Nationwide Study Finds No Short-Term Rise in Cancer After COVID-19 Vaccination

A nationwide analysis of more than 36 million South Koreans has found no detectable short-term change in cancer incidence trends following the country’s rapid SARS-CoV-2 vaccination rollout, directly challenging a recent individual-level study that reported increased cancer diagnoses within a year of vaccination. The new interrupted time series study, published in Biomarker Research, examined national insurance data spanning February 2011 to September 2025 and found that observed cancer incidence remained within the statistical bounds predicted by pre-vaccination trends in almost every month after the rollout began.

The study was prompted by a cohort study of Seoul residents that used Korean National Health Insurance data to report increased incidence of several cancers, including gastric, colorectal, lung, breast, thyroid and prostate cancer, among vaccinated individuals within one year of vaccination. The authors of the new study note that an excess of newly diagnosed cancers within so short an interval is difficult to reconcile with the induction periods generally observed in carcinogenesis, which are measured in years to decades. They also point out that observational cohort studies comparing vaccinated with unvaccinated individuals are susceptible to confounding and selection bias, including the healthy vaccinee effect, whereby people who choose vaccination differ from those who do not in baseline health, health-seeking behavior and access to care.

To avoid these pitfalls, the research team, led by Yewon Na, Munkhzul Radnaabaatar and Jaehun Jung, employed an interrupted time series design. Rather than comparing vaccinated with unvaccinated individuals, this quasi-experimental approach evaluates whether the incidence trend of an outcome in the whole population changes at the time of a population-wide intervention, using the pre-intervention trend as the counterfactual. Because vaccinated and unvaccinated people are never compared, selection on individual characteristics cannot bias the estimate, although the design remains sensitive to other events that coincide with the intervention.

The study population comprised 36,024,394 individuals born between 1955 and 2004, of whom 89.3 percent had received at least one dose of a SARS-CoV-2 vaccine, with coverage ranging from 84.7 to 93.8 percent across birth cohorts. The intervention point was set at August 2021, the month in which cumulative first-dose coverage in the general population exceeded 50 percent according to national vaccination statistics. The researchers chose this threshold because the Korean rollout was compressed rather than gradual: within each birth cohort, coverage rose by 33 to 72 percentage points in a single calendar month, so coverage curves approximate a step rather than a slowly rising function.

The primary outcomes were monthly incidence rates of 13 cancer categories defined by ICD-10 code ranges, calculated as newly diagnosed cases per 100,000 population at risk. Case ascertainment relied on registration in the national co-payment reduction program for critical illnesses, which requires certification of the diagnosis by a physician based on diagnostic test results such as histopathology, or imaging findings with a specialist’s confirmation when biopsy is not feasible. To reduce the inclusion of prevalent cases, the team applied a diagnosis-free washout period covering the full history available in the database before the start of observation, excluding anyone diagnosed with any cancer between January 2002 and January 2011. Four site-specific cancers, stomach, colorectum, lung and prostate, were additionally analyzed separately because each falls within a broader code range that could mask a change confined to one site.

For each cancer category, ARIMA or seasonal ARIMA models were fitted to the pre-intervention period of 126 monthly observations, more than ten complete annual cycles, and used to generate counterfactual forecasts for the 50 post-intervention months. Segmented regression using ARIMAX or SARIMAX models was then fitted to the full 176-month series, with terms representing an immediate level change and a change in the post-vaccination trend. With a prespecified significance threshold of P < 0.01, no immediate level change reached statistical significance for any of the 13 categories: relative risk estimates ranged from 0.923 for bone and articular cartilage to 1.072 for eye, brain and other parts of the central nervous system, and the relative risks for the monthly trend change ranged from 0.995 to 1.005. None of the four site-specific cancers showed a significant level or trend change either.

The observed monthly incidence rates lay within the 99 percent prediction intervals of the counterfactual forecasts in all but two of the 850 post-intervention cancer-months examined: eye, brain and other parts of the central nervous system in November 2021, and female genital organs in October 2022, both above the upper limit. In birth-cohort-stratified analyses across 5-year cohorts, no immediate level change met the threshold in any cohort, and only three of 260 estimates met it, all for the monthly trend change and all indicating slightly slower growth after the rollout. Across a lag grid in which the interruption was shifted to every month from 0 to 18 after the primary intervention point, 2.0 percent of estimates met the threshold, compared with 4.1 percent at 30 placebo points in the pre-pandemic period, when no vaccination had occurred.

The most notable departures involved female genital organ cancers, and the authors attribute them to the COVID-19 pandemic’s disruption of healthcare rather than to vaccination. Placing the interruption at the onset of the pandemic in February 2020 gave an immediate decrease of 16 percent, twelve months before the national vaccination program began. Incidence then rose above the 99 percent prediction interval in October 2022, and lag-grid estimates were elevated at placements between April and October 2022 before returning to unity by a lag of 17 months. The researchers note that participation in organized cervical cancer screening in Korea fell from 57.8 percent in 2019 to 52.2 percent in 2020, with a 49 percent drop in March 2020 alone, and that nationally the number of new cancer diagnoses fell in 2020 and rose by 10.8 percent in 2021 as access to care returned. They caution that because the rollout and the recovery of health care overlapped in time, their influences cannot be fully separated by this design.

The study also quantified the size of the effects it could exclude. At the primary intervention point, the 99 percent confidence limits ruled out immediate increases larger than 8.9 to 23.0 percent across the 13 cancer categories and larger than 8.1 to 23.4 percent for the four site-specific cancers. The authors argue that if the individual-level associations reported by the Seoul cohort study reflected an effect of vaccination that began at vaccination and persisted over the following year, the rapid uptake by nearly 90 percent of the population would have produced an increase in population incidence larger than these limits. The hazard ratio of 1.35 reported for thyroid cancer, for example, implies an increase of about 31 percent in population incidence at the observed coverage of 89 percent, well above the 16 percent excluded here. No such increase was observed.

The authors emphasize the limits of what their design can answer. This was an ecological study based on aggregate time-series data, so the findings represent population-level associations and cannot be used to infer causality or estimate individual-level risks. Individual-level risk factors such as smoking, obesity and genetic predisposition were not accounted for, and the number of doses, time since vaccination and vaccine type were not represented in the analysis. The 50-month post-intervention period is adequate for the short-term question addressed but not for effects with long induction periods, which for most solid cancers are measured in years to decades. Smaller changes cannot be ruled out, and effects that would emerge after the observation period are not addressed. Residual diagnostics also indicated departures from at least one model assumption for some models, and losses from the cohort through death or emigration were not incorporated into the monthly denominators. Within those constraints, the study concludes that the national SARS-CoV-2 vaccination rollout in South Korea was not accompanied by a detectable short-term, population-level change in cancer incidence trends over the 50 months that followed, and that the population-level increases implied by the reported individual-level associations were not observed.

Subject of Research: Population-level cancer incidence trends after the SARS-CoV-2 vaccination rollout in South Korea

Article Title: Impact of SARS-CoV-2 vaccination on cancer trends: a nationwide interrupted time series study

Article References: Na, Y., Radnaabaatar, M., & Jung, J. (2026). Impact of SARS-CoV-2 vaccination on cancer trends: a nationwide interrupted time series study. Biomarker Research, 14(1), Article 147. https://doi.org/10.1186/s40364-026-01027-6

Image Credits: AI Generated

DOI: 10.1186/s40364-026-01027-6

Keywords: SARS-CoV-2 vaccination, cancer incidence, interrupted time series, South Korea, National Health Insurance Service, COVID-19 pandemic, healthy vaccinee effect, ARIMA models, thyroid cancer, healthcare disruption, epidemiology, vaccine safety