Content uploaded by David Thurnham
Author content
All content in this area was uploaded by David Thurnham on Dec 17, 2013
Content may be subject to copyright.
Proceedings of the Nutrition Society (1999), 58, 289–293 289
Abbreviations:
CIC, conjunctival-impression cytology; L, lactulose; M, mannitol; VAD, vitamin A deficiency.
*Corresponding author:
Fiona McCullough, fax +44 (0)1265 324965, email f.mccullough@ulst.ac.uk
CAB International
The effect of vitamin A on epithelial integrity
F. S. W. McCullough*, C. A. Northrop-Clewes and D. I. Thurnham
Fiona M cCullough,
Northern Ireland Centre for Diet and Health (NICHE), University of Ulster, Coleraine BT52 1SA, UK
fax +44 ( 0)1265 324965, email f.mccu llough@ulst.ac .uk
Vitamin A is the generic term for a variety of fat-soluble substances including retinol,
retinyl palmitate and the provitamin A carotenoids such as all-
trans
-
β
-carotene. Vitamin
A is commonly known as the anti-infective vitamin and has an essential role in vision
and cellular differentiation, the latter providing a unique core mechanism helping to
explain the influence of vitamin A on epithelial barriers. Alterations in the epithelial
lining of vital organs occur early in deficiency, suggesting a potentially important role
for the barrier function. Vitamin A deficiency (VAD) is most commonly recognized in
the eye. The conjunctival-impression cytology test detects the presence of larger
irregular keratinized cells and the absence of mucous-secreting goblet cells, indicative
of VAD. The method is simple, quick and sensitive in populations where VAD is
present. In the respiratory tract, observational studies all show an association with VAD,
although vitamin A supplementation studies appear to have little effect on respiratory
disease. Organ-specific targeting may improve success rates. The dual-sugar intestinal-
permeability test allows the effect of vitamin A supplementation to be monitored on the
gastrointestinal tract. Two vitamin A supplementation studies were carried out recently
in Orissa State, India. Healthy infants of weaning age were administered orally eight
weekly doses of 5·0mg retinol equivalents and hospitalized infants received one large
oral dose 60 mg retinol equivalents in the form of retinyl palmitate. Improvements in gut
integrity and haematological status were observed in both studies. In summary, the
response of the eye to vitamin A supplementation is well established; the present review
highlights some of the more recent observations examining the effects of vitamin A.
Vitamin A: Conjunctival-impression cytology test: Dual-sugar intestinal-permeability test
‘
Vitamin A builds fences that keep germs out.’ E.V.
McCollum 1935
‘Vitamin A’ is a generic term for a variety of related
compounds, including retinol, retinyl palmitate, retinoic
acid and the pro-vitamin A carotenoids. The fat-soluble
substance is usually found in animal and dairy products as
the ester, retinyl palmitate. The alcohol retinol is formed
during hydrolysis in the small intestine where it is taken
up actively by the enterocytes. The most active provitamin
A carotenoid is all-
trans
-
β
-carotene, which is present in
yellow–orange fruits and dark-green leafy vegetables.
Intestinal cells convert provitamin A carotenoids to retinal-
dehyde, the majority of which is reduced to retinol.
Bioavailability of provitamin A carotenes varies widely,
depending on how the food is prepared and consumed.
Vitamin A is commonly known as the anti-infective
vitamin. At the beginning of the twentieth century, animals
fed on ‘pure’ diets containing only lard and carbohydrates
with no vitamins ceased to grow, lost weight and became
susceptible to infections. Administration of ‘accessory
factors’ found in dairy products and cod-liver oil reversed
the process. McCollum (McCollum & Davis, 1915) termed
the critical factor ‘Fat-soluble A’. By 1920 the clinical
manifestation of vitamin A deficiency (VAD), and its cause
and cure were well established.
Vitamin A has an essential role in vision and cellular
differentiation, the latter being particularly important in
growth, reproduction and immune response, and in main-
taining epithelial integrity. The critical function vitamin A
plays in regulating cellular differentiation provides a unique
‘core’ mechanism that would at least partly explain its influ-
ence on epithelial barriers, immune competence, healing,
resistance and recovery (Ross, 1992). No nutritional
deficiency is more synergistic with infection than vitamin A.
The two main mechanisms involved in the prevention of
disease are the effect of vitamin A on the immune system
AMA PNS0071.fm Page 289 Thursday, June 24, 1999 3:13 PM
290 F. S. W. McCullough et al.
and the effect on epithelial integrity, the subject of
the present review. All epithelial surfaces, including the
eye, skin, trachea, salivary gland, vaginal epithelium
and gastrointestinal tract, are involved. Alterations in the
epithelial lining of vital organs occur early in VAD, suggest-
ing a potentially important role for the ‘barrier function’.
Tissue integrity is vitally important for protection against
environmental pathogenic organisms.
Epithelial tissues
Eye
VAD is most commonly recognized in the eye. Night-
blindness and its successful treatment with animal liver was
known to the ancient Egyptians at least 3500 years
ago (Wolf, 1978), and was mentioned by Hippocrates.
Xerophthalmia (from the Greek: xero, dry; ophthalmia,
inflamed eye) is a constellation of ocular manifestations
long associated with VAD, representing the ‘classical’
presentation of this particular form of malnutrition.
Deficiency symptoms are progressive from night-blindness,
conjunctival xerosis, Bitot’s spots and corneal xerosis
to corneal ulceration (keratomalacia), corneal scar and
xerophthalmia fundus (World Health Organization, 1995).
Vitamin A status can be assessed by functional tests that
measure epithelial integrity. Keratinizing metaplasia is a
potentially practical index, and as early as 1925 the scrap-
ings of cornea, nose, mouth and vagina were recommended
as a diagnostic tool of VAD (Wolbach & Howe, 1925). The
early studies were neither qualitative nor validated against
any other index of vitamin A status. More detailed know-
ledge about the normal and deficient nature of conjunctival
epithelium, particularly the geographic distribution of goblet
cells (Kessing, 1968) and their response to vitamin A defi-
ciency (Sommer
etal
. 1982), created renewed interest in
the approach. The system of ‘simple conjunctival biopsy’
was described by Egbert
etal
. (1971), and Nelson
etal
.
(1988). Cellulose acetate filter paper was briefly applied
to the conjunctival surface to remove the superficial
layers of the conjunctival epithelium. The biopsy technique,
renamed conjunctival-impression cytology (CIC) provides
evidence of keratinizing metaplasia. Vitamin A-
unresponsive and -responsive Bitot’s spots have an identical
biological appearance; however, the CIC test detects the
absence of mucous-secreting goblet cells and the presence
of larger irregular keratinized cells. Initial trials demon-
strated markedly abnormal CIC from children with mild
xerophthalmia when compared with normal children or
following vitamin A treatment (Wittpenn
etal
. 1986). There
was a close correlation between the prevalence of abnormal
CIC and the severity of deficiency suggested by three other
indicators of vitamin A status (serum retinol, relative dose
response and modified relative dose response). Of the chil-
dren with clinically-responsive xerophthalmia and low
vitamin A levels 93% had abnormal cytology, and 6% of
children with normal eyes and serum retinol levels between
0·09 and 0·10
µ
mol/l (250–290
µ
g/l) had abnormal CIC
(Natadiastra
etal
. 1988). The validity of CIC has been
further confirmed by the following: comparison of liver
retinol concentrations and relative dose response in French
children with normal eyes but with hepatic disease
(Amedee-Manesme
etal
. 1988); comparisons of abnormal
CIC rates in xerophthalmic and non-xerophthalmic children
in India (Reddy
etal
. 1989); correlation of CIC with serum
and breast-milk retinol in Indonesian women (Miller
etal
.
1993) and by correlations of CIC with other indices of
vitamin A status (Tanumihardjo & Olsen, 1991).
CIC is of particular advantage under field conditions as it
does not require sophisticated laboratory analysis, storage or
transport. However, there are problems of compliance with
small children, who can be frightened of the paper being
set on their eye. CIC is not without its limitations and
several reports (for example, see Gadomski
etal
. 1989) have
questioned its validity and/or interpretation. In theory,
although not always in practice, and ideal marker would
detect subclinical deficiency, when stores are below ‘safe’
levels. By definition, a functional abnormality is evidence
that physiologically-significant deficiency already exists.
Hence, the CIC marker has been more successfully used
in epidemiological studies than as a diagnostic tool for
individuals, especially in areas of marginal deficiency
(Natadiastra
etal
. 1988). When using CIC Coutsoudis &
Way (1996) noted that allergic conjunctivitis could be
mistaken for VAD in South African children.
Respiratory tract
Observational studies (see Vijayaraghavan
etal
. 1990)
all show an association between VAD and respiratory
infection. However, a meta-analysis (Bates, 1995) showed a
lack of impact of vitamin A supplementation, although there
were some indications of reduced severity and possibly
reduced duration of respiratory symptoms. Interestingly,
some researchers suggest that increased cough following
supplementation may indicate greater mucus production
and removal of infective material (Herrera
etal
. 1996). A
promising method of treating respiratory infections using
organ-specific targeting was suggested by Biesalski (1996).
In rats direct inhalation of 0·9mg retinol equivalents as
retinyl esters, produced an increase in plasma retinyl ester
from 0·16
µ
mol/l (45
µ
g/l) to 0·66
µ
mol/l (190
µ
g/l) with no
side-effects.
It was suggested by Shenai
etal
. (1987) that vitamin A
supplementation in preterm infants was of benefit in
bronchopulmonary dysplasia. It was reported by Coutsoudis
etal
. (1993) that initial VAD in preterm infants does
not predispose them to respiratory distress syndrome or
pneumonia, but by 3 months post-partum vitamin A status
was adequate and similar to that of full-term infants.
Gastrointestinal tract
Until recently, mucosal damage could only be assessed by
endoscopy and/or biopsy of the small intestine. Using such
techniques a number of studies have shown a substantial
reduction in the number of goblet cells per duodenal villus
when compared with control and pair-fed mice (Ahmed
etal
. 1990), a reduction in luminal mucus (Meneghetti
etal
.
1989), and decreased cellular division preceding histo-
logical abnormalities in mildly-VAD animals (Zile
etal
.
1981).
AMA PNS0071.fm Page 290 Thursday, June 24, 1999 3:13 PM
Postgraduate symposium 291
Non-invasive techniques (Lunn
etal
. 1989; Northrop
etal
. 1990) have been introduced which estimate specific
aspects of mucosal function and integrity and allow repeated
measurements to be carried out, even under field conditions.
Absorption across healthy intestinal mucosa is highly selec-
tive. ‘Inert molecules’ pass across the absorptive surfaces
at rates determined by concentration gradient, surface area
and time, as well as the permeability of the intestinal
mucosa. The mucosa demonstrates marked discrimination
with respect to the molecular dimension and solubility of
inert molecules, this discrimination becoming impaired
with mucosal damage (Menzies
etal
. 1983). The xylose-
absorption test has been used as a preliminary screening
procedure for gastrointestinal integrity (Hubble &
Littlejohn, 1963), but has limitations. Gastric emptying,
intestinal transit, blood flow and renal function variables
reduce the correlation between xylose absorption or
excretion and the degree of mucosal damage (Lamabadu-
suriya
etal
. 1975). Dual-sugar absorption tests are not
affected to the same extent by such limitations. A variety
of markers have been utilized to indicate gastrointestinal
integrity, e.g. ingestion of lactulose and rhamnose showed
that during kwashiorkor, abnormal intestinal permeability
correlated with disease severity and improved only slowly
with nutritional rehabilitation (Brewster
etal
. 1997). Elia
etal
. (1987) investigated the factors affecting the intestinal
uptake and urinary excretion of mannitol (M), lactulose (L)
and Cr-EDTA in normal subjects and three patients
with ileostomy. They suggested that the uptake of L is
slower than that of M, and Cr-EDTA is readily absorbed
in the colon. L and M are two non-metabolizable sugars
that cross the mucosa paracellularly and transcellularly
respectively. The ratio L:M in blood is an indicator of
mucosal damage, L:M greater than 0·02 being considered
abnormal. L:M values discriminated well between subjects
with normal biopsy and those with villous atrophy. Fleming
etal
. (1996) successfully measured L and M by HPLC in
serum 90 min after intake, while other workers have used
automated enzymic assays (Lunn
etal
. 1989; Northrop
etal
.
1990). The HPLC results agree well with L:M values
determined enzymically following a 5h urine collection
(
r
0·88,
P
=0·05), hence the two methods can be used inter-
changeably.
Northrop-Clewes
etal
. (1997) investigated the role of
vitamin A on epithelial integrity using the L:M dual-sugar
intestinal-permeability test. This test uses a weight-
dependent dose of the two sugars (L and M) which are given
to the infant dissolved in water. A urine bag is used to
collect the urine produced during the following 5h. The
sugars are measured enzymically and the L : M value is an
indicator of epithelial integrity. A healthy mucosal surface
has villi which are long finger-like projections allowing
maximal absorption of M, while L, the larger molecule,
can only pass between the cells. Absorption of L is therefore
minimal, resulting in a low L:M value. Damage to villi
could include either a reduced surface area, e.g. by a flatten-
ing of the surface area, and/or damage caused by parasitic
infestation (e.g. hookworms) causing leakiness of the
gastrointestinal tract. In both situations the absorption of L
increases relative to that of M, resulting in an elevated L:M
value.
Faltering growth, infections, diarrhoea and anaemia
remain major problems in the developing world. Retinol and
Fe are transported in the bloodstream by negatively-charged
acute-phase proteins which decrease during infection, hence
retinol and Fe are ‘trapped’ in the liver and spleen during
disease. Foods are frequently contaminated with pathogens,
thus exposure to infections is increased in developing
countries, and the risk of infection may be exacerbated by
poor gastrointestinal integrity. Impaired gastrointestinal
integrity may facilitate the translocation of bacteria across
the mucosa. Studies carried out in rural Gambian infants
showed evidence of impaired intestinal integrity and
elevated acute-phase response, which accounted for up to
50% of the growth faltering noted during the first 2 years
of life (Northrop-Clewes
etal
. 1997). Seasonal data from
the same Gambian infants suggested that during the
mango (
Mangifera indica
) season there were improvements
in gastrointestinal integrity and growth, and the acute-phase
reponse normalized for a period of approximately 3 months
(Northrop-Clewes
etal
. 1997). Mangoes are a rich source of
β
-carotene and the main supply of vitamin A for the year in
this community. During infection, an acute-phase reaction is
initiated which results in reduced levels of plasma retinol.
Thus, increasing the availability of vitamin A throughout the
year may help infants to restore gastrointestinal integrity,
fight off infection and reduce the problem of faltering
growth. It is hypothesized that vitamin A acts by restoring
the physical and biochemical integrity of epithelial surfaces,
particularly the gastrointestinal tract, so that the risk of
infection and trauma will be reduced.
Effect of vitamin A supplementation on gastrointestinal
integrity.
Two vitamin A supplementation studies were
carried out in Rourkela, Orissa State, India. Low serum
retinol levels and growth faltering had previously been
reported in this region (Das etal. 1996). The first study was
carried out in four rural villages, where eighty infants
aged 4–18 months were recruited (Northrop-Clewes etal.
1997). The infants were randomly allocated to receive
either 5·0mg retinol equivalents as retinyl palmitate and
40mg
α
-tocopherol in 4ml groundnut oil weekly for
8 weeks or 40mg
α
-tocopherol in 4ml groundnut oil as
placebo. Baseline L:M values were 0·74 and 0·64 for the
treatment and placebo group respectively. L:M values were
reduced after 4 and 8 weeks in both groups, but only the
change in the treatment group was significant (P<0·05).
Mean haemoglobin concentration also showed slight but
non-significant improvement. However, the placebo group
displayed a significant (
P
<0·05) decrease in haemoglobin
from 91·95 to 88·33g/l after 8 weeks.
The second study investigated the effect of vitamin A
supplementation on hospitalized infants with diarrhoea and
respiratory infections (FSW McCullough, CA Northrop-
Clewes, BD Das and DI Thurnham, unpublished results).
Infants of weaning age (4–18 months;
n
100) admitted with
diarrhoeal disease or respiratory infection were randomly
assigned to one of three treatment groups as follows: treat-
ment 1, 60mg retinol equivalents as retinyl palmitate (30mg
retinol equivalents for infants under 12 months) and
40mg
α
-tocopherol on the day following admission;
treatment 2, 60mg retinol equivalents as retinyl palmitate
and 40mg
α
-tocopherol 5d post-admission; treatment 3,
AMA PNS0071.fm Page 291 Thursday, June 24, 1999 3:13 PM
292 F. S. W. McCullough et al.
placebo (no vitamin A, only 40mg
α
-tocopherol 5d post-
admission). Gastrointestinal integrity and haemoglobin
measurements were carried out at baseline, and 5, 10
and 30d post-admission. Gastrointestinal integrity was
measured using the dual-sugar intestinal-permeability test.
Haemoglobin was measured using the Hemocue device
(Hemocue, Angelholm, Sweden). Differences in L:M
and haemoglobin between baseline and day 5 were not
significant for any treatment. By day 10, however, both
L:M values and haemoglobin levels were significantly
(
P
<0·05) improved in the groups receiving the vitamin A,
whereas there was no improvement in the placebo group. By
day 30, the responses in the groups receiving vitamin A
were still better than those of the placebo group, although
not as significant (
P
>0·05).
Traditionally, micronutrient supplementation is pre-
scribed on admission of patients to Ispat General Hospital
(Rourkela City, Orissa State, India). The hospital policy
should be re-examined and more vitamin A supplied, since
the results of these studies suggest that gastrointestinal
integrity is severely impaired during illness, but responsive
to the additional vitamin A. Gastrointestinal integrity and
haemoglobin were significantly (
P
<0·05) more rapidly
improved in the vitamin A-treated groups than in the
placebo group. However, there were no statistical differ-
ences between the two vitamin A groups, therefore there
was no apparent advantage in giving vitamin A on admis-
sion. In addition, the majority of the children were
discharged within 5d.
In conclusion, marginal vitamin A status clearly impairs
the integrity of epithelial tissues. These effects are more
clearly seen in the eye as a result of the availability of the
CIC test. Use of the dual-sugar intestinal-permeability test
now opens up the possibility of examining more closely the
influence of vitamin A on the gastrointestinal tract.
Acknowledgements
The authors are grateful to Hoffmann-la-Roche Inc., Basel,
Switzerland who supplied the vitamin A and assisted
with fieldwork costs in India and Dr B. Dash, Director of
Health for permission to base the study at Ispat General
Hospital.
References
Ahmed F, Jones DB & Jackson AA (1990) The interaction of
vitamin A deficiency and rotavirus in the mouse. British Journal
of Nutrition 63, 363–373.
Alvarez JO, Salazar-Lindo E, Kohatsu J, Miranda P & Stephenson
CB (1995) Urinary excretion of retinol in children with acute
diarrhoea. American Journal of Clinical Nutrition 61, 1273–
1276.
Amedee-Manesme O, Luzeau R, Wittpenn JR, Hanck A &
Sommer A (1988) Impression cytology test detects subclinical
vitamin A deficiency. American Journal of Clinical Nutrition 50,
814–817.
Bates CJ (1995) Vitamin A. Lancet 345, 31–35.
Biesalski HK (1996) Efficient vitamin A supply through inhalation.
Report of the XVII IVACG Meeting, Guatamala City, Guata-
mala, p. 100. Washington, DC: IVACG Secretariat.
Brewster DR, Manary MJ, Menzies IS, O’Loughlin EV & Henry
RL (1997) Intestinal permeability in kwashiorkor. Archives of
Disease in Childhood 76, 236–241.
Coutsoudis A, Adhikari M & Coovadia HM (1993) Serum vitamin
A concentrations and association with respiratory disease in
premature infants. Journal of Tropical Pediatrics 41, 230–233.
Coutsoudis A & Way R (1996) An evaluation of the association
between external eye diseases and vitamin A status in black
South African children. South African Medical Journal 86, 871–
879.
Das BS, Thurnham DI & Das DB (1996) Plasma tocopherol, retinol
and carotenoids in children with falciparium malaria. American
Journal of Clinical Nutrition 64, 94–100.
Egbert PR, Lauber S & Maurice DM (1977) A simple conjunctival
biopsy. American Journal of Opthalmology 84, 798–801.
Elia M, Behrens R, Northrop C, Wraight P & Neale G (1987)
Evaluation of mannitol, lactulose and 51-Cr-labelled
ethylenediamine tetra acetate as markers of intestinal
permeability in man. Clinical Science 73, 197–204.
Fleming SC, Duncan A, Russell RI & Laker MF (1996)
Measurement of sugar probes in serum: an alternative to
urine measurement in intestinal permeability testing. Clinical
Chemistry 42, 445–448.
Gadomski AM, Kjolhede CL, Wittpenn J, Bulux J, Rosas AR &
Forman MR (1989) Conjunctival impression cytology (CIC)
to detect sub-clinical vitamin A deficiency: comparison of CIC
with biochemical assessments. American Journal of Clinical
Nutrition 49, 495–500.
Herrera MG, Fawzi WW, Nestel P, El Amin A & Mohamed KA
(1996) Effect of vitamin A on the incidence of cough, diarrhoea
and fever. Report of the XVII IVACG Meeting, Guatamala City,
Guatemala, p. 95. Washington, DC: IVACG Secretariat.
Hubble D & Littlejohn S (1963)
D
-Xylose excretion test in coeliac
disease in childhood. Archives of Disease in Childhood 38, 476–
480.
Kessing V (1968) Mucous gland system of the conjunctiva. Acta
Ophthalmologica 95, 119–133.
Lamabadusuriya SP, Packer S & Harries JT (1975) Limitations of
the xylose tolerance test as a screening procedure in childhood
coeliac disease. Archives of Disease in Childhood 50, 34–39.
Lunn PG, Northrop CA & Northrop AJ (1989) Automated
enzymatic assays for the determination of intestinal permeability
probes in urine. 2. Mannitol. Clinica Chimica Acta 183, 163–
170.
McCollum EV & Davis M (1915) The essential factors in the diet
during growth. Journal of Biological Chemistry 15, 23.
Meneghetti L, Ceriani T, Maccabelli G, Ferraris P, Gori S &
Ventura U (1989) Potential difference and mucus content in
the stomach of vitamin A deficient rats. European Journal of
Physiology 415, 30 Abstr.
Menzies IS, Laker MF & Pounder RE (1979) Abnormal intestinal
permeability to sugars in villous atrophy. Lancet ii, 1107–1109.
Miller KW, Stoltzfus RJ, Hakimi M & Rasmussen KM (1993)
Conjunctival impression cytology as an indicator of vitamin A
status in lactating Indonesian women. American Journal of
Clinical Nutrition 58, 167–173.
Natadiastra G, Wittpenn JR, Muhilal M, West KP Jr, Mele L &
Sommer A (1988) Impression cytology a practical index of
vitamin A status. American Journal of Clinical Nutrition 48,
695–701.
Nelson JA, Wiley CA, Reynolds-Kohler C, Reese CE, Mergaretten
W & Levey JA (1988) Human immunodeficiency virus was
detected in bowel epithelium from patients with gastrointestinal
symptoms. Lancet i, 259–262.
Northrop CA, Lunn PG & Behrens RH (1990) Automated
enzymatic assays for the determination of intestinal permeability
AMA PNS0071.fm Page 292 Thursday, June 24, 1999 3:13 PM
Postgraduate symposium 293
probes in urine. 1. Lactulose and lactose. Clinica Chimica Acta
187, 163–170.
Northrop-Clewes CA, Lunn PG & Downes RM (1997) Lactose
maldigestion in Gambian infants. Journal of Paediatric
Gastroenterology and Nutrition 24, 257–263.
Reddy V, Rao V, Arunjyothi & Reddy M (1989) Conjunctival
impression cytology for assessment of vitamin A status.
American Journal of Clinical Nutrition 50, 814–817.
Ross AC (1992) Vitamin A status: relationship to immunity and the
antibody response. Proceedings of the Society of Experimental
and Biological Medicine 200, 303–320.
Shenai JP, Kennedy KA, Chytil F & Stahlman MT (1987) Clinical
trial of vitamin A supplementation in infants susceptible to
bronchopulmonary dysplasia. Journal of Pediatrics 111, 269–
277.
Sommer A, Green WR & Kenyon KR (1982) Bitot’s spots
responsive and non-responsive to vitamin A. Clinicopathologic
correlations. Archives of Ophthalmology 99, 2014–2027.
Tanumihardjo SA & Olsen JA (1991) The reproducibility of the
modified relative dose response (MRDR) assay in healthy
individuals over time and its comparison with conjunctival
impression cytology (CIC). European Journal of Clinical
Nutrition 45, 407–411.
Vijayaraghavan K, Radhalah G, Prakasam BS, Sarma KVR &
Reddy V (1990) Effect of massive dose of vitamin A on
morbidity and mortality of Indian children. Lancet ii, 1342–
1345.
World Health Organization (1995) Division of Diarrhoeal and
Acute Respiratory Disease Control, Vitamin A Supplementation
and Childhood Pneumonia. Geneva: WHO.
Wittpenn JR, Tseng SCG & Sommer A (1986) Detection of
early xerophthalmia by impression cytology. Archives of
Ophthalmology 104, 237–239.
Wolbach SB & Howe PR (1925) Tissue changes following
deprivation of fat-soluble A vitamin. Journal of Experimental
Medicine 42, 753–777.
Wolf G (1978) A historical note on the mode of administration of
vitamin A for the cure of night-blindness. American Journal of
Clinical Nutrition 31, 290–292.
Zile MH, Bunge EC & DeLuca HF (1981) DNA labelling of rat
epithelial tissues in vitamin A deficiency. Journal of Nutrition
111, 777–788.
© Nutrition Society 1999
AMA PNS0071.fm Page 293 Thursday, June 24, 1999 3:13 PM