Resumo

ARRANZ CALZADO, Celeste et al. Reference criteria for insulin secretion indicators and of the lipid parameters in a hospital mixed population. Rev Cubana Endocrinol [online]. 2010, vol.21, n.1, pp. 1-12. ISSN 1561-2953.

INTRODUCTION: the clinic usefulness of many indicators for the diagnosis of insulin-sensitivity and resistance is studied and frequently applied using the same criteria of diagnostic interpretation proposed by authors. The differences among the particular features of the different populations and the methodology used for determination of glycemia and the insulinemia over the cut point used for diagnosis require the use of diagnostic criteria representative of that population. It is difficult to establish particular criteria because of the complexity and the cost of the conventional direct methodology using reference populations is inaccessible in general to individual laboratories. There is a simpler alternative, the estimation indirect method of reference ranks, based on segregation of hospital mixed populations in its gaussean components that could be more appropriate to obtain non-biased diagnostic criteria. OBJECTIVE: to estimate the reference limits of main indicators for insulin-resistance and the validity of results obtained using this method. METHODS: authors studied the statistic parameters and reference values of 8 indicators of insulin secretion, 3 indicators of resistance and 5 indicators of insulin-sensitivity, derived from relation between fasting glycemia and insulinemia from 1 000 consecutive samples of hospital mixed population seen in National Institute of Endocrinology. We applied this methodology to other parameters with an asymmetric distribution in population; by example, insulin and serum lipids (cholesterol and triglycerides) having reference values of universal acceptation used as an internal control to assess if the method applied offers non-biased results. We applied the FAO program called Fish Stock Assessment Tools (FISAT) to separate the value of population in its Gaussean components and to estimate the statistic parameters of component created by health population. We compared the results obtained using this methodology with cut value used in by this institution. RESULTS: the methodology used allowed us to determine the statistic parameters or "normal" component of study mixed population and provides reference ranks for glycemia (3,22-6,10 mmol/L), insulinemia (32-68 pmol/L), cholesterol (2,2-5,28 mmol/L) and triglycerides (0,65-1,97 mmol/L) similar to those generally accepted considered as a criterion of method validity. Results obtained in the insulin sensitivity and resistance indicators gave us a greater diagnostic sensitivity with a lower specificity. CONCLUSIONS: we concluded that the use of hospital mixed samples and of methodology described allows obtaining in a reliable and simple way reference ranks representative of the local conditions.

Palavras-chave : mixed sample; gaussean components; FISAT program; indicators; insulin resistance and sensitivity.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )