Suicide Risk in Oncology: Sex and Cancer Type Differences in a Case-Control Study

Background: Suicide risk among people with cancer may vary by sex and cancer site, reflecting different biological, psychosocial, and care-system pathways. We used statewide linked mortality and clinical data to examine whether cancer history and sex-specific cancer types differentially relate to pre-death suicidality and suicide mortality. Methods: We conducted a two-step study using population-level data from Utah linking suicide mortality, cancer, and electronic health records. First, we compared suicide decedents with versus without a prior cancer diagnosis (n=14,644) on suicidal ideation (SI), self-injurious behavior (SI), prior suicide attempts (SA), psychiatric and medical comorbidities. Second, we performed an age-and sex-matched case-control analysis of suicide decedents (cases; n=1,015) and living controls (n=9,173) to estimate adjusted odds of suicide death associated with any cancer history and specific cancer types, stratified by sex. Logistic regression models adjusted for prior suicidality, diagnosed mental and substance use disorders (SUD), and chronic medical morbidity. We also characterized the temporal sequencing of first-recorded encounter types (mental health, SUD, chronic medical, or cancer-related). Results: Among suicide decedents, those with any history of a cancer diagnosis had higher odds of pre-death SA (OR=1.27, 95% CI 1.09-1.49), SII (OR=1.34, 1.12-1.60), and SI (OR=1.29, 1.07-1.56) than decedents without cancer. Mental-health burden was substantially greater among female than male decedents (OR=7.90 vs 2.07). In case-control analyses, a history of any cancer was associated with lower overall odds of suicide death, but this aggregate effect masked heterogeneity by sex and cancer type. Among women, cervical cancer/dysplasia was over-represented in cases compared to controls (OR=1.53, 1.13-2.06), suggesting elevated risk in sex-specific, identity-salient cancers. Among men, prostate cancer was inversely associated with suicide death (OR=0.73, 0.59-0.91). First encounters for mental health and substance use were over-represented among cases of both sexes, while chronic-condition encounters suggest additional risk in men. Conclusions: In this study, any history of a cancer diagnosis was linked to greater pre-death suicidality but lower overall odds of suicide death, with important sex-and cancer-type specific differences. Patterns may support a dual-pathway model in which psychosocial/identity-related mechanisms may predominate among women with sex-specific cancers, whereas functional or disease-burden pathways may predominate among men with high-burden cancers. Tailored, sex-and cancer-type-specific suicide risk screening that leverages mental health and substance use encounter history may improve prevention in oncology settings.