Timing and mechanism of fractures in
human long bones and ribs
Willeke Langenhorst, 10145117
MSc in forensic science, University of Amsterdam
Research project thesis, 36 EC
January 2016 – July 2016
Presentation date: 25
thof August
Abstract
Determining the timing and trauma circumstances of skeletal trauma is a very important, but still a challenging task for forensic anthropologists. This study analysed the macroscopic fracture patterns of perimortem long bone (n=101) and rib (n=151) fractures and compared these to experimentally reproduced rib fractures of fresh and dry ribs. Additionally, the relations between the fracture pattern and other variables were statistically analysed. The five perimortem traits in long bones (layered breakage, wave lines, crushed margins, bone scales and flake/defects) previously described by Scheirs (2015) could be found in the new long bone specimens of this study and plastic deformation was added to the fracture pattern. The results of the statistical analyses showed that wave lines, plastic deformation and flake/defects were more likely to be found in the femur (p < 0.05). Moreover, plastic deformation was more likely to be found in younger individuals (p < 0.05).
The perimortem traits of long bones could not be found in the rib specimens. Nevertheless, six distinctive macroscopic traits were found in ribs that might provide information about the timing of trauma, fracture mechanism and/or trauma circumstances. These traits are peels, folds, differential fracture edges, incomplete fractures, plastic deformation and longitudinal lines. Peels, folds and plastic deformation might provide information about trauma timing. Folds and different fracture edges might provide information about the fracture mechanism. Statistical analyses showed that longitudinal lines, folds and incomplete fractures might provide information about the trauma circumstances and that age might have an influence on the presence of complete fractures, longitudinal lines and peels (p < 0.05). The new insights presented in this study might be valuable for forensic anthropologists in future trauma analysis.
Keywords: forensic anthropology, blunt force trauma, long bones, ribs, perimortem trauma, fracture pattern
Table of contents
1. Introduction...1
2. Material and methods...4
2.1 Sample...4
2.2 Experimental fracture reproduction...8
2.3 Defleshing and macroscopic examination...10
2.4 Statistical analyses...10
3. Results...11
3.1.1 Sample distribution long bones...11
3.1.2 Analysis of the fracture pattern in long bones...12
3.2.1 Sample distribution ribs...15
3.2.2 Analysis of the fracture pattern in ribs...16
4. Discussion...24
4.1 Fracture pattern in long bones...24
4.1.1 Trait distribution...24
4.1.2 Statistical analyses...25
4.2 Trauma analysis in ribs...26
4.2.1 Timing of trauma...27
4.2.2 Trauma circumstances...27
4.2.3 Manual versus mechanical CPR...28
4.2.4 Age and bone degeneration...28
4.2.5 Fracture mechanism...29
5. Conclusion...31
1. Introduction
One of the most important tasks of the forensic anthropologist in trauma analysis is to determine the time frame in which skeletal trauma was inflicted [1]. Determining whether trauma occurred before, during or after death (antemortem, perimortem and postmortem, respectively) is still one of the main challenges during the interpretation of skeletal remains [2, 3].
Antemortem trauma is defined as trauma that occurs prior to death and where signs of bone remodelling can be observed on the fracture edges [4, 5]. Evidence of the healing process can develop as early as one week after the injury. Between weeks one and three, the edges of the fracture will start to become remoulded and rounded, and by week six, bony calluses will begin to form [5, 6]. While this time frame is not important for the determination of the cause death, antemortem fractures are of great value in forensic anthropology, since they can contribute to the identification of skeletal remains [5]. Furthermore, antemortem trauma can be useful to document any history of abuse or accidental trauma[6].
Both peri- and postmortem trauma do not show any signs of healing, and the differentiation between perimortem trauma and postmortem taphonomic factors such as geological, biological or (un)intentional human alterations remains difficult [1–3]. The term perimortem is used in a different way between the fields of forensic pathology and forensic anthropology [5]. For forensic pathologists, the term ‘perimortem’ implies that an event occurred at or around the time of death. However, in forensic anthropology ‘perimortem’ implies that trauma was inflicted when the bone still had its organic components, causing it to behave as fresh bone [2, 5]. This means that when forensic anthropologists have to estimate the time of injury, the biomechanical conditions of the bone when it fractured (which are related to the bone composition) are more important than the ‘death event’ itself [5, 8].
Bone consists of both organic and inorganic material. Type 1 collagen comprises 90% of bone’s organic content [4]. The inorganic component of bone is a mineral composed of calcium phosphate called hydroxyapatite. Collagen occurs as long elastic fibres and gives bone its flexibility, whereas the mineral component gives bone its strength and rigidity. The combination of the organic and the mineral components makes the bone stiff and elastic. Dry bone lacks organic compounds which makes the bone stiff but brittle [8]. When fresh bone is submitted to a mechanical loading, it absorbs forces first through elastic deformation, after which the bone is able to return to its original shape. With greater force, the bone absorbs forces through plastic deformation, after which the bone will deform permanently, before it ultimately fails and fractures [8]. Dry bone lacks viscoelasticity and is unable to withstand as much elastic deformation as fresh bone. This causes dry bone to fracture immediately after the strength threshold is reached, and no plastic deformation is observed when dry bone is fractured. Due to the differences in fracture biomechanics between fresh and dry bones, the fracture patterns between these two different states are different [1–5, 9].
Fracturing a bone will result in different types of fractures, such transverse, oblique, butterfly, spiral, comminuted and buckle fractures [4]. Different fracture mechanisms produce these different fracture types, each type with its own characteristics. However, experience has taught us that most fractures have a combination of these fracture types and are therefore of a more complex nature. For a long time
will run along or perpendicular to the grain of the bone and the fracture edges are jagged. Fractures in fresh bone on the other hand will have more curved or V-shaped fracture lines and the fracture edges are smooth. Another often used trait is colour. If a fracture occurs postmortem, it results in a colour difference between the edge and the centre of a fracture due to different exposure to soil and other environmental factors. However, when no colour difference can be observed, this does not necessarily mean that the fracture can only be perimortem, since the colours could have been changed again due to exposure through the years [9].
Despite the literature about the differences between fractures in wet and dry bones, in practice there can be still doubt about the timing of a bone fracture [1–3, 7]. This is especially the case when the bones do not meet the extremely fresh or extremely dry conditions while fracturing that are described in literature. The existing literature gives primarily information about how to recognize a postmortem fracture pattern, but there is still very little information about a specific perimortem fracture pattern. From our knowledge, no existing studies have used perimortem human bone samples to optimize the estimation of the timing of bone trauma.
In order to overcome these problems, five new traits that characterize a perimortem fracture pattern in human long bones were recently reported by Scheirs (2015). According to her study, perimortem fractures are characterized by the presence of layered breakage, crushed margins, bone scales, wave lines and flakes with corresponding flake defects (Fig. 1) [10]. The presence of one or several of these traits is described as the perimortem fracture pattern. The main limitation of the study of Scheirs (2015) was the relatively small sample size (n=28), which allowed the author to only focus on the identification of the traits without statistically analysing the fracture pattern distribution.
Moreover, the study of Scheirs (2015) has focused on long bones. There is little to no knowledge about the application of these perimortem traits in other types of bone that are commonly related to forensic cases, such as ribs [9, 11–16]. Ribs and long bones have a different architecture, which might result in different fracture patterns. Despite the fact that rib fractures are very common in forensic cases, only a few studies have been performed on the exact fracture mechanism of ribs [12, 13, 15]. Furthermore, the studies that have been performed are contradictory and no fracture pattern is reported to distinguish perimortem rib fractures from postmortem rib fractures [12, 15].
This study is a follow-up research of Scheirs (2015) and will provide further insight into the characteristics of the perimortem fracture pattern of blunt force trauma in long bones and ribs. Specifically, this study investigates if there is a relation between the perimortem pattern, the kind of fracture, type of bone, trauma mechanism and several anthropometrical variables. Additionally, this study examines whether the perimortem fracture pattern in long bones also can be found on ribs. It
Fig 1 – The five traits of the perimortem fracture pattern. (a) Crushed margins are small fractured pieces of bone still attach to the cortical surface at the margin of the compression side of a fracture. (b) Wave lines are smooth undulations that have a gentle slope and a rapid drop, resembling a ‘wave’. (c) A flake is a superficial loss of bone. The imprint that is left on the bone is introduced as a flake defect. (d) Layered breakage is a pattern in the cortical of a long bone with very characteristic horizontal layers in the compression side. (e) Bone scales resemble fish scales and are a form of plastic deformation that is tangible. These pictures were used with permission from the copyright owner [10].
2. Material and methods
2.1 Sample
This study used perimortem fractured long bones and ribs from forensic autopsies of the Institut de Medicina Legal i Ciències Forenses de Catalunya (IMLCFC). These fractures were removed for complementary medico-legal investigation and were defined as perimortem fracture samples. The cause of trauma and anthropometrical variables were noted in the forensic autopsy report.
For the long bone specimens, 43 trauma cases were analysed which provided a total of 101 fractures (Table 1). Trauma circumstances included falls from a height, pedestrian train collisions and car and motorcycle accidents. Falls from heights will be noted as falls, pedestrian train collisions as collisions with train or train collisions and car and motorcycle accidents as traffic accidents. From the 101 fractures, 28 specimens were from the collection of Scheirs (2015). Additionally, 73 new specimens were collected during this six month research period.
Rib specimens were obtained from 25 trauma cases which provided a total of 151 fractures (Table 2). The trauma circumstances of the rib fractures were cardiopulmonary resuscitation (CPR, either manual or mechanical1), falls or train collisions. For convenience, CPR is referred to as slow loading and falls
and train collisions are referred to as fast loading. However, it should be kept in mind that it is not only the speed that is different between these two types of trauma circumstances.
The perimortem fracture patterns of perimortem autopsy rib samples were compared with the fracture patterns of experimentally fractured ribs. The Medical Anatomy Department of the Universitat Autònoma de Barcelona (UAB) provided 35 healthy, unfractured, fresh and dry anterior-lateral samples of 4th-6th ribs from 10 adults that donated their bodies to science (Table 3). This study was
approved by the Ethic Commission of Human and Animal Experimental work (CEEAH) of the UAB, in order to comply with the ethical requirements. Before the experiments, the fresh ribs from UAB were defleshed in such a way that only the periosteum and a thin layer of flesh were still attached to the bone.
Table 1 - Overview of available perimortem long bone fractures per case.
Case Type of bone Sex Age circumstancesTrauma
L01 Femur (2x)Tibia (2x)
Fibula (2x) Male 22 Traffic accident
L02 HumerusUlna
Femur Male 20 Traffic accident
L03 HumerusUlna
Radius Male 56 Traffic accident
L04 Femur Male 50 Traffic accident
L05 Humerus Female 41 Fall
L06 Humerus Female 30 Fall
L07 TibiaFibula Male 73 Traffic accident
L08 Humerus Male 83 Fall
L09 Femur Male 77 Fall
L10 Humerus Male 39 Fall
L11 Femur Male 34 Fall
L12 Femur (2x) Male 25 Traffic accident
L13 Femur Male 49 Fall
L14 Femur Female 63 Fall
L15 Femur Male 85 Fall
L16 HumerusUlna Male 39 Fall
L17 Radius Female 52 Fall
L18 Tibia Female 21 Fall
L19 Femur Female 48 Fall
L20 TibiaFibula Male 40 Fall
L21 UlnaRadius
Femur (2x) Male 65
Collision with train
L22 Femur Male 75 Fall
L23 Femur Male 78 Collision with train
L24 HumerusFemur Female 51 Fall
L27 Humerus Female 70 Fall
L28 TibiaFibula Female 91 Fall
L29 HumerusUlna Male 39 Traffic accident
L30 Humerus Tibia (2x)
Fibula (2x) Male 52
Collision with train
L31 Ulna Male 40 Collision with train
L32 Tibia Female 64 Fall
L33 Humerus (2x) Ulna Femur (2x) Tibia Fibula
Male 21 Collision with train
L34 UlnaFemur
Tibia Male 50 Fall
L35 UlnaRadius Male 35 Traffic accident
L36 Humerus Female 64 Fall
L37 HumerusRadius
Ulna Male 40
Collision with train
L38 HumerusUlna Male 23 Collision with train
L39 Femur Female 55 Traffic accident
L40
Humerus Femur Tibia (2x) Fibula (2x)
Male 41 Collision with train
L41 Fibula Male 30 Collision with train
L42
Humerus Ulna (2x)
Radius (2x) Male 70 Collision with train
L43 HumerusTibia
Table 2 - Overview of available perimortem rib fractures per case.
Case Sex Age Trauma
circumstances
Fractures
R01 Female 52 Fall 2
R02 Male 62 CPR (mechanical) 4
R03 Male 42 Collision with train 10
R04 Male 49 CPR (manual) 7 R05 Female 70 Fall 8 R06 Male 87 CPR (manual) 3 R07 Male 72 CPR (mechanical) 2 R08 Male 52 CPR (mechanical) 1 R09 Male 75 CPR (mechanical) 5 R10 Male 74 CPR (mechanical) 8 R11 Male 63 CPR (manual) 3 R12 Male 28 CPR (mechanical) 2 R13 Male 59 CPR (manual) 3 R14 Male 75 CPR (mechanical) 2 R15 Male 54 CPR (manual) 5
R16 Male 52 Collision with train 5
R17 Male 62 Collision with train 7
R18 Female 62 CPR (manual) 7
R19 Male 82 CPR (mechanical) 5
R20 Male 40 Collision with train 20
R21 Male 21 Collision with train 12
R22 Male 76 CPR (mechanical) 4
R23 Male 50 Fall 3
R24 Male 40 Collision with train 8
R25 Male 30 Collision with train 12
Total 151
Table 3 - Overview of available ribs from UAB used for the experimental fracture reproduction.
Case Rib condition Ribs Sex Age Speed of impact
Exp_R1 Dry 6 Male 63 160 mm/min
Exp_R2 Fresh 3 Male 45 160 mm/min
Exp_R3 Fresh 4 Male 48 160 mm/min
Exp_R4 Fresh 3 Male 57 160 mm/min
Exp_R5 Fresh 4 Female 45 160 mm/min
Exp_R6 Fresh 4 Male 27 160 mm/min
Exp_R7 Fresh 4 Male 35 160 mm/min
Exp_R8 Fresh 4 Male 61 3 m/s
Exp_R9 Fresh 2 Female 94 3 m/s
2.2 Experimental fracture reproduction
The ribs from the UAB donors were used to experimentally reproduce fractures. In this study, three-point bending tests were conducted in an experimental and controlled setting to describe the fracture patterns of ribs fractured by a slow and fast loading and trauma inflicted on dry and fresh ribs.
A servo-hydraulic testing machine (EM2/20 MicroTest) from research group GRABI was used with a U10M/25kN loading cell to simulate a slow loading (Fig. 2a+c). The test was controlled through the SCM3000 program. This low speed bending test was performed on 6 dry and 22 fresh human ribs at a strain rate of 160 millimetre per minute (mm/min) (Table 3). In order to approach real conditions, ribs were placed with the loading cell on the exterior side of the ribs. However, 4 of the 22 fresh ribs were fractured with the loading cell on the interior side of the rib.
In order to reproduce rib fractures caused by a fast loading, a devised drop test machine was used (Fig. 2b+d). This high speed bending test was performed on 7 fresh ribs at a strain rate of 3 metre per second (m/s) (equal to 180 000 mm/min) with the loading cell on the exterior side of the rib (Table 3). A rubber coating was attached to the loading cells as an attempt to simulate soft tissue and to prevent direct contact of the metal loading cell on the ribs.
Fig. 2 - (a + c) Servo-hydraulic testing machine EM2/20 MicroTest, with the U10M/25kN loading cell and two
supports to create a three-point bending, from UPC. (b) Drop test machine with vertical metal guide with a linear bearing carriage that freely falls along the vertical guide. On the carriage, a rubber coated loading cell is assembled. (d) On the sample holder, defleshed fresh rib segments were placed, which were fixed with two vices.
2.3 Defleshing and macroscopic examination
The bones obtained from autopsies and the fresh rib fractures resulting from the fracture reproduction experiments were defleshed using a method currently used at IMLCFC. The bone samples were placed in a water and detergent solution (1 cup of commercial degreasing detergent in 5 l of water) that was heated up to 100 °C. After the water reached 100 °C, the temperature was lowered to 90 °C and the bones were cooked for 2-3 hours, depending on the amount of flesh that had to be removed. After cooking, the remaining flesh was removed, using chirurgical tools. If needed, the bones were cooked again in the same solution in order to be able to remove all the flesh. Afterwards, they were cleaned with water and left to dry. If the bones were very greasy, they were cooked in a 2.5% ammonia solution in order to remove the grease. This method exposes the osseous surfaces in order to make the bones suitable for skeletal examination. After the bones were made suitable for skeletal examination, they first were examined for macroscopic traits by means of examination by the naked eye. In order to establish the presence of wave lines, long bones were examined under a microscope with 40x magnification.
2.4 Statistical analyses
For the perimortem long bone and rib specimens, analyses were carried out in order to establish whether there were statistically significant associations between the found traits, the kind of fracture, type of bone, trauma mechanism and anthropometrical variables. Binary logistic regression was used for analyses with dichotomous dependent variables. Multinomial logistic regression was used for analyses with categorical dependent variables. Odds ratios (ORs; with 95% confidence interval (CIs)) were calculated using the category with the lowest or highest percentage of trait appearance as a reference. The ORs were adjusted for sex, age (with young adults (20-40 years old) as reference) and the fact that in some cases more samples were obtained from the same individual. Moreover, for the perimortem rib samples the percentages of the traits for each individual case were calculated and were compared between groups of trauma caused by slow loading and fast loading using a Mann-Whitney test. P-values ≤ 0.05 were considered statistically significant. The statistical analyses were performed in IBM SPSS statistics software version 23.
3. Results
3.1.1 Sample distribution long bones
The fracture patterns of 101 long bone fractures caused by different trauma circumstances were studied for macroscopic perimortem traits. Table 4 shows that there was a predominance of cases involving young adults, males and fall injuries. However, if all types of traffic accidents were combined, the number of these cases was close to the number of fall injuries. Most fractures were obtained from young adults, males and train collisions. From some cases, more than one long bone fracture could be obtained, since trauma often involved multiple bones. In the statistical analyses this has been corrected for.
A large majority of the long bone fractures was comminuted (n=62), and there were only a few samples that were spiral, oblique or butterfly fractures (2, 4 and 6, respectively) (Table 5). The femur was the most frequently found type of bone (n=24), closely followed by the humerus (n=23). The amount of ulnas, tibias and fibulas are relatively similar (14, 15 and 17, respectively) whereas there were only 8 radius specimens.
Table 4 - Overview of sample distribution of the long bone samples and the amount of cases and fractures
within the sample set.
Cases Fractures Age
Young adults (20-40 years) 17 42
Mature (41-60 years) 13 38
Senile (>60 years) 13 21
Sex FemaleMale 1231 1685
Trauma circumstances
Fall 22 32
Train collisions 13 47
Traffic accidents 8 22
Total 43 101
Table 5 – Overview of each type of bone and each type of fracture.
Humerus Ulna Radius Femur Tibia Fibula Total
Transverse 2 4 3 3 1 2 15 Oblique 2 1 0 1 0 0 4 Butterfly 2 1 0 1 1 1 6 Comminuted Butterfly 2 1 3 3 3 0 12 Comminuted 14 7 2 15 10 14 62 Spiral 1 0 0 1 0 0 2 Total 23 14 8 24 15 17 101
Layered breakage, flakes and flake defects, wave lines, bone scales and crushed margins were found in the samples of this study (n=73). Additionally, the presence of plastic deformation was analysed and included as a trait. Layered breakage and flakes and flake defects were found in more than 50% of the cases, namely 87.1% and 70.3%, respectively (Fig. 3). These perimortem traits are followed by plastic deformation (46.5%), wave lines (35.6%), bone scales (16.8%) and crushed margins (14.9%).
Laye red br eaka ge Flak e/defe ct Plastic defo rmati on Wav e lin es Bone scale s Crus hed m argin s 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
P
er
ce
n
ta
ge
o
f
sa
m
p
le
s
Fig. 3 - Percentage of traits found in the long bone samples in descending order.
In order to decide which type of bone should be used as a reference for each trait in the binary logistic regression analysis, the percentages of each trait per bone were calculated (Fig. 4). The bone with lowest or the highest percentage of occurrence was used as a reference. The radius had the lowest appearance of layered breakage with 75%. For wave lines and plastic deformation, the femur had the highest percentage of the appearance (58.3% and 66.7%, respectively), while for crushed margins, the femur had the lowest percentage of appearance (12.5%). For the flakes and flake defects, the fibula had by far the lowest percentage of appearance (52.9%). Lastly, for bone scales the ulna had the highest percentage of appearance with 28.6%.
Laye red br eaka ge Wav e lin es Crus hed m argin s Plas tic de form ation Flak e/defe ct Bone scale s 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Humerus Ulna Radius Femur Tibia Fibula
P
er
ce
n
ta
ge
s
of
a
p
p
ea
ra
n
ce
in
e
ac
h
b
on
e
Fig. 4 - Percentages of all the perimortem traits within each type of bone with the reference bones indicated with
an asterisk.
The results of the binary logistic regression showed several statistically significant relations between the type of bone and the presence of certain traits (Table 6). Firstly, femurs were statistically significant more likely to have wave lines compared to the radius, tibia and fibula. Secondly, femurs were statistically significant more likely to have plastic deformation compared to the humerus, tibia and fibula. Lastly, femurs were statistically significant more likely to have flake and flake defects compared to the fibula. For the remaining traits and types of bone no statistically significant relations could be found.
With respect to the relation between age and the perimortem traits, the results of the binary logistic regression showed that young adults were statistically more likely to have plastic deformation compared to mature individuals (OR = 0.264, 95% CI [0.089 – 0.788], p = 0.017). For the remaining traits no statistically significant relation could be found with age. With respect to the relation between the perimortem traits, sex, the type of fracture and trauma circumstances, no statistically significant associations could be found (p > 0.05).
The results of the multinomial logistic regression showed that there were no significant associations between the type of fracture, trauma circumstances and age (p > 0.05).
*
*
*
*
*
Table 6 - ORs, 95% CIs and corresponding p-values of the binary logistic regression analyses with respect to the
relation between the type of bone and the fracture pattern.
Type of bone OR 95% CI p-value Wave lines (reference: femur) Humerus 0.32 0.086 – 1.2 0.089 Radius 0.048 0.004 – 0.55 0.014a Ulna 0.25 0.054 – 1.2 0.078 Tibia 0.17 0.036-0.81 0.026a Fibula 0.097 0.019 – 0.50 0.005a Plastic deformation (reference: femur) Humerus 0.19 0.047 – 0.80 0.023a Radius 0.14 0.020 – 1.0 0.053 Ulna 0.21 0.042 – 1.1 0.058 Tibia 0.19 0.040 – 0.89 0.033a Fibula 0.12 0.024 – 0.58 0.008a Flake/defect (reference: fibula) Humerus 2.8 0.65 – 12.1 0.161 Radius 1.41 0.21 – 9.6 0.723 Ulna 3.1 0.56 – 17.7 0.193 Femur 4.4 1.1 – 19.5 0.050a Tibia 4.6 0.84 – 25.6 0.078 Crushed margins (reference: femur) Humerus 0.76 0.12 – 4.7 0.764 Radius 1.2 0.085 – 17.1 0.890 Ulna 1.4 0.21 – 9.8 0.719 Tibia 0.83 0.11 – 6.3 0.855 Fibula 1.4 0.21 – 9.6 0.713 Layered breakage (reference: radius) Humerus 1.6 0.21 – 13.0 0.638 Ulna 2.1 0.20 – 22.7 0.531 Femur NAb NAb NAb Tibia 3.1 0.28 – 33.8 0.361 Fibula 2.9 0.28 – 30.8 0.369 Bone scales (reference: ulna) Humerus 0.69 0.13 – 3.6 0.658 Radius 0.52 0.044 – 6.0 0.596 Femur 0.88 0.18 – 4.3 0.878 Tibia 0.48 0.066 – 3.4 0.464 Fibula NAb NAb NAb
a Statistically significant results
b ORs of these types of bones could not be calculated for these traits due to the fact that bone scales
fractures were obtained from males, individuals of the senile age category and trauma caused by fast loading.
Table 7 - Overview of sample distribution of the rib samples and the amount of cases and fractures within the
sample set.
Cases Fractures Age
Young adults (20-40 years) 5 54
Mature (41-60 years) 8 36
Senile (>60 years) 12 61
Sex FemaleMale 223 13714
Trauma circumstances
Manual CPR 6 26
Mechanical CPR 9 32
Fast loading 10 93
Total 25 151
Table 8 - Occurrence of cases and fractures categorized by the different trauma circumstances and age. Age category Cases Fractures
Manual CPR 20-40 years 0 0 41-60 years 3 15 > 60 years 3 11 Mechanical CPR 20-40 years 1 2 41-60 years 1 1 > 60 years 7 29 Fast loading 20-40 years 4 52 41-60 years 4 20 > 60 years 2 21 Total 25 151
Table 8 lists the number of cases and fractures categorized by trauma circumstances and age. There were no cases of young adults with fractures caused by manual CPR and only 1 case of this age category that had rib fractures caused by mechanical CPR. Most cases and fractures caused by CPR were from senile individuals. The amount of cases of fast loading was nearly the same for young adults and mature individuals. A large proportion of the fractures was obtained from young individuals with trauma caused by fast loading.
specimens: peels, folds, longitudinal lines, differential fracture edges of the internal and external part of the rib, incomplete fractures and plastic deformation. The traits do not always have to be present in all fractures.
A peel is a structure in the fracture in which the cortex on one part of the rib fracture is ‘peeled off’ and is attached to the other part of the rib fracture (Fig. 5). This results in a peel defect on one part of the rib fracture and a thin, peel-like structure on the other part of the rib fracture.
Fig. 6 – Rib fracture with a fold indicated with the arrows.
Longitudinal lines are longitudinal cracks that follow the bone axis (Fig. 7).
Fig. 8 – Incomplete rib fracture with a disruption in the inner cortex.
The fracture edges can differ between the internal and external side of complete rib fractures (Fig. 10). Sometimes, one side of the fracture was very irregular and in some cases crushed, while the other side was more smooth and straight. This trait is independent of the morphology of the fracture and really only applies to the edges of a fracture. In all experimentally fractured samples in which this trait could be observed, the compression side showed an irregular and crushed edge and the tension side showed the more straight and smooth edge.
Fig. 10 – (A) irregular fracture edge on the external side and (B) smooth and straight fracture edge on the
internal side of a complete rib fracture.
Plastic deformation and different fracture edges on the internal and external side were found in more than 50% of the perimortem fractures, namely 57.6% and 51.7%, respectively. These traits are followed by peels (33.1%), longitudinal lines (21.2%), folds (16.6%) and incomplete fractures (15.6%) (Fig. 11). Plas tic de form ation Diffe rent f ractu re ed ges Peel Long itudi nal l ines Fold Inco mplet e frac tures 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
P
er
ce
n
ta
ge
o
f
sa
m
p
le
s
Table 9 gives an overview of the amount of observations and the percentages of the traits within each age category for each trauma circumstance in the perimortem samples obtained from autopsies. From the fractures that were caused by manual CPR, none were obtained from young adults. In rib fractures caused by manual CPR, no extreme differences could be observed in the presence of the traits between the mature and senile age group. In rib fractures caused by mechanical CPR, the described traits, except for plastic deformation, could only be found in senile individuals and no folds could be observed in any of the age categories. The largest proportion of the traits in rib fractures caused by fast loading could be found in young adults.
Table 10 gives an overview of the observations and percentages of all traits within their loading and age category for all the experimental rib samples. All dry samples show complete fractures and no
peels, folds or plastic deformation could be observed. The dry ribs did show longitudinal lines and differences in fracture edge appearance between the compression and tension side of the rib. The fresh ribs fractured by a slow loading were all incomplete fractures and only showed folds and plastic deformation. This also applied to the fresh ribs that were experimentally fractured with the loading cell on the interior part of the rib. The fresh ribs fractured with a fast loading all were completely fractured and showed, to a small extent, all the traits except for peels.
Table 9 - Overview of the amount of observations and the percentages of the traits within the age categories for each trauma circumstance in the perimortem samples obtained
from autopsies. For the total number of cases within each trauma circumstance and age category, please refer to Table 8.
Age category Incompletefractures n(%) Plastic deformation n(%) Peel n(%) Fold n(%) Longitudinal lines n(%) Interior/exterior different appearance n(%) Manual CPR 20-40 years NA NA NA NA NA NA 41-60 years (26.7 %)4 (40.0 %)6 (0.0 %)0 (13.3 %)2 (0.0 %)0 (40 %)6 >61 years (9.1 %)1 3 (27.3 %) 3 (27.3 %) 1 (9.1 %) 1 (9.1 %) 5 (45 %) Mechanical CPR 20-40 years (100 %)2 1 (50.0 %) 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 41-60 years (100 %)1 (0.0 %)0 (0.0 %)0 (0.0 %)0 (0.0 %)0 (0.0 %)0 >61 years (10.3 %)3 14 (48.3 %) 5 (17.2 %) 0 (0.0 %) 2 (6.9 %) 16 (55.0 %) Fast loading 20-40 years (17.4 %)9 39 (75.0 %) 32 (61.5 %) 20 (38.5 %) 23 (44.2 %) 33 (63.0 %) 41-60 years (0.0 %)0 10 (50.0 %) 7 (35 %) 2 (10.0 %) 4 (20.0 %) 6 (30.0 %) >61 years (9.5 %)2 14 (66.7 %) 3 (14.3 %) 0 (0.0 %) 2 (9.5 %) 12 (28.6 %)
Table 10 - Overview of the observations and percentages of all traits within their loading category for all the experimental rib samples. Experimental
loading Rib condition
Incomplete fractures n(%) Plastic deformation n(%) Peel n(%) Fold n(%) Longitudinal lines n(%) Interior/exterior different appearance (%) Slow Dry 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 4 (66.7 %) 5 (83.3 %) Fresh 22 (100 %) 16 (72.7 %) 0 (0.0 %) 19 (86.4 %) 0 (0.0 %) 0 (0.0 %) Fast Fresh 0 (0.0 %) 1 (14.3 %) 0 (0.0 %) 1 (14.3 %) 3 (42.9 %) 1 (14.3 %)
The results of the Mann-Whitney test showed statistically significant differences in the occurrence of longitudinal lines and folds when comparing trauma caused by slow loading and fast loading in the perimortem samples obtained from autopsies (Table 11). Rib fractures caused by fast loading are statistically significant more likely to get folds and longitudinal lines compared to rib fractures caused by slow loading. For the other traits, no significant differences could be found with the Mann-Whitney test.
Table 11 - Results of the Mann-Whitney test.
Mean% slow loading Mean% fast loading p-value
Peels 17.5 32.4 0.3
Folds 3.1 23.7 0.05a
Longitudinal lines 2.5 27.8 0.002a
Plastic deformation 40.3 61.7 0.09
Complete fractures 76.9 85.1 0.5
Different fracture edges 52.9 50.1 0.8
a Statistically significant results.
The results of the binary logistic regression in the analysis of the fracture pattern in ribs showed that young adults were statistically significant more likely to get peels compared to mature individuals (OR = 0.09, 95% CI [0.017-0.46], p = 0.004) and senile individuals (OR = 0.25, 95% CI [0.074-0.82], p = 0.02). Moreover, young adults were statistically significant more likely to get longitudinal lines compared to mature individuals (OR = 0.093, 95% CI [0.013 – 0.68], p = 0.02) and senile individuals (OR = 0.12, 95% CI [0.024 – 0.56], p = 0.007). The risk of having complete fractures was dependent on both the age of an individual and the trauma circumstances. Senile people were 9.6 times more likely to have complete rib fractures compared to young adults (95% CI [1.4 – 64.2], p = 0.02). Rib fractures that were caused by fast loading were 7.3 times more likely to be complete fractures compared to rib fractures that were caused by manual CPR (95% CI [1.1-48.7], p = 0.04). No statistically significant associations could be found between the presence of folds and plastic deformation when comparing groups categorized by trauma circumstances, age and sex (p > 0.05). An overview of the traits of the rib fracture pattern and their possible associations to other variables can be found be found in Table 12.
Table 12 - Overview of the traits of the fracture pattern in ribs and their possible implications.
Peel Fold Longitudinallines Incompletefractures deformationPlastic fracture edgesDifferent Timing Age Timing Circumstances Fracture mechanism Circumstances Age Circumstances Age Timing Fracture mechanism
4. Discussion
This study focussed on the timing and mechanism of blunt force trauma in human long bones and ribs. Despite the known traits that could make it possible to distinguish between peri- and postmortem trauma [3–5, 9, 12], trauma timing can still be doubtful [7]. In an effort to overcome this problem, Scheirs (2015) described a perimortem fracture pattern consisting of five distinctive macroscopic traits that might make it possible to assess the timing of skeletal trauma more accurate and reliable in the future [10]. These five traits are layered breakage, wave lines, crushed margins, flakes with corresponding defect and bone scales. Scheirs (2015) focused on the identification and description of the perimortem traits in long bones. This study took it a step further by increasing the sample size and investigating if there were associations between the traits, types of bone, types of fractures, trauma circumstances and anthropometrical variables. In addition, this study examined whether the perimortem fracture pattern described in long bones could also be found on ribs and if there were other (rib specific) traits that might provide information about trauma timing, mechanism or circumstances.
4.1 Fracture pattern in long bones
The five new perimortem traits described by Scheirs (2015) could also be found in the larger sample and no new perimortem traits could be observed in the long bone specimens. This implies that her description of a fracture pattern with newly found perimortem traits seems accurate and until now exhaustive. In order to optimize the perimortem fracture pattern in this study, the newly identified traits described by Scheirs (2015) were combined with a generally known perimortem trait: plastic deformation. This resulted in a perimortem fracture pattern with six traits. Fresh bone contains organic compounds, and the lack of organic compounds in dry bones makes dry bone fracture immediately after the strength threshold is reached without undergoing plastic deformation before fracturing. Therefore, plastic deformation can be considered as a reliable perimortem trait, since the presence of plastic deformation indicates that a bone is fractured in fresh, and thus perimortem conditions [1–5, 8]. The further implications of plastic deformation in forensic trauma analysis had not been described yet. In this study plastic deformation could be found more often in the long bones of young adults compared to the long bones of mature individuals. The finding that plastic deformation decreases with age might provide new insights in the exact changes of the biomechanical properties of long bones over age and the consequences on fracture biomechanics and the fracture pattern [18].
epiphysis is very thin, which could explain why layered breakage could not be observed in fractures in the epiphysis. Two samples did not show layered breakage, but were neither spiral fractures nor fractures in the epiphysis. These two samples had a very thin cortical bone in the diaphysis, which could explain, like fractures in the epiphysis, why no layered breakage could be found in these samples. Scheirs (2015) hypothesized that layered breakage does not require a certain fracture mechanism, except for the bone to be fresh while fracturing. The findings of this study support this hypothesis, but some additional fracture conditions are required in order to obtain layered breakage: the fracture needs to be in the diaphysis, should not be a spiral fracture and the cortical bone of the diaphysis should be thick enough in order to be able to exhibit layered breakage. The fact that layered breakage can be observed in the cortical area of the bone rather than the surface or the edges of a fracture might make the trait also more durable over a longer period of time compared to the other traits, which might make it possible to observe this trait after a longer time period of body decomposition. This all makes layered breakage a very important perimortem trait in long bones and it may be recommended to first look for the presence of layered breakage before the other perimortem traits in the estimation of the trauma timing during skeletal trauma examination.
Wave lines was the second to least frequently found trait in the sample of Scheirs (2015), but was at a higher position in the sample of this study. Scheirs (2015) already noticed that wave lines were present in butterfly, spiral and comminuted fractures with large pieces where the edges were smooth and long. In the newly collected samples, a large proportion of the fractures was comminuted with large pieces that had smooth and long edges. The increase in proportion of these types of fractures could explain why relatively more wave lines were found in this study.
Flakes are in both studies the second most frequently found trait and are found in more than half of the samples. Crushed margins, wave lines and bone scales are the least frequently found traits and it seems that these traits might require a very specific fracture mechanism to occur. Scheirs (2015) hypothesized that flakes, wave lines, crushed margins and bone scales require intra vitam (caused during life) conditions, caused by muscle contraction and surrounding flesh, to be present. This study has not been able to test this hypothesis and more research is required to obtain further knowledge about the exact mechanism that is needed for the occurrence of these specific traits. By doing this, the time gap between very fresh and very dry fractures could be shortened [1].
4.1.2 Statistical analyses
By increasing the sample size to n= 101 we were able to take the study of Scheirs (2015) a step further and statistically analyse the six macroscopic perimortem traits in long bones. The femur is the most frequently collected long bone in this study and is often involved in traffic accidents and falls [4, 9]. In this study, either the fibula or radius were for all traits the types of bone with the lowest percentage of occurrence. Wave lines, flakes with flake defects and plastic deformation were significantly more likely to be found in femurs compared to other types of long bones. For all these three traits, the fibula was one of the bones that was significantly less likely to have these traits compared to the femur, whereas for the radius this is only true for wave lines. The cortical bone of the fibula is very irregular, whereas the femur has a very straight, thick, round and regular cortical bone. Since all traits are present in or on the cortical bone, these differences in the cortex could explain why the traits are less frequently found in the fibula and more frequently found in the femur compared to other long bones.
would be expected that these two types of long bones get blunt force trauma in similar extents. Moreover, the anatomy of the ulna and the radius do not differ from each other in a high extent. Therefore it would be more likely that the fracture pattern of these two types of long bones would be more or less similar to each other. Nevertheless, the occurrence as well as the fracture pattern differed between these two types of bone. This might be explained by the differences in orientation in the arm and the geometry of the bones, which causes these two types of long bones to react different when subjected to blunt force trauma [9]. In order to rule out the possibility that the occurrence of the traits is influenced by the unequal distribution of the types of long bones, it is recommended to obtain a more homogeneous long bone sample in future studies in which all types of long bones and other variables are equally represented. This way, the relation between the occurrence of the traits and other variables can be assessed more reliable and accurate.
No statistically significant associations could be found between trauma circumstances, type of fracture and age. Since not all fractures were always recovered from the autopsies, it was unable to analyse whether there was a relation between the type of bone and trauma circumstances in this study. In general it is very difficult to make an interpretation about the influence of certain variables on bone fracture patterns, since each body reacts different under different trauma circumstances [9]. For example, individuals that have collided with a train could only suffer from internal bleeding or could be completely dismembered from the impact. This could be slightly influenced by anthropometrical variables, but external variables like the speed of impact and the shape of the vehicle seem to have a larger influence on the fracture pattern in these circumstances [4, 9]. To assess the exact influence of each variable, anthropometrical or external, it is recommended to obtain a homogeneous sample in which all variables are equally represented and to keep all but one variable equal during analysis. For example, if the effect of age on the fracture pattern needs to be assessed, the only variable that should vary between the samples is the age, and everything else that might have an influence on the fracture pattern should be kept equal. Nevertheless, it should be kept in mind that these kind of studies are dependent on the availability of human bone samples, which makes it a challenge to solve all limitations and to obtain a large sample with a homogeneous distribution in future studies.
4.2 Trauma analysis in ribs
Rib fractures are often involved in falls, traffic accidents and direct blows to the chest, which makes the rib cage important to examine for a forensic anthropologist [9, 11, 15, 16]. Interpreting rib fractures remains a challenge for forensic anthropologists, because the exact fracture mechanism of ribs is not known yet and so far no studies have described a perimortem fracture pattern in ribs [12,
causing a three point bending and that the impact speed and rib conditions have an influence on the fracture pattern. Therefore trauma was simulated on fresh and dry anterolateral individual rib samples that were fractured with three point bending tests with strain rates of 160 mm/min or 3 m/s. These were chosen to represent the different speeds of impact that caused blunt force trauma in the autopsy rib specimens.
4.2.1 Timing of trauma
Ribs and long bones have a differential architecture. Ribs are curved and have a relatively thin cortex compared to long bones [9, 12, 13, 15]. As a result of these differences in shape, the fracture biomechanics of these two types of bones are thought to be different from each other. All the traits described by Scheirs (2015) are very specifically present in or on the cortex of long bones. The differential fracture biomechanics and thin cortex of ribs could explain why the perimortem traits described by her could not be found in ribs.
Nevertheless, three distinctive traits were found in this study that might be useful in the estimation of timing in skeletal trauma analysis: peels, folds and plastic deformation. These three traits could not be found in any of the experimentally fractured dry ribs and folds and plastic deformation could only be observed in the experimentally fractured fresh ribs. Despite the fact that there were no dry ribs available for high speed experiments, it is hypothesised that the occurrence of these traits may be explained by the differences in bone composition between wet and dry bone, which influences the fracture biomechanics and thus the fracture patterns of these two states of bone. Considering the nature of these perimortem traits, it is likely that water and organic material are required to obtain these kind of structures in a rib fracture [8, 15, 19]. Peels could be observed in the autopsy samples, but could not be experimentally reproduced in any of the fresh or dry ribs. This suggests that peels are not only an indication of perimortem fracture conditions, but also could indicate that a rib was fractured in intra
vitam conditions. While the intra vitam traits in long bones are thought to be caused by both muscle
contraction and surrounding flesh [10], ribs do not really experience muscle contraction while fracturing [11, 13]. This might suggest that only surrounding flesh might have an influence on characteristics in rib fractures. This hypothesis deserves further research to obtain more in-depth knowledge about the exact mechanism that is needed for the occurrence of peels in rib fractures.
4.2.2 Trauma circumstances
Blunt force trauma in the autopsy rib samples was caused by either slow loading (manual CPR and mechanical CPR) or fast loading (falls, traffic accidents and train collisions). When comparing the mean percentages of the occurrence of the traits between perimortem fractures caused by these two types of loading, folds and longitudinal lines appeared to be more likely to be present in rib fractures caused by fast loading compared to slow loading. However, this significant association could not be found with the binary logistic regression analysis. The percentages of the occurrence of the traits were highly dependent on the amount of ribs recovered from an individual. Since these amounts differ per case, comparing percentages could give an erroneous interpretation. The results of the non-parametric test should therefore be interpreted with caution, especially when these results could not be found when using other statistical analysis methods. It has been described already that transverse rib
pattern with traits that could provide information about the trauma circumstances has not been described yet. Although the presence of longitudinal lines and folds might provide information about the trauma circumstances, further research should be performed in order to determine whether these traits are indeed reliable and useful in determining the trauma circumstances during skeletal trauma analysis.
The presence of complete fractures might provide information about the trauma circumstances as well, since our results show that it is more likely to have complete rib fractures when ribs are fractured by a fast loading compared to manual CPR. This might be due to the differences in speed and impact force between these two circumstances. The fact that all the experimental fresh ribs fractured with a slow loading were incomplete fractures and all the fresh ribs fractured with a fast loading were complete fractures supports this finding. The fact that a fracture is complete or incomplete is more a fracture characteristic rather than a trait, since a fracture is always either complete or incomplete. This characteristic is also not newly identified and has been described in the literature before [11, 13, 15, 20]. Nevertheless, in forensic medicine it is important to differentiate peri- or postmortem artefacts from injuries of possible legal significance [13–15]. Despite the fact that this fracture characteristic is generally known, linking complete or incomplete rib fractures to specific trauma circumstances could provide new insights to the influence of certain trauma circumstances in the forensic context.
4.2.3 Manual versus mechanical CPR
In this study, the rib fractures of slow loading were caused by two different types of CPR: manual and mechanical. Several studies have focused on the risk of obtaining rib fractures from manual CPR compared to mechanical CPR, but no study has focussed on the risk of having complete rib fractures with these two types of CPR [14, 20]. In case of a successful CPR, incomplete rib fractures are more favourable compared to complete rib fractures, since they have less severe repercussions [21]. In this study, no significant differences could be found in the risk of getting complete rib fractures between manual or mechanical CPR. Our finding could be of importance in the medical field in the deliberation whether to apply manual or mechanical CPR. Nevertheless, our finding suggests that there is no increased risk of getting complete rib fractures between manual CPR and mechanical CPR. In order to get a better understanding in the relation between CPR and rib fractures, future studies should look at both the risk of getting rib fractures and getting complete rib fractures to get more insights in what the risks are of applying manual CPR compared to mechanical CPR in cases of successful CPR.
whereas the study of Love and Symes (2004) did not find this difference in the risk of getting complete rib fractures between different age categories. All in all, the differences in fracture patterns in the different age categories might suggest that age and thus bone degeneration has an influence on fracture biomechanics. This change might have an influence on the fracture pattern. Nevertheless, it should be kept in mind that not every individual has the same bone degeneration rate with age, since this is also dependent on other factors like sex and health [18].
4.2.5 Fracture mechanism
Determining the exact fracture mechanism from a rib fracture pattern still remains a challenge for forensic anthropologists [12, 13]. In this study, two traits have been observed that might provide information about the fracture mechanism: folds and differences in fracture edges. Folds could not be observed in the experimentally fractured dry ribs, and in the experimentally fractured fresh ribs folds could only be observed on the compression side. Moreover, the dry ribs and the fresh ribs experimentally fractured with a fast loading showed differences in the fracture edges between the compression and tension side: the compression side showed an irregular surface, whereas the fracture edge on the tension side showed a smooth and regular surface. All fresh ribs that were experimentally fractured with a slow loading, including the ribs that were fractured with the loading cell on the interior part of the rib, were incomplete fractures with only a fracture in the cortex of the compression side. Since the cortex on the tension side was intact in all fresh specimens fractured with a slow loading, no distinction could be made between the fracture edges of the compression and tension side. The finding that folds were only found in the compression side of experimentally fractured ribs and the fracture edges of the compression and tension side of the experimentally fractured ribs show such distinctive structures might suggest that these traits could help in trauma analysis by indicating the direction of the impact. The smooth fracture edge at the tension side was earlier described by Kieser et al. (2013) in a study with 15 pig ribs. However, they did not describe the irregular compression side and could not find this particular trait in fractured dry bones. Moreover, no implications were provided by them on how this trait could help to interpret the fracture pattern in relation to the fracture mechanism during forensic trauma analysis [13]. The fact that this study uses a large sample of human bones instead of animal bone samples makes this study innovative and unique. We could find this trait in 78 of our human perimortem rib samples and provide an indication of the forensic implication of this trait, which takes the study of Kieser et al. (2013) a step further.
Folds and the differences in fracture edge appearance could also be observed in the samples obtained from autopsies. While in the experiments the force was applied directly to the anterolateral rib fragments, in the perimortem samples it could not be determined where the tension and compression exactly were during the traumatic event. Due to the shape of ribs and the rib cage, compression on the rib cage rarely causes fractures at the point at which the force directly is applied [9, 15]. For example, in case of manual CPR, force is applied on the sternum instead of directly on the rib, and this still causes ribs to fracture [14, 20]. Therefore, it is recommended to improve the experimental settings in a way that intra vitam fracture conditions could be reproduced, so real trauma circumstances can be simulated as accurate as possible. Furthermore, in the perimortem autopsy samples it appeared that it was always the external part of the rib fracture that was irregular and the internal part of the rib fracture that was smooth and regular. This might suggest that the architecture of the rib influences the presence of this trait instead of the direction of the impact. Further research should investigate if it is
Literature about the biomechanical properties of bone states that bone usually is stronger under compressive stress compared to tensile stress, which means that bone usually fractures at the tension side prior to the compression side [3, 5, 8]. However, in our study all experimentally fractured fresh rib samples, including the fresh ribs that have been fractured with the loading cell on the internal part of the rib, failed at the compression side prior to the tension side. Previous studies have already shown this, and incomplete fractures that failed at the compression side prior to the tension side were identified as ‘buckle fractures’ [12, 14, 15]. The finding that ribs also fractured first at compression when the loading cell was put on the interior part of the rib might suggest that ribs always fracture first at compression, regardless of the location of the compressive stress on the rib. In order to obtain further knowledge and to test this hypothesis, it is recommended for future research to use a larger sample size with improved the experimental settings in which real trauma conditions can be simulated.
5. Conclusion
Determining the timing and trauma circumstances of skeletal trauma is a very important, but still a challenging task for forensic anthropologists. This study provided further insight into the characteristics of the perimortem fracture pattern of blunt force trauma in long bones and ribs. The fact that this research is based on a large sample of human bones makes it innovative and unique.
The results of this study show that there is a fracture pattern with six distinctive macroscopic traits that might make it possible to distinguish perimortem from postmortem long bone fractures. These six perimortem traits are layered breakage, wave lines, flakes with corresponding defect, crushed margins, bone scales and plastic deformation. These traits are not equally distributed among the types of long bone: wave lines, flakes and plastic deformation are statistically more likely to be found in the femur compared to some other long bones. Furthermore, plastic deformation is more likely to be found in long bones of younger individuals compared to older individuals.
Traits found in long bones could not be found in ribs. On contrast, ribs show six distinct macroscopic traits that might provide information about rib trauma timing, circumstances and/or mechanism: plastic deformation, peels, folds, incomplete fractures, differential fracture edges and longitudinal lines. The presence of peels, folds and plastic deformation suggests that a rib was fractured in perimortem conditions. Moreover, peels, incomplete fractures and longitudinal lines might be related to the age of an individual and might provide further insight into the influence of aging on fracture biomechanics and the fracture pattern. Incomplete fractures and longitudinal lines might also provide information about the impact energy and thereby about the trauma circumstances, and differential fractures edges and folds might indicate the direction of impact and thereby provide information about the fracture mechanism.
The new insights on the fracture patterns of long bones and ribs that have been presented in this study might be of great value for forensic anthropologists in blunt force skeletal trauma analysis. Testing the influences of all variables on the fracture pattern accurately and reliably deserves further research. Improving the experimental setting, obtaining a more homogeneous sample and controlling the variables might allow us to test the functional implications of these traits in more depth in the future. 6. Acknowledgements
Special gratitude goes to my supervisor Assumpció Malgosa and co-supervisor Ignasi Galtés for sharing their knowledge, giving me advice about forensic anthropology and assisting me during this research project. Special thanks to Cristina Santos for her advice, sharing her knowledge and helping me with the statistical analyses. Many thanks to IMLCFC for providing samples from traumatic cases and to Marisa Ortega for providing fresh and dry ribs that were necessary for fracture reproduction. Special credits to David Sanchez and the rest of the research group of the UPC for their help in the rib fracture reproductions. My final thanks goes to Gabriel Font for his assistance in photo documentation and to Sarah Scheirs for her assistance and advice during the whole research project.
7. References
1. Wieberg DAM, Wescott DJ (2008) Estimating the timing of long bone fractures: Correlation between the postmortem interval, bone moisture content, and blunt force trauma fracture characteristics. J Forensic Sci 53:1028–1034
2. Symes SA, L’Abbé EN, Stull KE, et al. (2014) Chapter 13: Taphonomy and the Timing of Bone Fractures in Trauma Analysis. In: Pokines JT, Symes SA (eds) Manual of Forensic Taphonomy. CRC Press, Florida, pp 341–365
3. Christensen AM, Passalacqua N V, Bartelink EJ (2014) Forensic Anthropology: Current Methods and Practice. Academic Press, Oxford, p 447
4. Galloway A, Zephro L, Wedel VL (2014) Diagnostic Criteria for the Determination of Timing and Fracture Mechanism. In: Wedel VL, Galloway A (eds) Broken bones. Charles C Thomas, Springfield, pp 47-585
5. Dirkmaat DC (2012) A Companion to Forensic Anthropology. Wiley-Blackwell, Oxford, p 665 6. De Boer HH, van der Merwe AE, Hammer S, et al (2015) Assessing post-traumatic time
interval in human dry bone. Int J Osteoarchaeol 25:98–109
7. Cappella A, Amadasi A, Castoldi E, et al (2014) The difficult task of assessing perimortem and postmortem fractures on the skeleton: A blind text on 210 fractures of known origin. J Forensic Sci 59:1598–1601
8. Kieser J (2013) Biomechanics of Bone and Bony Trauma. In: Kieser J, Taylor M, Carr D (eds) Forensic Biomechanics. Wiley-Blackwell, Oxford, pp 35–70
9. Wedel VL, Galloway A (2014) Broken Bones. Charles C. Thomas, Springfield, p 479
10. Scheirs SF (2015) New insights in the analysis of blunt force trauma in human bones (master's thesis). University of Amsterdam. Accessed on January 18, 2016
12. Sirmali M, Türüt H, Topçu S, et al (2003) A comprehensive analysis of traumatic rib fractures: Morbidity, mortality and management. Eur J Cardio-thoracic Surg 24:133–138
13. Love JC, Symes SA (2004) Understanding Rib Fracture Patterns: Incomplete and Buckle Fractures. J Forensic Sci 49:1153–1158
14. Kieser JA, Weller S, Swain MV, et al (2013) Compressive rib fracture: Peri-mortem and post-mortem trauma patterns in a pig model. Leg Med 15:193–201
21. Smekal D, Lindgren E, Sandler H, et al (2014) CPR-related injuries after manual or mechanical chest compressions with the LUCASTM device: A multicentre study of victims after
unsuccessful resuscitation. Resuscitation 85:1708–1712
22. Nirula R, Mayberry JC (2010) Rib fracture fixation: Controversies and technical challenges. Am Surg 76:793–802
23. Rubertsson S, Silfverstolpe J, Rehn L, et al (2013) The Study Protocol for the LINC (LUCAS in Cardiac Arrest) Study: a study comparing conventional adult out-of-hospital
cardiopulmonary resuscitation with a concept with mechanical chest compressions and simultaneous defibrillation. Scand J Trauma Resusc Emerg Med 21:1-9
Supplementary material – Experimental fracture reproduction of two intact upper arms
One of the challenges in forensic anthropology is to find traits that are clearly related to intra vitam circumstances in order to shorten the time gap between very fresh and very dry fractures. The study of Scheirs (2015) experimentally fractured fresh long bones in order to find out whether the perimortem traits were also specific for intra vitam conditions. Crushed margins, bone scales and wave lines were thought to be related to axial compression. Therefore, it is hypothesized that these distinct traits are related to the presence of muscle contraction and the presence of surrounding flesh, and thus related to
intra vitam fracture conditions.
Originally the main study had one more aim: To improve the experimental fracturing of long bones in order to draw conclusions about the role of flesh and muscles in the found fracture patterns and to asses if there is a fracture pattern that is clearly related to intra vitam circumstances.
The Medical Anatomy Department of the Universitat Autònoma de Barcelona (UAB) provided 2 healthy, unfractured, fresh intact upper arms from 2 senile females of 84 and 91 years old that donated their bodies to science. This study was approved by the Ethic Commission of Human and Animal Experimental work (CEEAH) of the UAB, in order to comply with the ethical requirements. The samples were experimentally fractured using a servo-hydraulic testing machine (EM2/20 MicroTest) from research group GRABI with a U10M/25kN loading cell (Fig. 1). The test was controlled through the SCM3000 program. The arms were fractured with a strain rate of 160 mm/min and the experimental conditions of this study were similar to the experimental conditions of the study of Scheirs (2015) in order to be able to compare the results of both studies.
The humeri were fractured which resulted in one transverse and one oblique fracture. One of the bones only showed layered breakage, whereas the other bone did not show any perimortem trait at all. This could be explained by the fact that the age of the individuals was quite high and therefore the cortical bone was very thin. Moreover, only the role of the surrounding flesh could be assessed in this experiment. The main limitations of this experiment were the small sample size with only two humerus specimens of individuals with a high age and the slow speed of the testing machine, which did not simulate realistic trauma circumstances. In the future, we would like to perform these experimental fracture reproductions on a larger sample with different types of long bones in which all anthropometric variables are equally distributed. Furthermore, we would like to develop an experimental setup that simulates realistic trauma circumstances to monitor the influence of muscle contraction and surrounding flesh in the fracture pattern as accurate as possible. By doing this, the perimortem time frame between very fresh and very dry bone fractures could be reduced, so the time of injury could be estimated more accurately.
