Large Multiethnic Cohort Research Articles

Machine learning analysis of sex and menopausal differences in the gut microbiome in the HELIUS study

Sex differences in the gut microbiome have been examined previously, but results are inconsistent, often due to small sample sizes. We investigated sex and menopausal differences in the gut microbiome in a large multi-ethnic population cohort study, including 5166 participants. Using machine learning models, we revealed modest associations between sex and menopausal status, and gut microbiota composition (AUC 0.61–0.63). After adjustments for age, cardiovascular risk factors, and diet, a part of the associations of the highest-ranked gut microbes with sex were attenuated, but most associations remained significant. In contrast, most associations with menopausal status were driven by age and lost significance after adjustment. Using pathway analyses on metagenomic data, we identified sex differences in vitamin B6 synthesis and stachyose degradation pathways. Since some of sex differences in gut microbiome composition and function could not be explained by covariates, we recommend sex stratification in future microbiome studies.

Associations of circulating T-cell subsets with carotid artery stiffness: the multiethnic study of atherosclerosis.

Arterial stiffness measured by total pulse wave velocity (T-PWV) is associated with an increased risk of multiple age-related diseases. T-PWV can be described by structural (S-PWV) and load-dependent (LD-PWV) arterial stiffening. T-cells have been implicated in arterial remodeling, arterial stiffness, and hypertension in humans and animals; however, it is unknown whether T-cells are risk factors for T-PWV or its components. Therefore, we evaluated the cross-sectional associations of peripheral T-cell subpopulations with T-PWV, S-PWV, and LD-PWV. Peripheral blood T-cells were characterized using flow cytometry, and carotid artery stiffness was measured using B-mode ultrasound to calculate T-PWV at the baseline examination in a participant subset of the Multi-Ethnic Study of Atherosclerosis (MESA, n = 1,984). A participant-specific exponential model was used to calculate S-PWV and LD-PWV based on elastic modulus and blood pressure gradients. The associations between five primary (P-significance < 0.01) and 25 exploratory (P-significance < 0.05) immune cell subpopulations, per 1-SD increment, and arterial stiffness measures were assessed using adjusted linear regression models. For the primary analysis, higher CD4+CD28-CD57+, but not CD8+CD28-CD57+, T-cells were associated with higher LD-PWV (β = 0.04 m/s, P < 0.01) after adjusting for covariates. None of the remaining T-cell subpopulations in the primary analysis were associated with T-PWV or S-PWV. For the exploratory analysis, several memory and differentiated/senescence-associated CD4+ and CD8+ T-cell subpopulations were associated with greater T-PWV, S-PWV, and LD-PWV after adjusting for covariates. In conclusion, we highlight novel associations in humans between CD4+ and CD8+ memory and differentiated/senescence-associated T-cell subpopulations and measures of arterial stiffness in MESA. These results warrant longitudinal, prospective studies that examine changes in T-cell subpopulations and arterial stiffness in humans.NEW & NOTEWORTHY We investigated associations between T-cells and novel measures of structural and load-dependent arterial stiffness in a large multiethnic cohort. The primary analysis revealed that pro-inflammatory, senescence-associated CD4+CD28-CD57+ T-cells were associated with higher load-dependent arterial stiffness. An exploratory analysis revealed that multiple pro-inflammatory CD4+ and CD8+ T-cell subpopulations were associated with both higher structural and load-dependent arterial stiffness. These results suggest that pro-inflammatory T-cells may contribute to arterial stiffness through both arterial remodeling and elevated blood pressure.

Interplay between lifestyle factors and polygenic risk for incident coronary heart disease in a large multiethnic cohort

IntroductionThe objective of this study was to examine the interplay of polygenic risk and individual lifestyle factors (and a composite score of lifestyle) as antecedents of CHD in a large multiethnic cohort. MethodsWe used Genetic Epidemiology Resource in Adult Health and Aging (GERA) cohort participants free of CHD at baseline (n = 60,568; 67 % female; 18 % non-European). The individual and joint associations of smoking, Mediterranean diet pattern, level of physical activity and polygenic risk with incident CHD were assessed using Cox regression adjusting for genetic ancestry and non-mediating risk factors. Hazard ratios (HRs) and number needed to treat (NNT) were estimated according to these lifestyle factors and polygenic risk categories. Strengths included large sample size, long-follow-up, ethnic diversity, a clinically-validated polygenic risk score (PRS), and rich phenotype information. ResultsAfter 14 years of follow-up, there were 3159 incident CHD events. We observed no statistically significant interactions between individual lifestyle factors and polygenic risk (all p > 0.23). For individuals with a high genetic risk, moving from the worse lifestyle combination (no favorable lifestyle factors) to the best lifestyle combination (all three) is associated with 52 % lower rate of CHD. The NNT was highest in the low polygenic risk group (34), lowest in the high polygenic risk group [19] and in-between (Jin et al., 2011) [24] in the intermediate polygenic risk group. ConclusionsLifestyle and polygenic risk together influence the risk of incident CHD. Our results support consideration of polygenic risk in lifestyle interventions because those with high polygenic risk are likely to derive the most benefit.

Air Pollution and Breast Cancer Incidence in the Multiethnic Cohort Study.

Recent studies suggested fine particulate matter (PM2.5) exposure increases the risk of breast cancer, but evidence among racially and ethnically diverse populations remains sparse. Among 58,358 California female participants of the Multiethnic Cohort (MEC) Study followed for an average of 19.3 years (1993-2018), we used Cox proportional hazards regression to examine associations of time-varying PM with invasive breast cancer risk (n = 3,524 cases; 70% African American and Latino females), adjusting for sociodemographics and lifestyle factors. Subgroup analyses were conducted for race and ethnicity, hormone receptor status, and breast cancer risk factors. Satellite-based PM2.5 was associated with a statistically significant increased incidence of breast cancer (hazard ratio [HR] per 10 μg/m3, 1.28 [95% CI, 1.08 to 1.51]). We found no evidence of heterogeneity in associations by race and ethnicity and hormone receptor status. Family history of breast cancer showed evidence of heterogeneity in PM2.5-associations (Pheterogeneity = .046). In a meta-analysis of the MEC and 10 other prospective cohorts, breast cancer incidence increased in association with exposure to PM2.5 (HR per 10 μg/m3 increase, 1.05 [95% CI, 1.00 to 1.10]; P = .064). Findings from this large multiethnic cohort with long-term air pollutant exposure and published prospective cohort studies support PM2.5 as a risk factor for breast cancer. As about half of breast cancer cannot be explained by established breast cancer risk factors and incidence is continuing to increase, particularly in low- and middle-income countries, our results highlight that breast cancer prevention should include not only individual-level behavior-centered approaches but also population-wide policies and regulations to curb PM2.5 exposure.

Rare variant analyses validate known ALS genes in a multi-ethnic population and identifies ANTXR2 as a candidate in PLS

BackgroundAmyotrophic lateral sclerosis (ALS) is a neurodegenerative disease affecting over 300,000 people worldwide. It is characterized by the progressive decline of the nervous system that leads to the weakening of muscles which impacts physical function. Approximately, 15% of individuals diagnosed with ALS have a known genetic variant that contributes to their disease. As therapies that slow or prevent symptoms continue to develop, such as antisense oligonucleotides, it is important to discover novel genes that could be targets for treatment. Additionally, as cohorts continue to grow, performing analyses in ALS subtypes, such as primary lateral sclerosis (PLS), becomes possible due to an increase in power. These analyses could highlight novel pathways in disease manifestation.MethodsBuilding on our previous discoveries using rare variant association analyses, we conducted rare variant burden testing on a substantially larger multi-ethnic cohort of 6,970 ALS patients, 166 PLS patients, and 22,524 controls. We used intolerant domain percentiles based on sub-region Residual Variation Intolerance Score (subRVIS) that have been described previously in conjunction with gene based collapsing approaches to conduct burden testing to identify genes that associate with ALS and PLS.ResultsA gene based collapsing model showed significant associations with SOD1, TARDBP, and TBK1 (OR = 19.18, p = 3.67 × 10–39; OR = 4.73, p = 2 × 10–10; OR = 2.3, p = 7.49 × 10–9, respectively). These genes have been previously associated with ALS. Additionally, a significant novel control enriched gene, ALKBH3 (p = 4.88 × 10–7), was protective for ALS in this model. An intolerant domain-based collapsing model showed a significant improvement in identifying regions in TARDBP that associated with ALS (OR = 10.08, p = 3.62 × 10–16). Our PLS protein truncating variant collapsing analysis demonstrated significant case enrichment in ANTXR2 (p = 8.38 × 10–6).ConclusionsIn a large multi-ethnic cohort of 6,970 ALS patients, collapsing analyses validated known ALS genes and identified a novel potentially protective gene, ALKBH3. A first-ever analysis in 166 patients with PLS found a candidate association with loss-of-function mutations in ANTXR2.

Prevalence and target attainment of traditional cardiovascular risk factors in patients with systemic lupus erythematosus: a cross-sectional study including 3401 individuals from 24 countries

Systemic lupus erythematosus (SLE) is characterised by increased cardiovascular morbidity and mortality risk. We aimed to examine the prevalence of traditional cardiovascular risk factors and their control in an international survey of patients with systemic lupus erythematosus. In this multicentre, cross-sectional study, cardiovascular risk factor data from medical files of adult patients (aged ≥18) with SLE followed between Jan 1, 2015, and Jan 1, 2020, were collected from 24 countries, across five continents. We assessed the prevalence and target attainment of cardiovascular risk factors and examined potential differences by country income level and antiphospholipid syndrome coexistence. We used the Systemic Coronary Risk Evaluation algorithm for cardiovascular risk estimation, and the European Society of Cardiology guidelines for assessing cardiovascular risk factor target attainment. People with lived experience were not involved in the research or writing process. 3401 patients with SLE were included in the study. The median age was 43·0 years (IQR 33-54), 3047 (89·7%) of 3396 patients were women, 349 (10.3%) were men, and 1629 (48·1%) of 3390 were White. 556 (20·7%) of 2681 patients had concomitant antiphospholipid syndrome. We found a high cardiovascular risk factor prevalence (hypertension 1210 [35·6%] of 3398 patients, obesity 751 [23·7%] of 3169 patients, and hyperlipidaemia 650 [19·8%] of 3279 patients), and suboptimal control of modifiable cardiovascular risk factors (blood pressure [target of <130/80 mm Hg], BMI, and lipids) in the entire SLE group. Higher prevalence of cardiovascular risk factors but a better blood pressure (target of <130/80 mm Hg; 54·9% [1170 of 2132 patients] vs 46·8% [519 of 1109 patients]; p<0·0001), and lipid control (75·0% [895 of 1194 patients] vs 51·4% [386 of 751 patients], p<0·0001 for high-density lipoprotein [HDL]; 66·4% [769 of 1158 patients] vs 60·8% [453 of 745 patients], p=0·013 for non-HDL; 80·9% [1017 of 1257 patients] vs 61·4% [486 of 792 patients], p<0·0001 for triglycerides]) was observed in patients from high-income versus those from middle-income countries. Patients with SLE with antiphospholipid syndrome had a higher prevalence of modifiable cardiovascular risk factors, and significantly lower attainment of BMI and lipid targets (for low-density lipoprotein and non-HDL) than patients with SLE without antiphospholipid syndrome. High prevalence and inadequate cardiovascular risk factor control were observed in a large multicentre and multiethnic SLE cohort, especially among patients from middle-income compared with high-income countries and among those with coexistent antiphospholipid syndrome. Increased awareness of cardiovascular disease risk in SLE, especially in the above subgroups, is urgently warranted. None.

Polygenic risk and incident coronary heart disease in a large multiethnic cohort

ObjectiveMany studies support the notion that polygenic risk scores (PRS) improve risk prediction for coronary heart disease (CHD) beyond conventional risk factors. However, PRS are not yet considered risk-enhancing factor in guidelines. Our objective was to determine the predictive performance of a commercially available PRS (CARDIO inCode-Score®) compared with the Pooled Cohorts Equations (PCE) in a contemporary, multi-ethnic cohort. MethodsParticipants (n = 63,070; 67 % female; 18 % non-European) without prior CHD were followed from 2007 through 12/31/2022. The association between the PRS and incident CHD was assessed using Cox regression adjusting for genetic ancestry and risk factors. Event rates were estimated by categories of PCE and by low/intermediate/high genetic risk within PCE categories; risk discrimination and net reclassification improvement (NRI) were also assessed. ResultsThere were 3,289 incident CHD events during 14 years of follow-up. Adjusted hazard ratio (aHR) for incident CHD per 1 SD increase in PRS was 1.18 (95 % CI:1.14–1.22), and the aHR for the upper vs lower quintile of the PRS was 1.66 (95 % CI:1.49–1.86). The association was consistent in both sexes, in European participants compared with all minority groups combined and was strongest in the first 5 years of follow-up. The increase in the C-statistic was 0.004 (0.747 vs. 0.751; p < 0.0001); the NRI was 2.4 (0.9–3.8) for the entire cohort and 9.7 (7.5–12.0) for intermediate PCE risk individuals. After incorporating high genetic risk, a further 10 percent of participants at borderline/intermediate PCE risk would be candidates for statin therapy. ConclusionInclusion of polygenic risk improved identification of primary prevention individuals who may benefit from more intensive risk factor modification.

Nailfold capillaroscopy findings of a multicentric multi-ethnic cohort of patients with idiopathic inflammatory myopathies.

To assess nailfold video capillaroscopic (NVC) abnormalities and their association with clinical features, myositis-specific autoantibodies (MSA), and myositis-associated antibodies (MAA) in a large multi-ethnic cohort of patients with idiopathic inflammatory myopathies (IIM). We recruited 155 IIM patients from three centres in Mexico, Spain, and the USA. We evaluated the clinical and laboratory features of the patients and performed semiquantitative and quantitative analyses of the NVC. Each NVC study was defined as having a normal, non-specific, early systemic sclerosis (SSc), active SSc, or late SSc pattern. Twenty-three patients had at least one follow-up NVC when disease control was achieved. Quantitative variables were expressed as medians and interquartile range (IQR) and were compared with the Kruskal-Wallis, the Mann-Whitney U-test, and the Wilcoxon test for paired medians. Associations between qualitative variables were assessed with the χ2 test. Most patients were women (68.3%), Hispanic (73.5%), and had dermatomyositis (DM) (61.2%). Fourteen patients (9%) had a normal NVC. A non-specific abnormality pattern was the most frequent (53.9%), and was associated with joint involvement, interstitial lung disease, Jo1 autoantibodies, anti-synthetase syndrome, and immune-mediated necrotising myopathy. The SSc pattern was observed mostly in DM and overlap myositis and was associated with cutaneous features and anti-TIF-1g autoantibodies. After treatment, there was a decrease in the capillaroscopic score, the capillary diameter, and the number of avascular areas, and an increase in capillary density and bushy capillary number. NVC abnormalities are related to the diagnosis, clinical features, disease activity, and autoantibodies of patients with IIM.

Histologic and Clinical Factors Associated with Kidney Outcomes in IgA Vasculitis Nephritis.

Nephritis is a common manifestation of IgA vasculitis and is morphologically indistinguishable from IgA nephropathy. While MEST-C scores are predictive of kidney outcomes in IgA nephropathy, their value in IgA vasculitis nephritis has not been investigated in large multiethnic cohorts. Biopsies from 262 children and 99 adults with IgA vasculitis nephritis ( N =361) from 23 centers in North America, Europe, and Asia were independently scored by three pathologists. MEST-C scores were assessed for correlation with eGFR/proteinuria at biopsy. Because most patients ( N =309, 86%) received immunosuppression, risk factors for outcomes were evaluated in this group using latent class mixed models to identify classes of eGFR trajectories over a median follow-up of 2.7 years (interquartile range, 1.2-5.1). Clinical and histologic parameters associated with each class were determined using logistic regression. M, E, T, and C scores were correlated with either eGFR or proteinuria at biopsy. Two classes were identified by latent class mixed model, one with initial improvement in eGFR followed by a late decline (class 1, N =91) and another with stable eGFR (class 2, N =218). Class 1 was associated with a higher risk of an established kidney outcome (time to ≥30% decline in eGFR or kidney failure; hazard ratio, 5.84; 95% confidence interval, 2.37 to 14.4). Among MEST-C scores, only E1 was associated with class 1 by multivariable analysis. Other factors associated with class 1 were age 18 years and younger, male sex, lower eGFR at biopsy, and extrarenal noncutaneous disease. Fibrous crescents without active changes were associated with class 2. Kidney outcome in patients with biopsied IgA vasculitis nephritis treated with immunosuppression was determined by clinical risk factors and endocapillary hypercellularity (E1) and fibrous crescents, which are features that are not part of the International Study of Diseases of Children classification.

Noninvasive tests to identify liver fibrosis in metabolic dysfunction-associated steatotic liver disease are affected by race.

The objective of this study was to assess the performance of noninvasive tests (NITs) across different racial and ethnic groups in a large multiethnic cohort. Data were derived from the National Health and Nutrition Examination Survey (NHANES) 2017 through 2020. Participants without valid transient elastography measurements or with alternative etiologies of liver steatosis disease were excluded from the study. Among the 6359 adults included in the study, fatty liver index and nonalcoholic fatty liver disease liver fat scores performed well for the prediction of metabolic dysfunction-associated steatotic liver disease, without significant changes across racial and ethnic groups. However, significant differences were observed across racial and ethnic groups for the prediction of advanced fibrosis and cirrhosis. The fibrosis-4 (FIB-4) index, aspartate aminotransferase to platelet ratio index (APRI), and nonalcoholic fatty liver disease fibrosis score underperformed in non-Hispanic Black patients for the detection of cirrhosis. For the detection of advanced fibrosis, their performance was also numerically worse in non-Hispanic Black patients but only reached statistical significance for APRI. Using a cutoff point of 12 kPa for advanced fibrosis, both APRI and the FIB-4 index performed significantly worse in non-Hispanic Black patients. In a large, diverse national cohort, the performance of NITs was overall poor compared with transient elastography, and NITs showed differences across racial and ethnic groups. Given the widespread use of NITs, it is imperative that the scores are equitable across racial and ethnic groups.

Methods for mediation analysis with high-dimensional DNA methylation data: Possible choices and comparisons.

Epigenetic researchers often evaluate DNA methylation as a potential mediator of the effect of social/environmental exposures on a health outcome. Modern statistical methods for jointly evaluating many mediators have not been widely adopted. We compare seven methods for high-dimensional mediation analysis with continuous outcomes through both diverse simulations and analysis of DNAm data from a large multi-ethnic cohort in the United States, while providing an R package for their seamless implementation and adoption. Among the considered choices, the best-performing methods for detecting active mediators in simulations are the Bayesian sparse linear mixed model (BSLMM) and high-dimensional mediation analysis (HDMA); while the preferred methods for estimating the global mediation effect are high-dimensional linear mediation analysis (HILMA) and principal component mediation analysis (PCMA). We provide guidelines for epigenetic researchers on choosing the best method in practice and offer suggestions for future methodological development.

Novel Size-Based High-Density Lipoprotein Subspecies and Incident Vascular Events.

Background High-density lipoprotein (HDL) particle concentration likely outperforms HDL cholesterol in predicting atherosclerotic cardiovascular events. Whether size-based HDL subspecies explain the atheroprotective associations of HDL particle concentration remains unknown. Our objective was to assess whether levels of specific size-based HDL subspecies associate with atherosclerotic cardiovascular disease in a multiethnic pooled cohort and improve risk prediction beyond traditional atherosclerotic cardiovascular disease risk factors. Methods and Results Seven HDL size-based subspecies were quantified by nuclear magnetic resonance (LP4 algorithm; H1=smallest; H7=largest) among participants without prior atherosclerotic cardiovascular disease in ARIC (Atherosclerosis Risk in Communities), MESA (Multi-Ethnic Study of Atherosclerosis), PREVEND (Prevention of Renal and Vascular Endstage Disease), and DHS (Dallas Heart Study) cohorts (n=15 371 people). Multivariable Cox proportional hazards models were used to evaluate the association between HDL subspecies and incident myocardial infarction (MI) or ischemic stroke at follow-up (average 8-10 years) adjusting for HDL cholesterol and risk factors. Improvement in risk prediction was assessed via discrimination and reclassification analysis. Within the pooled cohort (median age 57 years; female 54%; Black 22%) higher H1 (small) and H4 (medium) concentrations were inversely associated with incident MI (hazard ratio [HR]/SD, H1 0.88 [95% CI, 0.81-0.94]; H4 0.89 [95% CI, 0.82-0.97]). H4 but not H1 improved risk prediction indices for incident MI. Increasing H2 and H4 were inversely associated with improved risk prediction indices for composite end point of stroke, MI, and cardiovascular death (HR/SD, H2 0.94 [95% CI, 0.88-0.99]; H4 0.91 [95% CI, 0.85-0.98]). Levels of the large subspecies (H6 and H7) were not associated with any vascular end point. Conclusions Two of 7 HDL size-based subspecies modestly improved risk prediction for MI and composite vascular end points in a large multiethnic pooled cohort. These findings support assessment of precise HDL subspecies for future studies regarding clinical utility.

Carotid intima-media thickness and incident hypertension: the Brazilian Longitudinal Study of Adult Health.

High blood pressure (BP) increases carotid intima-media thickness (CIMT). On the other hand, it is not clear whether the vascular abnormalities reflected in high CIMT may predict incident hypertension. The present study aims to investigate the association between CIMT and incident hypertension after 4 years of follow-up in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), a multiethnic sample of middle-aged adults from six Brazilian cities. We analyzed data from 6682 ELSA-Brasil participants (aged 35-74) without hypertension and with complete CIMT data at baseline. After 4 years of follow-up, we describe hypertension incidence, stratifying the sample according to age, sex, and race-specific CIMT quartiles. We also built crude and adjusted Poisson regression models to analyze the association between mean and maximal CIMT values and incident hypertension. We found incident hypertension in 987 (14.8%) participants. According to mean CIMT quartile groups, hypertension incidence varied from 10.2% (first quartile group) to 22.4% (fourth quartile group; P for trend <0.001). In fully adjusted models, 0.1 mm increments in mean CIMT values were associated with a 16% [relative risk (RR):1.16; 95% confidence interval (95% CI) 1.10-1.21; P < 0.001] higher risk of incident hypertension, respectively. Results were similar when maximal CIMT values were considered instead of mean CIMT values. CIMT values at baseline strongly predicted incident hypertension after 4 years of follow-up in this large multiethnic cohort. This highlights the relationship between CIMT and BP and may provide important insights into the significance of this ultrasound measurement.

179-LB: Should We Screen with a Fasting Blood Glucose Test? A Novel Two-Step Screening Strategy for Gestational Diabetes—A Large Tertiary Comparative Cohort Study

Background: Gestational diabetes mellitus (GDM) diagnosis is resource intensive and controversial. A modified COVID-19 GDM diagnostic method was introduced at Western Health Melbourne, an Australian tertiary hospital caring for a multi-ethnic population with a high GDM prevalence at 20-23%. In place of a universal 2 hour 75g OGTT, a novel two step diagnostic method was instituted. A fasting blood glucose (FBG) was performed with thresholds based on a secondary analysis of HAPO data. A FBG of ≥5.1mmol/L (91.8mg/dL) diagnosed GDM; women with FBG 4.7-5.0mmol/L (84.6-90.0mg/dL) proceeded to an OGTT (IADPSG diagnostic criteria); FBG ≤4.6mmol/L (82.8mg/dL) was considered to exclude GDM. Methods: A retrospective cohort study of singleton pregnancies birthing after 20 weeks’ gestation compared perinatal outcomes in 6469 women who underwent modified screening (MS) during the COVID-19 pandemic with 8669 women who had conventional screening (CS) prior to COVID-19. Results: GDM prevalence was the same using either MS or CS (22.9 vs 22.2% p=0.31) with similar rates of insulin use. There was no difference in the prevalence of large for gestational age infants (10.3 vs 9.4% p=0.06), pre-term birth (6.6 vs 6.4% p=0.53) or gestational age at birth (39.2 vs 39.1 weeks). Mean birth weight was increased with MS (3335g vs 3307g) and small for gestational age rate reduced (9.1 vs 10.5% p=0.01). The MS group had fewer inductions (34.2 vs 39.8% p&amp;lt;0.01). Primary caesarean section rate was unchanged (18.4 vs 18.4% p=0.91). MS reduced the need for an OGTT by 76%. Women with a FBG ≤4.6mmol/L by MS had non-inferior outcomes compared to women with a FBG ≤4.6mmol/L with CS irrespective of GDM diagnosis. Conclusion: In a large multi-ethnic cohort with a high prevalence of GDM, modified screening using FBG greatly reduced the number of resource intensive OGTTs with non-inferior perinatal outcomes. A prospective randomised trial comparing these methods is warranted. Disclosure J. Deitch: None. J. M. Said: None. G. R. Teale: None. D. Kevat: Advisory Panel; Novo Nordisk, Speaker's Bureau; Boehringer Ingelheim and Eli Lilly Alliance, Novo Nordisk, Abbott Diabetes, AstraZeneca. I. Lee: None. C. J. Yates: None. E. Kathpal: None. P. Lawton: None. I. Shahid: None. C. Steele: None. P. S. Hamblin: None. B. Krishnamurthy: None. Funding Australian Government Research Training Program; Sheppard Millar Lowe Research Trust Scholarship

Social determinants of health burden among adolescent and young adult cancer patients and impact on overall survival.

12017 Background: The 5-year survival rates for adolescents and young adults (AYA) with cancer have steadily improved, yet disparities by race/ethnicity persist. The drivers of these disparities remain unclear. Neighborhood socioeconomic deprivation has been linked to adverse health outcomes independent of the effects of individual-level measures of social determinants of health (SDOH). We evaluated the impact of neighborhood SDOH on overall survival (OS) in a diverse AYA cancer patient population treated in a single academic setting. Methods: This study utilized a multi-ethnic cohort of AYA cancer survivors (N=10,261) diagnosed between ages 15-39 seen at MD Anderson Cancer Center between 2000 and 2016. Patient demographics, zip code at presentation, cancer characteristics, and follow-up were obtained from our institutional tumor registry. Zip codes were linked to Area Deprivation Index (ADI) values, a validated neighborhood-level SDOH measure, with higher ADI representing worse SDOH. Results: The cohort included 6,361 (62%) non-Hispanic White, 1,112 (10.8%) non-Hispanic Black, and 2,133 (20.8%) Hispanic AYA cancer patients, with 655 (6.4%) of other ethnicities or missing information. The mean ADI value was significantly higher (p&lt;0.05) for Black (61.7) and Hispanic (65.3) patients, compared to White (51.2) patients. Analysis of ADI by cancer type showed significant differences, mainly driven by the significantly higher ADI in patients with cervical cancer (62.3) than other cancer types. In contrast, AYA patients diagnosed with breast cancer (53.0) and CNS tumors (52.9) resided in neighborhoods with the lowest ADI values. ADI was a significant predictor of OS overall. In multivariable models that included gender, age at diagnosis, cancer diagnosis, and race/ethnicity, risk of death for AYA cancer patients residing in neighborhoods with ADI in the highest quartile was greater than that for patients in the lowest quartile (HR: 1.09, 95% CI: 1.00-1.19, p=0.043). Conclusions: Our results in this large multi-ethnic cohort indicate that non-White AYA patients with cancer more commonly live in neighborhoods associated with higher ADI compared to their White counterparts. Even when treated in the same academic setting, AYA cancer patients with higher ADI experienced worse OS across cancers. Although the magnitude of the effect was moderate, the presence of this effect within a tertiary cancer center with existing barriers to care suggests that ADI is a meaningful risk factor impacting survival. Further analysis is warranted to determine the generalizability of these findings to the patient population more broadly. However, the results provide intriguing evidence for potential interventions aimed at supporting AYA cancer patients from disadvantaged backgrounds.

Rate and determinants of antiviral treatment initiation for patients with HBeAg-negative chronic hepatitis B.

Most clinic attenders with chronic hepatitis B (CHB) are serum HBeAg-negative, and a minority will require suppressive antiviral treatment. Expert guidelines propose schedules for the monitoring of untreated patients, but the recommended frequency of patient review does not reflect recognised demographic determinants of HBeAg-negative chronic hepatitis. Also, the impact of patient ethnicity on risk has not been defined. The aim of our study was to determine the rates and determinants of antiviral treatment initiation in a large multi-ethnic cohort of CHB patients attending a single centre. We undertook a retrospective study using entirely electronic sources of patient information. Treatment initiation dates were identified from electronic pharmacy records. Crude and time-dependent statistical analyses were undertaken to identify rate and risk factors for treatment initiation. Treatment was initiated for 232/1256 (18.5%) patients with rates of 23.2% and 33.2% at 5 and 10 years. An increased risk of treatment was associated with male sex (RR 1.803), older age at presentation (RR 1.027 per year increase) and with non-Black ethnicity (RR 1.654). Patient sex, baseline age and ethnicity also determined risk for treatment in the subset of patients with normal serum ALT and low HBV DNA at baseline, though overall treatment rate in this group was low (only 2% per annum). Thus, patient demographics permit risk stratification for treatment initiation and could determine to a significant extent the frequency of review required for untreated HBeAg-negative patients. Black ethnicity is associated with a significant reduction in risk of treatment initiation.

Lifetime healthcare expenses across demographic and cardiovascular risk groups: The application of a novel modeling strategy in a large multiethnic cohort study

ObjectiveTo understand the burden of healthcare expenses over the lifetime of individuals and evaluate differences among those with cardiovascular risk factors and among disadvantaged groups based on race/ethnicity and sex. MethodsWe linked data from the longitudinal multiethnic Dallas Heart Study, which recruited participants between 2000 and 2002, with inpatient and outpatient claims from all hospitals in the Dallas-Fort Worth metroplex through December 2018, capturing encounter expenses. Race/ethnicity and sex, as well as five risk factors, hypertension, diabetes, hyperlipidemia, smoking, and overweight/obesity, were defined at cohort enrollment. For each individual, expenses were indexed to age and cumulated between 40 and 80 years of age. Lifetime expenses across exposures were evaluated as interactions in generalized additive models. ResultsA total of 2184 individuals (mean age, 45±10 years; 61% women, 53% Black) were followed between 2000 and 2018. The mean modeled lifetime cumulative healthcare expenses were $442,629 (IQR, $423,850 to $461,408). In models that included 5 risk factors, Black individuals had $21,306 higher lifetime healthcare spending compared with non-Black individuals (P < .001), and men had modestly higher expenses than women ($5987, P < .001). Across demographic groups, the presence of risk factors was associated with progressively higher lifetime expenses, with significant independent association of diabetes ($28,075, P < .001), overweight/obesity ($8816, P < .001), smoking ($3980, P = .009), and hypertension ($528, P = .02) with excess spending. ConclusionOur study suggests Black individuals have higher lifetime healthcare expenses, exaggerated by the substantially higher prevalence of risk factors, with differences emerging in older age.

Abstract P244: Gestational Glucose Intolerance is Associated With Hypertensive Disorders of Pregnancy in a Large Hospital-Based Multiethnic Cohort

Introduction: Hypertensive disorders of pregnancy (HDP) are associated with higher future cardiovascular risk. Treatment of gestational diabetes (GDM) has been shown to reduce the risk of HDP. It is unknown if women with gestational glucose intolerance (GGI), defined as abnormal GDM screening but without a GDM diagnosis, are at higher risk of HDP. Currently, GGI is not treated. Hypothesis: We tested the hypothesis that GGI pregnancies have increased odds of HDP. Methods: This is a retrospective cohort study of 41,706 singleton pregnancies delivered at an academic center. Based on the results of GDM screening and diagnostic tests, pregnancies were categorized as normal glucose tolerance (NGT, glucose loading test &lt; 140 mg/dl), GGI (glucose loading test ≥140 mg/dl with 0 or 1 abnormal value on oral glucose tolerance testing [OGTT] thus without GDM), or GDM (≥ 2 abnormal values on OGTT). GGI was further classified into groups with zero (GGI-0) or 1 (GGI-1) abnormal OGTT values. We used laboratory, blood pressure, and delivery reports to capture HDP and exclude chronic hypertension. Generalized estimating equations for logistic regression were used to measure the risk of HDP after adjustment for age, 1st trimester BMI, parity, insurance, race/ethnicity, and marital status. A similar approach was used to test for associations between GGI and pre-eclampsia (PE). Results: A total of 84.5% of pregnancies were classified as NGT, 8.9% as GGI-0, 3.0% as GGI-1 and 3.6% as GDM. The unadjusted frequencies of HDP and PE were higher in GGI-1 (140 [11.2%] and 64 [5.1%] of 1,251) and GDM (165 [11.1%] and 94 [6.3%] of 1,485) compared to NGT (2,275 [6.5%] and 1,100 [3.1%] of 35,238) and GGI-0 (283 [7.6%] and 136 [3.6%] of 3,732). We observed greater odds of HDP in all 3 groups compared to NGT in adjusted models, and similar trends for PE (Figure 1). Conclusions: The risks of HDP and PE in GGI-1 are greater than for NGT pregnancies and similar to the risks in GDM. Our results suggest that women with GGI-1 should be considered for intervention, similar to GDM.

The Longitudinal Relationship Between Hearing Loss and Cognitive Decline

The Longitudinal Relationship Between Hearing Loss and Cognitive Decline

FHL2 Genetic Polymorphisms and Pro-Diabetogenic Lipid Profile in the Multiethnic HELIUS Cohort.

Type 2 diabetes mellitus (T2D) is a prevalent disease often accompanied by the occurrence of dyslipidemia. Four and a half LIM domains 2 (FHL2) is a scaffolding protein, whose involvement in metabolic disease has recently been demonstrated. The association of human FHL2 with T2D and dyslipidemia in a multiethnic setting is unknown. Therefore, we used the large multiethnic Amsterdam-based Healthy Life in an Urban Setting (HELIUS) cohort to investigate FHL2 genetic loci and their potential role in T2D and dyslipidemia. Baseline data of 10,056 participants from the HELIUS study were available for analysis. The HELIUS study contained individuals of European Dutch, South Asian Surinamese, African Surinamese, Ghanaian, Turkish, and Moroccan descent living in Amsterdam and were randomly sampled from the municipality register. Nineteen FHL2 polymorphisms were genotyped, and associations with lipid panels and T2D status were investigated. We observed that seven FHL2 polymorphisms associated nominally with a pro-diabetogenic lipid profile including triglyceride (TG), high-density and low-density lipoprotein-cholesterol (HDL-C and LDL-C), and total cholesterol (TC) concentrations, but not with blood glucose concentrations or T2D status in the complete HELIUS cohort upon correcting for age, gender, BMI, and ancestry. Upon stratifying for ethnicity, we observed that only two of the nominally significant associations passed multiple testing adjustments, namely, the association of rs4640402 with increased TG and rs880427 with decreased HDL-C concentrations in the Ghanaian population. Our results highlight the effect of ethnicity on pro-diabetogenic selected lipid biomarkers within the HELIUS cohort, as well as the need for more large multiethnic cohort studies.