Curt Johanson
EPI 222 HW 8
05 March 2019
1. What are 3 different ways to account for SES in a analytic models when investigating racial/ethnic health disparities? Briefly discuss the interpretations/implications of each approach as it relates to understanding health disparities by race/ethnicity.
I. DAG Models: I did not understand this completely before class today but one of the most important steps is to start out by making a DAG (diagrammatic Acyclic Graph) of all of the potential mediators and confounds of SES before you begin research and early in a manuscript. The diagram is essential for looking at spatial relationships and will hint at possible contextual vs. ecologic variable diffences and multilevel cross interactions as well that should be considered.Drawing one for the previous homework below:
Confounding: contextual exposure: Mediating: Outcome
Mom’s education 24 hr store density current occupation lower CHF OS Own Education current income for African & Latin Avail. & type of transpo
II. Confounding Factors: (Mom’s and Own education) occur before an exposure event (stores closing with 24 hr access) are also predictors of poor congestive heart failure survival and potentially the exposure event itself. This will help to analyze the SES confounding on the outcome effect size and could also show decreased likelihood that racial disparities have less chance of being explained by genetic or biological factors alone.
III. Mediating Factors: Occur after or close to the same time as the exposure and can give us the total effect of mediation of the exposure on the outcome. An outcome can be partially or fully mediated depending on if the effect size of the mediator is large enough to cancel out the exposure effect size on the outcome.
IV. Modification Effects: Multi level interaction effects between races and contextual variables such as lower CHF OS for latins and blacks in low density areas
2.
2. Describe a potential effect modifier, mediator, or contextual variable (for definition of contextual variable, see Diez-Roux reading) for an association of interest to you and relevant to health disparities. For example, for investigating the association between education and hypertension, I might be interested in evaluating whether the association between years of education and hypertension is different for Black men than for White men. Describe how you would study whether this relationship exists.
I am interested in seeing if exposure of adverse family relationship after coming out (a measure of LGBT rejection) leads to the adult outcome of greater promiscuity (by partner # and std incidence) differentially across White and Asian gay men and if family income level can intensify. My hypothesis is that Caucasian with adverse family experiences during coming out will have greater mediation effects of family income on promiscuity than Asian gay males.
3. Respond to one other person's post on the forum with a comment or suggestion.
HW8 Q3.
HI Mitzi,
I help with administration and data collection on a couple clinical trials on uterine fibroid treatments and I am intrigued by your question and wonder about differences in fibroids between women and men on various hormone therapy and post uterine transplant (See attached Lancet Article on Uterine Transplant success from live and post mortem organ donors). I had only heard of uterine fibroids before as a uncommon condition until working on studies here and have seen in the literature and in our studies higher prevalence rates, size and severity in Latin and African American patients. The fibroids can potentially lead to uterine cancers and they are extremely symptomatic and painful. For your study since cancer might be very difficult to have as an observational endpoint you could perhaps look at Fibroids as they develop very quick! I'm not an MD, but I seem to remember that estrogen seems to increase their growth and transxamic acid in some cases. Maybe a multi level analysis on trends across men and women with estrogen vs. t therapy ?