A physiometric investigation of inflammatory composites: Comparison of “a priori” aggregates, empirically-identified factors, and individual proteins

Most research testing the association between inflammation and health outcomes (e.g., heart disease, diabetes, depression) has focused on individual proteins; however, some studies have used summed composites of inflammatory markers without first investigating dimensionality. Using two different sam...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Daniel P. Moriarity, Lauren M. Ellman, Christopher L. Coe, Thomas M. Olino, Lauren B. Alloy
Formato: article
Lenguaje:EN
Publicado: Elsevier 2021
Materias:
Acceso en línea:https://doaj.org/article/8a75ef43c45f43a49fef478b8b00ddb6
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Descripción
Sumario:Most research testing the association between inflammation and health outcomes (e.g., heart disease, diabetes, depression) has focused on individual proteins; however, some studies have used summed composites of inflammatory markers without first investigating dimensionality. Using two different samples (MIDUS-2: N ​= ​1255 adults, MIDUS-R: N ​= ​863 adults), this study investigates the dimensionality of eight inflammatory proteins (C-reactive protein (CRP), interleukin (IL)-6, IL-8, IL-10, tumor necrosis factor-α (TNF-α), fibrinogen, E-selectin, and intercellular adhesion molecule (ICAM)-1) and compared the resulting factor structure to a) an “a priori”/tau-equivalent factor structure in which all inflammatory proteins equally load onto a single dimension (comparable to the summed composites) and b) proteins modeled individually (i.e., no latent variable) in terms of model fit, replicability, reliability, and their associations with health outcomes. An exploratory factor analysis indicated a two-factor structure (Factor 1: CRP and fibrinogen; Factor 2: IL-8 and IL-10) in MIDUS-2 and was replicated in MIDUS-R. Results did not clearly indicate whether the empirically-identified factor structure or the individual proteins modeled without a latent variable had superior model fit, but both strongly outperformed the “a priori”/tau-equivalent structure (which did not achieve acceptable model fit in any models). Modeling the empirically-identified factors and individual proteins (without a latent factor) as outcomes of medical diagnoses resulted in comparable conclusions. However, modeling individual proteins resulted in findings more robust to correction for multiple comparisons despite more conservative adjustments. Further, reliability for all latent variables was poor. These results indicate that modeling inflammation as a unidimensional construct equally associated with all available proteins does not fit the data well. Instead, individual inflammatory proteins or, potentially (if empirically supported and biologically-plausible) empirically-identified inflammatory factors should be used in accordance with theory.