Top Honors Awarded in First Annual BUSPH Research Day.

To demonstrate the scope of scholarship at the Boston University School of Public Health, the School launched its First Annual Research Day with a poster session designed to showcase the quality of research being conducted by students, post-doctoral fellows and junior faculty.

Associate Dean for Research Roberta “Bobbie” White discusses a participant’s entry.Event coordinators Roberta White, associate dean for research, and Deb Bowen, assistant dean for research, selected 33 abstracts that displayed the “diversity and excellence of research” at the School, with one common theme: a relevance to key issues in public health.

The event, held on Nov. 14, also featured a keynote by Kirsten Davison of the Harvard School of Public Health, who presented “Parent-Centered Community-Based Participatory Research and its Application to Childhood Obesity Prevention.”

Top honors were awarded in three categories, for best posters by faculty members, doctoral candidates, and current master’s students.

Due to the high number and quality of entries from doctoral candidates and post-docs, judges decided to grant four awards in that category. The winning entries are listed below, with an asterisk denoting the lead author of each study.

Faculty:

Exposure to Aircraft Noise and Hospital Admissions for Cardiovascular Diseases among Medicare Participants in the United States

Junenette Peters*; Andrew Correia; Jonathan I. Levy; Steven Melly; Francesca Dominici

Background: Aircraft noise has been associated with physiological and psychological reactions, and also hypertension outcomes, but its direct relation to cardiovascular health is less well established.

Objectives: We leveraged national data on U.S. Medicare population to evaluate associations between average residential aircraft noise exposure and hospital admission for cardiovascular-related diseases among individuals residing near airports in the contiguous U.S.

Methods: The Federal Aviation Administration provided noise contours for 89 airports in 2009 using the Integrated Noise Model. We aligned contours with zip code-resolution Medicare data constructing two exposure metrics: 1) population-weighted noise within each zip code tabulation area (ZCTA), and 2) 90th percentile of noise among populated census blocks within each ZCTA. We used hierarchical models to estimate risks around each airport and pooled across airports, also accounting for clustering. We ran a series of models controlling for individual data (age, gender, race), ZCTA-level socioeconomic status and demographics, temperature, PM2.5, and ozone.

Results: For the 6,000,000 individuals in 2,218 ZCTAs, we found a positive association between aircraft noise and cardiovascular disease hospitalization. The association was most pronounced using the 90th percentile of noise variable; this association persisted accounting for individual and ZCTA-level variables. We observed statistically significant heterogeneity in associations across airports, with higher values generally in the Midwest and lower in the West.

Conclusion: Despite limitations related to lacking key individual-level data (e.g., smoking), using ZCTA-level address information and potential exposure misclassification, results provide some support for the role of airport noise on cardiovascular health.

Doctoral Candidates and Post-Docs:

Predicting Discrepancy in Child Behavior Scores Between Mother-Child and Teacher-Child Ratings

Sarah L. Stone MPH*; Dr. Matthew L. Speltz, Ph.D.; Dr. Brent Collett, Ph.D.

Discrepancies between observers are common in studies of child behavior problems that rely on behavior ratings. Although modest concordance between informants is well documented, little is known about characteristics that predict discrepancies. In 477 children aged 5 to 12 years, indicators of socioeconomic status (SES; maternal age, maternal education, and family income) were evaluated regarding score discrepancies between the Parent-Child Behavior Checklist and Teacher Report Form on the Achenbach System of Empirically Based Assessments (ASEBA) for Total, Externalizing and Internalizing Problems. Multivariable logistic regression, controlling for maternal race, report of psychiatric medication during pregnancy and number of children under age 18 living at home, was used to identify the association between indicators of SES and score discrepancies. Family income <$35,000 was independently associated with discrepancy in scores. Mothers reported more child behavior problems than teachers by at least 10 points for Externalizing and Internalizing Problems [OR= 3.35, 95% CI (1.7 2 to 6.52) and OR=2.53, 95% CI (1.35 to 4.75), respectively]. Maternal age and education were not associated with discrepancies, but mothers with lower family income were almost three times as likely to have identified more problems in their child than the child’s teacher. These results suggest that studies that utilize only maternal or teacher report of child behavior may have misclassification of outcomes that is dependent on SES and could produce biased results.

Prediction of Fetal Hemoglobin in Sickle Cell Anemia Using a Genetic Risk Score

Jacqueline Milton*; Clinton T. Baldwin, PhD; Efthymia Melista; Victor R. Gordeuk, MD; Mehdi Nouraie, PhD; Gregory J. Kato, MD; Caterina Minniti, MD; James G. Taylor, MD; Andrew Campbell, MD; Lori Luchtman-Jones, MD; Sohail R. Rana, MD; Oswaldo L Castro, MD; Yingze Zhang, PhD; Mark T Gladwin, MD and Martin H. Steinberg, MD

Fetal hemoglobin (HbF) is the major genetic modifier of sickle cell anemia (SCA) and is a predictor of mortality. If it were possible to know at birth the HbF level likely to be present after stabilization, an improved prognosis might be given and HbF-inducing treatments better informed. We performed genetic risk prediction using a summary of risk alleles in a genetic risk score (GRS) that was used in a genetic prediction model. The GRS of HbF was developed using the Cooperative Study of Sickle Cell Disease cohort and its predictive value in three independent cohorts: PUSH, Walk-PHaSST and C-Data.

We used the results of a genome-wide association study of HbF in SCA. To build the GRS, we sorted SNPs by increasing p-value and generated a sequence of GRS by cumulatively adding the number of risk alleles a person had for each SNP. A GRS model including 14 SNPs had the best predictive value in all cohorts and explained 23.4% of the variability in HbF. The correlation between the predicted and observed HbF was 0.44, 0.28 and 0.39 in the validation cohorts. The GRS model also outperformed a model using currently used risk factors: gender, alpha thalassemia and HBB haplotypes.

Investigating occupational factors and biomarkers of kidney function among Nicaraguan workers

Rebecca Laws*; Daniel Brooks; Juan José Amador; Daniel Weiner; James Kaufman; Oriana Ramirez Rubio; Jose Marcell Sánchez Rodríguez; Michael McClean

Introduction: In Nicaragua, an excess prevalence of chronic kidney disease (CKD) with unknown etiology has been described primarily among young, male sugarcane workers. Our goals were to characterize the type of kidney damage and evaluate occupational factors.

Methods: Our study population included 284 sugarcane workers, 51 miners, 60 construction workers, and 53 port workers. Blood and urine samples were collected and analyzed for metals and biomarkers of kidney injury. We used linear regression models to investigate predictors of kidney injury.

Results: Estimated glomerular filtration rate (eGFR) was significantly different by sugarcane job and decreased by 6.4 mL/min/1.73 m2 in cane cutters as compared to factory workers (p=0.006). Similarly, NGAL was significantly different by job and increased most among cane cutters, by 19.2 μg/g creatinine, compared to factory workers (p=0.04). More workers than expected in other industries had eGFR <60 mL/min/1.73 m2, indicating CKD. Heavy metals were not associated with markers of kidney function, with the exception of arsenic; workers with the highest arsenic exposures had significantly lower eGFR (p=0.01).

Conclusions: In sugarcane workers, biomarkers of kidney injury and CKD were highest among field workers and lowest among factory workers, supporting the hypothesis that workers with the greatest heat exposure are at greater risk of developing disease. These data provide evidence of CKD among workers in other industries and indicate a tubulointerstitial disease. Finally, there is some evidence that high exposure to arsenic is associated with biomarkers of CKD.

The interplay between CD4 cell count, viral load suppression and duration of ART on mortality in a resource-limited setting

Alana T Brennan*; Dr. Mhairi Maskew and Dr. Ian Sanne

Objective: To examine the interaction between CD4 cell count, viral load suppression and duration of ART on mortality.

Methods: Cohort analysis of HIV-infected patients initiating antiretroviral therapy (ART) between April 2004-June 2011 at a large public-sector clinic in Johannesburg, South Africa. Log-linear models with Poisson distribution were used to estimate risk of death as a function of the interaction between current CD4 count, current viral load (VL) suppression and duration on ART in 12-month intervals. We calculated predicted mortality using estimated coefficients within combinations of predictors.

Results: Among 14,932 ART patients, 1,985 (13.3%) died. Current CD4 was the strongest predictor of death (<50 vs. >550 cells/mm3-RR:46.3; 95%CI: 26.8-80), while unsuppressed current VL vs. suppressed (RR:1.8; 95%CI: 1.5-2.1) and short duration on ART (0-11.9 vs. 66-71.9 months RR:1.7; 95%CI: 1.2-2.3) also predicted death. Our interaction model showed mortality was highest in the first 12-months on treatment across all CD4 and VL strata. As current CD4 and duration on ART increased and VL suppression occurred mortality dropped with CD4 count being the strongest predictor of death. In addition, the relative effect of current CD4 count varied strongly by VL and duration on ART (from 1.3 to 55). Lack of suppression increased the risk of mortality upwards of 6-fold depending on time on ART and current CD4.

Conclusions: Our findings show that while CD4 count is the strongest predictor of death, this effect is modified by VL and duration on ART. Assessment of risk should take into account all three factors.

Master’s Student

Influenza and Pneumococcal Vaccine Uptake of US Cancer Survivors

Geetika Kalloo*; Kathryn Kinzel; Kimberly Shea

Introduction: Cancer survivors are at increased risk for infections such as influenza and pneumococcal disease. Despite this, utilization of influenza and pneumococcal vaccines is lower than targeted levels.

Methods: We abstracted data from the 2009 U.S. Behavioral Risk Factor Surveillance System (BRFSS) survey to compare influenza and pneumococcal vaccine uptake in adult cancer survivors to respondents with no history of cancer. Logistic regression was used to estimate adjusted prevalence ratios and 95% confidence intervals (CI) between prior cancer diagnosis and receipt of influenza vaccine, pneumococcal vaccine, or both.

Results: Among 411,654 BRFSS respondents, 2.5% reported a prior cancer diagnosis, 37.4% reported having received an influenza vaccine in the past 12 months, and 22.9% reported having received a pneumococcal vaccine. Compared with respondents with no history of cancer, cancer survivors were 1.2 (95% CI 1.1-1.4) times as likely to have received an influenza vaccine, 1.5 times (95% CI 1.3-1.7), as likely to have received a pneumococcal vaccine, and 1.5 (95% CI 1.3-1.7) times as likely to have received both vaccines.

Conclusion: Although adult influenza and pneumococcal vaccination remain below targeted levels, cancer survivors report having received these vaccines more often than adults without cancer.