THE ANTHROPOMETRY OF BODY ACTION* By Wilfrid Taylor Dempster Defiartment of Anatomy, University of Michigan, Ann Arbor, Mich. Conventional anthropometry of the adult has been limited largely to meas- urements that may be made with the anthropometer, the tape measure, and the calipers. The dimensions are linear, and they relate to the body surface- heights of body landmarks, widths, circumferences, and spans between sur- face points. To be duplicable, the measurements must he made in the same manner on different individuals, and the subjects must assume a stereotyped static posture, usually standing, during measurement. Supplementary and equally stereotyped postures such as sitting, supine, arms outstretched, etc., may also be used. The measurements, of course, allow comparisons between individuals and between population groups. They provide information on the relative magnitude and variability of individuals, hut they are in no way functional measurements. If a dynamic anthropometry is to be developed that can cope with problems of movement and postural change, new methods must be sought, and the parameters of measurement must be extended to include angles, velocity, accel- eration, rhythmic patterns, space envelopes, and force. Body activity is in- finitely variable, and it is essential that there be a theoretical framework so that relevant and secondary features may be evaluated correctly. Similarly, it is important that the conditioning factors and inbuilt limitations of the body machinery he also understood. Simplifications and abstract concepts are neces- sary, but it is important that these should not be oversimplified relative to the specific problem at hand. Studies based on dynamic measurements should contribute to human com- fort, efficiency, convenience, and safety. One may envision the application of work-space information to the industrial worker, to the school room, to vehicles and machinery, and to military problems. There should be contributions to furniture design and to the architecture of such work spaces as kitchens and bathrooms. In addition, a better understanding of dynamic actions of the body mechanism should provide help in the coaching of athletic performance, in the design of work clothing, including shoes and gloves, in personnel selec- tion, and in the designing of prostheses. Braune and Fischer6 were pioneers in this field, and more recent work has been done by Lay and Fisher:6 Hooton,2l Randall ef al.," Chapanis et aE.,B Smith and his group,%M cFarland et aZ.,% and Dempster.l' Bone Functions in Body Mechanics Bones have many functions: they are reservoirs of calcium, phosphorus, and carbonate; they house the bone marrow, which produces the red blood cells, leucocytes, and platelets of the blood; they provide processes and expansions ' A study conducted under contracts with the Wright Air Development Center (Section of Anthropometry) Wright Patterson Air Force Base, Dayton, Ohio, and with the Unitcd States Office of NavalResearch, Washington, D.C. 559 56 0 Annals New York Academy of Sciences for muscle attachment and leverage; they grow and mature, and their form is responsive to diet, to hormones, and to persistent and intermittent forces. These features, however, may be ignored here as irrelevant to the mechanical essential of rigidity. Forces may be transmitted along or across the rigid sub- stance of bone. Bone has a tensile strength and modulus of elasticity in ten- sion like walnut or hickory wood. It is as strong as concrete in compression, and its compressive modulus of elasticity is comparable also.I* Within the range of ordinary behavior, bone may be considered as a com- petent rigid material. It spans the articulations and separates the joints by fixed distances. Moreover, the ends of the bones, where adjacent bones attach, are molded into specific configurations (FIGURE 1) that permit unique types and directions of movement. Bones form the essential levers of the body system. Although tensile or compressive stresses are transmitted along the substance of the bony material, the effective levers for rotatory actions follow straight lines between one joint center and the next in sequence. In the femur or radius, such straight lines between joint centers actually fall in part outside the limits of the curving shapes of the bony substance. Where both ends of a bone have rounded (i.e., convex) articulations, the line goes from center to center within the length of the bone. In certain locations such as the forearm and leg, the lines span from a joint center located in the more proximal bone, ie., in the humeral or femoral condyles, through the whole bony support of the segment in question to a center in the adjacent distal bone in the hand or foot, that is, through a chain of three bones. These spanning distances between adjacent joint centers are the func- tional equivalent of the “links” of the engineer. The Body Link Sysfem The engineer commonly regards links as parts of a two-dimensional system that might be diagrammed on paper. The articulating parts overlap and are joined by pins, which act as axes of rotation (sliding-plane and helical joints sometimes found in machinery may be ignored here as special classes that are of little concern in the body system). The link is a line of constant length that spans the pin axes. It is a functional dimension dependent upon the rigidity of the material that separates the centers of rotation. The metal parts of a machine per se are not links. A link is merely a line-an abstraction-that may execute only those angular displacements allowed by the joint elements. In the machinery of the engineer, the links move in relation to a framework, and this framework becomes, in reality, an additional link that interconnects the others. In order to transmit power, the linkages of machinery must form a closed system in which the motions of one link have determinate relations to every other link in the system. The closed system assures that forces are trans- mitted in positive predetermined ways and in no other.32 The study of the relative motions of the links of machinery forms the subject matter of kine- matics, a powerful tool for the design and improvement of mechanisms of all types. In engineering, a background of kinematics gives understanding. It should likewise have value in relation to body motions. Dempster : Anthropometry of Body Action 561 FIGURE1 . Left, human skeleton; right, plan of body links. The body links rarely overlap, and none is truly pin-centered. In the major limb joints, the contiguous articular surfaces slide relative to one another as the bones rotate about a center in one or the other of the bones. The surfaces, although pressed together by the variable forces of ligaments, muscle, and at- mospheric pressure, are ordinarily only partially congruent, since the curva- tures are not truly reciprocal. Nevertheless, the areas that do contact as the articular surfaces slide against one another either inhibit or permit free rotatory motions in particular directions. Most of the joints have a freedom for angu- 562 Annals New York Academy of Sciences lar motion in different directions, and power transmission is quite impossible without the accessory stabilization of joints by muscular action. Strictly pre- dictable, determinate motions of the linkages as found in closed-chain systems are found at very few joints, viz., radio-humero-ulnar, carpal, tarsal, and tho- racic joints. The open-chain (indeterminate) system, as FischerlS indicated, predominates in the body mechanism. ' Because of the basic differences between the links of the engineer and those of the body system, certain characteristics of the latter should be outlined. The body link is the straight axial line or core line that extends through a body segment. Its integrity depends upon the rigid substance of the bone, and it terminates in axes about which the adjacent links rotate. The skeletal ele- ments may be reduced to produce a stick figure like that of FIGURE 1. The anatomical specializations of bones for muscle attachment and joint articula- tions, important as they are for body functioning, may be ignored in consider- ations of simple link motions. The bone may flare out or bend away from the core line between adjacent joint centers. The femur, for instance, has its ec- centric greater trochanter, the scapula its blade, spine, and acromion, and the elbow its projecting olecranon process. These are important for the muscular control of link movements, but they are entirely secondary to the basic kine- matic system of the body segments. The complex structure of the pelvis with its visceral basin, its birth canal, and its wide areas for muscle attachment may be simplified to a single link though, in this instance, it is triangular, with articulations to both the right and left femoral links and to the vertebral chain of links. Although the thorax may be treated as a separate system of linkages for purposes of the study of the respiratory movements, it may, for gross postural purposes, be simplified to the vertebral links (perhaps even treated as a single link) and a bucket- handle type of upper sternocostal linkage bridging from the right to the left of the uppermost thoracic vertebra (FIGURE7 ). The latter permits linkage con- nection with the clavicular links of each side. Since the link concept has pertinence for the analysis of rotatory motions, it should be apparent that the type of analysis anticipated should determine the degree of reduction to be used. In FIGURE 1, certain obvious linkages have not been shown, as for instance finger and toe links, a forearm link between the radius and ulna, separate links between successive vertebrae, and linkages be- tween the ribs and sternum. The remaining links are probably adequate for most purposes of postural analysis. The system may be diagrammed on paper, or it may be treated as a three- dimensional stick figure. On paper, the angular motions of links may be dia- grammed with the aid of a pencil compass. The links can be laid out on frame- by-frame records from motion pictures of actual subjects or from stroboscopic photographs of body movements. The movements of parts are most easily analyzed as motions in a plane. Links are linear and their angular movements sweep over surfaces. If the latter are curved, or are not in the plane of refer- ence, they may be projected to suitable rectangular coordinates. Rotations about the longitudinal axes of the segments are permitted in the trunk and at the ball-and-socket joints of the hip and shoulder (glenohumeral Dempster : Anthropometry of Body Action 5 63 joint), at the shoulder girdle joints, at the forearm joints, and at the knee. These movements may be important functionally, and they can be analyzed further. In general, however, such movements are less gross than other rota- tions in the sense that they involve small moments of inertia. To this effect, the forearm (radio-ulnar joint system) may for many purposes be considered as a single link between the elbow and the wrist (FIGURE 1). A Plan of Body Linkages The “link” man of FIGURE 1 is presented as a simplified kinematic system subject to further extensions as required for specific problems. The 25 verte- bral links above the sacrum have been reduced to 3 links: cervical, thoracic, and lumbar. The head link articulates with the cervical. The triangular linkage of the pelvis connects with both the lumbar link and the two thigh (or femoral) links. The sternocostal bucket-handle linkage articulates with the clavicular links and these, in turn, with the scapular links. The latter linkage represents a straight line between the mean center of scapula rotation, relative to the clavicle (on the acromioclavicular and coracoclavicular joint system) and the mean center of glenohumeral rotation. Next comes the hu- meral link, the forearm link, and the hand link. In the lower limb, the chain of links consists of a femoral link, a shank link, and a foot link. The hand, foot, and head are end links beyond terminal joints. At the hand, the link might be variably referred to as the linkage between the wrist center and a functional contact point at the heel of the hand, the knuckles, the ball of the thumb, or the finger tip. The foot link might span the ankle center and the heel, or the ankle and the ball of the foot or toes, etc. In general, however, where free movements of the end members are concerned, the parts may be regarded as cantilever systems, and the link span may be regarded as the straight-line distance from the terminal joint to the center of gravity of the part. The center of gravity of the head lies at the midsagittal plane on a line that interconnects points anterior to the tragi of the ears. The head link extends from this point to the occipitocervical junction. The hand link passes from the wrist center to the “V” between the radial longitudinal and proximal trans- verse palmar creases in line with the third finger. For the foot, the link termi- nates halfway down along a line from the ankle center to the ball of the foot in the plane of the second toe. Mean link dimensions may be visualized as beginning and ending in functional joint centers (or centers of gravity of terminal members). Since these end points are not structural landmarks, the link lengths cannot be measured di- rectly on the static subject. The systematic manipulation of joints permits a fair approximation for the location of mean joint-center positions. More elaborate methods for locating joint centers, based upon cadaver joint prepa- rations and involving geometrical techniques (vide infra), measurements of bone dimensions, including articular curvatures and extrapolations for specific bone lengths, are possible (TABLE 1). But these are not truly routine anthro- pometric procedures. Fortunately, the linear dimensions are less significant than angular measurements of link movement. 5 64 Annals New York Academy of Sciences The link figure presents the essentials of the dynamic man. The links are the basic units to which the angular motions of the body parts may be referred in the multiplicity of body postures and motions. A dynamic anthropometry that is not consistent with their functioning cannot claim more than limited empirical value. Links and Dynamic Aspeds of the Body 'The body is predominately an open-chain system of links. The links rotate about joint centers and exhibit a variety of angular positions consistent with the range of motion permitted by the joints before restraining mechanisms im- pose limits. An open-chain system cannot produce determinate or strictly predictable motions. The end member may have an infinite number of posi- tions relative to the trunk allowed by the cumulative range of the more proxi- mal joints. There are advantages in such a system. A part may be placed at a varicty of point positions in space. Reach becomcs a meaningful function, and bal- ancing reflexes become an essential in the antigravity mechanism of postural maintenance. The body is permitted such motions as kicking, walking, throw- ing, and the use of tools. Muscular action may stabilize certain motions mo- mentarily at one or another of a chain of joints. Such stability allows the body to concentrate the prime motion of other muscles on specific joints and on speci- fic planes of action. By this mechanism, force applications, accurate throwing, the use of tools, writing, and other animal and human activities become pos- sible. The fingers of the two hands may be interlocked to interconnect thc right and upper limb links; the legs may be crossed for seated stability; the arms may be crossed or placed on the hips. In such actions as these, temporary approx- imations to closed chains are effected. Between certain of the links concerned, however, there are soft (nonrigid) tissues. These, in turn, are pressed together to a variable extent, so that interposed skin surfaces make a friction contact. Clothing, gloves, or shoes next to the skin surface will more often decrease than increase the friction contact. Link chains may be cross-connected, as in crossing the knees (uiz., pelvis and right and left thighs) or in placing the hand on the same or opposite shoulder. To the extent that these temporary closed chains approximate a triangular linkage, there is a degree of stability imparted even without muscular actions, but this is still approximate because of the interposed soft tissues. The closer the links approximate a closed triagular, or pyramidal, pattern, the less muscles are calIed upon for stabilizing action at joints. One may recognize many rest positions involving this principle: crossed arms, hands in pockets, or such sitting positions as crossed knees, ankle on opposite knee, elbow on knee, or head in hand. The dog sitting on his haunches and other quadrupeds show different rest positions. As additional joints come into the linkage, the accessory ten- sions of muscles become more and more important for stability or for directing forces in specific ways. Temporary closed chains that involve cxtrinsic environmental objects also may be recognized. For instance, the two arms using a pruning shears are Dempster : Anthropometry of Body Action 565 such a linkage. When a seated person operates a foot pedal, the closed chain involves the shoe sole, the soft tissues of the foot, the foot and limb links, the pelvis, the nonrigid buttock tissues, the seat, the floor, and the pedal. In the operation of machinery by a hand lever, there is a complex linkage involving the machine links, its frame, the floor, and the body links from foot to hand. These temporary closed chains can be only approximations to determinate systems, since there are nonrigid tissues between certain links. In addition, the variability of possible motions at most of the joints requires muscle actions that vary with the style of neuromuscular control that an individual may employ. The open-chain system of the body insures unpredictability of action. It allows a variety of selective and discriminatory actions in the transmission of forces. The accuracy and power of control is based upon neuromuscular physiology, and psychological factors cannot be excluded. The links are functional dimensions rather than structural entities. They represent deep-lying relationships of the joints, and the standard anthropo- metry involving surface landmarks is quite unrelated to them. Angular mo- tions of the links may be studied, and measurements of the space requirements of the tissue-enclosed kinematic system of the body may be devised. The planning of such measurements, however, requires an understanding well be- yond structure alone. Joints in Relation to Link Movement Descriptions of the anatomical structure of the joint elements that inter- connect the link members are readily available. The configuration of bones and their cartilages at the contacting regions of the joints predetermines the directions in which links may rotate. When muscle tensions or the superim- posed weight of the body acts across a joint, the contiguous surfaces are pressed together. The shape of the contact or the ligamentous bindings either allow or eliminate certain types of angular movement. At the elbow, the forearm joints, or the ankle (tibio-talar joint), movements are restricted to specific guided paths of rotation from which there is no deviation. At the elbow and ankle, the movements are generally referred to as flexion- extension movements, but in each instance there is a sequence of slight contin- gent movements not limited to the sagittal plane. Nevertheless, the move- xents are guided from instant to instant, and angular movement can continue or reverse along only the guided path provided by the joint structure. In such joints, extraneous movements are prevented by the wedging of the bony con- figuration, and there is but one degree of freedom for rotational movement allowed. In FIGURE 2, the members of the elbow and forearm joints have but one degree of freedom of motion. At the wrist, the hand may flex and extend or move at a right angle in ab- duction or adduction, but free rotation about a longitudinal axis is virtually eliminated. At the knee, both flexion and extension movements and rota- tional movements of the shank are possible, but lateral movements do not occur except for a limited range where the ligaments are momentarily lax. In each instance, the predominant movements involve two degrees of freedom (FIGURE 2). At the hall-and-socket joints at the hip and shoulder (glenohumeral joint). 5 66 Annals New York Academy of Sciences FIGWRE2. Link figure with globographic repreientatioiis of tlie joint ranges of the extremity links. Dempster : Anthropometry of Body Action 567 there may be inward or outward rotation about the axis of the femoral or hu- meral link, or there may be movements in the sagittal plane (flexion-extension), or in the transverse plane (abduction-adduction). There is no restriction to the direction of rotation except at the limits of joint range. One may speak of three degrees of freedom of rotational movement. In the subtalar joints of the foot, at the claviscapular joint system (the functional equivalent of the composite coracoclavicular and acromioclavicular joints), and at the sterno- clavicular joint, one finds three degrees of freedom at each joint. The analysis might be carried further to the finger and toe joints and to other joints of the trunk, but this is sufficient for illustrative purposes. Fischer18 spoke of the degrees of freedom at joints, but this convenient approach has seldom been followed by writers in the United States. One may note the cumulative pattern of degrees of freedom in a proximo- distal sequence: upper lirnb-3/3/3/1/1/2, and lower limb-3/2/1/3 (FIGURE 2, however, further simplifies ankle and foot movements in one globe). There are 13 degrees of freedom between the trunk and hand, and 9 degrees between the trunk and foot sole. The sequence in the pattern of motion and in the degree of freedom of succeeding joints differs in the two limbs. Thus the elbow bends on a transverse axis, the forearm rotates on a longitudinal axis, and the wrist has (virtually) no axial rotation. The knee permits bending on a trans- verse axis and rotation on a Iongitudinal tibia1 axis, the ankle axis is approxi- mately transverse, and the lower joints in the foot have rotational axes in several directions ?O Range of Joint Movement The type and amplitude of movement differs from joint to joint. All move- ments are rotational rather than translatory, and they are to be measured in degrees or radians. In each instance, ligaments, bony stops (elbow), tissue bulk (knee flexion), or muscle stretch define limits to the range of movement. The functional ranges are based upon anatomical structure, and they are vari- able from individual to individual. Measurable differences in flexibility are to be expected, due to age, sex, disease, and possibly race and occupation. There is a literature concerned with the systematizing of terminology and with techniques for measurement for clinical purposes. Instruments and methods have been described, but authors in general have been content with working values rather than with detailed descriptions of movements applicable to specific populations. Exceptions to the general pattern are: Sinelnikoff and Grigoro~itsch,3G~l anville and Kreezer,l9 and Dempster," but even these are incomplete as full coverages of the different joints. The dynamic anthropometrist should be able, in the future, to provide im- proved measurement values for the various joints. It would be inadvisable, however, to plan a battery of measurements without a sound functional know- ledge based upon experience with cadaver joints. Various authors have quantitatively studied the range of movement of the principal joints of the limbs on joint-ligament preparations of cadavers. Such data have been presented by: Strasser and Ga~sman,~Fi'c k,17 Shiin~B,r~au~s: 568 Annals New York Academy of Sciences P f ~ h lS,t~ei~nd le~-M,~~~l lier,*v~on Lanz and M'achsm~th,a~n~d Dempster." Most of the work in the literature applies to the shoulder, hip, and knee joints. The Albert-Strasser globographic presentation has been commonly used (Albert,2 Strasser, and Gassmans7). In this method, one of the bony members of a joint was clamped fast with the approximate functional center of the joint at the center of a large protractor system. The free member was swung to its limits in each direction along the protractor, and the maximum range of move- ment of a needle point marker on the moving part was recorded. This was repeated also for other meridians at measured angles from the first. These ranges were marked off at appropriate places on a globe provided with a scale of parallels and meridians. Where a joint presented three degrees of freedom of motion, the amplitude of movement along the long axis of the moving mem- ber was simply marked at points on the surface of the globe. The circum- scribed area on the surface of the globe, corresponding with a solid angle origi- nating at the globe center, described a bounding surface within which the mov- ing part was unhindered. This shape has been called the excursion cone of the joint25 or joint sinus.18 FIGUR2E s hows such globes at the principal joints of the limbs. Each globe may be regarded as fixed to one of the joint members, with the excursion cone so placed that the other member is provided with a guide or template that permits only a circumscribed amplitude of joint movement. (Except for the knee arid foot, where the globe is affixed to the distal member, the proximal part of the joint at the other globes is fixed with the distal part moving within the excursion cone.) Where there are three degrees of freedom, rotations of the moving member about its own axis are possible, but the extent of this has not been specified in FIGURE 2. At the ball-and-socket joints, there are contingent restraints to movement at the limits of the excursion cone. When the free member is twisted Qn its long axis to the limit in one direction or the other, the excursion cone for the remaining degrees of freedom may be characteristically widened or narrowed at one point or another of its boundary. Conversely, as shown in FTGURE 3C, if the free memher (humerus) is moved to the maximum limit of the excursion cone, the bone rotates so that its front surface comes to face in different direc- tions. FIGUR3E( ' shows movement at the glenohumeral joint (three degrees of freedom). FIGURE3B shows six degrees of freedom (glenohumeral plus claviwapular joint systems), arid FTGTJRE 3A shows nine degrees of freedom (glenohumeral, claviscapular, and sternoclavicular). At A, there is no one fixed joint center; note the rangc of movement of the humeral head. At B, the humeral head likewise shuttles about. Although it is less clear at C, there is, as will be shown later, likewise no fixed center. This illustration underscores the complexity of shoulder movements in a way that is not clear in globographic representations by Fick,'7 Rraus: and von Lanz and Wa~hsmuth.~*C orrect movements have been shown in this jointGb y the use of X, Y,a nd Z coordinates. The cadaver measurements referred to, although they have analytical value, are inadequate as representing living motions, and they are based on too few analyses. Measurements that have been made upon living subjects have like- wise been inadequate for most joints, since they have ignored the contingent