Incorporating longitudinal history of risk factors into atherosclerotic cardiovascular disease risk prediction using deep learning.

It is increasingly clear that longitudinal risk factor levels and trajectories are related to risk for atherosclerotic cardiovascular disease (ASCVD) above and beyond single measures. Currently used in clinical care, the Pooled Cohort Equations (PCE) are based on regression methods that predict ASCVD risk based on cross-sectional risk factor levels. Deep learning (DL) models have been developed to incorporate longitudinal data for risk prediction but its benefit for ASCVD risk prediction relative to the traditional Pooled Cohort Equations (PCE) remain unknown. Our study included 15,565 participants from four cardiovascular disease cohorts free of baseline ASCVD who were followed for adjudicated ASCVD. Ten-year ASCVD risk was calculated in the training set using our benchmark, the PCE, and a longitudinal DL model, Dynamic-DeepHit. Predictors included those incorporated in the PCE: sex, race, age, total cholesterol, high density lipid cholesterol, systolic and diastolic blood pressure, diabetes, hypertension treatment and smoking. The discrimination and calibration performance of the two models were evaluated in an overall hold-out testing dataset. Of the 15,565 participants in our dataset, 2170 (13.9%) developed ASCVD. The performance of the longitudinal DL model that incorporated 8 years of longitudinal risk factor data improved upon that of the PCE [AUROC: 0.815 (CI 0.782-0.844) vs 0.792 (CI 0.760-0.825)] and the net reclassification index was 0.385. The brier score for the DL model was 0.0514 compared with 0.0542 in the PCE. Incorporating longitudinal risk factors in ASCVD risk prediction using DL can improve model discrimination and calibration.

A Proposed Pediatric Clinical Cardiovascular Health Reference Standard.

INTRODUCTION: Clinical cardiovascular health is a construct that includes 4 health factors-systolic and diastolic blood pressure, fasting glucose, total cholesterol, and body mass index-which together provide an evidence-based, more holistic view of cardiovascular health risk in adults than each component separately. Currently, no pediatric version of this construct exists. This study sought to develop sex-specific charts of clinical cardiovascular health for age to describe current patterns of clinical cardiovascular health throughout childhood. METHODS: Data were used from children and adolescents aged 8-19 years in six pooled childhood cohorts (19,261 participants, collected between 1972 and 2010) to create reference standards for fasting glucose and total cholesterol. Using the models for glucose and cholesterol as well as previously published reference standards for body mass index and blood pressure, clinical cardiovascular health charts were developed. All models were estimated using sex-specific random-effects linear regression, and modeling was performed during 2020-2022. RESULTS: Models were created to generate charts with smoothed means, percentiles, and standard deviations of clinical cardiovascular health for each year of childhood. For example, a 10-year-old girl with a body mass index of 16 kg/m2 (30th percentile), blood pressure of 100/60 mm Hg (46th/50th), glucose of 80 mg/dL (31st), and total cholesterol of 160 mg/dL (46th) (lower implies better) would have a clinical cardiovascular health percentile of 62 (higher implies better). CONCLUSIONS: Clinical cardiovascular health charts based on pediatric data offer a standardized approach to express clinical cardiovascular health as an age- and sex-standardized percentile for clinicians to assess cardiovascular health in childhood to consider preventive approaches at early ages and proactively optimize lifetime trajectories of cardiovascular health.

US County-Level Variation in Preterm Birth Rates, 2007-2019.

Importance: Preterm birth is a leading cause of preventable neonatal morbidity and mortality. Preterm birth rates at the national level may mask important geographic variation in rates and trends at the county level. Objective: To estimate age-standardized preterm birth rates by US county from 2007 to 2019. Design, Setting, and Participants: This serial cross-sectional study used data from the National Center for Health Statistics composed of all live births in the US between 2007 and 2019. Data analyses were performed between March 22, 2022, and September 29, 2022. Main Outcomes and Measures: Age-standardized preterm birth (<37 weeks' gestation) and secondarily early preterm birth (<34 weeks' gestation) rates by county and year calculated with a validated small area estimation model (hierarchical bayesian spatiotemporal model) and percent change in preterm birth rates using log-linear regression models. Results: Between 2007 and 2019, there were 51 044 482 live births in 2383 counties. In 2007, the national age-standardized preterm birth rate was 12.6 (95% CI, 12.6-12.7) per 100 live births. Preterm birth rates varied significantly among counties, with an absolute difference between the 90th and 10th percentile counties of 6.4 (95% CI, 6.2-6.7). The gap between the highest and lowest counties for preterm births was 20.7 per 100 live births in 2007. Several counties in the Southeast consistently had the highest preterm birth rates compared with counties in California and New England, which had the lowest preterm birth rates. Although there was no statistically significant change in preterm birth rates between 2007 and 2019 at the national level (percent change, -5.0%; 95% CI, -10.7% to 0.9%), increases occurred in 15.4% (95% CI, 14.1%-16.9%) of counties. The absolute and relative geographic inequalities were similar across all maternal age groups. Higher quartile of the Social Vulnerability Index was associated with higher preterm birth rates (quartile 4 vs quartile 1 risk ratio, 1.34; 95% CI, 1.31-1.36), which persisted across the study period. Similar patterns were observed for early preterm birth rates. Conclusions and Relevance: In this serial cross-sectional study of county-level preterm and early preterm birth rates, substantial geographic disparities were observed, which were associated with place-based social disadvantage. Stability in aggregated rates of preterm birth at the national level masked increases in nearly 1 in 6 counties between 2007 and 2019.

Incorporating longitudinal history of risk factors into atherosclerotic cardiovascular disease risk prediction using deep learning.

Background: It is increasingly clear that longitudinal risk factor levels and trajectories are related to risk for atherosclerotic cardiovascular disease (ASCVD) above and beyond single measures. Currently used in clinical care, the Pooled Cohort Equations (PCE) are based on regression methods that predict ASCVD risk based on cross-sectional risk factor levels. Deep learning (DL) models have been developed to incorporate longitudinal data for risk prediction but its benefit for ASCVD risk prediction relative to the traditional Pooled Cohort Equations (PCE) remain unknown. Objective: To develop a ASCVD risk prediction model that incorporates longitudinal risk factors using deep learning. Methods: Our study included 15,565 participants from four cardiovascular disease cohorts free of baseline ASCVD who were followed for adjudicated ASCVD. Ten-year ASCVD risk was calculated in the training set using our benchmark, the PCE, and a longitudinal DL model, Dynamic-DeepHit. Predictors included those incorporated in the PCE: sex, race, age, total cholesterol, high density lipid cholesterol, systolic and diastolic blood pressure, diabetes, hypertension treatment and smoking. The discrimination and calibration performance of the two models were evaluated in an overall hold-out testing dataset. Results: Of the 15,565 participants in our dataset, 2,170 (13.9%) developed ASCVD. The performance of the longitudinal DL model that incorporated 8 years of longitudinal risk factor data improved upon that of the PCE [AUROC: 0.815 (CI: 0.782-0.844) vs 0.792 (CI: 0.760-0.825)] and the net reclassification index was 0.385. The brier score for the DL model was 0.0514 compared with 0.0542 in the PCE. Conclusion: Incorporating longitudinal risk factors in ASCVD risk prediction using DL can improve model discrimination and calibration.

Race- and Sex-Specific Factors Associated With Age-Related Slopes in Systolic Blood Pressure: Findings From the CARDIA Study.

BACKGROUND: Although blood pressure (BP) increases throughout young adulthood for most individuals, the age-related slope is not uniform. This study aimed to assess associations of demographic, clinical, behavioral, psychosocial, and neighborhood characteristics with age-related BP slope among 4 race-sex groups who participated in the Coronary Artery Risk Development in Young Adults study. METHODS: Individuals (n=3554) aged 18 to 30 years were included in this analysis if they had normal BP at baseline and ≥2 BP measurements during the years 1985/1986 to 2015/2016. Associations of exposure variables with systolic BP slope were assessed using multivariate linear models. RESULTS: Over a mean follow-up of ~30 years, greater decade increases in systolic BP were estimated among Black than White participants (mean difference between Black females and White females: 3.0 mm Hg/decade; between Black males and White males: 4.7 mm Hg/decade). The exposure risk factors associated with greater increases in systolic BP throughout adulthood varied by race and sex. None of these factors were associated with increases in systolic BP in all race-sex groups. Parent history of high BP was associated with a steeper positive slope among Black females (effect size per decade: 1.1 [95% CI, 0.6-1.6]; P<0.01), Black males (0.6 [95% CI, 0.02-1.2]; P<0.05), and White females (0.6 [95% CI, 0.2-1.0]; P<0.01). Other risk factors were associated with greater age-related yearly increases in systolic BP among 1 or 2 of the 4 race-sex groups or were not statistically significant. CONCLUSIONS: Culturally tailored BP reduction approach should be considered in conjunction with primordial prevention, to moderate increases in BP throughout adulthood.

The 2017 American College of Cardiology/American Heart Association Hypertension Guideline and Blood Pressure in Older Adults.

INTRODUCTION: The 2017 American College of Cardiology/American Heart Association blood pressure guideline redefined hypertension and lowered the blood pressure treatment target. Empirical data on the guideline's impact are needed. METHODS: Data were analyzed from Atherosclerosis Risk in Communities study participants who attended baseline pre-guideline (2016-2017) and post-guideline (2018-2019) visits with baseline systolic blood pressure between 120 and 159 mmHg. Participants were grouped according to baseline systolic blood pressure by change in classification under the new guideline as follows: not reclassified (120-129 mmHg), reclassified to Stage 1 hypertension (130-139 mmHg), and reclassified to Stage 2 hypertension (140-159 mmHg). Means and 95% CIs for systolic blood pressure changes between baseline and follow-up, changes in antihypertensive use, and percentages that achieved the post-guideline recommendation (systolic blood pressure <130 mmHg) were calculated. Analyses were performed in 2021-2022. RESULTS: Among 2,193 community-dwelling Atherosclerosis Risk in Communities participants aged 71-95 years at baseline, systolic blood pressure changes between baseline and follow-up visits differed among participants not reclassified (+4.1 mmHg, 95% CI=3.0, 5.3 mmHg), reclassified to Stage 1 hypertension (-1.1 mmHg, 95% CI= -2.2, 0.1 mmHg), and reclassified to Stage 2 hypertension (-5.7 mmHg, 95% CI= -6.8, -4.7 mmHg). Antihypertensive use changed from 77.3% to 78.4% (p=0.25) among participants reclassified to Stage 1 hypertension and from 78.3% to 81.4% (p<0.01) among participants reclassified to Stage 2 hypertension. At follow-up, 41.8% of the Stage 1 and 22.4% of the Stage 2 hypertension groups reached the systolic blood pressure <130 mmHg goal. CONCLUSIONS: There were small decreases in systolic blood pressure and increases in antihypertensive therapy among older adults reclassified to Stage 2 hypertension but not among those reclassified to Stage 1 hypertension by the 2017 American College of Cardiology/American Heart Association guideline.

Cardiovascular health trajectories from age 2-12: a pediatric electronic health record study.

PURPOSE: Many children have non-ideal cardiovascular health (CVH), but little is known about the course of CVH in early childhood. We identified CVH trajectories in children and assess the generalizability of these trajectories in an external sample. METHODS: We used data spanning 2010-2018 from children aged 2-12 years within the Chicago Area Patient-Centered Outcomes Research Network-an electronic health record network. Four clinical systems comprised the derivation sample and a fifth the validation sample. Body mass index, blood pressure, cholesterol, and blood glucose were categorized as ideal, intermediate, and poor using clinical measurements, laboratory readings, and International Classification of Diseases diagnosis codes and summed for an overall CVH score. Group-based trajectory modeling was used to create CVH score trajectories which were assessed for classification accuracy in the validation sample. RESULTS: Using data from 122,363 children (47% female, 47% non-Hispanic White) three trajectories were identified: 59.5% maintained high levels of clinical CVH, 23.4% had high levels of CVH that declined, and 17.1% had intermediate levels of CVH that further declined with age. A similar classification emerged when the trajectories were fitted in the validation sample. CONCLUSIONS: Stratification of CVH was present by age 2, implicating the need for early life and preconception prevention strategies.

Predictive Utility of Irritability "In Context": Proof-of-Principle for an Early Childhood Mental Health Risk Calculator.

OBJECTIVE: We provide proof-of-principle for a mental health risk calculator advancing clinical utility of the irritability construct for identification of young children at high risk for common, early onsetting syndromes. METHOD: Data were harmonized from two longitudinal early childhood subsamples (total N = 403; 50.1% Male; 66.7% Nonwhite; Mage = 4.3 years). The independent subsamples were clinically enriched via disruptive behavior and violence (Subsample 1) and depression (Subsample 2). In longitudinal models, epidemiologic risk prediction methods for risk calculators were applied to test the utility of the transdiagnostic indicator, early childhood irritability, in the context of other developmental and social-ecological indicators to predict risk of internalizing/externalizing disorders at preadolescence (Mage = 9.9 years). Predictors were retained when they improved model discrimination (area under the receiver operating characteristic curve [AUC] and integrated discrimination index [IDI]) beyond the base demographic model. RESULTS: Compared to the base model, the addition of early childhood irritability and adverse childhood experiences significantly improved the AUC (0.765) and IDI slope (0.192). Overall, 23% of preschoolers went on to develop a preadolescent internalizing/externalizing disorder. For preschoolers with both elevated irritability and adverse childhood experiences, the likelihood of an internalizing/externalizing disorder was 39-66%. CONCLUSIONS: Predictive analytic tools enable personalized prediction of psychopathological risk for irritable young children, holding transformative potential for clinical translation.

Cardiac Structure and Function Phenogroups and Risk of Incident Heart Failure (from the Multi-ethnic Study of Atherosclerosis).

Indices of cardiac structure and function, such as left ventricular (LV) mass and ejection fraction, have been associated with risk of incident heart failure (HF), but the clinical relevance of data-driven grouping of a comprehensive set of cardiac parameters is unclear. In Multi-Ethnic Study of Atherosclerosis participants, latent class analysis was applied in the sample stratified by gender to define phenogroups on the basis of cardiovascular magnetic resonance imaging parameters of right ventricular and LV structure and function at baseline. Cox proportional hazard models in gender-stratified analyses were used to assess the association between phenogroup membership and risk of HF subtypes adjusting for potential confounders. In the 4,204 participants (mean age 61 ± 10 years, 53% women), the mean follow-up time was 14 ± 4 years for men and 15 ± 4 years for women. For both genders, 4 distinct phenogroups were identified: (1) ideal cardiac mechanics; (2) higher output/hypertrophied LV; (3) impaired ejection fraction/dilated LV; and (4) higher output/hyperdynamic (LV). Men in phenogroups 4 (hazard ratio [HR] 2.91, 95% confidence interval [CI] 1.60 to 5.31, p = 0.0005), 3 (HR 3.52, 95% CI 1.90 to 6.53, p <0.0001), and 2 (HR 3.49, 95% CI 1.94 to 6.28, p <0.0001) had higher rates of incident HF than did men in phenogroup 1, in fully adjusted models. No significant associations were found between phenogroup membership and incident HF in women. In conclusion, phenogroup membership based on cardiac structure and function in men was significantly associated with incident HF. Integration of cardiac magnetic resonance imaging variables may help identify differential risk for HF in men.

Social Determinants of Cardiovascular Health in US Adolescents: National Health and Nutrition Examination Surveys 1999 to 2014.

Background Cardiovascular health (CVH) is suboptimal in US adolescents. Social determinants of health (SDOH) may affect CVH. We examined SDOH by race and ethnicity and assessed for associations between SDOH and CVH among US adolescents. Methods and Results We analyzed data from the National Health and Nutrition Examination Survey for 3590 participants aged 12 to 19 years from 1999 to 2014. SDOH variables were chosen and an SDOH score assigned (range, 0-7 points; higher=more favorable). CVH was classified according to American Heart Association criteria. We estimated population prevalence and used multivariable linear and polytomous logistic regression for associations between SDOH and CVH. SDOH varied by group, with the non-Hispanic White group (n=1155) having a higher/better mean SDOH score compared with non-Hispanic Black (n=1223) and Mexican American groups (n=1212). Associations between SDOH and CVH differed between racial and ethnic groups (interaction P<0.0001). For the non-Hispanic White group, each additional favorable SDOH variable was associated with a CVH score higher/better by 0.3 points (ß, 0.3, P<0.0001), 20% higher odds for moderate (versus low) CVH (odds ratio [OR], 1.2 [95% CI, 1.1-1.4]), and 80% higher odds for high/favorable (versus low) CVH (1.8 [1.5-2.1]). Associations between SDOH and CVH were more modest among the Mexican American group (ß, 0.12, P=0.001; OR 1.1 [1.0-1.2] for moderate CVH; OR, 1.3 [1.1-1.6] for high CVH) and were not significant among the non-Hispanic Black group (ß, 0.07; P=0.464). Conclusions SDOH and CVH were more favorable for non-Hispanic White adolescents compared with non-Hispanic Black and Mexican American adolescents. SDOH were strongly associated with CVH among the non-Hispanic White group. Racially and culturally sensitive public policy approaches may improve CVH in US adolescents.

Within-person decline in pregnancy smoking is observable prior to pregnancy awareness: Evidence across two independent observational cohorts.

Decreased consumption of nicotine and other drugs during pregnancy appears to be a cross-species phenomenon from which mechanism(s) capable of interrupting addictive processes could be elucidated. Whether pregnancy influences smoking behaviour independent of women's knowledge of the pregnancy, however, has not been considered. Using repeated measures analysis of variance (ANOVA), we estimated within-person change in mean cigarettes/day smoked across the estimated date of conception but prior to individually reported dates of pregnancy recognition using longitudinal smoking data from two independent observational cohorts, the Growing Up Healthy (GUH, n = 271) and Midwest Infant Development Studies (MIDS, n = 145). Participants smoked an average of half a pack/day in the month immediately before conception (M (SD) = 12(8.1) and 9.5(6.7) cigarettes/day in GUH and MIDS, respectively). We observed within-person declines in smoking after conception, both before (MGUH  = -0.9; 95% CI -1.6, -0.2; p = 0.01; MMIDS  = -1.1; 95% CI -1.9, -0.3; p = 0.01) and after (MGUH  = -4.8; 95% CI -5.5, -4.1; p < 0.001; MMIDS  = -3.3; 95% CI -4.4, -2.5; p < 0.001) women were aware of having conceived, even when women who had quit and women who were planning to conceive were excluded from analyses. Pregnancy may interrupt smoking-related processes via mechanisms not previously considered. Plausible candidates and directions for future research are discussed.

Exposure to Neighborhood-Level Racial Residential Segregation in Young Adulthood to Midlife and Incident Subclinical Atherosclerosis in Black Adults: The Coronary Artery Risk Development in Young Adults Study.

BACKGROUND: Neighborhood-level racial residential segregation has been linked to several cardiovascular disease risk factors and outcomes in Black adults, but its impact on subclinical atherosclerosis remains unknown. In addition, although the impact of segregation on health may vary over the life course, most studies have examined segregation exposure at a single point in time. This article takes a life course approach by examining associations of exposure to neighborhood-level racial residential segregation in young adulthood and patterns of exposure from young adulthood to midlife with coronary artery calcification (CAC) incidence. METHODS: We used data on 1125 Black CARDIA study (Coronary Artery Risk Development in Young Adults) participants free of CAC. Residential segregation was assessed using the Gi* statistic and measured when participants were young adults (18-30 years old, in 1985-1986) and as the pattern from young adulthood to midlife (15 years later). Poisson regression with generalized estimating equations models was used to measure CAC incidence. RESULTS: We found participants living in low segregation neighborhoods in young adulthood had 0.52 (rate ratio [95% CI: 0.28-0.98]) times lower risk of developing CAC compared with high segregation after adjusting for young adulthood sociodemographic characteristics and neighborhood poverty. Associations were attenuated and no longer statistically significant with adjustment for midlife CAC risk factors hypothesized to be on the causal pathway (rate ratio: 0.56 [95% CI: 0.29-1.09]). Findings for patterns of segregation over time suggest participants living in low segregation neighborhoods in young adulthood were less likely to develop CAC than those who started out in medium/high segregation neighborhoods, regardless of where they lived in midlife (rate ratio for increase from low to medium/high: 0.42 [95% CI: 0.19-0.95]; rate ratio for continuously low versus continuously medium/high segregation neighborhoods: 0.75 [95% CI: 0.31-1.83]). CONCLUSIONS: We found that participants living in more segregated neighborhoods in young adulthood were more likely to develop CAC due at least in part to differences in CAC risk factor burden accumulated over follow-up.

Association between county-level risk groups and COVID-19 outcomes in the United States: a socioecological study.

BACKGROUND: Geographic heterogeneity in COVID-19 outcomes in the United States is well-documented and has been linked with factors at the county level, including sociodemographic and health factors. Whether an integrated measure of place-based risk can classify counties at high risk for COVID-19 outcomes is not known. METHODS: We conducted an ecological nationwide analysis of 2,701 US counties from 1/21/20 to 2/17/21. County-level characteristics across multiple domains, including demographic, socioeconomic, healthcare access, physical environment, and health factor prevalence were harmonized and linked from a variety of sources. We performed latent class analysis to identify distinct groups of counties based on multiple sociodemographic, health, and environmental domains and examined the association with COVID-19 cases and deaths per 100,000 population. RESULTS: Analysis of 25.9 million COVID-19 cases and 481,238 COVID-19 deaths revealed large between-county differences with widespread geographic dispersion, with the gap in cumulative cases and death rates between counties in the 90th and 10th percentile of 6,581 and 291 per 100,000, respectively. Counties from rural areas tended to cluster together compared with urban areas and were further stratified by social determinants of health factors that reflected high and low social vulnerability. Highest rates of cumulative COVID-19 cases (9,557 [2,520]) and deaths (210 [97]) per 100,000 occurred in the cluster comprised of rural disadvantaged counties. CONCLUSIONS: County-level COVID-19 cases and deaths had substantial disparities with heterogeneous geographic spread across the US. The approach to county-level risk characterization used in this study has the potential to provide novel insights into communicable disease patterns and disparities at the local level.

Association of State Medicaid Expansion Status With Hypertensive Disorders of Pregnancy in a Singleton First Live Birth.

BACKGROUND: Incidence of hypertensive disorders of pregnancy is increasing in the United States. Early detection is important to prevent adverse maternal and offspring outcomes. This ecological study evaluated changes in rates of hypertensive disorders of pregnancy among states that expanded Medicaid compared with states that did not expand Medicaid. METHODS: A quasi-experimental analysis using difference-in-differences models compared changes in rates of hypertensive disorders of pregnancy in Medicaid expansion states relative to non-Medicaid expansion states from 2012 to 2019. Maternal data from singleton first live births to individuals aged 20 to 39 years were obtained from the National Center for Health Statistics. Outcomes of interest included age-adjusted rates of de novo hypertension in pregnancy (gestational hypertension or preeclampsia) and prepregnancy hypertension. RESULTS: Data from 7 764 965 individuals with a singleton first live birth were analyzed from 17 states and Washington, DC that expanded Medicaid and 15 states that did not. Rates of de novo hypertension in pregnancy increased over the study period in both expansion (54.34 [95% CI, 48.25-60.43] to 74.87 [95% CI, 71.20-78.55] per 1000 births) and nonexpansion states (68.32 [95% CI, 61.02-75.62] to 84.79 [95% CI, 80.67-88.91] per 1000 births). In adjusted difference-in-differences analyses, expansion status was associated with a greater increase in rates of de novo hypertension in pregnancy (difference-in-differences coefficient, +8.18 [95% CI, 4.00-12.36] per 1000 live births) but a decline in rates of de novo hypertension in pregnancy complicated by low birth weight (-7.20 [95% CI, -13.71 to -0.70] per 1000 births with hypertensive disorders of pregnancy). In adjusted difference-in-differences analyses, there were no significant changes in rates of prepregnancy hypertension in expansion relative to nonexpansion states (+1.13 [95% CI, -0.09 to +2.35] per 1000 live births). CONCLUSIONS: Between 2012 and 2019, states that expanded Medicaid had a significantly greater increase in rates of de novo hypertension, with some evidence of better outcomes among those with de novo hypertension diagnosed in pregnancy.

Longitudinal Analysis of Food Insufficiency and Cardiovascular Disease Risk Factors in the CARDIA study.

INTRODUCTION: Most previous studies on food insecurity and cardiovascular disease risk factors are cross-sectional. Without longitudinal data, it is unclear whether food insecurity precedes poor health and how exposure timing impacts these relationships. METHODS: Data from 2000 to 2001, 2005 to 2006, and 2010 to 2011 of the Coronary Artery Risk Development in Young Adults study were used. Food insufficiency-a screener measure related to food insecurity-was assessed in 2000-2001 and 2005-2006 using a single item. Cardiovascular disease risk factors were objectively assessed in 2010-2011. Impacts of food insufficiency patterns (food sufficient, food insufficient in 2000-2001 only, food insufficient in 2005-2006 only, food insufficient in both 2000-2001 and 2005-2006) on cardiovascular disease risk factors were estimated using inverse probability weighting of marginal structural models. Covariates that change over time were adjusted for using stabilized weights; baseline covariates were adjusted for in the marginal structural models. Analyses were conducted in 2020-2021. RESULTS: The baseline sample included 2,596 participants (56% women, 47% White). In unadjusted analyses, all food insufficiency patterns were associated with higher BMI, waist circumference, and blood pressure than food sufficiency. After accounting for covariates, estimates were attenuated but still consistent with adverse effects of food insufficiency, particularly among women. CONCLUSIONS: After covariate adjustment, food insufficiency was associated with several cardiovascular disease risk factors. Findings from this study should be replicated in other settings and populations. If verified, this evidence could provide justification for intervening in food insecurity to reduce future cardiovascular disease risk.

Impact of prenatal exposure characterization on early risk detection: Methodologic insights for the HEALthy Brain and Child Development (HBCD) study.

BACKGROUND: A major challenge in prenatal drug exposure research concerns the balance of measurement quality with sample sizes necessary to address confounders. To inform the selection of optimal exposure measures for the HEALthy Brain and Child Development (HBCD) Study, we employed integrated analysis to determine how different methods used to characterize prenatal tobacco exposure influence the detection of exposure-related risk, as reflected in normal variations in birth weight. METHODS: Participants were N = 2323 mother-infant dyads derived from 7 independent developmental cohorts harmonized on measures of exposure, outcome (birthweight), and covariates. We compared estimates of PTE-related effects on birthweight derived from linear regression models when PTE was categorized dichotomously based on any fetal exposure (30% exposed; 69% not exposed); versus categorically, based on common patterns of maternal smoking during pregnancy (never smoked 69%; quit smoking 16%; smoked intermittently 2%; smoked persistently 13%). We secondarily explored sex differences in PTE-birthweight associations across these categorization methods. RESULTS: When PTE was categorized dichotomously, exposure was associated with a - 125-g difference in birthweight (95% C.I. -173.7 - -76.6, p < .0001). When PTE was characterized categorically based on maternal smoking patterns, however, exposure was associated with either no difference in birthweight if mothers quit smoking by the end of the first trimester (B = -30.6, 95% C.I. -88.7-27.4, p = .30); or a - 221.8 g difference in birthweight if mothers did not [95% C.I. (-161.7 to -282.0); p < .001]. Qualitative sex differences were also detected though PTE x sex interactions did not reach statistical significance. Maternal smoking cessation during pregnancy was associated with a 239.3 g increase in birthweight for male infants, and a 114.0 g increase in birthweight for females infants (p = .07). CONCLUSIONS: Categorization of PTE based on patterns of maternal smoking rather than the presence or absence of exposure alone revealed striking nuances in estimates of exposure-related risk. The described method that captures both between-individual and within-individual variability in prenatal drug exposure is optimal and recommended for future developmental investigations such as the HBCD Study.

The Timing and Sequence of Cardiovascular Health Decline.

INTRODUCTION: Childhood declines in cardiovascular health have been linked to the development of subclinical atherosclerosis; however, less is known about the timing and sequence of the decline of the specific cardiovascular health components. The study objective is to identify the patterns of decline and associations with adulthood subclinical atherosclerosis. METHODS: Data were pooled from 5 cardiovascular cohorts. Clinical components of cardiovascular health (BMI, blood pressure, cholesterol, and blood glucose) were categorized as ideal or nonideal using American Heart Association definitions. Multitrajectory models simultaneously fitted the probability ideal for each factor. Adjusted associations between trajectory groups and carotid intima-media thickness were modeled. Data were pooled from December 1, 2015 to June 1, 2019; statistical analysis occurred between June 1, 2019 and June 1, 2020. RESULTS: This study included 9,388 individuals (55% female, 66% White). A total of 5 distinct trajectory groups were created: 1 maintained the ideal levels of all the 4 health factors, 2 had risk onset of a single factor in childhood, 1 had risk onset of multiple factors in childhood, and 1 had risk onset in adulthood. Those with childhood multiple risk onset had 8.1% higher carotid intima-media thickness (95% CI=0.067, 0.095) than those in the ideal group, childhood cholesterol risk onset had 5.9% higher carotid intima-media thickness (95% CI=0.045, 0.072), childhood BMI risk onset had 5.5% higher carotid intima-media thickness (95% CI=0.041, 0.069), and early adulthood multiple risk onset had 2.7% higher carotid intima-media thickness (95% CI=0.013, 0.041). CONCLUSIONS: Those who lost the ideal status of cardiovascular health in childhood and early adulthood had more subclinical atherosclerosis than those who retained the ideal cardiovascular health across the life course, underscoring the importance of preserving the ideal cardiovascular health beginning in childhood and continued into adulthood.