was about three times the mortality actually attributed to malaria on the basis of death certification’.¹⁹ For Grosseto, for example, where 23% of all deaths were directly attributed to malaria in 1882, that would imply that well over half of all deaths were in fact linked to malaria, directly or indirectly. Morbidity data, which were collected alongside the cause-of-death data for Grosseto in 1840–41, indicate that about 60% of all recorded illnesses in Grosseto were intermittent fevers.²⁰ This suggests that malaria might have been more frequent in Grosseto than all other diseases put together.²¹ In such contexts, it is easy to understand the equation of ‘fever’, puretÎß or febris, with malaria as the disease par excellence, as seen for example in the de medicina of Celsus. Malaria was the first disease discussed by Celsus in book 3 of his work, when he considered diseases one by one. He explicitly stated that malarial fevers were extremely common, for the benefit of any modern historians who find it hard to believe:
The next subject is the treatment of fevers, which both affect the entire body and are extremely common. One type of fever is quotidian, the second tertian, and the third quartan.²²
The data from Grosseto demonstrate the scale of the effect which malaria might be expected to have had on the population of at least some parts of the city of Rome in Galen’s time, when it was endemic (see Ch. 8 below). Since the indirect effects of malaria ¹⁸ Newman (1965: 76–9, 158–60). See most recently Jones (2000).
¹⁹ L. Molineaux in Wernsdorfer and McGregor (1988: ii. 976).
²⁰ Del Panta (1989: 48–9 n. 23). He stated that the effects of malaria were largely indirect as far as adult mortality in Grosseto was concerned. C. Fermi estimated that malaria was implicated in over 50% of all deaths in Sardinia as recently as 1900 (Brown (1986) ).
²¹ Similarly Desowitz (1997: 195) cited a report from the United States Public Health Service in 1919 which concluded that in the southern states of the USA malaria was more important than tuberculosis, typhoid fever, dysentery, and pellagra put together.
²² Celsus 3.3.1: sequitur vero curatio febrium, quod et in toto corpore et vulgare maxime morbi genus est.
Ex his una cotidiana, altera tertiana, altera quartana est.
122
Demography of malaria
are so important, indirect methods for estimating its effects on mortality, such as comparisons of overall mortality levels (most simply by crude death rates) or of the age-structures of populations (as between Grosseto and Treppio) are in fact a better guide than direct methods, such as the Italian national statistics for causes of death used by Shaw. Indirect methods are the chosen methods of professional malariologists.²³
In view of the quantitative importance of malaria in some areas, as shown by the data from Grosseto, it is not surprising that the popular explanations for malaria were sometimes transferred to other diseases, for want of anything better, but this only serves to demonstrate the all-pervasive influence of malaria in those areas where it occurred. Undoubtedly this also happened in antiquity, but the best example is the great epidemic of syphilis which swept across Europe in the years immediately following the return of Columbus from the New World. Leoniceno, in a booklet produced as part of the dispute at the court of Ferrara in 1497 concerning the nature of the epidemic, carefully discussed the types of diseases mentioned by classical authors and correctly concluded that the (for some) new disease was quite distinct from the elephantiasis or leprosy described in antiquity. Nevertheless he was unable to resist explaining it in Hippocratic terms and associating it with the massive Tiber flood in December 1495, even though he observed that the whole of Italy experienced very high rainfall in that particular year.²⁴ Similarly Fracastoro, in his poem on syphilis which gave the disease its modern name, advised people to avoid the ‘bad air’ of marshes and south winds:
First of all I would urge you not to be familiar with all types of air: avoid ²³ Dobson (1997: 134) on the use of crude death rates; L. Molineaux in Wernsdorfer and McGregor (1988: ii. 974).
²⁴ Leoniceno (1497) added two extra verses to the tetrastichon, which had been composed by another poet about the Tiber flood in 1495, to emphasize the high rainfall all over Europe in that winter: Tempore Alexandri sexti nonisq; decembris | Intumuit thybris bis senas circiter ulnas. |
Insula quaeque domus facta est. mediisque repente | Circunducta viis aequabat cymba fenestras. | Deucalion eo vix tantum tempore tellus | Diluvium passa est, latuit cum tota subundis. ((In the time of Pope Alexander VI, on the 5th of December, | the Tiber rose about a dozen arm-lengths | Each house became an island, and suddenly in the middle | of the streets a boat brought around reached the height of the windows. | Scarcely so much land was flooded at the time when Deucalion survived a flood that submerged everything.) On Leoniceno and the medical dispute at Ferrara see Arrizabalaga et al. (1997: 59–66, 72–7, also 195–6 on the various epidemics in Rome in the sixteenth century).
Demography of malaria
123
winds which always blow from the south, since they are loaded with dirt and the humidity of foul marshes.²⁵
Since syphilis occurred in many parts of Europe which had no marshes, no ‘bad air’, no sirocco wind, and no river Tiber, the environmental aspects of malaria are obviously of no value at all for explaining syphilis. It is testimony to the enormous influence of malaria that ideas for explaining it were transferred so readily to other, utterly dissimilar, diseases, like syphilis. However, malaria did not only interact with other diseases in the human mind at the level of ideas concerning disease causation. The various pathogens also interacted in reality.
5.2 I
It is interesting to go beyond statistics such as those from Grosseto in the nineteenth century to examine in more detail exactly how malaria interacts with other diseases, especially the gastro-intestinal diseases and the respiratory infections which were assigned the primary role in excess seasonal mortality in Rome in the fourth century by Shaw. Interactions between different species of pathogens can be synergistic, antagonistic, or neutral in terms, first, of their effects on each other and, secondly, of their effects on the human host. It is also possible, for example, that a human genetic mutation which confers a degree of resistance to one pathogen might also make the host more susceptible to another pathogen at the same time. An example of an antagonistic interaction between two pathogens is that between malaria and syphilis (see Ch. 2 above). The sequencing of the entire syphilis genome has shown that the syphilis spirochaete ( Treponema pallidum) cannot tolerate the intense body temperatures generated during malarial fevers because it lacks a factor responsible for transcription of heat-shock proteins in other bacteria and so cannot respond to heat shock.²⁶ The interaction between malaria and syphilis was artificially created in hospitals in Europe and North America as a treatment for syphilis in the central nervous system before the development of drugs against syphilis. However, it is probably not very important in nature, so to speak, in regions with endemic ²⁵ Fracastoro (1530), syphilis. sive morbus Gallicus 2.81–3: In primis ego non omni te assuescere coelo | Exhorter: fuge, perpetuo quod flatur ab Austro, | Quod coeno, immundaeque grave est sudore paludis.
²⁶ Fraser et al. (1998).
124
Demography of malaria
malaria, since syphilis is predominantly a sexually transmitted disease (although congenital transmission through the placenta also occurs) affecting mainly adults. In holoendemic regions people develop acquired immunity to malaria in infancy and childhood and consequently do not have severe malarial fevers during the years of sexual maturity when they are most likely to encounter syphilis. Whether or not syphilis existed in pre-Columbian Europe is a controversial question which need not be considered here.
A possible example of a ‘neutral’ interaction, according to the perceptions of contemporary observers, was that between malaria and
smallpox in early modern Rome. Lapi wrote that if women with tertian fevers touched or suckled children with smallpox they were liable to develop the relatively mild discrete form of smallpox (rather than the more dangerous confluent form). He maintained that in such cases the two different diseases would run their courses separately without getting confused with each other. Although Lapi seemed to believe that the two diseases were not interacting with each other, it is possible that what happened is that malaria depressed the immune systems of the women to the extent that they could develop secondary infections of smallpox (presumably following earlier primary attacks in their own childhood which generally conferred immunity).²⁷ Since smallpox certainly arrived in Italy during the Antonine ‘plague’ in the second century , if it was not present earlier, this interaction was possible from then onwards.²⁸ The focus of the rest of this section will be on definite synergistic interactions between malaria and other pathogens which endanger humans.
The gastro-intestinal diseases were indeed major causes of infant and child mortality in historical populations. In ancient Rome Celsus observed that most of the victims of diarrhoea and dysentery were children up to the age of ten.²⁹ However, the first point that needs to be made is that in European history populations with endemic malaria had higher overall mortality than populations with no experience of malaria (detailed demographic analysis in Ch. 5. 4 below). This can readily be seen by comparing the population of Florence in the thirteenth to fifteenth centuries with ²⁷ Lapi (1749: 48); Fenner et al. (1988: 51–2) on secondary smallpox infections. Endemic smallpox was finally eliminated from Italy in 1947.
²⁸ Duncan-Jones (1996) discussed the Antonine ‘plague’.
²⁹ Celsus 2.8.30.
Demography of malaria
125
nineteenth-century Grosseto. In Florence, which had no history of endemic malaria, gastro-intestinal diseases, operating principally from July to September inclusive, were indeed the leading cause of infant deaths (outside years of epidemics of bubonic plague), the sort of pattern noted by Shaw. In spite of the effects of gastro-intestinal diseases on children, Herlihy and Klapisch-Zuber reached the conclusion that the average duration of life in Florence in the period in question was as high as forty, if years of major plague epidemics are excluded from consideration.³⁰ In contrast, life expectancy at birth in Grosseto much more recently was only twenty. This comparison shows that children assaulted by gastro-intestinal diseases alone have better survival chances than when malaria is present as well. It is also important to note that infant mortality is virtually unrepresented in the funerary inscriptions from the city of Rome in the fourth century used by Shaw. Since direct mortality from P. falciparum malaria principally falls on infants and children where it is endemic, the bulk of its effects in terms of seasonal mortality would not be expected to be visible in those inscriptions in any case.
How does malaria exacerbate the effects of gastro-intestinal infections? P. vivax and P. malariae undergo the process of schizogony in erythrocytes in the peripheral blood vessels of the human body, but P. falciparum merozoites export proteins to the surface of infected erythrocytes which make them adhere to the capillaries of internal organs. Fixation inside internal organs in this way enables the P. falciparum parasites to undergo schizogony under a lower oxygen tension, which favours this process in this particular species of malaria. The most severely affected internal organs are the brain (causing cerebral malaria), the kidneys and the heart and the liver, and (most importantly for the discussion at this point), the placenta in pregnant women (especially primigravidae), in whom any acquired immunity to P. falciparum malaria tends to break down in the second and third trimesters. Lancisi noted that women were particularly badly affected during the malaria epidemic in Rome in 1695.³¹ In tropical Africa pregnant women are bitten twice as frequently as non-pregnant women by the mosquito Anopheles ³⁰ Herlihy and Klapisch-Zuber (1985: 83–4, 276–9). They refrained from constructing life tables for the population of Florence for reasons which will be considered in Chapter 11
below.
³¹ Lancisi (1717: 210–11).
126
Demography of malaria
gambiae, increasing their chances of being infected by malaria, perhaps because pregnant women produce a larger volume of exhaled breath on average.³² Since mosquitoes are generally believed to locate their prey by detecting chemicals emitted in human breath, European vectors of malaria may also have responded in a similar way to pregnant women in the past (although it is also possible that mosquitoes detect chemicals in human sweat). A new idea which is currently attracting attention among researchers in medicine is that placental malaria is caused by a specific population (or set of haploid clones) of P. falciparum with a special affinity for certain chemical receptors generally found only or predominantly in the placentae of pregnant women.³³ The presence of large numbers of plasmodia in the placenta frequently causes miscarriages in non-immune women, or fetal anaemia and intrauterine growth retard-ation in women with some degree of immunity, leading to low birthweight infants who are particularly susceptible to diseases in general and gastro-intestinal infections in particular. Desowitz, a specialist in tropical diseases, put it as follows:
The babies that are born to the malarious usually have a birth weight 200
grams or more below normal. A low-birth-weight baby who doesn’t ‘catch up’ has a puny immune system, and these children frequently die during childhood of all sorts of infections. The death certificate . . . may read diarrhoea or pneumonia, but the real cause of death was malaria of the mother.³⁴
This is how malaria can enormously exacerbate the effects of gastro-intestinal diseases on infant and child mortality without even being present in the children. Average birth weight of infants in fact increased significantly in areas of former malaria endemicity following modern eradication campaigns. Maternal infections with P. vivax may possibly reduce the severity of subsequent infections with P. falciparum during pregnancy, owing to immunological cross-reactions, but P. vivax itself causes birthweight reduction in all pregnancies, not just in primigravidae, in areas with endemic malaria. It has been estimated that up to 50% of low-birth-weight babies in some populations in tropical countries may be caused by ³² Lindsay et al. (2000).
³³ McFalls and McFalls (1984: 101, 107–8); Fried and Duffy (1996), discussed by Matteelli et al. (1997); Marchiafava (1931: 50–1) on placental malaria in Italy; Gilles and Warrell (1993: 46); Reuben (1993); Fried et al. (1998).
³⁴ Desowitz (1992: 118); Brabin (1992); Newby and Lovel (1995); and Brabin et al. (1996).
Demography of malaria
127
maternal malaria.³⁵ In Sardinia, in areas where the transmission rate of P. falciparum malaria was extremely high, fertility rates were higher than in areas where the transmission rate was lower. This proves a degree of adaptation by women to intense malaria, as a result of both acquired and inherited immunity (e.g. thalassaemia and glucose-6-phosphate dehydrogenase deficiency in Sardinia), but it has to be remembered that overall mortality was much higher and life expectancy much lower in the areas with a very high transmission rate. Higher fertility was required to balance higher mortality. Even with such increments in fertility, the mortality profile of populations exposed to endemic malaria was extremely poor.³⁶
Besides the indirect effects on infants caused by maternal infection, P. falciparum malaria can also be present in infants and children themselves alongside gastro-intestinal infections. In Macedonia during the First World War it was observed that malarial infections readily coexist with typhoid fever, paratyphoid fever, and amoebic dysentery, for example. Of particular interest is the observation that when patients in Macedonia were infected with both typhoid fever and P. falciparum malaria, as was proved by examination of blood smears for malaria and microbiological techniques for typhoid fever, the clinical symptoms observed were principally those of typhoid fever. The implication of this
with regard to ancient sources is that cases, for example in the books of Epidemics in the Hippocratic corpus, which look like cases of typhoid fever, could very easily have been infected with malaria as well, but this would not be apparent from the description given in the ancient text. Such dual infections are very likely to happen in areas where malaria is endemic. Malaria may also directly interact with gastro-intestinal infections, since there is some evidence that malaria suppresses the immune response to typhoid fever and other types of salmonella.³⁷ Malarial fevers do in any case strongly resemble the ³⁵ I. A. McGregor in Wernsdorfer and McGregor (1988: i. 757); Brabin and Piper (1997); Nosten et al. (1999) on P. vivax.
³⁶ Zei et al. (1990) observed that placental malaria is in fact more severe in women in areas of low transmission (and so less acquired/innate immunity) than in areas of high transmission (and so a higher degree of immunity).
³⁷ Armand-Delille et al. (1918: 79). Corvisier (1985: 117) lists six probable cases of typhoid fever in the Hippocratic Epidemics; Mabey et al. (1987). Urban et al. (1999) described one mechanism by means of which P. falciparum suppresses the human immune response to infection. Faccini (1984) discussed typhoid fever in early modern Italy.
128
Demography of malaria
fever of typhoid in the early stages of infection, before the classic periodicity is established (if it is established), but usually without the severe gastro-intestinal symptoms of typhoid fever. Before the true aetiology of malaria was established, this similarity between P. falciparum malaria and typhoid fever led many doctors to assimilate the two in the form of a syndrome, called typhomalarial fever, whose reality is no longer accepted by modern doctors. The similarity was noted in Rome, too. Baccelli stated that the malaria of Rome frequently took the form of what he called a subcontinua tifoide.³⁸ He reckoned that Lancisi had observed the same form of the disease in the Trastevere district of Rome during the great epidemic in 1695.³⁹ This was the most dangerous type of malaria for Baccelli.
Malaria and Rome: A History of Malaria in Ancient Italy Page 18
