A longitudinal analysis of nondaily smokers: the Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

Risk Factors for Cardiovascular Disease: Knowledge Gained from the Hispanic Community Health Study/Study of Latinos

PURPOSE OF REVIEW

The Hispanic Community Health Study/Study of Latinos (HCHS/SOL) has made important contributions on the prevalence of and factors associated with cardiovascular disease (CVD) risk factors among diverse Hispanic/Latino adults in the US. This article summarizes the knowledge gained thus far on major CVD risk factors from this landmark study.

RECENT FINDINGS

HCHS/SOL demonstrated the sizeable burdens of CVD risk in all major Hispanic/Latino groups in the US, as well as the marked variations in prevalence of hypertension, hypercholesterolemia, diabetes, obesity, and smoking by sex and background. It also identified sociodemographic, lifestyle, and sociocultural characteristics associated with risk factors. HCHS/SOL has yielded an expanding body of literature on characteristics associated with adverse CVD risk factors in this population. Long-term follow-up of this cohort will shed further light on the observed heterogeneity in CVD risk across Hispanic/Latino groups and identify specific risk/protective factors driving these variations.

Acculturation level and change in cigarette consumption behaviors among diverse Hispanics/Latinos: the Hispanic Community Health Study/Study of Latinos

PURPOSE

To study associations between language acculturation level and changes in cigarette consumption among the diverse and growing U.S.-based Hispanic/Latino population and inform culturally tailored smoking prevention and cessation strategies.

METHODS

In the Hispanic Community Health Survey/Study of Latinos cohort, we used cigarette consumption behaviors at baseline (2008-2011) and follow-up (2014-2017) and a modified Short Acculturation Scale for Hispanics (SASH) language subscale to measure associations of language acculturation (unidimensional) with changes in cigarette consumption and quitting rates. Weighted multivariable linear and logistic regressions were stratified by daily (n = 1397) and nondaily (n = 633) smoking, and either sex, educational attainment, or migration status.

RESULTS

Smokers at baseline (n = 2030) on average were aged 42 years old (SE = 0.5) with a mean SASH-language score of 2.3 (SE = 0.1; range = 1-5), indicating more Spanish language use. Among male daily smokers, we observed increases in smoked cigarettes-per-day (CPD) with unit increases in SASH-language score (1.08, 95% CI: 0.24-1.92). Associations with acculturation trended toward greater increases in CPD and lower odds of quitting as educational attainment increased.

CONCLUSIONS

Language acculturation level is an important determinant for increased smoking behaviors, particularly among men. Our findings are significant in informing smoking reduction programs for the Hispanic/Latino population.

Web-delivered Acceptance and Commitment Therapy (ACT) for smoking cessation: Is it engaging and efficacious for US Hispanic/Latinx adult smokers?

Hispanic/Latinx adult smokers in the United States (US) face barriers to receiving and utilizing evidenced-based cessation treatments compared with other racial/ethnic groups. The lack of efficacious and accessible smoking cessation treatments for this population further contributes to such smoking disparities. In a secondary analysis, we explored the efficacy of an Acceptance and Commitment Therapy (ACT)-based website (WebQuit.org) versus a US Clinical Practice Guidelines (USCPG)-based website (Smokefree.gov) for smoking cessation in a subset of Hispanic/Latinx adult participants enrolled in the WebQuit trial. Of the 2,637 participants who were randomized in the parent trial, 222 were Hispanic/Latinx (n = 101 in WebQuit, n = 121 in Smokefree). Smoking cessation outcomes were measured at 3, 6, and 12-months. The primary outcome was self-reported complete-case 30-day point prevalence abstinence (PPA) at 12-months. Treatment engagement and satisfaction, change in acceptance of urges to smoke, and commitment to quitting smoking were compared across conditions. Retention rate was 88% at 12-months. WebQuit participants had higher odds of smoking cessation compared to Smokefree participants at 12-months (40% vs. 25%; OR = 1.93 95% CI: 1.04, 3.59). Findings were similar using multiple imputation. WebQuit participants engaged more with the website than Smokefree participants through multiple indicators of engagement, including spending more time using the website (IRR = 2.32; 95% CI: 1.68, 3.20). Although WebQuit participants engaged more with the website than Smokefree participants, there was no evidence that differences in quit rates were mediated by engagement level. This study provides initial empirical evidence that digital interventions may be efficacious for helping Hispanic/Latinx adults quit smoking.

Hypothesized Explanations for the Observed Lung Cancer Survival Benefit Among Hispanics/Latinos in the United States

Hispanic/Latino ethnicity is associated with improved survival from non-small cell lung cancer compared to that for non-Hispanic Whites even though Hispanics/Latinos are more likely to potentially have inferior access-to-care and experience greater health disparities. To this end, we conducted a literature review to identify possible explanations for this survival benefit, including the role of chronic obstructive pulmonary disease and cardiovascular diseases, genetic variation, cultural influences, and immigration factors. Overall, intermittent smoking patterns, genetic variation, co-morbidities, and cultural influences were all factors likely to partially explain this survival benefit. On the other hand, immigration factors, acculturation, and access-to-care were less likely to support the survival advantage. Future research should analyze relevant Hispanic/Latino subgroups (e.g., Mexican, Puerto Rican, Cuban, Dominican, Central American, South American) and specifically focus on the relationship between Hispanic/Latino ethnicity and different lung cancer subtypes. If the Hispanic/Latino mortality benefit observed in lung cancer truly exists, a better understanding of the underlying mechanism(s) may help extend these benefits to other ethnic and racial groups.

Bias and fairness assessment of a natural language processing opioid misuse classifier: detection and mitigation of electronic health record data disadvantages across racial subgroups

OBJECTIVES

To assess fairness and bias of a previously validated machine learning opioid misuse classifier.

MATERIALS & METHODS

Two experiments were conducted with the classifier's original (n = 1000) and external validation (n = 53 974) datasets from 2 health systems. Bias was assessed via testing for differences in type II error rates across racial/ethnic subgroups (Black, Hispanic/Latinx, White, Other) using bootstrapped 95% confidence intervals. A local surrogate model was estimated to interpret the classifier's predictions by race and averaged globally from the datasets. Subgroup analyses and post-hoc recalibrations were conducted to attempt to mitigate biased metrics.

RESULTS

We identified bias in the false negative rate (FNR = 0.32) of the Black subgroup compared to the FNR (0.17) of the White subgroup. Top features included "heroin" and "substance abuse" across subgroups. Post-hoc recalibrations eliminated bias in FNR with minimal changes in other subgroup error metrics. The Black FNR subgroup had higher risk scores for readmission and mortality than the White FNR subgroup, and a higher mortality risk score than the Black true positive subgroup (P < .05).

DISCUSSION

The Black FNR subgroup had the greatest severity of disease and risk for poor outcomes. Similar features were present between subgroups for predicting opioid misuse, but inequities were present. Post-hoc mitigation techniques mitigated bias in type II error rate without creating substantial type I error rates. From model design through deployment, bias and data disadvantages should be systematically addressed.

CONCLUSION

Standardized, transparent bias assessments are needed to improve trustworthiness in clinical machine learning models.

Association of Diet Quality Indices with Longitudinal Changes in Kidney Function in U.S. Hispanics/Latinos: Findings from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

Background

Recent studies suggest an association between diet quality and incident CKD. However, Hispanics/Latinos were under-represented in these studies. We examined the relationship of diet quality with change in kidney function in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL).

Methods

Individuals who participated in HCHS/SOL visits 1 (2008-2011) and 2 (2014-2017) were analyzed (n=9921). We used Alternate Healthy Eating Index 2010 (AHEI-2010), Dietary Approaches to Stop Hypertension (DASH), and Mediterranean Diet (MeDS) scores as measures of dietary quality. Scores were calculated from two 24-hour dietary recalls administered at visit 1 and categorized into quartiles of each dietary score (higher quartiles correspond to a healthier diet). Kidney function was assessed at both visits using eGFR and urine albumin-creatinine ratio (UACR). Annualized change was computed as the difference in eGFR or UACR between visits divided by follow-up time in years. Weighted linear-regression models were used to examine the association between quartiles of each dietary quality index and annualized change in eGFR and UACR, adjusted for potential confounders.

Results

At visit 1, the mean (SD) age of participants was 41 (0.28) years, and 56% were female. The baseline mean eGFR was 107.1 ml/min per 1.73 m², and baseline median UACR was 6.1 mg/g. On average, eGFR declined by 0.65 ml/min per 1.73 m² per year, and UACR increased by 0.79 mg/g per year over a 6-year period. Lower AHEI-2010 quartiles were associated with eGFR decline in a dose-response manner (P trend=0.02). Higher AHEI-2010 quartiles showed a trend toward lower annualized change in UACR, but the result did not reach significance. Neither MeDS nor DASH scores were associated with eGFR decline or change in UACR.

Conclusions

Unhealthy diet, assessed at baseline by AHEI-2010, was associated with kidney-function decline over 6 years. Improving the quality of foods and nutrients according to the AHEI-2010 may help maintain kidney function in the Hispanic/Latino community.