Disease Source Material

Whether explicitly stated or not, the identification of parasites responsible for past epidemics is always indirect. The truth of this statement is dramatized by the 1918 influenza pandemic that killed millions of people in less than a year and a half (Jordan 1927). Despite early-twentieth-century advances in public health, microbiology, and immunology, an understanding of the variations of the influenza viruses was not attained until the 1970s (Mackenzie 1980; Stuart-Harris 1981). In 1918 and 1919 suggestions regarding the source of the disease ranged from Pheiffer's bacillus to swine influenza (Crosby 1976).

At least four methods are employed to reconstruct disease agents in historical epidemics: (1) evolutionary theory; (2) disease loads in nonimmune populations; (3) historical documents; and (4) skeletal biology. Typically, researchers rely on one or, at the most, two of these methods. To determine the strengths and weaknesses of any epidemiological reconstruction, the assumptions of each method must be evaluated.

Parasite Evolution and Ecology

Darwinian evolution is the theoretical framework for explaining the differential persistence of variation. Variation and natural selection are the principles that account for this persistence. Variation arises through such processes as mutation and genetic drift. Mechanisms of selection determine which variations persist or become extinct (Lewon-tin 1974; Levins and Lewontin 1985).

In the disease drama, individual variation between parasites and hosts determines, in part, which parasites become fixed in the disease load of the population. Parasites are predators seeking to maximize their reproductive fitness by colonizing hosts (Burnet and White 1972; Youmans, Paterson, and Sommers 1980). Not all predators are successful. Those that are take over cell machinery; cells then produce parasites at the expense of host cells.

The host's defenses, including skin, cilia, mucous membranes, and antibodies or immunoglobins, are the primary selective agents working to repel parasites. If the host's defense mechanisms are successful, there will be no predatory invasion. If they are not, illness and death of the host may ensue.

Besides explaining the relationship between predator and prey, Darwinian principles focus attention on how and why the relationship changes through time. If the parasite is specific in its reproductive requirements and kills the host before a new generation of parasites is released, the parasite faces death. A more successful strategy for the parasite is to inhibit or neutralize host defenses without causing the death of the host. As suggested by the European experience with epidemic or venereal syphilis (C. J. Hackett 1963; Hudson 1968; Crosby 1972), over time a kind of symbiosis evolves in which the host works for the parasite, but both host and parasite survive.

The pattern from initial virulence to relative quiescence has implications for the reconstruction of past infections. Quite simply, as I have stated elsewhere

(Ramenofsky 1987), present diseases cannot be accepted as analogs for diseases of antiquity. New parasites have become fixed by human populations; older parasites have evolved less destructive adaptive strategies; still other parasites, most likely unidentifiable, have died out. Although present processes of infection and transmission may be comparable to those of the past, the present human disease load does not mirror the past.

Epidemiologists with an interest in understanding past diseases are well aware of this problem of analogs. To overcome it, they have focused on reconstructing the human ecological settings most conducive to the fixing and persistence of human infectious parasites (Cockburn 1963, 1971; Black 1966, 1980; Fenner 1971, 1980). Immunology, population size and density, and population mobility are the key variables considered in this approach.

Most of the diseases considered in this essay have one trait in common: Recovery from the infection confers long-term or permanent immunity. Because of this protection, parasites require a continual pool of susceptible hosts to survive.

Large nucleated populations, such as those of cities, meet the criterion for such a pool. As documented archeologically, nucleated populations of any magnitude did not become a consistent feature of the human landscape until after the evolution of agricultural systems. Cities developed shortly thereafter, having a temporal depth in western Asia of approximately 5,000 years (Wenke 1984). Although it is likely that acute, infectious parasites periodically colonized individuals or populations before 5000 B.P., natural selection fixed these parasites only when population density passed a threshold necessary for parasitic survival. Even after 5000 B.P., the distribution of acute infections probably was uneven across Asia and Europe simply because population distribution was uneven.

Smaller nucleated populations that are sedentary or consist of those who practice a mobile settlement strategy are likely to be infected by other types of parasites. The list of potential parasites varies, depending on the degree of mobility, the presence or absence of herd animals or pets, and the size of the population. Both F. Fenner (1980) and F. L. Black (1980) think chronic or latent infections, including chickenpox (Varicella zoster) and Herpes simplex, are probable candidates for persistence in small populations. Zoonotic infections, including yellow fever (arbovirus) and tetanus (Clostridium tetani), that are transferred from animal reservoirs to humans by accidents of proximity are also likely.

On a global scale, these abstract treatments suggest that a variable set of diseases have infected human populations through time. Small mobile or sedentary populations are expected to have chronic or latent microbial flora. As population size or density of movement between populations increases, human groups are subject to acute infectious parasites that periodically erupt as epidemics. To determine whether these expectations hold on a local scale requires the use of other data sets.

In summary, Darwinism is crucial for any discussion of historical epidemics. Variability and selection explain the evolution and differential persistence of human diseases under differing environmental conditions. These simple principles account for the extinction of older diseases and the sudden onset of new ones. Ecological factors, including population distribution, nature of settlement, and immunology, are useful for stipulating the global distribution of chronic and acute infections.

Although important, Darwinian theory is also subject to limitations. Because of evolutionary changes in hosts and parasites, present diseases cannot be accepted as analogs for past diseases. In addition, theory, by definition, is ideational. Without testing ideas against the complexity of empirical variability, ideas about how evolution, persistence, or extinction occurred are empty.

Infectious Diseases in Population Isolates

The second method of reconstructing historical epidemics is to study the effects of introduced acute infections on living population isolates. Because of population size, distribution, or geography, these isolates, or virgin-soil populations, do not maintain acute infections endemically. These infections must be introduced from external sources. The temporal lapse between introductions of the same parasite coupled with the immunological responses of previously infected individuals determine whether the foreign parasite will have a minimal or maximal effect. Whereas temporal isolation of one year may be sufficient to cause an epidemic of the common cold (Burnet and White 1972), a six- to eight-year lapse may be necessary to cause a smallpox epidemic (Pitkanen and Eriksson 1984).

There are two approaches to the study of acute infectious outbreaks in nonimmune populations. During an epidemic, surveillance teams may continuously monitor the appearance of new victims and the recovery or death of older victims (e.g., Paul and Freeze 1933; Nagler, Van Rooyen, and Sturdy 1949; Christensen et al. 1953a,b; Peart and Nagler

1954; Phillip et al. 1959; Monrath et al. 1980; Brenneman, Silimperi, and Ward 1987; Gerber et al. 1989). Alternatively, antibody reactions can be measured in sputum or serological samples (Black and Rosen 1962; Adels and Gajdusek 1963; Brown, Gaj-dusek, and Morris 1966; Black et al. 1970; Brown and Gajdusek 1970; Neel et al. 1970).

Although both attack and mortality rates can be determined from disease notifications, the accuracy of the rates varies according to the severity of the outbreaks, the rapidity of spread, and the number of surveillance teams. In severe disease episodes, it is likely that both morbidity and mortality are greatly underestimated (Crosby 1976).

Because of immunological memory, antibody testing can be done at any time. Consequently, an entire population can be surveyed for the presence of measles antibody 15 years after the epidemic. Unfortunately, the resulting trends apply only to the survivors. Without independent evidence (e.g., death certificates, disease notifications), trends in mortality cannot be measured.

In addition to incomplete reporting, contemporary studies of epidemics in virgin-soil populations are subject to the problem of analogs. Although it is tempting to use modern morbidity and mortality data to account for what might have occurred 500 years ago, such an approach is erroneous for two reasons. First, as previously described, less deadly symbioses between parasites and hosts evolve through time. Second, where present, medical intervention will curtail the course of a disease outbreak. Because of these factors, contemporary responses of virgin-soil populations to acute infectious parasites are likely to be rather pale reflections of sixteenth-or seventeenth-century disease events.

Despite these limitations, knowledge of contemporary epidemics is crucial for understanding the infectious process. As demonstrated by the accidental introductions of measles to southern Greenlanders in 1951, stochastic contacts between infected and susceptible individuals still occur (Christensen et al. 1953a,b). In addition, these contemporary studies detail the complex interactions between the health of the hosts and primary and secondary invasions. Between 1942 and 1943 in the Yukon, a small band of Tlingit experienced six epidemics, including measles, malaria, whooping cough, diarrhea, and meningitis. With each episode, the health of the population deteriorated, allowing for the invasion of other parasites (Marchand 1943).

In summary, contemporary studies of the infectious process and disease loads of nonimmune populations are useful guides to understanding epidemic conditions in the past. These studies document the ease with which infectious parasites spread and the trauma that follows in the wake of epidemics. Rather than single epidemic events, it is likely that native Americans in the sixteenth and seventeenth centuries experienced waves of infections. The deterioration of health following an initial disease event established conditions appropriate for the invasion of other, allochthenous and autochthenous, parasites.

Historical Descriptions of Epidemics

Consulting historical documents is the third method of reconstructing past epidemics. Although lumped within a single heading, these sources vary from casual references to disease outbreaks to detailed descriptions of a single epidemic. Finding these documents demands great investments of time. Rarely are medical data indexed or otherwise distinguished in archives. Consequently, researchers may search assiduously through primary documents before coming across a single description of a disease event.

Descriptions of epidemics among native Americans during the Columbian period are subject to three sources of error. First, because contacts during the period were irregular, descriptions of any sort, including references to disease, are also irregular. Consequently, the first reference to a disease event may not mark the onset of introduced parasites. As late as 1800 contact between Europeans and native groups of the North American plains was still irregular. When Meriwether Lewis and William Clark visited the Upper Missouri tribes, the Mandan described a smallpox epidemic that had occurred 20 years earlier (Thwaites 1904).

Second, as discussed previously, the crudeness of European knowledge during the Columbian period affected the nature of early disease descriptions. For one thing, it is likely that observers failed to record many outbreaks. Moreover, a lack of familiarity with the people being described made it difficult for even the most conscientious explorers to estimate the number of deaths that diseases wrought. The Gentleman of Elvas (Bourne 1904), for instance, simply described the abandonment of villages visited by the expedition of Hernando DeSoto. Similarly, Daniel Gookin (1806) described an epidemic event from hearsay evidence that had decimated native groups of New England seven or eight years before the English established New Plymouth.

Third, there is confusion over disease names. Because Europeans were writing about native disease experiences, knowledge of what was known in Eu rope is crucial for understanding this ambiguity. C. Creighton's (1894) sourcebook on epidemics in Britain continues to be one of the finest and most readily available compendiums of evolving medical knowledge.

Before 1743, influenza was frequently confused with other agues or catarral fevers. Diphtheria was not defined as a separate illness until the eighteenth century. Before that time, it was lumped with scar-letina or the purples. Because measles and smallpox both produced rashes, the two terms were used interchangeably until 1593. By 1629, however, when the Bills of Mortality were established, smallpox and measles were consistently differentiated. Smallpox was noted as "flox or smallpox."

Because smallpox, measles, influenza, and diphtheria are acute and infectious, they all could have been introduced to the Americas at the time of first contact. Whether they were requires careful reading of documents informed by contemporary epidemiological knowledge. Historical documents by themselves are of marginal utility. The confusion over names, which they reflect, coupled with medical naïveté and unfamiliarity with the populations being described by observers, has led one microbiologist to state that documentary evidence "provides, at best, an accumulation of suppositions, since most accounts give insufficient details of symptoms or of epidemiological patterns for us to impute an aetiological agent" (Mackenzie 1980).

Despite these limitations, historical descriptions are crucial sources of information and must be integrated into any disease reconstruction. Documents not only highlight the presence of some infectious parasite, but force the historical epidemiologist to investigate what and why the disease event occurred.

Burial Populations

Irregularity of initial contact and the limitations of medical knowledge among observers during the Columbian period have been recognized as sources of error for some time. To compensate for these problems, other types of data have been deployed in reconstructing historical epidemics. Human skeletal remains are of some importance here (Milner 1980; Blakely 1989). G. R. Milner (1980) has argued that the analysis of mortuary populations from the earliest period of contact is of central importance in unraveling the timing and magnitude of aboriginal population loss from introduced infectious diseases.

Skeletal populations are employed in two ways. Analysis of the age-sex structure of a population can suggest whether its demographic profile reflects an epidemic episode (Blakely 1989; Blakely and Detweiler-Blakely 1989). Moreover, the frequency and type of skeletal pathologies can provide information on the presence, absence, and, perhaps, type of acute infectious agent (Steinbock 1976).

The use of reconstructed demographic profiles to infer past epidemics requires two assumptions, however. First, the mortuary population represents a single temporal interval comparable to an epidemic interval. Second, the burial sample is representative of those who lived. Both assumptions are difficult to support archeologically.

Although archeologists are skilled in controlling for time, the temporal resolution obtainable from the archeological record is rarely comparable to a single epidemic event. Only in unique depositional settings, such as the Tathum burial mound in west Florida (Mitchem 1989; Hutchinson 1990), does the weight of evidence suggest that the burial population represents an epidemic episode. Mortuary settings that incorporate 50 to 100 years are far more common.

Even if the temporal dimension is controlled for, the problem of the representativeness of the burial sample remains. Skeletal populations are typically biased samples of those who lived. Some age-sex classes are underrepresented; others may be over-represented (Buikstra and Cook 1980). Controlling for these biases is challenging, especially under epidemic conditions. As suggested by historical descriptions and rarity of burial populations during the Columbian period (Ramenofsky 1987), native populations may have ceased interment in some areas. This practice raises serious questions about the nature of available burial samples: Do they represent the victims or survivors of epidemics? Without controlling for sampling biases, accurate reconstruction of demographic profiles from cemetery data is, at best, difficult.

As with skeletal analyses, the investigation of osteological evidence of epidemic mortality has serious shortcomings. First, not all acute infections are expressed osteologically. In illnesses such as venereal syphilis, or tuberculosis, bone cells become involved only when the disease has progressed to a severe stage. If the individual dies before that point, the disease will not be reflected osteologically.

Second, bone cells have limited ways of responding to infectious agents: tissue loss, tissue gain, both loss and gain, or abnormal morphology (Ortner and Putschar 1981). The remodeling may be variously expressed, but different causative agents can result in the same osteological expression (Steinbock 1976;

Ortner and Putschar 1981). Although osteological remodeling may suggest the presence of an acute infection, stipulating the causative agent responsible for the remodeling is frequently impossible. Even in those rare instances where a specific acute infectious parasite is implicated as causing osteological remodeling (Jackes 1983), it is unlikely that the individual died from the infection. The process of osteological remodeling is of longer duration than the illness.

Because of inherent difficulties in isolating causative agents, bioarcheologists have begun to consider nonspecific stress indicators, especially osteomylitis and periostitus (Detweiler-Blakely 1989). If the health of native populations deteriorated after contact, the frequency or intensity of stress responses should have increased. To determine whether the suggested association holds requires the analysis of burial populations through time. Even if such a series were available and the appropriate pattern were documented, the cause or causes of the stress responses would remain elusive (Buikstra and Cook 1980). Periosteal responses, for instance, can be caused by anemia, treponemal infections, scurvy, trauma, or osteomylitis (Ortner and Putschar 1981).

In summary, inadequate control of the temporal dimension, biases present in cemetery populations, and nonspecific skeletal responses to disease limit the utility of skeletal populations as primary sources of data for reconstructing epidemic mortality in the Americas during the Columbian period. As a secondary resource, these investigations may provide information unobtainable through other sets of data.

The reconstruction of past epidemics is challenging because it is elusive. Instead of a single infallible method, historical epidemiologists must employ a suite of methods. Results of one investigation can then be weighed against those obtained from other inquiries. Even then, we can never be certain whether smallpox or malaria was the biggest killer of native Americans in the sixteenth century. At the same time, the integration of evidence from Darwinian theory, contemporary disease investigations of nonimmune populations, historical documents, and skeletal analyses can suggest the magnitude of the parasitic invasion that faced native peoples during the earliest centuries of European contact.

Your Heart and Nutrition

Your Heart and Nutrition

Prevention is better than a cure. Learn how to cherish your heart by taking the necessary means to keep it pumping healthily and steadily through your life.

Get My Free Ebook

Post a comment