Introduction

Toxoplasma gondii (T. gondii) is a protozoan parasite that causes infections in humans and animals [1]. Infections with T. gondii occur worldwide [2] and nearly one-third of humanity has been exposed to this parasite [3]. Transmission routes of T. gondii infection include ingestion of cysts in meat and meat derivates, milk contaminated with tachyzoites, oocysts in water, sand, and soil, and oocysts in raw fruits and vegetables [4]. Alternative routes for transmission are organ transplantation [5,6] and blood transfusion [5,7]. T. gondii can produce outbreaks [4]. Primary infection is usually subclinical but, in some patients, cervical lymphadenopathy or ocular disease can be present [8]. However, primary infections during pregnancies can cause blindness and mental retardation in congenitally infected children [3]. In addition, T. gondii infection can cause devasting disease in immunocompromised individuals [3]. T. gondii resides, in a latent form, in the central nervous system and drastically alters the behavior of rodents and is associated with the incidence of specific neuropsychiatric conditions in humans [9]. Infections with T. gondii have been associated with mental disorders as obsessive-compulsive disorder [10], schizophrenia [11,12], generalized anxiety disorder [13], mixed anxiety and depressive disorder [14], bipolar disorder [15], and depression [13,15]. Furthermore, high T. gondii antibody titer has been associated with suicide attempts [16,17].

There is scanty information about the association between of T. gondii infection and violent behavior. Infection with T. gondii was associated with nonfatal suicidal self-directed violence [18,19]. Aggression and impulsivity were associated with latent T. gondii infection in psychiatrically healthy subjects [20]. In another study, researchers found that T. gondii seropositivity and high blood kynurenine have a cumulative effect on the risk of nonfatal suicidal self-directed violence in schizophrenic patients [21]. In the present study, we investigated the question as to whether T. gondii infection might be associated with violent behavior in inmates. Therefore, we sought to evaluate the association between T. gondii seropositivity and violent behavior in a sample of inmates in the northern Mexican city of Durango.

Materials and methods

Violent behavior

Three different criteria for defining a violent behavior in inmates were used: 1) an evaluation based on the Historical, Clinical and Risk Management-20 (HCR-20) tool that assesses the risk of violence [22]; 2) an evaluation based on the type of the crime for which inmates were convicted; and 3) Assessment of aggression by the Buss-Perry aggression questionnaire (AGQ) [23–26]. HCR-20 assessments were performed at admittance of inmates to jail, and by using this tool the correctional facility classified inmates into three groups of violence risk: low (scores 0–11), medium (scores 12–21), and high (scores 22–30). Individual HCR-20 scores were not provided by the correctional facility but provided a two-group classification: violent (high HCR-20 scores) and non-violent (HCR-20 scores lower than 22). Analysis was thus performed considering these two groups. Concerning the evaluation based on the types of crimes, violent inmates included participants convicted for aggravated assault, murder, rape, rape attempt, domestic and non-domestic violence, possession of weapons, wounding, kidnapping, and indecent behavior with people. Whereas inmates were considered as non-violent when they were convicted for burglary/robbery/theft, sale and possession of drugs, entering in a property without permission, corruption of minors, and fraud. The AGQ includes items to assess four dimensions of aggression: physical aggression, verbal aggression, anger, and hostility [23]. An inmate was considered as violent when he/she obtained an AGQ overall score of 83 or more. Whereas inmates who scored 82 and lower were considered as non-violent. The use of these questionnaires was a minimal cost methodology. General sociodemographic characteristics of the violent and non-violent inmates is shown in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Sociodemographic characteristics of the study population.

https://doi.org/10.1371/journal.pone.0284202.t001

Results

Anti-T. gondii IgG antibodies were found in 6 (4.7%) of the 128 inmates studied. With respect to the HCR-20 criteria, 17 (13.3%) of the 128 inmates had high risk of violence. Whereas using the evaluation based on the type of the crime for which inmates were convicted, 72 (56.3%) of the 128 inmates were considered violent. Concerning the AGQ, 59 (46.1%) of the 128 inmates were considered violent. Results of the rates of T. gondii seropositivity and violence in inmates are shown in Table 2. Of the 128 inmates, 113 (88.3%) were considered as violent in at least one of the evaluations, 50 (39.1%) in at least 2 evaluations, and 4 (3.1%) in the 3 evaluations. Depending on the evaluation method to detect violence, the seroprevalence of T. gondii infection in violent inmates varied from 0% to 6.9%. No statistically significant difference in anti-T. gondii IgG seroprevalence between violent and non-violent inmates was found. This was the case when inmates were considered violent in at least one evaluation (HCR-20, crimes, or AGQ), in at least two evaluations or in the three evaluations.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Association between T. gondii seropositivity rates and violence groups.

https://doi.org/10.1371/journal.pone.0284202.t002

Table 3 shows the results of the AGQ scores found in inmates and the seroprevalence of T. gondii infection. Mean scores of the AGQ in T. gondii seropositive inmates (73.67 ± 29.09; 95% CI: 50.00–99.31) were similar to those (79.84 ± 25.00; 95% CI: 75.46–84.27) found in T. gondii seronegative inmates (P = 0.55). Stratification by the type of aggression showed that the mean scores of anger, psychical aggression, verbal aggression, and hostility in T. gondii seropositive inmates were similar to those found in T. gondii seronegative inmates.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Association between T. gondii seropositivity rates and AGQ scores.

https://doi.org/10.1371/journal.pone.0284202.t003

Violence and sociodemographic characteristics of inmates did not associate with T. gondii infection by logistic regression analysis (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Logistic regression analysis of characteristics of inmates and their association with T. gondii infection.

https://doi.org/10.1371/journal.pone.0284202.t004

Discussion

Little is known about the link between T. gondii infection and violence. Infection with T. gondii has been found associated with self-directed violence in people in Sweden [19], mothers in Denmark [18], and schizophrenic patients in the USA [21]. In addition, infection with T. gondii was associated with aggression and impulsivity in psychiatrically healthy adults in Germany [20]. However, the association between infection with T. gondii and violence has not been studied in inmates. Inmates may be considered as a population group with high rates of violence. Many inmates have been convicted for violent crimes. Therefore, in the present study we sought to determine the association between T. gondii seropositivity and violent behavior in a sample of inmates in a state correctional facility in Durango, Mexico. We found that the seroprevalence of T. gondii infection did not vary among violent and non-violent inmates as detected by the HCR-20, crimes committed, or AGQ. Results thus suggest that infection with T. gondii is not associated with violence in inmates in our setting. Apparently, our results conflict with those found in other studies on the positive association between T. gondii infection and violence. However, comparison of our results with those found in other studies should be interpreted with care since difference on the characteristics of the study populations, types of violence, and criteria for the evaluation of violence among the studies exists. Firstly, the positive association between T. gondii infection and self-directed violence found in people in Sweden [19], mothers in Denmark [18], and schizophrenic patients in the USA [21] cannot be fairly compared with the lack of association found in our study. This is because we studied violence of inmates against other people whereas other researchers studied self-directed violence [18,19,21]. In addition, our study population has different characteristics as gender and clinical diagnosis from the population groups of other studies including mothers [18], schizophrenic patients [21], and people with and without suicide behavior [19]. On the other hand, in a study of psychiatrically healthy adults, researchers evaluated aggression with the Questionnaire for Measuring Factors of Aggression (FAF) [20] whereas we evaluated aggression with the AGQ.

We found a low seroprevalence of T. gondii infection in inmates in general, and this finding was unexpected since a high (21.1%) seroprevalence of T. gondii infection in inmates was previously reported [28]. In addition, we found a low seroprevalence of T. gondii infection in violent inmates. This low seroprevalence in violent inmates is comparable with a 6.1% seroprevalence of T. gondii infection reported in the general population in Durango City, Mexico [27]. In addition, the low seroprevalence found in violent inmates in this study is lower than those reported in other populations groups in the same Durango City including waste pickers (21.1%) [29], schizophrenic patients (20%) [12], and female sex workers (15.4%) [30]. The lack of association between IgG seropositivity to T. gondii and violence found in this study does not necessarily mean that T. gondii infection might not be linked to violence. We assessed violence by using three methods and determined T. gondii infection by IgG serology but more methods to assess violence and to determine T. gondii infection including molecular methods can be used to determine the association between T. gondii infection and violence in inmates. Further studies to confirm or challenge our results are needed.

This study has some limitations including a small sample size, a low prevalence of T. gondii infection among the studied population and was performed in only one correctional facility. Further studies with larger sample sizes and in several correctional facilities to determine the association between T. gondii infection and violence in inmates should be conducted.

Conclusions

Results of this study using three types of evaluations of violence suggest that seropositivity to T. gondii is not associated with violence in inmates in Durango, Mexico. However, further studies with larger sample sizes and in several correctional facilities to determine the association between T. gondii infection and violence in inmates are needed.

Supporting information