Noninvasive Reflection Spectroscopy Measurement of Skin Carotenoid Score in Infants Is Feasible and Reliable

J Nutr. 2023 Jan 14;152(12):2966-2977. doi: 10.1093/jn/nxac182.


Background: Skin carotenoid measurement by reflection spectroscopy (RS) offers a noninvasive biomarker of carotenoid intake, but feasibility, reliability, and validity are not established in infants.

Objectives: In this study we aimed to determine the feasibility and reliability of 4-mo-old infant skin carotenoid score (SCS) measurement and its correlation with total carotenoid intake and plasma concentrations.

Methods: SCSs were measured in a prospective, observational study with a modified, portable RS device at the index finger and heel of the foot in 4-mo-olds (n = 21). Infant plasma, human milk, and formula carotenoid concentrations were measured by HPLC-photodiode array, and carotenoid intake was estimated from 7-d food diaries corrected for actual milk carotenoid content. Mean SCS, time to acquire measurements, replicate intraclass correlations, and bivariate correlations between SCS, carotenoid intake, and plasma carotenoids were examined. Exploratory analyses of returning 6- (n = 12) and 8-mo-old (n = 9) infants were conducted.

Results: Mean ± SD finger and heel SCSs in 4-, 6-, and 8-mo-olds were 92 ± 57 and 92 ± 51; 109 ± 41 and 119 ± 44; and 161 ± 89 and 197 ± 128 units, respectively. Replicate SCS measurements were reliable, with high intraclass correlation (≥0.70) of within-subject visit measurements. Finger SCSs in 4-mo-olds were correlated with carotenoid intake (ρ = 0.48, P = 0.0033), and finger and heel SCS were correlated with total plasma carotenoid concentrations (ρ = 0.71, P < 0.0001 and ρ = 0.57, P = 0.0006, respectively). Eight-mo-olds' finger and heel SCSs were correlated with total carotenoid intake (ρ = 0.73, P < 0.001; ρ = 0.58, P = 0.0014, respectively), whereas SCSs in 6-mo-olds, in transition from exclusive milk to complementary feeding, did not correlate with plasma carotenoid or dietary carotenoids, despite correlation between plasma and dietary carotenoid intake (ρ = 0.86, P = 0.0137). Mixed models suggest plasma total carotenoid concentration, age, carotenoid intake, and age × carotenoid intake, but not measurement site, are determinants of infant SCS.

Conclusions: Infant skin carotenoids are feasibly and reliably measured by RS and may provide a biomarker of carotenoid intake in 4-mo-olds. This trial was registered at as NCT03996395.

Keywords: Veggie Meter™; human milk; infants; nutrition assessment; nutritional biomarkers; reflection spectroscopy.

Publication types

  • Observational Study
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

  • Biomarkers
  • Carotenoids*
  • Diet
  • Humans
  • Infant
  • Prospective Studies
  • Reproducibility of Results
  • Spectrum Analysis, Raman* / methods


  • Carotenoids
  • Biomarkers

Associated data