Abstract: Reference values are proposed for the concentrations of As, Cd, Co, Cr, Cu, Fe, Hg, Mn, Mo, Pb, Se, and Zn in whole blood, blood serum, urine, milk, liver, and hair from adult human subjects. For F, I, and Ni, it was not possible to evaluate reference intervals for all the specimens mentioned above. For several elements, including Al, B, Br, Cs, Li, Rb, U, and V, the present status of the literature does not provide an adequate basis for formulating baseline concentrations; therefore, results from selected investigations are listed for information only. For elements such as Cu, Fe, and Zn, which are known to be homeostatically controlled, the concentrations in whole blood and blood serum follow a gaussian-like frequency distribution, and we could consider both median and mean values for evaluation. On the other hand, elements whose concentrations in tissues and body fluids are influenced by dietary availability (e.g., As and Se) or environmental factors (e.g., Cd, Hg, and Pb) show wide scatter. In these cases, the median appeared to be a better indicator of the central tendency than the mean, when different populations are involved. These points are illustrated.

Abstract: This is an equilibrium-type radioimmunoassay for the amino-terminal propeptide of type III procollagen (PIIINP), which overcomes the problem of nonparallelism between the standard and human serum samples encountered with earlier assays. Proper selection of antiserum and reaction conditions diminishes interference from degradation products of the propeptide in serum. Because a rapid solid-phase-bound second-antibody step is included, the assay takes only 3 h. The intra-assay and the interassay CVs are both about 5%. In infants and children the concentration of PIIINP in serum closely parallels the growth-velocity curve. For 88 presumably healthy adults, the PIIINP concentration was 1.7-4.2 micrograms/L, about a third that measured with the previously available commercial assay. This is because of lack of inhibition by small Col 1 domain-related degradation products.

Abstract: Microbiological assay is still widely used for estimating folic acid derivatives in serum and other biological samples. We describe here a modification of this procedure involving use of 96-well microtiter plates. This procedure, used with modern, computer-interfaced microtiter-plate readers and data-reduction software, greatly shortens the time and minimizes reagent costs for this assay. Under the conditions of our assay procedures, all folic acid derivatives tested gave equal growth response for Lactobacillus casei. Results for assays of rat liver extracts showed excellent agreement between the standard bioassay and the 96-well procedure.

Abstract: In the United States, coronary heart disease is the major cause of death and disability in women and in men. Despite this, little is known about the risk factors, including cholesterol and lipoprotein concentrations, for coronary disease in women. In this paper we review the determinants of cholesterol and lipoprotein concentrations in women, assess whether values for total cholesterol and lipoproteins (HDL and LDL) are associated with the occurrence of coronary heart disease in women, and evaluate the evidence that suggests that modifying the concentrations of lipids in women is associated with changing the risk of coronary disease. Besides genetic determinants, dietary cholesterol, dietary fat, total caloric intake, alcohol consumption, cigarette smoking, and physical activity are known to influence concentrations of lipids in women. Some of the strongest determinants of cholesterol and lipoprotein concentrations in women are sex hormones, including estrogen and progestin. Exogenous use of both of these hormones markedly influences HDL and LDL cholesterol; additional evidence suggests that endogenous sex hormones also influence lipid and lipoprotein concentrations. The few studies that have examined the association of total cholesterol with coronary heart disease occurrence and mortality in women have consistently shown that (a) women have much lower rates of coronary heart disease than men at the same values for cholesterol, and (b) clearly elevated risk for coronary heart disease in women is evident only at relatively high values of total cholesterol (i.e., greater than 260 mg/dL). There also appears to be an age effect, with total cholesterol concentrations being more predictive in older than in younger women.

Abstract: We describe a new clinical laboratory instrument, the IMx, used to automate immunoassay testing in the clinical laboratory. The IMx incorporates a novel technology called Microparticle capture Enzyme ImmunoAssay (MEIA) for assays of high-molecular-mass analytes, and fluorescence polarization immunoassay (FPIA) for hapten assays. A front-surface fluorometer is used to quantify the enzymatic generation of fluorescent product at a rate proportional to the concentration of the analyte in an MEIA, and a fluorescence polarization optical system is used to quantify results in an FPIA. The microprocessor-based instrument uses a robotic arm with two degrees of freedom and a rotating carousel to process the samples for assay. One assay can be done on each of 24 patients' specimens in 30 to 40 min with "walk-away" automation. Calibration curves are stable for at least two weeks. Instrument control involves software-labeled "command keys," a numeric keypad, and an interactive display. Results are output to a thermal printer or computer interface.

Abstract: We investigated the effects of storage and handling on measured values for carotenoids, retinol, and tocopherol in plasma. We found no significant differences in the concentrations of these analytes measured in plasma samples that were frozen immediately after separation as compared with replicate samples maintained at room temperature in the dark for 24 h. Analytes were stable in solvents for at least 18 h at 23 degrees C after extraction. Purging samples with nitrogen gas before freezing had no detectable beneficial effects. All analytes were stable in plasma stored at -70 degrees C for at least 28 months or at -20 degrees C for five months. By 15 months the concentrations of carotenoids were significantly less (P less than 0.05) in plasma stored at -20 degrees C than in plasma stored at -70 degrees C, while retinol and tocopherol concentrations were not significantly different. Concomitant with the decrease in carotenoids was the appearance of unidentified peaks in the ultraviolet. Adding ascorbic acid or butylated hydroxytoluene antioxidants to the precipitating solvent did not alter the losses of carotenoids or alter the appearance of unidentified peaks. Under appropriate conditions, plasma carotenoids, retinol, and tocopherol are stable for more than two years.

Abstract: We describe a rapid, sensitive method for the routine simultaneous determination of nicotine and cotinine in 1 mL of plasma. Extraction in 10-mL screw-capped Teflon tubes with methylene chloride after deproteinization with trichloroacetic acid eliminated emulsion formation. The extract, after evaporation and reconstitution in 30 microL of mobile phase, is injected into a reversed-phase C-18 ion-pair column of an isocratic high-performance liquid-chromatographic unit. Absorbance is monitored at 256 nm. The mobile phase is a citrate-phosphate (30 mmol each per liter) buffer mixture containing 50 mL of acetonitrile and 1 mmol of sodium heptanesulfonate per liter. 2-Phenylimidazole is the internal standard. The detection limit is 1 microgram/L for nicotine and 3 micrograms/L for cotinine. The standard curve is linear from 0 to 700 micrograms/L for both compounds. The average CV for nicotine in the concentration range 0-100 micrograms/L is 6.5%, and that for cotinine in the concentration range 50-700 micrograms/L is 4%.

Abstract: We quantified indoxyl sulfate in uremic serum by using internal-surface reversed-phase high-performance liquid chromatography. Its concentrations were markedly increased in chronic hemodialysis patients, and were significantly but weakly correlated with the concentrations of creatinine and beta 2-microglobulin in these patients' serum, and with the duration of their hemodialysis treatment. Indoxyl sulfate could not be removed effectively by conventional hemodialysis because of its strong binding to serum albumin. Equilibrium dialysis demonstrated that indoxyl sulfate inhibited the binding of salicylate to albumin, and that 3-carboxy-4-methyl-5-propyl-2-furanpropionic acid inhibited the binding of indoxyl sulfate to albumin. In conclusion, indoxyl sulfate was markedly accumulated in uremic serum, and inhibited drug binding.

Abstract: Lipoprotein cholesterol concentrations in plasma are routinely estimated by using the Friedewald formula, whereby very-low-density lipoprotein cholesterol (VLDL-C) is estimated to be one-fifth the plasma triglyceride concentration. Ordinarily, this formula is applied only to plasma sampled from patients in the fasted state. To determine whether lipoprotein cholesterol measurements are altered substantially in plasma sampled from nonfasting subjects, we obtained postprandial blood samples from 22 healthy subjects (nine men, 13 women, ages 22-79 years) fed a fat-rich meal (1 g fat per kilogram body wt.). The plasma triglyceride concentration increased postprandially in all subjects (233 +/- 16% of baseline at 3 h). The mean cholesterol concentration in plasma was essentially unchanged. High-density lipoprotein cholesterol (HDL-C) was significantly decreased (94 +/- 2% at 3 h, P less than 0.001). VLDL-C and low-density lipoprotein cholesterol (LDL-C), estimated by the Friedewald formula, were compared with measurements obtained by modified Lipid Research Clinics (LRC) methodology. As measured by either method, VLDL-C increased and LDL-C decreased significantly after the fat-rich meal. These postprandial changes were significantly greater (P less than 0.01) when estimated by the Friedewald formula than by LRC methodology. We conclude that (a) lipoprotein cholesterol concentrations measured in the fed subject differ significantly from those measured in the fasted subject, and (b) plasma must be obtained after at least a 12-h fast if an individual's risk of coronary heart disease is to be accurately assessed.

Abstract: Illicit-drug users may attempt to falsify results by in vitro adulteration of specimens. We investigated eight additives (NaCl, Visine, handsoap, Drano, bleach, vinegar, golden-seal tea, and lemon juice) claimed by drug users to invalidate enzyme immunoassay (EIA) drug assays. We also analyzed adulterated urine specimens to determine if they could be identified, adding adulterants at several concentrations to 222 EIA-positive specimens confirmed by gas chromatography and mass spectrometry (GC/MS) to contain illicit drugs. To identify adulterated urines, we monitored pH, relative density, and urine color and turbidity at adulterant concentrations that falsified EIA results. Specimens contaminated with NaCl had relative densities greater than 1.035. Liquid Drano, bleach, and vinegar shifted urine pH outside the physiological range. Golden-seal tea caused a dark appearance, and specimens containing liquid soap were unusually cloudy. Lemon juice had no effect on the assays. Visine was the only adulterant not detected. The adulterants interfered somewhat differently with each of the drug assays. EIA assays for illicit drugs can be invalidated by specimen adulteration producing false-negative results. Therefore, if urine drug testing is to be conducted, pH, relative density, and appearance should be assessed and suspect specimens should be rejected. Not all adulterants can be detected, so observed collection is strongly recommended.

Abstract: Age- and sex-specific reference intervals based on the 0.025 and 0.975 fractiles of data derived from a healthy pediatric population are presented for zinc, copper, selenium, iron, ferritin, retinol, alpha-tocopherol, and related analytes in serum. Age was an important covariate for copper, selenium, retinol, and tocopherol, and ferritin in boys. Strong correlations were found between retinol and retinol-binding protein, prealbumin (transthyretin), alpha-tocopherol, and selenium. Tocopherol was highly correlated with both cholesterol and triglycerides. We found no relationship between serum zinc and either retinol or retinol-binding protein. Despite exclusion of children in whom anemia, microcytosis, or variant hemoglobins were found, the 0.025 fractile for iron in several age groups was even less than the concentration considered to indicate poor iron nutritional status.

Abstract: Using the Ektachem-700 multilayer film analyzer, we defined age- and sex-specific reference intervals for 20 analytes in sera from a healthy population of neonates and children ages one to 19 years. Upper and lower normal reference intervals for each analyte were determined by nonparametric methods as the 0.975 and 0.025 fractiles, respectively. Newborns have lower concentrations of total protein and albumin, and higher concentrations of phosphate, bilirubin, and enzymes in serum than older children do. Concentrations of urea, glucose, calcium, phosphate, and bilirubin change rapidly postnatally. Outside the neonatal period, no significant age- or sex-related difference was found for plasma glucose, serum amylase, conjugated or unconjugated bilirubin, or lipase. There was no sex-related difference in reference intervals for albumin, total protein, calcium, phosphate, or urea. However, concentrations of uric acid and creatine kinase are much higher in postpubertal boys than in girls. Alkaline phosphatase values peak later in boys. Except for lactate dehydrogenase and gamma-glutamyltransferase, the reference intervals defined here do not differ strikingly from data derived with use of other analyzers. The age- and sex-related trends are independent of method. However, each laboratory should determine the degree to which these reference ranges can be directly applied to analyses performed with another analyzer.

Abstract: Slight albuminuria, an overnight albumin excretion rate (AER) greater than 30 micrograms/min in an "Albustix"-negative sample, predicts development of diabetic nephropathy. This study compares the AERs for 261 timed overnight urine collections with the albumin concentrations and albumin/creatinine ratios for the same specimens (equivalent to first morning specimens). Thirty-one specimens (11.9%) had AERs greater than 30 micrograms/min. Use of an albumin/creatinine ratio greater than 3.0 mg/mmol to predict an AER greater than 30 micrograms/min gave a sensitivity of 96.8%, a specificity of 93.9%, and a predictive value of 68.2%, with a correlation coefficient of 0.921. Use of an albumin concentration greater than 17 mg/L gave a sensitivity of 96.8%, a specificity of 90.9%, a predictive value of 58.8%, and a slightly poorer correlation (r = 0.904). Evidently either method is acceptable as an initial screening procedure, but determination of albumin concentration alone would be preferable because of lesser cost.

Abstract: Accurate laboratory measurement of serum cholesterol has become a national public health priority National proficiency testing surveys indicate that laboratory inaccuracy in cholesterol testing is more of a problem than precision Like precision, accuracy is a function of multiple pre-analytical and analytical sources of variation Controlling pre-analytical sources of variation helps minimize such sources of variation as intraperson biological, behavioral, and clinical differences, and variations caused by sample collection, handling, and shipping Standardization of analytical sources of variation helps to achieve and maintain desirable analytical performance, accurate reporting, and correct interpretation of a reported cholesterol result The intraperson total variation in lipoproteins and their constituents is of primary interest when one is interpreting a single result or a series of results from a single person The mean of multiple specimens from the same person is required if one is to obtain an accurate value for intraperson total cholesterol and minimize pre-analytical sources of variation Standardizing analytical sources of variation in some instrument systems requires standardizing results by using "fresh" patients' specimens

Abstract: We evaluated a recently developed commercial assay for quantifying thrombin-antithrombin III (TAT) complexes in human plasma. The assay is precise (within-assay CV less than 10%, between-assay CV less than 13%), and sensitive (detection limit 0.7 micrograms of TAT per liter of plasma). Measurements for healthy volunteers yielded a normal reference (95 percentile) interval of 0.8 to 5.0 micrograms/L (n = 50, mean 2.1 micrograms/L, range 1.1 to 7.5 micrograms/L). TAT concentrations were increased in 25 of the 41 patients who fulfilled the clinical criteria of disseminated intravascular coagulation (DIC, overall mean 15.8 micrograms/L) and in 30 of the 35 patients with deep-vein thrombosis of the leg (overall mean 9.4 micrograms/L). We assessed the accuracy of the TAT assay by comparison with established criteria for the laboratory diagnosis of DIC involving various cutoff values for antithrombin III, factor V, fibrinogen, platelet count, fibrin/fibrinogen degradation products, and activated partial thromboplastin time. The low specificity of the TAT assay with regard to some of these criteria indicates that the latter are probably insensitive.

Abstract: Measurements of carcinoembryonic antigen (CEA) in blood increased dramatically in some patients who were receiving injections of monoclonal antibody. CEA titers were measured with a monoclonal antibody-based double-determinant enzyme immunoassay in which untreated plasma specimens were diluted with an equal volume of buffer containing mouse serum. Increasing CEA titers were accompanied by the appearance and coincident increase in titers of human antibody against mouse Ig (HAMA). Adsorption of these sera with solid-phase anti-human IgG or Protein A restored antigen titers to pretreatment values; evidently the serum factor eliciting false-positive CEA titers was most probably HAMA. Neither addition of undiluted mouse serum to the assay mixture nor pretreatment by heating plasma specimens to 70 degrees C effectively abolished HAMA interference. By contrast, protein precipitation with polyethylene glycol (130 g/L) or heating plasma samples to 90 degrees C eliminated false-positive titers caused by HAMA, but did not reduce authentic CEA titers.

Abstract: S100 protein (S100) was assayed by particle counting immunoassay in serum samples from 50 healthy individuals, 325 patients with various neurological disorders, and 20 patients with malignant melanoma. The detection limit for this protein was 0.3 microgram/L. We detected none in healthy individuals or in 50 patients with multiple sclerosis, 23 with dementia, or 20 with meningitis. S100 was detectable in serum of only a few patients with meningoradiculitis (2/20), peripheral neuropathy (2/30), encephalitis (1/14), Guillain-Barre syndrome (1/25), or AIDS (2/20). In contrast, we observed high concentrations in 29 of 75 patients with tumors of the central nervous system, especially in meningioma (6/9), glioblastoma (9/23), and neurinoma (5/5). Values for S100 were mainly abnormally high (greater than 0.3 microgram/L) in serum from patients with cerebrovascular disorders (43/48) or with metastases of melanoma (9/11).

Abstract: This inexpensive method for fully automated amino acid analysis combines the advantages of automated precolumn derivatization with o-phthaldialdehyde and favorable analytical conditions to separate and quantify 30 amino acids found in normal plasma. The system can run unattended for almost four days, during which the data are processed automatically by a personal computer and a maximum of 76 samples and 19 standards can be processed (cycle time per analysis: 55 min). Only 1 microL of deproteinized plasma is required per analysis. Coefficients of variation for retention times and areas measured for all relevant amino acids are less than 1% and 3%, respectively. The system described is well suited for quick, sensitive operation in daily practice.

Abstract: This review summarizes physical and chemical properties of major subspecies of very-low-, low-, intermediate, and high-density lipoproteins. Hypotheses regarding the metabolic origins of these subspecies and evidence for their associations with risk of coronary artery disease are presented.

Abstract: Valproic acid (VPA) is widely used as an anticonvulsant, but therapy with the drug has been associated with hepatotoxicity, either reversible hepatic dysfunction or irreversible hepatic failure. Both clinical and experimental studies have revealed several VPA-related biochemical abnormalities in the liver: inhibition of the beta-oxidation and synthesis of fatty acids and inhibition of gluconeogenesis, urea synthesis, oxidative phosphorylation, and the glycine cleavage system. Other abnormalities noted include alteration in the protein conformation of the internal mitochondrial membrane, hyperammonemia, and increased bile flow. The mechanisms of such hepatotoxicity, whether mediated by VPA or by its metabolites, are still little understood. Susceptibility to VPA hepatotoxicity may be enhanced by such conditions as starvation, inborn errors of metabolism, additional neurological disease, and concomitant administration of enzyme-inducing drugs.

Abstract: We describe two independent HPLC procedures for the rapid, accurate analysis for ascorbic acid in human plasma. No sample extraction or phase separation is required. We also describe a procedure for preparing a human plasma reference material for use in clinical laboratory analysis for ascorbic acid. The ascorbic acid in plasma can be determined in 15 min, with as little as 50 microL of sample. Analytical recoveries are near 100% with direct injection of deproteinized plasma. Extensive stability data under several conditions, with dithiothreitol as a preservative (antioxidant), indicate that ascorbic acid remains stable in stored plasma for as long as 57 weeks. CVs for round-robin analysis of 11 normal human blood samples by two independent methods were between 0.1% and 5.3%. These clinical samples appear to be stable for at least 50 days under the described conditions of stabilization and sample treatment. Finally, because ascorbic acid prepared by the described procedures is stable at room temperature for at least 18 h, these methods can be readily adapted to clinical laboratory automation at room temperature.

Abstract: The traditional dose-response method of medication adjustment depends on several assumptions that are not met in the case of tricyclic antidepressants (TCAs), which makes therapeutic drug monitoring (TDM) particularly useful with these drugs. TDM can facilitate treatment by providing objective guidelines for dose adjustment. It provides a means of assessing compliance, ensuring an effective concentration, and avoiding toxicity. The latter is an often-overlooked benefit of therapeutic monitoring of TCAs and yet is just as important as improving response. The cardiac and central nervous system toxicity of TCAs is concentration-dependent and potentially life-threatening. Such toxicity will predictably occur in up to 5% of patients on standard antidepressant doses of TCAs when TDM is not used to rationally adjust the dose. Without TDM, such toxicity is difficult to detect early. A cost/benefit analysis supports the cost effectiveness of TDM as a standard part of TCA chemotherapy when doses in the 100-300 ng/day range are used.

Abstract: A procedure has been developed for measuring fluoxetine and its desmethyl metabolite, norfluoxetine, in serum by reversed-phase high-performance liquid chromatography (HPLC), with ultraviolet detection at 226 nm Fluoxetine and norfluoxetine are isolated from serum by liquid-liquid extraction They are then separated by HPLC and quantified, with reduced haloperidol as the internal standard Fluoxetine, norfluoxetine, and the reduced haloperidol are separated from all interfering peaks in about 15 min The standard curve is linear (r = 1000) for both fluoxetine and norfluoxetine concentrations over the range of 25 to 800 micrograms/L Between-run CVs for 60 and 200 micrograms/L controls (n = 8) were 68 and 41% for fluoxetine, and 88 and 62% for norfluoxetine, respectively In a study of 24 patients with depression who were being treated with 20-60 mg of fluoxetine per day, fluoxetine and norfluoxetine concentrations in serum, measured during the last three weeks of treatment, were 47-469 micrograms/L and 52-446 micrograms/L, respectively

Abstract: A modified agarose electrophoretic system for the separation of alkaline phosphatase (ALP, EC 3.1.3.1) isoenzymes is described. Bone, liver, high-molecular-mass, and intestinal ALP are separated with high reproducibility. The sensitivity of the agarose system is superior to cellulose acetate in detecting high-Mr ALP. Correlation is good between bone ALP fractions scanned before and after treatment with neuraminidase. Immunoglobulin-bound ALPs, the ALP-lipoprotein-X complex, and the additional ALP fraction observed in transient hyperphosphatasemia in children are detected by their peculiar electrophoretic mobility in the proposed system. Approximately 25% of the samples contained an additional fraction of intestinal-type ALP, as evidenced by neuraminidase treatment and use of polyclonal and monoclonal antibodies. Because the electrophoretic mobilities of this "intestinal variant" and of some immunoglobulin-bound ALP fractions are identical to those of bone and intestinal ALP, respectively, treatment of the samples with a polyclonal antibody that reacts with intestinal ALP is advised.

Abstract: A new fluorometric assay for determining dipeptidyl peptidase IV (DPP IV; EC 3.4.14.5) was developed. The synthetic substrate glycyl-L-proline-4-methoxy-2-naphthylamide (20 mmol/L), Tris buffer (50 mmol/L, pH 8.3), and serum (20 microL) are mixed and incubated. The reaction is stopped with citrate (100 mmol/L, pH 4.0) and the released 4-methoxy-2-naphthylamine is measured fluorometrically. The mean value of DPP IV activity in serum for 64 healthy subjects was 58 (SD 16) mumol of 4-methoxy-2-naphthylamine released per liter of serum per minute. The proposed procedure is sensitive, rapid, and accurate and can easily be automated.

Abstract: We compared the fructosamine activity in sera from healthy and diabetic subjects with the degree of protein glycation detected by a liquid-chromatographic method. The latter technique measures furosine as a specific product after hydrolysis of epsilon-amino-fructose-lysine. Our results indicate that the fructosamine assay measures the extent of glycation of purified human serum albumin correctly. On the other hand, we found no correlation between the two methods for sera from healthy subjects, although for diabetics' sera the values obtained with both methods were related. However, only about half of the reducing activity (fructosamine) was due to specific nonenzymatic glycation of proteins in healthy subjects and well-controlled diabetics. The remaining unspecific activity varied from serum to serum. It was not reducible with NaBH4 and was independent of the glycation of albumin, which normally accounts for about 80% of glycated serum proteins. The fructosamine assay is therefore of limited specificity for the exact measurement of glycated proteins in serum.

Abstract: We measured the excretion rates of six urinary enzymes that either originate from the proximal renal tubule, like alanine aminopeptidase (EC 3.4.11.2), alkaline phosphatase (EC 3.1.3.1), gamma-glutamyltransferase (EC 2.3.2.2), and N-acetyl-beta-D-glucosaminidase (EC 3.2.1.30), or that are typical low-molecular-mass proteins, like lysozyme (EC 3.2.1.17) and pancreatic ribonuclease (EC 3.1.27.5). These rates were compared with those of total protein and albumin in urine of 36 insulin-dependent diabetic men and 30 healthy men. Seventeen of the diabetics had "clinical proteinuria," defined as excretion of more than 7.5 g of protein per mole of urinary creatinine (group B). Group A comprised the 19 diabetics without proteinuria. Except for gamma-glutamyltransferase, the excretions of enzymes and proteins were significantly higher in diabetics than in controls and were greater in group B than in group A. N-Acetyl-beta-D-glucosaminidase was the analyte most often increased in group A (89%), followed by albumin and alkaline phosphatase (each 32%). All patients in group B showed increased excretion of N-acetyl-beta-D-glucosaminidase. We conclude from the comparative data that this enzyme may be useful as an early predictor of diabetic nephropathy.

Abstract: We measured the concentrations of 29 commonly measured analytes in fresh sera and in sera that had been stored as whole blood at seven different temperatures for 24 h. We determined the effect of storage temperature and prolonged contact with cell clot on the measured concentration of each analyte, with fresh serum as the control. Significant differences were observed for concentrations of creatinine, glucose, inorganic phosphorus, potassium, and both aminotransferases. The extent of these differences was temperature dependent. Values for the remaining 23 analytes examined were essentially unaffected by the storage.