Secondary Logo

Journal Logo

Research Report

Predicting Neurological Disorders in Infants with Extremely Low Birth Weight Using the Movement Assessment of Infants

Salokorpi, Teija, MD, PhD; Rajantie, Irmeli, PT; Kivikko, Irmeli, PT; Haajanen, Ritva, PT; Rajantie, Jukka, MD, PhD

Author Information



The Movement Assessment of Infants (MAI) was developed to assess motor behavior of infants at high risk for developmental disorders. The MAI includes measures of motor behavior in four areas: muscle tone, primitive reflexes, automatic reactions, and volitional movements. 1 The MAI is designed to not only identify motor disorders during the first year, but also to establish the basis for early interventions. 1,2 For infants born at term, the predictive validity of the MAI at the ages of four and eight months was demonstrated by Harris et al 2 and Swanson et al. 3 Moderate reliability of the MAI was demonstrated by Harris et al. 4 We decided to use the MAI at the corrected age of four months to identify the children with highest developmental risks to provide intensive physiotherapy. During the first time period from 1991 to 1994, the MAI was used for those infants who were suspected to have motor disorders at the time of the three-month neurological examination. During the second period from 1997 to 1998, the MAI was completed for a randomly selected sample of infants with extremely low birth weight (ELBW) born on odd-numbered days. The neurological outcome of the infants born between 1991 and 1994 was published earlier, 5–7 and the follow-up of the children born in 1997 and 1998 is still ongoing. The aim of this study was to calculate the predictive values of the MAI for cerebral palsy (CP) and minor neurologic disorders (MND) in a group of infants of ELBW at high risk for developmental disorders.



Subjects were infants of ELBW born from January l, 1991 to December 31, 1994 and from January 1, 1997 to December 31, 1998, who were admitted to the neonatal intensive care unit (NICU) at the Helsinki University Central Hospital. During the first period, 228 infants of ELBW were admitted to the hospital. Seventy-two (31%) of the neonates died during the primary hospitalization, and 143 children (91% of the survivors) completed the follow-up at the age of two years. 5 During this period, the MAI was completed on 13 infants with a mean birth weight of 879 g (range 685–970 g) and a mean gestational age at birth of 25 weeks (range 24–31 weeks). 5 During the second period, 97 neonates of ELBW were admitted to the hospital, and 23 (24%) of them died during the primary hospitalization. The MAI was performed on 21 infants randomly selected from among those who were born on odd-numbered days. Infants studied during the second period had a mean birth weight of 723 g (500–980 g) and a mean gestational age of 26 weeks (24–31 weeks). The two groups were statistically similar in the terms of birth weight, gestational age, sex, and the type of delivery.


In all, 34 infants were assessed at the corrected age of 4 months ± 2 weeks using the MAI. The assessments were performed by two of the three experienced physiotherapists working together, and every examination was videotaped for later detailed evaluation. The total high-risk points were calculated according to the manual. Scores less than10 were considered low-risk values, and scores greater than10 were considered high-risk values for CP. For MND, scores less than five were considered low-risk values, and scores greater than five were considered high-risk values.

Neurological follow-up examinations using the Munchener Funktionelle Entwicklungs Diagnostik (MFED) 8 took place at the corrected ages of three, six, nine, and 12 months, and a qualitative physiotherapeutic examination using a national method 9 was completed during the same follow-up visits. In addition to the MFED, the neurological assessment included examination of primitive reflexes, automatic reactions, and muscle tone. At all ages up to 12 months, the MFED includes assessment of the infant’s neurological development in six areas: Crawling, Sitting, Standing, Reaching, Perceiving, and Speaking. After nine months, Social Skills also are assessed. If the developmental age of the infant was lower than his or her corrected age in three to six fields, the result was considered to indicate a minor neurological disorder. A delay in all six or seven areas was considered to indicate mental retardation. If the neurological examination revealed abnormal muscle tone, exaggerated or constant primitive reflexes, delayed automatic reactions and constant delay in motor development, the child was considered to demonstrate CP. A neuropsychological examination took place at the corrected age of 24 months using the Bayley Scales of Infant Development (BSID). 10,11 The mental developmental index (MDI) was considered normal with values of more than 85 and MND was diagnosed with values of 70 to 85. MDI values of less than 70 indicated mental retardation. 11

Statistical Analyses

The predictability of the MAI was measured in terms of sensitivity, specificity, positive predictive value, and negative predictive value, which were calculated according to the following formulas:12

  • Sensitivity = Number of true-positives/Number of true-positives + Number of false-negatives
  • Specificity = Number of true-negatives/Number of true-negatives + Number of false-positives
  • Positive predictive value = Number of true-positives/Number of true-positives + Number of false-positives
  • Negative predictive value = Number of true-negatives/Number of true-negatives + Number of false-negatives

The scores of the infants who developed CP were compared with the scores of all the other infants, and the scores of the infants who developed MND were compared with the scores of the infants without neurological symptoms.


Among the infants entering the study during the first period (1991–1994) by the age of two years, a CP syndrome was identified in 27 (19%) of the children, MND in 36 (25%), mental retardation in 13 (9%), and normal neurological development in 67 (47%) of the children. 5 Among infants entering the study during the second period (1997–1998), one child was lost to follow-up. By the age of 24 months of the 73 children, six (8%) were diagnosed with CP, 20 (28%) with MND, and 47 children (64%) showed normal neurological development. None of the children was mentally retarded (Table 1).

Table 1
Table 1:
Outcome at Age Two Years (Period 1) and One to Two Years (Period 2) and the Number of Completed MAI Assessments

The MAI-test total risk scores (TRS) of the study infants are shown in Table 2. Of the nine infants with a TRS greater than 10, seven developed a CP syndrome. Whereas 10 of the 16 infants with a TRS less than five had an outcome considered normal, only one developed a CP syndrome (hemiplegia) later. All the infants with a TRS greater than 13 had a CP syndrome diagnosed later. Of the 16 children with a TRS less than five, 10 children showed typical neurological development and five children developed MND later (Table 2).

Table 2
Table 2:
The MAI Total Risk Scores of the 34 Children in Different Outcome Groups

The four MAI subcategories, muscle tone, primitive reflexes, automatic reactions and volitional movements, equally predicted typical development when scores were low (zero to two), but a high score (more than three) in the automatic reactions subcategory predicted CP 100% of the time (Table 3). In the case of MND, the subcategories muscle tone and primitive reflexes were found to be more important than automatic reactions and volitional movement. In the group of children with MND, high total risk scores (more than five) were associated with abnormalities (subcategory scores more than three) in muscle tone and primitive reflexes. In contrast to this, low total risk scores (less than five) of the infants with typical development correlated best with low sub scores (zero to two) in primitive reflexes and automatic reactions.

Table 3
Table 3:
Distribution of the Children (%) with Low Risk and High Risk in Subcategories of the MAI

The sensitivity for CP with the cut-off-point of more than 10 was 64%, and the specificity 91%. The positive predictive value was 78% and negative predictive value was 84%. With the cut-off-point more than five, sensitivity for MND was 44%, specificity 71%, and the corresponding positive and negative predictive values were 50% and 67%.


The MAI has already been shown to be both sensitive and specific for neurological disorders at four and eight months in infants born at term and in infants born preterm with low birth weight. 2,3,13 However, the infants in the Swanson et al 3 study were not extremely preterm, had a mean birth weight of 1289 g and a mean gestational age of 29.5 weeks. The aim of our study was to investigate the validity of the MAI in early detection of neurological disorders in infants of ELBW. This is a difficult task because many transient signs such as hypo- or hypertonus, asymmetric positioning, and exaggerated primitive reflexes are commonly present in infants born extremely premature. These signs, except for asymmetry, were shown as predictors for CP by Harris. 14 The same signs increased the total risk score in the MAI. 1 This result was also confirmed in our study. A total risk score of more than l0 was a strong predictor for CP, but also every single high subcategory score (≥three) of the four subcategories: muscle tone, primitive reflexes, automatic reactions, and volitional movements, was predictive for CP (Table 3).

Early detection of minor disorders is more complicated, especially in infants of ELBW who commonly show abnormalities in muscle tone and other transient signs at early age. In our study, as in previous studies, the MAI was found to predict typical development or CP better than minor disorders. There was a trend, however, showing that high scores in the subcategories of muscle tone and primitive reflexes predicted MND, whereas a high score in the subcategory of automatic reactions predicted CP when the total risk score was more than five.

Our results, thus, revealed very good specificity and acceptable sensitivity of the MAI in predicting CP at age four months. The values of specificity and sensitivity are naturally dependent on the cut-off-point of the scale. With the cut-off-point greater than 13, specificity for CP increased to 100%, but sensitivity decreased to 64%. Therefore, we decided to use the total risk score of 10 as the cut-off-point, as suggested in the manual. 1 For MND, no suggestions were given in the manual, and we chose the value of more than five to increase the sensitivity. The specificity value of 71% was acceptable, as at that time there were no other methods to detect minor disorders so early.

The criticism against early diagnostic examinations has been based widely on the controversial results of early intervention studies on infants born premature with a majority of reports showing no, or only minor, benefits of the intervention on the outcome. 15,16 However, there is evidence that interventions can decrease the impairments associated with CP if the therapy is started early. 16 A few reports have also shown that MND signs can he minimized too, and secondary social problems partly prevented with early intensive therapies. 17,18 We think that these reports justify the struggle toward early diagnosis and early interventions.


Early diagnosis of neurological disorders is very difficult in infants born extremely premature. That was the reason to investigate the validity of the MAI in infants of ELBW. For predicting CP, the MAI is highly specific and demonstrates acceptably sensitive. The negative and the positive predictive values are also good for CP. For minor disorders, however, this method has poor sensitivity and only acceptable specificity, which led us to search for other methods.


1. Chandler LS, Swanson MW, Andrews MS. Movement Assessment of Infants: A Manual. Rolling Bay, Washington: Chandler, Swanson, and Andrews; 1980.
2. Harris SR, Swanson MW, Andrews MS, et al. Predictive validity of the “Movement Assessment of Infants.” J Dev Behav Pediatr. 1984; 5: 336–342.
3. Swanson MW, Bennett FC, Shy KK, et al. Identification of neurodevelopmental abnormality at four and eight months by the movement assessment of infants. Dev Med Child Neurol. 1992; 34: 321–337.
4. Harris SR, Haley SM, Tada WL, et al. Reliability of observational measures of the movement assessment of infants. Phys Ther. 1984; 64: 471–475.
5. Salokorpi T. Neurological Disorders, Risk Factors, and Prevention in Tiny Preterm Infants. Helsinki: The University of Helsinki, 1999.Thesis.
6. Salokorpi T, Rajantie I, Viitala J, et al. Does perinatal hypocarbia play a role in the pathogenesis of cerebral palsy? Acta Paediatr. 1999; 88: 1–5.
7. Salokorpi T, Rautio T, Sajaniemi N, et al. Neurological development up to the age of four years of extremely low birth weight infants born in Southern Finland in 1991–94. Acta Paediatr. 2001; 90: 14.
8. Hellbrügge T, Lajosi F, Menara D, et al. Munchener Funktionelle Entwicklungsdiagnostik: Fortschritte der Sozialpädiatrie. Munich: Urban and Schwarzenberg; 1978.
9. Rajantie I, Topelius A. Alle 1000 g painoisena syntyneiden keskosten perusliikkumisen kehitys 18 kk korjattuun ikään asti. Helsingin yliopistollinen keskussairaala. Tutkimusjulkaisuja. 1993; 5: 93.
10. Bayley N. Bayley Scales of Infant Development: Manual. New York: The Psychological Corporation; 1969.
11. Bayley N. Bayley Scales of Infant Development. 2nd ed. San Antonio, Tex: The Psychological Corporation; 1993.
12. Hennekens CII, Buring JE, Mayrent SL, eds. Epidemiology in Medicine. Boston: Little, Brown & Co; 1987: 331–335.
13. Harris SR. Early detection of cerebral palsy: sensitivity and specificity of two motor assessment tools. J Perinatol. 1987; 7: 11–15.
14. Harris SR Early neuromotor predictors of cerebral palsy in low-birthweight infants. Dev Med Child Neurol. 1987; 29: 508–519.
15. Hur JJ. Review of the research on therapeutic interventions for children with cerebral palsy. Acta Neurol Scand. 1995; 91: 423–432.
16. Tumbull JD. Early intervention for children with or at risk of cerebral palsy. Am J Dis Child. 1993; 147: 54–59.
17. The Infant Health and Development Program. Enhancing the outcome of low-birth-weight, premature infants: a multisite, randomized trial. JAMA. 1990; 263: 3035–3042.
18. Matsuishi T, Ishibashi S, Kamiya Y, et al. Early intervention for low-birth-weight infants. Brain Dev. 1998; 20: 18–21.

infant; very low birth weight; neurologic examination; sensitivity and specificity,; cerebral palsy diagnosis

Copyright © 2001 Academy of Pediatric Physical Therapy of the American Physical Therapy Association