Solving the Mystery of an Ancient Roman Plague

Church records from the third century could help identify the disease that nearly killed the empire.

Jules-Élie Delaunay's painting "Plague in Rome"
Jules-Élie Delaunay's "Plague in Rome" (1869) (Photo Josse / Leemage / Getty)

The Plague of Cyprian, named after the man who by AD 248 found himself Bishop of Carthage, struck in a period of history when basic facts are sometimes known barely or not at all. Yet the one fact that virtually all of our sources do agree upon is that a great pestilence defined the age between AD 249 and AD 262.

Inscriptions, papyri, archaeological remains, and textual sources collectively insist on the high stakes of the pandemic. In a recent study, I was able to count at least seven eyewitnesses, and a further six independent lines of transmission, whose testimony we can trace back to the experience of the pestilence.

What is starkly lacking, however, is a Galen. The previous century’s dumb luck of having a great and prolific doctor to guide us has run out. But, now, for the first time, we have Christian testimony. The church experienced a growth spurt during the generation of the plague, and the mortality left a deep impression in Christian memory. The pagan and Christian sources not only confirm one another. Their different tone and timbre give us a richer sense of the plague than we would otherwise possess.

The lack of a medical witness like Galen is partly compensated by the vivid account of the disease in Cyprian’s sermon on the mortality. The preacher sought to console an audience encircled by unfathomable suffering. It took no mercy on his Christians.

“The pain in the eyes, the attack of the fevers, and the ailment of all the limbs are the same among us and among the others, so long as we share the common flesh of this age.” Cyprian tried to ennoble the victims of the disease, likening their strength in pain and death to the heroic intransigence of the martyrs. Cyprian conjured the symptoms for his hearers.

These are adduced as proof of faith: that, as the strength of the body is dissolved, the bowels dissipate in a flow; that a fire that begins in the inmost depths burns up into wounds in the throat; that the intestines are shaken with continuous vomiting; that the eyes are set on fire from the force of the blood; that the infection of the deadly putrefaction cuts off the feet or other extremities of some; and that as weakness prevails through the failures and losses of the bodies, the gait is crippled or the hearing is blocked or the vision is blinded.

Cyprian’s account is central to our understanding of the disease. The pathology included fatigue, bloody stool, fever, esophageal lesions, vomiting, conjunctival hemorrhaging, and severe infection in the extremities; debilitation, loss of hearing, and blindness followed in the aftermath. We can complement this record with more isolated and frankly uncertain hints from other witnesses. According to Cyprian’s biographer, the disease was characterized by acute onset: “carrying off day by day with abrupt attack numberless people, every one from his own house.”

The course of the infection and illness was terrifying. This impression is confirmed by another North African eyewitness, a Christian not far removed from the circle of Cyprian, who insisted on the sheer unfamiliarity of the disease. “Do we not see the rites of death every day? Are we not witnessing strange forms of dying? Do we not behold disasters from some previously unknown kind of plague brought on by furious and prolonged diseases? And the massacre of wasted cities?” The pestilence, he argued, was a manifest encouragement to martyrdom, since those who died the glorious death were spared the “common fate of others amidst the bloody destruction of ravaging diseases.”

The Plague of Cyprian was not just another turn through the periodic cycle of epidemic mortality. It was something qualitatively new—and the evocation of its “bloody” destruction may not be empty rhetoric, if hemorrhagic symptoms are implied.

The disease was of exotic origin and moved from southeast to northwest. It spread, over the course of two or three years, from Alexandria to other major coastal centers. The pandemic struck far and wide, in settlements large and small, deep into the interior of empire. It seemed “unusually relentless.” It reversed the ordinary seasonality of death in the Roman Empire, starting in the autumn and abating in the following summer. The pestilence was indiscriminate; it struck regardless of age, sex, or station. The disease invaded “every house.”

One account predictably blamed the “corrupted air” that spread over the empire. But another chronicle tradition, going back to a good contemporary historian in Athens, recorded that the “disease was transmitted through the clothes or simply by sight.” The observation is notable; in a culture without even a rudimentary sense of germs, the comment betrays a pretheoretical sense of contagion. The concern that the disease could be transmitted by clothing or eyesight suggests at least a dim awareness of an infectious origin. And it just might provide a further hint that the disease affected the eyes.

The ancients harbored plenty of eccentric notions about the powers of eyesight, among them that it was tactile, ejecting a flow of particulates from the eye of the looker. The bloody eyes of Cyprian’s victims may have presented a terrifying visage, in a culture where the eyes had the power to reach out and touch.

The death toll was grim. We have an intriguingly specific report from the bishop of Alexandria, who claimed that:

This immense city no longer contains as big a number of inhabitants, from infant children to those of extreme age, as it used to support of those described as hale old men. As for those from 40 to 70, they were then so much more numerous that their total is not reached now, though we have counted and registered as entitled to the public food ration all from 14 to 80; and those who look the youngest are now reckoned as equal in age to the oldest men of our earlier generation.

The reckoning implies that the city’s population had declined by about 62 percent (from something like 500,000 to 190,000). Not all of these need be dead of plague. Some may have fled in the chaos. And we can always suspect overheated rhetoric. But the number of citizens on the public grain dole is a tantalizingly credible detail, and all other witnesses agreed on the scale of the mortality. An Athenian historian claimed that 5,000 died each day. Witness after witness—dramatically if imprecisely—testified that depopulation was invariably the sequel of the pestilence. “The human race is wasted by the desolation of pestilence.”

These haphazard clues do not equip us well to identify the pathogenic agent of the Plague of Cyprian. But the range of suspects capable of causing a disease event of this scope is not large, and some possible agents can be almost certainly exculpated.

Bubonic plague does not fit the pathology, seasonality, or population-level dynamics. Cholera, typhus, and measles are remote possibilities, but each poses insuperable problems. Smallpox must be a serious candidate. The two-generation lapse between the episode under Commodus and the Plague of Cyprian means that effectively the entire population would have been susceptible again. The hemorrhagic form of the disease might also account for some of the features described by Cyprian.

But in all the case for smallpox is weak. A North African author claimed it was an unprecedented disease (though whether he would have had any memory of previous smallpox epidemics is of course questionable). None of our sources describe the full-body rash that is the distinctive feature of smallpox. In the church history of Eusebius, written in the early fourth century, an outbreak more like smallpox was recounted in AD 312–13. Eusebius both called this a “different illness” than the Plague of Cyprian and also distinctly described the pustular rash. The exotic origins of the third-century event, again from beyond the Roman Empire, do not suggest the eruption of a now-endemic pathogen. Finally, the putrescent limbs and permanent debilitation of the Plague of Cyprian are not a fit for smallpox. None of these clues are conclusive, but collectively they militate against the identification of smallpox.

Any identification must be highly speculative. We would offer two candidates for consideration. The first is pandemic influenza. The influenza virus has been responsible for some of the worst pandemics in human history, including the “Spanish flu” epidemic that carried off some 50 million souls at the end of World War I. The lack of clear evidence for influenza from the ancient world is puzzling, because the flu is old and it was undoubtedly not a stranger in the ancient world. Influenza is a highly contagious acute respiratory disease that comes in many forms. Most types are relatively mild, causing familiar cold-like symptoms. Other rare types of influenza are more menacing.

Zoonotic forms of the disease, especially those native in wild aquatic birds, can be pathogenic to other animals, including pigs, domestic fowl, and humans; when these strains evolve the capacity to spread directly between humans, the results are catastrophic. There have been four global outbreaks in the last century, and avian influenza (which includes some dreaded strains such as H5N1) remains a terrifying threat today.

Pathogenic zoonotic influenzas are viciously lethal. They induce an overheated immune response which is as dangerous as the viral pneumonia itself; hence, the young and healthy are paradoxically put at risk by the vigor of their immune response. The lack of any respiratory symptoms in the account of the Plague of Cyprian is a strike against the identification. But it is worth reading some observations of the 1918 pandemic.

Blood poured from noses, ears, eye sockets; some victims lay in agony; delirium took away others while living ... The mucosal membranes in the nose, pharynx, and throat became inflamed. The conjunctiva, the delicate membrane that lines the eyelids, becomes inflamed. Victims suffer headache, body aches, fever, often complete exhaustion, cough ... Often pain, terrific pain ... Cyanosis ... Then there was blood, blood pouring from the body. To see blood trickle, and in some cases spurt, from someone’s nose, mouth, even from the ears or around the eyes, had to terrify ... From 5 to 15 percent of all men hospitalized suffered from epistaxis—bleeding from the nose.

Pandemic influenza might indeed account for the horrifying experience of the Plague of Cyprian.

The winter seasonality of the Plague of Cyprian points to a germ that thrived on close interpersonal contact and direct transmission. The position of the Roman Empire astride some of the major flyways of migratory birds, and the intense cultivation of pigs and domestic fowl such as chickens and ducks, put the Romans at risk. Climate perturbations can subtly redirect the migratory routes of wild waterfowl, and the strong oscillations of the AD 240s could well have provided the environmental nudge for an unfamiliar zoonotic pathogen to find its way into new territory. The flu is a possible agent of the pestilence.

A second and more probable identification of the Plague of Cyprian is a viral hemorrhagic fever. The pestilence manifested itself as an acute-onset disease with burning fever and severe gastrointestinal disorder, and its symptoms included conjunctival bleeding, bloody stool, esophageal lesions, and tissue death in the extremities. These signs fit the course of an infection caused by a virus that induces a fulminant hemorrhagic fever.

Viral hemorrhagic fevers are zoonotic diseases caused by various families of RNA viruses. Flaviviruses cause diseases like yellow fever and dengue fever, which have some resemblance to the symptoms described by Cyprian. But flaviviruses are spread by mosquitoes, and the geographic reach, speed of diffusion, and winter seasonality of the Plague of Cyprian rule out a mosquito-borne virus.

The speed of diffusion points to direct human-to-human transmission. The belief that caring for the sick and handling the dead were fraught with danger underscores the possibility of a contagion spread between humans. Only one family of hemorrhagic viruses seems to provide a best match for both the pathology and epidemiology of the Plague of Cyprian: filoviruses, whose most notorious representative is the Ebola virus.

Filoviruses are millions of years old. Fragments of their genetic material are anciently embedded in mammalian genomes, and for millions of years they have infected bats, insectivores, and rodents. Yet filoviruses, like Ebola virus and Marburg virus, were only recognized in the second half of the 20th century during a series of small-scale outbreaks. The Ebola epidemic of 2014 brought further attention to the family. The natural host of the Ebola virus remains unconfirmed, although bats are suspected. Ebola virus grabs public attention because of its ghastly clinical course and extreme case-fatality rates.

To cause an epidemic, the Ebola virus must first leap from its host species to a human; this probably occurs when humans come into contact with infected bats or apes. Once infected, after a brief incubation period (on average four to 10 days, sometimes longer), victims suffer intense fever and a disease that breaks down multiple systems simultaneously, including gastrointestinal and vascular involvement. Conjunctival injection and severe hemorrhagic symptoms could well account for the disturbing reports of Cyprian. Tissue necrosis and permanent disfigurement of the limbs might reflect Cyprian’s description of extremities turning putrid and becoming irreversibly disabled.

Case-fatality rates, even with modern treatment, are grotesquely high: 50–70 percent. Death usually comes between days six and 16; survivors are thought to possess immunity. The Ebola virus is transmitted by bodily fluids, but not aerial droplets; it spreads easily within households. Caregivers are at special risk, and cadavers remain a potent source of infection. The observance of traditional burial rites has been a problematic risk factor even in recent outbreaks.

Retrospective diagnosis from anguished reports of nonmedical personnel across nearly 2,000 years is never going to offer great confidence. But the hemorrhagic symptoms, the shocked sensibilities, and the insistence on the novelty of the disease all fit a filovirus. An agent like Ebola virus could diffuse as quickly as the Plague of Cyprian, but because of its reliance on body fluids for transmission, it could exhibit the slow-burning, “unusually relentless” dynamics that so struck contemporary observers. The obsession with deadly corpses in the third-century pandemic strikes a profound chord, given the recent experience of the Ebola virus. The uncertainty lies in our profound ignorance about the deep history of pathogens like Ebola that never became endemic in human populations.

As historians, we understandably default to the familiar suspects. But our broadening awareness of the incessant force of emerging disease, at the frontier between human society and wild nature, suggests a place for significant disease events in the past, like the Plague of Cyprian, caused by zoonotic diseases that wreaked havoc and then retreated back to their animal hosts.

By the time of the Plague of Cyprian’s appearance in AD 249 there was much that was different. The empire’s stores of reserve energy were depleted. Perhaps this microbial enemy was just more sinister. In this event, the center could not hold. There is much that must remain uncertain about the Plague of Cyprian, but not this: In its immediate wake, anarchy was loosed on the world.

This post is adapted from Harper’s recent book, The Fate of Rome: Climate, Disease, and the End of an Empire.