Zinc status response to folic acid supplementation and the effect of level of zinc intake on folate utilization in human...


Material Information

Zinc status response to folic acid supplementation and the effect of level of zinc intake on folate utilization in human subjects
Physical Description:
xvi, 287 leaves : ill. ; 29 cm.
Kauwell, Gail P. Abbott, 1952-
Publication Date:


Subjects / Keywords:
Food Science and Human Nutrition thesis Ph. D
Dissertations, Academic -- Food Science and Human Nutrition -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1993.
Includes bibliographical references (leaves 253-285).
Statement of Responsibility:
by Gail P. Abbott Kauwell.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001938878
oclc - 30963523
notis - AKB5028
sobekcm - AA00004737_00001
System ID:

Full Text







Copyright 1993


Gail P. Abbott Kauwell

I dedicate this dissertation to all of the people who

were instrumental in helping me achieve this goal and to my

steadfast, loyal and loving companions, Chelsea and Curly.


It is with great difficulty that I write this section of

my dissertation because it is impossible to properly

acknowledge everyone who deserves to be recognized and

thanked. Your support has come in many forms: encouragement,

ideas, storage space, equipment, supplies, technical

assistance, helping hands, friendly smiles and more. I do

remember and sincerely appreciate all the ways that you have

supported me through the trials and tribulations of completing

this research project and dissertation.

Although I do not wish to slight anyone's contributions

by not including their name in this acknowledgment, I would be

remiss if I did not take this opportunity to highlight some of

the people to whom I am most thankful. I am most grateful to

Dr. Lynn Bailey, the chairman of my supervisory committee,

whose encouragement, support, guidance and friendship has been

limitless. I also appreciate the guidance and support

provided by my committee members, Dr. Robert Cousins, Dr.

Jesse Gregory, Dr. Claudia Probart and Dr. Rachel Shireman.

In addition to my committee members, I want to recognize and

thank Dr. Susan Percival, who provided me with the opportunity

to work in her lab, and Peter Johnson for helping me smile

along the way. The subjects who participated in this study


also deserve to be recognized for the dedication, interest and

enthusiasm they displayed throughout the duration of this

research project. Last, but certainly not least, I am

thankful for the support, encouragement and thoughtful advice

that has always been graciously provided to me by Dr. Richard

Gutekunst, Dean of the College of Health Related Professions.











Folate .
Chemistry .
Metabolism .
Biochemical Fun
Recommended D
Folate .

TURE . 5

. . 5
. . 5
. . 7
actions .. .. 17
dietary Allowances for
. . 21

Effects of High Doses of Folic Acid .
Sources/Distribution and Stability of
Folate in Foods . .
Assessment of Folate Status .
Folate Status in Special Population
Groups . .
Effects of Environmental Factors on
Folate Status . .
Consequences of Compromised Folate
Status . .
Methods Used to Assess Folate Status and
Metabolism in Humans .
Zinc . . .
Chemistry . .
Absorption . .
Transport, Distribution and Metabolism .
Excretion . .
Zinc Homeostasis . .
Biochemical Functions . .

. . iv

. . ix

. . .. x

. . .. xii

. . xv

Recommended Dietary Allowances for Zinc 88
Effects of High Doses of Zinc ... 92
Food Sources of Zinc, Usual Intakes and
Bioavailability ... 94
Factors Affecting Zinc Status ..... 99
Assessment of Zinc Status .. 101
Metallothionein . .. 110
Zinc Deficiency . 116
Folate-Zinc Interactions .. 118
Effect of Impaired Zinc Status on Folate
Absorption and Metabolism 118
Effect of Supplemental Folic Acid on Zinc
Status . .. .. 122


AND METHODS . .. 145

Subjects .. . 145
Recruitment and Selection 145
Description of Subjects 146
Experimental Design .. 147
Materials and Methods .. 151
Description of Diet and Supplements 151
Procedures Used to Foster, Monitor and
Assess Compliance . 156
Procedure Used to Assess Adequacy of
Blinding . .157
Procedures Used to Prevent Zinc
Contamination ... 157
Urine Collection and Processing
Procedures. . 160
Blood Collection and Processing
Procedures . ... 160
Preparation of Diet Composites .. 163
Determination of Dietary Zinc Content 164
Biochemical Analyses . .. 167
Plasma and Urine Zinc Concentrations 167
Erythrocyte Lysate Zinc Concentration 168
Protein Determination .. 169
Erythrocyte Metallothionein
Concentration .......... 170
Determination of Serum Ferritin
Concentration . 175
Determination of Serum, Whole Blood and
Urinary Folate Concentrations 176
Determination of Urinary Deuterium-
Labeled Folate . 180
Statistical Analysis .. 184


5 RESULTS . . .

Effect of Supplemental Folic Acid on
Zinc Status . .
Plasma Zinc . .
Erythrocyte Zinc . .
Serum Alkaline Phosphatase .
Erythrocyte Metallothionein .
Serum Ferritin . .
Urinary Zinc . .
Effect of Level of Zinc Intake on Folate
Utilization . .
Serum Folate . .
Erythrocyte Folate . .
Urinary Folate . .


Effect of Supplemental Folic Acid on
Zinc Status . .
Effect of Level of Zinc Intake on Folate
Utilization . .









S 198
S 201
S 203

S 206
S 207



S 227

S 231




S 253




Table page

2-1 Comparison of 1980 and 1989 Recommended Dietary
Allowances (RDA) for folate . ... .21

2-2 Physiologic functions of zinc . ... 88

4-1 Average nutrient content of three-day diet with and
without supplements .. ..... 152

4-2 Composition of protein shakes ... 153

5-1 Mean weights ( SD) of subjects on zinc-restricted
and zinc-adequate diets . .. 187

5-2 Mean ( SD) usual calorie, protein, folate and zinc
intakes of subjects on zinc-restricted or zinc-
adequate diets . ... .187

5-3 Molar ratios of zinc to folic acid for each
treatment combination . 189

5-4 Overall mean ( SD) values of response variables
used to assess the effect of supplemental folic
acid on zinc status in subjects consuming zinc-
restricted or zinc-adequate diets 191

5-5 Overall mean ( SD) values of response variables
used to assess the response to folic acid
supplementation in subjects consuming zinc-
restricted or zinc-adequate diets ... .209

5-6 Overall mean ( SD) deuterium-labeled folate
excreted expressed as a percentage of total folate
intake, total urinary folate and oral dose 215


Figure page

2-1 Structure of folic acid and folic acid
derivatives . . 6

2-2 Folate-mediated one-carbon metabolism 8

4-1 Experimental design . .. 148

4-2 Schematic of enzyme-linked immunosorbent assay used
for determination of erythrocyte metallothionein
concentrations . . 173

5-1 Effect of supplemental folic acid on plasma zinc in
subjects consuming zinc-restricted or zinc-adequate
diets . . ... .. 192

5-2 Effect of supplemental folic acid on erythrocyte
zinc in subjects consuming zinc-restricted or zinc-
adequate diets . ... 194

5-3 Effect of supplemental folic acid on serum alkaline
phosphatase in subjects consuming zinc-restricted
or zinc-adequate diets . ... .197

5-4 Effect of supplemental folic acid on erythrocyte
metallothionein in subjects consuming zinc-
restricted or zinc-adequate diets .. 200

5-5 Effect of supplemental folic acid on serum ferritin
in subjects consuming zinc-restricted or zinc-
adequate diets . ... 202

5-6 Effect of supplemental folic acid on urinary zinc
in subjects consuming zinc-restricted or zinc-
adequate diets . ... 205

5-7 Serum folate response to folic acid supplementation
in subjects consuming zinc-restricted or zinc-
adequate diets . .. 208

5-8 Erythrocyte folate response to folic acid
supplementation in subjects consuming zinc-
restricted or zinc-adequate diets .. 210

5-9 Urinary folate response to folic acid
supplementation in subjects consuming zinc-
restricted or zinc-adequate diets ... .212

5-10 Urinary deuterium-labeled folate response to folic
acid supplementation in subjects consuming zinc-
restricted or zinc-adequate diets 214

5-11 Percent of oral labeled folic acid plus dietary
folate excreted as deuterium-labeled folate in
subjects consuming zinc-restricted or zinc-adequate
diets .... . ... 216

5-12 Percent of total urinary folate excreted as
deuterium-labeled folate in subjects consuming
zinc-restricted or zinc-adequate diets ... 217

5-13 Percent of oral deuterium-labeled folic acid
excreted as deuterium-labeled folate in subjects
consuming zinc-restricted or zinc-adequate diets 218























atomic absorption spectrophotometry

bovine serum albumin


carbon 14


carbon 14-labeled


Dulbecco's phosphate-buffered saline

deuterium-labeled folic acid


deoxyribonucleic acid

ethylenediamine tetraacetic acid

enzyme-linked immunosorbent assay


folic acid

folate-binding protein

formimino-glutamic acid

Food and Drug Administration






























gas chromatography-mass spectrometry


hydrochloric acid

high-performance liquid chromatography

tritiated pteroylmonoglutamic acid

human papillomavirus 16

International Units









Medical Research Council



sodium chloride

sodium hydroxide

sodium azide





para-aminobenzoyl glutamate


PBS phosphate-buffered saline

ppm parts per million

RDA Recommended Dietary Allowances

RNA ribonucleic acid

rpm revolutions per minute

SD standard deviation

THF tetrahydrofolate

MCi microcurie

/g microgram

pL microliter

AM micromolar

Mmol micromole

v/v volume/volume

y year

Zn zinc


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Gail P. Abbott Kauwell

August 1993

Chairperson: Dr. Lynn B. Bailey
Major Department: Food Science and Human Nutrition

Changes in zinc status in response to folic acid

supplementation and the effect of level of zinc intake on

folate utilization were evaluated in human subjects. Twelve

healthy men were randomly assigned to consume a zinc-

restricted diet (3.5 mg zinc/d) or an identical diet

supplemented with 11 mg of zinc as zinc sulfate (14.5 mg

zinc/d total) for 25 days. Half of the subjects in each group

received 800 gg/d deuterium-labeled folic acid. The remaining

subjects received a placebo. After an 80-day washout period,

the study was repeated with the folic acid/placebo treatments

reversed. Fasting blood samples and 24-hour urine collections

were obtained at corresponding time points during both study


Plasma zinc, erythrocyte zinc, urinary zinc, serum

alkaline phosphatase and erythrocyte metallothionein

concentrations were measured to determine zinc status. No

differences in mean values within or between the zinc diet

groups were detected (P20.05) for these response variables.

A significant (P<0.001) time by supplement by diet interaction

was detected for erythrocyte metallothionein; however,

significant differences in the change in erythrocyte

metallothionein concentrations due to the supplemental effect

were not detected (P>0.05) within the zinc diet groups. No

significant differences due to the effect of the supplement

were detected (P!0.05) within or between the zinc diet groups

for the remaining response variables. These data suggest that

short-term supplementation with 800 pg folic acid/d does not

adversely affect zinc status in healthy men.

To determine the effect of zinc intake on folate

utilization, total urinary, serum and erythrocyte folate

concentrations, and urinary excretion of deuterium-labeled and

unlabeled folate were determined. Differences in mean serum,

erythrocyte, total urinary or urinary deuterium-labeled folate

concentrations between subjects fed 3.5 or 14.5 mg zinc/d were

not detected (P20.05). Similarly, significant differences in

the percent of total folate ingested or the percent of total

urinary folate excreted as deuterium-labeled folate were not

detected (P>O.05) between the groups. These data suggest that

utilization of supplemental folic acid in healthy men is not

influenced by the level of zinc intake.



The crucial roles of folate and zinc in protein and

nucleic acid metabolism and genetic expression have

contributed to the intense research interest centered around

these two nutrients. Changes in intake, bioavailability

and/or metabolism of either or both of these nutrients can

have deleterious effects in humans and animals. Understanding

the specific mechanisms of action and relationships between

zinc and folate is essential to improving our ability to make

efficacious nutritional recommendations and policies that will

enhance the health and well-being of the human race.

Zinc is an integral constituent of at least 60 different

enzymes (Cousins and Hempe, 1990) including deoxyribonucleic

acid (DNA) transferase, ribonucleic acid (RNA) polymerases,

alkaline phosphatase, gamma aminolevulinic acid dehydratase,

carbonic anhydrase, carboxypeptidase, alcohol dehydrogenase

and glutamic, lactic and malic acid dehydrogenases (Vallee and

Galdes, 1984). Of particular interest with regard to studies

investigating the relationship between zinc and folate

nutriture was the discovery (Silink et al., 1975) that bovine

hepatic folate conjugase is a zinc metalloenzyme. The fact

that zinc is a component of so many metalloenzymes explains


the multiplicity of physiologic functions attributed to this

nutrient. Examples of these functions include development and

maintenance of the body's immune system, prevention of lipid

peroxidation, metabolism of energy containing nutrients,

hormonal interactions, bone formation and the replication and

differentiation of cells (Cousins, 1985; Cunnane, 1988;

Hambidge et al., 1986).

"Folate" is the generic term used in reference to the

many different naturally occurring forms of pteroyl-

monoglutamic acid and pteroylpolyglutamic acid. These

compounds have nutritional properties and chemical structures

similar to their parent compound, folic acid. Folate is a

coenzyme for many one carbon reactions, and like zinc, is

required for cell replication. Specifically, folate coenzymes

are essential for the synthesis of the pyrimidine,

thymidylate, which is required for DNA synthesis. Other

metabolic processes requiring folate coenzymes include the

interconversion of serine and glycine, methionine synthesis,

histidine degradation, methylation of biogenic amines,

generation of format and purine biosynthesis (Brody, 1991).

Recent studies suggest that supplemental folic acid may be

important in reducing the incidence of neural tube defects

(Bower and Stanley, 1989; Czeizel and DudAs, 1992; Medical

Research Council Vitamin Study Research Group, 1991; Milunsky

et al., 1989; Mulinare et al., 1988; Smithells et al., 1981);

modulating cancer risk in a variety of tissues including the


lung (Heimburger et al., 1987; 1988), cervix (Butterworth et

al., 1982; 1992a; 1992b), esophagus (Jaskiewicz et al., 1988)

and colon (Lashner et al., 1989); and reducing elevated

homocysteine levels, an emerging independent risk factor for

coronary heart disease (Kang et al., 1986; 1987).

Despite the potential benefits associated with folic acid

supplementation, and the fact that it is widely believed to be

nontoxic to humans under normal circumstances (DiPalma and

Ritchie, 1977), concern has been expressed regarding the

relative safety of ingesting additional amounts of this

nutrient. The major safety issues as outlined by Butterworth

and Tamura (1989) focus on: the potential harm to users of

anticonvulsant medications; interference with the diagnosis of

vitamin B12 deficiency; the possibility of other unexpected

adverse health effects; and interference with zinc absorption

or metabolism. For the most part, all but the latter safety

issue have been adequately addressed.

The question of whether supplemental folic acid disturbs

zinc absorption or metabolism, and thus zinc status, has been

examined by numerous researchers (Butterworth et al., 1988;

Fuller et al., 1987; Ghishan et al., 1986; Keating et al.,

1987; Krebs et al., 1988; Milne, 1989; Milne et al., 1984;

Mukherjee et al., 1984; Simmer et al., 1987; Tamura et al.,

1992; Wilson et al., 1983). The results from these studies

have been equivocal. Of these studies, dietary intake was

controlled in only one research design (Milne et al., 1984);


consequently, the results and conclusions of these studies may

vary depending on the adequacy and comparability of intake

among subjects. Discrepant results may also be attributed to

the diversity of research designs and protocols used, as well

as the lack of a satisfactory index of zinc nutriture.

The present study was designed to overcome some of the

limitations of previous investigations by evaluating the

response of male human subjects to 0 Ag/d and 800 ig/d of

stable-isotopically labeled (deuterium) folic acid under

controlled dietary conditions. The objective was to determine

if supplemental folic acid affected zinc status in subjects

consuming zinc-adequate (i.e. 14.5 mg/d) or zinc-restricted

(i.e. 3.5 mg/d) diets. Additionally, the use of deuterium-

labeled folic acid provided the opportunity to study folate

utilization under conditions of marginal and adequate zinc

intakes. This is the first time that folate utilization using

a stable isotope has been studied under controlled dietary

conditions. Another significant aspect of this study was the

determination of erythrocyte metallothionein concentrations in

addition to traditional measures of zinc status. Unlike other

indices of zinc status, the concentration of erythrocyte

metallothionein has been shown (Grider et al., 1990) to

respond quickly to acute dietary zinc deficiency and to

supplementation and may therefore be a more reliable and

sensitive indicator of zinc status.




Folic acid, or pteroylglutamic acid (2-amino-4-hydroxy-6-

methyleneaminobenzoyl-L-glutamic acid pteridine), consists of

three distinct subunits: a pteridine moiety, para-aminobenzoic

acid and glutamic acid (Figure 2-1). The pteridine moiety is

linked by a methylene bridge to para-aminobenzoic acid, which

is then joined by peptide linkage to glutamic acid. Although

mammals can synthesize all the components of this vitamin,

they are not capable of de novo biosynthesis because they lack

the enzyme needed for coupling the pteridine molecule to para-

aminobenzoic acid (Cooper, 1984). De novo synthesis of

folates does occur in plants and bacteria.

Folic acid is yellow and has a molecular weight of 441.4.

It is only slightly soluble in water in the acid form, but is

quite soluble in the salt form (Brody, 1991). Folic acid

occurs only rarely in nature, although it is the form most

commonly found in vitamin supplements and is the parent

compound of the naturally occurring folate vitamin forms.

A 0







-4 r-4
0 A


r-4 0

0 t


4 4-








Compounds having nutritional properties and chemical

structures similar to those of folic acid have been assigned

the generic descriptor "folate." Natural folates occur in the

reduced 7,8-dihydro- and 5,6,7,8-tetrahydro- forms. Folates

may also contain one-carbon substituent groups (Figure 2-1).

These include: 5-methyl-, 10-formyl-, 5-formyl-, 5,10-

methenyl-, 5,10-methylene-, and 5-formimino-tetrahydrofolate

(Brody, 1991). Most of these naturally occurring folates can

be degraded by heat, oxidation and/or ultraviolet light,

although some tetrahydrofolate derivatives are more stable

than others. For example, N5-methyl-tetrahydrofolate is

relatively heat stable, but is destroyed by acid, as is N5'10-

methylene- tetrahydrofolate (Krumdieck, 1990; O'Brion et al.,

1975). For the most part, dietary folates occur in the form

of pteroylpolyglutamates (Halsted, 1979) containing three to

seven glutamic acid residues (Figure 2-1). Although the

principal pteroylpolyglutamate in food is N5-methyl-

tetrahydrofolate, over 150 different forms of folate have been

reported to exist (Sauberlich, 1987).


Absorption. The first stage of intestinal folate

absorption involves the hydrolysis of pteroylpolyglutamates to

pteroylmonoglutatmes (Butterworth et al., 1969). Hydrolysis

is performed by the gamma glutamylcarboxypeptidases, commonly

grouped together and referred to as "folate conjugase".



= I '-4x

= I

0 Z4 3>1
> 0 0

0 4 o oos4)
0 2 -)ELL --
O 0 u. u. u.'5 s
-u-z 3 .

o=0 o-=U a =

SM o0

Sz Oz I

I I 0

oiO omO 0

1 I 0

0 7Z 0I^u -Z r

z z z



Folate conjugase successively cleaves the gamma glutamyl

peptide bonds of pteroylpolyglutamates to the monoglutamate

form. Reisenauer and Halsted (1987) have estimated that the

activity of human jejunal brush border folate conjugase is

sufficient enough to preclude this step from limiting the rate

of absorption. However, research showing that the

bioavailability of folate polyglutamates is considerably less

than that of folate monoglutamates (Gregory et al., 1991)

suggests that hydrolysis of dietary polyglutamyl folates is a

rate-limiting step in absorption.

Two separate folate conjugases have been identified in

the human jejunum (Reisenauer et al., 1977). One of these is

soluble and located intracellularly, while the other is

membrane-bound and concentrated in the brush border. These

folate conjugases differ with respect to molecular weight, pH

optima and inhibition characteristics.

Brush border folate conjugase is a zinc-dependent

exopeptidase that sequentially cleaves polyglutamates to

monoglutamates and has a pH optimum near neutrality (Chandler

et al., 1986; Reisenauer et al., 1977). Chronic alcohol

consumption (Naughton et al., 1989; Reisenauer et al., 1989),

zinc deficiency (Tamura et al., 1978), mucosal damage

(Halsted, et al., 1986) or exposure to naturally occurring

inhibitors in food (Bhandari and Gregory, 1990) may exert a

negative effect on brush border folate conjugase activity and

subsequent folate absorption. Interestingly, brush border


folate conjugase activity does not appear to be affected by

the aging process (Bailey et al., 1984a). In contrast to

brush border folate conjugase, intracellular folate conjugase

is found in the lysosomes of intestinal cells and functions as

an endopeptidase with an acidic pH optimum (Wang et al.,

1986). The role of intracellular folate conjugase is unknown;

however, Wang et al. (1986) have proposed that its role in

cellular folate metabolism is unrelated to the digestion of

dietary folates.

Transport of folate across the brush border membrane is

the second stage of folate absorption. This is a complex

process that has not been entirely elucidated. Transport

occurs mainly in the jejunum and is believed to involve a

carrier system that is saturable, pH dependent (i.e. pH

optimum of 6.0), energy dependent and sodium dependent (Rose

et al., 1978; Said et al., 1987; Selhub et al., 1983). This

carrier-mediated system is thought to include a folate-binding

protein located in the jejunal brush border membrane. This

folate-binding protein is either the transport protein or is

an important component of the intestinal transport system

(Reisenauer, 1980). Data from competitive inhibition studies

(Said, et al., 1987; Selhub et al., 1984) support the

conclusion that the transport system may be the same for all

monoglutamate forms of folate. The carrier-mediated system

becomes saturated at luminal concentrations of 10-20 AM

(Selhub et al., 1984). In addition to carrier-mediated


transport, nonsaturable absorption involving passive diffusion

may also occur (Selhub et al., 1983). Absorption of folate by

this process occurs linearly at much higher folate


Decreased hydrolysis of folate polyglutamates and/or

interference with the transport of folate mononglutamates

across the intestinal brush border membrane can inhibit folate

absorption. Studies concerning the impact of dietary

components (i.e. conjugase inhibitors, dietary fiber, etc.);

dietary composition; the actual forms of folate ingested;

nutritional status; alcohol consumption; and the effect of

nutrient interactions on the extent of folate absorption have

been reviewed by Bailey (1988). The conclusions drawn from

these studies are often conflicting, confirming the need for

further research employing more consistent and appropriate

experimental designs and methodologies. A case in point is

the nutrient interaction that is hypothesized to occur between

zinc and folic acid.

Nondietary factors that may interfere with folate

absorption include altered gastrointestinal function and the

use of medications. Damage to the intestinal epithelial

cells, such as that which occurs in Crohn's disease (Hoffbrand

et al., 1968), celiac disease (Halsted et al., 1977; 1978;

Hoffbrand et al., 1970) and tropical sprue (Corcino et al.,

1976; Halsted, 1980b), or changes in the intraluminal

environment due to achlorhydria (Russell et al., 1986), may


adversely affect folate absorption. Medications that may

interfere with folate absorption include sulfasalazine

(Franklin and Rosenberg, 1973) and diphenylhydantoin (Gerson

et al., 1972; Hoffbrand and Necheles, 1968; Rosenberg et al.,

1968). Antacids and histamine receptor antagonists may also

adversely affect folate absorption by increasing the

intestinal pH to levels that exceed the optimum for folate

conjugase activity and carrier-mediated transport (Russell et

al., 1979). Folate analogs such as methotrexate, trimethoprim

and pyrimethamine, which are used primarily as dihydrofolate

reductase inhibitors, also suppress intestinal folate

absorption (Selhub et al., 1983).

Transport. Upon entry into the intestinal mucosal cells,

reduced dietary folates are converted, for the most part, to

N5-methyl-tetrahyrofolate monoglutamate (Cooper, 1984; Pratt

and Cooper, 1971). This is the predominant form in the portal

circulation under normal physiologic conditions. Reduction

and methylation of folic acid can also occur, although much of

this form of the vitamin appears in the portal blood unchanged

(Cooper, 1984).

Serum folate binding proteins are involved with the

transport and distribution of folates to the liver and other

tissues. Two types of serum folate-binding proteins have been

identified (Wagner, 1985), one is specific with a high-

affinity binding capacity; the other, which is thought to be

albumin (Soliman and Olesen, 1976), is nonspecific, with a


low-affinity binding capacity. Approximately two-thirds of

serum folate is protein-bound, with the majority bound to


A small amount (10 to 20%) of absorbed folates in reduced

substituted form is taken up by the liver on the first pass,

while the majority is distributed to other tissues (Steinberg,

1984). Since the liver is capable of secreting Ns-methyl-

tetrahydrofolate monoglutamate into the bile, the potential

exists for continual enterohepatic circulation of folate

(Steinberg, 1984). This idea is not supported, however, by

the results of an isotopic labeling study (Krumdieck et al.,

1978) that examined the routes of folate excretion following

administration of [14C]folic acid to a human subject. The

results of this study suggested that a significant portion of

biliary folate was not reabsorbed.

Contrary to the fate of reduced substituted forms of

folate entering the portal hepatic vein, folic acid entering

this vessel is taken up almost exclusively by the liver. Some

of this folic acid is used for polyglutamate synthesis, and

the rest is converted to N5-methyl-tetrahydrofolate and

secreted into the bile (Lavoie and Cooper, 1974). When a

large oral dose of folic acid is administered, most of it is

recovered in the urine in its original form (Brody, 1991).

Tissue deposition and storage. Transport of folate

across cellular membranes is an energy requiring process,

where anion gradients may serve as the energy source


(Yang et al., 1984). The N5-methyl-tetrahydrofolate

transported across cell membranes must be converted to the

polyglutamate form to assume its role as a functional

coenzyme. Polyglutamates cannot cross biological membranes,

so polyglutamation also serves to trap folates inside cells at

concentrations that are one to two orders of magnitude greater

than those of the extracellular fluid. Conversion to the

polyglutamate form requires methionine synthetase, a vitamin

B12-dependent enzyme, and pteroylpolyglutamate synthetase.

Methionine synthetase is responsible for removing the methyl

group, and pteroylpolyglutamate synthetase is responsible for

the addition of glutamyl residues. The resulting

tetrahydrofolate polyglutamates usually contain between four

to seven glutamyl residues.

The total body folate content is in the range of 5 to 10

mg. The liver is considered to be the primary storage organ

containing about 50% of the total body folate (Herbert and

Colman, 1988). A variety of folate derivatives with various

chain lengths is present in the liver and other tissues, with

one particular species usually dominating (Shane, 1990).

Alterations in the distribution of folate polyglutamates have

been observed under certain physiological and nutritional

conditions. For example, longer chain length folates tend to

accumulate under circumstances whereby cellular folate

concentrations are diminished (i.e. folate deficiency;

methionine deficiency; etc.) (Cook et al., 1987). The


significance of this particular change has not been

determined, although it has been reasoned that extending the

chain length does not affect the short-term regulation of one-

carbon metabolism since this change occurs very slowly (Shane,


Utilization of endogenous folate pools may provide a

mechanism for maintaining normal folate supplies to critical

tissues during acute or chronic folate deprivation. A small

pool of intracellular folate monoglutamates is thought to be

available for this purpose. As deprivation continues, the

liver and other "storage tissues" (i.e. the kidney) may

generate folate monoglutamates through the action of

folylpolyglutamyl hydrolase on folate polyglutamates.

Additionally, the amount of folate presented to the liver is

thought to decrease, thereby reducing hepatic monoglutamate

uptake and new polyglutamate synthesis. Thus, over time,

hepatic folate stores decline.

Folate derived from dying cells may be another potential

source of this nutrient during periods of prolonged

deprivation (Steinberg, 1984). For example, erythrocytes

could make an important contribution to folate homeostasis

because of their high folate content and limited life span. A

study (Hillman et al., 1982) of heat-treated, senescent

erythrocytes has shown that labeled folate can be recovered in

the liver and bile. Therefore, folate salvaged from expired


erythrocytes may be redistributed through the enterohepatic


Catabolism and excretion. Urinary excretion of intact

folates in well-nourished humans consuming a nutritionally

adequate diet is approximately 5-40 Ag per day (Herbert,

1987). The folate compounds identified in the urine include

Ns-methyl-tetrahydrofolate, N10-formyl-tetrahydrofolate, N5'10-

methenyl-tetrahydrofolate and N5-formyl-tetrahyrofolate

(Chanarin, 1979). Excretion of intact folates is limited due

to renal tubular reabsorption of this nutrient and the degree

of catabolism that occurs in vivo. Products of folate

catabolism occurring in the urine include pteridines, para-

acetamidobenzoyl glutamate and para-acetamidobenzoate (Anon.,

Nutrition Reviews, 1990; Chanarin, 1979). The latter two

catabolites appear to be the major excretory products,

suggesting that the principal route of catabolism occurs by

cleavage of the C9-N10 bond. The cleavage mechanism has been

the subject of much controversy. Current in vitro studies

suggest the existence of more than one mechanism for folate

cleavage (Anon., Nutrition Reviews, 1990).

Folate can also be excreted in the feces. The amount

excreted in the feces has been reported, in some cases, to be

higher than the estimated dietary intake, presumably because

of folate synthesis by colonic bacteria (Brody, 1991).

Consequently, fecal folate excretion is not considered a

reliable index of folate metabolism.

Biochemical Functions

Reduction of the pteridine ring to the tetrahydro-form,

elongation of the glutamyl side chain and acquisition of one-

carbon units at the N5 and/or N10 positions of the pteridine

ring system must occur in order to produce the intracellular,

metabolically active forms of folate. Pteridine ring

reduction is accomplished by the cytosolic enzyme, 7,8-

dihydrofolate reductase. Folic acid and dihydrofolic acid can

serve as substrates for this enzyme. Reduction to the

tetrahydrofolate form, and demethylation, in the case of N5-

methyl-tetrahydrofolate, must occur before elongation of the

side-chain or acquisition of one-carbon substituent groups can

proceed. Side-chain elongation is achieved by the action of

folate polyglutamate synthetase. This is an adenosine

triphosphate (ATP)-dependent enzyme which joins glutamyl

residues to the vitamin by peptide bonds in an oligo-gamma-

glutamyl linkage. Once these steps have been accomplished,

one-carbon units at the oxidation level of format,

formaldehyde or methanol can be added. The major contributor

of one-carbon groups is serine, although formimino-glutamate

(FIGLU), formylglutamate and format may also serve as the

single-carbon source. The resulting tetrahyrofolate

derivatives (NS-methyl-, N5-formyl-, N10-formyl-, N5'10-

methylene-, N5'10-methenyl- and Ns-formimino-) are used as

coenzymes and serve as donors and acceptors of one-carbon

units in a host of reactions involved in amino acid and


nucleotide metabolism. Examples of specific biochemical

functions of folate include: purine and pyrimidine

biosynthesis; the generation and utilization of format; and

amino acid interconversions including the catabolism of

histidine to glutamic acid, the interconversion of serine and

glycine and the conversion of homocysteine to methionine.

These reactions are collectively referred to as one-carbon

metabolism (Brody, 1991; Krumdieck, 1990).

The cyclical nature of intracellular folate metabolism is

exemplified by tracing the potential fates of tetrahydrofolate

polyglutamates (see Figure 2-2). In mammalian tissues, it

appears that the major cycle of one-carbon metabolism involves

the conversion of tetrahydrofolate polyglutamate to N5'10-

methylene-tetrahydrofolate by serine hydroxymethyl-

transferase. This one-carbon derivative is a key intermediate

that can be: oxidized to NS' 1-methenyl-tetrahydrofolate (by

N5,10-methylene-tetrahydrofolate dehydrogenase) for use in de

novo synthesis of purines; used for the synthesis of

thymidylate (by thymidylate synthetase); or reduced to N5-

methyl-tetrahydrofolate (by N5' 10-methylene-tetrahydrofolate

reductase) for use in the biosynthesis of methionine. The

latter two reactions are irreversible reactions that compete

for the N5'10-methylene-tetrahydrofolate intermediate.

Regardless of the metabolic path into which the N5'10-

methylene-tetrahydrofolate is directed, its eventual fate

(under normal physiologic conditions) is the loss of its one-


carbon substituent group, and its return to either dihydro- or

tetrahyro-folate polyglutamate (Brody, 1991; Krumdieck, 1990).

Regulation of one-carbon transfer reactions is highly

complex. Changes in the concentration of substrates, products

and cofactors serve as a means for quickly activating or

inhibiting certain folate-requiring reactions. An example of

this complex regulatory process is one that has been given

much attention and involves the fate of N5'10-methylene-

tetrahydrofolate. As noted above and depicted in Figure 2-2,

this intermediate can be used to synthesize several different

products. Methylene-tetrahydrofolate reductase, the enzyme

catalyzing the conversion of N5'10-methylene-tetrahydrofolate

to N5-methyl-tetrahydrofolate, is very highly regulated. When

insufficient amounts of methionine are available, inhibition

of the reductase enzyme is relaxed allowing for increased

production of methionine from homocysteine. The methionine

produced by this reaction is converted to S-adenosyl

methionine, which then serves as a methyl donor to form

various methylated products. However, when S-adenosyl

methionine accumulates, methylene- tetrahydrofolate reductase

activity is curtailed through feedback inhibition. Methylene-

tetrahydrofolate reductase is also inhibited by the

accumulation of dihydrofolate polyglutamates, a situation that

occurs when thymidylate synthesis increases. Inhibition of

this enzyme under these circumstances allows for continued


commitment of folate to nucleotide biosynthesis (Brody, 1991;

Krumdieck, 1990).

In addition to the regulatory influences posed by changes

in the concentration of substrates, products and cofactors,

evidence for a second type of regulation has been reviewed by

Krumdieck (1990). This newly proposed regulatory process is

thought to be a slow-response mechanism based on covalent

modification of the polyglutamyl chain length. The idea that

a second form of regulation existed was generated from in vivo

studies demonstrating that the glutamyl chain length of

cellular folates varies in response to physiological or

pathological stimuli (i.e. developmental age, tissue

regeneration, infection, starvation, alcohol ingestion and

methionine-choline deficiency) that alter the steady-state

equilibrium of one-carbon metabolism. Additionally, the

finding that chain-length distribution of folate

polyglutamates differs from organ to organ in the same species

adds credibility to this concept because it is unlikely that

the requirements for one-carbon transfer reactions are the

same for all organs. Since changes in polyglutamate chain

length are slow to develop and only respond to persistent

stimuli, it is thought that the purpose of this regulatory

mechanism is to correct prolonged deviations rather than brief

fluctuations that may occur in an otherwise steady-state

(Krumdieck, 1990).

Table 2-1. Comparison of 1980 and 1989 Recommended Dietary
Allowances (RDA) for folate.

Category/age 1980 RDA* 1989 RDA**

gg/d Ag/d
0.0-0.5 30 25
0.5-1.0 45 35
1-3 100 50
4-6 200 75
7-10 300 100
11-14 400 150
15-18 400 200
19-24 400 200
25-50 400 200
51+ 400 200
11-14 400 150
15-18 400 180
19-24 400 180
25-50 400 180
51+ 400 180
Pregnant women 800 400
Lactating women
First 6 months 500 280
Second 6 months 500 260
*Data from Food and Nutrition Board, 1980.
**Data from Food and Nutrition Board, 1989.

Recommended Dietary Allowances for Folate

The 1989 Recommended Dietary Allowances (RDA) for folate

(Food and Nutrition Board, 1989a) represent a significant


reduction from the previously recommended amounts (Food and

Nutrition Board, 1980) for all age, sex and special population

categories (see Table 2-1). The rationale for the reduction

of the RDA for folate was based on two types of data: the

quantity of folate required to invoke established physiologic

responses or replace daily losses, after adjusting for

bioavailability and individual variability; and estimates of

dietary folate consumption related to the prevalence of

deficiency in population groups (Bailey, 1990a; Bailey, 1992).

Although some researchers (Herbert, 1987; Reisenauer and

Halsted, 1987) support the view that the existing data are

sufficient to warrant a reduction in the recommended level of

folate intake, others (Bailey, 1992; Sauberlich et al., 1987)

cast doubt on the appropriateness of the new RDA. In a

critique of the research used to establish the 1989 RDA,

Bailey (1992) submits that this new RDA may be insufficient to

provide an adequate margin of safety for specific populations.

The rationale for her conclusion is based on the following

facts: 1) in some cases, the population surveys used to make

conclusions about folate requirements were not specifically

designed to assess folate status; 2) the studies used to

support the 1989 RDA were not always comparable in terms of

the forms of folate used (i.e. synthetic folic acid versus

dietary folate); 3) the correction factor used to account for

folate bioavailability is only an estimate based on limited

data from studies employing different experimental


methodologies; and 4) the food composition tables for folate

are incomplete and may not accurately reflect the actual

amount of folate ingested or available. This final problem is

due to the fact that estimates of the folate content of foods

vary due to the method of analysis; the type of food consumed;

the method of food processing, preparation, storage and

handling; and the effect of nutrient interactions.

Effects of High Doses of Folic Acid

Folate is considered nontoxic in small doses as well as

in doses that exceed the RDA several hundredfold (Brody, 1991;

Butterworth and Tamura, 1989; DiPalma and Ritchie, 1977;

Herbert and Colman, 1988). The water soluble nature of

folate, and the apparent requirement for attachment to

saturable folate-binding proteins as a condition for storage,

probably account for the relative nontoxicity of this vitamin.

These features provide a mechanism for rapid excretion of

folate when the serum- and tissue-binding capacity are

exceeded (Herbert and Colman, 1988).

Adverse effects of supplemental folic acid were not noted

in adult humans receiving 400 mg/d for 5 months, or after 10

mg/d for 5 years (Brody, 1991). Similarly, no adverse effects

were reported in a group of well-nourished women ingesting an

oral folic acid supplement (10 mg/d) for 4 months (Butterworth

et al., 1982). Contrary to these reports, insomnia and

irritability were noted in subjects consuming folic acid in


the amount of 15 mg/d (Hunter et al., 1970). This effect has

not been confirmed in subsequent studies (Alhadeff et al.,


Despite the apparent relative safety of supplemental

doses of folic acid in normal adult subjects, large doses of

this nutrient (5 mg/d) (Brody, 1991) can have deleterious

effects when administered to individuals with undiagnosed

and/or untreated pernicious anemia. Administration of folic

acid supplements obscures the diagnosis of pernicious anemia

by correcting the macrocytic anemia associated with this

condition, but fails to alleviate the concurrent neurologic

lesions. Consequently, neurologic damage progresses

unchecked. Assessment of vitamin B12 status prior to the

initiation of folic acid supplementation can prevent this

potentially harmful outcome.

Large doses of folic acid (100 or more times the RDA) may

also be harmful to individuals with epilepsy who are receiving

continuous phenytoin therapy. Supplementation at this level

may precipitate convulsions (Herbert, 1987). Lower doses of

folic acid (0.1 to 1 mg/d) have not been shown to impair

seizure control (Roe, 1989).

It has been suggested that high intakes of folic acid may

interfere with zinc absorption and/or metabolism (Ghishan et

al., 1986; Milne, 1989; Milne et al., 1984; Mukherjee et al.,

1984; Simmer, et al., 1987; Wilson et al., 1983). This topic


will be addressed in more detail in a subsequent section of

this dissertation.

In contrast to some of the potentially harmful effects of

supplemental folic acid, there is evidence to suggest that

additional amounts of this vitamin may be beneficial in

modulating cancer risk in a variety of tissues including

cervical dysplasia (Butterworth et al., 1982; Butterworth et

al., 1992a; Butterworth et al., 1992b), bronchial metaplasia

(Heimburger et al., 1987; Heimburger et al., 1988), and

neoplasms associated with the esophagus (Jaskiewicz et al.,

1988) and colon (Lashner, et al., 1989). Supplemental folic

acid has also been associated with a reduction in the risk of

occurrence/recurrence of neural tube defects (Bower and

Stanley, 1989; Czeizel and Dudds, 1992; Medical Research

Council Vitamin Study Research Group, 1991; Milunsky et al.,

1989; Mulinare et al., 1988; Smithells et al., 1981).

Sources/Distribution and Stability of Folate in Foods

Although there is a need for more complete data

concerning the folate content and bioavailability of foods, a

considerable amount of information concerning food sources of

folate and the forms of the vitamin present in these foods has

been generated. The most concentrated sources of folate

include liver, citrus fruits, raw broccoli and dark green

leafy vegetables such as raw spinach. Cooked greens,

including spinach, turnip and mustard greens also contain


folate, but in lower amounts. The reduction in folate content

due to cooking occurs primarily due to leaching of the vitamin

into the cooking water, although thermal and oxidative

destruction may also occur. Legumes are another highly

concentrated source of folate; however, it must be recognized

that they contain heat-activated conjugase inhibitors which

may decrease the availability of folate polyglutamates. This

is true for other folate containing foods such as cooked

cabbage. Other good sources of folate include fortified

breakfast cereals. Most fortified cereals supply at least 25%

of the United States Recommended Daily Allowance for this

nutrient. In addition to the more concentrated sources of

dietary folate, it is important to consider the potential

contribution that foods containing modest amounts of folate

can have on total intake if these foods are consumed

frequently and/or in large quantities. Examples of these

include ground beef and whole-grain breads and cereals, as

well as tea (Bailey, 1990a; Bailey, 1992).

Very little folic acid is naturally present in food,

although this is the form used in food fortification because

of its exceptional stability characteristics (Gregory, 1989).

Folates in food occur almost exclusively in reduced

polyglutamate forms, with the predominant forms being N5-

methyl-tetrahydrofolate, N10-formyl-tetrahydrofolate and

unsubstituted tetrahydrofolate polyglutamates of varying

chain-lengths (Gregory, 1989). Reduced folates, with the


exception of N5-formyl- tetrahydrofolate, are potentially more

labile and subject to oxidation under aerobic conditions,

especially in the presence of heat, light and/or metal ions

(Cooper et al., 1978; Gregory, 1989). The actual stability of

folate in foods has been difficult to determine, with reported

losses of folate activity being highly variable. This

variability may be due to differences in oxygen exposure

during cooking, the amount of cooking water present and/or

intrinsic differences in ascorbic acid content of different

foods (Gregory, 1989). When partial or full oxidation of

tetrahydrofolate derivatives occurs, they may be further

catabolized to yield compounds that are physiologically

inactive with respect to human nutrition. If these compounds

are capable of supporting growth responses in microorganisms

used to measure the folate content of foods, overestimation of

the biologically useful folate content of foods could occur.

Assessment of Folate Status

The nutritional status of an individual indicates the

degree to which physiological needs for nutrients are being

met. Nutritional status is often evaluated using dietary

history and intake data, and biochemical and clinical


Dietary intake. Estimation of the dietary folate content

of foods and the evaluation of folate intake by individuals

and population groups is complicated by many factors.


Researchers and practitioners need to recognize the

limitations of the available folate databases and the problems

inherent in collecting dietary intake and food frequency data

when using this information to evaluate nutritional status.

Examples of problems associated with folate food composition

data include: the use of different analytical techniques to

determine the folate content of foods; missing information for

foods that have not been analyzed; failure to incorporate

information on the bioavailability of folate from various

foods; and the effect of different methods of food

preparation, storage and handling (Bailey, 1990a; Bailey,

1992; Gregory, 1989).

The accurate recording and evaluation of individual

dietary intake and food frequency data is also problematic.

For example, it is often difficult for individuals to

correctly remember the types and/or amounts of foods they have

consumed. Food selection may also be affected as a result of

being asked to record food intake. Factors such as age, mood,

intelligence, attention span, frequency of exposure to the

process and perceived importance of the information can also

affect the ability to recall and/or record food intake

information (Blake et al., 1989; Karvetti and Knuts, 1985;

Lissner et al., 1989).

Biochemical and clinical measures of folate status. The

progressive changes in biochemical and clinical parameters

occurring during folate depletion were determined and


described by Herbert (1962) in a depletion study in which he

consumed a folate-deficient diet for four months. As a result

of this experiment, Herbert categorized folate depletion into

four stages: early negative folate balance; nutrient

depletion; biochemical nutrient deficiency; and clinical

nutrient deficiency.

During the initial stage of folate depletion, serum

folate values become depressed. Serum folate concentrations

are very responsive to recent dietary intake, with levels

becoming low after consuming a folate deficient diet for only

two to three weeks. A value of less than 3 ng/mL is

indicative of negative folate balance. Values from 3 to 6

ng/mL represent marginally negative folate balance, with

normal serum folate values ranging from 6 to 25 ng/mL. The

sensitivity of serum folate to recency of dietary folate

intake make it a poor indicator of the degree of folate

deficiency. To determine the severity of folate depletion,

parameters indicative of body stores and changes in metabolic

function need to be measured concurrently (Herbert, 1987;


The second stage of folate depletion is characterized by

a decline in body folate stores. The largest amount of folate

is stored in the liver. When folate intake is deficient,

normal liver folate stores can be maintained for approximately

four months. Coincidentally, the average life span of normal

erythrocytes is four months as well. Since the erythrocyte


folate concentration is actually a measure of folate status at

the time the erythrocyte was synthesized, erythrocyte folate

concentrations usually parallel liver folate stores. This

relationship, combined with the greater accessibility of

erythrocytes, has resulted in the routine use of erythrocyte

folate concentration as a means for determining tissue stores.

Thus, stage two is identified by measuring erythrocyte folate

concentrations, which decrease to less than 140 ng/mL when

tissue stores are low. Erythrocyte folate values between 140

to 160 ng/mL suggest marginal depletion, and concentrations

above 160 ng/mL indicate normal folate status (Herbert, 1987;


Severe depletion of folate stores, characterized by

impaired folate-dependent metabolism, represents the third

stage of depletion. Folate coenzymes are required for many

metabolic functions including the synthesis of thymidylate.

A severe folate deficiency retards the synthesis of this

nucleotide, and thus interferes with DNA synthesis.

Subsequently, deranged DNA synthesis results in morphologic

changes in erythrocytes and neutrophils. Neutrophils become

hypersegmented because more constriction bands are formed,

constricting the DNA into more lobes. Hypersegmentation of

neutrophils is identifiable early in the course of impaired

metabolism because of the short half-life of these cells. The

criterion used to define hypersegmentation of neutrophils is

a lobe average equal to or greater than 3.5 lobes per cell


(Herbert, 1987; 1990). This hematologic alteration is thought

to be a sensitive screening tool although it is an unreliable

indicator of folate status during pregnancy (Herbert et al.,

1975) and in a small percentage (1%) of otherwise normal

adults with congenital polymorphonuclear leukocyte

segmentation (Herbert, 1964).

The morphologic changes manifested in erythrocytes during

the third stage of depletion include an increase in size and

conversion to an oval shape. At this stage, morphologic

damage is confined to the youngest erythrocytes, which are not

yet in the majority because of their longer life span, so

macroovalocytosis and an increase in the mean corpuscular

hemoglobin concentration are not evident. Other changes that

occur during stage three of depletion include a further

decline in liver and erythrocyte folate concentration

(Herbert, 1990).

The final depletion stage is clinically manifested as

normochromic, macrocytic anemia. At this point, the majority

of the erythrocytes are larger than normal resulting in an

increase in the mean corpuscular volume. The hemoglobin level

also declines due to decreased erythropoiesis. Further

decreases in liver and erythrocyte folate concentrations may

also be noted (Herbert, 1990).

The type of anemia caused by a folate deficiency is

clinically indistinguishable from that caused by a deficiency

of vitamin B12. Anemia due to vitamin B12 deficiency is


thought to be caused by a secondary deficiency of folate. It

has been hypothesized that this secondary deficiency develops

as a result of trapping folate in the N5-methyl-

tetrahydrofolate form.

The predominant form of folate in the serum, liver, and

most likely, other body storage depots, is N5-methyl-

tetrahydrofolate. Another source of this form of folate is

that which is synthesized from NS10-methylene-tetrahydrofolate

when control of Ns'10-methylene-tetrahydrofolate reductase is

relaxed. Conversion of N5-methyl-tetrahydrofolate to

tetrahydrofolate is catalyzed by the vitamin B12-dependent

enzyme methionine synthetase. When vitamin B12 is present in

adequate amounts, the methyl group from N5-methyl-

tetrahydrofolate is removed resulting in the regeneration of

tetrahydrofolate. This reaction is important for two reasons:

1) there is no other mechanism for regenerating

tetrahydrofolate in human cells; and 2) tetrahydrofolate is a

precursor to many other folate coenzymes, including N5'10-

methylene-tetrahydrofolate, which is ultimately used for DNA

synthesis. According to the "methyl-folate trap" hypothesis

(Herbert and Zalusky, 1962), folate gets "trapped" in the N5-

methyl-tetrahydrofolate form when vitamin B12 is deficient.

This form of folate is not usable for any other folate-

requiring reactions, including the production of thymidylate.

Inadequate production of thymidylate interferes with the

synthesis of DNA and eventually causes the development of a


macrocytic anemia that is hematologically identical to that

caused by folate deficiency (Herbert, 1990).

The need to properly identify the underlying cause of

macrocytic anemia prior to initiating treatment is essential

in order to avoid deleterious consequences. For this reason,

concurrent assessment of vitamin B12 and folate status is

recommended. As an alternative, the deoxyuridine suppression

test can be used to distinguish between macrocytic anemia

caused by a deficiency of folate, vitamin B12 or both of these

nutrients. As previously discussed, in the absence of folate

and/or B12 deficiency, deoxyuridine is converted to thymidine

by thymidine synthetase. Thymidine is subsequently

incorporated into DNA. The deoxyuridine suppression test is

an in vitro test that measures the activity of thymidine

synthetase by comparing the amount of unlabeled versus labeled

thymidine incorporated into DNA, when labeled thymidine is

added to bone marrow cells or phytohemagglutinin-stimulated

lymphocytes. Labeled thymidine incorporation is suppressed

when folate and vitamin B12 are present in adequate amounts.

Conversely, when a deficiency of folate or vitamin B12 is

present, the conversion of deoxyuridine to thymidine is

reduced and more labeled thymidine is incorporated into DNA.

When suppression of labeled thymidine is low, the underlying

cause can be determined by adding Ns-methyl-tetrahydrofolate

to the medium. If folate is deficient, the suppression rate

will be increased when this coenzyme form is added. If the


cause is due to a vitamin B12 deficiency, the addition of N5-

methyl-tetrahydrofolate will have no effect (Brody, 1991;

Herbert, 1990).

In the past, problems associated with sample collection

and processing techniques have limited the application of the

deoxyuridine suppression test to laboratory research settings.

The development of a method using a whole blood (0.1mL)

lymphocyte culture instead of cultures of separated

lymphocytes has made this test more suitable for use in

clinical laboratories, and perhaps, survey studies (Das et

al., 1980); however, as a measure of folate status, it is no

more informative than erythrocyte folate concentration (Tamura

et al., 1990).

The histidine load test is another method that can be

used to assess folate status. Formimino-glutamate, a product

of histidine catabolism, is further catabolized to glutamate,

ammonia and carbon dioxide by tetrahydrofolate formimino-

transferase, a folate-requiring enzyme. When folate is

deficient, urinary excretion of FIGLU is elevated, with

excretion being particularly high after an oral dose of L-

histidine. Despite the sensitivity of this test to folate

deficiency, it is not specific for folate deficiency since a

deficiency of vitamin B12 will also cause an increase in FIGLU

excretion. Additionally, FIGLU excretion is affected by other

diseases and physiological conditions, so its application is

usually limited to scientific studies (Brody, 1991).


Measurement of plasma homocysteine has recently been

proposed as a new method for assessing folate status. Under

normal conditions, approximately 50% of the available

homocysteine is converted to methionine by the folate-vitamin

B12-requiring remethylation reaction catalyzed by methionine

synthetase (see Figure 2-2). Inhibition of this reaction due

to vitamin B12 deficiency or inborn errors of folate or

vitamin B12 metabolism, results in an accumulation of

homocysteine in the blood (Krumdieck, 1990). Kang et al.

(1987) and Stabler et al. (1988) investigated the potential

association of folate deficiency with homocysteinemia and

found a negative correlation between serum folate

concentrations and protein-bound homocysteine. Stabler et al.

(1988), reported elevated serum homocysteine concentrations in

18 of 19 folate-deficient patients. Compromised folate status

was attributed to nutritional inadequacy in 17 of these

subjects. However, hyperhomocysteinemia was also found to

occur in patients with a deficiency of vitamin B12 (Stabler et

al., 1988), so the total serum homocysteine concentration must

be used in combination with other parameters in order to

distinguish folate deficiency from a deficiency of vitamin

B12. Determination of serum methylmalonic acid concentration,

which is normal in patients with folate deficiency, has been

recommended for this purpose (Stabler et al., 1988).

Folate Status in Special Population Groups

Folate status is affected by physiological changes

occurring at different stages of the life cycle. The

potential impact of growth, development and maturation on

folate requirements, metabolism, and subsequently, status, in

selected population groups, is summarized below.

Infants. The rapid rate of growth that occurs during the

first year of life influences folate requirements. Folate

requirements are also affected by the developmental immaturity

of the infant. For example, low secretion of pancreatic

proteases and gastric and biliary secretions during the first

months of life may significantly affect the bioavailability of

food folate and thereby influence folate requirements and

status (Picciano, 1990).

It has been reported that blood folate values of infants

at birth are higher than values for pregnant and lactating

mothers (Ek and Magnus, 1979) or normal adults (Smith et al.,

1985; Vanier and Tyas, 1966), and that these values decline

when solid foods are introduced (Smith et al., 1985). A study

(Smith et al., 1985) designed to assess the folate status of

infants supports the concept that milk alone (human or

proprietary formulas) is an important dietary source of folate

during the first year of life, and inclusion of this food can

provide sufficient folate to maintain blood folate

concentrations within an acceptable range.


Preterm, low-birth-weight infants have greater folate

requirements than term infants (Rodriguez, 1978). Depressed

serum and erythrocyte folate concentrations and megaloblastic

anemia have been observed in this population subgroup (Gray

and Butler, 1965; Roberts et al., 1969; Strelling et al.,

1966). The efficacy of supplemental folate given to low-

birth-weight infants was demonstrated by Burland et al.

(1971). After comparing the serum and erythrocyte folate

values of supplemented infants to those of unsupplemented

infants during the first nine months of life, these

researchers concluded that folate supplementation should be

provided to all low-birth-weight infants, but that more

research was needed to determine the optimal route, dose and

duration of therapy. Dallman (1974) has stated that folate

supplementation at the level of 50 jg/d for well infants

weighing less than 2,000 g at birth is warranted. The

Subcommittee on Pediatric Parenteral Nutrient Requirements

(Greene et al., 1988) of the American Society for Clinical

Nutrition has recommended 56 4g/kg/d of folic acid for

preterm infants receiving parenteral nutrition.

Adolescents. The rapid growth experienced by adolescents

is characterized by increases in lean body mass, skeletal

tissue and blood volume. Folate requirements are elevated

during this accelerated growth period because of the role

folate plays in cell division. Superimposed on the extra

demands for folate due to growth, are the effects of other


factors that may adversely affect folate status. These

include poor diet, alcohol and drug use, smoking and the

possibility of pregnancy in the sexually mature female.

Compromised folate status is prevalent among adolescents,

particularly those living in low-income households. Bailey et

al. (1982a) reported that 45% of rural black and white

adolescent males and females living in low-income households

had erythrocyte folate concentrations below 140 ng/mL. Serum

folate concentrations were less than 6 ng/mL in 56% of these

subjects. A similar trend was found in adolescents from urban

low-income households, where 42% of the subjects had

erythrocyte folate concentrations less than 140 ng/mL, and 45%

had serum folate concentrations below 6 ng/mL (Bailey et al.,

1982b). When assessed as a function of sexual maturation,

these researchers (Bailey et al., 1982 a; 1982 b) found that

serum folate concentrations declined as sexual maturation

progressed. A similar relationship was reported by Daniel et

al. (1975). Based on erythrocyte folate concentration, Tsui

and Nordstrom (1990) found the prevalence of folate deficiency

to be 13% among males and 40% among females. In the same

study, analysis of seven day food records revealed that for

all race, sex and age groups, subjects who were folate

deficient had significantly lower folate intakes than those

with normal folate concentrations. The type of dietary

pattern that contributed to inadequate folate consumption in

this study was not addressed, but a study by Bailey et al.


(1984b) suggests that poor folate status among adolescents may

be due in part to their limited consumption of vegetables and


Pregnancy. The RDA for folate is doubled during

pregnancy. This large increase in the recommended intake

reflects the fact that cell division and multiplication is

occurring rapidly. Increased dietary folate, along with

increased amounts of other nutrients, are needed to support

the physiological and compositional changes occurring at this

time. Increased maternal erythropoiesis, uterine and mammary

tissue expansion, placental and fetal growth and greater

urinary folate losses all contribute to the increased demand

for folate during pregnancy (National Academy of Sciences,

1990). During the final weeks of the normal gestational

period, the rate of active transport of folate across the

placenta is elevated. This places further demands on maternal

folate stores and may explain why serum and erythrocyte folate

values are usually several-fold higher in neonates than in

pregnant and lactating mothers (Bailey, 1990b).

As with other population subgroups, estimates of dietary

folate intake during pregnancy by women residing in the United

States are limited. A recent study (Huber et al., 1988) of

566 pregnant women who were primarily white, middle class and

at least 20 years old, found that only 8.5% of the women

derived their folate intake entirely from the diet. The mean

folate intake of this group was 257 Ag/d. The remaining women


(91.5%) consumed folic acid supplements and had a mean intake

of 1087 Ag/d. The women who did not take folic acid

supplements had significantly lower serum and erythrocyte

folate concentrations compared to those who used supplements.

Data from the 1985 Continuing Survey of Food Intake (United

States Department of Agriculture, 1987) indicated that the

mean folate intake by women (nonpregnant) between the ages of

19 and 34 (all income levels) was 217 gg/d. These data

suggest that unless women are motivated to make dietary

changes, or are instructed or knowledgeable enough to

recognize the need to take a folic acid supplement during

pregnancy, the recommended allowance for folate may not be

met. This is particularly disconcerting in light of the

protective effect of folate against neural tube defects

(Anon., Morbidity and Mortality Weekly, 1992).

A high prevalence of folate deficiency has been suggested

by population studies of pregnant women in whom blood folate

concentrations were measured. Herbert et al. (1975) studied

110 low-income, predominantly black or Puerto Rican women

living in New York City. These researchers reported that 20%

of the subjects had serum folate values below 3 ng/mL, and 16%

of the subjects had erythrocyte folate values below 150 ng/mL.

Bailey et al. (1980) analyzed blood samples from low-income

women in Florida and found that 29% had erythrocyte folate

values below 140 ng/mL. Bailey did not attribute the low

erythrocyte folate concentrations found in these studies to


hemodilution because erythrocyte folate concentrations are

indicative of the folate available to precursor red cells in

the bone marrow at the time the currently circulating cells

were developed (Bailey, 1990b).

The value of folic acid supplementation in the prevention

of folate deficiency during pregnancy was demonstrated in a

study of African women (Colman et al., 1975). Pregnant women

receiving folic acid supplemented cereals experienced

significant increases in serum and erythrocyte folate

concentrations, whereas the erythrocyte concentrations in

unsupplemented women decreased by an average of 42 ng/mL

during the last month of gestation (33 days). Despite a

hemoglobin concentration of 11 g/dL or more at the start of

the study, the supplemented women experienced an increase in

hemoglobin concentration suggesting that hematopoiesis was

limited by folate deficiency. These researchers suggested

that this finding gives further credence to the need for folic

acid supplementation in this population.

Not all researchers agree that low serum folate

concentrations during pregnancy are associated with maternal

complications or congenital malformations of the fetus (Hall

et al., 1976). For example, Hall et al. (1976) examined over

2700 women at four time points and found progressive

reductions in serum folate values at each stage. Maternal

complications and fetal outcome were not reported for any of

these pregnancies. Women from low socioeconomic groups,


smokers, multigravidae and women with twin pregnancies had

greater declines in serum folate. With the exception of the

smokers, however, significant reductions in mean serum folate

concentrations were not detected. It was concluded that this

decline was due to plasma volume expansion and did not warrant

routine folate supplementation.

Lactation. The effect of lactation on maternal and

infant status and milk folate content in unsupplemented women

has been described by several researchers. Smith et al.

(1983) found that blood folate concentrations were lower in

well-nourished, unsupplemented lactating women compared to

supplemented lactating women and normal nonlactating controls.

While the erythrocyte folate concentration of the

unsupplemented women declined, the folate content of their

milk was comparable to that of the supplemented women. There

was no difference in the blood folate concentrations of

infants of supplemented or unsupplemented mothers. Similar

results were found in a study described by Metz (1970). In

this study, lactating women fed a controlled low-folate diet

experienced rapid reductions in serum folate concentrations,

while the folate content of their milk remained constant.

When folate status during pregnancy and lactation is

severe enough to cause megaloblastic anemia, oral

administration of supplemental folic acid to lactating mothers

has been shown to improve their milk folate content (Cooperman

et al., 1982). Milk folate concentrations have also been


increased in lactating women of low socioeconomic status when

folic acid supplements were administered (Sneed et al., 1981).

However, serum folate concentration appears to increase, while

milk folate content remains constant when maternal blood

folate levels are already within an acceptable range (Tamura

et al., 1980). These findings suggest that a regulatory

mechanism controls the level of milk folate secretion (Tamura

et al., 1980) with the concentration of folate in breast milk

being maintained at the expense of maternal reserves and

status (Smith et al., 1983). Folic acid supplementation

during pregnancy and lactation may help to protect the folate

status of lactating women (Smith et al., 1983).

Elderly. While some researchers (Baker et al., 1978)

have suggested that age related changes adversely affect

folate absorption, other investigators (Bailey et al., 1984a)

have concluded that folate absorption is not affected by the

aging process. Although alterations in other aspects of

folate metabolism may exist, as intimated by data showing

depressed erythrocyte folate uptake in elderly subjects

(Ettinger and Colman, 1985), assessment studies of population

subgroups of the elderly suggest that socioeconomic and

environmental factors are probably the most important

contributors to the development of compromised folate status.

The potential effect of socioeconomic level on folate

status in the elderly population is apparent when comparing

assessment studies conducted in population subgroups with


different income levels. The majority of high-income elderly

subjects participating in studies conducted in New Mexico

(Garry et al., 1984) and Florida (Wagner et al., 1981) had

normal serum and erythrocyte folate concentrations. Only 3%

and 6% of the participants in the New Mexico and Florida

studies, respectively, had erythrocyte folate values below 140

ng/mL. Conversely, Bailey et al. (1979) found that 60% of

elderly Floridians from very poor socioeconomic backgrounds

had erythrocyte folate concentrations below 140 ng/mL, as well

as evidence of macrocytic anemia.

Environmental factors that may place the elderly at

higher risk for the development of compromised folate status

have been reviewed by Bailey (1990b) and Sauberlich (1990).

These factors include: institutionalization; chronic use of

prescription and/or nonprescription medications; the presence

of disease; and the consumption of alcohol.

Although not specifically addressed for each stage of the

life cycle covered above, it is important to consider the

impact that cultural food habits and customs, food

preferences, educational level, state of mental and physical

health, socioeconomic factors and food availability have on

food selection and nutrient intake. While physiological

changes occurring throughout the life cycle can have a

profound affect on nutrient needs and metabolism, Sauberlich

(1990) has concluded that the most common cause of compromised

folate status is inadequate dietary intake.

Effects of Environmental Factors on Folate Status

Prescription and nonprescription medications. Certain

prescription and nonprescription medications interfere with

folate absorption and/or metabolism, and depending on the dose

and duration of use, may result in compromised folate status.

Examples of these drugs include folate antagonists,

anticonvulsants, histamine receptor blockers, antacids, anti-

inflammatory agents, aspirin and possibly oral contraceptive


Folate antagonists are used in cancer chemotherapy or to

treat infections such as malaria. Methotrexate is an example

of a folate antagonist. This drug interferes with folate

metabolism by inhibiting dihydrofolate reductase, resulting in

a functional folate deficiency and anemia. Methotrexate can

also cause reversible mucosal ulceration (Roe, 1989). The

resulting damage to enterocytes can impair the absorption of

folate and other nutrients.

Rosenberg et al. (1982) have noted that the frequent use

of antacids and histamine receptor blockers among the elderly

poses a theoretical risk to this population in terms of their

folate status. Histamine receptor blockers such as

cimetidine, and antacids like sodium bicarbonate, may impair

folate status by increasing the intestinal pH to levels that

exceed the optimum for folate conjugase activity and carrier-

mediated transport and/or passive diffusion (MacKenzie and

Russell, 1976; Russell et al., 1979). However, folate


bioavailability was not reduced by chronic bicarbonate

administration in a rat bioassay (Hoppner and Lampi, 1988).

Furthermore, evidence of folate deficiency due to use of these

medications is lacking (Rosenberg et al., 1982).

Sulfasalazine is an anti-inflammatory agent frequently

used in the treatment of inflammatory bowel disease. Franklin

and Rosenberg (1973) have demonstrated that this drug

interferes with folate absorption. It appears that

sulfasalazine interferes with the absorptive process through

competitive inhibition (Strum, 1981), and by inhibition of

jejunal brush border folate conjugase activity (Reisenauer and

Halsted, 1981). Sulfasalazine also has the potential to

disturb folate metabolism since it can inhibit several hepatic

folate-dependent enzymes (Selhub et al., 1978). Patients

taking daily therapeutic doses of sulfasalazine are at risk of

developing folate deficiency and should be provided with

therapeutic doses of folic acid in order to prevent a

deficiency of this nutrient.

The interaction between folate and diphenylhydantoin, an

anti-convulsant medication used for treatment of epilepsy, may

actually be a two-way interaction (Reynolds, 1973; Rivey et

al., 1984). Chronic use of this drug has been associated with

folate deficiency and anemia, although the progression of the

deficiency to megaloblastic anemia is rare (Gerson et al.,

1972; Hoffbrand and Necheles, 1968; Reynolds, 1973; Rivey et

al., 1984; Rosenberg et al., 1968) and administration of


phenytoin has not been shown to affect the kinetics of folate

excretion (Krumdieck et al., 1978). Conversely, long-term,

high-dose folic acid supplementation of folate-deficient

patients taking diphenylhydantoin may result in lowered serum

concentrations of this drug in selected patients, with the

potential for loss of control of the seizure disorder

(Reynolds, 1973; Rivey et al., 1984).

The mechanism responsible for the development of

compromised folate status in patients receiving

diphenylhydantoin is unclear, but several hypotheses have been

espoused. These hypotheses have been reviewed by Rivey et al.

(1984). They include: drug inhibition of brush border folate

conjugase (Hoffbrand and Necheles, 1968; Rosenberg et al.,

1968); elevation of the intraluminal pH resulting in

malabsorption of dietary folate (Benn et al., 1971); drug-

induced impairment of folate transport into tissues (Krumdieck

et al., 1978); and induction of folate-requiring metabolic

processes in the liver (Maxwell et al., 1972). Studies

employed to prove and/or disprove each of these hypotheses

have provided seemingly contradictory results. Rivey et al.

(1984) have suggested that diphenylhydantoin may affect folate

homeostasis by multiple mechanisms.

Alter et al. (1971) and Lawrence et al. (1984) have

reported that aspirin in therapeutic doses can reduce serum

folate concentrations. After aspirin is discontinued, serum

folate values rapidly increase. Individuals with rheumatoid


arthritis, as well as those taking therapeutic doses of

aspirin for the prevention of heart attacks, may be

particularly vulnerable to the effects of chronic aspirin

ingestion. Elderly individuals are more likely to use

aspirin for these purposes, which may make them more

susceptible to folate depletion depending on the amount and

frequency of use.

Several investigators have noted an association between

the use of oral contraceptive agents and low serum and

erythrocyte folate values (Butterworth et al., 1982;

Pietarinen et al., 1977; Smith et al., 1975). Other

researchers have found no difference in folate status of oral

contraceptive users versus nonusers (Paine et al., 1975; Ross

et al., 1976; Whitehead et al., 1973). Results of studies

comparing blood folate concentrations of women using oral

contraceptive agents within six months of conception to those

of nonusers have also been equivocal (Bailey, 1980; Martinez

and Roe, 1977). Factors such as the specific formulation of

the pill (i.e. level of estrogen), duration of pill use and/or

dietary intake may affect the body's response to oral

contraceptive agents. Based on observations that

megaloblastic changes in the cervical epithelium of women

taking oral contraceptive agents were reversed with oral folic

acid supplements (Whitehead et al., 1973), some researchers

have speculated that oral contraceptive use may lead to a

localized folate deficiency in the cervix (Butterworth et al.,


Smoking. The effects of smoking in nonpregnant women

were investigated by Wittier et al. (1982). The mean serum

and erythrocyte folate concentrations of the smokers who

participated in this study were lower than those of the

nonsmokers. No differences between smokers were detected,

however, with regard to the number of cigarettes smoked per


Heimburger et al. (1987; 1988) compared the folate status

of male smokers to that of a control group of nonsmokers and

found significantly lower serum and erythrocyte folate

concentrations in the smokers. Smokers with metaplasia had

lower blood folate concentrations than did smokers without

metaplasia, and folate concentrations appeared to decrease

with increasing severity of metaplasia. These studies suggest

that smoking may adversely affect folate status.

Alcohol. Compromised folate status is common in chronic

alcoholics and is probably caused by a combination of factors

including: poor dietary habits; intestinal malabsorption;

decreased hepatic uptake; and increased urinary folate

excretion (Halsted, 1980). The effects of binge drinking on

folate absorption have been illustrated by studies using

orally administered radio-labeled folic acid. Evidence of

folic acid malabsorption was provided by decreased plasma

concentrations of radioactivity after oral doses of tritiated


pteroylmonoglutamic acid ([3H]PteGlu) (Halsted et al., 1967)

and by reduced luminal disappearance of [3H]PteGlu during

jejunal perfusion (Halsted et al., 1971). The underlying

defect responsible for the limited jejunal uptake of folic

acid was thought to be due to inadequate intake of dietary

folate during chronic alcohol ingestion (Halsted et al.,

1971). As discussed by Halsted (1990), a subsequent human

study that examined the effects of a folate deficient diet on

the absorption of [3H]PteGlu (Halsted et al., 1973), and an

absorption study conducted with monkeys fed diets containing

50% of their calories as ethanol (Romero et al., 1981),

prompted researchers to hypothesize that folate malabsorption

in chronic alcoholism results from the combined effects of

folate deficiency and ethanol exposure.

Recent research has attempted to delineate the

pathogenesis of folate malabsorption occurring in alcoholics.

To separate the effects of poor diet from those of alcohol

exposure, studies using miniature pigs exposed to alcohol for

a short period of time and in the absence of folate deficiency

have been performed (Naughton et al., 1989; Reisenauer et al.,

1989). These studies suggest that inhibition of brush border

folate conjugase, resulting in decreased hydrolysis of folate

polyglutamates, may be the earliest functional lesion

contributing to folate malabsorption and deficiency in

alcoholism (Halsted, 1990).


Disease. Folate deficiency has been associated with a

variety of diseases including certain hemolytic diseases

(Brody, 1991), cancer of the head and neck (Brody, 1991),

inborn errors of metabolism (Brody, 1991) and diseases of the

intestinal mucosa (Corcino et al., 1976; Halsted et al., 1977;

1978; Hoffbrand et al., 1968; 1970). The proposed underlying

etiology of folate deficiency in these diseases is defective

absorption and/or altered folate metabolism. These defects

lead to an increased folate requirement. For example, the

increased damage to red blood cells occurring in hemolytic

diseases results in increased cell division in the bone

marrow. This increase in cell division is thought to elevate

the requirement for folate (Brody, 1991). Altered metabolism

may also be responsible for the folate deficiency that

develops in patients with certain types of cancer,

particularly cancer of the head and neck (Brody, 1991). The

effects of altered folate metabolism in these individuals is

independent of the effects of anticancer drugs such as

methotrexate which can also adversely affect folate status.

Patients with inborn errors of folate metabolism may have

lower levels of folate-dependent enzymes, as well as defective

folate absorption, thereby increasing their folate requirement

(Brody, 1991).

Malabsorption is thought to be the major defect

contributing to compromised folate status associated with

untreated gastrointestinal diseases such as celiac sprue and


tropical sprue. Luminal disappearance of [3H]PteGlu and

pteroyl [14C]-glutamylhexaglutamate ([14C]-PteGlu7) in patients

with celiac sprue (Halsted et al., 1977; 1978) and tropical

sprue (Corcino et al., 1976) is significantly less than

luminal disappearance in normal subjects. When appropriate

medical and nutritional management of these diseases is

instituted, luminal disappearance increases significantly

(Corcino et al., 1976; Halsted et al., 1977; 1978). Further

evidence corroborating the idea that malabsorption adversely

affects folate status, at least in terms of celiac sprue,

comes from two studies. A significant decrease in the

hydrolysis of perfused [14C]-PteGlu7 in patients with celiac

sprue was noted in one study (Halsted et al., 1977). The

other experiment showed that brush border folate conjugase

activity in jejunal biopsy specimens taken from patients with

celiac sprue was significantly lower than the activity of this

enzyme in specimens from normal subjects (Halsted et al.,


Consequences of Compromised Folate Status

Megaloblastic anemia is probably the most commonly

recognized clinical problem associated with folate deficiency,

but research is beginning to identify many new potential

consequences of compromised folate status. Negative outcomes

more recently associated with folate deficiency include neural

tube defects, compromised infant birth weight, increased


potential for neoplastic changes and altered immune function.

A brief review of the evidence linking folate deficiency to

these problems is presented in this section.

Neural tube defects. Neural tube defects, which include

spina bifida, anencephaly and encephalocele, are among the

most common severe congenital malformations. The idea that

folate deficiency contributed to the causation of fetal

malformations was introduced almost thirty years ago by

Hibbard (1964) and Hibbard and Smithells (1965). Smithells

et al. (1980; 1981) conducted the first intervention trial

which suggested that supplementation with folic acid (0.36

mg/d) or other vitamins near the time of conception might

reduce the risk of recurrence of birth defects categorized as

neural tube defects. The results of this study could not be

taken as definitive proof of a protective effect for folate,

however, due to the co-administration of folic acid with other

vitamins, the lack of randomization of subjects to the control

and treatment groups and the absence of a double-blind design.

Another early intervention trial was conducted by

Laurence et al. (1981). This was a small study that employed

a controlled, randomized, double-blind design to test the

effects of folic acid supplementation (4 mg/d) alone. The

data provided inconclusive results when analyzed using the

original randomization scheme; however, when the data for

women who did not take their supplements were transferred to

the control group, the supplemented subjects had a


significantly lower recurrence rate. Preliminary results of

a study using a similar design and a small number of subjects,

found a reduction in the recurrence rate after a

supplementation program had been initiated, compared to the

recurrence rate before a supplementation program had been

established (Holmes-Siedle et al. 1982).

These early studies supported the concept of a protective

role for folic acid, but their flawed designs precluded the

ability to make any definitive conclusions and

recommendations. The results of a recent study (Medical

Research Council (MRC) Vitamin Study Research Group, 1991),

however, have established that supplementation with folic acid

around the time of conception can decrease the risk of neural

tube defects in women who have previously given birth to an

affected infant. This study was a large, multi-center,

double-blind intervention trial that randomized participants

to one of four groups: 1) folic acid alone; 2) folic acid plus

other vitamins; 3) other vitamins alone; and 4) no

supplements. This design allowed the researchers to determine

if vitamins other than folic acid conferred a protective

effect. The results demonstrated that folic acid, rather than

the combination of other vitamins, was responsible for the

improved outcome.

While the MRC Vitamin Study Research Group (1991)

findings are positive with respect to folate's role in

reducing the risk of recurrent neural tube defects, they do


not address the association of folate intake with the

occurrence of neural tube defects. In the United States, most

(approximately 95%) of the neural tube defect-affected infants

and fetuses occur in pregnancies of women who have not

previously given birth to an infant with a neural tube defect.

Thus, in order to reduce the overall prevalence of neural tube

defects, folic acid supplementation must exert a protective

effect on the developing embryos of women who have no history

of a neural tube defect-affected pregnancy. The MRC study

also leaves open to question the minimum dose and form of

folate needed to confer a protective effect, since subjects

were given daily supplements containing 4 mg of folic acid.

Over the last five years, several observational studies

have been conducted to evaluate the impact of periconceptional

folic acid/vitamin supplementation on the occurrence of neural

tube defects. Three of these studies were case-control

studies (Bower and Stanley, 1989; Mills et al., 1989; Mulinare

et al., 1988) and one was a prospective cohort study (Milunsky

et al., 1989). All of these studies involved women who had no

history of a neural tube defect-affected infant/fetus. The

occurrence of neural tube defects in women who reported taking

multivitamins containing folic acid (approximately 0.4-0.8 mg

folic acid per day) for at least one month prior to conception

through the first trimester of pregnancy were compared to

women who did not take supplements. Dietary intake of folate

was also considered in some studies. A protective effect was


associated with multivitamin supplement use and/or higher

levels of dietary folate intake in three (Bower and Stanley,

1989; Milunsky et al., 1989; Mulinare et al., 1988) of the

four studies.

Two new studies have provided additional support in favor

of a protective effect of folate. One study was a randomized

controlled trial conducted in Hungary (Czeizel and Dudds,

1992). Subjects were 18-35 years old, were not pregnant at

the time of recruitment and had no history of infertility or

fetal death. The volunteers were randomized to receive either

a placebo or a multivitamin supplement (including 0.8 mg folic

acid). There were no cases of neural tube defects in the 2104

participants receiving the multivitamin, whereas, the placebo

group had six occurrences and 2046 unaffected pregnancies.

Although the criteria for inclusion did not specifically

exclude women with a history of a neural tube defect-affected

pregnancy, it is likely that most women falling into this

category would have already been recruited for the MRC study

and would not have been available to participate in the study

just described. Under this assumption, these results have

been interpreted as providing evidence that folic acid can

reduce the risk of occurrence of neural tube defects; and that

the quantity of folic acid needed to produce such an effect is

much less than the 4 mg daily dose used in the MRC trial.

The results of another study examining the rate of

occurrence of neural tube defects in folate supplemented and


unsupplemented women has recently been published (Werler et

al., 1993). This large case-control study was conducted in

Boston, Philadelphia and Toronto. The control group consisted

of 2615 infants with birth defects other than neural tube

defects and oral clefts. This group was compared to 443 cases

with neural tube defects. The prevalence of use of folic

acid-containing multivitamins during the periconceptional

period was compared between mothers of cases and controls.

Daily folic acid supplementation was found to reduce the risk

of occurrent neural tube defects by 60%. Since the most

commonly used dose of folic acid was 0.4 mg, these data were

considered to be consistent with the hypothesis that 0.4 mg

supplemental folic acid per day is sufficient to decrease the

risk of neural tube defects among pregnancies of women in the

general population.

In the wake of the findings of the MRC trial (1991) and

the studies by Czeizel and Dudas (1992) and Werler et al.

(1993) the data from observational studies have been

reevaluated. Taken together, the results have been

interpreted as providing support for the hypothesis that folic

acid will decrease the risk of occurrence, as well as the

recurrence, of neural tube defects. Accordingly, the United

States Public Health Service has recently recommended that:

All women of childbearing age in the United States who
are capable of becoming pregnant should consume 0.4 mg of
folic acid per day for the purpose of reducing their risk
of having a pregnancy affected with spina bifida or other
NTDs. Because the effects of higher intakes are not well
known but include complicating the diagnosis of vitamin


B12 deficiency, care should be taken to keep total folate
consumption at <1 mg per day, except under the
supervision of a physician. Women who have had a prior
NTD-affected pregnancy are at high risk of having a
subsequent affected pregnancy. When these women are
planning to become pregnant, they should consult their
physicians for advice. (Anon., Morbidity and Mortality
Weekly, 1992, p. 1)

Recognition of the importance of folate in reducing the

risk of bearing a child with a neural tube defect is certain

to stimulate research focused on understanding the mechanism

whereby folate exerts its protective effect. Particular

attention is likely to be directed at identifying the

existence of underlying defects in folate metabolism.

Compromised birth weight. Folate deficiency occurring

during pregnancy has been associated with low infant birth

weight. Several researchers have reported increased infant

birth weights when folic acid supplements were given to

folate-deficient pregnant women (Baumslag et al., 1970;

Iyengar and Rajalakshmi et al., 1975; and Rolschau et al.,

1979). Placental weights were measured in two of these

studies (Iyengar and Rajalakshmi et al., 1975; Rolschau et

al., 1979), and a positive association with birth weight was

noted. These findings suggest that folic acid supplementation

improves fetal outcome by improving nutrition via increased

placental size (Bailey, 1990b). Additional evidence

suggesting a positive relationship between folic acid

supplementation and infant birth weight has recently been

published by Goldenberg et al. (1992). These researchers

found that maternal serum folate concentrations, occurring


within a range suggestive of compliance with the

supplementation regimen, were associated with higher infant

birth weight and a decreased rate of fetal growth retardation.

Neoplastic changes. The potential relationship between

folate and cancer was identified as early as 1944 when several

researchers noted that large doses of folic acid directly

interfered with the growth of certain tumors. Subsequently,

it was found that administration of folate antagonists

produced beneficial effects in patients with certain forms of

cancer. This discovery, although it revolutionized the

treatment of childhood leukemia and other forms of cancer,

probably contributed to the paucity of further research

regarding the potential anticancer effect of folate

(Butterworth, 1991). Renewed interest in the anticancer

potential of folate has been sparked by recent findings

suggesting that inadequate dietary intake of folate and/or

compromised folate status may be associated with an increased

risk or prevalence of certain forms of cancer; and that

optimal folate intake may provide a protective effect.

The initiation of carcinogenic activity is thought to

occur due to altered regulation of genetic expression of both

endogenous and exogenous oncogenes. Although the role of

folate deficiency in carcinogenesis has not been completely

elucidated, the co-carcinogenic effect of a deficiency of this

nutrient is thought to be related to folate's role in the

maintenance of chromosome structure and function. It has been


well established that folate is important in the regulation of

methyl groups used for DNA methylation. A deficiency of

methyl groups due to folate deficiency results in

undermethylated DNA and substitution of uridylate for

thymidylate. Subsequently, this interferes with histone

binding and results in the transcription of genetic sequences

that would ordinarily be suppressed. Additionally, improper

histone binding results in increased exposure of DNA to attack

by endogenous nucleases, thereby raising the risk of

chromosome breaks and incorporation of viral genomes.

Chromosome breaks associated with folate deficiency occur at

specific heritable fragile sites and many folate-sensitive

breaks occur at positions known to be associated with

translocations seen in cancer. It is thought that these

translocations disturb regulatory patterns of contiguous

segments of genetic information (Butterworth, 1991; Eto and

Krumdieck, 1986).

Compromised folate status has been associated with

increased risk of dysplasia or cancer of the cervix, colon,

bronchus and esophagus. The research documenting these

associations is briefly reviewed in the following paragraphs.

The concept of "localized folate deficiency" was

introduced as a result of a study conducted by Whitehead et

al. (1973). These researchers reported the occurrence of

megaloblastic features in cervical epithelial cells from women

taking oral contraceptive agents. These cytologic


abnormalities were not related to hematologic changes or low

serum folate/vitamin B12 concentrations; however, after three

weeks of folic acid supplementation (10 mg/d), the abnormal

cytologic findings were reversed or improved. Later,

Butterworth et al. (1982) reported that daily oral folic acid

supplementation (10 mg/d) was associated with improvement in

the cytologic manifestations of dysplasia compared with

placebo-treated controls. The supplemented group also had

less severe biopsy readings after three months of

supplementation. On the basis of these findings, it was

suggested that folate deficiency either plays an integral role

in the dysplastic process or is occasionally misdiagnosed as

cervical dysplasia (Butterworth et al., 1982).

In a more recent study, Butterworth et al. (1992b)

examined the effect of high-dose oral supplements of folic

acid (10 mg/d for six months) on the course of cervical

dysplasia. Supplementation appeared to have no significant

effect on the course of established cases of dysplasia, with

a high rate of apparent regression occurring in both the

placebo-treated and the folic acid-supplemented groups. These

researchers attributed the difference between the findings of

this study and their earlier report (Butterworth et al., 1982)

to the use of a more adequate sample size, exclusion of

patients with atypia less than dysplasia and a longer period

of observation. Although they concluded that folic acid

supplementation does not alter the course of established


disease, they did not exclude the possibility that folate

deficiency played a role in carcinogenesis because they found

a higher prevalence of dysplasia associated with human

papillomavirus 16 (HPV-16) infection (an oncogenic strain

thought to cause cervical dysplasia) among women in the lower

two tertiles of red blood cell folate than in the highest

tertile. Based on this information, they concluded that

folate deficiency may act as a co-carcinogen during the

initiation of cervical dysplasia.

In a separate case-control study conducted by the same

research group (Butterworth et al., 1992a), infection with

HPV-16 was the strongest risk factor for cervical dysplasia,

and there was a statistically significant interaction between

low erythrocyte folate concentrations and the HPV-16 virus.

The conclusion reached by these investigators was that the

carcinogenic effect of HPV-16 infection is enhanced in women

with low concentrations of erythrocyte folate. Although the

mechanism of interaction between folate and HPV-16 was not

investigated in this study, the researchers suggested that a

folate deficiency might increase the possibility of

incorporating the viral genome into human DNA resulting in

transformation of the epithelial cells.

The potential role of folate deficiency in carcinogenesis

may not be limited to the cervix since premalignant lesions

occurring in other organs exhibit features similar to those

seen in cervical dysplasia (Heimburger et al., 1987). Studies


examining the relationship between folate status and bronchial

metaplasia suggest that folate deficiency may influence the

susceptibility of the bronchial mucosa to neoplastic

transformation. For example, Heimburger et al. (1987) found

that the serum and erythrocyte folate concentrations of men

who smoked an average of 25 cigarettes per day were

significantly lower than nonsmokers, with the lowest values

occurring in smokers with bronchial metaplasia. Serum and

erythrocyte folate concentrations were also significantly

lower in smokers with metaplasia compared to smokers without

metaplasia. In a subsequent double-blind intervention trial

(Heimburger et al., 1988), smokers with bronchial squamous

metaplasia were stratified according to smoking level and

randomly assigned to treatment with a placebo or 10 mg of

folic acid and 500 Ag of hydroxycobalamin for four months.

Reduction of atypia, as determined by direct cytological

comparison, was significantly greater in the supplemented

group. Unfortunately, it is impossible to determine if folic

acid alone was responsible for the favorable effect since

hydroxycobalamin was administered simultaneously. Never-

theless, these studies are provocative and provide fertile

ground for continued investigation in this area.

Compared with the general population, patients with

chronic ulcerative colitis are at greater risk for developing

cancer of the colon. It has been hypothesized that folic acid

supplementation may protect against the development of


dysplasia (a premalignant pathologic finding) and cancer in

ulcerative colitis. To test this hypothesis, Lashner et al.

(1989) conducted a case control study to examine the effect of

folic acid supplementation on the incidence of dysplasia or

cancer in 99 subjects with chronic ulcerative colitis. A 62%

lower incidence of neoplasia was associated with folic acid

supplementation compared to subjects not receiving supple-

mentation. Although this outcome was not statistically

significant, it did not change when adjustments were made for

known confounders, suggesting that this finding was not due

entirely to bias or confounding and that inadequate sample

size may have contributed to the lack of significance (Lashner

et al., 1989).

Folate deficiency may also play a role in the development

of esophageal cancer. Jaskiewicz et al. (1988) have examined

cytological specimens obtained by brush biopsy of the

esophagus from subjects living in an area of Africa where

squamous cell carcinoma of the esophagus is the most common

form of cancer. Biopsy results revealed morphologic features

similar to those seen with folate deficiency and dysplasia.

Assessment of folate status in this same population revealed

significantly lower erythrocyte folate concentrations in

subjects with dysplastic and cancer cells compared with

controls (van Helden et al., 1987). Additionally, the

erythrocyte folate concentrations of subjects living in high-

incidence districts was significantly lower than those in the


intermediate and low incidence districts. However, these

results must be interpreted cautiously since multiple nutrient

deficiencies, particularly in the high-incidence area, were


Altered immune function. Although the effect of

compromised folate status on immunocompetence has not been

widely researched, studies using folate-deficient animals and

observations of patients with megaloblastic anemia due to

folate deficiency suggest that alterations occur in both

humoral and cell-mediated immunity. This is not surprising,

since folate is essential for the synthesis of DNA, and a

deficiency of this nutrient would hinder the ability of the

sensitized cells to proliferate rapidly. Examples of the

effects of compromised folate status on immune function as

determined from human studies include delayed cutaneous

hypersensitivity and depressed peripheral lymphocyte response

to phytohemagglutinin. Similar changes in immune function

have been shown using animal models. In addition to these

changes, reductions in white blood cells, leukocytes,

granulocytes, antibody forming cells and T-cells have been

reported in various animal models subjected to an isolated

folate deficiency (Nauss and Newberne, 1981).

Methods Used to Assess Folate Status and Metabolism in Humans

Microbiological assay. The microbiological assay is

considered the most useful method for quantifying folate

levels in biological samples (Tamura, 1990). This method uses

the growth response of folate-sensitive microorganisms to

determine folate content. The assay is performed by comparing

the growth of a test organism added to folate-free media

containing an aliquot of the sample, to the growth of the same

organism added to folate-free media that has been enriched

with known concentrations of folic acid. Growth is measured

as turbidity following incubation under controlled conditions.

The microorganism used most often for this assay is

Lactobacillus case, although Streptococcus faecium or

Pediococcus cerevisiae may also be used.

The three species of bacteria that may be used for this

assay do not respond equally to the various forms of folate.

For example, S. faecium does not respond to N5-methyl-

tetrahydrofolate, nor to folate derivatives with more than two

molecules of glutamate. Pediococcus cerevisiae is even more

selective. It does not respond to N5-methyl-tetrahydrofolate,

and of the remaining folate derivatives, it can only use those

occurring as reduced monoglutamyl tetrahydrofolates (Tamura,

1990). Since neither S. faecium nor P. cerevisiae respond to

N5-methyl-tetrahydrofolate, they are not suitable test

organisms for determination of serum or erythrocyte folate

concentration. Lactobacillus case, however, responds to N5-

methyl-tetrahydrofolate, as well as other folate derivatives,

which makes it the best choice when analyzing biological

samples for total folate.


The fact that L. case, S. faecium, and P. cerevisiae

respond differently to folates with varying chain lengths,

one-carbon substituent groups and oxidation states offers

researchers the opportunity to estimate the amount of folate

derivatives present in a particular sample. This can be

useful when the quantity of sample available for analysis is

insufficient to perform column chromatography. For example,

if it is important to estimate the amount of folate

derivatives other than N5-methyl-tetrahydrofolate, L. case

can be used in combination with S. faecium. Similarly, P.

cerevisiae can be used in conjunction with L. case to

estimate folate derivatives other than Ns-methyl-

tetrahydrofolate and oxidized pteroylglutamic acid. This

deductive approach can also be used to determine whether

samples contain folate polyglutamates. If the activity of S.

faecium or P. cerevisiae increases after folate hydrolase

treatment, the sample contains folate polyglutamates with more

than two molecules of glutamate. Likewise, if L. casei

activity increases after enzyme treatment, the sample contains

polyglutamates with more than three molecules of glutamate

(Tamura, 1990).

Biological samples typically analyzed for folate content

include serum, whole blood, urine and tissues. Cellular

folates, such as those found in tissues, are usually in the

polyglutamate form and require cleavage to the monoglutamate

form before folate concentration using the microbiological


assay can be determined. Partially purified folate hydrolase

preparations from chicken pancreas, hog kidney and rat or

human plasma have been used for this purpose. Folate in the

erythrocytes is also in the polyglutamate form; however, the

presence of natural folate hydrolase in the blood precludes

the need for treating hemolyzed whole blood samples with an

exogenous folate hydrolase. Urine and serum samples do not

require prior treatment with folate hydrolase because the

folate present in these samples is in the monoglutamyl form.

Although the basic concept of the microbiological assay

has not changed over the years, several improvements have been

made. Examples of these improvements have been reviewed by

Tamura (1990) and include the addition of antioxidants (e.g.

ascorbic acid) to samples to protect labile reduced forms of

folate from oxidation and the use of a cryoprotected organism

to maintain constant growth-response curves. Recently, this

assay has been adapted to take advantage of the availability

of the 96-well microtiter plate reader (Horne and Patterson,

1988; Newman and Tsai, 1986; O'Broin and Kelleher, 1992).

Although the procedure followed with this adaptation is

similar in concept to the standard procedure, the microtiter

plate reader improves the efficiency of absorbance readings.

Rapid calculation of the results can be achieved by

interfacing the microtiter plate reader with a computer (Horne

and Patterson, 1988). These modifications can save a


tremendous amount of time and money without sacrificing

accuracy (Newman and Tsai, 1986; O'Broin and Kelleher, 1992).

Radiometric binding assay. The radiometric binding assay

is a competitive protein binding assay in which radio-labeled

folate and unbound folate in the sample compete for the

binding sites on folate binding proteins. This method is

usually performed by clinical laboratories using commercially

available radioassay kits (Brody, 1991; Tamura, 1990).

Compared with the microbiological assay, this competitive

binding assay is simpler to perform and is not affected by

bacterial contamination or the presence of antibiotics.

Despite these potential advantages, the usefulness of the

radiometric binding assay is limited because the proteins used

to bind folates do not have an equal affinity for all forms of

folate. Binding affinity seems to be influenced by the state

of oxidation, the one-carbon substituent group and glutamyl

chain-length. For this reason, the suitability of the

radioassay method for quantifying complex mixtures of

naturally occurring folates has been questioned. This method

may be acceptable for determining serum folate concentration,

however, because N5-methyl-tetrahydrofolate is the predominant

form of folate in serum, and this form of folate is tightly

bound by folate binding proteins. Conversely, if the

population being assessed is receiving large doses of folates

other than Ns-methyl-tetrahydrofolate, the radioassay may be

inappropriate (Brody, 1991; Tamura, 1990).

Folate bioavailability studies. Until recently, most of

the information known about the metabolic requirements and

bioavailability of folates in humans was derived from animal

bioassay studies, or from a limited number of human studies

which used plasma or urinary folate concentration in folate-

saturated human subjects as the response criteria. While

these studies have made an important contribution to our

understanding of folate metabolism, they have several

limitations which restrict their scope and applicability. A

potential limitation of animal bioassay studies is the use of

animal models that are not entirely appropriate for studying

human metabolism. For example, rats are not appropriate

models to use for folate polyglutamate absorption studies

because they exhibit little or no brush border folate

conjugase activity (Wang et al., 1985). Limitations of

bioassay studies conducted in human subjects include their

inability to provide information about in vivo metabolism and

the need to consume large quantities of tested foods or

purified folates, the later of which could change the rate of

digestion and absorption in a way that is different from that

which usually occurs (Gregory and Toth, 1990).

The development of new tools and techniques combined with

previous accomplishments has facilitated research in the area

of folate requirements and bioavailability. The synthesis of

various forms of folate, including folate polyglutamates

(Krumdieck and Baugh, 1969; Godwin et al., 1972), and the


ability to radio-label folate derivatives with tritium at

different locations in the molecule (Godwin et al., 1972),

provided valuable tools for initial studies of folate

metabolism in animals. Although radio-labeled folates have

been used in human studies, concern about their safety

precluded their use for long-term supplementation studies and

restricted the amount of labeled folate that could be

administered to no more than a tracer dose. The development

of stable-isotopically (deuterium) labeled folates (Gregory,

1990; Gregory and Toth, 1988a; 1988b) has circumvented these

problems, while providing the opportunity to trace the fate of

folate during normal metabolism and to compare this with the

fate of folate under altered conditions or physiological

states. Additionally, the availability of stable isotopes of

folate affords the luxury of evaluating several forms of

folate simultaneously and allows the researcher to determine

the kinetics of absorption and turnover of the administered

compound (Gregory and Toth, 1990). Other developments that

have facilitated research in this area include the development

of improved methods for separating various folate

monoglutamates using high-performance liquid chromatography

(HPLC) (Kashani and Cooper, 1985; Wilson and Home, 1984); the

development of affinity chromatography columns that allow for

easier separation of folates from the sample (Selhub et al.,

1980), which ultimately results in cleaner chromatographic

separation using HPLC (Selhub, 1989); and the development of


a mass spectrometric method to quantify the ratio of unlabeled

versus labeled folates (Toth and Gregory, 1988).

The first published studies of an in vivo application of

stable-isotopically labeled folates in human subjects were

performed by Gregory and coworkers (Gregory and Toth, 1988a;

Gregory et al., 1990). The initial report (Gregory and Toth,

1988a) described the results of a preliminary investigation of

gas chromatography/mass spectral (GCMS) analysis of urinary

folates. The second study (Gregory et al., 1990) evaluated

the adequacy of a saturation regimen of 2 mg/d of folic acid

and the effectiveness of simultaneous administration of two

forms of deuterium-labeled folate. Subjects participating in

this study received 2 mg/d unlabeled folic acid for one week

to enhance urinary excretion of absorbed folates. On the

morning prior to administration of the deuterium-labeled

folates, the subjects collected a 24 hour urine sample. The

next morning, after an overnight fast, subjects consumed apple

juice containing two different deuterium-labeled folates. The

subjects collected their urine for 48 hours following the

treatment. A constant diet was consumed beginning at the time

of the pre-dose urine collection through the end of the post-

dose urine collection. Dietary and total urinary folate

concentration was determined by microbiological assay using L.

casei. Urine samples were also prepared for mass spectral

analysis of labeled folates, after which the molar ratio of

each species of folate in the sample was calculated. The


results of these studies confirmed that stable-isotopically

labeled folates are well-suited for in vivo studies and that

the protocol outlined above is suitable for studying many

aspects of folate bioavailability and in vivo kinetics

(Gregory et al., 1990). Recently, Von der Porten et al.

(1992) used this methodology to study in vivo folate kinetics

in human subjects supplemented with deuterium-labeled folic

acid for four weeks. By using deuterium-labeled folate, these

researchers were able to determine rate constants and isotope

enrichment, which subsequently enabled them to assess tissue

uptake and equilibration, in vivo turnover rates and body pool

sizes of folate. It is expected that further application of

this technique will greatly enhance our knowledge of folate

metabolism and improve our ability to make appropriate

nutritional recommendations.



Zinc has an atomic weight of 65.37. It is a first series

transition element with the electronic configuration [Argon]

4s2 3d10. This configuration confers properties to zinc that

distinguish it from other transition metals. For example,

zinc is diamagnetic rather than paramagnetic, and although it

can exist in several valence states, it is resistant to

oxidation and is found almost universally as the divalent ion

(Zn2+). Another property of zinc that is relevant to its


specific biological functions is its ability to form stable

complexes with side chains of proteins. The high charge

density of Zn2+ allows this metal to function as a Lewis acid

to withdraw electrons from electron-rich functional groups of

ligands resulting in the formation of noncovalently bound

coordination complexes. These features make zinc ideally

suited for its involvement in enzyme function and structure

(Solomons, 1988; Williams, 1989).


The process of zinc absorption includes the uptake of

zinc by the intestinal mucosal cells, movement of zinc through

the mucosal cells and transfer into the portal circulation.

Paracellular movement of zinc to the portal circulation may

also occur. These aspects of zinc absorption are discussed in

the following paragraphs.

Although the study of zinc absorption has been given much

attention, the precise mechanism, location and control of this

process have not been fully delineated (Cousins and Hempe,

1990). Differences in experimental conditions may account for

the conflicting results of in vivo and in vitro experiments

designed to study zinc absorption, since a multitude of host

and environmental factors have been shown to influence this

process (Solomons and Cousins, 1984).

Zinc is absorbed throughout the small intestine, but the

segment with the highest capacity to absorb zinc has not yet


been determined (Cousins and Hempe, 1990). Rat studies used

to identify the major absorptive site have produced variable

results, with some studies (Davies, 1980; Methfessel and

Spencer, 1973; Van Campen and Mitchell, 1965) suggesting that

the greatest amount of zinc uptake occurs in the duodenum, and

others (Antonson et al., 1979; Emes and Arthur, 1975)

suggesting that greater zinc uptake occurs in the distal

portions of the intestine. Although Sahagian et al. (1966)

found the concentration of zinc from normal rat intestine to

be fairly uniform throughout each region, zinc uptake by

strips of rat intestine was greater in the duodenal and ileal

segments compared to the jejunal segment. These findings are

in contrast to those of an in vivo intestinal perfusion study

(Matseshe et al., 1980) conducted in humans. In this study,

Matseshe et al. found that more zinc left the distal duodenum

than what was contained in a test meal, suggesting that

endogenous secretions contributed to the total intraluminal

zinc concentration. For this reason, net duodenal zinc

disappearance, if any, could not be determined. Once the

intraluminal contents passed into the jejunum, zinc

disappeared gradually but incompletely. Ileal zinc absorption

was not determined in this study. In fact, all portions of

the small intestine may be functionally important in terms of

zinc absorption, with the duodenum having first access to

dietary and endogenous zinc, and subsequent sections having

the benefit of the action of digestive processes that may

increase the accessibility of this nutrient (Cousins and

Hempe, 1990; Lbnnerdal, 1989a).

The first phase of the absorptive process involves the

movement of zinc from the intestinal lumen into the mucosal

cell. Evidence that zinc absorption occurs rapidly and

involves a saturable, carrier-mediated component, as well as

nonsaturable diffusion has come from several rat studies.

Davies (1980) exposed ligated duodenal loops from rats to

different concentrations of zinc and found that these segments

exhibited saturation kinetics at low luminal zinc

concentrations. At higher luminal zinc concentrations, zinc

uptake was linear suggesting that zinc absorption also

occurred by diffusion. Menard and Cousins (1983) reported

similar results using isolated brush border membrane vesicles

from rats. Further evidence for a mediated component of zinc

absorption was derived from experiments with rats previously

fed a zinc-adequate or a zinc-deficient diet (Steel and

Cousins, 1985). Zinc absorption, as determined from zinc

accumulation in the portal perfusate, revealed that absorption

was saturable in both groups, and that zinc-deficient rats had

a more rapid rate of zinc absorption at all luminal zinc

concentrations. Zinc absorption appeared to involve both

mediated and nonmediated components, with most of the

absorption occurring by means of the mediated component in the

zinc-deficient group. These results suggest that the

saturable process is stimulated by zinc depletion, whereas the


nonsaturable process is unaffected by zinc deficiency and

proceeds in proportion to the intraluminal zinc concentration.

A more recent study conducted by Hoadley et al. (1987) is in

agreement with the findings of previous studies suggesting

that zinc absorption involves two kinetic processes.

Although there is general agreement that zinc uptake by

the intestinal mucosal cell involves a saturable, carrier-

mediated component, as well as nonsaturable diffusion, it is

not known if these processes represent components of a single

transport step occurring at the brush border membrane, or if

movement from the lumen to the portal circulation occurs by

two independent routes: transcellular and paracellular.

(Cousins and Hempe, 1990). Transcellular movement may involve

a protein that has recently been identified (Hempe and

Cousins, 1991), purified and partially characterized (Khoo and

Cousins, 1993). This protein, known as cysteine-rich

intestinal protein is a low molecular weight, zinc-binding,

cytosolic protein isolated from intestinal mucosal cells.

Cysteine-rich intestinal protein has two possible zinc binding

sites, and it is hypothesized that this protein serves as an

intracellular zinc carrier. This protein appears to bind more

zinc when intestinal metallothionein is not induced,

suggesting that intestinal metallothionein may interact with

cysteine-rich intestinal protein to regulate zinc absorption

and transport (Khoo and Cousins, 1993).

Movement of zinc from the intestinal lumen into the

mucosal cell may be affected by the intraluminal environment.

Zinc liberated from food matrices during the digestive process

may remain as the free ion or may form coordination complexes

with endogenous (pancreatic, biliary or mucosal) or exogenous

(dietary) intraluminal ligands (Solomons and Cousins, 1984).

Studies investigating the effect of intraluminal binding

ligands on zinc uptake have suggested that ligand binding may

affect zinc absorption by altering intestinal membrane

permeability and/or zinc gradients across intestinal cell

membranes (Cousins, 1985; Solomons and Cousins, 1984).

However, while there is little question that zinc-binding

ligands affect zinc absorption, it is doubtful that this

process is a prerequisite for movement of zinc across the

membrane surface (Cousins and Hempe, 1990).

Once zinc is inside the intestinal mucosal cell, it binds

to a variety of high molecular weight ligands, as well as to

intestinal metallothionein, a low molecular weight, cyteine-

rich metalloprotein. The amount of zinc bound to intestinal

metallothionein varies according to zinc status and intake.

Zinc binding to intestinal metallothionein restricts the

movement of zinc from the cell, thereby contributing to the

regulation of zinc absorption. Intestinal metallothionein

synthesis is induced at the transcriptional level by high

dietary zinc (Blalock et al., 1988; Menard et al., 1981). When

the rate of intestinal metallothionein synthesis is high, net

zinc release to the portal circulation is curtailed (Menard et

al., 1981). In zinc deficiency, intestinal metallothionein

concentration is low and more zinc is released into the portal

circulation (Hoadley, et al., 1988). Thus, it appears that

the induction of metallothionein synthesis in intestinal

mucosal cells could provide for the efficient regulation of

zinc transfer to the vascular compartment.

Based on the results of rat studies (Davies, 1980; Smith

and Cousins, 1980), transfer of zinc from the mucosal cell to

the portal circulation occurs more slowly than uptake and

accumulation within the cell. According to Davies (1980),

transfer of zinc into the portal circulation may occur in two

stages: 1) rapid transfer occurring over the first 30 minutes;

and 2) slower transfer occurring from 30 minutes to six hours

after receiving the dose, which may represent the release of

intracellularly bound zinc from zinc-binding proteins.

Consequently, movement of zinc from the intestinal mucosal

cell into the vascular compartment may be the rate-limiting

step in zinc absorption (Smith and Cousins, 1980). At low

luminal zinc concentrations, much of the zinc available is

released into the portal circulation; however, as the luminal

zinc concentration increases, less of this nutrient is

transported to the vascular system (Hambidge et al., 1986;

Smith and Cousins, 1980). An exception to this phenomenon

occurs when excessive loads of luminal zinc are available. In


this case, the ability to control the release of zinc into the

circulation is diminished (Smith and Cousins, 1980).

Transport. Distribution and Metabolism

Although it had been proposed that transferring was

responsible for portal zinc transport (Evans and Winter,

1975), several lines of investigation have shown that albumin

is the principal portal transport protein for zinc (Smith et

al., 1979). Smith et al. (1979) demonstrated the importance

of albumin in portal transport by comparing the percent of

65Zn absorbed when the vascular perfusate composition was

varied. The absence of albumin from the perfusate resulted in

negligible zinc transfer to the portal circulation, whereas

the isosmotic replacement of all plasma proteins with albumin

resulted in a twofold increase in zinc transfer over that of

the control perfusate. These data intimate that the extent of

zinc absorption may be affected by the concentration of

albumin in the blood (Cousins, 1985). Consequently, zinc

absorption may be impaired in disease states associated with

the development of hypoalbuminemia.

Zinc entering the portal circulation is rapidly

transported to the liver. The liver is the primary organ

involved in zinc metabolism, and a large portion of the zinc

in portal blood is exchanged with the liver (Cousins and

Hempe, 1990). In vitro experiments with rat liver parenchymal

cells demonstrated that zinc uptake was temperature and


energy-dependent, followed saturation kinetics and occurred in

two phases (Failla and Cousins, 1978a). The first phase was

characterized by rapid, saturable uptake. This was followed

by a slower phase comprised of both saturable and linear

components. Maximal uptake occurred at the normal plasma zinc

concentration (Cousins, 1989; Failla and Cousins, 1978a).

Zinc accumulation by these cells increased when physiological

concentrations of certain adrenal corticosteroids were added

suggesting that these hormones may perform an essential role

in the regulation of hepatic zinc metabolism. As a corollary,

it is possible that zinc has a role in glucocorticoid-mediated

alterations of hepatic metabolic processes (Failla and

Cousins, 1978b).

Distribution of zinc to the extrahepatic tissues occurs

mainly via the plasma. The plasma contains approximately 10

to 20% of the zinc in whole blood, whereas the erythron and

leukocytes contain the major portion of zinc in the blood.

Most of the plasma zinc is bound to albumin, although alpha-2-

macroglobulin, transferring, histidine and cysteine may also

transport small amounts of this metal. Zinc bound to albumin

is considered to be loosely bound and represents the

metabolically active, exchangeable zinc pool in the blood.

This pool is responsive to acute and chronic changes related

to stress, infection and dietary zinc. In contrast to plasma

zinc, the zinc in erythrocytes is mostly associated with

carbonic anhydrase, although small amounts may be associated


with superoxide dismutase and metallothionein (Cousins, 1989;

Cousins and Hempe, 1990; DiSilvestro and Cousins, 1983).

The rate of zinc incorporation into extrahepatic tissues

varies, as does the rate of zinc turnover in these tissues.

Zinc accumulation and turnover occurs rapidly in the kidney,

pancreas and spleen (Hambidge, et al., 1986). The rate of

zinc uptake by skeletal muscle and the central nervous system

is relatively slow, and the zinc in these tissues remains

tightly bound for long periods of time (Hambidge et al.,

1986). The zinc incorporated into hair is not available for

exchange either. Tissue-specific redistribution of body zinc

can occur during periods of zinc deprivation, stress and

infection/inflammation (Cousins and Leinart, 1988; Dunn and

Cousins, 1989; Giugliano and Millward, 1984; Huber and

Cousins, 1988; Jackson et al., 1982). For example, the zinc

concentration of muscle in zinc-deficient rats is protected,

while the concentration in bone, liver and plasma declines

(Giugliano and Millward, 1984; Jackson et al., 1982). This

redistribution may occur as a result of tissue-specific

induction of metallothionein.

In general, the largest concentration of intracellular

zinc is found in the cytosol, although smaller amounts are

present in the nuclear, microsomal and mitochondrial fractions

of cells (Hambidge et al., 1986). Zinc is also present in the

membrane and may enhance membrane stability (Bettger and

O'Dell, 1981). Within the cytosol, zinc is primarily bound to


large molecular weight proteins. These proteins may be zinc

metalloenzymes. In contrast, the amount of zinc bound to

metallothionein is relatively low under normal dietary

conditions; however, when dietary zinc is increased, the

metallothionein gene is induced and metallothionein synthesis

is elevated (Cousins and Lee-Ambrose, 1992; Cousins and

Leinart, 1988; Huber and Cousins, 1988). Consequently,

increased zinc binding to metallothionein in the liver,

pancreas, kidney and muscle occurs. Zinc may regulate the

expression of the metallothionein gene through a specific

nuclear metal-binding protein that also binds to unique DNA

sequences in the promoter region (Cousins and Hempe, 1990).

Synthesis of metallothionein, particularly in the liver, may

also occur in response to hormones, as well as physiological

stimuli such as stress, acute infection and shock (Cousins and

Hempe, 1990; Hambidge et al., 1986).


The major route of zinc excretion is via the feces. Zinc

in the feces is derived from unabsorbed dietary zinc, as well

as endogenously secreted zinc. Sources of endogenous zinc

include pancreatic, biliary and mucosal secretions, as well as

zinc present in desquamated mucosal cells. Usually,

endogenously secreted zinc is efficiently reabsorbed, but

intraluminal factors such as the presence of phytic acid may

decrease the efficiency of absorption (Cousins and Hempe,

1990). Additionally, dietary intake of zinc may affect

reabsorption, with more zinc of endogenous origin being

excreted in the feces at higher levels of zinc intake (Jackson

et al., 1984).

Urinary zinc losses account for a fairly small amount of

the zinc excreted under normal physiologic conditions. The

usual range of zinc excreted in the urine of healthy adults is

about 300 to 600 gg/d (Gibson, 1990). Urinary zinc excretion

has been shown to decrease in human subjects consuming a zinc-

deficient experimental diet (0.28 mg zinc/d) (Baer and King,

1984). Injury, burns, infection, acute starvation and

pathologic conditions resulting in excessive muscle catabolism

have been associated with clinically significant increases in

urinary zinc excretion. Hyperzincuria is known to occur in

conjunction with proteinuria due to kidney dysfunction and in

patients with sickle cell disease and cirrhosis of the liver

(Cousins and Hempe, 1990; Gibson, 1990).

Dermal losses due to sweating and to the sloughing of

epithelial tissue also account for some of the zinc lost from

the body. As reported in a review by Hambidge et al. (1986),

the total amount of zinc lost in sweat from adults living in

a temperate North American climate has been estimated to be

0.4 to 2.8 mg/d. Dermal zinc losses have been found to be

influenced by dietary zinc intake, with reduced losses

occurring when zinc intake is marginal, and losses increasing