Reliability of measures used in radiographic evaluation of the adult hip.

OBJECTIVE: The reliability of radiographic measurements has been studied in pediatric hips, but less has been published on the adult hip, and none have examined the reliability of measurements for the location of the center of rotation (COR) of the hip joint. We have investigated the reliability of various radiographic variables with a focus on the COR. MATERIALS AND METHODS: The study was carried out on a standardized format for anterior-posterior radiographs of the pelvis. The measured variables were; (A) the distance from a sagittal reference line to the COR, (B) the distance from the sagittal reference line to the proximal end of the lateral cortical line of the femur, (C) the distance from the sagittal reference line to the medial rim of the acetabulum, (D) the distance from the horizontal reference line to the roof of the acetabulum, and (E) the distance from the horizontal reference line to the COR. One observer (JAB) conducted the measurements twice separated by a time interval of 45-60 days to assess intra-observer reliability, and the first measurements of JAB were compared to those performed by another observer (OR) to assess inter-observer reliability. RESULTS: Intraclass correlation coefficients were above 0.98 for all measurements, and the minimum and maximum values that statistically include 95% of the observer differences were all within -3 to +3 mm. CONCLUSIONS: These measurements proved to have high reliability and agreement of both within the same observer and between two observers. They should therefore be reproducible in a clinical setting.

No difference in mortality between cemented and uncemented hemiprosthesis for elderly patients with cervical hip fracture. A prospective randomized study on 334 patients over 75 years.

INTRODUCTION: Laboratory and human mechanical studies indicated that chemical substances in bone cement had toxic and prothrombotic effects. Impaction of cement added a mechanical trauma to the reaming and broaching procedure and contributed to a substantial local and systemic thrombin generation. Case reports and materials have indicated bone cement as the immediate trigger of cardiorespiratory and vascular dysfunction, occasionally fatal, and described as the bone cement implantation syndrome. In spite of this knowledge, bone cement has gained popularity and is widely used for prosthesis fixation, possibly due to a lack of clinical evidence supporting the basic science indicating bone cement as a mortality risk factor. METHOD: This is a prospective, randomized study comparing cemented and non cemented hemiprosthesis on patients suffering a dislocated cervical hip fracture. Perioperative characteristics and 1 year mortality differences between the groups were estimated. PATIENTS: Hundred and thirty-four patients over 75 years were enrolled from two hospitals in Norway. Average age was 84 years, 75 % were female and 60 % had symptomatic comorbidities. RESULTS: We find no difference in mortality between cemented and uncemented hemiprosthesis up to 1 year (HR 0.77, 95 % CI 0.51-1.18, p = 0.233). However, statistically significant reduced operation time and blood loss were found in the non-cemented group. (mean difference of 13 min, 95 % CI 4-22, p = 0.004 and 92 ml 95 % CI 3-181, p = 0.043, respectively). CONCLUSION: Installation of non-cemented hemiprostheses in elderly with hip fracture may have benefits perioperatively regarding operation time and bleeding, and do not seem to influence 1 year mortality relative to cemented implants.

The epidemiology of shoulder dislocations in Oslo.

There are few previous studies on the incidence of shoulder dislocation in the general population. The aim of the study was to report the incidence of acute shoulder dislocations in the capital of Norway (Oslo) in 2009. Patients of all ages living in Oslo, sustaining a dislocation of the glenohumeral joint, were identified using electronic diagnosis registers, patient protocols, radiological registers of the hospitals, and the Norwegian Patient Register (NPR). The overall incidence rate was 56.3 [95% confidence interval (CI) 50.2-62.4] per 100,000 person-years, with rates of 82.2 (95% CI 71.7-92.8) and 30.9 (95% CI 24.5-37.3) in men and women, respectively. The incidence of primary dislocations was 26.2 (95% CI 22.1-30.4). The overall incidence of shoulder dislocations in Oslo was higher than previously reported incidences. The incidence of primary dislocations was also higher than that in previously reported studies for the general population but it was close to the incidence reported in Malmø, Sweden.

Systemic and local cytokine kinetics in musculoskeletal injury: a prospective study in patients with ankylosing spondylitis.

OBJECTIVE: Back surgery in patients with ankylosing spondylitis is a major trauma in individuals with tissue inflammation and joint destruction along the spine; we used surgery in these patients as a model in the study of systemic and local cytokine profiles in complicated trauma situations. MATERIAL AND METHODS: Blood was sampled before, during and after surgery in 10 patients operated on with extending osteotomy of the lumbar spine. Samples of arterial blood and local wound blood were analysed for proinflammatory and anti-inflammatory cytokines. RESULTS: Surgery induced no significant changes in systemic values of TNFalpha and IL-1beta. There were significant increments in systemic values of IL-6, IL-8 and sTNF-R1. A systemic increase in values of IL-10 was only noticed after 24 h. There were increments in local values of TNFalpha at 24 h and in local values of IL-1beta, IL-6, Il-8 and IL-10 at both 4 and 24 h postoperatively. The local values were in general significantly higher than the systemic values. CONCLUSIONS: This study indicates that a major musculoskeletal trauma principally is followed by significant increases in systemic levels of IL-6 with only modest systemic reactions in TNFalpha and IL-1beta, even in patients with an inflammatory disease. However, there are in general significantly increased local levels of IL-1beta, IL-6, IL-8 and IL-10, and our conclusion is that systemic cytokine levels might not reflect local reactions.

Systemic and local cytokine kinetics after total hip replacement surgery.

BACKGROUND: Trauma induces local and subsequent systemic inflammatory reactions. Aberrant reactions can lead to a systemic inflammatory response syndrome, with a potentially lethal outcome. Our aim was to investigate the early local cytokine kinetic compared to systemic changes in a standardized surgical trauma. METHODS: In 7 patients with total hip replacement, drained blood samples and venous blood samples were taken 3 times within the first day after the operation. Interleukin (IL) release was assessed by a multiplex antibody bead kit and compared to pre-operative values. RESULTS: The major findings were significant increases in systemic levels of IL-6 and in local levels of IL-6, IL-8 and IL-1 receptor antagonist (IL-1Ra), and that the local levels of these cytokines were significantly higher than the systemic ones after surgery. Besides, there were only modest changes in local and systemic levels of tumour necrosis factor alpha, IL-1 beta, IL-2, IL-2Ra, IL-4, IL-5, IL-7, IL-10, IL-12, IL-13, IL-15 and IL-17. CONCLUSIONS: The acute sterile inflammation after major orthopaedic surgery is principally characterized by significantly increased local and systemic levels of IL-6 and significantly increased local levels of IL-8 and IL-1Ra. Furthermore, the concentrations are higher at the local site of injury. Hence, we conclude that systemic cytokine levels might not reflect local reactions.

Investigation of lipopolysaccharide receptor expression on human monocytes after major orthopaedic surgery.

BACKGROUND: The innate immune system is suppressed after major orthopaedic surgery, implicating a risk of septic complications. Whole-blood ex vivo testing with lipopolysaccharide (LPS) has shown a depression of the tumour necrosis factor alpha (TNF-alpha) production until 12 days postoperatively. As part of the innate immune system, the Toll-like receptors TLR2 and TLR4 recognize antigens from Gram-positive and Gram-negative bacteria, respectively. The receptors CD14 and CD11b are involved in the LPS receptor complex, whereas human lymphocyte antigen (HLA)-DR binds endotoxin peptides. It is still uncertain whether the expression of all these receptors changes after major surgery. METHODS: In 6 patients undergoing hip arthroplasty, we investigated three times the display of TLR4, TLR2, CD14, CD11b, and HLA-DR on monocytes by fluorescence-activated cell sorting and white blood cell counts during 12 days postoperatively. At the same time, the plasma levels of interleukin (IL)-1beta, IL-4, IL-6, IL-10, IL-13, and TNF-alpha were measured. RESULTS: There was no significant change in the expression of TLR4, CD14, CD11b, HLA-DR, and TLR2. Monocyte count and cytokine analysis did not differ from the ones pre-operatively taken. CONCLUSIONS: After aseptic orthopaedic surgery, there is no change in the display of the LPS receptor complex on monocytes. Other mechanisms have to be investigated to gain insight into the decrease of the TNF-alpha production capacity postoperatively.

Fibre optic Bragg grating sensors: an alternative method to strain gauges for measuring deformation in bone.

Strain gauges are currently the default method for measuring deformation in bone. Strain gauges are not well suited for in vivo measurements because of their size and because they are difficult to use in bone. They are also unsuitable for repeated measurements over time since they cannot be left in the patient. The optical Bragg grating fibres behave like selective filters of light. As a result the structure will transmit most wavelengths of light, but will reflect certain specific wavelengths. If the Bragg grating is strained along the fibre axis, the wavelength will shift, and this change represents a measure of strain. The optical fibres are very thin, no thicker than a standard surgical suture and are easy to adhere to bone by use of the FDA approved polymethyl-methacrylate (PMMA) as bonding adhesive. Since they are made of biocompatible silica porous bioglass ceramics, it should also be possible to leave the fibres in the patient between and after measurements. We have shown that fibre optic Bragg grating sensors can be used as a measurement tool for bone strain by performing measurements both on an acryl tube and on an extracted sample of human femur diaphysis. On either of them we used four fibre optic sensors and four strain gauges, interspersed at every 45 degrees around the circumference. The standard deviation of the measurements on the acrylic tube for each of the sensors, both optical fibres and strain gauges, varied from 1.0 to 5.2%. Every sensor, both optical fibre and strain gauge, correlated significantly with all of the rest at the 0.01 level with a Pearson correlation coefficient r ranging from 0.986 to 1.0. The linearity for all of the sensors versus load was excellent, the lowest linearity of the eight sensors was 0.996 as expressed by r(2) (coefficient of determination), with no significant difference in linearity between optical fibres and strain gauges. Bone is not an ideal isotropic material, and we found that the strain readings of the sensors in the bone study showed less linearity than in the acryl tube study. The correlations between all sensors, both optical fibres and strain gauges, were less strong in the bone sample than acrylic tube with a Pearson correlation coefficient on the bone sample ranging from 0.629 to 0.999 and with a standard deviation of the measurements varying from 3.1 to 31.5%. However, no significant differences between strain gauges and optic fibres were found, neither in the acrylic tube measurements nor the bone sample measurements. Optical Bragg grating fibres are therefore well suited for dynamic measurements of strain in bone in vitro and should also be suitable for use on bone in vivo.

Immune modulation after total hip surgery. A prospective ex vivo study.

BACKGROUND: Major trauma affects the immune system, and immunosuppression may render the patients susceptible to septic complications. The purpose of this study was to investigate lipopolysaccharide (LPS)-induced releases of tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) in whole blood in patients undergoing total hip replacement. METHODS: Seven patients (6 females) who underwent elective total hip replacement were included. Ex vivo LPS-induced releases of TNF-alpha and IL-6 were measured in a whole blood assay at days 1, 6, 9 and 12 postoperatively, using low- and high-dose LPS incubation. At the same time, white blood cell counts were analyzed. RESULTS: The amount of TNF-alpha release was significantly reduced at days 6 and 12. Compared to monocyte counts, TNF-alpha release was significantly decreased also at day 9 in low- and high-dose LPS stimulation. IL-6 in plasma was significantly increased at day 1 and normalized thereafter. There were no differences in LPS-induced IL-6 levels compared to the levels before surgery. CONCLUSION: This in vivo/ex vivo study shows a reduced capacity of whole blood to release LPS-induced TNF-alpha at day 6 through to day 12 after major orthopedic surgery. Attenuated TNF-alpha release may contribute to an increased postoperative susceptibility to gram-negative sepsis.

Training increases the in vivo strength of the lower leg: an experimental study in the rat.

The effect of training on bone strength has been investigated in rats. After 7 weeks of training, fracture strength of the tibia in vivo during muscle contraction and after resection was assessed. A group of 30 male rats 11 weeks old were randomized to exercise on a treadmill, sedentary (ordinary caging), and inactivity (right-sided patellar tendonectomy) groups. The training group ran on a treadmill with a 10% inclination for 1 h per day. After 4 weeks the animals in all groups were anesthetized and the right lower legs loaded in three-point ventral bending until fracture during electrically induced muscle contraction. The contralateral tibiae were tested correspondingly after resection. Ultimate bending moment, energy absorption to failure, bending stiffness, and deflection were assessed for the in vivo and the resected tibiae. The body weight gain was 37% higher in the sedentary and 57% higher in the inactive animals than in the training group (P < 0.05), indicating a physiologic effect of the training. In the dissected tibiae there were no significant group differences in any of the mechanical parameters, indicating that neither training nor inactivity changed the structural capacity of the tibiae per se. In contrast, there were significant differences between the in vivo tibiae. Ultimate bending moment was 12% higher in the training group than in the sedentary and inactive groups (P = 0.03). Energy absorption in the training group was 11 and 12% higher (not significant) than in the sedentary and inactivity groups, respectively. Bending stiffness was 7 and 17% higher in the training group compared to sedentary and inactivity groups (P = 0.018).(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of hormonal conditions and drugs on both muscle and bone strength can be assessed in a single rat test.

Strength of both muscles and bone are important for fracture prevention in osteoporotic individuals. Therefore, drugs that are preclinically tested in animals for preventing or treating osteoporosis, and reducing fracture risk, should not only be checked for their effects on bone strength, but also for those on muscle strength. We developed a rat model to measure both in the same animal, using a single test. The model is based on an in vivo, ventral three-point bending test of the lower leg (Nordsletten L. and Ekeland A. J Orthop Res 11:299-304; 1993). This model was developed to test the contribution of triceps surae muscle contraction to the strength of the tibia. We hypothesized that this same test can be applied to determine bone and muscle strength independently, in an absolute sense. To investigate this possibility, the muscle contribution to bone stresses was estimated from mechanical analyses, based on direct assessment of muscle strength in a separate test. Sixteen mature female Wistar rats were used, half of which were ovariectomized. After 12 weeks, the rats were tested in vivo in three-point bending of the right lower leg during muscle contraction, and then the isolated triceps surae muscle strength in the left lower leg was measured separately, in another model. The rats were then killed, and the left nude shafts were tested mechanically in three-point bending in vitro to determine structural strength of the bone alone. Ultimate external bending moments of the in vivo and in vitro tests, maximal muscle force, and geometrical parameters formed the basis for the analysis. While contracting, the triceps surae loads the tibia in axial compression and bending. We found that the axial compressive stress on the bone due to muscle contraction was less than 2.5% of the bending stress this produced. This indicates that muscle contribution to lower leg strength is due almost entirely to the bending moment it produces, counteracting the external moment put on the leg by the testing device. Thus, the difference between the in vivo (lower leg) and in vitro (nude tibia) failure bending moments is approximately equal to the maximal muscle bending moment. This information can be applied to test the effects of hormonal conditions and drugs on both muscle and bone strength independently, in a single rat test, using the aforementioned procedure.

Generation of procoagulant (thromboplastin) and plasminogen activator activities in peripheral blood monocytes after total hip replacement surgery. Effects of high doses of corticosteroids.

Development of thromboplastin (tpl) and plasminogen activator (PA) activity in monocytes and the effects of high doses of corticosteroids (HCD) on these activities were studied in patients undergoing a standardized surgical trauma. Twelve patients who received uncemented total hip prostheses were divided into a nonsteroid group (n = 6) and a steroid group (n = 6). We found no significant differences between the two patient groups regarding tpl or PA activities of peripheral blood mononuclear cells (PBM) isolated during the postoperative phase. However, in the nonsteroid group there was a tendency for increased expression of procoagulant activity and decreased fibrinolytic activity on the 1st postoperative day, favoring the formation of fibrin in the monocyte microenvironment. Further, PBM isolated on the 1st and 2nd day after surgery were significantly less capable of generating tpl activity on endotoxin stimulation than cells isolated preoperatively. This was not the case in the steroid group. These patients had also a tendency for decreased fibrinolysis at the end of the 1st postoperative week, indicating increased imbalance towards a more thrombotic stage after surgery.

Platelet release reaction and plasma catecholamines during total hip replacement. No effects of high doses of corticosteroids.

Platelet activation and catecholamine levels during surgery and the effects of corticosteroids on these reactions were examined in fourteen patients operated by uncemented total hip replacement (THR). Beta-thromboglobulin (BTG), released from alpha granules during platelet activation, catecholamines and cortisol were examined in plasma before operation and in the early postoperative period. The patients were randomly divided into two groups, a corticosteroid group where the patients were treated by high doses of methylprednisolone (HDC) and a nonsteroid group. BTG increased about 200% during the operation, and thereafter, decreased to slightly supranormal values after 24 hours. There were no significant differences between the two patient groups. Catecholamine levels were low, and there were only minor changes following surgery. Cortisol increased following THR in the nonsteroid group. A standardized muscle-skeletal trauma in the form of THR caused a significant increase in platelet activation as evaluated by BTG increase during the first 24 hours after the operation. Catecholamines did not seem to be of importance for this activation, neither were the levels of BTG influenced by HDC.

Vascular and structural changes in rat femora following nailing and intramedullary occlusion.

The purpose of this study was to evaluate biomechanical, structural, and blood flow changes of the femoral canal in rats 12 weeks after intramedullary reaming, nailing, or occlusion. In one group, reaming alone was performed. In a second group, reaming was followed by use of a tight-fit steel nail. In a third group, reaming was followed by use of inert silicone that totally plugged the medullary cavity. A fourth group served as the control. The acute mechanical and vascular effects caused by reaming and nailing were determined in a separate group. Reaming and nail insertion reduced blood flow in femoral bone to about one-third. Reaming reduced bending moment by approximately 40%, whereas bending rigidity was unchanged. After 12 weeks, the cortical bone blood flow was significantly increased in both the nailed and the silicone-plugged bones compared with the reamed and control groups. The bending moment and energy absorption in the silicone group were inferior to those of the other groups. There were no differences in either the external or internal diameter or the medullary and net bone areas. In the silicone group, both the number and the area of large pores (larger than 10 microm) significantly increased in comparison with the other groups; hence, bone porosity was increased. This increment was confined to large pores. It is concluded that medullary occlusion contributes to structural and blood flow changes in bone.

Deformation across the zone of callotasis during loading. radiostereometric analysis in a patient with achondroplasia.

The present study demonstrates that high-resolution radiostereometric analysis (RSA) can be used to assess global longitudinal compressive deformation across the callotasis zone during loading. In an achondroplastic patient operated with bifocal lengthening of the tibia by use of the Ilizarov external fixator, the axial compressive intersegmental strain in the proximal lengthening zone under a load of 71% of body weight was 7.7 mm. The proximal lengthening zone was 51.0 mm, and accordingly the overall linear strain across the callotasis was 15.1%. This large strain value found in distraction osteogenesis 6 weeks after end of distraction is not consistent with classical theory of the magnitude of micromotion needed for adequate stimulation of bone formation in fracture healing. The increased axial displacement did not stimulate bone healing and delayed union was observed. This one single observation does not allow for any conclusions to be drawn about the relationship of strain to fracture healing, but further and refined use of the RSA method will certainly improve our understanding of the role of axial strains in distraction osteogenesis.

Effects of glucose-insulin-potassium infusion on glycogen levels in cardiac and skeletal muscles in rats.

The effects of loading with glucose-insulin-potassium on the glycogen contents of the heart and peripheral musculature were studied in rats. One group of animals was given intravenous infusion of high concentrations of glucose containing insulin and potassium for 3 days. A control group received the same volume of saline. The hearts and rectus femoris muscles were then analysed for levels of glycogen. It was found that loading with glucose-insulin-potassium induced a decrease in cardiac stores of glycogen while there were no changes in glycogen stores of peripheral musculature as compared to the control animals.

Comparison of the modifying effects of somatostatin and propranolol on morphine-induced changes in glucose, glucagon and insulin levels in fed rats.

Twenty-eight rats in four different groups were used. Catheters were implanted in the carotid artery and jugular vein one week before the experiments were performed. The rats were trained to the experimental situation daily, and allowed food and water ad libitum. One group of rats was used to establish control values; a second was injected with morphine (10 mg/ml, 1 ml/kg); a third group got morphine injection (10 mg/ml, 1 ml/kg) combined with somatostatin infusion (0.01 mg/ml, 1 ml/h); and the fourth group was injected with morphine (10 mg/ml, 1 ml/kg) combined with propranolol (0.4 ml, 1 mg/ml). Blood samples for venous glucose and arterial insulin and glucagon were drawn 15 min after start of stimulation. Glucose, insulin and glucagon levels were significantly higher in morphine treated than in control rats. When morphine was combined with somatostatin, the increase in glucose, insulin and glucagon was significantly reduced. However, after the morphine and propranolol stimulation the increase in glucose and glucagon was significantly reduced, whereas the insulin levels were as high as when morphine was given alone. The combined reduction of both glucagon and glucose after somatostatin or propranolol treatment in morphine exposed rats, points to glucagon as a potential link between opioid stimulation and hyperglycemia. Beta-receptor stimulation seems to contribute to the glucagon but not to the insulin release after morphine.

Increments in glucose, glucagon and insulin after morphine in rats, and naloxone blocking of this effect.

In awake rats adapted to experimental conditions and allowed food ad libitum, hyperglycemia was induced by the administration of morphine 10 mg/kg through indwelling catheters in the external jugular vein. High glucose values were measured at 5, 15 and 25 min. Glucagon values were high at 5 and 15 min, and again at basal level at 25 min. Insulin was increased after morphine both at 5, 15 and 25 min, whereas somatostatin levels did not change after morphine. When morphine was administered together with naloxone after an initial 10 min period of naloxone administration, there was no increment in glucose, insulin or somatostatin values; neither at 5, 15 or 25 min. There was a remarkable glucagon decrease after naloxone and morphine remaining from 5 to 25 min. Then, one of the possible mechanisms for the hyperglycemic response after morphine may be an opioid effect on pancreas, stimulating glucagon and thereby causing hepatic glucose output.

Tissue factor antigen and activity in serum of postoperatively shed blood used for autologous transfusion.

Orthopaedic surgery involves extensive dissection of connective and richly vascularised tissues rich in tissue factor (TF). It was, therefore, of interest to quantify the amount of TF antigen and activity in postoperatively drained, unwashed wound blood collected for the purpose of autologous transfusion. In nine young patients subjected to surgery for idiopathic thoracic scoliosis, samples were drawn postoperatively from collected shed blood, a pulmonary artery catheter and a radial arterial cannula prior to, during and after reinfusion of shed blood (10 ml/kg body weight), and analysed for TF antigen and activity. Preoperative arterial control samples contained 128 pg/ml TF antigen compared with 40 pg/ml postoperatively. During reinfusion of drained blood, arterial TF concentration rose to 96 pg/ml and dropped to 64 pg/ml after infusion. Arterial and mixed venous blood did not differ significantly in TF levels. Serum from drained blood contained high concentrations of TF antigen (773 pg/ml), but no TF activity was detected. It is concluded that the high concentrations of TF antigen in serum from postoperatively drained blood collected for autologous transfusion are devoid of procoagulant activity. The TF antigen in plasma of drained blood is suggested to be a soluble, proteolysed TF-apoprotein or a TF complex inactivated by the TF pathway inhibitor (TFPI).

Fibrinogen, fibrin and its degradation products in drained blood after major orthopaedic surgery.

The aim of this study was to evaluate the fibrinogen enzymatic conversion in blood collected postoperatively from a surgical wound. Ten otherwise healthy patients (aged 11-28 years) in need of surgical treatment for thoracic scoliosis were included in the study. Arterial blood preoperatively and at wound closure were compared with samples of drained blood from the wound at closure and from a collection system for autologous transfusion 2.8 +/- 1.1 h later. There was a decrease in the fibrinogen content in arterial blood from 2.17 +/- 0.35 g/l to 1.23 +/- 0.42 g/l, which followed a 40% haemodilution estimated from the blood loss of 1.6 +/- 0.9 l during the operation. Drained blood contained high concentrations of D-dimer (85 +/- 53 mg/l from the wound and 121 +/- 47 mg/l from the collection system), but no clottable fibrinogen. The Western immunoblots all visualized the same patterns; in drained blood there were split-products mainly from cross-linked fibrin, in contrast to arterial blood which contained only normal fibrinogen. This indicates a strong fibrinolysis in the surgical wound after closure, with concentrations of fibrin degradation products that may impair local coagulation, and if infused, might interfere with general haemostasis.

Basic concepts of femoral neck anteversion: comparison of two definitions.

Femoral neck anteversion is determined by the neck, shaft and knee axes. It is commonly assumed that the neck axis is a straight line connecting the centre of the femoral head, the centre of the femoral neck and the axis of the femoral shaft. However, the line through the centre of the femoral head and neck crosses anterior to the axis of the femoral shaft. A distinction has therefore to be made between two principal different definitions of anteversion, denoted ANTEV 1 and ANTEV 2 in this paper, depending on the precise definition of the neck axis. Previous studies have indicated substantial differences between these definitions. The present study, using precise measuring methods, shows that the definitions have a systematic but small spread of differences. The large individual variation previously found between the two definitions is therefore caused mainly by the measuring methods and to a lesser extent by real anatomical variations.