Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 May 1.
Published in final edited form as: Neurobiol Aging. 2017 Feb 3;53:76–82. doi: 10.1016/j.neurobiolaging.2017.01.018

Association between Toxoplasma gondii Seropositivity and Memory Function in Non-Demented Older Adults

Cynthia P Wyman a, Shawn D Gale a,b, Ariana Hedges-Muncy a, Lance D Erickson c, Eric Wilson d, Dawson W Hedges a,b
PMCID: PMC5482532  NIHMSID: NIHMS849285  PMID: 28235681

Abstract

Toxoplasma gondii (T. gondii) seropositivity may be associated with decreased memory in older adults. To further investigate the association between T. gondii seropositivity and memory in non-demented older adults, we obtained serum samples from 114 non-demented older adults evaluated by the Alzheimer’s Disease and Research Center at Washington University in St. Louis Missouri, USA. We determined T. gondii seropositivity and anti-T. gondii IgG antibody titer and examined associations with memory function while controlling for socioeconomic status, education level, age, and apolipoprotein E4 status. There were few associations between T. gondii seropositivity or anti-T. gondii IgG antibodies and memory, although there was some support suggesting an interaction between anti-T. gondii and sex. In the seropositive-only sample, there was an inverse relationship between anti-T. gondii titer and performance on the selective reminding test. Overall, we found little evidence of an association between impaired memory function and T. gondii seropositivity and anti-T gondii IgG antibodies in this sample of non-demented older adults.

Keywords: Toxoplasma gondii, toxoplasmosis, neurocognitive function, aging, memory

1. Introduction

Multiple factors are associated with cognitive function and cognitive aging in older adults including sociodemographic variables, lifestyle, and diseases such as hypertension and diabetes (Gifford, et al., 2013, Stewart and Liolitsa, 1999, Zhang, et al., 2015). Furthermore, many of these factors may increase or protect against the risk of developing dementia (Xu, et al., 2015). In addition, the association between various infectious diseases, cognitive function, and risk for cognitive impairment and even dementia in older adults has been increasingly investigated (Barnes, et al., 2014, Katan, et al., 2013, Watson, et al., 2013).

Among the infectious diseases implicated in affecting cognition is Toxoplasma gondii (T. gondii), a protozoan parasite whose definitive host is any member of the cat family including domestic cats. Affecting approximately 12% of the United States’ population (Jones, et al., 2014a), T. gondii can be transmitted to humans through the ingestion of oocysts (from cat feces) or through the ingestion of tissue cysts containing infectious bradyzoites from undercooked beef, pork, and other meats (Jones, et al., 2014b). Although T. gondii infection in humans often goes unnoticed in immunocompetent individuals, T. gondii tissue cysts can persist in human brain and muscle tissue (Montoya and Liesenfeld, 2004). T. gondii seropositivity increases with age, and 41% of older adults in the United States ages 60 to 90 years (median age = 69) may be infected (Mendy, et al., 2015). Further, T. gondii seropositivity appears to be associated with behavioral, cognitive, and neurological abnormalities in humans (Fekadu, et al., 2010, Flegr, 2015). Specifically, Kusbeci et al. (2011) found a significantly higher prevalence of T. gondii seropositivity in Alzheimer’s disease compared to non-demented controls, suggesting T. gondii seropositivity may possibly contribute to dementia. However, a subsequent study by Perry et al. (2016) found no association between Alzheimer’s disease and T. gondii seropositivity in a larger sample of older adults (114 control subjects and 105 Alzheimer’s disease subjects). Thus, the role of T. gondii in the development of neurological diseases affecting memory remains unclear.

In one of the few studies investigating the association between T. gondii seropositivity and cognitive function in older adults, Mendy and colleagues (2015) analyzed a database from the Centers for Disease Control in the U.S. that included 4,485 participants with a median age of 69 years and compared the performance of subjects T. gondii seropositive and seronegative on two brief memory measures, one of which was the recall of three words from the Mini-Mental State Examination and the other the recall of a brief story. Controlling for covariates including age, sex, race-ethnicity, and education, they found that T. gondii seropositivity was associated with worsened immediate but not delayed recall. Gajewski et al. (2014) administered a much more detailed cognitive battery to a group of 42 elderly adults (mean age = 69.8 years) T. gondii seropositive and 42 T. gondii seronegative (mean age = 70.8) and found that verbal learning and delayed recall for a 15-word list were lower in the seropositive group.

Based on previously reported associations between T. gondii seropositivity and deficits in immediate memory, delayed recall, and verbal learning in older adults and the widespread distribution of T. gondii seropositivity in older adults, we sought to further investigate associations between T. gondii seropositivity and memory function in a well characterized sample of older adults.

2. Methods

2.1 Study Sample

We obtained serum samples from 114 non-demented older adults evaluated by the Alzheimer’s Disease and Research Center (ADRC) at Washington University in St. Louis Missouri, USA, one of many Alzheimer’s Disease Centers across the United States funded by the National Institute of Aging and the same sample we used for a control group in a previous study (Perry, et al., 2016). With few exceptions, all subjects had clinical dementia rating (CDR) (Morris, 1993) scores of 0 (no dementia) and the cognitive status of each subject was not suggestive of dementia (McKhann, et al., 2011). Accordingly, all subjects in the current study had been defined as non-demented control subjects.

2.2 Determination of T. gondii Seropositivity

To identify T. gondii seropositivity, we assayed serum samples by quantitative ELISA for the presence of IgG anti-T. gondii antibodies using a commercially available ELISA kit (GenWay, BioTech). We did all assays according to the manufacturer’s recommended protocol and ran all samples in duplicate. In accordance with the manufacturer’s guidelines, we considered samples to be seropositive for T. gondii when the T. gondii IgG-class antibody titer was greater than 35 IU/mL. We considered samples below 30 IU/mL to be seronegative according to the manufacturer’s guidelines. We assayed samples in the equivocal range, measuring between 30 and 35 IU/mL, a second time. Positive controls included those provided by the kit manufacturer as well as an additional patient sample that was confirmed seropositive through the use of a kit produced by a different manufacturer (Abcam). In addition to positive controls, we assayed three to four patient samples from each plate on a subsequent plate to confirm reproducibility of the results. We diluted any samples measuring outside of the standard curve (generated on each plate) five and ten-fold and then assayed them again with results adjusted according to the dilution factor. The researcher conducting all assays was blind to all demographic variables. The same researcher performed all ELISA assays in an effort to reduce any potential user variation.

The presence of IgG anti-T gondii antibodies in the serum in the absence of IgM antibodies is indicative of a past infection with T gondii, whereas the presence of IgM antibodies is considered indicative of an acute or reactivated T. gondii infection. In an effort to distinguish between acute and a past infection with T. gondii, we assayed all samples seropositive or borderline positive for IgG antibodies to determine the titer of IgM anti-T. gondii antibodies using a commercially available IgM specific ELISA (GenWay, BioTech) following the protocol provided by the manufacturer. According to the protocol, we obtained the Toxoplasma IgM index by dividing the mean values of each sample by the calibrator mean value considering a Toxoplasma IgM index greater than 1 as positive for IgM antibodies to T. gondii, an index between 0.91 and 0.99 as equivocal, and an index below 0.90 as negative. As with the IgG assays, we ran all IgM assays in duplicate, and the researcher performing these assays was blind to all demographic data. We ran a positive and negative control in each experiment, and all results were within the index range for positive and negative controls.

2.3 Cognitive Function

Available measures of memory included the Logical Memory Story A (LM IA), Immediate and Delayed recall (LM IIA) from the Wechsler Memory Scale-Revised (Wechsler, 1987), Associate Learning from the Wechsler Memory Scale (Wechsler, 1973) and the selective reminding test (SRT) (Grober, et al., 1988). We included the total recall score from the SRT, which includes the number of items recalled on the three cued recall trials. All of these tests measure explicit memory.

2.4 Estimate of Socioeconomic Status

In this sample, the Hollingshead two-factor index of social position provided an estimate of socioeconomic status, with “1” on the scale representing higher socioeconomic status with the range of possible values on the scale being from 1 to 5 (Hollingshead and Redlich, 1958). For the analyses, we dichotomized this scale to high and low SES for ease of interpretation. Specifically, we considered scores from 1–3 as high and scores 4 and 5 as low.

2.5 Statistical Analyses

We used descriptive statistics to characterize the sample according to age, sex, socioeconomic status, educational attainment, apolipoprotein E4 (ApoE) status, average antibody titer, and performance on each of the memory tasks. We used a natural-logarithm transformation to normalize the distribution of the antibody titer and t-tests to compare scores on each measure between the T. gondii seropositive and seronegative groups. In the statistical analysis, we used two methods: linear regression and a post-hoc Bayesian analysis, both modeled without an intercept (Rencher and Schaalje, 2008). To ensure the assumptions of non-intercept regression were not violated, we centered both the age and antibody titers around zero.

We first used linear regressions to predict scores in each memory task using two models. The first model included the dichotomous measure of T. gondii seropositivity or seronegativity, and the second included the log-transformed, continuous measure of anti-T. gondii IgG antibody titers. Both models included sex, socioeconomic status, education level, age, and apolipoprotein E4 (ApoE) status (presence or absence of E4 allele) as control variables. We included ApoE status as a control variable due to its known association both with Alzheimer’s disease and cognitive decline (Dorey, et al., 2014). Specifically, Levy et al. (2004) found those carrying an E4 allele of ApoE had lower scores on Wechsler Memory Scale – Revised Logical Memory II. Given these findings, we wanted to control for the potential effect ApoE status may have on memory function and thus included it in our analyses. In each model, the differences in sample size are due to missing data on the dependent variables.

In addition, we analyzed the associations between the T. gondii seropositivity and performances on the memory tasks using the same models as mentioned previously but with an interaction included as we were interested in the effect of titer and sex. Also, in post-hoc analyses, we used the same linear regressions and Bayesian models to predict scores in each memory task using anti-T. gondii IgG antibody titers for T. gondii seropositive samples separately.

Because our sample size was relatively small, we performed a post-hoc power analysis for both the main effects models and the T. gondii seropositivity and sex interaction models with alpha set at .05 using G*power 3.1 software (University of Düsseldorf, http://www.gpower.hhu.de).

Because of the potential for alpha inflation due to multiple comparisons in linear regression, we also used a Bayesian analysis to further evaluate our findings. In Bayesian analyses, there is not a problem with alpha inflation from multiple comparisons because the parameters have a joint posterior distribution, and Bayesian analyses are constructed on understanding the limits of applicability instead of hypothesis testing (Gelman, et al., 2013).

In order to predict scores in each memory task, we fit a Bayesian generalized linear model using two sets of regressors, as we did in the above linear regression. The first set included the dichotomous measure of latent toxoplasmosis (Tx), and the second included the log-transformed, continuous measure of anti-T. gondii IgG antibody titers (LnTx). Both models included sex, socioeconomic status, education level, age, and ApoE status (presence or absence of E4 allele) as control variables. As we assumed the missing data was missing at random, the differences in sample size are due to taking out the missing variables.

Because each memory task is in a bounded region, we transformed each score y in the Bayesian analysis to be y∗=yb where b is the maximum score given the memory task. When the extreme scores (either a 0 or the maximum) was reached, the transformation for all scores was then y∗=(yb(n−1)+0.5)/n as suggested by Smithson and Verkuilen (2006). After the transformation, we have

y~Beta(μ,ϕ)

where μ is the mean and ϕ is the precision parameter. Additionally, for the generalized linear model, we used the linear predictor η =Xβ where β are the coefficients for the independent and control variables. Finally we used logit link function g(·);thus μ=g−1(Xβ)=eXβ1+eXβ

To finish specifying the Bayesian model, we put a normal prior on the β coefficients and an inverse gamma prior on the precision parameter ϕ. In particular, we had

βj~N(0,5)ϕ~IG(3,20)

We then used Metropolis-Hastings algorithm to sample from the joint posterior (Gelman, et al., 2013). Both trace plots and the stationarity test based on Heidelberger and Welch (1983) indicated convergence of the simulation.

To compare the models fit with either the dichotomous or continuous measure of T. gondii seropositivity as well as the differing link functions, we used the Deviance Information Criterion (DIC) metric, which is specifically useful for hierarchical Bayesian models (Spiegelhalter, et al., 2002). We first calculated the deviance of each posterior sample,

D(μ,ϕ)=−2log(Γ(ϕ)Γ(μϕ)Γ((1−μ)ϕ)yμϕ−1(1−y)(1−μ)ϕ−1)

The average of all the deviance scores is denoted D¯. Then the overall DIC is

DIC=2D¯−D(μ¯,ϕ¯)

where μ¯ and ϕ¯ are the mean of the posterior samples.

3. Results

Of the 114 subjects in this study, 38 (33.3%) were T. gondii seropositive. The average age of the seropositive group was 81.1 years, compared to an average of 79.0 years in the seronegative group, and all subjects included in the analysis were white. The two groups did not differ significantly by age (p = 0.1), sex (p = 0.5), education (p = 0.4), socioeconomic status (p = 0.3), or percent with at least one ApoE E4 allele (p = 1.0). As expected, the anti-T. gondii IgG concentration was significantly higher in the T. gondii seropositive group compared to the seronegative group (p < 0.001) (Table 1). Only one subject was seropositive for anti-T. gondii IgM-class antibodies, whom we removed from the analysis. The prevalence of hypertension, diabetes, heart disease, thyroid disease, alcohol abuse, tobacco use, stroke, and hypercholesterolemia did not significantly differ between the seropositive and seronegative groups.

WMS Logical Memory IA Immediate, WMS Logical Memory IIA Delayed, WMS Associate Learning, and the Total score from the SRT, uncorrected for any control variables, did not differ in the overall sample between subjects T. gondii seropositive and seronegative (Table 1). Additionally, scores uncorrected for any control variables on measures of memory did not differ significantly between males and females (Table 2).

3.1 Linear Regression Results

In the linear-regression analyses, T. gondii seropositivity was not associated with any of the memory measures in the overall sample in the model including control variables. However, statistical power ranged from 45 to 87 percent (Table 3) suggesting that the probability of false negative results of the tests was high.

In the seropositive-only interaction analyses, anti-T. gondii IgG titer was not associated with any of the memory measures for seropositive samples, nor was the titer-sex interaction (Table 4). Again, the power of the tests was very low (22% to 78%).

However, there was a significant interaction between anti-T. gondii antibody titer and sex such that the slope of anti-T-gondii antibody titer for males was more positive than the slope for females (i.e., on Logical Memory IA and Logical Memory IIA in the overall sample (Table 5). However, presence or absence of T. gondii seropositivity did not predict any memory test scores in the overall sample. There was low power for the SRT model. However, the power in the other models ranged from 55.1 to 84.8 percent.

3.2 Bayesian analysis results

The analyses presented in Table 5 were also estimated as Bayesian models. The DIC scores for the four outcomes by anti-T. gondii antibody titer and T. gondii seropositivity are presented in Supplemental Table 1. For the Logical Memory tests (both immediate (IA) and delayed recall (IIA)) and for the SRT, the model including anti-T. gondii antibody titer has a slightly better fit. On the other hand, the Associate Learning test had a better model fit for T. gondii seropositivity (Supplemental Table 1).

Parameters from the Bayesian analysis and their corresponding posterior density 95% credible interval are listed in the Supplemental Table 2. In particular, only the SRT test had an interaction between sex and anti-T. gondii antibody titer. For the SRT, higher titers in males was associated with higher scores with an odds ratio of 1.22, meaning there is a 22% increase in one unit of the scores for every one unit increase of titers of toxoplasmosis for males. Females, on the other hand, only had a 2% increase. That is, males had more of an improvement with higher titers. Note, this was when the entire sample (seropositive and seronegative) was included.

For the seropositive-only sample in the Bayesian analysis, the antibody titer for the SRT task had an odds ratio of .38, meaning there was a 62% decrease in one unit of the memory score for every one unit increase in titers of toxoplasmosis of those participants that were seropositive (Supplemental Table 3). However, there was no significant titer-sex interaction in the Bayesian seropositive only sample.

4. Discussion

In this well characterized sample of 114 older adults without dementia, there were few associations between memory function and T. gondii seropositivity in both linear regression and Bayesian models. Both seropositive and seronegative groups were well matched on sociodemographic and medical variables. In the linear regression models without interaction terms, there were no associations between either T. gondii seropositivity or anti-T. gondii antibody titer levels and any of the memory tasks. In the entire sample, there was a significant interaction, however, between anti-T. gondii antibody titer and sex, indicating either that males with higher anti-T. gondii IgG antibody titers had better performance than did males with lower titers or that males with lower titers had worse performance on the Logical Memory IA and IIA tasks. We did not find this association in the model looking at presence or absence of toxoplasmosis (i.e. seropositivity), nor was there an interaction between antibody titer (i.e. serointensity) and sex in the linear-regression models in the seropositive-sample only. Further, the results of the Bayesian models showed no main effects for T. gondii seropositivity or antibody titer on any of the memory tasks, although there was an interaction between sex and T. gondii antibody titer on the SRT task, with higher titers of anti-T. gondii antibodies in males associated with higher SRT scores. In the Bayesian model for the seropositive-only sample, higher antibody titer was associated with lower scores on the SRT task, suggesting that higher serointensity may be associated with worse performance on the SRT. However, we did not find any association in the Bayesian analyses in the seropositive-sample only and the other three memory tasks.

Although we had expected that T. gondii seropositivity would be negatively associated with memory function, our finding of limited support suggesting that T. gondii seropositivity is associated with better memory function on some of the memory tasks in males is consistent with a study of young adults where performance on a task of executive function modulated by dopamine was positively associated with T. gondii seropositivity (Stock, et al., 2014). In this regard, the SRT we used in our study has a strong executive component not present in the other three memory tests we used. However, we found different results in the regression models, where in males only increased anti-T. gondii antibody titers were associated with higher performance on the immediate (Logical Memory IA) and delayed recall (Logical Memory IIA) portions of the Logical Memory subtest of the Wechsler Memory Scale rather than on the SRT. These findings may have different interpretations, depending on whether higher anti-T. gondii antibody titer represents greater serointensity or whether lower titers of anti-T. gondii in seropositive subjects indicates a greater interval since the initial infection. If higher titers indicate higher serointensity, our findings may suggest limited evidence (i.e. on two of four memory measures) of an association between T. gondii and better memory function in males. If lower titers that are still in the range considered positive for the presence of latent toxoplasmosis reflect longer time since infection, there would be limited evidence of worse memory function in low anti T. gondii IgG males, that is, in males infected long before the onset of the present study. However, in the Bayesian analysis in the seropositive-only sample containing both females and males, there was an inverse relationship between antibody titer and performance on the SRT with higher titers associated with worse performance.

In addition to IgG seropositivity (i.e. presence of latent infection), some studies suggest higher anti-T. gondii titers are often associated with worse cognitive performance (Mendy, et al., 2015). Similarly, behavioral or psychiatric symptoms associated with chronic T. gondii infection are typically correlated with higher rather than lower titers (Groer, et al., 2011, Okusaga, et al., 2011, Pedersen, et al., 2012). Moreover, serointensity but not seropositivity differentiated between patients with schizophrenia and healthy controls (Emelia, et al., 2012) and also between patients with depression with or without a history of suicide attempts (Arling, et al., 2009). However, even though many studies suggest seropositivity for latent toxoplasmosis is associated with worse outcome, whether it be cognitive or psychiatric, other studies suggest that seropositivity may not be purely detrimental. For example, Flegr et al. (2012) analyzed males and females separately and found that while seropositive females did more poorly on one of several cognitive measures compared to seronegative females, seropositive males did equally well on most cognitive measures and actually did better than seronegative males on two of the cognitive measures. Similarly, Stock et al. (2014) found that seropositivity for T. gondii was associated with better performance on a measure of action control. Finally, there is another possibility to be considered. Havlicek et al. (2001) found that seropositive subjects with lower rather than higher titers were the ones that performed differently from seronegative controls and they suggested that lower titers in seropositive subjects reflect longer duration of latent T. gondii infection, an interpretation broadly consistent with a study by Konishi (1989), who demonstrated that longitudinally those with initially higher IgG antibody levels tended to demonstrate decreases in IgG antibody levels over time. Still, many of the subjects demonstrated no significant change over a few years, and some subjects with initially lower titer levels showed increased titers with time. Further, an anamnestic decrease in anti-T. gondii IgG titers could have led to false negatives in our study of older adults resulting in misclassification of T. gondii infection status, which could be a reason for our finding few if any differences in memory function associated with T. gondii seropositivity in older adults. Older subjects infected well before the study could have normal anti-T. gondii due to the anamnestic antibody response (Kodym P, et al., 2001). However, since our present data are cross-sectional, we cannot directly address chronicity of infection, and we cannot assume that seropositive subjects with lower titers are those with the longest duration of latent toxoplasma infection. Furthermore, we have no way to determine whether any of the subjects had either been exposed to the infection more than once or whether an existing infection had been reactivated at some point. Recent animal studies have demonstrated that situational factors such as stressors can reactivate a chronic T. gondii infection with conversion of T. gondii bradyzoites to tachyzoites (Shen, et al., 2016). In animals, an increase in tachyzoites in chronic T. gondii infection is associated with increases in IgG and IgA but not IgM antibodies (Singh, et al., 2010).

Possible reasons for the differences between our findings and those of Mendy et al. (2015) and Gajewski et al. (2014), who found a negative impact of toxoplasmosis on memory function in older adults, could be due to age differences between studies in that our sample had a mean age of 81.1 and 79.0 years for the seropositive and seronegative groups, respectively, compared to the median age in the study by Mendy et al. (2015) of 69 years and in the Gajewski et al. (2014) study a mean age of 70.8 and 69.8 years in the seropositive and seronegative groups, respectively. Because our sample was on average approximately ten years older, there could have been a higher death rate in the seropositive group compared to the seronegative group resulting in a loss of those subjects with potential memory deficits associated with T. gondii seropositivity. In this regard, we note that T. gondii seropositivity has been negatively associated with health status, particularly related to immune and digestive disorders (Flegr and Escudero, 2016, Flegr, et al., 2014) and has been associated with morbidity and mortality (Lykins, et al., 2016). Also, because we excluded participants with a diagnosis of dementia, the key feature of which is often memory impairment, we may have had a sample biased towards average memory performance. In addition, both Mendy et al. (2015) and Gajewski et al. (2014) used different tests to evaluate memory, which may have evaluated different aspects of memory function than did the tests of memory function we used. Still, the tests used in our study are well known and often used clinically for diagnostic purposes in the evaluation of age-related memory decline. Finally, the sample we used had relatively high levels of education, 15.3 years and 14.8 years of education in the seropositive and seronegative groups, respectively, which may have protected against any negative effects of T. gondii seropositivity on memory (Gale, et al., 2015).

How T. gondii seropositivity might account for both improved and worsened memory function in older adults is unclear. A possible mechanism may relate to the increased production of dopamine associated with T. gondii seropositivity, as dopamine has been associated with better working and verbal memory function (Henry and Sherwin, 2012). Specifically, Prandovszky et al. (2011) showed higher levels of dopamine in cyst-containing neural cells in mice infected with Toxoplasmosis, and McConkey et al. (2013) proposed that the parasite’s ability to increase dopamine also affects catecholaminergic neurons and the behaviors associated with them. Another study analyzing performance on a cognitive task modulated by the dopamine system found that T. gondii seropositivity in a group of 25-year olds was associated with better function (Stock, et al., 2014).

Although the subjects in this study were well characterized and were assessed with standard measures of memory function, several limitations in addition to those already discussed require consideration in interpreting these findings. Low statistical power may have led to false negatives in the analyses. In addition, the study design was cross-sectional, and we did not have information about the time of initial infection with T. gondii. Similarly, we did not have information about the length of time subjects had been T. gondii seropositive.

5. Conclusion

In conclusion, in this sample of 114 older female and male subjects, taking into account the limitations associated with this study including low statistical power, there was little evidence of an association between T. gondii seropositivity and anti-T. gondii IgG antibody titer and memory function. However, there was some evidence of an interaction between T. gondii and memory performance in males and also evidence that higher anti-T. gondii titers in the seropositive-only sample was associated with worse performance on the SRT.

Supplementary Material

supplement

Highlights.

  • Latent Toxoplasmosis has been associated with decreased memory in some studies.

  • In this sample, there were few associations between memory and latent toxoplasmosis.

  • In seropositive participants, higher anti-T. gondii titers were inversely associated with memory.

Acknowledgments

We wish to acknowledge and thank the ADRC (P50AG05681) for providing serum samples and cognitive data and Dr. John Kauwe’s assistance in this regard. We thank the Gerontology Program at Brigham Young University for providing a research grant award for processing the assays.

6. Appendix

Table 1.

Characteristics of the toxoplasmosis positive and negative groups.

Characteristics Positive Negative Total t [CI], p-value**
Sample Size 38 76 114
Age (years), mean, SD, [range] 81.1, 6.5, [70.4–94.3] 79.0, 7.6, [62.1–98.3] 79.7, 7.3, [62.1–98.3] −1.5 [−5.0, 0.7], 0.1
Women (%) 45 51 49 −0.7 [−0.1, 0.3], 0.5
Education (years), mean, SD 15.3, 2.9 14.8, 3.0 15.0, 3.0 −0.9 [−1.7, 0.6], 0.4
SES (Hollingshead Index), mean, SD 3.9, 1.0 3.8, 1.0 3.8, 1.0 −1.0 [−0.6, 0.2], 0.3
Prevalence of E4 allele (ApoE) (%) 26 27 27 0.0 [−0.2, 0.2], 1.0
Anti-T. gondii IgG Ab concentration IU/mL, mean, SD, [range] 293.5, 244.7, [35.2–940.8] 13.2, 7.0, [0.3–28.9] 106.6, 193.0, [0.3–940.8] −10.0 [−335.7, −224.8], <0.001
MMSE total score, mean, SD 28.5, 1.7 28.7, 1.4 28.7, 1.5 0.8 [−0.3, 0.8], 0.4
WMS Logical Memory IA Immediate, mean, SD 14.4, 4.0 13.9, 4.6 14.1, 4.4 −0.4 [−2.5, 1.6], 0.7
WMS Logical Memory IIA, mean, SD 13.5, 4.5 12.9, 5.0 13.1, 4.8 −0.5 [−2.8, 1.7], 0.6
WMS Associate Learning, mean, SD 14.6, 3.0 14.4, 3.9 14.5, 3.6 −0.3 [−1.6, 1.2], 0.8
Free and Cued Selective Reminding Test
 Total score, mean, SD 47.8, 0.5 47.7, 0.8 47.7, 0.7 −0.3 [−0.4, 0.3], 0.7
Open in a new tab
*

No significant differences between groups for hypertension, diabetes, heart disease, thyroid disease, alcohol consumption, tobacco use, stroke, hypercholesterolemia.

**

T-tests were performed between the groups which were seropositive or seronegative for T. gondii.

Table 2.

Memory test scores (M, SD) for males and females.

Memory measure Males Females Total t [CI], p-value**
WMS Logical Memory IA, Immediate, mean, SD 13.45, 0.7 14.7, 0.6 14.1, 1.0 −1.3 [−3.2, 0.7], 0.2
WMS Logical Memory IIA, mean, SD 12.3, 0.8 13.9, 0.7 13.1, 0.5 −1.5 [−3.7, 0.6], 0.1
WMS Associate Learning, mean, SD 13.8, 0.5 15.2, 0.5 14.5, 0.7 0.1 [−2.7, 0.0], 0.1
Free and Cued Selective Reminding Test
 Total score, mean, SD 47.7, 0.1 47.7, 0.1 47.7, 1.0 0.0 [−0.3, 0.3], 1.0
Open in a new tab
**

T-tests were performed between the male and female groups

Table 3.

Main effects models of anti-T. gondii antibody titer or presence of T. gondii seropositivity and each memory test score for the whole group: Unstandardized coefficients (standard errors) from linear regression.

Logical Memory IA Logical Memory IIA Associate Memory Selective Reminding Test
b SE b SE b SE b SE
Antibody titer   .030 (.02)   .033 (.02)   .000 (.02) −.005 (.02)
Male −.088* (.04) −.100* (.05) −.064 (.04) −.015 (.04)
SES (Hollingshead Index)
 High   .070 (.05)   .066 (.05)   .091* (.04)   .093* (.04)
Education
 More than High School   .580 *** (.03)   .542*** (.04)   .687*** (.03)   .951*** (.03)
Age −.002 (.02)   .005 (.02) −.005 (.02)   .009 (.02)
ApoE (presence of E4)   .032 (.05)   .049 (.05)   .029 (.05)   .092* (.04)
R2 .13 .16 .16 .08
N 81 81 107 91
Power .627 .735 .871 .453
Presence of Latent Toxoplasmosis   .021 (.04)   .027 (.05)   .014 (.04)   .010 (.04)
Male −.082 (.04) −.094* (.05) −.065 (.04) −.016 (.04)
SES (Hollingshead Index)
 High   .070 (.05)   .067 (.05)   .091* (.04)   .093* (.04)
Education
 More than High School   .574*** (.04)   .534*** (.04)   .682*** (.04)   .947*** (.03)
Age   .000 (.00)   .001 (.00) −.001 (.00)   .001 (.00)
ApoE (presence of E4)   .019 (.05)   .034 (.05)   .029 (.05)   .094* (.04)
R2 .11 .13 .16 .08
N 81 81 107 91
Power .543 .627 .871 .453
Open in a new tab

Abbreviations: SE = Standard error.

*

p< .05,

**

p < .01,

***

p < .001.

Table 4.

Interaction effects (titer-sex interaction) models of anti-T gondii antibody titer and each memory test score for seropositive samples only: Unstandardized coefficients (standard errors) from linear regression.

Logical Memory IA Logical Memory IIA Associate Memory Selective Reminding Test
b SE b SE b SE b SE
Antibody titer −.083 (.12)   .076 (.10) −.040 (.06) −.111 (.10)
Male −.045 (.07) −.036 (.07) −.061 (.04) −.068 (.05)
SES (Hollingshead Index)
 High   .118 (.08)   .134 (.09)   .201*** (.04) 195*** (.05)
Education
 More than High School   .204** (.02)   .188*** (.15)   .252*** (.01)   .353*** (.01)
Age −.004 (.01) −.006 (.05) −.008* (.00)   .001 (.00)
ApoE (presence of E4)   .058 (.09) −.058 (.10) −.102 (.05) −.037 (.07)
Titer−sex interaction   .126 (.14)   .154 (.15)   .012 (.08) −0.42 (.11)
R2 .19 .17 .47 .32
N 28 28 37 31
Power .245 .221 .781 .470
Open in a new tab

Abbreviations: SE = Standard error.

*

p < .05,

**

p < .01,

***

p < .001.

Table 5.

Interaction models (titer-sex, Toxoplasmosis) of the relationship between anti-T. gondii antibody titer or presence of T. gondii seropositivity and each memory test score for whole group: Unstandardized coefficients (standard errors) from linear regression.

Logical Memory IA Logical Memory IIA Associate Memory Selective Reminding Test
b SE b SE b SE b SE
Antibody titer −.010 (.03) −.012 (.03) −.020 (.03) −.030 (.03)
Male −.091* (.04) −.103* (.04) −.066 (.04) −.017 (.04)
SES (Hollingshead Index)
 High   .083 (.04)   .081 (.05)   .095* (.04)   .102* (.04)
Education
 More than High School   .573*** (.03)   .534*** (.03)   .684*** (.03) 945*** (.03)
Age   .000 (.02)   .008 (.02) −.004 (.02)   .008 (.02)
ApoE (presence of E4)   .044 (.05)   .062 (.05)   .032 (.05)   .099* (.04)
Titer-sex interaction   .081* (.04)   .094* (.04)   .040 (.04)   .053 (.04)
R2 .15 .18 .16 .08
N 81 81 107 91
Power .667 .762 .848 .422
Presence of Latent Toxoplasmosis −.043 (.06) −.052 (.07) −.014 (.06) −.021 (.06)
Male −.130* (.05) −.153* (.06) −.085 (.05) −.037 (.05)
SES (Hollingshead Index)
 High   .083 (.05)   .083 (.05)   .096* (.05)   .098* (.04)
Education
 More than High School   .593*** (.04)   .557*** (.04)   .689*** (.04)   .955*** (.03)
Age   .000 (.00)   .001 (.00) −.001 (.00)   .001 (.00)
ApoE (presence of E4)   .027 (.05)   .044 (.05)   .031 (.05)   .096* (.04)
Toxoplasmosis-sex interaction   .128 (.09)   .157 (.10)   .054 (.09)   .059 (.08)
R2 .1 .15 .16 .08
N 81 81 107 91
Power .551 .667 .848 .422
Open in a new tab

Abbreviations: SE = Standard error.

*

p < .05,

**

p < .01,

***

p < .001.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Suggested Reviewers:

We suggest the following individuals as potential reviewers of this manuscript:

Patrick Gajewski, Leibniz Research Centre for Working Environment and Human Factors, Ardeystr. 67, 44139 Dortmund, Germany. E-mail: gajewski@ifado.de

Angelico Mendy, Department of Epidemiology, College of Public Health, University of Iowa, S161 CPHB 105 River Street, Iowa City, IA, 52242, USA. E-mail: angelico-mendy@uiowa.edu

Ann-Kathrin Stock, Cognitive Neurophysiology, Department of Child and Adolescent Psychiatry, University of Dresden, Schubertstrasse 42, D-01307 Dresden, Germany. E-mail: ann-kathrin.stock@uniklinikum-ddresden.de

References

  1. Arling TA, Yolken RH, Lapidus M, Langenberg P, Dickerson FB, Zimmerman SA, Balis T, Cabassa JA, Scrandis DA, Tonelli LH, Postolache TT. Toxoplasma gondii antibody titers and history of suicide attempts in patients with recurrent mood disorders. J Nerv Ment Dis. 2009;197(12):905–8. doi: 10.1097/NMD.0b013e3181c29a23. [DOI] [PubMed] [Google Scholar]
  2. Barnes LL, Capuano AW, Aiello AE, Turner AD, Yolken RH, Torrey EF, Bennett DA. Cytomegalovirus Infection and Risk of Alzheimer Disease in Older Black and White Individuals. J Infect Dis. 2014 doi: 10.1093/infdis/jiu437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Dorey E, Chang N, Liu QY, Yang Z, Zhang W. Apolipoprotein E, amyloid-beta, and neuroinflammation in Alzheimer’s disease. Neurosci Bull. 2014;30(2):317–30. doi: 10.1007/s12264-013-1422-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Emelia O, Amal RN, Ruzanna ZZ, Shahida H, Azzubair Z, Tan KS, Noor Aadila S, Siti NA, Aisah MY. Seroprevalence of anti-Toxoplasma gondii IgG antibody in patients with schizophrenia. Trop Biomed. 2012;29(1):151–9. [PubMed] [Google Scholar]
  5. Fekadu A, Shibre T, Cleare AJ. Toxoplasmosis as a cause for behaviour disorders–overview of evidence and mechanisms. Folia Parasitol (Praha) 2010;57(2):105–13. doi: 10.14411/fp.2010.013. [DOI] [PubMed] [Google Scholar]
  6. Flegr J. Neurological and neuropsychiatric consequences of chronic toxoplasma infection. Current Clinical Microbiology Reports. 2015;2:163–72. doi: 10.1007/s40588-015-0024-0. [DOI] [Google Scholar]
  7. Flegr J, Escudero DQ. Impaired health status and increased incidence of diseases in Toxoplasma-seropositive subjects - an explorative cross-sectional study. Parasitology. 2016:1–16. doi: 10.1017/S0031182016001785. [DOI] [PubMed] [Google Scholar]
  8. Flegr J, Guenter W, Bielinski M, Deptula A, Zalas-Wiecek P, Piskunowicz M, Szwed K, Bucinski A, Gospodarek E, Borkowska A. Toxoplasma gondii infection affects cognitive function –corrigendum. Folia Parasitol (Praha) 2012;59(4):253–4. doi: 10.14411/fp.2012.014. [DOI] [PubMed] [Google Scholar]
  9. Flegr J, Prandota J, Sovickova M, Israili ZH. Toxoplasmosis–a global threat. Correlation of latent toxoplasmosis with specific disease burden in a set of 88 countries. PLoS One. 2014;9(3):e90203. doi: 10.1371/journal.pone.0090203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gajewski PD, Falkenstein M, Hengstler JG, Golka K. Toxoplasma gondii impairs memory in infected seniors. Brain Behav Immun. 2014;36:193–9. doi: 10.1016/j.bbi.2013.11.019. [DOI] [PubMed] [Google Scholar]
  11. Gale SD, Brown BL, Erickson LD, Berrett A, Hedges DW. Association between latent toxoplasmosis and cognition in adults: a cross-sectional study. Parasitology. 2015;142(4):557–65. doi: 10.1017/S0031182014001577. [DOI] [PubMed] [Google Scholar]
  12. Gelman A, Carlin JB, Stern HS, Dunson DB, Vehtari A, Rubin DB. Bayesian Data Analysis. 3rd. Chapman and Hall; 2013. [Google Scholar]
  13. Gifford KA, Badaracco M, Liu DD, Tripodis Y, Gentile A, Lu ZQ, Palmisano J, Jefferson AL. Blood Pressure and Cognition Among Older Adults: A Meta-Analysis. Arch Clin Neuropsychol. 2013;28(7):649–64. doi: 10.1093/arclin/act046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Grober E, Buschke H, Crystal H, Bang S, Dresner R. Screening for dementia by memory testing. Neurology. 1988;38(6):900–3. doi: 10.1212/wnl.38.6.900. [DOI] [PubMed] [Google Scholar]
  15. Groer MW, Yolken RH, Xiao JC, Beckstead JW, Fuchs D, Mohapatra SS, Seyfang A, Postolache TT. Prenatal depression and anxiety in Toxoplasma gondii-positive women. Am J Obstet Gynecol. 2011;204(5):433e1–7. doi: 10.1016/j.ajog.2011.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. HavlÍCek J, GaŠOvÁ Z, Smith AP, ZvÁRa K, Flegr J. Decrease of psychomotor performance in subjects with latent ‘asymptomatic’ toxoplasmosis. Parasitology. 2001;122(05) doi: 10.1017/s0031182001007624. [DOI] [PubMed] [Google Scholar]
  17. Heidelberger P, Welch PD. Simulation Run Length Control in the Presence of an Initial Transient. Oper Res. 1983;31(6):1109–44. doi: 10.1287/opre.31.6.1109. [DOI] [Google Scholar]
  18. Henry JF, Sherwin BB. Hormones and cognitive functioning during late pregnancy and postpartum: a longitudinal study. Behav Neurosci. 2012;126(1):73–85. doi: 10.1037/a0025540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hollingshead AB, Redlich FC. Social class and mental illness. John Wiley & Sons; New York, NY: 1958. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jones JL, Kruszon-Moran D, Rivera HN, Price C, Wilkins PP. Toxoplasma gondii seroprevalence in the United States 2009–2010 and comparison with the past two decades. Am J Trop Med Hyg. 2014a;90(6):1135–9. doi: 10.4269/ajtmh.14-0013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jones JL, Parise ME, Fiore AE. Neglected parasitic infections in the United States: toxoplasmosis. Am J Trop Med Hyg. 2014b;90(5):794–9. doi: 10.4269/ajtmh.13-0722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Katan M, Moon YP, Paik MC, Sacco RL, Wright CB, Elkind MS. Infectious burden and cognitive function: the Northern Manhattan Study. Neurology. 2013;80(13):1209–15. doi: 10.1212/WNL.0b013e3182896e79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kodym P, Malý M, Švandová M, Ležatková H, Bažoutová M, Vlčková J, Beneš Č, Zástěra . Toxoplasma in the Czech Republic 1923–1999: First case to widespread outbreak. In: Petersen E, Pollak A, Reiter-Owona I, editors. Int J Parasitol. 2001. pp. 125–32. [Google Scholar]
  24. Konishi E. Annual Change in Immunoglobulin-G and M-Antibody Levels to Toxoplasma-Gondii in Human-Sera. Microbiol Immunol. 1989;33(5):403–11. doi: 10.1111/j.1348-0421.1989.tb01988.x. [DOI] [PubMed] [Google Scholar]
  25. Kusbeci OY, Miman O, Yaman M, Aktepe OC, Yazar S. Could Toxoplasma gondii have any role in Alzheimer disease? Alzheimer Dis Assoc Disord. 2011;25(1):1–3. doi: 10.1097/WAD.0b013e3181f73bc2. [DOI] [PubMed] [Google Scholar]
  26. Levy JA, Bergeson J, Putnam K, Rosen V, Cohen R, Lalonde F, Mirza N, Linker G, Sunderland T. Context-specific memory and apolipoprotein E (ApoE) epsilon 4: cognitive evidence from the NIMH prospective study of risk for Alzheimer’s disease. J Int Neuropsychol Soc. 2004;10(3):362–70. doi: 10.1017/S1355617704103044. [DOI] [PubMed] [Google Scholar]
  27. Lykins J, Wang K, Wheeler K, Clouser F, Dixon A, El Bissati K, Zhou Y, Lyttle C, Rzhetsky A, McLeod R. Understanding Toxoplasmosis in the United States Through “Large Data” Analyses. Clin Infect Dis. 2016;63(4):468–75. doi: 10.1093/cid/ciw356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. McConkey GA, Martin HL, Bristow GC, Webster JP. Toxoplasma gondii infection and behaviour - location, location, location? J Exp Biol. 2013;216(Pt 1):113–9. doi: 10.1242/jeb.074153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR, Jr, Kawas CH, Klunk WE, Koroshetz WJ, Manly JJ, Mayeux R, Mohs RC, Morris JC, Rossor MN, Scheltens P, Carrillo MC, Thies B, Weintraub S, Phelps CH. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 2011;7(3):263–9. doi: 10.1016/j.jalz.2011.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mendy A, Vieira ER, Albatineh AN, Gasana J. Immediate rather than delayed memory impairment in older adults with latent toxoplasmosis. Brain Behav Immun. 2015;45:36–40. doi: 10.1016/j.bbi.2014.12.006. [DOI] [PubMed] [Google Scholar]
  31. Montoya JG, Liesenfeld O. Toxoplasmosis. Lancet. 2004;363(9425):1965–76. doi: 10.1016/S0140-6736(04)16412-X. [DOI] [PubMed] [Google Scholar]
  32. Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology. 1993;43(11):2412–4. doi: 10.1212/wnl.43.11.2412-a. [DOI] [PubMed] [Google Scholar]
  33. Okusaga O, Langenberg P, Sleemi A, Vaswani D, Giegling I, Hartmann AM, Konte B, Friedl M, Groer MW, Yolken RH, Rujescu D, Postolache TT. Toxoplasma gondii antibody titers and history of suicide attempts in patients with schizophrenia. Schizophr Res. 2011;133(1–3):150–5. doi: 10.1016/j.schres.2011.08.006. [DOI] [PubMed] [Google Scholar]
  34. Pedersen MG, Mortensen PB, Norgaard-Pedersen B, Postolache TT. Toxoplasma gondii infection and self-directed violence in mothers. Arch Gen Psychiatry. 2012;69(11):1123–30. doi: 10.1001/archgenpsychiatry.2012.668. [DOI] [PubMed] [Google Scholar]
  35. Perry CE, Gale SD, Erickson L, Wilson E, Nielsen B, Kauwe J, Hedges DW. Seroprevalence and Serointensity of Latent Toxoplasma gondii in a Sample of Elderly Adults With and Without Alzheimer Disease. Alzheimer Dis Assoc Disord. 2016;30(2):123–6. doi: 10.1097/WAD.0000000000000108. [DOI] [PubMed] [Google Scholar]
  36. Prandovszky E, Gaskell E, Martin H, Dubey JP, Webster JP, McConkey GA. The neurotropic parasite Toxoplasma gondii increases dopamine metabolism. PLoS One. 2011;6(9):e23866. doi: 10.1371/journal.pone.0023866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rencher AC, Schaalje GB. Linear models in statistics. 2nd. Wiley-Interscience; Hoboken, N.J.: 2008. [Google Scholar]
  38. Shen B, Yuan Y, Cheng J, Pan M, Xia N, Zhang W, Wang Y, Zhou Y, Zhao J. Activation of chronic toxoplasmosis by transportation stress in a mouse model. Oncotarget. 2016 doi: 10.18632/oncotarget.13568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Singh J, Graniello C, Ni Y, Payne L, Sa Q, Hester J, Shelton BJ, Suzuki Y. Toxoplasma IgG and IgA, but not IgM, antibody titers increase in sera of immunocompetent mice in association with proliferation of tachyzoites in the brain during the chronic stage of infection. Microbes and infection / Institut Pasteur. 2010;12(14–15):1252–7. doi: 10.1016/j.micinf.2010.07.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Smithson M, Verkuilen J. A better lemon squeezer? Maximum-likelihood regression with beta-distributed dependent variables. Psychol Methods. 2006;11(1):54–71. doi: 10.1037/1082-989X.11.1.54. [DOI] [PubMed] [Google Scholar]
  41. Spiegelhalter DJ, Best NG, Carlin BR, van der Linde A. Bayesian measures of model complexity and fit. J Roy Stat Soc B. 2002;64:583–616. doi: 10.1111/1467-9868.00353. [DOI] [Google Scholar]
  42. Stewart R, Liolitsa D. Type 2 diabetes mellitus, cognitive impairment and dementia. Diabet Med. 1999;16(2):93–112. doi: 10.1046/j.1464-5491.1999.00027.x. [DOI] [PubMed] [Google Scholar]
  43. Stock AK, Heintschel von Heinegg E, Kohling HL, Beste C. Latent Toxoplasma gondii infection leads to improved action control. Brain Behav Immun. 2014;37:103–8. doi: 10.1016/j.bbi.2013.11.004. [DOI] [PubMed] [Google Scholar]
  44. Watson AM, Prasad KM, Klei L, Wood JA, Yolken RH, Gur RC, Bradford LD, Calkins ME, Richard J, Edwards N, Savage RM, Allen TB, Kwentus J, McEvoy JP, Santos AB, Wiener HW, Go RC, Perry RT, Nasrallah HA, Gur RE, Devlin B, Nimgaonkar VL. Persistent infection with neurotropic herpes viruses and cognitive impairment. Psychol Med. 2013;43(5):1023–31. doi: 10.1017/S003329171200195X. [DOI] [PubMed] [Google Scholar]
  45. Wechsler D. Manual: Wechsler Memory Scale. Psychological Corporation; New York: 1973. [Google Scholar]
  46. Wechsler D. Manual: Wechsler Memory Scale-Revised. Psychological Corporation; San Antonio, Texas: 1987. [Google Scholar]
  47. Xu W, Tan L, Wang HF, Jiang T, Tan MS, Tan L, Zhao QF, Li JQ, Wang J, Yu JT. Meta-analysis of modifiable risk factors for Alzheimer’s disease. J Neurol Neurosurg Psychiatry. 2015;86(12):1299–306. doi: 10.1136/jnnp-2015-310548. [DOI] [PubMed] [Google Scholar]
  48. Zhang M, Gale SD, Erickson LD, Brown BL, Woody P, Hedges DW. Cognitive function in older adults according to current socioeconomic status. Aging, Neuropsychology, and Cognition. 2015:1–10. doi: 10.1080/13825585.2014.997663. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplement
NIHMS849285-supplement.docx (20.2KB, docx)

RESOURCES