Figure 1.
Scatter plot of observed longitudinal responses (MP measurements) in the less affected side for anterior putamen (left panel) and posterior putamen (right panel) with the superimposed non-linear curves derived from the random effects model for patients with an average age at symptom onset of 53 years.
Table 1.
Estimated nonlinear random effects model of responses (MP measurements) in the less affected side for anterior putamen and posterior putamen.
Figure 2.
Longitudinal model fits of responses (DTBZ measurements) in the more affected side for the average of the three putaminal regions in the younger and older groups (35 and 70 years, respectively).
Figure 3.
Longitudinal, bivariate model fits of standardized responses (FD and DTBZ measurements) in the more affected side averaged across the three putaminal regions.
This suggests that the effects of early compensation dissipate over time (see text).
Figure 4.
Longitudinal (all 3 visits) and cross-sectional (1st visits only) model fits of responses (FD measurements) in the more affected side averaged across the three putaminal regions for patients with an average age at symptom onset of 53 years.
There was little suggestion of bias due to selective loss of patients in the longitudinal approach. For example, the rate of decline in FD uptake at 15 years since symptom onset was about −0.0095 min per year from the longitudinal curve based on 3 visits and −0.0107 min
per year from the cross-sectional curve based on those patients who only provided a single visit.