OBSERVATIONS ON T H E NUTRIENT FORAMINA O F THE HUMAN RADIUS AND ULNA l S 2 S. S. SHUIiMAN Department of Anatomy, Dalhousie F,niversity, Halifax, N.B. SIX FIQUBES Although the topography and direction of the nutrient foramina on the long and elongated short bones are assumed to be constant, this is not invariably so in the human. Many anomalies have been reported in tetrapods (Hughes, ’52). The foramina on the human humerus and femur were studied in detail by Liitken ( ’50). It therefore seemed appropriate to continue the investigation on other human bones. This report is the first of a series based on a study of representative numbers of bones. MATERIALS AND METHODS The forearm bones from 82 cadavers (328 specimens of both radius and ulna) were examined for the number, the position, the direction and the incidence of asymmetry of the nutrient foramina. The recorded ages varied between 15 and 90 years. Arterial and venous foramina at the ends of the bones were ignored. Only well defined canals or foramina on the diaphysis were accepted. The position of a foramen was determined by measuring its distance from fixed points on the proximal and distal ends of the bone. The foraminal index was then calculated by using the formula I = (PE - F)/ TL X 100, where I = foraminal index. PE - F =distance * The major part of this investigation was done during 1954 in the Department of Anatomy, University of Cape Town. a Read by title at the Annual Meeting of the American Bssociation of Anatomists, Seattle, Washington, April 3, 1959. 685 686 S. S. SHULMAN from proximal end of the bone to nutrient foramen. T L = total length of the bone (Hughes, '52). The anatomical surface bearing the foramen was noted in each case. Foramina were classified as symmetrical if they were on the same anatomical surface with the difference between the two sites not exceeding 0.5 em. The degree of obliquity was assessed by measuring the visible portion of the canal in every instance. RESULTS A. Number of nutrient foraminla Radius. Of the 164 bones, the only exceptions to the general rule of one nutrient foramen were (a) three specimens (1.8%) which each presented two foramina, both foramina were on the proximal half of the diaphysis, the distance between the main and the proximal accessory foramen never exceeded 3 em. The duplication was found unilaterally only once on the right and twice on the left side (ages 48, 50 and 70 years). (b) Two specimens had no nutrient foramina (ages 50 and 90 years) ; secondary bone markings were not gross enough to suggest a causal relationship. Ulna. The exceptions to the general rule in these 164 bones, were (a) 14 specimens (8.6%) had duplication of the foramen and both were always on the proximal half of the diaphysis ; the accessory foramen was always proximal to the main one but never by more than 3 em. This duplication occurred bilaterally in two cadavers (ages 48 and 53 years) ; unilaterally in 10 cadavers, equally divided between the two sides (ages 23,23, 30,45, 50, 62, 65, 66, 69 and 70 years). (b) One left ulna had no nutrient foramen (age 71 years). B. Position of nutrient foramina They were all on the proximal half of the diaphysis. Assessing the site in terms of the proximal third of the shaft, the incidence within this area proved to be 19% and 38% for the ulna and radius respectively. Analysis of the foramina1 687 N U T R I E N T FORAMINA O F RADIUS AND U L N A indices (figs. 1 and 2) indicates that there is not any significant correlation with the known bone ages. This tends to support the theory that the development of the nutrient artery rather than the osseous development, is primarily responsible for the form of the nutrient canals (et seq.). EI. "1 0 . . I I0 to 20 30 40 50 60 70 ao 90 IOO AGE (in years) Fig. 1 Scattergram shows no significant relationship between the foraminal indices and ages of the left radius. %.I. = foraminal index. A similar absence of any significant trend line was found in the analysis of the right radius. *: .** . * . 30 10 I 10 1 20 30 40 M 60 70 80 90 I00 AGE (in vears) Fig. 2 Scattergram shows no significant relationship between the foraminal indices and ages of the left ulna. F.I. = foraminal index. A similar absence of any significant trend line was found in the analysis of the right ulna. Further confirmation of this, is suggested by the absence of a significant trend line in the graphs based on the ratios of paired radius and ulna lengths to the bone ages (figs. 3 to 6). When the foramina were assessed in relation to the anatomical surfaces of the bones, the findings were: Radius. On the anterior surface in 135 instances (82.3%), on the interosseous border in 25 instances (15.5?6), and on 688 6 . S. SHULMAN or within 1 cm of the lateral rounded border in 4 instances (2.4%). Ulna. On the anterior surface in 136 instances (83%), on the interosseous border in 24 instances (14.6%), and on the medial surfaces in 4 instances (2.4%). . . 140- R im- . 0 110 - I20 . *.. Io090 80 - 70 ' .. . * . . 0 .. . . - . . t 0.. * @ t - * . * 8 . ... . -: .. . . .. . . .. .. * . 0 0. . 8 . " 'a. 0 .* 20 ' 10 -t * Fig. 3 Scattergram showing absence of any significant relationship between the ratios of the paired bones and their ages, for the right forearm. R = ratio of total length of radius to total length of ulna. G. Direction. of nutrient carcals All the nutrient foramina on the diaphyses entered the compacta obliquely and were directed towards the ends which grow at the less rapid rate and which also fuse earlier than their opposite ends, i.e., all foramina were directed towards the elbow. The degrees of obliquity of the canals were assessed comparatively by measuring their visible portions on the surface of the bones. On the radius, these were always less than 0.5 em; on the ulna, they exceeded 0.5 cm in 6 pairs 689 NUTRIENT FORAMINA OF RADIUS AND U L N A (7.4%) whose ages were 21, 25, 30, 50, 53 and 63 years. No significant relationship was suggested between increased obliquity on the one hand and bone markings or distance from either end of the bone on the other hand. 190 R - e. 130120- - . . . . .... . . .. * . . . . . . .'.. .. . . .-@ a. 1100 .* 0 100- 0 a 0 90. 0 - 0 . ** 0 80. e.. 0 : 0. 0 0 0 e . 70 80 70. 0 10 Ib 1 i0 30 40 50 60 90 IM) AGE (in years) Fig. 4 Scattergram showing absence of any significant relationship between the ratios of the paired bones and their ages, for the left forearm. R = ratio of total length of radius to total length of ulna. D. Frequemy of symmetry This was assessed in respect of: 1. Duplicatiow of the foram.ew, which occurred in 17 bones; 15 were asymmetrical to either radius or ulna, and two were symmetrical. 2. Absewce of the foramen, which was so uncommon as to have no significance in the present series of results. 3. Positiow of the foramew, which was symmetrical on the radius specimens in 66 out of 82 pairs (132 bones-80.5%}. For the ulna the figures were 55 out of 82 pairs (66.7%). 690 5. S. SHULMAAN 4. Distance of foramert from either articular cad. The degree of asymmetry was assessed by labelling as asymmetrical a difference of more than 1 cm between such measurements on the left and right homonymous bone. I n the paired radius specimens asymmetry was found in 32 out of 82 pairs (64 bones -39%). For the ulna such asymmetry was found in 33 out of 82 pairs (40%). I n studying these asymmetrical figures, it becomes apparent that the asymmetry Mean of R% 110 --.o Ib io in ia io 6'0 i I-- 80 90 100 AGE. (in vears) Fig. 5 Histogram showing absence of any significant relationship between the bone ratios and their ages, for the right forearm. Mean R% =mean of ratios of radius to ulna. I20 Meanof 1 AGE (in yearb Fig. 6 Histogram showing abeence of any significant relationship between the bone ratios and their ages, for the left forearm. Mean R% = mean of ratios of radius to ulna. NUTRlENT F O R A M I N A O F RADIUS A N D U L N A 691 occurred simultaneously in respect of the radius and ulna of the same forearm, in 10 instances (33%). There was no indication that handedness, acting by virtue of better development of the muscles, had any significance ; on the right radius the asymmetry occurred 15 times (45.5%), on the right ulna 14 times (42.4%). DISCUSSION Most published work deals mainly with the direction of the nutrient canal. Havers (1691) described the directions and angles of entry f o r mammalian bones. I-Iumphrey (1861) was the first to offer a satisfactory explanation of the variations in direction. Not only did he confirm that long bones grow in length by accretion at the epiphyses (Hunter, 1837) : he also postulated that the bone increment in length occurs at a greater speed than the surrounding periosteum and that the latter grows interstitially both longitudinally and circumferentially. The firm attachment of periosteum to the epiphysis and the relatively loose attachment to diaphysis, would account for a relative sliding of diaphysial periosteum on the diaphysis during growth. The nutrient canal points to what would be the site of primary ossification, but this is modified to a greater or lesser degree by bone resorption forming the medullary canal. According to this theory of pcriosteal slip, the nutrient artery (which enters foetal long bones at a right angle to their long axes) would be angulated minimally or not at all if entering at the middle of the shaft, but angulation would increase the nearer the entry into the bone is to the epiphysis. I n the past, attempts have been made to confirm this theory by placing a wire loop through mctaphyseal periosteum and other similar loops through the diaphysis; Ollier (1873) worked on the upper end of the rabbit tibia, Lacroix ('51) on the metacarpal; both found that the nietaphyseal wire loop moved away from the diaphyseal loops during growth and they both interpreted this finding as cvidence of interstitial growth in periosteum. Hughes ('52) cri ticizps both experiments by assuming that thc metaphyseal 692 S. S. SHULMAN mire loop had entered the epipliyseal plate as well. If that had happened, however, the constant pressure on the growth plate would have caused obvious deformity or cessation of growth. If the experiments had been done as described by the authors, then interstitial growth of the periosteum should near the metabe equal in all its unit lengths-whether physis or near the middle of the shaft; whereas tension on the periosteal tube due to accretional growth at the epiphysis is greater than on that portion of periosteum surrounding the middle of the shaft because the periosteum is not an elastic tube. Other experiments (Lsacrois, ’51) indicate that the growth increments are greater in the subarticular area of the epiphysis than at the metaphysial aspect of the growth plate ; attachment of periosteum to an epiphysis is unusually firm and could not elongate other than by interstitial growth. The same almost certainly holds for the diaphysial periosteum as well, but the evidence is not conclusive as yet. Schwalbe (1871) had elaborated Humphrey’s hypotheses. He assumed that the periosteum is an elastic tube (which it is not) ; he then developed his argument on the basis of positive o r negative tensions being developed in the periosteal tube. He then postulated that the extent of the periosteum which is influenced by the growth tension of the diaphysis is directly proportional to the bone’s increase in length, so that the direction and slope of the nutrient artery should depend to a great extent on the original distance of the nutrient foramen from either epiphysis in foetal bone. Schwalbe’s examination of a large series of foetal and children’s bones led him to conclude that the nutrient canal (which he found to be at right angles to diaphysis in the foetus, because growth rate of opposite epiphyses is nearly equal and appositional increase of circumference of diaphysis is greater than postnatally) does retain the same relative position for its eztermal orifice on the shaft, whereas the endosteal orifice of the canal is the one which varies and decides obliquity as the result of resorption on this surface. Harris (’33) also put forward NUTRIENT FORAMINA O F RADIUS A K D U L N A 693 the view that the nutrient foramen maintains a constant position. Liitken ('50) found that the nutrient foramina on the femoral diaphysis all entered the shaft at or just below the lesser trochanter, and this finding convinces him that there is no basis for the theory that the foramen remains constant in position during growth. However, the difference in growth rate between the upper and lower femoral epiphyses is so great, and the chances of variations in the origin and direction of the nutrient artery (arising from the encircling medial circumflex artery) so much more favorable as compared with nutrient branches arising from arteries which are parallel to the diaphysis (Hughes, '52), that these two factors may be adequate to explain a shift of the foramen from midshaft to the lesser trochanter level. Some confirmation of such developmental explanation is suggested by other findings of Lutken ('50), viz., that 1%of these nutrient foramina on the femur in the adult entered the bone at right angles and a further 1%entered the bone in a distal direction. If one accepts the main premise that periosteum grows interstitially and at a rate slower than that at the epiphyses, then there still remains the question of unequal growth rate at opposite epiphyses of the same long bone. That this is true, has been proved by Payton ('32) and Bisgard and Bisgard ('35) on experimental animals, and by Gill and Abbott ('42) on clinical findings. The ratio of growth rates at the lower and upper epiphyses of the radius is 3 t o 1, for the ulna it is 4 to 1 respectively. Payton ('32) was also able to demonstrate that the rates of growth vary considerably from time to time in the same bones. Similar variations in growth rate have been reported in large numbers of mammalian and human bones (Schultz, '26 ; Martin, '28 ; Davenport, '33). All are agreed, however, that the greater increment occurs constantly in the same epiphysis of a long bone. These findings, therefore, do not explain those instances where there is a reversal of the direction of the canal. Lacroix ( '51) suggested that asymmetical muscular development could modify the traction forces acting on peri- 694 S. S. SHULMAN osteum, to a sufficient degree capable of causing even reversal of direction of the nutrient artery's entry into diaphysis. The bones reported in this article showed a remarkable degree of constancy and symmetry in this respect. If they were to be influenced significantly by muscles variation, then flexor pollicis longus and flexor digitorum profundus would seem to be the ones most likely since their attachments are placed nearest the foramina, i.e., on anterior surfaces of shafts and on interosseous membrane. Loth ('31) described a few differences in form for these muscles as h e and others found in white, black and yellow racial groups. No reliable reports of variations within the same racial group have been encountered. Nevertheless, this muscular theory cannot be regarded lightly in view of the known moulding which muscles exert so effectively on growing bone. The vascular theory appears to offer the most comprehensive explanation. Digby ('16) suggested that the obliquity of the nutrient canal may be decided by changes in direction of the nutrient artery at its origin. This presupposes that the direction of a blood vessel determines the form of its surrounding tissue. Now the embryonic branch which persists as the nutrient artery, does accompany the osteoblasts and osteoclasts on their invasion into cartilage; the vessel would therefore seem to precede bone. Hughes ('52) suggested that an artery grows by equal increments in its unit lengths, so that unequal growth of the two epiphyses would then cause the nutrient artery to be directed away from the end of quicker growth and consequently would determine the direction of the canal formed and reformed around the artery. There are also other suggestions that even developing veins can shape the form of the fascia1 hila (Salsbury, '39). The nutrient artery to the radius is a branch of the, volar interosseous, or of a muscular branch of the volar interosseous; that to the ulna, of the volar interosseous or of the proximal third of the ulnar artery or of a muscular branch of the ulnar artery (Piersol, '30). Since the volar N U T R I E N T FORAMINA OF RADIUS A K D U L N A 695 interosseous is the primitive axial artery which may be replaced by a series of anastomotic channels, a further explanation of anomalies in the nutrient artery to either forearm bone is possible on this basis. Certain surgical applications of the present findings deserve emphasis. The proximal halves of the forearm bones are well covered by muscle bellies which would ensure an adequate periosteal blood supply to the diaphyses. The distal halves, however, are devoid of any significant muscle attachments. The delayed union, and not infrequent nonunion, occuring in the middle shafts or lower, could be caused primarily by the normal anatomy of the nutrient arteries as described or by injudicious manipulations which could either increase tearing of the essential periosteal arteries or could injure the vital volar interosseous artery by tearing or tautening the interosseous membrane - particularly during forced pronation. The usual sites of entry of the arteries do not expose them to injury during the surgical exposures used for either bone, but careless retraction with periosteal elevators could easily damage them and adversely affect bone healing afterwards. SUMMARY The number, position, direction and symmetry of the nutrient foramina on 328 specimens of the human radius and ulna were investigated. All were present on the proximal half of the diaphysis and were directed towards the elbow. A very high incidence of symmetry in their positions was found. The periosteal, muscular and vascular theories are discussed; the vascular theory offers the best explanation of all reported anomalies as well as of the normal fashioning of the nutrient canals. Pertinent applications of the findings to orthopaedic surgery are indicated. 696 S. S . SHULMAN ACKNOWLEDGMENTS The author wishes to thank Professor AX. R. Drennan, Professor R. L. deC. H. Saunders and Professor Sir Solly Zuckerman for reading the script and for providing facilities at various stages of this work, and Professor A. J. Tingley for statistical advice. LITERATURE CITED ANSEROFF,N. J. 1937 Die Arterien des Skelets der Hand und des Fusses des Menschen. Z. Anat. Entwgesch., 206: 193-208. BISGARD,J. D., AND M. E. BISGARD1935 Longitudinal growth of long bones. Arch. Surg., 32: 568-578. DAVENPORT, C. B. 1933 The crural index. Am. J. Phys. Anthrop., 27: 333-353. DIGBY,K. H. 1916 The measurement of diaphysial growth in proximal and distal directions. J. Anat., London, 50: 187-188. GILL, G. G., AND L. C. ABBOTT, 1942 Practical method of predicting the growth of the femur and tibia i n the child. Arch. Surg., 45: 286-315. HARRIS, H. A. 1933 Bone Growth in Health and Disease. Humphrey Milford, London, pp. 239. HAVERS, C. 1691 Osteologia nova, or Some New Observations of the Bones. London. HUGHES, H. 1952 The factors determining the direction of the canal for the nutrient artery in the long bones of mammals and birds. Acta Anat., 25: 261-280. HUMPHREY, G. M. 1861 Observations on the growth of the long bones. Med.chir. Trans., 4 4 : 117-134. HUNTER, J. 1837 Experiments and observations on the growth of bones, from the papers of the late Mr. Hunter. I n : The Works of John Hunter, James I?. Palmer, Ed. Volume 4. Longmans Green, London. LACROIX, P. 1951 The Organization of Bones. J. & A. Churchill, London. LOTH,E. 1931 Anthropologie des parties molles. Masson et Cie, Paris, pp. 538. L ~ ~ T K EP. N , 1950 Investigation into the position of the nutrient foramina and the direction of the vessel canals in the shafts of the humerus and femur in man. Acta Anat., 9: 57-68. MARTIN,R. 1928 Lehrbuch dcr Anthropologie. 2nd ed. Gustnv Fischer, Jena, p. 414. OLLIER, L. 1873 Recherche8 experimentales sur le mode d 'accroissement des 08. Arch. physiol. norm., 5: 5 4 2 . PAYTOX, C. G. 1932 The growth in length of the long bones i n the madderfed pig. J. Anat., Lond., 66: 414-425. PIERSOL, G. A. 1930 Human Anatomy. 9th ed. G. C. Huber, Ed. J. B. Lippincott and Co., Philadelphia, p. 778. SALSBURY,C. R. 1939 The morphology of the saphenous opening. J. Anat., Lond., 73: 186-191. NUTRIENT FORAMINA O F RADIUS AND ULNA 697 SCHULTZ,A. H. 1926 Foetal growth in man and other primates. Quart. Rev. Biol., 1: 465-521. SCHWALBE, G. 1876 Ueber die Ernahrungskanale der Knochen und das Knochenwachstum. Zeitschr. f . Anat. 11. Entwgesch., 1 : 307-352.