312

DIAGNOSTIC PROCEDURES IN THYROIDDISEASE

By RUSSELL FRASER, M.D., F.R.C.P., D.P.M.Reader in Medicine, Postgraduate Medical School, London

Various diagnostic aids may be used to sup-plement clinical investigation when either thepatient is still only suspected of raised or loweredthyroid function, or when the pathology of histhyroid lump still remains in doubt. We shallconsider in this paper, the various tests availablefor assessing the level of thyroid secretory activity;of these there are three main tests and at least threesubsidiary indirect tests each with its own specialusefulness and its limitations. Since the adventof radioiodine and also of improved techniquesfor the micro-chemical estimation of iodine, testsof iodine metabolism have tended to dominatethis field, but they have not yet completelysupplanted either the older B.M.R. or the varioussubsidiary tests. Space does not permit con-sideration of the uses of needle biopsy of thethyroid or of thyroid scanning after radioiodine,procedures which are especially useful for thediagnosis of thyroid nodules and of suspectedlymphadenoid goitre. The finding of a raisedserum gamma-globulin is also a useful con-firmatory sign of a lymphadenoid goitre.Tests of Total Thyroid FunctionThe Basal Metabolic Rate (B.M.R.)The oldest clinical measure of total thyroid

function still provides the best index of theseverity of a disorder of thyroid function, andestimates the effect of the thyroid hormone onbody tissues. But as a measure of thyroidfunction it has some defects; firstly, there areother causes of an abnormal metabolic rate,secondly, it may be difficult to attain the basalstate in a patient, and thirdly, we may err in ourprediction of the normal B.M.R. to be expectedwith certain patients, i.e., with children or withvery fat or very thin patients. But proper clinicalexamination should reveal any other disease likelyto lead to an abnormal B.M.R.; and with dueprecautions a truly basal state should be attainablein nearly all patients.The B.M.R. is estimated from a measure of the

oxygen consumed by the patient when he is

'basal' (i.e. both fasted for I2 hours and com-pletely resting). It is usually expressed as apercentage of the oxygen consumption expectedfor a healthy person of similar size, age, and sex,obtainable from standard tables. With goodapparatus, the main problem is that of gettingthe patient in the completely resting state. Onlycareful expert technicians using one or two trainingruns can achieve this with most patients withoutthe help of sedation. This difficulty has even beenreflected in the B.M.R. standards. The commonlyused standards (Aub and Du Bois, 1917; Boothbyet al., I936) were derived from measurementsmade on the subject's first visit to the laboratory.Repeated B.M.R. estimations on the same subjecthave shown that the first two or three results tendto be higher than the level of values achievedthereafter (Vogelius, I945). Robertson and Reid(I952) collected more truly basal standards fromsubjects who had had preliminary training visitsuntil duplicate readings were attained agreeingwithin 5 per cent. These standards were foundto lie about Io per cent. lower than the previousand still commonly used standards. With askilled technician the same procedure can be usedclinically, and the results should be estimatedagainst these standards; which show a normalrange of + or 13 per cent.An alternative procedure which uses the same

standards is the estimation of the sleeping B.M.R.(Rapport, 1951; Fraser and Nordin, 1955). Inthis procedure the patient is given sedation whichinduces sleep but avoids the risks of anaesthesia;by giving either intravenously a slowly-actingbarbiturate such as pentobarbitone sodium, ororally three separate hourly doses of amylo-barbitone. To be certain that the patient is basal,the estimation must be made with the patientsleeping. The only real disadvantage of thisprocedure is the subsequent drowsiness of thepatient for the rest of the day; which can beminimized by a subsequent oral dose of a bar-biturate antagonist (Ibbertson et al., 1957). Thestandards for the sleeping B.M.R. are the same as

copyright. on A

pril 16, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.381.312 on 1 July 1957. D

ownloaded from

July 1957 RUSSELL FRASER: Diagnostic Procedures in Thyroid Disease 313

THE 5 MAIN TESTS FOR {Myxoedemo (untreated)Thyrotoxicosis( untreated)

(BY S. D. RANGESJ

NORIMAL RANGE

BM (slDeping:R&R standards) A 8I88-42A -2 - +35 +49/. AND OVER

' 'I Thyroid Uptoke(T)15 2 i2. 12 S7 AND OVEO

Plasma P-bd 'lodine. °0 13 2- 5' 7 10-4 12-9 IS5 AND OVER

Myxoedema ThyrotoxicosisA SEVERE 10 SEVERE 12 CRISIS

A MILD I 0 MODERATE 6

* MILO 10

FIG. i.-Results of three main tests in proved myxoedema and thyrotoxicosis. Each scale is matched instandard-deviation ranges of normals in order to compare tests' sensitivities. Nearly all B.M.R.s wererecorded sleeping, and their results are shown as percentage of appropriate mean Robertson and Reid(I952) standards. (From Fraser (1956), Lancet.)

those for the properly trained and basal B.M.R.,as Nordin and Fraser found. Further Benedictand Hendry (I921) found the B.M.R. of thesleeping adolescent to lie 7 to io per cent. belowthe then prevailing untrained standards; i.e., inthe same range as that found in the trained basalstate (Robertson and Reid, I952; Shock andSoley, I939; Bierring, 193i).A narrower range of normal B.M.R. values

which would permit a more sensitive estimationof deviations therefrom, would be attainable if theB.M.R. standards could also be corrected forabnormal body builds, as well as for size, age andsex. Most standards use surface area to estimatebody size; which is perhaps the best basis exceptin children where height may be better, or in fatand thin patients. For the latter, we should havesome additional correction incorporated to allowfor the abnormal proportion of their body sizewhich is metabolically active tissue, i.e., for theirproportion as lean body mass (Miller and Blyth,1953; Behnke, I953; Kleiber, 1956). Over awide range of animals the relation to size is

surprisingly consistent if the weight formulais used (B=3W'), (Kleiber, I947).Radioiodine TestsThese tests are capable of measuring the rate at

which some phase of the body's iodine cycle isoccurring and so of assessing thyroid cell activity.They cause minimal inconvenience to the patientsand involve a smaller radiation hazard than manycommon X-ray diagnostic procedures. Theirgreat advantage is their independence of thenervous state of the patient, and their disadvantageis their susceptibility to misinterpretation aftercertain drugs, and when the body's store of iodineis either depleted or over-filled.

After previous treatment to the thyroid or withdrugs containing iodine or thyroid substance, orafter depletion of the body's iodine stores by anabnormal diet or by a course of antithyroid drugs,these tests will reflect these circumstances as wellas thyroid secretory activity. This disadvantageis not overcome by any variation in the type ofradioiodine test, short of a 'supplemented

copyright. on A

pril 16, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.381.312 on 1 July 1957. D

ownloaded from

314 POSTGRADUATE MEDICAL JOURNAL July 1957

radioiodine test' (see below). This is becauseradioiodine tests measure the thyroid turnoverof iodine not in absolute units or micrograms, butin terms of the extracellular fluid's content ofiodide. Suspicion of these sources of error mayarise either from a history of recent drug adminis-tration, which should always be sought; or whenan abnormal radioiodine result is found whichdoes not seem compatible with the patient'sprevious clinical picture. Such suspicions shouldnot arise in more than about io per cent. of thepatients tested, and then some of the specialsupplementary radioiodine tests mentioned below,or other tests of thyroid function, need to be done.These will confirm or refute this possibility thatthe radioiodine test is not simply reflecting thyroidsecretory function, and also will measure thyroidfunction in the light of this suspicion.

131I of 8-day half-life is the usual form ofradioiodine used; but if it is desired to repeat thetest frequently, 132I of 2.4-hour half-life can beused with one of the rapid types of technique.The dose is either given intravenously or orallyon an empty stomach, and in either case rapidlymixes throughout the extracellular fluid. It isthence cleared at constant rates into both thethyroid and the urine, most of this being completednormally between 24 and 48 hours. Between thethird and the twenty-fourth hour, the radio-activity in the thyroid begins to be secreted backinto the plasma as thyroid hormone whichcirculates there bound to the plasma proteins.This simple theory provides an adequate basisfor the interpretation of clinical tests.

Several types of radioiodine test are available,which vary in their convenience or in theiraccuracy. But all of which have approximatelythe same sensitivity in reflecting the rate of iodinetransfer and so of thyroid function (Pochin, 1950;Myant, I952; Keating et al., 1950; Fraser et al.,I953; Wayne, 1954; Rall, I956). Except aftercertain drugs or when the stores of iodine areabnormal, it is safe to assume that the thyroiduptake and secretion rate are in equilibrium.All radioiodine tests are very sensitive to theincreased thyroid function, and segregate thethyrotoxic group .well away from the normalrange, though some non-toxic hyperplastic goitresshow similarly high uptakes. Severe or completemyxoedema is also fully distinguished but mildthyroid deficiency overlaps the normal rangeeven more on the radioiodine test than on the othermain tests of thyroid function. This is probablybecause thyroid failure leads to less conservingand storing of iodine in the body and so to aniodine deficiency. The rate of iodine transfercan be measured either 'at the uptake into thethyroid (Step i of the iodine cycle), which can be

assessed either from urine measurements orfrom measurements at the neck, or at its re-appearance in the plasma as protein-boundradioiodine (Step i + Step 2 of the iodine cycle).

Uptake measurements, whether from the urine orfrom neck measurements, are more sensitive andspecific when they are based on two or moremeasurements from which can be derived a ratereflecting the thyroid clearance of iodide. Suchrate measurements will then be independent of anyabnormality of renal function. Single measure-ments are however often used because of theirsimplicity; for clinical examination can usuallydetect any suspicion of renal disease. For therecognition of thyrotoxicosis, an early measure-ment, i.e., one done at any phase during the firstfour to six hours, is most suitable. A late measure-ment, i.e., one at or up to the 48th hour, is bestsuited to the recognition of thyroid deficiency;for in partial thyroid deficiency the early uptakemay be normal or even rapid but it is then poorlyretained in the thyroid. A common compromisefor a simple routine test is the measurement of the24-hour thyroid' collection or uptake. The besturine test involves collections in three periods(usually 0-8, 8-24 and 24-48 hours); these canthen be measured quickly and accurately, anderrors in collection are usually revealed fromoddities in the subdivision between the samples.The radioiodine content of this second collectionperiod (8 to 24 hours) usually gives a good reflectionof thyroid uptake; and from the three periodsa 'T' index of thyroid uptake can be calculated

(o-8 hr. %) x oo00

(8-24 hr. %) x (0-48 hr. %)normal values=2.8-I3.

This 'T' gives a good approximation to thethyroid clearance rate, expressed as a percentageof extracellular fluid cleared per hour (Fraser et al.,1953). It is, of course, really an index of extra-renal disposal rate, and so even in completethyroid failure its value falls only to about 1-2.When errors of urine collection are suspected, i.e.,in up to Io per cent. of tests, another repeatradioiodine test based on neck measurementsmust be given. Another simple but less preciseindex of thyroid clearance may be derived bymeasuring the neck/thigh ratio at an early phasesuch as two hours (Pochin, I950). This however,requires special apparatus and the values obtainedare not in absolute units but depend on theapparatus used. With slightly more elaboration,the thyroid clearance may be measured directlyduring the first or second hour preferably afterintravenous doses, by various procedures based onneck measurements made in association witheither thigh, plasma, or urine measurements to

copyright. on A

pril 16, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.381.312 on 1 July 1957. D

ownloaded from

July 1957 RUSSELL FRASER: Diagnostic Procedures in Thyroid Disease 315

index the falling extracellular concentration fromwhich the clearance is occurring (Pochin, 1950;Foote and Maclagan, I951; Coenegracht andFraser, 1955).Plasma tests, the remaining type of test, are

based on the measurement in single plasmasamples of the radioactivity which has reached thehormonal fraction, i.e., the protein-bound 131I(P.B.131I); the samples are usually taken at 48or 72 hours. Such clearly index the rate at whichthe radioiodine passes through the thyroid, butneed a sensitive counter and a larger dose ofradioiodine than uptake tests. For this reasonthey are not usually suited to the recognition ofthyroid failure but are well suited to the recogni-tion of increased thyroid function. Along withall radioiodine tests, they are however subject tothe general limitations already noted. They areconvenient in requiring only one sample whichcan easily be measured in a distant laboratory,but have little other advantage over the types of testbased on uptake measurements. The results maybe expressed either simply as a concentration ofP.B.131I (Goodwin et al., 1951; Ingbar et al.,1954) as a 'conversion-ratio,' the ratio of theP.B.131I to iodide-131I in the plasma (Sheline et al.,1951), or possibly more sensitively as a ratio ofP.B.131I to salivary iodide (Thode et al., 1954).This last procedure takes advantage of the fact thatthe iodide of the plasma is concentrated into thesaliva, normally about thirty-fold; but this is notso important since the really critical measurementof this test is that of the plasma P.B.'11I. Asmost of the radioiodine in the plasma at 48 hoursis hormonal, even the measurement of total plasmaradioactivity gives a good simple index.

Interpretation of Results. Whatever type of testis used, the result is first to be interpreted asindicating either increased, decreased or normalthyroid cell activity. Wherever possible thisshould be done against the normal values of thedistrict for the type of test used. Excluding theknown iodine deficient areas, most urban areasgive a similar normal range. A normal or lowresult in suspected hyperthyroidism is extremelyrare, unless the previous iodine intake has beenabnormal, and so this can usually exclude such adiagnosis. In perhaps io per cent. of non-toxicgoitres, a high uptake will be found; these arehyperplastic goitres, many of which are iodinedeficient and some of which may have subclinicalhyperthyroidism. A high result will usually befound at the conclusion of a course of antithyroiddrugs, since these will have induced an iodinedeficiency. In lymphadenoid goitres, all types ofuptake may be found according to the stage of thedisease. Subacute thyroiditis is characterizedby absent thyroid uptake despite clinical evidence

of normal or even increased thyroid secretion.The possibility of depressed uptakes due toconcurrently administered drugs, especiallyiodides, antithyroid drugs, or thyroid extractmust always be considered. Moderate hypo-pituitarism, without associated clinical signs ofcomplete hypothyroidism, may have a normalthyroid uptake of radioiodine; as may the lesserdegrees of primary thyroid failure. Some typesof cretinism show a rapid transfer of radioiodine;this is due to incomplete organic binding of theiodine in the thyroid (McGirr and Hutchison,I955; Stanbury et al., 1955). Immediately aftersub-total thyroidectomy, radioiodine uptake isincreased, probably from iodine deficiency. Forapproximately two months after radioiodinetreatment, the uptake will be depressed due totemporary radiation effects. Where the clinicalpicture does not appear to accord with the radio-iodine findings, the supplementary radioiodinetests mentioned below are indicated, to clarifywhether these findings had reflected abnormaliodine stores.

Supplemented Radioiodine TestsAdditions to the sensitive radioiodine tests which

can check on iodine deficiency or excess, are nowavailable to overcome most of the limitations ofthese tests.

(i) Associated Chemical Measurements of Iodine.There is no simple procedure for estimatingchemically the body's iodide pool and therebytransforming the radioiodine uptake into absolutefigures (Riggs, 1952). However, a simple chemicalmeasurement on urine can reveal excess iodide,but it is unreliable for detecting subnormalamounts of iodide and so for detecting iodinedeficiency (Fraser et al., 1953).

(ii) Drug Loading Tests. To distinguish highuptakes due to iodine deficiency; the iodine-repletion test and the triiodothyronine suppressiontest.

The iodide repletion test is based on the simpletheory that an adequate period of iodide adminis-tration can correct an iodine deficiency, while asubsequent four weeks gap will permit eliminationof the coincident iodide-flooding, so that thepatient may then be retested free of any suspicionof iodine deficiency. The standard iodide loadrecommended is io mg. of potassium iodide(K.I.) given as a pill for 14 days, and the retestingis done four weeks later. Burrell and Fraser (I957)have found this a valid way of distinguishinganxiety states or non-toxic goitres with highthyroid uptakes from thryotoxicosis; and also ofdiscriminating between clinically-remitted and un-remitted cases of thyrotoxicosis at the conclusionof a period of antithyroid drug treatment.

copyright. on A

pril 16, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.381.312 on 1 July 1957. D

ownloaded from

316 POSTGRADUATE MEDICAL JOURNAL July 1957

The avid thyroid uptake of iodine deficiency isknown to depend mainly on anterior pituitarystimulation, which itself arises from the consequentlowered thyroid secretion (Burrell and Fraser,1957). The administration of normal or slightlysupra-normal amounts of thyroid hormone shouldtherefore suppress this avidity. The administra-tion of similar amounts of thyroid hormone tothyrotoxic patients does not however suppress theiravid radioiodine uptake; for whatever its basis it isnot a lack of secreted thyroid hormone. This isthe basis of the triiodothyronine suppression testwhich thus also distinguishes the avid thyroiduptake of thyrotoxicosis from that of iodinedeficiency; the uptake in the supplemented testbeing suppressed both in normals and in the lattercondition but not in thyrotoxicosis. In this testthe patient is given o.18 mg. of triiodothyroninedaily for eight days and then given the final radio-iodine test on the seventh day. Alternatively6 to 9 gr. of thyroid extract may be given daily andthe test given on the fourteenth day (Greer andSmith, 1954; Werner and Spooner, 1955;Perlmutter and Slater, 1955; Higgins and Fraser,1957)-These tests are helpful for assessing un-

expectedly high thyroid uptakes of doubtfulclinical significance; and are particularly usefulwhenever a patient needs testing after a course ofantithyroid drug treatment.' While the iodide-repletion test is simple and reliable, it involves adelay of six weeks, which is rarely important withthese doubtful cases; and it does correct the basiciodine deficiency which the other test leavesuncorrected. When a quick answer is neededand the patient is not seriously ill, the triiodo-thyronine test may be preferred.

(iii) Tests after thyroid stimulating hormone(T.S.H.). This offers a means of assessingwhether a patient taking thyroid really needs to doso (the thyroid administration need not bestopped), and also of whether a thyroid deficiencyis or is not dependent only upon pituitary failure.Lowered thyroid uptake not due to primarythyroid failure may be restored after administeringTSH daily for one to five days, even though thepatient is still taking thyroid.Chemical Estimation of Protein-bound Iodine (127I).

This test measures mainly the concentration ofthyroid hormone circulating in the plasma.Since it is technically difficult because it measuresin micrograms and not in milligrams per Ioo ml.,the preceding tests are generally to be preferred.Its main usefulness is for the recognition ofpartial thyroid deficiency, and its use shouldprobably be reserved for those cases suspectedof such a disorder who do not show a clearlyabnormal result on the preceding tests, and also

for cases suspected of thyrotoxicosis factita. The.latter is characterized by a high P.B.127I, associatedwith a low 131I uptake.The true normal range is probably 3.5 to 7.5

Ig./ioo ml., and of this i to 2 fg. is non-hormonaland probably iodide. But the accuracy of theestimation is rarely better than ± i g./Ioo ml.and is often poorer so that estimations on normalscan easily fall outside the true normal range;as reflected in the published normal ranges, someof which start from 2.6 and some of which rise to11.3 gLg. per oo0 ml. Values within i jig. of thenormal limits are therefore never certain indicatorsof abnormal thyroid function; and similarly somemild cases of thyrotoxicosis and also a few withthyroid deficiency are found to show normal values(Blackburn and Power, 1955; Bierwalters, 1956;Fraser, 1956). Further sources of error to beremembered are the low values which are foundwhen the serum albumin is low or when thepatient has received mercury, the slightly highvalues found in normal pregnancy, and the highvalues which may persist for many months afterthe administration of organic iodine-containingdrugs.Subsidiary Indirect Tests

(i) Creatinuria is a feature of most thyrotoxicpatients (Treusch et al., 1944) but also of severalother conditions involving muscular wasting;and is associated with raised serum creatine levels.For the recognition of thyrotoxicosis, it has beensuperseded by the preceding tests.

(ii) Serum Cholesterol and Other Signs of ThyroidDeficiency. The recognition of the minor gradesof thyroid deficiency remains the Achilles heel ofthyroid tests. Most cases of thyroid deficiencyshow a raised serum cholesterol level, and itsdemonstration is an important corroboration of themeaning of a low B.M.R. But low levels are nota diagnostic feature of thyrotoxicosis. Withchildren suspected of myxoedema the X-rays canhelp by revealing epiphyseal dysgenesis which is aspecific sign of thyroid deficiency. Confirmationof the signs of myxoedema may also be had fromthe E.C.G. which in thyroid deficiency shows arather non-specific lowering of ' T' waves.When there is still doubt, a therapeutic trial withtwo weeks of thyroid extract, i g. per day may begiven; and the deficient patient will then showover 20 per cent. lowering of serum cholesterollevel and a rise in the height of the ' T' waveson the E.C.G.We may conclude that the tests at present

available can recognize all treatable thyrotoxicosis,and most treatable thyroid deficiency, providedtests are chosen which are appropriate to thepatient's total clinical state.Bibliography continued on page 321

copyright. on A

pril 16, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.381.312 on 1 July 1957. D

ownloaded from

July 1957 POCHIN: The Place of Radioactive Iodine in the Treatment of Thyroid Disease 321

size or vascularity may be more easily obtained.The treatment by these means of a retrosternalgoitre which is already causing pressure symptoms,involves the careful use of repeated small doses ofradioiodine to avoid the risk of oedema of theirradiated tissue in the days following the -dose,and an increase in obstruction. For these reasons,radioiodine has only a limited use in reducing thesize of ectopic thyroid deposits but may occa-sionally be of value.

REFERENCESI. BLOMFIELD, G. W., JONES, J. C., MACGREGOR, A. G.,

MILLER, H., WAYNE, E. J., and WEETCH, R. S. (x955),Brit. med. J., ii, 1223.

2. CHAPMAN, E. M., and MALOOF, F. (1955), Medicine(Baltimore), 34, 261.

3. CLARK, D. E. (I955), J.A.M.A., 159, 1007.

4. COLIEZ, R., TUBIANA, M., DUTREIX, J. M., andLAUGIER, A. (1956), Bull. Cancer, 43, 218.

5. DUFFY, B. J., and FITZGERALD, P. J. (1950), J. clin.Endocr., 10, 1296.

6. FRASER, R., ABBATT, J. D., and STEWART, F. S. (I954),Brit. J. Radiol., 29, 434.

7. HAHN, P. F. (1956), 'Therapeutic Use of Artificial Radio-isotopes' (New York: John Wiley & Sons, Inc.).

8. HAMILTON, J. G., and LAURENCE, J. H. (1952), J. lin.Invest., 21, 624.

9. HERTZ, S., and ROBERTS, A. (1942), J. din. Invest., 21, 624.xo. HILTON, G. (i956), Brit. J. Radiol., 29, 297.iI. MALOOF, F., VICKERY, A. L., and RAPP, B. (1956),

J. din. Endocr., x6, I.12. POCHIN, E. E. (1957), 'Moder Trends in Endocrinology'

(ed. Dr. H. Gardiner-Hill), in the press.13. POCHIN, E. E., CUNNINGHAM, R. M., and HILTON, G.

(I954), JY din. Endocr., 14, 1300.14. POCHIN, E. E., and HALNAN, K. E. (I957), Metabolism., 6,49S5. RAWSON, R. W., RALL, J. E., and ROBBINS, J. (I953),

Arch. intern. Med., 92, 299.16. SIMPSON, C. L., HEMPELMANN, L. H., and FULLER,

L. M. (1955), Radiology, 64, 840.

HOW TO GET THEREAn Address Book for the Medical profession, showing how to reach thevarious Colleges, Societies, Institutes and Hospitals in or near London

New (Fourth) Edition: 1954 Price 2s. 6d. (2s. 10d., post free)Published by the

FELLOWSHIP OF POSTGRADUATE MEDICINE60 Portland Place, London, W.I

Continued from page 316-Russell Fraser, M.D., F.R.C.P., D.P.M.BIBLIOGRAPHY

AUB, J. C., and DU BOIS, E. F. (1917), Arch. intern. Med., I9, 823.BEHNKE, A. R. (1953), Ann. N.Y. Acad. Sci., 56, i095.BEIERWALTES, W. H. (1956), Ann. intern. Med., 44, 40.BENEDICT, F. G., and HENDRY :M. F. (192I), Boston med.

surg. J., 184, 217.BIERRING, E. (193I), 'The Standard Metabolism of Boys (7 to

x8 years inclusive),' Levin and Munksgaard, Copenhagen.BLACKBURN, C. M., and POWER, M. H. (I955), J. din. Endocr.,

5I, I379.BOOTHBY, W. M., BERKSON, J., and DUNN, H. L. (1936),

Amer. J. Physiol., 116, 468.BURRELL, C. D., and FRASER, R. (I957). Quart. J. Med. To

be published.COENEGRACHT, J., and FRASER, R. (I955), J. Endocr., 12, 185.FOOTE, J. B., and MACLAGAN, N. F. (I95I), Lancet, i, 868.FRASER, R. (1956), Ibid., ii, 581.FRASER, R., HOBSON, Q. J. G., ARNOTT, D. G., and EMERY,

E. W. (I953), Quart. J. Med., N.S., 22, 99.FRASER, R., and NORDIN, B. E. C. (I955), Lancet, i, 532.GOODWIN, J. F., MACGREGOR, A. G., MILLER, H., and

WAYNE, E. J. (1951), Quart. J. Med., 20, 353.GREER, M. A., and SMITH, G. E. (1954), J. lin. Endocr., 14, 374.HIGGINS, H. P., and FRASER, R. (1957). To be published.IBBERTSON, K., JOPLIN, G., and FRASER, R. (1957), Brit.

med. J. In press.INGBAR, S. H., FREINKEL, N., HOEPRICH, P. D., and

ATHENS, J. W. (1954), J. din. Invest., 33, 388.

KEATING, F. R., Jr., HAINES, S. F., POWER, M. H., andWILLIAMS, M. M. D. (1950), J. clin. Endocr., Io, I425.

KLEIBER, M. (I947), Physiol. Rev., 27, 5I.KLEIBER, M. (1956), Ann. Rev. Physiol., 18, 35.MILLER, A. T., and BLYTH ,C. S. (i953),J. appl. Physiol., 5, 311.MYANT, N. B. (1952), Brit. med. Bull., 8, 14x.McGIRR, E. M., and HUTCHISON, J. H. (1955), 7. din. Endocr.,

15, 668.PERLMUTTER, M., and SLATER, S. (1955), J. Amer. med. Ass.,

158, 718.POCHIN, E. E. (1950), Lancet, ii, 41.RALL, J. E. (1956), Amer. J. Med., 20, 719.RAPPORT, R. L., CURTIS, G. M., and SIMCOX, S. J. (1951),

J. clin. Endocr., II, 1549.RIGGS, D. S. (1952), Pharmacol. Rev., 4, 284.ROBERTSON, J. D., and REID, D. D. (1952), Lancet, i, 940.SHELINE, G. E., MOORE, M. C., KAPPAS, A., and CLARK,

D. E. (1951), J. clin. Endocr., Ix, 91.SHOCK, N. W., and SOLEY, M. H. (1939), J. Nutr., 18, I43.STANBURY, J. B., OHELA, K., and PITT-RIVERS, R. (1955),

J. din. Endocr., 15, 54.THODE, H. G., JAIMET, C. H., and KIRKWOOD, S. (I954),

New Engl. J. Med., 25x, 129.TREUSCH, J. V., KEPLER, E. J., POWER, M. H., and HAINES,

S. F. (I944), Amer. J. med. Sci., 208, 310.VOGELIUS, H. (1945), Acta med. scand., Suppl. I65.WAYNE, E. J. (I954), Brit. med. J., i, 411.WERNER, S. C., and SPOONER, M. (1955), Bull. N.Y. Acad.

Med., 31, I37.

copyright. on A

pril 16, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.33.381.312 on 1 July 1957. D

ownloaded from