Abstract
Dear Editor:
If we now recognize that even the smallest or most routine neonatal care might seriously affect the subsequent development of premature infants, not too many studies deserve special attention on the part of neonatal practitioners as much as the enlightening article by Uygur et al. 1
More than 60 years ago, a tired mother asked her pediatrician, “Will cold milk hurt my baby?” In the following decades this subject was put on hold, pending clarification. Pioneering reports 2 showed no significant difference in child behavior or feeding intolerance, in neonates that were fed either warm or cold milk. They concluded that heating the feed was superfluous and carried no advantage to the infant. Most subsequent research was conducted with formula and described the physiological response of preterm infants to cold milk. These studies have found motility, sleep patterns, thermogenesis, and oxygen consumption to be independent of food temperature. In contrast, only one study found that larger gastric residuals were associated with feeds at room temperature when compared with feeds at body temperature. 3
As far as we know, the report by Uygur et al. 1 is the first to assess a wide range of clinical outcomes in preterm infants who are fed cold milk (mainly mother's own milk), to evaluate whether the existing routine of not monitoring feeding temperature may inadvertently contribute to infant morbidity. They report that infants fed with expressed breast milk at a temperature 32–34°C showed less frequent occurrence of apnea and smaller gastric residuals, and required less antireflux treatment or less time to regain birth weight, when compared with infants fed with milk at a temperature 22–24°C.
Even more, table 1 of the report contains a figure that should not be overlooked, that is, inotrope need. The significant finding in this study is that no infants (0%) fed with milk at a temperature 32–34°C needed inotropic drugs versus 17.5% of infants fed with milk at a temperature 22–24°C (p = 0.005). It is worth noting that table 1 of this report is different from data tables usually presented, as it displays not only basal characteristics, but also prematurity-related complications.
In line with figures from large databases, overall circa 9% of infants in Uygur's sample qualify for inotrope treatment, yet all of them belong to the cold milk group. It must be recalled that the use of inotropic drugs to treat neonatal hypotension is associated with higher risk of death and/or severe brain injury. Analysis of the Canadian Neonatal Network 4 demonstrated that this remained true even after blood pressure was included in the regression model, suggesting that it may be the inotrope use, rather than the presence of hypotension, which is harmful. Therefore, one of the most difficult challenges that clinicians face in the neonatal unit is to decide whether the hemodynamic changes are pathological or transitionally appropriate. Implementation of feeding closer to physiological conditions, with temperatures similar to those of breast milk, appears to be one of the most promising strategies for minimizing inotrope use while preventing failure of systemic perfusion. If, within this sample, it were possible to isolate all preterm infants who started feeds before needing inotropes, and on follow-up it was found that only babies fed cold milk were prescribed inotropes, it would strongly support this hypothesis. In contrast, if babies in the cold milk group need significantly more inotropes before starting feeds, not only this hypothesis must be discarded, but also the association between feeding temperature and apnea or feeding tolerance would be undermined. In any case, a fascinating aspect of Uygur's study is that it has the potential to provide a simple answer for a complex problem.