Subject characteristics

A total of 2,430,851 records of hospitalized women were found to be eligible for statistical analysis. The mean (standard deviation) age of the patient hospitalizations was 44.4 (16.0) years. Over half of the hospitalizations were non-Hispanic white followed by non-Hispanic black (17.4%) and Hispanic (13.4%). One-third of the hospitalized women (31.3%) had household income in the first quartile and primary payers were mostly private (39.4%), Medicaid (29.6%), or Medicare (23.6%). A limited proportion of hospitalized women (0.96%) were smokers with 3.4% alcohol users, 19.8% obese, and 8.9% with MS (Table 1). In the total cohort, 0.78% of hospitalizations (n = 18,926) had a diagnosis of HMB which included 15,180 (0.63%) hospitalizations without IM and 3746 (0.15%) with IM. Age was found to be a strong modifier for most CVD outcomes including diabetes mellitus (DM) (Additional file 1: Table S1) and thus results are presented for different age groups, separately.

Table 1 Sample characteristics of HMB among hospitalized women of age 18-70 years, NIS-2017

All the baseline characteristics were found to be different between HMB and NMC groups except for smoking status. HMB was strongly associated with age, black race, and private insurance. The proportion of obesity (23.1% vs. 19.8%, p < 0.001), contraceptive use (1.03% vs. 0.33%, p < 0.001), PCOS (1.64% vs. 0.55%, p < 0.001), infertility (0.52% vs. 0.04%, p < 0.001), anemia (15.7% vs. 8.5%, p < 0.001), NSAIDs (1.09% vs 0.71% p < 0.001), and leiomyoma uterus (51.5% vs. 1.3%, p < 0.001) were significantly higher in HMB hospitalizations than in NMC group whereas MS, individual components of MS, insulin use, IBD, and DM were significantly lower in HMB compared to NMC (Table 1). However, HMB hospitalizations with age ≤ 40 years were associated with all the risk factors including DM, MS, insulin use, IBD, and infertility (Additional file 1: Table S2). The proportion of MACE was found highest (10.2%, n = 247,464) after DM (19.3%, n = 468,304) in our selected cohort, and the lowest proportion was observed for stroke (1.5%, n = 36,004) and myocardial infarction (1.6%, n = 39,899). All the CVD outcomes were significantly higher in proportions in the HMB group compared to the NMC group among hospitalizations of women with age ≤ 40 years (Additional file 1: Table S3).

Unadjusted and propensity scores-matched associations of HMB diagnosis with CVD outcomes among hospitalizations of women with age ≤ 40 years

In the unadjusted analyses, the odds of MACE, CHD, HF, AF or arrhythmia, and MI were twofold higher in HMB hospitalizations compared to hospitalizations without menstrual disorders. A 70% or higher odds of stroke/ CVA and DM were also observed in HMB hospitalizations compared to hospitalizations with NMC (Fig. 1). In propensity scores-matched analysis, HMB was significantly associated with increased odds of CVD outcomes including MACE (OR = 1.61; 95% CI: 1.25, 2.08, p < 0.001), CHD (OR = 1.72; 95% CI: 1.10, 2.71, p = 0.019), stroke/CVA (OR = 1.95; 95% CI: 1.12, 3.40, p = 0.018), HF (OR = 1.53; 95% CI: 1.15, 2.03, p = 0.003), and AF/arrhythmia (OR = 1.84; 95% CI: 1.34, 2.54, p < 0.001) (Fig. 1).

Fig. 1

Unadjusted and adjusted associations of HMB with CVD outcomes including diabetes among hospitalized women of ages 18–40 years. OR, Odds ratio; CI, Confidence interval; CVD, Cardiovascular disease; MACE, Major adverse cardiovascular event; CHD, coronary heart disease; CVA, Cerebrovascular accident; HF, Heart failure; AF, Atrial fibrillation; MI, Myocardial infarction; DM, Diabetes mellitus; NMC, Normal menstrual cycle; HMB, Heavy menstrual bleeding. MACE was defined as the composite of myocardial infarction, stroke, and heart failure. The adjusted associations were carried out using propensity scores-matched analysis. The propensity scores model included race/ethnicity, household income quartile, primary payer, smoking, alcohol, contraceptive use, metabolic syndrome, non-steroidal anti-inflammatory drugs, leiomyoma uterus, obesity, polycystic ovary syndrome, anemia, and anticoagulants. Adjusted model additionally adjusted for age

Sensitivity analyses-adjusted association of HMB diagnosis with CVD outcomes among hospitalizations of women with age ≤ 40 years

After adjusting for age, race/ethnicity, household income quartile, primary payer, smoking, alcohol, contraceptive use, MS, NSAIDs, and leiomyoma uterus, all the outcomes including MACE, CHD, HF, AF, and DM were found to be significantly associated with HMB except for stroke/CVA and MI. The relative risk models also yielded the same pattern of effect sizes in the adjusted analyses (Additional file 1: Table S4). The adjusted association of HMB with each CVD outcome remained unchanged after additionally adjusting for fasting glucose, hypertension, high triglycerides, low HDL, obesity, pre-diabetes, DM, parity, anemia, and IBD (Additional file 1: Table S4). The associations obtained in primary propensity scores-matched analyses were unchanged after excluding hospitalization records with PCOS, leiomyoma uterus, or congenital heart disease (Additional file 1: Table S5) or removing anticoagulation use, NSAID use, or lifestyle cofactors (Additional file 1: Table S6) from the analyses. The associations were more pronounced after removing smoking/alcohol use and obesity (Additional file 1: Table S6).

Adjusted direct and indirect associations of HMB diagnosis with CVD outcome through mediators among hospitalizations of women aged ≤ 40 years

In the mediation analysis, the HMB was directly associated with MACE after accounting for mediating association of MS (OR = 1.45; 95%CI: 1.27, 1.66), obesity (OR = 1.41; 95%CI: 1.18, 1.67), hypertension (OR = 1.37; 95%CI: 1.17, 1.62), DM (OR = 1.46; 95%CI: 1.23, 1.74), and anemia (OR = 1.47; 95%CI: 1.24, 1.74). The indirect association of HMB with MACE was less pronounced compared to the direct association of HMB with MACE after adjusting for age, race/ethnicity, household income quartile, primary payer, smoking, alcohol, contraceptive use, MS, NSAIDs, and leiomyoma uterus (Additional file 1: Table S7).

Associations of all covariates with CVD outcome among hospitalizations of women aged ≤ 40 years

None of the lifestyle covariates had a significant interaction with HMB on most CVD outcomes (Additional file 1: Table S8). Most of the considered covariates were found to be associated with MACE except for infertility, IBD, and NSAIDs. In the adjusted analysis, anticoagulant use (OR = 5.31; 95%CI: 4.91, 5.74), black race/ethnicity (OR = 2.05; 95%CI: 1.94, 2.16), insulin use (OR = 2.53; 95%CI: 2.37, 2.71), contraceptive/hormone use (OR = 1.85; 95%CI: 1.46, 2.35), obesity (OR = 1.84; 95%CI: 1.76, 1.93), MS(OR = 1.80; 95%CI: 1.66, 1.96), smoking use (OR = 1.65; 95%CI: 1.42, 1.91), anemia (OR = 1.30; 95%CI: 1.22, 1.37), and alcohol use (OR = 1.11; 95%CI: 1.00, 1.23) were associated with increased odds of MACE events in addition to HMB (OR = 1.30; 95%CI: 1.10, 1.54) (Additional file 1: Table S9).

Unadjusted and adjusted associations of HMB with or without IM with CVD outcomes among hospitalizations of women with age ≤ 40 years

In the unadjusted analyses, HMB with or without IM was associated with all CVD outcomes except for MI. In the propensity scores-matched analyses, HMB without IM was associated with MACE (OR = 1.69; 95%CI: 1.30, 2.19, p < 0.001), CHD (OR = 1.67; 95%CI: 1.05, 2.66, p = 0.031), stroke/CVA (OR = 1.96; 95%CI: 1.11, 3.47, p = 0.021), HF (OR = 1.60; 95%CI: 1.11, 3.47, p = 0.002), and AF (OR = 1.84; 95%CI: 1.31, 2.57, p < 0.001). However, HMB with IM was only associated with CHD (OR = 1.95; 95%CI: 1.10, 3.48, p = 0.023) and AF (OR = 1.87; 95%CI: 1.17, 2.9, p = 0.009) (Table 2). These associations other than stroke/CVA remained consistent in adjusted regression models or after additionally adjusting for DM (Additional file 1: Table S10).

Table 2 Unadjusted adjusted associations of HMB with or without IM and CVD events including diabetes among hospitalized women of ages between 18-40 yearsUnadjusted and adjusted associations of HMB diagnosis with CVD outcomes among hospitalizations of women with age > 40 years

In the unadjusted analyses, HMB was inversely associated with all CVD outcomes. However, HMB was only found positively associated with MACE (OR = 1.19; 95% CI: 1.01, 1.42, p = 0.032), HF (OR = 1.19; 95% CI: 1.01, 1.40, p = 0.04), and MI (OR = 1.46; 95% CI: 1.04, 2.04, p = 0.027) in the propensity scores-matched analysis (Additional file 1: Table S11). These findings remained significant among hospitalizations of women aged > 55 years but not in those aged between 40 to 55 years. HMB was associated with AF/ Arrhythmia (OR = 1.96; 95%CI: 1.24, 3.09) only in ages between 40 to 55 years while HMB was also significantly associated with DM (OR = 1.12; 95%CI: 1.01, 1.24) among hospitalizations of women aged > 55 years (Additional file 1: Table S12).