Poster group
Details of individual items:
poster
Docosahexaenoic acid (DHA) is a long-chain polyunsaturated fatty acidfound in human milk. Depleted levels of DHA have been associated withlearning impairments in rats, and visual impairments in the offspring ofrhesus monkeys. These finding have suggested linkages between DHA andneural development, spawning numerous works aimed at establishing arelationship between DHA and intellectual functioning in humans. Sincecomparisons between breast versus bottle-fed infants are confounded byselection differences, studies have used formula-fed infants, and haverelied on comparisons between infants receiving formula which is eithersupplemented or unsupplemented with DHA. Such studies have resulted inequivocal results. The aim of the present study was to assess whether naturally occurringvariation in levels of DHA in human milk would predict infants' behavioralfunctioning. This methodology overcomes difficulties associated withselection differences. Further, this study assessed infant outcomesshortly after milk production had been established, but still within thepostpartum period, so that possible confounds associated with earlylearning could be minimized. Milk samples were collected from 21 lactating mothers, 7 - 10 (M8.98)days after delivery. The mothers ranged in age from 15 - 26 (M19.71)years. On average, they had a high school education, and were of lowersocio-economic status (3.81 on the Hollingshead scale). Infants were fullterm, with average hospital stays (M2.7 days), as well as acceptable two-(M7.67) and five-minute (M8.56) Apgar scores, birthweights (M3.21 kg),and length (M50.57 cm). A trained female experimenter visited the families at home approximatelyone week after delivery, during the morning. Mothers pumped the entiresupply of milk from one breast, using a Medela breast pump, while theexperimenter administered the Brazelton NBAS on the infant. DHA levelswere analyzed using gas chromatography, which yielded a mean value of0.08%, +/- 0.01. This level of DHA is lower than typically reported meansfor DHA, which normally range from 0.2 to 0.4% . Correlations were conducted between DHA levels in mothers' milk andinfants' performance on each of the six dimensions of the NBAS. Positivecorrelations were obtained between DHA and the State Regulation dimension(r.72, p<.001); and between DHA and the Motor dimension (r.38, p< .05). These data provide strong evidence of a dose-dependent relationshipbetween DHA and newborn humans' behavioral development. Specifically, theresults indicate that an enhanced level of DHA in mothers' milk isassociated with an infant showing greater ability to organize hisbehavioral state and better physical coordination. Whether links betweenDHA levels and behavioral functioning at the neonatal stage are predictiveof associations between DHA levels and intellectual performance, later ininfancy and early childhood, warrants further research using longitudinalparadigms
poster
Recently, researchers have studied how parents make decisions about feeding their infants and the subsequent effects (e.g., Pridham, 1997). One decision is how often to feed the infant. One approach is to feed-on-demand (FOD); the infant is fed whenever hunger 'cues' are observed (e.g., crying). Many parents, however, are opting for a feed-on-schedule (FOS) approach (Ezzo & Bucknam, 1995); the infant is fed every 3 or 4 hours. The American Academy of Pediatrics (AAP) guidelines (1994) suggest that newborns develop a eating routine of every 3 to 4 hours but does not specifically recommend that infants be fed at these intervals. There is little empirical data regarding these approaches to feeding infants. Studies that have examined FOS v. FOD have small sample sizes and are restricted to preterm infants. Saunders, Friedman and Stramoski (1991) found that weight gain was similar for FOD and FOS preterm infants. Collinge, Bradley, Perks, Rezny and Topping (1982) found that FOD preterm infants required fewer feedings per day and were discharged earlier. From the available literature, what is known about the effects FOD v. FOS is limited to preterm samples. The objective of the present study is to clarify the effect of infant feeding practices, FOD v. FOS, on full-term infants' weight gain from birth to 6 months. Fifty-seven infants (33 male; 24 female) were recruited from the waiting room of a pediatric clinic; infants were 12.2 months old (SD 3D 4.6) and primarily Caucasian (80%). Mothers completed a 14-item questionnaire about their infant's predominant eating routines (e.g., do you decide when the infant eats, or does the baby eat whenever she cries?) that classified them as FOD or FOS across the infant's first 6 months. Infants' weight in grams and percentile rank was recorded from past 2-, 4-, and 6-month-old clinic visits. Birth weight did not differ between groups. FOD infants averaged two more meals per day than did FOS infants, t(53) 3D 2.93, p <.005. Longitudinal data were analyzed in a one-way mixed-design ANOVA using Feeding Practice as the between-subjects factor and Age (2, 4 and 6 months) as the Within-subjects factor. No significant main or interaction effects emerged for infants' raw weight. However, when infants' weight was converted to percentile ranks for those infants with complete medical records, a significant interaction between feeding-practice and age emerged, F (2, 38) 3D 4.23, p < .02. It is concluded from these data that FOD v. FOS does not significantly affect the weight of the infant across the first 6 months of life despite more meals for FOD infants. When infant weight is evaluated as a percentile rank in comparison to age norms, however, the effect of FOD vs. FOS changes slightly with age. After the first 2 months, FOS infants have a higher weight ranking than FOD infants. By 4 months, however, FOD infants pass FOS infants in weight ranks, and this trend remains at 6 months. Implications include the long-term effects of early feeding practices on growth.
poster
Child health, nutrition, and physical activity are inextricably linked,with childhood obesity, unbalanced diet, and sedentary lifestyles allcontributing to coronary heart disease—America’s number one killer. Forlow income children younger than 5, the prevalence of overweight hasincreased from nearly 19% in 1983 to 22% in 1995, a relative increase of16% (Mei et al., 1998), with overweight rates higher for all ethnicitiesin infancy as compared to the preschool ages. The principle of energybalance suggests that weight gain results from an extended periodwherein energy intake must exceed energy expenditure (Goran, Reynolds &Lindquist, 1999). In other words, infant obesity may result fromoverfeeding relative to activity level. Recently, separateinvestigators have proposed that overfeeding by the mother may be aresponse to difficulty on the part of the infant (Carey, 1985; Wells etal., 1997). The aim of the present investigation was to examine therole of temperamental activity and irritability, and their possibleinteractions with feeding, in predicting infant weight. Eighty-six non-Hispanic white infants and their mothers were seen inour laboratory when the babies were 3-months of age. Activity level wasmeasured using a custom-built apparatus (McDonnell & Richards, 1984)that counted motor movements in all four limbs simultaneously. Motherscompleted a feeding history and questionnaires that tapped temperamentalactivity and difficulty (Medoff-Cooper, Carey & McDevitt, 1993;Rothbart, 1981). Difficulty was determined by summing the individualdimension scores for high withdrawal and intensity, poor adaptability,low regularity, and negative mood (Thomas, Chess & Birch, 1968). The 42 males (6609g) and 44 females (6164g) were of different (p<.01),but normal weights for 3-month-old infants. While females were rated asslightly more difficult than males, temperamental difficulty wasassociated with greater weight for males via a regression analysis.Formula feeding, as opposed to breastfeeding, was also related toweight, but only for males. In contrast, while males were rated andassessed as slightly more active than females, activity by females wasinversely related to weight. These results provide qualified support for the energy expenditure andtemperamental difficulty hypotheses that attempt to explain weight gainin infancy, as activity and difficulty were both shown to correlate withinfant weight, but differently by gender. As formula can more easilylend itself to overfeeding, its relation to weight is somewhatsurprising insofar as the association held only for males. Suchtemperament by gender interactions may play a provocative role in ourefforts to better understand the development of obesity in infancy.
poster
Feeding disorders in infants are increasingly common in affluent societies. Besides organic causes there is accumulating evidence for an interactional etiology. One central point in the diagnosis and treatment of feeding disorders is the analysis the mother-child interaction during the feeding situation. Most instruments currently available for this purpose are global rating scales, which rate the whole feeding session on a number of items. I.e. the Mother-Infant/Toddler Feeding Scale - Birth to three years (Chatoor et al.) rates 46 items on a 4-point scale and the Feeding Scale - Birth to one year (Barnard, 1978) rates 76 items on a 2-point scale. Besides a lot of advantages of these scales one shortcoming is their global character. The scales allow only for a summary index of the quality of the interaction. The specific dysfunctional interaction patterns which may be responsible for the persistence of the problem cannot be identified2E This is especially important for planning intervention, because every mother-infant dyad has its own interactional characteristics. Another disadvantage of the scales is that changes over time in the interaction cannot be registered. To avoid these problems a microanalytic classification system for the description and analysis of feeding disorders was developed. This is an event-based coding system, which allows coding to an accuracy of 40 to 33 msec., depending on the video system being used, i.e. PAL or NTSC. The system registers 39 classes of feeding behaviors of the mother and the child. These classes are summerized in six categories which comprise mutually exclusive classes. The categories for the mother are: 'feeding-behavior' (the motoric behavior like presenting food to the child or following the childB4s head turnings with the spoon), 'strategies' (behavior to make the child eating, like praise or call upon) and 'special behavior' (behavior which is not directly related to feeding itself like ignoring the child oder holding it). The classes for the child are: 'interest in eating' (singnals of hunger or trying to eat), 'oral activities concerning food' (anticipatory opening of the mouth, chewing or vomiting) and 'missing interest in eating' ( avoiding or provocating). What are the advantages of the microanalytic system and how reliable and valid is it (compared to global rating scales)? In an empirical investigation a systematic comparison was made between the Chatoor-Scales and the microanalytic classification system. The mother-infant feeding interactions of 60 feeding-disordered infants (age 5-24 months) were analysed both by independent trained observers with the Chatoor-Scales and with the microanalytic classification system. The mothers with their infants were referred to the Munich Interdisciplinary Research and Intervention Program for Fuzzy Babies because of feeding disorders. Their feeding-interactions during a regular mealtime was videotaped in the laboratory with two cameras through one-way mirrors. The infant was seated in a highchair. All meals included spoon feeding2E Reliability (inter-observer) and validity (concurrent) measures of the microanalytic classification system were obtained. The results of the analyses were compared according to their level of description in general, identification of specific interaction-patterns, and their potential value for interventions.
poster
Mother-infant feeding interaction style is a fairly global concept, and thus has been determined in numerous studies via overall rating scales. Although rating scales are easily applicable and an economic method for identifying interaction styles, they are far from unsusceptible to rater bias. Additionally, the criteria for the raters' estimation remain concealed to the investigator.Close observation of the sequence of feeding interaction, conversely, is a more objective approach, that not only permits a microanalytic analysis of the reciprocity of the mother's and the infant's behaviours during feeding, but also allows for changes in the dyadic process over time.Our objective was to differentiate dysfunctional feeding interaction patterns in feeding-disordered infants and toddlers by use of observational methods, focusing on the appropriateness of the mother's responses to her child's satiety signals.Subjects enrolled in our study were 30 children (9 boys and 21 girls), aged 5 to 24 months (mean age 12.3 B1 4.9 months), and their mothers (mean age 32.2 B1 4.2 years). All subjects were recruited from an outpatient sample referred to the Munich Interdisciplinary Research and Intervention Program for Fussy Babies because of non-organic feeding disorders. The children's average weight on admission was 9.5 B1 1.0 kg and 8.6 B1 1.3 kg for boys and girls, respectively. Their mean height was 77.6 B1 5.9 cm for boys and 74.4 B1 5.9 cm for girls.Video recordings of feeding interactions at a regular mealtime taken in our laboratory were subjected to observational analysis.The coding scheme used by our observers consisted of six sets of mutually exclusive and exhaustive codes, three sets for maternal and infantile behaviours each. The mother's interactive behaviours were categorised into the following sets: a) Feeding behaviours - seven classes of activities intended to stimulate the infant to consume food, like offering food or touching the infant's lips with the spoon; b) Strategic behaviours - seven codes, typically vocalisations, that encourage the child to eat, e.g. positive reinforcement, distraction, instructing or inviting the infant to have a spoonful; and c) Special behaviours - that is five classes of activities that facilitate the mother's endeavours to feed her child, for example coercion to eat or positioning the infant in a suitable posture to be fed against the child's reluctance.On part of the infantile behaviours the following categories were coded: a) Interest in eating - three codes that convey the child's readiness to eat, like focusing his / her attention on or reaching for the food; b) Oral activities - 11 classes of oral behaviours; examples are food consumption, anticipatory opening of the mouth, chewing or vomiting; and c) Missing interest in eating - four classes of activities that convey the infant's aversion to being fed, like avoidance of food intake, distraction, children's food refusal or noncompliant behaviour.Our observational data currently are being analysed sequentially, describing the interaction process with transitional probabilities. We expect our analyses to reveal clusters of typical dysfunctional feeding interaction patterns that could help clinicians to define individual and effective interventions for infantile feeding disorders.
poster
While there is some evidence suggesting that certain polyunsaturatedfatty acids, specifically docosahexaenoic (DHA) and arachidonic (AA),are vital for infant visual acuity development, other research shows norelationship. Even if there was consistent evidence that early visualacuity is compromised by a diet deficient in DHA and AA, there is noevidence that delays in visual acuity have later negative effects oninfant. Indeed evidence from one study suggests a negative effect onlanguage development for infants fed an enriched formula. One possibleexplanation for these contradictory results is that DHA status of thebreast fed reference groups differs for different studies, or that theDHA status of the infants entered into formula studies differs. Severalrecent papers have shown quite marked variation in DHA status at birthand that DHA status at birth is a major determinant of subsequentpostnatal changes in response to diet. The variations in DHA statusamong otherwise healthy preterm and term gestation infants is strictlydue to the maternal DHA supply. Furthermore, the range of DHA of breastfed babies overlaps that of the formula fed. Thus before attempting torationalize if supplementation of the formula fed baby has any positiveor negative advantage to neural development, it is imperative that therelation between the range of DHA and AA status and infant developmentin the breast fed baby is understood. In this research, we addressedthe question of whether levels of DHA and/or AA in infant blood plasmaand mother's milk are a factor in infant development. We tested breastfed infants whose mothers are either vegetarian or not - factors whichcould affect levels of DHA and AA both in the breast milk and in theinfantsED blood. Breast milk samples and infant blood samples were takenat 2 and 4 months. The analyses revealed that there was enoughvariation in levels of DHA and AA in mothers' milk and in infant bloodto investigate the possibility of differential effects of levels of DHAand AA on infant development. Infants were tested at 2, 4, 6, 9, 12, and 14 months of age. Herein,we present the data collected at 9 months of age. At 9 months of age,infants were tested on the visual task (the Fagan Test of InfantIntelligence), an object search task (A not B), and a speech perceptiontask (native and non-native speech discrimination). Statisticalanalyses included LISREL and mixed model ANOVAs. The data suggest arelationship between some of the developmental measures but there wasno indication that levels of DHA and AA were causally related todevelopmental status. For example, infants who consistently showed theA not B error (a less mature developmental pattern) were more likely todiscriminate the non-native contrast (also a less mature developmentalpattern but this was not related to levels of DHA and AA in infants'blood or mothers' milk. Implications of these results will bediscussed.
poster
A significant clinical problem for very low birth weight (VLBW) infants during early feeding is their coordination of suck-swallow-breathe mechanisms. To protect their airways, preterm infants often engage in a pattern whereby sucking and breathing alternate in a burst-pause pattern. When sucking bursts become prolonged or when breathing pauses are too short or become interrupted, infant physiologic regulation may be surpassed with resultant oxygen desaturations. Desaturation events during feedings, defined as SpO2 < 90%, have been found to continue through 52 weeks postconceptional age. Recurrent episodes of hypoxemia, such as may occur through repeated feedings throughout the preterm infant's day, may alter cerebral blood volume which has been implicated in neonatal cerebral injury. The purpose of this study was to examine the significance and pattern of oxygen desaturations during preterm infant feeding near the time of discharge from the neonatal intensive care unit (NICU) and to explore the relationship of prolonged sucking bursts with the occurrence of oxygen desaturation.Twenty-one VLBW infants were videotaped bottle-feeding by their mothers near the time of discharge from the NICU. Infants were A3 1500 grams birth weight (mean 1153, SD B1 286), varied in respiratory health (36.4% required supplemental oxygen at 36 weeks post-conceptional age), were of similar feeding skill (all were able to nipple feed 50% of their feedings), and differed in post-conceptional age at time of study (mean 36.5, SD B1 1.6). Mothers had nipple fed their infants at least 5 times prior to the study feeding (mean 13, SD B1 11). Throughout the feeding, at one-second intervals, SpO2 was recorded and infant and maternal feeding behaviors were coded (sucking/not sucking, engaged/non-engaged/withdrawing; holding nipple still/jiggling). Most of the infants (71.5%) spent greater than 26% of their feeding time in a state of oxygen desaturation. Feeding length (M 20.8 min., range 6 - 43.5) and number of desaturation events per feeding (M 10.8, range 1-28) varied. Desaturation events were ranked as mild (SpO2 85-89; 59.6%), moderate (SpO2 80-84; 18.4%) and severe (SpO2 <80; 21.9%) and events had a mean duration of 30.1 seconds (range 1-416 seconds). The majority of events occurred during feeding (64.9%) while the remainder occurred during a rest period. Events were evenly distributed across infants' feeding time with 30.7% occurring during the first third of the feeding, 32% occurring during the middle third, and 37.3% occurring during the final third. Infants averaged 7 prolonged sucking bursts per feeding (defined as greater than 10 consecutive seconds sucking). Prolonged sucking bursts were associated with 24% of the desaturation events. Additional analyses will be presented to explore the relationship of infant health, maturity, and experience feeding to frequency and severity of desaturation events and to explore the relationship of the length of sucking burst with the occurrence and severity of oxygen desaturation. Findings from this study extend the understanding of the risk of bottle-feeding for preterm infants' oxygenation. Further study is needed to evaluate feeding interventions to modulate preterm infant breathing patterns during feeding.
poster
Despite evidence that breastfed fullterm infants are healthier (Wright et al., 1989) and more mature neonatally (DiPietro et al., 1987), only about 60% of infants are breastfed (Ross Laboratories, 1990). Researchers have identified maternal characteristics that predict breast feeding (i.e., maternal age, marital status, education, race, and number of previous births; Grossman, et al. 1990; Kuan et al., 1999), as well as infant characteristics (i.e, gestational age, birth weight, and infant complications; DaVanzo et al., 1990; Kuan et al., 1999). Most of these studies, however, were conducted with fullterm infant samples. Because of the higher prevalence of infant complications and the different maternal demographics associated with preterm birth, it is likely that such factors play a different role in determining breastfeeding in preterm infants. Prolonged hospitalization in preterm infants also may negatively impact the likelihood to breast feed due to the need for maternal travel and to express milk. As breastfeeding has been shown to differentially benefit preterm infants (Lucas et al., 1992), it is important to identify the characteristics that predict breastfeeding in this population.Methods. The hospital medical charts of all preterm infants (<37 weeks gestational age, N 3D 106) born in 1997 and admitted to the Neonatal Intensive Care Unit of a regional, rural hospital were reviewed (Level II with an exception to treat infants up to 26 weeks not requiring surgery). Maternal and infant characteristics were recorded from infant charts (insurance status, race, prenatal smoking, maternal age, and percentage of days on which parents visited during hospitalization; Apgar scores, infant sex, intubation, total number of days on oxygen, total number of days on intravenous feeding). Gestational age was assigned based on charted estimated due dates and the Ballard score. Feeding information also was obtained from the charted daily hospital records. Infants were classified as breastfed (BRF, n 3D 53) if they received breast milk at 10% or more of their feedings while hospitalized, or else as formula fed (FF). It was hypothesized that the maternal and infant characteristics would predict feeding group membership. Linear discriminant function analyses were used to test these hypotheses. Results and Discussion. Maternal characteristics significantly predicted feeding group membership [Wilks' (3D .78; F(5,54) 3D 3.02, p <.02]. Variables that significantly discriminated among BRF and FF groups were insurance status [F(1,58)3D6.89, p<.02, R23D.11], maternal age [F(1,58)3D10.27, p<.01, R23D.15], and percentage of days during hospitalization that parents visited [F(1,58)3D7.08, p3D.01, R23D.11]. Mothers who breast fed their preterm infants were older, had medical insurance, and visited the infant for a greater percentage of the hospital stay. Infant characteristics also significantly predicted feeding group membership [Wilks' (3D .85; F(7,95)3D2.38, p<.03]. Only birthweight differentiated between BRF and FF groups [F(1,101)3D3.89, p3D.05, R23D.04], with BRF infants weighing less at birth. Although a smaller percentage of preterm infants were breast fed (50%) as compared to what is reported in fullterm samples, similar maternal variables, such as maternal age, predicted breast feeding in this preterm sample. Insurance status is likely a proxy for socioeconomic status, a robust predictor of breastfeeding. The results highlight the role of the hospital environment, as the percentage of parental visit days during hospitalization also predicted breastfeeding. Unlike fullterm infants, lower birthweight predicted breastfeeding in this preterm sample, perhaps related to parents' increased concerns with infant health or increased breast feeding promotion by health care workers. Identification of these characteristics can be used to target interventions to increase breastfeeding in this population.