Sargassum Inundations and the Risk of Hypertension Disorders Among Pregnant Women Living in the French Caribbean Island of Martinique.

The present study was conducted in accordance with the amended Declaration of Helsinki ( https://www.wma.net/what-we-do/medical-ethics/declaration-of-helsinki/ ; accessed on 20 September 2024) and Good Clinical Practice guidelines (GCP European Directive 2005/28/EC; accessed on 20 September 2024). Written informed consent was systematically obtained from all patients. The study was approved by the local Institutional Review Board of the University Hospital of Martinique (IRB 2023/033). The French Caribbean Island of Martinique (14.6415° N, 61.0242° W, surface area of 1128 km 2 ) is one of the most populated territories in the Caribbean basin (population size of 365,734 in 2023). The east coast of the island is bordered by the Atlantic Ocean, while its west coast faces the Caribbean Sea. Oceanic current and maritime trade winds off the Atlantic coast expose Martinique’s eastern coast to coastal sargassum stranding episodes, i.e., the onshore accumulation and compaction of large amounts of the seaweed. In contrast, the island’s Caribbean shoreline is protected by a mountainous relief, which shields the coastline from sargassum strandings. The sargassum invasion of the Atlantic coastline of Martinique was first noted in 2011, with massive invasions recurring annually or biannually since 2018. This single-center prospective observational study was conducted at the University Hospital of Martinique from 25 January 2016 to 31 July 2020. Pregnant women, residing in Martinique, were recruited and followed during the study period at the University Hospital’s obstetric center. The study population was restricted to women with pregnancy (determined by ultrasound scan). Other inclusion criteria were a minimal age of 18 years and a gestational age of 20 weeks or more. Exclusion criteria were as follows: inability or refusal to provide informed consent for study participation, pregnancy with more than two babies at a time, and women geographically relocating during the pregnancy period. The latter criterion makes it difficult to determine living/working distance from coastal sargassum stranding sites. Study data were retrospectively extracted from the administrative and medical databases of the University Hospital of Martinique using the Emergency DX Care software (DXCare 8.2021, Medasys, Dedalus, France). In addition to each subject’s home and work addresses, the data collected retrospectively for study purposes included the patients’ sociodemographic features (age, sex, body mass index, personal and medical history, as well as pregnancy complications and delivery characteristics). The patient’s status of exposure to sargassum gaseous emissions was determined based on the residential and work addresses provided. Island residents were considered to be exposed to ambient H 2 S emissions from decomposing sargassum if they lived and/or worked in areas along the Atlantic coast of Martinique, which are known to be impacted by sargassum influxes ( Figure S1 ). The levels of exposure were determined on the basis of the measured levels of ambient H 2 S and NH 3 levels in ambient air, which are gauged using a network of 16 ground sensors. These sensors have been deployed along the Atlantic coast of Martinique since 2016 by Martinique’s air quality observatory (Madininair) and certified by the French Ministry of Ecology, Energy, and Sustainable Development ( https://www.madininair.fr/ , accessed on 20 September 2024). Ground sensors are located in sensitive areas such as schools, colleges, health establishments, or densely populated areas close to sites of sargassum strandings. Accordingly, pregnant women identified as exposed to increased ambient H 2 S and NH 3 levels were living and/or working in areas within a distance of <2 km from the nearest ground sensor. For those pregnant women, mean individual H 2 S exposures were approximated by noting the daily concentrations (in ppm) of H 2 S provided by the nearest ground sensor to a patient’s living or working quarters and averaged over the whole pregnancy duration. In contrast, patients living and/or working in the island’s center or along the Caribbean shoreline, in areas located up to 2 km from sargassum stranding sites, were considered to be unexposed to H 2 S originating from decomposing sargassum. This assumption was made on the basis of the virtually nil levels of these gases registered by the mobile sensors deployed in these regions. As such, in light of the lack of sargassum exposure and any other natural, chemical, or industrial source of H 2 S emissions in these areas, populations not residing/working on the Atlantic coast of Martinique were thus considered unexposed. The outcome was hypertensive disorders of pregnancy, which were diagnosed according to the diagnostic criteria of the UK College of Obstetrics and Gynecology [ 35 ]. In brief, hypertensive disorders of pregnancy include gestational hypertension, pre-eclampsia (including chronic (pre-existing) hypertension with superimposed pre-eclampsia), and eclampsia. In this study, a history of chronic (pre-existing) hypertension was not considered as related to pregnancy. Air pollution data, collected by Martinique’s air quality observatory, were also analyzed in order to take into account the potential confounding effect of air pollution on pregnancy outcome. Levels of ozone (O 3 ), nitrogen dioxide (NO 2 ), sulfur dioxide (SO 2 ), particulate matter of ≤10 μm in diameter (PM10), and fine particulate matter of ≤2.5 μm in diameter (PM2.5) were considered over the study period. Crude daily data, originating from the different measure stations in Martinique, were obtained for each pollutant (readings provided by Madininair), and monthly aggregated mean concentrations of each pollutant were computed for the whole island. Computed means were then compared to limit values for the protection of human health, set by the French Ministry of Ecology, Energy, and Sustainable Development. Beyond these threshold values for each pollutant, a health risk was considered plausible. For all descriptive and inferential analyses, the assumption of normal data distribution was analyzed. Mean and 95% confidence intervals were reported for normally distributed variables and median and min-max range for non-normally distributed variables. Categorical variables were presented as absolute values and percentages. The following tests were used for group comparisons: Student’s t -test, chi-squared test, and Fisher’s exact test. The level of statistical significance was set at p < 0.05. Univariate and multivariate logistic regression models were fitted to assess the independent effect of predictors on pregnancy-related hypertension disorders. Variables with significant association in univariate analysis ( p < 0.25) were retained for backward stepwise multivariate regression analysis. Associations were quantified using odds ratio (OR) and 95% confidence intervals. All statistical analyses were conducted using the SPSS software 26.0 for Windows (IBM Corp., Armonk, NY, USA).

Several study limitations are to be noted in our exploratory research. Firstly, exposure to sargassum stranding was approximated by the closest ground H 2 S sensor, which might not reflect actual individual exposure. Unfortunately, a more accurate assessment of individual exposure was not possible with the current study design, notably by the implementation of personal gas detectors or the consideration of a proxy exposure indicator better modelling gas dispersion effects according to distance, landscape, and climatic parameters. Also, the sensor detection of gases emitted by decomposing sargassum seaweed was solely limited to ambient H 2 S and NH 3 , which might not reflect the overall patient exposure to the multitude of toxic gases contained in these emissions, as well as any potential physicochemical interaction between the different gases present. Of note, previous studies have detected heavy metals in sargassum (arsenic (As) and cadmium (Cd)) as well as other potentially toxic elements (lithium (Li), molybdenum (Mo), cesium (Cs), and uranium (U)) that may impact cardiovascular health. Specifically, arsenic exposure has been associated with hypertensive disorders of pregnancy. It is thus expected that, since arsenic is released by sargassum decomposition, chronic arsenic exposure could play a part in an adverse hypertensive outcome in pregnant women. Secondly, while remaining under threshold values defining optimal air quality, the association between air pollution (O 3 , NOx, NO 2 , SO 2 , and fine particle matter PM10 and PM2.5) and hypertensive disorders was not specifically evaluated. Numerous studies have, however, identified investigated the impact of air pollution on human health, which strongly suggest that exposure to fine particles and air pollutant gases is associated with the atherosclerosis process and onset of hypertensive disorders of pregnancy. Sargassum influxes can accompanied by specific climatic and air pollution episodes such as Saharan dust haze. During the study period of January 2016–July 2020, the only alert for Saharan dust haze was recorded in June 2018 and in September 2018, a few months after the mass stranding described that year. Thirdly, the retrospective nature of our study was responsible for the absence of consistent information (medical history) available in the medical patient file to adjust with pre-existing chronic hypertension. We acknowledge that individuals with pre-existing hypertension were at higher risks of developing gestational hypertension and pre-eclampsia during pregnancy. Also, whether women were exposed or not to sargassum emission before pregnancy was not evaluated. The lack of precise residential data did not allow to study the association with a specific sensor data, hence individual concentrations of H 2 S and NH 3 . Fourthly, confounding factors such as those related to sociodemographic and economic levels, housing, or maternal education were not considered. This is attributed to the retrospective nature of our study and the resulting lack of data reliability. In our study, we acknowledge that 49% of the patient’s report having employment, with 90% working near their residence.

In the last decade, massive sargassum seaweed beaching on the coasts of the Caribbean, Central America, and Brazil has become a real threat, causing major socio-economic, ecological, and health problems [ 1 ]. Several hypotheses are possible as to why sargassum has been increasingly washing to shore, including changes in hydrodynamic and wind conditions, increased nutrients from the Amazon due to deforestation and intensification of agriculture in these territories, and global warming of the tropical Atlantic [ 1 , 2 ]. The potential impacts on human health from sargassum influx events include emissions of potentially harmful gases, leaching of heavy metals, and exposure to potentially harmful bacteria and stinging organisms that co-occur with sargassum [ 1 , 3 , 4 , 5 ]. Toxic gas exposure typically happens during decomposition, approximately 48 h after sargassum mats wash ashore. During this process, sargassum releases hydrogen sulfide (H 2 S) gas and ammonia (NH 3 ), which can cause serious health problems, including neurological, digestive, respiratory, and ophthalmologic symptoms in populations exposed throughout weeks [ 1 , 4 ]. While the health effects of long-term and repeated exposure to this gaseous cocktail are largely unknown, previous studies in geothermal areas have observed that H 2 S exposure might increase morbidity for neurological, respiratory, and cardiovascular diseases [ 6 , 7 , 8 ]. Geothermal sources have the advantage over other ambient H 2 S-producing entities of not being known to produce other gases with the potential to confound results. Indeed, geothermal gases are mainly carbon dioxide and water vapor, with only small amounts of hydrogen, nitrogen, methane and carbon monoxide [ 8 ]. Most of the available knowledge about the health effects of chronic exposure to H 2 S comes from research conducted in Rotorua, New Zealand, which is home to the world’s largest community living over an active geothermal field [ 8 ]. In these studies, other possible air pollutants are vehicle emissions, which are limited to the main business street of Rotorua and can be easily identified as a confounding factor, hence allowing to study the health effects of chronic exposure to H 2 S. Of note, sargassum emissions are mainly related to H 2 S production at ranging doses similar to those reported in the Rotorua geothermal field [ 1 , 4 ]. Associations between the hypertensive disorders of pregnancy and gestational air pollution exposure have been previously reported [ 9 , 10 , 11 , 12 , 13 , 14 , 15 ]. Epidemiological studies and meta-analyses [ 11 , 14 ] provide strong evidence that various gaseous and particulate pollutants are associated with hypertensive disorders of pregnancy and pre-eclampsia. The impacts of nitrogen oxides (NO 2 , NO X ) and particulate matter (PM10, PM2.5) on hypertensive disorders of pregnancy were consistently observed [ 9 , 10 , 12 , 13 , 15 ], whereas the implication of ozone (O 3 ) exposure remains inconsistent [ 16 ]. H 2 S is present in the atmosphere as the result of industrial activities and volcanoes and geothermal vents, as well as being released from wetlands, salt marshes, and estuaries, where it is produced by bacteria during the anaerobic decay of organic sulfur compounds [ 8 , 17 ]. Surprisingly, atmospheric H 2 S has not been previously associated with hypertensive disorders of pregnancy. Evidence of both the protective role and deleterious effects of H 2 S in cardiovascular diseases, including hypertension, has been demonstrated [ 18 , 19 , 20 , 21 ]. Many reports have described the beneficial effects of H 2 S on cardiovascular cellular processes, including the modulation of inflammation, improved cell survival, cytoprotection against oxidative stress, as well as positive effects on mitochondrial metabolic function and biogenesis [ 18 , 19 , 20 , 21 ]. As a different concept, cardiovascular H 2 S toxicity has been attributed to the inhibition of cellular enzymes [ 18 , 19 , 20 , 21 , 22 ] and vasoconstriction [ 23 , 24 , 25 , 26 , 27 , 28 ]. The involvement of H 2 S in the pathophysiology of hypertensive disorders of pregnancy and pre-eclampsia is not clear. It has been reported that mRNA and protein expression of the enzyme cystathionine γ-lyase (CSE) are decreased in the placental tissue of pre-eclamptic women [ 29 ]. However, H 2 S plasma levels were found either decreased or increased in women with pre-eclampsia compared to healthy pregnant women [ 30 , 31 ]. To date, only a few studies, mainly from our research group, have objectively reported the clinical symptomatology associated with chronic exposure to sargassum gaseous emissions [ 4 , 32 , 33 ]. In the specific context of human H 2 S exposure to sargassum emissions, we have previously reported that the onset of pre-eclampsia occurred earlier in women living and/or working close coastal sargassum strandings [ 34 ]. Whether day-to-day variations in H 2 S levels are associated with increased risk of hypertensive disorders onset during pregnancy has not been previously investigated. The present study aims to analyze the potential relation between environmental exposure to H 2 S gas emitted by decomposing sargassum and pregnancy-related hypertensive disorders.

The study population was restricted to pregnant women attending the University Hospital of Martinique during the study period (2016–2020). Overall, 3020 eligible pregnant women with a complete dataset and available home/work addresses were included for analysis. The flow chart of inclusion is displayed in Figure 1 . The main clinical and biological characteristics of pregnant women according to pregnancy-induced hypertension disorders are presented in Table 1 . Compared to pregnant women without hypertension disorders, those with hypertension disorders (11.7%) were older and displayed a higher body weight index. Primiparity and primipaternity were more frequent in pregnant women with hypertension disorders compared to normotensive pregnant women. No significant differences were found for active tobacco use. Medical history of chronic hypertension, type 2 diabetes, sickle cell disease, and personal pre-eclampsia items were more frequently reported in pregnant women with hypertension disorders. Term pregnancy was shorter in pregnant women with hypertension disorders and those women were more prone to complications such as gestational diabetes, threat of premature labor, eclampsia, and HELLP syndrome ( Table 1 ). Compared to pregnant women without hypertension disorders, those with hypertension disorders displayed increased blood markers of liver dysfunction, inflammation, and leukocytosis. Newborn weight and Apgar score were also lower in pregnant women with hypertension disorders. Environmental exposure to active sargassum strandings tends to be more frequent in pregnant women with hypertension disorders but failed to reach statistical significance ( p -value = 0.073). The mean concentration of H 2 S over the whole pregnancy was found to be higher in women with hypertension disorders compared to normotensive pregnant women ( Table 1 ). The main characteristics of pregnant women (study population n = 3020) according to sargassum stranding exposure evaluated by the distance between their living/working places and coastline sargassum strandings are presented Table 2 . Compared to unexposed pregnant women (n = 2367), those exposed to sargassum strandings (n = 653) were older, while hypertensive disorders and HELLP syndrome were more frequent ( Table 2 ). On average, the mean H 2 S concentration over the whole period of pregnancy was 0.036 ± 0.236 ppm in exposed subjects. Pregnant women living/working on the Atlantic coast were exposed for 26 days to mean (daily) H 2 S levels of at least 1 ppm. In some areas, patients experienced a maximal number of 12 days of mean (daily) H 2 S concentrations of 5 pm or more. On average, the mean NH 3 concentration over the whole period of pregnancy was 0.052 ± 0.184 ppm in exposed subjects. None of the pregnant women living/working on the Atlantic coast were exposed to NH 3 levels up to 8.3 ppm. Of note, H 2 S and NH 3 recommendations and thresholds from the French High Council for Public Health (Haut Conseil Santé Public ( https://www.madininair.fr/ , accessed on 20 September 2024) are 1 ppm and 8.3 ppm, respectively. The main characteristics of pregnant women with hypertension disorders (n = 351) according to the distance between their living/working places and coastline sargassum strandings are presented in Table 3 . Pregnant women with hypertension disorders exposed to sargassum emission (living/working places and coastline within 2 km of sargassum strandings) have shorter-term pregnancies, gave birth to a newborn of lower weight, and were more prone to HELLP syndrome complication compared to unexposed pregnant women with hypertension disorders ( Table 3 ). Exposed and unexposed pregnant women with hypertension disorders had similar blood marker levels. The univariate and multivariate analyses of risk factors for hypertensive disorders in pregnant women are displayed Table 4 . The multivariate analysis further highlights this increased risk of hypertension disorders in pregnant women within a living/working distance of 2 km from coastline sargassum stranding sites (odds ratio (OR): 1.59 (1.09–2.34) p = 0.017). Significantly increased ORs were also observed for age, BMI, personal history of sickle cell disease, primipaternity, increased weight gain, and gestational diabetes ( Table 4 ). The continuous monitoring of air pollutants (O 3 , NOx, NO 2 , and SO 2 ) and ambient particulate matter during the study period indicated that these pollutants remained within threshold values defining optimal air quality (recommendations of the French Ministry of Ecology, Energy, and Sustainable Development). The only alert levels observed during the study period concerned ambient particulate matter PM10 concentrations in June 2018 and September 2018, distant time periods from the only massive stranding episode of that year (early March 2018). Because there were no differences in O 3 , NOx, NO 2 , SO 2 , and ambient particulate matters between exposed and unexposed participants, air pollution data were not considered in the logistic univariate and multivariate analyses.

The present pioneer work, all while confirming the previously reported general clinical syndrome associated with chronic exposure to sargassum emissions [ 1 , 3 , 4 ], further suggests a potential deleterious effect of these emissions on cardiovascular function in pregnant women. For the first time, we described that chronic exposure to sargassum H 2 S is potentially associated with the risk of hypertension disorders during pregnancy. In pregnant women, sargassum emission exposure was found an independent risk factor of hypertensive disorders in addition to the typical risk factors of hypertensive disorders of pregnancy such as age, body mass index, primipaternity, and gestational diabetes [ 35 ]. The accumulation of sargassum has been increasingly causing environmental and socio-economic challenges in recent years, particularly along the Caribbean coasts [ 1 , 2 , 3 , 4 ]. Despite active collecting, the inundation and compaction of large amounts of sargassum seaweed on shore result in their bacterial putrefaction, leading to the production of nonvolatile and volatile compounds, including hydrogen sulfide (H 2 S) and ammonia (NH 3 ). While the acute toxicities of H 2 S and NH 3 have been well established, the clinical symptomatology associated with chronic exposure to sargassum gaseous emissions have not been extensively studied. In our study, exposure to sargassum emissions was indirectly determined by the distance between their residence and/or workplace and sargassum strandings. Pregnant women identified as exposed to increased ambient H 2 S and NH 3 levels were living and/or working in areas within a distance of <2 km from the nearest ground sensor. H 2 S exposure was based on specific sensors deployed since 2016 along the Atlantic coast in anticipation of massive deposits. It was estimated in our study that pregnant women living/working in areas ≤2 km from sargassum stranding sites were exposed to a mean H 2 S concentration of 0.17 ± 0.49 ppm during their overall pregnancy duration. The observation of a potential higher risk of hypertensive disorders in pregnant women living/working near sargassum strandings is in contradiction with the large body of literature, which rather describes H 2 S as a vasodilation molecule with protective action in systemic arterial hypertension [ 18 , 19 , 20 , 21 ]. At the vascular level, H 2 S can induce endothelium-independent and endothelium-dependent vasorelaxation through mechanisms involving the activation of K ATP channels on vascular smooth muscle cells, voltage-dependent calcium channel inactivation, and release of the endothelium-derived hyperpolarizing factor [ 21 , 22 ]. On the contrary, H 2 S can induce vasoconstriction under certain conditions through mechanisms involving the quenching or inactivation of nitric oxide (NO) and the inhibition of endothelial cell NO synthase. NO-independent mechanisms have also been reported, including changes in calcium and cyclic adenosine monophosphate (cAMP) concentrations in vascular smooth muscle cells and Rho kinase signaling pathway activation [ 20 , 23 , 24 , 25 , 26 , 27 , 36 , 37 ]. Furthermore, it was recently shown that the single electron oxidation of H 2 S by oxyhemoglobin generates a hydrosulfide radical (HS • ), which causes vasoconstriction via an L-type calcium channel-dependent pathway and thus can exert systemic arterial hypertension [ 28 ]. Therefore, the cardiovascular effects of H 2 S are by nature bidirectional. In addition to the vascular dysfunction elicited by H 2 S, previous experimental studies have reported that inhaling H 2 S may induce central metabolic and hemodynamic changes, including reduced energy expenditure and hypothermia, and reduced cardiac output, while blood pressure and stroke volume remained unaffected [ 38 , 39 ]. It is, hence, possible that H 2 S may have limited the physiological increase in cardiac output during the course of pregnancy, which may induce an excessive vasoconstriction and increased blood pressure in exposed pregnant women. Such central hemodynamic maladaptation along with the direct vascular effects of H 2 S could be responsible for the pregnancy hypertensive disorders observed in our study. In our study, standard biochemistry markers, C reactive protein, blood count, and coagulation tests did not show significant differences between pregnant women exposed to sargassum emissions and those without. Due to the retrospective nature of our study, no attempt was made to explore cardiovascular function through cardiac biomarkers such brain natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) and vascular NO biodisponibility through the blood concentration of nitrites/nitrates (NO 2 /NO 3 ). Studies seeking to analyze adverse human health effects associated with chronic exposure to H 2 S have defined low-level exposures as H 2 S concentrations below 0.1 ppm, medium-level ones as between 0.1 and 1 ppm, and high-level exposures as above 1 ppm [ 8 ]. Ambient low H 2 S concentration may be related to a variety of sources including geothermal and volcanic emissions and anthropogenic sources, such as confined animal feeding operations, oil and gas facilities, paper mills, and waste-water plants. At low H 2 S concentrations, chronic exposure has been associated with either weak positive and negative associations with outcomes such as ischemic heart disease mortality, acute myocardial infarction mortality, and hospitalization for cardiovascular diseases. Pregnant women living/working in areas ≤2 km from sargassum stranding sites who developed hypertensive disorders were exposed to a mean H 2 S concentration of 0.17 ± 0.49 ppm, which may be considered a medium level of H 2 S exposure. The deleterious cardiovascular effects of sargassum emissions in our study are consistent with previous reports showing that exposure to low-medium H 2 S levels was associated with increased incidence of cardiovascular diseases, including hypertension and cardiovascular risk factors [ 40 , 41 , 42 , 43 , 44 ].

Our work highlights a potential association between exposure to sargassum emissions and the risk of hypertensive events in pregnant women. The present study results strongly support the deleterious effect on cardiovascular health of pregnant women chronically exposed to sargassum gas emissions. These preliminary findings emphasize the importance of implementing health measures by health authorities to prevent pregnant women from being in areas where sargassum strands ashore.