analysis of the possible effects of EE intervention on gross motor function, fine motor function, and cognitive development. Second, in this meta-analysis, we included all potential studies based on the latest diagnostic criteria for high-risk CP and included more high-quality RCTs to improve the accuracy and confidence of the results. In addition, we conducted subgroup analyses and found the optimal age effect of EE intervention on both motor and cognitive development. This study also had some limitations. One of the main limitations of this study is the heterogeneity in the baseline age of the participants and the age at which they were assessed, which is likely to have influenced the results of our study. Second, the limited quality of the evidence, particularly for outcomes rated as low, suggests that we have limited confidence in the effect estimates and that the true values may be quite different from the estimates.

Conclusion

This systematic review and meta-analysis suggests that EE should be considered an effective tool to improve motor development, gross motor function, and cognitive development in infants with or at high risk of CP. The period from 6 to 18 months of age may represent a critical window for promoting motor development, while cognitive development appears to benefit most from EE interventions administered between 6 and 12 months of age. It suggests that healthcare workers must consider the age of infants when developing intervention plans and emphasize the importance of early intervention for infants with or at high risk of CP. Further high-quality, large-sample, multicenter studies are needed to investigate the effects of EE intervention fully.

Abbreviations

CP

cerebral palsy

EE

environmental enrichment

COPCA

coping with and caring for infants with special needs

GAME

goals - activity - motor– enrichment

SPEEDI

supporting play exploration and early development intervention

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

PsycINFO

Psychological Information Database

SocINDEX

Sociological Index

BSID

Bayley Scales of Infant and Toddler Development

GMFM

Gross Motor Function Measure

PDMS

Peabody Developmental Motor Scales

RCTs

randomized controlled trials

RoB 2.0

The Cochrane Risk of Bias Tool 2.0

GRADE

Grading of Recommendations Assessment, Development and Evaluation Working Group

SMD

standard mean difference

CI

confidence interval

I²

I-squared

START-Play

Sitting Together and Reaching to Play

Authors’ contributions

QD and XL designed this study. XYZ, XL, and MDJ searched and screened the literature. JPL and XZ extracted the data. MDJ and JYZ were assessed for quality (including risk of bias and evaluation of the level of evidence). XYZ and XL analyzed the data and wrote the first draft. QLM and QD helped discuss the results and revise the manuscript. QLM and XL have accessed and verified the underlying data. All authors read and approved the final version of the manuscript. All authors have full access to all data in the study and are ultimately responsible for the decision to submit for publication.

Funding

This study was supported by the National Key R&D Program of China (2023YFC3604800) and the Xinhua Hospital-Shanghai Jiaotong University coordinate project of medical robot assignment (21XJMR03).

Data availability

The full data set used to support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.