# The inverse association between cancer and Alzheimers disease: comparing spurious and causal explanations to illuminate the causes of Alzheimers disease

> **NIH NIH RF1** · UNIVERSITY OF CALIFORNIA, SAN FRANCISCO · 2021 · $399,917

## Abstract

Abstract/Summary
Several studies report an inverse comorbidity between cancer and Alzheimer’s Disease (AD). Incidence rates
of AD are about 30% lower among cancer survivors than individuals with no history of cancer. Recent findings
indicate that differential survival after cancer cannot fully explain this inverse association, and biological
explanations are hypothesized. Neoplastic cell growth underlying cancer may be the flip side of the cellular
process that contributes to neuronal death in AD; for example regulation of apoptosis, immune response, or
DNA repair may account for the inverse comorbidity of cancer and AD. If the inverse association between
cancer and AD arises from a common physiologic process, explaining this association could reveal novel
insights into the pathophysiology of AD and highlight targets for preventive or therapeutic interventions. We
propose to evaluate competing explanations for the inverse cancer-AD association : (1) diagnostic bias:
individuals with a history of cancer are less likely to be diagnosed with AD; (2) competing risks: both conditions
increase mortality, so occurrence of either reduces lifetime risk of the other; (3) survival bias: factors that
improve survival of cancer patients are associated with lower AD risk, so cancer survivors are a biased sample
of all cancer patients; (4) inverse common causes: cancer incidence is reduced by genetic or environmental
factors which increase AD risk; (5) causality: physiologic or treatment responses to cancer reduce risk of AD.
We will systematically evaluate these 5 alternative explanations for the inverse comorbidity of cancer and AD,
with the aim of improving understanding of the biological events that initiate or maintain the Alzheimer’s
cascade. We use longitudinal analyses of two large cohorts (the Health and Retirement Study [HRS] and the
UK Biobank [UKB]), genetic quasi-experiments, and simulation models to evaluate the plausibility of competing
explanations. In AIM 1, we evaluate the link between cancer and longitudinal rate of cognitive change in HRS
and UKB. Only one prior study evaluated cancer and longitudinal cognitive change. We hypothesize that
cognitive decline will be slower both before and after cancer diagnosis, even for non-life-threatening cancers,
compared to people with no cancer diagnosis. In AIM 2, we test whether cancer shares genetic risk factors
with AD or cognitive change. We construct polygenic cancer risk scores both using genome-wide data and
using specific variants previously confirmed to influence cancer risk. We then assess whether these polygenic
cancer risk scores predict lower risk of AD. We also examine the reverse, whether polygenic AD risk scores
predict lower cancer risk. In AIM 3, we combine multiple sources of evidence on cancer type specific mortality
rates, genetic correlations, and associations with AD to specify simulation models. In combination, these
observations will demonstrate the most likely explanation for t...

## Key facts

- **NIH application ID:** 10465775
- **Project number:** 3RF1AG059872-01S1
- **Recipient organization:** UNIVERSITY OF CALIFORNIA, SAN FRANCISCO
- **Principal Investigator:** Medellena Maria Glymour
- **Activity code:** RF1 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2021
- **Award amount:** $399,917
- **Award type:** 3
- **Project period:** 2018-08-15 → 2023-08-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10465775

## Citation

> US National Institutes of Health, RePORTER application 10465775, The inverse association between cancer and Alzheimers disease: comparing spurious and causal explanations to illuminate the causes of Alzheimers disease (3RF1AG059872-01S1). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10465775. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*