Temporal variation in femoral cortical thickness 

 of North American Plains Indians 



Douglas W. Owsley 



Oortical bone growth during periods 

 of juvenile gain and later adult loss has 

 been documented in long-term studies 

 of populations in Central and North 

 America (Gam 1970). Several vari- 

 ables aft'ect tubular bone cortical thick- 

 ness including age, sex, and nutrition. 

 Changes in the bone envelope are 

 surface-specific and reflect the com- 

 bined response of subperiosteal apposi- 

 tion and endosteal resorption or apposi- 

 tion. 



Simple malnutrition slows the rate of 

 bone growth and can lead to the forma- 

 tion of less bone (Gam 1970,1972). 

 Subperiosteal growth depends more on 

 caloric sufficiency, protein seems to be 

 less of a limiting nutrient. The effect of 

 protein-energy malnutrition on cortical 

 thickness is seen primarily on the inner 

 bone surface. Endosteal resorption in- 

 duced by kwashiorkor or marasmus can 

 reduce the cortical wall to a thin shell 

 with an enlarged medullary cavity. As 

 much as 40% of the preformed bone 

 can be lost, even though the external 

 bone size remains relatively unaffected 

 (Gam 1970; Gam et al. 1964,1969). 

 Recovery-related catch-up growth or 

 surface repair through endosteal re- 

 placement is limited (Gam 1966). 



Although most studies of cortical 

 bone mass have focused on the living, 

 these observations can be applied to the 

 analysis of archeological samples. Bio- 

 archeological interpretations of past 

 subsistence patterns have used cortical 

 bone growth and thickness as an indica- 

 tor of nutritional status (Cashion 1987; 

 C(X)k 1979; Huss-Ashmore 1978; 



Zagreb Paleopathology Symp. 1*^88 



Hummert 1983; Hummert and Van 

 Gerven 1983; Keith 1981; Owsley 

 1985). Cashion (1987) and Owsley 

 (1985) have reported age-controlled 

 adult femoral midshaft cortical thick- 

 nesses in temporally sequential North 

 American Arikara Indians of South 

 Dakota. Comparison of bone cortex 

 data for villagers representing three ar- 

 cheological variants of the Plains Vil- 

 lage Coalescent Tradition (Extended. 

 A.D. 1550-1675; Post-Contact, a.d. 

 1675-1780; and Disorganized, a.d. 

 1780-1845) revealed statistically sig- 

 nificant differences. Relative to the ear- 

 lier Extended Coalescent and later Dis- 

 organized Coalescent samples, the 

 Post-Contact Variant sample showed 

 more cortical bone, presumably reflect- 

 ing greater success in meeting village 

 nutritional needs. 



This presentation examines the his- 

 torical patterning of these temporal dif- 

 ferences. Specific objectives are to de- 

 fine the surface-specific nature of the 

 differences reported between archeo- 

 logical variants and to apply greater 

 temporal control using chronological 

 periods of shorter duration. Cortical 

 thickness is a composite measure deter- 

 mined by both the medullary cavity and 

 the total subperiosteal diameters. Are 

 changes in cortical thickness caused by 

 changes in only one or both of these 

 dimensions? Greater temporal control 

 was made possible by obtaining data 

 for related sites and by sorting this 

 larger data base into smaller temporal 

 units representing Late Prehistoric 

 (A.D. 1600-1650), Eariy Protohistoric 



(1650-1740), Late Protohistoric 

 (1740-1795), and Historic (1795- 

 1832) period sites. Comparison of 

 these four samples provides a clearer 

 representation as to the timing of corti- 

 cal bone change during the Post- 

 Contact period. An appreciation of this 

 timing is essential for understanding 

 the nutritional impact of contact-related 

 historical events. 



Materials and methods 



Sample sizes and approximate dates of 

 the 12 archeological sites included in 

 this analysis are presented in Table 1 . In 

 order to avoid age-related cortical bone 

 involution, the analysis was limited to 

 femora of young adults aged 16-35 

 years, giving a total of 1 10 males and 

 134 females. The bones were x-rayed 

 in a standardized posterior-anterior pro- 

 jection using a Kodak single lanex, fine 

 screen X-omatic cassette. Only bones 

 in good condition were measured with 

 preference given the left side when 

 available. 



Two midpoint cross-sectional obser- 

 vations, total subperiosteal diameter 

 (T) and width of the medullary cavity 

 (M), were measured with a Helios dial 

 caliper to 0.1 mm (cf. Gam 1970). 

 Three composite variables were de- 

 rived from these measurements: corti- 

 cal thickness (C), Nordin's Index (NI), 

 and cortical area (CA). Cortical thick- 

 ness was calculated as C = T — M. The 

 value C represents the combined or net 

 thickness of the medial and lateral 

 walls. Nordin's Index (NI) or score was 



105 



