Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jul 27.
Published in final edited form as: J Natl Cancer Inst. 2025 Aug 1;117(8):1707–1716. doi: 10.1093/jnci/djaf122

Risk of Atherosclerotic Cardiovascular Disease after Cancer Diagnosis: Findings from Three Prospective Cohort Studies

Qiang Liu 1,2, Qiaoli Wang 3, Kai Wang 4, Han Han 2, Molin Wang 4,5,6, Jing Wang 1,*, Mingyang Song 2,4,7,8,*, Edward Giovannucci 2,4,*
PMCID: PMC12290291  NIHMSID: NIHMS2095411  PMID: 40445187

Abstract

Background

To determine the association between cancer diagnosis and subsequent risk of ASCVD, and to examine the trajectory of the association over time after cancer diagnosis.

Methods

We prospectively followed 108,689 women in the Nurses’ Health Study (NHS) (1984–2020), 113,569 women in the NHSII (1991–2019), and 45,328 men in the Health Professionals Follow-up Study (HPFS) (1986–2016) who were free of ASCVD and cancer at baseline. We conducted multivariable-adjusted time-varying Cox proportional hazards regression models to assess ASCVD risk following individual cancer diagnosis.

Results

During up to 36 years of follow-up, 4,334 new-onset ASCVD events among 49,603 incident cancer cases were documented. After adjusting for shared risk factors, cervical cancer (HR: 1.56; 95%CI: 1.06–2.29) and Hodgkin lymphoma (HR: 2.80; 95%CI: 1.89–4.15) was associated with increased risk of ASCVD incidence, while prostate cancer was associated with a lower ASCVD incidence (HR: 0.91; 95% CI: 0.85–0.97). Compared to cancer-free individuals, breast cancer survivors experienced lower ASCVD risk during the first 7.5 years but gradually increased afterwards (Pnon-linearity=0.01). The risk of ASCVD increased over time among patients with cancers of the colorectum (P=0.003), lung (P=0.002), and endometrium (P=0.04). No significant association with ASCVD risk was observed for cancers of the oral cavity and pharynx, sarcoma, melanoma, kidney, thyroid, leukemia, or ovary.

Conclusions

Cervical cancer or Hodgkin lymphoma was associated with an increased risk of new-onset ASCVD, independent of shared risk factors, while no increased risk was found with other cancers. ASCVD risk trajectories varied over time after diagnosis according to cancer types.

INTRODUCATION

Cardiovascular disease (CVD) is the leading cause of death worldwide[1]. Atherosclerotic CVD (ASCVD) accounts for four out of every five CVD deaths[3]. Many deaths among cancer patients are not attributed to cancer itself, but ASCVD[4, 5]. It has been hypothesized that both cancer and CVD share pathways of systematic inflammation and oxidative stress[6, 7] and that cancer treatment might increase CVD risk[8]. Recently, several studies suggested elevated risk of composite CVD (including heart failure [HF]) among cancer patients independent of shared risk factors[8, 9]. Such an association, if confirmed, has important implications for both clinical practice and basic medical sciences. It raises questions regarding CVD screening, treatment dosing options, and the inclusion of cancer in CVD risk models.

However, prior studies have important limitations. First, most prior studies were constrained by a retrospective design that compared the risk in cancer patients with matched controls[9–12], often lacking consistent adjudication of incident cancer and ASCVD. Second, studies did not always account for important confounders[8, 10, 13, 14], such as lifestyle factors, diet, family history of diabetes, family history of CVD, and preventive medication use for ASCVD. Third, most available studies were unable to distinguish CVD subtypes[10, 11, 14], particularly ASCVD and HF, which have distinct characteristics and pathophysiology. ASCVD, which mainly includes coronary heart disease and stroke, accounts for the majority of CVD cases[15], and results from plaque buildup in arterial walls causing the inside of the arteries to narrow over time[16]. By contrast, HF results from any structural or functional cardiac disorder that impairs the ability of the ventricle to hold or eject blood[17]. Certain cardiotoxic cancer treatments have been well-documented to increase the risk of HF by inducing myocardial injury or by inhibiting essential molecular pathways for normal cardiac function[18, 19], and the increased risk of HF among cancer survivors has been established[20]. However, the association between incident cancer and ASCVD (without HF) remains inconclusive[8]. Finally, no study has yet comprehensively evaluated the risk trajectory of ASCVD following an incident cancer diagnosis over time.

To address these knowledge gaps, we leveraged three well-established prospective cohorts featuring long-term follow-up, detailed lifestyle information, and comprehensive health attributes to evaluate further the association between incident cancer and new-onset ASCVD. This study has two primary objectives: (1) For the clinically relevant question, we assessed whether cancer patients experienced varied ASCVD risk over time following a cancer diagnosis and (2) for the etiological purpose, we evaluated further whether an incident cancer diagnosis is associated with ASCVD risk independent of shared risk factors.

PATIENTS AND METHODS

Study Population

The three prospective cohort studies encompassed in this analysis are the Nurses’ Health Study (NHS), the NHS II, and the Health Professionals Follow-up Study (HPFS). The NHS was initiated in 1976 and recruited 121,700 female nurses aged 30 to 55 years[21]. The NHSII, which began in 1989, included 116,429 female nurses aged 25 to 42 years[21] at baseline. The HPFS, established in 1986, involved 51,529 male health professionals aged 40 to 75 years[22] (Supplementary methods). Among participants who completed the baseline FFQ, we excluded those who reported cardiovascular disease or cancer at baseline, or missing age at baseline, resulting in a total of 267,586 participants in the final analyses (NHS, 108,689; NHSII, 113,569; HPFS; 45,328). The study protocol was approved by the institutional review boards of the Brigham and Women’s Hospital and Harvard T.H. Chan School of Public Health, and those of participating registries as required.

Ascertainment of Cancer diagnosis

Ascertainment of cancer diagnosis follows the methodology of our group’s previous studies[23]. Briefly, incident invasive cancers, excluding non-melanoma skin cancers, in-situ breast cancers, and stage T1a prostate cancers, were identified via biennial questionnaires or through regular National Death Index searches. The diagnoses of these cancers were subsequently confirmed by physicians through a thorough review of medical records, pathology reports, death certificates, or data from cancer registries. In this study, we only presented the individual cancer types with 20 ASCVD events or more during follow-up.

Ascertainment of ASCVD

Due to the large proportion of ASCVD within overall CVD cases[3], and the distinct characteristics between ASCVD and HF[15, 16], this study specifically focused on ASCVD. ASCVD ascertainment follows the methodology outlined in our previous studies[24]. Briefly, ASCVD is a composite endpoint that includes fatal and nonfatal coronary heart disease (CHD) (nonfatal myocardial infarction [MI]), fatal and nonfatal stroke, and coronary artery bypass surgery (CABG) /angioplasty/stent cases. Nonfatal MI was confirmed by physicians according to the World Health Organization criteria[25] (Supplementary Methods).

Ascertainment of Covariates

Updated information on participants’ medical history, disease family history, lifestyle, body mass index (BMI), and medication use was collected through biennial questionnaires. Among women, information on menopausal status and postmenopausal hormone use was assessed. We used the Alternative Healthy Eating Index-2010 (AHEI-2010) to quantify overall diet quality[26]. Physical activity was self-reported through a validated questionnaire[27]. Missing data on covariates ranged from 4% for postmenopausal hormone use and 16% for physical activity (Supplementary Table 1). We imputed missing covariate data by carrying forward non-missing data from prior questionnaire cycles for continuous variables and used missing indicator method to handle missing covariates, which has been shown be unbiased in our cohorts[28].

Statistical Analyses

To provide more accurate estimates by capturing exposure status changes during follow-up, we modelled cancer diagnosis status and cancer duration as time-varying exposures. For each participant, follow-up time was calculated from baseline to the date of ASCVD diagnosis, death date, date of return of last available follow-up questionnaire, or the end of follow-up (June 2020 in NHS, June 2019 in NHSII, and January 2016 in HPFS), whichever occurred first. Cancer duration was calculated as years that elapsed since cancer diagnosis. To better compare with prior studies[9], we conducted a matched cohort design in a secondary analysis (Supplementary methods).

Time-varying Cox proportional hazard regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between incident cancer diagnosis and subsequent risk of ASCVD outcomes, including total ASCVD, CHD, and stroke. All analyses were stratified by age in months, cohorts, and follow-up intervals, and multivariable models were further adjusted for race, time-varying covariates including smoking, menopausal status and postmenopausal hormone use (in women), multivitamin use, regular aspirin use, physical activity, family history of MI, family history of diabetes, BMI, total energy intake, alcohol intake, a modified AHEI-2010 (excluding alcohol), hypercholesterolemia, hypertension, and diabetes status.

In secondary analysis, restricted cubic spline analyses were employed to test for potential non-linear relationships between cancer duration and ASCVD risk (Supplementary methods). Several sensitivity analyses were conducted. First, to account for potential influence of screening, we additionally adjusted for physical examination and prostate-specific antigen (PSA) screening (prostate cancer). Second, we censored participants when they reached the age of 80 years as participants with low life expectancy might get underdiagnosed for cancer. Third, to explore whether the association could be largely explained by common risk factors, we employed a multivariate model only adjusting for age, BMI, smoking, and alcohol intake. Fourth, to account for potential selection bias that cancer diagnosis may influence the likelihood of undergoing CABG/angioplasty/stent, we repeated the analyses by excluding these cases from ASCVD outcome. Fifth, to account for the potential influence of cancer treatment, we excluded participants who died within 2 or 5 years after cancer diagnosis. Sixth, to assess the influence of peri-diagnosis lifestyle changes, we compared the results with the main results by stopping updating lifestyle and dietary information after cancer diagnosis. Seventh, the Fine-Gray subdistribution hazard model was used to account for potential competing risk by death. All statistical tests were 2-sided with a statistical significance level of .05 and were performed using SAS 9.4 (SAS Institute, Cary, NC).

RESULTS

Participant Characteristics

The age-standardized characteristics of participants are shown according to cancer-diagnosis status in Table 1. Overall, compared with participants without cancer, participants diagnosed with cancer were older, more likely to be past smokers, intake alcohol, use aspirin, use multivitamin, have hypertension, hypercholesterolemia, and diabetes. Similar BMI and diet were found for participants with and without cancer. Lung cancer survivors were more likely to be current and past smokers, have family history of MI (for women), and be less physically active.

Table 1.

Age-standardized baseline characteristics of participants according to major incident cancer

Cancer type All cancer All cancer Lung Colorectum Breast Prostate
Sex Non-cancer women Women with cancer Non-cancer men Men with cancer Women Men Women Men Women Men
Age, mean (SD), yrs 56.1 (12.5) 65.7 (11.3) 63.2 (11.3) 73 (9.1) 69.4 (9.7) 72.1 (9.2) 69.4 (10.9) 72.7 (9.7) 65.5 (11.1) 74.0 (8.4)
White, % 96.0 97.4 94.7 95.7 97.2 93.9 97.0 95.8 97.1 96.0
Body mass index, mean (SD), kg/m2 26.6 (5.9) 26.8 (5.9) 26 (3.8) 25.9 (3.7) 25.5 (5.3) 25.7 (4.0) 26.5 (5.6) 26.2 (3.6) 26.5 (5.4) 25.7 (3.5)
Physical activity, MET-h/wk 20.6 (27.4) 19 (25.1) 31.4 (32.1) 31.8 (31.7) 14.5 (19.9) 24.1 (26.2) 16.6 (23.4) 28 (30.2) 19 (24.0) 32.6 (32.2)
Alternative healthy eating index, mean (SD) 47 (9.9) 47.2 (9.4) 48.3 (10.1) 48.3 (9.9) 45.8 (9.0) 46.9 (9.5) 46.7 (8.8) 47.7 (10.0) 47.4 (9.4) 48.7 (9.7)
Alcohol consumption, mean (SD), g/day 4.8 (7.9) 5.5 (8.7) 11.1 (13.8) 11.7 (14) 7.9 (11.3) 15.0 (18.2) 5.6 (8.7) 13.3 (14.2) 5.7 (8.8) 11.7 (13.2)
Alcohol consumption, % 63.1 69.8 81.5 83.8 67.0 86.0 67.6 84.6 71.4 85.8
Past smokers, % 32.8 40.3 42.4 45.0 53.2 56.6 42.6 48.2 40.9 46.4
Current smokers, % 10.2 9.9 5.6 4.8 27.7 18.9 8.6 5.8 9.0 3.8
Aspirin use, % 35.4 39.9 36.6 37.3 44.3 32.0 41.2 35.9 40.3 40.2
Multivitamin use, % 51.2 56.0 54.0 59.5 51.3 54.2 51.7 54.5 57.4 61.0
History of Hypertension, % 30.9 38.9 37.0 43.3 39.6 39.3 40.3 40.3 37.4 46.3
History of diabetes, % 5.8 8.3 6.9 5.4 7.6 9.8 8.4 11.1 7.4 6.6
History of Hypercholesterolemia, % 40.8 48.7 41.9 50.6 48.4 41.7 46.9 48.8 48.8 54.1
Family history of MI, % 21.3 22.0 14.3 13.6 25.0 13.1 22.3 10.4 21.4 12.4
Family history of diabetes, % 29.7 28.8 20.6 19.9 25.4 22.2 29.4 24.2 28.0 18.1
Open in a new tab

Values are mean ±SD or % and are standardized to the age distribution of the study population. yrs=years. MI=myocardial infarction, MET=metabolic equivalent task; Data were presented using pooled dataset.

Cancer diagnosis and risk of ASCVD outcomes.

The longitudinal study design and matched cohort design are shown in Figure 1A, the objective of the study is illustrated in Figure 1B. Over 7,571,534 person-years of follow-up across the three cohorts, we documented 49,836 incident cancer cases and 34,219 new-onset ASCVD events. Among these, 4,334 new-onset ASCVD events occurred in cancer patients. The median follow-up times to ASCVD among cancer-free participants were 29.9 years and 9.8 years (from time of cancer diagnosis) among cancer survivors. In the age-adjusted model, overall cancer and most individual cancer types were not associated with ASCVD risk (Figure 2). However, compared to non-cancer participants, patients with Hodgkin lymphoma, cervical cancer, lung cancer, and colorectal cancer were associated with an increased risk of ASCVD, with the HRs ranging from 1.09 (95%CI:1.00–1.20) to 2.80 (95%CI: 1.89–4.15), while prostate cancer was inversely associated with ASCVD risk. For the etiological purpose, we employed a multivariable model further adjusting for demography, lifestyle, diet, BMI, and health attributes (Figure 2). We found the associations were minimally changed for Hodgkin lymphoma, cervical cancer, overall cancer, and prostate cancer, while the HRs attenuated to 1.09 (95%CI: 0.94–1.27) for lung cancer, and 1.03 (95%CI: 0.94–1.13) for colorectal cancer. By using the matched cohort design similar to prior studies[9], the results were largely consistent with the time-varying analysis, though the associations were slightly weakened (Figure 2). When considering specific ASCVD subtypes, the inverse associations with overall cancer, breast cancer, and prostate cancer were stronger for CHD risk, and no significant association was found with stroke risk (Supplementary Figure 1 and 2).

Figure 1. Longitudinal study design and objective illustration.

Figure 1.

Open in a new tab

A. Time varying design capturing exposure status changes during follow-up period. Black solid line indicates non-exposed person time before cancer diagnosis in time-varying design which was excluded in matched-control design. The red line denotes exposed person-time after cancer diagnosis. B. overall objective of the study. ASCVD: atherosclerotic cardiovascular disease.

Figure 2. Association of cancer diagnosis with ASCVD in both time-varying design and matched cohort design.

Figure 2.

Open in a new tab

Age-adjusted model adjusted for age. Multivariable (MV) model adjusted for age and race (White, non-White), pack-years of smoking), menopausal status and post-menopausal hormone use (premenopausal, never/past users, current users), multivitamin use (no, yes), regular aspirin use (no, yes), physical activity (<3, 3–9, 9–18, 18–27, 27–42, ≥42 metabolic equivalents/week), family history of myocardial infarction (no, yes), family history of diabetes (no, yes), marital status (married, widowed, divorced/separated), body mass index (<23, 23–24.9, 25–29,9, 30–34.9, ≥35 kg/m2), alcohol intake, and modified AHEI (in quintiles) and hypercholesterolemia (no, yes), hypertension (no, yes) and diabetes (no, yes). Events indicates the number of events among person-years diagnosed with cancer.

Risk Trajectory of ASCVD after cancer diagnosis

To address the clinically relevant question that whether cancer patients experienced varied ASCVD risk over time after cancer diagnosis, we employed spline analyses using age as the time scale (Figure 3). We found non-linear associations with ASCVD risk for overall, breast and bladder cancer. Linear relations were found for cancers in the colorectum, endometrium, lung, and prostate. For overall cancer, the risk of ASCVD initially decreased during the first 5 years following diagnosis, but it gradually evolved into an increased risk afterward. Breast cancer patients experienced a decreased risk during the first 7.5 years after cancer diagnosis, peaking around the 5th year, then gradually rising. Bladder cancer patients experienced an increased ASCVD risk between the around 5th and 12.5th year after diagnosis, peaking around the 10th year and then gradually attenuated to inverse association. For the etiological insight exploring how the independent effect of cancer diagnosis changed over time, we conducted spline analyses in the full multivariable model, the overall trend of ASCVD trajectories remained largely consistent to the age-adjusted model, while the linear association within lung and endometrial cancer patients attenuated to insignificant (Supplementary Figure 3). To assess whether the non-linear association for breast cancer was related to cancer treatment, we conducted a stratified analysis according to cancer treatments and found that the non-linear association was restricted to breast cancer patients who received hormone therapy (Figure 4, Supplementary Figure 4). Among these patients, the initial inverse association with CHD was even stronger, while the relationship with stroke seems different.

Figure 3. Risk of ASCVD over Time after Cancer Diagnosis According to Cancer Types.

Figure 3.

Open in a new tab

Adjusted for age. Time since cancer diagnosis was modeled as time varying exposure. The time duration was truncated based on the occurrence of 10 or more ASCVD events, assessed at each 5-year interval. Poverall indicates the test for overall significance of the curve. Pnon-linearity indicates the test for non-linear relationship between time since cancer diagnosis and ASCVD. Plinearity indicates the test for linear relationship between time since cancer diagnosis and ASCVD. If likelihood ratio test for non-linearity was non-statistically significant, linear model and corresponding P values were reported.

Figure 4. Risk of ASCVD, CHD, and Stroke over Time after Breast Cancer Diagnosis According to whether received hormone therapy.

Figure 4.

Open in a new tab

Adjusted for age and race, pack-years of smoking (continuous), menopausal status and post-menopausal hormone use, regular aspirin use, physical activity, family history of myocardial infarction, family history of diabetes, body mass index, alcohol intake, and AHEI and hypercholesterolemia, hypertension and diabetes. The time duration was truncated based on the occurrence of 10 or more events after breast cancer diagnosis, assessed at each 5-year interval. Poverall indicates the test for overall significance of the curve. Pnon-linearity indicates the test for non-linear relationship between time since cancer diagnosis and ASCVD. Plinearity indicates the test for linear relationship between time since cancer diagnosis and ASCVD. If likelihood ratio test for non-linearity was non-statistically significant, linear model and corresponding P values were reported.

Sensitivity and Subgroup Analyses

Results from the sensitivity analyses were similar to our main findings. When further adjusting for physical examination in the main multivariable model (Supplementary Figure 5) or censoring participants older than 80 years old (Supplementary Figure 6), the results were not changed appreciably. In a simpler multivariable model including BMI, alcohol intake and smoking (Supplementary Figure 7), the HRs (95% CIs) were similar to the full-adjusted multivariable model. Excluding CABG/angioplasty/stent cases did not materially change the results (Supplementary Figure 8). When excluding participants who died within 2 or 5 years after cancer diagnosis, the results remained robust (Supplementary Figure 9). When we stopped updating lifestyle factors after cancer diagnosis, the association attenuated slightly (Supplementary Figure 10). After accounting for the competing risk of death, the results remained similar to the main results (Supplementary Figure 11). To explore potential causes for the inverse association for prostate cancer, we further adjusted for PSA screening and stratified the population according to PSA screening status or cancer stage, The results showed minimal changes, but the inverse association was even stronger among advanced prostate cancer patients (Supplementary Table 2). Results were consistent in stratified analyses for major cancer types (Supplementary Table 3 & 4). Significant heterogeneity was only observed for colorectal cancer (Supplementary Table 5), while the meta-analyzed results were not materially changed.

Discussion

Contrasting with prior studies suggesting that cancer survivors are at an increased risk of composite CVD (including HF), our findings showed only specific cancer sites may be considered as independent risk factors in current equations to predict ASCVD and offered novel insights that certain cancer treatments following cancer diagnosis may have driven the cancer-specific patterns for ASCVD development. Clinicians managing cardiovascular risk for specific cancer survivors should be aware of the timing of risk, cancer treatments and CVD subtypes.

There are several differences between the present study and prior studies that analyze the association between incident cancer and ASCVD risk. Most of prior studies employed a retrospective matched cohort design and a composite CVD outcome (including HF)[9, 10]. Conceptually, the exposure status of the matched controls was assumed to remain unchanged during the follow-up period in these studies. However, cancer-free participants may be diagnosed with cancer after being matched to cases, potentially leading to misclassification. The present study applied the time-varying analysis and may thus provide more accurate estimates and minimize confounding by age[29, 30], which is the strongest risk factor for ASCVD. The only prospective study with 3,250 total cancer survivors, even fewer CVD events and 5.2 years median follow-up among cancer survivors suggested that several incident cancers were associated with increased risk of composite CVD incidence (including HF) independent of shared risk factors, but the association was largely driven by HF rather than ASCVD[8]. By contrast, our study, with a much longer follow-up duration among cancer survivors and a more than 10-fold larger sample size, found that most cancers were not associated with an increased risk of ASCVD (without HF). The observed association between lung and colorectal cancer and ASCVD can be attributed to shared traditional risk factors. Our findings have important clinical implications that caution should be made when stating cancer survivors are at increased risk of CVD without accounting for specific CVD subtypes.

The association between Hodgkin lymphoma[31, 32], prostate cancer[10, 33] and ASCVD was consistent with prior studies. The association between breast cancer and CHD was mixed in prior studies[10, 34, 35]. The CLUE II cohort similarly found an increased risk of CVD-related deaths only after 8 years following a breast cancer diagnosis[36]. We went further and found a non-linear association with CHD incidence among breast cancer survivors who received hormone therapy. Consistently, several studies showed that tamoxifen use, a commonly used hormone therapy, is associated with reduced risk of CHD by influencing cholesterol levels[37, 38] and hormone receptor-positive breast cancer patients typically take hormone therapy for 5–10 years after cancer diagnosis[39], which may corresponding to the inverse association peaked at around 5 years after cancer diagnosis and gradually arise afterward in the present study.

Cervical cancer, primarily caused by human papillomavirus (HPV) persistent infection, has recently been shown to be associated with over a 3.5-fold increased risk of ASCVD[40]. Although HPV has not yet been recognized as an independent risk factor for ASCVD in major consensus guidelines[41, 42], it could potentially underlie the increased risk. Our results suggested a biologically plausible relationship that could have important clinical implications; however, future studies are warranted to replicate this finding. Our findings of prostate cancer, though counterintuitive, aligned with prior studies. The association between ADT and CVD is still under debate, certain types of ADT have already been shown to associated with lower risk of CVD[14]. However, an alternative explanation for the inverse association might be selection bias due to screening, as screened individuals may more frequently access health care including CVD preventive measures involve with health care, though the results minimally changed after adjusting for PSA screening.

Our study has several strengths. First, the prospective longitudinal design and long-term follow-up reduced the potential for reverse causation and may provide more accurate risk estimates. Second, the repeated measurements of covariates, comprehensive and detailed collection of health attributes and risk factors minimized the potential for residual confounding. However, we acknowledge that residual confounding cannot be eliminated. Third, the rigorous confirmation of cancer cases and ASCVD endpoints and nearly complete follow-up based on three well established cohorts minimized selection and ascertainment biases. Our study is, to our knowledge, the largest prospective cohort study that evaluated the potential associations between incident cancer and subsequent new-onset ASCVD.

Some limitations should also be considered. First, the lack of detailed specific treatment data restricts our ability to explore associations between specific medications or treatment durations and ASCVD risk, and future studies are warranted. Second, even with a total of 267,586 participants at baseline, our power to detect associations with rare cancers was likely limited; therefore, larger studies particularly focusing on specific rare cancers are warranted. Finally, the generalizability of our findings may be constrained by having the predominantly being white, health professional demographic of the cohort, although some associations were consistent with prior studies conducted in more diverse populations[9]. It would be interesting for future studies with detailed treatment data to estimate potential area-under the curve effect, that a cancer treated early and present for a short period may have less impact compared to a longer-lasting cancer.

Conclusions

Cervical cancer or Hodgkin lymphoma was associated with an increased risk of ASCVD independent of shared risk factors, while other cancers showed no significant association after adjusting for confounders. Compared with cancer-free participants, cancer survivors experienced varied risk of ASCVD according to cancer types and this risk evolves over time since cancer diagnosis.

Supplementary Material

Supplementary Material

Acknowledgement

The authors would like to acknowledge the contribution to this study from central cancer registries supported through the Centers for Disease Control and Prevention’s National Program of Cancer Registries (NPCR) and/or the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program. Central registries may also be supported by state agencies, universities, and cancer centers. Participating central cancer registries include the following: Alabama, Alaska, Arizona, Arkansas, California, Delaware, Colorado, Connecticut, Florida, Georgia, Hawaii, Idaho, Indiana, Iowa, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Mississippi, Montana, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Puerto Rico, Rhode Island, Seattle SEER Registry, South Carolina, Tennessee, Texas, Utah, Virginia, West Virginia, Wyoming. Study sponsor(s) played no role in the design of the study; the collection, analysis, and interpretation of the data; the writing of the manuscript; and the decision to submit the manuscript for publication. This work has been presented in the AHA2024 Scientific Session as a moderated digital poster.

Funding information

This work was supported by grants from the National Institutes of Health (UM1 CA186107, P01 CA87969, R01 HL034594, R01 HL088521, U01 CA176726, U01 HL145386, U01 CA167552, R01 HL35464). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. EG is funded as an American Cancer Society Clinical Research Professor (CRP-23–1014041). JW reports grant from Natural Science Foundation of China (No. 82,173,328, No. 81,872,160).

Footnotes

Conflicts of interest

We declared no conflicts of interest.

Code availability

Code will be shared at the request of qualified investigators for purposes of replicating procedures and results. Further information including the procedures to obtain and access codes from the Nurses’ Health Studies and Health Professionals Follow-up Study is described at https://www.nurseshealthstudy.org/researchers (contact email: nhsaccess@channing.harvard.edu ) and https://sites.sph.harvard.edu/hpfs/for-collaborators/.

Data availability

Data will be shared at the request of qualified investigators for purposes of replicating procedures and results. Further information including the procedures to obtain and access data from the Nurses’ Health Studies and Health Professionals Follow-up Study is described at https://www.nurseshealthstudy.org/researchers (contact email: nhsaccess@channing.harvard.edu) and https://sites.sph.harvard.edu/hpfs/for-collaborators/.

References

  • 1.Timmis A, Vardas P, Townsend N, et al. European Society of Cardiology: cardiovascular disease statistics 2021. European heart journal 2022;43(8):716–799. [DOI] [PubMed] [Google Scholar]
  • 2.Sung H, Ferlay J, Siegel RL, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: a cancer journal for clinicians 2021;71(3):209–249. [DOI] [PubMed] [Google Scholar]
  • 3.Tsao CW, Aday AW, Almarzooq ZI, et al. Heart disease and stroke statistics—2023 update: a report from the American Heart Association. Circulation 2023;147(8):e93–e621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Afifi AM, Saad AM, Al-Husseini MJ, et al. Causes of death after breast cancer diagnosis: A US population-based analysis. Cancer 2020;126(7):1559–1567. [DOI] [PubMed] [Google Scholar]
  • 5.Zaorsky NG, Churilla T, Egleston B, et al. Causes of death among cancer patients. Annals of oncology 2017;28(2):400–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Adamo L, Rocha-Resende C, Prabhu SD, et al. Reappraising the role of inflammation in heart failure. Nature Reviews Cardiology 2020;17(5):269–285. [DOI] [PubMed] [Google Scholar]
  • 7.Panova-Noeva M, Schulz A, Arnold N, et al. Coagulation and inflammation in long-term cancer survivors: results from the adult population. Journal of Thrombosis and Haemostasis 2018;16(4):699–708. [DOI] [PubMed] [Google Scholar]
  • 8.Florido R, Daya NR, Ndumele CE, et al. Cardiovascular Disease Risk Among Cancer Survivors: The Atherosclerosis Risk In Communities (ARIC) Study. J Am Coll Cardiol 2022;80(1):22–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Strongman H, Gadd S, Matthews A, et al. Medium and long-term risks of specific cardiovascular diseases in survivors of 20 adult cancers: a population-based cohort study using multiple linked UK electronic health records databases. The Lancet 2019;394(10203):1041–1054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Armenian SH, Xu L, Ky B, et al. Cardiovascular Disease Among Survivors of Adult-Onset Cancer: A Community-Based Retrospective Cohort Study. J Clin Oncol 2016;34(10):1122–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Greenlee H, Iribarren C, Rana JS, et al. Risk of cardiovascular disease in women with and without breast cancer: the pathways heart study. Journal of Clinical Oncology 2022;40(15):1647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Matthews AA, Hinton SP, Stanway S, et al. Risk of cardiovascular diseases among older breast cancer survivors in the United States: a matched cohort study. Journal of the National Comprehensive Cancer Network 2021;19(3):275–284. [DOI] [PubMed] [Google Scholar]
  • 13.Matthews AA, Peacock Hinton S, Stanway S, et al. Risk of Cardiovascular Diseases Among Older Breast Cancer Survivors in the United States: A Matched Cohort Study. J Natl Compr Canc Netw 2021; 10.6004/jnccn.2020.7629:1–10. [DOI] [PubMed] [Google Scholar]
  • 14.O’Farrell S, Garmo H, Holmberg L, et al. Risk and timing of cardiovascular disease after androgen-deprivation therapy in men with prostate cancer. J Clin Oncol 2015;33(11):1243–51. [DOI] [PubMed] [Google Scholar]
  • 15.Arnett DK, Blumenthal RS, Albert MA, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation 2019;140(11):e596–e646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Shao C, Wang J, Tian J, et al. Coronary artery disease: from mechanism to clinical practice. Coronary Artery Disease: Therapeutics and Drug Discovery 2020:1–36. [DOI] [PubMed] [Google Scholar]
  • 17.Vaikunth SS, Lui GK. Heart failure with reduced and preserved ejection fraction in adult congenital heart disease. Heart Failure Reviews 2020;25(4):569–581. [DOI] [PubMed] [Google Scholar]
  • 18.Larsen CM, Arango MG, Dasari H, et al. Association of anthracycline with heart failure in patients treated for breast cancer or lymphoma, 1985–2010. JAMA network open 2023;6(2):e2254669–e2254669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Lyon AR, López-Fernández T, Couch LS, et al. 2022 ESC Guidelines on cardio-oncology developed in collaboration with the European Hematology Association (EHA), the European Society for Therapeutic Radiology and Oncology (ESTRO) and the International Cardio-Oncology Society (IC-OS) Developed by the task force on cardio-oncology of the European Society of Cardiology (ESC). European Heart Journal-Cardiovascular Imaging 2022;23(10):e333–e465. [DOI] [PubMed] [Google Scholar]
  • 20.Abdel-Qadir H, Thavendiranathan P, Austin PC, et al. The Risk of Heart Failure and Other Cardiovascular Hospitalizations After Early Stage Breast Cancer: A Matched Cohort Study. J Natl Cancer Inst 2019;111(8):854–862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bao Y, Bertoia ML, Lenart EB, et al. Origin, methods, and evolution of the three Nurses’ Health Studies. American journal of public health 2016;106(9):1573–1581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Rimm EB, Giovannucci EL, Willett WC, et al. Prospective study of alcohol consumption and risk of coronary disease in men. The Lancet 1991;338(8765):464–468. [DOI] [PubMed] [Google Scholar]
  • 23.Wang Q-L, Babic A, Rosenthal MH, et al. Cancer Diagnoses After Recent Weight Loss. JAMA 2024;331(4):318–328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Glenn AJ, Guasch-Ferré M, Malik VS, et al. Portfolio Diet Score and Risk of Cardiovascular Disease: Findings From 3 Prospective Cohort Studies. Circulation 2023;148(22):1750–1763. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Mendis S, Thygesen K, Kuulasmaa K, et al. World Health Organization definition of myocardial infarction: 2008–09 revision. Int J Epidemiol 2011;40(1):139–46. [DOI] [PubMed] [Google Scholar]
  • 26.Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly predict risk of chronic disease. The Journal of nutrition 2012;142(6):1009–1018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Rimm EB, Giovannucci EL, Stampfer MJ, et al. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. American journal of epidemiology 1992;135(10):1114–1126. [DOI] [PubMed] [Google Scholar]
  • 28.Song M, Zhou X, Pazaris M, et al. The missing covariate indicator method is nearly valid almost always. arXiv preprint arXiv:2111.00138 2021. [Google Scholar]
  • 29.Suissa S, Dell’Aniello S. Time-related biases in pharmacoepidemiology. Pharmacoepidemiology and drug safety 2020;29(9):1101–1110. [DOI] [PubMed] [Google Scholar]
  • 30.Selvaraj S, Bhatt DL, Claggett B, et al. Lack of Association Between Heart Failure and Incident Cancer. J Am Coll Cardiol 2018;71(14):1501–1510. [DOI] [PubMed] [Google Scholar]
  • 31.Van Nimwegen FA, Schaapveld M, Janus CP, et al. Cardiovascular disease after Hodgkin lymphoma treatment: 40-year disease risk. JAMA internal medicine 2015;175(6):1007–1017. [DOI] [PubMed] [Google Scholar]
  • 32.Tan MC, Yeo YH, Ibrahim R, et al. Trends and Disparities in Cardiovascular Death in Non-Hodgkin Lymphoma. American Journal of Cardiology 2024;210:276–278. [DOI] [PubMed] [Google Scholar]
  • 33.Shin DW, Han K, Park HS, et al. Risk of Ischemic Heart Disease and Stroke in Prostate Cancer Survivors: A Nationwide Study in South Korea. Sci Rep 2020;10(1):10313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Rugbjerg K, Mellemkjær L, Boice JD, et al. Cardiovascular disease in survivors of adolescent and young adult cancer: a Danish cohort study, 1943–2009. Journal of the National Cancer Institute 2014;106(6):dju110. [DOI] [PubMed] [Google Scholar]
  • 35.Jordan JH, Thwin SS, Lash TL, et al. Incident comorbidities and all-cause mortality among 5-year survivors of Stage I and II breast cancer diagnosed at age 65 or older: a prospective-matched cohort study. Breast cancer research and treatment 2014;146:401–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Ramin C, Schaeffer ML, Zheng Z, et al. All-Cause and Cardiovascular Disease Mortality Among Breast Cancer Survivors in CLUE II, a Long-Standing Community-Based Cohort. J Natl Cancer Inst 2021;113(2):137–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Khosrow-Khavar F, Filion KB, Al-Qurashi S, et al. Cardiotoxicity of aromatase inhibitors and tamoxifen in postmenopausal women with breast cancer: a systematic review and meta-analysis of randomized controlled trials. Ann Oncol 2017;28(3):487–496. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Grainger DJ, Schofield PM. Tamoxifen for the prevention of myocardial infarction in humans: preclinical and early clinical evidence. Circulation 2005;112(19):3018–3024. [DOI] [PubMed] [Google Scholar]
  • 39.Chlebowski RT, Anderson GL. Changing concepts: menopausal hormone therapy and breast cancer. Journal of the National Cancer Institute 2012;104(7):517–527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Cheong HS, Chang Y, Kim Y, et al. Human papillomavirus infection and cardiovascular mortality: a cohort study. European Heart Journal 2024:ehae020. [DOI] [PubMed] [Google Scholar]
  • 41.Arnett DK, Blumenthal RS, Albert MA, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Journal of the American College of cardiology 2019;74(10):e177–e232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Visseren FL, Mach F, Smulders YM, et al. 2021 ESC Guidelines on cardiovascular disease prevention in clinical practice: Developed by the Task Force for cardiovascular disease prevention in clinical practice with representatives of the European Society of Cardiology and 12 medical societies With the special contribution of the European Association of Preventive Cardiology (EAPC). European heart journal 2021;42(34):3227–3337.34458905 [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS2095411-supplement-Supplementary_Material.pdf (3.8MB, pdf)

Data Availability Statement

Data will be shared at the request of qualified investigators for purposes of replicating procedures and results. Further information including the procedures to obtain and access data from the Nurses’ Health Studies and Health Professionals Follow-up Study is described at https://www.nurseshealthstudy.org/researchers (contact email: nhsaccess@channing.harvard.edu) and https://sites.sph.harvard.edu/hpfs/for-collaborators/.

RESOURCES