Inter-population differences in acetabular senescence: relevance in age-at-death estimation

**Background:** Age-at-death estimation from adult skeletal remains remains challenging, particularly for pelvic age markers. While the acetabulum has gained recognition as a valid age marker, no prior study had investigated inter-population differences in acetabular aging rates. This study aimed to evaluate the relevance of geographical origin in acetabular senescence and the usability of the SanMillán-Rissech aging method across populations.

**Methods:** The study analyzed 826 European North Americans from the William Bass Donated Skeletal Collection (USA, deaths 1977-2013) and 611 Portuguese individuals from the Luis Lopes Collection (Portugal, deaths 1880-1975). Seven acetabular variables were assessed. Four experiments were conducted: (1) Portuguese tested with Portuguese reference, (2) North Americans tested with North American reference, (3) North Americans tested with Portuguese reference, and (4) Portuguese tested with North American reference. Mann-Whitney U tests evaluated inter-population differences in mean age per variable stage. Wilcoxon tests compared bias and absolute error between population-specific and non-specific reference conditions. Analyses were stratified by sex and age groups (<40, 40-64, >65 years).

**Key Results:** Inter-population analyses revealed that European North Americans reach consecutive acetabular stages at younger mean ages than Portuguese, indicating faster skeletal aging. This pattern was more pronounced in females, with significant differences (p<0.05) appearing across multiple variables and stages. For example, in females, variable 3 stages 3, 4, and 5 all showed significantly lower mean ages in North Americans (59.64, 69.39, 72.34 years) compared to Portuguese (63.49, 75.02, 81.39 years). Using population-specific reference samples generally produced better outcomes. For Portuguese males, global bias was -0.16 years with Portuguese reference versus -3.23 years with North American reference (p<0.001). For North American males, global bias was -0.17 years with North American reference versus 4.46 years with Portuguese reference (p<0.001). Absolute error for Portuguese males was 7.54 years (population-specific) versus 8.08 years (non-specific, p=0.301); for North American males, 7.30 versus 8.31 years (p=0.006). Notably, in the >65 age group, using the Portuguese reference for North Americans reduced bias (males: -0.53 vs -4.77 years; females: 0.89 vs -3.82 years). The meta-analysis showed acetabular mean errors of 7.09-9.65 years with population-specific references and <10.5 years with non-specific references, compared to pubic symphysis errors of 9.0-20.3 years and auricular surface errors of 7.55-16.47 years.

**Clinical Implications:** These findings demonstrate that population-specific reference samples significantly improve acetabular age estimation accuracy, though the SanMillán-Rissech method outperforms traditional pelvic methods even with non-specific references. The acetabulum's better preservation and superior accuracy support its prioritization over pubic symphysis and auricular surface in forensic and bioarchaeological contexts. Understanding population-specific aging rates can help anthropologists select appropriate reference samples to minimize bias. The study underscores that factors beyond chronological age—including genetics, lifestyle, socioeconomic status, and secular trends—influence skeletal aging and must be considered in age estimation.