Mortal factors in type 2 (NIDDM) diabetes mellitus

Five years after entry into a study prospective from diagnosis of non-insulin-dependent (Type 2) diabetes, 17 patients were known to be dead and 197 alive. On analysis of 136 patients (12 deaths) in whom there was complete information from clinical, metabolic and hormonal examination both before and 1 yr after start of treatment, an index of liability to death within this first 5 yr was calculated, to separate best the dead from the living. One index value correctly ascribed 83% of the dead and 90% of the living (89% of all predictions correct; Youden's "J" value = 0.73). Estimates from the less complete data on the other 78 patients (5 deaths) did not alter the predictive factors. Prediction is more successful with non-cardiac than with cardiac deaths. Five factors contributed to the predictive index for death. One was a long duration of symptoms as recalled at diagnosis. The other 4 factors all came from the 1-yr review (no death then). These were (a) greater glucose intolerance, as expressed by the KG rate constant for disappearance of i.v. injected glucose from the blood (more useful than fasting glucose concentration); (b) higher systolic blood pressure (more useful than diastolic); (c) less obesity, as expressed by the Body Mass Index, and (d) higher fasting blood glycerol concentration. As expected, hyper-glycaemia (as reflected in the KG value) is important, but adrenergic factors may contribute to (a), (b) and (d). From these results earlier diagnosis of this type of diabetes could well be associated with an improved outlook, and in addition this alone might ameliorate the other factors, but measures to normalise them post-diagnosis also require testing.