METHODS FOR IMPROVING THE EFFICIENCY OF ESTIMATING TOTAL OSTEON DENSITY IN THE HUMAN ANTERIOR MID-DIAPHYSEAL FEMUR

In order to preserve whole bone integrity and minimize destruction, pa leohistologists often rely on histomorphometric data obtained from sma ll areas (1.5-50 mm(2)) sampled within the anterior mid-diaphyseal fem ur. Because bone exhibits significant histological variation, the vali dity of results based on such sampling is questionable. The accuracy o f various subareas (columns, rows, squares approximating dimensions an d locations assessed by paleohistologists) in predicting total osteon density in the anterior mid-diaphyseal femur is assessed in the presen t study. Thirty-five specimens (12.7 mm wide, 100 mu m thick, average area 56.7 mm2) were chosen at random from a skeletal population of 94 Inuits and Pueblo agriculturists. The specimens were photographed and enlarged; an acetate grid (12 columns, 10 rows, 120 squares, square = 1 mm(2) of bone surface) was superimposed over the photograph; and sec ondary osteons and fragments were identified. Alternate columns (50% t otal area, T.Ar) predicted over 98% of entire section total osteon den sity. Two column combinations (15% T.Ar), separated by at least one co lumn, predicted 91 to 95% of total osteon density. Individual column ( 8% T.Ar) predictability ranged from 48 to 86%. Two row combination (32 to 40% T.Ar) predictability values ranged from 86 to 95%. Individual rows (<1 to 20% T.Ar) predicted from 45 to 92% of total variation. Com binations of squares approximating areas and locations assessed by oth er paleohistologists ranged in predictability values from 80 to 94%, T he results demonstrate that subareas of as little as 15% predict 95% o f variation in total osteon density in the entire anterior mid-diaphys eal femoral section. A minimization of histological area evaluated wit hout the loss of accuracy allows for a minimization of time invested i n data collection and the utilization of partially damaged specimens.