Remote dried blood spot collection for inflammatory markers in older adults is feasible, reliable, and valid

Abstract

Dried blood spots (DBS) provide a minimally invasive method to assess inflammatory markers and can be collected remotely at-home or in-person in the lab. However, there is a lack of methodological information comparing these different collection methods and in older adults. We investigated the feasibility (including adherence, yield, quality, and participant preferences) and measurement properties (reliability, validity) of remotely collected DBS inflammatory markers in older adults. Participants (N = 167, mean age = 72, range: 60-96 years) collected their own DBS (finger prick on filter paper) during three remote interviews over ∼ 6 months. Within 4-5 days on average of their last remote interview, a subset of 41 participants also attended an in-person lab visit that included a researcher-collected DBS sample, venous blood draw, and survey to assess participant preferences of DBS collection. DBS and venous blood were assayed for CRP, IL-6, and TNF-α. Adherence: 98% of expected DBS samples (493 out of 501) were completed and mailed back to the lab. Yield: 97% of DBS samples were sufficient for all assays. Quality: On average, 0.80 fewer optimal spots (60uL of blood that filled the entire circle) were obtained remotely vs. in-person (p = 0.013), but the number of useable or better spots (at least 30-40uL of blood) did not differ (p = 0.89). Preference: A slight majority of participants (54%) preferred in-person DBS collection. Reliability: DBS test-retest reliabilities were good: CRP (ICC = 0.74), IL-6 (ICC = 0.76), and TNF-α (ICC = 0.70). Validity: Inflammatory levels from DBS correlated strongly with levels from venous blood (r = 0.60-0.99) and correlated as expected with sociodemographic and physical health and function variables. Older adults can remotely collect their own DBS to acquire reliable and valid inflammatory data. Remote DBS collection is highly feasible and may allow for inflammatory markers to be assessed in larger, more representative samples than are possible with lab- or clinic-based research designs.

Keywords: Aging; Dried blood spots; Home-based; Inflammation; Self-sample.

Figure 1.

Scatterplot (and correlations) and Passing-Bablok regression line (95% confidence interval) evaluating agreement between in-person DBS and in-person venous blood samples for CRP (top, r=.99, p<.001), IL-6 (middle, r=.81, p<.001), and TNF-α (bottom, r=.60, p<.001). Y (Venous Blood) = intercept + slope × Method 1 (DBS).