Empirically Derived Dietary Patterns Using Robust Profile Clustering in the Hispanic Community Health Study/Study of Latinos

Prevalence and Risk Factors of Iron Deficiency and Anemia in Women of Reproductive Age in the Hispanic Community Health Study/Study of Latinos

Background

Women of reproductive age are at elevated risk of iron deficiency (ID) and anemia; in the United States, those of Hispanic/Latino background are at especially high risk. Causes of ID and anemia and variations in risk within Hispanic/Latino women of reproductive age are not well described.

Objectives

To characterize ID and anemia and their risk factors/markers in Hispanic/Latina women.

Methods

Data from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL) were used. HCHS/SOL enrolled participants in 2008-2011 at 4 sites: Bronx, Chicago, Miami, and San Diego; 5386 were non-pregnant women ages 18-49 y. Primary outcomes were ID (ferritin <30 μg/L) and anemia (hemoglobin <12 g/dL). Predictors were background/heritage, Short Acculturation Scale for Hispanics scores, years lived in the United States, and interview language, diet summarized in the Alternate Healthy Eating Index 2010, sociodemographic covariates, and study site. Main analyses used survey log binomial regressions adjusted for age, site, and smoking.

Results

Almost half (42%) of participants were of Mexican background, 15% of Cuban background, and <15% each were of Puerto Rican, Dominican, Central American, or South American background. ID prevalence was 34.4% overall but differed by site-background pairings. The lowest and highest prevalence were 26% and 42% among women of Cuban background in Miami and women of Mexican background in Chicago, respectively. Anemia prevalence was 16% and ranged from 8.9% (Central American background/Miami) to 22% (Dominican background/Bronx). Acculturation, sociodemographic, and diet variables examined did not explain observed prevalence differences by site/background.

Conclusions

Prevalence of ID and anemia were high among HCHS/SOL women and differed by field center and background. These differences highlight the importance of characterizing nutritional risk by background within Hispanic/Latino women.

Examining Generalizability across Different Surveys: Comparing Nutrient-Based Food Patterns and Their Cross-Sectional Associations with Cardiometabolic Health in the United States Hispanic/Latino Adults

Background

Ethnicity, cultural background, and geographic location differ significantly within the United States Hispanic/Latino population. These variations can greatly define diet and its relationship with cardiometabolic disease, thus influencing generalizability of results.

Objectives

We aimed to examine nutrient-based food patterns (NBFPs) of Hispanic/Latino adults and their association with cardiometabolic risk factors (dyslipidemia, hypertension, obesity, diabetes) across 2 United States population-based studies with differing sampling strategies.

Methods

Data were collected from Mexican or other Hispanic adult participants from 2007-2012 National Health and Nutrition Examination Survey (NHANES) (n = 3605) and 2007-2011 Hispanic Community Health Survey/Study of Latinos (HCHS/SOL, n = 14,416). NBFPs were derived using factor analysis on nutrient intake data estimated from 24-h dietary recalls and interpreted using common foods in which these nutrients are prominent. Cross-sectional associations between NBFPs (quintiles) and cardiometabolic risk factors, defined by clinical measures and self-report, were estimated using survey-weighted multivariable-adjusted logistic models, accounting for multiple testing.

Results

Five NBFPs were identified in both studies: 1) meats, 2) grains/legumes, 3) fruits/vegetables, 4) dairy, and 5) fats/oils. Associations with cardiometabolic risk factors differed by NBFP and study. In HCHS/SOL, the odds of diabetes were lower for persons in the highest quintile of meats NBFP (odds ratio [OR]: 0.73; 95% confidence interval [CI]: 0.58, 0.92) and odds were higher for those in the lowest quintile of fruits/vegetables (OR: 0.71; 95% CI: 0.55, 0.93) compared to those in the third (moderate intake) quintile. Those in the fourth quintile of dairy NBFP had higher odds of hypertension than those in the third quintile (OR: 1.31; 95% CI: 1.01, 1.70). In NHANES, the odds of hypertension were higher for those in the fourth quintile of dairy (OR: 1.88; 95% CI: 1.10, 3.24) than those in the third quintile.

Conclusions

Diet-disease relationships among Hispanic/Latino adults vary according to 2 population-based studies. These differences have research and practical implications when generalizing inferences on heterogeneous underrepresented populations.

Associations between sodium, potassium, and blood pressure: results from the Hispanic Community Health Study/Study of Latinos, a prospective cohort study

BACKGROUND

Although the subject of numerous studies, the associations between dietary sodium, potassium, and the ratio of dietary sodium to potassium with blood pressure are not clear-cut. In addition, there is a paucity of research on these relationships in prospective cohort studies with representation from diverse Hispanic/Latino adults.

OBJECTIVES

To evaluate the associations between dietary intake of sodium, potassium, and the ratio of dietary sodium to potassium and blood pressure in a diverse sample of Hispanics living in the United States.

METHODS

This analysis included 11,429 Hispanic/Latino participants of the prospective cohort Hispanic Community Health Study/Study of Latinos recruited between 2008 and 2011 in visit 1 who participated in a follow-up visit in 2014-2017. Dietary sodium and potassium intakes were averaged from 2 interviewer-administered 24-h diet recalls collected at visit 1. At both visits, blood pressure was measured 3 times in a seated position and averaged. We assessed the relationship between dietary sodium, potassium, and the sodium-to-potassium ratio with changes in systolic and diastolic blood pressure using survey-weighted multivariable-adjusted regression models.

RESULTS

At visit 1, the mean age was 41 y, and the mean sodium intake was 3203 mg/d. Each 500 mg/d sodium increment in intake was associated with an increase in systolic blood pressure (β: 0.35 [mmHg]; 95% confidence interval: 0.06, 0.63) and diastolic blood pressure (β: 0.45 [mmHg]; 95% confidence interval: 0.08, 0.82). Dietary potassium and the molar ratio of dietary sodium to potassium were not associated with changes in systolic or diastolic blood pressure.

CONCLUSIONS

Among a large sample of diverse United States Hispanic/Latino adults, higher sodium intake was associated with small increases in systolic blood pressure over 6 y. This research underscores the importance of dietary sodium reduction in maintaining lower blood pressure.

Identifying dietary consumption patterns from survey data: a Bayesian nonparametric latent class model

Dietary assessments provide the snapshots of population-based dietary habits. Questions remain about how generalisable those snapshots are in national survey data, where certain subgroups are sampled disproportionately. We propose a Bayesian overfitted latent class model to derive dietary patterns, accounting for survey design and sampling variability. Compared to standard approaches, our model showed improved identifiability of the true population pattern and prevalence in simulation. We focus application of this model to identify the intake patterns of adults living at or below the 130% poverty income level. Five dietary patterns were identified and characterised by reproducible code/data made available to encourage further research.

Collecting, Using, and Reporting Race and Ethnicity Information: Implications for Research in Nutrition Education, Practice, and Policy to Promote Health Equity

This report will describe approaches for collecting, analyzing, and reporting race and ethnicity information in nutrition education and behavior research, practice, and policy to advance health equity. Race and ethnicity information is used to describe study participants and compare nutrition and health-related outcomes. Depending on the study design, race and ethnicity categories are often defined by the research question or other standardized approaches. Participant self-reported data are more acceptable than researcher adjudicated identification data, which can add bias and/or error. Valid methods to collect, use, and report race and ethnicity information are foundational to publication quality, findings of value, contribution to the knowledge base, and health equity.