Intestinal β-carotene bioconversion in humans is determined by a new single-sample, plasma isotope ratio method and compared with traditional and modified area-under-the-curve methods

Abstract

The vitamin A value (bioefficacy) of provitamin A carotenoids is determined by absorption of the carotenoid (bioavailability) and its subsequent conversion to retinol (bioconversion). Here we show that intestinal bioconversion of β-carotene can be estimated based on analysis of a single plasma sample collected 6 h after subjects ingest a test dose of stable isotope-labeled β-carotene from the ratio of retinyl esters to retinyl esters plus β-carotene. Plasma isotope ratio predictions of bioconversion ranged from 50 to- 93% (mean 76%) for 45 healthy young adults with low vitamin A stores. Results were the same as predictions made by a traditional area-under-the-curve method calculated from 0 to- 8 h or a modified area-under-the-curve method calculated from 0 to- 12 h. The modified method may provide better estimates of bioconversion between 8 and 24 h after ingestion of a carotenoid dose when stable isotopes cannot be used due to cost or logistics. Furthermore, because the plasma isotope ratio method requires only one blood sample and no isolation of triglyceride-rich lipoproteins, its use will facilitate estimation of provitamin A carotenoid bioconversion in human subjects and especially children, in whom repeated blood sampling is not feasible.

Keywords: Area under the curve; Carotenoid bioconversion; Carotenoid bioefficacy; Vitamin A; β-Carotene.

Fig. 1

Observed data for concentration of tracers in plasma versus time. Shown are group mean μmol/L ± SEM (n = 45) for [¹³C₁₀]βC (circles connected by a solid line) and for [¹³C₅]RE_βCe (squares connected by a dashed line) (panel A). Arrows indicate when lunch and dinner were provided. Also shown are μmol/L versus time for [¹³C₁₀]βC (circles connected by a solid line), for [¹³C₅]RE_βCe derived from the βC dose (squares connected by a dashed line), and for the sum of these tracers (reflecting total absorbed [¹³C₁₀]βC; triangles and dotted lines) for subjects with high (85%; panel B) and low intestinal bioconversion (55%; panel C). Panel C inset shows tracer concentrations to 14 h for the subject with low bioconversion. Subjects whose data are illustrated in panels B and C were selected based on bioconversion calculated by AUC using data from 0 to -8 h. AUC, area under the curve; βC, β-carotene; RE_βCe, retinyl esters as molar β-carotene equivalents.

Fig. 2

Observed data for plasma tracer concentrations versus time. Shown are mean μmol/L ± SEM for [¹³C₁₀]βC (circles connected by a solid line) and for [¹³C₅]RE_βCe (squares connected by a dashed line) for subjects with the highest RE concentrations at 4 (n = 16), 6 (n = 22), 8 (n = 3) or 10 (n = 4) h (panels A–D, respectively). βC, β-carotene; RE_βCe, retinyl esters as molar β-carotene equivalents.

Fig. 3

Bioconversion estimated by an IR method and two AUC methods. Intestinal βC bioconversion calculated by the IR at 6 h versus AUC at 8 h is shown in panel A; IR at 6 h versus AUC_adj at 12 h is presented in panel C; n = 45. See text for methods of calculation. Least squares regression lines were y = 1.0x – 3.2 (R² = 0.91; P < 0.0001) (panel A) and y = 1.0x – 0.62 (R² = 0.95; P < 0.0001) (panel C). Rank-order correlations are shown in panel B for AUC (R = 0.96, P < 0.0001) and in panel D for AUC_adj (R = 0.98, P < 0.0001). AUC, area under the curve; AUC_adj, adjusted AUC; βC, β-carotene; IR, isotope ratio.

Fig. 4

Observed and estimated data for concentration of tracers in plasma versus time. Shown are group mean ± SEM (n = 45) observed μmol/L for [¹³C₁₀]βC (circles and solid line) and simulated data for βC in chylomicrons (squares and dashed line) versus VLDL (triangles and dotted line). Arrows indicate when lunch and dinner were provided. βC, β-carotene; VLDL, very low density lipoproteins.