(571) RNA profiles reveal patterns of circadian rhythm in pregnancy

This was a prospective, observational study of a cell-free RNA platform utilizing direct-to-participant recruitment efforts from July 2020 to December 2021. The IRB-approved study was open to patients ages 18-45 with a singleton pregnancy in the United States.

Samples were grouped by time of blood draw into morning (6am-10am, n=121), midday (10am-2pm, n=303) and afternoon (2pm-6pm, n=113). We ran differential gene expression (DGE) analyses to test the impact of time of day. A quasi-likelihood negative binomial generalized log-linear model was fit to count data using edgeR package (v. 3.38.1) and DGE were discovered with an empirical Bayes quasi-likelihood F-test (edgeR) in all 3 possible pairwise comparisons.

Results: When looking for DGE between the morning and afternoon groups we identify 5,729 genes (Figure 1) or 43% of all analyzed genes as significantly differentially expressed. For morning vs midday the number is 4,278 (32%) and finally midday vs afternoon has 15 (0.1%) DGE. The three sets have very high overlap, ranging from 73% - 87% (p < 10-30), and among the genes with highest separation we find those reported with circadian rhythm (PER1, PER3, DDIT4, FKBP5).

Conclusion: The cell-free RNA platform revealed that time of sample collection has a predictable impact on RNA expression patterns. This is the first evidence of this phenomenon demonstrated in a pregnant population, and particularly the morning group is differentiated from later blood draws. The effect of circadian rhythm is an important consideration in current and future development of biomarkers making the time of sample collection an important variable to collect and analyze.