Geography & Environmental Hazards Flashcards
McLaughlin, Stokes, Smith, & Nonoyama. (2007). Differential mortality across the United States: The influence of place-based inequality.
Gist: Mortality risk varies by geographic areas. Historically, cities were less healthy places to live, but rural-urban differentials have reduced over time. The current study seeks to examine the specific mechanisms by which neighborhood income inequality influences mortality.
1) Hypothesis: Income inequality will influence mortality mediated by other measures of local socioeconomic inequality, social conditions and safety, health service availability, and environmental risks.
2) Two major theories regarding mechanisms: (a) “reduced investment” perspective - higher levels of income inequality result in less investment in health, educational, & social services in a locality; (b) “psychosocial” perspective - those who have lower incomes in areas with high inequality are especially aware of their disadvantage & experience increased feelings of deprivation & isolation.
3) Income inequality: aggregate measure of how equally resources are distributed within a particular place.
4) Income inequality explains mortality risk, mediated through measures of socioeconomic status & rurality.
5) Other factors (social conditions & safety, health care availability, & environmental risks) don’t impact inequality-mortality relationship but do directly influence mortality outcome.
6) Controlling for spatial proximity may reduce influence of these factors on model.
Geronimus, Bound, & Ro. (2014). Residential mobility across local areas in the United States and the geographic distribution of the healthy population.
Gist: The overall question was whether population dynamics provided competing explanations to place effects for observed geographic patterns of population health. Focusing on the working age population (period of adulthood where health disparities are greatest), looked at whether residential mobility impacted health outcomes. Have three research questions: (1) extent of residential mobility flows into/out of a diverse set of local areas?, (2) extent residential mobility btw local areas associated with health-induced functional limitations or sociodemographic characteristics associated w/health more broadly, and (3) if residential mobility & health-induced functional limitation associated, does selection affect cross-sectional estimates of local area prevalence of health-induced functional limitation?
1) Health-selective migration hypothesis: Differences in health profiles across distinct areas may arise, in part, because the residentially mobile may be more healthy or less healthy than those who do not move.
2) Found little evidence that observed migration affected overall distribution of healthy or unhealthy populations across focal study areas. Those who moved out of focal areas tended to be younger, more highly educated, & in better health than those who stayed. However, those moving in also tended to be younger, more highly educated, & in better health. These two mobility streams roughly cancelled each other out, leaving cross-sectional estimates of local population health stable.
Nandi & Kawachi. (2011). Neighborhood effects on mortality.
Gist: This chapter has 3 aims: (1) Discuss mechanisms & processes by which exposure to neighborhood contexts influences adult mortality; (2) review of empirical, multilevel studies of contextual influences on mortality outcomes; & (3) outline challenges in field of research.
1) Neighborhood conceptualized along 3 dimensions: (a) physical environment - features local traffic patterns & concepts like “built environment” & “walkability”; (b) service environment - features of neighborhoods such as accessibility of health services, quality of schools; & (c) social environment - resources (e.g., social capital) & risks (e.g., exposure to violence).
2) Compositional (e.g., characteristics of the people) vs. contextual (e.g., characteristics of the environments) effects.
3) Issues of lag time btw exposure to environment & elevated risk of death; mortality mixes effects of disease incidence & prognosis; lack of specificity of outcome measure such as all-cause mortality.
4) Review of lit: (a) most consistently reported association btw neighborhood SES & mortality; (b) other associations include socioeconomic deprivation, social deprivation, social environment (e.g., collective efficacy, social support, perceived violence), racial composition (or racial segregation), physical hazards (few studies assessed this w/mortality).
5) Commonly accepted properties of neighborhoods: (a) spatially defined geographic areas of limited size; (b) have a name & recognized identity & carry symbolic significance for their residents; (c) are including physical, social, & service environments like mosaics.
6) Defining neighborhoods with 2 methods. Primary method is through predefined boundaries, but a few studies use alternative methods (e.g., residents’ perceptions of margins of neighborhoods).
7) Issues: (a) administrative boundaries are imperfect proxies for neighborhoods; (b) indices based on administrative data only indirect measures for underlying neighborhood attribute theorized as important to health; (c) when compared to bio/beh disease determinants, neighborhood environments relatively static.
Yang, Jensen, & Haran. (2011). Social capital and human mortality: Explaining the rural paradox with county-level mortality data.
Gist: The “rural paradox” refers to lower standardized mortality rates in rural areas considering the economic and structural disadvantages. This study introduces county-level measures of social capital and introduces models of spatial dependence to explain the paradox. The rural paradox was confirmed with the use of metro-nonmetro and RUCs; social capital does reduce the impact of residence on mortality after controlling for race, ethnicity, and socioeconomic covariates; the attenuation of social capital is greater when using spatial models (lag & error). Overall, social capital was negatively associated with mortality at the county-level, and rural areas had higher levels of social capital. This measure was an index originally developed by Putnam. Additionally, there is spatial dependence in mortality and in social capital that explains ~10% of additional variance.
Meijer, Rohl, Bloomfield, & Grittner. (2012). Do neighborhoods affect individual mortality? A systematic review and meta-analysis of multilevel studies.
Gist: Meta-analysis for whether residents in socially or physically deprived neighborhoods have higher mortality after adjusting for individual SES. Another review summarized processes through which neighborhoods contribute to health inequalities: (1) residential segregation by race, ethnicity, & SEP influence inequality of resource distribution in areas; (2) mutually reinforcing & reciprocal relationships suggesting that neighborhoods affect people & people affect neighborhoods. Overall, although found no clear associations btw mortality & income inequality or social capital at neighborhood level, there was evidence of associations for area-level SES, social coherence, & population density/urbanization & mortality. People living in areas with low SES have higher mortality than poeple living in higher SES areas, even after accounting for individual SES. Younger age groups, areal units with fewer inhabitants, & larger geographical areas are more susceptible to this affect of ALSES on mortality.
Cossman et al. (2007). Persistent clusters of mortality in the United States.
Gist: Explored 35 consecutive years of US mortality data at county level. Patterns of clusters changed little across time. Counties w/high mortality rates & counties with low mortality rates both experienced younger population out-migration, had economic decline, & were predominantly rural. Counties with low mortality in Upper Great Plans (former) was unexpected & in contrast, portions of South (including Southeastern US, Appalachia, & Mississippi Delta) had expected high mortality but experienced similar population loss & economic contraction.
Cossman, James, Cosby, & Cossman. (2010). Underlying causes of emerging nonmetropolitan mortality penalty.
“Gist: The nonmetropolitan mortality penalty is driven by declines in heart disease & cancer death rates in metropolitan areas.
1) Research indicates that historical metropolitan mortality penalty (i.e., higher rates of death in cities than in rural areas) has been reversed since mid-1980s.
2) Age-standardized mortality rates for all-cause & top 3 causes of death (heart disease, cancer, & stroke) provided. (Standardization by sex & race resulted in same pattern, so not used).
3) A non-metro penalty has been associated with strokes at least since mid-1960s.
4) All-cause nonmetro mortality penalty 1st appeared mid-80, coinciding w/heart disease nonmetro mortality penalty.
5) Cancer nonmetropolitan mortality penalty appeared in mid-1990s.
6) Underlying reasons for changes in trends associated with metro-nonmetro ratio are currently unknown.”
Sparks & Sparks. (2010). An application of spatially autoregressive models to the study of US county mortality rates.
Gist: US county mortality rates tend to be associated w/social & economic resources of counties & the unequal distribution of resources across space. The current article applies spatially autoregressive models of these rates that control for social & economic conditions. What has been missing from the literature is a theoretical explanation for the spatial patterning of mortality. The authors consider 2 spatial models: (1) spatial lag model & (2) spatial error model. They suggest the former relates to a spatially diffusive process, such as with infections or social networks. The latter specifies that there is a spatial pattern of the unmeasured independent variables in the model (i.e., not a diffusive process). It is the latter model that should better fit differences in county mortality rates. They also found spatial autocorrelation for all of the social & economic controls included in the models. The final warning is that spatial models cannot be selected only by diagnostic fit, but must also fit theoretically.
Sparks, McLaughlin, & Stokes. (2009). Differential neonatal and postneonatal infant mortality rates across US counties: The role of socioeconomic conditions and rurality.
Gist: Examined differences in correlates of neonatal & postneonatal infant mortality rates across counties & by degree or rurality. Generally, higher infant mortality rates have been found in rural than urban areas in the U.S. as well as in counties with higher levels of socioeconomic disadvantage. Overall, four findings were found: (1) relationships between infant mortality & rurality complex (although size of urban population in nonmetro counties not adjacent to metro counties is important for mortality); (2) After controlling for SES, health care availability, & environmental conditions, neonatal mortality rates were lower in 4 of 7 nonmetro categories than in largest metro counties (postneonatal mortality rates were still higher in the 2 most rural nonmetro counties than in large metro counties); (3) SES & other covariates and rurality differn in their association with neonatal & postneonatal mortality: (a) rurality is associated w/lower neonatal mortality while (b) rurality (when sig.) is associated w/higher postneonatal mortality than in largest metro counties; (4) variations in factors associated w/neonatal & postneonatal mortality are masked when rates aren’t examined separately. Overall, different patterns of association between rurality & mortality by the type of rate (neonatal vs. postneonatal). Regarding adjacency, postneonatal mortality rates were higher in nonmetro counties not adjacent to metro counties (largest disadvantage for most rural or isolated counties). Nonmetro counties adjacent to metro counties had similar postneonatal mortality rates as the most metro counties; one exception was adjacent counties with the largest urban pop. - had lower postneonatal mortality rates than largest metro counties.
Sparks, Sparks, & Campbell. (2013). An application of Bayesian spatial statistical methods to the study of racial and poverty segregation and infant mortality rates in the US.
Gist: The US has an infant mortality rate higher than what would be expected given its affluence. The current study uses spatial measures of racial segregation & found that when blacks live in close proximity to each other, this tends to increase the IMR. Results for poverty segregation suggest the same pattern on the county IMR. Interaction between blacks & whites and poor & non-poor residents is protective for infant mortality. Most studies apply regression models without spatial reference to neighboring values or attributes. Also, research on residential segregation & infant health outcomes must address the multidimensional nature of segregation & its inherently spatial nature. Bayesian regression models were used, which avoided extraneous variation due to differences in size of population at risk in IMRs.
Wight, Cummings, Karlamangla, & Aneshensel. (2010). Urban neighborhood context and mortality in late life.
Gist: The purpose was to examine contextual effects of urban neighborhood characteristics on mortality among older adults. Based on a number of theoretical perspectives, there was conceptual justification to investigate whether neighborhood socioeconomic conditions (including affluence), residential stability, age structure, & ethnic composition of the neighborhood were associated with mortality in late life. Overall, they found that neighborhood socioeconomic disadvantage was not associated with late-life mortality once individual-level variables were controlled. However, neighborhoods w/higher proportion Hispanic (after controls) related to higher late-life mortality (contradicts “Hispanic paradox” or possible ethnic enclave mortality benefit. This effect, however, is partly mediated by individual-level health variables. Risk of mortality was lower for affluent neighborhoods after controlling for individual variables. Overall, the presence of economic resources incumbent w/affluence may matter more for late-life mortality than the absence of resources associated with socioeconomic disadvantage.
Browning, Bjornstrom, & Cagney. (2011). Health and mortality consequences of the physical environment.
“Gist: This chapter reviews the current literature on relationship between the physical environment & adult mortality (and morbidity). The physical environment has 3 forms: (1) socially unconditioned - consequences of direct exposure to environmental harm are indiscriminate with respect to SEP, (2) socially conditioned - mortality from ““natural”” environmental harms exhibit differential impact due to SEP (e.g., Katrina), (3) socially produced - consequences of environment as environmental harms that follow from human actions. The chapter has 4 parts: (1) mediating & moderating factors, (2) review of socially produced health consequences, (3) review of socially conditioned health consequences, & (4) directions for future research.
(1) Pathways: health-related behavior, stress, social relationships, social capital.
(2) Socially produced: (a) built environment - alters likelihood of engaging in individual behaviors relevant to health, exposure to stress, individual social relationships, community social capital, exposure to toxins, & exposure to accidents; (b) design - land use, street layout, metropolitan sprawl (““dispersed & inefficient urban growth””); (c) consequences of urban sprawl - time spent in cars, traffic-related safety, exposure to toxins; (d) commercial establishments - access to quality food, prevalence of establishments selling alcohol; (e) public- & private-recreation facilities, public transportation; (f) quality - substandard housing, air pollution.
(3) Socially conditioned: (a) natural hazards include drought, biological hazards, floods, hurricanes & other coastal storms, earthquakes, wildfires, extreme temperature, volcanic eruption, & landslides; (b) vulnerability to natural disasters - ““characteristics of person or group & their situation that influence their capacity to anticipate, cope with, resist, & recover from impact of natural hazard””.
(4) Future research: (a) research on mortality remains limited; (b) delineation of mechanisms; (c) causality & selection concerns.”
Bond Huie, Hummer, & Rogers. (2002). Individual and contextual risks of death among race and ethnic groups in the United States.
“Gist: No existing research on adult mortality differentials distinguished by multiple Hispanic subgroups and explores role of nativity at both individual & contextual levels for small geographic areas.
1) Most of the research focus has been on individual & contextual effects on black-white mortality differentials.
2) Examine effects of individual & contextual factors on adult mortality differentials for blacks, whites, & Hispanic subgroups (Mexican Americans, Puerto Ricans, & other Hispanics).
3) Individual-level models have established link between race-ethnic mortality differentials & individual-level factors of education, income, marital status, & nativity.
4) Contextual-level models have found links between income inequality, population composition, & all-cause mortality.
5) Multilevel models (few) have found net of individual-level risk factors, community context affects individual’s risk of premature adult mortality.
6) Unique use of survey methodology data to create small-area estimates.
7) Overall findings: (a) Neighborhood-level variables reduces risk of mortality (controlling for individual-level) for blacks, Mexican Americans, Puerto Ricans, & other Hispanics compared to whites; (b) High neighborhood concentration of Hispanics is not protective of mortality (instead, related more to nativity); (c) Puerto Ricans, Mexican Americans, & other Hispanics age 45-64 lower mortality than age 18-44 due to contextual effects influenced by nativity; (d) Risks associated with some individual covariates of mortality also reduced by adding contextual-level variables.”
