Publications by Year: 2026

INTRODUCTION: Cardiovascular health (CVH) often declines in midlife. Diet during pregnancy, a similarly sensitive period involving physiological changes, may influence long-term CVH, but few studies have examined this relationship. We examined the associations between diet during pregnancy and CVH in midlife.

METHODS: Participants are 776 women enroled in Project Viva during early pregnancy (1999-2002, mean 32.7 years). Participants completed food frequency questionnaires during the first and second trimesters, from which we calculated three diet scores (higher scores = better diet): Dietary Approaches to Stop Hypertension (DASH), Mediterranean Diet (MD) and Dietary Inflammatory Index (DII × -1). We averaged scores across trimesters and generated internal z-scores. At the midlife visit (2017-2021, mean 51.1 y), we assessed CVH using the American Heart Association's Life's Essential 8 (range 0-100; higher scores = better CVH), including behavioural (diet, physical activity, sleep, smoking) and biomedical (BMI, blood pressure, lipids, glucose) subscores. We assessed associations using linear regression adjusted for age, race and ethnicity, nativity, education, parity, marital status and income at enrolment.

RESULTS: Most participants were non-Hispanic White (70%) and college graduates (76%). Mean midlife CVH score was 76.3 points. Higher (per 1-SD) DASH z-scores were associated with better overall (β 3.3; 95% CI 2.2-4.3), behavioural (3.5; 2.5-4.4) and biomedical (3.1; 1.5-4.7) CVH scores. Higher MD z-scores were associated with better overall CVH (2.4; 1.4-3.4), behavioural (2.6; 1.6-3.5) and biomedical (2.6; 1.0-4.2) scores. Higher DII (× -1) z-scores were associated with better overall CVH (2.6; 1.6-3.7), behavioural (2.7; 1.7-3.7) and biomedical (2.5; 0.9-4.1) scores. A better pregnancy diet was associated with higher CVH scores for diet, physical activity and BMI, and both DASH and DII with blood pressure.

CONCLUSIONS: Women who consumed healthier diets during pregnancy had better CVH two decades later. Dietary assessment in pregnancy may identify women at risk for both short-term and long-term health problems.

BACKGROUND: Orthostatic hypotension (OH) is common among older adults actively prescribed hypertension pharmacotherapy. Associations between antihypertensives and OH are inconsistent, due to differences in drug classes and OH assessment protocols.

METHODS: Three supine and six standing blood pressures (BPs) were measured over 6 minutes, timed to initiate immediately (0 minutes) and 1, 2, 3, 4, and 5 minutes after standing. OH was defined as a drop of systolic/diastolic BP ≥ 20/10 mmHg upon standing from the supine position. OH phenotypes were initial (immediately after standing), early (within 3 minutes), late (after 3 minutes), and sustained (during 6 minutes of standing). Participants were asked about orthostatic symptoms experienced during OH assessment. We investigated associations of antihypertensive classes with OH phenotypes and orthostatic symptoms via multivariable logistic regression.

RESULTS: Of 863 participants (mean age 84 years), 41% had initial, 24% early, 18% late, and 17% sustained OH; 81% used antihypertensive medications. First-line antihypertensives (renin angiotensin aldosterone system blockers, dihydropyridine calcium channel blockers, and thiazide diuretics) were not associated with OH or orthostatic symptoms. Among non-first-line, alpha blockers were significantly associated with initial OH (OR 1.80, 95%CI 1.07, 3.01) and orthostatic symptoms during postural change (OR 2.64, 95%CI 1.34, 5.23), whereas combined alpha-beta blockers were significantly associated with early, late, or sustained OH (2.24 [1.15, 4.37], 2.50 [1.21, 5.17], 2.99 [1.41, 6.30], respectively) and experienced symptoms while standing (3.08 [1.10, 8.65]).

CONCLUSIONS: First-line antihypertensives are not associated with OH among older adults, which does not support their discontinuation in the setting of non-neurogenic OH.

BACKGROUND: Orthostatic hypotension (OH) and orthostatic hypertension (OHT) are common among adults with hypertension and may complicate antihypertensive treatment selection.However, the impact of specific antihypertensive agents on orthostatic blood pressure (BP) responses and risk of OH or OHT remains incompletely characterized.

METHODS: We performed a secondary analysis of the Treatment of Mild Hypertension Study (TOMHS), a multicenter randomized clinical trial (1987-1993). A total of 902 adults aged 45-69 years with diastolic BP 90-99 mmHg were randomized to acebutolol, amlodipine, chlorthalidone, doxazosin, enalapril, or placebo. Orthostatic BP change (standing minus seated BP) was assessed at randomization, 1 month, and every 3 months up to 54 months. OH was defined as a systolic BP decrease ≥20 mmHg or diastolic BP decrease ≥10 mmHg; OHT was defined as a systolic BP increase ≥20 mmHg or diastolic BP increase ≥10 mmHg. Generalized estimating equations evaluated associations between agents and orthostatic BP changes overall, by baseline BP, and over time.

RESULTS: Among 897 participants with orthostatic BP measurements (mean age 54.8 years; 38.5% female; 19.6% Black), doxazosin increased odds of OH (OR 1.45; 95% CI, 1.05-2.02) and reduced odds of OHT (OR 0.66; 95% CI, 0.51-0.86) versus placebo. Chlorthalidone reduced OHT among participants with baseline OHT (P=0.018). Amlodipine and doxazosin attenuated orthostatic systolic BP changes across baseline orthostatic BP change strata. Over time, amlodipine and chlorthalidone reduced odds of OH (P=0.001, 0.044).

CONCLUSIONS: Specific antihypertensive agents demonstrated distinct effects on orthostatic BP regulation. Consideration of orthostatic effects during antihypertensive selection may help individualize therapy, improve hemodynamic stability, and optimize treatment for patients at risk of orthostatic dysregulation.

OBJECTIVE: This study aimed to examine associations of gestational weight gain (GWG) and postpartum weight retention (PPWR) with women's mid-life cardiovascular health.

METHODS: We studied 838 Project Viva participants enrolled 1999-2002 (mean age: 32.6 years). We categorized GWG as inadequate, adequate, or excessive using 2009 Institute of Medicine guidelines and calculated PPWR at 6 months, 1 year, and 2 years. At the Mid-Life Visit (2017-2021; mean age: 51.0 years), we derived Life's Essential 8 (LE8) scores (0-100; higher = better). We used linear regression models adjusted for pre-pregnancy BMI, age at enrollment, pregnancy smoking status, household income, and educational attainment.

RESULTS: Excessive versus adequate GWG was associated with worse weight status (β = -8.99, 95% CI: -12.81, -5.18) but not with the overall LE8 score (β = -1.93, 95% CI: -3.92, 0.07) or the biomedical domain score (β = -2.31, 95% CI: -5.66, 1.05). Each 1-kg higher 1-year PPWR was associated with worse biomedical domain (β = -0.49, 95% CI: -0.84, -0.14), weight status (β = -1.54, 95% CI: -1.90, -1.17), and blood pressure scores (β = -0.62, 95% CI: -1.15, -0.10). PPWR at 6 months and 2 years was also associated with worse glycemic traits scores.

CONCLUSIONS: Excessive GWG was associated primarily with worse mid-life weight status, whereas higher PPWR was associated with worse mid-life cardiovascular health.

The 2025 AHA/ACC blood pressure (BP) management guidelines recommend BP treatment for 10-year cardiovascular disease (CVD) ≥7.5% from the PREVENT equation, replacing the prior pooled cohort equations (PCE). We examined associations between risk in PCE and PREVENT and a composite of CVD or cardiovascular death in SPRINT. Among 1794 participants, 477 (26.6%) were at low risk in both, 119 (6.6%) were high-risk in PREVENT only, 89 (5.0%) were high-risk in PCE only, and 1109 (61.8%) were high-risk in both. Those at high risk in both equations receiving intensive treatment had a lower risk of the primary outcome [hazard ratio (HR): 0.46, 95% confidence interval (CI): 0.24, 0.89]. Intensive treatment did not affect the risk of any outcomes for those at high-risk in PREVENT only or PCE only due to low event rates. These findings confirm the neutrality of the change from PCE to PREVENT for BP treatment recommendations.

BACKGROUND: Assessment and treatment of orthostatic hypotension (OH) are recommended to prevent falls in older adults. However, the optimal timing of standing blood pressure (BP) assessments to identify clinically relevant OH remains unclear.

METHODS: We measured 3 supine BP readings after 5 minutes of rest at 30-second intervals and 6 standing BP readings (timed at 0, 1, 2, 3, 4, and 5 minutes after standing) during the ARIC study (Atherosclerosis Risk in Communities) visit 10. OH was defined as a drop in BP (systolic ≥20 mm Hg or diastolic ≥10 mm Hg) on standing from the supine position. We quantified the prevalence and predictors of initial OH (immediately after standing), early OH (within 3 minutes), late OH (after 3 minutes), and sustained OH (all 6 minutes). We examined their associations with orthostatic symptoms and falls using logistic regression and negative binomial models, respectively.

RESULTS: Of the 863 participants (mean age, 83.7 years), 41.3% had initial OH. Participants on antihypertensive medications had higher odds of all types of OH irrespective of controlled (<130/80 mm Hg) or uncontrolled (≥130/80 mm Hg) hypertension compared with those without hypertension (<130/80 mm Hg) and no antihypertensive use. Initial and early OH were strongly associated with orthostatic symptoms in the process of standing from the supine position (odds ratio, 2.34 [95% CI, 1.42-3.86]; odds ratio, 1.82 [95% CI, 1.11-2.98], respectively). Finally, participants with initial OH had higher fall rates (rate ratio, 1.45 [95% CI, 1.02-2.06]) than those without OH.

CONCLUSIONS: Measuring BP within 3 minutes of standing, particularly the first BP measurement immediately after standing, may provide clinicians with important information on falls risk and orthostatic symptoms.

INTRODUCTION: Previous research has shown narrowing racial disparities in self-reported health. However, these disparities may have worsened in recent years, particularly during and after the COVID-19 pandemic. Less is known regarding the current state of disparities in self-reported health.

METHODS: Participants are from the 2023 and 2024 U.S. National Health Interview Survey. Using survey-weighted, age- and sex-adjusted modified Poisson regression models, we examined associations of race, ethnicity, and socioeconomic status with fair or poor self-reported health, both overall and stratified by levels of education, income, and health insurance status.

RESULTS: Among 251,930,438 adults aged ≥ 18 years (unweighted N = 60,449), 14.9% (95% CI: 14.5, 15.3) reported fair or poor health. Compared to those with a bachelor's degree or higher, those who had not completed high school had a higher prevalence of fair or poor health (prevalence ratio [PR] = 3.79, 95% CI: 3.52, 4.08). Additionally, compared to people with household income ≥ 4 times the federal poverty line, those with household incomes below the federal poverty line had a higher prevalence of fair or poor health (PR = 3.94, 95% CI: 3.70, 4.19). Non-Hispanic Black (PR = 1.56, 95% CI: 1.46, 1.67) and Hispanic (PR = 1.49, 95% CI: 1.39, 1.59) adults were more likely than non-Hispanic White adults to report fair or poor health.

CONCLUSION: In this U.S. nationally representative study, fair or poor self-reported health was more prevalent among adults with lower socioeconomic status and among non-Hispanic Black and Hispanic adults. These findings underscore the need for policy interventions that simultaneously address socioeconomic inequality and structural racism to improve population health outcomes.

BACKGROUND: Successful trial recruitment depends on balancing initial interest with participants' commitment and eligibility. It is uncertain how different motivational advertisement themes affect recruitment of interested and eligible trial participants.

METHODS: The GoFresh trials examined how healthy groceries affect blood pressure among Black adults from Boston healthy food priority areas. Using Facebook's "A/B testing" feature, users 18 years and older in eligible ZIP codes were randomly shown one of four digital advertisement themes-science volunteerism, study incentives, blood pressure reduction, or wellness in the Black community-between November 2023 and December 2024. We tracked Facebook and website analytics by theme. 'Interest' was defined as submission of an online interest form. 'Commitment' as visit no-show rates, and 'Eligibility' as qualification following the in-person visit. The 'number needed to screen' was calculated as 1 divided by the proportion of inquiries who were eligible.

RESULTS: Advertisements reached 603,418 Facebook users and generated 32,168 clicks, 264 inquiries, 73 scheduled visits, and 23 eligible participants. The community theme generated the highest inquiry engagement rate (54%) but lower commitment (8 no-shows) and the highest number needed to screen (17). The blood pressure theme had the lowest inquiry engagement rate (32%) but yielded the highest commitment (2 no-shows) and the lowest number needed to screen (8).

CONCLUSION: Beyond generating study interest, motivational themes may select for participants with varying degrees of commitment and eligibility, which could affect the cost-effectiveness of recruitment and long-term retention. These findings have important implications for timely and representative recruitment for cardiovascular trials.

REGISTRATION: URL: https://clinicaltrials.gov/; Unique Identifier: NCT05121337 &NCT05393232.

BACKGROUND: Current clinical guidelines recommend home blood pressure monitoring (HBPM) to confirm hypertension diagnosis, guide therapy, and support long-term control. However, the optimal number of HBPM measurements needed for valid assessment is not established.

METHODS: At ARIC Visit 10 (2023), participants completed an eight-day HBPM protocol with morning and evening measurements and three readings per session. We evaluated the validity between the eight-day mean (reference) and streamlined protocols (fewer days, less than three readings per session, or only morning or evening measurements). In secondary analyses, we assessed the validity of these HBPM protocols relative to ambulatory BP.

RESULTS: Among 812 participants (median age 83 years, 24% Black adults, 40% male, 84% using anti-hypertensive medications), median BP varied by less than 3 mmHg across the eight-day protocol. The greatest improvement in HBPM validity occurred when extending the averaging window from one to two days. By three days, concordance correlation coefficients exceeded 0.95, and ≥97% of participants were within 10 mmHg of the reference for both systolic and diastolic BP. Averaging ≥2 readings per session, including both morning and evening measurements, yielded greater validity than a single reading or timepoint.

CONCLUSIONS: In this community-based cohort of older adults, HBPM provided stable BP estimates within three days (can be non-consecutive within an 8-day period) using morning and evening sessions with ≥2 readings each. While further studies are required to validate generalizability beyond this engaged cohort, these findings support shorter, less burdensome HBPM protocols in clinical guidelines for hypertension management in older adults.