Short-term safety of COVID-19 mRNA vaccines with respect to all-cause mortality in the older population in Norway

•Frail older individuals are not included in clinical trials of COVID-19 vaccination.•Previous studies have shown few local and systemic reactions among vaccinated older people; however, even mild adverse events following immunization could destabilize frail individuals.•There was no evidence of increased short-term mortality among vaccinated older individuals.•Short-term mortality was lower in the vaccinated group compared to the unvaccinated group, which could reflect a healthy-vaccinee effect.•Real-world studies give the possibility to monitor adverse effects of COVID-19 vaccination. There have been concerns about COVID-19 vaccination safety among frail older individuals. We investigated the relationship between COVID-19 mRNA vaccination and mortality among individuals aged ≥ 70 years and whether mortality varies across four groups of health services used. In this nationwide cohort study, we included 688,152 individuals aged ≥ 70 years at the start of the Norwegian vaccination campaign (December 27, 2020). We collected individual-level data from theNorwegian Emergency Preparedness Register for COVID-19. Vaccinated and unvaccinated individuals were matched (1:1 ratio) on the date of vaccination based on sociodemographic and clinical characteristics. The main outcome was all-cause mortality during 21 days after first dose of COVID-19 mRNA vaccination. Kaplan-Meier survival functions were estimated for the vaccinated and unvaccinated groups. We used Cox proportional-hazards regression to estimate hazard ratios (HRs) of death between vaccinated and unvaccinated individuals, with associated 95% confidence intervals (CIs), overall and by use of health services (none, home-based, short- and long-term nursing homes) and age group. Between December 27, 2020, and March 31, 2021, 420,771 older individuals (61.1%) were vaccinated against COVID-19. The Kaplan-Meier estimates based on the matched study sample showed a small absolute risk difference in all-cause mortality between vaccinated and unvaccinated individuals, with a lower mortality in the vaccinated group (overall HR 0.28 [95% CI: 0.24–0.31]). Similar results were obtained in analyses stratified by use of health services and age group. We found no evidence of increased short-term mortality among vaccinated individuals in the older population after matching on sociodemographic and clinical characteristics affecting vaccination and mortality.