The present study describes compound-specific δ13C and δ15N analyses of fatty acids and amino acids for improving the accurate authentication of organic milk (OM) against conventional milk (CM) collected in Korea. Most δ13Cfatty-acid and δ13Camino-acid values were lower in OM than in CM ( P < 0.05); however, most δ15Namino-acid values displayed weak discriminative power for OM authentication. Higher isotopic fractionation was observed in δ13Cfatty-acid than in δ13Camino-acid and δ15Namino-acid, with fractionation trends differing with individual amino acids. In particular, δ13Clinoleic-acid of -33.5‰ and δ13Cmyristic-acid of -28‰ were determined to be promising year-round threshold values for Korean OM authentication. The δ13Cbulk was highly correlated with δ13CAla ( r = 0.92) and δ13Coleic-acid, trans ( r = 0.77), and strong positive correlations were observed between δ13CVal and δ13CIle ( r = 0.98) and between δ15NThr and δ15NSer ( r = 0.90). Chemometric modeling for OM authentication produced a high quality model ( R2 X = 0.547, R2 Y = 0.865, and Q2 = 0.689) with reliable chemical markers, notably δ13Cmyristic-acid, δ13Clinoleic-acid, and δ13Cstearic-acid. Furthermore, the models developed for seasonal separation in OM ( Q2 = 0.954) and CM ( Q2 = 0.791) were of good quality. Our findings, based on compound-specific isotope data, improve the reliability of OM authentication in cases where bulk stable isotope ratio analysis alone is insufficient. They also provide valuable insight into the control of fraudulent OM labeling in Korea, with potential application in other countries.
Keywords: compound-specific isotope analysis; isotope fingerprinting; milk; multivariate model; organic authentication.