The PIN Kids Study obtained assessments of children’s performance on the Mullen Scales of Early Learning (MSEL) at 3, 12, and 36 months, though we limited our investigation to assessments obtained at 12 (Median: 12; IQR: 12-13) and 36 (Median: 36; IQR: 35-37) months. The MSEL was administered by trained PIN Study staff during follow-up visits that occurred at participants’ homes; staff were unaware of children’s exposure status at time of assessment. Administration of the MSEL takes approximately 30 minutes for children aged 12 months, and 45-60 minutes for children aged 36 months.
The MSEL is a comprehensive assessment of cognitive function, designed for children ages 0 to 68 months [120]. The MSEL yields measurements of the following cognitive domains: Expressive Language, Receptive Language, Visual Reception, Fine Motor, and Gross Motor (only assessed at 12 months); children’s scores in the first four domains can be summed to yield a Cognitive Composite Score, which reflects general cognitive function. The MSEL is a performance-based assessment, and each domain is assessed by requesting the child to perform a variety of age-specific tasks of increasing difficulty; for example, Gross Motor abilities are assessed by the child’s ability to stand, walk, and run, and the Visual Reception Scale is assessed by the child’s ability to match, sort, and nest cups. Children’s raw scores in each domain were converted to age-standardized T-Scores (Mean = 50; SD = 10), which are standardized to a nationally representative sample of 1,849 US children aged 2 days to 69
months. Cognitive Composite Scores were similarly converted to age-standardized scores based on the standardized population, which are standardized to a normal distribution with a mean of 100 and a standard deviation of 15. The MSEL has demonstrated internal consistency, test-retest reliability, and inter-rater reliability [120, 203], and scores correlate reasonably well with other developmental instruments [203].
Sample sizes and distributions of MSEL scores in the PIN Kids Study are presented in Table 8, and correlations between scores at 12- and 36-months are presented in Table 9. A total of 358 children were administered the 12-month assessment and 341 children were administered the 36-month assessment. In general, scores were higher in the 36-month assessment than the 12-month assessment, and correlations between domains approximately 24 months apart were moderate, ranging from 0.20 to 0.37.
Table 8. Sample sizes and distributions of MSEL T-scores in the PIN Kids Study.
Age at Assessment Domain n Mean SD
12-Month Composite 348 99 14 Expressive Language 354 53 9 Receptive Language 353 46 8 Visual Reception 345 49 11 Fine Motor 344 50 11 Gross Motor 355 50 12 36-Month Composite 315 107 20 Expressive Language 320 56 11 Receptive Language 331 52 12 Visual Reception 332 56 15 Fine Motor 337 48 14
Abbreviations: MSEL, Mullen Scales of Early Learning; PIN, Pregnancy, Infection, and Nutrition; SD, Standard Deviation.
Table 9. Pearson correlation coefficients between MSEL T-scores at 12 and 36 months.
Domain n Correlation Coefficient Composite 215 0.37 Expressive Language 224 0.21 Receptive Language 228 0.20 Visual Reception 220 0.25 Fine Motor 226 0.35
4.4.2 MacArthur-Bates Communicative Development Inventories (MB-CDI)
The PIN Kids Study obtained assessments of children’s performance on the MacArthur- Bates Communicative Development Inventories (MB-CDI), an early language assessment, at 12 months (Median: 13; IQR: 12-14) and 24 months (Median: 29; IQR: 27-35). The MB-CDI, a questionnaire, was provided to mothers during follow-up visits at their homes, and mothers were asked to return the questionnaires by mail; mothers were unaware of children’s exposure status at the time they completed the assessments. At the 12-month visit, mothers completed the Words and Gestures Short Form of the MB-CDI, which is designed for children aged 8- to 18- months. At 24 months post-delivery, mothers were mailed the Words and Sentences Long Form of the MB-CDI, which is designed for children age 16 to 30 months. Mothers were asked to return this assessment by mail, and if they had not done so by the 36 month visit, the assessment was completed at that time.
Both forms of the MB-CDI are designed to assess early language development in young children [199, 200]. The MB-CDI Words and Gestures Short Form includes 89 vocabulary items for which mothers select either “Understands” or “Understands and Says” for each item; the first response suggests the child understands the word but does not use it, whereas the second response suggests the child both understands and produces the word. The correlation between the short and long form of the assessment is very high [199]. The MB-CDI Words and
Sentences Long Form (for children aged 16 to 30 months) includes a vocabulary check-list of 680 words, as well as questions pertaining to how children use words (e.g., use of tense, ability to refer to absent objects), word endings (e.g., possessive forms, plurals), irregular word forms (e.g., mice, teeth), correct word endings (e.g., “blocks” vs. “blockses”, “drank” vs. “drinked”), and grammatical complexity. Good correlations between the Words and Gestures version and the Words and Sentences forms provide evidence of the instruments’ utility in assessing linguistic development [200].
In our analysis of MB-CDI scores from the 12 month assessment, we used the number of words the child understands (“Understands”) and the number of words the child understands and says (“Understands and Says”). To account for age at testing, we age-standardized these scores by dividing the raw scores by children’s age in months at assessment. In our analysis of MB-CDI scores from the 24 month assessment, we used the number of vocabulary items produced by the children (“Vocabulary”) and children’s scores on the grammatical complexity component of the examination (“Grammatical Complexity”). Here again, we age-standardized scores by dividing the raw scores on these two assessments by the child’s age in months at assessment.
A total of 401 mother-child pairs completed the 12-month assessment and 380
completed the 24-month assessment. Because the MB-CDI Words and Sentences is intended for children aged 18 to 30 months, and our assessments occurred between 21 and 47 months (Median: 29 months; IQR: 27-35), there are substantial ceiling effects and we subsequently restricted our analyses of this assessment to children who were assessed within the valid age range (n=244).