Article Figures & Data
Figures
- Figure 1.
Condition-general longitudinal increase in parietal activity over 4 years. A, Changes in brain activity over 4 years are observed in right inferior parietal cortex for maintenance and manipulation. Purple outlines, areas that were activate at baseline; red-yellow colors, significant increases (p < 0.05, FWE corrected). B, Corresponding bar graph for right and left inferior parietal cortex are based on peak activity in this region identified in a conjunction analysis of both conditions; cf. Table 1 for cluster statistics.
- Figure 2.
Condition-specific (manipulation–maintenance) decrease in brain activity over 4 years. Purple outlines, areas that were active at baseline; red-yellow color, significant decreases in left prefrontal cortex (p < 0.0001, uncorrected). Corresponding bar graphs for left lateral prefrontal cortex are based on peak activity in this region; cf. Table 1.
- Figure 3.
Dropout analysis. Increased levels of brain activity in right inferior parietal cortex (β values from the cluster identified in Table 1) for dropouts at baseline (A). Voxelwise analysis of higher brain activity during the maintenance condition in dropouts > returners shown in red-yellow (p < 0.0001, uncorrected for illustration), overlaid onto the purple outline of the brain areas implicated in maintenance in the dropouts.
- Figure 4.
Subgroup analyses. A, WM performance (hits – false alarms, manipulation–maintenance) by time point for subgroups decliners (n = 50) and stable (n = 50). B, fMRI activation (β) by subgroup, brain region, and time.
Tables
Contrast Location T k Threshold Condition-general (manipulation and maintenance) Increases 44 –52 30 (right parietal cortex)* 4.91 38 <0.05† –54 –52 –18 (left temporal cortex) 4.56 3 <0.05† –32 –62 34 (left parietal cortex)* 4.38 59 <0.0001 Decreases — Not significant Condition-specific (manipulation–maintenance) Increases — Not significant Decreases –46 28 22 (lateral PFC) * 4.20 440 <0.0001‡ Condition Task Control Maintenance Manipulation Baseline (n = 136) 8.63 (0.63) 8.37 (0.81) 7.35 (1.45) Follow-up (n = 136) 8.70 (0.65) 8.41 (0.94) 7.08 (1.44) Dropouts at baseline (n = 81) 8.57 (0.10) 8.12 (0.12) 6.77 (0.16) Variable OR SE p Bilateral parietal activation 1.69 0.17 <0.01 DLPFC activation 0.96 0.16 0.79 Global brain atrophy 0.81 0.18 0.24 Sex (reference category male) 1.95 0.31 0.03 Manipulation performance (hits – false alarms) 0.79 0.02 0.03 Age (years) 1.02 0.02 0.37 Dependent variable, dropout (yes/no). Predictors: bilateral parietal activation, z-scored baseline activation, averaged across manipulation and maintenance; DLPFC activation, z-scored baseline activation; global brain atrophy, global brain volum, z-scored, corrected for intracranial volume.
Line Data/dependent variable Type of test Statistic Confidence Methods: performance subgroups a Hits – false alarms baseline (manipulation–maintenance) t-test t = 0.50; DoF = 98 p = 0.62; CI = –0.30/0.50 b Hits – false alarms follow-up (manipulation–maintenance) t-test t = –10.42; DoF = 98 p < 0.001; CI = –2.57/–1.75 c Baseline age (years) t-test t = 0.28; DoF = 98 p = 0.78; CI = –2.12/2.81 d Education t-test t = –0.65; DoF = 98 p = 0.52; CI = –2.73/1.40 e Sum of correct responses n-back, baseline t-test t = –0.79; DoF = 98 p = 0.43; CI = –2.31/0.99 f Sum of correct responses n-back, follow-up t-test t = –2.01; DoF = 98 p = 0.05; CI = –2.94/–0.02 Results: FMRI second level voxelwise analyses using SPM12 g1 Contrast values, maintenance–control Multiple regression at baseline (cross-sectional age effect) p > 0.0001, no significant clusters g2 Paired t-test, baseline to follow-up (longitudinal effect) t = 5.21; DoF = 135 p = 0.006corrected g3 Multiple regression (time × age) No significant clusters g4 t-test (dropout vs. returners) t = 4.65 p < 0.03corrected h1 Contrast values, manipulation–control Multiple regression at baseline (cross-sectional age effect) p > 0.0001, no significant clusters h2 Paired t-test, baseline to follow-up (longitudinal effect) t = 5.11 DoF = 135 p = 0.009corrected h3 Multiple regression (time × age) No significant clusters h4 t-test (dropout vs. returners) t > 3.13 P < 0.001uncorrected i1 Contrast values, manipulation–maintenance Multiple regression at baseline (cross-sectional age effect) p > 0.0001, no significant clusters i2 Paired t-test, baseline to follow-up (longitudinal effect) cf. Table 1 i3 Multiple regression (time × age) No significant clusters j Contrast values, (1) maintenance–control; (2) manipulation–control Conjunction cf. Table 1 Results: post hoc/individual difference analyses k Right and left parietal β (maintenance–control, follow-up – baseline) Pearson’s correlation r = 0.83; DoF = 136 p < 0.01; CI = 0.58/0.92 l Right and left parietal β (manipulation–control, follow-up – baseline) Pearson’s correlation r = 0.68; DoF = 136 p < 0.01; CI = 0.35/0.85 m DLPFC β (manipulation–control) Paired t-test t = 2.50; DoF = 135 p = 0.01; CI = 0.02/0.19 n DLPFC β (maintenance–control) Paired t-test t = 0.97; DoF = 135 p = 0.33; CI = –0.11/0.04 o DLPFC β (manipulation–maintenance, follow-up – baseline), bilateral parietal beta (average manipulation and maintenance, follow-up – baseline) Pearson’s correlation r = 0.11; DoF = 136 p = 0.20; CI = –0.06/0.29 p Performance (hits – false alarms) by condition Two-way ANOVA F = 4.82; DoF = 135 p = 0.03; partial η2 = 0.03 q DLPFC β (manipulation–maintenance) Three-way ANOVA F = 5.63; DoF = 98 p = 0.02; partial η2 = 0.05 r Paired t-test t = 3.09; DoF = 49 p < 0.01; CI = 0.08/0.36 s Paired t-test t = 0.54; DoF = 49 p = 0.59; CI = –0.07/0.13 t Bilateral parietal β (maintenance–control, baseline) t-test t = –3.59; DoF = 215 p < 0.01; CI = –0.21/–0.03 u Bilateral parietal β (manipulation–control, baseline) t-test t = –2.70; DoF = 215 p = 0.01; CI = –0.19/–0.03 v Outcome: dropout (yes/no) Logistic regression cf. Table 3
In this issue
