BLOOD SUGAR TIME CURVES

Blood sugar time curves, or what are more generally, but less exactly, termed sugar tolerance curves, are now receiving wide clini-cal application. As a result of this, various disorders, other than dis-betes mellitus, are known to result in, or be associated with, a disturb-ance in the metabolism of carbohydrates. The response of the indi-vidual to glucose ingestion is probably the best of indices of such a disturbance. It is obvious that the "normal " type of curve must be recognized in order to detect pathological variations. Within the last few years, many different results, obtained in apparently normal individuals, have been recorded. A study of the literature however, shows that in obtaining these results, not only have various procedures been followed as to selection of the amount and type of sugar, and the periodic col-lection of blood and urine specimens, but the analytical methods for blood sugar estimation have been different. The literature on the subject has become voluminous, and for brevity no attempt at review will be made here, since it would serve no particular purpose in the present discussion. In spite of the many procedures and different technical methods, a statistical study of all the results would tend to give a fairly approxi-mate idea of what is the response of the normal individual to glucose ingestion. This view is based upon one of the theorems of statistics, namely, that the arithmetical mean of a large series of observed values, is the most probable value of the quantity measured. Such painstak-ing study has, recently, been made (1). In a comprehensive study, including a statistical analysis of 4000 individaal blood sugar deter-minations in 971 cases, Gray arrived at standards which are generally accepted. The values of the mean and the mode (the most frequen