All right. Let's look at a few additional examples of studies, and we'll make note of the study design type, and some of the types of data they're using as measurements for exposures and outcomes. But actually both of these have to do with neighborhood and health indicators. Interesting enough, but they're different study designs. So, the first one we're going to look at from an article published in the New England Journal of Medicine in 2011, the article is titled, Neighborhoods Obesity and Diabetes - A Randomized Social Experiment. So, the first sentence the article puts this out there, "Many observational studies have shown that neighborhood attributes such as poverty and racial segregation are associated with increased risks of obesity and diabetes, even after adjustment for observed individual and family-related factors". So, because the studies are of an observational nature, they had to make sure the groups they were comparing cause poverty, and some measures of racial segregation had to level the playing field in terms of other factors that may also be related to poverty and racial segregation, and also the outcomes of obesity and diabetes. Then, they go on additionally to say, "It is unclear though, whether neighborhood environments directly contribute to the development of obesity and diabetes. People living in neighborhoods with high poverty rates differs in many ways from those living in neighborhoods with low poverty rates. Only some of which can be adequately measured in observational studies. These unmeasured individual characteristics may be responsible for variations in health among different neighborhoods." So, here they're referring to that issue I talked about with observational studies that they may not have accounted to measured or even conceptualized in the observational studies additional confounders, ergo, there's always that nagging question is, did we put control for enough things in the observational study? Authors go on to say, "Inferences concerning the influence of neighborhood maybe more credible if they are based on randomized studies in which otherwise similar people are encouraged to live in different types of neighborhoods." So, they're going to use some data from an experiment that we'll describe in a minute called Moving to Opportunity, which is a large demonstration project intended to uncover the effects of neighborhood characteristics across a range of social and health outcomes and families. So, they use these data to examine the association of randomly assigned variations in neighborhood conditions with the outcomes of obesity and diabetes. So, let's just talk about Moving to Opportunity. This was done in the United States and there was a period where they randomized eligible participants to one of three groups. The randomization period happened back in 1994 to 98. The eligible participants for this study were families with children living in five cities in the United States including Baltimore, but also Boston, Chicago, Los Angeles, and New York City in selected public housing developments in census tracks with high poverty rates. Poverty rates of 40 percent or more in 1990 were eligible. So, the unit of observation here was a family who was living in public housing. Public housing in the US is government subsidized housing for persons who are living below selected income levels. So, the eligible participants, these families to randomized to one of three groups. They were given rent vouchers for private market housing but we required to use these in a census track with a low poverty rate, less than 10 percent in 1990. Or they were randomized to a group that was given rent vouchers for private market housing with no restrictions on where they rented. So, they weren't constrained to neighborhoods with low poverty. The third group was the control group, no assistance. So, they stayed extensively in their public housing units. The outcomes of interests that they wanted to look at were 10 to 16 years after the randomization took place from 2008 through 2010 as part of a long-term follow-up survey. They measured data indicating health outcomes on members of these families including weight, height, and level of glycated hemoglobin. So, these measures were actually quantify these for continuous measures the height, weight, and hemoglobin. Then, they used height and weight to create body mass index, and then declassified persons in the study as obese or not, for example. So, anyway, this type of study was initially a randomized study where participants were randomized to one of three groups, and then they were followed over time for extensively a fair amount of time after the randomization. Then, in a follow-up period, health outcomes were assessed on the members of the three randomization groups. So, this is an example of a longitudinal cohort study that is a randomized controlled trial. It's not of a medical variety here, it's a social experiment where people were randomized to potentially different living environments based on which of the three groups they ended up in. Let's look at another such study of observational nature. This was a study that was published in the December 2016 edition of The American Journal of Public Health. The title of the article, we're taking this from this, Neighborhood Disadvantage, Poor Social Conditions, and Cardiovascular Disease Incidence Among African-American results in the Jackson Heart Study. So, the objectives of this study, of this research as per the authors was to examine the impact of neighborhood conditions resulting from racial residential segregation on a cardiovascular disease risk in a socioeconomically diverse African-American sample. So, their study sample included 4,096 African-American women and men from 21 to 93 years old from the Jackson Heart Study. Larger study conducted in Jackson, Mississippi between 2000 and 2011. This was an observational perspective cohort study, there was no treatment here or randomization. What they did was reassessed neighborhood disadvantage with a composite measure of eight indicators from the 2000 US Census. So, they combine information from eight different indicators from the 2000 US Census into one scale and used it to rank neighborhoods in terms of disadvantage. So, extensively, they converted these eight indicators into some continuous scale that was used to rank relative disadvantage. They assess the neighborhood level social conditions including social cohesion, violence, and disorder with self-reported validated scales, as well. So, people answer questions, they got it, the underlying constructs of social cohesion, violence, and disorder, and they aggregated the questions related to each of these into scales that would measure these constructs. It's important to note here this is a prospective cohort study looking at the outcome of cardiovascular disease. So, the only included participants in these data with geocoded information so they could map them to neighborhoods and census tracks who resided in the Jackson Metropolitan area, and were free of cardiovascular disease at baseline. They want to look at factors. So, they have an exposure here at baseline which happens to be neighborhood level social conditions. So, the exposure is not randomized, "self-selected." It has a function of where people are living. So, unlike the previous experiment, persons in this study were not randomized to different living situations. Then, they followed these people who were free of cardiovascular disease over a fixed amount of time to see who developed cardiovascular disease in the follow-up period. So, we haven't the time to event outcome here as our outcome measure. So, again, the primary outcome was incidence of cardiovascular disease during the study follow-up period. They follow up participants with geocoding information free of cardiovascular disease at baseline from the time of their baseline examination from the period of 2000 to 2004. So, the date of their first cardiovascular event, death from another cause, or loss to follow up, or otherwise, through up to and including December 31, 2011. So, the outcome would be a CVD event in the follow-up period. Some people, it's density would have the outcome, others would be lost to follow-up, would make it to the end of the study window December 31, 2011 without having to follow-up, and may have died from other causes before they develop CVD, et cetera. So, all of those other groups aside from the group that had no CVD event were considered not to have had CVD during their time in the follow-up period. So, hopefully these additionally enlightened you to some of the ideas we'll be talking about throughout the term. Including that of study design and some of the types of measurements. We'll be looking at both for exposures and outcomes when we start getting into looking at outcomes across different groups, exposure groups.
