The modified relative dose response (MRDR) test distinguishes between vitamin A (VA) deficiency and sufficiency using the molar ratio of 3,4-didehydroretinol to retinol (DR:R) in serum 4–6 h after administering 3,4-didehydroretinyl acetate (DRA). Research in sows demonstrated that serum and milk DR:R are correlated. Two studies determined VA status in lactating women and investigated breast milk as a surrogate for serum in the MRDR test in VA-adequate women. A secondary outcome compared serum with milk carotenoids in US women. Lactating US (24–40 years old, n = 25) and Indonesian (22–40 years old, n = 18) women were given 8.8 µmol DRA. Subjects were allocated to three collections (3–48 h post-dose) for blood and milk ( n = 4–10/time point). DR, retinol, and carotenoids were determined by HPLC. Serum DR:R time-courses were evaluated in all women and DR kinetics analyzed by noncompartmental modeling in the US and VA-deficient Indonesian women. Indonesian women had a higher proportion of VA deficiency by MRDR (61%) than US women (0%). Milk DR concentration was higher than serum. In US women, serum and milk DR:R were correlated at 5 h ( r = 0.86, P < 0.01) but not at 6 or 7 h. Serum DR kinetics ( tmax = 8, t1/2 = 15.3 h) were similar to those in VA-adequate lactating sows ( tmax = 7.5 ± 1.9, t1/2 = 17.4 ± 4.5 h) but varied in milk (human: tmax = 12, t1/2 = 22.4 h; sow: tmax = 14.4 ± 9.8, t1/2 = 71.8 ± 51.2 h). Carotenoids in milk differed from serum ( P < 0.001), with β-cryptoxanthin enrichment. Serum but not milk VA kinetics were similar between women and sows. Milk DR:R may represent a simpler VA biomarker in lactating women vulnerable to VA deficiency, but requires further validation. Impact statement Vitamin A (VA) deficiency is a major health issue globally, and lactating women are particularly vulnerable due to increased needs for milk production. Accurate detection of VA deficiency is important; however, most population surveys measure VA status using serum retinol, which is affected by inflammation and lacks sensitivity. The modified relative dose response (MRDR) test qualitatively distinguishes between VA deficiency and sufficiency and could improve population surveys if completed in a randomly selected subsample of individuals in surveys. The original relative dose response test required two blood samples, while MRDR requires only one, a significant improvement in accessibility of the technique by decreasing burden on subjects and investigators. This work demonstrates significant deficiency in Indonesian women compared with US women. In combination with previous research using lactating sows, these human data support milk as a surrogate for blood in the MRDR, which may be less invasive, but requires further validation.
Validity of the relative-dose-response test and the modified-relative-dose-response test as indicators of vitamin A stores in liver
