Abstract
Background:
Herpes zoster is a potential risk factor for dementia. The effectiveness of the recombinant zoster vaccine for preventing dementia is uncertain.
Methods:
This retrospective cohort study used de-identified claims data from the Optum Labs Data Warehouse database from January 1, 2017, to December 31, 2022, to determine whether the recombinant zoster vaccine is associated with a reduced risk of dementia. Immunocompetent patients with ≥365 days of continuous enrollment were included, with the risk period starting upon age-eligibility for the recombinant zoster vaccination. Cox regression adjusted for time-fixed and time-updated measures every six months was implemented to estimate hazard ratios for dementia. Herpes zoster diagnosis and antiviral therapy were also assessed.
Results:
There were 4,502,678 individuals (median [IQR] age, 62 [54-71] years; 51% female) included in this study: 206,297 (4.6%) were partially vaccinated, and 460,413 (10.2%) were fully vaccinated. The incidence rate of dementia was 99.1 cases per 10,000 person-years in the fully vaccinated group, 108.2 cases per 10,000 person-years in the partially vaccinated group, and 135.0 cases per 10,000 person-years in the unvaccinated group. After adjustment, vaccination was significantly associated with a decreased risk of dementia for two doses (hazard ratio (HR): 0.68; 95% CI: 0.67-0.70; P<.001) and for one dose (HR 0.89; 95% CI: 0.87-0.92; P<.001). Having a herpes zoster diagnosis before the first vaccination dose was associated with an increased hazard of dementia (HR 1.47; 95% CI: 1.42-1.52; P<.001) compared to those with no diagnosis. Antivirals used to treat zoster infection were protective against dementia (HR 0.42; 95% CI: 0.40-0.44; P<.001).
Conclusions:
These findings suggest that the recombinant zoster vaccine is associated with a decreased risk of dementia and highlight an additional benefit of vaccination beyond preventing herpes zoster.
Keywords: herpes zoster vaccine, herpes zoster, vaccination, antiviral agents, dementia, recombinant zoster vaccine
Introduction
Dementia is a neurocognitive disorder characterized by a decline in the domains of cognition, such as executive function, perceptual-motor skills, or social cognition [1]. This disorder affects over 55 million people globally, with 10 million new cases diagnosed per year [2]. While there are no established treatments to reverse the effects of dementia after onset, some environmental and lifestyle factors may be controlled to reduce the risk of developing dementia. One potential risk factor is herpes zoster (HZ) infection, a neurocutaneous disease caused by the reactivation of the varicella-zoster virus (VZV), commonly known as shingles. In the United States, about one in three people will develop HZ in their lifetime [3].
Multiple studies have shown that HZ may be implicated in dementia risk, perhaps through neuroinflammation or cerebral vasculopathy [4]. Two population-based cohort studies from 2017 and 2020 found that HZ was associated with an increased risk of dementia, while antiviral therapy following HZ infection reduced the risk [5,6]. There is less clarity on whether HZ vaccination lowers the risk of developing dementia. Some prior studies have shown that HZ vaccination is associated with lower rates of dementia [7–12], but these primarily explored the effects of the zoster live vaccine (ZVL; Zostavax), which was discontinued in 2020. The recombinant zoster vaccine (RZV; SHINGRIX), released in late 2017, is now the approved and recommended vaccine for preventing HZ infection by the Centers for Disease Control and Prevention (CDC), with 97% efficacy in clinical trials [13,14].
The aim of our study was to determine the impact of RZV on developing dementia using a large claims database. Specific aims were to determine (1) whether vaccination with RZV has a protective effect on dementia and (2) how HZ infection and antiviral therapy are associated with the risk of developing dementia.
Methods
Data Source
This retrospective cohort study used medical and pharmacy claims data from the Optum Labs Data Warehouse (OLDW; Optum Labs, Eden Prairie, MN), a de-identified administrative claims and electronic health record (EHR) database consisting of enrollment information of commercial and Medicare Advantage enrollees of all ages since 1994 [15]. Medical claims comprised of diagnosis codes, i.e., International Classification of Disease 10th revision (ICD-10), Current Procedural Terminology (CPT) codes, dates of service, and provider specialty codes. Pharmacy claims included information such as the National Drug Code (NDC) of the medication, the name and strength of the medication, the quantity dispensed, the number of refills authorized, and the date of prescription fill.
Study Population
The index date was defined as when a patient met the eligibility criteria for the study, starting on January 1, 2018, when RZV became available. The study endpoint was defined as the date of dementia diagnosis, censoring, or the end of the study period (December 31, 2022). Patients were censored if they were disenrolled from the insurance plan, received the Zostavax (ZVL) vaccination, were diagnosed with an immunocompromising condition, started immunosuppressive medications (excluding steroids), or died.
We included individuals who met the following criteria: (1) age-eligibility for RZV, defined as reaching the age of 50 between 2018 and 2022, or already being over 50 years of age before 2018 and already enrolled in the OLDW database, and (2) having at least 365 days of continuous enrollment in the OLDW database. As exact birthdates were unavailable due to confidentiality constraints, patients’ ages were estimated based on their birth years. For the period spanning January 1 to December 31, 2018, individuals aged 50 and above with at least one year of continuous enrollment were included. From January 1, 2019, to December 31, 2022, only patients who turned 50 during this timeframe and had at least one year of continuous enrollment were included. The process of cohort selection is depicted in Figure 1. Baseline comorbidities were also recorded (eTable 1).
Figure 1.
Cohort selection flow diagram. ICD-10 = International Classification of Diseases 10th Revision; RZV = recombinant zoster vaccine.
Patients were excluded if they were immunocompromised or received immunosuppressive medications within the year before the index date. The conditions constituting immunocompromised status and specific medications are detailed in eMethods 1 and eMethods 2. Individuals with a prior dementia diagnosis during the baseline period were excluded (eMethods 3).
Exposure and Outcome
Receipt of the recombinant zoster vaccine (RZV) was ascertained through the Current Procedural Terminology (CPT) code 90750 or via pharmacy claims using the brand name SHINGRIX. Vaccination status was categorized into three groups: (1) Unvaccinated (up to 29 days after the first dose), (2) partially vaccinated (30 or more days after the first dose or up to 29 days after the second dose), and (3) fully vaccinated (30 or more days after the second dose). A 30-day window allowed for each dose to become effective. The primary outcome of interest was the initial diagnosis of dementia, identified using ICD-10 codes (eMethods 3).
Covariates
Baseline covariates included race/ethnicity, geographical region, and record of ZVL in the year before the index date (eMethods 4). Additionally, antiviral medication filled during the baseline or risk period was treated as a fixed covariate and classified as given in conjunction with an HZ diagnosis (prescription filled within seven days preceding or following an HZ diagnosis) or given for any other medical indication. Receipt of ZVL was identified by the CPT code (90736) or pharmacy claims by the brand name Zostavax. Antiviral medications included valacyclovir, acyclovir, and famciclovir. Categorical variables with missing data were categorized as “Other/Unknown.”
The time-varying covariates RZV vaccination status, age group, age-adjusted Charlson Comorbidity Index (CCI), and HZ infection were updated every six months. The 6-month interval was subdivided accordingly if the vaccination status changed during the period. Individuals contributed to unvaccinated time up to 29 days after the first dose, partially vaccinated time up to 29 days after the second dose, and fully vaccinated time thereafter.
Herpes zoster (HZ) diagnoses were identified using the ICD-10 code B02.xx and captured from any diagnostic position in provider and facility claims across various clinical settings (including inpatient hospital, long-term care, emergency department, outpatient hospital, office visits, and unclassified visits as defined by OLDW). The first diagnosis date between the baseline period and censoring date was recorded.
The status of HZ infection was subsequently classified into four groups: (1) No infection; (2) Infection before any RZV, up to 29 days after the first dose; (3) Infection 30 or more days after the first dose of RZV; and (4) Infection 30 or more days after the second dose of RZV. Patients were categorized under group (2) if they had an infection in the baseline period. After the HZ diagnosis date, infection status remained fixed for subsequent 6-month periods.
Statistical Analyses
Baseline covariates and comorbidities were summarized using descriptive statistics. Incidence rates of dementia were calculated per 10,000 person-years (PY) and stratified by vaccination status, HZ diagnoses, antiviral medications, previous receipt of ZVL, and age groups. We used a time-varying Cox proportional hazards model to estimate both unadjusted and adjusted hazard ratios for dementia, accounting for confounding bias with time-fixed and time-varying covariates. The 95% confidence intervals were estimated using robust standard errors. The number needed to vaccinate (NNV) to prevent a case of dementia was also calculated. To ensure robustness, we also conducted a stratified analysis using a time-varying Cox proportional hazards model by patients’ birth year without adjusting for age.
Statistical analyses were performed in R (Version 4.2.1, R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/). P-values less than 0.05 were considered statistically significant. This study was approved by the Institutional Review Board of the University of California, San Francisco and was conducted in adherence with the tenets of the Declaration of Helsinki.
Results
Table 1 shows the characteristics of the cohort, which consisted of 4,502,678 individuals. Individuals were included if they reached age eligibility for RZV of 50 years or older with at least one year of continuous enrollment before the eligibility date and were excluded if they had dementia or a condition that made them immunocompromised in the baseline period. Cohort selection is outlined in Figure 1. The cohort’s median age was 62 years (interquartile range (IQR): 54-71 years). Approximately 51% of the cohort was female, and 70% were of White race/ethnicity. The median ages at the first and second doses were 67 (IQR: 59-74) and 69 years (IQR: 60-75), respectively. The prevalence of comorbidities was similar across vaccination groups. Approximately 2.3% of the cohort was treated with antivirals in conjunction with an HZ diagnosis, 4.8% received antiviral treatment for other medical condition(s), and 1.3% received ZVL in the baseline period.
Table 1.
Baseline characteristics of the study cohort.
| Recombinant zoster vaccination status (RZV, at time of censor) | Unvaccinateda N=3835968 | Partially vaccinateda N=206297 | Fully vaccinateda N=460413 | Overall N=4502678 |
|---|---|---|---|---|
| Zostavax (ZVL, baseline) | ||||
| No | 3793078 (98.9%) | 201116 (97.5%) | 449154 (97.6%) | 4443348 (98.7%) |
| Yes | 42890 (1.1%) | 5181 (2.5%) | 11259 (2.4%) | 59330 (1.3%) |
| Antivirals taken | ||||
| Non-exposed | 3591406 (93.6%) | 183482 (88.9%) | 407220 (88.4%) | 4182108 (92.9%) |
| Exposed, HZ | 77927 (2.0%) | 7441 (3.6%) | 16983 (3.7%) | 102351 (2.3%) |
| Exposed, Other | 166635 (4.3%) | 15374 (7.5%) | 36210 (7.9%) | 218219 (4.8%) |
| Age (years) at index date | ||||
| Median [Q1, Q3] | 62.0 [53.0, 71.0] | 63.0 [54.0, 71.0] | 67.0 [58.0, 73.0] | 62.0 [54.0, 71.0] |
| Age group at index date | ||||
| 50-54 | 1123442 (29.3%) | 51747 (25.1%) | 75110 (16.3%) | 1250299 (27.8%) |
| 55-59 | 581628 (15.2%) | 30246 (14.7%) | 60316 (13.1%) | 672190 (14.9%) |
| 60-64 | 490122 (12.8%) | 30163 (14.6%) | 61331 (13.3%) | 581616 (12.9%) |
| 65-69 | 509608 (13.3%) | 31884 (15.5%) | 81320 (17.7%) | 622812 (13.8%) |
| 70-74 | 463397 (12.1%) | 29671 (14.4%) | 88550 (19.2%) | 581618 (12.9%) |
| 75-79 | 305196 (8.0%) | 18044 (8.7%) | 53839 (11.7%) | 377079 (8.4%) |
| 80-84 | 362575 (9.5%) | 14542 (7.0%) | 39947 (8.7%) | 417064 (9.3%) |
| Sex | ||||
| Female | 1916047 (49.9%) | 108351 (52.5%) | 253672 (55.1%) | 2278070 (50.6%) |
| Male | 1919105 (50.0%) | 97906 (47.5%) | 206657 (44.9%) | 2223668 (49.4%) |
| Other/Unknown | 816 (0.02%) | 40 (0.02%) | 84 (0.02%) | 940 (0.02%) |
| Race/ethnicity | ||||
| White | 2615255 (68.2%) | 149019 (72.2%) | 363148 (78.9%) | 3127422 (69.5%) |
| Asian | 149189 (3.9%) | 10483 (5.1%) | 21482 (4.7%) | 181154 (4.0%) |
| Black | 463360 (12.1%) | 20573 (10.0%) | 33241 (7.2%) | 517174 (11.5%) |
| Hispanic | 455527 (11.9%) | 19397 (9.4%) | 29029 (6.3%) | 503953 (11.2%) |
| Other/Unknown | 152637 (4.0%) | 6825 (3.3%) | 13513 (2.9%) | 172975 (3.8%) |
| Geographic region | ||||
| South | 1747252 (45.5%) | 87655 (42.5%) | 177835 (38.6%) | 2012742 (44.7%) |
| Midwest | 947174 (24.7%) | 59260 (28.7%) | 154089 (33.5%) | 1160523 (25.8%) |
| Northeast | 485124 (12.6%) | 23330 (11.3%) | 50403 (10.9%) | 558857 (12.4%) |
| West | 569789 (14.9%) | 35835 (17.4%) | 77881 (16.9%) | 683505 (15.2%) |
| Other/Unknown | 86629 (2.3%) | 217 (0.1%) | 205 (0.04%) | 87051 (1.9%) |
| Charlson comorbidity index (CCI, baseline) | ||||
| Median [Q1, Q3] | 2.00 [1.00, 3.00] | 2.00 [1.00, 3.00] | 2.00 [1.00, 3.00] | 2.00 [1.00, 3.00] |
| Comorbidities | ||||
| Alcohol or substance abuse | 89085 (2.3%) | 4276 (2.1%) | 6864 (1.5%) | 100225 (2.2%) |
| Asthma | 140647 (3.7%) | 9503 (4.6%) | 20431 (4.4%) | 170581 (3.8%) |
| Cerebrovascular disease | 199282 (5.2%) | 9946 (4.8%) | 22766 (4.9%) | 231994 (5.2%) |
| Chronic kidney disease | 297537 (7.8%) | 15128 (7.3%) | 34190 (7.4%) | 346855 (7.7%) |
| Chronic liver disease | 79388 (2.1%) | 4771 (2.3%) | 9590 (2.1%) | 93749 (2.1%) |
| Chronic lung disease | 230539 (6.0%) | 10540 (5.1%) | 22220 (4.8%) | 263299 (5.8%) |
| Type II Diabetes Mellitus | 763654 (19.9%) | 42270 (20.5%) | 87092 (18.9%) | 893016 (19.8%) |
| Heart disease | 573959 (15.0%) | 30054 (14.6%) | 70939 (15.4%) | 674952 (15.0%) |
| Hypertension | 1716233 (44.7%) | 98841 (47.9%) | 231230 (50.2%) | 2046304 (45.4%) |
| Anxiety or Depression | 554205 (14.4%) | 34463 (16.7%) | 66946 (14.5%) | 655614 (14.6%) |
| Parkinson’s Disease | 17702 (0.5%) | 762 (0.4%) | 1698 (0.4%) | 20162 (0.4%) |
| Obesity | 790509 (20.6%) | 46460 (22.5%) | 99036 (21.5%) | 936005 (20.8%) |
| Smoking (baseline or risk period) | ||||
| Never | 496721 (12.9%) | 35806 (17.4%) | 83928 (18.2%) | 616455 (13.7%) |
| Current/Former | 255591 (6.7%) | 14116 (6.8%) | 26862 (5.8%) | 296569 (6.6%) |
| Unknown | 3083656 (80.4%) | 156375 (75.8%) | 349623 (75.9%) | 3589654 (79.7%) |
| Number of comorbidities | ||||
| 0 | 1444058 (37.6%) | 65565 (31.8%) | 139516 (30.3%) | 1649139 (36.6%) |
| 1 | 793365 (20.7%) | 49825 (24.2%) | 120697 (26.2%) | 963887 (21.4%) |
| 2-3 | 1135642 (29.6%) | 67102 (32.5%) | 153189 (33.3%) | 1355933 (30.1%) |
| 4+ | 462903 (12.1%) | 23805 (11.5%) | 47011 (10.2%) | 533719 (11.9%) |
Values are reported as N (%) unless otherwise indicated. The index date was defined as the date an individual was eligible for study inclusion. Comorbidities included anxiety/depression, alcohol/drug dependence, asthma, cerebrovascular diseases, chronic kidney diseases, chronic lung diseases, chronic liver diseases, diabetes mellitus II, heart conditions, hypertension, Parkinson’s disease, obesity, and smoking status.
Abbreviations: CCI, Charlson Comorbidity Index; HZ, herpes zoster; Q1, first quartile; Q3, third quartile; RZV, recombinant zoster vaccine (Shingrix); ZVL, Zostavax.
“Unvaccinated”: No vaccination or within 29 days of 1st dose; “Partially vaccinated”: 30+ days after 1st dose or within 29 days of 2nd dose); “Fully vaccinated”: 30+ days after 2nd dose.
At the time of censoring, 3,835,968 (85.2%) individuals had no RZV vaccinations or had less than 29 days after the first dose of RZV with a median follow-up time of 24 months (IQR: 12-50), 206,297 (4.6%) individuals were partially vaccinated with 30 or more days after one dose only with a median follow-up time of 48 months (IQR: 27-60), and 460,413 (10.2%) individuals had 30 or more days after the second dose of RZV with a median follow-up time of 60 months (IQR: 37-60). Among those who were partially vaccinated by the end of the study, 8438 (4.1%) developed HZ before vaccination, and 1136 (0.6%) developed HZ after. Of the fully vaccinated individuals, 19,175 (4.2%) developed HZ before any vaccination, 411 (0.1%) developed HZ after the first vaccination, and 2137 (0.5%) developed HZ after the second vaccination.
There were 161,308 dementia cases with 12,247,881 person-years (PY), corresponding to an incidence rate (IR) of 131.7 cases per 10,000 PY. In the unvaccinated group, there were 144,867 cases with a follow-up time of 11,028,795 PY (IR: 135.0 cases per 10,000 PY), 4244 cases with 392,353 PY in the partially vaccinated group (IR: 108.2 cases per 10,000 PY), and 8197 cases with 826,733 PY in the fully vaccinated group (IR: 99.1 cases per 10,000 PY) (Table 2). Additionally, 133,887 individuals had an HZ diagnosis between the baseline period and end date, 5470 of whom developed dementia. Specifically, there were 5366 (98.1%) cases of dementia in those who had an HZ diagnosis before any RZV dose, 43 (0.8%) cases in those who had an HZ diagnosis while partially vaccinated, and 61 (1.1%) cases in those who were fully vaccinated with RZV. A detailed breakdown of HZ cases is provided in eTable 2.
Table 2.
Incidence of dementia by RZV vaccination status, HZ diagnoses, antivirals, previous ZVL, and age groups.
| No. of dementia cases | Person-years (PY) | Dementia rate / 10000PY | |
|---|---|---|---|
| Overall | 161308 | 12247881 | 131.7 |
| Recombinant zoster vaccine (RZV) status | |||
| Unvaccinateda | 148867 | 11028795 | 135.0 |
| Partially vaccinateda | 4244 | 392353 | 108.2 |
| Fully vaccinateda | 8197 | 826733 | 99.1 |
| Herpes zoster diagnosis (HZ dx) | |||
| No HZ dx | 155838 | 11930881 | 130.6 |
| HZ dx, Prior to any RZVb | 5366 | 311134 | 172.5 |
| HZ dx, Post 1st dose RZVb | 43 | 2579 | 166.7 |
| HZ dx, Post 2nd dose RZVb | 61 | 3287 | 185.6 |
| Antivirals | |||
| No Antivirals | 153858 | 11236665 | 136.9 |
| Antivirals, HZ related | 3707 | 362391 | 102.3 |
| Antivirals, other condition | 3743 | 648825 | 57.7 |
| Baseline Zostavax (ZVL) | |||
| No baseline ZVL | 159072 | 12081544 | 131.7 |
| Baseline ZVL | 2236 | 166337 | 134.4 |
| Age group (years, time-varying) | |||
| 50-54 | 1365 | 2349486 | 5.8 |
| 55-59 | 2121 | 1801812 | 11.8 |
| 60-64 | 4070 | 1645170 | 24.7 |
| 65-69 | 8796 | 1510032 | 58.3 |
| 70-74 | 21763 | 2056729 | 105.8 |
| 75-79 | 30941 | 1418808 | 218.1 |
| 80-84 | 53174 | 979371 | 542.9 |
| 85-89 | 39078 | 486474 | 803.3 |
Abbreviations: HZ dx, herpes zoster diagnosis; RZV, recombinant zoster vaccine (Shingrix); ZVL, Zostavax.
“Unvaccinated”: No vaccination or within 29 days of 1st dose; “Partially vaccinated”: 30+ days after 1st dose or within 29 days of 2nd dose); “Fully vaccinated”: 30+ days after 2nd dose.
“Prior to any RZV”: An HZ diagnosis prior to before 1st dose of RZV or within 29 days of it; “Post 1st dose RZV”: An HZ diagnosis after 30 or more days after the 1st dose of RZV or within 29 days of 2nd dose; “Post 2nd dose RZV”: An HZ diagnosis after 30 or more or more days after the 2nd dose of RZV.
Having one dose of RZV (≥30 days post-dose) was associated with a lower risk of dementia (unadjusted hazard ratio (HR) 0.75; 95% confidence interval (CI): 0.73-0.77; P<.001), while for two doses of RZV (≥30 days post-dose), the HR was 0.66 (95% CI: 0.65-0.68; P<.001). After adjustment for time-updated age groups, Charlson Comorbidity Index (CCI), RZV status, HZ diagnosis, antivirals, baseline Zostavax (ZVL), and demographics, one and two doses of RZV vaccination remained significantly associated with a reduced risk of developing dementia compared to individuals who did not receive any vaccination (HR: 0.89; 95% CI: 0.87-0.92; P<.001 and HR: 0.68; 95% CI: 0.67-0.70; P<.001, respectively) (Figure 2). The number needed to fully vaccinate with RZV to prevent one case of dementia was 45 patients.
Figure 2.
Adjusted HRs of RZV status, HZ diagnosis, antivirals, baseline Zostavax, and age on dementia. The model was adjusted for gender, race/ethnicity, geographic region, and the Charlson Comorbidity Index (CCI). RZV vaccination status was defined as: “Unvaccinated” = No vaccination or within 29 days of 1st dose; “Partially vaccinated” = 30+ days after 1st dose or within 29 days of 2nd dose); “Fully vaccinated” = 30+ days after 2nd dose. HZ diagnosis was defined based on vaccination status. CI = confidence interval; HZ = herpes zoster; RZV = recombinant zoster vaccination; ZVL = Zostavax.
Furthermore, HZ diagnosis, regardless of receipt or timing of RZV, was significantly associated with dementia after adjustment. Having an HZ diagnosis before the first dose of RZV or up to ≥30 days post-dose was associated with an HR of 1.47 (95% CI: 1.42-1.52; P<.001). An HZ diagnosis after the first or second dose also increased risk (HR: 1.41; 95% CI: 1.04-1.90; P=.03 and HR: 1.84; 95% CI: 1.43-2.37; P<.001, respectively). Additionally, antivirals were found to be protective, whether they were dispensed for an HZ diagnosis or other medical condition (HR: 0.42; 95% CI: 0.40-0.44; P<.001 and HR: 0.59; 95% CI: 0.57-0.61; P<.001).
The adjusted HRs for dementia increased with age groups, and the HR for males was lower than for females (0.72; 95% CI: 0.71-0.73; P<.001). Compared to White individuals, those identifying as Black or Hispanic showed slightly higher risks of dementia. These findings are summarized in Table 3. Hazard ratio estimates remained consistent with our primary analysis after stratification by birth year (eTable 3).
Table 3.
Unadjusted and adjusted hazard ratios showing associations between Recombinant Zoster Vaccine (RZV) and dementia.
| Unadjusted HR (95% CI) | P-valuea | Adjusted HR (95% CI) | P-valuea | |
|---|---|---|---|---|
| Partially vaccinatedb (ref: unvaccinated) | 0.75 (0.73, 0.77) | <.001 | 0.89 (0.87, 0.92) | <.001 |
| Fully vaccinated b | 0.66 (0.65, 0.68) | <.001 | 0.68 (0.67, 0.70) | <.001 |
| Herpes zoster diagnosis (HZ dx), Prior to any RZVc (ref: no HZ dx) | 1.28 (1.25, 1.32) | <.001 | 1.47 (1.42, 1.52) | <.001 |
| HZ dx, Post 1st dose RZV c | 1.18 (0.88, 1.60 | .27 | 1.41 (1.04, 1.90) | .03 |
| HZ dx, Post 2nd dose RZV c | 1.29 (1.002, 1.65) | .049 | 1.84 (1.43, 2.37) | <.001 |
| Antivirals (baseline or risk period, fixed), HZ related (ref: none) | 0.74 (0.72, 0.76) | <.001 | 0.42 (0.40, 0.44) | <.001 |
| Antivirals (baseline or risk period, fixed), other | 0.42 (0.41, 0.43) | <.001 | 0.59 (0.57, 0.61) | <.001 |
| Zostavax vaccination (baseline) | 1.02 (0.98, 1.06) | .35 | 0.88 (0.85, 0.92) | <.001 |
| Age Group: 55-59 (ref: 50-54) | 2.16 (2.02, 2.32) | <.001 | 1.96 (1.83, 2.10) | <.001 |
| Age Group: 60-64 | 4.59 (4.31, 4.88) | <.001 | 2.27 (2.13, 2.41) | <.001 |
| Age Group: 65-69 | 10.49 (9.91, 11.11) | <.001 | 4.31 (4.07, 4.56) | <.001 |
| Age Group: 70-74 | 20.13 (19.06, 21.27) | <.001 | 4.64 (4.39, 4.91) | <.001 |
| Age Group: 75-79 | 41.99 (39.77, 44.33) | <.001 | 8.43 (7.98, 8.90) | <.001 |
| Age Group: 80-84 | 98.65 (93.49, 104.1) | <.001 | 10.90 (10.33, 11.51) | <.001 |
| Age Group: 85-89 | 173.0 (164.8, 182.7) | <.001 | 17.60 (16.66, 18.60) | <.001 |
| Charlson Comorbidity Index (TV) | 1.702 (1.699, 1.705) | <.001 | 1.601 (1.599, 1.603) | <.001 |
| Gender Male (ref: Female) | 0.73 (0.72, 0.74) | <.001 | 0.72 (0.71, 0.73) | <.001 |
| Gender Unknown | 1.28 (0.94, 1.73) | .11 | 0.99 (0.73, 1.34) | .96 |
| Race/Ethnicity Asian (ref: White) | 0.66 (0.64, 0.68) | <.001 | 0.85 (0.82, 0.87) | <.001 |
| Race/Ethnicity Black | 1.41 (1.39, 1.43) | <.001 | 1.08 (1.06, 1.09) | <.001 |
| Race/Ethnicity Hispanic | 1.04 (1.03, 1.06) | <.001 | 1.05 (1.03, 1.07) | <.001 |
| Race/Ethnicity Unknown/Missing | 0.90 (0.87, 0.92) | <.001 | 1.08 (1.05, 1.11) | <.001 |
| Region Midwest (ref: South) | 0.85 (0.84, 0.86) | <.001 | 0.88 (0.86, 0.88) | <.001 |
| Region Northeast | 1.21 (1.20, 1.23) | <.001 | 1.03 (1.02, 1.05) | <.001 |
| Region West | 0.65 (0.64, 0.66) | <.001 | 0.92 (0.91, 0.94) | <.001 |
| Region Other/Unknown | 0.02 (0.015, 0.025) | <.001 | 0.076 (0.06, 0.098) | <.001 |
Abbreviations: HR, hazard ratio; HZ dx, herpes zoster diagnosis; RZV, recombinant zoster vaccine (Shingrix); ref, reference; TV, time-varying; ZVL, Zostavax.
P-values are calculated from Cox proportional hazards models.
“Unvaccinated”: No vaccination or within 29 days of 1st dose; “Partially vaccinated”: 30+ days after 1st dose or within 29 days of 2nd dose); “Fully vaccinated”: 30+ days after 2nd dose.
“Prior to any RZV”: An HZ diagnosis prior to before 1st dose of RZV or within 29 days of it; “Post 1st dose RZV”: An HZ diagnosis after 30 or more days after the 1st dose of RZV or within 29 days of 2nd dose; “Post 2nd dose RZV”: An HZ diagnosis after 30 or more or more days after the 2nd dose of RZV.
Discussion
In this population-based cohort study, we found that RZV was associated with a significantly decreased risk of dementia. Full vaccination with RZV conferred the most protection against dementia (32% decrease in risk; 95% CI: 30-33%; P<.001) compared to those classified as unvaccinated), but even partial vaccination was associated with an 11% decrease in risk (95% CI: 8-13%; P<.001) compared to being unvaccinated. The relatively high percentage of patients who received only one dose may be partly attributable to censoring, as some patients may not have completed the two-dose series before the end of the study period. This does not reflect the overall vaccination rate within the cohort, as some individuals could have received their second dose after being censored. HZ increased the risk of dementia regardless of the timing of infection with respect to RZV. Antiviral treatment was also a protective factor against dementia, especially if it was dispensed for an HZ infection. Antiviral agents may still have protective effects against zoster when used as treatment or as prophylaxis for other conditions, including herpes simplex virus infections or to prevent reactivation of VZV [16,17]. Future work could investigate the dose-dependent relationship between antivirals and dementia risk.
Our findings are consistent with previous research investigating the impact of the former live attenuated shingles vaccine (ZVL) on dementia risk. In a nested case-control study using the UK Biobank, Lophatanano et al. reported a significant decrease in dementia risk among individuals vaccinated with ZVL (odds ratio [OR]: 0.81; 95% CI: 0.66-0.99). While they considered HZ infection occurring more than three years prior to dementia diagnosis, they did not adjust for antiviral use [11]. Similarly, Schnier et al. found a reduced risk of dementia among ZVL recipients (adjusted hazard ratio [aHR]: 0.72; 95% CI: 0.69-0.75) after accounting for factors such as care home residency and comorbidities [18]. Notably, the association between ZVL and dementia was consistent regardless of HZ diagnosis, but antiviral use was not considered. Additionally, a study in preprint utilized a regression discontinuity design to evaluate the effect of ZVL on dementia risk over seven years. They observed a significant relative reduction of 19.9% in the probability of new dementia diagnoses among recipients of ZVL. However, that study did not account for HZ infections or antiviral treatments [19]. Our study also aligns with results from Harris et al.’s study, which looked at routine vaccinations from 2009 to 2019 and the occurrence of Alzheimer’s disease and related dementia risk and found greater protection from RZV compared to ZVL [12]. A recent study by Taquet et al. examined the association between recombinant zoster vaccine and dementia risk, with findings consistent with ours [20]. While their robust methodology offers valuable insights, our study complements their work by additionally examining the role of antiviral treatment following HZ infection.
There are several proposed mechanisms for the association between HZ and dementia. The causative agent of HZ is the varicella-zoster virus (VZV), a neurotropic virus known to establish lifelong latency in sensory ganglia following primary infection. One proposed mechanism involves the activation of innate immune cells in response to VZV reactivation. In the context of HZ reactivation, studies suggest that this response may trigger the release of pro-inflammatory cytokines and chemokines and play a pivotal role in neuroinflammation and neuronal damage, both of which are involved in the pathogenesis of dementia [4,21–23].
Additionally, VZV has been observed to directly infect neurons and glial cells, which can impair immune cells to effectively clear viral particles and mitigate neuroinflammation, contributing to the progressive neuronal loss characteristic of dementia [24,25]. Moreover, the literature suggests that VZV may induce vasculopathy, which can exacerbate neurodegeneration and cognitive decline through compromised cerebral blood flow, further linking HZ to the pathophysiology of dementia [18,26,27].
Vaccination with RZV may have indirect effects that are associated with decreased risk of dementia through inhibiting pro-inflammatory cytokines, such as interferon-gamma (IFN-γ), which are known to synergize with amyloid-beta (Aβ), a protein fragment derived from a larger protein called amyloid precursor protein (APP). This combination leads to the release of tumor necrosis factor-alpha (TNF-α) and reactive nitrogen species, which exhibit neurotoxicity. By reducing the levels of pro-inflammatory cytokines, vaccination could mitigate neuroinflammation triggered by the innate immune response, thereby reducing autophagy, apoptosis, and neuronal cell damage [24,28].
One of the strengths of this study is the large sample size, allowing high statistical power and the capability to adjust for potential confounders. Additionally, the time-varying analysis made it possible to capture evolving risk factors for dementia. We also controlled for cohort-specific effects that may influence vaccination efficacy and dementia risk, which yielded consistent estimates with the primary analysis, reinforcing the robustness of our findings. By stratifying by birth year, we accounted for variations in factors such as environmental exposures and healthcare access that potentially differ across ages. These findings highlight the need for a larger, extended randomized controlled trial evaluating the impact of RZV on outcomes such as dementia or comparing RZV with the live attenuated zoster vaccine (zoster vaccine live) to further investigate their respective effects on dementia risk; however, there are ethical concerns about withholding vaccinations, and zoster vaccine live is no longer available for use in the United States.
There are several limitations to this retrospective study, including outcome ascertainment using ICD-10 codes, which may have resulted in missing or misclassified patients with a diagnosis of herpes zoster or dementia. Selection bias and immortal time bias may also be present, as those who get vaccinated or who are treated for zoster may be more likely to have a dementia diagnosis captured. Additionally, differences in the vaccination groups could remain due to healthy vaccinee bias, even after adjusting for potential confounders. The maximum follow-up time in this study was five years, so the potential benefit of RZV on dementia may not have been captured yet in younger people. However, since we observed a robust protective effect with vaccination in a short follow-up period, results would be expected to be more pronounced over time.
Our study emphasizes the benefits of RZV beyond just preventing HZ, highlighting its additional potential protection against dementia. Despite the effectiveness of RZV in preventing HZ, full vaccination coverage rates remain suboptimal at 15.4% in adults 50 and older [29]. With Medicare Part D and most commercial insurances now fully covering RZV and removing financial barriers for many patients [30,31], highlighting these additional benefits may help promote higher vaccination rates among eligible adults.
Conclusion
In conclusion, this study from a large claims database found that RZV is associated with a lower risk of dementia, with the greatest protection in fully vaccinated individuals. Patients who had HZ had an increased risk of dementia, while those receiving antiviral treatment exhibited a lower risk, irrespective of the reason for receiving antivirals. These findings underscore the importance of promoting RZV as a preventive measure against dementia, regardless of prior zoster infection status.
Supplementary Material
Highlights.
Retrospective cohort study of recombinant zoster vaccine (RZV) on dementia risk
Use of medical and pharmacy claims in the Optum Labs Data Warehouse
Herpes zoster infection is a significant risk factor for dementia
Fully vaccinated individuals had a 32% reduced risk of dementia
Supports policy decisions to increase RZV vaccination rates for older adults
Funding/Support:
This study was funded by grant R01 EY028739 from the National Eye Institute and the Office of Research on Women’s Health at the National Institutes of Health (Dr. Acharya). The Department of Ophthalmology at the University of California, San Francisco, is supported by core grant EY06190 from the National Eye Institute and an unrestricted grant from the Research to Prevent Blindness Foundation.
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.
Footnotes
Conflict of Interest Disclosures: Dr. Acharya: AbbVie (drug donation for NIH-funded trial), Roche (consultant).
References
- 1.Emmady PD, Schoo C, Tadi P. Major Neurocognitive Disorder (Dementia). In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2024. [cited 2024 May 13]. Available from: http://www.ncbi.nlm.nih.gov/books/NBK557444/ [Google Scholar]
- 2.Dementia [Internet]. [cited 2024 May 13]. Available from: https://www.who.int/news-room/fact-sheets/detail/dementia
- 3.Shingles (Herpes Zoster) | CDC [Internet]. [cited 2024 May 13]. Available from: https://www.cdc.gov/shingles/index.html
- 4.Schmidt SAJ, Veres K, Sørensen HT, Obel N, Henderson VW. Incident Herpes Zoster and Risk of Dementia. Neurology [Internet]. 2022. Aug 16 [cited 2023 Oct 9];99(7):e660–8. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9484607/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen VCH, Wu SI, Huang KY, Yang YH, Kuo TY, Liang HY, Huang KL, Gossop M. Herpes Zoster and Dementia: A Nationwide Population-Based Cohort Study. J Clin Psychiatry. 2018;79(1):16m11312. [DOI] [PubMed] [Google Scholar]
- 6.Bae S, Yun SC, Kim MC, Yoon W, Lim JS, Lee SO, Choi SH, Kim YS, Woo JH, Kim SY, Kim SH. Association of herpes zoster with dementia and effect of antiviral therapy on dementia: a population-based cohort study. Eur Arch Psychiatry Clin Neurosci [Internet]. 2021. Aug 1 [cited 2024 Apr 26];271(5):987–97. Available from: 10.1007/s00406-020-01157-4 [DOI] [PubMed] [Google Scholar]
- 7.Wiemken TL, Salas J, Morley JE, Hoft DF, Jacobs C, Scherrer JF. Comparison of rates of dementia among older adult recipients of two, one, or no vaccinations. J Am Geriatr Soc [Internet]. 2022. Apr [cited 2023 Aug 3];70(4):1157. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9300193/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lehrer S, Rheinstein PH. Herpes Zoster Vaccination Reduces Risk of Dementia. In Vivo [Internet]. 2021. Nov 3 [cited 2023 Oct 9];35(6):3271–5. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8627719/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Scherrer JF, Salas J, Wiemken TL, Hoft DF, Jacobs C, Morley JE. Impact of herpes zoster vaccination on incident dementia: A retrospective study in two patient cohorts. PLOS ONE [Internet]. 2021. Nov 17 [cited 2023 Nov 8];16(11):e0257405. Available from: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0257405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Greenblatt CL, Lathe R. Vaccines and Dementia: Part II. Efficacy of BCG and Other Vaccines Against Dementia. J Alzheimers Dis [Internet]. 2024. Jan 1 [cited 2024 Mar 25];98(2):361–72. Available from: https://content.iospress.com/articles/journal-of-alzheimers-disease/jad231323 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lophatananon A, Mekli K, Cant R, Burns A, Dobson C, Itzhaki R, Muir K. Shingles, Zostavax vaccination and risk of developing dementia: a nested case–control study—results from the UK Biobank cohort. BMJ Open [Internet]. 2021. Oct 8 [cited 2023 Oct 9];11(10):e045871. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8504358/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harris K, Ling Y, Bukhbinder AS, Chen L, Phelps KN, Cruz G, Thomas J, Kim Y, Jiang X, Schulz PE. The Impact of Routine Vaccinations on Alzheimer’s Disease Risk in Persons 65 Years and Older: A Claims-Based Cohort Study using Propensity Score Matching. J Alzheimers Dis [Internet]. [cited 2024 May 15];95(2):703–18. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10578243/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Harbecke R, Cohen JI, Oxman MN. Herpes Zoster Vaccines. J Infect Dis [Internet]. 2021. Sep 30 [cited 2024 May 13];224(Suppl 4):S429–42. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8482024/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Himal Lal, Anthony L Cunningham, Olivier Godeaux, Roman Chlibek, Javier Diez-Domingo, Shinn-Jang Hwang, Myron J Levin, Janet E McElhaney, Airi Poder, Joan Puig-Barberà, Timo Vesikari, Daisuke Watanabe, Lily Weckx, Toufik Zahaf, Thomas C Heineman. Efficacy of an Adjuvanted Herpes Zoster Subunit Vaccine in Older Adults. N Engl J Med [Internet]. 2015. [cited 2024 Jun 24];372(22):2087–96. Available from: https://www.nejm.org/doi/full/10.1056/NEJMoa1501184 [DOI] [PubMed] [Google Scholar]
- 15.Optum Labs. Optum Labs and Optum Labs Data Warehouse (OLDW) Descriptions and Citation. Eden Prairie, MN: Optum Labs; 2024. Available at: https://www.optumlabs.com. Accessed August 30, 2024. [Google Scholar]
- 16.Seo HM, Kim YS, Bang CH, Lee JH, Lee JY, Lee DG, Park YM. Antiviral prophylaxis for preventing herpes zoster in hematopoietic stem cell transplant recipients: A systematic review and meta-analysis. Antiviral Res [Internet]. 2017. Apr 1 [cited 2024 Dec 9];140:106–15. Available from: https://www.sciencedirect.com/science/article/pii/S0166354216307458 [DOI] [PubMed] [Google Scholar]
- 17.Barnabas RV, Baeten JM, Lingappa JR, Thomas KK, Hughes JP, Mugo NR, Delany-Moretlwe S, Gray G, Rees H, Mujugira A, Ronald A, Stevens W, Kapiga S, Wald A, Celum C, for the Partners in Prevention HSV/HIV Transmission Study Team. Acyclovir Prophylaxis Reduces the Incidence of Herpes Zoster Among HIV-Infected Individuals: Results of a Randomized Clinical Trial. J Infect Dis [Internet]. 2016. Feb 15 [cited 2024 Dec 9];213(4):551–5. Available from: 10.1093/infdis/jiv318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Schnier C, Janbek J, Lathe R, Haas J. Reduced dementia incidence after varicella zoster vaccination in Wales 2013–2020. Alzheimers Dement Transl Res Clin Interv [Internet]. 2022. Apr 13 [cited 2024 May 13];8(1):e12293. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9006884/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Eyting M, Xie M, Heß S, Geldsetzer P. Causal evidence that herpes zoster vaccination prevents a proportion of dementia cases. medRxiv [Internet]. 2023. May 25 [cited 2024 May 3];2023.05.23.23290253. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10246135/
- 20.Taquet M, Dercon Q, Todd JA, Harrison PJ. The recombinant shingles vaccine is associated with lower risk of dementia. Nat Med [Internet]. 2024. Oct [cited 2024 Nov 27];30(10):2777–81. Available from: https://www.nature.com/articles/s41591-024-03201-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Gilden D, Nagel MA, Cohrs RJ, Mahalingam R. The Variegate Neurological Manifestations of Varicella Zoster Virus Infection. Curr Neurol Neurosci Rep [Internet]. 2013. Sep [cited 2024 May 1];13(9):374. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4051361/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gowrishankar K, Steain M, Cunningham AL, Rodriguez M, Blumbergs P, Slobedman B, Abendroth A. Characterization of the Host Immune Response in Human Ganglia after Herpes Zoster. J Virol [Internet]. 2010. Sep [cited 2024 May 1];84(17):8861–70. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2919016/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Nagel MA, Gilden D. Neurological Complications of VZV Reactivation. Curr Opin Neurol [Internet]. 2014. Jun [cited 2024 Nov 27];27(3):356–60. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4189810/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Sochocka M, Zwolińska K, Leszek J. The Infectious Etiology of Alzheimer’s Disease. Curr Neuropharmacol [Internet]. 2017. Oct [cited 2024 Apr 29];15(7):996–1009. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5652018/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Reichelt M, Zerboni L, Arvin AM. Mechanisms of Varicella-Zoster Virus Neuropathogenesis in Human Dorsal Root Ganglia. J Virol [Internet]. 2008. Apr [cited 2024 May 1];82(8):3971–83. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2292995/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gerada C, Campbell TM, Kennedy JJ, McSharry BP, Steain M, Slobedman B, Abendroth A. Manipulation of the Innate Immune Response by Varicella Zoster Virus. Front Immunol [Internet]. 2020. Jan 24 [cited 2024 May 1];11:1. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6992605/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Feller L, Khammissa RAG, Fourie J, Bouckaert M, Lemmer J. Postherpetic Neuralgia and Trigeminal Neuralgia. Pain Res Treat [Internet]. 2017. [cited 2024 May 1];2017:1681765. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5735631/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang WY, Tan MS, Yu JT, Tan L. Role of pro-inflammatory cytokines released from microglia in Alzheimer’s disease. Ann Transl Med [Internet]. 2015. Jun [cited 2024 May 2];3(10):136. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4486922/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Vaccination Coverage among Adults in the United States, National Health Interview Survey, 2021 | CDC [Internet]. 2023. [cited 2024 Jun 25]. Available from: https://www.cdc.gov/vaccines/imz-managers/coverage/adultvaxview/pubs-resources/vaccination-coverage-adults-2021.html
- 30.Qato DM, Romley JA, Myerson R, Goldman D, Fendrick AM. Shingles Vaccination in Medicare Part D After Inflation Reduction Act Elimination of Cost Sharing. JAMA [Internet]. 2024. Jun 18 [cited 2024 Jul 8];331(23):2043–5. Available from: 10.1001/jama.2024.7348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cost & Coverage | SHINGRIX (Zoster Vaccine Recombinant, Adjuvanted) [Internet]. [cited 2024 Jul 8]. Available from: https://www.shingrix.com/get-shingrix/shingles-vaccine-cost-coverage/ [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.