Abstract TMP47: Food Insecurity And Long Term Mortality Rates After Stroke In The US

Abstract

Background: Around 50 million people in the US have inadequate access to nutritious food. Food insecurity is linked with chronic conditions and worse overall health. Research has shown that living in a lower SES neighborhood is linked to increased post-stroke mortality, but little is known about the effect of food insecurity on mortality. Methods: This study used the National Health and Nutrition Examination Survey (NHANES) dataset 1999/2000-2013/2014 with linked mortality files through 2015. Participants included those ≥ 20 years with a complete survey, physical examination, and mortality information. Stroke was self-reported. Food security was characterized as food secure (full, marginal) or food insecure (low, very low). We evaluated the relationship between food insecurity and all-cause, cardiovascular (CV) and stroke mortality in the general and stroke population using Cox proportional hazards regression models, adjusting for sociodemographic factors and comorbidities. Results: From 1999 to 2014, 40,777 people participated in NHANES and 1,530 reported a history of stroke. Prevalence of food insecurity was 11.6% in the general population and 15.5% in the stroke subset. Among the general population, there were 5394 deaths due to all-causes, 1152 CV deaths, and 216 stroke deaths. Among the stroke population, there were 650 deaths due to all-causes, 176 CV deaths, and 41 stroke deaths. In the general population, food insecure individuals had higher adjusted all-cause mortality (aHR 1.15, CI 1.01-1.32) and a trend towards higher CV (aHR 1.14, 0.85-1.52) and stroke mortality (aHR 1.23, 0.65-2.34). In participants with prior stroke, food insecurity had no impact on all-cause, CV or stroke mortality. Discussion: Although prior studies have demonstrated that living in a low SES neighborhood results in worse post-stroke mortality, this data suggests the effect is not directly linked with food insecurity. A potential explanation is that individuals may receive greater access to food resources after a stroke. There is also the possibility that the low frequency of deaths among stroke survivors limited the ability to detect an effect. Further study is needed to explore the underpinnings of worse post-stroke outcomes in lower SES neighborhoods.