Postural Reactions in Infants with Down Syndrome Relationship to Motor Milestone Development and Age STEPHEN M. HALEY

I examined the emergence of automatic postural reactions (righting, equilibrium, and protective reactions) in a group of infants with Down syndrome and in a developmentally matched group of nonhandicapped infants. I assessed the relationship of postural reactions to chronological age and to motor milestone development (as measured by the Bayley Scales of Infant Development). The relationship of postural reactions to chronological age was weaker in the infants with Down syndrome; a regression analysis revealed a slower rate of postural reaction development evident at an age of about five months and older in infants with Down syndrome. The association between postural reactions and motor milestone performance was remarkably similar for both infant groups in degree of association and pattern across the span of motor development tested. The results indicated that, even though the rate of development differed, the associ­ation between postural reactions and motor milestones was similar in normal and delayed development. These results support the current rationale that promotes intervention efforts focusing on the facilitation of postural reactions to enhance motor milestone development in infants with Down syndrome. Key Words: Child development, Down's syndrome, Equilibrium, Infant, Motor skills,

Posture.

The investigation of movement behavior in infants with Down syndrome has concentrated mainly on the description and extent of gross motor milestone delays, such as sitting and walking.1-3 Surprisingly, the relationship of these gross motor delays to other motor dimensions has not been exam­ined systematically in the infant and young child with Down syndrome. One clinically relevant motor dimension with which to compare the development of gross motor milestones is the system of automatic postural reactions.

The emergence of automatic postural reactions is an im­portant prerequisite for the attainment of motor skills in the course of early development. According to Bobath, the se­quence of appearance of the postural reactions "underlies" and "explains" the ontogeny of motor behavior in infancy.4

Milani-Comparetti and Gidoni have identified the postural reactions (righting, equilibrium, and protective reactions) as the underlying responses that are related most functionally to motor milestones.5 Postural reactions provide automatic sup­port and stability to the head, trunk, and extremities and facilitate normal weight shifts and mobility. They function to keep the head aligned vertically during movement, to provide balance and postural shift mechanisms when the center of gravity is disturbed, and to establish protective responses to prevent a fall (luring rapid movements.6,7

The well-documented gross motor delays in infants with Down syndrome have generated an interest in therapists and educators to develop training programs to stimulate mobility goals. A number of clinical reports8,9 and early motor inter­vention efforts10-12 have been published concerning the motor delays caused by Down syndrome. Even though intervention strategies often include facilitation of postural reactions, little attention has been given to the nature of postural reaction deficits and their possible association with the delay of motor milestone achievement in infants and children with Down syndrome.

Cowie, in an extensive longitudinal study, found a delayed integration of primitive reflexes and a delayed emergence of righting and equilibrium reactions in a sample of 79 infants with Down syndrome.13 The strength of the relationship of these delays of postural reactions with motor milestone delays, however, was not examined. Molnar has examined longitu­dinally the relationship of postural reactions to motor delays in a group of young retarded children.1415 Her results indi­cated a strong temporal relationship between the emergence of postural reactions and the subsequent acquisition of gross motor milestones in motorically delayed young children. None of her subjects were reported to be children with Down syndrome.

The purpose of this study was to examine the system of postural reactions as a component that may have a strong interrelationship with the development of gross motor achievements. I predicted that the development of postural reactions would be associated strongly with the development of motor behavior in infants with Down syndrome, even though the motor behavior may be delayed considerably when compared with nonhandicapped infants.

Specifically, I expected to find the following relationships on examination of postural reactions, variables of motor milestone development, and chronological ages of infants with Down syndrome and nonhandicapped infants: 1) The rela­tionship between age and postural reactions in infants with

Dr. Haley was a student in the doctoral program, Department of Educational Psychology, College of Education, University of Washington, Seattle, WA, when this study was conducted. He is now Assistant Professor, Department of Physical Therapy, Sargent College of Allied Health Professions, University Rd, Boston, MA 02215 (USA).

This study was conducted in partial fulfillment of the requirements for the PhD degree, College of Education, University of Washington.

This project was made possible with funding from the Washington Associa­tion for Retarded Citizens. This paper was adopted from a presentation at the Combined Sections Meeting, American Physical Therapy Association, Hous­ton, TX, February 1984.

This article was submitted April 18, 1985; was with the author for revision three weeks; and was accepted July 11, 1985.

Volume 66 / Number 1, January 1986 17

TABLE 1 Age Grouping of Total Infant Sample

Groups by Age

1 2 3 4

TOTAL

n

6 5 4 5

20

Down Syndrome

range 2-7 7-12

12-18 18-24 2-24

Age (mo)

4.39 10.23 15.34 19.68 11.88

s 1.03 1.26 2.31 1.31 6.12

n

10 10 10 10 40

Nonhandicapped

range 2-4 4-6 6-8 8-10 2-10

Age (mo)

2.70 4.91 7.02 9.27 5.97

s 0.56 0.59 0.56 0.57 2.50

Down syndrome would be moderate; however, the relation­ship would be stronger in nonhandicapped infants; 2) the regression equation describing the relationship between age and postural reactions would have a smaller rate of change for the Down syndrome group than for the nonhandicapped group, thus, indicating a delayed development of postural reactions in infants with Down syndrome; 3) a strong, positive relationship would exist between postural reactions and motor milestone level as given by the Bayley Motor Score in infants with Down syndrome; this relationship would be similar in nonhandicapped infants; and 4) the regression equation de­scribing the relationship between postural reactions and the Bayley Motor Score would have a similar rate of change and level for both infant groups, indicating that, at any motor performance level, postural reaction development would be similar across infant groups.

METHOD

Subjects

Twenty infants with Down syndrome, ages 2 to 24 months, were recruited from four early intervention programs located in the greater King County, Washington, area. The majority of the infants with Down syndrome were enrolled in the Training Research and Intervention Program for infants and parents at the Child Development and Mental Retardation Center at the University of Washington. Forty nonhandi­capped infants, ages 2 to 10 months, were recruited mainly from local community college parent-infant classes to serve as a normative comparison group. All were full-term infants (greater than 37 weeks gestational age). Male infants com­posed 35% of the infants with Down syndrome and 58% of the nonhandicapped infants. The age groupings of the infants are presented in Table 1. The demographic characteristics of parents in both infant groups were very similar; the nonhand­icapped group had a slightly higher proportion of parents who had achieved a college education than the parents of the infants with Down syndrome. All parents signed a consent form approved by the Human Subjects Committee at the University of Washington before participation in the project.

For proper assessment of the association of postural reac­tions and motor milestone development, both infant groups ideally should span the same developmental motor stages with equal distributions of infants sampled across all motor devel­opmental levels. In nonhandicapped infants, postural reac­tions begin to appear around the age of 2 months and are nearly all present by age 10 months. To incorporate the same postural reaction and motor milestone developmental span in the Down syndrome group, I extended the age range to 24

Fig. 1. Regression lines: Postural reactions with age.

months. The sample shows an equal spread of nonhandi­capped infants across the age groupings of 2 months; the spread of infants with Down syndrome across the age span was closely approximated (Tab. 1). A sample selection of infants that did not spread equally across the developmental ages possibly could bias results at either extreme of the con­tinuum.

Procedure

All nonhandicapped infants and infants with Down syn­drome were seen for one test session by one of three exam­iners. At that time, each infant was assessed with the Bayley Motor Scale and a series of tests for postural reactions. The parents were allowed to be present during the test sessions. Frequent breaks were needed for the very young infants during the sessions. Most of the nonhandicapped infants were tested in a small clinic room at the Child Development and Mental Retardation Center at the University of Washington; a majority of the infants with Down syndrome were tested in their own early intervention classroom during their regularly

18 PHYSICAL THERAPY

RESEARCH

scheduled program time. A small number of the infants (less than 10%) had to be tested in the home because of scheduling or transportation problems. Previous reports have not pro­vided evidence that a setting difference significantly alters infant test scores.16

Measures The measures reported here include an assessment of pos­

tural reactions modified from the Movement Assessment of Infants (MAI)17 (copies are available from the author on request), an assessment of motor milestone development us­ing the Bayley Scales of Infant Development,18 and chrono­logical age.

Postural reactions were defined as a combination of partial and full responses to 12 items: four righting reactions, five equilibrium reactions, and three protective reactions. Two of the equilibrium tests, supine and all-fours positions, are not part of the MAI, but were included here for comparison with other studies that have examined these responses.5,14,15 The dependent measure used was the total number of the partial and full postural reaction responses.

The motor section of the Bayley Scales of Infant Develop­ment also was administered to each infant. These scales have been recommended specifically for the assessment of motor development in the Down syndrome population.19 The inter-rater reliability of the Bayley raw score has been reported to be high (r = .90).20

Because many of the scores of the infants with Down syndrome are greater than 3.00 standard deviations below the mean for their age, the standard scores could not be calculated. Therefore, the raw score of the Bayley Motor Scale was used in the analysis of estimates of overall motor milestone devel­opment.

Interrater agreement for the postural reactions was calcu­lated using a form of the intraclass correlation coefficient (ICC).21 For a sample of 10 infants examined throughout the study, reliability for the Bayley Motor Score was .98; reliability for the postural reactions was .94.

Data Analysis

Measures included in the analyses were the raw score from the Bayley Motor Scale, a summary score for the number of

postural reactions present, a postural reaction subscore in three areas (righting, equilibrium, and protective reactions), and chronological age.

To examine the degree of association among the variables, I calculated Pearson product-moment correlations within each infant group. The difference in the magnitude of the correlations between groups was tested for significance using Fisher's Z transformation.22

I also used an analysis of covariance (ANCOVA) to exam­ine infant group differences in the proposed relationships of postural reactions, age, and motor milestones. Regression lines for each infant group were developed to compare age and motor milestone achievement with the number of pos­tural reactions. These pairs of regression lines were tested for homogeneity of slope. The regression lines were tested using an ANCOVA if the slopes were similar or using a computer program to determine regions of significance if the lines were different and the slopes were heterogeneous.23

RESULTS Nonhandicapped infants tended to have a somewhat

stronger degree of relationship between age and postural reactions (r - .91) than infants with Down syndrome (r = .77), although the difference between the two groups did not meet a level of significance at the p = .05 level. The differences in the rate of change of postural reactions to age in both infant groups are illustrated in Figure 1. The regression solu­tions are given in Table 2. The two regression lines describing the relationship between age and postural reactions are con­sidered different at approximately the age of 4.6 months (Tab. 3).

The association between postural reactions and scores on the Bayley Motor Scale for both infant groups was high and positive. The magnitude of correlation for the Down syn­drome group (r = .97) was significantly greater than that for the nonhandicapped group (r = .90), (Z = 2.05, p = .04). The ANCOVA regression lines for the summary postural reactions score and the raw score on the Bayley scale are given in Figure 2. In addition, the results of the ANCOVA using the subtotals for the three components of the postural reactions score (righting, equilibrium, and protective reactions) for both in­fant groups and the Bayley Motor Score as the covariate are in Table 3. No difference was seen in the adjusted means of

TABLE 2 Adjusted Means and Regression Solutions for Group Comparisons of Postural Reactions with Bayley Score and Age

Comparison

Postural reactions with age

Postural reactions with Bayley

Righting reactions with Bayley

Equilibrium reactions with Bayley

Protective reactions with Bayley

Infant Group

DSa

NHb

DS NH DS NH DS NH DS NH

Mean Postural

Reactions

22.90 23.52 22.90 23.53 9.70 9.85 7.70 8.27 5.40 5.40

Adjusted Mean

Postural Reactions

18.70 25.62 22.65 23.65 9.62 9.88 7.61 8.31 5.31 5.44

r

11 .90 .96 .89 .88 .85 .95 .75 .96 .89

Regression Equation

Y = 14.4 + 0.71X Y = 10.9 + 2.12X Y = 7.91 + 0.52X Y = 9.14 + 0.51X Y = 5.07 +0.16X Y = 5.34 +0.16X Y = 2.59 +0.17X Y = 3.37 +0.17X Y = 0.32 +0.17X Y = 0.42 +0.17X

a Down syndrome. b Nonhandicapped.

Volume 66 / Number 1, January 1986 19

TABLE 3 Analysis of Covariance Summary: Group Comparisons of Postural Reactions with Bayley Score and Age

Comparison Postural reaction with age

Postural reaction with Bayley

Righting reaction with Bayley

Equilibrium reactions with Bayley

Protective reactions with Bayley

Source within + residual age group age x group within + residual Bayley Motor Score group Bayley x group within + residual Bayley Motor Score group Bayley x group within + residual Bayley Motor Score group Bayley x group within + residual Bayley Motor Score group Bayley x group

SS 499.87 704.76 435.43 370.91 311.02

1686.60 13.26

.09 56.72

163.93 .94 .01

101.07 194.93

6.56 .01

37.02 199.15

.21

.01

df 56

1

a Region of significance for covariate was age >4.6 months.

Fig. 2. Regression lines: Postural reactions with Bayley Motor Score.

postural reactions (Tab. 3) or the slope of the regression lines between the two infant groups (Fig. 2). In general, nonhand-icapped infants tended to have a higher level of postural reactions for each corresponding Bayley Motor Score across the entire range of scores, although the difference was not significant. The regression lines between infant groups for each subsection of the postural reactions and Bayley Motor

Score all had homogeneity of slope. For each type of postural reaction (righting, equilibrium, and protective), the nonhand-icapped infants tended to have a higher level of reactions at corresponding Bayley raw score levels, although none of the differences between adjusted mean values reached signifi­cance. The equilibrium responses tended to differ the most between infant groups across the range of Bayley scores (Tab. 3).

DISCUSSION

As predicted, the magnitude of the association between postural reactions and chronological age was considerably less in the infants with Down syndrome than in the nonhandi-capped group. Furthermore, the regression lines describing the association between postural reactions and age had a slower rate of change (smaller slope) in the infants with Down syndrome, which indicates a slower development of these reactions than a similar developmental group of nonhandi-capped infants. This finding is consistent with previous reports that have indicated delays in motor development in infants with Down syndrome increase after the first six months.13 It further suggests that delays in development become much more obvious when the expected development of many of the postural reactions between four to six months does not occur in infants with Down syndrome. Harris24 reports that certain precocious behaviors may exist in some infants with Down syndrome during the early months, such as early rolling from prone to supine positions, in part because of the lack of development of righting reactions. In my study, the regression lines indicating the relationship between postural reactions and age can be considered significantly different at age 4.6 months and older.

The relationship between postural reactions and the Bayley Motor Score actually is higher in infants with Down syndrome than in nonhandicapped infants. This may indicate a lack of

20 PHYSICAL THERAPY

1 1

56 1 1 1

56 1 1 1

56 1 1 1

56 1 1 1

MS 8.92

704.76 435.43 370.91

5.55 1686.60

13.26 .09

1.01 163.93

.94

.01 1.80

194.93 6.56

.01

.66 199.15

.21

.01

F

78.95 48.78 41.55

303.67 2.38

.10

161.83 .93 .01

108.00 3.63

.01

301.19 .32 .01

P

.001a

.001

.001

.001 NS NS

.001 NS NS

.001 NS NS

.001 NS NS

RESEARCH

variability of expression of postural reactions at each motor milestone level in infants with Down syndrome. The very high correlation supports the notion that postural reactions and milestone achievements have an exceedingly high corre­spondence throughout early motor development.

The regression lines describing the relationship between postural reactions and the Bayley Motor Scale further support the notion that nearly similar levels of postural reactions are present at corresponding motor milestone levels. The number of equilibrium reactions, however, tends to be slightly less across the motor milestone spectrum for infants with Down syndrome. Even though the difference in equilibrium reac­tions between groups is not significant, an important pattern emerges. This pattern suggests that infants with Down syn­drome develop only those equilibrium reactions necessary for the acquisition of a motor milestone phase with very little intrasubject variability. At corresponding motor phases, non-handicapped infants develop a wide variety of equilibrium reactions at higher levels and with more intrasubject variabil­ity than infants with Down syndrome. This project, however, did not record many of the abnormal and exaggerated pos­tures that many infants with Down syndrome use to maintain or assume certain milestone skills. For example, Lydic and Steele8 describe a typical pattern of coming to sit from the prone position in infants with Down syndrome as using exaggerated hip abduction and poor active trunk rotation. This prone-to-sitting pattern appears to be an effective com­pensation for the diminished postural reactions seen in infants with Down syndrome. These abnormal patterns may block or delay further the development of postural reactions and the development of a normal pattern of motor milestone progress.8

In general, the data provide strong evidence that postural reactions are delayed in infants with Down syndrome. The delays reinforce the argument that postural reactions are not related strongly to age but develop in close correspondence to motor milestones. The data also support the hypothesis that the levels of postural reaction development nearly are similar at corresponding phases of motor milestone development for both infant groups.

The data in this paper have both theoretical and practical implications. The results are consistent with the premise that postural reactions are strongly associated with motor mile­stone development. The confirmation of this relationship in infants with Down syndrome provides further support for the generalization of this premise across both normal and delayed motor development. Clinically, an important implication can be mentioned. Although the evidence of an association be­tween postural reactions and motor milestones does not imply strictly that remediation of one behavior will augment the other, the findings are consistent with that notion. Further­more, the delays in postural reactions are consistent with the delays observed in motor milestones in infants with Down syndrome. This suggests that educational and therapeutic intervention programs designed to promote the stimulation of postural reactions may have considerable impact in altering the severe motor delays in this population.

The cross-sectional nature and the small available sample size of the infants with Down syndrome impose caution on the interpretations and implications of this study. Verification of the results presented here should be attempted with longi­tudinal data. Longitudinal data could help establish temporal

relationships between postural reactions and developmental milestones. An increased sample size of both infant groups would improve the power of detecting group differences in correlations of postural reactions with motor milestones and in mean postural reactions. Further studies examining the descriptive data of the sequential nature of postural reactions and their profiles with motor milestones are needed.

In this study, the postural reactions are defined as those behaviors that can be elicited in a formal testing session. Although this procedure is designed to provide a standardized format for infant behavioral observation, not all reactions may be assessed accurately in such a procedure. For example, equilibrium reactions in the supine and prone positions are observed rarely in the test situation, yet they may still be present (albeit difficult to elicit). The results of this study should be interpreted with the knowledge that postural reac­tions are recorded only for the behaviors noted in the formal testing situation with examiner handling.

Although numerous clinical reports are available that dis­cuss possible underlying mechanisms related to motor mile­stone development, few reports have attempted to validate empirically these associations. This study has attempted to validate the notion that postural reactions are related closely to motor milestone development in a normal and a delayed motor development sample of infants. Future research en­deavors examining the underlying movement components of milestone behaviors will promote more accurate conceptual­izations of motor development in nonhandicapped infants and infants with Down syndrome.

CONCLUSIONS

The development of postural reactions was examined in a group of infants with Down syndrome and in a group of nonhandicapped infants. A series of hypotheses was tested to determine if the fundamental associations between postural reactions and motor milestones in infants with Down syn­drome were similar to nonhandicapped infants.

The overall results suggest that the interrelationship be­tween postural reactions and motor milestones is similar in both nonhandicapped infants and those with Down syn­drome. Postural reactions are related more closely to age in nonhandicapped infants than in infants with Down syn­drome. The delays in postural reaction development are seen in conjunction with the delays in motor milestone achieve­ment in infants with Down syndrome. These results are consistent with current rationales shaping the educational and therapeutic intervention techniques for infants with Down syndrome. These intervention techniques attempt to promote motor milestone progress in infants with Down syndrome by following a model of normal motor development. This model suggests that postural reactions are underlying components for the expression of motor milestones.

Acknowledgments. I express my appreciation to Carol Gray and Susan Harris for their support and guidance in this project. Special thanks to Wendy Tada, Kathy Washington, and Sharon Diamond for their data collection efforts. The project was supported by the Washington State Association for Retarded Citizens.

Volume 66 / Number 1, January 1986 21

REFERENCES

1. Carr J: Mental and motor development in young mongol children. J Ment Defic Res 14:205-220, 1970

2. Hanson MJ: A Longitudinal, Descriptive Study of the Behaviors of Down's Syndrome Infants in an Early Intervention Program. Eugene, OR, University of Oregon Books, 1981, pp 1-49

3. Kopp CB: Perspectives on infant motor development. In Bornstein MH, Kessen W (eds): Psychological Development from Infancy: Image to Inten­tion. Hillsdale, NJ, Lawrence Erlbawn Associates Inc, 1979, pp 9-35

4. Bobath B: Abnormal Postural Reflex Activity Caused by Brain Lesions, ed 2. London, England, Heinemann Medical Books, 1971

5. Milani-Comparetti A, Gidoni EA: Routine developmental examination in normal and retarded children. Dev Med Child Neurol 9:631-638, 1967

6. Barnes MR, Crutchfield CA, Heriza CB: The Neurophysiological Basis of Patient Treatment: Reflexes in Motor Development. Morgantown, WV, Stokesville Publishing Co, 1978

7. Chandler L: Gross and fine motor development. In Cohen MA, Gross PJ (eds): The Developmental Resource—Behavioral Sequences for Assess­ment and Program Planning. New York, NY, Grune & Stratton Inc, 1979, vol 1, pp 119-153

8. Lydic JS, Steele C: Assessment of the quality of sitting and gait patterns in children with Down's syndrome. Phys Ther 59:1489-1494, 1979

9. Griffiths Ml: Development of children with Down's syndrome. Physiother­apy 62:11-15, 1976

10. Connolly BH, Morgan S, Russell FF: Evaluation of children with Down syndrome who participated in an early intervention program. Phys Ther 64:1515-1519,1984

11. Harris SR: Effects of neurodevelopmental therapy on motor performance of infants with Down syndrome. Dev Med Child Neurol 23:477-483, 1981

12. Hayden AH, Haring NG: Early intervention for high risk infants and young children: Programs for Down's syndrome children. In Tjossem TD (ed):

Intervention Strategies for High Risk Infants and Young Children. Balti­more, MD, University Park Press, 1976

13. Cowie VA: A Study of the Early Development of Mongols. Oxford, England, Pergamon Press Ltd, 1970

14. Molnar G: Motor deficit of retarded infants and young children. Arch Phys Med Rehabil 55:393-398, 1974

15. Molnar G: Analysis of motor disorders in retarded infants and young children. Am J Ment Defic 18:212-222, 1976

16. Horner TM: Test-retest and home-clinic characteristics of the Bayley Scales of Infant Development in nine- and fifteen-month-old infants. Child Dev 51:751-758,1980

17. Chandler L, Andrews M, Swanson M: Movement Assessment of Infants. Rolling Bay, WA, Movement Assessment of Infants, 1980

18. Bayley N: The Bayley Scales of Infant Development. New York, NY, Psychological Corporation, 1969

19. Eipper DS, Azen SP: A comparison of two developmental instruments in evaluating children with Down's syndrome. Phys Ther 58:1066-1069, 1978

20. Werner E, Bayley N: The reliability of Bayley's revised scale of mental and motor development during the first year of life. Child Dev 37:39-50, 1966

21. Winer BJ: Statistical Principles in Experimental Design, ed 2. New York, NY, McGraw-Hill Book Co, 1971, pp 185-204

22. Ferguson GA: Statistical Analysis in Psychology and Education, rev ed 5. New York, NY, McGraw-Hill Book Co, 1981, pp 194-196

23. Karpman MB: The Johnson-Neyman technique using SPSS or BMDP. Educational and Psychological Measurement 43:137-147, 1983

24. Harris SR: Down syndrome. In Campbell SK (ed): Pediatric Neurologic Physical Therapy. New York, NY, Churchill Livingstone Inc, 1984

22 PHYSICAL THERAPY