Journal of biological and health sciences http://biotecnia.unison.mx
Universidad de Sonora
ISSN: 1665-1456
Original Article
1 Universidad de Sonora, Departamento de Medicina y Ciencias de la Salud. Hermosillo, México.
2 Instituto Tecnológico de Sonora, Departamento de Ciencias del Agua y Medio Ambiente. Ciudad Obregón, México.
3 Universidad de Sonora, Departamento de Matemáticas. Hermosillo, México.
⁴ El Colegio de Sonora, Centro de Estudios en Salud y Sociedad. Hermosillo, México.
Determinantes sociales y ambientales de la fiebre manchada por Rickettsia rickettsii a nivel censal en las ciudades de Hermosillo y Obregón en Sonora, México
Rocky Mountain Spotted Fever (RMSF) is a disease caused by the Rickettsia rickettsii bacteria, transmitted by the bite of infected Ixodidae hard ticks. In 2023 the cumulative inci- dence (40.1 per million population) and the case fatality rate (55.4 %) of RMSF presented in Sonora, led national statistics. Objective. To better understand the geographic distribution of the disease in two endemic cities in Mexico, we used spa- tial analysis techniques to estimate the correlation between physical and social determinants with the geographic distri- bution of RMSF incidence at the census-tract level. Methods. An ecological design was carried out to study the correlation between the Land Surface Temperature (LST) and the Index of Social Marginalization (ISM) with the distribution of the RMSF incidence in Hermosillo and Ciudad Obregón, Sonora, which are the cities with the largest burden of the disease within Sonora. Results. Overall, a heterogeneous spatial dis- tribution of RMSF with hot spots randomly dispersed across both cities was observed. Furthermore, there were hot spots characterized by a high incidence of RMSF, a high ISM, and a LST above 31.6 °C. Conclusions. Our findings show that the spatial analysis of physical and social determinants can be used to identify hot spots of RMSF at the census-tract level to which specific measures of prevention and control can be directed.
La fiebre manchada por Rickettsia rickettsii (FMRR) es una enfermedad zoonótica transmitida a los humanos por la mordedura de garrapatas duras infectadas de la familia Ixodi- dae. En 2023 la incidencia acumulada (40.1 por millón de ha- bitantes) y la tasa de letalidad (55.4 %) de FMMR de Sonora, lideraron las estadísticas nacionales. Objetivo. Para entender mejor su distribución espacial en dos ciudades endémicas de México, usamos técnicas del análisis espacial para estudiar
*Author for correspondence: Gerardo Álvarez-Hernández e-mail: gerardo.alvarez@unison.mx
Received: February 20, 2024
Accepted: February 13, 2025
Published: March 7, 2025
la correlación entre determinantes físicos y sociales con la incidencia de FMRR a nivel de área censal. Métodos. Se llevó a cabo un diseño ecológico para estimar la correlación entre la temperatura de la superficie terrestre (TST) y el índice de marginación social (IMS), con la incidencia de FMRR en Her- mosillo y Ciudad Obregón, que son las ciudades con la mayor carga de la enfermedad en Sonora. Resultados. En general, se observó una distribución espacial heterogénea de FMRR con puntos calientes dispersos aleatoriamente en ambas ciuda- des. Además, hubo puntos calientes caracterizados por una alta incidencia de FMRR, un IMS alto y una temperatura de la superficie terrestre superior a 31.6° C. Nuestros hallazgos muestran que el análisis espacial de determinantes físicos y sociales puede utilizarse para identificar zonas críticas de FMRR en áreas urbanas de pequeño tamaño a las que se pueden dirigir medidas específicas de prevención y control. Palabras clave: Fiebre manchada por Rickettsia rickettsii; análisis espacial; México
Rocky Mountain Spotted Fever (RMSF) is a disease caused by the Rickettsia rickettsii bacterium, transmitted in Mexico mainly through the bite of infected Rhipicephalus sanguineus hard ticks, a common parasite of domestic dogs (Biggs et al., 2016). Although RMSF can be prevented and treated, it rema- ins as the most lethal of the spotted fever group rickettsioses (Biggs et al., 2016; Álvarez-Hernández et al., 2017; Treadwell et al., 2000).
Multiple biological, environmental and social deter- minants can be associated with the burden of the disease, despite understimation of its actual magnitude and impact (Álvarez-Hernández et al., 2024a) . In Mexico the most afflic- ted are biological and socially disadvantaged groups, such as children younger than 10 years, older adults, and people living in poverty (Álvarez-Hernández et al., 2017; Álvarez- Hernández et al., 2024b). Despite uncertainty on the causal mechanism, factors such as socioeconomic status, individual
Volume XXVII
DOI: 10.18633/biotecnia.v27.2262
behaviors, and interaction between human and host reser- voirs (i.e. dogs), among others, have been associated with heavy exposure to ticks and increased risk of being infected (Parola et al., 2008; Atkinson et al., 2012; Tomassone et al., 2018). Moreover, variability of air temperature and relative humidity have an impact on the density and spatial distribu- tion of hard ticks (Dantas-Torres, 2010; Ogden 2021; Nutall, 2022).
In Mexico, the annual cumulative incidence (CI) of RMSF fluctuates between 3.2 and 9.8/1 000 000 inhabitants (Álvarez-Hernández et al., 2024b). Sonora is a northwestern Mexican state, with an upward trend of morbidity and mortality of the disease since its reemergence in the early 2000’s. Here, the incidence of RMSF is 26 times higher than the national average, with case-fatality rates (CFR) above 30 %, mainly in children and socially underserved groups (Álvarez-Hernández et al., 2017; Álvarez-Hernández et al., 2024). Within the state of Sonora, the cities of Hermosillo and Obregon together account for 53 % of the annual volume of cases (Álvarez-López et al., 2021).
Overall, there is little evidence about the potential link between temperature and the geographic distribution of RMSF at the small-area level in urban settings. In Mexico, scarcity of data from meteorological stations in urban areas and limited access to laboratory techniques to confirm cases hamper real-time studies. However, by using Land Surface Temperature (LST) as a proxy for air temperature variations, can serve to approaching variability of environmental tem- perature, which is important because survival of ticks may be influenced by the formation of microclimates within a city (Suss et al., 2008); therefore, LST can help in identifying areas that can be suitable for the abundance of R. sanguineus ticks. The aim of this research was to describe the spatial distribution of RMSF at the census tract level in two cities in the state of Sonora, Mexico, by examining the correla- tion between the CI of the disease with an index of social marginalization (ISM), and LST. The study of the geographic distribution of RMSF at the small area level can provide un- derstanding to design effective public health interventions
to deal with the problem.
An ecological study was carried out in Hermosillo and Obre- gon (see Figure 1a), the most populated cities in the state of Sonora, Mexico. Both cities comprise the highest incidence of RMSF within the state (Álvarez-López et al., 2021). Hermo- sillo is the state capital and had a population estimated to be 936,263 in the year 2020, and 530 urban census-tracts, known in Mexico as Basic Geostatistical Areas (BGA) [INEGI, 2020], which is the unit of analysis of this study. Its average annual air temperature is 24.9° C (17.4º C-34.0º C) [Navarro et al., 2018]. On the other hand, Obregon had an estimated po- pulation of 436,484 and 244 urban census tract [INEGI, 2020]. It has an average annual air temperature of 25.1° C (17.6° C-33.9° C) [Navarro et al., 2018]. The annual precipitation in
both cities depends on the influence of the North American monsoon which generates around 70 % of the rainfall during the months of July - September. In the urban area, the his- torical precipitation in Hermosillo is 356 mm per year and in Obregon is 386 mm per year [Navarro-Estupiñán et al., 2018].
The socioeconomic status of each census tract is based on ISM, a population indicator widely used in Mexico to cate- gorize the social welfare of a defined population. The ISM was from the year 2020; it is based on nine socioeconomic varia- bles, three of them hypothetically related with tick exposure (i.e., percentage of illiterate adult population, proportion of homes having dirt floors, and proportion of employed popu- lation earning up to two minimum wages). Statistically, the ISM is built on a principal components technique to classify into five strata the degree of social marginalization: very high (worst level), high, intermediate, low, and very low (best le- vel). Each census tract has an ISM that allows the stratification into one of the five strata (Gobierno de México, 2020). Thus, Hermosillo grouped 68 % and Obregon 73 % of their census tracts in the intermediate, low, and very low strata. See figure 1b-c.
All cases of RMSF included in the study corresponded to the period 2006-2018. Each case is systematically registered in an electronic database from the Sonora Public Health Department, from which they were retrieved. A confirmed case of RMSF, code A77.0 of the 10th Rev of the International Classification of Diseases (WHO, 2019), was as a patient with two or more acute clinical manifestations (≤7 days) including fever, headache, malaise, rash, diarrhea, vomit, and history of tick exposure with a single positive confirmatory test, which included detection of R. rickettsii DNA using real-time or end- point polymerase chain reaction (PCR) from a whole blood specimen. To identify the Rickettsia genus, the gltA gene was used with an amplification product of 401 base pairs (Labru- na et al., 2004). For the rickettsii species, the A1G_04230 gene was identified, which codes for the hypothetical protein through an amplification product of 153 base pairs (Kato et al., 2013). We also confirmed a case in clinically suspected patients with detection of antibodies IgG (>1:64) specific against R. rickettsii in an Indirect Immunofluorescence Assay (IFA) from a single blood sample. All the diagnostic tests were performed in the Sonora State Laboratory of Public Health. Only cases with a known address were included.
The study database included 428 cases, 202 (48.1 %) from Hermosillo, and 226 (51.9 %) from Obregon, who were linked geographically to their corresponding census tract. By using Google Earth®, we retrieved the street, number, inter- sections, and neighborhood of each case and then exported to a Geographic Information System (GIS) without identi- fying any of the subjects’ addresses to safeguard privacy and confidentiality. All study procedures were approved by the Bioethics Committee of the Department of Medicine and Health Sciences, University of Sonora (D-124/24.08.2019).
Figure 1. Official land use map (INEGI, 2010) at a 30-m resolution using the 2010 city limits. (a) Location of Hermosillo and Obregon. Index of Social Marginalization by census-tract in Hermosillo (b) and Obregon (c). 304x228mm (300 x 300 DPI)
Figura 1. Mapa oficial de uso del suelo (INEGI, 2010) con una resolución de 30 m utilizando los límites de la ciudad de 2010. (a) Ubicación de Hermosillo y Obregón. Índice de Marginación Social por área censal en Hermosillo (b) y Obregón (c). 304x228mm (300 x 300 DPI)
It has been referred that because of the scarcity of climate stations within the city limits, many cities are unable to generate spatial maps and temporal resolutions about the variability of air temperature (Tsou et al., 2017; Yang et al., 2017). Thus, satellites can measure the radiation coming from the Earth-Ocean-Earth-Atmosphere systems, including the solar radiation reflected and the infrared radiation emitted, which can be used as a proxy for air temperature variations (Sorek-Hamer et al., 2016).
In our study, data was obtained through the thermal in- frared sensor (TIRS) of the satellites Landsat 5 and Landsat 8; remote sensing images were used to estimate the Land Sur- face Temperature (LST). Three thermal bands at a resolution of 30 meters were used, with images available approximately every 16 days at 11:00 am local time (Navarro-Estupiñán et al., 2018). To calculate the LST, data from the Landsat 5 were used for the period 2006 - 2012, while the Landsat 8 images were utilized for the period from 2013 - 2018. A total of 428 satellite images were downloaded, from which 100 were eliminated due to cloud cover (Figure 2).
We described the geographic distribution of cases by gen- der, age, and the census tract assigned category of social
marginalization. The georeferenced census tracts were the basis for depicting the spatial distribution of cases by using the point density tool in ArcGIS TM (ESRI, 2011). The statistical differences in categorical variables between the cities were examined using the chi-squared test; a two-tailed hypothesis was used, and p values < 0.05 were considered statistically significant. We utilize the StataSE 12.0 TM software to conduct the statistical analysis.
Cumulative incidence (CI) of RMSF per 10,000 inhabi- tants was calculated for each census tract; because of hete- rogeneity in the population size of the census tract, the CI of RMSF was smoothed through the Bayes empirical method (Besag et al., 1991) by using the GeoDa ™ software version
1.14.0. Once the CI was smoothed, maps about its geogra- phic distribution were created for both cities using ArcGIS™ version 10.3.
The Moran’s I statistic measured the global spatial correla- tion between the cumulative incidence of the disease and the ISM. The observed Moran’s I value ranges from -1 to +1. For interpretation purposes, a value close to +1 indicates a strong positive spatial correlation, meaning a clustering of cases in a geographic area of high social marginalization and high rates of the event; a value close to -1 indicates strong
Figure 2. Annual average of spatial distributions of land surface temperature (LST) from 2006-2018 for: (a) Hermosillo and (b) Obregon. Figura 2. Promedio anual de distribuciones espaciales de la temperatura superficial terrestre (TST) de 2006 a 2018 para: (a) Hermosillo y (b) Obregón.
negative spatial correlation, that is, a dispersion pattern; and a value of zero implies a random pattern (Siabato y Guzmán- Manrique, 2019). The Local Indicators of Spatial Association (LISA) method was used to identify statistically significant local clusters and local spatial outliers (Celemín, 2009). Maps were obtained through the GeoDa™ software version 1.14.0. To assess the spatial correlation between the CI of RMSF,
ISM and LST, each variable was normalized into a common scale (0 to 1). In Hermosillo, average LST higher than 31.6° C received a value of 1, while a CI ˃ than 3.8 cases per 10,000 inhabitants (80th percentile) was considered to have a value of 1; for Obregon, LST higher than 31.8 ° C and a CI > 8.16 cases per 10 000 inhabitants corresponded to a value of 1. Furthermore, each stratum of the ISM was assigned with a value ranged from 1 (very high social marginalization) up to
0.2 (very low)
Lastly, the Getis-Ord (Gi*) method was applied to find statistically significant “hot” and “cold” spots in the heat risk map. Gi* is a Z-score, with large values implying a more inten- se clustering of high values and negative values mean cold spots as shown in the equation:
Where Gi* (d) is the local G statistic for a feature i within a distance (d), and Wij (d) is the spatial weight for the target- neighbor i and j pair (Peeters et al., 2015). Ord and Getis (Ord y Getis, 1995) developed a z-transformed form of Gi* to improve the statistical testing. The statistical significance and degree of clustering is evaluated according to the confidence level and Z-scores. If Z(G *) has a positive and significant value, it means that the pixel has a relatively high probability of being a hot spot area. Otherwise, if Z(G*) has a negative and significant value, the pixel has a high probability of being a cold spot area (Ord y Getis, 1995).
i
i
The study sample consisted of 428 cases of RMSF, 200 (46.7
%) of them were confirmed through PCR, 104 (24.3 %) by detection of IgG ≥ 1:64; in 84 (19.6 %) patients both IgM and IgG titles ≥ 1:64 were found, while in 40 (9.4 %) only IgM was detected. The overall cumulative incidence (CI) for both cities was 4.7 per 10,000 inhabitants, while in Hermosillo was of 3.5 cases per 10,000, and for the city of Obregon was of 6.9 per 10,000. The observed case fatality rate (CFR) was of 18.0 %
G* (d)=
𝑛𝑛
∑
𝑗𝑗=1
𝑊𝑊
𝑖𝑖𝑗𝑗
(𝑑𝑑)𝑥𝑥
𝑗𝑗
(1)
(77/428), 56 (CFR 29.1 %) of the deaths occurred in Hermosi- llo, 21 (CFR 9.8 %) took place in Obregon. Ninety-five percent
∑
𝑛𝑛
i 𝑗𝑗=1
𝑥𝑥
𝑗𝑗
(407) of the subjects were linked geographically to their
census tract of residence. Out of these, 44.8 % (192) resided in Hermosillo and 50.2 % (215) in Obregon.
A higher proportion (59.5 %) of the cases were women, even more (67.9 %) in Ciudad Obregon than in Hermosillo (50.0 %) (p < 0.001). The cases were more frequent (167; 41.0
%) among children and adolescents, followed by adults aged 30 - 49 years (128; 31.5 %). A significant difference (p = 0.004) in the age distribution of the cases was observed. Regarding the degree of social marginalization, the highest proportion (33.7 %) of cases corresponded to the intermediate stratum, while only 13.3 % was into the strata of high and very high social marginalization, such differences were significant (p = 0.008) (Table I).
On the other hand, the census tract having a high de- gree of social marginalization had the highest smoothed CI of RMSF, followed by those with very high degree of social marginalization. This pattern was observed for both localities, despite the largest number of census tracts corresponded to neighborhoods with low (n = 130) or very low (n = 259) level of social marginalization (Table 2).
Tabla I. Características seleccionadas de casos incidentes de fiebre manchada por Rickettsia rickettsii. Sonora, México. 2006-2018-
Table I. Selected characteristics of incident cases of Rocky Mountain spotted
fever. Sonora, Mexico. 2006-2018
Variable | N (%) Hermosillo Obregon Total (n = 192) (n = 215) (n = 407) | P* | ||
Sex * | ||||
̶ Male | 96 (50.0) | 69 (32.1) | 165 (40.5) | |
̶ Female | 96 (50.0) | 146 (67.9) | 242 (59.5) | 0.000 |
Age group (years)* | ||||
< 10 | 41 (21.4) | 31 (14.2) | 72 (17.7) | |
10 - 19 | 52 (27.1) | 43 (20.0) | 95 (23.3) | |
20 - 29 | 27 (14.1) | 30 (13.9) | 57 (14.0) | |
0.040 | ||||
30 - 49 | 49 (25.5) | 79 (36.7) | 128 (31.5) | |
50 - 59 | 13 (6.8) | 24 (11.2) | 37 (9.1) | |
> 60 | 10 (5.2) | 8 (3.7) | 18 (4.4) | |
Degree of social | ||||
marginalization a* | ||||
̶ Very high | 7 (3.7) | 4 (1.9) | 11 (2.7) | |
̶ High | 30 (15.6) | 13 (6.0) | 43 (10.6) | |
̶ Intermediate | 67 (34.9) | 70 (32.6) | 137 (33.7) | 0.008 |
̶ Low | 47 (24.5) | 67 (31.2) | 114 (28.0) | |
̶ Very low | 41 (21.4) | 61 (28.4) | 102 (25.0) | |
a Based-on data from the National Population Council 2015. The Index of social marginalization is composed of nine variables: % of illiterate population >15-years-old; % of population >15-years-old without elementary; education; % of occupants in homes without drainage or toilets; % of occupants in homes without electricity;
% of occupants in homes without piped water; % of homes with certain level of overcrowding; % of occupants in homes with dirt floors; % of population in towns with less than 5000 inhabitants; and % of employed population earning up to two minimum wages.
*Based on a χ2 statistic
Tabla II. Incidencia de fiebre manchada por Rickettsia rickettsii por índice de marginación social a nivel de área censal de las localidades de estudio. Sonora, México. 2006-2018.
Table II. Incidence of RMSF by index of social marginalization according to
census tract in study localities. Sonora, Mexico. 2006-2018.
Variable | Hermos | illo | Obregon | |
Incidence of RMSF * | IQR | Incidence of RMSF * | IQR | |
Degree of social marginalization a | ||||
̶ Very high (n= 25) 2.2 | 1.8 - 2.4 | 10.0 | 10.0 - 10.4 | |
̶ High (n= 20) 5.6 | 2.3 - 9.5 | 7.2 | 6.9 - 13.0 | |
̶ Intermediate (n= 107) 2.1 | 0.8 - 4.0 | 6.9 | 6.1 - 9.1 | |
̶ Low (n= 130) 1.6 | 0.8 - 3.8 | 6.7 | 5.8 - 8.2 | |
̶ Very low (n= 259) 1.3 | 0.8 - 2.2 | 6.5 | 5.6 - 6.8 | |
a Based-on the index of social marginalization with data from the National Population Council, 2015.
* Smoothed rate per 10,000 population. IQR= Interquartile Range.
Overall, a heterogeneous geographic pattern of the RMSF cases throughout both cities was observed, with no evident clustering. However, when using the density of cases per km2, a higher case-density in scattered census tracts in Her- mosillo was observed, reaching clusters of 12 to 14 cases per km2, mainly in the north of the city. In Obregon, clusters of 23 to 27 cases per km2 were found in the northeast and the southwest of the city, as shown in the Figure 3a-b.
Regarding the spatial distribution of the smoothed CI, the census tracts with the highest CI of RMSF in Obregon are coincidently located at the northeast side of the city, whereas in Hermosillo are placed at the east and north, with higher dispersion towards the city suburbs (Figure 3c-d).
Broadly, a weak (Moran’s I = 0.27) although positive (p < 0.05) spatial autocorrelation between the smoothed CI of RMSF and the ISM was observed in Hermosillo. Additionally, in the local bivariate analysis, census tracts having high incidence of RMSF as well as high ISM were detected throughout the four cardinal points of the city. A similar value (Moran’s I = 0.29) for the geographic autocorrelation was found in Obregon. Despite that, census tracts with high incidence of RMSF and high social marginalization were identified at the northeast and southeast sides of this city (Figure 4).
Lastly, by using the Gi* method several RMSF hot spots were detected in Hermosillo, in such areas coincided a high smoothed CI (> 3.88 cases per 10 000 inhabitants), high de- gree of social marginalization, and land surface temperature
> 31.6° C; these clusters were statistically significant and can be qualified as of high risk for transmission of the disease within the city. Furthermore, in Obregon two hot spots were clearly defined at the northeast and southwest sides of the city. Such patterns are displayed in Figure 5.
Figure 3. Spatial distributions of the density of RMSF cases for (a) Hermosillo and (b) Obregon and cumulative incidence by census- tract for (c) Hermosillo and (d) Obregon.
Figura 3. Distribución espacial de la densidad de casos de FMRR para (a) Hermosillo y (b) Obregón, e incidencia acumulada por área censal para (c) Hermosillo y (d) Obregón.
Figure 4. Bivariate local spatial clustering by census-tract: incidence of RMSF-index of social marginalization, in (a) Hermosillo and (b) Obregon.
Figura 4. Agrupamientos espaciales locales bivariados por área censal: incidencia de FMRR-índice de marginación social, en (a) Hermosillo y (b) Obregón.
Figure 5. Hot and cold spot areas with 99 %, 95 % and 90 % confidence that combine cumulative incidence of RMSF, index of social marginalization and land surface temperature in (a) Hermosillo and (b) Obregon.
Figura 5. Puntos calientes y fríos con 99 %, 95 % y 90 % de confianza que combinan incidencia acumulada de FMRR, índice de marginación social y temperatura superficial terrestre en (a) Hermosillo y (b) Obregón.
Since early 2000, RMSF has remained as a serious public health problem in Sonora and other regions of Mexico, with growing evidence about its geographical spread across the country, particularly along the US-Mexico border (Drexler et al., 2017; López-Castillo et al., 2018; Maynez-Prieto et al., 2021; Zazueta et al., 2021; Estrada-Mendizabal et al., 2023). In this context, our findings can contribute to gain understanding on the geographic distribution of the disease in urban set- tings. More specifically, in determining neighborhoods that are at high-risk and where interventions can be implemen- ted. Overall, we found that RMSF followed a heterogeneous spatial distribution within the study cities, albeit several high-risk areas with a high incidence of RMSF and a high de- gree of social marginalization were identified by using spatial interpolation and clustering techniques.
Although there is no certainty about the urban distri- bution of RMSF in Mexico, our findings are consistent with evidence from Brazil in which the spatial distribution of the disease correlates with socioenvironmental variables and tick-host interaction (Ribeiro et al., 2020). In this sense, we cannot conclude that there is an individual effect of socioe- conomic status over the risk of RMSF, but an aggregated effect may exist since the higher incidence of the disease was positively correlated with worse living-conditions at the census-tract level in both cities.
In RMSF endemic regions, it is pivotal to relate the population wellbeing with the geographic patterns of the
disease. In Mexico, the ISM is an influential indicator and can be helpful to understand the spread of the disease, mainly because it includes indicators about sanitary conditions of households, such as the quality of build materials, access to sanitation, disposal of garbage, among others. Some of these characteristics can help to depict epidemiological patterns of the disease, as many determinants of the physical space can foster growth, multiplication, and development of the brown dog tick. For instance, the makeshift building materials are directly related with sites suitable for breeding and hiding of ticks (i.e., holes in walls or floors, yards with dirt floors) around and indoor of households, which increases the risk of cases and outbreaks of the disease (Demma et al., 2005; Álvarez-Hernández et al., 2020). Additionally, demography and movement patterns of dogs, as related with density of homes, should be considered when analyzing urban patterns of the disease (López-Pérez et al., 2020).
We also found that high incidence of RMSF correlate with a high degree of social marginalization, similar to that observed in endemic regions of Brazil (Oliveira et al., 2017). In Sonora, there are no studies specifically addressing the geographical distribution of RMSF at the census-tract level, however, previous findings at the individual level highlight the need to improve preventive strategies for the disease, particularly in socially vulnerable neighborhoods (Reyes- Castro et al., 2021).
On the other hand, across the Americas RMSF shows a seasonality with epidemic peaks when environmental temperature and humidity are higher (Biggs et al., 2016). In Sonora that rise occurs during the period of June through November, thus environmental conditions may increase the risk for RMSF in the outskirts of localities where a higher land surface temperature (LST) significantly correlate with higher incidence of RMSF. However, a more refined analysis at the regional level is needed, for instance differentiating the LST by time of day could determine if this effect is associated with patterns of feeding, growth, or development of the brown dog tick and how this could explain local patterns of disease transmission.
Furthermore, by using the Gi* method for joint analysis we found significant clusters of RMSF, where there is an overlap between high levels of social marginalization, high cumulative incidence, and high LST. Although we did not assess the role of relative humidity, our finding is comparable with previous evidence, in which a higher level of poverty, a higher average of land temperature (> 35.0 °C), and increase of relative humidity, were associated with higher incidence of RMSF (Raghavan et al., 2016).
For a tick to transmit the disease, the pathogen R. ric- kettsii must survive the molting of every stage of the tick, which means that the pathogen greatly depends on the development, survival, and reproduction of the tick, and on the relationships with the environment (Parola et al., 2008; Tomassone et al., 2018). Therefore, factors like urban poverty and climate can be critical for the spread of RMSF, because such determinants can provoke microclimates that affect the life cycle of hard ticks (Estrada-Peña y Venzal, 2007; Gray et al., 2009). Soil temperature is correlated with evaporation, so the presence of uncut brush, scrubs, and other shelters in-and-around homes (i.e. cracks in walls and floors) may provide a suitable microclimate for the development of ticks, increasing the risk of human exposure to R. rickettsii and other rickettsial pathogens (Dantas-Torres, 2010).
Finally, our study has limitations that should be carefully considered. First, our ecologic design could bias the global effect of social and environmental factors on the geographic pattern of the disease, which may flaw the correlation bet- ween the actual incidence of RMSF and ISM, therefore future studies can explore that relationship with epidemiologic designs using individual units of study. Importantly, misclas- sification of RMSF cases may have arisen because 50 % of study subjects were confirmed with a single blood sample processed by IFA, which is notoriously unable to distinguish between SFGR and RMSF (McQuiston et al., 2014); in addition, we used regular PCR targeting a 401 base pairs of Rickettsiae from human blood samples, which could have a relatively low sensitivity, prompt to false negatives. Undoubtely, there is an urgen need to improve the methods of diagnosis of RMSF in Mexico. Despite this, still there is a strong indication that R. rickettsii is the only pathogen associated with the oc- currence of RMSF cases in northwest Mexico (Álvarez-López et al., 2021; Zazueta et al., 2021; Estrada-Mendizabal et al.,
Volume XXVII
2023) so our findings can be orientative for studying urban distribution of the disease in endemic regions.
The global distribution of RMSF is heterogeneous in the study cities, with areas of high incidence randomly spread. Moreover, at the global level, a weak spatial correlation between the incidence of RMSF and the ISM was identified, but socially deprived census tracts within the cities showed high incidence of the disease, and thus, can be considered hot spots for public health interventions. The land surface temperature can be a useful indicator to detect hot spots of the disease, given that the survival and development of hard ticks, such as R. sanguineus sensu lato, are associated with ecosystems having warmer surface temperatures (Parola et al., 2008; Ogden et al., 2021; Nutall, 2022).
Our findings show that spatial analysis is a useful tool for the identification and stratification of risk areas for RMSF at the small-area level. In scenarios like ours, its routinary use can be cost-effective and, through data management systems, can improve the quality of information sources. This has the potential for better analysis and real-time results that would help with timely decision making for the prevention and control of the disease.
We are grateful for National Council of Humanities, Science and Technology, Mexico (CONAHCYT) support to A.A.E.C. through a graduate scholarship. This material was supported by the Secretary of Health of the State of Sonora by the data provided. Data for LandSat 5 and 8 can be obtained at http:// earthexplorer.usgs.gov.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Álvarez-Hernández, G., Gutiérrez-Roldán, J.F., Milan-Hernández, N.S., Lash, R.R., Barton-Behravesh, C. and Paddock, C.D. 2017. Rocky Mountain spotted fever in Mexico: past, present, and future. Lancet Infect Dis 17:e189-e196. https://doi. org/10.1016/S1473-3099(17)30173-1
Álvarez-Hernández, G., Paddock, C.D., Walker, D.H., Valenzuela, J., Calleja-López, J.R.T., Rivera-Rosas, C.N., et al. 2024a. Rocky Mountain spotted fever is a neglected tropical disease in Latin America. PLoS Negl Trop Dis 18 (7):e0012276 https:// doi.org/10.1371/journal.pntd.0012276
Álvarez-Hernández, G., López-Ridaura, R., Cortés-Alcalá, R., García-Rodríguez, G., Calleja-López, J.R.T., Rivera-Rosas, C.N., et al. 2024b. Rocky Mountain spotted fever in Mexico: a call to action. Am J Trop Med Hyg 111 (5): 1070-1077 https://doi. org/10.4269/ajtmh.24-0265
Álvarez-Hernández, G., Drexler, N., Paddock, C.D., Licona- Enriquez, J.D., De la Mora-Delgado, J., Straily, A., et al. 2020.
Community-based prevention of epidemic Rocky Mountain spotted fever among minority populations in Sonora, Mexico, using a One Health approach. Trans R Soc Trop Med Hyg. 114: 293-300. DOI:10.1093/trstmh/trz114
Álvarez-López, D.I., Ochoa-Mora, E., Heitman, K.N., Binder, A.M., Alvarez, G. and Armstrong, P.A. 2021. Epidemiology and clinical features of Rocky Mountain spotted fever from enhanced surveillance, Sonora, Mexico: 2015-2018. Am J Trop Med Hyg 104 (1): 190-197. DOI:10.4269/ajtmh.20-0854
Atkinson, S.F., Sarkar, S., Aviña, A., Schuermann, J.A., Schuermann,
J.A. and Williamson, P. 2012. Modelling spatial concordance between Rocky Mountain spotted fever disease incidence and habitat probability of its vector Dermacentor variabilis (American dog tick). Geospat Health 7 (1):91-100. DOI:10481/ gh.2012.108
Besag, J., York, J., Mollié, A. 1991. Bayesian image restoration, with two applications in spatial statistics. Ann Inst Stat Math 43:1-20.
Biggs, H.M., Barton-Behravesh, C., Bradley, K.K., Dahlgreen, F.S., Drexler, N.A., Dumler, J.S., et al. 2016. Diagnosis and management of tickborne rickettsial diseases: Rocky Mountain spotted fever and other spotted fever group rickettsioses, ehrlichiosis and anaplasmosis – United States. A practical guide for health care and public health professionals. MMWR Recomm Rep 65 (2): 1-48. DOI: 10.15585/mmwr.rr6502a1.
Celemín, J.P. 2009. Spatial autocorrelation and local indicators of spatial association: importance, structure and application. Rev Univ Geogr 18:11-31. [In Spanish] https://www.redalyc. org/articulo.oa?id=383239099001
Dantas-Torres, F. 2010. Biology and ecology of the brown dog tick, Rhipicephalus sanguineus. Parasit Vectors 3:26. https:// doi.org/10.1186/1756-3305-3-26
De Oliveira, S.V., Willeman, M.C.A., Gazeta, G.S., Angerami, R.N., Gurgel-Goncalves, R. 2017. Predictive factors for fatal tick- borne spotted fever in Brazil. Zoonoses Public Health 64 (7): e44-e50. DOI:10.1111/zph.12345
Demma, L.J., Traeger, M.S., Nicholson, W.L., Paddock, C.D., Blau, D.M., Eremeeva, M.E., et al. 2005. Rocky Mountain spotted fever from an unexpected tick vector in Arizona. N Engl J Med 353: 587-594. https://doi.org/10.1056/NEJMoa050043
Drexler, N.A., Yaglom, H., Casal, M., Fierro, M., Kriner, M., Murphy, B., et al. 2017. Fatal Rocky Mountain spotted fever along the United States-Mexico border, 2013-2016. Emerg Infect Dis 23(10): 1621-1626. DOI: 10.3201/eid2310.170309
Eremeeva, M.E., Zambrano, M.L., Anaya, L., Beati, L., Karpathy, S.E., Santos-Silva, M.M., et al. 2011. Rickettsia rickettsii in Rhipicephalus ticks, Mexicali, Mexico. J Med Entomol 48 (2): 418-421. https://dx.oi.org/10.1603/ME10181
ESRI. How point density works. ArcGIS Desktop (10.3), 2011 [Internet] United States. Environmental Systems Research Institute. Available from: https://webhelp.esri.com/ arcgisdesktop/9.3/index.cfm?topicname=how_point_ density_works. Access: Feb 20, 2024
Estrada-Mendizabal, R.J., Tamez-Rivera, O., Vela, E.H., Blanco- Murillo, P., Alanis-Garza, C., Flores-Gouyonnet, J., et al. Rickettsial disease outbreak, Mexico, 2022. Emerg Infect Dis 2023;29(9):1944-1947. DOI:10.3201/eid2909.230344
Estrada-Peña, A. and Venzal, J.M. 2007. Climate niches of tick species in the Mediterranean region: modeling of occurrence data, distributional constraints, and impact of
climate change. J Med Entomol 44:1130-1138. https://doi. org/10.1603/0022-2585(2007)44
Gobierno de México. Índices de marginación 2020. [Internet], México. Consejo Nacional de Población y Vivienda. Available from: https://www.gob.mx/cms/uploads/attachment/ file/848423/Indices_Coleccion_280623_entymun-p_ ginas-1-153.pdf. Access: Feb 20, 2024.
Gray, J.S., Dautel, H., Estrada-Peña, A., Kahl, O. and Lindgren,
E. 2009. Effects of climate change on ticks and tick-borne diseases in Europe. Interdiscip Perspect Infect Dis 2009: 593232. https://doi.org/10.1155/2009/593232
Gobierno de México. Panorama sociodemográfico de Sonora. Censo de Población y Vivienda 2020. [Internet]. México. Instituto Nacional de Estadística y Geografía. Available from: https://www.inegi.org.mx/contenidos/productos/ prod_serv/contenidos/espanol/bvinegi/productos/nueva_ estruc/702825197995.pdf. Access: Feb 20, 2024.
Kato, C.Y., Chung, I.H., Robinson, L.K., Austin, A.L., Dasch, G.A. and Massung, R.F. 2013. Assessment of real-time PCR assay for detection of Rickettsia spp. and Rickettsia rickettsii in banked clinical samples. J Clin Microbiol 51:314-317. https:// doi.org/10.1128/JCM.01723-12
Labruna, M.B., Whitworth, T., Horta, M.C., Bouyer, D.H., McBride, J.W., Pinter, A., et al. 2004. Rickettsia species infecting Amblyomma cooperi ticks from an area in the state of Sao Paulo, Brazil, where Brazilian spotted fever is endemic. J Clin Microbiol 42:90-98. https://doi.org/10.1128/JCM.42.1.90-
López-Castillo, D.C., Vaquera-Aparicio, D., González-Soto, M.A., Martínez-Ramírez, R., Rodríguez-Muñoz, L. and Solórzano- Santos, F. 2018. Rocky Mountain spotted fever: experience from 5 years of active surveillance in a pediatric hospital in northeast Mexico. Bol Med Hosp Infant Mex 75: 303-308. DOI:10.24875/BMHIM.M18000034. [Article in Spanish]
López-Pérez, A.M., Orozco, L., Zazueta, O.E., Fierro, M., Gomez, P. and Foley, J. 2020. An exploratory analysis of demography and movement patterns of dogs: New insights in the ecology of endemic Rocky Mountain spotted fever in Mexicali, Mexico. PLoS One. 15(5):e0233567. DOI: 10.1371/ journal.pone.0233567
Maynez-Prieto, C., Rodríguez-Alarcón, C., Garza-Hernández, A., Laredo-Tiscareño, S., Espino-Solís, G. and Adame-Gallegos,
J. 2021. Rickettsiosis and other tick-borne diseases across the border: regional focus in Chihuahua. Southwestern Entomologist 46 (1): 167-178. DOI:10.3958/046.0116.
McQuiston, J.H., Wiedeman, C., Singleton, J., Carpenter, L.R., McElroy, K., Mosites, E., et al. 2014. Inadequacy of IgM antibody tests for diagnosis of Rocky Mountain spotted fever. Am J Trop Med Hyg 91: 767-70. DOI: 10.4269/ajtmh.14-0123
Navarro-Estupiñan, J., Robles-Morua, A., Díaz-Caravantes, R. and Vivoni, E. 2020. Heat risk mapping through spatial analysis of remotely sensed data and socioeconomic vulnerability in Hermosillo, México. Urban Clim, 31, 100576. https://doi. org/10.1016/j.uclim.2019.100576
Navarro-Estupiñan, J., Robles-Morua, A., Vivoni, E.R., Zepeda, J., Montoya, J. and Verduzco, V.V. 2018. Observed trends and future projections of extreme heat events in Sonora, Mexico. Int J Climatol 38:5168-5181. https://doi.org/10.1002/ joc.5719
Nutall, P.A. 2022. Climate changes impacts on ticks and tick- borne infections. Biologia 77: 1503-1512 https://doi. org/10.1007/s11756-021-00927-2
Ogden, N.H., Beard, C.B., Ginsberg, H.S. and Tsao, J.I. 2021. Possible effects of climate change on ixodid ticks and the pathogens they transmit: predictions and observations. J Med Entomol 58 (4): 1536-1541. DOI:10.1093/jme/tjaa220
Ord, J.K. and Getis, A. 1995. Local spatial autocorrelation statistics: Distributional issues and an application. Geogr Anal 27 (4):286-306. https://doi.org/10.1111/j.1538-4632.1995. tb00912.x
Parola, P., Socolovschi, C., Jeanjean, L., Bitam, I., Fournier, P.E., Sotto, A., et al. 2008. Warmer weather linked to tick attack and emergence of severe rickettsioses. PLoS Negl Trop Dis 2 (11):e338. https://doi.org/10.1371/journal.pntd.0000338
Peeters, A., Zude, M., Käthner, J., Ünlü, M., Kanber, R., Hetzroni A., et al. 2015. Getis–Ord’s hot- and cold-spot statistics as a basis for multivariate spatial clustering of orchard tree data. Comput Electron Agric 111:140-150. https://doi. org/10.1016/J.COMPAG.2014.12.011
Raghavan, R.K., Goodin, D.G., Neises, D., Anderson, G.A. and Ganta R.R. 2016. Hierarchical Bayesian spatio-temporal analysis of climatic and socio-economic determinants of Rocky Mountain spotted fever. PLoS One 11:e0150180. https://doi.org/10.1371/journal.pone.0150180
Reyes-Castro, P.A., Ernst, K.C., Walker, K.R., Hayden, M.H. and Alvarez-Hernandez, G. 2021. Knowledge, attitudes, and practices related to Rocky Mountain spotted fever in Hermosillo, México. Am J Trop Med Hyg 104:184-189. https://doi.org/10.4269/ajtmh.20-0181
Ribeiro, C.M., da Costa, V.M., de Carvalho, J.L.B., Mendes, R.G., Bastos, P.A.S, Katagiri, S., et al. 2020. Brazilian spotted fever: a spatial analysis of human cases and vectors in the state of Sao Paulo, Brazil. Zoonoses Public Health 67 (6): 629-636. https://doi.org/10.1111/zph.12742
Siabato, W. and Guzmán-Manrique, J. 2019. La autocorrelación espacial y el desarrollo de la geografía cuantitativa. Cuad Geogr Rev Colomb Geogr 28(1):1-22. (In Spanish). https:// doi.org/10.1544/rcdg.v28n1.76919
Sorek-Hamer, M., Just, A.C. and Kloog, I. 2016. Satellite remote sensing in epidemiological studies. Curr Opin Pediatr 28:228-
234. https://doi.org/10.1097/MOP.0000000000000326
Süss, J., Klaus, C., Gerstengarbe, F.W. and Werner, P.C. 2008. What makes ticks tick? Climate change, ticks, and tick-borne diseases. J Travel Med 15 (1):39-45. https://doi.org/10.1111/ j.1708-8305.2007.00176.x
Tomassone, L., Portillo, A., Nováková, M., de Sousa, R. and Oteo,
J.A. 2018. Neglected aspects of tick-borne rickettsioses. Parasit Vectors 11:263. https://doi.org/10.1186/s13071-018- 2856-y
Treadwell, T.A., Holman, R.C., Clarke, M.J., Krebs, J.W., Paddock,
C.D. and Childs, J.E. 2000. Rocky Mountain spotted fever in the United States, 1993-1996. Am J Trop Med Hyg 63:21-26. DOI:10.4269/ajtmh.2000.63.21
Tsou, J., Zhuang, J., Li, Y. and Zhang, Y. 2017. Urban heat island assessment using the Landsat 8 Data: A case study in Shenzhen and Hong Kong. Urban Sci 1 (1):10. https://doi. org/10.3390/urbansci1010010
World Health Organization. 2019. International Classification of Diseases and Health Related Problems. [Internet]. https:// www.who.int/standards/classifications/classification-of- diseases
Yang, C., He, X., Yan, F., Lingxue, Y., Bu, K., Yang, J., et al. 2017. Mapping the influence of Land Use/Land Cover changes on the urban heat island effect—A case study of Changchun, China. Sustainability 9 (2): 312. https://doi.org/10.3390/ su9020312
Zazueta, O., Armstrong, P., Márquez-Elguea, A., Hernández- Milan, N., Peterson, A., Ovalle-Marroquín, D., et al. 2021. Rocky Mountain spotted fever in a large Metropolitan Center, Mexico-United States Border, 2009-2019. Emerg Infect Dis 27 (6): 1567-1576. DOI: 10.3201/eid2706.191662.
Volume XXVII