Malaria and Rome: A History of Malaria in Ancient Italy

book is simply that in order to understand the demography of the ancient Roman world, it is necessary to pay much more attention to medical history than has been done in the bulk of research carried out so far into Roman demography.
Karl Julius Beloch, the pioneer of Roman demography, did not confine himself to antiquity. He intensively investigated Italian population history in more recent periods to provide a comparative foundation for his analysis of the scanty evidence from antiquity.
A large proportion of the output of Roman historians over the last ¹ Hopkins (1966: 246).
² Hopkins (1966: 264).
2

Introduction
thirty years or so has failed to follow his example. Yet there is a wealth of evidence there, waiting to be exploited, which provides food for thought. The starting-point of the argument here is an extremely interesting recent discussion, by two leading Italian historical demographers, del Panta and Rettaroli, of an earlier study by one of them (del Panta) of the demography of Grosseto in the last century.³ They observed that the age-structure of the population of Grosseto in the 1840s does not match closely any of the Coale and Demeny life-tables, the set of model life-tables for the age-structures of human populations most widely employed by demographers, which are so popular with some ancient historians.
Nor did the age-structure of the population of Grosseto match any of the other available sets of model life-tables based on data for modern European populations. However, the demography of Grosseto in the last century does resemble something else that is completely different. What it resembles more than anything else, as they pointed out, is the demography of certain parts of Africa today where malaria is endemic.⁴ Of course the demographic patterns are not, and never could have been, precisely identical, because of numerous profound differences in the environment between central Italy and tropical Africa.
Nevertheless the degree of similarity in the demographic patterns between Grosseto and tropical Africa is striking. Malaria has an awesome power as a determinant of demographic patterns. It is estimated to have caused about a million deaths in Africa in 1995, and it also has far-reaching indirect effects, as will be seen later.⁵ In addition, it had the ability in the past to overcome the ecological barriers between tropical Africa and the Mediterranean world and so bring its power as a determinant of human demographic patterns to the ancient Mediterranean world, but this has been ignored by most Roman historians interested in ancient demography.⁶
³ Del Panta and Rettaroli (1994: 201–3); del Panta (1989).
⁴ Root (1999) discussed the relationship between differences in the intensity of malaria transmission and differences in spatial mortality patterns in various parts of Africa.
⁵ Snow et al. (1999).
⁶ A few historians have noted malaria as a possible contributing factor to mortality in antiquity. For example, J. C. Russell (1985: 88–9, 228, 234) rightly associated malaria with extremely high adult mortality, but regarded tuberculosis as more important. His suggestion that the ‘Antonine plague’ in the second century  was P. falciparum malaria is exceedingly improbable (it was almost certainly smallpox). Similarly the suggestion (Karlen (1996: 69–70); Cartwright and Biddiss (2000: 9–10) ), that the epidemic which struck Rome (Cassius Dio 66.23.5; Suetonius, Divus Titus 8.3–4) after the eruption of Vesuvius in  79 was the first


Introduction
3
The crux of the matter, as a problem in demography, is the relationship between infant mortality and adult mortality. This balance is affected by the differing effects of malaria on different age groups.
These effects of malaria on different age groups depend, in turn, on its transmission rate. In general, European populations in the past where malaria was endemic had higher levels of adult mortality relative to infant and child mortality than otherwise similar populations unaffected by malaria, as well as much higher total mortality.
The overall effect of malaria in terms of increasing mortality has indeed long been known, at least in principle (and, as will be seen later, the essence of the matter was also known in antiquity). Angelo Celli, one of the pioneers of the important Italian school of malariology, wrote at the end of the nineteenth century as follows: ‘the average life of the worker in malarious places is shorter, and the infant mortality higher, than in healthy places’.⁷ However, at that time modern historical demography based on such techniques as family reconstitution had not yet developed. It is only more recently that it has become possible to appreciate in quantitative terms the sheer scale of the phenomenon. Del Panta and Rettaroli compared the demography of Grosseto in the last century with the demography of Treppio, an Appennine community located northeast of Pistoia which was not affected by malaria. The population of Treppio had a life expectancy at birth of 37, but the population of Grosseto had a life expectancy at birth of only 20, with corresponding substantial differences in the age distribution of mortality in the two populations. The effect of the presence of malaria was to nearly halve life expectancy at birth (see Ch. 5. 4 below for further discussion).
Given such enormous effects, it is not surprising that Italian historians, scientists, and politicians devoted a great deal of attention to malaria so long as it continued to be a major public health problem in Italy. As Bonelli put it, ‘malaria was for centuries . . .
appearance of P. falciparum malaria in Rome, has nothing to recommend it, even if the emperor Vespasian himself did die from malaria in that year at Aquae Cutiliae in the Sabine region (Cassius Dio 66.17.1: nos&saß . . . pureto∏ß (he became ill with fevers); Suetonius, Divus Ves-pasianus 24.1). Scobie (1986: 422) mentioned malaria alongside gastro-intestinal diseases; Morley (1996: 43) also mentioned malaria; Salmon (1992) is rather more detailed. Nevertheless its importance has in the main not been appreciated by historians interested in ancient demography. A. Wear (1995: 226), writing about the early modern period, asserted that ‘the falciparum form of malaria did not exist in Europe’, an