Identifying and characterizing zinc deficiency
Zinc deficiency and altered food intake
Critical periods or behavioral teratology experiments in animals
Concurrent experiments in animals
Developmental/concurrent studies in monkeys
Studies in humans
Mechanisms of zinc deprivation effects on behavior
Caveats for integrating conclusions from animal and human studies
Acknowledgments
Literature cited
MARI S GOLUB,*3 CARL L KEEN,* M ERIC GERSHWIN* AND ANDREW G. HEIYDRICKX$§
1Prepared for the International Dietary Energy Consultative Group (lDECG) Task Force workshop on malnutrition and behavior at the University of California, Davis, CA, December 6-10, 1993. This workshop was supported by IDECG, the Nestle Foundation, Kraft Foods, and the International Union for Nutritional Science. Guest editor for this supplement publication was Ernesto Pollitt, Department of Pediatrics, University of California, Davis, CA, 956 16.
2Supported by NIH grants HD14388 and RR00169.
3To whom correspondence should be addressed: California Regional Primate Research Center, University of California, Davis, Davis, CA 95616.
*Department of Internal Medicine, Department of Nutrition, California Regional Primate Research Center, §Department of Anatomy, University of California, Davis, Davis, CA 95616
ABSTRACT The majority of studies of developmental zinc deficiency and behavior were conducted in laboratory animals, primarily rats and rhesus monkeys. Effects on food intake c' complicate interpretation of experiments using severe zinc deficiency. Severe zinc deficiency in rats during the period of rapid brain growth has similar effects to protein calorie malnourishment during this period, including altered emotinality and food motivation. When behavior is tested durig a period of zinc deprivation in immature animals, lethrgy (reduced activity and responsiveness) is a prominet characteristic, but learning, attention and memory are also affected. The few supplement studies available in children did not report effects on behavior. Although zinc has multiple roles in brain function, considerable brain ~ paring occurs in zinc deficiency, and peripheral mechanisms of altered behavior also need to be considered.. Nutr. 125: 2263S-2271S, 1995. INDEXING KEY WORDS: zinc deficiency
children animals behavior development |
Zinc is an essential trace element contained in protein-rich foods such as whole grains, meat, fish and shellfish. Zinc deficiency often accompanies protein/calorie malnourishment. Periods of rapid growth such as late pregnancy, infancy adolescence are most susceptible to induction of dietary zinc deficiency (Favier 1992). Breast milk zinc content is not sensitive to maternal zinc intake (Lonnerdal 1996) but can be deficient in generally malnourished populations (Lehti 1990)
In addition zinc deficiency can occur in the absence of frank malnourishment in populations with poor quality diets (limited animal protein! or with traditional foods and food preparations that limit zinc bioavailability. Most widely recognized in limiting biouvailability is the phytate content of predominately vegetable diets and those that use unleavened grain staples (Torre et al. 1991). Infants and children also experience zinc deficiency in conjunction with parenteral feeding (Dahlstrom et al. 1986), malabsorption syndromes and a genetic syndrome, acrodermatitis enteropathica (Van Wouwe 1989)
Zinc deficiency syndromes have been reported in adolescents, infants and, more recently, in school-age children. Adolescents in the Middle East were one of the first populations identified as zinc deficient due to a striking syndrome of arrested sexual maturation that was reversed by supplements containing zinc (Halsted et al. 1972). Infants in the United States and France were studied in relation to clinical trials of zinc supplements in growth faltering (Walravens et al. 1989, Walravens et al. 1992). Reports in school-age children include clinic populations of short children (Buzina et al. 1980, Ghavami-Maibodi et al. 1983) and surveys of nutrient intake in populations worldwide (Ferguson et al. 1989, Gibson et al. 1991, Udomkesmalee et al. 1992). Growth, and in particular linear growth, is the main focus of these studies (Allen 1993), although information on behavior is sometimes anecdotally reported. Only a handful of studies, all in school-age children, directly addressed the effects of compromised zinc nutriture on behavior (Cavan et al. 1993a, Cavan et al. 1993b, Gibson et al. 1989, Wachs et al. 1995).
In contrast to research on iron
deficiency, the majority of available studies of zinc deficiency and behavior, as reviewed
here, were conducted in animal models and include work with monkeys. The finding in 1966
that severe zinc deficiency is teratogenic in rats (Hurley and Swenerton 1966) stimulated
a large body of research in laboratory animals on the effects of developmental zinc
deficiency including effects on the nervous system as evaluated through behavioral
studies. Developmental zinc deficiency was also widely studied in domestic animals,
particularly food production animals, and was an important consideration in optimizing
dietary aspects of husbandry (Lueke 1984). An unusual feature of research on developmental
zinc deficiency is a fairly large number of studies in nonhuman primates. Nonhuman primate
models are especially important in generalizing the results obtained in laboratory and
domestic animals to human populations, given the paucity of data from humans.
A major factor impeding studies of zinc deficiency
in human populations is lack of a suitable marker for zinc status. Zinc deficiency can be
detected in plasma, hair and tissue (WBC, RBC) (Meadows et al. 1981) zinc values, activity
of circulating zinc-dependent enzymes (Ruz et al. 1992) and plasma metallothionein (Grider
et al. 1990). Severe zinc deficiency syndromes, characterized by clinical signs such as
parakeratosis, dermatitis, anorexia and weight loss, are rare. Clinical signs of moderate
deficiency include hypogeusia and immune function suppression. Each of these markers is a
valid reflection of zinc status as demonstrated in controlled human and animal studies
using dietary zinc deprivation. However, all these markers are multidetermined, and, in
complex and dynamic situations of malnourished populations, they are less reliable.
There is no central, recruitable body store of zinc whose depletion would reflect dietary deprivation. However intracellular zinc, as well as zinc content of some tissues such as brain, are tightly regulated by a number of mechanisms. Sensitive signs of zinc deficiency are likely to be detected in homeostatic responses to inadequate zinc supplies. Indeed, two common symptoms of zinc deficiency, anorexia and growth retardation, serve to normalize plasma zinc and zinc supply to tissues. Changes in behavior, in particular reduced activity and responsiveness, may be one of the components of the zinc regulatory system that provides a sensitive index of inadequate dietary supplies. For example, we showed that during a period of accelerated post weaning growth, zinc-deprived monkeys who develop behavioral deficits continued to grow, whereas infants with arrested growth demonstrated normal behavior (Golub et al. 1985).
Zinc deficiency is usually
characterized as severe, moderate, or marginal/mild (Prasad 1988). Severe zinc
deficiency occurs with near complete absence of absorbable zinc in the diet and is
manifested in symptoms such as dermatitis and anorexia. In moderate zinc
deficiency, sensitive signs include reduced plasma zinc, retarded growth and depressed
immune response. Marginal or mild deficiency is a borderline state in which zinc
deficiency signs occur only in conjunction with other stress-ors (for example, rapid
growth). In rodents, dietary zinc contents associated with marginal/mild deprivation are
10 Dg/g, with moderate deprivation, 5-7 Dg/g,
and with severe deprivation, < 1-2 Dg/g. In monkeys, a
zinc-adequate diet contains ~ 50 DZn/g, marginal deprivation
occurs at 4 Dg/g, moderate deprivation at 2 Dg/g
and severe deprivation at < 1 Dg/g (Keen et al. 1993). Diets
used in experimental induction of zinc deficiency in humans provide 3-4 mg/d (Prasad et
al. 1978, Ruz et al. 1991, Sandstead et al. 1980), as compared with the RDA of 15 mg/d,
and the deficiency was characterized as mild. This points up the relatively wide margin
between optimal and inadequate dietary content and the potentially narrow margin between
marginally adequate and severely deficient diets.
An important consideration in planning and
interpreting studies of zinc deficiency and behavior is the induction of hypogeusia,
altered food preferences, decreased food use efficiency and frank anorexia by dietary zinc
deprivation.
Hypogeusia is a well-known effect of concurrent zinc deficiency in adults (Henkin 1984) that has also been described in children (Buzina et al. 1980, Gibson et al. 1989, Hambidge et al. 1971). In animals, hypogeusia is well characterized in rats (Catalanotto 1978) and mice (Liu et al. 1991). We showed both hypogeusia (increased acceptance of quinine solution) (Golub et al. 1984b) and altered food preferences (Golub et al. 1990) in young monkeys concurrently deprived of zinc as well as hypogeusia in adults (Golub et al. 1982).
Anorexia was widely demonstrated in laboratory animals, nonhuman primates and domestic animals only in conjunction with severe deficiencies. Anorexia was not well studied in humans and was not noted in the few experimental studies in humans, which did not employ severe deprivation. Anorexia has been associated with zinc status in malnourished children (mean age, 11.6 y) with childhood onset anorexia nervosa, but causative relationships are unclear (task et al. 1993).
Several hypotheses have been advanced to explain anorexia, including behavioral hypotheses concerning bait shyness and specific nutrient aversions/preferences (Cannon et al. 1988, Christensen et al. 1974) and biochemical hypotheses involving alterations in neurotransmitter systems due to amino acid imbalance (Wallwork et al. 1981) or to responsiveness of the endogenous opiate system (Essatara et al. 1984). A striking characteristic of zinc-induced anorexia in rats is its cyclic pattern (Cheaters and Quarterman 1970), which suggests buildup of a toxic metabolic product or a metabolic deprivation that periodically stimulates feeding/food aversion. A toxic metabolite is suggested by the finding that anorexia is less prominent when zinc-deficient diets have a low protein content (Reeves and O'Dell 1980). Zinc itself may provide the signal for cyclic feeding because muscle catabolism can serve to normalize zinc status (Masters et al. 1983), and zinc supplements can restore appetite and food intake in anorexic animals (Greeley 1984). However- cyclicity may also be an artifact of the laboratory situation in which a constant supply of one food is available. Use of alternate foraging strategies and food selections might instead be seen if different foods were available at varied times in a number of locations. Chafetz et al. (1984) showed that zinc-deprived rats made more errors in a radial maze task when only zinc-deficient food rewards were available, whereas performance was similar to controls if zincadequate rewards or a mixture of deficient and adequate rewards were available. Cannon et al. (1988) showed that animals who reject a zinc-deficient diet will eat a diet with dissimilar organoleptic properties.
Because of the anorexia induced by severe zinc deprivation, there has been a recurring question of whether developmental effects in animal studies can be ascribed entirely to reduced food intake rather than lack of zinc. This question is somewhat circular in that food-deprived animals will also be zinc deficient, whereas severely zincdeficient animals will be food deprived because of anorexia. In any event, pair-fed controls are almost an obligatory part of experiments using severe zinc deprivation. Pair feeding usually involves providing the same amount of food consumed by zinc-deprived animals to a control group (food restriction). In many cases, pair-fed animals show effects that are similar but of lesser magnitude than those shown by zinc-deficient animals. However, it was argued that the pattern of food intake is also important, and group differences would be reconciled if the pattern, as well as the amount, of food intake were identical in zinc-deficient and pair-fed groups. Approaches such as artificially feeding a zinc-deprived diet when spontaneous food intake is halted or providing zinc supplements to pair-fed animals proved problematic.
Moderate and mild zinc deprivation do not induce anorexia in animals and, in fact, frequently produce increased food intake accompanying decreased food use efficiency (Golub et al. 1984a, Haynes et al. 1987). Pairfed controls are not necessary in animal studies with moderate, marginal or mild deprivations. Hypogeusia, which occurs in moderate zinc deprivation, is probably not sufficient reduce food intake appreciably (Mattes 1985).
Altered food intake is interesting
in its own right as a behavioral manifestation of zinc deficiency. From a practical point
of view, altered food intake can have secondary impacts on general nutritional status. In
structured behavioral testing of animals, the value of food rewards and the motivation to
perform food-rewarded tasks could be altered.
In parallel to work with protein/calorie
malnutrition, animal studies of zinc malnutrition most frequently employed deprivations
during the period of most rapid brain growth. This type of study design is referred to
here as a critical periods design but is usually termed behavioral teratology in
connection with drug/toxicant administration. After the major developmental period for
brain growth, the treatment is discontinued and the animals are rehabilitated. Behavioral
testing is conducted later in adulthood (sometime after sexual maturation). Studies with
this design and either behavioral or other brain function end points are outlined in Table
1.
A number of critical period studies of developmental zinc deficiency in rats were conducted by Halas and collaborators (Halas 1983). Severe zinc deprivation was induced in pregnant rats during the last third of gestation (gestation d 14-20) subsequent to the major period of organogenesis. With this regimen, malformations were not induced, but brain growth could be affected. Initial studies (Hales and Sandstead 1975, Lokken et al. 1973) found effects on performance in shuttlebox shock avoidance and the Tolman Honzik maze, two standard tasks using negative (shock) and positive (food) reinforcers, respectively.
The studies using shock suggested enhanced response to stress, as demonstrated in a sudden drop in performance of shock-motivated learning tasks after initial learning. Unpublished data (Hales also suggested a greater susceptibility of the zinc-deficient (and pair-fed) offspring to stress-induced ulcers. Enhanced stress response was also seen as the cause of enhanced shockelicited aggression in females, but not males, as demonstrated in later experiments (Hales et al. 1975, Halas et al. 1977). In both cases' pair-fed controls were also affected but not to the same degree as zinc-deprived groups. Both these effects were attributed by the authors to increased emotionality. Using the same deprivation regimen, enhanced aggression in male offspring was demonstrated by enhanced aggressive behavior and reduced affiliation in dyadic encounters between zinc-deprived animals and controls (Peters 1979a).
In studies of maze learning with food reinforcement, rats severely zinc deprived during lactation and tested at 44 d of age made more errors than control or pair-fed groups (Lokken et al. 1973). In this early study, as in others published before 1980 (Caldwell et al. 1973), the current practice of considering the litter as the unit of analysis was not used. There were only two litters represented in each group, thus complicating interpretation of statistical results. The authors mentioned that altered food motivation might be responsible for the results, and later experiments addressed this issue directly using a bar-pressing task (Hales et al. 1980). Zinc-deprived animals could not be distinguished from pair-fed controls in reduced food motivation, and the effect was attributable to the smaller body size of previously malnourished animals. Other investigations of the partial reinforcement extinction effect and the negative contrast effect, which probe motivation by varying amount of food rewards, also demonstrated identical patterns of effects in zincdeprived and pair-fed animals and interpreted these effects in terms of enhanced incentive value of the food reward (Peters 1979b).
This pattern of effects, enhanced emotionality and enhanced food motivation, is very similar to effects reported for protein/calorie malnutrition in similar paradigms. For experiments that use severe developmental zinc deficiency, effects on shock and food motivation complicate the interpretation of tasks that employ these reinforcers and are intended to assess learning and memory. They also raise the issue of whether lack of zinc or reduced food intake occasioned by zinc deficiency are responsible causal variables.
Another hypothesis guiding research with developmental zinc deficiency was based on the knowledge that zinc is concentrated in the hippocampal area of the brain. In the nerve terminals of some hippocampal neurons (mossy fiber system) zinc ions are stored in vesicles that release their contents into the synaptic cleft during neurotransmission (Slomianka 1992). Some research indicated that severe zinc deficiency in adult rats led to behavioral changes resembling those produced by hippocampal lesions (Hesse et al. 1979). Hence, investigators looked at behaviors thought to be mediated in hippocampus (especially memory formation) in connection with
TABLE 1
Critical period studies: effects on behavior
recorded after a limited period of zinc deprivation during development'
Species extent and time of deprivation |
Effects in pair-fed controls |
Age at evaluation/results (relative to ad lib controls) |
Reference |
Rats |
ZD < PF = AL |
45 p |
Lokken et al. 1973 |
Severe, 14-20 g |
More errors Tolman-Honzik maze |
||
Rats |
ZD < PF = AL |
60 p |
Halas and Sandstead 1975 |
Severe, 14-20 g |
Impaired active avoidance |
||
Rats |
ZD > PF = AL |
105 p |
Peters 1979a |
Severe, 14-20 g |
ZD = PF > AL |
Enhanced aggression Reduced affiliation |
|
Rats |
ZD = PF > AL |
50P |
Peters 1979c |
Severe, 14-20 g |
Enhanced partial reinforcement and negative contrast effects |
||
Rats |
ZD = PF < AL |
38 p |
Peters 1979b |
Severe, 14-20 g |
Impaired relearning of spatial discrimination |
||
Rats |
ZD > PF > AL |
75 and 105 p |
Halas et al. 1975, 1977 |
Severe, 14-20 g |
Increased shock induced aggression |
||
Rats |
ZD < PF = AL |
10-20 p |
Halas et al. 1979 |
Severe, 0-21 p or 14-21 g |
Poorer memory (conditioned suppression) with 0-20 p deprivation but not 14-21 g
deprivation |
||
Rats |
ZD = PF > AL |
100 p |
Halas et al. 1980 |
Severe, O-21 p |
Better discrimination reversal learning (appetitive task) |
||
Mice |
ZD < PF = AL |
70 p |
Golub et al. 1983 |
Moderate and marginal, 16- 15 p |
Impaired passive avoidance, no change in hippocampal zinc, staining, no change in
plasma corticosterone |
||
Rats |
ZD < PF = AL |
90 p |
Halas and Sandstead 1982 |
Mild, 0-21 p |
Impaired working memory (radial maze, female) |
||
Rhesus monkeys |
2/4 PF dams aborted and 1/4 had a neonatal death |
Lactation |
Sandstead et al. 1978 |
Severe deficiency, 110- 150 g |
Reduced activity, exploration, play; more time with mom |
Strobel and Sandstead 1984 |
|
Post-weaning |
|||
No effect on discrimination reversals, impaired learning set |
Abbreviations used: g, days of gestation; gestation, 21 d in rats, 18 d in mice, 165 d in monkeys; p, days postnatal; sexual maturation, 35-40 d in mice, 50-60 d in rats, 3-4 y in monkeys. Conclusions ate those of the authors zinc deficiency. These studies did not generally support the hypothesis that zinc deprivation effects on behavior are mediated by depletion of hippocampal zinc.
One experiment in rats studied the the late nursing period when the hippocampus and long-term memory mature in rats (Halas et al. 1979). Tone-shock conditioning trials were given at 11, 14, 17 or 20 d of age, and behavioral suppression in response to the tone was measured 42 d later when the offspring were adults. Rats severely deprived postnatally, but not those deprived prenatally, exhibited poorer memory of d 17 training as adults.
In other studies, offspring of rats severely deprived of zinc from d 14 to 20 of gestation were trained in a Tmaze spatial discrimination learning task at 24 d postnatal (Peters 1979c). Both zinc-deprived and pair-fed offspring showed deficits relative to controls in relearning the task at 38 d of age but not at 84 d of age. This was attributed to poorer memory at the earlier interval (none of the groups demonstrated memory of original learning at 84 d of age).
Effects of severe lactational and mild/marginal gestational and lactational zinc deprivation were assessed in adults (90-100 d of age) using the radial maze, a task designed to differentiate working memory, short-term memory and long-term memory in rats. Preliminary reports of these studies describe inferior performance of the zinc-deprived offspring relative to controls (Halas and Sandstead 1980, Halas and Sandstead 1981, Halas and Sandstead 1982, Hunt et al. 1984)
A study in mice (Golub et al. 1983) assessed performance of passive avoidance, a task known to be sensitive to hippocampal damage and also to corticosteroids. Mice mildly or moderately deprived of zinc from d 1 6 gestation to d 15 postnatal were tested at 70 d of age. Passive avoidance latencies were lower in both zinc-deprived groups and intermediate in the pair-fed group. Hippocampal zinc content (as determined by histological stains) was not apparently altered by zinc deficiency.
TABLE 2
UC Davis studies evaluating behavior of zinc
deprived monkeys during development (infant, juvenile, adolescent)
Concurrent, cross-sectional study (zinc-deprived vs. control); dams fed marginally zinc-deficient diet from conception and offspring fed same diet after weaning
Age at evaluation |
Test |
Result |
Reference |
Newborn |
Neurobehavioral test battery |
Lower postural muscle tone |
Golub et al. 1984a |
No effect on reflexes |
|||
1 mo |
Activity level reflex and motor maturation |
Less activity (males) |
Golub et al. 1985 |
No effect on maturation |
|||
3-4 mo |
Mother-infant interaction |
Less activity |
Golub et al. 1985 |
No effect on interaction |
|||
7-12 mo |
Delayed response task |
Enhanced delayed response performance; impaired reversal performance |
Golub et al. 1985 |
Discrimination reversals |
|||
12 mo |
Open field exploration |
Less exploration |
Golub et al. 1985 |
Taste sensitivity |
Enhanced quinine acceptance |
||
20-24 mo |
Novel food preference |
Reduced novel food preference |
Golub et al. 1990 |
30-42 mo (adolescent) |
Discrimination reversals |
Impaired learning and reversal performance |
Golub et al. 1988 |
Concurrent longitudinal study (zinc-deprived vs. zinc-adequate diet periods);
young monkeys (20-24 mo of age) fed moderately zinc-deficient diets and zinc-adequate diet
for 15 wk |
|||
Test |
Result |
Reference |
|
Spontaneous motor activity (during testing) |
Less activity |
Golub et al. 1994 |
|
Continuous performance test |
Impaired performance |
Golub et al. 1994 |
|
Delayed match to sample test |
Slower improvement in performance at longer intervals |
Golub et al. 1994 |
In addition to rodent studies, one
study with a critical periods design in rhesus monkeys reported reduced activity,
exploration, and play, less time spent separated from the dam (Sandstead et al. 1978) and
impairment of learning sets (Strobel and Sandstead 1984) in offspring after a severe zinc
deficiency imposed during the third trimester. This study is weakened by lack of data from
the pair-fed control groups. Of four pair-fed dams, two had abortions and one had a
neonatal death. This was attributed to poor tolerance of pair feeding to zinc-deprived
animals who manifested severe anorexia.