Faculty Bibliography

Low serum magnesium has been associated with kidney function decline in persons with diabetes as well as cardiovascular disease in the general population. As the association of serum magnesium with incident kidney disease in the general population is unknown, we assessed this in 13,226 participants (aged 45-65) in the Atherosclerosis Risk in Communities study with baseline estimated glomerular filtration rate of at least 60 ml/min per 1.73 m(2) in years 1987-89 and followed through 2010. The risks for incident chronic kidney disease (CKD) and end-stage renal disease (ESRD) associated with baseline total serum magnesium levels were evaluated using Cox regression. There were 1965 CKD and 208 ESRD events during a median follow-up of 21 years. In adjusted analysis, low serum magnesium levels (0.7 mmol/l or less) had significant associations with incident CKD and ESRD compared with the highest quartile with adjusted hazard ratio of 1.58 (95% CI: 1.35-1.87) for CKD and 2.39 (95% CI: 1.61-3.56) for ESRD. These associations remained significant after excluding users of diuretics and across subgroups stratified by hypertension, diabetes, and self-reported race. Thus, in a large sample of middle-aged adults, low total serum magnesium was independently associated with incident CKD and ESRD. Further studies are needed to determine whether modification of serum magnesium levels might alter subsequent incident kidney disease rates.

BACKGROUND:Epidemiologic data for cardiac abnormality predating decreased kidney function are sparse. We investigated the associations of high-sensitivity cardiac troponin T (hs-cTnT) and N-terminal pro-brain natriuretic peptide (NT-proBNP) with end-stage renal disease (ESRD) risk in a community-based cohort. STUDY DESIGN/METHODS:A prospective cohort study. SETTING & PARTICIPANTS/METHODS:10,749 white and black participants at the fourth visit (1996-1998) of the Atherosclerosis Risk in Communities (ARIC) Study with follow-up through 2010. PREDICTOR/METHODS:hs-cTnT (3, 6, 9, and 14ng/L) and NT-proBNP (41.6, 81.0, 142.5, and 272.5pg/mL) levels were divided into 5 categories at the same percentiles (32th, 57th, 77th, and 91th; corresponding to ordinary thresholds of hs-cTnT), with the lowest category as a reference. OUTCOMES/RESULTS:Incident ESRD defined as initiation of dialysis therapy, transplantation, or death due to kidney disease. MEASUREMENTS/METHODS:Relative risk and risk prediction of ESRD according to hs-cTnT and NT-proBNP levels based on Cox proportional hazards models. RESULTS:During a median follow-up of 13.1 years, 235 participants developed ESRD (1.8 cases/1,000 person-years). hs-cTnT and NT-proBNP levels were associated with ESRD risk independently of each other and of potential confounders, including kidney function and albuminuria (adjusted HRs for highest category, 4.43 [95% CI, 2.43-8.09] and 2.28 [95% CI, 1.44-3.60], respectively). For hs-cTnT level, the association was significant even at the third category (HR for 6-8ng/L of hs-cTnT, 2.74 [95% CI, 1.54-4.88]). Their associations were largely consistent even among persons without decreased kidney function or history of cardiovascular disease. hs-cTnT and NT-proBNP levels both significantly improved ESRD prediction (C statistic differences of 0.0084 [95% CI, 0.0005-0.0164] and 0.0045 [95% CI, 0.0004-0.0087], respectively, from 0.884 with conventional risk factors). LIMITATIONS/CONCLUSIONS:Relatively small number of ESRD cases and single measurement of hs-cTnT and NT-proBNP. CONCLUSIONS:hs-cTnT and NT-proBNP levels independently predicted ESRD risk in the general population, with more evident results for hs-cTnT. These results suggest the involvement of cardiac abnormality, particularly cardiac injury, in the progression of reduced kidney function and/or may reflect the useful property of hs-cTnT as an end-organ damage marker.

BACKGROUND & AIMS/OBJECTIVE:Patients with non-alcoholic fatty liver disease (NAFLD) are thought to be at increased risk of cardiovascular morbidity and mortality. However, the relationships between NAFLD and subclinical myocardial injury or structural heart disease are unknown. METHODS:We conducted a cross-sectional analysis of 8668 participants from the Atherosclerosis Risk in Communities (ARIC) Study, who showed no clinical evidence of cardiovascular disease. We used levels of liver enzymes (alanine aminotransferase [ALT], aspartate aminotransferase [AST] and gamma-glutamyl transpeptidase [GGT]), in the context of no history of elevated alcohol consumption as non-invasive surrogates of NAFLD. We used highly sensitive cardiac troponin T (hs-cTnT) and N-terminal pro-Brain natriuretic peptide (NT-proBNP) as biomarkers of myocardial damage and function. RESULTS:In this population-based study (mean age 63 years, 60% women, 78% white), higher levels of ALT, AST, and GGT, even within the normal range, were significantly and independently associated with detectable (hs-cTnT >3 ng/L) and elevated (hs-cTnT ⩾14 ng/L) concentrations of hs-cTnT. The adjusted odds ratios (95% confidence interval) for elevated liver enzymes (vs. normal levels) with elevated hs-cTnT were: 1.65 (1.28-2.14) for ALT, 1.90 (1.36-2.68) for AST, and 1.55 (1.13-2.12) for GGT. Furthermore, there was evidence for inverse associations of ALT and AST with NT-proBNP. CONCLUSIONS:Our results suggest that elevated liver enzyme levels in the absence of elevated alcohol consumption may be associated with subclinical myocardial injury. The inverse association between NT-proBNP and both ALT and AST supports the recently described metabolic role of natriuretic peptides.

Racial/ethnic disparities in the prevalence of type 2 diabetes mellitus (T2DM) are well documented and until recently, research has focused almost exclusively on individual-based determinants as potential contributors to these disparities (health behaviors, biological/genetic factors, and individual-level socio-demographics). Research on the role of neighborhood characteristics in relation to racial/ethnic disparities in T2DM is very limited. Therefore, the aim of this research is to identify and estimate the contribution of specific aspects of neighborhoods that may be associated with racial/ethnic disparities in T2DM. Data from the Boston Area Community Health III Survey (N = 2764) was used in this study, which is a community-based random-sample survey of adults in Boston, Massachusetts from three racial/ethnic groups (Black, Hispanic, and White). We applied two-level random intercepts logistic regression to assess the associations between race/ethnicity, neighborhood characteristics (census tract socioeconomic status, racial composition, property and violent crime, open space, geographic proximity to grocery stores, convenience stores, and fast food, and neighborhood disorder) and prevalent T2DM (fasting glucose > 125 mg/dL, HbA1c >/= 6.5%, or self-report of a T2DM diagnosis). Black and Hispanic participants had 2.89 times and 1.48 times the odds of T2DM as White participants, respectively. Multilevel models indicated a significant between-neighborhood variance estimate of 0.943, providing evidence of neighborhood variation. Individual demographics (race/ethnicity, age and gender) explained 22.3% of the neighborhood variability in T2DM. The addition of neighborhood-level variables to the model had very little effect on the magnitude of the racial/ethnic disparities and on the between-neighborhood variability. For example, census tract poverty explained less than 1% and 6% of the excess odds of T2DM among Blacks and Hispanics and only 1.8% of the neighborhood variance in T2DM. While the findings of this study overall suggest that neighborhood factors are not a major contributor to racial/ethnic disparities in T2DM, further research is needed including data from other geographic locations.

Policies supporting breastfeeding vary by state, but little is known about the geographical aspects of this variation. This study describes state breastfeeding licensing and administrative regulations targeting child care settings, compares regulations with national standards, and examines the spatial patterning and clustering of these regulations throughout the United States (US). We compared regulations for child care centers (centers) and family child care homes (homes) with national standards for: (1) general breastfeeding support; (2) designated place for breastfeeding; (3) no solids before infants are four months of age; and (4) no formula for breastfed infants without parent permission. We scored state regulations as 0 = standard not addressed, 1 = standard partially addressed, and 2 = standard fully addressed. We considered each regulation individually, and also summed scores to provide an overall rating of regulations by state. We mapped regulations using geographic information systems technology, and explored overall and local spatial autocorrelation using global and local variants of Moran's I. Five states had regulations for centers and two for homes that addressed all four standards. Mean regulation scores were 0.35, 0.20, 0.98, 0.74 for centers, and 0.17, 0.15, 0.79, 0.58 for homes. Local Moran's I revealed that New York and Pennsylvania had substantially stronger regulations than their adjacent states, while Florida had weaker regulations than its neighbors. Overall, few states had regulations that met breastfeeding standards. We identified some patterns of spatial correlation, suggesting avenues for future research to better understand distributions of regulations across the US.

The effect of years of residence in the US on the weight of adolescents is unclear. We examined the association between generation (i.e. 1st, 1.5, 2nd, and 3rd) and weight indicators among Boston adolescents. Data are from a sample of 1,420 9-12th grade public school students in Boston, Massachusetts. We used self-reported information to calculate generation and weight characteristics (i.e., body mass index (BMI), BMI z-score, overweight status), and ran multivariate analyses to estimate the association between generation and weight characteristics, adjusting for race/ethnicity, gender, age and school. In pooled multivariate models, 1.5 generation, second generation, and third generation youth had significantly higher mean BMI scores and mean BMI z-scores than first generation youth. Second (RR 1.87, 95% CI 1.13-3.12) and third generation youth (RR 2.06, 95% CI 1.21-3.50) were also significantly more likely to be overweight than first generation youth. In multivariate models stratified by sex, this pattern persisted for females only. There is a positive, linear trend in BMI by generation that differs by gender. Mechanisms underlying this association should be addressed.

Background: We aimed to develop a clinically practical, multi-domain tool for measuring and monitoring self-reported and caregiver-reported symptoms of older patients. Most existing tools are disease specific, single-domain, or too lengthy for clinical practice. SymTrak was developed to be brief, clinically actionable, sensitive to change, broadly applicable to multiple chronic conditions, culturally sensitive, and easily understood. Methods: SymTrak was developed from multidisciplinary expert panels, existing data, extant instruments, and focus groups. Results: Items tapped psychological, functional, cognitive, pain, sleep, fatigue, and other physical symptoms. Focus groups preferred 3 to 5 item response options, but were neutral regarding frequency versus severity format. Four frequency options (never, sometimes, often, always) were chosen for all items to balance clinical brevity with sensitivity to change. Physician and nurse practitioner focus groups highly valued instrument performance: administrable within 5 minutes; easily retrievable visual graphics from medical records; viewable at item, domain and total-score levels. SymTrak was perceived as more useful for tracking than screening. Clinicians preferred a single brief physical symptom domain instead of multi-item pain, sleep, and fatigue domains. Patient and caregiver focus groups valued item wording: simple language; and applicability regardless of roles. They provided numerous helpful item revisions during "think aloud" interviewing, held subsequent to focus group sessions. They also rated SymTrak highly useful on an 8-item usability scale and were enthusiastic about its use as a communication aid with providers. Version 1.0 (25 items) was finalized and is currently being psychometrically tested. Conclusions: Focus groups of clinicians, patients, and caregivers were helpful in developing Version 1.0 of SymTrak and evaluated it to be useful for tracking symptoms in primary care

OBJECTIVE: Maternal prepregnancy obesity is associated with several poor infant health outcomes; however, studies that investigated motor development have been inconsistent. Thus, maternal prepregnancy weight status and infants' gross motor development were examined. METHODS: Participants consisted of 4,901 mother-infant pairs from the Upstate KIDS study, a longitudinal cohort in New York. Mothers indicated dates when infants achieved each of six gross motor milestones when infants were 4, 8, 12, 18, and 24 months old. Failure time modeling under a Weibull distribution was utilized to compare time to achievement across three levels of maternal prepregnancy BMI. Hazard ratios (HR) below one indicate a lower "risk" of achieving the milestone and translate to later achievement. RESULTS: Compared to infants born to thin and normal-weight mothers (BMI < 25), infants born to mothers with obesity (BMI > 30) were slower to sit without support (HR = 0.91, P = 0.03) and crawl on hands and knees (HR = 0.86, P < 0.001), after adjusting for maternal and birth characteristics. Increased gestational age was associated with faster achievement of all milestones, but additional adjustment did not impact results. CONCLUSIONS: Maternal prepregnancy obesity was associated with a slightly longer time for infant to sit and crawl, potentially due to a compromised intrauterine environment or reduced physically active play.

Limited investigations have been conducted on syndemics and HIV continuum of care outcomes. Using baseline data from a multi-site, randomized controlled study of HIV-positive injection drug users (n = 1,052), we examined whether psychosocial factors co-occurred, and whether these factors were additively associated with behavioral and HIV continuum of care outcomes. Experiencing one type of psychosocial problem was significantly (p < 0.05) associated with an increased odds of experiencing another type of problem. Persons with 3 or more psychosocial problems were significantly more likely to report sexual and injection risk behaviors and were less likely to be adherent to HIV medications. Persons with 4 or more problems were less likely to be virally suppressed. Reporting any problems was associated with not currently taking HIV medications. Our findings highlight the association of syndemics not only with risk behaviors, but also with outcomes related to the continuum of care for HIV-positive persons.