Skin total carotenoids predict plasma carotenoid levels during a 28‐week experimental feeding study with varying levels of vegetables and fruit

Abstract

Objective biomarkers are needed to assess adherence to vegetable and fruit intervention trials. This study compared plasma carotenoid concentrations to non‐invasive skin carotenoid assessments. Thirty participants consumed a low‐carotenoid diet (6 wk, Phases 1 & 3), a provided diet containing 6 c/day vegetables and fruit (8 wk, Phase 2) and their usual diet (final 8 wks, Phase 4). Skin was measured 5x/wk during Phase 2 and at least 2x/wk for the duration. Plasma carotenoids were measured at baseline and mid and end of each phase.Skin and plasma carotenoid levels decreased (p<0.01) by a third from baseline to end of phase 1, then increased (p<0.01) over 300% at end of phase 2. Plasma carotenoids returned to baseline concentrations at the end of phase 3 and skin carotenoid levels by end of phase 4.At baseline, skin and plasma total carotenoid values were significantly correlated (p <0.01). Skin carotenoid status significantly predicted plasma values using a mixed linear model including all 9 time points (p < 0.001), indicating that changes in skin carotenoid status closely follow changes in plasma across a broad range of intakes. At the individual level, skin carotenoids predicted plasma values as strongly (p < 0.001).This study supports the use of skin carotenoid status as an objective indicator for interventions involving carotenoids or fruit/vegetable intake. Support: USDA‐ARS