Bone loss markers in the earliest Pacific Islanders

Abstract

Kingdom of Tonga in Polynesia is one of the most obese nations where metabolic conditions, sedentary lifestyles, and poor quality diet are widespread. These factors can lead to poor musculoskeletal health. However, whether metabolic abnormalities such as osteoporosis occurred in archaeological populations of Tonga is unknown. We employed a microscopic investigation of femur samples to establish whether bone loss afflicted humans in this Pacific region approximately 3000 years ago. Histology, laser confocal microscopy, and synchrotron Fourier-transform infrared microspectroscopy were used to measure bone vascular canal densities, bone porosity, and carbonate and phosphate content of bone composition in eight samples extracted from adult Talasiu males and females dated to 2650 BP. Compared to males, samples from females had fewer vascular canals, lower carbonate and phosphate content, and higher bone porosity. Although both sexes showed evidence of trabecularised cortical bone, it was more widespread in females (35.5%) than males (15.8%). Our data suggest experiences of advanced bone resorption, possibly as a result of osteoporosis. This provides first evidence for microscopic bone loss in a sample of archaeological humans from a Pacific population widely afflicted by metabolic conditions today.

Figure 1

Summary of methods and key findings in the present study. Posterior view of two right archaeological femora from Talasiu individuals estimated as female (♀, ID: Sk3.1) and male (♂, ID: Sk3.2) shows the sectioning location (black dashed line) and approximate sample removed (red dashed box). Methodological steps included synchrotron sourced infrared microspectroscopy to measure bone mineral composition, histomorphometry to estimate Haversian canal densities (red dots), and laser confocal scanning of bone topography to provide a qualitative illustration of intra-cortical bone porosity producing trabecularisation.

Figure 2

A summary of histomorphometric procedures for estimating densities of Haversian canals and abnormal intra-cortical bone porosity. (a, b): localised histology (60 ×) viewed under linearly polarised light showing secondary osteon lamellae (white arrows) surrounding Haversian canals (oval features), which were obscured by patches of diagenesis (grey arrows). Cement lines (black arrow in b) were inconsistently preserved (examples shown are from individuals BG2 (a) and Sk9.3 (b). Haversian canals were counted manually form thresholded images (c–f): (c) shows numbered canals, (d) shows manual segmentation of cortical wall from abnormal intra-cortical porosity (red dashed line), while the solid line marks whole section area in individual BG2; (e) illustrates final counts of Haversian canals, while (f) shows ‘extracted’ abnormal bone porosity.

Figure 3

Intra-cortical surfaces scanned using a three-dimensional laser confocal topography microscope Olympus OLS5000 with applied false-coloured heat map to illustrate extreme differences in porosity in archaeological posterior midshaft femur samples from Talasiu individuals estimated as female (♀, ID: Sk3.1) and male (♂, ID: Sk3.2). (a) shows topography viewed superiorly from the bone surface, whereas (b) is reverted upside down to show the inferior view of the same scan as (a).

Figure 4

Graphical summary of results illustrating descriptive comparisons of data between archaeological Tongan individuals of estimated male (♂) and female (♀) sex. The boxplots indicate that females present with: (a) smaller femoral midshaft (dm: diameter), with lower Haversian canal density (per mm²), but higher intra-cortical porosity producing trabecularisation (%); (b, c) lower phosphate content; and (d–f) lower carbonate, and carbonate:phosphate ratios. Outliers are marked by asterisks and circles. *p < 0.05; ***p < 0.0001.

Figure 5

Statistically significant Mann Whitney U ranking comparisons of the sFTIRM data for phosphate peaks (a), carbonate peaks and integrated area (b–e), and ratios computed using carbonate v₂: phosphate area (f). A strong negative correlation between cortical area and trabecularised cortical bone in the whole sample is shown (g).