Further studies on thyroxin-binding globulin-dependence in equilibrium dialysis assays of free thyroxin.

is between 0.40 and 0.70 mmol/L. Granted, our patients are type I rather than type U diabetics, but both are allegedly prone to have increased concentrations of free glycerol in plasma. 3. We serve as the clinical laboratory for the University of Minnesota's hyperlipidemia clinic and I see patients in that clinic. This clinic's patient population is principally referral patients with more severe hyperlipi-demias. Numerous patients have tri-glycerides >3.4 mmoL'L and it is not unusual to see triglycerides >10.0 mmol/L. Many of these patients with grossly increased triglycerides are type U diabetics,similar to those studied by Uprichard et a!. in the gemS-brozil trial. Until the summer of 1989, our laboratory routinely measured free glycerol on all clinical specimens, subtracted it from an unblanked tri-glyceride, and reported a true triglyc-eride. We virtually never saw free glycerol values >0.30 mmol/L in any of these outpatients, even those with markedly high concentrations of trig-lycerides. Many of the hyperlipidemic outpatient results shown in Figures 1 and 2 of our original report (1) were from those being seen in this clinic. Parenthetically, I would add that the main reason that we began performing free glycerol measurements about seven years ago was that it is absolutely necessary to blank for free glycerol to get the " right " answer for the proficiency-testing specimens from the Centers for Disease Control/Na-tional Heart, Lung, and Blood Institute. These specimens often have markedly increased free glycerol (0.60-1.00 mmol/L), which I believe was added either intentionally or unintentionally (glycerol is used as a wetting agent in the microbiological filters used to sterilize many quality-control sera). For outpatient specimens ,I personally do not think blank-ing for free glycerol is necessary, a!-though for someinpatients it may be. I suspect that the marked differences in the free glycerol concentrations for type II diabetics reported by Uprichard et al., compared with our experience and that reported by Schwertner et al., may be related to a methodologicalproblem. Uprichard et al. did not actually measure free glyc-erol, as Schwertner et al. and we did. Rather, they inferred it from the difference in results using two methods for triglycerides: one with known interference from free glycerol, and one supposedto be free of glycerol interference .Absolute calibration of triglycer-ide methods is a problem, and exactly how a method is blanked is especially important for turbid specimens (3). The two methods usedby Uprichard et a!. may have subtle differences in …