Another important challenge in the study of neighborhood heath effects is the measurement of neighborhood exposures. Two important questions illustrate the scope of this challenge: (1) What are the relevant features of neighborhoods important for health? (2) How does one obtain information on these features? With regard to the first question, most studies examining neighborhoods in relation to health investigate the socioeconomic characteristics of neighborhoods. These census-derived indicators are readily available and easily linked to health outcomes data and have been studied in relation to CVD risk factors and outcomes. Specifically studies have consistently documented associations between neighborhood disadvantage or deprivation and increased CVD risk, morbidity, and mortality (Diez Roux et al, 2001; Sundquist et al, 2004). However, documenting associations between neighborhood socioeconomic indicators and health provides few clues as to the underlying important health-relevant features. It also does not shed light on the pathways by which neighborhoods impact a particular health outcome (Diez Roux, 2001). Establishing evidence of these pathways is important to drawing causal inferences and to identifying important interventions.
The specific features of neighborhoods that are most health relevant are likely to differ for each health outcome. For this reason, beginning with a clear conceptual framework is an important pre-requisite for delineating these features and the pathways by which they impact health. Figure 24.1 shows a conceptual framework of neighborhood environments in relation to CVD. The framework highlights features of the physical and social features of neighborhood environments and the hypothesized pathways by which they may impact CVD risk. For example, neighborhood availability and relative cost of health foods may impact diet quality which in turn may impact more proximate biological risk factors such as BMI and hypertension. Alternatively, neighborhood crime may have a direct impact on
CVD through stress processes. Such frameworks have been an important omission in the literature. An increased usage of conceptual frameworks may serve to address important critiques regarding the atheoretical and data-driven nature of some research in this area (Macintyre et al, 2002; O'Campo, 2003).
Consistent with the aforementioned conceptual framework, researchers have begun to measure potentially health-relevant features of neighborhoods. These include measures of physical environment such as access to health-enriching resources including the density and accessibility of opportunities for engaging in physical activity and purchasing affordable healthy foods (Mujahid et al, 2007). Measures of the built environment have also been measured such as land use patterns (mix of commercial and residential spaces), the presence and amount of green space, street connectivity, and housing density (Handy et al, 2002). Finally measures of the presence of ill-health promoting facilities (i.e., the presence and density of liquor stores and fast food restaurants) and features of the social environment such as collective efficacy, social cohesion, and informal social control have also been considered (Mujahid et al, 2007). These novel measures have been increasingly studied in relation to health in general and cardiovascular health in particular (Auchincloss et al, 2009; Augustin et al, 2008; Moore et al, 2009; Mujahid et al, 2008b).
A major challenge in measuring specific health-relevant features of neighborhood environments involves the second issue, obtaining data on these features. Administrative data sources can sometimes be used to obtain information on specific health-relevant features. For example, proprietary data sources can provide information on the presence of businesses and facilities in a given area. The linkages of these geo-referenced data sources to health study data are made possible through Geographical Information System (GIS) technology (Rushton, 2003). In addition to linking data sources, GIS can be used to create sophisticated measures of spatial accessibility including density measures such as kernel densities which allow for smoothing densities over space to create neighborhood specific measures for various spatial scales and various neighborhood definitions.
Neighborhood environments can also be assessed using data from individuals. There are three general approaches to this type of investigation. The first two approaches involve asking individuals to report on neighborhood characteristics via a self-administered survey. In the first approach, each health study participant provides an assessment of his or her neighborhood. These measures of neighborhoods can then be examined in relation to a particular health outcome. Information on neighborhood environments can also be obtained from an informant sample of individuals who reside in the same areas as the health study participants but who themselves are not participants (Mujahid et al, 2007). Through administration of a neighborhood survey, information can be obtained on various dimensions of neighborhood environments and can then be linked to study participants. While the first approach has the benefit of feasibility, it is limited by the potential for same-source bias, the process by which obtaining information on self-reported neighborhood features and self-reported outcomes may result in spurious associations if both types of reports are affected by underlying propensities of the individual. Moreover, the outcome may affect an individual's assessment of the neighborhood exposure. For example, individuals who are physically inactive may report few opportunities for physical activity provided by the neighborhood, irrespective of the actual availability. The use of an information sample can reduce this source of bias. An additional benefit of the second approach is that there can be a more detailed assessment of the neighborhood environment. The often broad scope of health studies usually limits the ability of an in-depth investigation of any given area. Moreover, the use of informants allows denser sampling and multiple responses can be aggregated across areas to obtain more reliable and valid estimates of neighborhood properties. (Mujahid et al, 2007).
A final approach to obtaining information on neighborhood environments is the use of systematic social observation by which a team of trained investigators assess the neighborhood environments based on a set of systematic criteria (Sampson and Raudenbush, 1999). While this approach may allow for more objective and systematic characterizations of neighborhood environments, it can be very time and labor intensive limiting its feasibility and use in the public health literature. Additionally, this approach may not provide an accurate assessment of the social environments of neighborhoods because it is very difficult as an outsider to observe features such as neighborhood social cohesion or crime.
Recent studies have begun to compare associations between specific features of neighborhood environments in relation to health using multiple methods of collecting neighborhood measures (Moore et al, 2008). For example, Moore and colleagues examined measures of neighborhood food environments in relation to diet quality. They found that the food environment as measured by a proprietary data source, perceptions of health study participants, and perceptions of an informant sample produced comparable results in terms of the direction and magnitude of associations with diet quality.
As more researchers are developing neighborhood measures based on individual assessments (i.e., informant sample or systematic social observation), advanced methods have been developed to examine the measurement properties of these approaches. Because these measures are obtained via questionnaire, neighborhood measures are constructed by aggregating survey items within a particular construct to create a summary measure. Traditional psychometric assessments can only characterize the validity and reliability of measures based on two levels of data (i.e., scale items nested within individuals). For example, the internal consistency has been used to assess the extent to which responses to scale items are consistent within individuals. However, ecometrics, developed as an extension of psychometrics, considers the additional nested structure of the data (i.e., scale items nested within individuals, nested within neighborhoods) (Raudenbush and Sampson, 1999). Thus, these methods can allow investigators to additionally assess the extent to which individuals within the same neighborhood agree on their assessment of neighborhood conditions (Raudenbush and Sampson, 1999). These approaches are increasingly being used in the literature (Mujahid et al, 2007).
Was this article helpful?
Your heart pumps blood throughout your body using a network of tubing called arteries and capillaries which return the blood back to your heart via your veins. Blood pressure is the force of the blood pushing against the walls of your arteries as your heart beats.Learn more...