(P14-050-23) Associations of Self-Reported Fruit and Vegetable Intake With Serum and Red Blood Cell Carotenoids

Objectives: Traditional methods for dietary assessment can be burdensome and their accuracies may be compromised due to self-reporting. Serum carotenoids are an objective and widely used biomarker for fruits and vegetables (FV) intake. Notably, red blood cells (RBC) also contain carotenoids, and potentially may be a better indicator of long-term carotenoids intake than serum carotenoids. However, RBC carotenoids have not yet been investigated as biomarker of FV intake. Studies on RBC carotenoid changes in response to β-carotene supplementation show mixed results. The aim of this study was to determine the relative strength of the associations of FV intake with serum vs. RBC total carotenoids.

Methods: We examined the associations between FV intake (servings/d) over the past 1 y measured by the Block FFQ and serum and RBC carotenoids measured by HPLC. Data from a previous study in our lab in 97 healthy, non-smoking adults (41M/56F; aged 29.8±12.4 y, BMI 24.5±4.0 kg/m²) were used. Variables not normally distributed were log-transformed prior to analysis.

Results: The median (Q1, Q3) FV servings was 3.9 (1.8, 5.5) per d. Median (Q1, Q3) serum and RBC carotenoids were 132 (99, 170) and 72 (49, 127) µmol/L, respectively. Serum and RBC carotenoids were significantly and positively associated (β=0.499; p=0.003). Significant and positive associations were observed between total serum carotenoids and fruit (β=0.134; p=0.026), vegetable (β=0.193; p=0.003), and FV (β=0.229; p=0.002) intake. After controlling for covariates (age, sex, race, percentage body fat, percent energy intake from alcohol), associations between total serum carotenoids and fruit (β=0.122; p=0.030) and also FV (β=0.158; p=0.033) intake remained significant, though the association between vegetable servings and total serum carotenoids became marginally nonsignificant (β=0.127; p=0.058). Associations between RBC carotenoids and fruit (β=0.164; p=0.107), vegetable (β=0.165; p=0.171), and FV (β=0.213; p=0.111) intake were non-significant (values with covariates shown).

Conclusions: RBC carotenoids may not be as strong a biomarker of FV intake assessed by FFQ as compared to serum carotenoids. Further studies are needed to examine individual carotenoids and other assessment methods (e.g., 24-h dietary recall) in larger sample sizes.

Funding Sources: NIDDK and Purdue University