EFFECTS OF INANITION ON THE BODY AS A WHOLE 87 



only 43 cases, or 12.9 per cent, was the index above the normal average. The 

 remaining 87 per cent are below the normal average to a variable extent. 



This brings us to the difficult problem of determining, if possible, the point 

 below which the weight-height index should be considered abnormal, denoting 

 pathological emaciation. It must be remembered that in general at any age, 

 and in each sex of a given race, the individual heights and weights are normally 

 variable, the frequencies at various intervals above and below the average being 

 distributed roughly according to the "probability curve." Robertson ('23) 

 estimates that in children of a given age the normal variability in weight varies 

 (at different ages) from 11-20 per cent, and that in stature from 4-6 per cent. 

 This necessarily affects the normal weight-height index, which Matusiewicz 

 ('14), Pfaundler ('16), and Gerber ('21) have shown to be subject to marked 

 individual variation. A low index therefore does not necessarily indicate 

 malnutrition in the individual case, as Pfaundler ('21) and others have repeatedly 

 emphasized. On the other hand, malnutrition may exist in some cases without 

 marked depression of the weight-height index. Nobecourt ('16) describes two 

 types of infantile denutrition,one ("cachexie") with emaciation and one ("hypo- 

 trophie") without. With reference to the racial factor, Dublin and Gebhart 

 ('23) have shown that the Wood -Woodbury weight-height tables fail as an index 

 of malnutrition in Italian children of New York. Davenport ('23) concludes 

 that body-build is determined largely by heredity, probably acting through the 

 endocrine glands. 



Of the various indices proposed, aside from Bornhardt's index (or others 

 involving the thorax), the weight-height index is perhaps the most convenient 

 and reliable single physical index of the nutritional condition of the body, but 

 no single index can be conclusive for every individual case. A low weight-height 

 index is presumptive evidence of malnutrition, but it needs confirmation by 

 clinical evidence. The various physical signs and symptoms of malnutrition in 

 children have recently been summarized by Roberts ('23) and Goldberger ('23). 

 As a result of clinical experience, Holt ('18) proposed an arbitrary zone of 10 

 {or 12) per cent or more below the normal weight-height relation as an indication 

 of malnutrition, while Emerson and Manny ('20) and Emerson ('22) draw the 

 line at 7 per cent (below the Boas-Burk norm). Baldwin ('24) finds in healthy 

 school-children an average variability of 6-9 per cent in weight, according to 

 height, age and sex. Any such boundary line is necessarily arbitrary and mis- 

 leading, however, unless the limitations above mentioned are kept clearly in 

 mind. Thus in a series of 506 healthy children, Clark, Sydenstricker and Collins 

 ('23) found 13 per cent of the individuals more than 10 per cent underweight 

 according to Dreyer's standard (stem length and chest circumference tables) ; 

 20 per cent were more than 10 per cent underweight according to the Wood 

 standard (height-weight-age tables); and 17 per cent were subnormal accord- 

 ing to Pirquet's standard, having a " pelidisi " of 94 or less. None of the indices 

 so far proposed appears to be very closely correlated with clinical findings 

 (Huth '21; Van der Loo '22; Clark, Sydenstricker and Collins '23; Helmreich 

 and Kassowitz '23 and others). The subject needs further investigation {cf. 

 Taylor '22). 



