Table 2. Regression Coefficients From Linear Mixed Models to Investigate Change Over Time in Cognitive Domainsa.
| Model | Global cognition | Memory | Executive functions | Verbal fluency | Visuospatial functions | Attention/processing speed | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β (95% CI) | P value | β (95% CI) | P value | β (95% CI) | P value | β (95% CI) | P value | β (95% CI) | P value | β (95% CI) | P value | |
| Total sample (n = 330) | ||||||||||||
| Linear time model | ||||||||||||
| Time | −0.03 (−0.06 to −0.01) | .01b | −0.10 (−0.14 to −0.05) | <.001b | 0.02 (−0.02 to 0.06) | .28 | −0.01 (−0.06 to 0.03) | .53 | 0.01 (−0.04 to 0.05) | .72 | 0.01 (−0.04 to 0.05) | .79 |
| Discrete time model | ||||||||||||
| Year 1 | −0.00 (−0.06 to 0.05) | .97 | 0.02 (−0.08 to 0.11) | .72 | 0.02 (−0.05 to 0.10) | .55 | 0.01 (−0.08 to 0.11) | .79 | 0.04 (−0.05 to 0.14) | .39 | 0.04 (−0.05 to 0.13) | .36 |
| Year 2 | −0.10 (−0.17 to −0.03) | .008b | −0.26 (−0.39 to −0.14) | <.001b | −0.04 (−0.14 to 0.07) | .47 | 0.01 (−0.12 to 0.14) | .87 | −0.07 (−0.20 to 0.06) | .27 | 0.08 (−0.04 to 0.19) | .20 |
| Year 3 | −0.13 (−0.24 to −0.01) | .03 | −0.35 (−0.55 to −0.16) | <.001b | 0.26 (0.10 to 0.43) | .002b | −0.16 (−0.36 to 0.04) | .11 | −0.02 (−0.21 to 0.18) | .87 | −0.08 (−0.26 to 0.10) | .37 |
| Year 4 | 0.00 (−0.19 to 0.20) | .97 | −0.16 (−0.50 to 0.18) | .37 | −0.00 (−0.29 to 0.28) | .98 | 0.08 (−0.26 to 0.43) | .64 | 0.44 (0.09 to 0.79) | .01 | −0.03 (−0.35 to 0.29) | .85 |
| MMSE ≥26 after 2 y (n = 43) | ||||||||||||
| Linear time model | ||||||||||||
| Time | −0.05 (−0.09 to −0.01) | .01b | −0.10 (−0.17 to −0.02) | .01 | −0.02 (−0.08 to 0.04) | .49 | −0.02 (−0.09 to 0.05) | .56 | −0.01 (−0.07 to 0.05) | .76 | −0.09 (−0.16 to −0.03) | .006a |
| Discrete time model | ||||||||||||
| Year 1 | −0.01 (−0.11 to 0.09) | .92 | 0.09 (−0.11 to 0.29) | .37 | −0.06 (−0.23 to 0.10) | .46 | −0.00 (−0.20 to 0.19) | .98 | −0.02 (−0.18 to 0.14) | .79 | −0.04 (−0.22 to 0.13) | .63 |
| Year 2 | −0.06 (−0.17 to 0.04) | .22 | −0.21 (−0.42 to −0.01) | .04 | −0.04 (−0.21 to 0.13) | .63 | 0.06 (−0.14 to 0.26) | .57 | −0.07 (−0.24 to 0.09) | .38 | −0.07 (−0.25 to 0.11) | .45 |
| Year 3 | −0.26 (−0.42 to −0.10) | .002b | −0.31 (−0.63 to 0.00) | .06 | 0.05 (−0.21 to 0.31) | .73 | −0.22 (−0.53 to 0.09) | .17 | −0.27 (−0.52 to −0.01) | .04 | −0.51 (−0.78 to −0.24) | <.001a |
| Year 4 | −0.09 (−0.31 to 0.13) | .44 | −0.17 (−0.60 to 0.27) | .45 | −0.20 (−0.56 to 0.15) | .27 | −0.09 (−0.52 to 0.34) | .68 | 0.39 (0.05 to 0.74) | .03 | −0.32 (−0.69 to 0.06) | .10 |
Abbreviation: MMSE, Mini-Mental State Examination.
a
Linear time models are defined as linear models with time as a continuous variable for each cognitive domain separately. Discrete time models are separate linear models with time as a dummy variable for each cognitive domain to investigate the trajectories for every year separately. All models include a random intercept, and were adjusted for sex, age, education, and vision and hearing capacities. The models on verbal fluency were only adjusted for hearing capacities. A random slope was not included.
b
Significant after correction for multiple testing (listed P values are uncorrected for multiple comparisons).