Neurobehavioral Development and Environmental Exposures: Measures for the National Children’s Study Workshop  

 

This meeting was held in conjunction with the National Children’s Study, which is led by a consortium of federal agency partners: the U.S. Department of Health and Human Services (including the National Institute of Child Health and Human Development [NICHD] and the National Institute of Environmental Health Sciences [NIEHS], two parts of the National Institutes of Health, and the Centers for Disease Control and Prevention [CDC]) and the U.S. Environmental Protection Agency (EPA).

Background

Alterations in neurobehavioral development can be expressed through a number of different structural and functional end points; therefore, assessments of multiple end points are necessary to adequately assess and characterize the effects of environmental agents on neurobehavioral development (Adams et al., 2000). Fewer than 25% of neurodevelopmental disabilities that affect 3-8% of 4 million babies born each year in the U.S. have known causes (Weiss and Landrigan 2000). In addition to profound immediate effects, there is evidence that neurotoxic effects can occur after long latent periods, sometimes as a result of prenatal exposure to chemical substances that cause irreversible damage or progressive effects in adulthood. Furthermore, there is a need to determine the role of developmental insult in the etiology of human psychiatric disorders that are manifested in adolescence and young adulthood (Adams et al., 2000). A longitudinal study like the National Children’s Study (Study) is the optimal way to develop the information necessary to link early exposures with later onset neurobehavioral abnormalities. The structural and functional outcomes consequent to these exposures will depend on how well the exposures are characterized (timing, duration, level) and when the outcomes are measured. This workshop was designed to address the issue of what outcomes are important to measure for detecting the effects of environmental exposures on neurobehavioral development, and to make recommendations for measures to be included in the Study. This workshop was jointly sponsored and funded by the National Children’s Study/NICHD and the National Center for Environmental Assessment/EPA. ¹

---

¹ The views expressed in this report are those of the participants and authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency.

Purpose and Charge for the Workshop
Carole Kimmel, Ph.D., National Center for Environmental Assessment, EPA

Dr. Kimmel welcomed the participants and reviewed the history of the National Children’s Study and the status of protocol development. The steering committee for the workshop included Drs. Stanley Barone, Jr. (NCEA/EPA), David Bellinger (Harvard), Peter Scheidt (National Children’s Study/NIH), Tracey Thomas (NCEA/EPA), and Carole Kimmel (chair).

Dr. Kimmel indicated that the purpose of the workshop was to focus on:

• Measures within neurobehavioral domains (what, how, why)
• Longitudinal nature of assessments and analyses of developmental outcomes
• Critical timing of exposures.

Dr. Kimmel presented the following charge to the workshop participants:

• Propose a set of neurobehavioral measures/test batteries to evaluate normal neurodevelopment and assess the effects of exposure within four broadly defined neurobehavioral domains:
  • Cognitive
  • Motor
  • Sensory
  • Social/emotional and psychiatric
• Within each domain, propose:
  • Age of assessment/measurement
  • What, how, when
  • Assessment tool (preferred)
  • Reliability, predictive validity, specificity
  • Length of time required to administer? In what setting?
  • Cost
• Integrate measures across domains
• Consider the longitudinal nature of assessments and analyses:
  • Need for evaluation of developmental trajectories (normal and altered)
  • Linking of exposures with early and latent neurobehavioral effects
  • Linking of altered neurobehavioral effects at different developmental stages
  • Analysis of longitudinal development, repeated measures, linking exposures and outcomes
• Identify outcome measures important for particular exposures:
  • Agent
  • Critical timing of exposure
  • Biological and environmental samples, timing of sample collection
  • Critical time, age, and sources of exposure
  • Known or potential confounders
  • Known, suspected, or presumed mechanisms
  • Outcomes of concern, age at measurement.

Background Materials

Draft white papers on neuropsychological outcomes (White et al., 2005), psychiatric outcomes (McClellan et al., 2005), and motor development assessment (Rosenbaum et al., 2005) generated via a contract from the National Children’s Study Program Office served as background material prior to and during the workshop. In addition, a report from a previously held workshop on genes, environment, and behavior was made available to participants (Knox, 2004).

To aid the workshop participants in making recommendations concerning the appropriate measures that could be used to assess outcomes linked to environmental exposures, the steering committee along with the subgroup co-chairs identified a number of exposures thought to be important in causing abnormal neurobehavioral development. These exposures were tabulated and distributed to the participants prior to the workshop (see Appendix I). The workshop participants suggested other exposures of concern at the workshop, and also discussed issues concerning biological mechanisms, critical timing of exposures, persistent versus non-persistent exposures, study constraints, and the need for reality-based approaches. The exposures discussed during the workshop are reflected in Appendix I as well.

References

Adams, J., Barone, S. Jr,, LaMantia, A., Philen, R., Rice, D.C., Spear, L., Susser, E. (2000). Workshop to identify critical windows of exposure for children’s health: neurobehavioral work group summary. Environ. Health Perspect., 108 (Suppl 3):535-44.

Knox, S.S. (2004). Report from the National Children’s Study Workshop: Gene/Environment Interactions and the Regulation of Behavior. Available at http://www.nationalchildrensstudy.gov/research/workshops/.

McClellan, J., Susser, E., Bresnahan, M.A. (2005). The National Children’s Study: White paper outlining psychiatric assessments. Developed under contract with Battelle.

Rosenbaum, P. (2005). Social-emotional protocol, NCS. Developed under contract with Battelle.Weiss, B., Landrigan, P.J. (2000). The developing brain and the environment: an introduction. Environ. Health Perspect., 108 (Suppl 3):373-4.

White, R.F. (2005). National Child Study White Paper: Neuropsychological Outcomes. Developed under contract with Battelle.

Breakout Groups

Participants were assigned before the workshop to one of four breakout groups on cognitive, motor, sensory, and social/emotional and psychiatric outcomes. The remainder of the first day was spent in discussions in these groups. The groups were asked to discuss outcome measures appropriate at each age of assessment/measurement; what measures to include; what assessment tools were preferred; the reliability, predictive validity, and specificity of measures; the time required to administer them; what setting is required for measurement; and the cost. In addition, they were asked to think about measures that could be used at multiple ages for a longitudinal analysis. Following are the reports from each of the groups.

Report of the Cognitive Function Workgroup
Susan Schantz, Ph.D., Jane Adams, Ph.D., Brenda Eskenazi, Ph.D., and Patricia Rodier, Ph.D.

The Cognitive Function Group identified the following functional domains for assessment: general cognitive ability; language processing and development (including measures of expressive and receptive language as well as fluency); visual-perceptual processing; learning and memory (including measures of verbal and non-verbal memory and immediate and delayed recall); executive function (including measures of planning, cognitive flexibility, working memory and response inhibition); and attention (including measures of sustained and selective attention); and school readiness and achievement.

Tests that could be used to assess these functions in infants, preschoolers, elementary school-age children, and adolescents were discussed, and for each recommended age of assessment, a suggested test battery was proposed after considering comments and suggestions from members of other working groups and gathering additional information after the meeting. A number of important considerations that could impact the final selection of tests were also discussed. These included practical considerations, such as the availability of tests in languages other than English, the extent to which translated tests have been standardized in appropriate cultural groups, the level of training necessary to reliably administer the test, the degree to which examiner effects play a role in administration and scoring, the ease of scoring the test including whether a computerized scoring program exists, and subject-based time constraints that will be especially limiting at the younger ages of assessment. The group also discussed computer-based administration and the important role it could play in reducing training time, examiner effects, and scoring time. The group recognizes that some of the tests that have been recommended may not satisfy all of these criteria.

Recommended ages for assessment were 1 year, 2 years, 4-5 years, twice during elementary school, once each during early and late adolescence, and at age 20-21. Exact ages for the elementary school-age and adolescent assessments were not settled upon. We considered the value of having Brazelton neonatal assessments, but decided that the training demands made this impracticable.

One and Two Years of Age

It is important to recognize that the range of ages over which individual subjects are evaluated must be narrower for infants than for older children because development is quite rapid and differences may be found over just a few weeks. Thus, for greater accuracy and validity, it is recommended that 12-month-olds be tested between 47 to 52 weeks of age, and 24-month-olds be tested between 95 to 100 weeks of age.

General Mental Ability, Language Processing, and Visual-Perceptual Processing

For the assessments at one and two years of age, it is recommended that either the Bayley III, or the Mullen Scales of Early Learning (Mullen, 1995), be administered as a measure of general as well as specific mental abilities. While the Bayley Scales have a longer history of use, in recent years the Mullen (1995) has become the leading instrument for neurodevelopmental assessment in research settings. The Bayley Scales of Infant Development II (Bayley, 1993), which represent the currently available version, have scoring and interpretation complications associated with the item sets, particularly in the case of developmentally delayed or premature infants with uneven profiles (Gauthier et al, 1999; Ross and Lawson, 1997). It is not yet known whether this feature will be corrected in the coming Bayley III. Both measures provide information on motor and mental development through the use of similar items. The Mullen is advantageous due to the ease of administration and scoring, applicability from birth to 68 months of age, and assessment of the child’s general mental ability as well as strengths and weaknesses across 5 areas of functioning. The five scales assess expressive and receptive language, visual perception, and gross and fine motor development. Thus, the Mullen provides information on general mental ability, language processing, and visual-perceptual processing. The Mullen scoring system provides scaled scores, percentile ranks, and age equivalents for each scale as well as a composite standard score and percentile rank. This breakdown allows the child’s strengths and weaknesses to be assessed at early ages, thus providing better longitudinal assessment of neuropsychological development. Having the five subscale scores also allows interpretation and characterization of the child’s abilities in a way that is relevant to qualification for early intervention programs, thus facilitating referral when indicated. While the Mullen appears to be a better instrument for cognitive assessment, the Bayley is argued to be superior for the measurement of motor development (see Motor group report). Thus, the Bayley III may be recommended as a means of satisfying the needs of both groups. If the Bayley III does not change and improve the administration rules regarding item sets administered as starting points on the mental scale of the current Bayley II; however, we recommend that a standardized protocol be adopted that assures that infants of the same age will all be scored according to the administration of the same item sets.

An additional assessment of language development such as the Preschool Language Screen (PSL-4; Zimmerman et al., 2002) or the MacArthur Communicative Development Inventories (Fenson et al., 2003) is also recommended. The PSL-4 covers birth through 6 years 11 months of age, thus providing the opportunity for longitudinal use of the same test instrument at later ages. At ages one and two, the test items target interaction, attention, and vocal gestural behaviors and require 20-45 minutes for administration to the infant. The PSL-4 is also available in a Spanish version with Spanish norms (Zimmerman et al., 2002). However, the PSL-4 is recommended for use by skilled Speech Pathologists and may demand increased training of examiners to properly administer.

The MacArthur Communicative Development Inventories have recently been renamed the MacArthur-Bates Communicative Development Inventories (CDI) and are parent report forms for assessing communication skills in infants and young children up to 30 months of age. Administration to the primary caregiver requires approximately 20-40 minutes. The advantage of this measure is that administration to the primary caregiver alleviates demands on infant testing; however, use of this measure at later ages would not be possible. The CDI: Words and Gestures (Infant form) is designed for use with 8- to 16-month old children. The CDI: Words and Sentences (Toddler form) is designed for use with 16- to 30-month old children and can also be used with older children with developmental delays. These forms are also available in Spanish versions with appropriate norms (Jackson-Maldonado et al., 2003).

Executive Functions, Memory, and Attention

While there are no standardized tests that assess these specific abilities in infants, there are tests of attention, memory, and executive function that have been well-researched and replicated, yielding robust and consistent findings. These tests are advantageous because they tap cognitive abilities that are also tapped in later tests of intelligence. In contrast, infant tests of general intelligence such as the Bayley Scales assess mainly sensorimotor abilities at younger ages. Some of the tests proposed here are superior to infant tests of general intelligence in predicting performance on intelligence tests later in childhood. Predictive validity is especially robust in the case of high-risk infants (preterm and drug exposed infants, as well as infants with disabilities such as Down syndrome). In addition to their ability to tap specific cognitive abilities and their predictive validity, these tests are relatively easy to administer and score and do not require expensive equipment or materials to implement. The disadvantages are that these paradigms were not developed with the goal of being used as clinical tools and a lot of work remains to be done in terms of establishing norms. We present multiple options below, although selections need to be made to create a testing session of a length compatible with infant tolerance. All three methodologies are appropriate for use at 12 and 24 months of age and all three benefit from videotaping to improve scoring accuracy and options. Each has unique features and measures specific aspects of processing. Should only one be possible given time constraints, we would recommend that the focused-attention paradigm be used.

To assess attention in infants, it is recommended that the focused-attention paradigm developed by Ruff and colleagues be administered (for a review of this work and its potential uses in clinical assessments, see Lawson and Ruff, 2001). Ruff and colleagues argue that measures of focused attention, which include the selective as well as the intensive aspects of attention, add important information to developmental assessments of infants, because they provide information about the infant’s current learning style and quality and also predict later cognitive status. Ruff and colleagues’ paradigm assesses infants’ focused attention during active, "hands on" exploration of objects. This active exploration and play with objects figures prominently in the behavior of infants in the second half of the first year of life and throughout the second year. Focused attention is distinct from casual attention even in young infants. Focused attention takes priority over other behaviors when infants are presented with a novel object; that is, infants are more likely to examine the novel object with deliberate visual and tactile exploitation of the novel object’s properties, before engaging in less attentive behaviors such as mouthing or repetitive banging, shaking, and pushing. Focused attention increases with the presentation of a novel object, declines as the object becomes more familiar, and then increases when another novel object is introduced. More casual looking or attention does not show these systematic changes in relation to novelty. Infants are also less distractible when focusing on an object than when casually engaged in it. Focused attention, but not casual attention, has been found to be predictive of later IQ (Lawson and Ruff, 2004; Ruff, 1988; Ruff and Dubiner, 1987; Ruff et al., 1990), with greater levels of focused attention in infancy being associated with better performance in later tests of intelligence.

In the focused-attention paradigm, infants are tested in rooms that present minimal distractions and routine procedures for testing have been developed. At 12 and 24 months, administration time requires 15-25 minutes. The infant sits on the lap of a familiar caregiver and toys are presented at a table at midline and within easy reach. Uniform instructions ask the caregiver to allow the child to play independently in any way with the toys. The session is videotaped for later scoring. Procedures for presentation of toys vary depending on the age of the infant to match the infant’s attentional and object manipulation skills (see Lawson and Ruff, 2004). At 12 months of age, for example, a single toy and then a tray of multiple toys in a specified layout are presented, each for 2 minutes. The presentation of multiple toys allows for measures of the infant’s capacity to shift attention and to concentrate in the face of competing targets. The toys used are also age-appropriate and should be of moderate interest; toys that are fascinating or boring to the infant should not be used as they tend to generate a range of attention that is too limited to provide a good basis for comparison across infants. Ratings of infants’ attentional style during the session are on a 5-point scale, with 1 being assigned to infants showing relatively little engagement and no signs of concentration, and 5 to infants showing an exceptionally high level of object engagement, with clear and prolonged periods of absorption in the object. The researchers have developed detailed written guidelines and videotapes of play episodes to aid in training individuals to rate focused attention (see Lawson and Ruff, 2004).

To assess memory abilities in infants, it is recommended that visual recognition memory paradigms used by Rose, Orlian, and colleagues be used (see Rose and Orlian, 2001, for a review of their research and a discussion of potential clinical uses of their paradigm). The researchers argue that there is important evidence linking infant recognition memory to the same brain substrate that underlies explicit memory in adults. Explicit memory is involved in the conscious recollection of past events, and it is one of the cognitive skills measured in tests of general intelligence beyond infancy. There is considerable evidence that poorer performance on tests of visual recognition memory in infancy is associated with risk for cognitive delays. Among the groups studied are infants with Down syndrome (Miranda and Fantz, 1974), prenatal exposure to chemicals such as PCBs (Jacobson et al., 1985), prenatal exposure to drugs such as cocaine (Jacobson et al., 1996), malnourishment (Rose, 1994; Singer and Fagan, 1984), and prematurity (Caron and Caron, 1981; Rose, 1980; Sigman and Parmelee, 1974). Furthermore, there is growing evidence that measures of recognition memory in infancy have substantial predictive validity for performance in intelligence tests in childhood (see Rose and Orlian, 2001, for a brief review of this literature). This evidence indicates that infants who require lesser amounts of time to habituate to and to later recognize a visual stimulus tend to perform better at both non-verbal and verbal measures of IQ later in childhood.

At 12 and 24 months of age, it is recommended that a tactual-visual, cross-modal transfer task be used to assess recognition memory. This task requires 15-25 minutes. In this task, infants are first habituated to a shape tactually; that is, they are given the opportunity to palpate and manipulate an object and then it is hidden from their view in a box or some other sort of cover. The hidden object is removed after a short period of time and infants are then presented with the same object that they had previously manipulated together with a new object that they haven’t seen before and are permitted to visually inspect both of them. Evidence of recognition memory is indicated by a significant visual preference for the novel over the tactually familiar object. For details on procedures to administer and score this task see the review chapter by Rose and Orlian (2001).

There are at least three potential advantages to using visual-tactile cross-modal transfer tasks instead of visual-only recognition tasks. First, the cross-modal task requires the infant to construct a fairly abstract mental representation, given that the visual information presented in the test must be compared to a representation stored in a different modality (tactual). The level of abstraction required by the cross-modal task might be more akin to the skills assessed in later cognitive tests, possibly making the cross-modal task a better predictor of later cognitive performance than the visual-only task. Second, the greater complexity of the cross-modal task may avoid ceiling effects among older infants (12 and 24 months of age), making it a more sensitive tool for assessing individual differences late in infancy. Third, because the cross-modal task is more challenging than the visual-only task, it might keep older infants interested and motivated longer, thus facilitating data collection.

Tests of executive function for infants 2 years and younger measure the infants’ ability to use information in working memory to inhibit a prepotent response. The task, known as the A-not-B task, requires the infant to retrieve an object that is hidden in one of two identical locations. The performance of infants 8 to 12 months of age in this task has been extensively examined in research that has spanned the last 50 years. Although less researched, the task can be modified for use with infants between 13 and 24 months of age (Berger, 2004; McKenzie and Bigelow, 1986; Spencer et al., 2001; Rieser et al., 1982) so that inhibition of prepotent responses can also be assessed with infants at these older age groups. At 12 and 24 months, the task requires approximately 30 minutes.

In the typical A-not-B task used with infants 8 to 12 months, the infant is seated at a table and presented with two potential hiding locations (e.g., two cups, or two wells on the surface of the table) at midline and within easy reach. The infant then watches as the experimenter hides a toy in one of these two locations—location A. After a brief delay (1-12 seconds, depending on the age of the infant), the infant is coaxed to reach and retrieve the hidden toy. After the infant has successfully retrieved the toy from location A in two consecutive trials, the infant watches as the toy is hidden in location B, and again is coaxed to search for the toy after a brief delay. With the right length of delay (longer delays are needed as infants get older), infants will typically continue to search for the toy at location A, despite having watched it being hidden in the new, B location. The location (correct vs. incorrect) towards which the infant reaches in this reversal trial is the central measure of A-not-B performance, as it has been used by developmental psychologists (see Marcovitch and Zelazo, 1999; and Wellman et al., 1986, for reviews of the extensive research using this type of A-not-B paradigm).

Infants’ perseverative tendencies (the tendency to continue to search at A after they see the object being moved to B—the A-not-B error) can be manipulated and assessed by varying the length of the delay between the time the toy is hidden and the time the infant is allowed to reach. Work by Diamond and others has revealed that with delays of less than 1 second, most infants, 8 months of age and older, should correctly retrieve the toy in the reversal trial. At 9 months, infants should succeed in the reversal trial with delays of 2 seconds, but commit the A-not-B error when the delay is increased to 5 seconds. At 12 months, infants should succeed in the reversal trial with delays close to 10 seconds. Developmental researchers have argued that the assessment of infants’ delay tolerance in the A-not-B task can provide a good behavioral marker for the development and integrity of infants’ executive functions supported by the prefrontal cortical system (Bell and Fox, 1992; Diamond, 1990; Diamond et al., 1997; Matthews et al., 1996).

At 24 months, the manual retrieval A-not-B task becomes too easy for infants, and it is no longer a good test of executive function. However, the task can be made more cognitively complex to still allow for the study of perseverative tendencies at this age, as indexed by A-not-B type of errors, (see Berger, 2004; McKenzie and Bigelow, 1986; Spencer et al., 2001; and Rieser et al., 1982, for ways in which the A-not-B task can be adapted for use with infants older than 12 months).

Although the bulk of empirical work with the A-not-B task has not focused on potential clinical uses of the test, there is good evidence that the task can be a robust measure of individual differences. Significant differences have been found in A-not-B task performance between: infants with Down Syndrome and their controls (Rast and Meltzoff, 1995); autistic children and their controls (Dawson et al., 1998); PKU infants and their controls (Diamond et al., 1997); and cocaine-exposed infants and their controls (Noland et al., 2000). There have been no demonstrations of predictive validity of the A-not-B task as a measure of individual differences, but there have been follow-up studies of at-risk groups. For instance, infants with early-treated PKU performed more poorly than controls on the task, and when they were tested 4 years later, they were still impaired on tests of frontal lobe functioning (Diamond et al., 1997).

Unlike most infant cognition tasks, the A-not-B task does require infants to search for the target on dozens of trials and, when multiple reversal trials are used, to remain motivated even after repeated failures. Thus, experience working with this age group is extremely important in helping the examiner to establish a context in which the trials, failed or passed, are enjoyable play. There is no established version of the procedure. Published versions have varied (Bell and Fox, 1992; Diamond, 1985; Matthews et al., 1996), and the published method sections are not as extensive as the manuals for standardized tests. Thus, familiarity with the A-not-B task literature and consultation with researchers on methods would be necessary to develop standards of testing and minimize between-subject procedural variations that could potentially affect performance. It is anticipated that the training requirements for experimenters may be somewhat higher for this task than for the attention or memory tasks described above.

Four to Five Years of Age

General Mental Ability, Language Processing, and Visual-Perceptual Processing

At 4-5 years of age it is recommended that general and specific mental abilities be assessed either by repeating the Mullen Scales or by using the WPPSI-III (Wechsler, 2002; Harcourt Assessment, Inc). The Mullen Scales of Early Learning are appropriate through 68 months of age and would allow within scale longitudinal data. The WPPSI-III is divided into two age bands, 2:6-3:11 years and 4:0-7:3 years. Beginning at age 4.0, the subtests measure general mental ability, verbal comprehension, perceptual organization, and processing speed abilities. The only concern about this measure is the possibility of floor effects in lower functioning 4 year-old children, thus arguing for use at a slightly older age. Administration requires 40-55 minutes. A Spanish version is also in use.

At this age, additional measures of language processing are recommended. The PLS-4 is appropriate from birth through 6 years 11 months and is recommended for use at this time.

Learning and Memory

The Children’s Memory Scale (CMS; Cohen, 1997) is a widely used tool for the measurement of learning and memory in both verbal and visual formats and under both immediate and delayed recall conditions. The CMS is appropriate for use from age 5 to 16 years and requires approximately 30 minutes. Our group also discussed the value of the Bead Memory test from the Stanford-Binet IV as a measure of non-verbal memory, but decided that non-verbal memory could be adequately assessed using the CMS. The memory subtests from the McCarthy Scale of Children’s Ability were also considered; however, since they are normed only for 2.5 to 8.5 year olds, we felt that the broader age span of the CMS provided more value across the longitudinal ages at assessment.

Attention

Beginning at 4-5 years of age the group also recommended that children be assessed on a computerized continuous performance task (CPT) to assess sustained attention. Several CPT tests (e.g. the Conners CPT; Conners et al., 2003) are commercially available, but, in general, they rely on the child’s ability to recognize letters or numbers and thus are generally not appropriate for 4-5 year old children. However, several CPT tasks for 4-5 year olds have been developed and shared among researchers. For example, the Catch the Cat test originally developed by Streissguth et al. (1984) for use in fetal alcohol children has been adapted and used in various studies of PCB exposure (e.g., see Jacobson et al., 1992; Stewart et al., 2003). In this computerized CPT task, an image of a house with three windows where stimuli appear (an apple, butterfly, or cat) is shown on the computer screen. The cat is the target stimulus and the other two stimuli are not targets. The test is divided into three consecutive blocks of presentations. During each block the three stimuli have an equal chance of appearing and are displayed randomly in the three windows. The stimuli are presented on a variable interval schedule. The child is instructed to "catch the cat" by pushing the button on the joy stick as quickly as possible each time the cat appears.

Executive Functioning

Beginning at this age, it is also recommended that children’s executive functioning be assessed with respect to different components of executive function including planning, cognitive flexibility, inhibitory control, and working memory. Several test batteries or behavior assessment scales that specifically assess executive functions have been developed, but most have not been used in children this young.

Some computerized batteries include tests of executive functions and can be employed in children as young as four years of age. Reviewing all of them is beyond the scope of this report. However, the CANTAB stands out because it has been validated in functional neuro-imaging studies and in various clinical populations including children with autism (Hughes et al., 1999), ADHD (Kempton et al., 1999), prematurity (Luciana et al., 1999), and lead exposure (Canfield et al, 2004), as well as in adults with Parkinson’s Disease, Alzheimers Disease, various neuropsychiatric conditions, and various localized brain lesions (Robbins, 1996). In addition, although the sample size is small, normative data for individuals 8-64 years of age have been published (De Luca et al., 2003). The CANTAB battery has been used successfully to assess executive functions in children as young as 4 years of age (Luciana and Nelson, 1998; Canfield et al, 2004) and the subset of tests from the CANTAB that assess executive function (working memory and planning battery) can be administered in about an hour. These include tests of spatial span, spatial working memory, an intra- and extra-dimensional shift task that shares many features with the Wisconsin Card Sorting Task, and a planning task similar to the Tower of London task. If preferred, a smaller subset of these tests could be administered. Use of the CANTAB would allow for continuity of assessments across a wide range of ages, as well as computerized administration and scoring.

Elementary School Age

School Readiness and Academic Achievement

Our group felt that it was important scientifically, and as a benefit to participants, to assess school readiness at the start of the school years. We also thought that a later assessment of achievement would be beneficial. For this purpose, we recommend the Woodcock-Johnson III Test of Achievement. This battery contains multiple subtests that are appropriate from age 2 to adulthood. The specific subtests of value may vary depending on the precise ages selected for evaluation.

General Mental Ability, Language Processing, and Visual-Perceptual Processing

Depending on the exact age of assessment, either the WPPSI-III or the WISC-IV may be the best selection. The WPPSI-III could be used provided that testing occurred prior to 7 years 3 months of age. Alternatively, the WISC-IV (Wechsler, 2003), which is designed for use beginning at age 6, could be administered; however, the group expressed some concern about floor effects that may affect lower functioning children if evaluated at that time. All endorsed the use of the WISC-IV after age 7 to assess general and specific cognitive abilities. The WISC-IV is appropriate for children aged 6 to 16 years, thus allowing for longitudinal assessment at later ages. The WISC-IV provides information on general mental ability, verbal processing, perceptual processing, memory, and processing speed. A Spanish version is available, but level of standardization is unclear.

Learning and Memory

Repeated administration of the Children’s Memory Scale is recommended.

Attention and Executive Functioning

The Conners CPT has been a widely used measure of visual attention (Conners et al., 2003); however, it is now somewhat outdated technically. A newer measure, the Integrated Visual and Auditory (IVA) CPT (Tinius, 2003) assesses attention in both the visual and auditory modalities, is highly automated (the computer "speaks" the test instructions to minimize test variability due to administration differences), and includes several different conditions which allows for the assessment of a number of interesting and potentially important variables which are not assessed in the more traditional CPTs, including the Conners. The test is normed in individuals age 5 and up. The 13-minute test assesses impulsivity and inattention in a counter-balanced design and in both the auditory and visual modalities. The test requires the subject to click the mouse only when he or she hears or sees the target (the number "1") and not to click when he or she hears or sees the non target item (the number "2"). The first block of 100 trials consists of 50 trials in the auditory modality and 50 trials in the visual modality. This is a measure of impulsivity using a ratio of targets to non-targets of 5.25:1. The second block of 100 trials consists of 50 trials in each modality but the ratio of targets to non-targets is reversed, providing a measure of inattention. The test provides six global composite quotient scores and 22 raw scale scores that allow an in-depth analysis of the subject’s pattern of responding. A normative database of individuals from age 5-90 exists. The test-retest reliability is good and the predictive validity for correctly identifying children with ADHD is approximately 90%.

Although we recommend the repeated use of the same executive function battery employed at 4-5 years of age, there are additional options for testing of executive function at later ages. The Delis-Kaplan Executive Function System (D-KEFS; Delis et al., 2004) is a battery of tests adapted from existing executive function tests such as the Trail Making Test (Reitan, 1968). The tests themselves are not computerized, but a computerized scoring program is available. The test can be used with children as young as 8 years of age. If all items are administered, test time is approximately 90 minutes. The test was first published in 2001, and at this time there is very little information about its available in the literature, although much information is available regarding the broadly used individual test components upon which the tests in the battery are based. The Behavior Rating Inventory of Executive Function (BRIEF) is a parent questionnaire that assesses parental observations of behaviors related to executive function (Gioia et al., 2000) The scale can be used with children 5-18 years of age. There are 86 items in 8 non-overlapping clinical scales and two validity scales. These form two broader scales of behavioral regulation and metacognition as well as an overall global executive composite score. The test-retest reliability is good and the convergent validity with other scales that measure ADHD is good (Mahone et al., 2002), but BRIEF scores are not correlated with performance-based measures of executive function (Mahone et al., 2002).

Early Adolescence to Age 20

School Achievement

We recommend repeated use of the Woodcock-Johnson.

General Mental Ability, Language Processing, and Visual-Perceptual Processing

The WISC-IV can be used through 16 years of age. After this, we recommend use of its adult counterpart, the Wechsler Adult Intelligence Scale-III.

Learning and Memory

The Children’s Memory Scale can be used through 16 years of age, after which its adult counterpart, the Wechsler Memory Scales (WMS-III; Wechsler, 1997) can be employed. The adult scale is appropriate from age 16 through 89 years.

Attention and Executive Functioning

All of the childhood executive function batteries discussed above can be used into adulthood and we recommend retesting on the same battery through adolescence and in young adulthood. We also recommend use of measures demonstrated to have sensitivity to the cognitive developmental changes that occur during adolescence. Certain subtests of the CANTAB appear to be sensitive to changes from 11-14 years versus 15-19 years (De Luca et al., 2003). Our group recommends consideration of the continued use of these measures, while remaining open to measures that may emerge in the coming years.

References

Bayley, N. (1993). Bayley Scales of Infant Development—Second Edition. San Antonio, TX: Psychological Corporation.

Berger, S. E. (2004). Demands on finite cognitive capacity cause infants’ perseverative errors. Infancy, 5(2):217-238.

Bell, M.A., and Fox, N. A. (1992). The relations between frontal brain electrical activity and cognitive development during infancy. Child Development, 63:1142-1163.

Canfield, R.L., Gendle, M.H., and Cory-Slechta, D.A. (2004) Impaired Neuropsychological functioning in lead-exposed children. Developmental Neuropsychology, 26:513-540.

Caron, A. J., and Caron R. F. (1981). Processing of relational information as an index of infant risk. In S. L. Friedman and M. Sigman (Eds.), Preterm birth and psychological development (pp. 219-240). New York: Academic Press.

Cohen, M. (1997). The Children’s Memory Scale. San Antonio, Harcourt Assessment, Inc.

Conners, C.K., Epstein, J.N., Angold, A., Klaric, J. (2003). Continuous performance test performance in a normative epidemiological sample. J. Abnorm. Child Psychol., 31(5):555-62.

Dawson, G, Meltzoff, A. N., Osterling, J., and Rinaldi, J. (1998). Neuropsychological correlates of early symptoms of autism. Child Development, 69(5):1276-1285.

Delis, D., Kaplan, E., Kramer, J. (2001). D-KEFS (Delis-Kaplan Executive Function System) Examiners Manual. The Psychological Corporation; 2001.

De Luca, C. R., Wood, S. J., Anderson, V., Buchanan, J. A., Proffitt, T. M., Mahony, K. and Pantelis, C. (2003). Normative data from the CANTAB. I: development of executive function over the lifespan. J Clin Exp Neuropsychol, 25, 242-54.

Diamond, A. (1985). Development of the ability to use recall to guide action, as indicated by infants performance on A not B. Child Development, 56:868-883.

Diamond, A. (1990). The development and neural bases of memory functions as indexed by the A not B and delayed response tasks in human infants and infant monkeys. In A. Diamond (Ed.), The development and neural bases of cognitive functions (pp. 276-317). New York: New York Academy of Sciences.

Diamond, A., Prevor, M. B., Callender, G., and Druin, D. P. (1997). Prefrontal cortex cognitive deficits in children treated early and continuously for PKU. Monographs of the Society for Research in Child Development, 62(4).

Fenson, L., Dale, P. S., Reznick, J. S., Thal, D., Bates, E., Hartung, J. P., Pethick, S., and Reilly, J. S. (1993). The MacArthur Communicative Development Inventories: User’s guide and technical manual. San Diego, CA: Singular Publishing Group.

Gauthier, S.M., Bauer, C.R., Messinger, D.S., Closius, J.M. (1999). The Bayley Scales of Infant Development. II: Where to start? J. Dev. Behav. Pediatr., 20(2):75-9. (Erratum in J. Dev. Behav. Pediatr., Jun;20(3):197, 1999.)

Gioia GA, Isquith PK, Guy SC, Kenworthy L (2000). Behavior rating inventory of executive function. Neuropsychol Dev Cogn C Child Neuropsychol., Sep;6(3):235-8.

Hughes, C., Plumet, M. H., and Leboyer, M. (1999). Towards a cognitive phenotype for autism: increased prevalence of executive dysfunction and superior spatial span amongst siblings of children with autism. J. Child Psychol Psychiatry Allied Discipl, 40:705-718.

Jackson-Maldonado, D., Thal, D., Marchman, V., Newton, T., Fenson, L, and Conboy, B. (2003). MacArthur Inventarios del Desarrollo de Habilidades Comunicativas. User´s Guide and Technical Manual. Brookes, Baltimore.

Jacobson, J.L., Jacobson, S.W., Padgett, R., Brumitt, G. and Billings, R. (1992) Effects of prenatal PCB exposure on cognitive processing efficiency and sustained attention. Developmental Psychology, 28:297-306.

Jacobson, S. W., Fein, G. G., Jacobson, J. L., Schwartz, P. M., and Dowler, J. K. (1985). The effect of intrauterine PCB exposure of visual recognition memory. Child Development, 56, 853-860.

Jacobson, S. W., Jacobson, J. L., Sokol, R. J., Martier, S. S., and Chiodo, L. M. (1996). New evidence for neurobehavioral effects of in utero cocaine exposure. Journal of Pediatrics, 129, 581-590.

Kempton, S., Vance, A., Maruff, P., Luk, E., Costin, J., and Pantelis, C. (1999). Executive function and attention deficit hyperactivity disorder: stimulant medication and better executive function performance in children. Psychol. Med., 29:527-538.

Lawson, K. R., and Ruff, H. A. (2001). Focused attention: Assessing a fundamental cognitive process in infancy. In L. T. Singer and P. S. Zeskind (Eds.), Biobehavioral Assessment of the Infant. New York: Guilford Press.

Lawson, K. R., and Ruff, H. A. (2004). Early attention and negative emotionality predict later cognitive and behavioural function. International Journal of Behavioral Development, 28(2):157-165

Luciana, M., Lindeke, L., Georgieff, M., Mills, M., and Nelson, C. A. (1999). Neurobehavioral evidence for working-memory deficits in school-aged children with histories of prematurity. Dev. Med. Child Neurol., 41:521-533.

Luciana, M. and Nelson, C.A. (1998) The functional emergence of prefrontally-guided working memory systems in four- to eight-year-old children. Neuropsychologia, 36:273-293.

Mahone, E.M., Cirino, P.T., Cutting, L.E., Cerrone, P.M., Hagelthorn, K.M., Heimenz, J.R., Singer, H.S., Denckla, M.B. (2002). Validity of the behavior rating inventory of executive function in children with ADHD and/or Tourette syndrome. Arch. Clinical Neuropsychology 17:643-662.

Matthews, A., Ellis, A. E., Nelson, C. A. (1996). Development of preterm and full-term infant ability of AB, recall memory, transparent barrier detour, and means-end tasks. Child Development, 67:2658-2676.

McKenzie, B. E., and Bigelow, E. (1986). Detour behavior in young human infants. British Journal of Developmental Psychology, 4:139-148.

Marcovitch, S., and Zelazo, P. D. (1999). The A-not-B error: Results from a logistic meta-analysis. Child Development, 70, 1297-1313.

Miranda, S. B., and Fantz, R. L. (1974). Recognition memory in Down’s syndrome and normal infants. Child Development, 48, 723-738.

Mullen, E. (1995). Mullen Scales of Early Learning. Circle Pines, MN: American Guidance Service.

Noland, J. S., Singer, L. B., Mehta, S. K., and Hoang, B. H. (2000). Delay tolerance on the A-not-B task as a measure of executive functioning in cocaine-exposed infants. Poster presented at the biannual International Conference on Infant Studies in Brighton, UK.

Rast, M., and Meltzoff, A. N. (1995). Memory and representation in young children with Down syndrome; exploring deferred imitation and object permanence. Developmental Psychopathology, 3:137-162.

Reitan, R. M. (1971). Trail Making Test Results for Normal and Brain-Damaged Children. Perceptual and Motor Skills, 33, 575-581.

Rieser, J. J., Doxey, P. A., McCarrell, N. J., and Brooks, P. H. (1982). Wayfinding and toddlers’ use of information from an aerial view maze. Developmental Psychology, 18:714-720.

Robbins, T. W. (1996). Dissociating executive functions of the prefrontal cortex. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 351:1463-1471.

Rose, S. A. (1980). Enhancing visual recognition memory in preterm infants. Developmental Psychology, 16:85-92.

Rose, S. A. (1994). Relation between physical growth and information processing in infants born in India. Child Development, 65:889-902.

Rose, S. A., and Orlian, E. K. (2001). Visual information processing. In L. T. Singer and P. S. Zeskind (Eds.), Biobehavioral Assessment of the Infant. New York: Guilford Press.

Ross, G. and Lawson, K. (1997). Using the Bayley II:unresolved issues in assessing the development of prematurely born children. Journal of Developmental and Behavioral Pediatrics, 18:109-11.

Ruff, H. A. (1988). The measurement of attention in high-risk infants. In P. Vietze and H. G. Vaughan (Eds.), Early identification of infants with developmental disabilities (pp. 282-296). Philadelphia: Grune and Stratton.

Ruff, H. A., and Dubiner, K. (1987). Stability of individual differences in infants’ manipulation and exploration of objects. Perceptual and Motor Skills, 64:1095-1101.

Ruff, H. A., Lawson, K. R., Parrinello, R., and Weissberg, R. (1990). Long-term stability of individual differences in sustained attention in the early years. Child Development, 61:60-75.

Sigman, M. and Parmelee, A. H. (1974). Visual preferences for four-moth-old premature and full-term infants. Child Development, 45:959-965.

Singer, L. T., and Fagan, J. F. (1984). The cognitive development of the failure-to-thrive infant: A three-year longitudinal study. Journal of Pediatric Psychology, 9:363-384.

Spencer, J. P., Smith, L. B, and Thelen, E. (2001). Tests of a dynamic systems account of the A-not-B error: The influence of prior experience on the spatial memory abilities of two-year-olds. Child Development, 72(5):1327-1346

Stewart, P., Fitzgerald, S., Reihman, J., Gump, B., Lonky, E., Darvill, T., Pagano, J., and Hauser, P. (2003). Prenatal PCB exposure, the corpus callosum and response inhibition. Environmental Health Perspectives, 111:1670-1677.

Streissguth, A. P., Martin, D. C., Barr, H. M., Sandman, B. M., Kirchner, G. L., and Darby, B. L. (1984). Intrauterine alcohol and nicotine exposure: Attention and reaction time in 4-year-old children. Developmental Psychology, 20:533-541.

Tinius, T. P. (2003). The Integrated Visual and Auditory Continuous Performance Test as a neuropsychological measure. Archives of Clinical Neuropsychology, 18:439-454.

Wellman, H. M, Cross, D., and Bartsch, K. (1986). Infant search and object permanence: A meta-analysis of the A-not-B error. Monographs on the Society for Research in Child Development, 51 (3).

Wechsler, D. (1997). The Wechsler Adult Intelligence Scale III. San Antonio, TX: Psychological Corporation, Harcourt Brace and Company.

Wechsler D. 2002. Wechsler Intelligence Scale for Children - IV. San Antonio, TX:Psychological Corporation, Harcourt Brace and Company.

Wechsler D. 2003. WISC-IV Spanish. San Antonio, TX:Psychological Corporation, Harcourt Brace and Company.

Zimmerman, I. L., Steiner, V. G., and Pond, E. P. (1993). Preschool language scale-3, Spanish edition. San Antonio, TX: Harcourt Brace.

 

Report of the Motor Function Workgroup
Suzanne McMaster, Ph.D., Carole Kimmel, Ph.D., Kim Dietrich, Ph.D., Jane Hammond, M.A.

The workgroup charge was to identify appropriate measures of motor activity for inclusion in the National Children’s Study. Our initial discussion was structured around 6 questions:

• Which motor deficits/disorders are we trying to identify?
  • All?
  • Most prevalent?
  • Those clearly developmental in nature?
  • Those with identified/proposed links to environmental exposures?
• Are expected frequencies in line with the size of the study population?
• Is a tiered approach appropriate?
  • Parent-report
  • Motor assessment
  • Clinical evaluation
• When would measures be made?
  • Age
  • Frequency
• What instruments are recommended?
• Are any recommended measures useful for other aspects of the National Children’s Study (Study)?

In the process of answering these questions, recommendations for the study design were developed. Assessment recommendations, including test instrument, age at administration, time to administer, and appropriate test setting are provided in matrix form in Table 1. A summary of the discussion that resulted in these recommendations is provided here.

Which Motor Deficits/Disorders Are We Trying to Identify?

Our first concern was the need to define the universe of motor deficits/disabilities that should be considered for inclusion in the Study. We agreed that some categories of motor disorders were either too rare in occurrence (e.g., Duchenne muscular dystrophy) or too unlikely to be related to environmental exposures (e.g., cerebral palsy) to warrant inclusion in the Study (Missiuna et al., 2001). It was further agreed that the primary focus for motor activity evaluation should be on deficits/disorders with identified or proposed links to environmental exposures. Specific chemical classes of concern were discussed, including metals, solvents and pesticides (Rees et al., 1990; Rodier, 2004; Jacobson et al., 1984).

Are Expected Frequencies in Line with the Size of the Study Population?

The recommended evaluations are all designed to detect developmental delays or deficits that occur with frequencies well within the range expected in groups the size of the Study cohort.

Is a Tiered Approach Appropriate?

A tiered approach is not recommended. The use of a first tier parent questionnaire as a screening tool, to be followed by a second tier motor assessment, and a third tier clinical evaluation was proposed for purposes of discussion. The group agreed that use of a parent questionnaire, such as the Ages and Stages (Squires et al., 1995) posed too many potential problems to make it practical. Issues of greatest concern for this approach were parent literacy level and lack of availability of a standardized interview in multiple languages. The recommended alternative is evaluation of all infants and follow-up on all of them through age 20.

When Would Measures be Made?

The group agreed that evaluations of motor tone and reflexes at birth, and of motor development at 6-month intervals through age 2 (i.e., 6, 12, 18, and 24 months), tests of motor proficiency at 4, 6, and 10 years, and of vestibular-proprioceptive function biannually between ages 6 and 20 (i.e., 6, 8, 10, 12, 14, 16, 18, and 20 years).

What Instruments are Recommended?

Many assessment tools for infants, toddlers, young and older school-age children, and adolescents were considered by the workgroup. Tests recommended for evaluating motor development in the Study are provided in Table 1.

For the youngest infants, we recommend the New Ballard Score (Ballard, 1991) with a few additional basic reflex measures drawn from the Neonatal Behavioral Assessment Scale (NBAS) (Brazelton and Nugent, 1995) at birth. This approach will provide anatomical and functional measures of gestational maturity. No comparable measures were identified. The group also considered both the NBAS and the NICU Network Neurobehavioral Scales (NNNS), but felt that these were not feasible because of:

• The extensive amount of training involved, including the need for assessment of inter-examiner reliability across sites
• Uncertain availability of qualified personnel at some sites
• The need for repeat assessments (the NBAS was designed as a measure of recovery from the birth process and accompanying complications, requiring at least 2-3 assessments)
• The lack of an appropriately controlled environment in which to examine the neonate.

For toddlers (6-24 months) the workgroup recommended use of the Psychomotor Development Scale of the Bayley III (no reference yet for Bayley-III; Bayley, 1993) with parental queries included for failed items. The parental queries will provide additional data based on parental knowledge of the child’s performance outside of the clinical environment that would supplement objective observations made by the examiner. This approach is preferable to the questionnaire methodology because the response demands are much simpler (i.e., the caretaker will be present and actually witness the motor skill the examiner is trying to elicit). Examinations can be scored using only those items the child was observed to pass (as is ordinarily done) as well as combining objectively passed items with parental reports. We also carefully considered the Child Development Inventory (Ireton and Glascoe, 1995), but determined that the additional time required in contrast to the Bayley was not justified. When our recommendation was presented to and discussed with the full group of workshop participants in the final session, we found that the cognitive group had recommended using the Mullen Scales of Early Learning (Mullen, 1995) rather than the cognitive assessment in the Bayley II. The motor group felt that the Bayley is a better choice for measuring motor function, because there are wide scoring age ranges in the Mullen and a small number of items for each age-item set. However, it was agreed that differences between the two scales are not sufficient to warrant administering both the Bayley and the Mullen. Assuming appropriate changes are made in the Bayley III to be satisfactory for cognitive evaluation of developmentally delayed and premature infants, the motor workgroup felt that the Bayley would be preferable to use in the Study (see cognitive group report).

Our initial recommendation for 4-10 year-old children included either the Movement Assessment Battery for Children (Henderson and Sugden, 1992) or the Bruininks-Oseretsky Test of Motor Proficiency (BOTMP) (Bruininks, 1978). Further investigation shortly after the workshop refined this recommendation; it was determined that the Movement ABC is not equivalent to the BOTMP. Therefore, it is the workgroup’s recommendation that the Peabody Developmental Motor Scales-II (PDMS-II) (Folio and Fewell, 2000) be administered at 4 years and the BOTMP at ages 6 and 10 years. In addition, the Purdue pegboard test should be administered at 6 and 10 years and finger tapping at 10 years.

Our final recommendation is to include a measure of vestibular-proprioceptive function every other year beginning at age 6 and continuing through age 20. The Postural Sway test (Bhattacharya et al., 1995) is the preferred approach to obtain these data.

In summary, this group makes the following recommendations:

• Evaluate all study participants fully using the instruments described above and listed in Table 1.
• Begin evaluation of motor development at birth and follow through age 20.
• Adjust test instruments employed in the study across time to capture developing skills and/or deficits.

References

Ballard, J.L., Khoury, J.C., Wedig, K., Wang, L., Eilers-Walsman, B.L., Lipp, R. (1991). New Ballard Score, expanded to include extremely premature infants. J. Pediatr., 119(3):417-23.

Bayley, N. (1993). Bayley Scales of Infant Development—Second Edition. San Antonio, TX: Psychological Corporation.

Bhattacharya, A., Shukla, R., Dietrich, K., et al. (1995). Effect of early lead exposure on children’s postural balance. Developmental Medicine and Child Neurology, 37:861-878.

Brazelton, T.B., Nugent, J.K. (1995) Neonatal Behavioral Assessment Scale, 3 ^rd Edition , London: Mac Keith Press.

Bruininks, R.H. (1978). Bruininks-Oseretsky Test of Motor Proficiency. Circle Pines, MN: American Guidance Service.

Dubowitz, L., Dubowitz, V. (1981). The neurological assessment of the preterm and full-term newborn infant. Clinics in Developmental Medicine No. 79. Londons: Spastics International Medical Publications.

Folio, M.R., Fewell, R.R. (2000) Peabody Developmental Motor Scales—Second Edition. Austin, TX: Pro-Ed, Inc.

Henderson, S.E., Sugden, D.S. (1992). Movement Assessment Battery for Children. San Antonio, TX: Psychological Corporation.

Ireton, H., Glascoe, R. (1995). "Assessing children’s development using parents’ reports: The Child Development Inventory." Clinical Pediatrics, 34:248-255.

Jacobson, J.L., Fein, G.G., Jacobson, S.W., Schwartz, P.M., Dowler, J.K. (1984). The transfer of polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs) across the human placenta and into maternal milk. Am. J. Public Health, 74(4):378-379.

Missiuna, C., Mandich, A.D., Polatajko, H.J., Malloy-Miller, T. (2001). Cognitive orientation to daily occupational performance (CO-OP): part I--theoretical foundations. Phys. Occup. Ther. Pediatr., 20(2-3):69-81.

Mullen, E.M. (1995). Mullen Scales of Early Learning. Circle Pines, MN: American Guidance Service, Inc.

Purdue Grooved Pegboard. LaFayette Instrument Company. Lafayette, IN.

Rees, D.C., Francis, E.Z., Kimmel, C.A. (1990). Qualitative and quantitative comparability of human and animal developmental neurotoxicants: a workshop summary. Neurotoxicology, 11(2):257-269.

Rodier, P.M. (2004). Environmental causes of central nervous system maldevelopment. Pediatrics, 113(4 Suppl):1076-1083.

Squires J, Bricker D, Potter L. (1997). Revision of a parent-completed development screening tool: Ages and Stages Questionnaires. Pediatrics, 99(3):501-2; author reply 502-3.

 

Table 1. Motor Activity Measures

Report of the Sensory Function Workgroup
Deborah Rice, Ph.D., Stan Barone, Ph.D., Claire Coles, Ph.D., Tom Burbacher, Ph.D.

Clinical assessments of abnormalities of the visual and auditory systems are typically part of infant examination. A test of acuity (to detect the need for corrective lenses) is often performed near school age, and a simple hearing test may also be included. However, testing of all five senses is not standard procedure. Moreover, neither detailed testing of primary sensory function nor assessment of higher-order sensory processing is included in childhood health assessment. Although some studies of the effects of chemical exposure during development have specifically included testing of primary sensory function (most notably the Faroe Islands study of methylmercury exposure), such assessment is unusual. Therefore, the effects of chemical exposure on sensory system function are largely unknown. In addition, children with autism may have serious abnormalities in processing all modalities, and there is mounting evidence that individuals with dyslexia have trouble processing rapid changes in visual, auditory, and tactile stimuli (Hari et al., 2001). It is therefore extremely important to assess sensory function in a detailed and comprehensive manner in the National Children’s Study (Study). Early sensory deficits may prove to be predictors of later impairments, including but not limited to clinically defined syndromes.

The goals for testing sensory system function should be three-fold. First, all five senses should be assessed at appropriate and relevant ages, although not all senses will necessarily be tested at each age. Second, primary sensory function should be tested in the most sensitive and rigorous manner possible at each chosen age, recognizing the temporal and financial constraints imposed by such an ambitious undertaking. Third, it is essential that higher-order sensory processing be assessed. There is evidence, for example, that children with language and reading difficulties have abnormalities in auditory or visual processing, so that deficits that are labeled as cognitive may in fact be sensory. In addition, deficits in higher-order processing are more likely to be affected by exposure to xenobiotics than is primary sensory function.

Specific ages for testing of sensory function are suggested in this proposal. However, for the most part the ages at which testing should take place is flexible, so that the age at which sensory system testing is performed can accommodate other requirements, such as the need to test specific cognitive functions (or use specific instruments for cognitive testing) at specific ages. Nonetheless, the following recommendations were developed taking into account the developmental trajectory of specific sensory systems, as well as the ability of the child to accurately and consistently report stimulus detection or change.

Birth

Assessment of otoacoustic emissions (OAEs) and brainstem auditory evoked responses (BAERs) are often included in standard neonatal screening, although both are not necessarily included. It is recommended that both types of tests be standard for all children in the Study, since they measure different aspects of the intactness of the auditory system (see Cunningham et al., 2003, for recommendations for auditory screening at various ages, by the American Academy of Pediatrics). Inclusion of both transient evoked OAEs (TEOAEs) and distortion product OAEs (DPOAEs) should be considered. Specific stimulus parameters are recommended by the NIH Multi-center Consortium of Identification of Neonatal Hearing Impairment (Norton et al., 2000). It must be remembered, however, that neither of these techniques measures hearing, but only lower portions of the auditory pathway. Examination of the eye is also part of the typical clinical examination at birth. This should include ophthalmologic examination, ocular motility, and red reflex. The visual system (at least spatial visual function) of the newborn is relatively less well developed than the auditory system. Evoked potentials could be used to assess visual function at birth (Oliveira, 2004); however, this test is not recommended by the sensory group because of time constraints. These electrophysiological tests are not recommended later in infancy because the infant has to be sedated.

A powerful technique for assessing nervous system function in infants is changes in heart rate in response to stimuli (Porges et al., 1973). While this technique is more variable in newborns, it has been shown to be useful in detecting response to stimuli. This technique can be used to examine cognitive as well as sensory function. Cardiac orienting responses to stimulus presentations are used to explore the information processing skills of infants, and offer the opportunity for a more detailed examination of infant sensory and information processing. Orienting responses are characterized by decelerations in heart rate (HR) after the onset of a novel stimulus and are interpreted to reflect the infant’s inhibition of competing psychophysiological responses to allow for encoding of novel information. Such responses to stimuli have been found to be related to behavioral measures of infant attention (Richards and Casey, 1991; Richards, 1989; Richard, 1988; Richards, 1985) with characteristic changes in the pattern of the HR corresponding to changes in behavioral attention (Richards, 1995). Evidence suggests such characteristics of the cardiac orienting response may provide insights into the information processing skills of infants or, more specifically, in the development of the attentional system by permitting examination of the components of this process, including the speed of encoding stimuli, overall magnitude of the response, duration of the inhibition of the heart rate, and the recovery of HR after stimulus offset. This technique can be adapted to assess infant perception of a variety of stimuli. At this time point, it is recommended that hearing thresholds be determined at two frequencies: a speech frequency (1-2 kHz) and a higher frequency. A suprathreshold stimulus should be used to test habituation, as a test of cognitive function (learning).

Another technique that can be used for a variety of purposes is the infant’s sucking response to a nipple connected to a pressure transducer. It is recommended that this technique be used at birth to assess detection of a change in sweetness (babies will increase their sucking rate in response to a sweeter stimulus). One of the purposes of the Study is to identify early predictors of later problems. Autistic children have abnormalities in gustatory response (P. Rodier, personal communication); therefore, testing gustatory detection and preference at birth may prove particularly useful. This technique has also been used to assess early cognition. A baby will suck harder for the opportunity to listen to its mother’s voice versus a strange voice, for example. If the sucking response is used, it is recommended that it be used for testing other functions in addition to taste.

A simple test of olfaction should also be included in the sensory system test at birth. A normal infant will turn toward the smell of its mother’s colostrum or other pleasant smell (vanilla has been used) and away from an unpleasant odor. Heart rate response could also be used to assess olfactory stimulus detection. As is the case for taste, autistic children have abnormal olfactory recognition.

To assess somatosensory (and motor) integrity, it is recommended that nerve conduction velocity be determined in arm or leg.

Six Months

For the auditory system, OAEs and BAERs are often repeated as part of the standard six-month checkup. These should be part of the protocol for the Study. Tympanometry should be included at this age as well. In addition, heart rate response should be used to test actual hearing: threshold testing at two frequencies (speech and a higher frequency) as well as difference thresholds (ability to detect a change in frequency or amplitude). Changes in frequency and amplitude, and the temporal patterning of these changes, is what comprises human speech. Whether such early testing is predictive of later problems is of central interest to the Study. In addition, infants at this age should respond differentially to the speech sounds ba/da, an early test of language recognition (Kable and Coles, 2004).

By six months, visual function can also be tested using heart rate response (Richards, 1995). This technique can be used to test other systems, such as the auditory system, and includes visual thresholds for both spatial and temporal (motion) vision, as well as changes in stimulus presentation. There is some evidence, for example, that the fast, or transient, systems for audition, vision, and somatosensory function are impaired relative to sustained systems in children who have problems with language and reading. Therefore testing motion vision is important. It is important to keep in mind, however, that not all infants will be able to maintain the state of focused attention necessary to carry out habituation protocols so that such procedures often have a high rate of "drop out" at this age. In addition, even children who are cooperative can only sustain a limited period of involvement in such procedures so measurement should be targeted. That is, not all systems can be assessed at a particular age.

At six months, Teller Acuity Cards may also be used to assess visual acuity. This is a standard instrument that can be used at subsequent ages to assess activity longitudinally ($3,000 per instrument that can be used on multiple children, less than 10 minutes, good validity and normed). There are commercially available instruments to test visual fields at this age as well. This is recommended if time permits.

The sensory system group also recommends that occulomotor function be tested if possible. This would assess motor and sensory functions, as well as sensorimotor integration. Deficits in such reflex circuits are associated with autism and ADHD, among other neurodevelopmental disorders (Sweeney et al., 2004). Occulomotor tracking can be tested longitudinally starting at 6 months of age. Developing this procedure for routine testing may not be possible due to the time and effort required in training personnel and collecting and analyzing data.

It is not recommended that other senses be tested at six months.

Twelve Months

The tests of auditory and visual function that were assessed at six months can be repeated at 12 months. This will allow the developmental trajectory of the child to be assessed. At this age, the speech sounds bi/di should be substituted for ba/da, since babies should be past the age when ba/da is a difficult discrimination.

An additional, or alternate, measure of hearing at 12 months (through 30 months) is visual reinforced audiometry (VRA). In this procedure, the baby is conditioned to turn toward an interesting moving visual stimulus at the sound of the tone. This can be used for auditory threshold testing. If earphones are used, monaural thresholds can be determined. However, this is more labor intensive than the heart rate response paradigm, which can also be used to measure more than just pure tone thresholds. Therefore, the heart rate response is recommended over VRA.

Early Childhood (to 4 Years)

Pure tone thresholds can be assessed by age 2 using play audiometry. In this test, the child drops a toy in a box at the sound of a tone, for example.

With respect to visual system function, a battery of standard tests for detecting visual abnormalities has been recommended in a policy statement for professional medical groups (American Academy of Pediatrics et al., 2003a, b). For children from age 3 onward, distant visual acuity, ocular alignment, and ocular media clarity are to be assessed, and the policy statement makes recommendations concerning instruments and procedures. It should be noted, however, that numerous researchers also recommend instruments other than the Snellen test chart (e.g. Simmers et al., 1997; VIP study). The choice of the best test requires further examination. It is also noted that a multicenter National Eye Institute study is currently underway (Vision in Preschoolers - VIP) to determine the best set of instruments for screening for visual abnormalities in preschool children. Recommendations concerning the best instruments are already available, and several peer-reviewed papers have been published (Vision in Preschoolers Group 2004a, b; 2003a, b, c). However, additional tests should be administered in the Study. A color vision test should be included. An age-appropriate version of the Ishihara, which uses figures (boat, dog) (Cotter et al., 1999) may be useful.

A screening assessment of somatosensory function should also be performed at this age. Children can be tested for two-point discrimination on foot and hand. The ability to discriminate sandpaper of different grits (i.e., identify which is finer) has been successfully used in studies in adults in remote locations where more standard instrumentation cannot be used. Alternately, the ability to discriminate balls of slightly differing sizes can be used to test touch sensitivity.

School Age (6-7 Years)

At this age, standard pure tone audiometry testing is appropriate. In addition, language processing and auditory discrimination ability should be assessed using the Comprehensive Test of Phonological Processing (CTOPP) (price $300 per child, 30 minutes to administer).

Color vision can also be assessed at this age, using the standard Ishihara cards, and visual acuity should be measured using the best instrument (note results of the VIP study). The Electronic-Visual Acuity test should also be considered (Cotter et al., 2003; Beck et al., 2003), particularly since it provided automatic scoring and good control of stimuli. Contrast sensitivity should also be measured at this age. Visual acuity only measures one point on the spatial visual function: high frequency-high contrast detection. The visual system analyzes spatial vision as the composition of spatial frequencies (Fourier transformation). Therefore, it is important to measure visual thresholds at low and intermediate frequencies. Degradation of the ability to detect middle frequencies may result in an inability to interpret facial expression, for example, and might be interpreted as a social impairment. At this age, it is proposed that two additional points on the contrast sensitivity curve be assessed, in addition to high-frequency (acuity). The Pelli-Robson test can be used to assess contrast sensitivity for a low frequency, as long as the child can recognize letters. To determine contrast sensitivity at a middle frequency, the Cambridge or Lea low-contrast gratings can be used.

To test somatosensory function, the sandpaper grit and ball could again be used. Conversely the Pediatric Environmental Neurobehavioral Test Battery (PENTB) designed by ATSDR for identifying chemical-induced deficits contains a test of vibrotactile function (Amler et al., 1996).

Additionally, postural sway assesses vestibular, proprioceptive, and motor function. The methodology is standardized and validated, and has proven sensitive to environmental chemical exposure.

Later Childhood (10-11, 14, 15, and 18-20 Years, Exact Three Ages to be Chosen to Conform to Other Testing Requirements)

The same battery of tests is proposed for all three ages. This will allow longitudinal assessment of sensory function using the same instruments, thereby eliminating concerns about the comparability of different tests. Longitudinal assessment will provide important information not captured by only doing comparisons in a cross-sectional manner.

For audition, pure-tone audiology should again be tested. In addition, higher-order auditory processing should be included. Deficits in auditory (and visual) processing may be interpreted as deficits in cognition or social competency, or could conceivably contribute to them. Testing must be performed by an audiologist, and should include measures of the child’s ability to process complex auditory stimuli, including in noise and when degraded, and binaural processing. Test norms for these tests typically begin at 8 or 9, so younger children cannot be assessed using standard clinical instruments. Tests may include subtests of the SCAN-C (auditory figure-ground, competing words, and competing sentences), pitch, pattern test, and the random gap detection test. The last test assesses the inter-stimulus interval detection threshold, which assesses fast-response auditory processing. There are also various low-redundancy speech tests available, which assess the ability to understand degraded speech or speech in noise. Binaural speech tests may measure attention as well as audition, and should be included. In addition, deficits on temporal order judgment tasks, both auditory and visual, have been reported in dyslexics (Tallal, 1980; Hood and Conlon, 2004) and may be predictive of early reading development. Individuals with reading difficulties are also impaired in tests of fine frequency discrimination in both the auditory and visual systems (Amitay et al., 2002). An expert (or experts) in auditory testing for children should be consulted to select the best tests from the various choices, but at least 40 minutes should be devoted to higher-order auditory processing. Results can be compared to auditory processing assessment using heart-rate response during infancy, to determine predictability of early tests.

Color vision should be tested using both Farnsworth-15 and the Lanthony-15 color caps. The Farnsworth 100 is more sensitive than the Farnsworth-15 to serious (usually hereditary) color vision loss, but it is too lengthy. The Lanthony-15 is less saturated than the Farnsworth, and is sensitive to acquired color vision loss. For example, it has been used successfully to detect color vision loss from solvent exposure in adults. Dynamic contrast sensitivity and acuity should be tested using electronic presentation of spatial and temporal sine wave stimuli. This assesses both spatial and temporal (motion) vision. There is evidence, for example, that both dyslexia and schizophrenia (Butler et al., 2001) are associated with an inability to process magnocellular-biased stimuli, although other deficits are probably also present. For example, contrast sensitivity, both static and dynamic, is impaired in dyslexia (see Stuart et al., 2001 for references). The use of automated equipment is necessary for this testing, and allows automatic data collection.

Higher-order visual processing should also be assessed at all three ages. There are various tests of higher-order visual processing, such as various subtests of the Woodcock-Johnson. Tests should assess figure-ground, form constancy and spatial relations, and speed of information processing. Again, experts in the use of these batteries should be consulted for the best choices given the time constraints. In addition, such tests are often included in "cognitive" batteries, so that these tasks should be chosen in conjunction with the goals of cognitive testing. It may be preferential, however, to test for object detection and object recognition using a reaction time task, allowing carefully controlled computer-generated stimuli that are presented for short durations (e.g. 10-150 msec) and automatic data recording (see papers by K. Grill-Spector and colleagues). The CANTAB battery includes a test of rapid visual information processing, although the sensory group does not know the validity or sensitivity relative to other tests. (The CANTAB in general has the advantage of being fully automated. We suggest that it be researched for inclusion in the Study for cognitive assessment as a possible substitution for pencil-and-paper testing.)

More sophisticated somatosensory testing can be performed at these ages. Vibrotactile sensitivity at the tip of the middle finger of the preferred hand can be measured using automated equipment, and thresholds should be determined at two frequencies. It is recognized that vibration sensitivity does not assess all modalities (pain, temperature, touch, deep pressure) but the stimuli are easy to present and control, and vibration sensitivity is sensitive to lead and methylmercury exposure in monkeys (Rice and Gilbert, 1995). Vibrotactile assessment is included in the CANTAB battery, along with contrast sensitivity. Postural sway should also be assessed at all three ages.

Olfactory function should be assessed at these ages, and include both threshold and identification testing. A butanol threshold is used at the Connecticut Chemosensory Clinical Research Center. Threshold testing of one substance may not be generalizable to the thousands of other odors detectable by humans. A commercial kit (Cross-Cultural Smell Identification Test or University of Pennsylvania Smell Identification Kit) is also available for odor identification, consisting of a multiple-choice identification of 12 odors presented in a scratch-and-sniff format (5 minutes). Olfactory tests have been used in studies of pesticide exposure in adults (Dick et al., 2001; Steenland et al., 2000).

Gustatory perception should also be measured. The standard test consists of putting each of the five tastes—salt, sweet, sour, bitter, glutamate (or only the first four)—on each of the four quadrants of the tongue, requiring self-report of detection. All tastes should not be detectable in all quadrants. This method could also be used for threshold testing. Children with autism have abnormal gustatory perception (P. Rodier, personal communication). Electronic stimuli can also be presented on the tongue, and the individual asked to report the taste perceived.

 

References

American Academy of Pediatrics Committee on Practice and Ambulatory Medicine Section on Ophthalmology, American Association of Certified Orthoptists, American Association for Pediatric Ophthalmology and Strabismus, and American Academy of Ophthalmology. (2003a). Policy Statement: Eye examination in infants, children, and young adults by pediatricians. Ophthalmology 110:860-865.

American Academy of Pediatrics Committee on Practice and Ambulatory Medicine Section on Otolaryngology and Bronchoesophagology, American Association of Certified Orthoptists, American Association for Pediatric Ophthalmology and Strabismus, and American Academy of Ophthalmology. (2003b). Policy Statement: Eye examination in infants, children, and young adults by pediatricians. Pediatrics 111:902-907.

Amitay, S., Ben-Yehudah, G., Banai, K., and Ahissar, M. (2002). Disabled readers suffer from visual and auditory impairments but not from a specific magnocellular deficit. Brain, 125:2272-2285.

Amler, R.W., Gibertini, M., Lybarger, J.A., Hall, A., Kakolewski, K., Phifer, B.L., and Olsen, K.L. (1996). Selective approaches to basic neurobehavioral testing of children in environmental health studies. Neurotoxicology and Teratology, 18:429-434.

Beck, R.W., Moke, P.S., Turpin, A.H., Ferris, F.L. III, Sangiovanni, J.P., Johnson, C.A., Birch, E.E., Chandler, D.L., Cox, T.A., Blair, R.C., and Kraker, R.T. (2003). A computerized method of visual acuity testing: Adaptation of the early treatment of diabetic retinopathy study testing protocol. American Journal of Ophthalmology, 135:194-205.

Butler, P.D., Schechter, I., Zemon, V., Schwartz, S.G., Greenstein, V.C., Gordon, J., Schroeder, C.E., and Javitt, D.C. (2001). Dysfunction of early-stage visual processing in schizophrenia. American Journal of Psychiatry, 158:1126-1133.

Cotter, S.A., Chu, R.H., Chandler, D.L., Beck, R.W., Holmes, J.M., Rice, M.L., Hertle, R.W., Birch, E.E., and Moke, P.S. (2003). Reliability of the electronic early treatment diabetic retinopathy study testing protocol in children 7 to <13 years old. American Journal of Ophthalmology, 136:655-661.

Cotter, S.A., Lee, D.Y., and French, A.L. (1999). Evaluation of a new color vision test: "Color Vision Testing Made Easy". Optometry and Vision Science, 76(9):631-636.

Cunningham, M., Cox, E.O., Committee on Practice and Ambulatory Medicine, and the Section on Otolaryngology and Bronchoesophagology. (2003). Hearing assessment in infants and children: Recommendations beyond neonatal screening. Pediatrics, 111:436-440.

Dick, R.B., Steenland, K., Krieg, E.F., and Hines, C.J. (2001). Evaluation of acute sensory-motor effects and test sensitivity using termiticide workers exposed to chlorpyrifos. Neurotoxicology and Teratology, 23: 381-393.

Hari, R., Renvall, H., and Tanskanen, T. (2001). Left minineglect in dyslexic adults. Brain, 124:1373-1380.

Hood, M., and Conlon, E. (2004). Visual and auditory temporal processing and early reading development. Dyslexia, 10:234-252.

Kable, JA, and Coles, CD (2004). The impact of prenatal alcohol exposure on neurophysiological encoding of environmental events at six months. Alcoholism: Clinical and Experimental Research, 28:489-496.

Norton, S.J, Gorga, M.P., Widen, J.E., Folsom, R.C., Sininger, Y., Cone-Wesson, B., Vohr, B.R., and Fletcher, K.A. (2000). Identification of neonatal hearing impairment: Summary and recommendations. Ear and Hearing, 21:529-535.

Oliveira, A.G.F. (2004). Contrast sensitivity threshold measured by sweep-visual evoked potential in term and preterm infants at 3 and 10 months of age. Brazilian Journal of Medical and Biological Research, 37:1389-1396.

Porges, S.W., Arnold, W.R. and Formbes, E.J. (1973). Heart rate variability: An index of attentional responsivity in human newborns. Developmental Psychology, 8: 85-92.

Rice, D.C. and Gilbert, S.G. (1995). Effects of developmental methylmercury exposure or lifetime lead exposure on vibration sensitivity function in monkeys. Toxicology and Applied Pharmacology, 134:161-169.

Richards, J.E. (1989). Development and stability in visual sustained attention in 14, 20, and 26 week old infants. Psychophysiology 26:422-430.

Richards, J.E. (1988). Heart rate offset responses to visual stimuli in infants from14 to 26 weeks of age. Psychophysiology, 25:278-291.

Richards, J.E. (1985). Respiratory sinus arrhythmia predicts heart rate and visual responses during visual attention in 14 and 20 week old infants. Psychophysiology, 22:101-109.

Richards, J.E. (1995). Infant cognitive psychophysiology: Normal development and implications for abnormal developmental outcomes. Advances in Clinical Child Psychology, 17:77-107.

Richards, J.E., and Casey, B.J. (1991). Heart rate variability during attention phases in young infants. Psychophysiology, 28:43-53.

Simmers, A.J., Gray, L.S., and Spowart, K. (1997). Screening for amblyopia: A comparison of paediatric letter tests. British Journal of Ophthalmology, 81:465-469.

Steenland, K., Dick, R.B., Howell, R.J., Chrislip, D.W., Hines, C.J., Reid, T.M., Lehman, E., Laber, P., Krieg, E.F. Jr., and Knott, C. (2000). Neurologic function among termiticide applicators exposed to chlorpyrifos. Environmental Health Perspectives, 108:293-300.

Stuart, G.W., McAnally, K.I., and Castles, A. (2001). Can contrast sensitivity functions in dyslexia be explained by inattention rather than a magnocellular deficit? Vision Research, 41:3205-3211.

Sweeny, J.A., Takarae, Y., Macmillan, C., Luna, B., and Minshew, N.J. (2004). Eye movements in neurodevelopmental disorders. Current Opinion in Neurology, 17:37-42.

Tallal, P. (1980). Auditory temporal perception, phonics, and reading disabilities in children. Brain and Language, 9:182-198.

Vision in Preschoolers Study Group. (2004). Comparison of preschool vision screening tests as administered by licensed eye care professionals in the Vision in Preschoolers Study. Ophthalmology, 111:637-650.

Vision in Preschoolers Study Group. (2004). The Electronic Visual Acuity Tester: Testability of preschool children. Optometry and Vision Science, 81:238-243.

Vision in Preschoolers Study Group. (2003). Testability of preschoolers on stereotests used to screen vision disorders. Optometry and Vision Science, 80:753-757.

Vision in Preschoolers Study Group. (2003). Visual acuity results in school-aged children and adults: Lea Symbols versus Bailey-Lovie Chart. Optometry and Vision Science, 80:650-654.

Vision in Preschoolers Study Group. (2003). Threshold visual acuity testing of preschool children using the Crowded HOTV and Lea Symbols acuity tests. Journal of Pediatric Ophthalmology and Strabismus, 7:396-399.

Report of the Psychiatric/Social-Emotional Workgroup
Stephen Buka, Sc.D., David Bellinger, Ph.D., Alice Carter, Ph.D., Sarah Knox, Ph.D., Virginia Rauh, Sc.D.

Preliminary Considerations

The selection of outcome measures by the Psychiatric/Social-Emotional Endpoints workgroup was guided by two general principles. The first is that it would be important to assess both aberrant or pathological infant, child and youth functioning, as well as variations in the normal range of social and emotional behavior. This is important for several reasons, including: a) to increase understanding of environmental contributions to social/emotional functioning across the full range of variation observed in the population (rather than just yielding knowledge of environmental determinants of extreme behaviors); and b) to determine the extent to which early variations within the normal range of behavior are predictive of later social-emotional and mental health outcomes. This implies that a dimensional approach to measurement is preferable to one that is limited to dichotomous classifications such as "normal/abnormal," especially during the youngest assessment phases. The second principle is that, to an extent greater than sensory, motor, and even cognitive function, a child’s social-emotional status is likely to be context-dependent. This implies that a more veridical description of a child will be obtained if data are obtained from multiple informants who observe the child in different settings such as the home, day care center, and school, and if observations include the parent-child context.

As a result of our discussions, the workgroup formulated several methodological recommendations that should be high priorities for the National Children’s Study (Study). The first is that samples of behavior should be "banked" or archived as a resource to support future analyses just as biological samples will be archived. This could be done by videotaping both standardized assessments and routine clinical encounters with digital cameras and storing the digital media files in the Study database. For example, videotapes of the proposed cognitive assessments could be coded later for persistence and frustration tolerance. Similarly, a videotape of a child’s receipt of a routine immunization could be coded later for behavioral dimensions such as reactivity, state regulation, and aspects of parent-child interaction (such as positivity/warmth, support, and intrusiveness). In general, videotaped sessions should be briefly coded at the time of collection, and then banked for later used. Depending on the final study design, this procedure would also accommodate smaller site-specific or population-specific sub-studies in which such videotaped data may be analyzed to test focused hypotheses. In addition, such data could be used in later years in case-control analyses constructed to search for early behavioral markers of disorders emerging later in childhood or adulthood (e.g., ADHD, depression, schizophrenia). For example, close analyses of videotapes of the first-year birthday parties of children later diagnosed with autism have revealed behavioral abnormalities that distinguish these children from controls (e.g., Osterling, Dawson, and Munson, 2002).

Another methodological recommendation is that, to the extent possible, new and emerging technologies be applied in data collection and processing. Given that many questionnaires will be self-administered to adults and, eventually, to the children, loading these tools into tablet computers or PDAs and having respondents enter their answers directly would reduce the amount of paper generated by the Study, minimize data entry costs and, to the extent that some data processing can be programmed into the tablets, reduce the frequencies of scoring errors. Another innovation is systems such as the audio CASI (Computer Assisted Structured Interviews), which provide both auditory (headphones) and visual (computer) presentation of a questionnaire. This has been shown to be helpful for respondents with lower literacy skills and for the administration of sensitive items (e.g., risk-taking behaviors, substance use).

Suggested Assessments

6 Months (Clinic)

The work group decided that a neonatal assessment was unlikely to provide sufficiently valid information regarding an infant’s psychiatric/social-emotional status to warrant the costs and logistical challenges that would be involved. It was the group’s assumption that sensory, motor and cognitive assessments would take place during the first 6 months of life and that these would include some neurological screening items that would be relevant for subsequent psychiatric and social/emotional outcomes. A remaining challenge for the Study is to determine how to assess subtle neurological conditions (e.g., soft signs) for the full cohort. Whether these would be performed by well-trained lay assessors, clinicians, or some combination (through a screening process) requires additional discussion and decisions. In terms of the emphasis of this workgroup, and the planned assessment schedule of the Study, it was decided that 6 months of age would be an appropriate time to begin collecting data on social/emotional functioning.

Temperament. This was considered to be a particularly important domain to measure. This multi-dimensional construct refers to qualities such as an infant’s emotional reactivity, attention regulation, activity, soothability, etc. While many parent-completed temperament questionnaires are available, most are quite lengthy. A reduced version of the Revised Infant Behavior Questionnaire, developed by Rothbart and colleagues, was recommended (Gartstein and Rothbart, 2003).

Parent-Child Interaction. A sample of parent-child interaction under conditions that mildly stress the dyad was recommended. This could be as brief as 10 minutes and involve such paradigms for which coding schemes are well-validated. One such example is the "still face" paradigm in which the parent is instructed to play with their infant for 2-3 minutes, then to maintain a still and neutral face while looking at their infant for 2-3 minutes, and finally, to return to play for 2-3 minutes (c.f., Adamson and Frick, 2003). Alternatively, the parent could be asked to teach his/her child a task that is chosen so as to be slightly beyond the child’s present capabilities or to engage in structured play (paper bags with toys) for an equivalent period of time. The Nursing Child Assessment Teaching Scales, which is completed immediately following the episode, is a widely-used scheme designed to be used to code such interactions. To increase the efficiency of data collection, videotapes of a child participating in other Study activities, such as cognitive, motor, or sensory examinations, could be videotaped and later scored by an observer for behaviors pertinent to temperament. Parent-child interactions would provide an important sample of the child’s behavior within the primary caregiving context as well as providing a sample of parenting and dyadic interaction behaviors.

It would also be useful to observe a child’s behavior without the presence of a parent. Kagan and colleagues have developed a brief (ca. 10 minute) protocol involving a set of auditory and visual stimuli that, on the basis of the infant’s reactions, can be used to identify children who are shy and behaviorally inhibited (Kagan and Snidman, 1991). Such behaviorally inhibited children have been shown to be at increased risk of developing a variety of social-emotional disturbances in subsequent years. This task was developed for use with 4-month-old infants. Although it seems likely that it would also be valid for use with 6 months, this needs to be confirmed.

Measurement of neurophysiological correlates of temperament would be useful although standardized protocols might not be available. Insofar as measurement of heart-rate variability is planned as part of the sensory examination, it should be possible to analyze these data in ways that would inform the psychiatric/social-emotional examinations, as well. Specifically, patterns of high stable heart rate have been associated with the temperamental style of behavioral inhibition and reduced heart rate variability has been shown to be a reliable indicator of infant stress vulnerability.

The natural history of an infant’s crying might be a useful early behavioral marker of temperament. Previous studies have shown that parents are able to provide valid "cry diaries." The project might ask parents to maintain such cry diaries and/or code information on the child’s cry during the structured assessment. The work of Barry Lester and Ron Barr was cited in this regard.

Parents will be relied on to provide much of the information that will initially be available regarding infants’ psychiatric/social-emotional status. This can be problematic insofar as a parent’s characteristics might influence his or her ratings of the infant. To help in this regard, data should also be gathered on a parent’s mood and mental health at the time of assessment. Four dimensions are particularly important: depression (which could be assessed using the CES-D or other symptom measures), anxiety (Beck Anxiety Inventory), hostility/aggression (Cook-Medley), and somatization (SCL-90). In addition, the short form of the Parenting Stress Inventory should be completed. Finally, a parent’s reading skills should be determined insofar as many of the questionnaires assume at least a 5 ^th grade reading level. This list is not intended to cover the multiple domains of parent/caregiver functioning that should be considered as potential social environment risk factors, but rather as a core set of information that will facilitate the analysis and interpretation of the adult-provided data on child behavior.

12 Months (Clinic)

The assessments proposed at this age do not require direct observation, but only parent-report. Therefore the questionnaires could be mailed to the parents, provided to the parents on a tablet in the home, or conducted by telephone.

To assess temperament, the age-appropriate version of the family of questionnaires developed by Rothbart and colleagues, the Toddler Behavior Questionnaire, was recommended.

To assess an infant’s social-emotional status, the Brief Infant Toddler Social and Emotional Assessment (BITSEA) was recommended (Briggs-Gowan, Irwin, Cicchetti, Wachtel and Carter, 2004). This requires 5—7 minutes for the parent to complete and includes items from four dimensions: externalizing behaviors (aggression, defiance, peer aggression, impulsivity/overactivity), internalizing behaviors (anxiety, depression/social withdrawal, inhibition, separation/distress), dysregulation (sleep problems, eating problems, negative affectivity, sensory sensitivities), and competence (attention, empathy, compliance, imitation/play skills, prosocial peer relations, mastery motivation). The BITSEA also includes low frequency behaviors that are commonly observed among infants and toddlers on the autism spectrum.

In the case of an infant who screens positive for problems in one of the two BITSEA scales (Problems or Competence), the parent should be asked to complete the entire ITSEA, which takes an additional 20 minutes (Carter, Briggs-Gowan, Jones, and Little, 2003). In addition, parents of a 10% random sample of infants who screen negative on the BITSEA should be asked to complete the ITSEA in order to evaluate the false negative rate of the BITSEA. The group recommended that for all subsequent assessments, all participants who receive the more lengthy protocol (i.e., those who screen positive and this random 10% sample) also receive the full potential protocol at all subsequent waves of assessment. This will ensure detailed data throughout development for those judged to be at risk and a representative sample of the full cohort. In addition to the domains and scales noted above, the ITSEA also includes three indices of deviant behaviors, including atypical behaviors and social relatedness skills important in screening for autism, and low frequency maladaptive behaviors that may reflect trauma, movement disorders, and severe behavior problems. The BITSEA/ITSEA have been shown to have good test-retest reliability, inter-rate reliability, and intra-scale reliability. Preliminary data on predictive validity appear promising. For instance, in one study, 88% of infants who screened positive on the BITSEA and whose language development was poor were rated as having a social-emotional or learning problem in kindergarten.

The measures of parent/respondent characteristics described at 6 months (depression, anxiety, hostility/aggression, somatization) should be repeated, as should the brief Parenting Stress Inventory.

18 Months (Clinic)

It is recommended that temperament should again be assessed using the Toddler Behavior Questionnaire. The Laboratory Assessment of Temperament and Behavior (Goldsmith) could also be used to provide additional information on temperament. Including two emotion modules of the LABTAB would require 5 minutes of administration time. This assessment, which could be videotaped for later scoring, uses probes to elicit emotional responses from children (e.g., fear, anger, frustration).

To assess social-emotional status, it was recommended that the BITSEA again be used as a screening tool and that children who screen positive be rated by their parents on the ITSEA. The ITSEA should also be completed by the same random 10% sample that was assessed at age 12 months. Because parent report may at times be biased, children should also be screened for autism in a direct assessment, using selected items from the Autism Diagnostic Observation Schedule (ADOS). Such items might include observing the child’s response to her name and observing whether she follows a pointing finger. For children who fail this screen, the Autism Diagnostic Interview-Revised (ADI-R) and the ADOS should be completed. Children should be referred for a full clinical evaluation if indicated by the ADI-R and ADOS results.

Although it is not feasible to assess attachment at the 18-month evaluation using the Strange Situation (due to the training, administration, and coding burdens), some standardized assessment of parent-child interaction/engagement is recommended. A brief attachment assessment is apparently being developed and if it does become available, it should be considered (the Study staff should consult with the investigators of the Eccles B national birth cohort study on this matter). Another possibility would be to make a checklist out of the top and bottom 10—15 items of the Attachment Q-Sort. A trained observer could actually code these directly after or even during part of another scheduled interview in the home. An alternative method of assessing dimensions relevant to the attachment system would be to assess qualities of the parent-child relationship using videotaped teaching, play, and/or feeding episodes. Dimensions such as positivity/warmth, negativity, and joint attention could be coded using schemes developed for the NICHD Child Care Study.

Interpretation of the psychiatric/social-emotional data collected at this age would be enhanced if the cognitive assessments conducted during this period included assessments of attention and early dimensions of executive functions. For similar reasons, the measures of parent/respondent characteristics (depression, anxiety, hostility/aggression, somatization) should be repeated, as should the brief Parenting Stress Inventory.

24 Months (Home)

It was recommended that this visit be conducted in the child’s home. No direct observations of the child would be conducted, however.

The instruments used to assess temperament and social-emotional status at 18 months should be used at this time as well. The same protocol for following up children who screen positive on the BITSEA should be followed, including follow-up of the 10% random sample of children who initially screened negative.

Assessment of the quality of the home environment should be conducted using the Home Observation for Measurement of the Environment (H.O.M.E.). In addition, neighborhood characteristics should be assessed using some form of "systematic social observation" of the child’s neighborhood environment. For instance, the Study might adopt methods developed in the Project on Human Development in Chicago Neighborhoods, where interviewers walked around the block of each participant and completed coding forms for items such as levels of graffiti, broken bottles, recreational areas, and the like. The measures of parent/respondent characteristics (depression, anxiety, hostility/aggression, somatization) should be repeated, as should the brief Parenting Stress Inventory.

Starting at some point during the toddler period, an effort should be made to obtain any additional information on developmental or psychosocial services received by the child as a result of referrals by parents, preschool teachers, clinicians, or other service providers. In particular, it will be increasingly important to obtain data on medications taken as a result of these referral services (e.g., mental health, psychological counseling, or behavioral interventions).

36 Months (Clinic)

Temperament should be assessed using the Children’s Behavior Questionnaire (Rothbart et al., 2001), the age-appropriate instrument in the family of tools developed by Rothbart and colleagues. Social-emotional status should be evaluated using the same protocol as at 12, 18, and 24 months. In the case of children who screen positive on the ITSEA, however, an additional follow-up diagnostic interview with the parent should be conducted using the Preschool Age Psychiatric Assessment (PAPA) (Egger and Angold, 2004). Specific modules can be administered, depending on the clinical indications. Among the modules available for children this age are disruptive behavior disorders, anxiety/affective disorders, post-traumatic stress disorder (PTSD), and sleep disorders, all of which we expect would be emerging with a moderate level of prevalence at this age. It would also be valuable to have the BITSEA completed by a non-parent caregiver, such as a day care provider, who spends 9 or more hours per week (i.e., 3 mornings per week) with the child. This would provide the perspective of another informant on the child’s behavior, which might differ from a child’s behavior with the parent.

Children should be re-screened for autism using the appropriate items on the BITSEA with direct assessment using selected items on the ADOS, followed, if indicated, by the ADI-R and ADOS. Children should be referred for a full clinical evaluation if indicated by the ADI-R or ADOS results.

Several behavioral observations were recommended. A child’s impulse control should be assessed using tests of effortful control. These might include computerized Go-No Go tasks or naturalistic assessments of the ability to delay gratification (e.g., resist opening a wrapped package when given the opportunity). Behavioral inhibition could be assessed by noting the latency to the child’s first utterance to a non-caregiver in the unfamiliar clinic environment. Computerized emotional recognition tasks should be used to assess a child’s social cognition skills (e.g., recognition of emotions in others). If this involves observing pictures of facial expressions, it is important that a thorough evaluation of the child’s vision also be completed. Finally, the child could be asked to complete a boring task as an assessment of compliance/obedience.

Parent-child interaction during a teaching task or in free play (10 minutes total) should be videotaped and coded using the NICHD Child Study coding schemes for attachment-related dimensions such as positivity/warmth, negativity, and joint attention.

48 Months (Clinic)

At this age, it will be possible to rely on the child him or herself as an informant regarding information about the child that a parent might not know. The Berkeley Puppet Interview is a method that does so in a non-threatening manner and that might be adapted for large-scale use (Measelle, Ablow, Cowan, and Cowan, 1998). Among the issues about which the child should be queried are anxiety, depression, inhibition, mother/father closeness, peer/social competence, and marital conflict.

Parent-child interaction should again be videotaped and scored for the same dimensions germane to a child’s attachment that were assessed at 18 and 36 months.

6 Years (Clinic)

Diagnostic evaluations should be conducted for all children for the first time at this evaluation, based on parent/caregiver report. Several screener modules of the Diagnostic Interview Schedule for Children-Parent (DISC-P) should be administered to a parent, including ADHD, Conduct Disorder/Oppositional Defiant Disorder, Anxiety/Depression, and PTSD. In addition, parents should complete the dimensional Behavior Assessment System for Children-Parent (BASC-P) and be queried about the child’s utilization of mental health services (to detect psychiatric conditions that may not be detected by these standardized measures.

The child should also be interviewed, ideally with an age-appropriate technique such as the Berkeley Puppet Interview, with special focus, as at 48 months, on anxiety, depression, inhibition, mother/father closeness, peer/social competence, and marital conflict. In addition, children should be queried about fire setting, stealing, and cruelty to animals, a classic triad of behaviors that identify a child at risk for later antisocial behavior.

Schedule for Assessments

Table 2 indicates the schedule recommended for assessments conducted when children are 6 to 18 years of age. Note that diagnostic interviews with parents, using the DISC-P, will be conducted every three years during this period. Parents will also complete the BASC-P yearly. Diagnostic interviews of children, using the DISC-Children (DISC-C) will be conducted every 3 years beginning at age 12. Children will also complete the BASC-Self-Report yearly beginning at age 9. For both parents and children, only selected modules of the DISC would be administered. It is recommended that the anxiety, PTSD, depression, oppositional defiant disorder, conduct disorder, and ADHD modules be administered to all children as these are likely to have the highest rates of occurrence.

Table 2. Schedule for Assessments at 6-18 years of age

6  9  10 11 12 13 14 15 16 17 18 DISC-P X X     X     X     X BASC-P X X X X X X X X X X X DISC-C         X     X     X BASC-C   X X X X X X X X X X

In addition, beginning at age 9, information should be collected yearly on substance use and risky behaviors, and on a child’s mental health service utilization. Obtaining teacher reports on children’s behaviors (e.g., using the BASC-Teacher) and school records (for data on disciplinary actions, absenteeism, contacts with counselors) would be helpful, although logistically difficult. If possible, reports, treatment and outcome data from Child Protective Service agencies should be matched and integrated into the data set, especially information on temporary foster or kinship care arrangements.

Additional Considerations

• In some cases, the workgroup was uncertain about the availability of proposed instruments in Spanish and other languages, as well as about the cultural appropriateness of instruments for minority groups. Although time did not allow a lengthy discussion of issues related to cultural differences, we recognize that important sources of child health disparities include differential exposures, susceptibilities, diagnosis, and access to care. These last two categories are especially relevant to the work of our subgroup because the Study will be collecting information on endpoints from parents, teachers, clinicians, and other informants—all of whom will reflect, to some degree, the values and constraints of their cultures and social contexts. For both data provided by informants and administrative data on services received, children may be treated, perceived, and labeled differently as a function of race/ethnicity and socioeconomic status. We recognize that some instruments are simply not appropriate for all populations, and recommend that all instruments and procedures selected for the study be pilot-tested or at least subjected to rigorous review for their validity across the populations to be studied.
  
  
• While obtaining teacher report and school records would be very informative, the feasibility must be evaluated. It can be difficult to achieve a response rate that justifies the resources required.
  
  
• Institutional sources of information about children, such as the juvenile court system, protective services, would also be valuable, but accessing such records might not be feasible.
  
  
• The work group did not address in detail issues regarding the assessment of social competence. The white paper provided on this topic is currently being reworked. When it is completed, the assessment protocol will need to be modified in light of its recommendations.
  
  
• As noted at the outset, information from multiple sources is of particular value for these domains of assessment. Additional work is required to specify which parent/caregiver should be assessed at various ages. When possible, the parent assessments described above should be administered to multiple primary caregivers.

References

Adamson, L.B. and Frick, J.E. (2003). The still face: A history of a shared experimental paradigm. Infancy, 4(4):451-473.

Briggs-Gowan, M.J., Irwin, J., Cicchetti, D.V., Wachtel, K. and Carter, A.S. (2004). The Brief Infant-Toddler Social and Emotional Assessment: Screening for social-emotional problems and delays in competence. Journal of Pediatric Psychology, 29(2):143-155.

Carter, A.S., Briggs-Gowan, M., Jones, S.M., and Little, T. (2003). The Infant-Toddler Social and Emotional Assessment (ITSEA): Factor structure, reliability, and validity. Journal of Abnormal Child Psychology, 31(5):495-514.

Gartstein, M. A., and Rothbart, M. K. (2003). Studying infant temperament via the Revised Infant Behavior Questionnaire. Infant Behavior and Development, 26(1):64-86.

Kagan J, Snidman N. (1991). Temperamental factors in human development. American Psychologist, 46(8):856-62.

Measelle, J.R., Ablow, J.C., Cowan, P.A., and Cowan, C.P. (1998). Assessing young children’s self-perceptions of their academic, social and emotional lives: An evaluation of the Berkeley Puppet Interview. Child Development, 69:1556-1576.

Osterling, J.A., Dawson, G., Munson, J.A.(2002). Early recognition of 1-year-old infants with autism spectrum disorder versus mental retardation. Development and Psychopathology, 14(2):239-251.

Rothbart, M. K., Ahadi, S. A., Hershey, K., and Fisher, P. (2001). Investigations of Temperament at three to seven years: The children’s Behavior Questionnaire. Child Development, 72(5):1394-1408.

Data Analysis for Longitudinal Studies
Christopher Cox, Ph.D., NICHD, NIH, DHHS

Dr. Cox introduced himself as an applied statistician who tends to be very practical. He described two issues for collecting and analyzing longitudinal data of child development:

• How to measure the same construct/concept longitudinally
• How to analyze repeated measures on the same subject.

In considering measuring the same construct longitudinally, Dr. Cox explained that it is imperative to measure the same concept over time, even though the tests used may change with an individual’s age. He offered the example of measuring IQ at 6 months, 1 year, 5 years, 10 years, and so on. The challenges lie in comparing different scores from different age-appropriate tests. Dr. Cox posed the following questions:

• If researchers are measuring the same concept, such as IQ, is it being measured in such a way that the scores can be analyzed longitudinally?
• Can the scores of different tests at different ages be combined and compared in a meaningful manner?
• Do the tests allow researchers to detect whether the measured value of a concept such as IQ has increased or decreased over time?
• Is it meaningful to average two such values?
• Is it meaningful to say that one value indicates a higher level of IQ?

Longitudinal analyses can be powerful research tools because they reveal trends over time that would not be detected in cross-sectional analyses. The repeated measures of longitudinal studies provide sensitivity in detecting differences in trajectories and rates of developmental changes. The challenge, however, is in dealing with correlated data. Dr. Cox briefly discussed historical approaches to analyzing longitudinal data, including repeated analyses of variance, correlation between repeated measurements, subject-to-subject variation, and compound symmetry. Although the correlation of this type of data is still valid, it tends to work better in shorter time frames of experimental settings. In longer time frames (over years and years), the power of serial correlations diminishes over time.

More recent analytical approaches are based on maximum likelihood and restricted maximum likelihood methods of estimations that allow incorporation of repeated measures types of correlation and serial correlations in the same model. These techniques—unbalanced, mixed model techniques, when the outcomes can be assumed to be normally distributed—are useful and effective, and new diagnostic tools now allow better determination of correlation structures.

Another area of analysis involves outcomes that are not normally distributed, such as repeated binary outcomes or repeated counts (often labeled "GEE"). Dr. Cox recommended that researchers who intend to gather longitudinal data should consult and collaborate with statisticians who are familiar with these various analytical techniques and approaches.

Dr. Cox noted that analyzing longitudinal data can be challenging, and he said that these challenges are not always recognized. He cited Henry Scheffé’s 1959 book The Analysis of Variance, in which the author stated that it is much more difficult to model variances than to model means. Dr. Cox commented that the results of analyses of variance are sensitive to the assumed correlation structure, which can be difficult to "get right." In Dr. Cox’s opinion, longitudinal analyses should be performed after cross-sectional analyses have been completed. He noted that cross-sectional analyses still have value.

Another challenge in longitudinal analyses involves independent variables that change with time, or what Dr. Cox called "time-varying covariates." Such situations can be tricky, for example, when outcomes at a particular time (T) affect one or more independent variables at a subsequent time (T + 1). Dr. Cox explained that analyzing a clinical trial with repeated outcomes is one scenario, whereas it is quite another scenario to analyze an observational study with a longitudinal component. These scenarios are not comparable, with the second one much more difficult to "get right." He added that any multiple regression analyses should be performed with care. Dr. Cox said that statistical methodologies such as SAS and SAS Plus are now available in standard, commercially available software packages. He concluded by noting that longitudinal studies present difficult measurement problems that are not necessarily statistical problems.

Integration of Measures Across Time/Domains; Longitudinal Analysis of Data
Stan Barone, Jr., Ph.D., and Carole Kimmel, Ph.D., National Center for Environmental Assessment, EPA

The final session of the workshop was focused on integrating neurobehavioral measures across time and domains. Participants were asked to consider the following issues in their attempts to come up with recommendations for the National children’s Study that would cover all neurobehavioral domains:

• Logistics
• Practicality/feasibility
• What measures the Study absolutely needs
• Assessment tools that can be used across domains
• Time necessary to capture/collect data (not including data coding or decoding)
• Critical timing of measurements
• Home versus clinical settings
• Need for controlled conditions during assessments
• Quality control/quality assurance
• Resource limitations at Study centers
• Training and practice of personnel performing assessments/using tools
• Study constraints (that is, parameters of duration of testing time, patient burden, costs, and logistics).

The group began by developing a matrix of measures for different age groups integrated across domains (Table 3). Recommendations for testing at different ages through 4 years of age were discussed by all groups, but there was insufficient time to fully discuss testing at older ages. Table 3 has been updated based on the final individual workgroup reports that were further developed after the workshop itself. The workgroup reports should be consulted for more detail at these ages and for suggested measurements at older ages. Because the tools available for neurobehavioral evaluation are expected to advance with time and further research, it will be necessary to revisit the testing most appropriate for different age groups at intervals throughout the Study.

Table 3. Matrix of Integrated Measures Recommended by Age - 1 ^st Four Years

Participants

Jane Adams, Ph.D., University of Massachusetts, Boston
Stanley Barone, Jr., Ph.D., National Center for Environmental Assessment, EPA
David C. Bellinger, Ph.D., M.Sc., Harvard Medical School
Arthur M. Bennett, M.E.A., B.E.E., NICHD, NIH, DHHS
Ruth A. Brenner, M.D., M.P.H., NICHD, NIH, DHHS
Rebecca Brown, M.P.H., M.E.M., National Center for Environmental Assessment, EPA
Stephen L. Buka, Sc.D., Harvard University School of Public Health
Thomas M. Burbacher, Ph.D., University of Washington
Alice Carter, Ph.D., University of Massachusetts, Boston
Claire D. Coles, Ph.D., Emory University School of Medicine
Christopher Cox, Ph.D., NICHD, NIH, DHHS
Kim N. Dietrich, Ph.D., University of Cincinnati College of Medicine
Brenda Eskenazi, Ph.D., University of California, Berkeley
Alexa Fraser, Ph.D., Westat
Jane A. Hammond, M.A., RTI International
Sandra Jacobson, Ph.D., Wayne State University
Carole A. Kimmel, Ph.D., Office of Research and Development, EPA
Sarah S. Knox, Ph.D., NICHD, NIH, DHHS
Susan S. Lundquist, Office of Environmental Information, EPA
Suzanne B. McMaster, Ph.D., Office of Research and Development, EPA
Barbara Montwill, M.Sc., Center for Food Safety and Applied Nutrition,
    U.S. Food and Drug Administration
Monica Pourrat, M.D., children’s National Medical Center
Virginia A. Rauh, Sc.D., Columbia University
Deborah C. Rice, Ph.D., Maine Bureau of Health
Patricia M. Rodier, Ph.D., University of Rochester
Susan L. Schantz, Ph.D., University of Illinois, Urbana-Champaign
Peter C. Scheidt, M.D., M.P.H., NICHD, NIH, DHHS
Sherry G. Selevan, Ph.D., National Center for Environmental Assessment, EPA
Tracey W. Thomas, Ph.D., Office of Research and Development, EPA
Fumika Yamaguchi, M.D., Japan Science and Technology Agency

Appendix I. Matrix for Some Environmental Exposures and Outcomes for the National Children’s Study

4. Polychlorinated Biphenyls and Related Compound* Samples Needed Timing of Sample Collection Critical Time/Age, Source(s) Known or Potential Confounders Known, Suspected or Presumed Mechanism (s) Outcome (s) Age at Measure- ment Maternal blood prenatal Maternal diet (primarily seafood). Point sources. Diet; Parental IQ; Home environment; SES; ethnicicty; maternal smoking, drug and ETOH consumption; ETS; Obstetrical history. Disruption of central nervous system morphoregulation due to interference with thyroid and estrogenic hormonal activities. neuro- behavioral assessments focusing on neuromotor and sensory capacities, state, autonomic regulation and tone neonatal cord blood birth measures of sensorimotor and cognitive developmental abilities and milestones infancy breast milk early postnatal measures of activity levels, neuromotor coordination, attention, executive functions, social behavior and adjustment childhood meconium birth audition

5. Environmental Tobacco Smoke* Samples Needed Timing of Sample Collection Critical Time/Age, Source(s) Known or Potential Confounders Known, Suspected or Presumed Mechanism (s) Outcome (s) Age at Measure- ment serum cotinine child Prenatal, early postnatal, Postnatal; air Diet; Parental IQ; Home environment; SES; ethnicity; prenatal tobacco, drug and ETOH exposure; Obstetrical history. Active parental smoking, lead, nutrition, psychosocial stress, other air pollutants. Potential effect modification by social or physical factors (as above) Delays/deficits in neurite outgrowth and synaptogenesis (nicotine effects on neuro- transmission) ---excitatory. Alteration of receptor-mediated cell signaling in the brain; anti-estrogenic effect; induction of P450 enzymes; DNA damage resulting in activation of apoptotic pathways, and/or agents that bind to receptors for placental growth factors resulting in decreased exchange of oxygen and nutrients. Suspected mechanism for effect modification may involve arousal regulatory mechanism (e.g., impairment of threshold for activation of catecholamine and norepinephrin arousal system) cogntive developmental abilitites Early childhood Maternal blood 1 st, 2 nd, 3 rd trimester academic achievement   cord blood   social-behavioral adjustment   infant blood   audition       much as above (Hg, PCBs)       Fetal growth (birth weight, length, head circumference, length of gestation)       Early school performance deficits       Attention deficits

6. Repeated Exposure to Phthalates Samples Needed Timing of Sample Collection Critical Time/Age, Source(s) Known or Potential Confounders Known, Suspected or Presumed Mechan- ism(s) Outcome (s) Age at Measure- ment cord blood birth prenatal/ postnatal/ prepubertal; food, dust, indoor air Concurrent exposure to pesticides, metals and/or other chemicals ? temperament, attention, motor activity infant spot urines, bloods and environmental, i.e. residential, monitoring may be sufficient to get several samples over a range of developmen- tal stages, i.e. infant, toddler, pre-school, school entry, adolescent neuro- behavioral endpoints: attention, motor development, speech, cognition, socialization preschool and older

7. Repeated Exposure to Brominated Flame Retardants Samples Needed Timing of Sample Collection Critical Time/Age, Source(s) Known or Potential Confounders Known, Suspected or Presumed Mechan- ism(s) Outcome (s) Age at Measure- ment cord blood birth Prenatal/ postnatal/ prepubertal; Food, dust, indoor air Concurrent exposure to pesticides, metals and/ or other chemicals ? temperament, attention, motor activity infant spot urines, bloods and environmental, i.e. residential, monitoring may be sufficient to get several samples over a range of developmen- tal stages, i.e. infant, toddler, pre-school, school entry, adolescent neuro- behavioral endpoints: attention, motor development, speech, cognition, socialization preschool and older

8. Repeated Exposure to Perfluorinated Compounds (e.g., PFOS, PFOA) Samples Needed Timing of Sample Collection Critical Time/Age, Source(s) Known or Potential Confounders Known, Suspected or Presumed Mechan- ism(s) Outcome (s) Age at Measure- ment cord blood birth Prenatal/ postnatal/ prepubertal; Food, dust, indoor air Concurrent exposure to pesticides, metals and/ or other chemicals ? temperament, attention, motor activity infant spot urines, bloods and environmental, i.e. residential, monitoring may be sufficient to get several samples over a range of developmental stages, i.e. infant, toddler, pre-school, school entry, adolescent neuro- behavioral endpoints: attention, motor development, speech, cognition, socialization preschool and older