Table 2.
Model parameters, including modifiers of performance trajectories. Values are model estimates ± SE.
| Model estimates without modifiers | |||
|---|---|---|---|
| ϕ1 | ϕ2 | ϕ3 | |
| Impairment: UEFM Df = 336 |
45.97 ± 2.47*** | 3.90 ± 2.85 | 13.02 ± 0.77*** |
| Capacity: ARAT Df = 337 |
35.64 ± 2.62*** | 19.79 ± 2.86*** | 9.28 ± 0.47*** |
| Performance: Use ratio Df = 326 |
0.65 ± 0.04*** | −5.60 ± 4.93† | 13.51 ± 3.17*** |
| Performance: Paretic hours Df = 326 |
4.27 ± 0.34*** | 4.90 ± 2.97 | 16.53 ± 3.02*** |
|
Contributions of modifiers to Use Ratio model Values are interactions of the modifier and parameter indicating the amount by which the logistic parameter changes with a one unit change in the moderator. | |||
| ARAT | 0.01 ± 0.0009*** | −0.55 ± 0.13*** | -- |
| Concordance | 0.17 ± 0.07* | 10.8 ± 3.91** | -- |
| Age | −0.005 ± 0.004 | -- | -- |
| Upper limb rehabilitation | −0.01 ± 0.02 | 3.53 ± 1.14** | -- |
| CES-D | −0.008 ± 0.004* | 0.09 ± 0.20 | -- |
| MoCA | 0.008 ± 0.005 | −0.7 ± 0.27 | -- |
| Unstructured Mesulam | −0.001 ± 0.007 | −0.21 ± 0.36 | -- |
|
Contributions of modifiers to Paretic Hours model Values are interactions of the modifier and parameter indicating the amount by which the logistic parameter changes with a one unit change in the moderator. | |||
| ARAT | 0.08 ± 0.01*** | −0.43 ± 0.11*** | -- |
| Concordance | 1.65 ± 0.66* | 6.57 ± 4.13 | -- |
| Age | −0.05 ± 0.03 | -- | -- |
| Upper limb rehabilitation | −0.24 ± 0.04 | 2.59 ±0.92** | -- |
| CES-D | −0.09 ± 0.03** | −0.24 ± 0.26 | -- |
| MoCA | 0.09 ± 0.05 | −0.66 ± 0.29* | -- |
| Unstructured Mesulam | −0.04 ± 0.07 | −0.68 ± 0.62 | -- |
ϕ1: upper asymptote in units of the scale; i.e. the average value into which the curve flattens, and does not increase further over time.
ϕ2: inflection point in days; i.e. the average time when the curve starts to slow its growth. Note that some individual participant curves fit best only to the upper part of the curve, such that the inflection point occurs on or before the time of stroke.
ϕ3: “growth rate” best viewed as a scaling factor for how quickly the asymptote, or flattening of the curve is achieved.
ARAT: Action Research Arm Test; CES-D: Center for Epidemiologic Studies – Depression Scale; Df: degrees of freedom; MoCA: Montreal Cognitive Assessment; UEFM: Upper extremity Fugl-Meyer Assessment.
--: models that tried to estimate these interaction parameters did not converge.
p ≤ 0.05
p < 0.01
p < 0.001 Note that the statistical significance of each parameter reflects whether or not the value is different from zero, not the fit of the model.
The negative value for ϕ2 indicates that the best model places the inflection point of the curve prior to the stroke.