Exploring the association between precipitation and population cases of ocular toxoplasmosis in Colombia.

BACKGROUND: Previous studies suggest a relationship between precipitation and ocular toxoplasmosis (OT) reactivation and congenital toxoplasmosis infection. We aimed to investigate the relationship between precipitation and the frequency of new OT cases in Colombia from 2015 to 2019.
METHODOLOGY: This retrospective cohort study analyzed data obtained from a claims-based database created by the Colombian Ministry of Health and national registries of precipitation of the Institute of Hydrology, Meteorology, and Environmental Studies. We estimated the daily number of OT cases, interpolating data from the average number of annual cases from 2015 to 2019. Then, we compared exposures (mean daily precipitation) in the case period in which the events (interpolated OT new cases) occurred by a quasi-Poisson regression, combined with a distributed lag non-linear model to estimate the non-linear and lag-response curve. PRINCIPAL
FINDINGS: In the 5-year analysis, there were 1,741 new OT cases. Most of the cases occurred in 2019, followed by 2015 and 2018. New OT cases among departments were significantly different (P< 0.01). The cumulative exposure-response curve was decreasing for most departments. Nevertheless, in Chocó, Bogotá, Cesar, Cauca, and Guajira, when a certain amount of precipitation accumulates, the relative risk (RR) increases, which was contrary to the pattern observed in the other regions. The response curves to the one-day lag showed that precipitation influences the RR; however, the trends vary by department. Finally, an increasing trend in the number of cases was directly proportional to precipitation in Guajira, Atlántico, Norte de Santander, Santander, Caquetá and Quindío (r = 0.84; P< 0.05).
CONCLUSIONS: Precipitation influenced the RR for new OT cases. However, varying trends among geographical regions (departments) lead us to hypothesize that other sociodemographic, behavioral, and environmental variables, such as wind and water contamination, could influence the RR.

Introduction

Toxoplasma gondii (Tg), an intracellular protozoan belonging to the Apicomplexa family, is responsible for toxoplasmosis [1]. The life cycle of Tg allows it to survive in humid conditions. Warm-blooded animals, including humans, are intermediate hosts and felines are the definitive hosts of Tg. Unsporulated oocysts are present in the cats’ feces. Oocysts take 1–5 days to sporulate and become infective in the environment. Intermediate hosts become infected after ingesting soil, water, or plant materials contaminated with oocysts. Inside the host’s gastrointestinal tract, oocysts become tachyzoites, which then migrate to the neural and muscle tissues and develop into tissue cyst bradyzoites. Cats become infected after consuming intermediate hosts harboring tissue cysts or sporulated oocysts. Humans become infected by eating undercooked meat of animals with tissue cysts, consuming food or water contaminated with oocysts, blood transfusion, organ transplantation, and transplacentally from mother to fetus. The parasites form tissue cysts in the human host, most commonly in the skeletal muscle, myocardium, brain, and retina [2]. In humans, the infection can be congenital or postnatally acquired [3]. The clinical spectrum varies from asymptomatic to local manifestations, such as ocular toxoplasmosis (OT) or systemic disease [4].

Although OT is one of the leading causes of infectious posterior uveitis worldwide [5], it is considered a neglected tropical disease. In South America, retinochoroiditis reactivation episodes are more frequent and severe than those in Europe and North America due to the virulent serotypes in our region [6]. The annual incidence of OT was calculated previously as three new episodes per 100,000 inhabitants in Colombia [7]. Furthermore, Tg is the leading infectious cause of visual impairment in immunocompetent patients [8] and Colombia’s second most common cause of congenital blindness [9].

A previous study has shown that the mean annual rainfall significantly correlates with the incidence of congenital toxoplasmosis [10]. Furthermore, another study found a positive correlation between the annual precipitation and toxoplasmic retinochoroiditis reactivation rate [11]. This correlation could exist because oocyst survival is higher in humid, warm soil than under dry conditions [12]. Quantitative estimation of the viability of Tg oocysts in soil demonstrated that, after 100 days, 7.4% and 43.7% of oocysts survived under dry and damp conditions, respectively [12]. Tg oocysts survive up to 54 months in cold water [13].

The literature describing the effect of precipitation and risk of OT is limited; however, it is hypothesized that higher rainfall is associated with increased rates of congenital toxoplasmosis and OT recurrences [10,11]. Therefore, considering that Colombia is one of the rainiest countries worldwide [14], this study aimed to investigate the relationship between precipitation and new cases of OT from 2015 to 2019 in our country. We hypothesize that higher precipitation is a risk factor for developing OT.

Methods

Ethics statement

This study adheres to the ethical principles for human research established by the Helsinki Declaration, Belmont Report, and Colombian Resolution 008430 of 1993. According to the risks contemplated in resolution 8430 from 1993, this investigation is considered without risks. The information in the databases used in this article is freely accessible and is available for research purposes. In the same way, their coding system ensures data confidentiality.

Design

In this retrospective cohort study, we analyzed the correlation of the daily mean of new OT cases among Colombian patients with the daily average precipitation in Colombia between 2015 and 2019.

Population data

Colombia has a tropical forest and a tropical monsoon climate based on the Köppen–Geiger classification system; thus, it has a huge climatic diversity [15]. Colombia is administratively and politically divided into 32 geographical regions with different population densities that share cultural and economic characteristics, which are called “departments.” During the study’s observation period, Colombia had between 46,313,898 and 49,395,678 inhabitants. The distribution in each of the departments is shown in Table 1.

Table 1

Distribution of precipitation, population, and ocular toxoplasmosis cases by department.

Department	Precipitationa	Populationb	OT casesc	OT cases per inhabitants
	Daily mean in mm (SD) T = Total precipitation in 5 years in mm.	Mean 5 years [range]	Annual mean (SD)	Average cases in 5 years per million inhabitants
Amazonas	nd	75,144.2 [72,485–77,753]	nd	nd
Antioquia	8.02 (16.5) T = 104,436	6’320,083.2 [6’134,953–6’550,206]	88.2 (19.0)	13.96
Arauca	nd	255,352.6 [239,772–280,109]	3.40 (3.85)	13.31
Atlántico	2.67 (9.94) T = 7,141	2’492,540 [2’393,557–2’638,151]	6.40 (5.22)	2.57
Bogotá, DC.	2.39 (5.28) T = 26,641	7’383,413.8 [7’273,265–7’592,871]	106 (38.5)	14.36
Bolívar	7.23 (18.2) T = 21,462	2’048,620.6 [1’993,760–2’130,512]	9.20 (6.22)	4.49
Boyacá	3.21 (7.65) T = 128,436	1’209,827 [1’193,206–1’230,910]	8.80 (6.22)	7.27
Caldas	9.78 (19.5) T = 15,049	993,907.2 [984,360–1’008,344]	39.2 (15.6)	39.44
Caquetá	nd	401,559.6 [398,725–406,142]	22.2 (15.9)	55.28
Casanare	nd	412,039.2 [396,320–428,563]	3.40 (1.14)	8.25
Cauca	5.40 (8.71) T = 13,723	1’449,288 [1’420,313–1’478,407]	20.6 (7.13)	14.21
Cesar	4.65 (13.4) T = 49,399	1’173,191 [1’114,269–1’252,398]	23.4 (17.2)	19.95
Chocó	12.8 (23.8) T = 44,973	525,824.2 [509,240–539,933]	4.40 (2.97)	8.37
Córdoba	4.00 (12.0) T = 88,767	1’765,363.2 [1’726,287–1’808,439]	13.0 (3.54)	7.36
Cundinamarca	3.25 (8.07) T = 142,975	2’794,656.2 [2’543,338–3’085,522]	37.4 (15.4)	13.38
Guainía	nd	46,370 [43,291–49,473]	0.40 (0.55)	8.63
Guajira	1.61 (8.06) T = 72,608	855,974 [803,092–927,506]	4.00 (4.00)	4.67
Guaviare	nd	80,822.8 [77,328–84,716]	1.60 (2.51)	19.8
Huila	3.87 (9.43) T = 175,723	1’086,841 [1’061,405–1’111,844]	26.0 (8.75)	23.92
Magdalena	4.91 (13.2) T = 13,553	1’319,255.8 [1’268,980–1’388,832]	9.20 (7.92)	6.97
Meta	nd	1’021,131 [987,232–1’052,125]	16.6 (12.7)	16.26
Nariño	nd	1’621,210 [1’608,726–1’628,981]	25.4 (16.8)	15.67
Norte de Santander	4.45 (12.3) T = 96,860	1’467,692 [1’409,900–1’565,362]	8.20 (6.91)	5.59
Putumayo	nd	340,939.8 [327,856–353,759]	5.20 (5.59)	15.25
Quindío	nd	535,620 [526,484–547,855]	13.0 (7.97)	24.27
Risaralda	nd	936,713 [923,443–952,511]	24.6 (10.9)	26.26
San Andrés	4.05 (12.7) T = 10,677	61,567 [61,406–62,482]	0.00 (0.00)	0
Santander	5.39 (12.6) T = 175,976	2’157,188.6 [2’097,069–2’237,587]	28.2 (8.84)	13.07
Sucre	3.56 (11.7) T = 32,638	893,516.6 [867,701–928,984]	14.8 (9.31)	16.56
Tolima	4.95 (12.0) T = 167,612	1’327,187.4 [1’320,911–1’335,313]	27.0 (10.9)	20.34
Valle del Cauca	nd	4’445,393.2 [4’397,194–4’506,768]	109 (50.6)	24.52
Vaupés	nd	39,940.8 [37,638–42,721]	nd	nd
Vichada	nd	105,304.2 [100,392–110,599]	nd	nd
Not defined a	na	na	54.6 (54.9)	nd

nd: no data, na: not applicable

aAvailable data from the National Institute of Hydrology, Meteorology, and Environmental Studies (IDEAM) about daily average precipitation from January 01, 2015 to December 31, 2019.

bData from the retroprojections of the National Administrative Department of Statistics.

cMean of cases between 2015 and 2019 by department, which are calculated as follows: total of records of OT in the department from 2015 to 2019 divided in 5 years. Cases of OT in Colombia by departments between 2015 and 2019 were retrieved from the System of Information of Social Protection.

We obtained the study data from the national database created by the Colombian Ministry of Health, which is known as the System of Information of Social Protection (SISPRO) [16]. Its function is to store, process, and systematize Colombian citizens’ information to make decisions that support the development of effective policies and monitoring in sectors, such as health, pensions, occupational risks, and social promotion. Health data are collected and codified by medical staff during each medical contact (inpatient or outpatient) from private and public health providers and insurers using the International Statistical Classification of Diseases 10th Revision (ICD-10). The Individual Records of Health Service Provision (RIPS) groups all demographic and clinical data [17]. According to a recent report, the Colombian Health System has one of the most prominent coverages in Latin America, encompassing 50 million inhabitants, which represents 98.88% of the population in 2021 [18].

We searched demographic data of patients grouped using the RIPS of the SISPRO database to obtain the epidemiological descriptions of all the OT registers (coded as ICD-10: B58.0) [16]. We applied a filter by year (2015–2019) and type of diagnosis (i.e., new confirmed, repeated confirmed, or not confirmed) to delimit the results. We only used the “new confirmed” filter for every month of each year (2015–2019), assuming that they were the new cases diagnosed in that period by physicians following international standards. We confirmed that the resulting dynamic table only reported consultations with the unique identifier in a single column filter to ensure no repeated patients.

Precipitation data

Additionally, we requested precipitation data for all Colombia’s departments from January 1, 2015, to December 31, 2019, via corporate email to the National Institute of Hydrology, Meteorology, and Environmental Studies (IDEAM) [19], which has been gathering information on precipitation from 1,567 weather stations distributed in the natural regions of Colombia since 1981. They sent a total of 2,157 files, from which we selected 825 files. We excluded files with no information on daily precipitation for the years 2015 through 2019. Each file corresponded to a meteorological station with the hourly precipitation measurement data. We further averaged these hourly data into daily measurements and calculated the daily average. We obtained data from most departments of Colombia (Table 1).

Statistical analysis

In this study, we compared the exposure (daily average precipitation) in the period when events (cases of OT) occurred with exposures in nearby referent periods to examine the differences in exposure that may contribute to the differences in the daily number of OT cases by departments. A time-stratified case-crossover design was adopted to regulate potential confounders (e.g., age and sex) using self-control and excluding the long-term impact of precipitation by stratification of time by departments. We used the calendar month as the time stratum to control the effects of long-term trends, seasonality, and day of the week.

Some departments, such as Amazonas, Vaupés, and Vichada, were excluded from the analysis because they did not have information about precipitation and cases of OT. Moreover, other departments, including Arauca, Caquetá, Casanare, Guainía, Guaviare, Meta, Nariño, Putumayo, Quindío, Risaralda, and Valle del Cauca, were excluded due to the lack of data from the meteorological stations.

We performed a quasi-Poisson regression, controlling for the over-scattering problem, combined with a distributed lag non-linear model (DLNM) to estimate the precipitation’s non-linear and lag influence on the appearance of OT in Colombia. The DLNM class is used to describe the associations where the dependency between an exposure and an outcome is delayed in time, it means, the lag. The DLNM is based on the definition of “cross-based,” two-dimensional space of functions to reflect the non-linear exposure responses and delay structure of the association [20,21]. First, a base model with natural cubic splines was fitted to optimize control for confounding factors, and DLNM was built as follows:

A quasi-Poisson regression model combined with a time-stratified case-crossover design and DLNM was built as follow: day of observation (t); the count of OT cases on t (Y); the disease count at time t (disease); α is an intercept; precipitation, as the ith precipitation concentration on t; and ns(t) represents the cross-basis function with natural spline df2 = 7 for precipitation.

Although the Tg can survive until months in humid conditions, to capture the complete lag–response curve, the maximal lag of precipitation was set in our model to 14 days, for the sake of simplification and without loss of generalization; meanwhile, this maximal lag was assigned to the length of the case and control periods. In addition, a 3-day duration was specified to be the maximal lag of meteorological factors. The df and maximum lag days for precipitation determination referred to the Akaike information criterion for quasi-Poisson (Q-AIC), which could produce the relatively superior model.

To identify the influence of precipitation, we calculated for and presented the relative risk (RR). We calculated the single-day lag influence and cumulative lag influence (lag0–1, lag0–6, lag0–8, lag0–10, lag0–12, lag0–13, and lag0–14) to effectively depict the characteristics of the association between precipitation and OT cases.

We computed the incidence rates using a spatial generalized linear mixed model for unit area data, where the response variable is Poisson, for the precipitation map, fitted using log-linear regression (a log link and error distribution of Poisson) [22-24]. With the Poisson spatial model, the exponents of the β coefficients are equal to the incidence rate ratio (RR). The β coefficient was estimated using as covariate the exposure to precipitation. For this analysis, we used the data of the new cases of OT reported in 2019 and the precipitation data published by the IDEAM for 2019 [25]. Additionally, we performed a Pearson correlation test using the available data for 2019 in the departments evaluated previously by Gomez-Marin et al. [10] (Guajira, Atlántico, Norte de Santander, Santander, Caquetá, and Quindío) to evaluate if their results were reproducible for other types of Tg infection, such as OT. A larger temporal subanalysis could not be done due to the lack of data in the previous years. All analyses were conducted using R version 3.5.1 with the package for fitting the DLNM and GNM package for conditional quasi-Poisson regression and CARBayes [26].

Bias control

Selection bias may occur due to the large number of filters that can be applied to diagnostic data in SISPRO and the existence of ICD codes, such as H30.9 (unspecified chorioretinal inflammation), where some retinochoroidal toxoplasmosis could be misclassified, leading to underestimation. To control bias, we only included patients with a new diagnosis of OT (“new confirmed”). Additionally, it is essential to note that ophthalmologists most commonly use the B58.0 code, SISPRO-based studies using ICD-10 have shown 83.4% of concordance with the medical record, and ICD-10 has demonstrated acceptable accuracy in studies on uveitis using big data [27-29].

Results

There were 1,741 new cases of OT in Colombia between the years 2015 and 2019. The departments with the highest annual average number of OT cases were Valle del Cauca (109; standard deviation [SD] 50.6), Bogotá (106; SD 38.5), and Antioquia (88.2; SD 19.0). More detailed information regarding the number of OT cases registered in each department is presented in Table 1. We found a significant difference in the distribution of OT cases among different departments, which implies that each department behaves differently (P< 0.01). Additionally, regarding the average precipitation in millimeters of water (mm), the departments with the highest daily average precipitation were Chocó (12.8 mm; SD 23.8), Caldas (9.78 mm; SD 19.5), and Antioquia (8.02 mm; SD 16.5). The summary statistics for precipitation of departments for which data were available are shown in Table 1 [19].

Cumulative exposure–response curves

The cumulative exposure–response curves show the cumulative influence of precipitation as the independent variable and the probability of having OT expressed as RR as the dependent variable for each department. The data from most departments showed a downslope cumulative exposure–response curve (Fig 1). Due to the size of the information presented and the variability of the data, we decided to present the results of our analysis of the seven departments that represent the most characteristic patterns in Fig 2. The departments of Guajira and Bogotá showed a directly proportional relationship between precipitation and greater risk of new OT development. In the department of Cauca, although the increase in precipitation is not directly proportional to the risk of new OT, it is accompanied by an increase in risk that becomes constant over time. Regarding Chocó, although a protective effect is evident in lag0–6 days, from lag7–14 days, there is evidence of a directly proportional relationship between increased precipitation and risk of new OT.

Fig 1

Cumulative exposure–response curves for the association between precipitation and the new cases of ocular toxoplasmosis (OT) and its distribution in Colombia.

A. Evidence of the effect of precipitation in the northern departments. B. Effect of precipitation in the southern departments. Due to the climatic diversity secondary to the country’s geography, it is not possible to completely segment the northern and southern regions of the country in any of the Köppen–Geiger classification system [15]. *The seven departments of interest are (1) San Andrés, (2) Sucre, (3) Chocó, (4) Guajira, (5) Cauca, (6) Bogotá, and (7) Huila. Map is from https://d-maps.com/carte.php?num_car=4095&lang=es.

Fig 2

Cumulative exposure–response curves of precipitation on ocular toxoplasmosis (OT) new cases at lag0–14, using a DLNM from 2015 to 2019 in seven selected departments of Colombia.

The solid green line represents the RR. Shaded areas represent 95% confidence interval (CI). Each lag refers to a period of the day (morning and afternoon); therefore, two lag equals 1 day. The cumulative exposure–response curves have a J-shape for Chocó and Bogotá, indicating that when a certain amount of precipitation accumulates, the RR increases. Similarly, the curve shows a growing trend for Guajira, meaning that the accumulative RR of OT increases proportionally with the precipitation exposure. In contrast, Huila’s curve decreases indicating that the accumulated risk decreases while exposure increases (precipitation). The curve for Cauca shows an inverse U-shape with a transient increase in the RR, but the curve for Sucre has a similar pattern without a clear association because it crosses the zone of no effect. Finally, the graph for the department of San Andrés serves as a control case because it has no reported cases and maintains a constant null risk.

Contrary to the abovementioned data, the characteristic pattern of the other regions showed an inversely proportional relationship between precipitation and the risk of new OT, as in the case of Huila. Furthermore, in Sucre, precipitation increase the risk during the first 12 days, but then this effect decreases. Finally, San Andrés acted as a control chart, because no OT cases were recorded during this period, based on the Colombian Ministry of Health database. More information from all the departments is presented in S1 Fig.

Single-day lag–response curves

The single-day lag–response curves showed daily precipitation as the independent variable and the probability of having OT, expressed as a RR, as the dependent variable, for each department (Fig 3). The curve for Bogotá demonstrated a variable effect at different lag days, reaching the highest RR on the lag10 day. Contrary to the Chocó curve, the Cauca and Sucre curves showed a decrease in RR; consequently, as time progresses, exposure loses its effect on the risk of developing OT, crossing the line of no effect at lag12 and lag7 days, respectively. Furthermore, Guajira showed a decreasing curve that remains over the no effect line, indicating precipitation as a constant risk factor for OT. Contrarily, Huila showed an increasing curve and a peak of RR near the lag10 day that always remains under the line of no effect as a protective factor. The San Andrés’ graph showed a continuous or straight line, working as a control. Information on the other departments is available in S2 Fig.

Fig 3

Single-day lag–response curves on ocular toxoplasmosis for precipitation at a lag0–15 model in seven departments of Colombia from 2015 to 2019.

This figure represents the effect of a single-day precipitation for 15 consecutive days, among the total cases, using the DLNM model in seven departments of Colombia from 2015 to 2019.

Table 2 summarizes the cumulative influence of precipitation on different lag days for all new cases of OT in Bogotá, Cauca, Chocó, Guajira, Huila, San Andrés, and Sucre. The results showed that high precipitation could significantly increase the risk of OT cases, and the accumulated influence increased proportionally in some departments. The maximum RR value was 1.005 (95% CI 1–1.01), appearing in lag13–15 days in Guajira. Contrarily, in Chocó, the incidence rate seems to be the lowest, but it can be seen in the temporal analysis that the risk increases when the exposure increases (Table 2). Information regarding RR in other departments is available in S1 Table.

Table 2

Influence of precipitation in the risk of ocular toxoplasmosis on different lag days.

Department	lag 0–6Mean of relative risk RR (95% CI)	lag 6–10Mean of relative risk RR (95% CI)	lag 10–12Mean of relative risk RR (95% CI)	lag 13–15Mean of relative risk RR (95% CI)
Bogotá	1.001(0.998–1)	1.001(0.999–1)	1.001(1–1)	1.001(0.999–1)
Cauca	1.002(1–1)	1(0.999–1)	1(0.999–1)	0.9998(0.998–1)
Chocó	0.9998(0.998–1)	1(0.999–1)	1.001(1–1)	1.001(1–1)
Guajira	1.007(1–1.01)	1.005(1–1.01)	1.005(1–1.01)	1.005(1–1.01)
Huila	0.9955(0.994–0.997)	0.9975(0.997–0.998)	0.9983(0.997–0.999)	0.998(0.997–0.999)
San Andrés	1(0.984–1.02)	1(0.992–1.01)	1(0.992–1.01)	1(0.991–1.01)
Sucre	1(0.998–1)	1(0.999–1)	0.9996(0.999–1)	0.9991(0.998–1)

In the analysis by the mean precipitation in 2019 (Fig 4A), Chocó and Antioquia and the departments bordering the Andean mountains have a higher precipitation frequency than the others. However, this does not correspond with the departments with the highest risk of OT in the same period, as the highest incidence rates were found in Caquetá and Caldas (Fig 4B). Nevertheless, since the average of precipitation tends to vary over time, the risk could also change. This analysis supported the use of DLNM to observe the dynamic effect of exposure over time. Additionally, we estimated a positive β coefficient (0.03) with P< 0.05 between precipitation and OT risk overall, which suggests that precipitation significantly influences the response variable (risk of OT) in all the evaluated departments. The analysis for other years could not be carried out since we did not have their daily data.

Fig 4

Precipitation in Colombia and incidence rate of ocular toxoplasmosis (OT).

(A) The precipitation map of the departments in Colombia, which shows the average annual value of precipitation in each department for the year 2019; we were able to collect the information for all the departments from the National Institute of Hydrology, Meteorology, and Environmental Studies (IDEAM)[25]. (B) The estimated relative risk of OT with the fitted conditional autoregressive (CAR) model for the data available in 2019. Caquetá, Caldas, and Quindío have a higher relative risk, and departments with less precipitation, such as Atlántico and La Guajira, have a lower relative risk. Map was created based on https://cran.r-project.org/web/packages/leaflet/index.html.

Validation of a previous hypothesis

We performed a Pearson correlation test for the available data to compare with the results of Gómez-Marin et al.[10] on the correlation between precipitation and Tg infection rates in some Colombian departments. Considering that IDEAM did not provide sufficient raw data from Caquetá and Quindío, we collected the average precipitation data for these departments from the secondary data available for 2019 [25]. This analysis showed a correlation between precipitation and incidence of OT of 0.84 (P< 0.05), supporting the results of their hypothesis. Fig 5A and 5B demonstrates the assumption.

Fig 5

The mean number of cases of ocular toxoplasmosis (OT) and precipitation by department in 2019.

(A) Represents the average number of cases of ocular toxoplasmosis (OT) in salmon bars and average precipitation by the department in the blue line (For Caquetá and Quindío only the information for 2019 was available). (B) The graph shows an incremental trend in the number of OT cases directly proportional to the precipitation in most cases (Pearson correlation test 0.84; P< 0.05). The data for this subanalysis were taken from IDEAM [25].

Discussion

Some studies relate rainfall with congenital toxoplasmosis development, OT relapses, and outbreaks. Regarding congenital infection, Gómez-Marin’s et al. study [10] found a statistically significant association between annual rainfall and congenital toxoplasmosis. With respect to OT relapses, a study carried out in Argentina found that the frequency of toxoplasmic reactivation episodes increases when precipitation increases and that the mean annual rainfall could be a predictor of the frequency of reactivations; for every mm of rainfall, there was a 2% increase in the reactivation episodes (OR = 1.002, 95% CI = 1.000–1.003, P = 0.019) [11].

As in other foodborne infectious diseases, the incidence of OT is influenced by environmental factors. For example, OT outbreaks have been associated with contaminated water and undercooked food [30-35]. Regarding outbreaks and rainfall, a study from British Columbia found that the increased precipitation between 1994 and 1995 was associated with a waterborne outbreak of toxoplasmosis. They noted an increase in the number of OT cases, either congenital or acquired [34]. In Coimbatore, India, in September 2004, 178 patients had de novo retinochoroidal lesions and positive antibodies for Tg. All cases were closely related to the Siruvani reservoir, which supplies water to the city of Coimbatore and received rainwater due to increased rainfall that year, which could be related to the development of new OT cases [35]. Nevertheless, these studies did not find a statistical correlation between rainfall and OT.

Our results and those of the abovementioned studies could be directly related to the viability of Tg oocysts in soil. In one study, after 100 days, 7.4% and 43.7% of oocysts survived under dry and damp conditions, respectively [12]. Tg oocysts survive up to 54 months in cold water [13]. Furthermore, water contaminated with the parasite can potentially cause large toxoplasmosis outbreaks [13]. Small parasites can be filtered out by coagulation, flocculation, and settling used in most municipal water treatment systems in developed countries. However, developing countries do not always have these complex systems; therefore, oocysts approximately 8–12 μm in size could not be filtered in the water [13,36,37].

In the current study, we conducted a temporary analysis, which has never done before on the subject, evidencing a clear benefit to establishing the impact of different precipitation levels over time and revealing findings that cannot be appreciated in the clinical setting immediately after rainfall episodes. This type of analysis gives a clearer perspective of the effect of a climatic event on the population. We found a significant difference in the number of cases per department that correlates with what was found in the patterns of single-day lag–response curves and cumulative exposure–response curves. We can state that rainfall in Colombia is a factor that influences the presentation of OT (as a risk or protective factor), probably depending on the population behavior since, in some cases, unknown variables significantly reduce this risk, as shown in S1 and S2 Figs.

We suspect that the orographic barrier that surrounds the Andes mountains and induces the formation of local and regional high complexity climates in Colombia affects precipitation behavior, turning it highly variable [38]. Furthermore, the influence of thermodynamic processes in the Atlantic and Pacific oceans through the El Niño, La Niña-Southern Oscillation cycle can influence the climate and precipitation differently in each region across the country [39]. Both phenomena explain the variability of precipitation in our data by departments.

The cumulative exposure–response curve showed a downward trend for most of the departments with available data (Fig 1), indicating that the influence of precipitation on OT decreased as the exposure accumulated. However, the single-day lag–response curves were highly variable between departments, with no clear predominant trend (shown in S2 Fig). The cumulative exposure–response curves and Table 2 show that Chocó, Bogotá, Cesar, Cauca, and Guajira have a pattern that is contrary to that observed in the rest of the country, with an increase in the RR for OT directly proportional to the mean precipitation. This supports the theory that precipitation is a risk factor. Chocó is one of the areas with the highest rainfall in the world, with an average annual precipitation of 8,000 to 13,000 mm. The data demonstrated that precipitation in this department increased the risk of OT from lag0–6 (RR 0.99) to lag13–15 days (RR: 1.01) [14].

It is also important to consider how cultural habits could explain the variability in RR of OT in the proposed model. In Chocó, when the rainy season begins, the RR may decrease because the population does not go out to collect water or work in the rivers. Previous studies have shown that behaviors tend to vary significantly in populations with a higher risk of floods and rising rivers [40,41]. However, over time the risk tends to increase in Chocó. It may also be due to the presence of contaminated waters, as Chocó is one of the country’s poorest regions with the most limited access to health, food, and aqueduct; in fact, only 35% of the population have access to this essential service [42-44].

As for Bogotá, a direct relationship between precipitation and OT risk was observed. This finding provides strong data, as this is the capital city with better health access and reporting systems. Therefore, the general decreasing risk in most departments in Colombia, one of the endemic countries, may be strongly affected by the underreporting of the disease throughout the country, suggesting that the risk may be even greater than expected, for example, inadequate reporting technique could explain that San Andrés does not have new cases reported [45,46].

Our results suggest that many other environmental variables could influence the relationship between OT and precipitation, and they must be assessed locally and through a temporal analysis. For example, the wind is an ecological variable involved in oocyst sporulation and dispersal (contamination) in water sources, soil, and food [47-49]. Studies have even proposed that oocysts could aerosolize. An epidemiological study showed an outbreak in people from an equine stable, suspecting that the cause was aerosolized oocysts. However, their attempts to isolate the parasite in different samples did not show the presence of oocysts at any time [48]. Interestingly, in the Guajira, the average wind speed values are higher than nine m/s. In the Andean region, a wind corridor reaches average wind speed values above five m/s [50,51]. Therefore, this and other environmental factors should be analyzed in further studies to confirm an association. We propose it as an example of variables that confound the effect of precipitation on the risk of OT.

Limitations

Given that our study was a population-based one, we did not have access to the clinical records of patients and could not confirmed the serological status. However, we only included new confirmed cases with code (B 58.0), which ophthalmologists commonly use. Additionally, data from the national databases may have inherent coverage and content errors that underestimate the number of new cases. Moreover, most of the ophthalmology and tertiary care centers are located in capital cities, such as Bogotá, which can generate some bias in the analysis. Furthermore, this explains why departments with a high incidence of OT, such as Quindío, did not have too many cases. Finally, due to limitations in the data provided by IDEAM, some departments could not be evaluated in detail and therefore were not included in the analysis. From the 2,157 files that they have sent for the specified period, only 825 (38%) files had the required information. For instance, in the data obtained from the National Registry in 2017, the department of Chocó had 30 municipalities, of which only 8 (26.7%) reported information on the monitoring of the quality of water for human consumption [52].

Conclusions

In the present study, our temporal analysis showed the impact of different precipitation levels over time on the RR for OT. It revealed findings we could not otherwise appreciate in the clinical setting immediately after rainfall episodes. There were variable tendencies according to the department evaluated. Depending on the region, the precipitation increased, decreased, or showed no relationship with RR for OT. We hypothesize that other sociodemographic and environmental variables, such as wind, could influence the RR. Additionally, the correlation analysis shows a directly proportional relationship between the increased number of cases and the increase in precipitation. Data regarding precipitation should be analyzed individually for each region and particular context. The influence of other environmental, behavioral, and sociodemographic factors should be examined in the departments with a higher risk.

Cumulative-response curves for the association between precipitation and new cases of ocular toxoplasmosis in Colombia, 2015–2019.

(DOCX)

Click here for additional data file.

Single-day lag-response curves for the association between precipitation and new cases of ocular toxoplasmosis in Colombia, 2015–2019.

(DOCX)

Click here for additional data file.

Mean of Relative risk (RR) and 95% confidence intervals (95% CI) in each department of Colombia of ocular toxoplasmosis cases.

(DOCX)

Click here for additional data file.

21 May 2022

Dear Dr de-la-Torre,

Thank you very much for submitting your manuscript "Exploring the association between precipitation and population cases of ocular toxoplasmosis in Colombia" for consideration at PLOS Neglected Tropical Diseases. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Christine A. Petersen

Deputy Editor

PLOS Neglected Tropical Diseases

Christine Petersen

Deputy Editor

PLOS Neglected Tropical Diseases

***********************

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: There is a clear objective for this study, but no state hypothesis from the authors. Additionally the authors need to make clear if there was any missing precipitation data and the potential impacts of missing precipitation data. the authors also need to account for population differences in the geographical regions of interest presented in this study.

Reviewer #2: (No Response)

--------------------

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: the results followed and analyses discussed in the methods section. Figure 1 is not clearly presented. Supplemental material 1 and 2 should be included as in-text tables for this manuscript.

Reviewer #2: (No Response)

--------------------

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: the authors conclusions are supported by the results. Additionally the authors do a nice job discussing the limitations of their work and data.

Reviewer #2: (No Response)

--------------------

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

--------------------

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: This study evaluated the relationship between precipitation and ocular toxoplasmosis incidence. I have some comments below on your manuscript:

a. Introduction: The authors should consider adding more information on the protozoan lifecycle and biology that allows it survive in conditions outside of its hosts. Additionally the authors should discuss routes of infection and what can influence infection.

b. Methods: the authors should note what months and years they gathered data for in regards to their outcome data (population data section) in a similar way it is mentioned in the precipitation data.

c. Methods: The authors state “825 files” were selected on line 148. Do these files represent data from 825 unique stations for the four years they were interested in researching? Or is it 825 unique files with data from all weather stations available?

d. Methods: the authors should state if there was missing data in regards to precipitation data, and if there was missing data how was in handled?

e. Methods: the authors should make remarks on how they accounted for population size differences between geographic regions that could account for some differences they see in counts of OT per region.

f. Results: Supplementary Material 1 and 2 should be included as in-text tables as they describe the authors data used for their results

g. Results: In line 209, the authors describe different departments. I believe departments is referring to geographical regions of Columbia. The authors should make this more clear throughout the manuscript and provide greater discussion in their methods that these are analyses are performed for each geographic region.

h. Results: the authors should give a greater description of what a characteristic pattern is for line 217.

i. Results: The legend in figure 4B would indicate that precipitation is associated with a decrease in risk since RR < 1. Is the legend presented in this figure accurate?

j. Results: in line 273-275 the authors should state there is an association between precipitation and increased OT rather than the causal language currently used. Direct causality is difficult to prove and assess in a population study.

k. Results: For data presented in supplementary Material 2, is this average of daily averages, or the average of hourly precipitation recorded from weather stations? Authors should make this clearer.

l. Results: Figure 1 is difficult to read the graphs. Either splitting the map into northern and southern compartment to make a two part figure would be helpful. Or placing an asterisk (or some denoting feature) on the 7 characteristic patterns would provide improved readability of these graphs. Or decreasing the size of the map and increasing the graphs.

m. Results / Discussion: Authors should discuss why in some regions there is large variability in RR (Cauca, Sucre, San Andres, etc.) where in others there is very little RR variability

Reviewer #2: In this manuscript, the authors describe the association between precipitation and occurrence of ocular toxoplasmosis (OT) cases in different location in Colombia. OT is a serious clinical problem in South America, and studies to understand the interaction between condition and its environment are more than welcome to enable efficient public health measures on a local level. As previous studies (cited in the manuscript) showed, precipitation is one of the important factors. This kind of study is well adapted for publication in PLOS Neglected Tropical Diseases. The senior author has a solid publication list in OT clinical and basic science. The authors collected a huge data set to explain different ways of how precipitation affects OT cases, more than enough to justify an article. Unfortunately, the manuscript is written in a language and style apparently destined to meteorology experts, including a lot of formulas, techniques and interpretations specific to this area. In my opinion, an article in PLOS NTD should be accessible to readers trained in infectiology and tropical health science. Being such a reader, I considerably struggled to understand the methods, results and conclusions of the study. So, to be able to evaluate the manuscript correctly, I suggest the following points:

1 Clearly explain the methods used, and the terms employed for interpretation (lag time…). For example, the authors state that the filter … was applied for each year (l. 140), but the term ‘year’ never appeared in the results. In l. 154, it should be explained what exactly means ‘case period’ and ‘nearby referent periods’. The two following sentences are, as generally all this statistical analysis (= methods sensu stricto), very difficult to understand for the average infection/tropical health reader. The second paragraph of this section seems to be important, but should be clearly explained. This does, of course, not exclude some technical formulas, when well explained.

2 The results are also difficult to decipher. The main findings should be explained in clear statements for each figure. The legend to Fig.3 is very difficult to understand, to give just one example.

3 The text treating Fig4 and Table 1 (l.270-277) is quite difficult to follow, for different reasons: The methods employed are not clearly explained to the average reader (cf. point 1), the language of some sentences lack themselves clarity, e.g. l.272-273, and finally, the principal findings of the Table and Fig. are not clearly resumed in the text.

4 The discussion also lacks clear interpretation of data, stating basically that rainfall can both be risky and protective, for unknown ‘population behavior’ (l. 311f) reasons. Apart from the wind, at the end of the discussion, none is really detailed, and maybe taken as example for specific districts.

In summary, I think that the manuscript contains valuable information, but should be explained in a more commonly understandable fashion and language. I have just some more general points:

5 Most of the cited articles relate high precipitation levels and toxoplasmosis outbreaks to failures of the filtration system in public water supply. Here, it should be at least discussed if more rain acts through longer oocyst survival or more oocysts leaking into the public water supply.

6 I admit that I did not entirely understand the meaning of the term ‘delay’ throughout the manuscript, but if this means that OT cases were observed with a delay of a few days after rainfall episodes, this has to be explained. Ocular affection usually takes time to develop following infection. Moreover, can new detected OT cases serve as measure for new general infections?

7 Do the data allow to confirm the results in Ref 9 (Gomez-Marin et al.) of a correlation between rainfall and T. gondii infection rates in Colombian districts?

--------------------

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

12 Jul 2022

Submitted filename: Answer for reviewers.pdf

Click here for additional data file.

15 Aug 2022

Dear Dr de-la-Torre,

We are pleased to inform you that your manuscript 'Exploring the association between precipitation and population cases of ocular toxoplasmosis in Colombia' has been provisionally accepted for publication in PLOS Neglected Tropical Diseases.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Christine A. Petersen

Section Editor

PLOS Neglected Tropical Diseases

Christine Petersen

Section Editor

PLOS Neglected Tropical Diseases

***********************************************************

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: The objective of the study is clearly articulate with a testable hypothesis. the study design, data, and analysis are clearly stated. The population is clear and is appropriate for the hypothesis. Sample size is sufficient for adequate power.

Reviewer #2: (No Response)

**********

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: The analysis presented matches the analysis plan described in the methods. The results are clear and completely presented and show a vast improvement over the previous draft. In particular figures and figure legends are more clear.

Reviewer #2: (No Response)

**********

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: Conclusions are supported by the data presented and limitations are clearly presented.

Reviewer #2: (No Response)

**********

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: Large improvement over previous submission. Methods and results are more clear and reader friendly.

Reviewer #2: According to the reviewers’ remarks, the manuscript has been ameliorated. I have just still a few remarks:

1 When I mentioned ‘language’, it did not mean the English proficiency, but the technical language used throughout, which is actually quite difficult to understand for an average expert in infectious diseases. So, every finding and the implication for oocyst infectivity should be explained in usual epidemiologic terms, in addition to the already present technical terms, of course. This should also include some statements why the effect is seen so fast after a single episode of rainfall, when the authors state that oocysts are viable for several months in humid conditions, as found in Colombia.

2 Fig.5: It is good that this analysis has been included, but it seems to me that the type of graph is not adapted to a correlation analysis. Furthermore, ‘mean of total cases’ and ‘precipitation’ should be specified.

3 The number of OT cases should also be calculated per inhabitants, as the departments vary considerably in population density.

4 Just one remark on (English) language, l. 419: ‘Lousy reporting technique’ should be substituted by a more neutral term.

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

15 Sep 2022

Dear Dr de-la-Torre,

We are delighted to inform you that your manuscript, "Exploring the association between precipitation and population cases of ocular toxoplasmosis in Colombia," has been formally accepted for publication in PLOS Neglected Tropical Diseases.

We have now passed your article onto the PLOS Production Department who will complete the rest of the publication process. All authors will receive a confirmation email upon publication.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any scientific or type-setting errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript. Note: Proofs for Front Matter articles (Editorial, Viewpoint, Symposium, Review, etc...) are generated on a different schedule and may not be made available as quickly.

Soon after your final files are uploaded, the early version of your manuscript will be published online unless you opted out of this process. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Shaden Kamhawi

co-Editor-in-Chief

PLOS Neglected Tropical Diseases

Paul Brindley

co-Editor-in-Chief

PLOS Neglected Tropical Diseases