TABLE 4.
HRs (95% CIs) of Alzheimer disease (AD) events over 26 y of follow-up based on a 5-y cutoff between AD diagnosis and updated flavonoid intake data1
| Flavonoid intake percentile category | |||||
|---|---|---|---|---|---|
| Flavonoid class | ≤15th | >15th to 30th | >30th to 60th | >60th | P-trend2 |
| Flavonols | (n3 = 419/32) | (n = 421/29) | (n = 840/53) | (n = 1120/44) | |
| Model 14 | 1.00 (ref.) | 0.94 (0.56, 1.57) | 0.90 (0.57, 1.43) | 0.54 (0.33, 0.90) | 0.005 |
| Model 25 | 1.00 (ref.) | 0.87 (0.51, 1.47) | 0.84 (0.53, 1.35) | 0.47 (0.28, 0.79) | 0.001 |
| Model 36 | 1.00 (ref.) | 0.90 (0.52, 1.55) | 0.86 (0.52, 1.43) | 0.50 (0.28, 0.88) | 0.004 |
| Flavones | (n = 419/27) | (n = 421/15) | (n = 838/43) | (n = 1122/73) | |
| Model 14 | 1.00 (ref.) | 0.63 (0.34, 1.19) | 0.65 (0.40, 1.05) | 0.83 (0.52, 1.34) | 0.96 |
| Model 25 | 1.00 (ref.) | 0.64 (0.34, 1.20) | 0.60 (0.37, 1.00) | 0.79 (0.49, 1.29) | 0.93 |
| Model 36 | 1.00 (ref.) | 0.67 (0.36, 1.28) | 0.63 (0.38, 1.06) | 0.86 (0.51, 1.45) | 0.84 |
| Flavanones | (n = 418/22) | (n = 422/15) | (n = 840/35) | (n = 1120/86) | |
| Model 14 | 1.00 (ref.) | 0.54 (0.28, 1.04) | 0.53 (0.31, 0.91) | 0.85 (0.52, 1.39) | 0.28 |
| Model 25 | 1.00 (ref.) | 0.53 (0.27, 1.03) | 0.51 (0.29, 0.88) | 0.81 (0.49, 1.34) | 0.39 |
| Model 36 | 1.00 (ref.) | 0.55 (0.28, 1.08) | 0.53 (0.30, 0.94) | 0.85 (0.51, 1.44) | 0.33 |
| Flavan-3-ols | (n = 419/29) | (n = 421/30) | (n = 840/39) | (n = 1120/60) | |
| Model 14 | 1.00 (ref.) | 1.10 (0.65, 1.85) | 0.71 (0.43, 1.16) | 0.76 (0.47, 1.22) | 0.23 |
| Model 25 | 1.00 (ref.) | 1.06 (0.63, 1.80) | 0.62 (0.37, 1.04) | 0.68 (0.41, 1.11) | 0.13 |
| Model 36 | 1.00 (ref.) | 1.01 (0.58, 1.74) | 0.61 (0.36, 1.03) | 0.69 (0.41, 1.14) | 0.21 |
| Anthocyanins | (n = 420/38) | (n = 420/28) | (n = 839/58) | (n = 1121/34) | |
| Model 14 | 1.00 (ref.) | 0.59 (0.36, 0.97) | 0.62 (0.40, 0.95) | 0.28 (0.17, 0.46) | <0.001 |
| Model 25 | 1.00 (ref.) | 0.55 (0.33, 0.90) | 0.60 (0.39, 0.93) | 0.23 (0.13, 0.38) | <0.001 |
| Model 36 | 1.00 (ref.) | 0.51 (0.31, 0.86) | 0.54 (0.34, 0.86) | 0.20 (0.11, 0.36) | <0.001 |
| Flavonoid polymers | (n = 419/26) | (n = 421/30) | (n = 840/45) | (n = 1120/57) | |
| Model 14 | 1.00 (ref.) | 1.02 (0.59, 1.74) | 0.74 (0.45, 1.24) | 0.68 (0.41, 1.14) | 0.10 |
| Model 25 | 1.00 (ref.) | 1.00 (0.58, 1.71) | 0.65 (0.39, 1.10) | 0.61 (0.36, 1.02) | 0.05 |
| Model 36 | 1.00 (ref.) | 0.98 (0.56, 1.72) | 0.66 (0.38, 1.14) | 0.64 (0.37, 1.10) | 0.11 |
| Total flavonoids | (n = 420/24) | (n = 420/22) | (n = 840/51) | (n = 1120/61) | |
| Model 14 | 1.00 (ref.) | 0.80 (0.45, 1.45) | 0.94 (0.57, 1.57) | 0.78 (0.46, 1.32) | 0.40 |
| Model 25 | 1.00 (ref.) | 0.77 (0.42, 1.39) | 0.85 (0.50, 1.43) | 0.68 (0.39, 1.16) | 0.19 |
| Model 36 | 1.00 (ref.) | 0.72 (0.39, 1.34) | 0.87 (0.51, 1.50) | 0.71 (0.40, 1.25) | 0.31 |
Mean follow-up time of 19.7 y.
P values for the test of linear trend across percentile categories were based on linear regression models with the median intake of each percentile category assigned to individuals with intake in that category, and this percentile median variable was used as a continuous measure in regression models.
n represents the number of participants/AD events per percentile category for model 1.
Model 1 adjusted for age, sex, education, apoE ε4, and total energy intake; total n = 2800; n AD events = 158.
Model 2 adjusted for model 1 covariates + physical activity index, smoking status, BMI, prevalent stroke, diabetes, hypertension, and hypercholesterolemia; total n = 2796; n AD events = 155.
Model 3 adjusted for model 1 and model 2 covariates + overall dietary quality, vitamin and mineral supplement use, and dietary intakes of alcohol, ω-3 fatty acids (EPA and DHA), lutein, and zeaxanthin; total n = 2794; n AD events = 153.