Materials and Methods Ancient: The Accounts of Dionysius and Livy
Firstly, from Dionysius of Halicarnassus [
2]: So, it [the disease] meandered among shepherds and settlers gradually throughout the region, ultimately invading Rome as well. It is not easy to recount how many servants, how many mercenaries, and how many of the indigent class perished. At first, their corpses were carried in heaps on chariots: but then those of the less respectable were thrown into the current of the [Tiber] river. Counting, a fourth of the senators perished, and with them two consuls, and most of the tribunes. That disease began around the early part of September. (According to the dating in use, the month of the beginning of the epidemic, indicated by Livy (see below), is August. And continued for a year backward, investing and consuming every month and age… [
3,
4].
Livy reports the same events, in the third book of his
Histories, as recorded in the following passage
[3,4,5,6,7,8,9,10,11,12,13]:
It was the sickly season and chanced to be a year of pestilence both in the city and in the country, for beasts as well as men; and the people increased the virulence of the disease, in their dread of pillage, by receiving flocks and country-folk into the city. This conflux of all kinds of living things distressed the citizens with its strange smells, while the country people, being packed into narrow quarters, suffered greatly from the heat and want of sleep; and the exchange of ministrations and mere contact spread the infection. The Romans could scarce endure the calamities which pressed hard upon them, when suddenly envoys from the Hernici [a people allied to Rome] appeared, announcing that the Aequi and the Volsci had joined forces and established a camp in their territory, from which base they were devastating their land with an enormous army. Not only did the reduced numbers of the senate {Dionysius also notes stricken senators too weak or lethargic to stand attempting to carry on the business of the Senate while “barely alive on litters”[
4]} show their allies that the [Roman] nation was prostrated by the pestilence, but they also returned a melancholy answer to their suit, that the Hernici, namely, with the help of the[ir allies the…] Latins must defend their own possessions; for the City of Rome, in a sudden visitation of divine displeasure, was being ravaged by disease. If there should come any respite from their suffering, they [the Romans] would help their friends, as they had done the year before and on every other occasion ...From that time onward, little by little, both because of the peace obtained from the gods and the gradual exhaustion of the unhealthy season, the bodies in which the course of the disease had run its course began to return to health, while minds turned to the problems of the State.”
The accounts themselves along with a supporting wealth of consilient interdisciplinary analyses strongly implicate
P. falciparum in the 463 BCE outbreak. The disease clearly was part of, and in many ways fit well within, an established annualized pattern that was characteristic, then and later, of the late summer/early fall malarial “sickly” or “fever season”. This was something that would have been well understood by both Livy and Dionysius centuries later. References to this phenomenon indeed constitute a common signifying trope that formed part of the broader cultural and literary heritage of the ancient Greco-Roman Mediterranean world with examples ranging from Homer’s
Iliad to Roman poetic traditions. [
6,
7]. The more prosaic words of the 1st Century CE Roman physician Celsus in his work
De Medicina “lethal autumn” also reflect specific awareness of this disease phenomenon within Greco-Roman medicine [
8].
Such annualized outbreaks of sickness in Rome had, however, centered, up to 463 BCE, around malarial fevers caused by the less deadly plasmodium parasites
P. vivax and
P. malariae [
8]. Yet, the epidemic event of 463 BCE was different than the norm; both ancient accounts document large numbers of deaths not confined to children or the elderly but also including large numbers of presumably healthy adults. The high death rates indicate an at least partially epidemiologically naive population and a more virulent form of the disease. As noted,
P. falciparum statistically causes 90% of human malaria fatalities [
1] and fits the retrospective diagnostic parameters for 463 BCE, in a multifaceted sense, extremely precisely. Celsus (as well as other Greco-Roman physicians) accurately categorized malarial fever types (which we now know are caused by the plasmodium parasites
P.
malariae,
P.vivax and
P.falciparum) with a sound understanding of comparative pathologies, enabling environments and seasonality–peak August-October [
9], pp. 55-62 and p. 71 if
not of underlying (including vector/secondary host epidemiology) causation. 463 BCE thus seems to have marked a true watershed moment for ancient Rome in terms of increased disease impact from malaria.
While recitations of the elements of pathology in the accounts regarding the 463 BCE outbreak are limited, descriptions indicative of fatigue/lethargy or weakness (noted above) as a primary symptom and protracted illnesses from which numerous sufferers eventually recovered as well as the disease’s seasonality are also strongly consistent with the
P. falciparum pathology and epidemiology [
1]. The described apparent rural to urban epidemiological progression of the outbreak is also consistent with malaria and, notably, inconsistent with other candidate pathogens such as tuberculosis, dysentery or other gastro-intestinal bacterial disorders. Additionally, the outbreak occurred during a time of Roman contact and conflict with the Volsci, whose coastal territories (then and later after their eventual incorporation into the expanding Roman state) constituted one of ancient Italy’s primary malaria outbreak hotspots []9, p.179. The described unusual influx of animals and their herders (as a symptom of the prevailing (A) political, economic and security/military contexts that were characterized by an overall high level of disorder, uncertainty and insecurity, into the city) likely also greatly augmented and concentrated mosquito– including
Anopheles vector/secondary host
- presence in the city of Rome. This may well have been a key epidemiologically enabling factor that increased the scale of the outbreak and overcame the formerly countervailing epidemic malarial stochasticity that was linked to certain limitations on
Anopheles presence [
8]. A 2018 epidemiological study conducted in Indonesia, in fact, strongly “highlights the importance of livestock [presence] for amplified “malaria risk [by a factor of 2.8] rather than prophylaxis” in such proximity settings [
10]. Another holistically exacerbating factor for the epidemic is that Romans would have been naive to
P. falciparum pathology. Unlike Africans, they would not have long before developed variants such as the sickle cell trait “which is common in African populations and protects against fatal
P. falciparum malaria.” [
11,
12].
References in both ancient accounts to a preceding disease outbreak that devastated cattle and other animals in the 5th Century BCE also lend at least indirect (if unfortunately, also somewhat circular) support to the diagnosis of malaria through indirect possible implication of stagnant pools of contaminated water in that preceding outbreak. The now eradicated RNA single strand
paramyxoviridae rinderpest virus is clearly a possible candidate pathogen in that instance, though the evidence in this regard is somewhat thinner and less compelling. Northern rinderpest strains had an extraordinary morbidity rate of nearly 100% [
13], which is, however, consistent with both ancient accounts [
2,
5]. Rinderpest was “very fragile” in terms of aerosolization but that was not its primary path of transmission [
14]. Stagnant pools (which could also have also been fertile environments for
Anopheles mosquitoes that are the secondary host/primary vector of malaria) of contaminated water were. Ancient writers such as Livy seem to have erroneously conflated the epidemics [
5]; modern science has additionally found no evidence of zoonotic transmission of rinderpest to humans [
14]. and WGS analysis indicates a later to much later genetic (and zoonotic) emergence of measles from rinderpest; with these genetically linked viruses, however, continuing to co-exist in a temporal sense [
15].
According to a relaxed clock Bayesian phylogenetics study with an HPD confidence level of 95%, the eleventh and twelfth centuries are the likely temporal window for measles emergence from rinderpest– the highly relevant early edge of that temporal window would be between 437 and 678 CE [
16]. This would be far too late to be the cause of the 5th Century BCE epidemic or even of the much later (2nd-3rd Centuries CE) Antonine or Cyprian pandemics (see below). The described seasonality of the 463 BCE ancient outbreak also does not match that of measles [
17], and therefore that alternative can be firmly ruled out. Though rinderpest itself was likely already present in the 5th Century BCE in ancient Italy. Rinderpest was certainly established in the Roman Empire sometime prior to the 4th Century CE. The detailed description contained in
De Mortibus Boum (alternatively
Carmen bucolicum de virtute signi crucis domini) written by the 4th century CE Roman poet Severus Sanctus Endeleichus strongly indicates regular and long-standing outbreaks of epizootic (and endemic) rinderpest disease in the European territories of the ancient Roman Empire [
13,
14].
More broadly, our proposed timeline and implicated pathogen, also (while not rising to the level of scientific conformation) fits seamlessly in terms of more holistically considered retrospective-diagnosis or deductive analysis, like a precisely shaped missing piece of a puzzle, into what we already know or can surmise about malaria and infectious disease history more broadly in ancient Italy. As to the far less likely possibility of more exotic/less common/unlikely pathogens that also
nonetheless fit the annualized seasonal pattern, we reference the famous phrase of Dr. Theodore Woodward traditionally used to train medical students, “When you hear hoofbeats behind you, don’t expect to see a zebra” [
18]. Dates for the establishment and spread of malaria as an annualized/endemic (peak late summer early Fall seasonality) ancient disease in Europe do, however, vary somewhat in the established scholarship. Such timelines have been particularly disputed, and we hope that our conclusions here help clarify this question, for
P. falciparum; though, it seems to have been well on the way to becoming endemic in nearby northern Greece by 500 BCE. General agreement also exists that
P. vivax arrived earlier in Italy than
P. falciparum [
8]. Malaria outbreaks occurred from the 6th Century BCE onwards with the effects of endemic disease eventually becoming “considerable” in terms of killing or debilitating people [to the extent that malaria] altered the age structure of human populations” in the ancient Mediterranean World [
8]. As noted above, we believe that our evidence, when considered holistically and consiliently, demonstrates that the epidemic of 463 BCE marked a key moment in this progression.
Based on a highly detailed description of the pathology of semitertian fevers [
19] from Celsus in the early 1st Century CE, one can safely conclude that the more pathogenic form of the disease
P. falciparum had long been endemic in Italy before then [
8]. Less detailed but telling references to annually be escaping, deadly “bad air”, an allusion to the Greco-Roman humoral medical concept of disease-causing miasmas, in the late summer/early fall around the town of Tarquinia (with nearby apparently
Anopheles-ridden malarial coastal marshes) indeed date as far back as the 2nd and even 3rd Centuries BCE. These are again indicative of the evolutionary lineage of
P. falciparum causing endemic disease [
8] by that, even earlier, period. As noted, Roman
(B) culture, more broadly, reflected and shaped perceptions and at times outcomes (through inducing changes in behavior
) the power of the annual late summer/early fall pestilential season’s fear and menace. The First Century BCE famed poet Horace, in his
Odes and
Epistles, uses the phrase “fever season” not only to describe specific conditions in the late summer-early fall but also to represent periods of difficulty, illness, or even historical events that have caused suffering to humanity – a clear signifier of the trope’s cultural recognition, power and relevance for ancient Romans [
20].
Overall, one can conclude that the process of malaria becoming an endemic disease throughout non-alpine Italy was steady and significant but also slow and extended (hindered initially by the lack of consistent/sufficient presence of
Anopheles mosquitoes) and took place between 700 and 100 BCE [
8]. In fact, malaria reached the Italian north much later– only by around 1000 CE [
19]; though the colder and drier prevailing climate there likely also inhibited the spread of the disease. The Pontine Marshes near Rome itself, however, were a suitable spot for the disease to become strongly established. By the First Century BCE, Julius Caesar was making the first of many historical attempts to drain the Pontine Marshes [
9], pp.192-201. Horace, like many upper-class Romans, fled Rome during these unpleasant and dangerous months to avoid the muggy climate and the malarial disease that often spread through the city. This was common behavior among the well-to-do, who preferred to take refuge in villas and summer residences in the countryside in drier and cooler locations [
20]. Besides the evidence from a multitude of literary sources, ancient Roman
P. falciparum malaria infections and deaths, from corpses found at the Umbrian so-called “Infant Cemetery” from the 5th Century CE (occurring against the later backdrop of the Western Roman Empire’s collapse) at the archaeological site of Lugnano, have been scientifically confirmed through PCR/DNA analysis [
8]. That same 5th Century CE outbreak also highlighted the differences in terms of epidemiology and pathology for a naive population (in this case Attila’s invading Huns whom the outbreak affected so severely as to likely drive them from Italy) and one that had been living with
P. falciparum for nearly a millennium. The victims at Lugnano were (typically) children [
21], p. 97. Lacking basic germ theory and modern blood tests to confirm and identify parasite species [
22], a wide range of antimalarial drugs, ACT therapies used by modern medicine to treat uncomplicated forms of P.falciparum infection or the parenteral antimalarial therapies used to combat more severe cases [
1], surviving adults at Lugnano desperately turned to magical remedies involving ravens’ talons, “stones in the mouths of the small children’s corpses” and pots of ash [
21], p.97 The intensification of malaria with the onset of Italian
P.falciparum infections that seems to have begun in 463 BCE indeed had significant and long term civilizational-level consequences for ancient Rome.
More broadly,
(C) changing climatic conditions clearly also played a major role in the expanding epidemiological footprint of a variety of types of malaria in ancient Italy. A combination of written records and a range of consilient scientific data makes deducing relevant climate trends generally reliable from the 5th Century BCE onward. This was the early part of the era often characterized by historians as the Roman Warm Period (RWP), alternatively known as the Roman Climate Optimum (RCO). There are strong signs of warmer and wetter climate trends from around the 6th Century BCE onwards [
23]. The Tiber River, for example, seems to have changed course by the 6th Century BCE due to, amongst other factors, steadily warmer temperatures. Wetter conditions also prevailed more generally, in a partially overlapping manner, between 800 to 400 BCE across the non-alpine regions of Italy [
23]. The overall picture though is complex in more than one sense, often regionally varied and at times strongly interrupted. In 426 BCE, for example, an unknown stratospheric volcanic eruption took place that led to sharply cooler temperatures for the following three years. More consistent “warmer, wetter or more stable” climate conditions began to prevail around 200 BCE [
23]. Malaria’s arrival and expanding presence in ancient Italy was also enabled and exacerbated by additional
(D) anthropogenic environmental factors: Increasing deforestation, interrelated practices of agriculture, increasing urbanization and other forms of human activity. The water table also rose in Italy by a meter between the 6th and 4th Centuries BCE [
9], pp. 101-105. Over time, under these changed climate and environmental realities, malaria became more widespread and significant as an endemic disease in the emerging Roman Empire. Indeed, malaria (in all its three present types-
P. malariae, P.vivax and P.falciparum), tuberculosis and many forms of dysentery were the most prominent diseases in ancient Rome. Chickenpox, diphtheria, mumps, and whooping cough also occurred in childhood, with less frequent attacks on adults [
24].
Yet, it was malaria, augmented by the more deadly
P. falciparum legacy that likely began in 463 BCE, that was the most consistently significant infectious disease threat for ancient Rome. Robert Sallares in
Malaria and Rome:
A History of Malaria in Ancient Italy (Oxford, 2002) noted that ancient Rome’s holistic interaction with malaria (particularly
P. falciparum) shows that “Malaria has an awesome power as a determinant of demographic patterns” [9, p.2]. These include population health, structure and life expectancy, and in Rome’s case, malaria (again, particularly
P. falciparum) was also a contributor to an extremely high rate of infant mortality. Especially if we agree with the complex systems theory-based analysis of the Club of Rome’s Ugo Bardi that the Roman Empire of the 1st through 3rd Centuries was “larger, better organized and better managed than anything that had existed before” [
25],p.13, malaria and its destructive demographic consequences (particularly those caused by
P. falciparum) [
9], p.2 constituted an annually-present key Bardian ‘damping” [
25], p.2 feedback and, largely unaddressed flaw in the holistic complex functioning of empire and a continuing quiet but damaging blow to Rome’s civilizational resilience. The event of 463 BCE thus seems to have functioned as a key inflection point after which the threat to Roman civilization from malaria was afterwards amplified. Even more fundamentally, malarial disease impacts human populations at a deep biological level; “variants in the human genome that are associated with resistance to
Plasmodium infection disease are estimated to be thousands of years old” in Africa [
26], pp.283-304.
During the late 19th Century, however, the areas around Rome contributed to the foundational modern medical understanding of malaria (through the pioneering work of 19th-20th Century French physician Alphonse Laveran) and saw efforts in the use of engineering to eradicate malaria– through the successful draining of the Pontine Marshes [
9], pp. 14-16. The influx of a new and deadlier form of malaria in 463 BCE also occurred against the developing
(E) perceptions, practices, knowledge and capabilities of the prevailing medical system and (F) the existing holistic underlying state of the health of the affected population. Largely urban-based ancient Greco-Roman physicians associated the epidemiology/presence of malaria generally with marshes but not specifically with mosquito bites. Though, contrastingly, rurally based Roman agricultural experts such as Varro (2nd Century BCE) and Columella (1st Century CE) do seem to have been at least vaguely aware of the mosquito vector for the various malarial fevers [
19]. The likely reason for this oversight by the physicians, along with the prevailing and distorting lens of Greco-Roman medicine’s humoral theory (e.g., the famed second century CE physician Galen believed that severe quotidian fever was caused by an excess of phlegm [
27]), is that the
Anopheles secondary host/vector is complex and highly species dependent. Many of the relevant species are indistinguishable without the modern technology of a microscope [
9], p.26. Thus, the prevalence of situations characterized by “lots of mosquitos but no malaria” likely explains why “ancient Greek and Roman physicians failed to notice the connection between the periodic intermittent fevers of malaria and mosquito bites” [
9], p.45. The dangers of
P. falciparum, a modern transdisciplinary combination of archaeology and medicine informs us, were also likely exacerbated “by [underlying] moderate degrees of malnutrition” amongst the Roman lower classes during both the Republican and Imperial periods [
9], p.146.
Greco-Roman medicine also lacked a consistently applied or effective remedy for severe malaria that extended beyond basic palliative care. The resulting intermixing of mystical elements within (noticeable in the accounts of the 463 BCE outbreak) Greco-Roman medicine– conceptualization of malaria as a demon, magic, cults of
Dea Febris (Goddess of Fevers) and amulets also didn’t help matters. Regarding pathology, Greco-Roman physicians such as Celsus and Galen were on somewhat firmer ground. They correctly classified, for example, the intermittent (at times recurring over months or even years and associated with gastro-intestinal disorders and miscarriages) but generally non-fatal milder “quartan” fevers (those which we know today are caused by
P. malariae) as different from the more dangerous semitertian or “quotidian” fevers associated, we know now, with
P. falciparum. [
9], pp.125-135. Malaria seems to have depressed Roman life expectancy overall in a way at least somewhat, if not exactly, analogous to
Yersinia pestis infection during the later Middle Ages [
9], pp.272-273. The physician Asclepiades of Bithynia described more severe forms associated with
P. falciparum (semitertian or quotidian) and
P. vivax (tertian) as “common” in ancient Italy by the 3rd to 2nd Centuries BCE [
9], pp.219-220. One of the main symptoms of the more severe fevers was lethargy. The brilliant Galen, in the 2nd century CE, also inferred the concept of acquired immunity, noting the vulnerability of children and newcomers to these fevers [
9], pp.219-220. He thus indirectly anticipated the modern medical conclusion that, “In low transmission areas, all ages are at risk due to low immunity” [
1]. The Romans also had enough engineering capability and understanding of the importance of the threat from malaria to build, at the empire’s height in the 2nd Century CE, mitigating drainage systems in the Campagna around Rome that helped limit the threat of malaria to the city itself for centuries. These complex mitigation/control systems, however, collapsed completely along with the Western Roman Empire during the 5th Century CE [
19].
However, “during the 20th Century, malaria was eradicated from many temperate areas including the whole of Europe” [
28]. Sophisticated detection and eradication methods have proved effective, and in today’s Europe, malaria is largely a disease of travelers. Though some autochthonous cases of human-to-human transmission do occur [
28]. While, as in ancient times, climate change (now more exclusively anthropogenic as opposed to naturally cyclic
and anthropogenic) may mean that vector-borne malaria may once again become endemic to southern Europe [
29], modern medical understanding and continued diligent use of a wide variety of eradication and control measures are likely to mitigate against renewed expansion of non-zoonotic malaria in the southern European lands that it once haunted in ancient times. Such expansion of mosquito vector-borne disease threats in Europe during an era of anthropogenic climate change and environmental disturbance is also, however, not limited to
Anopheles-vector malaria. Invasive
aedes albopictus mosquitoes, linked to the spread of tropical diseases such as dengue fever, chikungunya and the Zika virus have now been found in no less than 18 European countries.
Aedes aegypti mosquitoes, linked to the spread of yellow fever, have recently become established in Cyprus [
30].