Measuring allostatic load: Approaches and limitations to algorithm creation

Abstract

Objective: Allostatic load literature has proliferated over the past three decades, and a growing body of research demonstrates that higher levels of allostatic load are associated with a wide range of negative physical and mental health outcomes. However, there remain significant challenges with operationalization of the concept. A scoping review of the methods employed to create an allostatic load algorithm was conducted and recommendations for future research with an orientation towards advancing clinical application of the theory are discussed.

Methods: A search of seven electronic databases (PubMed, PsycINFO, Social Work Abstracts, Social Service Abstracts, Social Sciences Citation Index (Web of Science), Sociological Abstracts, Scopus) was completed with the search term "allostatic load." Studies were reviewed, and if they met the inclusion criteria, data was extracted, complied, and presented in the narrative, table, and figures.

Results: The initial searches yielded 5280 results with the final sample of 395 non-duplicate articles that met the inclusion criteria. More than half (52.5%) of all included publications employed biomarker cutoffs based on the high-risk quartiles of the sample distribution, 11.1% employed the sum of at-risk clinical scores, and the remainder of studies utilized a range of different algorithms.

Conclusion: Allostatic load literature has grown at an exponential rate in recent years, but researchers continue to operationalize the concept via algorithms that may have limited utility moving forward. More nuanced statistical approaches are emerging and should be considered, as should a shift towards an approach that can provide additional clinical utility.

Keywords: Allostatic load; Biological dysregulation; Biomarkers; Scoping review.