Letters

LETTERS

The best strategy for reducing iron deficiency

I very much enjoyed "Helping toddlers eat right" (March). Dr.Eden appropriately brings to our attention how difficult it is for parents--andpediatricians--to encourage and provide appropriate nutrition to this agegroup. Dr. Eden documents the seriousness of this problem and offers wonderfulsuggestions.

One idea that he avoids, however, as does the AAP, is the easiest andbest strategy for providing optimal nutrition, iron, and calcium to 1- and2-year-olds. That strategy is to encourage children in this age group tocontinue consuming either breast milk or iron-fortified formula until theyare at least 2 years old. This tactic has proven effective in reducing irondeficiency in children younger than 1. What is it that happens after thefirst birthday that makes most pediatricians prefer cow's milk from thattime on?

The excuse I hear most often for not suggesting continuation of breastfeedingor iron-fortified formula past the age of 1 is the added expense of formula.With breastfeeding cost is not an issue, of course. As for formula, it costsabout $1.80 a day more than cow's milk, and if you add the cost of multivitamin/irondrops to the cost of cow's milk, the difference is about $1.60. Strong backingfrom the AAP might provide an incentive for extra federal funding to theSpecial Supplemental Nutrition Program for Women, Infants, and Children(WIC) to help those who cannot afford this modest increase in food cost.Encouraging continuation of breastfeeding or breastfeeding complementedby infant formula would reduce costs even more.

Harry Pellman, MD

Fountain Valley, CA

The author replies: With Dr. Pellman's letter we now know thatat least two pediatricians are concerned about the iron status of 1- and2-year-olds in the United States. While I agree that continuing an iron-fortifiedformula through the second year of life would be an ideal strategy to preventiron deficiency, it would be much cheaper for WIC simply to supply iron-fortifiedvitamin drops to their 1- to 2-year-olds (at a cost of less than five centsa day). This is a rather small investment to help protect against the possibleloss of IQ points during this vulnerable period of rapid brain growth.

Alvin N. Eden, MD

Brooklyn, NY

More about misshapen skulls

I read with interest "Infants with misshapen skulls: When to worry"(February). As a clinical geneticist who focuses on craniofacial disorders,I was pleased to see a paper for the general pediatrician on this relativelycommon medical problem. In many ways the article was excellent, since itreviewed some important information about evaluation and management thatevery pediatrician should know. Furthermore, I was happy to see so muchabout the genetic basis for this group of disorders, which has been subjectto so many new findings in the past few years.

I found errors related to these genetics issues, however. With regardto incidence of craniosynostosis, for example, estimates have been as highas 1/1900, but not the 1 to 2.5/1000 stated in the article. The authorscite chromosome 4p, 5q, and 10q mutations as causes of craniosynostosissyndromes (the list of which was itself inaccurate). These are not chromosomalabnormalities, but mutations in FGFR3, MSX2, and FGFR2, respectively. Chromosomalalterations, not "mutations," are microscopically visible andinvolve the deletion or duplication of millions of base pairs. The causesof these syndromes are gene mutations at the single base-pair level.

In addition, the authors do not mention FGFR1 mutations, a cause of Pfeiffersyndrome; the FGFR3­ associated coronal craniosynostosis syndrome, themost common genetic form of craniosynostosis and by far the most importantfor pediatricians to know about; and TWIST mutations, the cause of Saethre-Chotzen,which is erroneously attributed to 10q. Last, using the phrase "deformationalmalformations" is like saying "noninfectious infection."A deformation is an abnormality caused by external or internal forces actingon a structure that was normally formed--that is, the genetic program forthe structure was normal. The leg position in frank breech babies is anexample of this phenomenon. A malformation refers to an abnormal structurecaused by an abnormality in the development of the structure, such as acleft lip. These two types of abnormalities obviously have far differentimplications for prognosis, recurrence risk, and risk of associated anomalies.

Nathaniel H. Robin, MD

Cleveland, OH

The article on misshapen skulls was very informative. It does, however,portray a common misconception, which the article states as follows: "Thesutures are largely ossified by the time a child is 8, and bone union iscompleted by age 20." This is not true. The only sutures that wouldfuse are those that are pathologic, synostotic, or intraosseous, such asthe four components of the occiput, which fuse in early childhood. The adultskulls we have all studied in anatomy lab have intact sutures. A small percentageof adults have a patent metopic suture.

Quite a bit of research has been done on the functional reason for thepatency of these sutures into adulthood. An excellent anatomical reviewon this topic came out of research at Michigan State University: The Craniumand Its Sutures by E. W. Retzlaff (now out of print). A distorted skullshape has an effect on venous and lymphatic drainage from the head and neckarea and associated conditions, such as otitis media, sinusitis, colic,and gastrointestinal symptoms in infants. The cranial base is often distortedalong with the vault in children whose heads are distorted. Any number ofsymptoms related to the physiology of the cranial nerves that exit the baseof the skull could be caused by the direct mechanical effect of the distortionon these nerves, their blood supply, and their corresponding function.

Robert Beckman, DO

Muskegon, MI

The authors reply: With regard to the incidence figures Dr. Robincites, Dr. Michael Cohen, who has published extensively on the genetic basisfor craniosynostosis, estimated in one of his early papers a minimum incidenceof 0.4/1000, recognizing that less severe and some late-onset cases of synostosisdo not come to medical attention (Cohen MM: Birth Defects 1979;15[5B]:13).With the evolution of pediatric craniofacial surgery, we certainly are seeingmore cases, many of which are managed without surgery. Nevertheless, evenin publications in earlier decades there have been reports of incidenceof 10/1000 (Gordon IR: Clin Radiol 1970; 21[I]:19) and even 16/1000 basedon skeletal collections (Bennett KA: Am J Phys Anthropol 1967; 27[1]:1).

Regarding the term "deformational malformations," we agreethat it is deceptive. In our article, we make a great effort to distinguishbetween deformations and malformations for the pediatrician. In our introductionwe state, "whereas synostosis implies a congenital developmental malformation,skull deformations ...arise from an alteration in the normal forces actingon the growing cranium." We go on to describe the need for pediatriciansto differentiate between these two entities and compare the theoreticaletiologies, diagnosis, and treatment of primary synostotic malformationswith secondary synostoses and skull deformations.

Dr. Robin's knowledge of the genetic basis for craniosynostosis is commendable.We opted not to mention the FGFR1 mutation as a possible cause of Pfeiffersyndrome, the FGFR3 association with coronal synostosis, and the TWIST mutationso that this multifaceted topic could be summarized for general review.Our table, which we agree would be better titled Craniosynostosis and GeneticMutations, is also in keeping with the object of providing a general reviewfor the pediatrician by examining basic anatomy, physiology and etiologictheories, ordinary clinical findings, and common treatment modalities.

Dr. Beckman raises some important questions about the cause of craniosynostosisand its management. Certainly, sutural physiology is very complex, and ourstatements about ossification and bone union attempt to simplify a difficultissue for a general review. Dr. Lawrence Becker and Dr. David Hinton havepublished their findings on sutural development. They describe the boneedges of the wide suture at birth as covered by an osteoblastic layer oflow cellularity. By 3 months of age, this layer has increased numbers ofosteoblasts with very increased cellular activity. Between 9 and 12 monthsof age, the suture narrows with a return to lower cellularity of this osteoblasticlayer. By 2 years of age, this layer is no longer apparent. The authorsdescribe woven bone across the suture consisting of a sclerotic band witha thin lucent center at 20 years of age. Only at 60 years of age is bonefusion appreciated and a lucent line just focally identified (Becker LEet al: Pediatr Neurosurg 1995;22[2]:104 and Hinton et al: J Neurosurg 1984;61:333).One can see that the time assignment of "sutural union" is vagueand may depend on whether one is addressing structural closure or functionalclosure.

Even more important, Dr. Beckman mentions the role of the skull basein craniosynostosis. This, too, is a complex issue currently under investigation,and we did not cover it thoroughly in our article. Distortion of the skulland skull base in craniosynostosis could certainly result in a wide rangeof symptoms stemming from altered mechanics. Indeed, the cranial base haseven been implicated as harboring the primary anomaly in craniosynostosis,with premature sutural fusion a secondary effect (Moss ML: Childs Brain1975;1:22).

Annie Jill Rohan, NNP/PNP

Sergio G. Golombek, MD

Stony Brook, NY

Alan D. Rosenthal, MD

Bronx, NY

When jaundice appears

The article "The jaundiced newborn" (April) is an excellentreview of the subject. It calls the pediatrician's attention to many importantissues.

However, there are some statements made that I disagree with and thatcould be quite misleading to your readers. For example, the assertion thatnewborns reach a "mean peak level of approximately 5 to 6 mg/dL onthe third day of life" is based on Dr. Gartner's own studies in whichthe population was entirely formula fed and hospitalized. Very large studiesin all racial groups that include a significant percentage of breastfedbabies (65% to 70%) followed as outpatients indicate that the mean bilirubinlevel is about 8 to 9 mg/dL and this occurs at about 96 hours. Furthermore,in these populations, the peak bilirubin level does not decline significantlysoon after that, but may remain elevated for as long as seven days. It maywell be that in certain breastfed populations (where nursing is frequentand highly effective) a higher bilirubin level may occur. Nevertheless,this is certainly not the rule in most parts of the world.

The authors note that when infant formula is substituted for breastfeeding,a rapid decrease of bilirubin level to one half its previous value can beexpected. In the only controlled clinical trial ever done on this subject,we showed that a far less spectacular decline in bilirubin levels is morelikely to occur. In a group of 26 breastfed infants whose bilirubin levelsaveraged 17.8 mg/dL at entry into the study, breastfeeding was discontinuedand formula supplemented. The average decrease in bilirubin after 48 hourswas 2 mg/dL (hardly a 50% decrease) and the level fell below 13.5 mg/dLby 48 hours in only five (19.2%) of the 26 infants (Pediatrics 1993;91:470).

The authors also state that physicians and nurses are capable of assessingthe severity of jaundice. A study done 58 years ago showed that this wasnot the case (Davidson et al: Am J Dis Child 1941;61: 958) and a 1998 studyby Moyer has confirmed this. The ability of even experienced physiciansand nurses to distinguish between mild, moderate, and severe jaundice ismarginal at best and, in some studies, is no better than guesswork.

The article also indicates that studies are needed of infants with "serumbilirubin concentrations >5 mg/dL in the first 24 hours of life and oldernewborns with serum bilirubin concentrations >12 to 14 mg/dL." Thereference quoted to support this opinion is a 1990 article by Newman inwhich he specifically debunks the idea of investigating such bilirubin levelsand shows that the overwhelming majority of such testing, including bloodsmears, hematocrits, hemoglobins, or reticulocyte counts yields almost nouseful information and is a waste of time and money. A similar study thatI reported in 1988 came to the same conclusion (Pediatrics 1988;81;505).

The authors go on to say that "measurement of glucose-6-phosphatedehydrogenase [is] rarely useful in the absence of anemia and can be performedselectively." It is true that measurement of G6PD is probably not indicatedin most populations, but in those few infants who have developed extremelyhigh bilirubin levels associated with G6PD deficiency, anemia almost neveroccurs. In fact, recent studies suggest that the hyperbilirubinemia associatedwith G6PD deficiency is primarily associated with a failure of bilirubinconjugation and not with hemolysis. These infants are rarely anemic andcannot be distinguished from the normal population by measurements of hematocritor reticulocyte count. Thus, the presence or absence of anemia is of nohelp in providing a clue to those babies who need screening for G6PD deficiency(Kaplan M et al: Clin Perinatol 1998;25:575).

The issue of phototherapy lamps is also addressed. Here a potentiallydangerous statement is made that "halogen lamps...are probably equallyefficient if the light is placed as close to the infant as is safely possible."Halogen lights cannot be safely placed close to the infant because theygive off considerable heat. Thus, when used as recommended by the manufacturers,they are not equal in efficiency to special blue fluorescent lamps. Also,it is not true that fiberoptic pads are efficient. They may help to preventa rising bilirubin level in an infant who has a modest elevation, but theycannot be recommended for infants whose bilirubin levels are above 20 mg/dL.Under these circumstances, they are much less efficient than special bluelamp phototherapy.

M. Jeffrey Maisels, MD
Royal Oak, MI

The authors reply: We thank Dr. Maisels for adding his very helpfulcomments to our recent review of neonatal jaundice. We would like to replyand elaborate further on these issues.

The "mean peak level of approximately 5 to 6 mg/dL on the thirdday of life" is certainly valid for the artificially fed infant, asour studies demonstrated previously and as Dr. Maisels agrees. The samemean peak value is also valid for the optimally breastfed infant, as demonstratedby studies from the United States, Japan, and Italy (summarized in GartnerLM: Semin Perinatol 1994;18;502). The optimally breastfed infant startsbreastfeeding during the first hour of life and is fed for an unlimitedtime at each breast with a frequency determined by the infant's early hungercues, usually 10 to 12 times per 24 hours during the early days of life.Continuous rooming-in and good education of the mother about breastfeedingare essential to achieve this optimal pattern. Under these circumstances,weight loss is almost always less than 7% of birth weight and weight gainbegins by the end of the first week of life. Unfortunately, the majorityof American infants are not breastfed in this manner and suffer from somedegree of nutritional deficiency. It is this degree of "starvation"that results in the increased bilirubin levels seen in many breastfed babies,as noted in Dr. Maisels' letter.

In our article, we provided data from one of Dr. Maisels' own studiesin which breastfed infants had elevated levels of bilirubin and greaterweight loss than formula-fed infants. Unfortunately, this study did notdefine or describe the breastfeeding techniques and timing used in the hospital.In recent years, clinical studies of breastfeeding have included data ontime of initiation, frequency and duration of nursing, use of supplementalwater, and other factors that may modify the establishment of successfullactation with optimal caloric intake for the infant. Thus, breastfeedingcomes in many different forms and these forms alter the pattern of neonataljaundice. Suboptimal breastfeeding certainly is a cause of increased serumbilirubin levels and prolongation of the time to initial peak. This is whatwe have called "breastfeeding jaundice" and this is presumablywhat Dr. Maisels is observing in his patients. Our article went to greatlengths to elaborate on this situation and the mechanism of it as well ason methods for prevention of such exaggerated bilirubin levels. The readershould not have been misled.

Dr. Maisels makes an important distinction. In our article, for the sakeof brevity, we lumped together breastfeeding-jaundice and breast milk­jaundicebabies when discussing the options for management. The reduction by halfin bilirubin levels when formula is substituted for breast milk appliesto breast milk­jaundice babies, those having their elevations of 20mg/dL or greater in the second or third weeks of life. Dr. Maisels is correctin pointing out that the reductions are not as great in the younger infantswith breastfeeding jaundice, the population studied in his report.

With regard to the ability of physicians and nurses to judge the intensityof clinical jaundice, we and Dr. Maisels are on the same page. Our statementwas, "With experience, physicians, nurses, and even parents can betrained to assess whether jaundice is severe enough to require laboratoryevaluation." We certainly did not wish to imply that anyone can assessjaundice with accuracy comparable to the laboratory. Yet, we all rely onour clinical judgment of jaundice to determine which infant requires laboratoryevaluation.

Dr. Maisels' comments on the need for laboratory workup in the jaundicedinfant are appreciated, but the fact remains that clinicians need guidelinesfor management that are readily available. The use of serum bilirubin levelsis one such guideline. Of course, other clinical data, including history,may modify these guidelines. We agree that most laboratory studies of jaundicefail to demonstrate an etiology, yet there is a need to identify the relativelyfew infants diagnosed with hemolytic disease because management of suchinfants is different.

In our article, we did not suggest that halogen lamps or any other lampshould be placed closer to the infant than is recommended by the manufacturerand is safe. The length of the article did not allow a full discussion ofthe application of phototherapy or relative advantages of one type of unitover the other. We refer the reader to Dr. Maisels' very detailed commentaryon the subject (Maisels MJ: Pediatrics 1996; 98;283).

Rajeev Dixit, MD

Chicago, IL

Lawrence M. Gartner, MD

Valley Center, CA

Do you have a comment on an article, a question to raise, an opinionto express? Let us hear from you!
Write to:

Letters to the Editor
Contemporary Pediatrics
5 Paragon Drive
Montvale NJ 07645-1742

Fax: 201-358-7260
E-mail address: cpletters@medec.com

We reserve the right to edit letters for clarity and length.

Iris Rosendahl. KLETSEPT. Contemporary Pediatrics 1999;9:13.