by Julius H. Hess, M.D.
The early appearance of rachitic manifestations in premature infants has been noted by many observers, especially associated with spasmophilia and anemia. Most prematurely born infants become rachitic -- the lower the weight, the more certainly -- and even human milk is not an absolute protection against this.
Huenekens  was able to collect 70 cases of prematures and twins, of which 58 developed definite signs of rachitis (82 per cent). The time of occurrence was interesting, inasmuch as of 33 cases seen for the first time at or before four months, 27, or 81 per cent, showed evidence of rachitis at that time. The first symptom usually noted was craniotabes, which in 3 instances was already present at six weeks. Langstein  observed it frequently in the third to the fourth month of life and not much less often was the tendency to convulsions (hyperirritability of the nervous system of these infants).
Ylppö observed commonly a megacephalus in connection with rachitis of the skull, which often left marks permanent for all life. These have to be regarded as characteristics of the prematures and not, as unfortunately is often the case, as signs of special "constitutional degeneration." Along with this megacephalus with its somewhat large, plump skull, there is asymmetry, which is not congenital but is produced in a mechanical way by the pressure of the infant's head in the first months of life and the softness of the skull.
The narrow thorax with its more or less marked signs of rachitis may also be regarded as a peculiarity characteristic of the smallest prematures but not of those of greater weight. This is not to be confused with the early functional, funnel chest which can be demonstrated in the first weeks of life and is due to the softness of the ribs in the smaller prematures. This leads to further deformity of the thorax, as the marked contraction of the lower half, which is the result of the congenital softness of the ribs and the rachitic affections later developing. The constriction around the chest is best seen about the insertion of the diaphragm.
The rachitic rosary is very prominent in prematures and is explained on the basis of the constant respiratory movements leading to deformities and marked enlargement of the epiphyses of the ribs.
The long cylindrical bones, however, only exceptionally show enlargements of the epiphyses in prematures, although rachitic changes appear in these bones very early. The process in these bones results rather in bone absorption and fringing of the epiphyses than in marked proliferation, which is the rule in strong full-term rachitic infants. The explanation of this feature in prematures may be in the fact that the rachitis appearing very early is already at an end by the time the infant learns to walk, whereas in the full-term infant the hyperplastic epiphyseal enlargement occurs as a compensatory process in the period when the lower extremities are called upon to support the weight of the body. In the absence of special rachitic curvatures and epiphyseal enlargements of the long bones, we cannot therefore exclude rachitis in the premature. Histological examination shows a characteristic picture in the absence of marked external manifestations.
Etiology. -- The etiology of rachitis in premature and full-term infants has been the subject of much discussion. Huenekens believes that the explanation may be found in that the chemical constitution of prematures is abnormal. Their salt content is far below normal. Birk  found that a four-months fetus contained 14 gm. of ash, at six months 30 gm., at nine months 100 gm., showing that two-thirds of the minerals were taken on during the last three months of fetal life. In the new born fully 75 per cent of this ash is made up of calcium and phosphate, the chief constituents of the bones. Huenekens believes, therefore, that the more premature the infant, the greater will be the deficiency of calcium and other minerals, so that by the third or fourth month of extra-uterine life the supply is entirely exhausted and rachitis results.
Underfeeding is another factor in the development of rickets in the premature. The low calcium content of human milk and the difficulty of metabolizing even this food in sufficient quantities to prevent drawing on the inherited supply may be an active factor. The artificially fed are especially prone to develop severe rickets. If the diet contains sufficient milk the tendency to develop the disease is less than when fed mainly on cereals and proprietary cereal foods with only small amounts of milk. A diminished calcium retention (negative calcium balance) exists in the florid stage of rickets, even though the intake is ample. A deficiency of calcium in the diet while important in itself is probably not the precipitating factor. It has been shown experimentally in puppies that a diet containing an abundance of calcium does not prevent rickets when the diet is deficient in other factors.
The average inorganic phosphorus concentration in the serum is about 5 mg. per 100 cc. Howland and Kramer  found that in all patients in the active stage of rickets the concentration of inorganic phosphorus in the blood serum was low and that in all children under two and a half years of age, in whom an inorganic phosphorus content of the serum of 3 mg. or less was found, active rickets was present. With the healing of the process in the bones that occurred after cod-liver oil medication, the phosphorus rose gradually to normal. These facts led them to consider the presence of a low percentage of inorganic phosphorus in the serum of a young child as nearly conclusive evidence of active rickets. They believe that there is constantly a marked and for the causation of pathological lesions, and important deficiency in inorganic phosphorus. To this deficiency they ascribe the failure of calcium deposition.
The phosphorus content of the blood can be increased by feeding phosphorus by mouth. Marriott , working with artificial blood, found that by small increases in the phosphorus content, a precipitate resembling in composition the salts of bone was formed. Phemister  applied these experiments to children and noticed by roentgenogram studies that phosphorus affected the normal bones of children as it did Wegner's  animals and that the accumulation of calcium and overproduction of bone in the metaphysis continued for some time even after the administration of phosphorus was discontinued. He has more recently reported similar results in rachitic infants.
McCollum and his associates [8, 9], in a study of the effect on the growth and development in rats, came to the conclusion that the etiological factor is to be found in an improper dietetic regimen. Their experiments showed that the majority of young rats developed pathological conditions of the skeleton having a fundamental resemblance to rickets when fed upon diets low in both fat soluble vitamines and phosphorus. When they modified this diet so that the deficiency in phosphorus is compensated for by the addition of a complete salt mixture, containing the phosphate ion, the deficiency in fat soluble factors still existing, no pathological changes of a rachitic nature developed. They, therefore, concluded that a deficiency in this vitamine cannot be the sole cause of rickets. In summarizing, they state that the phosphate ion in the diet may be a determining influence for or against the development of rickets, but that these findings should not exclude the absence of fat soluble vitamine from consideration as an etiological factor in the production of rickets and kindred diseases, since the level of blood phosphate is, in all probability, determined in part by the amount of fat soluble vitamine available for the needs of the organism.
Summarizing, it appears that rickets is a nutritional disturbance especially affecting the osseous and muscular system, with resulting lesions which prevent the bones and muscles from utilizing calcium, thus leading to a diminished retention of this element, although there is plenty of it in the food intake and in the blood. Phosphorus probably plays an intermediate role in influencing the formation and deposition of the lime salts in bone. Whether the diet plays its role by directly interfering with the calcium and phosphorus metabolism due to lack of an antirachitic factor or indirectly by causing an underlying nutritional disturbance is open to conjecture.
Hygiene is an important factor in that improper hygiene results in impaired metabolism with a resulting inability to utilize the dietetic constituents even when properly balanced. Infections play a similar role. Impairment of the body functions also directly affects the glands of internal secretion with secondary disturbances following such dysfunction. Therefore, while an impairment of mineral metabolism precipitates the clinical symptoms, one or several of the secondary factors may have an important relation to the utilization of phosphorus and calcium.
Treatment. -- Our therapy is along the same lines as in full-term children with special stress on the feeding of human milk. Fresh air and sunshine in the older children and the observation of careful hygiene for all are without doubt highly important.
Diet must receive very careful consideration. In the very young the ideal food is, of course, mother's milk. Where artificial feeding must be instituted the amount of cow's milk should be minimal, and cereals and vegetables started early. Orange juice diluted with water should be given in small amounts from the second or third months (one to four teaspoonfuls daily). After the first month the diluent in the milk mixtures should be a cereal water (one tablespoonful of whole barley or oatmeal to the quart of water -- and not the dextrinized cereal flours). From the third month cereal should be fed. After the fifth month vegetable soups should be given, substituting an ounce of soup for an equal amount of bottle feeding. By the sixth or seventh month a milk-feeding should be replaced by a vegetable-soup meal.
Cod-liver oil with phosphorus in a preparation containing 0.0003 gm. (1/200 gr.) to each 4 cc (1 dr.) oleum morrhuae, is a most practical mixture and can be administered to most infants by the fourth to the sixth week, beginning with 1/2 cc doses twice daily and increasing to 4 cc twice daily by the fifth month. The work of Schloss  has shown that the addition of a calcium salt to cod-liver oil with phosphorus further enhances the value of the mixture. Such a preparation is the tricalcium phosphate C. P. (10 per cent) in emulsion of cod-liver oil U.S.P.
Closely associated with rachitis in premature infants is an anemia, which develops quite regularly and strikingly during the first three months of life. In our previous discussion of the physiology of the blood we noted from the work of Kunckel, Lichtenstein, Lande and others that in contrast with full-term infants, in the premature there is a greater number of nucleated red blood corpuscles, a more frequent appearance of myeloblasts and myelocytes during the first days of life, a lesser development of absolute and relative leucocytosis, and a greater number of immature leucocyte forms. There is also a distinct and very early hemoglobin impoverishment of the blood, which reaches its maximum in about the third to the fourth month.
Etiology. -- Kunckel  believed that this anemia appearing regularly in the first three months of life was physiological and was of the chlorotic type. His children improved in the second half year of life, but if infection was present any time the infants developed a severe secondary anemia much more readily than full-term infants. His opinion was that the anemia did not rest on an alimentary basis but was due to an insufficiency in hemoglobin metabolism, besides a deficient iron storage.
Pfaundler  felt that the anemia was closely related to a lack of fresh air and sunshine.
Lichtenstein  fixes the early anemia in the first three months of life as a hypoplastic condition resulting through insufficiency of the hematopoietic system. The later oligochromemia, after spontaneous retrogression of the oligocythemia, he considers as a sequel of the impoverished iron storage. He opposes the hypothesis of alimentary anemia of Czerny and Kleinschmidt. The theory of the harmful action of milk on the hematopoietic apparatus he asserts is disproved by the excellent results attending the feeding of human milk and the administration of small amounts of ferrous lactate.
Lande  is in accord with the opinions of Kunckel and Lichtenstein. As evidence in favor of the importance of iron storage he emphasizes the fact that eighth-month infants in the course of the second quarter year of life show a higher percentage of hemoglobin and erythrocytes than do seventh-month prematures.
The examination by Lande of the bone marrow in ten prematures disclosed no decisive picture except an insufficiency of the granulocyte system. Thus he disproves the theory that the basis of the anemia rests with a defective erythropoietic system.
Lichtenstein feels that there is no marked difference in the blood picture of artificially and breast-fed prematures. Examination of twenty-eight cases artificially fed, many of whom were born of nephritic, anemic, and tuberculous mothers, showed no great differences in the blood picture from those breast-fed.
Symptoms. -- The most marked symptom observed by Lande was pallor of the skin, which he saw with great regularity. It appeared especially early as a fore-runner of icterus, which in prematures is constantly present. The question arises as to whether the anemia is promoted by the icterus or both icterus and anemia are bound up with a third factor -- the maturity of the infant.
One can differentiate various grades of pallor, which is earliest and most clearly seen in the face and well agrees in general with the degree of pathological blood change. The most marked form of anemia gives the infant a bluish, transparent appearance, or a waxen, yellowish color, somewhat akin to the infants with severe congenital syphilis or chronic pyelitis. The picture is accentuated by the outstanding bluish veins, especially prominent on the skull and abdomen. The ears are transparent with hardly the vestige of a rosy hue and the mucous membranes are very pale.
The pallor after open-air treatment has been noted to give way to a rosy hue, but only in a few cases is there a parallel permanent increase in hemoglobin. However, with the increase in hemoglobin and erythrocytes in the fourth to sixth months, the color simultaneously improves.
In Lande's series the appetite of the infants was in general satisfactory. There was no stupor, especially in the more anemic. There were no elevations of temperature, as described by some observers, present with the marked blood changes.
Marked glandular and splenic swelling was not observed by Kunckel or Lande, but Lichtenstein states that splenic tumor was present in two-thirds of his children.
Lichtenstein finds that the blood pictures in the well breast-fed prematures and those showing alimentary disturbances are both of the chlorotic type and differ mainly in degree. He also believes that the clinical picture described as pseudoleukemic anemia is a severe form of secondary anemia and is not a distinct clinical entity.
Treatment. -- For the general and hygienic treatment of primary and secondary anemia, the suggestions made for general measures in the care of rachitis should be followed. The infants must above all be given the advantage of a good environment, plenty of fresh air and sunshine.
Iron therapy for the purpose of increasing the iron content of the tissues and the hemoglobin has met with individual success. It should be started early. Among the iron compounds to be recommended are ferri carbonas saccharatus 0.25 to 0.5 gm, ferri et ammonii citratis 0.06 to 0.12 gm. or ferri lactis 0.12 to 0.25 gm. one to three times daily.
Small doses of liquor potassii arsenitis 0.03 to 0.06 cc may be given one or two times daily for short periods. The infant should be observed carefully for evidence of arsenic intoxication.
In the presence of congenital syphilis, mercurial therapy is imperative and may be combined with the arsenic treatment to good advantage.
Lande suggests the use of intramuscular injections of normal human blood. The blood is drawn from the vein of a healthy adult with a Wassermann needle and allowed to flow into a flask containing small glass beads. It is shaken about five minutes and thus defibrinated and before injection is passed through a double thickness of sterile gauze. In individual cases the result may be very good, however, in a series of thirteen cases he was unable to demonstrate a marked increase in hemoglobin or red corpuscles.
It is of the greatest importance to bear in mind that, as in the case of rachitis, the treatment for anemia should be started early. It is our custom to begin the prophylactic treatment of both of these conditions in the first weeks of life.
Besides anemia and rachitis, spasmophilia is one of the most interesting clinical peculiarities of premature infants. The term spasmophilia is used here in the sense of designating the manifestations occurring in the nursling, while tetany refers to the older child. Typical tetany with all its characteristic symptoms, the phenomena of Erb, Chvostek and Trousseau, carpopedal spasms, tonic and clonic convulsions and laryngospasm, has rarely been observed in the premature newborn. While spasmophilia is by no means rare in artificially fed, full-term infants, it is found quite frequently in prematures, and not only in those artificially fed, but also in the infants fed on human milk. With breast-feeding, however, spasmophilic manifestations occur only exceptionally and lead to convulsions usually only in connection with infections.
Etiology and Symptoms. -- There are other predisposing factors besides feeding in the development of spasmophilia in prematures. We may not call these factors constitutional, since in all probability they depend on the various noxae of the extra-uterine life. These lead to a hyperirritability of the nervous system. The nervous system in all premature infants is extraordinarily lowered during the first weeks of life against all possible stimuli, including the electrical.
On the other hand, however, it is a fact that feeding plays a very important role in the development of spasmophilia and above all in the appearance of convulsions. Not uncommonly one observes prematures in whom the spasmophilic manifestations remain latent as long as breast-feeding is continued, but appear shortly after the institution of artificial feeding. Langstein  reported a case of twins in whom convulsions always appeared after artificial feeding was added to human milk. There were individual differences between the twins in the ease with which the convulsions could be produced. In one infant they developed within seven to twelve days after the addition of artificial food, in the other within eighteen to twenty days.
If we systematically examine the electrical irritability in a large number of prematures we can determine that in infants fed on human milk the electrical irritability may increase to such an extent to be C.O.C. less than 5 milliampères at the age of six to ten weeks. Rosenstern  studied the spasmophilic diathesis in premature infants and noted individual differences in the electrical hyperirritability, which appeared very early and frequently in breast-fed prematures. He was able to demonstrate spasmophilia in the form of electrical hyperirritability in 76 per cent of the prematures and debilitated infants that he studied.
In Ylppö's series the electrical hyperirritability was not as frequent. Among the 42 premature infants in whom he was able to determine the electrical reaction systematically during the first three to six months of life, only 15 infants (35 per cent) showed C.O.C. less than 5 milliampères. In 3 cases electrical hyperirritability was already present in the second month. One infant was on human milk-feeding and the other on mixed feeding. In the third month electrical hyperirritability appeared in 3 additional infants. It was most frequently present in the fourth month. From this data it seems that spasmophilia appears earlier in premature than in full-term infants.
Ylppö  also noted the interesting fact that the great tendency to electrical hyperirritability and convulsions, which he determined in many artificially fed premature infants, three to four months old, gradually disappeared in the fifth to sixth months without any treatment, while the feeding remained the same. In other children it often took months before the electrical hyperirritability disappeared.
In premature infants we frequently find very interesting deviations from the generally recognized symptoms of spasmophilia. It is not at all infrequent that the cardinal symptom of spasmophilia (Erb's symptom), the electrical hyperirritability of peripheral nerves, may be absent, in spite of manifest signs of the disorder. To know this is very important, because we know that there exist conditions in premature infants in which the electrical reaction remains increased for months, although no convulsions occur. This lack of electrical hyperirritability in spasmophilic convulsions in prematures exists not only after convulsions have taken place -- which could easily be accounted for by exhaustion of the nervous system -- but also before the appearance of convulsions.
In individual cases there may be pathologically increased electrical hyperirritability, even when the electrical reaction does not go below 5 milliampères for C.O.C. Rosenstern called attention to this fact and pointed out that the value below 5 milliampères for C.O.C., which is regarded as pathognomonic for the spasmophilic diathesis, was determined by Mann  and Thiemich  only for the age of eight weeks. From this it follows that this value is not to be regarded as a limit for younger infants, at least not for the younger prematures.
In the majority of cases the disappearance of the spasmophilic tendency in prematures occurs at the same time at which anemia and craniotabes begin to improve. Thus it becomes more and more apparent that the three symptoms, anemia, rachitis, and spasmophilia are in a certain relationship. It may very well be that the same factors that damage the activity of the hematopoietic organs in the first months of life and also the growth and the normal calcification of the bones in such a high degree produce in some manner unknown to us changes in the nervous system.
Calcium metabolism. -- A calcium deficiency in the tissues has been demonstrated by numerous investigators, more especially in the brain and blood. The earlier investigations on the blood have more recently been confirmed by Howland and Marriott , who found the calcium of the blood serum to be low in this condition, averaging 5.6 mg. per 100 cc of serum in a group of 18 cases, the lowest being 3.5 mg. per 100 cc of serum, the average normal amounts being 10 to 11 mg per 100 cc. They found a normal calcium content in the serum in convulsive disorders due to other causes. These same authors found the magnesium content in the serum to be within normal limits even in the presence of active spasmophilia. The relation of calcium to the symptoms of spasmophilia has been studied extensively, especially its influence on the electric excitability. Physiologists have shown that certain mineral ions exert a specific effect on muscle-nerve irritability. Rosenstern  and Sedgwick  reduced the electric irritability in spasmophilic infants by administering large doses of calcium by mouth. Loeb's  findings indicate that Na and K increases the threshold for excitation, while Ca and Mg tend to decrease this. This muscle nerve irritability is the function of the quotient (Ca + Mg / Na + K) as designated by Reiss . During a diarrhea Holt  has demonstrated there is a much greater loss of Na and K than Ca and Mg in the stools. Diuresis and catharsis often cause an improvement in the spasmophilic symptoms. Consequently, there is much clinical and experimental evidence that spasmophilia is much influenced by the relationship between the Ca-Mg and Na-K group of ions.
Accidental removal of the parathyroid gland in humans and experimental excision of these glands in animals have both resulted in a tetany that resembles in its clinical manifestations the spasmophilia of infants. Following the animal experiments Howland and Marriott  have demonstrated a diminution in the calcium content of the blood. These findings have been verified by MacCallum and his co-workers , who also found a decreased calcium content in the brain and an increased excretion.
Greenwald , in his experimental studies, found that the phosphorus excretion in the urine of his animals was greatly decreased (to as low as 8 per cent of the normal) shortly after operation. He also found an increase of the phosphorus content of the blood before the appearance of tetany. There was also a sodium and potassium retention. He believes that following the extirpation of the parathyroid there is a decreased excretion through the kidneys and an abnormal retention in the tissues of the alkali phosphates, which is followed by a decreased retention and an increased excretion through the kidneys as soon as the spasms develop.
There is, however, great question as to the relationship of parathyroid function and tetany in the infant. Pathological studies lead us to believe that parathyroid lesions in infantile tetany are the great exception. Parathyroid lesions have been described in patients who have shown no evidence during life of the pathognomonic findings of tetany.
In summarizing the pathogenesis we may state that a diminution of calcium salts in all probability is the most important factor in the development of this condition. However, the possibility of an absolute or relative excess of sodium and potassium salts, particularly the phosphates, playing an important role cannot be overlooked. The relationship of disturbance in parathyroid function to the diminution of calcium tissue content must be made the subject of further study before its importance can be fixed.
Diagnosis. -- The differential diagnosis of spasmophilic convulsions in prematures is very difficult. Among the conditions to be considered are hydrocephalus, congenital syphilis and tuberculosis, epilepsy, infections, brain injuries, asphyxia and pulmonary atelectasis. Tetanus neonatorum is rarely seen today. Meningitis and encephalitis are the most important of the infectious processes, and the primary focus often is unknown. Perhaps the best test after careful history and physical examination is the determination of the electrical reactions.
It must not be forgotten that given an injured brain and a marked tendency to spasmophilia, this leads, in the premature, in the first place to convulsions and other manifest phenomena of this diathesis.
There are only a few cases in the literature where special attention has been devoted continuously from birth to the later years to the condition of spasmophilic infants. Ylppö's material enabled him to fill this gap to a certain extent. He was able to show positively that spasmophilia in premature infants very frequently occurred after a preceding injury to the brain, and this injury rather than spasmophilia, causes the later brain changes. Spasmophilic convulsions may, however, produce extensive damage to the brain, and may result in various defects of intelligence and other cerebral disturbances.
Treatment. -- The treatment of spasmophilia is largely prophylactic and embraces the therapy of rachitis and anemia. With the early institution and the continuation of these hygienic, dietetic and medicinal measures, the development of spasmophilic convulsions will be very unusual.
If convulsions appear the infant must be kept absolutely quiet and warm. Narcotics are usually employed, the best being chloral hydrate 0.25 to 0.5 gm. per rectum and calcium bromide 0.5 to 1 gm. per day.
Lumbar puncture with the drawing-off of 5 to 15 cc of spinal fluid may give relief from repeated convulsions.
The use of general anesthesia and morphine derivatives to control convulsions is to be avoided except as measures of last resort.
Magnesium sulphate in sterile 8 per cent solution has been used in subcutaneous injections, 5 to 15 cc repeated once or twice within forty-eight hours, to control convulsions. Because of its marked depressive action on the nervous system, the infant must be very closely watched for collapse. Although this method has been much used in some clinics our experience has not warranted its use in preference to our preceding measures.
For the acute manifestations the calcium salts in maximal doses (preferably calcium lactate, 0.3 to 0.6 gm., three times daily, in solution or suspension) in our experience have been more valuable than the magnesium salts.
Administration of cod-liver oil and phosphorus or tricalcium phosphate in emulsion of cod-liver oil, as recommended in the treatment of rachitis, should be started at the same time and continued indefinitely, in doses varying from 1/2 to 4 cc twice daily, dependent upon the age and the indications.
Where feeding by mouth is difficult, catheter administration must be resorted to together with inert fluids per rectum. If human milk is not obtainable the best substitute is albumin milk, which is poor in whey and rich in calcium. Where stimulating treatment becomes necessary that which has been previously mentioned may be employed.
In all our measures extreme gentleness must be used, as any rough handling or violence, in case of asphyxia, has a very harmful influence.
Fig. 177. Rickets - first stage.
Fig. 178. Rickets - second stage.
Fig. 179. Spasmophilia. Infant in state of "tetany."
Fig. 180. Spasmophilia - double fracture of both forearms following prolonged carpal spasm. Premature, aged six months.
 Jour. Lancet, 1917, 37, 804.
 Ztschr. f. Kinderh., 1916, 15, 49.
 In Monatsschr. f. Kinderh., 1910, 1, 644.
 Jour. Biol. Chem., 1920, 43, 35.
 Report of Thirty-second Meeting of Am. Ped. Soc., Arch, Ped., 1920, vol. 37.
 Effects of Phosphorus on Growing Normal and Diseased Bones, Jour. Am. Med. Assn., 1918, 70, 1737.
 Virchow's Arch. f. Path. Anat., 1872, 55, 9.
 McCollum, Simmonds, Parsons, and Shipley: Jour. Biol. Chem., 1921, 45, 333.
 Shipley and Park, McCollum and Simmonds: Johns Hopkins Hosp. Bull., 1921, 32, 160.
 Zur Therapie der Rachitis, Jahrb. f. Kinderh., 1914, 79, 194.
 Ztschr. f. Kinderh., 1915, 13, 101.
 Verhand. d. Ges. f. Kinderh., Breslau, 1904, 21, 24.
 Svenska, La Karesa as Kapets Handlingor, 1917, No. 4, 43.
 Ztschr. f. Kinderh., 1918, 22, 299.
 Kassowitz, Festschrift, Berlin, 1912.
 Ztschr. f. Kinderh., 1913, 8, 171.
 Ztschr. f. Kinderh., 1919, 24, 1.
 Monatsschr. f. Psych. u. Neurol., 1900, 7, 14.
 Jahrb. f. Kinderh., 1900, 51, 99, 222.
 Quarterly Jour. Med., 1917-1918, 11, 289.
 Jahrb. f. Kinderh., 1910, 72, 154.
 St. Paul Med. Jour., 1912, 14, 497-519.
 Oppenheimer's Handbuch der Biochemie.
 Ztschr. f. Kinderh., 1911, 3, 1.
 Am. Jour. Dis. Child., 1915, 9, 213.
 Trans. Am. Ped. Soc., 1916, 28, 200.
 MacCallum and Voegtlein: Jour. Exp. Med., 1909, 11, 118.
 Jour. Biol. Chem., 1913, 14, 370.