The effect of shingles vaccination at different stages of the dementia disease course

Abstract

Using natural experiments, we have previously reported that live-attenuated herpes zoster (HZ) vaccination appears to have prevented or delayed dementia diagnoses in both Wales and Australia. Here, we find that HZ vaccination also reduces mild cognitive impairment diagnoses and, among patients living with dementia, deaths due to dementia. Exploratory analyses suggest that the effects are not driven by a specific dementia type. Our approach takes advantage of the fact that individuals who had their eightieth birthday just after the start date of the HZ vaccination program in Wales were eligible for the vaccine for 1 year, whereas those who had their eightieth birthday just before were ineligible and remained ineligible for life. The key strength of our natural experiments is that these comparison groups should be similar in all characteristics except for a minute difference in age. Our findings suggest that live-attenuated HZ vaccination prevents or delays mild cognitive impairment and dementia and slows the disease course among those already living with dementia.

Keywords: Alzheimer’s disease; dementia; herpes zoster; immunization; mild cognitive impairment; natural experiment; quasi-experiment; vaccination; varicella zoster virus.

Figure 1.. The abrupt change in the probability of receiving the HZ vaccination at the September 2, 1933, date-of-birth eligibility threshold

“Baseline” refers to the start date of the HZ vaccination program (i.e., September 1, 2013). Linear regression lines were drawn in the mean squared error-optimal bandwidth only. Gray dots show the mean value for each 10-week increment in week of birth. The gray shading of the dots is in proportion to the weight that observations from this 10-week increment received in the analysis. Abbreviations are as follows: HZ, herpes zoster; Coef., coefficient; CI, confidence interval; Sept, September. See also Figure S10.

Figure 2.. The effect of HZ vaccination on new diagnoses of MCI and deaths due to dementia

Dots show the point estimate and horizontal bars the 95% CI. New diagnoses of MCI were analyzed among a study cohort of patients who did not have any record of cognitive impairment prior to the start date of the HZ vaccination program. Deaths due to dementia were analyzed among a study cohort of patients who had received a diagnosis of dementia prior to the start date of the HZ vaccination program. Abbreviations are as follows: MCI, mild cognitive impairment; CI, confidence interval. See also Figures S2–S4.

Figure 3.. No significant effect of being eligible for HZ vaccination on the leading causes of morbidity and mortality (other than dementia) nor on indicators of preventive health services uptake

Baseline refers to the start date (September 1, 2013) of the HZ vaccination program. Dots show the point estimate and horizontal bars the 95% CI. Among patients without a record of cognitive impairment prior to the start date of the HZ vaccination program, the leading causes of morbidity and mortality were the ten (other than dementia) leading causes of DALYs and mortality among adults aged 70+ years in Wales, as estimated by the Global Burden of Disease Project. Among patients with a diagnosis of dementia made prior to the start date of the HZ vaccination program, the leading causes of mortality were the ten (other than dementia) leading causes of mortality among adults aged 70+ years in Wales, as estimated by the Global Burden of Disease Project. Among patients without a record of cognitive impairment prior to the start date of the HZ vaccination program, a new diagnosis of a given condition was defined as the diagnosis being recorded for the first time in our electronic health record data or as an underlying or contributing cause on the death certificate. Among patients with a diagnosis of dementia made prior to the start date of the HZ vaccination program, a death was considered to be caused by a given condition if it was recorded as the underlying cause of death on the death certificate. Influenza vaccination was defined as receipt of influenza vaccination at any time in the 12 months after the start date of the HZ vaccination program. Statin and antihypertensive medication use was defined as any prescription of these medications during the 12 months after the start date of the HZ vaccination program. Breast cancer screening and diagnoses were analyzed among women only. Breast cancer screening was defined as a record of referral to, attendance at, or a report from “breast cancer screening” or mammography at any time after the start date of the HZ vaccination program. The Read and ICD (International Classification of Diseases) codes used to define each variable shown in this figure are provided in Data S1. Abbreviations are as follows: COPD, chronic obstructive pulmonary disease; CI, confidence interval.

Figure 4.. The effect of HZ vaccination on new diagnoses of MCI and deaths due to dementia separately by gender

Dots show the point estimate and horizontal bars the 95% CI. New diagnoses of MCI were analyzed among a study cohort of patients who did not have any record of cognitive impairment prior to the start date of the HZ vaccination program. Deaths due to dementia were analyzed among a study cohort of patients who had received a diagnosis of dementia prior to the start date of the HZ vaccination program. Abbreviations are as follows: MCI, mild cognitive impairment; CI, confidence interval. See also Figures S5–S7.

Figure 5.. The effect of HZ vaccination on deaths due to dementia, non-dementia deaths, and all-cause mortality

Dots show the point estimate and horizontal bars the 95% CI. Deaths due to dementia were analyzed among a study cohort of patients who had received a diagnosis of dementia prior to the start date of the HZ vaccination program. Non-dementia deaths were defined as deaths for which dementia was recorded as neither the underlying nor a contributing cause of death in the death certificate. Abbreviation is as follows: CI, confidence interval. See also Figures S8 and S9.