Age estimation of adult skeletal remains is still a very uncertain procedure. This study
represents an attempt to decrease the estimation error associated with adult age estimation
in a forensic context, by using alternative bone indicators and statistical approaches. A
sample of 91 known-age adult skeletons (47 females and 44 males) from the Spitalfields
skeletal collection housed at the Natural History Museum in London was used. Five
age-related variables were assessed on the sternal end of the clavicle: microporosity
(<1mm); macroporosity (>1mm), sclerosis; new bone formation and marginal lipping.
Ages where categorized in ten-year intervals and multinomial regression was used to
estimate the probability of an individual being in a certain age interval conditional on
the observed data (skeletal indicators). Results presented here are considered preliminary and illustrative of alternative skeletal indicators for age estimation in adults using a Bayesian approach