Accessibility settings

This paper is in the following e-collection/theme issue:

Formative Evaluation of Digital Health Interventions (5207) Clinical Informatics (2306) Electronic Health Records (1749) Advanced Data Analytics in eHealth (357) Dementia and Cognitive Decline (890) Epidemiologic Studies and Surveys in Elder Care (189) Shared Electronic Health Records and Health Information Systems (114)

Published on in Vol 10 (2026)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/83028, first published .
Alternative text does not exist

Association of Herpes Simplex Virus Type-1 With Dementia Outcomes: Longitudinal Retrospective Cohort Study Using Real-World Electronic Health Record Data

Association of Herpes Simplex Virus Type-1 With Dementia Outcomes: Longitudinal Retrospective Cohort Study Using Real-World Electronic Health Record Data

Authors of this article:

Muhammad Muinul Islam1 Author Orcid Image ;   Randi Foraker1 Author Orcid Image ;   Md Saber Hossain1 Author Orcid Image ;   W David Arnold2, 3 Author Orcid Image ;   Dima Dandachi4 Author Orcid Image ;   Abu Saleh Mohammad Mosa1, 5 Author Orcid Image
Article Authors Cited by Tweetations Metrics

    1Department of Biomedical Informatics, Biostatistics, and Medical Epidemiology, School of Medicine, University of Missouri, 1 Hospital Drive, Columbia, MO, United States

    2Department of Physical Medicine and Rehabilitation, NextGen Precision Health, University of Missouri, Columbia, MO, United States

    3Department of Neurology, NextGen Precision Health, University of Missouri, Columbia, MO, United States

    4Department of Medicine, Division of Infectious Diseases, University of Missouri, Columbia, MO, United States

    5Department of Biomedical Informatics and Data Science, Heersink School of Medicine, The University of Alabama at Birmingham, Birmingham, AL, United States

    Corresponding Author:

    Muhammad Muinul Islam, MPA, MSS, PhD


    Background: Global dementia cases, currently exceeding 55 million, are projected to triple by 2050. In the absence of disease-modifying therapies, identifying modifiable risk factors is critical. Preclinical studies show that herpes simplex virus type-1 (HSV-1) is neurotropic and can drive amyloid-beta accumulation and tau hyperphosphorylation. Epidemiologic findings remain inconsistent, partly because many studies lack standardized designs for real-world data.

    Objective: To quantify the association between clinically coded HSV-1 diagnosis and incident cognitive impairment or dementia among patients receiving care in a health-system electronic health record (EHR) network.

    Methods: We assembled a retrospective cohort within an Observational Medical Outcomes Partnership (OMOP)–mapped EHR data lake (2010‐2024), comparing patients with a first HSV-1 diagnosis (n=6274) to those without HSV-1 codes but with parallel baseline criteria (n=379,975). Eligibility required at least 365 days of prior observation and absence of baseline cognitive, HIV, selected neurotropic viral, or recent transplant codes. The outcome combined mild cognitive impairment and Alzheimer disease and related dementias (ADRD). A high-dimensional propensity score (PS) incorporating approximately 17,000 baseline covariates was estimated with regularized logistic regression; 4 strata weights were applied in a Cox model. Sensitivity analyses examined equipoise trimming, doubly adjusted models, fixed 1-, 5-, and 10-year censoring horizons, and a dementia-only outcome subset. Negative-control outcomes (n=271) were prespecified to assess empirical calibration.

    Results: The HSV-1 cohort contributed 23,186 person-years (PY) and 469 events; the non-HSV-1 cohort contributed 1,519,827 PY and 24,764 events. Crude incidence was 20.23 (95% CI 18.44‐22.14) per 1000 PY in the HSV-1 cohort and 16.29 (95% CI 16.09‐16.50) in the non-HSV-1 cohort, an absolute difference of 3.94 events per 1000 PY. PS-stratified analysis yielded a hazard ratio (HR) of 1.14 (95% CI 1.03‐1.25; P=.007). Estimates were comparable after equipoise trimming (HR 1.12, 95% CI 1.01‐1.25, P=.04), doubly adjusted modeling (HR 1.13, 95% CI 1.02‐1.24, P=.01), and fixed 5-year (P=.008) and 10-year follow-up (P=.004) (both HR 1.15). The 1-year censored model showed HR 1.00 (95% CI 0.84‐1.20; P=.96). The dementia-only endpoint mirrored the primary analysis (HR 1.14, 95% CI 1.03‐1.25; P=.008). No postindex events occurred for any negative-control outcome, so empirical calibration could not be performed and all estimates are uncalibrated.

    Conclusions: Within an OMOP-standardized EHR cohort, an HSV-1 diagnosis was associated with a small elevation in the hazard of subsequent cognitive impairment or dementia compared with individuals with no recorded HSV-1 diagnosis. Given the ubiquity of HSV-1, even a modest elevation in individual hazard could translate to a meaningful population-level increment if the association is causal. Confirmation in external data sets with laboratory viral typing, antiviral treatment records, and mortality linkage is warranted.

    JMIR Form Res 2026;10:e83028

    doi:10.2196/83028

    Keywords

    HSV-1; herpes simplex virus; dementia; Alzheimer disease; electronic health records; propensity score; aging 


    Dementia represents a major public health challenge, with prevalence trends exceeding current therapeutic capacities. Current estimates indicate 55 million individuals live with dementia worldwide, with this figure expected to nearly triple by 2050 as populations age [1-4]. In the United States alone, approximately 7 million adults aged 65 years or older are diagnosed with Alzheimer disease and related dementias (ADRD) [5-8], a number forecast to nearly double by midcentury without effective interventions. While several factors are strongly implicated, the full causal pathway and the degree to which these factors interact over time remain active areas of investigation. In addition, because disease-modifying therapies remain elusive, elucidating modifiable risk factors is essential for shaping effective prevention strategies.

    Herpes simplex virus type-1 (HSV-1), a neurotropic virus, establishes lifelong latency in peripheral sensory ganglia and reactivates periodically throughout adulthood [2,5,9-23]. Global seroprevalence estimates indicate widespread exposure, with antibody positivity rates exceeding 65% among individuals older than 50 years [16]. Experimental studies demonstrate that HSV-1 can penetrate the blood-brain barrier, induce microglial activation, and promote amyloid-beta aggregation and tau hyperphosphorylation, which are central features of Alzheimer disease (AD)–related neuropathology [9,19,24-39]. Experimental findings further suggest that recurrent HSV-1 reactivation may trigger or accelerate neuropathological mechanisms underlying mild cognitive impairment and AD, though these relationships have not been directly tested in human epidemiologic studies [30,40-43].

    Epidemiological evidence on the relationship between HSV-1 and cognitive outcomes has been mixed [44-54]. Early serological and smaller cohort studies yielded inconsistent associations [42,55,56], possibly due to variations in exposure definitions, outcome ascertainment, and insufficient control for confounders. Large-scale administrative database studies have yielded conflicting results despite improved statistical precision, potentially attributable to variability in antiviral treatment patterns, diagnostic coding practices, and challenges distinguishing between HSV-1 and HSV-2 infections [40,41,55,57-62]. Recent methodological appraisals highlight persistent limitations in the existing literature, emphasizing the need for large-scale, well-controlled longitudinal studies [5,12,15]. A systematic review focusing specifically on the association between HSV-1 and dementia has been registered and its protocol published (PROSPERO CRD42024516789) [12]; however, the completed review had not appeared in the indexed literature at the time of this writing, and well-designed individual studies using structured clinical data with explicit confounding control continue to constitute the primary evidence base.

    To address these gaps, we conducted a retrospective cohort study using Electronic Health Record (EHR) data standardized to the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) to evaluate whether a clinically documented HSV-1 diagnosis is associated with an increased hazard of incident cognitive impairment or dementia. The analysis used a new-user (incident-exposure) cohort design with an encounter-anchored comparator and up to 15 years of longitudinal follow-up, high-dimensional propensity score (PS) stratification incorporating approximately 17,000 baseline covariates spanning medication exposures, comorbidity profiles, health care usage patterns, and demographic characteristics, and a prespecified program of sensitivity analyses [12,14]. The present analysis is designed to address several of the methodological gaps anticipated by the systematic review protocol noted above: it uses formal new-user cohort construction with symmetric eligibility and an encounter-anchored comparator, applies high-dimensional PS adjustment, uses an OMOP-standardized phenotype for both exposure and outcome, and isolates HSV-1 using type-specific diagnostic concept sets, features that were absent from most prior EHR-based studies of this association. We hypothesized that coded HSV-1 exposure would be associated with an elevated hazard of cognitive impairment or dementia, consistent with mechanisms proposed in prior laboratory and translational research.


    Data Source

    We analyzed data from the University of Missouri Health Care Research Data Lake, a repository encompassing approximately 2.2 million patients with clinical encounters recorded between January 1, 2010, and December 31, 2024. The Data Lake integrates structured inpatient, outpatient, emergency department, and specialty clinic records standardized to the OMOP Common Data Model (CDM) version 5.0, enabling consistent phenotype ascertainment across heterogeneous clinical settings [63]. Cohort definitions were implemented in ATLAS version 2.12 (Observational Health Data Sciences and Informatics [OHDSI]) [64,65] and executed in R version 4.4.1 (R Foundation for Statistical Computing) using the OHDSI packages CohortMethod version 4.2.3 [66,67], FeatureExtraction version 3.7.2 [68], and EmpiricalCalibration version 3.1.3 [69,70], together with supporting dependencies. The analytic package (ATLAS JSON cohort definitions, concept sets, executable R scripts, model objects) is archived in a version-controlled public repository.

    Study Design

    We conducted a retrospective cohort study to evaluate whether a recorded HSV-1 diagnosis is associated with subsequent incident cognitive impairment or dementia. We used a new-user, incident-exposure cohort design in which both cohorts entered at incident eligibility, had a clean baseline period, and followed parallel exclusion rules. We do not describe this as an active-comparator design because, in pharmacoepidemiology, an active comparator typically includes new users of alternative treatments for the same indication, helping control confounding by indication and healthy-user or frailty bias at the design stage [71]. Comparator entry was anchored to a health-system encounter, defined by an observation period and any recorded condition occurrence. This ensured that comparator members were health-system users with measurable baseline covariates and a defined time zero, rather than untreated general-population members. Thus, the design aligned data provenance, eligibility, and health-system engagement across groups, while control of confounding by indication and related host factors relied on the high-dimensional PS approach, as discussed in the Limitations section. Eligible observation time spanned 2010-01-01 through 2024-12-31. Each participant was required to have at least 365 days of continuous observation prior to cohort entry (baseline wash-in). Follow-up for outcome assessment began 1 day after the cohort index date and continued until the earliest of (1) qualifying outcome, (2) end of continuous observation in the Data Lake, or (3) 2024-12-31 (administrative study end). This approach ensured that follow-up periods were standardized between groups, independent of subsequent diagnostic activity. A schematic of the cohort construction and time-at-risk is provided in Multimedia Appendix 1 (Figure A: analytic workflow diagram and Figure B: cohort logic flow diagram).

    Cohort Definitions

    HSV-1 Exposure Cohort

    The HSV-1 exposure cohort was defined as individuals with a first recorded diagnosis of HSV-1 during the study period (January 1, 2010, through December 31, 2024). Exposure was operationalized strictly on the basis of diagnostic codes, not downstream sequelae. This operationalization therefore identifies clinically coded HSV-1 disease rather than true biological infection status. A coded HSV-1 diagnosis most plausibly reflects a subset of infected individuals whose disease was symptomatic, recurrent, severe enough to prompt medical attention, or clinically recognized by a provider, and who received a corresponding diagnostic code during the encounter. It does not capture asymptomatic primary infection or clinically silent reactivation, which accounts for the majority of HSV-1 carriage in the general population. Given the high background seroprevalence of HSV-1 in the general adult population, the non-HSV-1 comparator therefore represents an absence of coded HSV-1 rather than a confirmed absence of infection and likely includes a substantial proportion of latently infected but never-coded individuals. This nondifferential exposure misclassification is expected to bias effect estimates toward the null and is discussed further in the Limitations section. Codes were drawn from a curated OMOP concept set including ICD-10-CM B00.x (herpesviral infections), ICD-10-CM A60.x (anogenital herpes attributed to HSV-1), and ICD-9-CM 054 .x, and were restricted to HSV-1 specific descendants or mapped concepts; HSV-2 specific and nontyped herpes simplex codes were not used for exposure qualification. A full listing of included concepts is provided in Multimedia Appendix 2. To capture incident HSV-1 exposure and minimize confounding, hierarchical exclusion criteria were applied as detailed below.

    Participants were excluded if they had a recorded diagnosis of HIV at any time prior to or on the index date. In addition, individuals with documented infection by other neurotropic viruses including varicella zoster virus, cytomegalovirus, Epstein-Barr virus, or human herpesvirus 6 or 7 within the 365-day baseline period were excluded. The cohort also excluded patients with a history of solid-organ or tissue transplantation within the 90 days preceding the index date, to avoid inclusion of immunosuppressed individuals with elevated risk for neurological complications. To ensure outcome temporality, individuals with any cognitive impairment diagnosis within the 365-day baseline window were excluded. Only the earliest qualifying HSV-1 diagnosis per person was retained for cohort entry. When multiple qualifying records occurred on the same date, they were merged into a single episode using an era-based approach (EraPad =0 day). Because systemic immunomodulatory therapies (eg, high-dose systemic corticosteroids, calcineurin inhibitors, antimetabolites, biologic DMARDs, or cytotoxic chemotherapy) can modify HSV-1 reactivation risk, we did not exclude such patients a priori; instead, baseline medication exposures including immunomodulators were captured in the high-dimensional covariate set and controlled via the PS model (Multimedia Appendix 3). Follow-up began one day after the index event (postindex day 1) and continued until the end of continuous observation.

    Non-HSV-1 (No Recorded HSV-1 Diagnosis) Comparator Cohort

    The non-HSV-1 comparator cohort was drawn from the same Data Lake and calendar frame. Cohort entry events were generated from observation periods within the study dates combined with any recorded condition occurrence to ensure an encounter-anchored index (full listing in Multimedia Appendix 4). The earliest qualifying event per person was used. Inclusion required: (1) no recorded HSV-1 diagnosis at any time prior to index; (2) no HIV infection at or before index; (3) no specified other neurotropic viral infection from 365 days before through index; (4) no solid-organ or tissue transplant from 90 days before through index; and (5) no cognitive-impairment codes in the 365-day baseline. Follow-up and exit rules matched the exposure cohort (end of continuous observation). Because comparator entry was anchored to a health-system encounter rather than an HSV-1 diagnosis, the comparator cohort shared data provenance, baseline eligibility, and health-system engagement with the HSV-1 cohort, but not a common clinical indication. We therefore characterize it as an encounter-anchored, no-recorded-HSV-1 comparator rather than an active comparator; the implications for residual confounding are addressed in the Limitations section.

    Outcome Definition

    The primary endpoint was incident all-cause cognitive impairment or dementia, ascertained using a curated OMOP concept set of more than 120 ICD-9-CM and ICD-10-CM codes spanning mild cognitive impairment, ADRD (vascular, frontotemporal, Lewy body), amnestic syndromes, and nonspecific cognitive deficit codes; the full listing is provided in Multimedia Appendix 5. Individuals with any code from this set during the 365-day baseline were excluded from both cohorts. The earliest qualifying postindex diagnosis defined the outcome event; both inpatient and outpatient settings were permitted.

    Covariates

    "Covariates for PS modeling were identified using a high-dimensional feature extraction process implemented through the FeatureExtraction package version 3.7.2 (Observational Health Data Sciences and Informatics [OHDSI], coordinating center at Columbia University, New York, NY, United States) [68].". Approximately 17,000 candidate baseline covariates were considered, capturing clinical features from two distinct observational windows prior to the index date: a long-term period (day −365 through day −31) and a short-term period immediately preceding cohort entry (day −30 through day −1). These covariates represented 5 broad domains, including medication use, clinical procedures, condition history, aggregated clinical measurements, and indicators of health care usage. In addition, patient demographic characteristics encompassing age categories, sex, race, ethnicity, and calendar timing (index month and year) were explicitly included. To prevent redundancy within the large covariate set, intermediate clinical measurement windows and standardized composite indices such as Charlson and CHADS2 scores were deliberately excluded. Rare covariates were retained without imposing frequency thresholds, and regularized logistic regression was utilized to handle sparse data. No imputation was performed for missing values; race or ethnicity data classified as “unknown” or “not recorded” in the source records were retained as distinct categories. Continuous covariates were converted to categorical or binary indicator variables per default FeatureExtraction processing methods.

    Propensity Score Estimation

    A PS for HSV-1 exposure was estimated using regularized logistic regression implemented in CohortMethod with a Laplace (L1) prior with variance 1 and tuned by ten-fold cross-validation. All 17,154 candidate baseline covariates were used in the model; 826 retained nonzero coefficients after shrinkage. Coefficients ranged from −6.56 to +2.62 (median −0.006; IQR −0.12 to +0.11). The full coefficient vector with OMOP covariate identifiers is available in Multimedia Appendix 6 to permit external reconstruction of the PS.

    PS Implementation and Balance Assessment

    Individuals were divided into four PS strata using CohortMethod defaults (strata defined to contain equal numbers of target persons). Stratum weights were applied in downstream effect estimation; no trimming was used in the primary analysis. Baseline balance was evaluated using the CohortMethod and covariateBalance output. For each of the 17,154 covariates we computed prevalence (or mean for continuous measures) by cohort and the absolute standardized mean difference (|SMD|) before and after PS stratification. Residual imbalance was considered acceptable if |SMD| was less than 0.10. Summary diagnostics appear in Figure 1 (domain-stratified Love plot) and Figure 2 (covariate-level balance scatterplot).

    ‎
    Figure 1. Domain-stratified covariate balance plot (Love plot) before and after PS weighting. PS: propensity score.
    ‎
    Figure 2. Covariate balance before versus after propensity score weighting. Each point represents one baseline covariate (n≈17,000). The dashed 45° line marks no change; points below the line indicate improved balance. Postweighting, >99 % of covariates lie below |SMD|=0.10, confirming that the propensity-score procedure effectively eliminated measurable baseline differences between cohorts.

    Attrition and Analytic Sample

    Eligibility was applied sequentially using the criteria described above. The composition of the resulting analytic cohorts is summarized in Table 1 (attrition table) and illustrated in Figure 3 (cohort flow diagram). Person-time and outcome counts are reported in the Results section.

    Table 1. Cohort attrition from candidate pools to analytic sample.
    Exclusion stepHSV-1 a candidates remainingHSV-1 removed at stepComparator candidates remainingComparator removed at stepStep attrition % (HSV-1)Step attrition % (Comp)Cumulative attrition % (HSV-1)Cumulative attrition % (Comp)
    Candidate population6900—b450,628—————
    First qualifying event and ≥365 days prior observation69000450,62800000
    No prior outcome (365-days washout)6673227446,20744213.313.31
    ≥1 day at risk postindex6274399379,97566,232614.89.115.7
    Final analytic cohort6274—379,975———9.115.7

    aHSV: herpes simplex virus type-1.

    bNot available.

    ‎
    Figure 3. Cohort attrition flow diagram. Sequential application of eligibility and exclusion criteria to derive the HSV-1 and non-HSV-1 cohorts from the University of Missouri Research Data Lake (2010‐2024). Numbers at right show exclusions at each step; final analytic counts were 6274 HSV-1 and 379,975 non-HSV-1. For further details related to HSV-1, see Table 1. HSV: herpes simplex virus type-1.

    Primary Effect Estimation

    The primary association between HSV-1 diagnosis and incident cognitive impairment or dementia was estimated using a Cox proportional hazards model stratified by PS stratum. Time zero was index date plus 1 day. Individuals were followed until outcome, end of continuous observation, or 31 December 2024, whichever occurred first. Hazard ratios with 95% CIs are reported in the Results section.

    Negative-Control Outcomes and Calibration

    We prespecified 271 negative-control outcomes (conditions with no plausible causal relationship to HSV-1, eg, psoriasis, osteoarthritis; complete list in Multimedia Appendix 7) to assess residual systematic error. Each control outcome was analyzed through the same PS-stratified Cox workflow used for the primary endpoint. Empirical null distributions were to be estimated from these controls to calibrate effect estimates following the approach of Schuemie et al [69,70]; the results of this calibration step are reported in the Results section.

    Statistical Power

    Minimum detectable relative risk was calculated using CohortMethod power diagnostics, assuming 80% power and two-sided α=.05. The calculation was applied to observed person-time and event counts from the PS-adjusted analytic sample; these values are reported in the Results section in Table 2.

    Table 2. Cohort person-time, events, crude incidence rates, and minimum detectable relative risk.
    MetricHSV-1a (Target)Non-HSV-1a (Comparator)
    Subjects6274379,975
    PYb23,1861,519,827
    Events46924,764
    Incidence rate/1000 PYb20.2316.29
    Minimum detectable RRc (80% power)—d1.15

    aHSV-1: herpes simplex virus type-1.

    bPY: person-years

    cRR: relative risk.

    dNot applicable.

    Software and Reproducibility

    Analyses were conducted in R version 4.4.1 (R Foundation for Statistical Computing) using the open-source packages CohortMethod version 4.2.3 [66,67], FeatureExtraction version 3.7.2 [68], and EmpiricalCalibration version 3.1.3 [69,70], all developed and maintained by Observational Health Data Sciences and Informatics (OHDSI), together with related dependencies. All ATLAS (developed by OHDSI) JSON cohort definitions, concept set files, analysis settings, and executable code required to reproduce the results are available in a public GitHub (Microsoft Corporation) repository [72].

    Ethical Considerations and Data Governance

    This secondary analysis of deidentified University of Missouri (MU) Health system EHR data was reviewed by the University of Missouri Institutional Review Board and determined to be non-human subjects research (IRB #2114987 MU); informed consent was waived. Analyses were conducted within a secure University of Missouri analytic environment compliant with (Health Insurance Portability and Accountability Act) HIPAA and institutional data governance policies. This study adheres to the (Strengthening the Reporting of Observational Studies in Epidemiology) STROBE guidelines; the completed checklist is provided in Checklist 1.


    The following results refer to the analytic cohorts assembled as described in the Methods section. Unless otherwise specified, counts and person-time are unweighted, and “HSV-1 cohort” refers throughout to individuals meeting the type-specific diagnostic code definition described above.

    Cohort Accrual

    We identified 6900 candidate HSV-1 index events and 450,628 potential comparator candidates in the University of Missouri Research Data Lake for the 2010‐2024 study window. After applying the continuous prior observation requirement (at least 365 days), exclusion criteria (HIV, specified other neurotropic infections, transplant), and the 365-day baseline washout for prior cognitive diagnoses, the analytic cohorts included 6,274 HSV-1 and 379,975 non-HSV-1 (no recorded HSV-1 diagnosis) individuals (about 60:1 comparator to HSV-1). Baseline attrition was 9.1% in the HSV-1 cohort and 15.7% in the comparator cohort.

    Follow-Up Time

    The distribution of follow-up times was right-skewed in both cohorts, reflecting a minority of individuals with extended observation periods. Median follow-up was 1111 days (IQR 413‐2,631; 90th percentile 2924; maximum 5460) in HSV-1 participants and 956 days (interquartile range 327‐3165; 90th percentile 3708; maximum 5478) in non-HSV-1 participants (Table 3). These distributions informed the prespecified administrative censoring horizons evaluated in sensitivity analyses (1 year, 5 years, and 10 years).

    Table 3. Distribution of individual follow-up time (days) by cohort.
    CohortMinP10P25MedianP75P90Max
    HSV-1(Target)a11404131111263129245460
    Non-HSV-1 (Comparator)171327956319537065476

    aHSV: herpes simplex virus type-1.

    Person-Time, Events, and Statistical Power

    Among participants retained in the analytic sample, the HSV-1 cohort contributed 23,186 PY and 469 incident all-cause cognitive impairment events; the non-HSV-1 cohort contributed 1,519,827 person-years and 24,764 events. Given these observed rates and sample sizes, the study achieved 80% power to detect HRs of 1.15 or greater (two-sided α=.05). Crude incidence rates are shown in Table 2.

    Crude Incidence Rates of All-Cause Cognitive Impairment

    Unweighted crude incidence of the primary outcome was 20.23 per 1000 PY (95% CI 18.44‐22.14) in HSV-1 and 16.29 per 1000 person-years (95% CI 16.09‐16.50) in non-HSV-1 participants. The absolute rate difference was about 3.94 events per 1000 PY (one additional event for every approximately 254 PY of follow-up among HSV-1 participants compared with non-HSV-1). These crude findings were consistent in direction with the adjusted HR.

    Baseline Characteristics and PS Balance

    Before PS adjustment, the HSV-1 cohort had a higher proportion assigned female at birth (Table 4) and a lower proportion aged 0‐4 years than the comparator cohort, with 72.1% (4976/6900) vs 52.8% (132,122/250,000) female and 2.7% (189/6900) vs 11.3% (28,370/250,000) aged 0‐4 years. Four-stratum PS weighting reduced these imbalances. With an absolute |SMD| threshold of less than 0.10, all demographics displayed in Table 4 met balance after weighting except age 0‐4 years (|SMD| 0.18) and race (Black |SMD| 0.11; White |SMD| −0.13). These residual differences were addressed in a doubly robust sensitivity model. Global balance across approximately 17,000 covariates is shown in the domain-stratified balance plot (Figure 1), and nearly all postweighting covariates had |SMD| less than 0.10. Baseline characteristics before and after weighting are summarized in Table 4, and full covariate-level results are provided in Multimedia Appendix 3.

    Figures 1 and 4 depict the preference score distribution & domain-stratified covariate balance plot.

    Empirical overlap in the preference score distribution (min{PS, 1 - PS}) was visualized with density plots; trimming thresholds used in sensitivity analyses are marked at 0.10 and 0.90.

    Table 4. Baseline characteristics before and after propensity score (PS) weighting.
    CharacteristicBefore targetBefore compBefore SMDaAfter targetAfter compAfter SMD
    Age 0‐4, years2.711.3−0.3418.4120.18
    Age 5‐9, years2.85.6−0.147.85.50.09
    Age 10‐14, years2.24.8−0.145.24.70.02
    Age 15‐19, years4.86.2−0.066.86.30.02
    Age 20‐24, years12.115.4−0.112.514.8−0.07
    Age 25‐29, years8.66.20.094.86.1−0.06
    Age 30‐34, years8.55.50.1245.4−0.07
    Age 35‐39, years8.14.70.144.34.7−0.02
    Age 45‐49, years6.74.40.14.54.50
    Age 50‐54, years6.74.70.093.34.7−0.07
    Age 55‐59, years6.24.80.063.14.9−0.09
    Age 60‐64, years6.94.90.094.84.9−0.02
    Age 65‐69, years64.40.0744.4−0.02
    Age 70‐74, years4.53.50.0533.4−0.03
    Age 75‐79, years3.24.5−−0.074.64.50
    Age 80‐84, years1.73.1−0.093.130
    Age 85‐89, years1.11.8−0.061.51.7−0.02
    Gender: women72.152.80.4151.953.9−0.04
    Race: Asian1.82.2−0.033.72.20.09
    Race: Black or African American8.97.10.0710.27.20.11
    Race: White85.284.70.0180.385.1−0.13
    Race: Native Hawaiian or other Pacific Islander00.1−0.010.10.10
    Race: refuse to answer0.30.6−0.040.70.60.01
    Ethnicity: Hispanic or Latino2.92.903.530.03
    Ethnicity: not Hispanic or Latino96.491.20.2290.592.3−0.06

    aSMD: standardized mean difference.

    ‎
    Figure 4. Kernel-smoothed preference score distributions for HSV-1 (red) and non-HSV-1 (blue) cohorts. HSV: herpes simplex virus type-1.

    PS Diagnostics

    PS density plots (Figures 4 and 5) showed a high concentration of HSV-1 participants at higher exposure probabilities and non-HSV-1 participants at lower probabilities, with limited overlap between about 0.30 and 0.70, consistent with strong model discrimination (c-statistic approximately 0.92). The full distribution of preference scores in exposure and comparator cohorts is provided in Multimedia Appendix 8. However, trimming observations with preference scores below 0.10 or above 0.90 retained 89 % (5584/6274) of HSV-1 and 84 % (319,179/379,975) of non-HSV-1 participants and was evaluated as a sensitivity analysis.

    Figures 2 and 5 show preference score density distributions by cohort and covariate balance scatterplot before vs after weighting.

    ‎
    Figure 5. Preference score density distributions by cohort. HSV: herpes simplex virus type-1.

    Primary Effect Estimate: PS-Stratified HR

    In the PS-stratified Cox proportional hazards model, HSV-1 diagnosis was associated with a 14% higher hazard of incident all-cause cognitive impairment (HR 1.14; 95% CI 1.03‐1.25; P=.007).

    Figures 6-8 show Kaplan-Meier survival curves, cumulative absolute risk difference, and annual survival probabilities.

    ‎
    Figure 6. Kaplan-Meier curves for remaining free of all-cause cognitive impairment up to 10 years, weighted by propensity score (PS) strata; ribbons show 95 % CI. HSV: herpes simplex virus type-1
    ‎
    Figure 7. Cumulative absolute risk difference (non-HSV-1 minus HSV-1) over 10 years. Day-by-day absolute difference in survival probability. Positive values indicate excess risk in HSV-1. HSV: herpes simplex virus type-1.
    ‎
    Figure 8. Annual Kaplan-Meier survival probabilities with 95% CI for HSV-1 (orange) and non-HSV-1 (blue) cohorts at years 1‐10 postindex. HSV: herpes simplex virus type-1.

    Cumulative Incidence Over Time

    Kaplan-Meier curves weighted by PS strata showed gradual divergence over follow-up, with clearer separation emerging over extended follow-up (Figure 6 for KM curves; Figure 7 for absolute risk difference; Figure 8 for annual survival probabilities). At 1-year postindex, cognitive disorder-free survival was about 97.7% in HSV-1 and 97.8% in non-HSV-1 participants, a difference of less than 0.1 percentage points. The gap widened to about 2.3 percentage points near year 8 and narrowed in late follow-up (about 1‐2 percentage points by year 10 (Figure 8); wider confidence limits because of attrition). The cumulative absolute risk difference curve (non-HSV-1 minus HSV-1) and annual survival probability plot illustrate these patterns.

    The absolute risk difference curve (Figure 7) increased steadily from cohort entry to a peak of approximately 2.8 percentage points near day 3048 (approximately 8.3 years of follow-up), then declined in the final observational period. This pattern reflects primarily the depletion of the HSV-1 at-risk pool in the tail rather than a true convergence of cumulative risk. At day 3000, 539 HSV-1 participants remained at risk compared with 63,554 comparator participants; by day 3500, these counts were 143 and 45,025 respectively, a proportional depletion approximately 2.5 times greater in the HSV-1 group. This asymmetry arises because most HSV-1 individuals entered the cohort after 2010, anchored on their incident diagnosis, and the 2024 administrative end date therefore limits maximum observable follow-up disproportionately for those enrolled later. With fewer than 150 HSV-1 participants at risk beyond day 3500, the Kaplan-Meier estimator becomes unstable: individual outcome events produce large step-function changes in the HSV-1 survival estimate, and the derived absolute difference fluctuates accordingly. Additionally, individuals still event-free at more than 8 years represent a highly selected survivor population, which may attenuate the apparent relative difference independent of statistical noise. For these reasons, point estimates of the absolute risk difference beyond year 8 should be interpreted with caution; the 5- and 10-year administratively censored sensitivity analyses (Table 5) provide more stable comparisons over those defined horizons.

    Table 5. Summary of sensitivity analyses (preference score trimming, doubly robust model, administrative censoring, dementia-only endpoint).
    AnalysisHorizonHRa95% CIP valuebNotes
    Preference-score trimming<0.10 or>0.90 removed1.121.01‐1.25.04Sensitivity
    Doubly robust (add age 0‐4 years, race White, race Black)—c1.131.02‐1.24.01Sensitivity
    Admin censor 1 yearPostindex1.000.84‐1.20>.99Reconstructed PS-stratd Cox
    Admin censor 5 yearsPostindex1.151.04‐1.28.008Reconstructed PS-strat Cox
    Admin censor 10 yearsPostindex1.151.04‐1.26.004Reconstructed PS-strat Cox
    Dementia-only outcomeADRDe subset1.141.03‐1.25.008Sensitivity

    aHR: hazard ratio.

    bNote: Two-sided P value computed from HR and 95% CI as described above; values are rounded to three significant digits.

    cNot available.

    dPS: propensity score.

    eADRD: Alzheimer disease and related dementias.

    Negative-Control Outcomes

    In the MU Data Lake extract, no postindex events were observed in either cohort for any of the 271 prespecified negative-control outcomes (0 events across all 271 outcomes in both the HSV-1 [n=6274] and non-HSV-1 [n=379,975] groups; full listing in Multimedia Appendix 7). Because no events occurred, an empirical null distribution could not be constructed, and effect calibration was not performed; all reported HRs are therefore uncalibrated. This finding likely reflects the rarity of the selected control conditions within the single-institution study window and the relatively modest size of the HSV-1 cohort, rather than true absence of these conditions in the broader population.

    Sensitivity Analyses

    Sensitivity analyses demonstrated consistent findings across multiple analytic strategies. After applying preference-score trimming, excluding participants with preference scores below 0.10 or above 0.90, the adjusted HR was 1.12, (95% CI 1.01‐1.25). In a doubly adjusted model accounting for residual imbalance in the age 0‐4 years and race (White; Black or African American) covariates, the estimated HR was 1.13 (95% CI 1.02‐1.24). Administrative censoring at one-year postindex showed no measurable association (HR 1.00, 95% CI 0.84‐1.20), a finding consistent with the low absolute number of outcome events accruing within the first year of follow-up and indicating that the primary signal emerges over extended follow-up rather than in the immediate postindex period. Censoring at five-year and ten-year horizons yielded HRs consistent with the primary finding (HR 1.15, 95% CI 1.04‐1.28, and HR 1.15, 95% CI 1.04‐1.26, respectively), confirming that the association becomes statistically detectable only as cumulative follow-up time increases. Restricting the outcome definition to dementia attributable to Alzheimer disease produced a HR identical to the primary estimate (HR 1.14, 95% CI 1.03‐1.25), reinforcing the stability of findings across these analytic scenarios.


    Principal Findings

    Among 386,249 participants followed for up to 15 years, individuals with a clinically documented HSV-1 diagnosis experienced a 14% higher hazard of incident cognitive impairment or dementia than comparator participants without coded HSV-1 diagnoses (HR 1.14, 95% CI 1.03‐1.25, P=.007). Throughout this analysis, “HSV-1 exposure” refers specifically to a coded HSV-1 diagnosis in structured EHR data, which most plausibly reflects symptomatic or clinically recognized disease rather than latent or asymptomatic infection; the clinical phenotype captured is therefore a subset of the larger infected population identifiable only by serology. The absolute incidence was approximately 4 events per 1000 PY higher in the HSV-1 cohort (Table 2). PS-weighted survival curves showed gradual divergence that became progressively more apparent and sustained across extended follow-up (Figures 6-8). This pattern is consistent with the sensitivity analyses, in which the 1-year administratively censored model yielded no measurable association, whereas the 5-year and 10-year censored models produced HRs consistent with the primary estimate, together indicating that the association accumulates over extended follow-up rather than being concentrated in the immediate postindex period. The high seroprevalence of HSV-1 in the general adult population means that, if the observed association reflects a causal relationship, even a 14% increase in individual hazard could translate to a meaningful population-attributable increment in dementia cases, a point elaborated in the implications section below.

    Comparison With Prior Work

    Prior epidemiological investigations have reported heterogeneous results, likely reflecting differences in case definitions, study design, and confounding adjustment. A nested case-control study using Swedish registry data reported elevated odds of AD in individuals with prior HSV-1 infection, suggesting a temporal link between infection and diagnosis [61]. A Taiwan-based retrospective cohort study using national claims data found increased dementia incidence among patients with HSV diagnoses, though that study combined HSV-1 and HSV-2 under a single code [17]. Our HR is directionally concordant but smaller, possibly reflecting our narrower HSV-1 concept set, finer confounder adjustment via high-dimensional PS stratification, and the single health-system EHR source [61,73]. A parallel analysis by Chen et al [59] examining antiviral treatment in herpetic infections suggested a protective association, but that study did not isolate HSV-1 and did not use PS methods [73].

    The biological plausibility of the observed association is supported by mechanistic evidence. Experimental models show that HSV-1 can induce tau hyperphosphorylation and amyloid-beta aggregation in neuronal cultures [11,74,75], disrupt tau exon splicing in human neuron preparations [76,77], and accelerate microglial activation in murine AD models [78,79]. A prospective study within the Multidomain Alzheimer Preventive Trial cohort detected HSV-1 DNA in peripheral blood mononuclear cells and found seropositivity associated with abnormal phosphorylated tau and Abeta42/40 ratios [2,40,61]. The temporal association between HSV-1 diagnosis and cognitive impairment in our cohort is consistent with, but does not directly test, these mechanistic hypotheses; biomarker and laboratory data that would be required to evaluate them were unavailable in this study.

    The magnitude of the observed association in this study (HR 1.14) is smaller than estimates from several earlier investigations, and misclassification of exposure is the most parsimonious explanation for this discrepancy. Studies reporting HRs or odds ratios of 1.5 or greater for HSV and dementia have generally used seroprevalence data or seroconversion designs [25,60], in which the exposed group includes all seropositive individuals regardless of whether they received a clinical diagnosis code. In contrast, EHR-based studies restricted to coded diagnoses, as in the present analysis, identify only a small, clinically visible fraction of the truly exposed population. The non-HSV-1 comparator group therefore includes a large number of latently infected individuals, diluting the between-group difference and attenuating the observed HR toward the null. Rather than indicating that the true association is smaller than previously reported, our estimate is best interpreted as a lower bound of the true association under the assumption of nondifferential misclassification. This interpretation is consistent with the systematic evidence gap that Hong et al [12] identified when publishing their systematic review protocol in 2025, namely that the field lacks adequately powered, methodologically rigorous longitudinal studies using structured clinical data with explicit confounding control, a gap the present analysis is specifically designed to address.

    Implications for Practice and Research

    The present findings carry several specific implications for clinical practice, health system design, and future research, even in the absence of causal confirmation. Before describing these, we emphasize that a HR of 1.14, while statistically meaningful and directionally consistent across sensitivity analyses, represents a modest effect at the individual patient level and does not, in isolation, justify any change to current clinical management of HSV-1 infection or to existing cognitive-screening recommendations. The findings are intended to guide future research rather than to inform immediate changes in care. At the clinical level, the results suggest that a coded HSV-1 diagnosis, already routinely captured in structured EHR data, may serve as a readily available indicator for an elevated likelihood of subsequent cognitive impairment that clinicians could factor into decisions about cognitive screening. This does not imply that every patient with a cold sore diagnosis requires neurological surveillance; the absolute incidence difference of approximately 4 events per 1000 PY is modest. However, in older patients with additional factors associated with dementia, including advanced age, apolipoprotein E. (APOE) epsilon4 carrier status, and cardiovascular disease burden, a history of clinically documented HSV-1 infection may justify earlier cognitive screening or enrollment in prevention programs. Health systems with OMOP-standardized data could operationalize such surveillance without additional data collection.

    At the population level, the estimated seroprevalence of HSV-1 exceeding 65% among adults over 50 years means that even a HR of 1.14 translates to a nontrivial population-attributable fraction of late-life cognitive impairment. If the association is causal, a meaningful proportion of late-life cognitive decline may be linked to this common, potentially modifiable exposure. Antiviral therapies that suppress HSV-1 reactivation, including acyclovir and valacyclovir, are safe, inexpensive, and widely available. Several observational studies have suggested that treated patients have lower dementia incidence than untreated patients [59,61,80], and a registry-based cohort study reported an association between antiherpetic therapy and reduced AD incidence [61]. These reports are hypothesis-generating and are subject to confounding by indication, healthy-user effects, and adherence-related bias; they do not by themselves establish that antiviral therapy modifies dementia risk. Taken together with the mechanistic and epidemiologic evidence summarized above, our findings support continued investigation into whether HSV-1 reactivation contributes meaningfully to cognitive decline and whether antiviral strategies targeting reactivation warrant formal evaluation in appropriately designed prospective studies. The present analysis does not, on its own, provide a basis for recommending antiviral prophylaxis for dementia prevention.

    This study provides an analytical template, including OMOP-standardized cohort definitions, open-source ATLAS and CohortMethod code, and a replicable PS framework, that could support replication across multiple OMOP-standardized health systems prior to trial design. Prospective studies that distinguish HSV-1 from HSV-2 serologically, quantify viral reactivation frequency, capture antiviral treatment patterns, and incorporate genetic susceptibility markers, including APOE genotype and markers of immune response such as IGDCC4 [81], are needed to determine whether antiviral therapy or immunomodulation could modify dementia trajectories.

    Large, multisite studies integrating laboratory-confirmed infection data, pharmacy claims, and vital status linkage will be necessary to validate these findings and to clarify whether virologic suppression is associated with reduced dementia incidence at the population level.

    Strengths and Limitations

    Strengths

    The methodological design of this study is specifically constructed to address the threats to validity that have limited prior EHR-based investigations of HSV-1 and cognitive outcomes, and each design element corresponds to a known weakness in the existing literature. The new-user design with symmetric eligibility and an encounter-anchored comparator is the design element that most clearly separates this study from prior EHR-based analyses, which have generally used prevalent-user or simple-incidence designs. By requiring both HSV-1 and comparator cohort members to have at least 365 days of prior continuous observation free of the outcome, and by anchoring comparator entry on an encounter-based event rather than arbitrary calendar assignment, the design ensures that baseline covariate measurement, outcome washout, and follow-up rules are symmetric across groups. This approach eliminates immortal time bias and reduces index event confounding, two threats that prevalent-user designs cannot adequately address.

    The use of OMOP CDM standardization with type-specific HSV-1 concept sets represents a substantive advance over prior studies that used broad herpesvirus codes or administrative claims codes that cannot distinguish HSV-1 from HSV-2. The curated concept set, restricted to ICD-10-CM B00.x, A60.x (HSV-1-attributed anogenital herpes), and ICD-9-CM 054 .x descendants, and explicitly excluding nontyped codes, produces the most phenotypically specific coded HSV-1 exposure definition achievable from structured EHR data without laboratory linkage.

    The high-dimensional PS, incorporating approximately 17,000 baseline covariates spanning medication use, condition history, procedures, clinical measurements, and health care usage, addresses a limitation that affects smaller studies: residual confounding from unmeasured or partially measured baseline differences. The L1-penalized model ensures parsimony while retaining flexibility, and the full coefficient vector is publicly archived to support reproducibility. The resulting covariate balance, with absolute |SMD| below 0.10 for the large majority of covariates after stratification, is substantially better than what is achievable with manually specified propensity models in this clinical domain. The prespecified sensitivity analysis program, encompassing preference-score trimming, doubly adjusted modeling, three fixed administrative censoring horizons, and a dementia-only outcome restriction, was designed prior to examination of results and provides a transparent record of estimate stability across design perturbations. The concordance of all 6 analyses with the primary estimate supports the robustness of the finding. In addition, statistical power calculations confirmed adequate precision to detect HRs of approximately 1.15, and 271 negative-control outcomes were prespecified to evaluate residual systematic bias, although no postindex events were observed for any control outcome and empirical calibration could not be performed.

    Finally, the complete analytic package, including ATLAS JSON cohort definitions, concept set files, CohortMethod R scripts, propensity model objects, and supporting code, is publicly archived and executable, enabling direct replication in any OMOP-standardized data network. This supports federated replication across OHDSI network sites and supports use of the present analysis as a reference for multi-database validation.

    Limitations

    A central limitation of this study is that exposure was defined entirely from clinical diagnosis codes, which capture clinically recognized HSV-1 disease rather than true biological infection status. Because HSV-1 establishes lifelong latency and reactivates without producing clinically attended symptoms in most infected individuals, only a small fraction of carriers ever receive a diagnostic code. Population-based serologic surveys estimate that approximately 65% of adults over 50 carry HSV-1 antibodies, yet clinical codes identify fewer than 5% of the seropositive population in EHR-based studies; the 6274 coded individuals in a data lake of 2.2 million patients are consistent with this gap. The practical consequence is that the comparison is not between infected and uninfected individuals, but between individuals with a coded HSV-1 diagnosis and a comparator group that, given the high background seroprevalence, almost certainly contains a large number of latently infected but never-coded individuals. This contamination of the unexposed group weakens the true exposure contrast and constitutes nondifferential exposure misclassification, which is expected to bias the HR toward the null. As discussed in the Comparison With Prior Work section, the observed HR is therefore attenuated and is best interpreted as a lower bound on the association between HSV-1 and cognitive impairment rather than as a direct estimate of the effect of infection itself. For the same reason, the coded-exposure design cannot separate the effect of HSV-1 infection from the effect of clinically apparent or recurrent HSV-1 disease, and the causal interpretation of this contrast should be made with corresponding caution. Restriction to HSV-1-specific ICD codes reduced, but did not eliminate, this misclassification; definitive resolution would require serologic or virologic confirmation of infection status in both cohorts, which was not available in this data source and is identified above as a priority for replication.

    Beyond misclassification, a coded HSV-1 diagnosis may function in part as a marker of underlying host characteristics rather than as a direct causal exposure. Individuals who receive a clinical HSV-1 diagnosis are disproportionately those in whom the virus has reactivated with sufficient clinical severity to prompt medical evaluation, which is itself influenced by age-related immunosenescence, frailty, comorbidity burden, and relative immune dysregulation from other chronic conditions. Each of these features is independently associated with an elevated risk of cognitive impairment. Although the high-dimensional PS incorporated approximately 17,000 baseline covariates capturing medication use, condition history, clinical procedures, and health care usage, residual confounding by unmeasured dimensions of host susceptibility and immune status cannot be excluded. Accordingly, part of the observed association between coded HSV-1 and subsequent cognitive impairment may reflect shared antecedents of both outcomes rather than a direct effect of the virus on neurodegeneration. This consideration reinforces the need for replication in cohorts with richer immune, genetic, and frailty phenotyping before causal claims can be advanced.

    Several additional limitations should be noted. Educational attainment and APOE genotype were not captured and could not be included in the PS model. Antiviral treatment patterns could not be systematically ascertained from diagnosis codes alone. Death and other competing outcomes were incompletely recorded, and the observed HR may therefore be subject to competing risks bias. The absence of postindex events for any of the 271 prespecified negative-control outcomes prevented empirical calibration; all reported estimates are uncalibrated, and residual systematic error cannot be formally quantified.

    Preference-score distributions showed limited overlap between groups, with a c-statistic of approximately 0.92, indicating substantial baseline differences and supporting interpretation primarily within the region of covariate overlap. Although preference-score trimming yielded estimates consistent with the primary analysis, the comparator was encounter-anchored and defined by the absence of a recorded HSV-1 diagnosis rather than by a shared clinical indication. The high-dimensional PS improved balance across measured covariates, with absolute |SMD| below 0.10 for the large majority of covariates; however, residual confounding by unmeasured determinants of both HSV-1 diagnosis and cognitive impairment, including care-seeking behavior, symptomatic burden, frailty, and host susceptibility, cannot be excluded. The absolute risk difference curve narrows after approximately day 3000, primarily because of differential depletion of the HSV-1 at-risk pool rather than true convergence of cumulative incidence; tail-region estimates should be interpreted cautiously.

    Eligibility criteria excluded individuals with recorded HIV, other selected neurotropic viral infections, or recent solid-organ or tissue transplantation in order to reduce confounding from established immunocompromising conditions. As a consequence, the findings may not generalize to immunocompromised populations, including people living with HIV, transplant recipients, and individuals on systemic immunosuppression, in whom HSV-1 reactivation frequency, clinical presentation, and central nervous system involvement are known to differ. Dedicated studies would be required before extrapolating these findings to such groups.

    Finally, all data were drawn from a single Midwestern academic health system. Diagnostic coding intensity, patient demographics, and follow-up completeness may differ across institutions and regions. Replication in external OMOP-standardized networks with more diverse populations will be necessary to assess the generalizability of these findings.

    Conclusions

    In a fifteen-year EHR cohort, patients with any HSV-1 diagnosis code experienced a 14% higher hazard of incident cognitive impairment, including dementia subtypes, than comparator patients with no recorded HSV-1 diagnosis (adjusted HR 1.14, 95% CI 1.03‐1.25). Absolute incidence differed by about four events per 1000 PY (Table 2), and PS-weighted survival curves diverged gradually, with separation emerging over extended follow-up (Figures 6-8). Although the estimated association is smaller than some earlier reports, it is consistent with estimates from recent longitudinal studies that applied extensive adjustment. Measurement limitations, incomplete death data, and potential residual confounding preclude causal inference; nevertheless, the consistent direction across sensitivity analyses supports a modest association within the observed data environment. A HR of 1.14 represents a modest effect on the individual level, and these findings alone do not support changes to current clinical practice for HSV-1 management or cognitive screening. Given the high prevalence of HSV-1, if the observed association is causal, even this modest relative excess could contribute meaningfully to population dementia burden. Future studies should integrate laboratory-confirmed HSV-1 typing, antiviral treatment histories, and genetic data to clarify whether virologic suppression is associated with reduced dementia incidence. If a causal relationship is confirmed, sustained antiviral suppression and future HSV-1 vaccines could become useful tools for reducing dementia incidence at the population level.

    Acknowledgments

    The authors declare the use of generative AI in the research and writing process. According to the GAIDeT taxonomy (2025), the following tasks were delegated to GAI tools under full human supervision:

    - Proofreading and editing

    - Recommendations

    The GAI tool used was: Claude (Anthropic).

    Additional note: Use was limited to grammar and language checking and to editorial recommendations on structure and clarity. All content, data analysis, interpretation, and conclusions were produced and verified by the authors.

    Funding

    This work was supported in part by the National Institutes of Health Clinical and Translational Science Award program (NIH 2UL1TR002345) and by the PCORnet Technical Enhancement Award (PCORI RI-MISSOURI-01-PS6) for the development and deployment of OMOP Common Data Model and ATLAS. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, PCORnet, or the authors’ institutions. We thank the NextGen Biomedical Informatics Center (BMI) at the University of Missouri for technical and administrative support, including sustained access to the Research Data Lake, data-governance support, and deployment of the ATLAS application. We thank Vasanthi Mandhadi for technical assistance in investigating and resolving source-data discrepancies identified during cohort construction, including corrections to the PCORnet CDM-to-OMOP extract-transform-load process underlying the ATLAS instance. The funders had no role in study design, data collection, data analysis, interpretation, or the decision to submit the manuscript.

    Data Availability

    Data are available from the author upon reasonable request.

    Authors' Contributions

    Conceptualization: MMI (lead), ASMM (equal).

    Data curation: MMI (lead), MSH (equal), ASMM (supporting).

    Formal analysis: MMI (lead), MSH (supporting), WDA (supporting), ASMM (supporting).

    Investigation: MMI.

    Methodology: MMI (lead), WDA (equal), DD (equal), ASMM (equal).

    Resources: MMI (equal), ASMM (equal).

    Software: MSH (lead), MMI (equal), ASMM (supporting).

    Supervision: ASMM (lead), RF (equal).

    Visualization: MMI.

    Writing – original draft: MMI.

    Writing – review & editing: MMI (lead), RF (supporting), MSH (supporting), WDA (supporting), DD (supporting), ASMM (supporting).

    Conflicts of Interest

    The authors declare no conflicts of interest related to this work. No external funding was received for the conception, analysis, or writing of this manuscript. Outside the submitted work, DD reports research funding and honoraria from Gilead Sciences and ViiV Healthcare. These relationships are unrelated to the present study. All other authors report no relevant financial relationships.

    Multimedia Appendix 1

    Analytic workflow diagram and cohort logic flow diagram.

    DOCX File, 1426 KB

    Multimedia Appendix 2

    HSV-1 cohort concept sets.

    XLSX File, 22 KB

    Multimedia Appendix 3

    Covariate balance.

    XLSX File, 175 KB

    Multimedia Appendix 4

    Non-HSV-1 cohort concept sets.

    XLSX File, 22 KB

    Multimedia Appendix 5

    Incident cognitive impairment or dementia cohort concept sets.

    XLSX File, 14 KB

    Multimedia Appendix 6

    Propensity model.

    XLSX File, 42 KB

    Multimedia Appendix 7

    Negative control list.

    XLSX File, 19 KB

    Multimedia Appendix 8

    Distribution of preference scores in exposure and comparator cohorts.

    DOCX File, 21 KB

    Checklist 1

    STROBE checklist.

    DOCX File, 20 KB
    1. Hyde VR, Zhou C, Fernandez JR, et al. Anti-herpetic tau preserves neurons via the cGAS-STING-TBK1 pathway in Alzheimer’s disease. Cell Rep. Jan 2025;44(1):115109. [CrossRef]
    2. Powell-Doherty RD, Abbott ARN, Nelson LA, Bertke AS. Amyloid-β and p-Tau anti-threat response to herpes simplex virus 1 infection in primary adult murine hippocampal neurons. J Virol. Apr 16, 2020;94(9):e01874-19. [CrossRef] [Medline]
    3. Global status report on the public health response to dementia. World Health Organization (WHO). Sep 1, 2021. URL: https://www.who.int/publications/i/item/9789240033245 [Accessed 2026-08-20]
    4. Dementia. World Health Organization (WHO). 2023. URL: https://www.who.int/news-room/fact-sheets/detail/dementia [Accessed 2026-08-20]
    5. Nichols E, Vos T. The estimation of the global prevalence of dementia from 1990‐2019 and forecasted prevalence through 2050: an analysis for the Global Burden of Disease (GBD) study 2019. Alzheimers Dement. Dec 2021;17(S10):e105-e125. [CrossRef]
    6. Alzheimer’s Assocation. 2025 Alzheimer’s Disease Facts and Figures: Special Report on American Perspectives on Early Detection of Alzheimer’s Disease in the Era of Treatment. Vol 21. Alzheimer’s Dement; 2025. [CrossRef]
    7. Alzheimer’s Association. 2024 Alzheimer’s Disease Facts and Figures. Vol 20. Alzheimer’s & Dementia; 2024:3708-3821. [CrossRef]
    8. Looker KJ, Magaret AS, May MT, et al. Global and regional estimates of prevalent and incident herpes simplex virus type 1 infections in 2012. PLoS One. 2015;10(10):e0140765. [CrossRef] [Medline]
    9. Duarte LF, Farías MA, Álvarez DM, Bueno SM, Riedel CA, González PA. Herpes simplex virus type 1 infection of the central nervous system: insights into proposed interrelationships with neurodegenerative disorders. Front Cell Neurosci. 2019;13:46. [CrossRef] [Medline]
    10. James LM, Stratigopoulos G, Georgopoulos AP. Human herpes viruses are associated with steeper age-dependent increases of serum biomarkers for dementia in cognitively unimpaired women. Sci Rep. Jul 15, 2025;15(1):25475. [CrossRef] [Medline]
    11. Itzhaki RF. Overwhelming evidence for a major role for herpes simplex virus type 1 (HSV1) in Alzheimer’s disease (AD); underwhelming evidence against. Vaccines (Basel). 9(6):679. [CrossRef]
    12. Hong AT, Luu IYK, Lin F, et al. Association between herpes simplex virus 1 and dementia: a systematic review protocol. BMJ Open. Apr 19, 2025;15(4):e088632. [CrossRef] [Medline]
    13. Feng S, Liu Y, Zhou Y, et al. Mechanistic insights into the role of herpes simplex virus 1 in Alzheimer’s disease. Front Aging Neurosci. 2023;15:1245904. [CrossRef] [Medline]
    14. Itzhaki RF. Herpes and Alzheimer’s Disease: subversion in the central nervous system and how it might be halted. J Alzheimers Dis. Oct 18, 2016;54(4):1273-1281. [CrossRef] [Medline]
    15. Linard M, Garrigue I, Vellas B, et al. Association between herpes simplex virus infection and Alzheimer’s disease biomarkers: analysis within the MAPT trial. Sci Rep. Jan 18, 2025;15(1):2362. [CrossRef] [Medline]
    16. De Chiara G, Piacentini R, Fabiani M, et al. Recurrent herpes simplex virus-1 infection induces hallmarks of neurodegeneration and cognitive deficits in mice. PLoS Pathog. Mar 2019;15(3):e1007617. [CrossRef] [Medline]
    17. Tzeng NS, Chung CH, Lin FH, et al. Anti-herpetic medications and reduced risk of dementia in patients with herpes simplex virus infections-a nationwide, population-based cohort study in Taiwan. Neurotherapeutics. Apr 2018;15(2):417-429. [CrossRef] [Medline]
    18. Harfouche M, AlMukdad S, Alareeki A, et al. Estimated global and regional incidence and prevalence of herpes simplex virus infections and genital ulcer disease in 2020: mathematical modeling analyses. medRxiv. Preprint posted online on Jun 3, 2024. [CrossRef]
    19. Gopinath D, Koe KH, Maharajan MK, Panda S. A comprehensive overview of epidemiology, pathogenesis and the management of herpes labialis. Viruses. Jan 13, 2023;15(1):225. [CrossRef] [Medline]
    20. Itzhaki RF. Herpes simplex virus type 1 and Alzheimer’s disease: increasing evidence for a major role of the virus. Front Aging Neurosci. 2014;6:202. [CrossRef] [Medline]
    21. Itzhaki RF. Herpes simplex virus type 1 and Alzheimer’s disease: possible mechanisms and signposts. FASEB J. Aug 2017;31(8):3216-3226. [CrossRef] [Medline]
    22. Agostini S, Clerici M, Mancuso R. How plausible is a link between HSV-1 infection and Alzheimer’s disease? Expert Rev Anti Infect Ther. Mar 2014;12(3):275-278. [CrossRef] [Medline]
    23. Piacentini R, De Chiara G, Li Puma DD, et al. HSV-1 and Alzheimer’s disease: more than a hypothesis. Front Pharmacol. 2014;5:97. [CrossRef] [Medline]
    24. Readhead B, Haure-Mirande JV, Funk CC, et al. Multiscale analysis of independent Alzheimer’s cohorts finds disruption of molecular, genetic, and clinical networks by human herpesvirus. Neuron. Jul 11, 2018;99(1):64-82. [CrossRef] [Medline]
    25. Lövheim H, Gilthorpe J, Johansson A, Eriksson S, Hallmans G, Elgh F. Herpes simplex infection and the risk of Alzheimer’s disease: a nested case-control study. Alzheimers Dement. Jun 2015;11(6):587-592. [CrossRef] [Medline]
    26. Steiner I, Kennedy PGE, Pachner AR. The neurotropic herpes viruses: herpes simplex and varicella-zoster. Lancet Neurol. Nov 2007;6(11):1015-1028. [CrossRef] [Medline]
    27. Menendez CM, Carr DJJ. Defining nervous system susceptibility during acute and latent herpes simplex virus-1 infection. J Neuroimmunol. Jul 15, 2017;308:43-49. [CrossRef] [Medline]
    28. Albaret MA, Textoris J, Dalzon B, et al. HSV-1 cellular model reveals links between aggresome formation and early step of Alzheimer’s disease. Transl Psychiatry. Mar 10, 2023;13(1):86. [CrossRef] [Medline]
    29. Itzhaki RF, Lin WR, Shang D, Wilcock GK, Faragher B, Jamieson GA. Herpes simplex virus type 1 in brain and risk of Alzheimer’s disease. The Lancet. Jan 1997;349(9047):241-244. [CrossRef]
    30. Eimer WA, Vijaya Kumar DK, Navalpur Shanmugam NK, et al. Alzheimer’s disease-associated β-amyloid Is rapidly seeded by herpesviridae to protect against brain infection. Neuron. Jul 11, 2018;99(1):56-63. [CrossRef] [Medline]
    31. Campos EMN, Rodrigues LD, Oliveira LF, Dos Santos JCC. Dementia and cognitive impairment in adults as sequels of HSV-1-related encephalitis: a review. Dement Neuropsychol. 2021;15(2):164-172. [CrossRef] [Medline]
    32. Bearer EL, Wu C. Herpes simplex virus, Alzheimer’s disease and a possible role for rab GTPases. Front Cell Dev Biol. 2019;7:134. [CrossRef] [Medline]
    33. Marcocci ME, Napoletani G, Protto V, et al. Herpes simplex virus-1 in the brain: the dark side of a sneaky infection. Trends Microbiol. Oct 2020;28(10):808-820. [CrossRef] [Medline]
    34. Wozniak MA, Shipley SJ, Combrinck M, Wilcock GK, Itzhaki RF. Productive herpes simplex virus in brain of elderly normal subjects and Alzheimer’s disease patients. J Med Virol. Feb 2005;75(2):300-306. [CrossRef] [Medline]
    35. Harris SA, Harris EA. Herpes simplex virus type 1 and other pathogens are key causative factors in sporadic Alzheimer’s Disease. J Alzheimers Dis. 2015;48(2):319-353. [CrossRef] [Medline]
    36. Sait A, Angeli C, Doig AJ, Day PJR. Viral involvement in Alzheimer’s disease. ACS Chem Neurosci. Apr 7, 2021;12(7):1049-1060. [CrossRef] [Medline]
    37. Naughton SX, Raval U, Pasinetti GM. The viral hypothesis in Alzheimer’s disease: novel insights and pathogen-based biomarkers. J Pers Med. Jul 29, 2020;10(3):74. [CrossRef] [Medline]
    38. Cairns DM, Itzhaki RF, Kaplan DL. Potential involvement of varicella zoster virus in Alzheimer’s disease via reactivation of quiescent herpes simplex virus type 1. J Alzheimers Dis. 2022;88(3):1189-1200. [CrossRef] [Medline]
    39. Kayed R. Infectious etiology and amyloidosis in Alzheimer’s disease: the puzzle continues. J Biol Chem. Aug 2021;297(2):100936. [CrossRef] [Medline]
    40. Linard M, Letenneur L, Garrigue I, Doize A, Dartigues JF, Helmer C. Interaction between APOE4 and herpes simplex virus type 1 in Alzheimer’s disease. Alzheimers Dement. Jan 2020;16(1):200-208. [CrossRef] [Medline]
    41. Dobson CB, Itzhaki RF. Herpes simplex virus type 1 and Alzheimer’s disease. Neurobiol Aging. 1999;20(4):457-465. [CrossRef] [Medline]
    42. Warren-Gash C, Forbes HJ, Williamson E, et al. Human herpesvirus infections and dementia or mild cognitive impairment: a systematic review and meta-analysis. Sci Rep. Mar 18, 2019;9(1):4743. [CrossRef] [Medline]
    43. Liu Y, Johnston C, Jarousse N, Fletcher SP, Iqbal S. Association between herpes simplex virus type 1 and the risk of Alzheimer’s disease: a retrospective case–control study. BMJ Open. May 2025;15(5):e093946. [CrossRef]
    44. Duarte LF, Gatica S, Castillo A, et al. Is there a role for herpes simplex virus type 1 in multiple sclerosis? Microbes Infect. Jun 2023;25(5):105084. [CrossRef] [Medline]
    45. Romanescu C, Schreiner TG, Mukovozov I. The role of human herpesvirus 6 infection in Alzheimer’s disease pathogenicity-a theoretical mosaic. J Clin Med. May 29, 2022;11(11):3061. [CrossRef] [Medline]
    46. Piotrowski SL, Tucker A, Jacobson S. The elusive role of herpesviruses in Alzheimer’s disease: current evidence and future directions. NeuroImmune Pharm Ther. Sep 2023;2(3):253-266. [CrossRef] [Medline]
    47. Schmidt SAJ, Veres K, Sørensen HT, Obel N, Henderson VW. Incident herpes zoster and risk of dementia. Neurology. Aug 16, 2022;99(7). [CrossRef]
    48. Ijezie EC, Miller MJ, Hardy C, et al. Herpes simplex virus-1 infection alters microtubule-associated protein tau splicing and promotes tau pathology in neural models of Alzheimer’s disease. Brain Pathol. Sep 2025;35(5):e70006. [CrossRef] [Medline]
    49. Cantero JL, Atienza M, Sastre I, Bullido MJ. Human in vivo evidence of associations between herpes simplex virus and cerebral amyloid-beta load in normal aging. Alzheimers Res Ther. Apr 3, 2024;16(1):68. [CrossRef] [Medline]
    50. Dutton AJ, Hayes CK, Leib DA, Akhtar LN. Herpes simplex virus neurovirulence across the human lifespan. J Infect Dis. Jul 30, 2025;232(1):14-24. [CrossRef] [Medline]
    51. Laval K, Enquist LW. The potential role of herpes simplex virus type 1 and neuroinflammation in the pathogenesis of Alzheimer’s disease. Front Neurol. 2021;12:658695. [CrossRef] [Medline]
    52. Khokale R, Kang A, Buchanan-Peart KAR, Nelson ML, Awolumate OJ, Cancarevic I. Alzheimer’s gone viral: could herpes simplex virus type-1 be stealing your memories? Cureus. Nov 27, 2020;12(11):e11726. [CrossRef] [Medline]
    53. Frölich L. Alzheimer’s disease - the “microbial hypothesis” from a clinical and neuroimaging perspective. Psychiatry Res Neuroimaging. Dec 30, 2020;306:111181. [CrossRef] [Medline]
    54. Carbone I, Lazzarotto T, Ianni M, et al. Herpes virus in Alzheimer’s disease: relation to progression of the disease. Neurobiol Aging. Jan 2014;35(1):122-129. [CrossRef] [Medline]
    55. Torniainen-Holm M, Suvisaari J, Lindgren M, Härkänen T, Dickerson F, Yolken RH. The lack of association between herpes simplex virus 1 or Toxoplasma gondii infection and cognitive decline in the general population: an 11-year follow-up study. Brain Behav Immun. Feb 2019;76:159-164. [CrossRef] [Medline]
    56. Rizzo R. Controversial role of herpesviruses in Alzheimer’s disease. PLoS Pathog. Jun 2020;16(6):e1008575. [CrossRef] [Medline]
    57. Eisenstein M. Do infections have a role in Alzheimer’s disease? Nature. Apr 2025;640(8059):S8-S10. [CrossRef] [Medline]
    58. Abbott A. Are infections seeding some cases of Alzheimer’s disease? Nature. Nov 4, 2020;587(7832):22-25. URL: https://www.nature.com/articles/d41586-020-03084-9 [Accessed 2026-08-27] [CrossRef] [Medline]
    59. Chen VCH, Wu SI, Huang KY, et al. Herpes zoster and Dementia. J Clin Psychiatry. Feb 21, 2018;79(1):16m11312. [CrossRef]
    60. Vestin E, Boström G, Olsson J, et al. Herpes simplex viral infection doubles the risk of dementia in a contemporary cohort of older adults: a prospective study. J Alzheimers Dis. 2024;97(4):1841-1850. [CrossRef] [Medline]
    61. Lopatko Lindman K, Hemmingsson E, Weidung B, et al. Herpesvirus infections, antiviral treatment, and the risk of dementia—a registry‐based cohort study in Sweden. Alzheimers Dement (N Y). Jan 2021;7(1). [CrossRef] [Medline]
    62. Letenneur L, Pérès K, Fleury H, et al. Seropositivity to herpes simplex virus antibodies and risk of Alzheimer’s disease: a population-based cohort study. PLoS One. 2008;3(11):e3637. [CrossRef] [Medline]
    63. Jampani Y, Rana K, Hossain S, et al. HABITAT: evolution of a cloud-based, mesh-enabled research data ecosystem at an academic medical center. medRxiv. Preprint posted online on Jul 4, 2026. [CrossRef]
    64. Software tools. OHDSI. 2025. URL: https://www.ohdsi.org/software-tools [Accessed 2026-08-20]
    65. Hripcsak G, Duke JD, Shah NH, et al. Observational health data sciences and informatics (OHDSI): opportunities for observational researchers. Stud Health Technol Inform. 2015;216:574-578. [Medline]
    66. Schuemie M, Reps J, Black A, et al. Health-analytics data to evidence suite (HADES): open-source software for observational research. Stud Health Technol Inform. Jan 25, 2024;310:966-970. [CrossRef] [Medline]
    67. CohortMethod: new-user cohort method with large scale propensity and outcome models. GitHub. 2024. URL: https://github.com/OHDSI/CohortMethod [Accessed 2026-08-31]
    68. OHDSI/FeatureExtraction. GitHub. 2025. URL: https://github.com/OHDSI/FeatureExtraction [Accessed 2026-08-20]
    69. Schuemie MJ, Hripcsak G, Ryan PB, Madigan D, Suchard MA. Empirical confidence interval calibration for population-level effect estimation studies in observational healthcare data. Proc Natl Acad Sci U S A. Mar 13, 2018;115(11):2571-2577. [CrossRef] [Medline]
    70. Schuemie MJ, Ryan PB, DuMouchel W, Suchard MA, Madigan D. Interpreting observational studies: why empirical calibration is needed to correct p-values. Stat Med. Jan 30, 2014;33(2):209-218. [CrossRef] [Medline]
    71. Lund JL, Richardson DB, Stürmer T. The active comparator, new user study design in pharmacoepidemiology: historical foundations and contemporary application. Curr Epidemiol Rep. Dec 2015;2(4):221-228. [CrossRef] [Medline]
    72. Islam MM. GitHub repository for the study “Association of Herpes Simplex Virus Type-1 With Dementia Outcomes: Longitudinal Retrospective Cohort Study Using Real World EHR Data”. GitHub. URL: https://github.com/mizq7/IDCR [Accessed 2026-09-08]
    73. Alghamdi A, Koen E, Helmantel M, et al. The association between herpes zoster, antiherpetic therapies, and alzheimer’s disease: a comprehensive systematic review. Pharmaceuticals (Basel). May 14, 2025;18(5):722. [CrossRef] [Medline]
    74. Wozniak MA, Mee AP, Itzhaki RF. Herpes simplex virus type 1 DNA is located within Alzheimer’s disease amyloid plaques. J Pathol. Jan 2009;217(1):131-138. [CrossRef] [Medline]
    75. Wozniak MA, Frost AL, Itzhaki RF. Alzheimer’s disease-specific tau phosphorylation is induced by herpes simplex virus type 1. J Alzheimers Dis. 2009;16(2):341-350. [CrossRef] [Medline]
    76. Zambrano A, Solis L, Salvadores N, Cortés M, Lerchundi R, Otth C. Neuronal cytoskeletal dynamic modification and neurodegeneration induced by infection with herpes simplex virus type 1. J Alzheimers Dis. Jul 2008;14(3):259-269. [CrossRef] [Medline]
    77. Goldhardt O, Freiberger R, Dreyer T, et al. Herpes simplex virus alters Alzheimer’s disease biomarkers ‐ a hypothesis paper. Alzheimers Dement. May 2023;19(5):2117-2134. [CrossRef] [Medline]
    78. Piacentini R, Li Puma DD, Ripoli C, et al. Herpes simplex virus type-1 infection induces synaptic dysfunction in cultured cortical neurons via GSK-3 activation and intraneuronal amyloid-β protein accumulation. Sci Rep. Oct 21, 2015;5(1):15444. [CrossRef] [Medline]
    79. Li Puma DD, Colussi C, Bandiera B, et al. Interleukin 1β triggers synaptic and memory deficits in herpes simplex virus type-1-infected mice by downregulating the expression of synaptic plasticity-related genes via the epigenetic MeCP2/HDAC4 complex. Cell Mol Life Sci. Jun 1, 2023;80(6):172. [CrossRef] [Medline]
    80. Schnier C, Janbek J, Williams L, et al. Antiherpetic medication and incident dementia: observational cohort studies in four countries. Eur J Neurol. Jun 2021;28(6):1840-1848. [CrossRef] [Medline]
    81. Duggan MR, Yang S, Gomez GT, et al. Proteomic signatures of corona and herpes viral antibodies identify IGDCC4 as a mediator of neurodegeneration. Sci Adv. May 30, 2025;11(22):eadt7176. [CrossRef] [Medline]

    ‎
    AD: Alzheimer disease
    ADRD: Alzheimer disease and related dementias
    APOE: apolipoprotein E.
    CDM: Common data model
    EHR: electronic health record
    HIPAA: Health Insurance Portability and Accountability Act
    HR: hazard ratio
    HSV-1: herpes simplex virus type-1
    ICD-10-CM: International Classification of Diseases, Tenth Revision, Clinical Modification
    ICD-9-CM: International Classification of Diseases, Ninth Revision, Clinical Modification
    OHDSI: Observational Health Data Sciences and Informatics
    OMOP: Observational Medical Outcomes Partnership
    PS: propensity score
    PY: person-years
    RR: relative risk
    SMD: standardized mean difference

    Edited by Javad Sarvestan; submitted 27.Aug.2025; peer-reviewed by Knoo Lee, Philippe Thompson-Leduc; final revised version received 26.Jun.2026; accepted 30.Jun.2026; published 11.Sep.2026.

    Copyright

    © Muhammad Muinul Islam, Randi Foraker, Md Saber Hossain, W David Arnold, Dima Dandachi, Abu Saleh Mohammad Mosa. Originally published in JMIR Formative Research (https://formative.jmir.org), 11.Sep.2026.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on https://formative.jmir.org, as well as this copyright and license information must be included.