Heterogeneity in proportional cardiovascular co-listing on cancer death certificates in the United States: a national multiple cause-of-death analysis, 1999-2020.

BACKGROUND

Cardiovascular disease (CVD) and cancer share common risk factors, and cancer survivors experience elevated long-term cardiotoxicity. Traditional risk-based metrics require person-time denominators that are unavailable in aggregate public health databases. This study introduces a proportional co-listing framework to detect population-level cardiovascular surveillance signals on death certificates across cancer types and demographic strata.

METHODS

We conducted a national cross-sectional surveillance analysis using CDC WONDER Multiple Cause-of-Death data from 1999 to 2020 for decedents aged ≥ 15 years. Utilizing an "any-mention" framework, we calculated the Proportional Mortality Ratio (PMR), defined as the ratio of CVD co-listing on cancer-related death certificates to CVD listing on non-cancer certificates. Temporal trends were evaluated using joinpoint regression to estimate the average annual percent change (AAPC). Analyses were stratified by age, sex, race/ethnicity, geography, and index cancer subtypes (lung, colon, and leukaemia).

RESULTS

Of 56,014,102 death certificates analyzed, 13,917,133 (24.8%) listed a malignant neoplasm, of which 3,997,238 (28.7%) co-listed CVD. Across the lifespan, the overall PMR followed a reverse J-shaped curve. Proportional co-listing was highest in the youngest cohort (ages 15-24: PMR 1.828, 95% CI: 1.778-1.879), declined to a nadir at ages 55-64 (PMR 0.336, 95% CI: 0.335-0.337), and rose modestly in older adulthood. From 1999 to 2020, PMRs increased significantly across all adult age groups, peaking at ages 55-64 (AAPC 1.69%, 95% CI: 1.52-1.83, p < 0.001). Young-adult PMR elevations were consistently higher in male, metropolitan, and Hispanic/Latino decedents. Among subtypes, leukaemia demonstrated the highest relative CVD co-listing in young adults (PMR 2.174), whereas colon cancer showed no statistically significant elevation (PMR 1.173, 95% CI: 0.951-1.447).

CONCLUSIONS

Cardiovascular co-listing on US cancer death certificates is highly patterned by age, demographics, and cancer subtype. While elevated PMRs in young adults are consistent with patterns previously reported in cardio-oncology literature, sub-unity PMRs in middle-aged and older adults may reflect under-documentation of cardiovascular comorbidities in oncology settings alongside denominator effects arising from frequent CVD listing on non-cancer certificates. Improving comorbidity documentation on death certificates may enhance the utility of national administrative data for cardio-oncology surveillance and research.