Table 3. . Linear LASSO-PCR results for 28 continuous HCP attributes.
| Model response (Significant CV results italicized) | Sample size | Significant correlation with intracranial volume | Training correlation (measure of model fit) | Observed vs. CV prediction correlation | Confidence interval [lower, upper] | Correlation p value | |
|---|---|---|---|---|---|---|---|
| Age (in years) | 841 | Yes | 0.1430* | 0.0311 | −0.0378 | 0.1007 | 0.1776 |
| Handedness | 841 | No | 0.5581* | −0.0594 | −0.1208 | 0.0017 | 0.9475 |
| Total household income | 836 | Yes | 0.1604* | −0.0029 | −0.0753 | 0.0632 | 0.5181 |
| Years of education completed | 840 | No | 0.4377* | 0.0729* | 0.0127 | 0.1343 | <10E-4 |
| Body mass index | 840 | No | 0.4976* | 0.2736* | 0.2067 | 0.3421 | <10E-4 |
| Mean hematocrit sample | 740 | Yes | 0.4348* | 0.1324* | 0.0654 | 0.1939 | <10E-4 |
| Diastolic blood pressure | 830 | No | 0.2058* | 0.0615 | −0.0154 | 0.1378 | 0.0331 |
| Systolic blood pressure | 830 | Yes | 0.3596* | 0.1396* | 0.0745 | 0.2076 | <10E-4 |
| Systolic-diastolic blood pressure ratio | 830 | Yes | NA** | −0.0240 | −0.0926 | 0.0474 | 0.7457 |
| Hemoglobin A1C | 566 | No | 0.2130* | 0.0098 | −0.0794 | 0.1071 | 0.4165 |
| Pittsburgh Sleep Quality Index | 841 | No | NA** | −0.0314 | −0.0966 | 0.0415 | 0.8277 |
| NIH Picture Sequence Memory Test | 840 | No | 0.5964* | 0.0977* | 0.0290 | 0.1618 | <10E-4 |
| NIH Dimensional Change Card Sort Test | 839 | No | 0.2381* | −0.0299 | −0.0945 | 0.0379 | 0.8071 |
| NIH Flanker Inhibitory Control and Attention Test | 841 | Yes | 0.1285* | −0.0001 | −0.0706 | 0.0651 | 0.5161 |
| Penn Progressive Matrices: Number of correct responses | 838 | Yes | 0.2027* | 0.0849* | 0.0187 | 0.1502 | <10E-4 |
| Penn Progressive Matrices: Total skipped items | 838 | Yes | 0.2090* | 0.0733* | 0.0120 | 0.1383 | <10E-4 |
| Penn Progressive Matrices: Median reaction time for correct responses | 838 | Yes | 0.1078* | 0.0086 | −0.0619 | 0.0754 | 0.4075 |
| NIH Oral Reading Recognition Test | 841 | Yes | 0.1665* | 0.0008 | −0.0702 | 0.0660 | 0.4748 |
| NIH Picture Vocabulary Test | 841 | Yes | 0.5206* | 0.0481 | −0.0187 | 0.1142 | 0.0781 |
| NIH Toolbox Pattern Comparison Processing Speed Test | 841 | No | 0.1814* | −0.0569 | −0.1260 | 0.0061 | 0.9390 |
| Delay Discounting: Area under the curve for discounting of $200 | 838 | Yes | 0.3010* | 0.0275 | −0.0311 | 0.0891 | 0.2202 |
| Delay Discounting: Area under the curve for discounting of $40,000 | 838 | No | 0.2056* | 0.0802* | 0.0132 | 0.1527 | <10E-4 |
| Variable Short Penn Line Orientation: Total number correct | 838 | Yes | 0.4490* | 0.0951* | 0.0279 | 0.1589 | <10E-4 |
| Variable Short Penn Line Orientation: Median reaction time divided by expected number of clicks for correct | 838 | Yes | 0.4449* | −0.0572 | −0.1302 | 0.0141 | 0.9520 |
| Variable Short Penn Line Orientation: Total positions off for all trials | 838 | Yes | 0.4695* | 0.0014 | −0.0621 | 0.0735 | 0.4741 |
| Penn Word Memory Test: Total number of correct responses | 838 | No | 0.2382* | 0.0474 | −0.0228 | 0.1189 | 0.0764 |
| Penn Word Memory Test: Median reaction time for correct responses | 838 | No | 0.2354* | −0.0391 | −0.0965 | 0.0191 | 0.9034 |
| NIH List Sorting Working Memory Test | 841 | Yes | 0.4140* | 0.0793* | 0.0097 | 0.1540 | <10E-4 |
*
The observed-predicted correlation was statistically significant after applying the FDR correction for multiple comparisons.
**
Training correlation could not be computed when the full HCP training set yielded no nonzero LASSO coefficients for ICV or LCF PCs.