The dawn of disease

The threat of global pandemics looms large as antibiotic resistance continues to persist. But how did it all begin? Dr Guillaume Fournié thinks the early development of farming could have promoted the emergence of one of today’s most common bacterial zoonosis 

About 13,000 years ago, our ancestors’ diet consisted of the wild plants they picked and meat from the wild animals they hunted. Nowadays, almost all of us are fed from farming systems. The development of agriculture during the Neolithic period (13,000-5,000 years ago) did not only drastically modify our diet, but it also created conditions for the exponential growth of human populations. This had far-reaching repercussions on the environment and our social organisations.

Can we consider agriculture as evidence of progress that improved the quality of life of early farmers by allowing them to forgo the hunter-gatherer lifestyle for the better? Archaeological evidence suggests the contrary: it made their life harder and shorter compared with those of their contemporary hunter-gatherer neighbours. According to Jared Diamond in Guns, germs, and steel, the progressive sedentarisation of human populations and greater reliance on plant and animal production was the worst mistake in human history, bringing great misery to mankind: malnutrition and famine, social and gender inequality, despotism and warfare, and a new set of infectious diseases.

The rise of pathogens coincides with the change in early humans from hunter-gathers to farmers.

The close proximity between high density animal and human populations promoted the cross-species transmission of pathogens and their adaptation to humans. As human populations were large and interconnected, conditions were created for pathogens, such as smallpox and measles viruses, to cause major epidemics, and be sustained within these populations, despite acute infections resulting in the death or immunisation of their hosts. This would not have been possible in small hunter-gatherer groups. An outbreak would ultimately fade out in smaller populations, resulting in the extinction of the newly transferred pathogen. The epidemiology of zoonotic pathogens which are not transmitted between humans, but exclusively from animals to humans, was also impacted by the Neolithic transition.

Common infection Brucella melitensis is the main agent responsible for human brucellosis, today’s most common bacterial zoonosis in the world. Humans become infected through ingestion of unpasteurised dairy products and the management of infected animals – primarily sheep and goats, the main reservoir of the bacteria. Before their domestication, wild small ruminants were hunted by hominid species, and these contacts likely resulted in sporadic human brucellosis cases, as suggested by lesions suggestive of brucellosis on a 2.4-2.8 million year old hominin Australopithecus africanus skeleton. However, the development of animal farming is likely to have further enhanced the risk of human infection by increasing the prevalence of infection among in-contact animal populations, and the frequency of contact between humans and infected animals through, for instance, the emergence of milk exploitation.

On the path to domestication Sheep and goats were among the first domesticated animal species. The long evolution from the hunting of wild small ruminants to the management of captive populations started in multiple locations of the Fertile Crescent about 10,000-11,000 years ago. This region of the Near East – shaped as an arc and curving from the Levant through south-eastern Turkey, western Iran, northern and eastern Iraq, into the Persian Gulf – was one of the main and earliest centre of the Neolithic transition. Several plant and animal species, which are now key agricultural resources, were domesticated there, before their farming spread to the entire world. On the path towards the creation of domestic species, morphologically and genetically distinct from their wild ancestors, the first impact of the human management of animal populations was the modification of their demographic profiles.

Some archaelogical research has sought indications of such human management of animal populations through the study of skeletal remains

While wildfowl may live several years, there is no need, from a productionist perspective, to keep feeding chickens intended to become meat more than a few weeks – the time required for them to reach their optimal weight. Early farmers discovered that only a few male animals were needed to ensure the reproductive continuity of a sheep or goat flock, and that it would be more productive to slaughter young males for consumption while retaining females for reproduction. Some archaelogical research has sought indications of such human management of animal populations through the study of skeletal remains. As long bones, such as proximal and distal radius, humerus, tibia, calcaneus, metacarpal and metatarsal, fuse at different ages and exhibit sexual dimorphism, counting the number of fused and unfused bones allows archaeologists to estimate sex-specific survival curves, and to reconstruct the demographic profiles of Neolithic animal populations. Evidence of managed goat populations could thus be identified in multiple locations, in particular in the Zagros region, across Iraq and Iran, where goat was the only domestic food animal for around a millennium.

Early management One of the earliest and most accurately dated demographic profiles suggestive of a managed goat population was at Ganj Dareh in the Zagros highlands, where possible brucellosis lesions were identified in a human skeleton. The spread of goat husbandry can then be followed to nearby lowland zones in Ali Kosh and Jarmo, a few centuries later. Reconstructed goat demographic profiles on these three sites showed different features, illustrating the diversity in management strategies during earlier phases of animal husbandry. Of particular interest were the variations in the intensity of selective harvesting of young males: the estimated proportion of females among adult goats was 57% and 62% in Jarmo and Ali Kosh, but reached 80% in Ganj Dareh. While these alterations of goat population structures were likely to be motivated by an attempt to improve meat, and potentially milk, production, its consequences on the epidemiology of brucellosis remains uncertain. To address this question, Robin Bendrey (University of Edinburgh), Dirk Pfeiffer (City University of Hong Kong and Royal Veterinary College, London) and I developed a mathematical model simulating the spread of Brucella melitensis within goat populations in Neolithic settlements comparable to the sites of Ganj Dareh, Ali Kosh and Jarmo.

The spread of goat husbandry can then be followed to nearby lowland zones in Ali Kosh and Jarmo, a few centuries later

This work has been published in the Royal Society Open Science journal. The results suggest that, following its incursion in a settlement, the pathogen could be maintained for low levels of transmission in goat populations of which the sizes were within the estimated ranges for the three investigated Neolithic sites. Conditions were present in these early domestic goat populations for the establishment of endemicity of the pathogen, which could have thus acted as a potential permanent reservoir for human infection.

A real networker The intensity of sex-biased harvesting was found to be highly influential on pathogen transmission. Indeed, preferential harvesting of males increased the proportion of females in the population, and therefore, their proportion among newly infected goats. Infectious material excreted from the vaginal tract of infected goats following abortion or full-term parturition is the main source of infection for susceptible goats. Adult females being responsible for pathogen transmission then, such a population structure would promote the transmission of Brucella melitensis. 

The pathogen would also circulate at a higher prevalence in populations with high male-biased harvesting

The pathogen could thus persist in populations with high male-biased harvesting at disease transmission levels that did not allow it to persist in comparable populations of the same size, but for which harvesting was not biased towards males. The pathogen would also circulate at a higher prevalence in populations with high male-biased harvesting. Moreover, Neolithic communities are known to have been linked by a broad range of regional interactions, including exchange networks amongst diverse social relationships. Such contacts between human settlements may have promoted the spread of certain animal management practices, but also the direct transmission of Brucella melitensis among their animals. Even low levels of mixing between goat populations could re-introduce the pathogen in populations in which it faded out, increasing its potential to become endemic.

Same again? The Royal Veterinary College’s study suggests that the increase in livestock densities may not be the only feature resulting from the early development of farming that promoted disease invasion and maintenance. The alteration of goat population dynamics and contact patterns promoted the exposure of humans to a zoonotic pathogen at an early stage of farming development. The development of an understanding of the interrelationship between disease dynamics and population characteristics will allow future osteological and genetic research to focus on those areas most likely to produce direct evidence for the emergence of livestock-related zoonotic disease.

These findings further support the view that the transition from food collection to production during the Neolithic transition, while allowing for larger human population sizes, resulted in significant adverse effects on human health and wellbeing. Along this progressive path leading towards the full reliance of human societies on food production, people were not able to appreciate the consequences of their decisions, as these were not directly observable. However, even if they recognised this causal relationship, it is uncertain whether they would have renounced the perceived benefits of more effective food production. Nowadays, we are facing another agricultural transition, with the rapid intensification of production, drastic increases in the sheer number of domestic food animals, and growing interconnections between animal production systems across the world.

Contrary to our Neolithic ancestors, we can now anticipate the possible epidemiological consequences, which include increased risk of the emergence of human pandemic viral strains and the development of antimicrobial resistance. Mitigating these risks by altering farming practices remains, however, a major challenge.

Author: Dr Guillaume Fournié is a veterinarian and epidemiologist at the Royal Veterinary College.

He is currently involved in projects focusing on avian influenza, peste des petits ruminants, joint economic and epidemiological modelling, and the role of live animal trade networks in the spread of animal and zoonotic diseases.