Background: Breast milk banks (BMBs) provide human donor breast milk to preterm infants when mothers are not able to provide breast milk themselves (Arslanoglu et al., 2010:20). Breastfeeding is the single most effective intervention to saving the lives of millions of children in developing countries (Bhutta & Labbok, 2011:378-380). BMBs form an integral part of the millennium development goals (MDGs) (Dempsey & Miletin, 2010:2) on reducing infant mortality and morbidity and are being implemented all over the world (Hartmann et al., 2007:667, Arslanoglu et al., 2010:20, Eidelman & Schanler 2012:827). Problem statement: BMBs can be established more effectively with an implementation tool in place. Currently, there is no tool available to guide the implementation of a BMB in South Africa; the only guidelines that could be found are those describing the operation of BMBs in other countries. Some of the BMBs already established in South Africa were implemented with the help of the South African Breast Milk Bank Reserve (SABR), but without a formalised guideline to provide implementation guidance. Aims and objectives: The purpose of this study was to adapt a current Kangaroo mother care (KMC) progress-monitoring tool and to adjust it according to the BMB setting in South Africa in order to provide a final suggested implementation tool for the implementation of BMBs. Method: An explorative, descriptive design with multiple phases was used. Different data sets were audited; including patient files, written reports, working files, research articles and policies. Observations were also made with regards to available equipment and designated BMB space. During phase one, a critical analysis was performed on research articles and websites regarding BMBs. The qualitative data was analysed by using content analysis. This information was adapted and contextualised in phase two. This phase entailed applying the adapted tool to the BMB setting of the selected regional hospital, which then led to phase three, in which the audited data was incorporated into the final suggested implementation tool resulting from this study. Results: The suggested tool, the North West Province BMB implementation tool, shared the six main constructs with the KMC progress-monitoring tool. The constructs are creating awareness, adopting the concept, taking ownership, evidence of practice, evidence of routine and integration and sustainable practice. However, some of the progress markers and instrument items were specifically relevant to the KMC setting and others to the BMB setting. On the other hand, some constructs overlapped, being relevant to both settings: for instance awareness by management, conscious decision to implement, mobilisation of resources (human, space and equipment), information about mothers’ other resources such as budget, patient records, staff orientation, evidence of discharge, policies and other written documents, and 1-2 year audit evidence and staff development. Conclusion: Adapting the KMC progress-monitoring tool led to the development of a suitable tool to guide BMB implementation. The tool is called the North West Province BMB implementation tool. This tool could be used to guide the implementation of a BMB in other hospitals in South Africa. / M (Dietetics), North-West University, Potchefstroom Campus, 2015

Background: Breast milk banks (BMBs) provide human donor breast milk to preterm infants when mothers are not able to provide breast milk themselves (Arslanoglu et al., 2010:20). Breastfeeding is the single most effective intervention to saving the lives of millions of children in developing countries (Bhutta & Labbok, 2011:378-380). BMBs form an integral part of the millennium development goals (MDGs) (Dempsey & Miletin, 2010:2) on reducing infant mortality and morbidity and are being implemented all over the world (Hartmann et al., 2007:667, Arslanoglu et al., 2010:20, Eidelman & Schanler 2012:827). Problem statement: BMBs can be established more effectively with an implementation tool in place. Currently, there is no tool available to guide the implementation of a BMB in South Africa; the only guidelines that could be found are those describing the operation of BMBs in other countries. Some of the BMBs already established in South Africa were implemented with the help of the South African Breast Milk Bank Reserve (SABR), but without a formalised guideline to provide implementation guidance. Aims and objectives: The purpose of this study was to adapt a current Kangaroo mother care (KMC) progress-monitoring tool and to adjust it according to the BMB setting in South Africa in order to provide a final suggested implementation tool for the implementation of BMBs. Method: An explorative, descriptive design with multiple phases was used. Different data sets were audited; including patient files, written reports, working files, research articles and policies. Observations were also made with regards to available equipment and designated BMB space. During phase one, a critical analysis was performed on research articles and websites regarding BMBs. The qualitative data was analysed by using content analysis. This information was adapted and contextualised in phase two. This phase entailed applying the adapted tool to the BMB setting of the selected regional hospital, which then led to phase three, in which the audited data was incorporated into the final suggested implementation tool resulting from this study. Results: The suggested tool, the North West Province BMB implementation tool, shared the six main constructs with the KMC progress-monitoring tool. The constructs are creating awareness, adopting the concept, taking ownership, evidence of practice, evidence of routine and integration and sustainable practice. However, some of the progress markers and instrument items were specifically relevant to the KMC setting and others to the BMB setting. On the other hand, some constructs overlapped, being relevant to both settings: for instance awareness by management, conscious decision to implement, mobilisation of resources (human, space and equipment), information about mothers’ other resources such as budget, patient records, staff orientation, evidence of discharge, policies and other written documents, and 1-2 year audit evidence and staff development. Conclusion: Adapting the KMC progress-monitoring tool led to the development of a suitable tool to guide BMB implementation. The tool is called the North West Province BMB implementation tool. This tool could be used to guide the implementation of a BMB in other hospitals in South Africa. / M (Dietetics), North-West University, Potchefstroom Campus, 2015

The Healthy People 2010 Objectives for the population of the United States include the objective for women to sustain breastfeeding at a higher rate than they are currently.The American Academy of Pediatrics (AAP) now recommends that all infants be fed human milk for the first year of life instead of the previously recommended first six months of life. Breastfeeding initiation rates are rising, but breastfeeding duration rates continue to fall. It was hypothesized that the support system of the women may play a role in the duration of breastfeeding. The purpose of this research study was to investigate the support system and breastfeeding rates of mothers during the first six weeks after delivery and to determine if income level had an effect on these systems and rates. Mothers were interviewed during their hospitalization for childbirth regarding their perception of their support system currently in place. A questionnaire was administered regarding support from friends, family members, co-workers, and health care professionals and the responses were quantified. Mothers were contacted weekly to determine if breastfeeding was being sustained and who was the most supportive person to them. At the end of six weeks or when breastfeeding was discontinued, the mothers were interviewed a second time to determine if the support system had changed. The results of this study showed that income level affected the duration of breastfeeding and the mother's perception of her support system. Additionally, the results showed that the evaluation of the support system changed only slightly over time. The results showed preliminary ability of the study tool to predict breastfeeding at six weeks by income.This document was created using Microsoft Word 2003. The statistical package used for data analysis was JMP version 6.

Le lait maternel est considéré comme l’aliment de référence pour le développement des nouveau-nés. Cependant dans le cas des grands prématurés qui ont des besoins nutritionnels particulièrement élevés, le lait maternel ne permet pas toujours d’apporter des quantités optimales pour certaines molécules, en raison de déficiences intrinsèques ou des conditions d’administration.Nous avons étudié les relations entre l’alimentation actuelle des mères et la composition en acides gras polyinsaturés de leur lait. Les laits maternels du Sud de la France et de la région Nord/Ouest présentent un déséquilibre de la composition en défaveur des acides gras polyinsaturés (AGPI) n-3 du lait, notamment en acide docosahexaénoïque (DHA). Ceci est en relation avec une consommation faible en poissons et produits de la mer, ainsi qu’une consommation importante d’AGPI n-6. Une deuxième partie du travail a porté sur la relation entre certaines molécules bioactives du lait et le développement intestinal des nouveau-nés. Le protocole de prise en charge nutritionnelle des nouveau-nés prématurés conduit à l’administration d’une faible quantité de molécules bioactives (AGPI n-3, phospholipides et plasmalogènes, sCD14, sphingomyélinase acide) ne permettant pas de couvrir les besoins, notamment en DHA, et probablement pas d’assurer une santé intestinale optimale pendant les premières semaines de vie. Un délai nécessaire pour atteindre la maturité digestive plus court est associé à la consommation d’AGPI pendant les deux premières semaines de vie, et de phospholipides (sphingomyéline et plasmalogènes) à partir de la deuxième semaine de vie. La consommation de sCD14 chez les enfants ne consommant que du lait de don est inversement corrélée au délai de maturité digestive. Nous avons constaté que cette molécule influence la réponse inflammatoire de cellules intestinales fœtales à une stimulation bactérienne chronique.Nous avons aussi étudié certaines propriétés physico-chimiques de laits d’espèces animales potentiellement intéressantes pour l’amélioration de la qualité des substituts ou des suppléments de lait et il apparait que le lait de chamelle présente des teneurs en sphingomyéline et plasmalogènes comparables à celles du lait humain. L’utilisation des traitements technologiques, comme la pasteurisation et l’homogénéisation, altère les propriétés du lait au niveau des teneurs en molécules bioactives (sCD14, sphingomyélinase acide) ainsi qu’au niveau de la structure du globule lipidique.En conclusion, ce travail ouvre la voie à une réflexion sur la formulation de nouveaux substituts et de suppléments de lait maternel à élaborer pour une meilleure prise en charge des nouveau-nés. / Human milk is considered as the gold standard to newborn development. However, in case of very preterm infants who have particularly high nutrients needs, it does not always provide optimal quantities of several bioactive molecules, due to composition or supply conditions.We investigated the relationship between current maternal nutrition and polyunsatured fatty acids (PUFA) composition of human milk. Human milk from mothers of South and North-West of France had imbalance PUFA composition to the detriment of n-3 PUFA, especially docosahexaenoic acid (DHA). This was linked to low consumption of fish and seafood products, and to an important n-6 PUFA intake.A second part of our work was focused on bioactive compounds of human milk and newborn intestinal development. Nutritional care of preterm infants leads to low intake of bioactive compounds (PUFA n-3, phospholipids and plasmalogens, sCD14, acid sphingomylinase), insufficient to cover the needs of HAD and to ensure optimal intestinal health during the first weeks of life. A shorter time to reach digestive maturity was associated with PUFA consumption during the first two weeks of life, and with phospholipids (sphingomyelin and plasmalogens) from the second week. sCD14 intake by preterm fed only with donor human milk was inversely correlated with digestive maturity time. This molecule influences the inflammatory response of fœtal intestinal cells to chronic bacterial stimulation.We had also investigated the physicochemical properties of milk from different animal species potentially interesting for improving infant formula or milk fortifier. Camel milk showed sphingomyelin and plasmalogen contents close to human milk. Technological treatments, such as pasteurisation or homogenisation, lead to alteration of milk properties, either in bioactive molecules content (sCD14, acid sphingomyelinase) or in fat globule structure.In conclusion, this work provides new possibilities for the conception of infant formula and fortifier to improve newborn care.

Objetivo: Descrever o crescimento e a evolução clínico laboratorial de recém-nascidos de muito baixo peso (RNMBP) alimentados com leite de banco de leite humano (BLH) segundo o valor calórico e protéico durante a internação e no 6 o mês de idade gestacional corrigida (IGC). Métodos: foram incluídos no estudo 40 RNMBP sem restrição do crescimento intra-útero, divididos em 2 grupos.Grupo I (GI) constituído por 10 RNMBP alimentados com o leite da própria mãe e Grupo II (GII) com 30 RNMBP alimentados com pelo menos 60% de leite de BLH escolhido segundo o valor calórico e protéico. Durante a internação foram analisadas a aceitação e progressão alimentar, as intercorrências clínicas (incidência de sepse, enterocolite necrosante e displasia broncopulmonar) dosados cálcio, fósforo (no soro e urina), uréia e creatinina e Hb / Hct. Construíram-se curvas de crescimento a partir dos dados relativos ao peso, comprimento e perímetro cefálico sendo estabelecidas as equações de regressão não linear que melhor se ajustavam aos dados dos parâmetros de crescimento individuais para cada RNMBP. Resultados: O início da dieta enteral foi, em média, de 1, 2 dias para o G I e 1, 3 dias para o GII e a recuperação do peso de nascimento foi, em média, de 7,5 dias para G I e 11 dias para o GII. A dieta enteral plena foi atingida, em média, no G I em 5,5 dias e 10 dias no GII e o período para atingir 2kg de peso foi, em média, para o GI 7,3 semanas e GII 7,8 semanas. Verificou-se que a incidência de sepse foi no GI 30% e GII 23%, de ECN no GI não encontrada e GII 10% sendo 6,5 % 1 A e 3,5 % 1B. O GI apresentou cálcio urinário > 4 mg/L em 1/10 (10%), fósforo urinário <1mg/L em 10/10 (100%) e relação Ca/Cr > 0,6 em 1/10 (10%) dos casos; no G II nenhuma criança apresentou alterações nos valores do cálcio urinário e na relação Ca e Cr; 19/30 (63%) apresentaram fósforo urinário <1mg/L. Quanto ao crescimento, durante a internação verificou-se no GI, entre a 30 a e 39 a semana de IGC, no percentil 50 , em média, ganho ponderal de 12,1g/dia, comprimento de 0,75cm/semana e perímetro cefálico de 0,74 cm/semana. No G II, da 28ª até a 39ª semana de IGC, em relação ao percentil 50, em média, o ganho ponderal foi de 15,8 g por dia, comprimento de 1,02cm /semana e o perímetro cefálico de 0,76 cm/semana. No momento da alta hospitalar o índice de aleitamento materno exclusivo foi no GI 9/10 (90%) crianças e GII 25/30 (83,3%). No sexto mês de idade gestacional corrigida somente uma criança apresentou-se com escore Z < -2 de P/I (tabela NCHS-2000). Conclusões: O leite de BLH previamente selecionado quanto ao valor calórico e protéico proporcionou aos RNMBP crescimento satisfatório durante a internação e adequação nutricional no 6 o mês de IGC. Verificou-se poucas complicações clínicas à internação e boa aceitação da dieta; no entanto, deve-se monitorar deficiência de fósforo. / Objective: To describe the growth and clinical-laboratory evolution of very low birth weight (VLBW) infants fed with milk from a Human Milk Bank (HMB) according to the caloric and protein value during the hospital stay and in the 6th month of the Corrected Gestational Age (CGA). Method: 40 VLBW infants were included in the study with no intra-uterus growth restriction, divided into 2 groups. Group I (GI) consisted of 10 VLBW infants fed with their own mother\'s milk and Group II (GII) was made of 30 VLBW infants fed with at least 60% of the milk from an HMB according to the caloric and protein value. During the hospital stay, the acceptance and food progression were analyzed, the clinical intercurrences (sepsis incidence, necrotizing enterocolitis, and bronchopulmonary dysplasia) dosed calcium, phosphorus (in the serum and urine), urea and creatinine and Hb/ Hct. Growth curves were built based on the data relating to the weight, length and cephalic perimeter, being established the nonlinear regression equations that better fit the data of the individual growth parameters for each VLBW infant. Results: The beginning of the enteral diet was, on average, of 1.2 days for G I and 1.3 days for GII and birth weight recovery was, on average, of 7.5 days for G I and 11 days for GII. The full enteral diet was reached, on average, in 5.5 days by G I and 10 days by GII and the period to reach 2kg weight was, on average, of 7.3 weeks for GI and 7.8 weeks for GII. It was verified that GI had a 30% sepsis incidence and GII, 23%. NEC was not found in GI and GII had 10%, being 6.5% 1 A and 3.5% 1B. GI presented urinary calcium > 4 mg/L in 1/10 (10%), urinary phosphorus <1mg/L in 10/10 (100%) and Ca/Cr >0.6 ratio in 1/10 (10%) of the cases; in G II no children presented alterations in the values of the urinary calcium and in the Ca and Cr ratio; 19/30 (63%) presented <1mg/L urinary phosphorus. As for growth, during the hospital stay there was in GI, between the 30 th and 39 th CGA week, in the 50 percentile, on average, a verified ponderal gain of 12.1g / day, a length of 0.75cm / week and a cephalic perimeter of 0.74 cm / week. In G II, from the 28th to the 39th CGA week, in relation to the 50 percentile, on average, a ponderal gain of 15.8 g/day, a length of 1.02cm / week and a cephalic perimeter of 0.76 cm / week. At the moment of hospital discharge the index of exclusive maternal breast feeding was of 9/10 (90%) of the children in GI and 25/30 (83.3%) in GII. In the sixth month of the CGA only one child presented score Z < -2 of W/L (table NCHS-2000). Conclusions: The milk from HMB previously selected for its caloric and protein value afforded VLBW infants with a satisfactory growth during hospital stay and suitable nutrition in the 6th CGA month. Few clinical complications were verified during stay and there was good acceptance of the diet; however, phosphorus deficiency should be monitored.

INTRODUÇÃO: A Doença Metabólica Óssea é uma patologia caracterizada por alterações da mineralização esquelética decorrente do acréscimo deficiente do conteúdo mineral ósseo do recém-nascido pré-termo (RNPT). A densitometria óssea tem sido o método de escolha para avaliar o conteúdo mineral ósseo, particularmente no RNPT alimentado ou não com leite humano. OBJETIVOS: 1. Comparar os resultados do conteúdo e densidade minerais ósseos obtidos do corpo inteiro através da densitometria óssea em RNPT adequados para a idade gestacional (AIG) com os valores obtidos em recém-nascidos de termo (RNT), alimentados com leite humano exclusivo; 2. Comparar os parâmetros bioquímicos ósseos sanguíneos (cálcio, fósforo e fosfatase alcalina) entre os RNPT e RNTAIG; 3. Analisar a excreção urinária de cálcio e fósforo no RNPTAIG, alimentado com leite humano exclusivo. MÉTODOS: Estudo coorte prospectivo realizado na Unidade Neonatal da Divisão de Clínica Pediátrica do Hospital Universitário da Universidade de São Paulo (USP). Durante o período de julho de 2006 a setembro de 2008 foram estudados 28 RN sendo 14 RNPT com idade gestacional inferior a 34 semanas e 14 RNT, em 3 momentos: 40 semanas de idade pósconcepcional corrigida, 3 e 6 meses de idade pós-natal corrigida, alimentados com leite humano exclusivo da própria mãe ou procedente do banco de leite. Foi realizada a densitometria óssea através da técnica de dupla emissão de fonte de Raio-X (DXA) em um aparelho da marca Hologic, modelo Discovery A, no laboratório de metabolismo ósseo da Faculdade de Medicina da USP, sendo avaliados: conteúdo mineral ósseo (BMC), densidade mineral óssea (BMD) e massa magra. Foram também realizados: cálcio, fósforo e fosfatase alcalina séricos nas idades de 40 semanas pósconcepcionais (RNPT e RNT) e com 6 meses de idade pós-natal corrigida (RNPT). Além disso, analisou-se a concentração de cálcio e fósforo na urina de 6 horas dos RNPT entre a 3ª e a 4ª semanas de vida. RESULTADOS: O BMC do RNPT foi inferior ao obtido no RNT na mesma idade de 40 semanas pós-concepcionais (p<0,001), o mesmo ocorreu com o BMD (p<0,001) e com a massa magra (p=0,047). Destacamos que houve uma importante aceleração para que os pré-termos atingissem os valores normais encontrados nos recém-nascidos de termo aos 6 meses de idade corrigida, ie, BMC (137,1422,46 vs. 152,8620,92 g, p=0,054), BMD (0,220,02 vs. 0,230,02 g/cm2, p=0,618) e massa magra (5.750,13765,11 vs. 5.745,88726,52g, p=1,00) que foram comparáveis nos dois grupos. Os parâmetros bioquímicos séricos foram semelhantes entre os RNPT e os RNT. Entre os RNPT, três (21,4%) apresentaram exames urinários sugestivos de Síndrome da Deficiência de Fósforo sendo indicado o uso da solução oral de cálcio e fósforo. CONCLUSÕES: A avaliação da mineralização óssea através da DXA do corpo inteiro pode ser utilizada em RNPT constituindo exame de excelência, com precisão elevada e exposição baixa à radiação para detecção de alterações ósseas quando há risco de Doença Metabólica Óssea. Neste estudo, a alimentação com leite humano seja da própria mãe ou do banco de leite pode proporcionar um crescimento ósseo adequado, no RNPT, semelhante ao do RNT nos primeiros 6 meses de vida. Os parâmetros bioquímicos urinários (cálcio e fósforo) foram de utilidade para detecção precoce de hipofosfatúria e risco de deficiência mineral nos RNPT, entretanto requerem coleta de urina de 6 horas, tornando difícil a investigação após a alta hospitalar. Mesmo os RNPT, com exames urinários normais, apresentaram na densitometria óssea um BMC inferior aos valores encontrados nos RNT mostrando que a DXA se constitui em um método sensível e que poderá ser utilizada para o seguimento destes RNPT. / BACKGROUND: Metabolic Bone Disease is a pathology characterized by skeletal mineralization changes resulting from insufficient increase of the bone mineral content in preterm newborns (PTNB). Bone densitometry has been the method of choice to evaluate the bone mineral content, particularly in the PTNB that is or is not fed with human milk. OBJECTIVES: 1. To compare the bone mineral content and density results obtained from the whole body using bone densitometry in appropriate-for-gestational-age (AGA) PTNB with the values obtained from full-term newborns (FTNB), exclusively fed with human milk; 2. To compare blood bone biochemical parameters (calcium, phosphorus, and alkaline phosphatase) obtained from PTNB and AGA FTNB; 3. To analyze calcium and phosphorus urinary excretion in AGA PTNB exclusively fed with human milk. METHODS: Prospective cohort study conducted at the University of Sao Paulo (USP) University Hospital Pediatric Clinic Neonatal Unit. From July 2006 to September 2008, 28 NB were studied, with 14 of them being PTNB of gestational age inferior to 34 weeks and 14 being FTNB, in 3 timepoints: 40 weeks of corrected post-conception age, and 3 and 6 months of corrected postnatal age, exclusively fed with human milk from their own mothers or from a human milk bank. Bone densitometry was performed by using the dual-energy X-ray absorption (DXA) technique in a Hologic device, Discovery A model, at USP Medical School bone metabolism laboratory, and the following were evaluated: bone mineral content (BMC), bone mineral density (BMD), and lean mass. Also, the following were performed: serum calcium, phosphorus, and alkaline phosphatase at 40 weeks post-conception (PTNB and FTNB) and 6 months of corrected postnatal age (PTNB). In addition, calcium and phosphorus concentrations in PTNBs 6-hour urine were analyzed at 3 and 4 weeks of life. RESULTS: The BMC of preterm newborn was lower than in fullterm at the same 40 weeks postconceptional age (p<0.001), the same occurred with the BMD (p<0.001) and the lean mass (p=0.047). Interestingly, there was an important acceleration to achieve the normal values seen in fullterm at six months age, ie, BMC (137.14+-22.46 vs. 152.86+-20.92 g, p=0.054), BMD (0.22+-0.02 vs. 0.23+-0.02 g/cm2, p=0.618) and lean mass (5,750.13+-765.11 vs. 5,745.88726.52g, p=1.00) were comparable in both groups. PTNB and FTNBs serum biochemical parameters were similar. Among PTNBs, three (21.4%) had urinary exams suggesting Phosphorus Deficiency Syndrome, with the use of oral calcium and phosphorus solution being indicated. CONCLUSIONS: The bone mineralization evaluation using whole-body DXA may be used in PTNBs, being an optimal exam of high precision and low radiation exposure to detect bone changes when there is a risk for MBD. In this study, being fed with human milk, either from the own mothers or from a human milk bank is able to provide appropriate bone growth in PTNBs, similar to that seen in FTNBs in the first 6 months of life. Urinary biochemical parameters (calcium and phosphorus) were useful for the early detection of hypophosphaturia and mineral deficiency risk in PTNBs; however, the collection of 6-hour urine is required, as the investigation becomes difficult after hospital discharge. Even PTNBs with normal urinary exams showed, upon bone densitometry, a BMC inferior to the values found in FTNBs, which supports that DXA presents a sensitive method that can be used to follow up these PTNBs.

O objetivo geral deste estudo foi investigar o efeito do leite humano na redução de respostas biocomportamentais de dor em RNPTs submetidos ao exame de fundo de olho para diagnóstico precoce da retinopatia da prematuridade, em comparação com a sacarose 25%. Trata-se de ensaio clínico randomizado, realizado na unidade neonatal de um hospital universitário de Ribeirão Preto-SP, com 48 prematuros distribuídos randomicamente em dois grupos: leite humano e sacarose, administrados por via oral 2 minutos antes do procedimento doloroso. As variáveis dimensionadas foram mímica facial da escala de dor NFCS, choro, frequência cardíaca e cortisol salivar. A coleta de dados foi realizada em cinco períodos para a mímica facial e o NFCS: basal (05 minutos), olho direito e olho esquerdo, recuperação imediata (05 minutos após o procedimento) e recuperação tardia (05 minutos após a recuperação imediata); quatro períodos para a frequência cardíaca e o choro: basal (05 minutos), procedimento, recuperação imediata e tardia. O cortisol salivar foi coletado antes do início do exame de fundo de olho (basal), após 30 (cortisol resposta) e 60 (cortisol recuperação) minutos do término do exame. O projeto foi aprovado pelo Comitê de Ética em Pesquisa da instituição e foi registrado no Australian and New Zeland Clinical Trials. O Teste de Kolmogorov-Smirnov e o Teste t-student mostraram que os grupos são comparáveis. Para análise das variáveis de desfecho, utilizou-se a Análise de Variância com Medidas Repetidas (RM-ANOVA). Após testar a homogeneidade das variáveis (teste de Levene), foi avaliado o pressuposto da esfericidade das matrizes de variância (teste de Mauchly). Assumindo-se a esfericidade, adotaram-se a análise multivariada, com valor de épsilon < 0,75, e a estatística de Wilk´s Lambda. Para valores de épsilon > 0,75, adotaram-se a análise univariada e a estatística de Hunh-Feldt (HF) ou Greenhouse-Geisser. Para as variáveis categóricas, foi utilizado o teste Qui-quadrado ou Teste exato Fisher. O efeito do leite humano no alívio da dor, decorrente do exame de fundo de olho, não diferiu estatisticamente da sacarose 25%, ao avaliarmos a duração da presença de fronte saliente [Wilk\'s Lamba=0,918, F=0,960, p=0,241], do sulco nasolabial [Wilk\'s Lambda=0,883; F=1,43; p=0,241] e do choro [Wilk\'s Lambda=0,964; F=3,44; p=0,656], e o escore médio do NFCS adaptado ao considerarmos essas duas ações da mímica facial [Wilk\'s Lambda=0,924; F=0,884; p=0,481], a FC [Wilk\'s Lambda=0,977; F=3,44; p=0,792] e a concentração de cortisol salivar [Wilk\'s Lambda=0,978; F=2,45; p=0,612]. A comparação intrasujeitos mostrou diferença significativa nos períodos de coleta dos dados, com aumento nos escores do NFCS, na duração da fronte saliente e do sulco nasolabial, durante o exame do fundo de olho direito e esquerdo, como também no aumento do choro, durante o procedimento doloroso, em comparação com os períodos basal, recuperação inicial e tardia. A concentração do cortisol salivar aumentou significativamente aos 30 e 60 minutos após o procedimento, em comparação com os valores do cortisol basal; não houve diferença entre o cortisol resposta (30 minutos) e recuperação (60 minutos). A duração do sulco nasolabial, do choro e os escores do NFCS no basal não diferiram significativamente dos valores encontrados na recuperação inicial e tardia. Na FC, houve diferença no basal em relação à recuperação imediata, e na duração da fronte saliente houve diferença nas duas recuperações. Concluímos que o efeito do leite humano, na redução das respostas biocomportamentais de dor em prematuros submetidos ao exame de fundo de olho para diagnóstico da retinopatia da prematuridade, não é superior ao efeito obtido com a sacarose 25%, rejeitando-se a hipótese alternativa. / The general aim of this study was to investigate the effect of human milk on reducing biobehavioral pain responses in PN\'s undergoing the ocular fundus examination for the early diagnosis of retinopathy of prematurity, compared with 25% sucrose. This is a randomized clinical trial conducted in the neonatal unit of a university hospital of Ribeirão Preto-SP, with 48 premature newborns randomly distributed into two groups: human milk and sucrose, orally administered 2 minutes before the painful procedure. The variables investigated were the facial movements of the NFCS pain scale, crying, heart rate (HR) and salivary cortisol levels. Data collection was conducted in five periods for the facial movements and NFCS: baseline (05 minutes), right eye and left eye, immediate recovery (05 minutes after the procedure) and delayed recovery (05 minutes after immediate recovery); and in four periods for heart rate and crying: baseline (05 minutes), procedure, immediate and delayed recovery. Salivary cortisol was collected before the start of the fundus examination (baseline), and at 30 minutes (cortisol response) and 60 minutes (cortisol recovery) after the end of the examination. The project was approved by the Research Ethics Committee and was registered in the Australian and New Zealand Clinical Trials. The Kolmogorov-Smirnov test and Student\'s t-test showed that the groups were comparable. For the analysis of the outcome variables, the Analysis of Variance with Repeated Measures (RM-ANOVA) was used. After testing the homogeneity of the variables (Levene\'s test) the assumption of sphericity of the variance matrices was evaluated (Mauchly\'s test). Assuming sphericity, multivariate analysis was adopted, with a < 0.75 value of epsilon and Wilk\'s Lambda statistic. For values of epsilon > 0.75, univariate analysis and the Hunh-Feldt (HF) or Greenhouse-Geisser statistic were adopted. For the categorical variables, the chi-square test or Fisher exact test was used. The effect of human milk on pain relief, arising from the ocular fundus examination, did not differ statistically from 25% sucrose, when the duration of the presence of brow lowering [Wilk\'s Lambda=0.918; F=0.960; p=0.241], of deepening of the naso-labial furrow [Wilk\'s Lambda=0.883; F=1.43; p=0.241] and of crying [Wilk\'s Lambda=0.964; F=3.44; p=0.656] were evaluated, as well as the mean score of the NFCS adapted to consider these two facial movement actions [Wilk\'s Lambda=0.924; F=0.884; p=0.481], the HR [Wilk\'s Lambda=0.977; F=3.44; p=0.792] and the salivary cortisol concentration [Wilk\'s Lambda=0.978; F=2.45; p=0.612]. The intra-subject comparison showed a significant difference in the data collection periods, with an increase in the NFCS scores, in the duration of brow lowering and naso-labial furrow deepening, during the examination fundus of the right and left eyes, as well as increased crying during the painful procedure, compared with the baseline period, immediate and delayed recovery. The concentration of salivary cortisol increased significantly at 30 and 60 minutes after the procedure, compared with the baseline cortisol values; there was no difference between cortisol response (30 minutes) and recovery (60 minutes). The duration of naso-labial furrow deepening, of crying and the NFCS scores at the baseline did not differ significantly from the values found in the immediate and delayed recovery. In the HR, there was no difference between the baseline and the immediate recovery, and in the duration of brow lowering there was no difference in the two recoveries. We conclude that the effect of human milk, on the reduction of biobehavioral responses to pain in preterm infants undergoing the ocular fundus examination for the diagnosis of retinopathy of prematurity, is not greater than the effect obtained with 25% sucrose, rejecting the alternative hypothesis.

Objetivo: Descrever o crescimento e a evolução clínico laboratorial de recém-nascidos de muito baixo peso (RNMBP) alimentados com leite de banco de leite humano (BLH) segundo o valor calórico e protéico durante a internação e no 6 o mês de idade gestacional corrigida (IGC). Métodos: foram incluídos no estudo 40 RNMBP sem restrição do crescimento intra-útero, divididos em 2 grupos.Grupo I (GI) constituído por 10 RNMBP alimentados com o leite da própria mãe e Grupo II (GII) com 30 RNMBP alimentados com pelo menos 60% de leite de BLH escolhido segundo o valor calórico e protéico. Durante a internação foram analisadas a aceitação e progressão alimentar, as intercorrências clínicas (incidência de sepse, enterocolite necrosante e displasia broncopulmonar) dosados cálcio, fósforo (no soro e urina), uréia e creatinina e Hb / Hct. Construíram-se curvas de crescimento a partir dos dados relativos ao peso, comprimento e perímetro cefálico sendo estabelecidas as equações de regressão não linear que melhor se ajustavam aos dados dos parâmetros de crescimento individuais para cada RNMBP. Resultados: O início da dieta enteral foi, em média, de 1, 2 dias para o G I e 1, 3 dias para o GII e a recuperação do peso de nascimento foi, em média, de 7,5 dias para G I e 11 dias para o GII. A dieta enteral plena foi atingida, em média, no G I em 5,5 dias e 10 dias no GII e o período para atingir 2kg de peso foi, em média, para o GI 7,3 semanas e GII 7,8 semanas. Verificou-se que a incidência de sepse foi no GI 30% e GII 23%, de ECN no GI não encontrada e GII 10% sendo 6,5 % 1 A e 3,5 % 1B. O GI apresentou cálcio urinário > 4 mg/L em 1/10 (10%), fósforo urinário <1mg/L em 10/10 (100%) e relação Ca/Cr > 0,6 em 1/10 (10%) dos casos; no G II nenhuma criança apresentou alterações nos valores do cálcio urinário e na relação Ca e Cr; 19/30 (63%) apresentaram fósforo urinário <1mg/L. Quanto ao crescimento, durante a internação verificou-se no GI, entre a 30 a e 39 a semana de IGC, no percentil 50 , em média, ganho ponderal de 12,1g/dia, comprimento de 0,75cm/semana e perímetro cefálico de 0,74 cm/semana. No G II, da 28ª até a 39ª semana de IGC, em relação ao percentil 50, em média, o ganho ponderal foi de 15,8 g por dia, comprimento de 1,02cm /semana e o perímetro cefálico de 0,76 cm/semana. No momento da alta hospitalar o índice de aleitamento materno exclusivo foi no GI 9/10 (90%) crianças e GII 25/30 (83,3%). No sexto mês de idade gestacional corrigida somente uma criança apresentou-se com escore Z < -2 de P/I (tabela NCHS-2000). Conclusões: O leite de BLH previamente selecionado quanto ao valor calórico e protéico proporcionou aos RNMBP crescimento satisfatório durante a internação e adequação nutricional no 6 o mês de IGC. Verificou-se poucas complicações clínicas à internação e boa aceitação da dieta; no entanto, deve-se monitorar deficiência de fósforo. / Objective: To describe the growth and clinical-laboratory evolution of very low birth weight (VLBW) infants fed with milk from a Human Milk Bank (HMB) according to the caloric and protein value during the hospital stay and in the 6th month of the Corrected Gestational Age (CGA). Method: 40 VLBW infants were included in the study with no intra-uterus growth restriction, divided into 2 groups. Group I (GI) consisted of 10 VLBW infants fed with their own mother\'s milk and Group II (GII) was made of 30 VLBW infants fed with at least 60% of the milk from an HMB according to the caloric and protein value. During the hospital stay, the acceptance and food progression were analyzed, the clinical intercurrences (sepsis incidence, necrotizing enterocolitis, and bronchopulmonary dysplasia) dosed calcium, phosphorus (in the serum and urine), urea and creatinine and Hb/ Hct. Growth curves were built based on the data relating to the weight, length and cephalic perimeter, being established the nonlinear regression equations that better fit the data of the individual growth parameters for each VLBW infant. Results: The beginning of the enteral diet was, on average, of 1.2 days for G I and 1.3 days for GII and birth weight recovery was, on average, of 7.5 days for G I and 11 days for GII. The full enteral diet was reached, on average, in 5.5 days by G I and 10 days by GII and the period to reach 2kg weight was, on average, of 7.3 weeks for GI and 7.8 weeks for GII. It was verified that GI had a 30% sepsis incidence and GII, 23%. NEC was not found in GI and GII had 10%, being 6.5% 1 A and 3.5% 1B. GI presented urinary calcium > 4 mg/L in 1/10 (10%), urinary phosphorus <1mg/L in 10/10 (100%) and Ca/Cr >0.6 ratio in 1/10 (10%) of the cases; in G II no children presented alterations in the values of the urinary calcium and in the Ca and Cr ratio; 19/30 (63%) presented <1mg/L urinary phosphorus. As for growth, during the hospital stay there was in GI, between the 30 th and 39 th CGA week, in the 50 percentile, on average, a verified ponderal gain of 12.1g / day, a length of 0.75cm / week and a cephalic perimeter of 0.74 cm / week. In G II, from the 28th to the 39th CGA week, in relation to the 50 percentile, on average, a ponderal gain of 15.8 g/day, a length of 1.02cm / week and a cephalic perimeter of 0.76 cm / week. At the moment of hospital discharge the index of exclusive maternal breast feeding was of 9/10 (90%) of the children in GI and 25/30 (83.3%) in GII. In the sixth month of the CGA only one child presented score Z < -2 of W/L (table NCHS-2000). Conclusions: The milk from HMB previously selected for its caloric and protein value afforded VLBW infants with a satisfactory growth during hospital stay and suitable nutrition in the 6th CGA month. Few clinical complications were verified during stay and there was good acceptance of the diet; however, phosphorus deficiency should be monitored.

INTRODUÇÃO: A Doença Metabólica Óssea é uma patologia caracterizada por alterações da mineralização esquelética decorrente do acréscimo deficiente do conteúdo mineral ósseo do recém-nascido pré-termo (RNPT). A densitometria óssea tem sido o método de escolha para avaliar o conteúdo mineral ósseo, particularmente no RNPT alimentado ou não com leite humano. OBJETIVOS: 1. Comparar os resultados do conteúdo e densidade minerais ósseos obtidos do corpo inteiro através da densitometria óssea em RNPT adequados para a idade gestacional (AIG) com os valores obtidos em recém-nascidos de termo (RNT), alimentados com leite humano exclusivo; 2. Comparar os parâmetros bioquímicos ósseos sanguíneos (cálcio, fósforo e fosfatase alcalina) entre os RNPT e RNTAIG; 3. Analisar a excreção urinária de cálcio e fósforo no RNPTAIG, alimentado com leite humano exclusivo. MÉTODOS: Estudo coorte prospectivo realizado na Unidade Neonatal da Divisão de Clínica Pediátrica do Hospital Universitário da Universidade de São Paulo (USP). Durante o período de julho de 2006 a setembro de 2008 foram estudados 28 RN sendo 14 RNPT com idade gestacional inferior a 34 semanas e 14 RNT, em 3 momentos: 40 semanas de idade pósconcepcional corrigida, 3 e 6 meses de idade pós-natal corrigida, alimentados com leite humano exclusivo da própria mãe ou procedente do banco de leite. Foi realizada a densitometria óssea através da técnica de dupla emissão de fonte de Raio-X (DXA) em um aparelho da marca Hologic, modelo Discovery A, no laboratório de metabolismo ósseo da Faculdade de Medicina da USP, sendo avaliados: conteúdo mineral ósseo (BMC), densidade mineral óssea (BMD) e massa magra. Foram também realizados: cálcio, fósforo e fosfatase alcalina séricos nas idades de 40 semanas pósconcepcionais (RNPT e RNT) e com 6 meses de idade pós-natal corrigida (RNPT). Além disso, analisou-se a concentração de cálcio e fósforo na urina de 6 horas dos RNPT entre a 3ª e a 4ª semanas de vida. RESULTADOS: O BMC do RNPT foi inferior ao obtido no RNT na mesma idade de 40 semanas pós-concepcionais (p<0,001), o mesmo ocorreu com o BMD (p<0,001) e com a massa magra (p=0,047). Destacamos que houve uma importante aceleração para que os pré-termos atingissem os valores normais encontrados nos recém-nascidos de termo aos 6 meses de idade corrigida, ie, BMC (137,1422,46 vs. 152,8620,92 g, p=0,054), BMD (0,220,02 vs. 0,230,02 g/cm2, p=0,618) e massa magra (5.750,13765,11 vs. 5.745,88726,52g, p=1,00) que foram comparáveis nos dois grupos. Os parâmetros bioquímicos séricos foram semelhantes entre os RNPT e os RNT. Entre os RNPT, três (21,4%) apresentaram exames urinários sugestivos de Síndrome da Deficiência de Fósforo sendo indicado o uso da solução oral de cálcio e fósforo. CONCLUSÕES: A avaliação da mineralização óssea através da DXA do corpo inteiro pode ser utilizada em RNPT constituindo exame de excelência, com precisão elevada e exposição baixa à radiação para detecção de alterações ósseas quando há risco de Doença Metabólica Óssea. Neste estudo, a alimentação com leite humano seja da própria mãe ou do banco de leite pode proporcionar um crescimento ósseo adequado, no RNPT, semelhante ao do RNT nos primeiros 6 meses de vida. Os parâmetros bioquímicos urinários (cálcio e fósforo) foram de utilidade para detecção precoce de hipofosfatúria e risco de deficiência mineral nos RNPT, entretanto requerem coleta de urina de 6 horas, tornando difícil a investigação após a alta hospitalar. Mesmo os RNPT, com exames urinários normais, apresentaram na densitometria óssea um BMC inferior aos valores encontrados nos RNT mostrando que a DXA se constitui em um método sensível e que poderá ser utilizada para o seguimento destes RNPT. / BACKGROUND: Metabolic Bone Disease is a pathology characterized by skeletal mineralization changes resulting from insufficient increase of the bone mineral content in preterm newborns (PTNB). Bone densitometry has been the method of choice to evaluate the bone mineral content, particularly in the PTNB that is or is not fed with human milk. OBJECTIVES: 1. To compare the bone mineral content and density results obtained from the whole body using bone densitometry in appropriate-for-gestational-age (AGA) PTNB with the values obtained from full-term newborns (FTNB), exclusively fed with human milk; 2. To compare blood bone biochemical parameters (calcium, phosphorus, and alkaline phosphatase) obtained from PTNB and AGA FTNB; 3. To analyze calcium and phosphorus urinary excretion in AGA PTNB exclusively fed with human milk. METHODS: Prospective cohort study conducted at the University of Sao Paulo (USP) University Hospital Pediatric Clinic Neonatal Unit. From July 2006 to September 2008, 28 NB were studied, with 14 of them being PTNB of gestational age inferior to 34 weeks and 14 being FTNB, in 3 timepoints: 40 weeks of corrected post-conception age, and 3 and 6 months of corrected postnatal age, exclusively fed with human milk from their own mothers or from a human milk bank. Bone densitometry was performed by using the dual-energy X-ray absorption (DXA) technique in a Hologic device, Discovery A model, at USP Medical School bone metabolism laboratory, and the following were evaluated: bone mineral content (BMC), bone mineral density (BMD), and lean mass. Also, the following were performed: serum calcium, phosphorus, and alkaline phosphatase at 40 weeks post-conception (PTNB and FTNB) and 6 months of corrected postnatal age (PTNB). In addition, calcium and phosphorus concentrations in PTNBs 6-hour urine were analyzed at 3 and 4 weeks of life. RESULTS: The BMC of preterm newborn was lower than in fullterm at the same 40 weeks postconceptional age (p<0.001), the same occurred with the BMD (p<0.001) and the lean mass (p=0.047). Interestingly, there was an important acceleration to achieve the normal values seen in fullterm at six months age, ie, BMC (137.14+-22.46 vs. 152.86+-20.92 g, p=0.054), BMD (0.22+-0.02 vs. 0.23+-0.02 g/cm2, p=0.618) and lean mass (5,750.13+-765.11 vs. 5,745.88726.52g, p=1.00) were comparable in both groups. PTNB and FTNBs serum biochemical parameters were similar. Among PTNBs, three (21.4%) had urinary exams suggesting Phosphorus Deficiency Syndrome, with the use of oral calcium and phosphorus solution being indicated. CONCLUSIONS: The bone mineralization evaluation using whole-body DXA may be used in PTNBs, being an optimal exam of high precision and low radiation exposure to detect bone changes when there is a risk for MBD. In this study, being fed with human milk, either from the own mothers or from a human milk bank is able to provide appropriate bone growth in PTNBs, similar to that seen in FTNBs in the first 6 months of life. Urinary biochemical parameters (calcium and phosphorus) were useful for the early detection of hypophosphaturia and mineral deficiency risk in PTNBs; however, the collection of 6-hour urine is required, as the investigation becomes difficult after hospital discharge. Even PTNBs with normal urinary exams showed, upon bone densitometry, a BMC inferior to the values found in FTNBs, which supports that DXA presents a sensitive method that can be used to follow up these PTNBs.

O objetivo geral deste estudo foi investigar o efeito do leite humano na redução de respostas biocomportamentais de dor em RNPTs submetidos ao exame de fundo de olho para diagnóstico precoce da retinopatia da prematuridade, em comparação com a sacarose 25%. Trata-se de ensaio clínico randomizado, realizado na unidade neonatal de um hospital universitário de Ribeirão Preto-SP, com 48 prematuros distribuídos randomicamente em dois grupos: leite humano e sacarose, administrados por via oral 2 minutos antes do procedimento doloroso. As variáveis dimensionadas foram mímica facial da escala de dor NFCS, choro, frequência cardíaca e cortisol salivar. A coleta de dados foi realizada em cinco períodos para a mímica facial e o NFCS: basal (05 minutos), olho direito e olho esquerdo, recuperação imediata (05 minutos após o procedimento) e recuperação tardia (05 minutos após a recuperação imediata); quatro períodos para a frequência cardíaca e o choro: basal (05 minutos), procedimento, recuperação imediata e tardia. O cortisol salivar foi coletado antes do início do exame de fundo de olho (basal), após 30 (cortisol resposta) e 60 (cortisol recuperação) minutos do término do exame. O projeto foi aprovado pelo Comitê de Ética em Pesquisa da instituição e foi registrado no Australian and New Zeland Clinical Trials. O Teste de Kolmogorov-Smirnov e o Teste t-student mostraram que os grupos são comparáveis. Para análise das variáveis de desfecho, utilizou-se a Análise de Variância com Medidas Repetidas (RM-ANOVA). Após testar a homogeneidade das variáveis (teste de Levene), foi avaliado o pressuposto da esfericidade das matrizes de variância (teste de Mauchly). Assumindo-se a esfericidade, adotaram-se a análise multivariada, com valor de épsilon < 0,75, e a estatística de Wilk´s Lambda. Para valores de épsilon > 0,75, adotaram-se a análise univariada e a estatística de Hunh-Feldt (HF) ou Greenhouse-Geisser. Para as variáveis categóricas, foi utilizado o teste Qui-quadrado ou Teste exato Fisher. O efeito do leite humano no alívio da dor, decorrente do exame de fundo de olho, não diferiu estatisticamente da sacarose 25%, ao avaliarmos a duração da presença de fronte saliente [Wilk\'s Lamba=0,918, F=0,960, p=0,241], do sulco nasolabial [Wilk\'s Lambda=0,883; F=1,43; p=0,241] e do choro [Wilk\'s Lambda=0,964; F=3,44; p=0,656], e o escore médio do NFCS adaptado ao considerarmos essas duas ações da mímica facial [Wilk\'s Lambda=0,924; F=0,884; p=0,481], a FC [Wilk\'s Lambda=0,977; F=3,44; p=0,792] e a concentração de cortisol salivar [Wilk\'s Lambda=0,978; F=2,45; p=0,612]. A comparação intrasujeitos mostrou diferença significativa nos períodos de coleta dos dados, com aumento nos escores do NFCS, na duração da fronte saliente e do sulco nasolabial, durante o exame do fundo de olho direito e esquerdo, como também no aumento do choro, durante o procedimento doloroso, em comparação com os períodos basal, recuperação inicial e tardia. A concentração do cortisol salivar aumentou significativamente aos 30 e 60 minutos após o procedimento, em comparação com os valores do cortisol basal; não houve diferença entre o cortisol resposta (30 minutos) e recuperação (60 minutos). A duração do sulco nasolabial, do choro e os escores do NFCS no basal não diferiram significativamente dos valores encontrados na recuperação inicial e tardia. Na FC, houve diferença no basal em relação à recuperação imediata, e na duração da fronte saliente houve diferença nas duas recuperações. Concluímos que o efeito do leite humano, na redução das respostas biocomportamentais de dor em prematuros submetidos ao exame de fundo de olho para diagnóstico da retinopatia da prematuridade, não é superior ao efeito obtido com a sacarose 25%, rejeitando-se a hipótese alternativa. / The general aim of this study was to investigate the effect of human milk on reducing biobehavioral pain responses in PN\'s undergoing the ocular fundus examination for the early diagnosis of retinopathy of prematurity, compared with 25% sucrose. This is a randomized clinical trial conducted in the neonatal unit of a university hospital of Ribeirão Preto-SP, with 48 premature newborns randomly distributed into two groups: human milk and sucrose, orally administered 2 minutes before the painful procedure. The variables investigated were the facial movements of the NFCS pain scale, crying, heart rate (HR) and salivary cortisol levels. Data collection was conducted in five periods for the facial movements and NFCS: baseline (05 minutes), right eye and left eye, immediate recovery (05 minutes after the procedure) and delayed recovery (05 minutes after immediate recovery); and in four periods for heart rate and crying: baseline (05 minutes), procedure, immediate and delayed recovery. Salivary cortisol was collected before the start of the fundus examination (baseline), and at 30 minutes (cortisol response) and 60 minutes (cortisol recovery) after the end of the examination. The project was approved by the Research Ethics Committee and was registered in the Australian and New Zealand Clinical Trials. The Kolmogorov-Smirnov test and Student\'s t-test showed that the groups were comparable. For the analysis of the outcome variables, the Analysis of Variance with Repeated Measures (RM-ANOVA) was used. After testing the homogeneity of the variables (Levene\'s test) the assumption of sphericity of the variance matrices was evaluated (Mauchly\'s test). Assuming sphericity, multivariate analysis was adopted, with a < 0.75 value of epsilon and Wilk\'s Lambda statistic. For values of epsilon > 0.75, univariate analysis and the Hunh-Feldt (HF) or Greenhouse-Geisser statistic were adopted. For the categorical variables, the chi-square test or Fisher exact test was used. The effect of human milk on pain relief, arising from the ocular fundus examination, did not differ statistically from 25% sucrose, when the duration of the presence of brow lowering [Wilk\'s Lambda=0.918; F=0.960; p=0.241], of deepening of the naso-labial furrow [Wilk\'s Lambda=0.883; F=1.43; p=0.241] and of crying [Wilk\'s Lambda=0.964; F=3.44; p=0.656] were evaluated, as well as the mean score of the NFCS adapted to consider these two facial movement actions [Wilk\'s Lambda=0.924; F=0.884; p=0.481], the HR [Wilk\'s Lambda=0.977; F=3.44; p=0.792] and the salivary cortisol concentration [Wilk\'s Lambda=0.978; F=2.45; p=0.612]. The intra-subject comparison showed a significant difference in the data collection periods, with an increase in the NFCS scores, in the duration of brow lowering and naso-labial furrow deepening, during the examination fundus of the right and left eyes, as well as increased crying during the painful procedure, compared with the baseline period, immediate and delayed recovery. The concentration of salivary cortisol increased significantly at 30 and 60 minutes after the procedure, compared with the baseline cortisol values; there was no difference between cortisol response (30 minutes) and recovery (60 minutes). The duration of naso-labial furrow deepening, of crying and the NFCS scores at the baseline did not differ significantly from the values found in the immediate and delayed recovery. In the HR, there was no difference between the baseline and the immediate recovery, and in the duration of brow lowering there was no difference in the two recoveries. We conclude that the effect of human milk, on the reduction of biobehavioral responses to pain in preterm infants undergoing the ocular fundus examination for the diagnosis of retinopathy of prematurity, is not greater than the effect obtained with 25% sucrose, rejecting the alternative hypothesis.