Hypertension, Cardiovascular Risk Factors, and Uterine Fibroid Diagnosis in Midlife

Key Points Question Are study-measured blood pressure, antihypertensive treatment, and cardiovascular risk factors (anthropometry and biomarkers) associated with incidence of fibroids? Findings In this cohort study of 2570 individuals at midlife, participants with untreated and new-onset hypertension had increased risk of newly self-reported fibroids, whereas those taking antihypertensive treatment had lower risk. Anthropometric factors and blood biomarkers were not associated with risk of newly reported fibroids. Meaning These findings motivate investigation into the mechanisms underlying fibroids and may lead to new strategies for their prevention.

CRP was measured using the Human High Sensitivity CRP ELISA (R&D Systems, DCRP00), a quantitative sandwich enzyme immunoassay.The LLD for this assay was 0.010 ng/ml (or 0.00001 mg/L).
All values above the LLD for the 3 original assays were calibrated to the high sensitivity ELISA by simultaneously assaying 600 paired samples (representing the full range of results from each of the original 3 assays) and the high sensitivity assay; thus, there were 200 paired samples for each original vs. high sensitivity ELISA.After calibration, correlations between the first 3 CRP assays and the fourth, high sensitivity ELISA were >0.94.
Total cholesterol.From baseline to visit 7, cholesterol was measured using an automated cholesterol oxidase assay on a Hitachi 747-200 clinical analyzer using RAICHEM reagents and calibrators with target values assigned by the Centers for Disease Control Lipid Section.At visits 9, 12, and 13 the cholesterol esters were hydrolyzed by cholesterol esterase to cholesterol and free fatty acids.The cholesterol was converted to cholest-4-en-3-one by cholesterol oxidase in the presence of oxygen to form hydrogen peroxide.A colored complex was formed from hydrogen peroxide, 4-aminoantipyrine and phenol under the catalytic influence of peroxidase, and the absorbance of the complex was measured at 505/694 nm.
LDL cholesterol was calculated at al visits using the Friedewald equation (total cholesterol -(triglycerides / 5) -HDL cholesterol) 1 when triglycerides were below 400 mg/dL.HDL cholesterol.From baseline to visit 7, HDL cholesterol was measured following precipitation of LDL and VLDL with heparin and manganese chloride by the modified Lipid Research Clinics procedure.
HDL cholesterol in the supernate was measured by an automated cholesterol oxidase assay on a Hitachi 747-200 clinical analyzer using RAICHEM reagents, a low cholesterol level calibrator with target values assigned by the Centers for Disease Control Lipid Section, and including controls for low level cholesterol in addition to controls for HDL cholesterol.At visits 9, 12, and 13 HDL cholesterol was measured without prior separation, based on procedures developed by Izawa, Okada and Matsui. 2 Cholesterol from non-HDL particles was released and eliminated in the first step of the reaction, and cholesterol in HDL particles was released in the second step by detergent in R2.The HDL cholesterol was measured by a Trinder reaction.
Triglycerides.From baseline to visit 7, triglycerides were measured using an automated glycerol kinase enzymatic assay on a Hitachi 747-200 clinical analyzer using Technicon reagents and with controls with target values assigned by the Centers for Disease Control Lipid Section.Glycerol blanks were performed for any samples with triglycerides greater than 300 mg/dL.At visits 9, 12, and 13, triglycerides were measured using the ADVIA assay method: triglycerides were converted to glycerol and free fatty acids by lipase, and the glycerol was then converted to glycerol-3-phosphate by glycerol kinase followed by its conversion by glycerol-3-phosphate-oxidase to hydrogen peroxide.A colored complex was formed from hydrogen peroxide, 4-aminophenazone and 4-chlorophenol under the catalytic influence of peroxidase.
The absorbance of the complex was measured as an endpoint reaction at 505/694 nm.At each study visit, discrete-time survival models tested the association between exposure and outcome, adjusting for visit-specific covariate levels, and pooled across visits.eTable 3. Longitudinal Discrete Survival Models Estimate the Risk of Newly Diagnosed Fibroids Associated With Each Cardiovascular Risk Factor in the Study of Women's Health Across the Nation (SWAN) Cohort.The "Symptoms and screening" column represents regression models that are identical to the Healthcare-adjusted model but restricted to participants who reported a pelvic exam, abnormal bleeding, or pelvic pain since the prior visit (63-80% of eligible participants per visit (average 70%)).1.01 (0.94, 1.09) 1.02 (0.92, 1.13) 1 Minimally-adjusted model adjusted for baseline value of the exposure, age, education, race/ethnicity, site; timevarying smoking status, menopausal status, parity, body mass index.Healthcare-adjusted model additionally adjusted for time-varying times spoken to a healthcare provider, multivitamins, calcium supplements, mammogram since prior visit.Blood pressure models additionally adjusted for blood pressure medication; cholesterol, triglyceride models additionally adjusted for statins. 2 Among participants eligible to use blood pressure medication (n=624 at baseline).

Characteristic
3 Hypertension status: "Never hypertensive" group has no history of antihypertensive use or measured hypertension; "pre-existing hypertension" group previously reported antihypertensive use or had measured hypertension; "new-onset hypertension" group reported antihypertensive use or had measured hypertension for the first time at the current visit.eTable 5. Longitudinal Discrete Survival Models Estimate the Risk of Newly Diagnosed Fibroids Associated With Each Cardiovascular Risk Factor in the Study of Women's Health Across the Nation (SWAN) Cohort.The "Includes post-menopause" column represents regression models that are identical to the Healthcare-adjusted model but without censoring at menopause (approximately 44% of observations were censored due to menopause in main models, ranging from 2% (visit 1) to 99% (visit 13)).
3 Hypertension status: "Never hypertensive" group has no history of antihypertensive use or measured hypertension; "pre-existing hypertension" group previously reported antihypertensive use or had measured hypertension; "new-onset hypertension" group reported antihypertensive use or had measured hypertension for the first time at the current visit.

eReferences 1 .
Friedewald WT, Levy RI, Fredrickson DS.Estimation of the concentration of low-densiy lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.Clinical Chemistry.1972;18(6):499-502.2. Izawa S, Okada M, Matsui H, Horita Y.A new direct method for measuring HDL-cholesterol which does not produce any biased values.J Med Pharm Sci 1997;37:1385-1388.eFigure 1. Study Timeline And Schematic.The eligible population at each visit (i.e., pre-and perimenopausal without a prior fibroid diagnosis) is depicted in light grey squares.The number of eligible participants reporting a fibroid diagnosis at each visit is shown to the left.The number of participants eligible to receive antihypertensive medication is shown to the right.

eFigure 2 .
Directed Acyclic Graph (DAG) Showing the Relationships Between Each Confounder, the Exposure (in This Example, Blood Pressure) and the Outcome (Self-Reported Fibroid Diagnosis).The exposure, outcome, and some confounders update at each study visit.

Fibroids at Baseline Baseline characteristics Ever had a fibroid at baseline (n=670) Never had a fibroid at baseline (n=2570)
Characteristics of Participants Eligible to Receive Antihypertensive Treatment (n=624) Longitudinal Discrete Survival Models Estimate the Risk of Newly Diagnosed Fibroids Associated With Each Cardiovascular Risk Factor in the Study of Women's Health Across the Nation (SWAN) Cohort.The "No fill" column represents regression models that are identical to the Healthcare-adjusted model but without carrying forward observations to fill random or structural missings (on average, 24% to 36% of values in main models were carried forward values, varying by cardiovascular exposure).
Longitudinal Discrete Survival Models Estimate the Risk of Newly Diagnosed Fibroids Associated With Each Cardiovascular Risk Factor in the Study of Women's Health Across the Nation (SWAN) Cohort.The "Not adjusted for BL" column represents regression models that are identical to the Healthcareadjusted model except that they are not adjusted for baseline values of the exposure.