Y E AR BOO BOOK K OF PH Y SI CA L AN TH ROPOL OG OGY Y 40:139–1 –17 70 (1997)
Trauma Tr auma Analy Analysis sis in Paleopa Paleopatholog thology y NANCY NA NCY C . LOVELL Departm ent of Anth ropo Departm ropology logy,, U ni ve versity rsity of Alberta, Edmonton, AB T6G 2H 4, Canada
KEY WORDS
fra cture; dislo dislocation; cation; vio violenc lence; e; head injury injury
This paper reviews reviews the mechanisms mechanisms of injury a nd the types of fra ctures tha t most commonly commonly affect affect th e huma n skeleton, skeleton, presents presents descriptive protocols for crani al and po st cranial fractures adapted from c linical and forensic medic medicine, ine, and summa rizes an a tomic tomicaa lly the injuries most likely likely to be found in archaeological skeletons along with their most common causes and compl co mplic icat at io ions. ns. Mec Mecha ha nisms of injury are catego categorize rizedd as direc directt an d indire indirect ct trauma, stress, and fracture that occurs secondary to pathology. These are considered to be the proximate, or most direct, causes of injury and they are influencedd by intrinsic bio influence biolo logic gical al factors such as age and sex sex,, a nd extrinsi extrinsicc environme envi ronmenta nta l factors, both both physi physical cal a nd so soci cioc ocultural, ultural, th at may be thought of as the ultimate, or remote, causes of injury. Interpersonal conflict may be one of those causes but the skeletal evidence itself is rarely conclusive and must therefore be evalua evalua ted in its individua l, populationa populationa l, soci socioc ocultura ultura l, and physical physic al co context. ntext. A cautio cautiona na ry ta le regarding pa rry fractures is presented as Wiley-L L iss, I nc nc.. a n illust ra tion. Yrbk P hys Anth ropol 40:1 40:139–1 39–170, 70, 199 1997. 7. 1997 WileyABSTRACT
Tr a u m a m a y b e d efi efi n e d m a n y w a y s b ut ut conventiona conventio na ll llyy is understood understood to r efe eferr to a n injury to living ti ssue that is caused by a force or mechanism extrinsic to the body. The an at omic mical al imp importance ortance and soc socio iocu culltural imp impli liccation ationss of t rauma in a ntiqui ntiquity ty long have been recognized and the description tio n of tra uma in human ske skele leta ta l remains and the ide identificatio ntification n and comparis mparisoon of tra uma pa tterns a mong ancient ancient popul populat at io ions ns therefo there fore re have a le lengthy ngthy his history tory.. As the discipline of palaeopathology has developed, the objectives of traumatic injury analysis have shifted from a fo focus cus on on t he identification tio n a nd descriptio description n of the ea rlie rliest st a nd t he most unusual pathological specimens to the interpreta tion of the social, social, cultura l, or or environmenta ronme nta l causes of tra umat ic injury; injury; their relationship to biological variables, such as sex and age, tha t ma y ha ve soci social al or cultural cultural relevance; and their temporal an d spatial var iat io ion. n. Thus, Thus, interpreta tions of of the cause of tra uma in ant iquity range from from inter-an d intra group conflict (e.g., (e.g., Angel, Angel, 1974; 1974; Ha mperl, 1967; J anssens, 1970; J urmain, 1991;
19 1997 WIL EY-L I SS SS,, INC.
Liston and Ba ke Liston ker, r, 19 1996 96;; Shermi Shermis, s, 19 1984 84;; St ewa rt , 197 1974; 4; Wa lker, 198 1989; 9; Woo oodd-JJ ones, 1910 19 10;; Zivanovic, Zivanovic, 1982 1982;; an d others ) to environmentally or occupationally facilitated misadventure and accident (e.g., Angel, 1974; B urrell et a l., 198 1986; 6; Cybulski, 199 1992; 2; Gra uer an d Robe Roberts rts , 1996; 1996; Kelley a nd Angel, 1987 1987;; Lovejoy and Heiple, 1981; Wells, 1964; and others). Although great advances have been made in pale paleoopathol patholoogi giccal diagno diagnosis sis and interpreta interpre ta tion in rec recent ent years, inc inconsi onsistenstencies in descriptions and interpretations of t r a u m a i n t h e l it it e ra ra t u r e , p a r t i cu cu la la r l y a s they a ff ffec ectt our understa nding of of the na ture an d extent of interpersonal violenc violencee in an tiquity, have made it difficult to compare the results res ults of dif diffe ferent rent studi studies es and to acc accep eptt with co confi nfi dence some conclusio conclusions. ns. The purpose of this paper, therefore, is to review types of f ractures and the mechani sms of injury, critiq critiq ue protocols protocols for fra cture description, an d consider consider t he problems problems of interpreting t he cau ses of injury. injury. Although Although a n import a n t s ou ou r ce ce of d a t a f or or t h e s t u dy d y of t h e history of medicine, medicine, a discussio discussion n of skeleta l
140
[Vol. [V ol. 40, 1997
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
TAB LE 1. Vari ation in the cate catego gori ri zation of of traum atic inju ri es by differe different nt authors
Knowles (19 1983 83)) F r a ct u r e s D is is l oca ti t i on s Trephinat io ion n a nd amputation We a p on on w o u n ds ds E x o s t os e s S c h mo mor l ’s n od od e s Osteochondritis dissecans
Merbs (19 1989 89a) a)
Ortner and Put schar (19 1981 81))
Steinbock Steinbo ck (19 1976 76))
F r a ct u r e s D is is l oca ti t i on s S ur ur ge ge r y
F r a ct u r e s D is is l oca ti t i on s Tr ep ep h in in at a t io ion
F r a ct u r e s D is is l oca ti t i on s S ha ha rp r p I ns ns t ru ru m e n ts t s2
We a p on on w o u n ds ds S ca lp lp i n g Dental trauma 1
We a p on on w o u n ds ds S ca lp lp i n g Deformation 1 Pregnancy-related
G r o w t h a r r e s t l in in e s C r u s h i n g i n ju r i e s
Roberts Rob erts an d Manchester (19 1995 95)) F r a ct u r e s3 D is is l oca ti t i on s Osteochondritis dissecans
This study Fractures 4 D is is l oca ti t i on s
1Includes
cranial deformation, filing of teeth, and other modifications performed for aesthetic purposes. surgery a nd wea pon wounds. 3Includes piercing injuries cau cau sed by knife and sw ord cuts, scalping, an d projectile projectile points (i.e., (i.e., surgery a nd wea pon wounds). 4Includes piercing injuries injuries caused by knife a nd sword cuts, sca lping, and projectile points points (i.e., surgery a nd wea pon wounds), and crush fractures caused by foot foot binding and by cranial binding and fl att ening. 2Include Includess
indicators indic ators of surgi surgiccal prac practic ticee and other medicaa l intervent io medic ion n is beyond th e scope scope of this pa per per.. Scholars Scho lars have cat ego egorize rizedd t ra umat ic injuinjur ie ie s i n a v a r ie ie t y of w a y s (Ta b le le 1), b ut ut generally refer to both accidental and intentional tio nal tra uma, t he former former usua ll llyy including including mostt fra ctures and dislocatio mos dislocations ns a nd the lat ter usually including examples of surgical intervention interve ntion and w eapo eapon n wounds. It ma y be more prudent, however, to fi rst sort injuries according to their predominant characteristic, either f r a c t u r e1 (an y break in the co contintinuity of a bone) or di slo slocation cation (the displacement of one or or more bones at a jo joint), int), rat her than to cl assif y injuries in a manner that implies impli es ca ca usa tion or or intent . DISLOCATIONS
Tra uma tic injuries injuries to joints joints m ay result in partial or complete dislocations. A dislocation, tio n, or lux luxat at io ion, n, occurs occurs when the ar tic ticular ular surfa ce cess of a joint joint a re tota lly displa displa ce cedd from one another. Asubluxation results when the articular articu lar surf surface acess are partiall partiallyy dis displ place acedd but do reta in some conta conta ct. Although dislocadislocations and subluxa tions may be congenita congenita l or or spontaneous in origin, they are most commonly ca ca used by tra uma a nd in such such cases it is not unco uncommon mmon for t he joint joint displaceme displacement nt 1 ‘‘Infracture’’ and ‘‘infraction’’ are alternative terms for ‘‘fracture,’’ particularly undispl ture,’’ undisplaced aced fra ctures, according according t o medical dictionaries dic tionaries (e.g., (e.g., Stedman , 19 1982 82). ). Although Although ra rely used in paleopathology paleo pathology,, these terms ma y be seen in the literature w ith a different meaning: ‘‘infra ‘‘infra ction,’ ction,’’’ for example, has been defined as an inco incomplete mplete fracture. In the interests of developing developing clear clear and standard terminology, these alternative terms are not used in this review.
to be associated with a fracture. Since displaceme pl acement nt cannot occu occurr w itho ithout ut dama ge to the joi oint nt capsul capsulee a nd li ligaments, gaments, co compl mplic icaations tio ns such a s t he ossifi ossifi catio cation n of membrane, ligament, li gament, a nd tendo tendon n att achme achments nts to bone bone ma y ensue. Persistent inst a bil bility ity of the joi joint nt also may r esult, part ic icular ular ly in the shoulder shoulder and ankle, a lthou lthough gh this compli complicatio cation n cannott be easi no easily ly ide identified ntified in archaeo archaeolo logic gical al skeleta skele ta l remains. Osteoart hritis is one of the more common, and recognizable, complications tio ns a nd results from from da mage to the a rticular cartilage itself or from prolonged incongruence of the joint surfaces. For joint displacements to be recognizable in dry bone the injury must have occurred some time before the death of an individual a nd rem a ined unr educed (i.e., (i.e., not ‘‘‘‘set’’ set’’)) long enough for for bone modificat io ions ns t o ta ke pla pla ce ce.. Some dislo dislocations ations,, suc such h as of the dig digits, its, usually can be relat relat ive ively ly quickly quickly a nd ea sil silyy reduced (but see D reier, 1992) 1992),, wh ile oth oth ers, such as of the vertebra vertebra e, may cause immediat e death. In either case no evidence evidence of of the injury will be observable in archaeological skeletons. Dislocatio Dislo cations ns tend to be more frequent in young an d middle-a middle-a ged ad ults, since in subad ults a similar force instea d causes epiphyepiphyseal se al se separation paration a nd in older older a dul dults ts causes fracture of osteoporotic bones. The glenohumeral joint is a common site of dislocation, the sha ll llowness owness of the shoulder shoulder jo joint int making it particularly susceptible to displacement. Tra Tra uma tic dislocation dislocation of the femoral head from the acetabulum, in contrast, requiress consi quire nsiderabl derablee fo forc rcee and this si site te is
140
[Vol. [V ol. 40, 1997
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
TAB LE 1. Vari ation in the cate catego gori ri zation of of traum atic inju ri es by differe different nt authors
Knowles (19 1983 83)) F r a ct u r e s D is is l oca ti t i on s Trephinat io ion n a nd amputation We a p on on w o u n ds ds E x o s t os e s S c h mo mor l ’s n od od e s Osteochondritis dissecans
Merbs (19 1989 89a) a)
Ortner and Put schar (19 1981 81))
Steinbock Steinbo ck (19 1976 76))
F r a ct u r e s D is is l oca ti t i on s S ur ur ge ge r y
F r a ct u r e s D is is l oca ti t i on s Tr ep ep h in in at a t io ion
F r a ct u r e s D is is l oca ti t i on s S ha ha rp r p I ns ns t ru ru m e n ts t s2
We a p on on w o u n ds ds S ca lp lp i n g Dental trauma 1
We a p on on w o u n ds ds S ca lp lp i n g Deformation 1 Pregnancy-related
G r o w t h a r r e s t l in in e s C r u s h i n g i n ju r i e s
Roberts Rob erts an d Manchester (19 1995 95)) F r a ct u r e s3 D is is l oca ti t i on s Osteochondritis dissecans
This study Fractures 4 D is is l oca ti t i on s
1Includes
cranial deformation, filing of teeth, and other modifications performed for aesthetic purposes. surgery a nd wea pon wounds. 3Includes piercing injuries cau cau sed by knife and sw ord cuts, scalping, an d projectile projectile points (i.e., (i.e., surgery a nd wea pon wounds). 4Includes piercing injuries injuries caused by knife a nd sword cuts, sca lping, and projectile points points (i.e., surgery a nd wea pon wounds), and crush fractures caused by foot foot binding and by cranial binding and fl att ening. 2Include Includess
indicators indic ators of surgi surgiccal prac practic ticee and other medicaa l intervent io medic ion n is beyond th e scope scope of this pa per per.. Scholars Scho lars have cat ego egorize rizedd t ra umat ic injuinjur ie ie s i n a v a r ie ie t y of w a y s (Ta b le le 1), b ut ut generally refer to both accidental and intentional tio nal tra uma, t he former former usua ll llyy including including mostt fra ctures and dislocatio mos dislocations ns a nd the lat ter usually including examples of surgical intervention interve ntion and w eapo eapon n wounds. It ma y be more prudent, however, to fi rst sort injuries according to their predominant characteristic, either f r a c t u r e1 (an y break in the co contintinuity of a bone) or di slo slocation cation (the displacement of one or or more bones at a jo joint), int), rat her than to cl assif y injuries in a manner that implies impli es ca ca usa tion or or intent . DISLOCATIONS
Tra uma tic injuries injuries to joints joints m ay result in partial or complete dislocations. A dislocation, tio n, or lux luxat at io ion, n, occurs occurs when the ar tic ticular ular surfa ce cess of a joint joint a re tota lly displa displa ce cedd from one another. Asubluxation results when the articular articu lar surf surface acess are partiall partiallyy dis displ place acedd but do reta in some conta conta ct. Although dislocadislocations and subluxa tions may be congenita congenita l or or spontaneous in origin, they are most commonly ca ca used by tra uma a nd in such such cases it is not unco uncommon mmon for t he joint joint displaceme displacement nt 1 ‘‘Infracture’’ and ‘‘infraction’’ are alternative terms for ‘‘fracture,’’ particularly undispl ture,’’ undisplaced aced fra ctures, according according t o medical dictionaries dic tionaries (e.g., (e.g., Stedman , 19 1982 82). ). Although Although ra rely used in paleopathology paleo pathology,, these terms ma y be seen in the literature w ith a different meaning: ‘‘infra ‘‘infra ction,’ ction,’’’ for example, has been defined as an inco incomplete mplete fracture. In the interests of developing developing clear clear and standard terminology, these alternative terms are not used in this review.
to be associated with a fracture. Since displaceme pl acement nt cannot occu occurr w itho ithout ut dama ge to the joi oint nt capsul capsulee a nd li ligaments, gaments, co compl mplic icaations tio ns such a s t he ossifi ossifi catio cation n of membrane, ligament, li gament, a nd tendo tendon n att achme achments nts to bone bone ma y ensue. Persistent inst a bil bility ity of the joi joint nt also may r esult, part ic icular ular ly in the shoulder shoulder and ankle, a lthou lthough gh this compli complicatio cation n cannott be easi no easily ly ide identified ntified in archaeo archaeolo logic gical al skeleta skele ta l remains. Osteoart hritis is one of the more common, and recognizable, complications tio ns a nd results from from da mage to the a rticular cartilage itself or from prolonged incongruence of the joint surfaces. For joint displacements to be recognizable in dry bone the injury must have occurred some time before the death of an individual a nd rem a ined unr educed (i.e., (i.e., not ‘‘‘‘set’’ set’’)) long enough for for bone modificat io ions ns t o ta ke pla pla ce ce.. Some dislo dislocations ations,, suc such h as of the dig digits, its, usually can be relat relat ive ively ly quickly quickly a nd ea sil silyy reduced (but see D reier, 1992) 1992),, wh ile oth oth ers, such as of the vertebra vertebra e, may cause immediat e death. In either case no evidence evidence of of the injury will be observable in archaeological skeletons. Dislocatio Dislo cations ns tend to be more frequent in young an d middle-a middle-a ged ad ults, since in subad ults a similar force instea d causes epiphyepiphyseal se al se separation paration a nd in older older a dul dults ts causes fracture of osteoporotic bones. The glenohumeral joint is a common site of dislocation, the sha ll llowness owness of the shoulder shoulder jo joint int making it particularly susceptible to displacement. Tra Tra uma tic dislocation dislocation of the femoral head from the acetabulum, in contrast, requiress consi quire nsiderabl derablee fo forc rcee and this si site te is
Lovell]
141
TRAUMA ANALYSIS
TAB LE 2. Summ ary of mechani mechani sms of of in jur y and assoc associated iated types types of of fractur es
M e c h a n i s m of I n j u r y D ir ir ec ect t ra r a u ma ma
Ty p e of f r a ct u r e P en e n et et ra r a t in in g C o mm mm i n ut ut e d Tr a n s v e rs rs e C ru r u sh sh D e pr pr es es s io ion C om om pr pr es es s io ion P r es es su su re re S pi pi r a l
I nd n d ir ir ec ect t ra r a u ma ma
O bl bl iq iq u e Torus/greenstick I m pa pa ct ct ed ed B u r st st Comminutedd Comminute Av ul ul si si on on S t r es s S ec econ da da r y t o p a t ho h ol og y
C om m e n t s P a r ti t i a l or com pl pl e t e p e n et et ra r a t io ion of b on e cor te t ex B o n e i s b r ok ok en en i n m or or e t h a n t w o p ie ie ce ce s ; m os os t co mm mm on on i n l on on g bone diaphyses F or or ce ce a p p lili ed ed i n a l in in e p er er p en en d i cu cu l a r t o l on on g a x i s o f t h e b on on e M os t com mo mon i n ca nc n ce l l ou s b on e C r u sh sh in in g f or or ce ce on on e s id id e of t h e b on on e C r u sh sh in in g f or or ce ce on b ot ot h s id id es es F or or ce ce a pp p p l i e d t o g ro row in in g b on on e R ot ot a t io ion a l a n d l on gi gi t ud u d in in a l s tr t r es es s on l on g a xi x i s ; of te te n con fu fu se se d with oblique fracture R ot ot a t i on on a l a n d a n g u la la r s t r es es s on l on on g a x is is ; of t en en con f us us ed ed w i t h spiral fracture Bend ing of of the bone bone due to longitudinal compressio compression; n; common common in children B on on e e nd nd s a r e d ri r i ve ve n i nt nt o e a ch ch ot h er er F ou ou n d i n t h e s pi pi n e d ue u e t o v er er t ic ica l com pr pr es es si si on on Forcee spl Forc splits its in seve several ral direc directions tions and fo forms rms a ‘‘T’’or ‘‘Y ‘Y’’’sha pe F r a ct ct u r e d u e t o t e ns ns io ion a t l ig ig a m en en t or t e nd nd on on a t t a c h me me nt nt D u e t o r e p e t i t i v e f or ce , u s u a llll y p e r p e n d i cu l a r t o l on g a xi xi s May be confused confused with direct direct tra uma t ransverse fracture S ec econ da d a r y t o l oca li l i z ed ed or s ys ys te t e m ic ic d is is ea ea s e t ha h a t h a s w ea e a ke k e n ed ed t he he bone
more commonl commonlyy assoc associated iated with co congenital ngenital dislocations. FRACTURES
A fracture consists of an incomplete or complete break in the continuity of a bone. The most common types of fract ures, such a s transverse, spiral, oblique, and crush fractures, result from direct or indirect trauma. Two a dditional t ypes of fractur es, those resulting from stress and those secondary to pathology, are less common and have distinctt eti tinc etioolo logi gies es.. Fracture type typess and thei theirr assocciated mec asso mechanisms hanisms of inj injury ury are reviewed below below (a nd summa rized in Ta ble 2), 2), followed by discussions of fracture healing an d complic complicat at io ions. ns. 2 Mechanisms of injury and types of fractures D i r ec t t r a u m a . When a break occurs at
the po int of impact it is referred to as a direct trauma injury (Miller and Miller, 1979) and the resulting fracture may be transverse,, pe verse penetra netra ting, co comminuted, mminuted, or crush (Fig. 1). 1). A tra nsverse fra cture results from f or or ce ce a p pl pl ie ie d i n, n, a n d a p pe pe a r s a s , a l in in e perpendicular perpe ndicular t o the longitud longitud inal a xis of of the bone. bo ne. Clini Clinicall callyy, this inj injury ury of often ten resul results ts 2 The information prov provided ided here has bee been n co compile mpiledd from a variety of sources, including Adams (1987), Apley and Solomon (1992 (1 992), ), Gust ilo (1991 (1991), ), Har kess a nd R am sey (1991) (1991),, a nd S chultz (1990).
from a ha rd kick to the shin a nd is often seen seen am ong soccer soccer pla yers. Typic Typicaa lly lly,, t ra nsverse fractures are caused by a relatively small forcee delivered forc delivered to a s ma ll area . Pa rtial or compl mplete ete penetrat penetrat io ion n of the bone bo ne cortex by cutting, pie piercing, rcing, drilling, or scraping, such as the excision of pieces of crania l vault bones in the practice of of trephina tion, or or the a mputa tion of of a limb segment is classed classed as a direc directt tr aum a injury. injury. Penetra ting fractur es typica typica lly are caused by applica applica t i on on of a l a r g e fo f or ce ce t o a s m a ll ll a r ea ea . I n archa eo eolo logic gical al co contexts, ntexts, pe penetra netra tio tion n co could uld be caused by a proj projec ectile tile point, point, t he blad e of an axe or sword, or a musket ball (Blair, 1983; Butler, 1971). Wounds from arrow or spear points often can be identified with certainty only if the point remains embedded in th e bone and healing wo uld not be evident if such wounds were linked to the deat h of the individua l. The The huma n rema ins in ma ny histo historic ric ceme cemeterie teriess show the t raumatic results results of conflic nflicts ts with bul bulle lets ts and other proj projec ectiles tiles (e.g., (e.g., G ill ill,, 1994; 1994; L a rsen et a l., 1996; 1996; Ow sley et a l., 1991) 1991).. Ea rly ca ses of penetrat pe netrat ing pro projjec ectile tile w ou ounds nds can be expected pec ted to sh ow subsequent infec infection tion a nd/or pronounced prono unced deformity deformity in th e absence of stabili bi lizat zat io ion n or rest of the injured injured part. Some penetrating fractures may also be comminuted, which occurs when the bone is broken in more than two pieces. In clinical cases,
142
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
[Vol. 40, 1997
Fig. 1. Fractures caused by direct trauma . From left to right: tran sverse, penetrat ing, comminuted, and crush.
high velocity bullets a nd blunt force tra uma to the cranium typically cause comminuted fractures. Crush fractures most commonly occur in cancellous bone a nd r esult from th e application of a direct force to the bone, whi ch collapses on itself. Three types of crush fractures are recognized: depression, compression, and pressure. The first refers to crushing on one side of the bone (especia lly common on t he ectocra nium) while th e second refers to a crushing force tha t originat es on both sides of the bone. The incomplete penetration of a bone by a low velocity projectile may result in a crush fracture, such as depressed fractures caused by the impact of musket balls (Liston and Baker, 1996) or shotgun pellets (Swan and Swan, 1989). Blunt trauma, such as that produced by a bludgeon, fi st, or ha mmer, or when a n object is dropped on the hand or foot, results in crush fractures. The third type of crushing injury results when developing bone responds to the application of direct force. Examples of thi s last type are culturally ma nda ted bone altera tions, such as th e sha ping of immature cranial and foot bones by var ious types of binding for bea utifi cation. Rarely, direct trauma injury that bruises a joint may fra cture ar ticular cart ilage, and
sometimes the subchondral bone as well, causing separation of a fragment from the margin of the articular surface. This resulting lesion may be confused with osteochondritis dissecans, which is caused by a septic necrosis an d is usua lly seen a s th e complete or incomplete sepa ra tion of a portion of joint cartilage a nd subchondral bone, most commonly on the femoral condyles. Osteochondritis dissecans is usually recognized in dry bone a s a pit, often 2 to 5 mm in diam eter, in the subchondral bone, although new bone formation may partially or completely fill the defect or may produce a deposit t hat exceeds the level of the normal articular surface. The etiology of osteochondritis disseca n s i s u n ce rt a i n b ut i nd ir ect t r a u m a i s thought t o play a t least a contr ibutory role. When a fractur e occurs in a pla ce other th an the point of impact it is said to result from indirect trauma (Miller and Miller, 1972). Oblique, spira l, greenstick, impacted, burst, and avulsion fractures are consequences of indirect trauma (Fig. 2). An oblique fra cture, w here t he line an gles across the longitudinal a xis, is indica tive of a combined a ngula ted/rota ted force (Ha rkess a nd Ra msey, 1991). If th e fractur e is well healed, this break is easily confused I n d i r ec t t r a u m a .
Lovell]
TRAUMA ANALYSIS
143
Fig. 2. Fractures caused by indirect trauma . From left to right: oblique, spiral, greenstick due to an gular force, greenstick due to compression, impaction, a nd a vulsion.
w i t h a s pi ra l l in e. A s pi ra l f r a ct u r e l in e winds down a round a long bone shaft due to a rota tiona l and downwa rd loading stress on the longitudinal axis. In some cases, such a force a pplied to the tibia result s in a fra cturedislocat ion of the ankle rat her tha n a spiral fracture (Harkess and Ramsey, 1991); at other times, the force results in an associat ed proximal fi bular fracture. Torus or gr eenstick fractur es result from bending or buckling of bone when stress is applied. These often are due to indirect t r a u m a a n d a r e m os t com m on ly s ee n i n children, whose bones are still pliable and hence less likely t o break, inst ead producing a localized bulging on the bone. An example is a greenstick fracture of the clavicle that results during childbirth when the child’s biacromial breadt h is too lar ge to pass ea sily through the mother’s pelvic outlet. Greenstick fra ctures a re also chara cterized by a n incomplete fracture involving only the convex side of a bone tha t ha s been subjected t o bending stress. In adults the ribs are commonly a ffected. Less common fractures resulting from indirect trauma are impacted, avulsion, a nd burst fr a ctures. An impacted fra cture occurs when the bone ends at a fracture site are driven into ea ch other by the force of injury.
Clinica lly, this is often seen in th e proximal h um er u s a s t h e r e su lt of a f a ll on t o a n outstretched hand, and in the metacarpals a s a r es ult of t r a um a t o t h e fi s t w h en punching. An avulsion fracture is caused when a joint capsule, ligament, or tendon is stra ined an d pulls awa y from its atta chment to the bone, tear ing a piece of bone with it. A particular type of avulsion fracture leaves a transverse fracture line: a transverse fracture ma y occur to the ulnar olecra non process or the patella if the extensor muscles contract forcefully while the joint is flexed, and in extreme cases the bone fragments will separ at e and ma y heal without uniting. A burst fracture is located in the spine. It results from a vertical compression that ruptures the intervertebral disc through the vertebra l end plate, forcing disc tissue into the vertebral body (Fig. 3). A mild form of this injury is often seen in archaeological specimens as a small, localized, typically circular, depression in the end plate that is usua lly called a ‘‘Schm orl’s node.’’ Comminuted fra ctures may be due to indirect tra uma a s well as t o direct tra uma. The indirect comminuted fracture is patterned like a ‘‘T’’ or ‘‘Y,’’ and is produced by a force tha t passes through th e bone, splitting it in several directions (P erkins, 1958). Crush
144
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
[Vol. 40, 1997
trauma transverse injuries. Commonly a stress fracture will be visible as a nondisplaced line or crack in the bone, called a hairline fracture, which is not detectable ra diologically until a bony callus ha s formed over th e break. F r a c t u r es sec o n d a r y t o p a t h o l o g y. Frac-
tures often occur secondarily to a disease already present in the body. Systemic diseases such as metabolic disturbances a nd nutritional deficiencies leave bone vulnerable to spontaneous fracture or to fracture from minor tr a uma . For exa mple, postmenopausal females ma y suffer fra ctures if their bones h a ve been wea kened by osteoporosis. Other skeletal markers of specific disease may aid in a ttributing cause to the fra cture: neoplast ic fra ctures are seen when the break is through or a djacent t o a t umor th at is in, or of, bone, and the c ollapse of vertebral bodies is not an uncommon consequence of tuberculosis in the spine (Pott’s disease). Fracture healing D u r a t i o n o f h ea l i n g . Fractures begin to Fig. 3. Burst fra ctures of the lumbar vertebrae in a young a dult ma le from the historic Fur Tra de period in Alberta.
fractures al so can be found as a result of indirect trauma, suc h as in th e calcaneus aft er a person ha s jumped from a height. Repetitive force ca n result in a str ess or fat igue fracture. The usual ar eas of occurrence are the meta ta rsa l, calca neus, a nd tibia (Wilson and Ka tz, 1969). St ress fra ctures in the meta ta rsa ls ar e sometimes referred to as ‘‘marching’’ fractures, since they are often diagnosed in military cadets. Those in the tibia have been known for some time to affect dancers, while the increased interest in jogging and aerobic dancing in recent decades as well as the adoption of alternative religious practices has led to their higher prevalence in other segments of western society (Burrows, 1956; Cohen et al., 1974). The fracture line is usually perpendicular to t he longitudinal axis, therefore problems may arise in trying to distinguish between stress and direct S t r e ss f r a c t u r es .
heal immediately after the bone is broken, but the process differs for cancellous and tubular bone. Most investigators identify fi ve overlapping sta ges in tubular bone hea ling (Adams, 1987; Apley and Solomon, 1992; Paton, 1984). Table 3 summarizes the activities of these stages and the approximate t i me a f t er i nju r y t h a t e a ch i s ob se rv ed . Healing normally is not visible on radiographs until approxi matel y 2 to 3 weeks after the injury, w hen a callus of woven bone, the result of cell proliferation from the periosteum, marrow cavity, and surrounding connective tissue, appears around the site of injury. The callus internally and externally bridges the gap caused by the fracture and stabilizes the fractured ends. Consolidation of this woven bone into matur e lamellar bone occurs subsequently, but the dur a tion of the process depends upon th e nature of the fracture and the type of bone involved. In a phalanx, a solidly united fracture ma y develop in less tha n 1 month, while the same tra nsforma tion may t ake up to 6 month s in a tibia or femur. Bones of the
Lovell]
145
TRAUMA ANALYSIS
T A B L E 3 . T h e p r o ces s a n d d u r a t i o n of f r a c t u r e h ea l i n g i n t u b u l a r b on es (Adams, 1987, Apley and Solomon, 1993, and Paton, 1984)
H ea lin g st a ge
H ea lin g processes
H a e m a t o m a f or m a t i on
B l oo d f r om t o rn v es s el s s e ep s ou t a n d f or m s a h a em a t o m a Fra ctured bone ends die due t o lack of blood supply Osteoid is deposited around each fragment by osteoblasts of periosteum a nd endosteum and pushes ha ematoma a side Fra cture is bridged; visible in dry bone Callus of woven bone forms from mineralization of osteoid and acts as a splint for periosteal and endosteal surfaces Visible radiologically M a t u r e l a m e ll a r b on e f or m s f r om ca l l u s p r ecu r s or a n d r e s ul t s in a solidly united fracture area
Cellular proliferation Callus formation C o n so li d a t i on R e mo de ll in g
G r a d u a l r e m od e ll in g o f b on e t o i t s o ri g in a l f or m , s t r e n gt h ening along lines of mecha nical stress Increased density on radiographs marks the fracture site on adult bones
upper limb tend to heal faster than do those of the lower limbs, and spiral and oblique f r a ct u r es h ea l f a s t er t h a n d o t r a n s ve rs e fractures. In contra st t o compact tubular bone, cancellous bone ha s a meshlike structure with no medullary canal. This provides a much larger area of contact between fracture fragments, which facilitates healing. In addit i on , t h is m es h ca n b e m o r e e a s il y p en etrated by bone-forming t issue than can compact bone, so the union occurs directly between bone fragments instead of indirectly via the periosteal and endosteal callus. The initial haematoma is penetrated by proliferating bone cells which grow from opposing fracture surfaces. The developing t i ss u es f us e w h en t h ey m ee t a n d s u bs equent ly calcify t o form w oven bone. Healing is thus simpler an d fa ster in can cellous bone tha n in compact bone. Because there is a delay before healing is visible m acroscopica lly or ra diographically, it may be difficult to distinguish some postmortem breaks from unhealed premortem fractures. Perimortem fra ctures is th e term g iv en t o s u ch i nju r ie s, w h i ch m a y h a v e occurred in the recent antemortem period (i.e., up to 3 weeks before death) an d are therefore unhealed, or that alternatively may have occurred in a postmortem period that is of indeterminate length (perhaps weeks or month s) but dur ing wh ich th e bone is still relatively fresh and its organic components not yet deteriorat ed. Otherwise, distinguishing between a ntemortem/perimort em t r a u m a a n d t h a t w h i ch cl ea r l y occu r r ed
D ur a t ion 2 4 h o ur s 3 weeks 3 to 9 weeks Var ies by skeleta l element from a few weeks to a few months 6 to 9 years
after death is predicated upon the different fra cture properties associat ed with bone tha t retains its viscoelastic nat ure and bone tha t does not, and upon the different appearances of bone surfaces after various postmortem interva ls (B uikstra a nd U bela ker, 1994; Maples, 1986; Mann and Murphy, 1990; Ubelaker and Adams, 1995). Antemortem or perimortem fra ctures can be identified by 1) any evidence of healing or inflammation; 2) the un iform presence of sta ins from w at er, soil, or vegetation on broken and adjacent bone su rfa ces; 3) the presence of gr eenstick fractures, incomplete fractures, spiral fractures, a nd depressed or compressed fra ctures; 4) oblique angles on fracture edges; an d/or 5) a pat tern of concentric circula r, radiating, or stellate fracture lines. Postmortem fractures, in co ntrast , tend to be characterized by 1) smaller fragments; 2 ) nonuniform colorat ion of the fracture ends and the adjacent bone surface, especially light-colored edges; 3) squa red fracture edges; and 4) absence of fracture patterning due to th e increased tendency of dry, brittle, bone to shat ter on impa ct. C o m p l i c a t i o n s o f h e a l i n g . Complica-
tions should be assessed when examining fra ctures because they ma y provide informa tion regarding mobility, morbidity, mortalit y, a n d m ed ica l t r ea t m e nt or t h e l a ck thereof. In additi on to the fracture types described a bove, the relat ionship of the fra cture to surrounding tissue is referred to as ‘‘closed’’ or ‘‘open.’’ When the fractured bone does not come into contact with the outer
146
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
surface of the skin, the fracture is termed closed. An open fracture, also known as a compound fracture, is when the bone protrud es through th e skin or th e skin is broken to the level of th e bone, as in a crushing or penetra ting w ound. Open fractures ar e prone to infection, which hinders the union of the fracture and creates instability. A pathogen, Staphylococcus aureus in a bout 90%of clinical cases (Ortner an d P utschar, 1981), ma y be introduced to t he body thr ough an open fra cture from surface conta mina tion or from a penetrating instrument or contamin ant. Although th ere is a tendency to rega rd localized infections as related to observed fractures, but to interpret nonlocalized infections a s unrelated to fracture, posttrauma tic infection ma y in fa ct be present either a s a localized condition or, due to h emat ogenous disseminat ion of the pat hogen, a s a syst emic infection. Whether localized or systemic, if the body’s immune sy stem is u na ble to combat the infection successfully bony response is usua lly visible in the form of periostitis (an infla mma tion of the periosteum) or osteomyelitis (a more severe bone infection that involves the medullar y cavity ). P eriostitis is usua lly cha ra cterized by foca l periosteal bone d ep os it i on t h a t m a y e ve nt u a l ly f or m a plaq uelike sheet over th e cortex. Osteomyelitis is identifi ed by a t hickened contour in th e area of the fra cture and the bone may feel hea vier. P a th ognomonic evidence of osteomyelitis results from the development of subperiostea l abscesses tha t deprive the bone of its blood supply and lead to necrosis (the dead bone forms a sequestr um). The periosteum continues to produce new, hypervascular bone a round the sequestrum, forming a shell of bone called involucrum . The su bperiosteal pus must escape t hrough the involucrum to the skin surface, however, and in doing so forms one or more sinu ses (cloaca e) in the involucrum for pus drainage. In dry bone the sequestrum, lying under the involucrum, ma y be visible thr ough a cloacal opening. Postt ra uma tic osteomyelitis is most commonly observed in the cranium and long bones of archaeological skeletons. Fra ctures inevita bly result in the rupture of minor blood vessels but this is not usually a serious complica tion. In some ca ses, however, bone displacement can compress or
[Vol. 40, 1997
twist blood vessels a nd lead to ischemia. This will delay the healing process and could lead to bone death if unrelieved. Avascular necrosis n orma lly occurs near the ar ticular ends of bones where the blood supply to subchondral bone is limited. Death of the t i ss u e b eg in s a w e ek a f t er t h e n u t r ie nt supply is reduced and may continue for up to 4 years. D uring this t ime the bone loses its trabecular structure, becomes granular, and begins to disintegra te due to muscle stress or body w eight. The adja cent a rticular ca rtilage a lso dies as a r esult of deficient nourishment, usually resulting in osteoarthritis. Nerve injuries also may be associated with fra ctures. Three ty pes of nerve injuries are generally recognized. Damage is slight in neurapraxia and results in t emporary impairment that corrects itself within a few weeks. In contra st, the interna l nerve architecture is preserved but axons are badly damaged in axonotmesis, resulting in peripher a l d eg en er a t ion t h a t m a y t a k e m a n y months to heal. Such a lesion may result from pinching, crushing, or prolonged pressure. The most serious type of n erve injury, neurotmesis, involves complete division of a nerve, either through severing or severe scarring, and requires surgical repair. The conseq uences of these ty pes of nerv e injuries range from loss of sensation to loss of function. Usually the loss is t emporary, but muscle atrophy may result a nd if the nerve loss is prolonged or permanent the bones will display signs of disuse atrophy as well. This sequel would be most likely in archaeological cases of neurotmesis. In addition, if there is loss of innervation to the fracture site, the individual wil l not feel pain and may therefore continue to use the broken bone, impairing healing. Fracture of the vertebral column may result in dama ge to the spinal cord or spina l nerves, wit h para lysis below the level of the injury a possible outcome. Depressed skull fractures w ith endocranial displacement are usually associated with signifi cant brain injury, which must also be considered in cases of linear fractures of the cra nial vault. Another complica tion is posttra uma tic ossifi cation of a haematoma, w hich results when absorption of the haematoma is prevented by excessive stress placed on the
Lovell]
TRAUMA ANALYSIS
periosteum. Asmooth mass of bone is macroscopica lly visible aft er 2 months, w ith ca lcifi cation being visible ra diologically a few weeks after the injury. Although usually benign, movement ma y be restricted if there is joint in volvement. If joint function is affected by traumatic injury, osteoarthritis may develop as a complicat ion. S tiffness caused by fi brous a dhesions or joint swelling may lead to prolonged disuse of the joint or limb. Shortening or angulation may result in some loss of normal function in the affected limb or in the joints directly a bove and below the fra cture; this may be difficult to interpret since unu s ua l b iom ech a n i ca l s t r es s a t a joi nt i n which a fractured bone part icipates may cause osteoarthritis, but it is also possible for a jo int on an uninjured limb to be affected. The la tt er might occur, for exa mple, if weight bearing was shifted in order to favor the injured leg. Premat ure deterioration of artic ular c artilage and subsequent deterioration of subchondral bone are common complications of breaks affecting the joint sur face itself, since cart ilage repair is a very slow process. Such fractures also can result in a nkylosis of th e joint. Three final complications of fractures are delayed union, nonunion, a nd malunion. In clinical settings the union of a fracture is defined as delayed if it has not occurred in the time expected for that skeletal element, age, and sex of the i ndividual, and it may eventua lly be classed as a nonunion. In dry bone specimens, of cours e, delay ed but eventually successful union cannot be distinguished from undelayed union. Several factors may impede the process of healing, but overa ll poor healt h a nd/or nut rition in a ncient populations may be a largely unrecogn i ze d con t r i b ut o r (G r a u e r a n d R ob er t s , 1996). The dia gnosis of nonun ion is a pplied wh en the fracture fragments fail to unite and the mar row cavity seals. Ra diologically, nonunion may be identified by sclerosis at the bone ends. After a prolonged period of time, the fragments take on a ro unded appearance at their ends, which are connected by fi brous t issue. Nonunion may result from inadequate bone healing due to infection, inadequate blood supply, insufficiency of vi-
147
tamin D or C or of calcium, excessive movement between bone fra gments during hea ling, soft tissue being caught bet ween the fragment ends, ina dequate contact between the fragments, presence of foreign material, or from the dest ruction of bone due to pat hology or the injury itself (Altner et al., 1975; Ka rlstrom a nd Olerud, 1974; Sevitt, 1981; Stewart, 1974; Urist et al., 1954; Yamigashi an d Yoshimura , 1955). If there is persistent movement between the ununited ends, a pseudarthrosis, or false joint, may form, although this complication is relatively rare (Stewa rt, 1974). S tudies of modern huma n populations indicate a frequency of pseuda rthrosis of less than 5%(Heppenstall, 1980; Urist et al., 1954), while an examination of dat a from tempora lly, geographically, an d cultural diverse archaeological populations reveals an average frequency of 2%(Burrell et al., 1986; J imenez, 1994; Lovejoy et al., 1981; Stewart, 1974). Among alloprimates, da ta from B ra mblett (1967), J urm a in (1989), Lovell (1990), a nd Schu ltz (1937, 1939) a lso give an average pseudarthrosis frequency of a pproxima tely 2%. A malunion consists of a fracture that heals leaving a deformity. This may occur when a fracture h as n ot been reduced or when reduction was not maintained, leaving the fragments to heal grossly angulated or excessively shortened. S hortening is caused by overlap, substantial angulation, crushing, or gross bone loss. Injuries to growing bone th a t affect the epiphyses a nd lead to premature fusion of t he growth plate may result in shortening, as may bone infarction resulting from sickle cell disease, which most often affects the epiphyses of the growing skeleton, especially those of the proxima l femur. The presence of a sh ortened bone is most detrimental to the lower, weigh tbearing limbs, a lthough a difference of up to 20 mm is consider ed by clinica l pra ctitioners to be tolerable. A greater loss in length can lead to backache from pelvis tilting and lateral and rotational spinal deviation. Recent st udies have interpret ed minimal deformity in bones tha t ar e likely to be severely affected when fractured as evidence for immobilization of the injured par t a nd possible m ed ica l t r ea t m e n t (G r a u er a n d R ob er t s , 1996).
148
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
DESCRIPTIVE PROTOCOLS FOR FRACTURES
Proper description of an injury is the first step in tra uma an a lysis (Ortner and Putschar, 1981; Steinbock, 1976) and is the basis for determining the mechanism, or proximate cause, of the injury. In t urn, a n understa nding of the proximate cause is crucial for the i de nt i fi ca t i on of t h e u lt i ma t e ca u s e of trauma, usually behavior. Proper description of observed lesions a lso provides other scholars wi th an o pportunity to agree or disagree w ith the diagnosis and /or inferences tha t a re ma de about the sociocultura l or environmental context of the injury. Although severa l models proposed recently have made great strides in standardizing descriptive protocols (e.g., Buikstra and U belaker, 1994; Da stug ue and G erva is, 1992; Grauer and Roberts, 1996; Roberts, 1991), paleopathologists have not yet reached a con s e n s us o n d e s cr i pt i v e s t a n d a r d s f or trauma and many are not always familiar with the underlying mecha nisms of injury. Ideally, a ny method of fra cture description will recognize two main sources of confusion in interpretation: the variation in appearance expressed by fractures caused by the same mechanism of injury, as wel l as t he similarities in a ppea ra nce displayed by fractures caused by different mechanisms of injury. Ultimately, proper fracture description should seek to improve the accuracy and reliability of interpretation without exceeding t he limits of inference tha t ar e set by the descriptive dat a t hemselves. Although fracture types a re here subsumed under their proximate cause, when describing a nd interpreting injury th e fra cture type is usua lly recognized first . Identifi cation of the mechanism of injury then follows logically, a nd t he third step in tra uma ana lysis involves interpreta tion of the ultimate cause of the injury. For example, an impacted fracture of the distal radius with posterior displacement of the distal fragment ma y be recognized as a Colles’fra cture due to its chara cteristic locat ion an d deformity. The proxima te cau se of the injury m a y then be identified a s indirect tra uma. Int erpreting the ultima te cause may be difficult, but a fall onto the outstretched ha nd w ould
[Vol. 40, 1997
be a logical conclusion. If the fracture w as observed in an older female, the possibility of the fra cture occurring seconda ry t o osteoporosis could a lso be considered. The principal aim of most protocols has been to establish standardized descriptions for fractur es observed in dry bone, a lthough ad ditional objectives, such a s the evalua tion of evidence for treatment of traumatic injuries are sometimes a lso stated (e.g., Gra uer an d Roberts , 1996; Roberts, 1991). Three recently developed protocols are outlined here. The repatriation of Native American prehistoric and historic skeletal remains drove the development of sta nda rds for da ta collection tha t includes procedures for documenting fra ctures (B uikstra a nd U belaker, 1994). These procedures recognize n ine t ypes of fra ctures and eight varieties of shape cha ra cteristics. All types and varieties are not mutually exclusive, but may have restricted application. Sha pe chara cteristics, for example, describe lesions caused by blunt or sharp force and by projectiles, as well as radiating fra ctures and amputations. P erimortem fractures are identi fied at a third level of description, followed by sequelae such as healing sta tus a nd var ious complica tions. Dislocations are classed separately. The recommended data collection forms and descriptive protocol do not provide for ma lunion as a component of fracture description specifi cally, but ra ther under t he pa thology category of ‘‘a bnormality of shape,’’ in w hich ma lunion w ould be identifi ed as either ba rely discernable or clear ly discerna ble a ngula tion. Asecond method wa s designed specifica lly to describe fractures in a way that would provide the information necessary to examine the t echnology and knowledge of tr eatments in past societies (G ra uer an d Roberts, 1996; Robert s, 1991). The meth od d escribes the location an d type of fractur e and empha sizes evaluation of the success of long bone h e a l in g . M a c r os cop ic a n d r a d i o g r a p h ic mean s a re employed to a ssess complica tions of shortening and deformity, and sequelae such as infection and osteoarthritis. Skull fractures are described as resulting from blunt or sh arp force and are ev al uated in terms of healing as well as evidence for t r e pa n a t i on . Th e n e ed f or r a d i o gr a p h i c
Lovell]
TRAUMA ANALYSIS
149
Fig. 4. Common fractures of the cranial vault. From left to right: simple linear fracture due to blunt trauma, comminuted depressed fracture due to blunt trauma, and comminuted penetrating fracture from a high velocity projectile.
evalua tion of fra ctures in order t o determine the a mount of healing a nd th e particular s of deformity a nd/or displacement is str essed, a s i s t h e i mp or t a n ce of r a d i og r a ph y f or detecting and interpreting well-remodelled fractures (Grauer and Roberts, 1996; Roberts, 1991). Un fortuna tely, ra diogra phic equipment is not alwa ys available, especially in fi eld settings, and the i nterpretation of ra diographs ma y be mad e difficult by postmortem a ltera tions common in ar chaeologica l contexts, such a s soil inclusions t ha t affect density or the differential identification of osteoporosis versus diagenetic bone loss (Roberts, 1991). Finally, a third system concerns cranial vault injuries, categorizing them as piercings, depressions, gashes, cuts, and slices (Filer, 1992). The first category is described as consistent with a penetra ting injury, the second with blunt force trauma, and the last thr ee a s r esulting from edged/bladed implements, including sha rp projectiles. The ma jority of these lesions were interpreted as resulting from interpersonal violence, an assessment not inconsistent with the apparent culture-historical context of the remains. It is likely t hat no one system of fracture description will suit all investigators, since some will be more or less concerned w ith t he affected body part, specific complications, or possible causat ive behaviors. Most protocols, however, sha re similar basic cat egories of description. The method for fracture description tha t is presented below incorporates these categories in a system adapted from clinical an d forensic medicine. I t is predicated on identification of the skeletal element(s)involved and the type of injury, as
well as detailed descriptions of deformation and of any a ssociat ed nontrauma tic lesions that may indicate causality or postinjury complications. The informa tion thus obta ined t hen serves a s a basis for inferences about t he mechanism of injury, which can in turn provide clues as to the social, cultural, or environmental associations of the injury. The method outlines descriptive features for cranial a nd long bone fractures since these predomina te in th e paleopat hological literature. Description of cranial fractures
The interpretation of the mechanism of injury of cra nia l fractures relies on a va riety of char acterist ics of the fra cture, such as t he bones involved, pat terning of fracture lines, a n d p r es en ce of d e for m a t i on (G u r d j ia n , 1975; G ustilo, 1991; H ooper, 1969; for a comprehensive discussion of lesions of the calvar ium, see Ka ufma n et a l., 1997). Str ess fractures an d fra ctures seconda ry t o pat hology are uncommon in the cranium. The most common fractures of the cranium affect the v a ult a n d a r e ca u s ed b y d ir ect t r a u ma . These can be described according to their basic type, usually linear, crush, or penetra ting (Fig. 4), wh ich a re not necessarily mu tually exclusive. Although vault fractures are most common, the base, ma xillae, nasa l bones, orbits, a nd/or zygoma e ma y be fr actured alternatively or additionally, and the temporomandibular joint may be tra umat ically disloca ted. Low velocity, bl un t trauma to the head may result in simple linear fractures or depressed (crush) fra ctures. The kinetics involved ma y rela te to a ccelera tion injuries,
150
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
in which the hea d is struck by an object an d set in motion, or deceleration injuries, in wh ich th e moving hea d suddenly comes to a halt. In either case, the curve of the skull at the point of impact tends t o fla tten out, a nd as a result the force of the impact is distributed over a relatively la rge a rea. The bone surrounding the area of impact bends outwa rd, an d, if the deformity of the cran ium is great enough, fracture lines begin, usually in the areas subjected to bending outward. The areas of bending a re not uniformly circular, since the degree and direction to which the fracture lines extend depends upon both the magnitude of the applied force an d the local bony a rchitecture. Penetrating injuries of the cranium are characterized by a small area of impact with a localized area of distortion and a re usually caused by sharp-edged objects or projectiles. With higher velocity impact, the inbending of the skull remains localized but the depth of penetration increased. As a general rule, w h en t h e a r e a of im pa c t d ecr ea s es t h e stresses are more localized but greater in magnitude and the stresses in surrounding areas diminish. The severity of impact in direct cran ial tra uma is usually determined from the extent and separation of lin ear fra ctures, by th e extent of comminution of a localized fractur e, or by the displa cement of bone fragments in penetra ting w ounds. Indirect trauma injuries are relatively rare, but may result from vertical loading forces transmitted from the feet or buttocks when a person falls from a height. A basilar ‘‘ring’’fr actur e around the fora men magn um is an example of such an injury; it reflects impact forces transmitted up through the cervical spine an d occipita l condyles. B as ilar fractures through the petrous bones a nd fractures of the mandibular condyles have been observed to result from impact to t he chin (Ha rvey a nd J ones, 1980). Description of long bone fractures
In contra st to fractures of fla t a nd irregular bones, fractures of appendicular long bones (and, by extension, short bones) often require more comprehensive description since their positions in the skeleton and their functions ma ke them more susceptible to a variety of forces. Long bone fractures
[Vol. 40, 1997
from archaeological contexts can be described in a ma nner ada pted from tha t used in clinical orthopedics (e.g., Gustilo, 1991; Harkess and Ramsey, 1991; Schultz, 1990) and can be first classified as intraarticular (involving a joint, including the metaphyseal region) or extraarticular. Intraarticular fra ctures a re described as either linea r, comminuted, or impacted. Extraarticular fractures ar e described a s linear, comminuted, or segmental. Linear fractures fall into thr ee subtypes, tra nsverse, oblique, and spira l, all of w hich have been previously described. Comminuted fractures are categorized according to th e size of the fr a gment s (multiple or ‘‘butt erfly’’) and the percentage of th e sha ft (50% or 50%) tha t is involved. A butt erfl y fra cture is formed from a combination of compression a nd t ension stresses th at result in the separation of a triangular fragment of bone. Segmental fra ctures a re identified by the multiple fracture lines that divide the b on e i nt o a t l ea s t t w o s eg m en t s a l on g a longitudinal axis. The location of the fracture should be noted as occurring at the proximal end, dista l end, or sha ft (either th e proximal, middle, or dista l third of the sha ft or one of the junctions t hereof). The final components of long bone fractur e description a re length , apposition (shift), rota tion, a nd an gulat ion (a lignment), identified by the acronym, LARA. Convention decrees tha t w hen describing the four components t he distal f ragment is measured in relat ion to th e proximal fra gment. The principal a ims here a re to describe fra ctures so that the mec hanism of injury can be deduced, and to distinguish fractures with no or slight deformity from those with marked deformity. Length of t he bone is measured with an osteometric board and the maximum length is recorded (per Bass, 1987). Length is recorded a s n orma l, distracted, or shortened, an d is determined by compar ing the injured bone to its counterpart, if possible. Distract i on i s a l en g t h en in g of t h e b on e a n d i s caused by the separation of bone fragments, often due to muscular forces. Bones themselves ma y distract a fracture, however, such as when an intact ulna pulls apart the fragment ends of a fra ctured radius or wh en
Lovell]
TRAUMA ANALYSIS
a fra ctured tibia is associated with a n inta ct fi bula. Distraction a lso may be caused w hen tissue is caught betw een fra gment ends. In contra st, shortening results w hen muscular forces pull the fragments over each other. This typically occurs when broken bones ha ve not been set, often d ue to severe pain or muscle spasm, or if a fracture reduction failed because of inst ability. Apposition is the percentage of bony conta ct betw een fragment ends in fresh injuries an d is mea sured on ra diographs. Apposition from an x-ray is measured using a ruler and is expressed as a percentage, the horizontal displacement being a function of the sur face a rea of bone. Ther efore, if th ere is no horizonta l displacement betw een the fra ctured bone ends when hea led, tha t is, the bone ends are in perfect alignment, the bone is 100% apposed. In dry bone, however, shifting of the distal fragment in relation to the proximal end can be recorded in the a bsence of radiographs. If t he bone is viewed in a nat omical position, a medial or l ateral shift may be seen; if viewed in a lat era l position, ant erior or posterior displacement may be observed. The shift in both the an teroposterior (AP) and lateral planes should be noted, as the bone can be displaced in both directions. Rotation occurs when the dista l fragment has turned relative to the proximal fragment. There is no measurement, but the dista l portion is recorded a s being interna lly or externally rotated. This is usually easily identifia ble in dry bone, especially if the affected bone can be compar ed to the contr a latera l element. If rotation is observed, the adjacent joint surfaces should be examined since rotation may result in osteoarthritis, or in ankylosis of a joint if ligaments were torn in the injury. A n g u l a t i o n a t t h e f r a c t u r e s i t e i s m e a sured in degrees with a goniometer. This measurement is easily obtained from a ra diograph but also may be obtained from th e bone. One end of the goniometer is pla ced on the midline of the proximal fragment’s longitudina l axis, the other end on the a xis of the distal fra gment with the center of the goniometer directly over the fracture site. The number of degrees the distal fragment has displaced in relation to the midline of the proximal fragment is the angulation. The
151
direction of movement must also be noted. In the AP view, t he di stal portion of the distal fragment will move medially (varus) or laterally (valgus). In the lateral view, ant erior a ngulation refers t o the distal portion of the distal fragment moving anteriorly so tha t the fra cture site a ppea rs posteriorly bowed. Posterior angulation refers to t h e d is t a l p or t ion of t h e d is t a l f r a g me nt moving posteriorly; the fra cture site appears anteriorly bowed. Degree and direction of angulation should be measured in the AP a n d l a t er a l p os it i on s a s b ot h p la n e s a r e often affected. Examples of long bone fractures
The value and a pplica tion of sta nda rdized fra cture descriptions is illustr a ted here with the description of radiographs from four clinical cases at the University of Alberta Hospital in Edmonton. With known mechanisms of injury, treatment, and follow-up, these cases unambiguously illustrate the skeletal effects of trauma and their variability of expression. For each set of radiographs, fra ctures were noted for t heir type, the bone(s) involved, a rea of involvement, degree of healing, length, apposition, rotation, and angulation. Apposition data are reported t o 5%. The mea sur ement of an gulation was found to be the most problematic and consequently a ll an gulation dat a are presented t o 2° . 3 The sex a nd age of each patient were recorded although other personal, identifying information was not revealed. Although the degree of deformity is s om et i m es u s ed t o a s s es s t h e e xi st e n ce and/or qua lity of medical treat ment in th e past, these examples reinforce the observation that the association is not always direct (Gr auer an d Roberts, 1996). F ra cture injuries often improve with t ime but conversely t h ey m a y d et e r ior a t e a n d i nd iv id u a l r esponses to fra cture var y w idely. A direct trauma transverse fracture resulted when a 20-year-old male was kicked in the shin while playing soccer. Figure 5a is a lateral view radiograph t aken immedi3 In order to evaluate interobserver error all clinical cases were independently scored by N. Lovell and C. Prins. The error in measured length as 1 mm; in apposition 5%; and in a ngula tion 2°; all adequate for distinguishing between none, slight, and marked deformity.
152
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
[Vol. 40, 1997
F i g . 5 . Tr a n s v er s e f r a c t u re d u e t o d i r ec t t r a u m a . a: Radiograph taken immediately postinjury, showing a tran sverse midshaft fracture without fi bular involvement. b: Radiograph taken more than 5 months later, showing callus around the fracture site.
ately postinjury that shows a transverse, midshaft tibial fracture wi th no fi bular involvement. The bone was perfectly aligned on both the i nitial x-rays and those ta ken more than 5 mon ths l at er (Fig. 5b). T he degree of callus formation may appear to those who are inexperienced with clinical cases to be excessive, given the apparent lack of angular deformity, displacement, or comminution, but t his example is fairly ty pical of such injury. Not all transverse fractures h eal a s nicely, however, an d nonunion, despite good alignment, may often occur in the t ibia a nd/or fi bula du e to their inherent i n s t a b i li t y w h e n s u p por t i n g t h e b od y ’s weigh t in locomotion. In some cases fra ctured bones hea l well in one dimension , only to det eri orate in another. In itial ra diographs of a 24-year-old male injured in a motor vehicle a ccident showed a transverse fracture at the junction of the mid and distal t hirds of the left tibia, with tw o fibular fractures, one at the same
lev el a s in t h e t i bia a n d a n ot h er a t t h e proximal end of the shaft. Both tibia an d fi bula were in perfect alignm ent wh en viewed ant eroposteriorly immediately a fter t he injury, although 3° of posterior angulation of t h e t i b ia w a s n ot e d i n t h e l a t e r a l v i ew . Figures 6a and 6b were taken more than 4 months after the injury and the fracture lines are still visible. In the AP view, the tibia has now shifted laterally by about the width of the bone cortex, a nd shows 4° of valgus an gulat ion. The midsha ft fi bular fra cture a lso displays 4° of va lgus in the AP plane. The la tera l view, however, now shows the tibia in good alignment. Although the effects of high velocity vehicular accidents may appear to have little relevance to archaeological remains, multiple fractures of this t ype ha ve been noted in h istoric cases of i nju r y i n h or s e-d r a w n ca r t a n d ca r r i a g e accidents, and this c ase has obvious relevance for modern forensic investiga tions of dry bone lesions.
Lovell]
153
TRAUMA ANALYSIS
Fig. 6. Single tibial fracture with double fibular fra cture. These radiographs w ere taken more tha n 4 months postinjury an d the fra cture lines are still visible. a: Anterior view. b: Latera l view.
Figure 7a is an immediate postinjury xray of a 54-year-old ma le who fell and suffered oblique, midsha ft fra ctures of the right tibia and fi bula. I n th e AP view there is 11° of valgus angulation in the tibia and 13° of valgus angulation in the fibula. The lateral view shows 10 mm of shortening in both t he tibia and fibula and 1° of posterior angulation in the tibia. Both bones have sh ifted anteriorly about the width of the bone cortex. Figures 7b and 7c were taken 6 months later. The fracture lines are still very evident, little callus is seen, and the fra gment ends are rounded, suggesting nonunion in both bones. S hortening has lessened to 6 mm in both t he tibia an d fi bula. On t he AP view the tibia reta ins 11° of valgus a ngulation but the fibula now displays only 4° of valgus. In t he latera l view, posterior angu lation in th e ti bi a has i ncreased to 4 °, 1° of posterior angulation i s seen in the fibula, an d both bones reta in their ant erior shifting. A good example of a rotation injury w ith no fibula r involvement is a spira l fracture of the dista l third of the right tibia in a 12-yearold male who fell off his bicycle (Fig. 8). Ca llus is evident since the radiogra phs were
ta ken about 2 month s postinjury, a nd on the AP view there is 5° of valgus angulation. The tibia is not shortened, probably because the i nt a c t fi b ul a h el pe d i t m a i n t a i n n o r ma l length. On the lat eral v iew, t here i s 6° of ant erior a ngulation in the tibia. ANATOMICAL SUMMARY OF FRACTURES AND DISLOCATIONS COMMONLY SEEN IN ARCHAEOLOGICAL BONE
To aid the diagnosis of t ra uma according to the mechanism of injury, this section is organized as an at las and descri bes those fra ctures a nd disloca tions commonly seen in archa eological bone according to t heir a natomica l loca tion. The possible complicat ions of the injury a nd th eir affects on healing a lso ar e described. Cranium
Fra ctures of the bones of the cranium va ry considerably, but perha ps t he most commonly described are those involving the flat bones of the vau lt. Typica lly, th e pat terning of fra cture lines on th e cranium is correlated with the severity o f the force: whether a
154
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
[Vol. 40, 1997
Fig. 7. Oblique, midsha ft fra ctures of the right tibia a n d fi b u l a . a: Immedia tely postinjury. Anterior view (b) and lateral view (c) 6 months postinjury. Nonunion is evident in both bones.
blow lands on the frontal, occipital, or parietal region, a single linear fracture line indicates less force than does a pattern of concentric and radiating stellate fracture lines (Gurdjian et al., 1950). The position of fracture lines can sometimes be used to identify the point of impact. Stellate, or
sta r-sha ped, fra cture lines form at the point of impact, for example, and radiating fracture lines run laterally, away from the point of impact. C oncentric heaving fra ctures ar e caused by shearing forces a nd ha ve cha racteristics, such as bevel an gle, tha t can dist ing u is h b et w e e n h i g h v el oci t y a n d b lu n t
Lovell]
TRAUMA ANALYSIS
Fig. 8. Spira l fractur e, 2 months postinjury. Anterior view is on the left an d latera l view is on the right.
tra uma injury (Berryman and Ha un, 1996). With blunt trauma the concentric fractures are caused by fo rc e from outside the cranium, which leads to beveling on the inner table, whereas with high velocity projectile tra uma t he fractures are caused by pressure from within the cranium, which produces beveling on th e outer t able. Identifi cation of the point of impact and the direction of the force becomes increasingly difficult with more severe trauma but the sequence of multiple impacts usua lly ca n be determined since a subsequently produced fracture will not cross a preexisting one. D i r ect t r a u m a i nju r ie s t o t h e cr a n i um often occur wh en the head is struck by a moving object. Tra uma from h igh velocity objects, such as bullets and motorized vehicles, is seen commonly in clinical cases, but tha t from lower velocity objects (e.g., bricks, rocks, bludgeons, push carts, wagons) is also observed toda y a nd un doubtedly occurred in the past. Direct trauma to t he cranium a lso occurs if t he head strikes the ground after a fall or jump from a height or
155
when balance is lost after l anding on the feet. These low velocity impacts usually result i n linear fractures. Li near fracture lines tend to sweep a round the thick, bony buttresses of the cranium (i.e., the petrous bones, mast oid process, etc.) unless they approach t hese area s perpendicularly. Since the structurally weak areas of the cranium ar e most prone to develop fracture lines, the unfused cranial sutures in children will readily separa te t o a ccommodate t he forces of impact. Alterna tively, in very young children the cranial bones may bend inward without fra cturing and the depressed deformity ma y persist. Clinically, blunt trauma injuries to the cranium usually cause linear fractures of the vault and the appearance of these fracture lines may help identify the point of impact and the mechanism of injury. B lunt trauma to t he frontal bone, for example, produces fracture lines tha t r adia te thr ough the frontal sinus, the cribriform plate, and the orbital roofs, although transverse fracture lines affecting the temporal regions may also a ppear. Anterior temporal impact leads to fracture lin es that radiate down, across either the orbital plate or the sphenoid-temporal region. In contrast, lateral or posterior temporal impa ct produces fra cture lines that ra diate downwa rds either in front of or b eh in d t h e p et r ou s p or t i on of t h e temporal bone and extend across the cranial base. Impact to the occipital bone usually produces fracture lines tha t r adia te down to the foramen magnum or the jugular foramen, and tha t ma y extend a nteriorly across the cranial ba se. Tra uma to the cranial base must be severe in order to cause a fracture, since the bone here is heavily buttressed. A base fra cture is th erefore considered to represent a severe injury. After vault fractures, sphenoid fractures ar e the m ost common clinical result of blunt trauma to the cranium (Unger et al., 1990). Unfortunately, sphenoidal structures are very fragile and thus prone to postmortem damage as well as to fatal consequences of fracture a nd t herefore it may be difficult to identify sphenoid fractures in ar chaeological skeletons. Facial fractures, either due to direct or indirect trauma , are often very complex but commonly heal adequately with-
156
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
[Vol. 40, 1997
Fig. 9. Well-healed, crush fracture of the right parieto-temporal region.
out medical treatment. S ince the zygoma, maxilla, and orbital margin are mutually supportive, a fracture of one of these bones usually involves a fracture of at least one of the o thers. Fractures of the nasal bon es, wh ile usua lly not severe, a re not uncommon. These a re often called depressed fra ctures (e.g., Filer, 1992) alt hough this description refers to the observed deformity of the na sa l bridge, not the type of injury. Clinically, interpersonal violence often produces sma ll fra ctures of the nasa l and zygoma tic bones. Crush frac tures of the cranial vault are commonly seen in archaeological human r em a i ns a n d a r e ca u s ed b y l ow v el oci t y direct trauma (Fig. 9). Lesser force is indicated by the lack of displacement of bone fragments, w hile great er force is chara cterized by inward displacement. A portion of bone might be completely detached if great force is applied, particularly if the object has a sm a ll striking surfa ce, but more often seen in archaeological remains is the incomplete detachment of the bone (Fig. 10). The fracture line on the ectocranium is usually irregular and may be comminuted, producing a cobweb or mosaic pattern. The depressed area indicates the point of impact, from which linear fractures r adia te. Clinically, blows from ha mmers, fi replace pokers, and
the butt ends of axes are commonly responsible for incompletely detached depression fractures, as a re falls onto the sharp edge of furniture or concrete steps (Polson et al., 1985). Penetrating injuries of the cranium are caused by pointed and edged objects (e.g., knives, swords) or by bullets. Heavy cutt ingedged w eapons th at ar e used in a chopping manner will produce crush injuries in addition t o penetrat ion, a nd furth er injury ma y be caused if the embedded weapon is removed with a twisting motion. This damage is often indicated by splintering of the bone with outward displacement near the initial impact site. The t ype a nd size of w ound produced by a projectile depends upon the size of the projectile, the speed at which it strikes the bone, and the distance it travels. Historical skeletons ma y exhibit evidence of gunshot trauma , although these injuries would be less severe in terms of bone fragmentation and destruction than those typically found in a metropolitan trauma center today. E arly musket balls, for example, had low velocity chara cteristics due to their spherical sha pe and the poor q uality of gun powder (Butler, 1971). High velocity bullets (3,000 ft/sec) were not developed until a l m os t 1900, a d a t e t h a t u s ua l ly p la c es
Lovell]
157
TRAUMA ANALYSIS
Fig. 10. Depression fra cture of the cranial va ult, showing radiating a nd concentric fracture lines. P robably due to low velocity blunt tra uma .
h u ma n r em a i n s i n a f or en s ic r a t h e r t h a n archaeological context. Details of the interpretation of modern gunshot wounds can be found in many textbooks on forensic medicine. P ossible complica tions of cra nial fr a ctures include displacement of bone fra gment s (ma lunion), indirect tra uma injuries elsewhere on the cranium due to the transmission of impact force, and soft tissue damage. The location of the impact d etermines th e subsequent consequences of the injury due to the different anatomical structures in the cranium. Linear fractures usually involve both the inner and outer tables of the cranium but do not involve displacement or depression of the bone and thus are often not considered a s serious a s those injuries resulting from greater force. Complications can arise, however, due to transmission of the
force of impact, such as when direct trauma to the ba ck of the cra nium produces indirect effects on the orbital plates. The consequences of cranial fractures can be fatal if the blood vessels running along the inner tables of the cranium (e.g., middle meningeal a rteries) a re torn, alt hough this complication is unlikely to be detected with certainty in archaeological remains. Mandible
The mandible forms what is essentially part of a ‘‘ring’’ structure, and therefore a fracture on one side is commonly accompanied by a balancing fracture on the other side. Usua lly the fra cture affects t he horizontal ramus or angle on one side and the condyle on the opposite side. Fractures at the a ngle very often communicat e with the
158
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
[Vol. 40, 1997
roots of the dist a l molar s. Very few ma ndibular fractures in an cient skulls have been described but fractures of the ascending ram us, mandibular a ngle, and condylar processes have been reported (Alexandersen, 1967). Asymmetr ica l tooth w ear an d osteoa rthritis at the temporomandibular joint are possible complications of jaw fractures. Hyoid
Although t he hyoid bone is not al ways recovered during archaeological excavations, a perimortem hyoid fracture is considered strongly suggestive of interpersonal violence t hrough str a ngula tion (Maples, 1986). Vertebrae
The most common fractures of the vertebrae a re due to indirect tra uma, preexisting disease, or st ress. A very distinctive vertebral fracture is the traumatic separation of the neural a rch from the vertebral body at t he p a r s i n t er a r t i cu l a r i s (known as spondylolysis; Fi g. 11), which appears to be a common consequence of habitual physical str ess (J imenez, 1994; Mer bs, 1989a, 1989b, 1995, 1996). Although th e spondylolysis seen most often in clinical settings is complete separ at ion, comprehensive surveys of a rchaeologica l skeletons indicat e tha t t he condition begins as incomplete stress fractures in adolescents that may heal or, conversely, may progress to complete lysi s by young a dult hood (Merbs, 1995). The condition m a y be initiated by a n a cute overload event t hat causes microfra ctures, but it is generally ag reed tha t th e determining fa ctor is chronic trauma, with repeated stressing promoting nonunion of the microfractures. These fatigue fractures appear to have the greatest populationa l frequency (a pproa ching 50%) am ong ar ctic-ad apt ed peoples following tra ditional lifewa ys, but clinically they are most often observed among athletes and laborers whose activities involve frequent and large stress reversals between lumbar hyperextension and lumbar flexion (Merbs, 1989b, 1996). Reported sex differences in the prevalence of th e condition ma y be a ctivityrelated (Merbs, 1989b). Spondylolysis may be unilateral or bilateral in expression, but predominates in the lumbosacral region, particularly L5 and, to a lesser degree, L4
Fig. 11. Lumbar spondylolysis. This separat ion of the neural arch and the body at the p a r s i n t er a r t i cu l a r i s is usually attributed to a fatigue fracture.
(Merbs, 1996). Complete separation is frequent ly accompan ied by ant erior slippa ge of the vertebral body (spondylolisthesis), but functional complications of this are rare (Merbs, 1989b). A fracture with possibly a similar origin is the traumatic separation of the t ip of the spinous process of the seventh cervical or fi rst thoracic vertebra. Referred to as ‘‘clay-shoveller’s fracture’’(Roberts and Manchester, 1995), it may result from the strenuous muscle action associated with shovelling clay, cement, or r ocks. More common in a rcha eologica l vertebra e than stress fractures are in di rect trauma injuries, such as Schmorl’s nodes. These result from bulging of the disc’s nucleus pulposus, w hich puts pressure on the vert ebral end plat e a nd leads to bone resorption in the affected area. Herniation of the disc tends to occur gradually in adults because the nucleus has lost resiliency, whereas it may occur suddenly in younger individuals
Lovell]
TRAUMA ANALYSIS
159
in whom the nucleus still quite gelatinous (B ullough a nd B oachie-Adjei, 1988). Ind irect vertebra l dama ge also can occur in a fa ll or jump onto the feet, since the force o f impact is carried up from the lower limbs thr ough the spine. Fra ctures secondar y t o pat hology a re also common in the vertebral column. The best known clinical example is that of biconcave vertebrae, which result when intervertebral disks expand into the superior or inferior surfaces of vertebral bodies t hat have been weakened by osteoporosis.4 Similar ly, compression flattening of vertebral end plates due to sparse, c oarse trabeculation in the vertebra l body is a classic feature of sickle cell disease. Due to the greater strength of the rims of the vertebral end plates, they ma y be spar ed even wh en the vertebral body is compress ed. Although direct trauma injuries to the vertebra e are ra re, hairline tra nsverse fractures on, or posterior t o, the superior a rticular processes of the second cervical vertebra are worth noting since they can result from strangulation (Maples, 1986). Ribs and sternum
Ribs are known t o incur str ess fractures, usually a s a result of occupational or simil a r ly h a b it u a l l a bor b ut s om et i me s a s a conseq uence of persisten t coughing or vomiting. Most often, however, rib fractures result from direct tr auma , such a s a blow or a fall against a hard object (Adams, 1987). Clinically, rib f ra ctures are the most common type of thoracic injury a nd ar e observed in 60 to 70%of individua ls a dmitt ed to hospital w i th b lu n t ch es t t r a u m a (C a r r er o a n d Wayne, 1989). The direction of the impact usua lly ca n be determined from the loca tion of the fracture, i.e., ribs are usually fractured near the angle if the force is applied from th e front; beside th e spine if th e force is applied from the back; and beside both the spine and the st ernum if t he force is applied from the sides. Th e fi f t h t o n i n t h r ib s a r e m os t of t en fractured (Fig. 12). Fracture of the first to third r ibs and/or the sternum indicat es tha t 4 Abiconcave vertebra should not be confused with a ‘‘butterfly vertebra,’’which is a congenital malformation.
Fig. 12. Multiple healed fractures of the left ribs. The location of the fractures, near the a ngle, suggests tha t the force wa s applied from the front.
the mecha nism of injury wa s a high kinetic force. D ue t o the flexibility of the rib cage, particularly in the anteroposterior dimension, the degree of inward displacement at impact may have been much greater than tha t discernable postinjury, a nd thus soft tissue damage may have been more severe than might be inferred from the damage to the bones themselves. Such soft tissue da mage includes laceration of the pleura, lungs, or intercostal vessels, which would have been largely untreatable i n earlier times and thus ma y point to a possible cause of deat h. Pn eumothorax (the presence of air in the pleural cavity) and hemothorax (the presence of blood in th e pleural ca vity) may be caused by rib fra ctures at a ny level in the t h or a c ic ca g e a n d m a y b e s im il a r ly l if ethr eat ening. Another serious complica tion of rib fractur e occurs w hen there ar e at least t w o b r ea k s i n on e r ib . Th is p rod u ce s a free-floating fragment and thus a risk of
160
[Vol. 40, 1997
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
interna l dama ge, referred to a s ‘‘fl ail’’ injury. U n f or t u n a t el y, i f t h e s of t t i ss u e d a m a g e caused by rib fractures is serious enough to cause death, the bony injuries observed in the a rchaeologica l skeleton w ould be identifi able only a s perimortem fra ctures. Hea led rib fra ctures in t he a rchaeological record thus probably represent non-life-thr eat ening trauma . Clavicle
Cla vicular fra ctures are most often caused by a fall onto the shoulder but occasionally result from a fall onto an outst retched han d. The break tends to occur at the junction of the middle and lateral thi rds, with downwa rd a nd medial displacement of the latera l fragment being a common complicat ion. Since clinical treatment of clavicular fractures is usually limited to the use of a sling f or 1 or 2 w e ek s f or p a in r el ie f, h ea l ed fra ctures often exhibit some deformity. Malaligned clavicular fractures in ar chaeological skeletons therefore do not necessarily point to an a bsence of medica l treat ment. Scapula
Scapular fra ctures are uncommon but a re usually the result of direct tra uma. B oth the flat and irregular portion s of the scapula may be involved. Four types are commonly described in the clinical literature: 1) fracture of th e scapular body, whi ch may be comminuted but ra rely displaced because of the la rge muscles holding t he bone in pla ce; 2) fracture of the neck, which may lead to downward displacement of the glenoid; 3) and 4) fractures of the acromion and coracoid processes, respectively, which ra nge from simple cracks to comminution and which may be associated with downward displacement. Although none of these injuries is usually considered serious, a possible complica tion of scapula r fra cture, especially i f t h e i nju r y occu r s on t h e l ef t s id e, i s pneumothorax (Carrero and Wayne, 1989). Humerus
The most common humeral fractures in adults affect the neck, great er t uberosity, and shaft . Neck fra ctures are most common in older women in whom osteoporosis has weakened the bone. Indirect t ra uma from a
fall onto an outstretched hand is the usual cause and in more than 50%of these cases the fra cture is self-sta bilized through impa ction. Direct tr a uma , in the form of a fall onto the shoulder or a blow, ma y cause fra cture of the greater tuberosity. S haft fractures are most common in th e middle third of the bone and may be due to direct or indirect trauma. The proximal half of the shaft is a common site of fra cture seconda ry to pat hology, such a s w h en t h e b on e h a s been in va d ed b y metastatic disease. Complicat ions of humeral shaft fractures include displacement, nonunion, an d injury to the radia l nerve. I n con t r a s t t o t h e p a t t er n ob se rv ed i n adults and described above, humeral fractures in children tend to occur at the dista l end, affecting the supracondylar, epicondylar, and condylar regions. Supracondylar fra ctures are indirect tra uma injuries caused b y a f a ll o n t o a n ou t s t re t ch ed a r m , w i t h displacement occurring posteriorly. Complications include maluni on, damage to the b ra c h ia l a r t e r y, a n d t e a r in g of t h e joi nt capsule with hemorrha ge into the joint a nd surrounding tissues. Epicondylar fractures a r e u s ua l ly m ed ia l a n d m a y r es u lt f r om direct trauma but more often from avulsion by flexor muscles i n a fall. If the avulsed fragment enters the joint, which occurs frequently in children, complications in terms of reduced function and osteoar thritis usually ensue. This injury may also cause damage to the ulnar nerve. Condylar fractures ar e uncommon but a lso tend t o result from a fall. In contr a st t o the medial involvement in epicondyla r fra ctures, the lat eral portion, or capitulum, is usually involved in condylar fractures. D isplacement is n ot uncommon and this injury is often complicated by deformity, nonunion, a nd osteoart hritis. Ulna
Ulnar fract ures are not especially common clinically, but when they occur t hey usually affect the olecranon or the shaft. Olecranon fractures are more common in ad ults an d result from the direct tra uma of a fall onto the point of the elbow. S everity ranges from a simple crack to comminution and the injury ma y be complicat ed by n onunion and osteoarthritis. D iaphyseal fractures can result from either direct or indi-
Lovell]
161
TRAUMA ANALYSIS
Fig. 13. Healed oblique fracture of the ulnar shaft. This type of fracture suggests tha t the mechanism of injury w as indirect tra uma (e.g., falling onto an outstretched ha nd) rather t han direct t rauma (e.g., parrying a blow).
rect trauma (Fig. 13). They are prone to severe displacement, malunion, nonunion, a nd t o infection becaus e of the bone’s proximity to the sur face of the skin. Fra cture of the proximal sha ft of the ulna is often a ssociat ed with the dislocation of the radial head. This injury is referred to as a Monteggia fra cturedislocation. It is usua lly ca used by a fa ll onto an outstretched hand with forced pronation, but it ma y a lso be caused by a blow to the back of the upper forearm (i.e., a ‘‘parry’’ fracture). Deformity commonly chara cterizes this injury in archaeological skeletons since the fracture cannot be properly reduced without sur gery. Radius
Another forearm injury, the G a leazzi fracture-dislocation of the radius, is more common clinically than is the Monteggia fracture-dislocation. It is also usua lly caused by a f a ll o n t o t h e h a n d , a n d s im il a r ly i t i s difficult to rea lign without su rgical intervention. The fracture occurs near the junction of the middle and di st al th irds of the radial shaft and is accompanied by dislocation of th e inferior ra dio-ulna r joint . The most comm on r a d i us f r a ct u r e o ccu r s a t t h e d is t a l sha ft a nd is ca lled the Colles’fra cture. Clinically, it is the most common of all fractures i n a d u lt s ov er t h e a g e of 40, e sp eci a ll y females, and is nearly a lway s caused by the indirect t ra uma of a fa ll onto the ha nd. The break usua lly occurs about 2 cm a bove the distal articular surface of the radius, an d
the distal fragment is posteriorly displaced and usually impacted. This injury may be associated with fracture of the styloid process of the ulna. Although fractures of the distal ra dius, such as t his, are one hundredfold more frequent tha n a re fractures of the proximal radius (Knowelden et al., 1964), fracture of the radial head also c an result from a fall on an outstretch ed hand. Observed mainly in young adults, it usually appears as a crack without di spl acement. Comminution is possible, however, and in such cases th e injury ma y be complica ted by osteoar thritis. In contra st to these indirect trauma injuries, a simple fra cture of t he radial sh aft, somewhat less common than the fr a cture-dislocation injuries, usua lly results from direct trauma. Malunion is the most common complicat i on of r a d i us f r a ct u r es , a n d a b s en ce of medical treatment cannot be presumed if deformity is observed since redisplacement is very common clinically within a week of fracture reduction. In archaeological specimens, at least one study ha s reported tha t ra dius fractures ra rely healed wit hout deformity (Gra uer a nd Roberts, 1996). Pelvis
Isolat ed fra ctures of the pelvic bones most commonly appear on the superior and/or inferior ischio-pubic ramus and the wing of the ilium. These fractures ar e q uite benign unless displacement occurs. More serious pelvic fractures a re t hose tha t disrupt t he
162
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
pelvic ring through the r ami or a t t he pubic symphysis, with associated dislocat ion of th e sacroilia c joint. The mecha nism of injury in th ese ca ses tends to be a nterior-posterior crushing, latera l compression, or vertical shearing force. Complications are usually serious a nd w ould likely be life-thr eat ening in the absence of modern medical trea tment . Afracture-dislocation of the hip occurs when the head of the femur is driven through the floor of the a ceta bulum. This is usua lly t he r es u lt of a h ea v y b low u pon t h e l a t e r a l femur, due to a serious fall or a similar impact (e.g., vehicular tra uma in clinical cases). The injury tends to comminution and serious complica tions. A poorly underst ood lesion tha t may result from t he incomplete and tempora ry dislocat ion of th e hip is the acetabular fl ange lesion, which a ppears a s a fla ttening of th e superoposterior rim of the aceta bulum (Knowles, 1983). Osteoar thr itis is a common sequel to any injury that involves the aceta bulum. Femur
Clinical data indicate tha t fra ctures of the femoral neck and trochanteric region are very common i n the elderly an d are seriously disabling. Femoral neck fractures are often a consequence of osteoporosis and therefore appear most often among older females, although physical activity during the reproductive years (i.e., age 15 to 45) diminishes signifi cantly the long-term risk of femoral neck fractures (Åstro¨m et al., 1987). When s econda ry t o ost eoporosis, fem ora l neck fractur es may r esult from very mild tra uma , such as a st umble. They usua lly a re due to rotational force and cause lateral rotat ion and upward displacement of t he sha ft. Ava scular necrosis is a serious complication of femoral neck fracture a nd is caused by dam ag e to vessels in the neck tha t supply t h e f em or a l h ea d . Th e r es u lt i s t h a t t h e blood supply t o the hea d ma y rely overly on vessels in the ligamentum teres, which is only one of th ree routes of blood supply an d w h i ch i s u s u a l ly i na d e q ua t e on i t s o w n . Necrosis is usually sufficiently adva nced within 2 to 6 months postinjury that th e head collapses. Unless rigidly immobilized, femoral neck fractures are prone to nonunion, with avascular necrosis being the
[Vol. 40, 1997
most importa nt contributor. Osteoar thritis is another common sequel and ma y be due to mechanical damage at the ti me of injury, impairment of blood supply to the basal layer of art icular cartilage, a nd/or to m alalignment of a united fra cture. An impacted abduction fracture of the femoral neck occurs less commonly, but it usually unites without surgic al intervention and may be complica ted only by s light short ening of the limb a nd possibly by art hritis. Fractures of the trochanteric region (i.e., roughly between the grea ter a nd lesser t rochanters) are common clinically but are not likely to be observed in archaeological populations because they are a lmost a lway s seen in adults over 75 years of age. Decreased physical activity among older individuals is likely a contributing factor (Nilsson et al., 1991), further limiting their occurrence in past populations who followed traditional, nonmechanized, lifeways. Femoral shaft fractures a re often due t o severe direct or indirect trauma. They may occur at any location on the shaft and may be of any fracture type. They are complicated by simultaneous hip dislocation; arterial damage that can compromise the viability of t he limb; nerve dama ge, especially to the sciat ic nerve; an d delay ed, mal-, or nonunion. Four months is considered the average healing time in clinical cases of shaft fra ctures. Associat ed deformity of shaft fra ctures tends t o be shortening or a ngulation, the latter having a tendency to facilitat e development of osteoart hritis a t t he knee. Supracondylar fractures of the femur are more or less transverse an d located just above the epicondylar region. Ma lunion is not a common sequel in clinical cases w here there has been proper medical treatmen t but is a possible complication in earlier times if t he knee could n ot be kept immobilized for 2 to 3 weeks. Condyla r fra ctures a re uncommon, but when these do occur they are a lmost a lways due to direct tr auma . The severity of the f ra cture can range from a n undisplaced crack t o complete separ at ion of a condyle with marked upwa rd displacement. In the absence of treatment, a displaced fractur e is likely to heal out of a lignment a nd osteoart hritis of the knee joint is a
Lovell]
163
TRAUMA ANALYSIS
proba ble consequence. Occasionally a tra nsverse supracondylar fracture combines with a condylar fracture a nd forms a T-shaped fracture line that splits a part the t wo condyles. Patella
Fra ctures of the pat ella ma y be caused by direct or indirect trauma. Indirect trauma causes an avulsion fracture, usually a clean, tra nsverse sepa ra tion of the bone, due to the sudden a nd violent contr action of the qua driceps muscle. In contra st, direct tra uma from a fall or a blow onto the patella tends to ca u s e a cr a c k f r a c t ur e or a com m in u t ed fracture. Undisplaced crack fractures tend to hea l w ithout complicat ion because t he fragments are held in position by the aponeurosis of the q ua driceps muscle, but fra ctures tha t involve separa tion of the fra gments a nd those that ar e comminuted will produce an irregular articular surface unless surgically repaired, a nd osteoar thritis is then a n obvious sequel. Tibia and fibula
Injuries t o the knee joint most commonly involve th e menisci a nd liga ment s, not bones, and therefore there may be no evidence of trauma other t han soft tissue ossification after ligament stra in or tear, or a vulsion of the tibial spine from injury to the cruciate ligaments. When fractures do occur around lower limb joints t hey t end t o a ffect the ankle, not the knee. Clinically, the bones forming t he ankle a re injured more often tha n an y other bone except the dista l radius. Isolated fractures of the malleol us of the tibia or fibula are especially common, and occur with or without dislocation of the talus. The usual mechanism of injury for these is either a bduction a nd/or latera l rotation for fractures of the lateral malleolus of the fi bula; a nd a dduction for fra ctures of the medial malleolus of the tibia. Simultaneous fractures of both malleoli are less common. Vertical compression forces can lead to fracture of the anterior margin of the distal tibia, or, i f severe, f ragmentati on of the distal tibial articular surface, the latter injury being prone to osteoa rth ritis. Fra ctures and dislocations at the ankle are often complicated by ligament damage, which could
be identifiable in archaeological skeletons due to soft tissue ossifi cation. Most diaphyseal fractures of the leg involve both the tibia a nd fi bula. If the mecha nism of injury is an a ngula r force, it w ill lead to transverse or short oblique fractures of the shafts at roughly the same level. If the injury is due to a rotat ional force, spiral fractures will result and will occur at different levels in the two bones. Distal tibial sha ft fra ctures, i.e., above the medial ma lleolus, ar e commonly a ccompanied by proximal fibular shaft fractures. Conservative treatment of tibial shaft fractures is to manua lly reduce the fra cture an d to minimize weightbearing on the limb. Immobiliza tion for 2 to 3 weeks is recommended in cases of sta ble fra ctures, but for up to 6 weeks if the injury is likely to be displa ced; full union ma y ta ke as long as 4 months. Malunion is rare in clinical cases but common archaeologically beca use of the gr eat er likelihood of fra ctures remaining unreduced and the difficulty of immobilizing th e leg. Beca use it is so close to the surfa ce of the skin, the tibial shaft is the most common site of an open (compound) fracture and hence infection from contamination. Infection can also lead to nonunion. A fi bular sha ft fra cture is not considered serious by most clinica l pra ctitioners because it unites rea dily and is of such little functiona l importance that surgical removal of a portion of the sha ft is not only t olerable but is advocated when the bone acts as a strut, d is t r a ct i ng t h e f r a g m en t e nd s of a t i bi a l fra cture an d promoting nonunion. Thus, fi bular sha ft fra ctures are expected to be rar e in the a rcha eologica l record but wh en observed they ma y be healed with deformity. Fractures of the tibia at the knee joint are uncommon. Clinically, the most common is fracture of the lateral tibial condyle, caused by a lateral force aga inst the knee such a s experienced by football players or by pedestrians struck by a vehicle bumper. If t he articular surface is fragmented in the injury then osteoart hritis is a predictable sequel. Hand, wrist, foot, and ankle
According to clinical evidence, the irregular bones most commonly fra ctured are the scaphoid and t riquetral in the hand, and t he calcaneus in the foot (Adams, 1987). Frac-
164
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
ture of th e scaphoid tends to occur in young adults, usua lly due to indirect t ra uma from a fa ll onto an outstr etched hand . The typical i nju r y i s a t r a n s ve rs e b r e a k t h r ou g h t h e ‘‘waist’’ of the bone, and should not be confused with a congenita lly bipart ite sca phoid. Scaphoid fractures are accompanied by a high frequency of complications, including delayed union, nonunion, avascular necrosis, and osteoar thritis. Injury to th e triquet r a l a l s o u s ua l ly r es u lt s f r om a f a ll , b u t causes a chip fracture on the dorsal surface of the bone. Almost all calcaneal fractures are caused by a fall from a height onto the heels, such as occurs as an occupational hazard among builders and window cleaners (Wells, 1976) and a result of misad venture among parachutists, hikers, and rock climbers, for example. Afa ll from a t ree, cliff, or roof of a dwelling is a possible cause of such injury in past populations, although it ha s been suggested tha t th ese injuries would be uncommon in falls of less than 4 m (Wells, 1976). The fracture may be observed as a split or crack in the subtalar tuberosity, but more often th e art icular surfa ce of the calcaneus fails to withstand the stress and results in a crush injury. Fracture lines may radiate to the front a nd a ppear a lso on t he calca neo-cuboid joint. Associat ed crush fra cture of lower thoracic or upper lumbar vertebrae may be noted. The usual complication of calcaneal fra cture is osteoar thritis. Fra ctures of the talus a re relatively uncommon. Metacarpals, metata rsals, and phalanges ar e common sites of tra uma tic injury. Meta carpals are often fractured due to longitudinal compression impact such as from boxing. If the fracture line enters t he joint t hen osteoarthritis is a likely complication. The neck and dista l shaft of meta carpals also are prone to transverse or oblique fractures, often complicated by displacement. Manual phalanges tend to exhibit spi ral or transverse fractures of the sha ft or oblique fra ctures of the base. Comminution is most likely in t he dista l pha langes. D islocat ions of pha lan ges ar e mainly due t o forced hyperextension, but at least one ar chaeological example appears to result from forced adduction with associated t earing of the medial collateral ligament of the interphalangeal joint t ha t ma de reduction impossible and led
[Vol. 40, 1997
Fig. 14. Healed fracture of the left fifth metat arsa l, with the un affected contralatera l element for comparison. Although t he fra cture line is not clear, th e rotat ion of the hea d indicates a spira l fracture, probably due to a combined an gular /rotat ional force.
to mala lignment of the joint a nd subsequent ankylosis (Drier, 1992). Metatarsal shafts may exhibit tra nsverse or oblique fra ctures (Fig. 14). The base of the fi fth m eta ta rsa l is a common site of avulsion fracture caused by a twisting injury. The pedal phalanges, particularly that of the great toe, often suffer comminuted crushing injuries from direct trauma. INTERPRETING THE ULTIMATE CAUSE OF INJURY
Once the injury has been described, its proximat e cause determined, a nd a ny complications identified, the ultimate cause of the trauma can be evaluated. This evaluation must consider three types of information: 1) the characteristics of the fracture itself; 2) the skeletal pattern of trauma in the individual and th e population; and 3)t he
Lovell]
TRAUMA ANALYSIS
social, cultu re hist orical, a nd/or environmental context of the human remains, including the presence of artifacts. Clinical research ha s considered the role of ma ny va ria bles in t r a u m a ca u s a t i on (e .g . , Ag a r w a l , 1 98 0; Åstr o¨m et a l., 1987; Ba rber, 1973; B jo¨rn st ig et al., 1991; Buhr and Cooke, 1959; Busch et al., 1986; Cogbill et al., 1991; Donaldson et a l., 1990; Fife and B ar a ncik, 1985; Fife et al., 1 984; Ga rraw ay et a l., 19 79; Grimm, 1980; J oha nsson et a l., 1991; J o´nsson et a l., 1992; J ones, 1990; Knowelden et al., 1964; Madhock et al., 1993; Nilsson et al., 1991; P rince et a l., 1993; Ra lis, 1986; Sa hlin, 1990; Sha heen et a l., 1990; Zylke, 1990) and provides valuable aids for the interpretation of fractures in antiquity, particularly with regard to skeleta l patterning a nd the contexts of injury. The use of the characteristics of the fract u r e i t se lf t o i de nt i fy t h e m ech a n i sm of injury an d point t o logical causes of fractur e can be illustrated by reference to a particularly problematic interpreta tion of tra uma, that of the ‘‘parry’’ fracture. Although this injury has been described va riously as a s i mp le m i d sh a f t f r a c t u r e (L ov e joy a n d Heiple, 1 981), the result of a blow to an upra ised arm dur ing a fi ght (J a nssens, 1970), or t h e f r a c t ur e o f b ot h r a d i us a n d u ln a (Wells, 1964), the term is interpreted by many scholars to identify the involved bone (the ulna ), indicate t he locat ion of the injury (on the shaft), and imply specific social and cultural circumstances, i.e., interpersonal conflict. A true parry fracture would indeed be caused by a direct blow to the forearm, but a proper diagn osis must rely on descriptions of the fra cture location, fra cture type, bones involved, a pparent direction of force, and evidence of any complications. Although simple transverse fractures of the ulna are conventiona lly viewed a s pa rry fractures, a considerat ion of fracture ty pes a nd mechanisms of injury indicates that a parry fracture could be indicated by eith er a transverse line or a comminution; t he fra cture could be closed or open; and the radius could be affected in a ddition to the ulna . The term ‘‘par ry fr a cture’’ does not a ppea r in t he clinical literature on fracture description, etiology, a nd trea tment (e.g., Ada ms, 1987), a nd other terms commonly used in orthopedic
165
medicine should therefore be used to describe relevant injuries. An example is the Monteggia fra cture of the ulna with associat ed dislocation of the proximal ra dius, described previously in this paper. Although this term describes the type of fracture, it does not specify a cause. In fact, this injury can be caused by a direct blow to the posterior ulnar shaft or by the indirect t ra uma of a forced pronat ion injury. A fat igue fra cture might also manifest as a simple transverse break, a s illustrated by the case of a midshaft ulnar fra cture caused by the repeat ed physical stress of forking manure (Kitchin, 1948). Thus, although the prevalence of ‘‘par ry fra ctures’’ is often reported in pa leopathological studies it is not always clear whether all or any of the injuries actually resulted from parr ying a blow. These problems wit h th e diagn osis of parr y fra ctures a re compounded in paleopa thological interpreta tions of social order, par ticularly vis a´ vis gender relations (for a critique, see Mafart , 1991). Without other supporting evidence, Wood-J ones (1910) ascribed female sex to ancient Nubian skeletons that displayed forearm fractures since he thought the cause of the fract ures was spousal a buse by Nubian men a rmed with heavy staffs. Wells referred to forearm fractures among the a ncient Nubian s as indicat ing ‘‘short tempers and aggressive conduct’’ that implied wife beating or a generally low status of women (Wells 1964). His accompanying fracture descriptions, however, refer only to the fact that the injury appeared in t h e m i d - or l ow e r s h a f t of t h e u l n a , or a f fe ct e d t h e u ln a a n d t h e r a d iu s . U l n a r fra ctures among th e Cro-Ma gnon were similarly interpreted to indicate an aggressive nature (Zivanovic, 1982); however, the accompanying photograph and x-ray of one such injury shows a spiral fra cture, with rotation in the distal aspect. Spiral fractures, in particular, are unlikely to result from a direct blow, but ra ther from a tw isting, forced pronation injury, such as in a fall. While features of the fracture itself may identify the mechanism of injury, the skeletal pattern o f fractures in an individual may also help to clarify the probable causes of trauma. Clinicall y, th e ul na and radi us are fractured more commonly than are any
166
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
other skeletal elements, but the cause of injury va ries and ra rely is due to assault. A comprehensive study of long bone fra ctures in Medieval Britain also indicates that the radius and ul na are the bones most c ommonly affected, likely due to falls and other mishaps ra ther t han violence (Grauer a nd Roberts, 1996). Ind eed, fractur es of the cra n iu m , r i bs , or h a n d s a r e m or e l ik el y t o indicate trauma due t o int erpersonal viol en ce t h a n a r e f r a ct u r es t o t h e f or ea r m . Injuries that are considered to have a high specificity for a clinical diagnosis of assault are fractures of the skull (especially the nasal and zygomatic bones and the mandible); and posterior rib fractures, vertebral spinous process fractures, and fractures of hand and foot bones, which can result from the direct tr a uma of punches or kicks. Occasionally the palmar surfaces of the manual phalanges will exhibit healed or unhealed cutma rks, originat ing a s defensive wounds incurred a s a victim of a knife or sword att ack. Although fra ctures tha t pass along sutur e lines are common on subad ult skulls, th e exposure of diploe¨ in int ersut ura l perforations is a key indicator of antemortem trauma. In addition, metaphyseal lesions, transverse fractures of the scapular acromion, and sternal fractures are considered indicative of child abuse in clinical cases and may result from shaking forces ra ther tha n direct blows (Appleton, 1980; Brismar and Tuner, 1982; Fisher et a l., 1990; Fonseka , 1974; Gayford, 1979; Kleinman, 1987; Shepherd et a l., 1990). According to clinical pra ctitioners, the most specific finding of physical abuse in both adults and children is multiple injuries at different stages of healing (Maples, 1986; Walker et al., 1997; Wilkinson and Van Wagenen, 1993). 5 Multiple but simultaneous fractures, in contrast, ma y represent a ccidental trauma , as demonstra ted by th e ca se of a Neolithic flint miner in Belgium who suffered numerous fra ctures an d appa rently died of his injuries wh en th e roof of the mine collaps ed on top of him (Knowles, 1983). 5 For further descriptions a nd discussions of the skeletal evidence for interpersona l confl ict, the reader is referred t o Troubl ed
Times: Osteological and Archaeological Evid ence of Violence,
edited by D. Ma rtin a nd D. Fra yer and forthcoming from Gordon and Breach publishers.
[Vol. 40, 1997
Patterns of fractures within a population may also be informative. Clinically, older, postmenopausal females have more fractures t ha n do a ny other a ge/sex group due to the influ ence of osteoporosis a s a predisposing factor in fracture from minor, often indirect, tra uma, a nd t his pattern is paralleled in an early 20th century skeletal collection (Mensforth and Latimer, 1989). When older fema les ar e excluded, clinica l fra cture rates are greatest among individuals younger than 26 years of age and tend to be determined by their a ctivities. Age also influences the skeletal pattern of involvement. Femoral neck fractures, for example, occur commonly in older adults but rarely in children. Fea tur es of the physica l environment a lso have been shown to influence the frequency and n at ure of tra uma. For example, adverse weather conditions (e.g., snow and ice) and irregular landscapes increase fracture risk from falls, while reduced winter daylight hours in northern latitudes increase fracture risk from mishaps due to limited visibility. Decreased sunlight also ma y impair calcium absorption and lead to fractures secondary to osteoporosis or rickets, a nd dieta ry ina dequa cies of vita min C or ca lcium may increase the risk of pathological fractures. The sociocultural context of injuries must also be considered. Clinical evidence overwhelmingly indicates, for example, that most fractures are due to daily activity rather t h a n i n t er p er s on a l v iol en ce o r u n u s u a l events. Tra ditionally, most fr a ctures in females occur in the home while most fract u r es i n m a l es occu r a t w or k or d u ri ng sports, although this trend varies according to country, income level, occupat ion, an d ag e. High fra cture risks exist in occupations tha t h ave been genera lly restricted to men, such as agriculture, mining, forestry, and construction. In developing nations, household work such a s carrying wa ter an d loads of fi r e w ood p os e h ig h f r a ct u r e r is ks f or women, as do many f armin g activi ties enga ged in by both sexes. Technology-based transportation, either mechanized (e.g., automobiles) or unmechan ized (e.g., bicycles and horse-draw n wa gons), also carries its own fra cture risk. Given circumsta nces such
Lovell]
167
TRAUMA ANALYSIS
as these, the trauma caused by ac tivity o r occupation may be indistinguishable from that due to interpersonal violence (Smith, 1996; Wa kely, 1996). Finally, historical records and an understanding of the sociocultural context of the injuries can benefit their interpretation, such as the explanation of fractures due to occupational a ccident du ring t he 1856 constr uction of t h e G r a n d Tr u n k R a i lr oa d i n C a n a d a (J imenez, 1994). Sever a l cases of interpr eta tion ha ve successfully relied on such informa tion, such as t he identifi cation of stra ngulation from hyoid fra ctures and of hanging a nd decapitation from vertebral injuries (Angel and Caldwell, 1984; Waldron, 1996), and the comprehensive analyses of trauma etiology in an aborigina l populat ion of the Ca nadia n northwest coast (Cybulski, 1992), in prehistoric populations of the American midwest (Milner, 1995), and among soldiers of the 14th century battle of Visby (Ingelmark, 1939; Knowles, 1983). Often the most compelling evidence for interpersonal violence lies in the presence of artifacts or other physical evidence of violence. Pr ojectile points embedded in bone or recovered from the abdominal cavi ty are diagnostic of at least some tra umat ic events, a nd ma y point logically to the cause of other injuries as well (e.g., Bennike, 1985; J urmain, 1991; Smith, 1996; Wa lker, 1989; a nd several papers in Owsley a nd J a ntz, 1994). CONCLUSIONS
This review of fracture types and of the proximate and ultimate causes of injury indicates t ha t sta nda rdized descriptive protocols can be us ed to improve paleopat hological anal ysi s and interpretation of trauma. These pr otocols should include identifi cation of t he skeleta l element(s) involved; the location of the injury; its appearance; and any evidence for complications of the injury. These descriptions th us serve a s a basis for inferences about the mechanism of injury, from which social, cultural, or environmenta l associations may then be examined. Attention to skeletal patterns, biological factors such a s a ge and sex, and varia bles such as physical and sociocultural environments also improves the dia gnostic accura cy of an y interpretat ions. In part icular, the trauma tic
effects of violence ma y be difficult t o distinguish from those of high-risk activities or occupations solely on the basis of skeletal e vi d en ce . A lt h o ug h t h e a c t u a l c a u s e of trauma displayed by an archaeological skeleton may remai n unkno wn, a quantifiable description based on specific t erminology, supplemented by photographic and radiographic images of the fracture, and clearly placed in a n ind ividual, popula tional, sociocultural, a nd physical context w ill enable others t o evalua te a resear cher’s interpretation of the injur y. ACKNOWLEDGMENTS
I thank Carolyn P rins, who selected a nd evaluated t he clinical case studies an d w ho conducted some of the library research; the University of Alberta Hospital, which provided the clinical fracture data; Margaret J udd, who provided helpful discussion and numerous references; Ben Lees, who supplied useful information on firearms; and Scott Haddow, who assisted with printing of photographs. I have also benefitted greatly from discussions with Robert J urma in a nd Lynne Kilgore and from suggestions made by the reviewers of this paper. This research was supported by a gran t from the Soci al Sciences and Humanities Research Council of Canada . LITERATURE CITED Adams J C (1987) Outline of F ractures. Edinburgh: Longman. Aga rwa l RP (1980) P at tern of bone injuries in a h ill area (P aur i-Ga rhw al). J . India n Med. Assoc. 7 4: 6 5–66. Alexandersen V (1967) The evidence for injuries to the jaws. I n DE Brothwell a nd AT Sandison (eds.): Diseases in Antiquity. Springfield: CC Thomas, pp 623– 629. Altner P C, G ra na L , and G ordon M (1975)An experimenta l study on th e significan ce of muscle tissue interposition on fra cture healing. Clin. Orth op. 111: 2 69–273. Angel J L (1974) Patterns of fracture from Neolithic to modern times . Anth ropologia i K o¨z leme´n yek 18: 9–18. Angel J L, a nd C aldwell PC (1984) Death by stra ngulation: Aforensic anthropological case from Wilmington, D e la w a r e . I n TA R a t h b u n a n d J E B u i k st r a (e d s. ): Human Identification: Case Studies in Forensic Anthropology. Springfield: CC Thomas, pp 168–175. Apley AG, and Solomon L (1992) C oncise S ystem of Orthopaedics and Fractures. Oxford: B utterworths. Appleton W(1980) The battered woman syndrome. Ann. Emerg. Med. 9 : 8 4–91. Åstro¨m J , Ahlq vist S, B eertem a J , an d J o´n sson B (1987) Phy sical a ctivity in w omen susta ining fracture of the neck of the femur. J . Bone J t. Su rg. [Br.] 69 : 3 81–383. Ba rber HM (1973) Horse-play: S urvey of accidents wit h horses. Br. Med. J . 3 : 5 32–534.
168
YEARBOOK OF PHYSICAL ANTH ROPOLOGY
Bass WM (1987) Human Osteology. 3rd ed. Columbia: Missouri Archaeological S ociety. Bennike P (1985) Paleopathology of Danish Skeletons. Copenhagen: Akademisk Forlag. Berryman HE, and Haun SJ (1996) Applying forensic techniques to interpret cranial fracture patterns in an ar chaeological specimen. I nt. J . Osteoarchaeol. 6 : 2 –9. ¨ rnehult L (1991) Injuries Bjo¨r nst ig U, E rikss on A, an d O caused by animals. Injury 2 2: 2 95–298. Bla ir C (ed.) (1983) Pollar d’s H istory of Firea rms. Middlesex: Country Life. Bramblett CA (1967) Pathology in the Darajani baboon. Am. J . Phys. Anthropol. 2 6: 3 31–340. B r i s m a r B , a n d Tu n e r K (1 98 2) B a t t e r ed w o m en : A surgical problem. Acta Chir. Scan d. 148: 1 03–105. B u h r AJ , a n d C o ok e AM (1 95 9) F r a c t u r e p a t t e r n s. Lancet 1 : 5 31–536. Buikstra J E, and Ubelaker DH (eds.) (1994) Standa rds for Data Collection From Human Skeletal Remains. Fayett eville: Arkansa s Archaeological Survey, Research Series No. 44. Bullough P G, a nd B oachie-Adjei O (1988)Atlas of Spina l Diseases. P hiladelphia: J B Lippincott. B u r r el l L L , M a s s M C , a n d Va n G e r ve n D P (19 86) P at terns of long-bone fractur e in two Nubian cemeteries. Hum. E vol. 1 : 4 95–506. Burrows J H (1956) Fatigue infraction of the middle of t h e t i bi a i n b a l le t d a n c er s . J . B o n e J t . S u r g . [B r . ] 38: 83–94. Busch H M, Cogbill TH, La ndercasper J , a nd La ndercasper BO (1986) Blun t bovine and equ ine tra uma . J . Tra uma 26 : 5 59–560. Butler DF (1971) U.S. Firearms: 1776–1875. New York: Winchester. Car rero R, and Wayne M (1989) Chest t rauma . Emerg. Med. Clin. Nort Am. 7 : 3 89–418. Cogbill TH, S teenlage ES, Landercasper J , a nd S trutt P J (1 99 1) D e a t h a n d d i sa b i li t y f r om a g r i cu l t ur a l injuries in Wisconsin: A 12-year experience with 739 pat ients. J . Tra uma 31: 1632–1637. Cohen HR, Becker MH, and G enieser NB (1974)F at igue fractur e in Har e Krishna converts. N. Y. Sta te J . Med. 74: 1998–1999. Cybulski J S (1992)AGreenville Burial Ground. Archaeological Survey of C ana da Mercury Series No. 146. Hull: Canadian Museum of Civilization. Dast ugue J , a nd Gervais V (1992) Pa leopathologie du S q u e l e t t e H u m a i n . P a r i s : S o ci e´t e´ N o uv e ll e d e s E´ dit ions B oube´ e. Donaldson LJ , C ook A, a nd Thomson RG (1990) Incidence of fractur es in a geogra phically defined population. J . Epidemiol. Community Hea lth 4 4: 2 41–245. Dreier FG (1992)The palaeopathology of a finger dislocation. Int . J . Osteoarchaeol. 2 : 3 1–35. F i fe D , a n d B a r a n ci k J I (1 98 5) N o r t h ea s t e r n O h io trauma study III: Incidence of fractures. Ann. Emerg. Med. 3 : 2 44–248. Fife D, Ba ran cik J I, and Ch at terjee BF (1984)Northeast e r n O h i o t r a u m a s t u d y I I : I n j u r y r a t e s b y a g e , s e x, an d cause. Am. J . Public Hea lth 7 4: 4 73–478. F i le r J (1 99 2) H e a d i n ju r i es i n E g y p t a n d N u b ia : A comparison of skulls from Giza and Kerma. J . Egypt. Archaeol. 7 8: 2 81–285. Fisher EB, Kraus H, and Lewis VL (1990) Assaulted women: Maxillofacial injuries in rape and domestic violence. P last . Reconstr. Sur g. 8 6: 1 61–162. Fonseka S (1974)A study of wife-beating in t he Ca mberwell area. B r. J . Clin. Pra c. 2 8: 4 00–402. Ga rraw ay WM, Sta uffer RN, Kurla nd LT, an d O’Fallon WM (1979) Limb fra ctures in a defin ed population 1: Frequency a nd distribution. May o Clinic Pr oc. 54 : 7 01– 707.
[Vol. 40, 1997
Ga yford J J (1979) Ba ttered wives. Br. J . Hosp. Med. 22: 496–503. Gill GW (1994) Skeletal injuries of pioneers. In DW Owsley a nd RL J antz (eds.): S keletal Biology in t he Great Pla ins. Washington, DC: Smithsonian Inst itution Press, pp 159–172. Grauer AL, and Roberts CA (1996) Paleoepidemiology, h e a l in g , a n d p o s si bl e t r e a t m e n t o f t r a u m a i n t h e Medieval cemetery population of St. Helen-on-theWalls, York, Engla nd. Am. J . P hys. Anthropol. 100: 531– 544. Grimm H (1980) Sex differences in t he frequency of bone fra cture in historic and prehistoric times. In I Schwidezkef, B Chiarelli, and O Necrason (eds.): Physical Anthropology of European Populations. The Hague: Mouton, pp 347–349. Gurdjian ES (1975) Impact Head Injury. Sprinfield: CC Thomas. Gurdjian ES, Webster J E, and Lissner HR (1950) The mechanism of skull fractur e. Radiology 5 4: 3 14–339. Gustilo RB (1991) The Fracture Classification Manual. St. Louis: Mosby. Ha mperl H (1967) The osteological consequences of scalping. In DR Brothwell and AT Sandison (eds.): Diseases in Antiquity. S pringfield: CC Thomas, pp 630–634. Har kess J W, and Ramsey WC (1991) P rinciples of fractures a nd dislocations. In CARockwood, DP G reen, and RW Bucholz (eds.): Fractures in Adults. 3rd ed. P hiladelphia: J B Lippincott, pp 1–180. Ha rvey FH, an d J ones AM (1980) ‘‘Typical’’ basa l skull fractur e of both petrous bones: An unr eliable indicat or of head impa ct site. J . Forensic Sci. 2 5: 2 80–286. Heppenstall RB (1980) Delayed union, nonunion, and pseudoarthrosis. In RB Heppenstall (ed.): Fracture Treatment and Healing. Philadelphia: WB S aunders, pp 80–96. Hooper R (1969) Pa ttern s of Acute H ead Injury. London: Edward Arnold. Ingelmark BE (1939) The Skeletons. In B Thordeman (ed.): Arms and Armour from the Ba ttle of Wisby, 1361. St ockholm: Vitterh ets Historie Och Antikvitets Akademien , pp 149–209. J anssens PA (1970) P aleopathology. London: Curw en. J i m en e z S B (1 99 4) O cc up a t i on a l h a z a r d s i n 1 9t h ce n tu r y U p pe r C a n a d a . I n A H e r r in g a n d L C h a n (eds.): Str ength in Diversity. Toronto: Cana dian Scholar s, pp 345–364. J ohanss on C, Mellstro¨m D, Milsom I, Rundgr en Å, an d E k e lu n d P (1 99 1) P r e v a l en ce of f r a ct u r e s a m o n g 10,000 women from the 1900 to 1940 birth cohorts resident in G othenberg. Maturita s 1 4: 6 5–74. J o n es M W ( 19 90 ) A s t u d y o f t r a u m a i n a n Am i sh communit y. J . Tra uma 30 : 8 99–902. J o´n sson B , G a¨ rdsell P, J ohnell O, R edlund -J ohnell I, a nd Sernbo I (1992) Differences in fracture pattern between a n urba n a nd rura l population: A compara tive populat ion-based st udy in south ern Sw eden. Osteoporosis Int. 2 : 2 69–273. J urma in RD (1989) Tra uma , degenera tive disease, a nd other pat hologies a mong th e G ombe chimpanzees. Am. J . Phys. Anthropol. 8 0: 2 29–237. J urma in RD (1991) Pa leoepidemiology of t ra uma in a prehistoric centra l Ca lifornia population. In DJ Ortner and AC Aufderheide (eds.): Human Paleopathology: Current syntheses a nd future options. Washington, DC: Sm ithsonian I nstitu tion Press, pp 241–248. Karlstrom G, and Olerud S (1974)Fractures of the tibial shaft: A critical evaluation of treatment alternat ives. Clin. Orthop. 105: 82–115. K a u f m a n M H , Wh i t a k er D , a n d M cTa v i s h J (1 99 7) Differential dia gnosis of holes in t he calva rium: Appli-
Lovell]
TRAUMA ANALYSIS
cation of modern clinical data to palaeopathology. J . Archaeol. Sci. 2 4: 1 93–218. Kelley J O, and Angel J L (1987) Life stresses of slavery. Am. J . Phys. Anthropol. 7 4: 199–211. Kitchin ID (1948) Fa tigue fra ctures of the ulna . J . Bone J t. Surg. [Br.]. 3 0: 6 22–623. Kleinman PK (ed.) (1987) Diagnostic Imaging of Child Abuse. Baltimore: Williams and Wilkins. Knowelden J , Buh r AJ , a nd D unba r O (1964) Incidence of fractures in persons over 35 year s of age. B r. J . Prev. Soc. Med. 1 8: 1 30–141. Knowles AK (1983)Acute tra uma tic lesions. In G D H ar t (ed.): Disea se in Ancient Ma n. Toronto: C larke I rwin, pp 61–83. Larsen CS, H uyn HP, and McEwa n BG (1996) Death by gunshot: biocultural implications of trauma at Mission San Luis. In t. J . Osteoarcha eol. 6 : 4 2–50. Liston MA, a nd Ba ker B J (1996) Reconstructing t he massa cre at Fort William Henry, New York. In t. J . Osteoarcheaol. 6 : 2 8–41. Lovejoy C O, and Heiple KG (1981) The ana lysis of fractures in skeletal populations w ith a n example from the Libben Site, Ottowa County Ohio. Am. J . P hys. Anthropol. 5 5: 5 29–541. L ov el l N C (1 99 0) S k e le t a l a n d d e n t a l p a t h ol og y of free-ranging mountain gorillas. Am. J . Phys. Anthropol. 8 1: 3 99–412. Madh ock R, Melton LJ III , Atkinson E J , O’Fallon WM, and Lewallen DG (1993) Urban vs. rural increase in hip fractur e incidence. Acta Or thop. Scand . 5 8: 3 8–42. Mafa rt B-Y (1991) Apport de l’e´t ude des fractur es osseuses pour la connaissan ce des populations anciennes. Actes des 5e J ourn e´es Anthr opologiques, Dossier de Documentat ion Arche´ologique no. 14. P ar is: E´ ditions du CNRS. Man n RW, and Mur phy SP (1990)Regional Atlas of Bone Disease. Springfield: CC Thomas. Maples WR (1986) Trauma ana lysis by the forensic anthropologist. In KJ Reichs (ed.): Forensic Osteology. Springfield: CC Thomas, pp 218–228. Mensforth RP, a nd Latimer BM (1989) Ha mann-Todd collection aging studies: Osteoporosis fracture syndrome. Am. J . Phys. Anthropol. 8 0: 4 61–479. M er b s C F (1 98 9a ) Tr a u m a . I n M Y I s c a n a n d K AR Kennedy (eds): Reconstruction of Life From t he S keleton. New York: Ala n R. Liss, pp 161–189. Merbs CF (1989b) Spondylolysis: Its na ture a nd a nth ropological signifi cance. Int . J . Anthropol. 4 : 1 63–169. Merbs C F (1995) Incomplete spondylolysis a nd h ealing: A study of ancient Canadian Eskimo skeletons. Spine 20: 2328–2334. Merbs CF (1996) Spondylolysis and Spondylolisthesis: A cost of being an erect biped or a clever adaptation? Yrbk. P hys. Anthropol. 3 9: 2 01–228. Miller M, a nd Miller J H (1979) Orth opaedics an d Accidents. London: The English U niversities Press. Milner G R (1995)An osteologica l perspective on prehistoric wa rfa re. In LA Beck (ed.): Regional Approaches to Mortuary Analysis. New York: Plenum, pp 221–244. Nilsson R, L o¨fma n O, B erglund K , La rsson L, a nd Toss G (1991)I ncrea sed hip-fra cture incidence in the county of O¨ stergo¨t land, Sw eden, 1940–1986, w ith forecas ts up to t he yea r 2000: An epidemiological study. In t. J . Epidemiol. 2 0: 1 018–1024. Ortner D J , a nd Putschar WGJ (1981) Identification of Pa thological Conditions in H uman Skeletal Remains. Smithsonian Contributions to Anthropology No. 28. Washington, DC: Smithsonian Inst itution P ress. Ows ley DW, an d J ant z RL (eds.) (1994) Skelet a l Biology in the Grea t P lains: Migrat ion, Warfa re, Health, and Subsistence. Washington, D C: Smithsonian Institution Press.
169
Owsley D W, Ma nn RW, a nd Mur phy S P (1991) Injuries, s u rg ica l ca r e , a n d d is ea s e . I n S P f ei ff er a n d R F Williamson (eds.): Snake Hill: An Investigation of a Military Cemetery from th e War of 1812. Toronto: Dundurn, pp 198–226. Paton D (1984) Notes on Fractures. London: Churchill Livingstone. Perkins G (1958) Fractures and Dislocations. London: Athalone Pr ess of the University of London. Polson C, G ee DJ , a nd Knight B (1985) Essentials of Forens ic Medicine. 4th ed. Toronto: Per ga mon. Prince RL, K nuiman MW, a nd Gulland L (1993) Fracture prevalence in an Australian population. Aust. J . Public Health 1 7: 1 24–128. Ralis ZA(1986) Epidemics of fractures during periods of snow a nd ice. Br. Med. J . 293: 4 84. Roberts C (1991) Trauma and treatment in the British Isles in the Historic Period: A design for multidisciplinary research. In DJ Ortner and AC Aufderheide (eds.): H uman Pa leopathology: C urrent Syntheses and Future Options. Washington, DC: Smithsonian Inst itution P ress, pp 225–240. Roberts C, and Manchester K (1995) The Archaeology of Disease. Ithaca: Cornell University P ress. Sahlin Y (1990) Occurrence of fractures in a defined populat ion: A1-year study. In jury 2 1: 1 58–160. Schultz AH (1937) P roportions, varia bility and a symmetries of the long bones of the limbs an d th e clavicles in man a nd apes. Hum. Biol. 9 : 2 81–328. Schultz AH (1939) Notes on diseases and healed fractures of wild apes. Bull. Hist. Med. 7 : 5 71–582. Schultz RJ (1990) The Language of Fractures. 2nd ed. Ba ltimore: Williams an d Wilkins. Sevitt S (1981) Bone Repair and Fracture Healing in Man. Edinburgh: Longman. S h a h e e n M AE , B a r d AA, Al -K h u d a i r y N , K h a n F A, Mosalem A, an d Sa bet N (1990) P at terns of accidental f r a ct u r e s a n d d i sl oca t i on s i n S a u d i Ar a b i a . I n ju r y 21: 347–350. S h e ph e r d J P , S h a p la n d M , P e a r ce N X, a n d S c ul ly C (1990) Pattern, severity, and aetiology of injuries in victims of a ssa ult. J . R. S oc. Med. 8 3: 1 61–162. Shermis S (1984) Domestic violence in two skeletal populat ions. Ossa 9–11: 143–151. Smith MO (1996) ‘Parry’ fractures and female-directed interpersonal violence: Implications from t he late Archaic P eriod of w est Tennessee. Int. J . Osteoarchaeol. 6 : 8 4–91. Stedma n TL (1982) Stedman ’s Medical D ictiona ry, 24th Edn . Ba ltimore: William s a nd Wilkins. Steinbock RT (1976) Paleopathological Diagnosis and Interpretation. Springfield: CC Thomas. Stewart TD (1974) Nonunion of fractures in antiquity, with descriptions of five cases from the New World involving the forearm. Bull. N. Y. Acad. Sci. 50 : 8 76– 891. S w a n K G , a n d S w a n R C (19 89) G u n sh ot Wou n ds : Pa thophysiology a nd Man agement. C hicago: Year B ook Medical Publishers. Ubelaker DH , a nd Adams B J (1995) Differentiation of p er i m or t e m a n d p os t m or t e m t r a u m a u s in g t a p h onomic indicators. J . Forensic Sci. 4 0: 5 09–512. Unger J M, G entry LR, and Grossman J E (1990) Sphenoid fractures: P revalence, sites, a nd significance. Radiology 175: 1 75–180. Urist MR, Mazet R, and McLean FC (1954) The pathog e n es i s a n d t r e a t m e n t o f d e l a y e d u n i on a n d n o nunion. J . Bone J t. Surg. 3 6: 9 31–967. Wakely J (1996) Limits t o interpretation of skeleta l trauma : Two case studies from Medieval Abingdon, England. Int. J . Osteoarchaeol. 6 : 7 6–83. Waldron T (1996) Legalized t rauma . Int. J . Osteoarchaeol. 6 : 114–118.
