Verification and Clarification of Patterns of Sensory Integrative Dysfunction

Review of the Literature

The early formulation of ASI included the development of a set of measures for the systematic investigation of patterns of sensory integrative function and dysfunction. The Sensory Integration and Praxis Tests (SIPT; Ayres 1989), and their earlier version, the Southern California Sensory Integration Tests (Ayres, 1972a), as well as measures in areas such as tactile defensiveness, hyperactivity, distractibility, and language, were included as variables in multiple studies. A series of factor and, later, cluster analysis studies conducted between 1963 and 1989 with samples of typically developing children, children with learning disorders, and children with suspected sensory integration problems yielded fairly consistent patterns of sensory integrative dysfunction:

Other patterns, such as those involving language functions or associations between two or more areas of sensory perception, were found in some studies (Ayres, 1969, 1972b, 1977, 1989).

The patterns that emerged supported an understanding of disorders in sensory integration as multidimensional. Clarity of the associations or subtypes developed as the measures of sensory integration functions became greater over time. Other researchers have also contributed to the understanding of patterns of sensory integration dysfunction. Lai, Fisher, Magalhaes, and Bundy (1996) proposed that praxis might be a possible underlying construct for both bilateral integration and sequencing deficits as well as for somatodyspraxia. Mulligan (1998) conducted a large-scale factor analysis on a sample of >10,000 children and found patterns similar to those found in Ayres’ earlier studies (1965, 1966a, 1966b, 1969, 1972b, 1977, 1989), including problems with bilateral integration and sequencing, somatosensory deficits, visual–perceptual deficits, and dyspraxia. Her study also found that although subtypes emerged from the data, the patterns were highly correlated with one another. Mulligan suggested that they had a common underlying construct and reinforced the concept of integration of multiple sensory inputs from multiple sensory systems as being important for neurological functioning. The factor loadings found in the studies by Mulligan (1998) and Ayres (1965, 1966a, 1966b, 1969, 1972b, 1977, 1989) also provided insight into specific relationships among functions, such as the role of somatosensory perception in relation to motor planning and the relationships between vestibular, postural–ocular, and bilateral functions.

On the basis of SIPT scores, Ayres (1989) identified six cluster patterns that emerged from a mixed sample of typically developing children and children with sensory integration problems. Three of the six cluster groupings were thought to represent patterns not specifically associated with sensory integration dysfunction. Two of the groupings were characterized by sensory integration functions in the average range (low-average and high-average sensory integration and praxis). The third pattern involved language processing (dyspraxia on verbal command) that Ayres (1989, p. 182) suggested might reflect left cerebral hemisphere functions rather than sensory integration dysfunction.

The three clusters characterized by recognizable patterns of sensory integration were Bilateral Integration and Sequencing, Visuo-Somato Dyspraxia, and Generalized Sensory Integration Dysfunction (Ayres, 1989). Mulligan’s (2000) cluster analysis of a large data set of children tested with the SIPT revealed that children fell into one of five cluster groupings: average sensory integration and praxis, moderate sensory integration dysfunction, severe sensory integration dysfunction and dyspraxia, dyspraxia, and low-average bilateral integration and sequencing. The results of this study suggested that in addition to the type of dysfunction, severity of dysfunction was an important consideration.

Other researchers (Dunn 1997, 1999; Parham & Ecker, 2007; Miller-Kuhanek, Glennon, & Henry, 2007) have focused more attention on patterns related to responses to sensory experiences than have previous researchers. Some of these studies revealed strong associations between various sensory patterns, again reinforcing the interrelationships of sensory integration constructs.

The body of research by Ayres (1965, 1966a, 1966b, 1969, 1971, 1972b, 1977, 1989) and Mulligan (1996, 1998, 2000), showing consistent patterns of sensory integration dysfunction and the strong ability of the SIPT and its precursors to discriminate between typically developing children and children with learning, behavior, attention, or suspected sensory integration deficits has established the tests’ validity. However, previous studies examining patterns of sensory integration dysfunction have had some limitations or weaknesses. Some of the earlier studies included variables that were measured by observation or other nonstandardized procedures. Later, in more rigorous studies, certain measures, especially those related to overresponsiveness to sensation (e.g., gravitational insecurity, tactile defensiveness), were omitted in Ayres’ factor analysis (1977, 1989). However, over- and underresponsiveness to sensation have always been recognized in practice (Parham & Mailloux, 2010) and have been studied with both parent questionnaires and physiological measures (Brett-Green, Miller, Schoen, & Nielson, 2010; Dunn, 1999; Parham & Ecker, 2007; McIntosh, Miller, Shyu, & Hagerman, 1999; Parush, Sohmer, Steinberg & Kaitz, 1997; Schaaf, Miller, Seawell, & O’Keefe, 2003; Reynolds & Lane, 2009). Mulligan’s (1998) work used a convenience sample from a stored database, and knowledge about the characteristics of the children included in the analysis was limited. Moreover, Mulligan’s research did not include a measure of over- or underresponsiveness to sensory experiences. In a an evidence review of subtypes of problems in sensory integration, Davies and Tucker (2010) identified the need for “comprehensive assessment of sensory function and sensory-based motor performance that includes sensory perception, discrimination, modulation, and praxis in a single study” (p. 399).

In this study, we aimed to improve on these methodological challenges by examining data from a large clinical sample in which (1) the test examiners were known to have had extensive training in the use of the tests and in ASI theory and its application in occupational therapy practice, (2) the background and presenting problems of the children were known, and (3) parent report measures regarding overresponsiveness to sensory experiences and the presence or absence of difficulties with attention were available in addition to SIPT scores.

Another challenge faced by researchers who have conducted studies examining patterns of dysfunction using the SIPT concerned the management of data from the Postrotary Nystagmus (PRN) test. PRN, 1 of the 17 SIPT components, is a measure of vestibular function. PRN is unique in that it is the only SIPT measure in which both a high and a low Z score are indicative of dysfunction (for all other tests, only low scores are indicative of dysfunction). Moreover, a low PRN score is believed to be associated with a type of problem different from a high PRN score. This distribution of high and low scores in children with sensory integrative challenges is suspected to have diminished this measure’s clarity in relation to the other measures (such as bilateral coordination and balance), particularly when mean scores and correlation statistics have been applied.

To verify and clarify patterns of sensory integrative dysfunction, with consideration of previous research findings and challenges regarding variables, we reexamined associations identified in previous studies by including comprehensive measures of sensory integration functions (SIPT) and items from a parent report of responses to sensory experiences and parent reports of problems with attention. In addition, we explored methods for managing the PRN test, which vary from those for other sensory integration tests because both high and low scores are reflective of dysfunction.

Hypotheses

We hypothesized that the population under study would demonstrate patterns of sensory integration dysfunction similar to those found in previous studies. The specific hypotheses were that the following associations and patterns would emerge:

Method

Results

The sample included 193 boys (70.7%) and 80 girls (29.3%) ranging in age from 4 to 9. Most of the children (78.1%) had no diagnosis on file and were referred for evaluation because of occupational performance problems suspected to be the result of underlying sensory integration dysfunction. Diagnoses or conditions that were reported included sensory integration disorder (n = 29; 10.6%), attention deficit disorders (n = 20; 7.3%), and speech–language or learning disorders (n = 17; 6.2%).

The results of the factor analysis yielded six factors with eigenvalues >1. However, because two factors included only one variable, we felt that a four-factor solution best fit the data, accounting for 48% of the variance. The four-factor solution with accompanying factor loadings is shown in Figure 1. All factors in this study were significantly related to one another (Table 3).

OPEN IN VIEWER

Figure 1.

Primary loadings for four-factor solution.

The first factor, named Visuodyspraxia and Somatodyspraxia, is characterized by high loadings on the tests of visual perception and visuopraxis functions, including Design Copy, Construction Praxis, and Space Visualization. These three tests were all previously associated with factors generally named Visuopraxis or Visuodyspraxia (Ayres, 1965, 1966a, 1966b, 1969, 1972b, 1977, 1989). The presence of Postural Praxis, Praxis on Verbal Command, and Manual Form Perception on this factor suggests a mild association with a pattern of Somatodyspraxia. Postural Praxis (formerly called Imitation of Postures) consistently loaded on factors characterized by an association between tactile perception and praxis tests in previous studies (Ayres, 1965, 1966a, 1966b, 1969, 1972b, 1977, 1989). In addition, Praxis on Verbal Command does not have a visual component, and the Manual Form Perception Test involves tactile processing and visual perception. Therefore, in relation to the patterns that were hypothesized, this factor represents Visuodyspraxia with some secondary aspects of Somatodyspraxia.

The second factor, Vestibular and Proprioceptive Bilateral Integration and Sequencing, was consistent with the second hypothesized pattern. Many previous studies demonstrated a bilateral integration and sequencing pattern, associating bilateral integration with vestibular and proprioceptive functions. This pattern was characterized by high loadings on Oral Praxis, Standing and Walking Balance, Postrotary Nystagmus, Bilateral Motor Coordination, Graphesthesia, Motor Accuracy, and Sequencing Praxis. Of the tests loading on this factor, Oral Praxis, Standing and Walking Balance, Bilateral Motor Coordination, Graphesthesia, and Sequencing Praxis all loaded on the Bilateral and Sequencing factor reported in the SIPT manual (Ayres, 1989). All of these tests were noted to involve bilateral and sequential actions. In this study, the addition of Motor Accuracy to this group is hypothesized to reflect the vestibular and bilateral functions needed for the coordination of eye, head, and hand movements; postural adjustments; and crossing of body midline needed in this test. Another significant finding in this study was the presence of the PRN test on this factor, which had been hypothesized as an important measure of this pattern in the past but had not loaded with similar factors in previous work. It is likely that the distribution of scores in this sample (relatively few children with abnormally high PRN scores) contributed to this finding. Also noteworthy was that the test of Kinesthesia, as would be expected, loaded on this factor, although modestly. The presence of both Standing and Walking Balance and Kinesthesia led to adding the term “proprioceptive” to the name of this factor.

The third factor, Tactile and Visual Discrimination, was consistent with the third hypothesis and was characterized by two tests of tactile discrimination and one test of visual discrimination. This pattern is similar to sensory system factors identified in previous work (Ayres, 1989; Mulligan 1998), which revealed associations between tests of tactile function and tests measuring another sensory system, without the accompanying loadings of praxis, bilateral, or postural measures.

The fourth factor, Tactile Defensiveness and Attention Problems, was consistent with the hypothesized pattern and with the relationship between inattention and hyperactivity and tactile defensiveness described in Ayres’ (1965, 1966a, 1966b, 1969, 1972b) early studies.

Discussion

The importance of this study lies in the verification of patterns of sensory integrative dysfunction that have been studied for nearly 50 years. Ayres found factors related to praxis that she named Apraxia and Dyspraxia (later, Visuodyspraxia and Somatodyspraxia). In the 1989 data from the SIPT manual, these independent patterns were notably stronger (i.e., first-order factors) in the samples that had large numbers of typically developing children. A cluster analysis of the same data revealed a Visuodyspraxia and Somatodyspraxia cluster group that included more children with learning disabilities or sensory integration issues (n = 22) than children from the normative group (n =13). It is possible that visuo- and somato-based praxis functions overlap more in samples of children with identified problems, as was the case with this study’s sample.

The second factor finding that is especially important for both therapists and researchers is the demonstrated link between the PRN test and the low scores on tests of vestibular and bilateral functions. Although the relationship of hyporesponsiveness to vestibular input, evidenced by low PRN scores, has been hypothesized, this study marks the first time this relationship has been supported by empirical data. The small percentage of cases with prolonged nystagmus (8.4% [n = 23] vs. 35.9% [n = 98] with shortened-duration nystagmus) in this sample avoided the averaging effect that occurs when both high and low PRN scores are prevalent in a sample. Thus, these findings reinforce a long-held view of the significance of low PRN scores and validate Ayres' (1989) theoretical assumption that bilateral integration problems and other signs of vestibular inefficiencies are associated with a shortened duration of postrotary nystagmus. The added factor loading of the Motor Accuracy Test (which requires vestibular- and proprioception-related coordination of head, eye, and hand movements and crossing the body midline) and a moderate loading on Kinesthesia also reinforce the underlying commonality of vestibular and proprioceptive functions in this group. Although the hypothesis related to the association of low PRN with bilateral and sequencing measures was supported, an association between the measures of directionality, crossing midline, and laterality with this pattern was not found. The subscores used to measure these functions may not have had adequate biometric properties, or these measures may not reflect functions specific to this pattern.

The absence of the Dyspraxia on Verbal Command pattern (characterized by moderately high PRN and low Praxis on Verbal Command in previous studies) is an additional finding in this study that may be related to the nature of the sample. The Ayres (1989) standardization study that identified this pattern, as well as Mulligan (1998), most likely included more children with learning disabilities and speech–language deficits, a group more likely to have both prolonged PRN and speech–language problems than the sample in this study.

Finally, the hypothesized association between attention problems and heightened sensitivity to touch rather than to tactile discrimination and practic functions was supported by empirical data. At this time, fully exploring the hypothesis that other forms of sensory overresponsiveness (e.g., auditory defensiveness and gravitational insecurity) are associated with problems in attention was not possible because the number of items representing those variables was smaller and less specific to overresponsiveness than the items for tactile defensiveness. Continued clarification of the relationships between generalized sensory overresponsiveness and specific sensory responsiveness is still needed. Efforts to clarify underlying physiological functions associated with over- and underresponsiveness will also provide insight into the impact of these problems on behavior, learning, and participation (Reynolds & Lane, 2009). The significant relationship between the factors in this study and Ayres’ (1989) studies reinforces the long-held idea that the patterns revealed in these studies show specific associations among sensory, motor, and praxis; these associations reflect that sensory systems function in a synergistic, rather than isolated, way.

Identifying and understanding underlying problems that affect health, performance, and participation are important first steps for practice and research in occupational therapy. The results of this study confirm the presence of sensory integration patterns of dysfunction in children commonly referred for an occupational therapy evaluation. Knowledge of these patterns allows therapists to understand the nature of difficulties interfering with performance, contributes to more precise intervention planning, and provides a means by which to determine the interventions’ effectiveness. Using standardized assessments, such as the SIPT, in conjunction with caregiver reports will enhance participant selection in efficacy studies of occupational therapy applying a sensory integrative intervention approach.

Limitations

This study’s limitations include the subjectivity of factor-analytic studies, the subjectivity of the tools used to measure attention and tactile overresponsiveness, and the degree of representation in the sample. First, as with all factor analysis, the interpretation of the factors has a level of subjectivity (e.g., the decision to use the four-factor solution and the naming of the factors). However, given the breadth of previous studies on which to base decisions and the many researchers who collaborated throughout the interpretation process, subjectivity was minimized. Second, the methods for measuring attention and tactile overresponsiveness in this study do not have established reliability and validity data; however, other means for assessing these variables were not available in this retrospective sample. Finally, data were retrieved from only two private practices and may therefore have limitations in generalizability. However, the similarity with previous findings and hypothesized constructs reduces this potential concern.

Future Research

Future research should continue to identify and enhance our understanding of the underlying neurodevelopmental problems that affect health, performance, and participation. Measures for children younger and older than the range included in the SIPT normative sample are needed to determine whether these patterns of dysfunction hold true for these age groups and are also needed to conduct outcome studies. To conduct outcome studies, changes in sensory integration and praxis functions must correlate with needed and desired changes in health, performance, and participation of the child with identified sensory integration and praxis difficulties, the family, and community. Although this study confirmed and clarified the presence of various subtypes of sensory integrative dysfunction, current assessment tools are still limited in several areas, including aspects of vestibular and proprioceptive processing, postural ocular control, and ideational praxis. Validity studies are also needed to enhance our understanding of low- and high-duration PRN. Finally, future research that includes confirmatory factor analysis of the model with a new data set to determine whether the model holds true is suggested.

Conclusion

ASI is one of the most developed and distinctive frames of reference to emanate from the profession of occupational therapy. Occupational therapists with specialized knowledge of ASI have a unique understanding of the ways in which functions such as vestibular processing, tactile perception, and praxis contribute to daily life participation and success. Understanding the different patterns of sensory integration dysfunction allows occupational therapists to be better equipped to design, implement, and study intervention programs to alleviate challenges and, ultimately, to support occupational performance.