Search

Summary: In observational cohort studies, there is frequently interest in modeling longitudinal change in a biomarker (ie, physiological measure indicative of metabolic dysregulation or disease; eg, blood pressure) in the absence of treatment (ie, medication), and its association with modifiable risk factors expected to affect health (eg, body mass index). However, individuals may start treatment during the study period, and consequently biomarker values observed while on treatment may be different than those that would have been observed in the absence of treatment. If treated individuals are excluded from analysis, then effect estimates may be biased if treated individuals differ systematically from untreated individuals. We addressed this concern in the setting of the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), an observational cohort study that employed a complex survey sampling design to enable inference to a finite target population. We considered biomarker values measured while on treatment to be missing data, and applied missing data methodology (inverse probability weighting (IPW) and doubly robust estimation) to this problem. The proposed methods leverage information collected between study visits on when individuals started treatment, by adapting IPW and doubly robust approaches to model the treatment mechanism using survival analysis methods. This methodology also incorporates sampling weights and uses a bootstrap approach to estimate standard errors accounting for the complex survey sampling design. We investigated variance estimation for these methods, conducted simulation studies to assess statistical performance in finite samples, and applied the methodology to model temporal change in blood pressure in HCHS/SOL. [ABSTRACT FROM AUTHOR]

Background: Associations between preconception cardiometabolic markers and birth outcomes have been noted, but data are scarce for Hispanics/Latinos. We examined the association between preconception cardiometabolic markers, birthweight and preterm birth among U.S. Hispanic/Latina women. Materials and Methods: The Hispanic Community Health Study/Study of Latinos is a cohort study of U.S. adults 18–74 years of age, including 3,798 women of reproductive age (18–44 years) from four field centers representing Hispanic/Latino backgrounds of Cuban, Dominican, Mexican, Puerto Rican, Central American, and South American. A baseline clinic examination (2008–2011) and a second clinic examination (2014–2017), including ascertainment of birth outcomes, allowed for identification of 517 singleton live births between the exams. Preconception cardiometabolic markers included abdominal obesity (waist circumference ≥88 cm), body mass index >30 kg/m2, high blood pressure (systolic ≥120 mmHg and diastolic ≥80 mmHg), elevated triglycerides (≥150 mg/dL), low high-density lipoprotein cholesterol (<50 mg/dL), elevated fasting glucose (≥100 mg/dL), and insulin. Complex survey linear regression modeled the association between cardiometabolic markers and birthweight-for-gestational age z-score; complex survey logistic regression modeled the association with preterm birth. Analyses adjusted for Hispanic/Latina background, field center, years between baseline and birth, age, and nulliparity. Results: In adjusted linear regression models, elevated fasting glucose was associated with higher birthweight z-scores (β = 0.56, 95% confidence interval [95% CI] 0.14 to 0.99), even after further adjustment for maternal percent body fat (β = 0.53, 95% CI 0.10 to 0.95). In adjusted logistic regression models, high blood pressure (odds ratio [OR] = 2.57, 95% CI 1.13 to 5.88) and increased insulin (OR = 1.50, 95% CI 1.06 to 2.14, for a 10 mU/L increase) were associated with higher odds for preterm birth. Conclusions: Infant birthweight and preterm birth may be influenced by selected cardiometabolic risk factors before pregnancy among Hispanic/Latina women. [ABSTRACT FROM AUTHOR]

Missing data are common in longitudinal cohort studies and can lead to bias, particularly in studies with informative missingness. Many common methods for handling informatively missing data in survey samples require correctly specifying a model for missingness. Although doubly robust methods exist to provide unbiased regression coefficients in the presence of missing outcome data, these methods do not account for correlation due to clustering inherent in longitudinal or cluster‐sampled studies. In this work, we developed a doubly robust method to estimate the regression of an outcome on a predictor in the presence of missing multilevel data on the outcome, which results in consistent estimation of regression coefficients assuming correct specification of either (1) the probability of missingness or (2) the outcome model. This method involves specification of separate hierarchical models for missingness and for the outcome, conditional on observed auxiliary variables and cluster‐specific random effects, to account for correlation among observations. We showed this proposed estimator is doubly robust and derived its asymptotic distribution, conducted simulation studies to compare the method to an existing doubly robust method developed for independent data, and applied the method to data from the China Health and Nutrition Survey, an ongoing multilevel longitudinal cohort study. [ABSTRACT FROM AUTHOR]

Background: United States (US) Hispanic/Latinos experience a disproportionate burden of obesity, which may in part be related to demographic or sociocultural factors, including acculturation to an US diet or inactive lifestyle. Therefore, we sought to describe the association between adulthood weight histories and demographic and sociocultural factors in a large diverse community-based cohort of US Hispanic/Latinos.Methods: We estimated the effect of several factors on weight gain across adulthood, using multivariable linear mixed models to leverage 38,759 self-reported current body weights and weight histories recalled for 21, 45 and 65 years of age, from 15,203 adults at least 21 years of age at the baseline visit of the Hispanic Community Health Study/Study of Latinos (2008-2011).Results: The average rate of weight gain was nearly 10 kg per decade in early adulthood, but slowed to < 5 kg a decade among individuals 60+ years of age. Birth cohort, gender, nativity or age at immigration, Hispanic/Latino background, and study site each significantly modified the form of the predicted adulthood weight trajectory. Among immigrants, weight gain during the 5 years post-migration was on average 0.88 kg (95% CI: 0.04, 1.72) greater than the weight gain during the 5 years prior. The rate of weight gain appeared to slow after 15 years post-migration.Conclusions: Using self-reported and weight history data in a diverse sample of US Hispanic/Latinos, we revealed that both demographic and sociocultural factors were associated with the patterning of adulthood weight gain in this sample. Given the steep rate of weight gain in this population and the fact that many Hispanic/Latinos living in the US immigrated as adults, efforts to promote weight maintenance across the life course, including after immigration, should be a top priority for promoting Hispanic/Latino health and addressing US health disparities more broadly. [ABSTRACT FROM AUTHOR]

OBJECTIVE: To describe the individual and joint associations of baseline factors with glycemia, and also with differential effectiveness of medications added to metformin. RESEARCH DESIGN AND METHODS: Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) participants (with type 2 diabetes diagnosed for <10 years, on metformin, and with HbA1c 6.8–8.5%; N = 5,047) were randomly assigned to a basal insulin (glargine), sulfonylurea (glimepiride), glucagon-like peptide 1 agonist (liraglutide), or dipeptidyl peptidase 4 inhibitor (sitagliptin). The glycemic outcome was HbA1c ≥7.0%, subsequently confirmed. Univariate and multivariate regression and classification and regression tree (CART) analyses were used to assess the association of baseline factors with the glycemic outcome at years 1 and 4. RESULTS: In univariate analyses at baseline, younger age (<58 years), Hispanic ethnicity, higher HbA1c, fasting glucose, and triglyceride levels, lower insulin secretion, and relatively greater insulin resistance were associated with the glycemic outcome at years 1 and/or 4. No factors were associated with differential effectiveness of the medications by year 4. In multivariate analyses, treatment group, younger age, and higher baseline HbA1c and fasting glucose were jointly associated with the glycemic outcome by year 4. The superiority of glargine and liraglutide at year 4 persisted after multiple baseline factors were controlled for. CART analyses indicated that failure to maintain HbA1c <7% by year 4 was more likely for younger participants and those with baseline HbA1c ≥7.4%. CONCLUSIONS: Several baseline factors were associated with the glycemic outcome but not with differential effectiveness of the four medications. Failure to maintain HbA1c <7% was largely driven by younger age and higher HbA1c at baseline. Factors that predict earlier glycemic deterioration could help in targeting patients for more aggressive management. [ABSTRACT FROM AUTHOR]

OBJECTIVE: To compare the effects of insulin sensitivity and β-cell function over time on HbA1c and durability of glycemic control in response to dual therapy. RESEARCH DESIGN AND METHODS: GRADE participants were randomized to glimepiride (n = 1,254), liraglutide (n = 1,262), or sitagliptin (n = 1,268) added to baseline metformin and followed for mean ± SD 5.0 ± 1.3 years, with HbA1c assessed quarterly and oral glucose tolerance tests at baseline, 1, 3, and 5 years. We related time-varying insulin sensitivity (HOMA 2 of insulin sensitivity [HOMA2-%S]) and early (0–30 min) and total (0–120 min) C-peptide (CP) responses to changes in HbA1c and glycemic failure (primary outcome HbA1c ≥7% [53 mmol/mol] and secondary outcome HbA1c >7.5% [58 mmol/mol]) and examined differential treatment responses. RESULTS: Higher HOMA2-%S was associated with greater initial HbA1c lowering (3 months) but not subsequent HbA1c rise. Greater CP responses were associated with a greater initial treatment response and slower subsequent HbA1c rise. Higher HOMA2-%S and CP responses were each associated with lower risk of primary and secondary outcomes. These associations differed by treatment. In the sitagliptin group, HOMA2-%S and CP responses had greater impact on initial HbA1c reduction (test of heterogeneity, P = 0.009 HOMA2-%S, P = 0.018 early CP, P = 0.001 total CP) and risk of primary outcome (P = 0.005 HOMA2-%S, P = 0.11 early CP, P = 0.025 total CP) but lesser impact on HbA1c rise (P = 0.175 HOMA2-%S, P = 0.006 early CP, P < 0.001 total CP) in comparisons with the glimepiride and liraglutide groups. There were no differential treatment effects on secondary outcome. CONCLUSIONS: Insulin sensitivity and β-cell function affected treatment outcomes irrespective of drug assignment, with greater impact in the sitagliptin group on initial (short-term) HbA1c response in comparison with the glimepiride and liraglutide groups. [ABSTRACT FROM AUTHOR]

OBJECTIVE: To describe rescue insulin use and associated factors in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE). RESEARCH DESIGN AND METHODS: GRADE participants (type 2 diabetes duration <10 years, baseline A1C 6.8%–8.5% on metformin monotherapy, N = 5,047) were randomly assigned to insulin glargine U-100, glimepiride, liraglutide, or sitagliptin and followed quarterly for a mean of 5 years. Rescue insulin (glargine or aspart) was to be started within 6 weeks of A1C >7.5%, confirmed. Reasons for delaying rescue insulin were reported by staff-completed survey. RESULTS: Nearly one-half of GRADE participants (N = 2,387 [47.3%]) met the threshold for rescue insulin. Among participants assigned to glimepiride, liraglutide, or sitagliptin, rescue glargine was added by 69% (39% within 6 weeks). Rescue aspart was added by 44% of glargine-assigned participants (19% within 6 weeks) and by 30% of non-glargine-assigned participants (14% within 6 weeks). Higher A1C values were associated with adding rescue insulin. Intention to change health behaviors (diet/lifestyle, adherence to current treatment) and not wanting to take insulin were among the most common reasons reported for not adding rescue insulin within 6 weeks. CONCLUSIONS: Proportionately, rescue glargine, when required, was more often used than rescue aspart, and higher A1C values were associated with greater rescue insulin use. Wanting to use noninsulin strategies to improve glycemia was commonly reported, although multiple factors likely contributed to not using rescue insulin. These findings highlight the persistent challenge of intensifying type 2 diabetes treatment with insulin, even in a clinical trial. [ABSTRACT FROM AUTHOR]

The nutritional status of women in the preconception period is of paramount importance due to its role in reproduction. Our aim was to assess overall diet quality during the preconception period and its association with infant birth weight adjusted for gestational age (GA). This is an observational longitudinal cohort of Hispanic people living in the United States. Data are from the Hispanic Community Health Study/Study of Latinos baseline (2008-2011) and second clinic examinations (2014-2017). Included are the first 497 singleton live-born infants among the 2,556 women (younger than 45 years) who attended the second visit. Field sites were located in Miami, FL; Bronx, NY; Chicago, IL; and San Diego, CA, and represent individuals with heritage from Cuba, Dominican Republic, Mexico, Puerto Rico, and Central and South America. Diet assessment included two 24-hour recalls from baseline. The 2010 Healthy Eating Index (HEI-2010) was used to measure diet quality, with higher scores indicating better quality. Complex survey linear regression estimated the association between HEI-2010 scores (continuous variable and categorized into tertiles) and birth-weight z score and birth weight for GA percentile. Mean (standard deviation) age of women was 25.8 (5.2) years and 36.4% were classified as underweight or normal weight, 30.0% were overweight, and 33.6% had obesity at baseline. Mean (standard deviation) HEI-2010 score was 56.5 (13.4), and by weight classifications was 54.4 (14.1) for underweight or normal weight and 57.7 (12.8) for overweight or obesity. Median (interquartile range) birth-weight z score was 0.5 (interquartile range [IQR], –0.2 to 1.3) overall and 0.2 (IQR, –0.5 to 1.0), 0.6 (IQR, –0.2 to 1.3), and 0.5 (IQR, –0.2 to 1.4) for the first, second, and third HEI-2010 tertile, respectively. Median birth weight for GA percentile was 68.2 (IQR, 40.2 to 89.7) overall, and 56.8 (IQR, 29.6 to 85.0), 71.5 (IQR, 42.8 to 90.0), and 70.1 (IQR, 42.9 to 91.2) by HEI-2010 tertile. In adjusted models, the highest tertile of the HEI-2010 score was associated with a higher birth-weight z score and birth weight for GA percentile, and the continuous HEI-2010 score was only associated with birth weight for GA percentile. Preconception body mass index (calculated as kg/m2) did not modify these associations. Overall diet quality, as measured by the HEI-2010, in the preconception period is associated with infant birth weight adjusted for GA among US Hispanic and Latina women. [ABSTRACT FROM AUTHOR]

Design‐based analysis, which accounts for the design features of the study, is commonly used to conduct data analysis in studies with complex survey sampling, such as the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). In this type of longitudinal study, attrition has often been a problem. Although there have been various statistical approaches proposed to handle attrition, such as inverse probability weighting (IPW), non‐response cell weighting (NRCW), multiple imputation (MI), and full information maximum likelihood (FIML) approach, there has not been a systematic assessment of these methods to compare their performance in design‐based analyses. In this article, we perform extensive simulation studies and compare the performance of different missing data methods in linear and generalized linear population models, and under different missing data mechanism. We find that the design‐based analysis is able to produce valid estimation and statistical inference when the missing data are handled appropriately using IPW, NRCW, MI, or FIML approach under missing‐completely‐at‐random or missing‐at‐random missing mechanism and when the missingness model is correctly specified or over‐specified. We also illustrate the use of these methods using data from HCHS/SOL. [ABSTRACT FROM AUTHOR]

Although early sexual initiation has been linked to negative outcomes, it is unknown whether these effects are causal. In this study, we use propensity score methods to estimate the causal effect of early sexual initiation on young adult sexual risk behaviors and health outcomes using data from the National Longitudinal Study of Adolescent to Adult Health. We found that early sexual initiation predicted having two or more partners (for both males and females) and having a sexually transmitted infection in the past year (females only) but did not predict depressive symptoms in the past week (for either gender). These results underscore the importance of continued programmatic efforts to delay age of sexual initiation, particularly for females. [ABSTRACT FROM AUTHOR]

Background: The comparative effectiveness of glucose-lowering medications for use with metformin to maintain target glycated hemoglobin levels in persons with type 2 diabetes is uncertain.Methods: In this trial involving participants with type 2 diabetes of less than 10 years' duration who were receiving metformin and had glycated hemoglobin levels of 6.8 to 8.5%, we compared the effectiveness of four commonly used glucose-lowering medications. We randomly assigned participants to receive insulin glargine U-100 (hereafter, glargine), the sulfonylurea glimepiride, the glucagon-like peptide-1 receptor agonist liraglutide, or sitagliptin, a dipeptidyl peptidase 4 inhibitor. The primary metabolic outcome was a glycated hemoglobin level, measured quarterly, of 7.0% or higher that was subsequently confirmed, and the secondary metabolic outcome was a confirmed glycated hemoglobin level greater than 7.5%.Results: A total of 5047 participants (19.8% Black and 18.6% Hispanic or Latinx) who had received metformin for type 2 diabetes were followed for a mean of 5.0 years. The cumulative incidence of a glycated hemoglobin level of 7.0% or higher (the primary metabolic outcome) differed significantly among the four groups (P<0.001 for a global test of differences across groups); the rates with glargine (26.5 per 100 participant-years) and liraglutide (26.1) were similar and lower than those with glimepiride (30.4) and sitagliptin (38.1). The differences among the groups with respect to a glycated hemoglobin level greater than 7.5% (the secondary outcome) paralleled those of the primary outcome. There were no material differences with respect to the primary outcome across prespecified subgroups defined according to sex, age, or race or ethnic group; however, among participants with higher baseline glycated hemoglobin levels there appeared to be an even greater benefit with glargine, liraglutide, and glimepiride than with sitagliptin. Severe hypoglycemia was rare but significantly more frequent with glimepiride (in 2.2% of the participants) than with glargine (1.3%), liraglutide (1.0%), or sitagliptin (0.7%). Participants who received liraglutide reported more frequent gastrointestinal side effects and lost more weight than those in the other treatment groups.Conclusions: All four medications, when added to metformin, decreased glycated hemoglobin levels. However, glargine and liraglutide were significantly, albeit modestly, more effective in achieving and maintaining target glycated hemoglobin levels. (Funded by the National Institute of Diabetes and Digestive and Kidney Diseases and others; GRADE ClinicalTrials.gov number, NCT01794143.). [ABSTRACT FROM AUTHOR]

Confounding present in observational data impede community psychologists' ability to draw causal inferences. This paper describes propensity score methods as a conceptually straightforward approach to drawing causal inferences from observational data. A step-by-step demonstration of three propensity score methods-weighting, matching, and subclassification-is presented in the context of an empirical examination of the causal effect of preschool experiences (Head Start vs. parental care) on reading development in kindergarten. Although the unadjusted population estimate indicated that children with parental care had substantially higher reading scores than children who attended Head Start, all propensity score adjustments reduce the size of this overall causal effect by more than half. The causal effect was also defined and estimated among children who attended Head Start. Results provide no evidence for improved reading if those children had instead received parental care. We carefully define different causal effects and discuss their respective policy implications, summarize advantages and limitations of each propensity score method, and provide SAS and R syntax so that community psychologists may conduct causal inference in their own research. [ABSTRACT FROM AUTHOR]

Objective: We evaluated the effect of optimizing metformin dosing on glycemia and body weight in type 2 diabetes.Research Design and Methods: This was a prespecified analysis of 6,823 participants in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) taking metformin as the sole glucose-lowering drug who completed a 4- to 14-week (mean ± SD 7.9 ± 2.4) run-in in which metformin was adjusted to 2,000 mg/day or a maximally tolerated lower dose. Participants had type 2 diabetes for <10 years and an HbA1c ≥6.8% (51 mmol/mol) while taking ≥500 mg of metformin/day. Participants also received diet and exercise counseling. The primary outcome was the change in HbA1c during run-in.Results: Adjusted for duration of run-in, the mean ± SD change in HbA1c was -0.65 ± 0.02% (-7.1 ± 0.2 mmol/mol) when the dose was increased by ≥1,000 mg/day, -0.48 ± 0.02% (-5.2 ± 0.2 mmol/mol) when the dose was unchanged, and -0.23 ± 0.07% (-2.5 ± 0.8 mmol/mol) when the dose was decreased (n = 2,169, 3,548, and 192, respectively). Higher HbA1c at entry predicted greater reduction in HbA1c (P < 0.001) in univariate and multivariate analyses. Weight loss adjusted for duration of run-in averaged 0.91 ± 0.05 kg in participants who increased metformin by ≥1,000 mg/day (n = 1,894).Conclusions: Optimizing metformin to 2,000 mg/day or a maximally tolerated lower dose combined with emphasis on medication adherence and lifestyle can improve glycemia in type 2 diabetes and HbA1c values ≥6.8% (51 mmol/mol). These findings may help guide efforts to optimize metformin therapy among persons with type 2 diabetes and suboptimal glycemic control. [ABSTRACT FROM AUTHOR]

Background: HCHS/SOL showed that a sizeable proportion of Hispanics/Latinos (80% of men, 71% of women) had at least 1 major CVD risk factor (RF), with marked variations by ethnic background. Little is known about changes in CVD RF profiles over time in this population. Objective: To describe ~6-year changes in CVD RF profiles and examine associations with demographic and socioeconomic/sociocultural factors. Methods: HCHS/SOL is a multi-center prospective community-based study of 16,415 Hispanic/Latino adults of Cuban, Dominican, Mexican, Puerto Rican, Central American, and South American backgrounds, aged 18-74 at Visit 1 (2008-11). Visit 2 (2014-17) is ongoing and 8,413 persons (~60% of the cohort to be studied) attended as of Sept. 2016. Analyses included 7,789 men and women with complete data. CVD RF profiles were defined as having 0 (0RF) or any 1 or more (1+RF) of the following: hypercholesterolemia, hypertension, obesity, diabetes, and smoking (see definitions in Table). Adjusted percent increases in number of RFs were computed. Multinomial logistic regression was used to examine associations of Visit 1 characteristics with change in RFs, adjusted for sociodemographic, sociocultural, and lifestyle factors. Results: After 5.8 years, 29% of men and 27% of women had increases in number of RFs; changes occurred more frequently in persons with 1+ RF at Visit 1 than in those with 0RF and varied by background (Table). Significantly higher odds of increase in number of RFs (vs. 0RF at both visits) were seen with older age (OR=1.07, 95% CI=1.06-1.08 per 1 yr) and male sex (1.74, 1.37-2.21); lower odds with higher education (0.60; 0.44-0.83 for > vs. < high school) and income (0.56, 0.38-0.81 for >$50,000 vs. <$20,000); acculturation did not relate to RF changes. Conclusions: In just a few years, a large percent of US Hispanic/Latino adults had an increase in number of adverse RFs, which varied by background; age, sex, education, and income were associated with RF increases. Greater efforts are needed to prevent CVD RFs in this population. [ABSTRACT FROM AUTHOR]