Subclinical Hypothyroidism in Families Due to the Chronic Consumption of Nitrate-contaminated Water in Rural Areas of Durango, Mexico

Edgar García-Torres Universidad Juárez del Estado de Durango: Universidad Juarez del Estado de Durango Rebeca Pérez-Morales Universidad Juárez del Estado de Durango: Universidad Juarez del Estado de Durango Alberto González-Zamora Universidad Juárez del Estado de Durango: Universidad Juarez del Estado de Durango Esperanza Yasmín Calleros Rincón (  esperanza.calleros@gmail.com ) Universidad Juarez del Estado de Durango https://orcid.org/0000-0001-9341-8366


Background
The main source of potable water, groundwater is a stable and reliable resource that meets the demands of human consumption (Carrard et al. 2019) and a wide range of processes involved in the development and industrialization of urban and rural areas (Velis et al. 2017). However, in recent years, the contamination of these groundwater reserves has increased due to the increased levels of nitrogenated compounds, speci cally nitrate (Shukla & Saxena 2018), resulting from their lixiviation, which is mainly caused by anthropogenic activities and soil use, such as the following: intensive agriculture (Serio et al. 2018); intensive livestock farming (Sahoo et al. 2016); the use of nitrogenated chemical fertilizers (Yu et al. 2020); the use of animal manure as an organic fertilizer (Galindo et al. 2019); the production and disposal of urban wastewater (Shalev et al. 2015); and, the disposal of septic waste (Wang et al. 2017). Said contamination is also affected by the following speci c characteristics of the area: geological characteristics (Wongsanit et al. 2015); pluvial uctuations (Kawagoshi et al. 2019); interaction with surface water (Lasagna et al. 2016a); changes in phreatic levels (Rhymes et al. 2016); and, the presence of biological processes such as denitri cation (Lasagna et al. 2016b). The foregoing result in levels exceeding the permissible limits in water for human consumption, as established by the World Health Organization (WHO), of 50 mg/l nitrate ion, the equivalent of 11.3 mg/l N-NO 3 − (nitrate nitrogen) (WHO 2017).
Nitrate is a nitrogenated compound with a wide natural distribution (Takai 2019) via plants, soil, air, and water (Gassara et al. 2016). Notable among its characteristics are that it is colorless, odorless, tasteless, and highly soluble, making it di cult to detect in water used for human consumption (Almasi et al. 2016). As a result of the action of nitrifying algae and bacteria, nitrite and ammonia are easily converted via oxidation into nitrate, which is the most abundant stable nitrogenated form (Elisante & Muzuka 2016).
Nitrogen is a component of many biomolecules, for which reason it can also be obtained via an endogenous pathway that metabolizes nitric oxide via L-arginine-nitric oxide synthase. In the exogenous pathway, nitrogen is obtained via the diet (Lundberg & Weitzberg 2018), from the consumption of vegetables, green leafy vegetables, processed foods containing nitrogen as an additive, and water contaminated with high nitrate concentrations. Once ingested, nitrogen is absorbed via the gastrointestinal mucosa and the proximal duodenum, reaching a bioavailability of over 90% (Harper et al. 2017). It is then disseminated around the entire body via systemic circulation, with approximately 25% absorbed via salivary gland uptake and then transported to the stomach by means of the enterosalivary circulation. Finally, 65% of the nitrate ingested is eliminated in urine in the form of nitrate, ammonium, or urea (Lundberg and Weitzberg 2017).
The exogenous nitrate-nitrite-NO pathway plays an important role due to its precursor action in the formation of nitrite and, in turn, of the genotoxic N-nitrose compounds present in endogenous nitrosation (Schullehner et al. 2018). Moreover, the ingestion of high nitrate concentrations in water used for human consumption has been found to be related to the following: the appearance of methemoglobinemia A study undertaken on women living in this area found alterations to their metabolic parameters and thyroid hormone pro les, as well as genotoxic damage and an increased incidence of SCH (Gandarilla et al., in review). Furthermore, inhabitants report more than one member of their family experiencing thyroid hormone alterations, as detected in the region's primary care health centers. In light of the foregoing, the present study sought to analyze the cases identi ed in the abovementioned study and broaden the research to the majority of the family members of the index case in order to identify the presence of more cases in the family. Moreover, the study aimed to identify the environmental, genetic, and immunological factors that are present in these populations and may explain the increased incidence of SCH.

Study population
A familial, observational, descriptive, and cross-sectional study was conducted on individuals previously identi ed with thyroid dysfunction and residing in rural areas of the municipality of Lerdo, Durango. The study population comprised families resident in properties that form the villa (village) Juan E. García, the ejido (communally-held indigenous land) San Jacinto, the ejido 21 de marzo, the villa La Loma, and the villa Nazareno. The sample comprised 102 subjects forming part of 26 families resident in the abovementioned communities. The inclusion criteria were the following: voluntary participation; signed informed consent; any sex; any age; more than a year's residence in the locations of interest; consumption of water from the municipal network; and, not receiving medication with either anti-thyroid effects or compounds containing nitrate salts. All those ful lling the inclusion criteria were tested for exposure biomarkers (plasma and urinary nitrite), an effect biomarker (percentage of methemoglobin), and thyroid function (thyroid pro le). The genotypes corresponding to the single nucleotide polymorphisms (SNPs) rs965513 and rs1867277 of the FOXE1 gene were also determined. The presence of anti-thyroid peroxidase antibody (TPO-Ab) was determined in those subjects who were found to have SCH. The variables of interest comprised the subject's age, sex, body mass index (BMI), and thyroid pro le, as well as the exposure and effect biomarkers, and the levels of nitrate exposure via drinking water.

Biological sample collection
Study participants were asked to fast for eight hours prior to the collection of a 10 ml peripheral blood sample and a 10 ml sample of the rst urine void in the morning, with both samples collected in sterile containers. The blood sample collection was conducted using BD Vacutainer® tubes containing 6 ml coagulation activator for the recovery of the serum to be used for establishing both, the thyroid pro le and TPO-Ab levels. Tubes containing 4 ml ethylenediamine tetraacetic acid (EDTA) were also used to measure plasma nitrite and methemoglobin levels and for the puri cation of genomic DNA. Once obtained, the samples were kept at -70ºC until processing.

Nitrate exposure via drinking water
Once the nitrate levels in the water for human consumption had been established, the study locations were then strati ed according to their exposure level. A low level was classi ed as 4.7 ± 3.3 mg/l (ejido 21 de marzo), while a medium level was classi ed as 32.1 ± 3.7 mg/l (ejido San Jacinto and Juan E. García) and a high level as 56.9 ± 14.7 mg/l (La Loma and Nazareno).

Measurement of methemoglobin percentage
The percentage of MetHb was established via the spectrophotometry method reported by Sakata et al. (1982), using 100 µl of the blood collected in the EDTA tubes and following the manufacturer's instructions (FAR Diagnostics Verona, Italy) for measurements conducted via four readings at 630 nm (HACH-DR5000). The rst reading (D1) is a hemolyzed preparation and the second reading (D2) is conducted via the addition of sodium azide, while the third (D3) corresponds to a hemolyzed preparation to which potassium ferricyanide has been added, to which, in turn, sodium azide is nally added for the fourth reading (D4). The percentage of methahemoglobin is calculated via the following equation: % x 100 with values of over 1.5% total methahemoglobin considered to be above the reference values. The reagents used complied with high purity standards (Merck KGaA, Darmstadt, Germany and/or its a liates).

Nitrite measurement
The nitrite concentration was determined, as a stable metabolite, for the plasma and urine samples, which were then deproteinized using absolute ethanol at a 1:7 ratio and centrifuged at 3500 rpm for 15 minutes. The nitrite levels in the plasma and urine were measured by subjecting 100 µl of supernatant to the Griess colorimetric method (Merck KGaA, Darmstadt, Germany and/or its a liates), as reported by Miranda et al. 2001, using spectrophotometric quanti cation (HACH-DR5000) at 540 nm. A standard curve was generated using sodium nitrite for the quantitative determination.

Measurement of TPO-Ab concentration
For those individuals presenting TSH levels above the normal range, 100µl blood serum was used to conduct a sandwich-type enzyme-linked immunosorbent assay (ELISA), using a commercial TPO-Ab detection kit in accordance with the manufacturer's instructions (Merck KGaA, Darmstadt, Germany and/or its a liates). The levels of TPO-Ab + were determined once values of > 35 UI/ml were observed.

DNA extraction and genotypi cation of FOXE1
Genomic DNA extraction was conducted using whole blood following the protocol described by Lahiri

Statistical analysis
Measures of central tendency were analyzed for the description of the data and Pearson's correlation for quantitative variables with normal distribution. The information for 16 variables was compiled for 102 individuals, with the database including numerical and categorical variables. The variables were standardized by subtracting the average corresponding to the variable from the value of each data point and then dividing by the standard deviation. The multivariate analysis comprised a correlation analysis applied solely for the 11 numerical variables, given that principal component analysis (PCA) requires that the variables used present some level of correlation. For this reason, only those variables with a level of signi cance greater or equal to 0.90 were considered. Once the variables presenting a correlation had been identi ed, a Mahalanobis distance analysis was undertaken in order to discard the data considered atypical for this set of variables. The PCA was carried out to show, rstly, the distribution of the individuals based on their subclinical condition and, secondly, on their nitrate exposure according to the levels detected in the communities in which they live. A heatmap analysis was carried out in order to group the individuals according to the 11 numerical variables and, at the same time, show the levels of correlation of each of the variables for each individual. The cluster analysis was undertaken using the Euclidean distance index and the Ward algorithm in order to construct the groupings, while genotype, subclinical condition, sex, and the N-NO 3 − level to which they were exposed were mapped for each individual in the heatmap. Moreover, in order to reassign those individuals who had been classi ed with a non-hypothyroid subclinical condition prior to the analysis, a discriminant function analysis was conducted to establish the percentage of correct classi cations, which was then statistically corroborated via the Wilkins test. Finally, post hoc and chi-squared Kruskal-Wallis tests were performed to determine the differences between the different exposure scenarios, where applicable. All statistical and multivariate analysis was undertaken using the R 6.6.1 software, a p value of < 0.05 was considered as statistically signi cant.

Ethical considerations
This study has been approved by the Ethics Committee of the Faculty of Chemical Sciences at the Juarez University of the State of Durango, with a unique assigned registration number of R-2017-123301538X0201-026. The present study followed the guidelines, as pertaining to health research, stipulated in articles 13, 14, 17, 21, and 22 of the General Health Law in Mexico.

Results
The study population initially comprised 26 female subjects, who had been previously diagnosed with SCH and who described other cases in their families. The analysis was conducted on the highest number of the subject's family members possible, in order to understand the potential causes behind the increased incidence of SCH in these rural populations, which had been exposed to a possible endocrine disruptor.
The total sample comprised 102 people with an average age of 22 years, 69% of whom were women, 15% had no schooling and 36% had been educated up to lower secondary level, while 58% were overweight or presented some degree of obesity and the majority worked in the home or were students. With regard to water consumption, 85.5 % consumed water drawn from the public water system or from private wells, while only 14.5 % occasionally consumed water from water puri cation machines. The measurement of MetHb found 79% of subjects to fall outside the reference limit, while the average levels of nitrite in plasma and nitrite in urine were 24.62 µmol/mL and 3.46 µmol/mL, respectively. In terms of the thyroid function parameters, 45% of subjects presented TSH levels over the reference limits (SCH), 49% presented reduced total T3 levels, and 19.6% presented a reduced total T4 levels. The detection of TPO-Ab found that 15.7% of the SCH cases were positive for the presence of antibodies.  Table 3 presents the genotypic and allelic distribution of the SNPs rs965513 and rs1867277 for each of the levels established by the present study for nitrate exposure in water used for human consumption.
The SNP rs1867277 was not found in Hardy-Weinberg equilibrium (Ji 2 = 0.012) and its genotypic distribution across the exposure scenarios was found to be different (Ji 2 = 0.019).  Table 4 presents a summary of the alterations observed in the following variables, in light of the different level of chronic exposure to nitrates in the water used for human consumption: exposure and effect biomarkers for nitrates (via the measurement of nitrites); thyroid dysfunction (SCH); the genotypic frequency of the SNPs; and, the presence of TPO as an autoimmune factor. Differences were observed in the levels of both nitrates and the exposure (nitrite in urine p = 0.001) and effect (methemoglobin p = 0.042) biomarkers, while the development of SCH was observed for all levels of chronic exposure (Ji 2 = 0.817). The SNP rs1867277 only presented in the low exposure scenario, corresponding to 4% (Ji 2 = 0.043), and, nally, the TPO-Ab percentage was highest in the medium exposure scenario, with 14.2%. and age and T3F (r= -043, r 2 = 18.49, p = 0.001). Analysis of the correlation between the effect and exposure biomarkers revealed no statistically signi cant correlation, while a weak negative correlation was observed between the effect biomarkers (MetHb) and thyroid pro le. However, a statistically signi cant correlation was found between the MetHb percentage and the levels of T3T and T3F (r= -0.43, r 2 = 18.49, p = 0.001, and r= -0.34, r 2 = 11.56, p = 0.001, respectively). A statistical signi cance was observed for the level of nitrite in the blood and its correlation with the thyroid hormones, but only for T4F and T3T (r= -0.20, r 2 = 4, and p = 0.045, and r = 0.23, r 2 = 5.29, and p = 0.022, respectively). In terms of the correlation between BMI and thyroid hormone levels, a negative relationship was found across almost the entire thyroid pro le with the exception of TSH, while only a weak negative, but statistically signi cant, correlation was also found between BMI and T3F (r= -0.31, r 2 = 0.096, p = 0.01).
The PCA revealed that the three rst components represent 63.3% of the variability, wherein the variables that most contributed to the analysis of the rst component were T3T, T3F, and T4T, while the main variables in the second component were age and BMI [ Figure 2A near here]. The majority of the individuals presenting a SHC condition were identi ed by increases in the NO 2 − levels in the blood, BMI, age, TSH level, and MetHb levels, while a greater relationship was found between those individuals with a non-SCH clinical condition and increases in the variables T3T, T4T, T3F, and, mainly, T4F. When the nitrate exposure scenarios were classi ed, the PCA analysis did not identify a clear condition, but did establish that high levels of MetHb are more related to those study locations presenting high levels of nitrate in potable water.
Additionally, both multivariate conglomerate and heatmap analysis was undertaken to group subjects according to the numerical variables and, at the same time, present the levels of correlation for each of the variables for each individual. The conglomerate analysis was carried out using the using the Euclidean distance index and the Ward algorithm to construct the groupings, while the heatmap analysis mapped the subject's genotype and subclinical condition and the level of N-NO 3 − to which they had been exposed.
Finally, the discriminant function analysis revealed 81% accuracy in the classi cation of the subjects.

Discussion
The present study is the rst to seek evidence as to the possible direct relationship between chronic nitrate exposure via the consumption of contaminated water and alterations in thyroid function in rural areas of Durango.
As widely described in detail in the literature, human exposure to nitrates and nitrites occurs endogenously due to the metabolism of NO (Lundberg, Weitzberg & Gladwin 2008  The lactating population is more prone to developing this condition as a result of the low level of methemoglobin reductase activity, while the ease with which fetal hemoglobin is oxidized and its more acidic gastric pH enable the intestinal microbiota to reduce the amount of nitrate ingested into nitrite. However, a notable correlation is not observed between the levels of nitrites in the blood and urine, while the high MetHb percentages observed is notable, wherein 79% of individuals analyzed presented levels over 1.5%, with an average of 2.80 ± 1.88. Methemoglobinemia is a condition characterized by an abnormally high level (> 40%) of MetHb, which is mainly produced by acute exposure to oxidizing compounds, pharmaceuticals, and most chemical agents (Alanazi 2017). However, in the present study, the highest MetHb percentage, of 12.35%, was obtained from a study population which indicated, via the questionnaire, to having had no prior contact with any medication or chemical agent that could explain these high MetHb levels.
Another health aspect of interest to the present study is the state of thyroid function in the presence of chronic nitrate exposure. Nitrate, together with other sodium-iodide symporter inhibitors, is able to alter carcinoma. Moreover, while do not they concur with the reduced TSH levels and increased thyroid hormone levels found by the foregoing studies, they do show altered thyroid function after chronic nitrate exposure via potable water, ranging from 4.7 ± 3.3 mg/l to 56.9 ± 14.7 mg/l. Said alteration found in the present study corresponds to SCH and occurs due to the increased levels of TSH (up to 45%) and a reduction in T4T and T3T (up to 49% and 19.6%, respectively), a nding concurring with that reported by While previous studies have indicated the relationship between high levels of nitrates in potable water and altered thyroid function, others have shown that said relationship is not completely established and have not found structural or functional changes in the thyroid gland (Hunault et al. 2007). In contrast, the present study identi ed nitrate concentrations in potable water of up to 56.9 ± 14.7 mg/l, which is considered to be chronic exposure (≥ 1 year), further to identifying the presence of other conditioning factors for the development of SCH.
The prevalence of SCH found by the present study was 45%, which exceeds the national and global prevalence of 8% and 10%, respectively (Bruneel et   Increased body weight is one correlation plausibly connected to the presence of SCH, in that hypothyroidism and SCH provide the conditions for reduced metabolic function, alterations in thermogenesis, and processes involving the metabolism of both glucose and lipids. This translates into weight gain over the long-term, once the level of thyroid hormone synthesis begins to decrease (Sanyal & Raychaudhuri 2016). Given the foregoing, it is possible that the correlation between age and BMI found in the present study may reveal that the older a subject, the greater their exposure to nitrates and, consequently, the greater the alteration in metabolic processes, thus resulting in weight gain. However, we also consider that said correlation would be affected by other factors such as nutritional variation, sedentarism, and, even, the inherent relationship with growth, without underplaying the importance of said correlation in itself.
It should be noted that some studies indicate a relationship between increased BMI and increased thyroid In terms of the relationship observed between age and thyroid pro le, the results of the present study provide evidence of a negative and statistically signi cant correlation. However, the production of hormones regulated by the endocrine system, including the thyroid hormones, decreases due to aging, which causes general morphological and physiological changes (Barbesino 2019). Nevertheless, the importance of the chronic presence of nitrate as an inhibitor of iodine uptake in the thyroid gland cannot be understated.
It is of the utmost importance to highlight that the following environmental characteristics of the rural region in which the present study was carried out are a problem with real impacts on human health: intermittent rainfall; soil/aquifer contamination via inorganic nitrate and other compounds; overfertilization; over-use of land for forage crops; and, intensive livestock and agricultural practices. Not only are said problems present in rural zones of Durango, Mexico, but are also, today, without a doubt, a public health and environmental problem in large parts of the world (Shukla & Saxena, 2019).
As discussed above, the development of SCH may have a multifactorial origin (Ibañez, 2017). However, Table 4 clearly shows that SCH presents in each of the exposure scenarios, thus suggesting that, further to the high nitrate levels, it is chronic nitrate exposure that provides the conditions for alterations in thyroid function.

Conclusions
The results of the present study suggest that environmental factors have a direct and signi cant in uence, via chronic exposure to nitrate in potable water, on thyroid dysfunction in general and speci cally the disruption of the synthesis and secretion of thyroid hormones. Therefore, the high frequency of SCH in the study population is mainly due to the nitrate contamination of potable water and not as a response to an additive effect and other causal factors.
The present study draws attention to the nitrate concentrations that may be found in potable water and suggests that exposing individuals to such concentrations during critical periods of their development may have signi cant long-term consequences. There is an evident need to ascertain the role played by nitrate as an endocrine disruptor via potable water. Our ndings contribute to the implementation of public health measures aiming to ensure access to water resources of adequate quality, further to raising awareness in the rural population of Durango as to the dangers of consuming water with high nitrate content.

Declarations
Funding: The present study was funded by the Consejo de Ciencia y Tecnologia del Estado de Durango Con icts of interest/Competing interests: The authors declare no type of con icts of interest.
Availability of data and material: The data that support the ndings of this study are available from the corresponding author upon reasonable request.   Methaemoglobin; NO2-, Nitrite. The cluster analysis was performed using the Euclidean distance index and using Ward's algorithm to construct the clusters; n the Heatmap analysis, the genotype presented by each individual, the subclinical condition, and the level of N-NO3-to which they are exposed were mapped.