Inflammatory Proteins are Highly Variable During a Natural Stressor: A Call for Intensive Longitudinal Blood Microsampling.

Although research into how inflammation mediates the effects of stress on disorders such as heart disease and depression has grown rapidly, standard longitudinal designs often have measurement lags of months or years between assessments. This fact is at odds with evidence that stress and inflammatory biology are both highly dynamic-requiring much shorter time lags to quantify change as it naturally occurs, with minimal risk of confounding or effects diluting over time. To investigate the value of intensive longitudinal collection of inflammatory proteins, we collected high-frequency (every 3 days) blood microsamples over 22 days as 86 incoming college students transitioned onto campus (total observations = 622 blood samples). After data cleaning, the number of observations ranged from 176 (22 participants) to 280 (35 participants), depending on the protein. Samples were analyzed for CRP, TNF-α, IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12p70, and IL-17A. Reliability was quantified using both average 3-day correlations and intra-class correlation coefficients (ICCs) given their interpretive differences. Average 3-day correlations ranged from r=.09-.86 across the proteins. ICCs more consistently indicated low reliability, with all proteins except CRP and IL-10 being characterized by less than 6% stable between-person differences. These results provide compelling evidence that high frequency immune assessment is likely necessary to capture natural variability in these proteins.