The Nutritional Link with Slave Infant and Child Mortality in Brazil

In 1970, a World Health Organization consultant wrote that “[B]eriberi has never been endemic in the large areas in the tropics and subtropics where rice is not a staple food.”¹ This study has three aims: (1) to prove the consultant wrong by demonstrating through the “discovery” process of beriberi in Brazil that it was, in fact, endemic there in the nineteenth century and earlier, especially among the slaves although rice was not an important part of their diet; (2) to explain how this could have been the case; and (3) to indicate the implications of widespread beriberi in Brazil for slave infant and child mortality and thus for slave demography. Among the methods I use is one of shining present-day medical knowledge on the past, which is fraught with some danger. Disease descriptions of yesterday can frequently mislead, and therefore “presentism” of this sort is sometimes frowned on by medical historians. On the other hand, beriberi has a number of distinctive symptoms, and thus, if Brazilian diets should have produced the disease, and if the symptoms of that disease were widespread, then it seems a reasonable supposition that the disease itself was rampant in Brazil.

Beriberi is normally associated with the rice-eating cultures of Asia because the basic cause of beriberi is thiamine or vitamin B1 deficiency, and the practice of stripping away the rice husk to prevent spoilage also strips away almost all of the thiamine that the grain contains. Thus, persons whose diets center too closely on rice become thiamine deficient, and, as this occurs, they lose their appetites, further exacerbating the deficiency and rendering them even more beriberi-prone. Women with bodies nutritionally depleted by pregnancy or the process of giving birth and males doing hard physical labor are the main adult victims.² Yet in Brazil, save for isolated spots in the Northeast, rice played little part in the diet until this century. Thus, it was something of a shock for nineteenth-century physicians there to discover the disease in their midst. That discovery was largely the work of Dr. José Francisco de Silva Lima, one of the members of what has come to be known as the Bahian School of Tropical Medicine.³

Such a discovery was complicated by more than the absence of rice in the diet. Because beriberi takes a “wet” form, so called because of the edema that the victim develops along with cardiac symptoms, and a “dry” form manifested in peripheral nerve lesions that affect different muscular systems often causing paralysis, it was frequently considered two distinct diseases, or even three when the wet and dry symptoms combined. In 1866, Silva Lima began a series of articles in the Gazeta Médica da Bahia on an unknown disease (or diseases) with these symptoms that was rife among the people of Bahia, and had been at least since 1863.⁴ About a year later, in the last 1867 issue of the Gazeta, Silva Lima first ventured the opinion that all of these symptoms stemmed from a single disease entity, namely beriberi. Few of Silva Lima’s Brazilian colleagues were instantly persuaded, but at least one foreigner was. The Siglo Médico of Madrid had reprinted the articles of Silva Lima, and these had been noticed in turn by Dr. A. Leroy de Méricourt, a distinguished French physician and an expert on beriberi who had already identified the disease in the West Indies. His subsequent article in the Archives de Médicine Navale was titled (in translation) “Beriberi Is Not a Disease Exclusive to the East: It Occurs in the Antilles and Brazil as Well.”⁵

In the study, Méricourt pointed out that although beriberi was thought to be restricted geographically to Asia, French physicians had reported epidemics of the illness among blacks on the sugar plantations of the French Antilles, and Cuban physicians had reported its occurrence on Cuba’s ingenios, where during annual appearances it carried off numerous slaves and Chinese contract workers. He next proceeded to analyze the work of Silva Lima, whose clinical observations had been so carefully set forth and who had correctly identified all three forms of the disease. Méricourt joined Silva Lima in pronouncing the disease in Brazil beriberi, which had clearly escaped its eastern confines.⁶

Brazilian physicians, however, were far from convinced, for while beriberi in the Caribbean had been linked to a rice diet, no such linkage was possible in Brazil. Even to the extent that they accepted the existence of beriberi, Brazilian authorities, according to August Hirsch, as late as 1887 were “most decided and almost unanimous in disputing all connection in the pathogenesis with deficient or improper diet.” Hirsch then echoed the difficulties expressed by Brazilian physicians in determining the etiology of beriberi by stating that “there has been no general diffusion of it [beriberi] until recent times, while there has been no such change in the state of the food of the people as would render its phenomenal outburst intelligible.”⁷ The outburst was indeed phenomenal, for by 1899 a medical graduate of Rio de Janeiro had shown beriberi to be rampant in Brazil, with a “notable presence” in Amazonas, Pará, Maranhão, Bahia, Espírito Santo, Rio de Janeiro, São Paulo, Paraná, Santa Catarina, and Matto Grosso.⁸

Interestingly, in 1887, the same year that Hirsch had written, Hernani da Silva Pereira, a Bahia medical graduate, inadvertently addressed the question of beriberi’s etiology in Brazil while investigating regional eating patterns in a chapter of his thesis entitled “The Most Common Foods in Our Provinces.”⁹ He wrote that in Rio de Janeiro and São Paulo provinces the diets were stratified between the well off and the poor, with the former consuming much fresh meat and beans, while those “without great financial resources” substituted dried meat for the fresh and replaced beans with cornmeal or manioc flour. In Minas Gerais, the diet was “more uniform,” and both rich and poor consumed a great deal of pork, cornmeal, and beans. On the other hand, in the far south, in Rio Grande do Sul, the diet of all classes centered on fresh meat, cereals, and vegetables. Finally, in the Northeast, the base of the diet was dried meat and manioc flour. Fresh meat there was “in general tired, very expensive, and what is worse, bad.” Thus, a little dried beef and a lot of manioc flour were the normal fare for the working classes of that region.¹⁰

The regional differences are revealing where beriberi is concerned. One of the constants, and certainly the most important of the culprits, is manioc flour, which is even lower in thiamine than milled rice (although this is not widely known). Rut another, more interesting, constant is the dried meat. In the Northeast, where the diet of most centered on it along with manioc meal, beriberi was at its most virulent. In Rio de Janeiro and São Paulo, where the poor made up the greater proportion of beri-beri victims, only the diet of the poor centered on dried meat. In Minas Gerais, where all classes consumed pork, there was little beriberi, and in Rio Grande do Sul, where the diet consisted of fresh meat, there was no beriberi.¹¹ Rio Grande do Sul was the center of Brazil’s meat-drying industry, and Pereira found it significant that while the people there produced dried beef for the whole of the country they would not eat it themselves. He explained that the drying process consisted of first salting the beef and then placing the beef slabs on racks to dry in the sun for a minimum of two months—thus providing one of the major reasons for beriberi in Brazil.¹²

All living things contain some thiamine, and though animal tissue, save for pork, is not a particularly good source of the vitamin, most victims of beriberi in Asia were individuals whose diet contained very little meat or fish. In Brazil some meat seems to have been eaten by practically everyone, and still beriberi was widespread. Significantly, however, those who consumed fresh meat did not get the disease, only those who ate the dried meat, and the latter came down with it for a number of reasons that would only emerge as the science of nutrition grew more sophisticated.¹³ One important reason is the destruction of thiamine in the meat: whereas salting and drying are hard on the thiamine content of tissue, irradiation destroys it. Moreover, salting and drying meat make it tough. Consequently, it is normally prepared by soaking it to remove the salt, and then cooking it for a lengthy period; this process would have eradicated any thiamine that might have survived the two or more months of exposure to sunlight. Thiamine is highly soluble in water, therefore much would have been thrown away in the salty water that was discarded after the soaking process was completed. Finally, thiamine is sensitive to heat, and is destroyed in prolonged cooking. To put it plainly: had humans deliberately tried to devise a method of thiamine destruction in beef it is doubtful that they could have invented a better system than salting, sun drying, then soaking and cooking.¹⁴ Nor would all of this have produced merely a neutral effect, i.e., the denial of the original thiamine in the meat to the consumer. To be properly utilized, the B complex must be in balance. Sunlight also destroys riboflavin, and thus the beef would have undersupplied both thiamine and riboflavin relative to niacin and thereby raised the normal vitamin B requirements of the consumer considerably.¹⁵

To be sure, few subsisted solely on a dried beef and manioc regimen, and other foods bearing thiamine such as beans and maize must have found their way into diets from time to time. But there is still one more deadly feature of the dried beef to take into account which would have assured that in many cases those foods could not have prevented beriberi. For, in drying meat, a good deal of the fat is removed, and the diet of many in Brazil was already extraordinarily low in fats. When the intake of fats is low, the body adjusts by utilizing carbohydrates as the major energy source. But while fats do not require thiamine for metabolization, carbohydrates do. In other words, a low fat diet significantly increases the thiamine requirements of the body.¹⁶

We have thus far established that the diet of many Brazilians was seriously thiamine deficient, and that beriberi, the major deficiency disease which lack of thiamine produces, was “discovered” to be a serious problem of health during the last half of the nineteenth century. The next question is: how pervasive was beriberi in Brazil before that discovery? The answer would seem to be that it was quite pervasive, for there are numerous glimpses of the disease through the mist of a still more distant past. Aristides Moll asserts that Brazil was plagued by it from early colonial times, although the first real description of what appears to have been beriberi in Brazil seems to have originated from the pen of Willem Piso. This physician, in the country during the Dutch occupation, made a number of medical observations which he published in his Historia naturalis Brasiliae (1648). Among them was his report of a disease that attacked the nerves and produced a “profound torpor in the limbs.” He called it “chronic and common.”¹⁷

In the middle of the eighteenth century, João Cardoso de Miranda, in writing about the diseases of Brazil, devoted a chapter to “stupor and paralysis.” Significantly, he wrote that this condition could be cured by fresh meat. The Brazilian-born naturalist, Alexandre Rodrigues Ferreira, who had studied at the University of Coimbra and was sent back to Brazil by the crown in 1783 to study its flora and fauna, encountered and reported on a strange but widespread disease which he actually called beriberi.¹⁸ The French physician Joseph François Xavier Sigaud, writing in the nineteenth century, recalled an epidemic that had raged in Rio in 1780, which he called the “grippe,” but which Brazilian physicians much later came to believe was beriberi based on the symptoms Sigaud had described. Visitors to Salvador in 1819 remarked that the city had more “dropsy” cases than any other they had visited, while Hirsch pronounced beriberi “endemic” to Brazil, from at least the 1820s, and still another observer denounced it as the “most notorious” of Brazil’s nineteenth-century ailments.¹⁹

Clearly then, the literature seems to contain ample justification for the suspicion that beriberi was causing people to swell and hobble, not to mention die, long before the physicians of Bahia identified it, and this means that the slaves (euphemistically the poor mentioned so often) had to be among the chief sufferers. For historically the disease has been especially attracted to closed populations who had little control over their diet, such as prison inmates, soldiers on campaign, and sailors confined to ships. Similarly, slaves on plantations often had little control over their diets, and the literature leaves no doubt that during the nineteenth century, at least, those diets in many cases amounted to little more than manioc and dried beef. Before the nineteenth century when dried beef became a mainstay, the slave diets of the Northeast seem to have focused heavily on manioc, and whenever and wherever there was insufficient supplementation, in the form of whale meat, shellfish, chickens, beans, and so forth, beriberi would have quickly appeared. Thus, a historian of Medicine in Pernambuco mentions beriberi as one of the diseases of slaves from the sixteenth century onward, and Santos Filho, perhaps the most famous of Brazil’s medical historians, believes that the disease was present among the slaves from the very onset of slavery.²⁰ Unfortunately, neither of these authorities tells us how widespread it was, nor what guise or guises it assumed before its detection in Brazil, which suggests the need to probe some of the more mysterious, as well as the more notorious, of the slave illnesses, even though such a quest leads straight into a semantic jungle of Brazilian nosology.

One heretofore unexplained slave disease was called “cansaço” (fatigue or exhaustion), which Stuart Schwartz discovered among about 6 percent of the slaves on some of the sugar estates of Bahia. A Bahian physician, writing on the foods of his region in 1845, commented that a great number of slaves on the engenhos were victims of this ailment “that the vulgar call cansaço” but that was properly termed opilação, which in his day meant inward “obstruction” diagnosed on the basis of outward swellings. Significantly, he proceeded to blame the symptoms of swellings and general weakness on precisely what would cause them—the slave diet. He deplored the fact that manioc, although the “bread of Bahia,” was not a healthy nutriment, and, when united with dried beef, brought on opilação. To clinch his case, he reported that on one engenho he knew, where the master substituted cornmeal for manioc flour, the slaves were more robust and opilações were rare.²¹ A few years later, a physician in Rio de Janeiro wrote about opilação on the engenhos of that region. He considered that manioc might be the culprit, but rejected the idea because, even though that flour was the slaves’ principal food, he knew of slaves with the illness whose masters provided them with wheat flour. He could not know that which beriberi researchers later discovered—that a diet consisting of mostly wheat flour which has not been enriched will also cause the disease, especially when combined with the dried beef.²²

In Pernambuco, opilação was often called “frialidade,” but here the connotation was a disease of the liver and stomach which reportedly killed many slaves. Importantly, gastrointestinal and liver complaints are both early signals of beriberi.²³Frialidade sometimes seems to have meant weakness or laziness, and students of Brazilian slavery have pointed out the frequent complaints by masters of the laziness and lassitude of their slaves. Though few would suggest that slavery was a highly motivating institution there may have been more at work in many of the instances than simply laziness or indifference. In fact, in the past many individuals in prisons or doing outside forced labor who have been accused of malingering actually turned out to be suffering from beriberi.²⁴

One extreme form of suspected slave malingering in which death almost invariably claimed the victim was called “banzo” in Brazil. Planters thought it a form of suicide and knew its symptoms well. The slave developed a melancholic appearance, sank into lethargy, and refused to eat until he or she quite literally starved to death. Because one of the functions of thiamine is to maintain the normal urge to eat, and because one of the first symptoms of thiamine deficiency is anorexia, along with “mental changes” and “easy exhaustion,” the suspicion is sharpened that many purported cases of banzo were actually cases of beriberi.²⁵

Other slave diseases may also have camouflaged beriberi, especially leprosy and elephantiasis. But one more mysterious disease that deserves mention because it seems to have frequently been beriberi in disguise was called the mal d’estomach. Just to add to semantical chaos, it was also called cansaço, opilação, hypoemia intertropical, and cachexia africana by different physicians in different parts of the country, although the descriptions make it clear that, no matter what it was called, the intention was to give a disease label to the practice of geophagy that was reportedly widespread (but secretly done) among the slaves.²⁶

After hookworm was discovered in Brazil, this infection gradually received credit for slave dirt eating as well as many of the opilações or obstructions already mentioned. There was good reason for this, since many of the symptoms of beriberi such as lassitude, shortness of breath, anorexia, and swelling of the legs could also be those of hookworm disease and vice versa. Yet when Rockefeller Foundation physicians began to investigate hookworm in Brazil they discovered that blacks there, as in the West Indies and the southern United States, possessed a relative immunity to hookworm disease and thus, as a rule, would not have exhibited the abnormal appetites that whites and those of Asian descent so often did with heavy hookworm loads. Moreover, it should be understood that dirt eating rarely kills anyone, and hookworm disease, if it kills at all, only does so slowly and in conjunction with intercurrent infections. Yet the various mysterious ailments that have been discussed frequently killed in an epidemic fashion compatible with a diagnosis of beriberi but not with hookworm disease. Finally, it is important to note that physicians continued to observe that where the slave diet was based on a core which contained a substitute for either the dried beef or the manioc flour, deadly ailments characterized by “obstructions” were rare. Similarly, these symptoms were seldom a problem in those parts of Brazil where the diet was sufficiently thiamine rich to keep beriberi at bay.²⁷

There was, however, still another kind of beriberi to be discovered in Brazil. In 1916, Bahian physicians had the opportunity to read in the Gazeta of the infantile variety, and the article must have surprised many, for it had previously been believed that beriberi was rare or nonexistent in the young.²⁸ But researchers now know that thiamine deficiency is the only serious nutritional deficiency that can be passed along from mother to nursing infant and child. In fact, human milk is so low in thiamine to begin with that, even with well-nourished mothers, the baby normally has a “slender margin of supply.” When the mother is thiamine deficient, her milk and consequently her nursing infant will also be deficient.²⁹

That deficiency may be one that the mother has lived with and will continue to live with for many years without showing outward signs of beriberi, for thiamine deficiency is best thought of as a spectrum with overt beriberi symptoms as signals in only the most severe cases. But even though women can live with thiamine deficiency, their infants, with no other source of food but their mothers’ milk, will almost certainly die. Similarly, children kept too long on the breast—and the slave young were nursed upwards of three years—without adequate dietary supplementation run the risk of their mothers becoming thiamine deficient during this prolonged period although healthy at the time they gave birth. That beriberi kills far more infants than adults can be seen from the case of the Philippines, where of the 25,000 deaths annually from beriberi during the 1950s, close to 80 percent were infant victims.³⁰

In light of the experience of the Philippines with beriberi, I would suggest for Brazil in the first place, that the mortality precipitated by beriberi in adults represented only the most severe cases of thiamine deficiency— cases which generally became manifest during that time of the year when supplemental foods were in shortest supply. But these cases were only the tip of an iceberg of thiamine deficiency, because estimates suggest that about 90 percent of those suffering from beriberi display mild or subacute symptoms, and for every case of overt beriberi in a population, hundreds or even thousands of individuals are suffering from thiamine deficiency to some extent. From this it follows, secondly, that because thiamine deficiency was widespread in Brazilian adults, infantile beriberi had to be a major killer of their infants as it was in Manila at the turn of this century, where it has been estimated that the disease killed about 40 percent of all babies born, before they reached six months of life.³¹

Clearly then, Brazil’s problems with beriberi were hundreds or even thousands of times greater than physicians initially suspected, and because of a limited diet the disease doubtless fell with much greater weight on the slaves than on the free population. That beriberi would have been virulent among this group, in turn, has important implications for slave demography in Brazil. Many who have studied slave fertility there have found it to be high. Yet, at the same time, most report that the slave population was always naturally decreasing even in the last decades of slavery.³² High mortality rates are generally viewed as the most important reason for the failure of Brazil’s slave population to grow, and I do not quarrel with that, but I believe that adult mortality was far less important in this regard than infant and child mortality, which demographic investigators have found to be “very high” and which one investigator termed “appalling. ”³³

In fact, I would argue that beriberi, as a major killer of slave infants and children, must bear much of the blame for the failure of Brazil’s slave population to grow by natural means. In Minas Gerais, where a recent study has indicated that the slave population may have done fairly well demographically, the diet was varied and included pork which is a good source of thiamine. But where the slave diet concentrated on manioc and dried beef, one may catch glimpses of the disease doing its deadly work of forestalling population growth. For example, José Pereira Rego, who was scrutinizing the causes of infant mortality in Rio during the 1870s, speculated that the infants were dying of “poor milk” from badly nourished mothers, and thus may have come tantalizingly close to pinpointing the etiology of infantile beriberi. Because of the nature of the “poor milk” in question, a characteristic of the disease is that it runs in families, meaning that a thiamine-deficient mother tends to lose one child after another in infancy. In this connection, Emília Viotti da Costa quotes slave owners who insisted that one white child was more likely to survive infancy than three or even four black babies because of the blacks’ “greater fragility.”³⁴

In Salvador, an investigator who has examined the records of the Santa Casa da Misericórdia located there discovered that patients often “came without speech,” suggestive of the aphonia of adult beriberi. In turn, the prevalence of adult beriberi during colonial times would help immensely in explaining another researcher’s depiction of slave infant and child mortality in the region as “horrendous.” During 1842 in Pernambuco, a disease was reported that was caused by “aberrations” in the diet and was killing an increasing number of infants as well as women in pregnancy or who had just delivered a baby. Later, a regional historian wrote of nine epidemics of beriberi between 1871 and 1885—all this in a region where Maria Graham and others had reported the death rate of the slave young to be well over 50 percent. In Rio de Janeiro, where the slave diet in 1849 was reported to consist primarily of manioc flour and dried salted beef and “obstructions” were “so common among the slaves,” a physician denounced that diet as responsible for the great mortality of the slave infants and children.³⁵

How great that mortality may have been can be inferred from statistics on the mortality in Rio de Janeiro for 1845-47, which revealed that infants accounted for 30.2 percent of the total mortality. And when these deaths were combined with the deaths of those aged one to ten years (mostly in the one-to-three-year-old group), they accounted for 52 percent of the total deaths. This mortality picture was termed “excessively horrible” and “unnatural” by a local physician who blamed it on a variety of ailments that, when brought together, fell mostly under the rubrics of tetanus, gastrointestinal complaints, hepatitis, lung diseases, marasmus, and mesenteric tuberculosis.³⁶³⁷ It is interesting to note that, aside from neonatal tetanus (which invariably occurred during the first two weeks of life), the normal course of infantile beriberi would produce symptoms that fit all of these categories. Severe vomiting and diarrhea (or constipation), which fall under the gastrointestinal rubric, are almost always present in infantile beriberi. As a noted beriberi researcher points out, “gastrointestinal derangement” has often concealed infantile beriberi in the statistics. So has bronchitis, also normally present in infantile beriberi cases, which in the Rio de Janeiro data would have been subsumed under the lung ailment category. Hepatitis was a term used by Brazilian physicians when they thought they were confronting liver problems. Babies with infantile beriberi tend to retain their urine and develop a peculiar waxy appearance” that would surely have been taken to mean liver ailment and have been called hepatitis.³⁷

Even the diagnosis of mesenteric tuberculosis suggests strongly that physicians were actually confronting infantile beriberi. In the nineteenth century, the diagnosis of tuberculosis in adults was a haphazard affair at best, and even today, in the opinion of experts, the “diagnosis of tuberculosis in infants and children, except in the presence of obvious clinical and/or radiographic evidence, is difficult.. . .”³⁸ What was it then that physicians in Rio de Janeiro were calling mesenteric tuberculosis? If by the term mesenteric they meant complaints of the abdominal cavity, it merely indicates another variation of gastrointestinal ailments. If, on the other hand, they meant the term to more specifically connote what today is called tuberculosis of the serous membranes (involving the pleura, peritoneum, and cerebral meninges), then they may have been diagnosing on the basis of symptoms of meningitis, and this could easily have been the pseudomeningeal form of infantile beriberi which strikes infants at between eight and ten months of age.³⁹

Infantile beriberi thus presents a variety of symptoms, and given the diet of the slaves and the poor generally it is difficult to believe that the disease was not deeply implicated in the holocaust in which the infants and children of Rio de Janeiro were caught up. And a holocaust it was, according to a physician who wrote of the great number of slave infants who died, implying that they dominated the infant death lists.⁴⁰ However, there were obviously a number of white infants who died as well, and physicians blamed many of these deaths on the practice engaged in by white mothers of turning their infants over to slave women to wet-nurse with “little thought given to their fitness for the task.” One doctor reported dourly that he and his colleagues still heard “too many times” parents express joy on the death of their infants, for now they had become angels, and he pointedly referred to the wet-nurses as “angel makers.”⁴¹ How terribly ironic that malnourished slave women may have inadvertently killed the infants of their masters because of that malnutrition.

Author notes