Margaret J Weinbergerm. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, UK: Cambridge University Press, 2000.

Broadly stated, pica is the term given to the compulsive consumption of substances not generally considered food. However, a precise definition of pica is somewhat elusive because understandings of what constitutes “food,” what symptoms signify pica, and explanations of what causes the condition vary with historical and cultural context. A more specific definition of pica is “the chronic, compulsive eating of non-foods such as earth, ashes, chalk, and lead-paint chips …” (Hunter 1973: 171), but it may also include a “false or craving appetite” or “deliberate ingestion of a bizarre selection of food,” as well as the compulsive ingestion of nonnutritive or nonfood items such as ice and ice water (Parry-Jones and Parry-Jones 1994: 290).

Pica, in various forms, has been widely noted historically and geographically, primarily in medical texts and anthropological writings (see, for example, Laufer 1930; Cooper 1957; Anell and Lagercrantz 1958). Its practice, although not considered a disease, is of medical concern because ingestion of some substances may result in disease. Additionally, there are types of pica that have been linked by medical researchers to the correction of mineral deficiencies (see, for example, Coltman 1969; Crosby 1971; Hunter 1973). Pica is classified by the DSM-III-R (American Psychiatric Association 1987) and the ICD-10 of the World Health Organization (1992) as an eating disorder, along with anorexia nervosa, bulimia, and infant rumination. Various forms of pica have also been associated with mental retardation.

The incidence of pica in any particular population is difficult to determine because of a tendency to conceal eating behavior that may be considered abnormal or deviant within the cultural context. In addition, the varying definitions of pica we have taken note of also contribute to difficulty in documenting the scope of the behavior. Thus, although pica has been widely observed (both historically and geographically) and has been the subject of much research in many disciplines, it remains poorly understood.


Pica is generally classified according to the type of substance consumed. Names for subclassifications of pica are comprised of the Greek word for the ingested substance and the suffix from the Greek word “phagein,” meaning “to eat” (Moore and Sears 1994). Cross-culturally, the most commonly noted and explored type of pica is geophagy or geophagia, the consumption of earth and especially clay. Other types include ingestion of ice or ice water (pagophagia); laundry starch (amylophagia); hair (trichophagia); gravel, stones, or pebbles (lithophagia); leaves, grass, or other plants (foliophagia); feces (coprophagia); and unusual amounts of lettuce (lectophagia), peanuts (gooberphagia), and raw potatoes (geomelophagia). Paint, plaster, coal, chalk, cloth, pepper, coffee grounds, paper, cigarette butts, and other household items are also commonly consumed by those engaged in pica (Feldman 1986: 521).

There is some agreement historically and cross-culturally that the populations most prone to pica are young children, pregnant women, persons with mental illness, and the mentally retarded. There is also similarity cross-culturally in the types of items most frequently consumed. These include coal, ice, chalk, plaster, and various types of earth, in particular clay.

As a rule, children who engage in pica are under the age of 6 (Castiglia 1993). The things they consume are restricted by proximity to their grasp and normally tend to be relatively harmless items such as cloth, dirt, leaves, sand, and small rocks or pebbles (Parry-Jones and Parry-Jones 1992). Children who chew furniture or eat paint or plaster can be harmed, and plumbophagia, the ingestion of lead paint, is an important cause of lead poisoning. In fact, the practice of plumbophagia has led to the banning of lead-based paints for interior use in homes.

Trichophagia refers to the ingestion of hair and is one of the types of pica found most often among children. It is especially associated with the habit of girls chewing on long hair and is believed to be related to other, somewhat common, behaviors such as chewing one’s fingernails and chewing pencils (Higgins 1993). Trichophagia is of medical concern when substantial amounts of hair result in the formation of “hair balls” within the intestinal tract.


The term “pica” was first coined by the French physician Ambroise Paré in the sixteenth century, although references to the syndrome predated him by centuries. Aristotle and Socrates both wrote about “earth eating,” and during the classical periods of Greece and Rome, red clay lozenges from Lemnos were believed by physicians such as Galen to be antidotes for poison and cures for illnesses. They were also believed to facilitate childbirth. The lozenges were called terra sigillata (sealed earth) and stamped with the seals of the goddesses Artemis and Diana. As Christianity spread, these seals were replaced with Christian symbols and the lozenges, blessed by monks, were traded throughout western Europe and the Mediterranean region with the approval of the Roman Catholic Church (Hunter and De Kleine 1984).

The word “pica,” and its older variant, cissa, come from the Latin word for “magpie,” a bird thought to have a not very discriminatory appetite where edible and nonedible substances were concerned. Nineteenth-century medical texts describe both the eating behavior of magpies and that of humans with pica as consisting of an appetite for unusual edible and nonedible items (see, for example, Hooper 1811). The misconception that the magpie consumed earth and clay was likely based on observations of magpies collecting clay to build nests.

Pica was classified in Greek and Roman medical texts as a form of morbid or depraved appetite. In 1638, M. H. Boezo distinguished pica, the consumption of nonfoods, from “malacia,” a voracious appetite for “normal” foods. He attributed malacia in pregnant women to mental changes thought to occur in pregnancy. (As early as the sixth century A.D., pica was thought to result from the cessation of menstruation during pregnancy [Cooper 1957]). Today, malacia, or the craving and binge eating of specific foods, is considered a form of pica (Castiglia 1993). (For the early literature on pica, see Cooper 1957 and Halsted 1968.)

In mid-sixteenth-century England, pica was associated with coal eating among pregnant women and children. But within western Europe and the United States from the sixteenth century through the late nineteenth century, pica was commonly understood as the consumption by young women of substances such as lime, coal, vinegar, and chalk so as to achieve a pale complexion and otherwise improve on appearance (Parry-Jones and Parry-Jones 1994: 290). Historically, this condition was said to be accompanied by “chlorosis” or “green sickness” in prepubescent girls and young women. Chlorosis, a disease recognized from the sixteenth century through the late nineteenth century, was characterized by a loss of menses or irregular menstruation and was accompanied by symptoms such as listlessness, pallid skin, loss of appetite, and weight loss.

It is interesting to note that the debate in early medical literature about the causes of chlorosis prefigures the current debate about pica and iron deficiency in terms of cause and effect. For example, the consumption of nonfoods by young women in order to achieve a pale complexion could easily have resulted in iron-deficiency anemia or chlorosis. However, iron-deficiency anemia can cause cravings for nonfoods, and chlorotic females ingested large amounts of unusual foods such as pepper, nutmeg, and raw corn, as well as nonfoods such as plaster. In addition, psychological reasons such as sexual frustration and nervous conditions were considered possible causes of both chlorosis and pica (Loudon 1980). In the twentieth century, pica among young women has been manifested by excessive consumption of real foods such as fruit and vegetables and nonfood substances like ice (Parry-Jones and Parry-Jones 1992). (For an extensive historical account of pica within Western industrialized cultural contexts, see Parry-Jones and Parry-Jones 1994).

In various regions of the world, especially in tropical zones, pica most often takes the form of geophagy. Harry D. Eastwell has noted that geophagy is associated with the “world’s poor or more tribally oriented people” (1979: 264). Other investigators have characterized such groups as constituting “subsistence” societies (Hunter 1973), although in this chapter, the term “nonindustrialized” societies is used. In these societies, geophagy has been observed for many centuries and variously attributed to religious, cultural, and physiological causes (Hunter 1973; Hunter and De Kleine 1984; Parry-Jones and Parry-Jones 1992). Geophagy, or “dirt eating,” was also thought to be a peculiar affliction of enslaved Africans and, later, lower-income African-Americans and whites in the southern United States, who were characterized as “dirt eaters” (Forsyth and Benoit 1989).


Explanations of pica are numerous, reflecting the diversity of items consumed and the geographical regions within which the compulsive consumption occurs. In addition to the historical and cultural practices just mentioned, explanations also include psychiatric disorders and psychological and physiological needs, satisfaction of oral needs, behavioral disorders, responses to physiological or psychological stress, and the use of nonfoods for medicinal or pharmacological purposes (see, for example, Talkington et al. 1970; Hunter 1973; Crosby 1976; Eastwell 1979; Hunter and De Kleine 1984; Prince 1989; Horst 1990; Reid 1992). In fact, the practice of pica is so widespread both geographically and historically that one might be tempted to question its characterization as an abnormal practice.

Nonetheless, at least in the West, pica is considered an aberrant behavior warranting medical or psychological treatment. Within the medical literature, pica is discussed in terms of possible biological causes and their negative consequences. The psychological literature characterizes pica as a pathological behavior linked to other eating disorders or found among populations of children and the mentally retarded.

However, this tendency to view pica as a pathological eating practice, an idiosyncracy of “primitive” peoples, or an affliction primarily of the rural and impoverished has resulted in a failure to recognize the nutritional, medicinal, and cultural importance of geophagy in nonindustrialized cultural contexts.

Geophagy (Geophagia)

Geophagy, or the consumption of earth substances, is the most widely observed and researched type of pica. The term “geophagy” was first coined by Aristotle and means “dirt eating.” Geophagia has also been termed allotriophagia by Sophocles, erdessen in medieval Germany, mal d’estomac in French, citta in Latin, and cachexia africana—literally meaning a “wasting away of Africans”—a phrase employed by slave owners and physicians in the West Indies and southern United States.

Despite the historical evidence for various forms of pica throughout the world, geophagy is the only type to be discussed extensively from a cross-cultural perspective, especially in ethnographic anthropological and geographical literature. In transcultural perspective, both anthropologists and geographers have described it as a socially acceptable custom, with specific cultural meanings and functions. Such literature reports geophagy to be most frequently practiced in tropical areas of Africa, Latin America, and the Caribbean, as well as in the southern United States. It is also practiced widely in parts of Iran, India, and China, and in tropical areas of Indonesia and Oceania (Anell and Lagercrantz 1958; Hunter 1973).

In the Andes Mountains of Peru and Bolivia, two dozen comestible or edible earth substances have been found listed in pre-Columbian Incan sources, and about six different comestible earths have been discovered archaeologically in pre-Incan contexts. In fact, evidence that the practice of eating earth is at least some 2,500 years old in the Andean region was discovered when a specimen of comestible earth was recovered from a Bolivian site dating from 400 B.C. (Browman and Gundersen 1993). Although cultural explanations have been popular, physiological explanations of geophagy as an adaptive human behavior are also important (Hunter 1973; Hunter and De Kleine 1984; Johns and Duquette 1991; Browman and Gundersen 1993).

Some of the first domesticated plants contained substances toxic to humans and were treated in processes involving earth substances to make them less toxic. The absorptive properties of clay were well known to ancient physicians who included terra sigillata and terra silesiaca in their pharmacopoeias for the treatment of poison (Dannenfeldt 1984). Several authors note the example of a condemned man, during the sixteenth century, who elected to swallow a lethal dose of mercury if he were first allowed to ingest clay. He reportedly survived an amount of mercury three times the normal lethal dose and was granted a pardon for having contributed to the medical knowledge of the time (Halsted 1968; Dannenfeldt 1984). In our times, kaolinite is the common type of clay used in medicines and the primary ingredient in commercially marketed Kaopectate.

Clay is also used medicinally in nonindustrialized societies where hookworm is a common ailment. This intestinal parasite causes gastric distress that is frequently alleviated with clay. Clay is also employed as a treatment for diarrhea, heartburn, and intestinal gas and has been used to relieve nausea and vomiting in pregnant women (Anell and Lagercrantz 1958; Hunter 1973). But authorities caution that the practice of geophagy can also introduce intestinal parasites into the body (Castiglia 1993) as well as cause intestinal blockage and excessive wear of dental enamel.

John M. Hunter (1973) hypothesized that a connection exists between mineral deficiencies (particularly those resulting from increased nutritional requirements during pregnancy) and the cultural practice of geophagy in Africa. Moreover, along with Renate De Kleine, he suggested that clay eating in Central America may be a “behavioral response to a physiological need” created by various mineral deficiencies, particularly during pregnancy (Hunter and De Kleine 1984: 157). Similarly, Donald E. Vermeer investigated geophagy among the Tiv of Nigeria and the Ewe of Ghana and discovered that clays were consumed during pregnancy as a treatment for diarrhea as well as for the minerals that they contain (Vermeer 1966; Vermeer and Frate 1975).

Hunter, who described geophagy in Africa as “common among children and adults” (1973: 171), acknowledged that the practice also has a cultural basis. For example, earth, taken from a shrine or holy burial site, is eaten for religious purposes or to swear oaths. However, earth eating is most commonly viewed as a remedy. Syphilis, diarrhea, and gastrointestinal discomfort caused by parasitic diseases (such as hookworm) are all conditions treated with geophagy by the general population, but in Africa, the practice of consuming clay is most frequently found among pregnant women. In a field study conducted in the Kailahun District of Sierra Leone, Hunter (1984) found that 50 percent of pregnant women ate clay from termite mounds and 7 percent ate vespid mud or clay from mud-daubing wasps’ nests. In both cases, the mud was cooked over a fire until dried and blackened. An analysis of the clay to determine mineral content and availability for humans led Hunter to conclude that the practice of geophagy was “sensible and appropriate behavior”—as the clays made a significant contribution to the calcium, manganese, iron, and other mineral requirements of pregnant women (1984: 11).

In other parts of Africa, well-known and highly regarded clays are “extracted, processed, and passed from producers through middlemen to retailers and reach a wide consumer public through a network of periodic markets” (Hunter 1973: 173). Analysis of the nutritional content of a sample of such clays from Ghana revealed that clays of distinctive shapes (indicating a sort of brand) provided mineral supplementation in distinctive, varying proportions. For example, Hunter wrote:

a distinctive end-to-end, truncated cone-shaped clay (sample six), is richest in two macronutrients, potassium and magnesium, but is poor in calcium. Another distinctive clay, sample 12, the slim, fingerlike bar, is richest in two micronutrients, zinc and manganese, but is weak in iron and copper. (1973: 177)

Hunter concluded that in view of the deficiencies inherent in the diets of many Africans, clays supplement those diets with a wide range of minerals important for pregnant women.

Moving to the other side of the world, Hunter and De Kleine (1984) also evaluated the nutritional properties of clay used to make tablets sold at religious shrines in Belize. Samples were tested for 13 minerals, and the conditions of digestion were simulated to determine the bioavailability of those minerals. Although the tests showed great variation in the amount of minerals available, in all cases the mineral content (and mineral availability) was sufficiently strong to suggest that the clay did indeed provide nutritional supplementation. A daily dose of from 1 to 6 tablets delivered 11 minerals in varying degrees, depending on the clay sampled. For example, 1 tablet yielded 73 milligrams of calcium and 68 micrograms of copper, among other minerals. A sample dose of 6 tablets provided 10 milligrams of iron, more than 200 micrograms each of nickel and manganese, and 12 micrograms of cobalt. When compared with U.S. Recommended Dietary Allowances and recommended supplements during pregnancy, a single tablet provided about 9 percent of the recommended iron and calcium.

Yet some types of clay are suspected of having properties that inhibit the absorption of minerals from the gastrointestinal tract, and there is no consistent agreement that iron from clay is useful in correcting anemia. Still, Hunter (1973), De Kleine (Hunter and De Kleine 1984), and David L. Browman and James N. Gundersen (1993) have asserted that clays can serve as a culturally acceptable and nutritionally functional source of mineral supplementation within some nonindustrialized societies.

In addition, there is an economic factor. Vermeer (1966, 1971) and Hunter (1973) have both documented how clay is incorporated into trade practices and how its processing has been the basis for local cottage industries.

Cultural Diffusion: Case Studies

Perhaps the best approach to understanding the phenomenon of geophagy involves the merging of nutritional, cultural, and economic explanations. In other words, from an African standpoint it might be viewed as a physiologically based adaptive behavior supported by religious and other cultural beliefs and institutionalized within the local economy.

Moreover, many have noted the diffusion of geophagy via the slave trade with the result that the practice was frequently observed among enslaved Africans in the West Indies, South America, and the southern United States. In addition, forms of geophagy continue to be documented among African-Americans in the southern United States and among those who migrated to northern cities and took the practice with them. It was common to have dirt or clay from a particular site in the South sent north by relatives. However, more recently, amylophagia (consumption of laundry starch) seems to have replaced clay eating, and geophagy generally seems to have declined significantly in recent decades.

During the last century, cachexia africana was of concern to physicians and plantation owners, who viewed it as an important cause of death among enslaved Africans. The condition they described seems similar to the syndrome of chlorosis we have already described. Both chlorosis and cachexia africana were characterized by a seemingly uncontrollable desire for the substances that are eaten. Although the reasons advanced for the aberrant appetites of young white women and African slaves are different, in both cases questions of cause and effect arise. David Mason (1833) was one of the first to suggest that obsessive dirt eating or atrophia a ventriculo (stomach atrophy), as he termed it, was actually a consequence of disease, and not its cause. He wrote:

The train of symptoms that progressively arise from atrophy of the stomach and dirt-eating are indigestion and emaciation; a bloated countenance; a dirty-yellow tinge in the cellular tissue of the eyelids; paleness of the lips and ends of the fingers; whiteness of the tongue; great indolence, with an utter aversion to the most ordinary exertion; palpitation of the heart; difficult, or rather frequent and oppressed, respiration, even during moderate exercise, which never fails to induce a rapid pulse; habitual coldness of the skin; and occasional giddiness of the head, attended with a disposition to faint, sometimes causing a state of stupor. (1833: 292)

In addition, he noted a change in the color and density of blood, the appearance of skin ulcers, polypi in the heart, and pathologies of the liver and gallbladder. Mason also reported that the obsession with nonfood items extended to eating “cloth, both linen and woolen …” (1833: 291).

Mason and later authors (Laufer 1930; Anell and Lagercrantz 1958) discussed a variety of conflicting explanations for the disease. Plantation owners often considered it either a kind of addiction or a means of escaping work, or both, and punished geophagy with confinement, beatings, and the use of metal mouth-locks. But others believed that homesickness and depression or abusive conditions led to the consumption of dirt; Berthold Laufer (1930) and Bengt Anell and Sture Lagercrantz (1958) mention the conviction, held by some, that geophagy was a deliberate, slow form of suicide among the enslaved Africans. Mason, by contrast, foreshadowed the current general debate about iron deficiency and pica in his belief that dirt eating was “rather a consequence than a cause of the disease” and suggested that the earth eaten contained “useful ingredients mixed up with much hurtful matter” and that “iron and alkalis are of great efficacy in this disease” (1833: 289, 292). He recommended exercise, proper diet, cleanliness, and proper clothing, as well as purely medical treatment with emetics, purgatives, and tonics. Unlike geophagy more generally, cachexia Africana appeared to afflict both men and women in large numbers.

Hunter (1973) has noted similarities between geophagy in Africa and that practiced more recently by African-Americans in the United States, leading him along with others (see, for example, Vermeer and Frate 1975) to the view that geophagy arrived in the United States with enslaved Africans in a process of cultural diffusion. In both geographical locations the practice is found most often among women (especially pregnant women), and clays are gathered from special sites such as anthills, termitariums, or river banks and often referred to by the location (Forsyth and Benoit 1989). In Africa and in the southern United States, clays are shaped, baked, or sun-dried before eating and are claimed to have health-giving properties, as well as providing satisfaction. Reasons cited by southern women (black and white) reflect folk beliefs about the medicinal qualities of earth eating, while also indicating its character as an addiction:

I craves it. I eat dirt just the same way you would smoke a cigarette. I crave something sour like the taste of clay. It seems to settle my stomach. I know I shouldn’t eat it.When I go up in Jasper County I get it, but can’t find any good dirt here. This Biloxi dirt ain’t no good, so I gets my sister in Birmingham to send it to me. I never heard of a man eating dirt. They not got the same taste a woman has. My mother eats it because she be’s in the change and they say it will help her. When I was a child I was coming home from Sunday School and it had rained. I could smell the dirt on the bank and started to eat it then. Have kept it up. I would eat more dirt than I do, but I have a hard time getting it. (quoted in Ferguson and Keaton 1950: 463)

In moving to non-Western case studies of pica, we can encounter one of the few of a psychological nature. This one concerns an “epidemic” of pica (geophagy) that took place in Aboriginal coastal towns of northern Australia during 1958 and 1959 (Eastwell 1979). As diarrhea was presumed to be the cause of the earth eating, white nurses in the area treated patients with placebo tablets and with Kaopectate. The “epidemic” was ended, but the cessation of geophagy produced a community-wide disorder of hypochondriasis, and another outbreak of geophagy occurred among Aboriginal women who were past the years of childbearing.

Eastwell (1979), who discussed the epidemic, believed that the hypochondriasis and the earth eating of postmenopausal women had deep sociocultural causes. He argued that within the aboriginal hunting-and-gathering mode of subsistence, geophagy was originally practiced for medicinal purposes, but this function disappeared after these small-scale societies were colonized by the British. As the traditional hunting and gathering came to an end, the practice of geophagy was transformed into an indigenous cultural statement of Aboriginal status. Thus, bereft of its traditional medicinal function, the practice of pica among Aboriginal women became not only a type of protest, reflecting gender status, but also constituted a psychological adjustment strategy for women, whose economic importance diminished during the shift from a nomadic to a sedentary existence.

In a separate example, involving cultural diffusion and social change, Hunter and De Kleine (1984) provided a case study demonstrating the interaction of cultural and nutritional aspects of geophagia in Belize, where the indigenous population was resettled by the Spaniards in Santiago de Esquipulas, a pre-Columbian town of economic and religious importance to the Maya. “Esquipulas was noted for its shrine, health-giving earth, and sulfurous springs; thus it served as a place of spiritual significance and healing activity” (Hunter and De Kleine 1984: 157). The Spaniards built a chapel at the site to house a crucifix carved of balsam and orangewood. The crucifix darkened over time with the burning of incense and candles and became known as Nuestro Senor de Esquipulas, el Cristo Negro, or “the Black Christ.” The Black Christ was worshiped by the Indians and symbolized a cultural fusion of Christian and Mayan beliefs by bringing an “Indian saint” into what was a new religion for the Mayas. The Black Christ also became known for its miraculous cures, which focused attention on the healing properties of the spring and mud at the site.

In the 1700s, the site was formally recognized by the Catholic Church, and a sanctuary for the Black Christ was constructed. The shrine continues to be of religious importance; pilgrims visit it, particularly on January 15, the Day of Esquipulas, and during Lent. Prior to the institution of border regulations requiring passports and visas, an estimated 100,000 pilgrims annually visited the shrine to be cured of ailments including “leprosy, blindness, muteness, insanity, paralysis, rabies, yellow fever, malaria, tetanus, and hemorrhages” (Hunter and De Kleine 1984: 158).

Many of the alleged cures were credited to the tierra santa (holy clay) at the site, which is believed to have health-giving properties and is blessed by the Roman Catholic Church. Tierra santa was (and is) sold at the shrine in the form of clay pressed into small cakes, stamped with images of the Virgin Mary, the Black Christ, and other saints. The clay tablets or benditos (blessed ones) (either eaten or dissolved in water and drunk) are believed to alleviate diseases of the stomach, heart, and eyes, to ease menstrual difficulties, and to facilitate pregnancy and childbirth.

As the cult of the Black Christ and the reputation for cures spread throughout Central America, new shrines were built that also became associated with curative, blessed earth. Indeed, by the end of the eighteenth century there were shrines in at least 40 towns where supposedly curative earths were available for consumption.

Pagophagia and Amylophagia

Within Western industrialized societies, recent medical literature on pica is dominated not so much by geophagy as by pagophagia and amylophagia. The term “pagophagia,” first used in 1969 by Charles Colt-man, a U.S. Air Force physician, refers to the compulsive consumption of ice and other frozen substances. Some, however, do not view pagophagia as a form of pica because ice (and ice water) consumption can be a positive measure in controlling body weight and addictions such as the use of tobacco. In addition, of course, chewing ice is more socially acceptable within industrialized countries than eating dirt. Like geophagy, pagophagia is strongly associated by many researchers with iron and other mineral deficiencies.

Amylophagia, by contrast, is the eating of laundry starch, which is associated almost exclusively with women. It was first observed in rural areas of the southeastern United States, where it was thought by some to have replaced dirt eating. As we have already noted, Hunter described a process of cultural diffusion and change occurring as geophagy from Africa was brought first to the southern United States:

Next came the northward migration of blacks to the urban ghettos of Cleveland, Chicago, New York, and Detroit. Such migrants ask their southern relatives to mail them boxes of clay for consumption during pregnancy. At this state, however, the forces of culture conflict come to the fore: lack of local clay in the concrete jungles of the North, pressures of poverty, and stress on kinship ties with the South lead to the consumption of laundry starch replacing traditional geophagy. But micronutrient minerals are totally lacking in the starch. Calories apart, nutritional inputs are zero; gastric irritation is caused. A cultural practice is now divorced from nutritional empiricism; cultural adjustment to socioenvironmental change has broken down, and atrophy and decay are the result. (1973: 193)

Perhaps significantly, a similarity between dirt and laundry starch in texture (although not in taste) has been noted by investigators. The reasons cited by women for consuming laundry starch include the alleviation of nausea and vomiting associated with pregnancy and various folk beliefs, found largely among African-Americans, that consuming starch during pregnancy helps the baby “slide out” during delivery, promotes a healthy baby, or a whiter (or darker) baby (O’Rourke et al. 1967).

Like pagophagia and geophagia, as well as other forms of pica, amylophagia has also been associated with iron deficiency. Deleterious consequences include impacted bowels and intestinal obstructions.

Pica during Pregnancy

Pica has been associated since classical times with pregnant women. Until the twentieth century, pregnancy was commonly believed to cause mental instability—manifested, for example, in unusual food cravings. More recent studies of food preferences during pregnancy, however, report that changes in these, as well as the onset of specific cravings, are not universal phenomena.

Much of the research on pica among pregnant women in the United States has focused on those living in rural areas. The prevalence of pica among women considered at risk seems to have declined by about half between 1950 and 1970, but it has remained fairly constant from 1970 to the present. Nonetheless, it was estimated that pica is practiced by about one-fifth of pregnant women in the United States who are considered at “high risk” for this behavior.”High-risk” factors include being African-American, living in a rural area, having a family history of pica, and having practiced pica during childhood (Horner et al. 1991). Pregnant black women are over four times more likely than their white counterparts to engage in pica behavior. Additionally, pregnant women living in rural areas are more than twice as prone to pica as those living in urban areas.

Although some investigators have found no significant association between age and pica among pregnant women (Dunston 1961; Butler 1982), others have observed that pregnant women who practice pica tend to be relatively older than those who do not. It is interesting to note that women who report consuming clay tend to be older than those who report consuming starch (Vermeer and Frate 1979). One study, reanalyzing data from previous research, indicated that pregnant women who did practice pica were six times as likely to have a history of childhood pica than pregnant women who did not. Women who practice pica during pregnancy are also more likely to report pica behavior among family members, particularly their mothers and grandmothers (Lackey 1978). Little evidence of pica among white and upper-income women may reflect a lack of research among these populations (Keith, Brown, and Rosenberg 1970; Horner et al. 1991).

Among pregnant women in the United States, the three forms of pica that occur most frequently are geophagia, amylophagia, and pagophagia (Horner et al. 1991). Although, as we have noted, some researchers believe that as African-American women migrated to northern urban areas, laundry starch became a substitute for the more traditional clay eaten in the South (Keith et al. 1970), other research indicates that consumption preferences themselves might be changing, with younger women preferring starch over clay. In one study of rural women in North Carolina, participants indicated a preference for starch, even though their mothers had consumed both clay and starch (Mansfield 1977). Explanations of pica during pregnancy, like those of pica in general, range from the psychological through the cultural, to the nutritional (Horner et al. 1991; Edwards et al. 1994).

A recent study of eating habits and disorders during pregnancy mentions the case of a woman who, at 32 weeks of pregnancy, developed a craving for coal, reporting she found it “irresistibly inviting” (Fairburn, Stein, and Jones 1992: 668). Two other participants in the study developed a taste for eating vegetables while still frozen, which indicates something of the difficulty involved in determining pica incidence. The consumption of frozen vegetables, although not defined as pica by these researchers, would surely be considered a type of pagophagia by others.

In terms of medical consequences, pica has been related to anemia and toxemia among pregnant women and newborn infants (Horner et al. 1991). In some cases, pica reportedly contributed to dysfunctional labor (through impacted bowels) and maternal death (Horner et al. 1991). Pica during pregnancy has also been associated with a “poor” functional status of fetuses and infants, perinatal mortality, and low birth weight.

The authors of a report on pica in the form of baking-powder consumption that caused toxemia during pregnancy have pointed out that previous investigators discovered a significant correlation between toxemia and geophagia, but not between toxemia and amylophagia (Barton, Riely, and Sibai 1992). The case involved a 23-year-old black woman with anemia and hypokalemia who admitted to a one-and-a-half-year history of consuming up to 7 ounces of Calumet baking powder daily. The baking powder was considered a family remedy for gas discomfort. Ingestion of baking powder, comprised of 30 percent sodium bicarbonate with cornstarch, sodium aluminum sulfate, calcium acid phosphate, and calcium sulfate, is known to increase blood pressure. In this case, liver dysfunction and hypokalemia also resulted.

The psychological aspects of pica among pregnant women are similar to those of other pica practitioners. In addition to reporting a craving for the ingested substance, pregnant women exhibiting pica commonly say that they feel anxious when the substance is unavailable yet experience a sense of considerable satisfaction during and after eating the substance (Horner et al. 1991).

Pica and Iron Deficiency

Although no definitive connection has been established between pica and nutritional deficiencies, many have consistently linked pica with iron deficiency and its consequent anemia. Indeed, some have estimated that upward of 50 percent of patients with iron-deficiency anemia practice pica (Coltman 1969; Crosby 1976). It is interesting to note that the correlation of pica with anemia dates back to medieval times, and that iron therapy was prescribed as a cure even then (Keith et al. 1970).

As pointed out previously, pica behavior during pregnancy has also been strongly associated with iron-deficiency and iron deficient anemia. What remains unresolved is a problem of cause and effect. As Dennis F. Moore and David A. Sears wrote: “Some authors have suggested that the habit may induce iron deficiency by replacing dietary iron sources or inhibiting the absorption of iron. However, considerable evidence suggests that iron deficiency is usually the primary event and pica a consequence” (1994: 390). Although some insist that ingested starch inhibits iron absorption, Kenneth Talkington and colleagues (1970) have reported that this is not the case. These authors concluded that iron deficiency and anemia result from amylophagia only when laundry starch replaces nutritional substances in the diet.

Turning to clay ingestion, studies have found that its effect on iron absorption varies and depends upon the type of clay ingested. Some clays impair iron absorption, whereas others contain large amounts of iron. However, as already mentioned, there is no consistent agreement that iron from clay is useful in correcting anemia (Coltman 1969; Keith et al. 1970; Crosby 1976).

Coltman (1969), who first used the term pagophagia, was also one of the first to link the practice with iron deficiency. Indeed, he reported that the compulsive consumption of ice could be stopped within one or two weeks with iron treatment, even in instances where iron supplementation was not sufficient to correct iron-deficiency anemia. This dovetails with the work of William H. Crosby (1976), who has noted that although ice neither displaces other dietary calories nor impairs iron absorption, it is still the case that pagophagia is diminished when treated with iron supplements. Moreover, other cases of pica involving unusual ingested substances (e.g., toothpicks, dust from venetian blinds, and cigarette ashes) also respond positively to iron supplements (Moore and Sears 1994).

Perhaps even more powerful support for iron deficiency as a cause of pica comes from findings that intramuscular injections of iron diminish the habit of pica in children. But there is also evidence that intramuscular injections of a saline solution have the same effect, suggesting that the additional attention paid to children with pica behavior may help to reverse the condition (Keith et al. 1970).

Countering this theory, however, are two cases of childhood pica in which parental attention was apparently not a factor. One involved a 6-year-old boy with a 2-year history of ingesting large amounts of foam rubber, whereas the second case was that of a 2-year-old boy with a 6-month history of eating plastic and rubber items. Both of these cases of pica behavior were resolved through the administration of iron supplements, even though there was no increase in parental attentiveness (Arbiter and Black 1991).

In 1970, Louis Keith, E. R. Brown, and L. Rosenberg summarized the medical questions surrounding pica that required further investigation and clarification. These were:

  1. Is iron-deficiency anemia a direct adverse consequence of pica?
  2. Is iron-deficiency anemia an indirect result of nutritional replacement by unnatural substances, allaying the appetite for nutritional foods by filling?
  3. Are so-called cures of the habit of pica among children the result of increased attention or of injections of iron or saline solution?
  4. If those “cures” among children are the result of therapy, should the therapy consist of iron injections, saline injections, or an adequate diet high in iron content?
  5. Is the mechanism of pica among children different from that among adults, especially pregnant women? Would injections of iron reverse the habit in pregnant women?
  6. Does the coexistence of amylophagia and anemia adversely affect the pregnant woman, or are these two separate and distinct unrelated concomitant adverse conditions? (1970: 630)

As we have seen, although almost three decades have elapsed, there is still no consensus within the medical community regarding the answers to these questions. Instead, they are still being asked.

The physiological mechanism linking iron deficiency and pica behavior is not known. As Crosby noted, “Somewhere in our emotional circuits iron deficiency can sometimes cross the wires” (1976: 342). Somewhat more scientifically, it has been suggested that pica cravings are generated by a functional disorder of the hypothalamus, which is sensitive to changes in iron levels (Castiglia 1993).

As for pagophagia, Mary Elks (1994) has made two observations. She noted that even in industrialized nations, both geography and culture play a role in determining how pagophagia is viewed. For example, in England and other European countries compulsive ice eating is considered pathological behavior, perhaps indicative of disease. However, in the warmer climate of the southern United States, ice eating may be regarded as normal. In addition, she reported a case involving an entire family practicing pagophagia, including a 14-month-old girl for whom, Elks believed, ice consumption should not be assumed to be a learned behavior. She suggested that in some cases, familial or heritable factors cause pagophagia that is independent of other types of pica and probably not correlated with iron deficiency.

Psychological Explanations of Pica

Pica, as we have observed, is included as an eating disorder (along with anorexia nervosa, bulimia, and rumination in infancy) in the classification systems of the DSM-III-R of the American Psychiatric Association (APA) (1987) and the ICD-10 of the World Health Organization (1992). Pica is defined by the APA as the repeated consumption of nonnutritive substances for a period of at least one month, when the behavior is not attributable to another mental disorder.

It is interesting to note that psychological literature on eating disorders discusses pica in early childhood as a risk factor for bulimia in adolescence (Marchi and Cohen 1990). Moreover, some aspects of pica, such as excessive consumption of ice, ice water, lemon juice, and vinegar, are linked with anorexia nervosa (Parry-Jones 1992; Parry-Jones and Parry-Jones 1994). Pica has also been associated with rumination in children and persons with mental retardation; in such cases, the behavior may be interpreted as a regressive behavior reflecting oral needs that have not been met (Feldman 1986). Poor feeding and weaning practices are more frequently observed in children with pica than in those without (Singhi, Singhi, and Adwani 1981). In addition, comparisons of children who have iron-deficiency anemia and who practice pica and children with anemia who do not practice pica show that the former score higher on measures of stress, including that caused by maternal deprivation, child abuse, and parental separation (Singhi et al. 1981).

It is important to understand that psychological explanations themselves tend to reflect cultural beliefs in the Western industrialized world, in contrast to religious and spiritual beliefs of cultures in other areas. To view all types of pica behavior as pathology risks a failure to recognize other important issues, such as cultural variation in food preference, indigenous medicinal and nutritional knowledge, and the very real question of the effects of nutritional deficiencies.


Although explanations of pica vary with the type of substance consumed and with the cultural context, there are several consistent themes in the literature. One is that whereas men do practice pica, the behavior is most frequently associated with women and children. Second, regardless of cultural context and whether pica is considered acceptable among adults, there seems to be a uniform concern about the practice of pica by children.

A third has to do with similarities shared by the substances most frequently consumed. They tend to be brittle, dry, and crunchy. Moreover, the smell of the clay and soil is cited as important in a variety of cultural contexts, as is the location from which the clay or earth is obtained. These sites are frequently the homes of living things such as termites or cray-fish. There is also a pattern, historically and geographically, of consumption of earth or clay from sites of special significance. Such sites may be religious, as in Latin America and Africa, locally distinctive, as in the southern United States, or they may be places of burial, as in Asia, Africa, and Europe. Finally, there is a cross-cultural consistency in the debate concerning the relationship of pica to iron and other mineral deficiencies, and in the debate over whether the practice is psychologically or physiologically based.

One question that remains unexplored (in fact, it is barely mentioned in the literature) is why the definitions and syndromes associated with pica are not extended to the many practices in which men rather than women more typically engage, such as chewing tobacco and cigars or pipe stems, using snuff, and chewing toothpicks, betel nut, and chewing gum. While these habits may not generally involve consumption, they are not so dissimilar from the practices of pagophagia. Certainly, it appears that the relationship between the forms of pica we have discussed and other cravings and sources of oral satisfaction is an area in need of further investigation.