Abstract
SIGNIFICANCE: Despite the numerous benefits of breastfeeding, many mothers quit breastfeeding early due to understudied lactation problems. Optical and photonic techniques provide promising, objective, and non-invasive methods for human milk analysis but require accurate characterization and modeling of light scattering in human milk. AIM: We aimed to describe the influences of sample-specific properties on the forward scattering coefficient of human milk samples. APPROACH: We composed an experimental dataset of 50 human milk samples, containing the scattering coefficient, fat concentration, serum refractive index, milk fat globule (MFG) refractive index, MFG size distribution, and refractive index and size distribution of non-fat particles per sample. We reconstructed scattering coefficients using Mie theory with sample-specific properties and compared them to experimental values. RESULTS: This study confirmed a moderate correlation between experimental and reconstructed scattering coefficients when only the sample-specific fat concentration was considered (Spearman ρ = 0.62 ). Including the sample-specific MFG size distribution significantly improved the correlation ( ρ = 0.82 ), while the sample-specific scattering contribution of non-fat particles and refractive index of serum and MFGs had a negligible influence. CONCLUSIONS: The sample-specific fat concentration and MFG size distribution are the most important parameters to consider when developing light scattering models of human milk.