点击显示 收起
Johns Hopkins University School of Medicine
600 North Wolfe Street
Brady 320
Baltimore, MD 21287
E-mail: saavedramail{at}earthlink.net
Dear Sir:
We thank Falk et al for their comments and pertinent questions, which should help us to clarity the validity of our study results, better understand their potential application, and adequately assert the contribution of this study to the probiotic literature in infant nutrition.
Of course, we agree with Falk et al that breastfeeding is a major contributor to the immunologic health of infants, even when provided in limited amounts, and that this is a confounding variable. Nevertheless, exclusive formula feeding in this population would shorten the time that infants receive the formula, limit the range and total time of observation, and raise ethical concerns. Because the primary goal of this trial was to assess safety and tolerance in this population of infants attending day care, a design was considered that maximized the number of infants and length of follow up.
The children were included in the study on the basis of their parents' willingness to participate in the trial after being approached by the research team. Although still far from ideal, breastfeeding rates and length of breastfeeding in the United States have increased. Given that the longest possible length of follow-up of infants consuming study formulas is essential to reasonably assess their safety and potential health effects, a compromise in design was reached to allow the entry of infants who were already in the process of weaning from breastfeeding (ie, had started taking formula) before being approached for participation. Breastfeeding 3 times/d was a study exclusion criterion. For obvious ethical reasons, the study did not discourage the amount or the frequency of breastfeeding, which was determined entirely by the parents. This design allowed infants in the process of weaning to participate, thus increasing our recruitment rate, allowing for a longer duration of follow-up during the weaning period and beyond, and resulting in observations that were actually closer to a "real-life situation" for most of the infants.
Careful randomization yielded homogeneous groups relative to all baseline anthropometric and demographic characteristics. There were differences between the groups in the percentage of infants being breastfed at the time of randomization: 23% in the high-supplement group, 28% in the low-supplement group, and 15% in the placebo group. These differences were not statistically significant. However, we also analyzed the potential for breastfeeding as a confounder in several other ways.
First, we assessed in each study group the average number of times that infants who were still being breastfed were being put to the breast at the time of enrollment: 2.2, 2.7, and 2.2 times for the high-supplement, low-supplement, and placebo groups, respectively. Second, we analyzed in each group the average number of days of continued breastfeeding (duration of weaning) during study participation: 56.2, 53.3, and 49.5 d for the high-supplement, low-supplement, and placebo groups, respectively. Finally, we analyzed the average actual number of times per day that each infant in each group was being put to the breast while participating in the study, until they weaned completely to formula: 1.8 (range: 1.5–2.2), 2.2 (1.0–2.6), and 1.8 (1.0–2.1) times/d for the high-supplement, low-supplement, and placebo groups, respectively. None of these differences were significant.
In total, the infants studied were at least partially breastfed for a grand total of 1070 subject-days, or only 4.3% of the total 24 830 subject-days of follow-up. Although, ultimately, the true effect of breastfeeding cannot be ascertained, the effect on the outcomes observed could be reasonably considered minimal if at all contributory to any differences in outcomes.
With regard to colic and irritability, Falk et al are correct that definitions of colic vary, as does the age at which infants can be diagnosed as having colic. The purpose of the inclusion of "colic or irritability" in the weekly questionnaire was to identify—as perceived and reported by the parents–any potential and apparent abdominal discomfort not attributable to other changes in the child's routine. The ultimate reason for this was to identify any potential gastrointestinal intolerances (in conjunction with the other questions related to upper and lower gastrointestinal symptoms). Given the age span of the group in the study, we preferred to use the term "colic or irritability" to better describe the responses given by the parents to the standardized questionnaire. This is how the results and conclusions are reported.
Regarding antibiotic use, although a 1-d difference in antibiotic use per 100 subject-days may not appear to some to be "clinically significant," for a child in their first 2 y (730 d) of life, this equates to 7 d of antibiotic use or one standard course of antibiotic treatment for common illnesses requiring antimicrobials. Although this is obviously an extrapolation of the results, a decrease of one course of antibiotic treatment in a child receiving a formula containing probiotics in their first 2 y of life is not only clinically significant and relevant but is of potentially great epidemiologic effect, particularly in our current environment of indiscriminate antibiotic prescription and growing antibiotic resistance.
This is the most detailed and carefully controlled study yet to follow infants consuming a probiotic-supplemented formula for an extended period of time and to document safety and tolerance. The purpose of this single study was not to recommend changes across the board in standard clinical practice but to add to the growing body of evidence of safety and potential beneficial effects of specific agents for their use as probiotics in infant nutrition.