Background: Breast milk iodine concentration (BMIC) may be an indicator of iodine status during lactation, but there are few data comparing different analytical methods or timing of sampling. The aims of this study were: (i) to optimize a new inductively coupled plasma mass spectrometry (ICP-MS) method; and (ii) to evaluate the effect of analytical method and timing of within-feed sample collection on BMIC.
Methods: The colorimetric Sandell-Kolthoff method was evaluated with (a) or without (b) alkaline ashing, and ICP-MS was evaluated using a new (129)I isotope ratio approach including Tellurium (Te) for mass bias correction (c) or external standard curve (d). From iodine-sufficient lactating women (n = 97), three samples were collected within one breast-feeding session (fore-, mid-, and hind-feed samples) and BMIC was analyzed using (c) and (d).
Results: Iodine recovery from NIST SRM1549a whole milk powder for methods (a)-(d) was 67%, 24%, 105%, and 102%, respectively. Intra- and inter-assay coefficients of variation for ICP-MS comparing (c) and (d) were 1.3% versus 5.6% (p = 0.04) and 1.1% versus 2.4% (p = 0.33). The limit of detection (LOD) was lower for (c) (0.26 μg/kg) than it was for (d) (2.54 μg/kg; p = 0.02). Using (c), the median [95% confidence interval (CI) obtained by bootstrap] BMIC (μg/kg) in foremilk (179 [CI 161-206]) and in mid-feed milk (184 [CI 160-220]) were not significantly different (p = 0.017), but were higher than in hindmilk (175 [CI 153-216]; p < 0.001). In foremilk using (d), BMIC was 199 ([CI 182-257]; p < 0.001 vs. (c)). The variation in BMIC comparing (c) and (d) (13%) was greater than variation within feeding (5%; p < 0.001).
Conclusions: Because of poor recoveries, (a) and (b) should not be used to measure BMIC. Compared with (d), (c) has the advantages of higher precision and a lower LOD. In iodine-sufficient women, BMIC shows low variation within a breast-feeding session, so timing of sampling is not a major determinant of BMIC.