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


Genser, B; Fischer, JE; Figueiredo, CA; Alcântara-Neves, N; Barreto, ML; Cooper, PJ; Amorim, LD; Saemann, MD; Weichhart, T; Rodrigues, LC; (2016) Applied immuno-epidemiological research: an approach for integrating existing knowledge into the statistical analysis of multiple immune markers. BMC Immunol, 17 (1). p. 11. ISSN 1471-2172 DOI: 10.1186/s12865-016-0149-9

[img] Text - Published Version
License:

Download (1MB)

Abstract

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. 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. 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.

Item Type: Article
Faculty and Department: Faculty of Epidemiology and Population Health > Dept of Infectious Disease Epidemiology
PubMed ID: 27206492
Web of Science ID: 379303900001
URI: http://researchonline.lshtm.ac.uk/id/eprint/2550761

Statistics


Download activity - last 12 months
Downloads since deposit
17Downloads
70Hits
Accesses by country - last 12 months
Accesses by referrer - last 12 months
Impact and interest
Additional statistics for this record are available via IRStats2

Actions (login required)

Edit Item Edit Item