Effects of field size when using Kerley's histological method for determination of age at death.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 58:123-125(1982) Effects of Field Size When Using Kerley’s Histological Method for Determination of Age at Death SAMUEL D. STOUT A N D SARAH J. GEHLERT Department of Anthropology, University of Missouri, Columbia, Missouri 65211 KEY WORDS Osteon, Age estimation, Field size, Histology, Bone ABSTRACT The field size at which a bone is read affects the results obtained when using Kerley’s histological method for age estimation, even after applying the recommended correction factor. Whereas there is no tendency for any one of three field sizes tested to consistently underestimate or overestimate age, a field size closest to that used by Kerley in his original study had significantly lower variances for its age estimates, and thus provides greater reliability. This particular field size yields more precise estimates because it is sampling a pattern and number of structures more similar to that of Kerley. Correction factors cannot equalize the counts of osteons and osteon fragments because of spatial variations in the distributions of these histological structures. A field size similar to that used by Kerley in gathering the data from which he developed his regression equations must be used to assure that the same pattern and number of structures is being sampled. For this reason, we suggest a field size as close to 2.06 mm2as possible be used when employing Kerley’s method. It has been well established that differences in field size can introduce error when employing Kerley’s (1965) histological aging method (Ubelaker, 1977.) Since the field size that is observable at the same magnification, but with different oculars and objectives can vary considerably, a correction factor has been recommended that will make the observed field size comparable to the original field size used by Kerley in developing his regression formulas (Kerley and Ubelaker, 1978). A great deal of spatial variation has been reported within the cortex of bone (Amprino and Marotti, 1964; Frost, 1969). Merely adjusting osteon counts in proportion to differences in field size, therefore, may be of limited value. The purpose of this study is to investigate the limitation of the field size correction factor currently recommended for Kerley’s histological aging method. MATERIALS AND METHODS A sample of twenty femoral sections from twenty individuals was used in this study; all were obtained from dissecting room specimens. These individuals were well 0002-9483/82/5802-0123$01.~~ . 1982 ALAN R. LISS, INC documented as to age, sex, and cause of death. The sample consisted of nine females and eleven males. The age range was 13-95 years, with a mean of 55 years. A Nikon microscope with a l o x objective was used with three different l o x oculars, each of which produced a different sized field of view. The area of each field was measured with a stage micrometer. Following the procedure suggested by Ubelaker (1977),correction factors were derived by dividing each of the field sizes into Kerley’s estimated original field size of 2.06 mm2.Our first field, using a Nikon wideangle ocular, had a diameter of 1.90 mm. This yielded an area of 2.83 mm2which, when multiplied by 0.73, the correction factor for this ocular, gives the desired corrected field area of 2.06 mm2(Kerley and Ubelaker, 1978).The second ocular, an aus Jena wide-angle,had a measured field diameter of 1.70 mm, and an area of 2.27 mm’. The correction for this ocular was 0.91. The third ocular was a Merz, with a square reticulated grid, which measured 1.03 Received October 2. 1980; accepted September 14. 1981 124 S.D. STOUT AND S.J. GEHLERT mm2 with an area of 1.06 mm2. I t s correction factor was 1.94. The number of osteons, osteon fragments, and non-Haversian canals was counted at each field size using the methods described by Kerley (1965).Each of these counts was multiplied by the appropriate correction factor. Age estimates were computed from each of the corrected counts using Kerley’s revised regression equations (Kerley and Ubelaker, 1978). Corrected counts of osteons, osteon fragments, and non-Haversian canals were first compared. Differences in the corrected counts among the three field sizes were determined for each variable for the 20 femora. The mean and standard deviation of the differences for each of the variables was computed, and a t-test applied to determine whether the mean difference in corrected counts for any field size differed significantly. The same procedure was used to compare the differences in age estimates derived from each field size, and to determine whether any of the age estimates derived from each field size, differed significantly from the known age. RESULTS Results of the t-test of mean differences between corrected raw counts for the 20 femora showed that one of nine variables differed significantly from zero (see Table 1).This mean difference was of the number of osteon fragments between the 2.83 mm’ and the 1.06 mm’ field sizes. None of the mean differences between age estimates were found to be significantly different for any of the three field sizes (see Table 1). When estimated age and known age were compared for each count at each field size, three of the nine estimates differed significantly from known age (see Table 2). Those differing significantly were ages estimated from the non-Haversian canal regression equations for all three field sizes. Variances were lowest for all counts taken at the 2.27 mmz field size. Two of these counts, those based on the number of osteons and the number of osteon fragments, had significantly lower variances, using the F statistic, than did the variances for all other counts (see Table 2). DISCUSSION Field size differences of the magnitudes represented in this study have little effect upon age prediction using osteon and osteon fragment counts when the correction factor described by Kerley and Ubelaker (1978)is introduced. Non-Haversian canal count, however, significantly underestimated actual at all three field sizes. Kerley’s equation for this measurement truncates at 58.39 years, and, therefore, cannot be applied to older ages. Since 40% of our sample is over the age of 58 years, this probably acounts for the observed tendency towards underestimation. Although none of three field sizes showed a tendency to consistently underestimate or overestimate age, the field size closest to that reportedly used by Kerley in his original study exhibited a significantly lower variation in differences between known age and age predicted from osteon and osteon fragment counts. Therefore, for any given case, there is a significantly greater reliability for age prediction when a field size similar to that used by Kerley (1965) is employed, even when the proposed correction for field size difference is introduced. Within the cortex of bone, structures are not uniformly distributed i.e., there is a great deal of spatial variation (Amprino and Marotti, 1964; Frost, 1969).This spatial variation may be the reason that the field size closest to Kerley’s original size gives more precise age estimates when using his regression equations. If all structures, such as osteons and osteon fragments, were evenly distributed throughout the cortex, one could read at any field size. Since this is not the case, it is necessary to use a field size similar to that which was used to determine the original regression equations in order to assure that the same patterns and structures are being sampled. This suggests that care should be taken when using Kerley’s method to use a field size as close as possible to 2.06 mm2. Spatial variation may also account for why the 1.06 mm2 field size showed significantly higher osteon fragment counts than the other two field sizes. This field size is smaller than the others. It thus samples an area which is both smaller, and more closely situated at the periosteal edge of the cortex. This evidence may indicate that the area closer to the periosteum contains a higher concentration of osteon fragments than other areas of the outer third of the cortex. This observation may also be due in part to the older age distribution of our sample. In younger bone samples, the relative number of fragments would tend to be re- 125 FIELD SIZE AND AGE ESTIMATES FROM BONE TABLE 1. Mean differences between corrected counts and age estimates from fields of different sizes‘ 2.83 mm2 field Count Osteons mean std. dev. t-value’ Osteon fragments mean std. dev. t-value Non-Haversian canals mean std. dev. t-value 2.27 mm2 field 2.83 mm’ field less less less 2.27 mm’ field 1.06 mm2 field 1.06 mm’ field Counts Ages - 1.03 Counts Ages Counts Ages 15.19 - 0.30 0.20 12.05 0.07 1.77 16.35 0.48 0.20 12.14 0.08 0.74 14.25 0.23 0.41 11.27 0.16 - 2.11 - 2.23 - 7.45 - 4.49 11.49 - 0.87 - 5.34 12.46 - 1.92 - 2.27 12.17 0.77 10.84 - 0.94 15.11 - 2.20* - 1.47 - 0.66 1.07 5.32 0.90 - 2.37 1.25 - 0.92 4.83 0.85 0.47 7.65 0.85 - - 13.64 1.59 5.40 1.31 1.54 8.07 0.27 - ‘Sample size is 20 for all measures. ’The t-value for difference of mean from zero. *Significance level of P < 0.06. TABLE 2.Mean differences between actual ages and those estimated from counts of three different histological structures at three different field sizes (in years)‘ Age estimated from counts of Osteons mean std. dev. t-value’ Osteon fragments mean std. dev. t-value Nan-Haversian canals mean std. dev. t-value 2.83 mm2 field - 2.15 2.27 mm* field - 2.35 1.06 mm’ field - 2.56 15.13 0.64 - - 4.29 - 10.61 - 1.81 - - - 8.62 - 9.08 12.54 - 3.07;; - 3.19;; - 7.55 14.37 - 2.35* 6.43t 1.63 2.07 6.46t 1.43 - 14.38 0.79 0.20 12.19 0.07 12.74 ‘sample size is 20 for dl measures. ’The t-value for mean difference from zero. ?The squares of these standard deviations, i.e.. the variances of these measures, are. by the F statistic, smaller than those for all other measures at the P = 0.05 level. *Significance level of P < 0.05. **Significance level of P < 0.01. duced near the periosteum due to the presence of circumferential lamellae. ACKNOWLEDGMENTS B~~~ specimens used in this study were made through the courtesy Of Dr’ Roy Department Of Anatomy and Neurobiology, Washington University School of Medicine, and Dr. William Goodge, Departmerit of Anatomy, University of ~ i ~ ri-(=oumbia School of Medicine. T L research was funded in part by grant AM 22118 from the National Institute of Health. LITERATURE CITED Amprino, R, and Marotti G (1964) A topographic study of bone formation and reconstruction. In H J J Blackwood (ed):Bone and Tooth Symposium. New York: MacMilian Company, pp. 21-23. Frost, HM (1969) Tetracycline-based histological analysis of hone remodeling. Calc. Tiss. Res. 3.211-327. Kerley, ER (1965) The microscopic determination of age in human hone. Am. J. Phys. Anthrop. 23149-164, Kerley. ER, and Ubelaker, DH (1978) Revisions in the microscopic method of estimating age a t death in human cortical ~ bone. Am. ~ J. Phys. ~ Anthrop. ~ 49:545-546. Ubelaker. DH 11977) Problems of the microscopic deter mination of age a t death. Paper presented to the 29th annual meeting of the American Academy of Forensic Sciences, San Diego, California.