Shahadut Hossain

Bayesian adjustment for covariate measurement errors: A flexible parametric approach

In most epidemiological investigations, the study units are people, the outcome variable (or the response) is a health-related event, and the explanatory variables are usually environmental and/or socio-demographic factors. The fundamental task in such investigations is to quantify the association between the explanatory variables (covariates/exposures) and the outcome variable through a suitable regression model. The accuracy of such quantification depends on how precisely the relevant covariates are measured. In many instances, we cannot measure some of the covariates accurately. Rather, we can measure noisy (mismeasured) versions of them. In statistical terminology, mismeasurement in continuous covariates is known as measurement errors or errors-in-variables. Regression analyses based on mismeasured covariates lead to biased inference about the true underlying response-covariate associations. In this paper, we suggest a flexible parametric approach for avoiding this bias when estimating the response-covariate relationship through a logistic regression model. More specifically, we consider the flexible generalized skew-normal and the flexible generalized skew-t distributions for modeling the unobserved true exposure. For inference and computational purposes, we use Bayesian Markov chain Monte Carlo techniques. We investigate the performance of the proposed flexible parametric approach in comparison with a common flexible parametric approach through extensive simulation studies. We also compare the proposed method with the competing flexible parametric method on a real-life data set. Though emphasis is put on the logistic regression model, the proposed method is unified and is applicable to the other generalized linear models, and to other types of non-linear regression models as well.

Innovative Bayesian Methods for Biostatistics and Epidemiology

Complex data and models now pervade biostatistics and epidemiology. Increasingly, Bayesian methods are seen as desirable tools to tame this complexity and make principled inferences from the data at hand. In this chapter we try to convey the flavor of what Bayesian methods have to offer, by describing a number of applications of Bayesian methods to health research. We emphasize the strengths of these approaches, and points of departure from non-Bayesian techniques.

Ethnic and gender differences in perceptions of mortality risk in a Canadian urban centre

Background: Women reportedly do not perceive heart disease (HD) as a major threat to their health; however, men’s perceptions are rarely studied. Purpose: We explored gender and ethnic differences in risk perception of HD mortality. Methods: The survey was completed by 976 people 40+ years of age, in metropolitan Vancouver, Canada. Results: Men, compared with women, were more likely not to know the answer to a question about whether HD is the most common cause of death for women; however, women were more likely not to know the answer to a question about whether HD is the most common cause of death for men. Chinese-Canadian and South Asian-Canadian participants were more likely than participants of other ethnic groups not to know the answer to either question, and the Chinese-Canadian participants were more likely to disagree that HD is the most common cause of death for women. Conclusion: There is a need to educate the Chinese-Canadian and South Asian-Canadian communities about HD as a first step in promoting health behavior change. Men and women must be educated about the other gender’s risk of HD because all adults play integral roles in making decisions about the prevention of and early intervention for HD.

Depression and Anxiety as Possible Mediators of the Association between Smoking and Attention Deficit Hyperactivity Disorder.

The objective of this study was to investigate the association between depression and anxiety and adolescents’ smoking status, and to determine whether depression or anxiety mediate the association between Attention Deficit Hyperactivity Disorder (ADHD) and smoking. A cross-sectional survey of tobacco use was conducted in regional school districts located outside the Greater Vancouver area of British Columbia, Canada. The sample included 6,943 students. Having taken medications for depression, anxiety, or ADHD; ethnicity/race; and parental, peer, and sibling smoking status were significantly associated with smoking status. ADHD was significantly associated with depression and anxiety. In multivariate analysis, although depression/anxiety and ADHD were found to be significant correlates of regular smoking when each variable was independently included in a multinomial logistic regression model, ADHD did not remain significant when it was included with depression/anxiety. Depression and anxiety may mediate the relationship between ADHD and smoking.