Table 2.
Crude and adjusted indicators of reliability and inter- and intra-individual variability for each antioxidant biomarker in premenopausal women a,b.
| Biomarkers of Antioxidant Status | n c | ICCunadj [95% CI] | CVG | CVT | II | RCV%,pos | RCV%,neg | nβ,10% | nβ,20% | ICCadj [95% CI] d |
|---|---|---|---|---|---|---|---|---|---|---|
| Serum bilirubin (µmol/L) | 42 | 0.567 [0.380, 0.710] | 0.343 | 0.299 | 0.874 | ±83.0 | 7 | 3 | 0.558 [0.393, 0.737] | |
| Serum uric acid (µmol/L) e | 44 | 0.693 [0.540, 0.801] | 0.170 | 0.113 | 0.663 | +36.6 | −26.8 | 4 | 2 | 0.682 [0.535, 0.810] |
| Plasma vitamin C (µmol/L) e | 44 | 0.515 [0.326, 0.668] | 0.127 | 0.123 | 0.970 | +40.6 | −28.9 | 9 | 4 | 0.529 [0.344, 0.704] |
| Plasma α-tocopherol (µmol/L) e | 44 | 0.700 [0.548, 0.806] | 0.169 | 0.110 | 0.652 | +35.6 | −26.2 | 4 | 2 | 0.541 [0.360, 0.711] |
| Plasma γ-tocopherol (nmol/L) e | 43 | 0.628 [0.442, 0.753] | 0.409 | 0.310 | 0.757 | +131.6 | −56.8 | 6 | 3 | 0.549 [0.377, 0.724] |
| Plasma retinol (nmol/L) e | 44 | 0.514 [0.325, 0.667] | 0.191 | 0.186 | 0.972 | +66.7 | −40.0 | 9 | 4 | 0.435 [0.239, 0.633] |
| Plasma α-carotene (nmol/L) e | 40 | 0.793 [0.673, 0.869] | 0.547 | 0.266 | 0.487 | +106.4 | −51.5 | 3 | 1 | 0.783 [0.676, 0.885] |
| Plasma β-carotene (nmol/L) e | 41 | 0.875 [0.796, 0.924] | 0.617 | 0.217 | 0.352 | +81.2 | −44.8 | 2 | 1 | 0.832 [0.747, 0.906] |
| Plasma lutein (nmol/L) e | 43 | 0.681 [0.509, 0.792] | 0.336 | 0.227 | 0.675 | +86.1 | −46.3 | 5 | 2 | 0.414 [0.224, 0.627] |
| Plasma zeaxanthin (nmol/L) e | 44 | 0.514 [0.325, 0.668] | 0.333 | 0.323 | 0.316 | +139.7 | −58.3 | 9 | 4 | 0.421 [0.222, 0.621] |
| Plasma β-cryptoxanthin (nmol/L) e | 41 | 0.758 [0.626, 0.846] | 0.378 | 0.209 | 0.552 | +77.2 | −43.6 | 3 | 2 | 0.781 [0.669, 0.877] |
| Plasma lycopene (nmol/L) e | 44 | 0.699 [0.547, 0.805] | 0.385 | 0.248 | 0.644 | +96.9 | −49.2 | 4 | 2 | 0.447 [0.251, 0.643] |
| Serum coenzyme Q10 (nmol/L) e | 44 | 0.760 [0.627, 0.846] | 0.300 | 0.166 | 0.554 | +58.1 | −36.7 | 3 | 2 | 0.578 [0.405, 0.739] |
| eCAT (kU/gHb) | 40 | 0.658 [0.493, 0.776] | 0.083 | 0.060 | 0.721 | ±16.7 | 5 | 2 | 0.602 [0.432, 0.771] | |
| eGPx (U/gHb) e | 40 | 0.765 [0.634, 0.850] | 0.236 | 0.130 | 0.549 | +43.1 | −30.1 | 3 | 2 | 0.801 [0.698, 0.892] |
| Blood glutathione (µmol/L) | 43 | 0.098 [0.000, 0.286] | 0.053 | 0.161 | 3.030 | ±44.6 | 83 | 37 | 0.072 [0.000, 0.313] | |
| Plasma TEAC (mmol/L) f | 43 | 0.000 [0.000, 0.175] | 0.000 | 0.070 | / | ±19.3 | / | / | 0.000 [0.000, 0.235] | |
Abbreviations: ICC, intraclass correlation coefficient; CVG, inter-individual variation; CVT, total intra-individual variation; II, index of individuality; RCV, reference change value; nβ,10%, number of required measurements to limit the attenuation in regression coefficient to 10%; nβ,20%, number of required measurements to limit the attenuation in regression coefficient to 20%; eCAT, erythrocyte catalase; Hb, hemoglobin; eGPx, erythrocyte glutathione peroxidase; TEAC, Trolox equivalent antioxidant capacity. a For each considered subject, three mean values of the respective blood biomarker were available. b Except for the adjusted ICC, variance components were estimated by restricted maximum likelihood method via one-way random effect model. c Sample size after excluding pre-identified outliers by Cochran’s test and Reed’s criterion. The original sample size was 44 subjects for all biomarkers except for eCAT and eGPx, for which the original sample size was 40 subjects. d Adjusted ICC was calculated using the variance components estimated by restricted maximum likelihood method via a linear mixed-effects model using the following fixed effects: time elapsed since the first day of the last menstrual period (d), body mass index (kg/m2), fasting duration (h), sleep duration of the last night (h), use of contraceptives (no vs. yes), serum cholesterol concentration (mmol/L) and physical activity level. In addition, energy intake (MJ/d) or, if applicable, nutrient intake levels (i.e., purine intake (mg/d) for serum uric acid, vitamin C intake (mg/d) for plasma vitamin C, tocopherol equivalents (mg/d) for both plasma tocopherol parameters, retinol equivalents (mg/d) for plasma retinol and β-carotene (mg/d) for plasma carotenoids) were included in the model, respectively. For each of these fixed effects, data from all three measuring time points were considered. Sample size was reduced by one subject because one subject provided no dietary record in the third assessment. e Biomarker values were logarithmically transformed (natural logarithm). f Model with a singular fit.