Poster #69 - Where You Live Matters: Visualizing Environmental Effects on Reading Attainment

• In Event: Poster Session 08

Abstract

Background. The effect of socioeconomic status (SES) on the etiology of developmental traits has been explored in many phenotypes, including reading achievement. Past work has found that genetic influences are suppressed under conditions of socioeconomic deprivation and more fully realized under conditions of socioeconomic advantage: a gene-SES interaction. In addition to SES influencing the etiology of a trait, past work has also pointed towards the presence of gene-location interactions, with etiologies varying across geographic regions of the same country/state. To what degree the magnitude of the gene-SES interaction is uniform across geographical areas is unknown and was tested in this study, using an example of one U.S. state, Florida. Specifically, we examined how the level of gene-SES interactions on reading achievement varied by where a child lives in Florida. We hypothesized that geographical locations with larger socioeconomic inequity and increased variability in the socioeconomic and cultural factors would draw the gene-SES interplay out more than areas with fewer socioeconomic variabilities, expecting more pronounced gene-SES interactions in the former areas as compared to the latter. Method. The sample included 2,135 twin pairs from the Florida Twin Project on Reading. Twins ranged in age from 10.71 to 17.77 years old (mean=13.82), and had data available on reading achievement, SES , and home addresses. Reading achievement was operationalized as reading comprehension scores from the 2013-2014 Florida-wide-school-administered reading test. SES was assessed using household income (U.S. Census). Home addresses were converted to distances from locations in which the etiology of reading and gene-SES interactions were to be assessed. First, we used a GxE interaction analysis to examine a gene-SES interplay in reading. Next, we used an adapted version of spatial twin analysis to visualize geographical variation of gene-SES interaction. Results. As hypothesized, a gene-SES interaction was found, with heritability of reading being suppressed in lower (23%) versus higher SES homes (78%). On average, the magnitude of the moderating parameter for the genetic component was βx = 0.0568 (Figure 1), but this magnitude was not consistent across the state. Rather, the size of this moderation effect varied by location, suggesting a gene-location-SES interaction. As shown in Figure 2, the magnitude of the βx parameter was lower (gold) in large urban areas with vaster socioeconomic disparities, highlighting suppressed genetic influences on reading in those areas. For smaller areas and those with less socioeconomic variability, magnitude of the βx parameter was high (pink) giving rise to more fully realized genetic influences on reading there. Conclusions. Our results imply SES significantly shapes reading variability. However, they also suggest a child’s home location matters in both the overall etiology and how strongly SES moderates said etiologies. These results point towards the presence of multiple significant environmental factors that simultaneously, and inseparably, influence the development and underlying etiology of reading achievement. Further, this study provides the first empirical analysis applying our extended twin spatial analysis model, highlighting that said model can be used to better understand how environmental factors differentially impact developmental traits on the basis of geographic location.

Authors

• Jeffrey Shero, Florida State University
  Presenting Author

• Florina Erbeli, Texas A&M University - College Station

• Zoe E Reed, University of Bristol

• Rasheda Haughbrook, Florida State University

• Oliver S.P. Davis, University of Bristol

• Sara Hart, Florida State University

• Jeanette Taylor, Florida State University