Association of Seropositivity to Borrelia burgdorferi With the Risk of Neuropsychiatric Disorders and Functional Decline in Older Adults: The Aging Multidisciplinary Investigation Study

Key Points

Question  Is there an association between seropositivity to Borrelia burgdorferi and incidental neuropsychiatric disorders and functional decline in older adults?

Findings  In this cohort study that included 689 participants, no significant association between seropositivity for Borrelia burgdorferi and any of the studied outcomes (ie, cognitive decline, incident dementia, depressive symptoms, and functional decline) was found.

Meaning  This study suggests that exposure to Borrelia burgdorferi is not a risk factor for certain incidental neuropsychiatric disorders or functional decline.

Importance  Exposure to Borrelia burgdorferi (Bb) has been reported to be associated with certain neuropsychiatric disorders.

Objective  To establish the association between seropositivity to Bb and incidental neuropsychiatric disorders (eg, cognitive decline, incident dementia, and depressive symptoms) as well as functional decline.

Design, Setting, and Participants  This prospective, 6-year follow-up cohort study was conducted in a rural southwestern region of France and included 689 retired farmers 65 years or older randomly recruited from the Farmer Health Insurance System who agreed to submit a blood sample and were participants in the Aging Multidisciplinary Investigation study, an ongoing epidemiological prospective study of aging initiated in 2007. The data were analyzed from April to May 2019.

Exposures  Borrelia burgdorferi serology testing was performed in a 2-tiered approach. During the follow-up period, cognitive decline, incident dementia, depressive symptoms, and functional decline were repeatedly assessed.

Main Outcomes and Measures  Diagnosis of dementia relied on a 3-step procedure; cognitive decline was determined using the Mini-Mental State Examination and depressive symptomatology was assessed using the Center for Epidemiologic Studies Depression scale. For disability, scores on instrumental and basic activities of daily living were investigated.

Results  Of 689 participants, 432 (62.2%) were men and the mean (SD) age was 75.8 (6.4) years. The seroprevalence rate of Bb was 6.5%. At baseline, compared with Bb− participants, those who were Bb+ were older, predominantly men, and had lower depressive symptoms. No association between seropositivity and any of the studied outcomes (ie, cognitive decline, depressive symptoms, or functional decline) was found in the crude analysis and after adjusting for confounding variables. Regarding incident dementia, no increased risk was found among Bb+ participants (hazard ratio, 0.42; 95% CI, 0.1-1.17; adjusted for diverse confounders).

Conclusions and Relevance  To our knowledge, this is one of the few longitudinal studies exploring the risk of neuropsychiatric disorders and functional decline associated with exposure to Bb. Despite its limitations (eg, a lack of information if clinical manifestations of Lyme borreliosis existed, date of exposure, or treatment received), this study suggests that seropositivity to Bb is not a risk factor for incidental neuropsychiatric disorders and functional decline.

As a result of the demographic transition worldwide, certain neuropsychiatric disorders, such as depression, neurocognitive disorders, and disability, are undergoing an increase in their prevalence. Lyme borreliosis is currently the most common vector-borne disease in the northern hemisphere and its causative agent is a group of spirochetes from the Borrelia burgdorferi (Bb) group.^1-3 Epidemiological data suggest that in countries with a high frequency of tick-borne diseases, such as Lyme borreliosis, the seroprevalence (previous exposure) to Bb increases for adults with age.^4-6 However, a seropositive status does not necessarily represent disease.⁷

Lyme borreliosis may manifest as a systemic condition, with multiorganic involvement (including neurological manifestations), and has been described in 3 clinical stages: early-localized, early-disseminated, and late-disseminated disease.^2,7 The nervous system can be affected in early-disseminated and late-disseminated disease and some neuropsychiatric disturbances, such as cognitive decline, dementia, and depression, have been reported as potential manifestations of the disease.^8,9

However, until now, the studies investigating the association between Bb exposure and neuropsychiatric disorders were scarce and faced important methodological limitations, most of them being cross-sectional.¹⁰ In addition, to our knowledge, important outcomes for older populations, such as disability, have not been considered in previous studies.

This study was conducted within a prospective cohort study of older farmers from southwestern France. Our aim was to determine the relationship between previous exposure (seropositivity) to Bb and incidental neuropsychiatric outcomes such as cognitive decline, dementia (major neurocognitive disorder), depression, and disability in older adults.

The participants of the study were recruited from the Aging Multidisciplinary Investigation study cohort, an ongoing epidemiological prospective study of aging initiated in 2007. The methods have been previously published.¹¹ Briefly, at baseline, 1002 retired farmers 65 years or older who were living in a rural setting in southwestern France were randomly recruited from the Farmer Health Insurance System (Mutualité Sociale Agricole). Written informed consent was obtained from all participants, and the ethics committee of the University Hospital of Bordeaux approved this research according to the principles embodied in the Declaration of Helsinki. The 689 participants of the Aging Multidisciplinary Investigation cohort who agreed to provide a blood sample at inclusion (plasma, serum, red blood cells, and DNA) were included in this study.

Participants were interviewed at home in 2007, 2010, 2013, and 2016 by trained psychologists. Age, sex, marital status, educational level, income, and former occupational sector were collected along with a range of environmental, social, neuropsychological, functional, and medical information at each wave using standardized questionnaires. Medical administrative data were also provided by the French Farmers Health Insurance System.

At baseline (2007), blood samples were collected at participants’ homes (34 mL of blood), transported at 4°C, frozen, and stored at −80°C. Tubes were thawed in 2016 and aliquots were used to perform tests for the present study. Detection of anti-Bb IgG was performed by a commercial ELISA test (Enzygnost Borreliosis; Siemens). For any positive or doubtful result, a confirmation was made using a commercial Western blot assay (Euroline RN AT IgG; Euroimmun) by the French National Reference Center for Borrelia in Strasbourg. Individuals with positive and probable serology results were defined as previously exposed participants (Bb+) and those with negative and dubious serology results (laboratory criteria) as nonexposed participants (Bb−).

The diagnosis of dementia relied on a 3-step procedure. After the interview and neuropsychological evaluation, the psychologist completed a criteria checklist for dementia using the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III R) checklist. Individuals who met dementia criteria were reevaluated at home by a neurologist or geriatrician to confirm or exclude the diagnosis. Finally, the diagnosis of dementia and its etiology were reviewed by an independent committee of neurology experts according to current standards.¹¹

Cognitive decline was determined using the Mini-Mental State Examination (MMSE) as an evaluation for global cognitive performance and its change over time (0-30 points; lower scores indicate worst cognitive performance).¹² Depressive symptomatology was assessed using the Center for Epidemiologic Studies Depression (CES-D) scale, a 20-item self-reported scale (ranging from 0-60 points; a higher score indicates a higher presence of depressive symptomatology).¹³

For disability, scores on instrumental and basic activities of daily living were investigated. Instrumental activities of daily living were assessed using the Lawton–Brody scale, including the ability to perform telephone calls, shopping, transportation, handling medications, managing finances, doing the laundry, preparing meals, and housekeeping.¹⁴ Basic activities of daily living were assessed using the Katz index, including bathing, dressing, using a toilet, transferring from a bed to a chair, and eating.¹⁵ For each domain of disability, if participants indicated that they were unable to perform 1 or more activities without human help, they were considered as having a disability.

Sociodemographic variables included age, sex, educational level (highly educated [HE] or low education level [LE], defined by a threshold of 7 years of schooling), main occupational sector (agricultural, tertiary, or both), marital status (married, widowed, single, divorced, or separated), living condition (alone, in a couple, with a family member, or non–family cohabitation), and subjective health (excellent or good vs fair, bad, or very bad). Health-associated variables included long-term diseases (LTD; ie, 30 conditions with prolonged treatment and expensive therapy for which the French government offers patients an exemption from copayment),¹⁶ apolipoprotein E (ApoE) ε4 status, and cardiovascular diseases (examined from the second wave of the study data collection and onwards), including history of stroke, myocardial infarction, arrhythmia, hypertension, angina pectoris, or peripheral artery disease.

Bb+ and Bb− participants were compared using the Student t test and χ² tests. The associations between Bb serological status and cognitive decline, depressive symptoms, and functional decline over the 6-year follow-up period were investigated using latent process mixed models. This procedure allows studying non-Gaussian longitudinal outcomes and models the change of the latent process over time by considering the correlation between the repeated and multiple measurements in each individual. Splines link functions were used to describe the change over time of the outcomes. Models were adjusted for potentially confounding association of age, sex, educational level, number of LTDs, ApoE ε4, cardiovascular diseases, and baseline outcomes scores. In addition, the association between Bb serological status and incident dementia was also investigated using a Cox proportional hazard model. Participants with prevalent dementia at baseline were excluded from the analysis. The number of years since baseline was used as the basic time scale in the analysis. Multivariate analyses were adjusted for age, sex, educational level, number of LTDs, ApoE ε4, cardiovascular diseases, and baseline MMSE score.

For all analyses, a P value of .05 or less was considered statistically significant and the 95% confidence intervals were provided. Analyses were performed using SAS software (version 9.3; SAS Institute) and R software, version 3.2.3 (R Foundation; for the lcmm package, version 1.7.9).

Of the 689 participants tested for Bb, 34 (4.9%) had results confirmed by Western blot as seropositive, 11 (1.6%) as probable, and 10 (1.5%) as dubious. This resulted in a seroprevalence rate of 6.5%, considering positive and probable results as seropositivity.

Sociodemographic and health-associated variables and the comparative analysis between Bb+ and Bb− participants at baseline are presented in Table 1. Compared with Bb− participants, Bb+ were older (mean [SD], 77.8 [6.6] years vs 75.7 [6.4] years; P = .02), more likely to be men [(86.7% vs 60.5%; P < .010), had lower mean (SD) CES-D scores (2.6 [4.2] vs 4.6 [6.2]; P < .010), and had a higher number of LTDs (>4 LTDs, 11.5% vs 3.1%; P < .010). There were no statistically significant differences respecting the other variables between groups (Table 1).

Results from the latent process mixed models are presented in Table 2. No association between serological status (Bb+) and any of the studied outcomes of interest (ie, cognitive decline, depressive symptoms, and disability) was found in the crude analysis or after adjusting for potentially confounding variables.

Regarding incident dementia, after excluding the participants with prevalent dementia at baseline, no increased risk of incident dementia was found among Bb+ participants (nonadjusted hazard ratio, 0.65; 95% CI, 0.24-1.76; P = .39 and hazard ratio, 0.43; 95% CI, 0.10-1.82; P = .25 adjusting by age, sex, educational level, cardiovascular diseases, number of LTDs, ApoE ε4 status, and baseline MMSE score).

This study shows no association between seropositivity to Bb and incident neuropsychiatric disorders or functional decline among older adults living in a rural environment in the southwestern region of France. These results should inform available literature on the subject. They strengthen the hypothesis that seropositivity to Bb is not a risk factor for developing further neuropsychiatric disorders or disability. Such results are important because neuropsychiatric manifestations in Lyme disease or in people exposed to Bb are still an ongoing field of investigation, especially in the context of the condition termed chronic Lyme disease. Although chronic Lyme disease has not been consensually defined, it is often associated with symptomatology such as chronic fatigue, myalgia, arthralgia, headache, affective or cognitive symptoms, in many cases without a confirmed history of borreliosis or neuroborreliosis.^10,17,18 This symptomatologic association (an assumed association between exposure to Bb and symptoms) may lead to negative outcomes, such as repeated antibiotic exposure or multiple consultations because of a perceived incomplete treatment response. A systematic review by Dersch et al¹⁹ describes how less rigorous definitions of disease may lead to misattributions of symptoms and consequently to unnecessary treatment. That could also partially explain the perceived incomplete response to antibiotic treatment experienced by some patients.

Moreover, the presence of nonspecific conditions, such as fatigue, affective symptoms, and cognitive complaint, are relatively frequent in the general population, and currently there is no evidence that extended antibiotic treatment for Lyme borreliosis may benefit patients in terms of cognitive or affective symptomatology.^20-22 This is also suggested by a recent randomized clinical trial in which participants had persistent symptoms attributed to Lyme borreliosis and were given a 2-week open-label intravenous treatment before a 12-week masked oral regimen (treatment vs placebo). Patients performed better in several cognitive domains during the follow-up (40 weeks), but the effect was not specific to a treatment group.²³

The results of this study are aligned with previously reported cross-sectional data suggesting that a seropositivity status for Bb (exposure to the Bb spirochetes) was not a risk factor for the previously mentioned outcomes in a population that is easily exposed to the tick and the presence of Bb.^24,25 Regarding the observed seroprevalence of Bb, it is lower compared with other reports from studies focused on agricultural workers but similar to general population data.^26-30

Limitations of this study include the lack of clinical data, such as time of exposition, manifestations of disease (if there were any), or if patients received antibiotic treatment in the case of clinical manifestations. On the other hand, despite being standardized and well-studied instruments, only 1 of each was reported for certain categories of interest (eg, MMSE for cognitive decline and CES-D for depression), this may be a limit to the interpretation of clinical manifestations because of their more general nature; however, it can be partially compensated by the fact that a specialized follow-up was made for each patient. In addition, the age and geography of participants in this study may limit the application of the results to younger persons and regions with different predominant subspecies of Bb (also to seronegative cases). Finally, in terms of follow-up and cases, there is a possibility that clinical manifestations may be different after longer periods and the number of probable and seropositive patients was small.

However, several strengths should be underlined. The study relies on a population-based sample composed of participants from a region with tick burden and whose lifestyle expose them to the tick and Bb infection. The participants had a systematic follow-up specifically designed for the diagnosis of dementia, assessment of neuropsychiatric symptoms, and disability in activities of the daily living, which, to our knowledge had never been studied in association with Bb exposure.

Exposure to Bb is an obligatory risk factor for developing Lyme borreliosis or its clinical manifestations. Previous exposure “alone” (seropositivity) to the causative agent of Lyme borreliosis is not enough as a risk factor for incidental neuropsychiatric outcomes, such as cognitive decline, dementia, depression, and even disability in older adults living in a rural environment in the southwestern region of France.

Accepted for Publication: August 4, 2019.

Corresponding Author: Virgilio Hernández Ruiz, MD, Department of Geriatrics, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Vasco de Quiroga #15, PC 14080, Tlalpan, Mexico City, México (virgilioh@gmail.com).

Published Online: September 27, 2019. doi:10.1001/jamaneurol.2019.3292

Author Contributions: Dr Pérès had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Hernandez Ruiz, Edjolo, Roubaud-Baudron, Benoît Jaulhac, Dartigues, Amieva, Peres.

Acquisition, analysis, or interpretation of data: Edjolo, Roubaud-Baudron, Avila-Funes, Amieva.

Drafting of the manuscript: Hernandez Ruiz, Edjolo, Benoît Jaulhac, Avila-Funes.

Critical revision of the manuscript for important intellectual content: Edjolo, Roubaud-Baudron, Benoît Jaulhac, Avila-Funes, Dartigues, Amieva, Peres.

Statistical analysis: Edjolo, Avila-Funes, Amieva.

Obtained funding: Dartigues, Peres.

Administrative, technical, or material support: Benoît Jaulhac, Dartigues, Amieva, Peres.

Supervision: Avila-Funes, Dartigues, Amieva.

Other - Integration of each of the coauthor’s contributions : Hernandez Ruiz.

Conflict of Interest Disclosures: Dr Jaulhac reported grants from Inserm, France during the conduct of the study. Dr Dartigues reported grants from Roche outside the submitted work. No other disclosures were reported.

Funding/Support: The AMI project was funded by AGRICA (CAMARCA, CRCCA, CCPMA PREVOYANCE, CPCEA, AGRI PREVOYANCE), the Mutualité Sociale Agricole (MSA) de Gironde, and the Caisse Centrale de la MSA. Laboratory testing conducted by the French National Reference Center for Borrelia was funded by the National Institute of Health and Medical Research–Aviesan ITMO Santé Publique. Dr Hernández Ruiz is supported by the Fundación para la Salud y la Educación Dr. Salvador Zubirán in Mexico.

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.