Table 1. Significant dependencies among neurological phenotypes (Bayley) and growth parameters (Growth) with 2-way (top sub-table) and 3-way effects (middle and bottom sub-tables) were observed.
| Bayley | Growth / Bayley | Growth | P-values |
|---|---|---|---|
| Cognitive | alpha | 1.76E-05 | |
| Adaptive | lambda | 3.29E-04 | |
| Social-emotional | alpha | 4.7E-04 | |
| Language | linf | alpha | 2.7E-04 |
| Social-emotional | linf | alpha | 1.293E-03 |
| Adaptive | linf | alpha | 1.905E-03 |
| Cognitive | linf | alpha | 2.257E-03 |
| Adaptive | alpha | lambda | 2.416E-03 |
| Social-emotional | Adaptive | linf | 2.22E-06 |
| Social-emotional | Adaptive | lambda | 3.97E-06 |
| Language | ITSEA | linf | 5.69E-06 |
| Adaptive | CBCL | lambda | 7.96E-06 |
| Language | Social-emotional | alpha | 8.87E-06 |
| Language | Adaptive | alpha | 1.02E-05 |
Note that the change in statistical significance after multiple hypothesis correction here is small. The composite Cognitive Bayley scale score and the growth rate parameter alpha showed the strongest 2-way dependencies, as measured by the p-value (described in Methods section 4.4.) There was a clear relationship between robust head growth and the Bayley phenotypes for both 2-way and 3-way dependencies. With the exception of the two shown in the bottom sub-table, the other child-specific neurological phenotype data such as Infant Toddler Social Emotional Assessment (ITSEA) and Child Behavior Check List (CBCL) were observed in 3-way dependency with weaker significance levels (p-values > 10−4) and are not reported here.