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

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.

Electronic supplementary material

The online version of this article (doi:10.1186/s12865-016-0149-9) contains supplementary material, which is available to authorized users.

Risk factors and immunological pathways for asthma and other allergic diseases in children: background and methodology of a longitudinal study in a large urban center in Northeastern Brazil (Salvador-SCAALA study)

BACKGROUND

The prevalence of asthma and allergic diseases has increased in industrialised countries, and it is known that rates vary according whether the area is urban or rural and to socio-economic status. Surveys conducted in some urban settings in Latin America found high prevalence rates, only exceeded by the rates observed in industrialised English-speaking countries. It is likely that the marked changes in the environment, life style and living conditions in Latin America are responsible for these observations. The understanding of the epidemiological and immunological changes that underlie the increase in asthma and allergic diseases in Latin America aimed by SCAALA studies in Brazil and Ecuador will be crucial for the identification of novel preventive interventions.

METHODS/DESIGN

The Salvador-SCAALA project described here is a longitudinal study involving children aged 4-11 years living in the city of Salvador, Northeastern Brazil. Data on asthma and allergic diseases (rhinitis and eczema) and potential risk factors will be collected in successive surveys using standardised questionnaire. This will be completed with data on dust collection (to dust mite and endotoxin), skin test to most common allergens, stool examinations to helminth and parasites, blood samples (to infection, total and specific IgE, and immunological makers), formaldehyde, physical inspection to diagnoses of eczema, and anthropometric measures. Data on earlier exposures when these children were 0-3 years old are available from a different project.

DISCUSSION

It is expected that knowledge generated may help identify public health interventions that may enable countries in LA to enjoy the benefits of a "modern" lifestyle while avoiding--or minimising--increases in morbidity caused by asthma and allergies.