Table 3.
Association between serum interleukin-6 concentrations and flavonol intake during the trial, stratified by baseline interleukin-6 concentrations (n = 872)
| Interleukin-6 (T0) Tertiles* (in pg/mL) | Flavonol Intake Quartiles (in mg/d)* | ||||
|---|---|---|---|---|---|
| Q1: < 21.1 | Q2: 21.1–29.6 | Q3: 29.7–40.0 | Q4: >40.0 | ||
| Median (IQR) | Median (IQR) | Median (IQR) | Median (IQR) | P-value‡ | |
| IL-6 (T0) <1.53 * | |||||
| Sample size | 65 | 71 | 79 | 76 | |
| Trial (T1, 3‡) | 1.47 (1.15–1.84) | 1.49 (1.19–1.90) | 1.35 (1.01–1.72) | 1.44 (1.07–1.75) | 0.22 |
| Change (T1,3–T0‡) | 0.34 (0.00–0.58) | 0.36 (0.02–0.69) | 0.20 (−0.05–0.49) | 0.30 (0.01–0.58) | 0.15 |
| IL-6 (T0) 1.53–2.37 * | |||||
| Sample size | 72 | 73 | 73 | 73 | |
| Trial (T1, 3‡) | 2.07 (1.68–2.52) | 2.12 (1.79–2.69) | 1.93 (1.62–2.49) | 1.90 (1.52–2.50) | 0.29 |
| Change (T1,3–T0‡) | 0.22 (−0.17–0.51) | 0.33 (−0.22–0.76) | 0.12 (−0.31–0.49) | −0.08 (−0.37–0.42) | 0.13 |
| IL-6 (T0) >2.37 * | |||||
| Sample size | 81 | 74 | 66 | 69 | |
| Trial (T1,3‡) | 3.39 (2.45–4.66) | 3.14 (2.15–4.90) | 2.72 (2.09–4.11) | 2.63 (1.77–3.56) | 0.01 |
| Change (T1,3–T0‡) | −0.26 (−1.15–0.66) | −0.90 (−2.31–0.26) | −0.76 (−1.91–−.15) | −0.75 (−1.65–0.01) | 0.05 |
Participants were grouped in tertiles by their baseline IL-6 concentrations (T0) and in quartiles (Q1–Q4) by mean flavonol intake during the first 3 trial years.
P values for differences in medians among the flavonol intake quartiles within baseline IL-6 tertiles were calculated based on the Kruskal-Wallis test. The P values for trend using a multiple regression model adjusting for age tertiles (<58, 58–66, >66 yrs), sex, average BMI (<25, 25.0–29.9, ≥30 kg/m2), current smoking status, and average energy intake (continuous) during the first 3 trial years were 0.008 for Trial (T1,3) and 0.02 for Change (T1,3–T0) for the highest baseline IL-6 tertile.
Trial (T1,3) = geometric mean of year 1 and 3 values. Change (T1,3–T0) = T1,3 minus baseline values.