Table 7.
Model 1* (N=1182) |
Model 2 (N=1104) |
Model 3 (N=1180) |
|||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
β | 95% CI | P | Overall | β | 95% CI | P | Overall | β | 95% CI | P | Overall | ||
Time | − 0.12 | − 0.28 – 0.04 | .14 | − 0.16 | − 0.35 – −0.03 | .10 | − 0.11 | − 0.27 – 0.05 | .16 | ||||
Middle Category (MeDi 4–5), baseline testing | − 0.02 | − 0.11 – 0.06 | .59 | .54 | 0.01 | − 0.07 – 0.10 | .72 | .75 | − 0.02 | − 0.10 – 0.06 | .65 | .52 | |
High Category (MeDi 6–9), baseline testing | 0.02 | − 0.07 – 0.12 | .60 | 0.04 | − 0.06 – 0.14 | .45 | 0.03 | − 0.06 – 0.12 | .53 | ||||
Middle Category × time | − 0.01 | − 0.03 – 0.009 | .27 | .19 | − 0.02 | − 0.04 – 0.0007 | .06 | .08 | − 0.01 | − 0.06 – 0.001 | .27 | .18 | |
High Category × time | − 0.02 | − 0.04 – 0.002 | .07 | − 0.02 | − 0.05 – −0.0007 | .04 | − 0.02 | − 0.05 – 0.001 | .06 |
FCSRT | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Model 1 |
Model 2 |
Model 3 |
||||||||||
β | 95% CI | P | Overall | β | 95% CI | P | Overall | β | 95% CI | P | Overall | |
Time | 0.58 | − 0.74 – 1.90 | .39 | 0.04 | − 1.49 – 1.57 | .96 | 0.55 | − 0.77 – 1.87 | .41 | |||
Middle Category (MeDi 4–5), baseline testing | − 0.34 | − 1.17 – 0.49 | .42 | − 0.51 | − 1.37 – 0.35 | .25 | − 0.33 | − 1.16 – 0.50 | .44 | |||
.25 | .18 | .25 | ||||||||||
High Category (MeDi 6–9), baseline testing | 0.37 | − 0.56 – 1.29 | .44 | 0.27 | − 0.70 – 1.23 | .59 | 0.38 | − 0.55 – 1.31 | .42 | |||
Middle Category × time | 0.18 | 0.006 – 0.35 | .04 | 0.17 | − 0.007 – 0.35 | .06 | 0.18 | 0.002 – 0.35 | .048 | |||
.07 | .08 | .08 | ||||||||||
High Category × time | 0.19 | 0.001 – 0.39 | .048 | 0.21 | 0.01 – 0.41 | .04 | 0.19 | − 0.003 – 0.39 | .05 |
IST | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Model 1 |
Model 2 |
Model 3 |
||||||||||
β | 95% CI | P | Overall | β | 95% CI | P | Overall | β | 95% CI | P | Overall | |
Time | 0.27 | − 1.31 – 1.86 | .73 | 0.56 | − 1.33 – 2.44 | .56 | 0.22 | − 1.37 – 1.82 | .78 | |||
Middle Category (MeDi 4–5), baseline testing | 1.42 | 0.20 – 2.65 | .02 | 1.22 | − 0.06 – 2.50 | .06 | 1.34 | 0.11 – 2.57 | .03 | |||
.05 | .13 | .06 | ||||||||||
High Category (MeDi 6–9), baseline testing | 0.36 | − 1.01 – 1.74 | .60 | 0.26 | − 1.19 – 1.70 | .72 | 0.27 | − 1.10 – 1.65 | .70 | |||
Middle Category × time | − 0.05 | − 0.26 – 0.16 | .62 | 0.01 | − 0.20 – 0.24 | .89 | − 0.05 | − 0.27 – 0.16 | .62 | |||
.79 | .69 | .79 | ||||||||||
High Category × time | 0.02 | − 0.22 – 0.25 | .89 | 0.10 | − 0.15 – 0.35 | .43 | 0.01 | − 0.22 – 0.26 | .90 |
BVRT | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Model 1 |
Model 2 |
Model 3 |
||||||||||
β | 95% CI | P | Overall | β | 95% CI | P | Overall | β | 95% CI | P | Overall | |
Time | 0.43 | − 0.04 – 0.90 | .07 | 0.62 | 0.05 – 1.19 | .03 | 0.42 | − 0.05 – 0.89 | .08 | |||
Middle Category (MeDi 4–5), baseline testing | 0.04 | − 0.19 – 0.29 | .71 | − 0.002 | − 0.25 – 0.25 | .98 | 0.05 | − 0.19 – 0.29 | .70 | |||
.90 | .98 | .89 | ||||||||||
High Category (MeDi 6–9), baseline testing | 0.06 | − 0.21 – 0.33 | .66 | 0.02 | − 0.26 – 0.30 | .91 | 0.06 | − 0.21 – 0.33 | .65 | |||
Middle Category × time | 0.03 | − 0.03 – 0.10 | .30 | 0.042 | − 0.02 – 0.11 | .20 | 0.03 | − 0.03 – 0.09 | .31 | |||
.48 | .41 | .50 | ||||||||||
High Category × time | 0.002 | − 0.07 – 0.07 | .94 | 0.017 | − 0.06 – 0.09 | .65 | 0.002 | − 0.07 – 0.07 | .95 |
Abbreviations: MeDi Mediterranean diet; MMSE Mini-Mental State Examination; IST Isaacs Set Test; BVRT Benton Visual Retention Test; FCSRT Free and Cued Selective Reminding Test; CI Confidence Intervals
Linear mixed models: the beta coefficient for MeDi score represents the association of MeDi with baseline mean cognitive scores and the beta coefficient for the MeDi interaction with time (i.e., the slope) represents the association of MeDi with the change in cognitive scores over time. For MMSE × time, a negative beta coefficient indicates that a higher MeDi score was associated with fewer MMSE errors over time, whereas for FCSRT, IST and BVRT, a positive beta coefficient indicates slower decline in cognitive performance.
Model 1: Model adjusted for age, sex, education, marital status, total energy intake, practice of physical exercise, taking ≥ 5 drugs/d, Center for Epidemiological Studies-Depression scale score and Apolipoprotein E genotype and their interaction with time
Model 2: Model 1 + additional adjustment for Body Mass Index, hypertension, hypercholesterolemia, diabetes and tobacco and their interaction with time
Model 3: Model 1 + additional adjustment for stroke and its interaction with time