The effect of herpes zoster vaccination on the occurrence of deaths due to dementia in England and Wales

Abstract

AbstractBackgroundThe United Kingdom (UK) has used date of birth-based eligibility rules for live-attenuated herpes zoster (HZ) vaccination that have led to large differences in HZ vaccination coverage between individuals who differed in their age by merely a few days. Using this unique natural randomization, we have recently provided evidence from Welsh electronic health record data that HZ vaccination caused a reduction in new dementia diagnoses over a seven-year period. Based on this, we hypothesized that HZ vaccination may have slowed the dementia disease process more generally and, thus, already reduced deaths with dementia as their underlying cause even though the UK’s HZ vaccination program commenced as recently as September 2013. Using country-wide death certificate data for England and Wales, this study, therefore, aimed to determine whether eligibility for HZ vaccination caused a reduction in deaths due to dementia over a nine-year follow-up period.MethodsAdults who had their 80thbirthday shortly before September 1 2013 were ineligible for HZ vaccination in the UK’s National Health Service and remained ineligible for life, whereas those who had their 80thbirthday shortly after September 1 2013 (i.e., born on or after September 2 1933) were eligible for one year. Akin to a randomized trial, this date-of-birth threshold generated birth cohorts who are likely exchangeable in observed and unobserved characteristics except for a small difference in age and a large difference in HZ vaccination uptake. We used country-wide data from death certificates in England and Wales on underlying causes of death from September 1 2004 to August 31 2022 by ICD-10 code and month of birth. Our analysis compared the percentage of the population with a death due to dementia among the month-of-birth cohorts around the September 2 1933 eligibility threshold using a regression discontinuity design. The primary analyses used the maximal available follow-up period of nine years.ResultsThe study population included 5,077,426 adults born between September 1 1925 and August 31 1941 who were alive at the start of the HZ vaccination program. The month-of-birth cohorts around the September 2 1933 eligibility cutoff were well balanced in their occurrence of all-cause and cause-specific deaths (including deaths due to dementia) prior to the start of the vaccination program. We estimated that over a nine-year follow-up period, eligibility for HZ vaccination reduced the percentage of the population with a death due to dementia by 0.38 (95% CI: 0.08 to 0.68, p=0.012) percentage points, corresponding to a relative reduction of 4.8%. As in our prior analysis, this effect was stronger among women (−0.62 [95% CI: −1.06 to −0.19] percentage points, p=0.004) than among men (−0.11 [95% CI: −0.51 to 0.28] percentage points, p=0.574). The reduction in deaths due to dementia likely resulted in an increase in remaining life expectancy because we found that HZ vaccination eligibility reduced all-cause mortality but had no effect on deaths not due to dementia. An effect on deaths due to dementia at the September 2 date-of-birth eligibility threshold existed only since the year in which the HZ vaccination program was implemented.ConclusionsOur findings indicate that HZ vaccination improved cognitive function at a fairly advanced stage of the dementia disease process because most individuals whose underlying cause of death was dementia during our nine-year follow-up period were likely already living with dementia at the start of the HZ vaccination program. By using a different population, type of data, and outcome than our prior study in Welsh electronic health record data, this analysis adds to the evidence base that HZ vaccination slows, or potentially even prevents, the natural history of dementia.