Brucella species are gram negative bacteria causing brucellosis, a zoonotic disease of global significance1. The species most often involved in human infection are B. melitensis, B. abortus, B. suis, and B. canis2. The most common clinical findings in human brucellosis are fever and a flu-like illness, followed by osteoarticular manifestations3. Lymphadenopathy and a broad palette of genitourinary, neurological, pulmonary and gastrointestinal complications are also possible3,4. Human brucellosis has an overall mortality rate of 5%, most commonly linked to endocarditis3. In livestock, brucellosis causes abortion, infertility and subsequent loss of milk production5. In addition to the high disease burden in animals and humans, brucellosis can cause substantial economic impact due to reduced animal production and patient time lost from work and routine activities6,7. B. abortus and B. melitensishave the greatest economic impact5. Animal to human transmission occurs through direct contact with placenta, amniotic fluid or aborted foetuses or during slaughtering of infected animals, as well as through consumption of unpasteurized dairy products from infected animals4. In Latin America, Brucella spp. have been isolated from humans, domestic animals (cattle, goats, pigs, sheep, dogs and horses) as well as wildlife (buffalos, foxes, grey weasels, capybaras and ferrets)8. Most human isolates have come from Argentina, Peru and Mexico; whereas in Colombia, Brazil, Chile and Cuba, Brucella spp. was predominantly isolated from cattle8.

There are limited reports on the seroprevalence of Brucella spp. in humans from South America. One study on the population of an urban slum in Brazil found sero-positivity of 13% for B. abortus and 4.6% for B. canis9. A study from Northwest Ecuador found an overall seroprevalence of only 1.9%10. Although, the seroprevalence for participants in the high-risk group (occupational exposure to animals) was 4.8% compared to 1.4% in the low-risk group (no occupational exposure to animals)10.

Serological studies in animals from South America have primarily been conducted in Argentina and Brazil. In Argentina, Brucella spp. exposure has been reported in dogs (14.7%)11, cattle (3.7%)12, water buffalo (6.4%)13, foxes (17.8%)14, minks (9.2%)15 and armadillos (16.0%)16. In Brazil, Brucella spp. exposure has been reported in cattle (2.2% - 3.7%), buffaloes (4.8% - 6.8%)^17,18, equines (1.3%)19 and rams (2.9%)^20,21. One additional study from Ecuador also reported exposure in cattle (16.7%)22,23.

In this study we hypothesized to find a high seroprevalence of Brucella spp. in the two indigenous communities, the Wiwa and Koguis, residing in the Sierra Nevada de Santa Marta, Colombia. Due to the limited contact these communities have with the outside world, little is known about their healthcare needs and disease burden. It was expected that these communities would have a high burden of infectious disease due to their limited access to medical services, poor socioeconomic status, inadequate access to sanitation and clean drinking water and simple housing (clay huts with unsealed floors)24,25. These communities live in close contact with their livestock (cattle, small ruminants, pigs and poultry), indicating a potential susceptibility to zoonotic pathogens including Brucella spp.

The samples included were from participants aged from 1-90 years, with a median age of 22 years (IQR 11 - 37), with 57% being female. Of the 539 serum samples screened by the Serion ELISA, one sample was ELISA positive and one was equivocal. In the screening phase, thirteen samples (including the positive and equivocal sample from the Serion ELISA) demonstrated an OD greater than the additional cut-off calculated based on the negative results. These 13 samples underwent confirmatory testing by the BrucellaCapt test (performed at the certified laboratory IFIK, Bern, Switzerland). One sample, from a Koguis participant was found positive by the BrucellaCapt test – leading to an overall seroprevalence of 0.2% (95% CI 0.0 – 1.0).

Exposure data in the Wiwa people (n = 147) identified specific practices associated with zoonotic disease transmission: animal slaughtering (84.4%), assisting with birth (23.8%) and milking (18.4%) including cattle, goats and sheep (Table 1). One investigator reported similar practices among the Koguis people. All Wiwas and Koguis questioned reported regular exposure to and handling of livestock and animals (cattle, goats and sheep), thereby limiting the unequivocal identification of risk factors.

A very low seroprevalence of <1% was identified for Brucella spp. in over 539 sera from two indigenous communities residing in the Sierra Nevada de Santa Marta, Colombia. While studies are limited on brucellosis in Colombia, two case reports on Brucella spp. infections in humans describe; a farmer with antibodies to Brucella spp. (not speciated) believed to have been infected through contact with cattle32 and a merchant diagnosed with B. melitensis by blood culture, whose infection was suspected to be linked to consumption of unpasteurized milk33. A molecular study in the Colombian Caribbean region found 22.2% (6/27) of cheese samples made from unpasteurized cow milk to be positive for Brucella spp. DNA34. B. abortus is the only species previously identified in livestock from Colombia, with herd-level seroprevalence reported between 22.0% – 27.5%3536373839. The only other species identified in Colombia is B. canis, found in humans (9.0%) and dogs (2.0% - 15.0%) in Bogotá and the Antioquia region404142. There is evidence for the absence of Brucella spp. exposure in regions of Colombia in two serological studies; one in domestic pigs (n = 350, in 23 farms), and another in collared peccaries (Pecari tajacu, n = 58) and feral pigs (n = 15) finding no seropositivity for Brucella spp.43,44.

The serion ELISA used in our study is coated with native extract from B. abortus, containing genus-specific antigens for the detection of antibodies against human pathogenic Brucella spp31. The manufacturer reported a high sensitivity (>99%) and specificity (99.3%)31. An independent evaluation, however, found a lower sensitivity of 84%, subsequently there is a possibility of false negatives with the ELISA45. Although, we do not believe this to affect the findings of this study as the samples with the highest ELISA OD readings were confirmed negative by the BrucellaCapt test performed at the certified IFIK, Bern, Switzerland. The BrucellaCapt test covers the three smooth lipopolysaccharides (LPS) containing human pathogenic Brucella spp. (B. melitensis, B. abortus and B. suis) with a reported 100% sensitivity and specificity45,46. B. canis, however, has a rough LPS and is often not adequately diagnosed by routine serological tests47. The reports of B. canis in Colombia may warrant further epidemiological investigations to determine the exposure to dogs and B. canis among the indigenous tribes of the Sierra Nevada de Santa Marta.

While we found limited evidence of Brucella spp. amongst the Wiwa and Koguis tribes, this should not discount the potential risk of zoonotic diseases of these remote, vulnerable communities, with notably highly limited access to health care facilities.

Zoonotic diseases are responsible for substantial morbidity and mortality, particularly in low resource communities with high dependence on livestock48. Globally, zoonoses account for 60.3% of emerging infectious diseases48. The major drivers of their emergence or re-emergence include expansion of cities or farmland with deforestation, increasing human population and urbanization, globalization of food systems, climate change and environmental contamination484950. Remote tribes having been protected through their isolation and with high proportions of immunologically naïve inhabitants may be particularly vulnerable to these drivers and exposure to new pathogens. Given the high level of exposure to animals in the populations of this study, it was unexpected to find such a low seroprevalence for this normally common zoonotic disease. The low rates could be attributed to the isolated raising and maintenance of livestock with small herd numbers, reducing the contact rate and transmission risk51. Introduction of infected animals poses a risk to these vulnerable populations and discussions with the community on potential surveillance programs are warranted. One Health surveillance systems provide a means of early detection to prevent outbreaks in both animal and human hosts52. These systems are particularly effective for zoonoses that have limited human-to-human transmission risk, as is the case with brucellosis, which can be detected and controlled in the animal host preventing transmission to humans53,54.

A limitation of this study is that despite testing more than double the calculated sample size, the sample may not reflect adequately the geographical distribution in the Sierra Nevada de Santa Marta, Colombia. The Wiwa and Koguis participants accounted for 1.5% and 1.7% of their entire populations, respectively – and the sampled villages were located to the East of the Sierra Nevada de Santa Marta. We cannot discount the possibility of differential exposure in other Wiwa and Koguis villages. Further, is that we do not have the risk factor data for the Koguis participants, including the Brucella spp. positive individual. Consequently, we cannot comment on the possible route by which this individual became exposed to the bacteria.

The evidence of Brucella spp. exposure in the Wiwa and Koguis tribes was found to be very small with a seroprevalence of <1% by screening/confirmatory testing. Consultation with these communities for the implementation of systematic One Health investigations for the detection and prevention of potential emerging zoonotic threats are recommended due to their isolation and limited access to healthcare.