Identificação precoce de predisposição à alergia à proteína do leite de vaca e ganho ponderal por espectroscopia no infravermelho.

Abstract

Introduction: Food allergy comprises hypersensitivity reactions to dietary components, most often proteins. In this group, cow's milk protein allergy (CMPA) is a frequent diagnosis in infants and children with an overall prevalence of 2-7.5%. It presents a wide range of presentations with gastrointestinal, cutaneous, systemic and respiratory manifestations, with non-specific symptoms or overlapping with other pathologies. The diagnosis is based on the oral feeding challenge and is performed when the patient is already symptomatic, delaying possible therapeutic interventions. Thus, it is important to research earlier diagnostic methods in order to anticipate conduct and improve the quality of life of affected patients. Objective: To evaluate the applicability of molecular absorption spectroscopy in infrared with Fourier transform as a method of identifying cases with predisposition to CMPA. Article Abstract I: Meconium FTIR: a new approach for early identification of cow's milk protein allergy? Objective: To evaluate the potential of infrared spectroscopy (FTIR) in meconium samples to identify cases with a predisposition to CMPA. Method: A prospective longitudinal cut study in which 67 newborns were initially evaluated using a questionnaire with prenatal and birth data, in addition to the collection of meconium samples and FTIR analysis. Patients were followed up for 6 months and assessed for the development of symptoms of APLV using the CoMiSSTM® questionnaire. Results: CoMiSSTM scores ranged from 0 to 20 with a mean of 7 (GI ≤ 7, n = 61; GII > 7, n = 6). Significant differences in the spectral profile between the groups occurred in the bands of 2000-1700, 1650-1500, 1300-1250 and 850-800 cm-1. The best discrimination occurred at 3016 cm-1 (p = 0.0013) and 905 cm-1 (p = 0.0037). An ATR/OPLS model elaborated with spectral data transformed by the first derivative, in the range 4000-650 cm-1, 1 orthogonal signal correction component and 2 latent variables made it possible to predict the scores with RMSECV < 0.1 and determination coefficient R > 0.999. Conclusions: The meconium ATR/FTIR technique associated with the OPLS algorithm makes it possible to predict CoMiSSTM scores in infants up to six months of age and presents itself as a promising method for early identification of predisposition to the development of symptoms for CMPA, which may in the future compose point technology of care for neonatal screening of patients. Abstract of Article II: Prediction of weight outcomes in children in the first three months of life by meconium FTIR: A pilot study. Objective: To evaluate the ability of meconium Fourier transform infrared spectroscopy (FTIR) as a method of assessing weight gain in newborns in the first three months of life. Methods: Prospective longitudinal pilot study with newborns categorized into different weight outcomes in the first three months of life according to final weight (P23) and relative weight gain (GP23%) and average weekly weight gain (GP23m). Meconium samples were collected, with preference for the child's first evacuation and analyzed according to metabolic fingerprints by Attenuated Total Reflection (ATR). Chemometric models were developed by multivariate regression via partial least squares with orthogonal projections (ATR/OPLS) correlating spectral information with newborn weight outcomes. Results: Trend measures of P23, GP23% and GP23m reached 5747 ± 985 g, 71.8 ± 31.8 % and 32.4 ± 10.8 g/week, respectively, in 53 children investigated. The 2nd derivative treatment indicated that several wave numbers contributed to significantly discriminate children with low and high responses to the investigated weight parameters. The transformation of the data with the 1st derivative followed by the application of 1 component of orthogonal signal generated ATR/OPLS models with 2 latent variables with high robustness. Root-mean cross-validation error square values (RMSECV) equal to 53 g, 0.11 % and 0.06 g/week and correlation coefficients (Rcv) > 0.998 were obtained for the prediction of P23, GP23% and GP23m, respectively. Conclusions: The results obtained in the study demonstrate the potential use of FTIR spectral data in meconium samples to predict weight gain in newborns, enabling early interventions in children with a predisposition to low weight gain.