Introduction

Leptospirosis is a worldwide bacterial zoonotic disease causing around 1,000,000 cases and 60,000 deaths each year [1]. Leptospirosis has long been considered as a disease of poverty and precariousness because of the association between low incomes, poor sanitation or housing and leptospirosis past exposure assessed by serological survey or leptospirosis incidence as demonstrated in numerous studies especially conducted in Brazilian urban slums [2–4]. French Guiana (FG) is a French overseas territory located on the northeastern coast of South America bordering Brazil and Suriname. French Guiana has a population of approximately 300,000 inhabitants, most of whom live in urban or suburban areas along the Atlantic coast. Over the past decade, demographic growth in informal settlements has been particularly increasing around Cayenne, the main city and could concern up to 20% of inhabitants [5,6]. This phenomenon occurs in a context of growing precariousness for many inhabitants, as well as continuous migratory flow from the neighboring countries but also from more distant countries such as Haiti, Syria or Afghanistan. A large part of the French Guianese population lives in a high level of precarity across multiple dimensions: social, financial, residential but also sanitary [7–9]. This has serious implications for health, including the local epidemiology of leptospirosis. Between 2016 and 2022, the annual incidence more than doubled compared with 2007–2014, highlighting an increased disease burden [10]. Approximately one third of the affected individuals lacked health insurance, and a similar proportion lived in informal settlements—both figures higher than those observed in the general population. This overrepresentation of uninsured individuals and residents of informal settlements among leptospirosis cases suggests that the disease may disproportionately affect socially and economically vulnerable populations in French Guiana. Poor living conditions in urban informal settlements, particularly limited access to safe water and adequate sanitation, likely contribute to this increased risk. With few exceptions, residents living in informal settlements generally do not have access to running water for drinking, food preparation, and other domestic uses. In these areas, inadequate wastewater and solid waste management leads to environmental contamination, including rudimentary disposal sites, insufficient garbage collection, and high rat populations. These factors provide a favorable setting for the emergence and transmission of leptospirosis. However, few data are available on the population living in these areas, apart from studies conducted by organizations involved in public health prevention, such as the Red Cross [11,12]. Especially, no data exist in French Guiana on leptospirosis exposure among people living in informal settlements. Collecting local data, whether through risk assessment or serological surveys, is a necessary first step to guide public health strategies in these communities.

The primary objective of the study was to assess the prevalence of prior Leptospira exposure in a population living in informal settlements near Cayenne using MAT serological analysis, and to investigate factors associated with seropositivity. The secondary objective was to describe risk factors for leptospirosis exposure among residents of informal settlements in French Guiana.

Materials and methods

Results

The initial Hantavirus survey was conducted in 274 volunteers, which corresponded to roughly 30% of the inhabitants in study areas based on previous estimations [12]. Among them, 266 (97.1%) accepted to participate in this ancillary study. Finally, investigators were able to administer the complementary questionnaire to 190/266 (71.4%) participants during individual Hantavirus results delivery.

Fig 1 details the flow chart of the reported leptospirosis ancillary survey.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow chart of participants in leptospirosis serological survey.

MAT Microscopic agglutination test. *Complementary questionnaire was administered during the delivery of Hantavirus serology results to 190 participants.

https://doi.org/10.1371/journal.pntd.0013764.g001

Median age of participants was 42 (34.4-51.7). Age distribution is presented in supplementary S1 Fig. Women accounted for 54.0% (143/265) of cases. Participants born in mainland France or French Guiana accounted for 3.9% (9/233) of participants. Participants were born mostly in Haiti (82.8%, 193/233). Sociodemographic data are given in Table 1. Participants born abroad had been living in French Guiana for more than two years before the survey took place in 92.6% (174/188) of cases. Participants resided in the study area for less than one year in 7% (13/186) of cases and more than 1, 2 and 5 years for 93.0% (173/186), 81.7% (152/186) and 43.6% (81/186) of them respectively. Distribution of the duration of living in French Guiana (supplementary data S2 Fig) and in the study area (supplementary data S3 Fig) are provided.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Sociodemographic data and residence characteristics.

https://doi.org/10.1371/journal.pntd.0013764.t001

Half (54.5%, 145/266) participants lived in a carbet defined by informal housing made of metal sheets and wood. However, 83.1% of participants (246/264) had tiling or concrete as primary flooring material. Housing conditions are summarized in Table 1.

With regard to domestic freshwater use, most participants frequently used the prepaid water standpipe in their neighborhood (78.6% 209/266).

More than 90% of respondents to the complementary survey used an alternative source in addition to the neighborhood collective water distribution standpipe. Among them, nearly half used a well or collected rainwater. Half of the well users did not use protection from flooding. Two thirds of the users of rain water collector protected it with a lid. Table 2 details the water collection sources and usages of participants.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Water collection sources and usages.

https://doi.org/10.1371/journal.pntd.0013764.t002

Over half (54.9% 146/266) reported a professional activity. Construction worker (30.8%), odd jobs (15.8%), housekeeping/ cleaning (14.4%) and commerce (13.0%) were the most frequent occupations. Table 3 details occupational and non-occupation individual exposures. Among, those reporting professional activity, 67.8% (99/146) were estimated to have an occupation at risk for leptospirosis.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Occupational and other individual exposure factors.

https://doi.org/10.1371/journal.pntd.0013764.t003

Overall the median (IQR, range) number of cumulative risk factors investigated by the questionnaires was 7 (6–8; 2–12), considering only the participants who responded to the ancillary survey (n = 190).

MAT analysis revealed that sera were reactive for 7.5% 95% CI [4.7-11.4] (20/266) participants. Among the 20 reactive sera, 5 presumptive infective serogroups were identified: Ballum (n = 9), Icterohaemorrhagiae (n = 6), Panama (n = 3), Australis (n = 1) and Tarassovi (n = 1), with a median [IQR] titer of 100 [100–200]. Results of MAT analysis are presented in Table 4.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Microscopic agglutination test results.

https://doi.org/10.1371/journal.pntd.0013764.t004

Bivariate analysis revealed that walking barefoot or with open shoes during heavy rainfall or in stagnant water was more common among seroreactive participants (93.3%, 14/15) than non-reactive participants (64.9%, 96/148), with an odds ratio of 7.6 (95% CI: 1.1–326.7, p = 0.039).

The median (IQR, range) number of cumulative risk factors was 7 (6–9; 5–11) and 7 (6–8; 2–12) for seroreactive and non-reactive participants, respectively (p = 0.491). Logistic regression analysis showed that an increase of the number of risk factors was not associated with a higher risk of positive MAT test, with an OR of 1.12 (95% CI: 0.86–1.45) per one additional risk factor (see Table 3).

Discussion

Leptospirosis at-risk exposures were very common among surveyed participants of three informal settlements of French Guiana, especially rodent vicinity. This proximity with rodent combined with housing type characteristics, appears to be a relevant criterion to estimate the risk of severe leptospirosis. A study conducted in urban slum populations in Salvador, Brazil, revealed that signs of rodent infestation close to the household as well as unplastered walls were risk factors for the incidence of severe leptospirosis [17]. However, the magnitude of association between rodent proximity, rat abundance or Leptospira excretion rate by rats and Leptospira seroprevalence has been debated. In fact, environmental factors such as inadequate source of water, sewage, flooding risk and sanitation infrastructure as well as other environmental characteristics have been reported as even more important drivers for leptospirosis transmission [18,19]. Nevertheless, Rattus spp. remains the main reservoir of Leptospira and is involved in the transmission cycle of human leptospirosis and rat exposure has been associated with Leptospira seroprevalence at various degrees [20–22]. The only available data on the rate of Leptospira excretion by invasive rodents in French Guiana are outdated and provide little informative value. [23]. In Brazil, reports have provided rates of Leptospira excretion by rats of 31–92% [24]. Measuring the prevalence of Leptospira excretion in locally collected rats could provide valuable data to better assess environmental circulation in the area. Most participants frequently used a source other than the collective prepaid water standpipe. While it seems reasonable to assume that the standpipe-collected water provides free-from-Leptospira water, a regular use of the standpipe may not be totally harmless [25]. In fact, residents often go to the standpipe wearing open shoes or even barefoot and can be in contact with contaminated soil. Indeed, standpipes are often located at the bottom of residential areas where water can drain or accumulate and increase the risk of leptospirosis [22,26,27]. In addition, standpipes are also often close to waste deposits at the entry of the settlements (see supplementary S4 Fig). Regarding the use of unsafe water, different degrees of risk may exist depending on the type and provenance of water source used [28]. The use of wells or rainwater collection were reported by half of the participants whereas they are both risk factors for leptospirosis especially when not protected [14,29–31]. Providing information and protective measures to protect the collected water from contamination should be a priority and be considered as an actionable measure to decrease the risk of leptospirosis in informal settlements [32].

A previous leptospirosis infection was found in 7.5% (95% CI [4.7-11.4]) of patients, a lower rate than expected given the high level of exposure. In a study conducted in Cali, Colombia, seroprevalence (MAT, titer threshold 1: 100) was 12% (95% CI [10–14]), slightly higher than the present study despite most of participants had access to running water [33]. In another study conducted in an urban community of Puerto Rico, a higher seroprevalence (MAT, titer threshold 1: 50) of 27.2 (95% CI [21.2-33.9]) was found [34]. Delight et al. recently reported a large study conducted among urban slum residents of Salvador, Brazil, and found a seroprevalence (MAT, titer threshold 1: 100) of 11.3% (95% CI [9.1–13.8]) again slightly higher but in the same magnitude as the reported study [35]. Many factors were associated with seropositivity with a differential effect depending on gender (see below). These three seroprevalence studies were conducted in response to occurrence of cases locally [33,35] or because environment was favorable for leptospirosis transmission as done in French Guiana [34].

So far, the unique study of human leptospirosis seroprevalence conducted in French Guiana has been performed among illegal goldminers operating in FG and investigated mostly men (73%). The seropositivity (MAT, titer threshold 1: 50) rate was notably higher than in the present study population in 2015 (31.0%, ORPAL1 study) and 2019 (28.1%, ORPAL2 study). Since a different titer threshold was used, comparison may be questionable. Thus, to better contextualize our findings, we reanalyzed the ORPAL1 and ORPAL2 data on MAT seroprevalence in collaboration with the ORPAL study investigators (Prof. Maylis Douine, personal communication). Applying a 1:100 cutoff, the seropositivity rate was 19.2% and 16.2% for 2015 and 2019 respectively. Both were still significantly higher than the seropositivity rate observed in informal settlements. A hypothesis to explain this higher seroprevalence would be a high level of mud exposure during gold mining activity associated with a high level of environmental contamination of gold mining sites. A contamination of illegal goldminers outside of gold mining sites is also a credible situation given the precariousness of this population [36].

The moderately elevated seropositivity rate observed in the study may suggest two hypotheses. First, if we consider that there is a high level of environmental contamination it should imply that preventive measures (not investigated) are taken to avoid the penetration of the bacteria. The role of individual preventive measure has already been reported [29]. This would imply that the supposed exposure factors, which were frequently reported and cumulated among informal settlements inhabitants may not actually be as risky as expected in this context. On the contrary, if we acknowledge that the level of cumulative exposure factors was indeed elevated, this could imply that the environment of study sites may possibly be less contaminated than anticipated. Repeating surveys in other settings could help to explore these hypotheses as well as exploring environmental contamination through soil and surface water analysis.

The sole factor associated with MAT seropositivity in the reported study was a feet skin exposure to wet environments during heavy rain or in stagnant water. Although the confidence interval is wide, this factor appears plausible given known leptospirosis transmission pathways and the aforementioned risk of transmission around standpipes. In contrast, male gender was the only factor associated with seropositivity among illegal goldminers working in French Guiana. A recent study conducted across four informal settlements at-risk communities in Salvador, Brazil used an original gender-based analysis to investigate differential effects on risk perception, exposure, and at-risk activities for leptospirosis [35]. Male gender and increasing age were the unique factors related to positive Leptospira serostatus when investigating the whole study population. Interestingly, men who perceived leptospirosis as extremely serious had lower odds of being seropositive and engage in at-risk activity such as walking through sewage and walking barefoot outside. The impact of diseases severity perception was different among men and women. Besides, at-risk occupation increased the risk for seropositivity in men while age increased the seropositivity rate among women. This suggests a clear differential effect of gender on the modalities of leptospirosis exposure.

This study has several limitations, particularly the fact that many patients did not answer to the second questionnaire generating missing data and its non-random sampling design. All participants were recruited after being sensitized through outreach campaigns, either several days before sampling or on the same day just prior to participation. Consequently, we cannot exclude the possibility that this pre-intervention sensitization about a rodent-borne disease (HPS) may have biased some of the questionnaire responses. Nevertheless, more than 90% of sensitized inhabitants expressed their intention to participate, and nearly 80% eventually did so [12]. The attrition between sensitization and sampling was therefore limited, suggesting that inclusion bias among the sensitized population was minimal. The main limitation relates to representativeness and generalizability, since the outreach campaigns reached only about one-third of the estimated population across the three sites, despite the investigators’ extensive efforts. To better estimate the seroprevalence of leptospirosis and associated factors across French Guiana, larger and random sampling should be considered as performed for other zoonotic diseases [37,38]. Additionally, the time and place of acquisition of the previous exposure to leptospirosis cannot be ascertained since MAT remains positive for several years [39]. Unfortunately there is no data regarding the epidemiology of leptospirosis in Haiti to discuss the seroprevalence of the participants born in Haiti. Since 1965 the sole published study of leptospirosis cases from Haiti is a case series of three patients who needed air medical transport [40]. Most people born in Haiti who migrate to French Guiana travel through Brazil or Suriname. There are no published data on the epidemiology of leptospirosis in Suriname. In Brazil, numerous studies have investigated leptospirosis across various regions and contexts. However, the specific regions and duration of transit for each migrant participant were not assessed. For these reasons, further analyses on this aspect were not feasible. Another limitation is that the study focused on peri domestic exposures, without extensively investigating extra-domestic activities - apart from occupational and agricultural activities but without detailing them. Many potentially relevant risk factors for leptospirosis-such as socioeconomic status, health literacy, prior disease knowledge, exposure to animals other than rodents, and local wastewater management-were scarcely or not assessed.

Finally, despite the methodological limitations of the study, the estimation of seroprevalence in understudied informal settlements and the identification of a local risk factor for exposure represent relevant findings for the local management of leptospirosis. Although their impact may be modest, these findings could support intervention campaigns around reported cases as well as broader awareness initiatives.

Conclusion

This study confirmed a high rate of leptospirosis exposure factors but a rather limited seroprevalence among the surveyed population of three informal settlements of French Guiana. This may suggest that despite at-risk exposures the effective transmission of leptospirosis remains limited within the study area because of individual protective measures or a limited contamination of environment. Broader serological surveys and environmental studies should clarify the areas of at-risk leptospirosis transmission in French Guiana.

Supporting information

Acknowledgments

The authors would like to thank people among Agence Régional de Santé Guyane (Mathilde Ballet, Mathilde Hangard, Alban Lemonnier, Franky Mubenga, Adrien Ortelli, Solène Wiedner-Papin), Institut Pasteur de la Guyane (Damien Donato, Dominique Rousset), Centre National de Référence de la Leptospirose Institut Pasteur de Paris (Pascale Bourhy, Sofia Pombo, Farida Zinini, Céline Lorioux, Jean-François Mariet), La Croix-Rouge francaise (Karl Kpossou, Diana Perez-Aranguena, Luxamarre Saint Hervé, Marckenson Therasse) and Cayenne Hospital (Frédégonde About, Amadou Oury Baldé, Yoris Demars, Olivier Lesens, Camille Meunier, Marina Palomar, Déborah Porez, Mathilde Zenou), Santé Publique France Guyane (Luisiane Carvalho, Sophie Devos, Marion Guyot, Tiphanie Succo), Guyane Promo Santé (Jean-Luc Bauzat) and others who contributed to the Hantavirus survey and the ancillary study on leptospirosis. The authors acknowledge Prof. Maylis Douine (Cayenne University Hospital) for sharing the MAT data of ORPAL1 and ORPAL2 studies.