Challenges Our Facts -- Scientist Replies To Criticism
15 March 2001
received criticism from a reader that our recent interview with
Professor Robert Reynolds contained inaccuracies. In particular,
the reader said "the B6 level was set using more than one baby,"
and called this "an example of one inaccurate anecdote" in our interview
of 14 February 2001 where Professor Reynolds says:
1989 RDA for vitamin B6 in infants is based on a single infant.
And if you read the paper carefully, it looks like that infant
was born to an abnormal mother. But [the Institute of Medicine]
went and did an RDA anyway."
though the Institute of Medicine issued a new edition for the B
vitamins in 2000, the 1989 RDA is still used on labels today. The
new guidelines will not be used until the Institute completes its
next report (on macronutrients).
asked Professor Reynolds about our reader's challenge to his facts.
His reply is both detailed and thorough. First, he quotes the 1989
(10th edition) RDA book, which states on pages 145-146:
vitamin B-6 content and vitamin-to-protein ratio is generally
low in milk from nonsupplemented women, and there is evidence
of vitamin B-6 deficiency symptoms in infants breastfed by women
whose intakes are less than 2.0 mg/day and whose milk contains
less than 0.1 mg of vitamin B-6 per day (Kirksey et al, 1981;
Kirksey and Udipi, 1985; McCoy et al, 1985). In healthy babies,
vitamin B-6 intakes of 0.3 mg/day protected against abnormal excretion
of tryptophan metabolites following a test load (Bessey et al,
1957). General experience with proprietary formulas suggests that
metabolic requirements are satisfied if the vitamin is present
in amounts of 0.015 mg/g of protein or 0.04 mg/100 kcal (AAP,
1976; McCoy, 1978). The present subcommittee maintains the vitamin
B-6 recommendations of the ninth edition of the RDA -- 0.3 mg/day
during the first 6 months of infancy and 0.6 mg/day for older
he looks at the references cited, one at a time. The first is Kirksey
et al (1981), which is a chapter entitled "The Vitamin B-6 Content
of Human Milk" in a book that Reynolds co-edited. Reynolds says:
that chapter, Kirksey gives lots of data, most of which is devoted
to the vitamin content at different stages of lactation and effect
of maternal dietary intake of the vitamin on the vitamin content
in her milk, but there is nothing related to the vitamin requirements
for infants between 0 and 3 months of age!"
second reference is Kirksey and Udipi (1985), which is a chapter
entitled "Vitamin B-6 in Human Pregnancy and Lactation" in another
book that was co-edited by Reynolds. Reynolds says:
that chapter, they look at Apgar scores at delivery, at p50 values
at delivery, at the distribution of the different forms of the
vitamin in human milk, and the relationship between maternal dietary
intake of the vitamin and how that is reflected in the amount
of the vitamin delivered to her breast-fed infant, and how that
compares to the 1980 RDA (not how it compares to the requirement
of her infant, but to the RDA!). However, there is nothing
about the requirements of the infants!"
third citation, McCoy et al (1985), also appears in the 1985 book
that Reynolds co-edited. This chapter is entitled "Vitamin B-6 Intake
and Whole Blood Levels of Breast and Formula Fed Infants: Serial
Whole Blood Vitamin B-6 Levels in Premature Infants." Reynolds says:
off, notice that these were premature infants -- not full
term infants. We know that premies have lots of problems, not
just with vitamin B-6, and it's not clear how to separate each
of the confounding factors from one another. But again, there
is nothing about the requirements of infants 0 to 3 months
of age -- premature delivery or full term delivery. This paper
discusses only the effects of supplemental vitamin intake by the
mothers and how this is reflected in the blood of the infant.
This is not requirements! This is biochemistry."
other paper (1978) is also a chapter in a book (one that Reynolds
did not co-edit). Reynolds says:
page 258, [McCoy] states, 'A careful survey of the literature
failed to reveal experimental studies linking vitamin B-6 requirements
to protein intake in infants or children.' He then goes on to
discuss two early papers -- one by Snyderman et al (1953) and
one by Bessey (the one cited above in the 1989 RDA, to be discussed
Snyderman paper places mentally retarded infants (or as they called
them, severely mentally defective) on vitamin B-6 devoid
diets and sees when they begin to show symptoms of deficiency.
It's a rather gross paper to read, especially if you love kids.
(This experiment could not be done today.) In that paper,
they describe what happened to one subject: 'On the 76th day of
deprivation, subject M.L. suffered a series of severe convulsions
. . . (a description of the convulsions follows). Aside
from oxygen during seizure, no other therapy was given until he
received 50 mg of pyridoxine hydrochloride intravenously. Within
three hours he started to show signs of improvement.'
maintain that giving a single 50 mg bolus of PN does not constitute
determining the requirement of that infant. It was a classic
depletion/repletion study -- not a study designed to determine
the requirement of the infant."
AAP 1976, Reynolds says:
AAP citation refers to the vitamin B-6/protein ratio and is not
relevant to any basic research (of their own) on the requirement
of the vitamin for infants -- just the most appropriate vitamin
to protein ratio."
then examines the last of the cited papers, that of Bessey et al
(1957), and says:
studied two groups of infants. The first was formula-fed infants
with seizures -- where the vitamin in the formula was accidentally
destroyed by over-sterilization. The second group was 'Two breast-fed
infants with seizures (Cases 6 and 7) who had received neither
vitamin supplements nor solid foods in significant amounts. In
one infant (Case 6), convulsions ceased after the diet was changed
to an evaporated milk mixture; in the other, the same response
followed the therapeutic administration of pyridoxine hydrochloride;
in both, there was biochemical evidence of deficiency of vitamin
B-6. Analysis of the human milk from Case 7 indicated an extremely
low content of vitamin B-6.'"
states that Bessey et al also studied three other groups, but these
are not relevant to this discussion because they were controls and
older children. Reynolds continues:
et al gave increasing amounts of the vitamin to infants, then
gave them a tryptophan load test to see where their urinary excretion
of the tryptophan metabolites 'normalized' (whatever that means).
They stated '. . . until a level of intake of the vitamin
was found which just prevented excretion of xanthurenic acid after
a test dose of DL-tryptophan of 0.54 gm/kg.'"
doesn't know why they felt this criterion would indicate the infant's
requirement for B-6 -- but that is the subject of a different discussion.
Concerning the Bessey et al study, Reynolds continues:
relevant infants (data contained in Table II) are Cases 6 and
7, as mentioned above. Case 6 had no measure of the intake
of the vitamin! However, when it was given a formula which
provided 0.26 mg vitamin B-6 per day, convulsions ceased. For
Case 7, they did determine that this breast-fed infant was receiving
0.067 mg of the vitamin per day (from analysis of the mother's
milk), and its convulsions ceased when given the same formula
of 0.26 mg/day.
my math is way off, I count a grand total of one infant
in which the initial intake was determined (and which had a clinical
symptom of a vitamin B-6 deficiency) and then a single level
of the vitamin which caused the deficiency symptom to disappear.
So, all we know is that the requirement is somewhere between
0.067 and 0.26 mg/day, and those numbers come from one
infant. In its infinite wisdom, the RDA committee took this
0.26 mg/day and rounded it up to 0.3 mg/day. Voila! The RDA for
infants 0 to 3 months of age."
hopes this clarifies the issue raised by the reader, and says, "It's
too bad when otherwise responsible persons fail to carefully read
the literature, especially persons on the RDA subcommittes."
and safe upper levels are more political than scientific."
Interview with Professor Robert Reynolds of the University of Illinois