The importance of early screening, assessment, and intervention planning is crucial in all educational and health-related fields, including occupational therapy. This focus on prevention and early intervention affords the opportunity to identify and intervene as soon as possible to ensure optimal lifelong occupational performance. Early intervention (EI) services provided to children from birth to age 3 yr, especially for those who are at risk for or identified with a developmental delay, usually consist of educational, therapeutic, and social services aimed at minimizing negative and supporting positive effects on development (American Occupational Therapy Association [AOTA], 2010). However, even with increasing awareness of and emphasis on EI services, many children, especially those with more subtle conditions or from families with limited resources, are not identified with developmental or learning concerns until they become disruptive to the family, teacher, or other caregivers. Included among the more subtle problems that can occur in early development are irregularities in sensory perception, sensory regulation, and motor skills, collectively known as sensory integration dysfunction (Ayres, 2005). Many types of sensory integration difficulties are associated with problems in attention, learning, and behavior (Mailloux et al., 2011; Parham & Mailloux, 2010) and are often observed in people with additional diagnoses such as autism and attention deficit hyperactivity disorder (Baranek, Roberts, et al., 2008; Baranek, Wakeford, & David, 2008; Mailloux & Smith Roley, 2010; Miller et al., 1999). Because children with underlying sensory and motor irregularities are not easy to identify without specific testing or specialized professional training (Bodison & Mailloux, 2006), these challenges can easily be missed or misinterpreted.

Despite increasing research showing the benefits of early identification and intervention (Karoly, Kilburn, & Cannon, 2009), economic and budgetary challenges in the United States, as well as in other countries, have limited the programs and services for families of children at risk for developmental problems (Johnson, Oliff, & Williams, 2011), especially those without a clear medical diagnosis. Identifying methods for early assessment of sensory integration functions has been a challenge for occupational therapy practitioners and researchers because of the limited availability of standardized assessments in this area for infants and toddlers (Williamson, Anzalone, & Hanft, 2000). Therefore, the purpose of this study was to explore the use of the postrotary nystagmus (PRN) test for infants and young children ages 2 mo–47 mo. The PRN test, currently standardized on children ages 4.0–8.11 yr (Ayres, 2004), provides a noninvasive, quick, and easy-to-use assessment of the vestibular–ocular reflex (VOR) that provides insight into sensory and motor irregularities.

Mulligan (2010)  noted that few easily administered and standardized evaluation tools are available for assessing vestibular function in children overall, and almost none are available for infants and toddlers. Therefore, therapists rely on clinical observations of a child’s gross motor performance to make inferences about vestibular functioning. Given the importance of the vestibular system in early development (Angelaki, Klier, & Snyder, 2009), an attempt to assess its function at as early an age as possible is prudent. One of the most common ways to evaluate the integrity of the vestibular system and its components is to test the VOR, a three-neuron arc reflex from the vestibular nucleus to the oculomotor nuclei that was first described by Lorente de Nó in 1933 (Cohen & Raphan, 2004; Gonçalves, Felipe, & Lima, 2008; Highstein, Popper, & Fay, 2004). Rotational and translational head movements stimulate the VOR, which responds with an eye movement that is opposite to head movement direction (Amin & Konrad, 2008; Highstein et al., 2004). The three common tests used to measure the VOR are electronystagmography (Barber & Stockwell, 1980; Ganança, Caovilla, & Ganança, 2010), caloric testing (Zajonc & Roland, 2005), and the rotational chair (Fife et al., 2000). The rotational chair paradigm was introduced as early as 1907 by the Austro-Hungarian otologist and Nobel Prize winner Robert Barany (Kingma, 2009) and served as a basis for the Southern California Postrotary Nystagmus Tests (SCPNT) as originally published by Ayres (1975) . This test, originally published individually and standardized for children ages 5–9 yr, is now 1 of 17 tests included as part of the Sensory Integration and Praxis Tests (SIPT; Ayres, 2004), which is designed to assess sensory integration and praxis functions in children ages 4–8.11 yr (Ayres, 2004).

The PRN test evaluates the integrity of the VOR after rotation of the head in space (Ayres, 2004). This reflex reflects the role of the vestibular system in ocular mechanisms. One important function resulting from this relationship is the vestibular system’s role in stabilizing the eye muscles in response to head movements, which allows maintenance of a stable visual field when the head is moving. The VOR allows for the eyes to turn in the opposite direction of head movement to preserve visual field stability and object fixation (Lundy-Ekman, 2008). This important neurological relationship assists in maintaining orientation in space while navigating through the world and in daily tasks such as eye tracking when looking up and down, for example, in an activity such as copying something from a whiteboard to a notebook.

Critical to human function, the vestibular sensory system (Highstein et al., 2004) includes separate sensory organs that perceive the basic attributes of movement (i.e., angular and linear acceleration, angular and linear velocity, and position with regard to gravity). Evolving and continued research has led to the development of new tools, techniques, and concepts for revealing more about the role of the vestibular system’s importance in the development and functions of balance, awareness of head position in relation to gravity, ocular stabilization (maintenance of a stable visual field when the head moves), postural adjustments, and autonomic nervous system functions (Highstein et al., 2004; Lundy-Ekman, 2008).

Successful active engagement in most daily life activities requires intact vestibular system functions such as coordination of head and eye movements, adequate postural control, balance and postural awareness, and coordination of the two sides of the body (Ayres, 1972, 2005; Parham & Mailloux, 2010). These body functions provide the basis for coordinated movement and allow children to explore, play, develop fine and gross motor skills, and move efficiently in their multiple environments in their daily lives (Mulligan, 2010). Therefore, vestibular functioning, measured in part through the assessment of PRN, is important in the evaluation of children who are at risk for neurodevelopmental, learning, or behavioral difficulties (Mulligan, 2010).

Both shortened and prolonged duration of PRN, in comparison with normative data, are associated with learning and developmental concerns (Ayres, 2004). Studies in the United States, Australia, and South America have suggested that at least 50% of all children with learning or language problems show signs of vestibular dysfunction, indicated by an atypically shortened duration of PRN (Ayres, 2005). In a case report study, Ayres and Mailloux (1981)  found a relationship between expressive language and vestibular processing and, later, another pattern characterized by prolonged-duration PRN and language-related test scores was also reported by Ayres (2004) . Ayres (1965, 1966a, 1966b, 1969, 1972, 1977, 2004) repeatedly demonstrated patterns of difficulties in postural, ocular, bilateral integration, and sequencing functions that she hypothesized were associated with vestibular dysfunction, and current research by Mailloux et al. (2011)  verified that shortened-duration PRN is, in fact, associated with a pattern characterized by bilateral integration and sequencing. Vestibular processing problems are also commonly reported in children with other conditions, such as developmental coordination disorder (Mulligan, 2010; Przysucha, Taylor, & Weber, 2008), autism (Mailloux, 2001), attention disorders (Mulligan, 1996), hearing impairment (Valente & McCaslin, 2011), and learning disorders (Ayres, 1978; Mulligan, 2010). These findings suggest that healthy vestibular function is important in meeting the demands of schoolwork and other life tasks and that it is vital to assess vestibular functions in children because of the incidence of problems in this sensory system among children who are struggling in various aspects of childhood occupations (Ayres, 2005).

We conducted two clinical studies to explore the adaptation and use of the PRN test for children younger than age 4 yr as a possible screening tool for early signs of vestibular-based sensory dysfunction. This measurement may provide one way in which to address the need for early and easy measures that identify salient developmental concerns for the purpose of ensuring EI and prevention of later disabilities. This project was based on retrospective de-identified chart review and, as such, received University of Southern California institutional review board approval as an exempt study.

To explore the feasibility of adapting the PRN test for children younger than age 4 yr, it was first necessary to determine whether holding the child during rotation would have an effect on the PRN reflex. This positioning change is critical for testing younger children who are not able to sit and maintain the correct position. Specifically, this project aimed to examine whether there is a difference in duration of PRN as a function of whether the child sits independently on the PRN board or whether the child is held on a larger board by the child’s parent or therapist. The adaptation includes using a larger rotary board (the standardized board is approximately 1 ft × 1 ft square, whereas the adapted board is approximately 2 ft × 2 ft square) with an adult sitting on the board and holding the child in place. This adapted administration of the PRN test was previously used by Ayres in clinical practice, as well as by subsequent practitioners, because some children, even within the standardized age group, were unable to remain in the required position for testing. However, the potential effect of this change on PRN duration has not been formally studied.

Study participants were 37 children ages 4–9 yr. The group tested included children who had diagnoses such as autism and learning disabilities as well as those who were receiving occupational therapy. Also tested were their typically developing peers and siblings who accompanied them to therapy sessions.

Duration of PRN was tested in both standard (C1) and adult-held (C2) administrations. Each child was tested in both conditions (C1 and C2) with a random assignment to C1 or C2 administered first.

The differences between C1 and C2 PRN durations were computed using a two-tailed paired Student’s t test. Significant differences between C1 and C2 were determined at p < .05.

A two-tailed t test indicated no significant differences between the duration of PRN in the standard independent position (C1) and the adult-held position (C2) for both clockwise and counterclockwise rotations (see Table 1).

The results of this study indicate that there is not a significant difference in duration of PRN as a function of whether the child is sitting independently or is held in the lap of an adult. This finding indicates that either position may be used to assess the duration of the PRN reflex and supports clinical practice that has used this method to assess PRN in children unable to maintain the standard, independent sitting position. In addition, this adaptation may result in more reliable results for children who have difficulty following standard test procedures, such as those with poor postural control or those who have difficulty following the directions for maintaining head position.

This study examined the three following questions:

Study participants were 44 typically developing children younger than age 4 yr (23 girls and 21 boys, ages 2–47 mo). Of the 44 children, 15 were participating in an Early Head Start program for typically developing, low-income children and families. Testing PRN in this group was part of regular program procedures exploring potential assessments for pretest and posttest measures to monitor and track the children’s development, risk for delays, and response to participation in the program. Scores for these children were identified by reviewing Early Head Start program records to identify children who had been administered the PRN as part of routine practice. The remaining 29 children were typically developing infants and toddlers who were participating in community programs or were children of therapists or their friends who participated in practice sessions conducted in preparation for developing clinical application of this test for younger children. Developmental delays were determined to be an exclusion criterion; however, none of the records for the children indicated any concerns or delays and, as such, we made no exclusions for the records reviewed and included in this exempt study.

The test was adapted with the children sitting on the lap of an adult and being held in the proper position on a larger board. The children’s heads were positioned by the examiner in 30° of forward flexion as indicated in the standardized administration for the PRN test. The child’s head position was maintained by the person holding the child. The duration of the children’s PRN was recorded after two separate administrations of 10 rotations in 20 s (the first 10 rotations in a counterclockwise direction in 20 s and the second 10 rotations in a clockwise direction in 20 s) in accordance with the standardized procedures of PRN administration (Ayres, 2004). At the conclusion of each rotation, the person holding the child lifted the child’s head to a neutral position while the examiner observed the PRN response.

For all participants, the procedure was explained to each parent or caregiver, all of whom expressed interest in and willingness to have their child tested. The parents and caregivers who felt they could tolerate being rotated on the board while holding their child did so (n = 14), and for those who did not feel they would be able to tolerate the movement themselves, an occupational therapist held the child during the test (n = 30). The children’s and parents’ responses to the testing were noted during and after each test administration.

We used the SCPNT (Ayres, 1975), the precursor to the PRN test now published as part of the SIPT, because it reports mean group scores in number of seconds of duration for this reflex (with standard deviations). The tests of the SIPT are computer scored, and the means and standard deviations for the tests were not available in the manual (Ayres, 2004). Although the SIPT and the tests that were precursors to it were standardized on children ages 4.0–8.11 yr, the SCPNT was standardized on children 5.0 yr–8.11 yr. All the other skill-based tests of the SIPT show an age progression, with older children performing better than younger children. However, the PRN test as a reflex measure shows a flat age trend such that no variation in the reflex response is seen in the normative data for children ages 4.0 yr–8.11 yr. Similarly, the mean number of seconds of duration of PRN (and standard deviations) reported in the SCPNT manual were very similar across the 5–9 yr age range and also between boys and girls. Therefore, we determined that comparing the PRN scores of the youngest norm group in the SCPNT manual (Ayres, 1975)—that is, 44 children age 5 yr—with those of the current project sample of 44 children ages 2–37 mo would be a reasonable process for the initial comparison of infants and toddlers with older children on this measure.

The first two research questions regarding the feasibility of using the PRN test with younger children were answered by reviewing the children’s and parents’ responses to the testing. The third question was addressed through comparison of infants’ and toddlers’ responses on the PRN test with the available published normative means and standard deviations for older children. We analyzed PRN duration, recorded after rotation in both directions (clockwise and counterclockwise), using Student’s t test for comparing two independent samples, that is, a comparison of 44 typically developing infants and toddlers (ages 2–47 mo) with the 44 children (age 5 yr) whose group mean scores were published in the SCPNT manual (Ayres, 1975). We explored whether a correlation existed between infant and toddler age and duration of PRN using Pearson’s correlation coefficient and examined age trends using linear regression.

Once we determined that holding a child in position did not significantly alter the PRN response, it was necessary to determine whether it would be feasible to hold infants and toddlers and view the PRN reflex. Initially, the adults sat in an office chair to determine whether holding an infant in their laps on this type of chair would be a comfortable and stable way to administer the test. However, the reduced adult hip flexion when the adult was seated in a chair compared with when the adult was seated on a rotational board resulted in a less defined lap, which made holding the infant or toddler and keeping the child’s head in 30° of forward flexion while being rotated more difficult. Therefore, we determined that the rotational board was the preferred choice. Another consideration was whether the adults would be able to tolerate the movement themselves. Some of the parents and caregivers did have difficulty with this because of the common intolerance of the rotatory stimuli among adults, and a few parents did not find sitting on a board very comfortable for their physical size. However, some parents (approximately 30%) were able to tolerate the rotations. The alternative solution of having a therapist, other adult, or older sibling who could tolerate movement hold the child remedied this issue for those parents who could not.

Although a few children (5 of 44) cried while being positioned on the board, all but 1 child stopped crying when the rotation of the board began. The adults were able to hold the child in place and lift the child’s head on stopping so the reflex could be observed. The children did not close their eyes; thus, administration of the PRN test and observation of the reflex using this method was determined to be feasible.

An initial concern was whether the parents and caregivers would understand the usefulness or meaning of the PRN test and its relevance to their child’s development. However, the parents appeared to make easy comparisons between the explanation of the vestibular system and its role in early development and their children’s behavior in daily life activities. Comments such as “Yes, he really likes movement” were frequent, with an occasional comment such as “She is not very comfortable with swings.” All the parents expressed that they were comfortable with the procedure, which took <1 min, and seemed to feel it was safe for their children, even when some of the children cried while being positioned. The children were able to tolerate the test and had no adverse effects such as vomiting, nausea, or change in state of alertness. The most significant issue was initial hesitation to being held by one of the test administrators instead of their caregiver. However, in almost every instance, the children stopped crying once the rotation was initiated; in most cases, the children appeared to enjoy the movement. Thus, the procedure appeared acceptable to both parents and young children.

We found no significant differences in PRN duration between infants and toddlers in comparison with the SCPNT norms for either boys or girls (boys+girls: counterclockwise PRN, p = .78; clockwise PRN, p = .38; counterclockwise + clockwise PRN, p = .52) (Table 2). Within our sample, we also found no significant differences in PRN duration between boys and girls (counterclockwise PRN, p = .35; clockwise PRN, p = .37; counterclockwise + clockwise PRN, p = .29). However, we found a low but significant correlation, r(42) = −.34, p = .02, and a negative relationship between child’s age in months and combined duration of PRN, such that for every month increase in age between 2 and 47 mo, combined duration of PRN decreased by approximately 0.14 s (β = −0.14, p = .02; see Figure 1).

We found significant differences between the ≤17-mo age group and the 18- to 47-mo age group (counterclockwise + clockwise, p = .03) and also a natural age cutoff to separate walkers from non- or early walkers (Capute, Shapiro, Palmer, Ross, & Wachtel, 1985; Group & de Onis, 2006). When we separated these two age groups, we found weaker correlations, r(17) = −.20, p = .41, and r(23) = −.13, p = .54, respectively, and the linear relationship between child’s age and duration of PRN almost disappeared (β= −0.16, p = .41, and β = −0.08, p = .53; respectively; see Figure 2). However, larger studies would need to confirm this finding.

Determination of validity of adapting the PRN test to accommodate children who cannot sit independently was of interest because this method of altering the original test has been used clinically for many years. Affirming that this adaptation has no effect on PRN duration is a useful validation of a common clinical practice. The comparison of PRN between children ages 2–47 mo with older norms published in the SCPNT manual (Ayres, 1975) was of interest on several accounts. The fact that we found no significant differences between the younger children and the published norms in the SCPNT for 5-yr-olds suggests that the current SIPT norms can be used as a comparative sample for children who are younger. This information is useful in light of the current absence of normative data on PRN for children younger than age 4 yr. Closer inspection of the data revealing some evidence of a relationship between child’s age and nystagmus duration in the 2- to 47-mo age range was also noteworthy. It is possible that slightly higher PRN scores are found among young infants (0–17 mo) than among children ages 18 mo or older, although the sample size in this study was too small to make the comparison. Although both shortened and prolonged PRN scores (i.e., less than or greater than 1.0 standard deviation from the mean) have been associated with patterns of sensory integration problems in children older than age 4 yr, interpretation of these findings in young children warrants further study. Validity studies examining the predictive and discriminatory ability of this test in infants and toddlers will be particularly useful. However, the overall results of this study do suggest that this test can be used with children younger than age 4 with valid comparison with the current published norms for the 4.0- to 8.11-yr age group. With children of any age, the PRN test is meant to be considered within the context of comprehensive assessment and is not used as a singular determinant of function or dysfunction.

Because early assessment is a critical first step toward EI, identification of sensitive, reliable, and valid tools is an important consideration for pediatric occupational therapists serving young children and their families. The PRN test has been a simple, quick, and low-cost means of testing one aspect of vestibular functioning as a part of occupational therapy assessments for children for nearly 4 decades. Occupational therapists with advanced training in evaluation of and intervention for sensory integrative concerns have grappled with the means to provide similar comprehensive assessments for younger children. The results of this project suggest the following implications for occupational therapy practice:

Quick and easy early screening tools predictive of developmental concerns are needed by health care providers to ensure early identification and intervention services for families of children with developmental concerns. Observations such as head lag have been shown to be salient in identifying developmental concerns, including a later diagnosis of autism (Flanagan, Landa, Bhat, & Bauman, 2012). Supplementing these observations with an objective measure of neurological integrity, such as a labyrinthine response as measured by the PRN test, will inform the understanding of sensory integration in early development by occupational therapy practitioners who work with young children. Given the importance of the vestibular system in early development, consideration of the PRN test for infants and toddlers seems well worth further study. Studies that examine the ability of the PRN test to discriminate typical from atypical development and occupational performance in young children will be especially important. Examination of the possibility of predicting future performance from results of this early reflex testing will also be useful. Because both shortened and prolonged duration of PRN are associated with different types of concerns in children ages 4–8 yr (Ayres, 2004; Mailloux et al., 2011), it will also be interesting to explore ways of examining whether similar patterns can be identified at younger ages.

We thank the following people for their assistance with de-identified chart review and project coordination: Angee Dowdy, Gina Coleman, Joaquim Faias, Paulo Fernandes, Shay McAtee, Aja Roley, Mary Singer, Kerstin Starzer, and Maria Joao Trigueiro. Portions of this project were completed in partial fulfillment of advanced study requirements for Marco Leão at the Escola Superior de Tecnologia da Saude do Porto, Porto, Portugal.