Associations of Oral Contraceptive Use With Cardiovascular Disease and All-Cause Death: Evidence From the UK Biobank Cohort Study

Abstract

Background The associations of oral contraceptive (OC) use with cardiovascular disease (CVD) and all-cause death remains unclear. We aimed to determine the associations of OC use with incident CVD and all-cause death. Methods and Results This cohort study included 161 017 women who had no CVD at baseline and reported their OC use. We divided OC use into ever use and never use. Cox proportional hazard models were used to calculate hazard ratios and 95% CIs for cardiovascular outcomes and death. Overall, 131 131 (81.4%) of 161 017 participants reported OC use at baseline. The multivariable-adjusted hazard ratios for OC ever users versus never users were 0.92 (95% CI, 0.86-0.99) for all-cause death, 0.91 (95% CI, 0.87-0.96) for incident CVD events, 0.88 (95% CI, 0.81-0.95) for coronary heart disease, 0.87 (95% CI, 0.76-0.99) for heart failure, and 0.92 (95% CI, 0.84-0.99) for atrial fibrillation. However, no significant associations of OC use with CVD death, myocardial infarction, or stroke were observed. Furthermore, the associations of OC use with CVD events were stronger among participants with longer durations of use (P for trend<0.001). Conclusions OC use was not associated with an increased risk of CVD events and all-cause death in women and may even produce an apparent net benefit. In addition, the beneficial effects appeared to be more apparent in participants with longer durations of use.

Keywords: all‐cause death; cardiovascular disease; oral contraceptive use; women's health.

Figure 1. Subgroup analysis for the associations between OC use and risk of CVD events (A), coronary heart disease (B), all‐cause death (C), and CVD death (D).

Analyses were conducted with a multiple imputation approach for missing data. OC never users are included as referent (1.00). Hazard ratios were adjusted for age, academic degree (university/college degree or others), ethnicity (White or others), baseline systolic blood pressure, baseline diabetes (yes or no), baseline total cholesterol level, age at menarche, parity (0, 1, or ≥2), family history of cardiovascular diseases or stroke (yes or no), menopause status (yes, no, hysterectomy, or other reason), aspirin use (yes or no), lipid‐lowering medication (yes or no), antihypertensive drug (yes or no), smoking status (never, past, or current), drinking status (never, past, or current), BMI, and metabolic equivalent. BMI indicates body mass index; CRP, C‐reactive protein; CVD, cardiovascular disease; and OC, oral contraceptive.

Figure 2. Subgroup analysis for the associations between OC use and risk of myocardial infarction (A), heart failure (B), atrial fibrillation (C), and stroke (D).

Analyses were conducted with a multiple imputation approach for missing data. OC never users are included as referent (1.00). Hazard ratios were adjusted for age, academic degree (university/college degree or others), ethnicity (White or others), baseline systolic blood pressure, baseline diabetes (yes or no), baseline total cholesterol level, age at menarche, parity (0, 1, or ≥2), family history of cardiovascular diseases or stroke (yes or no), menopause status (yes, no, hysterectomy, or other reason), aspirin use (yes or no), lipid‐lowering medication (yes or no), antihypertensive drug (yes or no), smoking status (never, past, or current), drinking status (never, past, or current), BMI, and metabolic equivalent. BMI indicates body mass index; CRP, C‐reactive protein; and OC, oral contraceptive.