TL;DR
This paper introduces a novel method for estimating the age of older adults (40+ years) by examining degenerative changes on the sternal end of the clavicle. By scoring surface topography, porosity, and osteophyte formation, the researchers developed a composite score and regression models that can refine age estimates beyond the typical "mature adult" category, offering a new tool for bioarchaeology.
What Problem Does This Paper Address?
Accurately determining the age of deceased individuals is crucial for understanding past populations, including their life courses, health, and social structures. While methods exist for aging younger adults, estimating the age of older adults (often broadly categorized as 46+ years) remains a significant challenge in bioarchaeology. This lack of precision can lead to the "invisibility" of the elderly in the archaeological record, potentially skewing our understanding of past demographics and life expectancies. Existing macroscopic aging techniques often plateau in accuracy around age 60, leaving a gap for identifying and differentiating within the older adult spectrum. This study aims to fill this gap by investigating the sternal end of the clavicle, a bone that fuses later in life, making it a potential indicator of age-related degeneration in older individuals.
Methods
The researchers developed an ordinal scoring system to quantify three macroscopic degenerative traits on the sternal end of the clavicle:
- Surface topography: Describing the texture and relief of the bone surface, ranging from smooth to completely broken down.
- Porosity: Quantifying the presence and severity of small (micro) and large (macro) perforations in the bone surface.
- Osteophyte formation: Assessing the degree of bony outgrowths (lipping) along the joint rim.
Each trait was scored on a scale, and these scores were summed to create a composite score. This composite score was then used to develop linear regression models to predict age. The study also included rigorous testing for intra- and inter-observer error to ensure the reliability of the scoring system. A blind test was conducted on a separate skeletal collection to evaluate the accuracy of the developed methods.
Datasets
The study utilized four documented skeletal collections for training and developing the aging criteria:
- Hamann-Todd Human Osteological Collection (HT): 20th-century U.S. skeletal collection.
- Pretoria Bone Collection (PBC): Late 20th-century South African skeletal collection.
- St. Bride's Documented Skeletal Collection (SB): 18th-19th century British skeletal collection.
- Coimbra Human Identified Osteological Collection (CC): Late 19th-early 20th century Portuguese skeletal collection.
A total of 564 clavicles from individuals aged 40+ years were analyzed.
For blind testing and validation, a sample of 56 individuals from the Christ Church Spitalfields Documented Skeletal Collection (Natural History Museum, London) was used.
Key Insights and Ideas
- The sternal end of the clavicle is a promising indicator of age-related degeneration beyond epiphyseal fusion. Unlike other skeletal markers that plateau, the sternal clavicle shows progressive changes well into older adulthood.
- Surface topography is the most reliable indicator of age. This feature showed the strongest correlation with chronological age among the three traits examined.
- A composite score of multiple degenerative traits offers a more robust age estimation. Combining topography, porosity, and osteophyte formation provided a better predictor of age than any single trait alone.
- The developed methods show high accuracy in blind tests. This suggests the criteria are generalizable and can be applied to new, unknown skeletal samples.
- Sex differences in aging patterns exist, but can be accounted for. While females tended to be slightly older for a given composite score, the regression models can incorporate sex as a variable or be applied separately.
Key Results
- Spearman's rank correlation coefficients showed significant correlations between all three degenerative traits (surface topography, porosity, osteophyte formation) and known age across all study samples.
- Surface topography consistently exhibited the strongest correlation with age (e.g., male $r_s = 0.762$, female $r_s = 0.665$ in the combined sample).
- Composite scores also showed strong correlations with age (e.g., male $r_s = 0.755$, female $r_s = 0.644$ in the combined sample).
- In blind tests on 56 individuals, both the composite score method and the regression formulae achieved 96.4% accuracy in placing individuals within a 95% confidence interval.
- The regression equation for predicting age, incorporating topography and porosity, was developed. For example, one prediction model is: $27.21 + 8.70 \times \text{Topo (right)} + 2.00 \times \text{Poro (right)} - 3.65 \times \text{Male}$.
Limitations and Future Work
- Current criteria are limited to individuals aged 40+ years. The study did not investigate younger adults or the initial stages of fusion.
- The 95% confidence intervals for age estimates are broad. This means the methods are best suited for refining age ranges rather than providing precise single-year estimates, especially in forensic contexts.
- The role of occupation and handedness needs further investigation. While generally stable across populations, the side differences observed in the Coimbra sample suggest these factors might influence degeneration.
- Secular changes in epiphyseal union and degeneration warrant further study.
- Interobserver agreement for porosity and osteophyte formation was moderate. The authors suggest that improved descriptions or 3D casts might be beneficial for more reliable trait identification.
- Future research could explore more advanced statistical approaches such as Bayesian methods or transition analysis to improve accuracy and precision.
- The study acknowledges potential limitations in the accuracy of "known ages" in some historical collections, though the consistency across multiple samples mitigates this concern.
Summary updated