Serum Triglycerides and Fatty Acid Incorporation into Human Adipose Tissue (FIAT)

ABSTRACT. Fatty acid incorporation into adipose tissue (FIAT), the metabolic process assimilating plasma triglyceride fatty acids liberated by lipoprotein lipase, was recently found to be lower in hyper‐than in normotriglyceridaemia. In the present report, the relation of FIAT to glucose tolerance and adipose tissue morphology and fatty acid composition has been studied in a population of men with normo‐ and hypertriglyceridaemia, using needle biopsy specimens. In addition, the associations between plasma triglyceride concentration and these factors as well as FIAT were examined by statistical methods. FIAT and GLIAT (glucose incorporation into adipose tissue) activities per cell were positively correlated with fat cell diameter but not with fat cell number. FIAT activities per cell and per unit surface area were lower in hyper‐ than in normo‐triglyceridaemic subjects. The k‐value of the i.v. glucose tolerance test and glycerol release from adipose tissue did not correlate with FIAT or GLIAT activities. The proportion of stearic acid in adipose tissue was negatively correlated with the serum triglyceride level and with fat cell diameter, but positively correlated with FIAT. Linolenic acid in adipose tissue correlated positively with the k‐value. The negative correlation between serum triglycerides and FIAT remained when the other variables which were significantly correlated with FIAT or the serum triglycerides were entered in partial correlation analysis. These results suggest that although low FIAT activity is related in part to other characteristics, it occurs in hypertriglyceridaemia independent of glucose tolerance or various characteristics in fat. With serum triglyceride concentration as dependent variable, stepwise regression analysis was performed, entering all other variables as independent ones. The highest multiple k‐value was 0.76 (p<0.001) and it was obtained with three adipose tissue parameters: FIAT (or GLIAT), content of linolenic acid and of stearic acid. The other parameters did not give rise to any further improvement in the prediction of the serum triglyceride concentration which is better than 50% (R2=0.57).