Both social isolation and cognitive decline are urgent threats to public health, as they
predispose persons to Alzheimer’s disease and related dementias. Air pollution and the built
environment are community-based factors that have been shown to adversely affect cognitive
function. Individual factors such as social isolation and economic disadvantage further
contribute to cognitive risk. Rural settings have more limited opportunities for social
engagement, when compared to urban settings. These conditions converge to form a perfect
storm of social isolation and accelerated cognitive decline, yet prior studies have not focused on
rural, disadvantaged, ethnic minority residents. Using a multi-method approach, we propose to
demonstrate how a uniquely vulnerable rural Florida population (N = 1087) are at risk for social
isolation and decline in cognitive function due to the unique characteristics of the physical and
social environment. We will recruit community-dwelling, non-clinical, dementia-free, middle aged
and older adults from 5 communities in the Lake Okeechobee area of Florida for a 5-year study
incorporating time-series individual social and cognitive measures and community-level
measures of the physical and built environment. Apple watches will be used by a subsample of
120 participants representative of 5 communities to continuously monitor sensory data, daily
routine and predefined activities for 2 months. Ecological momentary assessment (EMA) will be
used to collect data daily over periods of agricultural burning and no burning.
Using a combination of primary data collection, secondary data analysis, subsample continuous
monitoring and EMA, we will examine the following aims over 36 months among rural,
racially/ethnically, and linguistically diverse underserved community-dwelling adults aged 45
years and older:
Aim 1: Examine the contribution of smoke-related PM2.5 exposures to SI and cognitive
function, through multilevel growth modeling.
Aim 2. Determine the effects of the built (e.g., retail destinations, park space) and social
environment (e.g., crime SES) on social isolation and cognitive function through mixed linear
modeling.
Aim 3. Contextualize social isolation and cognitive function among residents from different
racial/ethnic groups using EMA and sensor-derived behavior models with a subsample of 120
stratified by Lake O communities during burn and non-burn seasons.
