Applied immuno-epidemiological research: an approach for integrating existing knowledge into the statistical analysis of multiple immune markers.

BMC Immunol 2016 05 20;17(1):11. Epub 2016 May 20.

London School of Hygiene & Tropical Medicine, London, UK.

Background: Immunologists often measure several correlated immunological markers, such as concentrations of different cytokines produced by different immune cells and/or measured under different conditions, to draw insights from complex immunological mechanisms. Although there have been recent methodological efforts to improve the statistical analysis of immunological data, a framework is still needed for the simultaneous analysis of multiple, often correlated, immune markers. This framework would allow the immunologists' hypotheses about the underlying biological mechanisms to be integrated.

Results: We present an analytical approach for statistical analysis of correlated immune markers, such as those commonly collected in modern immuno-epidemiological studies. We demonstrate i) how to deal with interdependencies among multiple measurements of the same immune marker, ii) how to analyse association patterns among different markers, iii) how to aggregate different measures and/or markers to immunological summary scores, iv) how to model the inter-relationships among these scores, and v) how to use these scores in epidemiological association analyses. We illustrate the application of our approach to multiple cytokine measurements from 818 children enrolled in a large immuno-epidemiological study (SCAALA Salvador), which aimed to quantify the major immunological mechanisms underlying atopic diseases or asthma. We demonstrate how to aggregate systematically the information captured in multiple cytokine measurements to immunological summary scores aimed at reflecting the presumed underlying immunological mechanisms (Th1/Th2 balance and immune regulatory network). We show how these aggregated immune scores can be used as predictors in regression models with outcomes of immunological studies (e.g. specific IgE) and compare the results to those obtained by a traditional multivariate regression approach.

Conclusion: The proposed analytical approach may be especially useful to quantify complex immune responses in immuno-epidemiological studies, where investigators examine the relationship among epidemiological patterns, immune response, and disease outcomes.

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12865-016-0149-9DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4875650PMC
May 2016
28 Reads

Publication Analysis

Top Keywords

statistical analysis
12
immunological mechanisms
12
immune markers
12
immune
9
analysis multiple
8
analytical approach
8
summary scores
8
correlated immune
8
cytokine measurements
8
multiple cytokine
8
immunological summary
8
immunological
8
immuno-epidemiological studies
8
markers
6
multiple
5
scores
5
children enrolled
4
large immuno-epidemiological
4
818 children
4
enrolled large
4

References

(Supplied by CrossRef)

B Hellriegel et al.
Trends Parasitol 2001

S Ozkara et al.
Eur Arch Otorhinolaryngol 2012

S Bennett et al.
J Immunol Methods 1992

D McGuinness et al.
J Immunol Methods 1997

HW Uh et al.
BMC Immunol 2008

N Ballenberger et al.
PLoS One 2012

PH Eilers et al.
BMC Immunol 2012

Similar Publications