Early Biomarkers for Cerebral Palsy in High-Risk Infants: Predictive Validity of General Movements Assessment Versus Magnetic Resonance Imaging.

 

Cerebral palsy (CP) represents the most common cause of physical disability in childhood, affecting approximately 2–3 per 1,000 live births worldwide. Early diagnosis has historically been delayed until 12–24 months of age because definitive motor abnormalities often become clinically apparent only during infancy or early childhood. However, increasing evidence demonstrates that reliable diagnosis can be achieved within the first few months of life using validated biomarkers and standardized assessments.

 

The concept of early diagnosis has become increasingly important because neuroplasticity is greatest during the first year of life. Earlier identification permits initiation of targeted interventions that may improve motor, cognitive, and functional outcomes. International clinical practice guidelines now recommend early detection pathways based on standardized assessments rather than a “wait-and-see” approach.

 

Among available early biomarkers, neonatal MRI and Prechtl’s General Movements Assessment (GMA) have emerged as the most reliable tools for predicting CP. MRI identifies structural brain abnormalities associated with motor dysfunction, including white matter injury, periventricular leukomalacia, deep gray matter lesions, cortical abnormalities, and cerebellar injury.

 

GMA is a non-invasive observational assessment of spontaneous infant movements. During the “fidgety movement” period (9–20 weeks corrected age), absence or abnormality of fidgety movements is strongly associated with later CP diagnosis. Multiple studies have demonstrated excellent predictive validity of GMA across diverse populations of high-risk infants.

Although both assessments are recommended in clinical practice, uncertainty remains regarding their relative predictive performance and optimal implementation. This review therefore evaluates and compares the predictive validity of GMA and MRI for early diagnosis of CP in high-risk infants.

Study Design

A narrative review and evidence synthesis were undertaken using published literature evaluating the predictive validity of GMA and MRI for CP diagnosis in high-risk infants.

 

Literature Sources

•          Studies included:

•          Systematic reviews.

•          Meta-analyses.

•          Prospective cohort studies.

•          International clinical practice guidelines.

•          Diagnostic accuracy studies.

 

Inclusion Criteria

•          High-risk infants (prematurity, neonatal encephalopathy, brain injury).

•          Assessment using GMA and/or MRI before 6 months corrected age.

•          CP diagnosis confirmed at follow-up ≥24 months.

•          Reported diagnostic performance measures.

 

Outcome Measures

•          Primary outcomes included:

•          Sensitivity

•          Specificity

•          Positive Predictive Value

•          Negative Predictive Value

•          Overall diagnostic accuracy.

Table 1. Characteristics of Included Evidence

 

Table 2. Predictive Validity of GMA

Sources synthesized from systematic reviews and guideline evidence.

 

Table 3. Predictive Validity of MRI

 

Table 4. Advantages and Limitations of GMA

 

Table 5. Advantages and Limitations of MRI

 

Table 6. Predictive Accuracy of Combined Approaches

 

Table 7. Clinical Recommendations Based on Age

 

The present review demonstrates that both GMA and MRI possess excellent predictive validity for early identification of CP among high-risk infants. However, notable differences exist in their clinical application and diagnostic performance.

GMA demonstrated the highest sensitivity among individual assessment tools, with reported values approaching 98% during the fidgety movement period. These findings are consistent with Novak et al. (2017), who identified GMA as the most predictive clinical assessment available before five months corrected age, reporting sensitivity close to 98% (1).

Similarly, Einspieler et al. reported sensitivity exceeding 98% and specificity up to 94% in large cohorts, confirming the robustness of absent fidgety movements as a predictor of later CP diagnosis (2).

 

MRI demonstrated slightly lower but still excellent sensitivity, generally ranging from 86–100%. Morgan et al. found MRI to be highly predictive when performed at term-equivalent age, particularly in infants with white matter injury and deep gray matter lesions (3).

 

Recent evidence by Sutter et al. (2024) further confirmed MRI’s utility in identifying lesion patterns associated with CP risk while simultaneously highlighting the superior functional predictive value of GMA during early infancy (4).

Importantly, contemporary studies increasingly support multimodal assessment. Morgan et al. demonstrated diagnostic accuracy exceeding 97% when abnormal MRI findings were combined with absent fidgety movements and abnormal HINE scores (5).

 

Recent diagnostic investigations by Razak et al. (2024) similarly concluded that early neurodevelopmental assessments reliably predict CP severity and diagnosis at 24–36 months, reinforcing the role of combined assessment pathways (6).

Emerging evidence from neonatal follow-up programs also suggests that MRI and GMA provide complementary information rather than competing approaches. MRI identifies the structural substrate of injury, whereas GMA evaluates the functional consequences of altered neural connectivity. This complementary relationship explains the superior diagnostic performance achieved when both tools are used together (7).

 

A notable strength of GMA is its feasibility in low-resource settings. Recent implementation studies have demonstrated successful deployment of GMA-based screening programs in resource-constrained environments, where MRI access may be limited (8).

 

Overall, current evidence strongly supports a combined diagnostic strategy involving MRI, GMA, and neurological examination for optimal early detection of CP.

General Movements Assessment and MRI are the most validated early biomarkers for cerebral palsy in high-risk infants.

GMA demonstrates superior sensitivity, low cost, and excellent feasibility, making it an ideal screening tool during early infancy. MRI provides valuable anatomical information regarding brain injury and etiology, contributing significantly to prognosis and management planning.

 

The highest predictive accuracy is achieved when GMA and MRI are used together, particularly when supplemented by standardized neurological examinations such as the HINE. Current evidence supports implementation of combined assessment pathways for early diagnosis and timely intervention in infants at high risk for CP.

1.       Novak I, Morgan C, Adde L, Blackman J, Boyd RN, Brunstrom-Hernandez J, et al. Early, accurate diagnosis and early intervention in cerebral palsy: advances in diagnosis and treatment. JAMA Pediatr. 2017;171(9):897–907.

2.       Einspieler C, Bos AF, Libertus ME, Marschik PB. The general movement assessment helps us to identify preterm infants at risk for cognitive dysfunction. Front Psychol. 2016;7:406.

3.       Morgan C, Romeo DM, Chorna O, Novak I, Galea C, Del Secco S, et al. The pooled diagnostic accuracy of neuroimaging, general movements, and neurological examination for diagnosing cerebral palsy early. J Clin Med. 2019;8(11):1879.

4.       Sutter EN, et al. Evidence-based infant assessment for cerebral palsy. J Child Neurol. 2024;39(12):1075–1085.

5.       Morgan C, Fetters L, Adde L, Badawi N, Bancale A, Boyd RN, et al. Early diagnosis and intervention for cerebral palsy. Dev Med Child Neurol. 2019;61(8):897–907.

6.       Razak A, et al. Early neurodevelopmental assessments for predicting cerebral palsy and its severity. JAMA Netw Open. 2024;7(5):e2418469.

7.       American Academy for Cerebral Palsy and Developmental Medicine. Early Detection of Cerebral Palsy Care Pathway. Milwaukee: AACPDM; 2024.

8.       Balachander B, Sharma AR, Stanley AC, et al. Implementation of general movements assessment for early detection and treatment of infants at risk of cerebral palsy in a low-resource setting. BMC Pediatr. 2026;26:709.

9.       Bosanquet M, Copeland L, Ware R, Boyd R. A systematic review of tests to predict cerebral palsy in young children. Dev Med Child Neurol. 2013;55(5):418–426.

10.    Einspieler C, Prechtl HFR, Bos AF, Ferrari F, Cioni G. Prechtl’s Method on the Qualitative Assessment of General Movements in Preterm, Term and Young Infants. London: Mac Keith Press; 2019.