Frail or hale: Skeletal frailty indices in Medieval London skeletons

Frail or hale: Skeletal frailty indices in Medieval London skeletons

By Kathryn E. Marklein and Douglas E. Crews

PLoS ONE, Vol.12:5 (2017)

London Black Death victim

Skeleton discover in London with Black Death – Photo by Robby Whitfield/Crossrails

Abstract: In this paper, we apply a recently proposed methodology for assessing frailty using a skeletal frailty index (SFI, see [1]) and discuss its applicability to archaeological samples. We developed the SFI to assess frailty in skeletal materials following extensive observations on biomarkers of frailty in modern living humans [2–7]. Originally, we proposed a SFI including 13 skeletal biomarkers observable on human skeletons, representing four categories of stressors: trauma, nutrition/disease, physical activity, and growth disruptions. The strength we advocated for in this index was its broad range of measurable and observable biomarkers that capture aspects of both childhood and adult frailty.

Unfortunately, in many skeletal assemblages, available materials are not sufficient to assess all 13 biomarkers included in the original SFI. Using identical samples from Medieval London [1], we examine how versions of the SFI constructed of fewer biomarkers may increase available sample size while maintaining results comparable to the more comprehensive 13-variable SFI. We expect reducing the number of contributing assessments will substantially increase available sample sizes. Our goal is to determine the fewest number of biomarkers needed in a SFI to reproduce results in the Medieval sample equivalent to the original 13-variable SFI, i.e., the lower limits for a robust SFI with statistical similarities in explained variance and significance. Our aim is to balance the maximization of frailty information (number of biomarkers) with population representation (sample size). After exploring applicability and reliability of reduced-variable SFIs (<13 biomarkers) we assess two modified SFIs (4- and 6-biomarkers) for associations with lifestyle (determined by cemetery/location), age, and sex in a larger monastic and nonmonastic Medieval sample than assessed previously to develop the original SFI.

Click here to read this article from PLoS ONE

Sign up to get a Weekly Email from

* indicates required

medievalverse magazine