Effect of moderate improvement in metabolic control on magnesium and lipid concentrations in patients with type 1 diabetes.

OBJECTIVE: To evaluate the effect of clinically obtainable improvements in metabolic control in patients with type 1 diabetes on biochemical cardiovascular risk factors. RESEARCH DESIGN AND METHODS: Blood and 24-h urinary samples were obtained from 49 patients with type 1 diabetes before and after a run-in period and after 3 months of intervention, with frequent adjustment of insulin dosage according to measured blood glucose concentrations. RESULTS: The intervention caused a mean insulin dosage increment of 10%, a 20% decrease in fasting plasma glucose concentration, a 10% decrease in albumin corrected serum fructosamine, and a somewhat lesser decrease in HbAlc.A 14% decrease in the renal excretion of magnesium (Mg) was observed, but without a change in average serum Mg concentration. Serum HDL cholesterol increased 4%, and serum triglycerides decreased 10% as an average. Looking at individual patients, the decrease in serum triglycerides correlated with both the change in serum total Mg concentration and with the increase in insulin dosage. Using the change in serum total Mg concentration and in insulin dosage as independent variables in a multiple regression analysis, the coefficient of correlation with the decrease in serum triglycerides was 0.52. CONCLUSIONS: Moderate but clinically obtainable improvement of metabolic control in patients with type 1 diabetes seems to reduce the loss of Mg, increase serum HDL cholesterol, and decrease serum triglycerides. The decrease in serum triglycerides was associated with the change in serum total Mg concentration. These reductions in Mg loss and serum triglycerides might reduce the risk of developing cardiovascular disease in patients with type 1 diabetes.

Muscle sodium, potassium, and [(3)H]ouabain binding in identical twins, discordant for type 2 diabetes.

A reduced functional capacity of the sodium (Na), potassium (K) pump might reduce energy expenditure, inducing obesity and type 2 diabetes. Consequently, the Na and K content and [(3)H]ouabain binding capacity of skeletal muscle were measured in 10 monozygotic twin pairs discordant for type 2 diabetes and in 10 obese controls. Muscle [(3)H]ouabain binding capacity was reduced by approximately 20% in type 2 diabetes. Removing the genetic component by looking at differences within twin pairs, the difference in waist/hip ratio was associated with the difference in [(3)H]ouabain binding (r = -0.85; P < 0.002). Except for the type 2 diabetic twins in the basal state, both basal and insulin-stimulated energy expenditure were associated with the muscle K/Na ratio in the twins. In controls, the 2-h plasma glucose concentration during an oral glucose tolerance test was associated with the change in both muscle and plasma K induced by a euglycemic, hyperinsulinemic clamp. In conclusion, environmental factors related to the waist/hip ratio reduce the muscle [(3)H]ouabain binding capacity in type 2 diabetes. Without proving causality, the muscle K/Na ratio is associated with energy expenditure in individuals genetically predisposed to the development of type 2 diabetes.

Potassium and magnesium distribution, ECG changes, and ventricular ectopic beats during beta 2-adrenergic stimulation with terbutaline in healthy subjects.

OBJECTIVE: To study the effect of intravenous (i.v.) terbutaline on potassium (K) and magnesium (Mg) distribution, ECG changes, and prevalence of ventricular ectopic beats in healthy subjects. DESIGN: Randomized double-blind, placebo-controlled crossover. Subjects received either placebo or terbutaline (bolus, 0.25 mg; maintenance dose, 5 micrograms/min). SETTING: University Department of Clinical Chemistry. PARTICIPANTS: Ten healthy male volunteers. Mean age was 24.1 (range, 20 to 31) years. MAIN OUTCOME MEASURES: Serum potassium and magnesium muscle potassium and magnesium, and muscle sodium-potassium pump number. Urinary excretion of potassium and magnesium. ECG changes (T-wave and QTC interval) and the number of ventricular ectopic beats. MAIN RESULTS: Terbutaline produced an immediate decrease in serum potassium level from 4.17 (4.04 to 4.30) mmol/L to a nadir of 3.32 (3.06 to 3.58) mmol/L (p < 0.001). The urinary excretion of potassium decreased from 0.077 mmol/min (0.052 to 0.102) to 0.038 mmol/min (0.025 to 0.051) (p < 0.01). There was an increase in the number of sodium potassium pumps from 1,104.1 nmol/kg dry weight (1,030.6 to 1,177.5) to 1,273.3 nmol/kg dry weight (1,193.5 to 1,353.2) (p < 0.01), but no measurable change in muscle potassium. The QTC interval increased from 395 (385 to 405)ms to 449 (432 to 466) ms (p < 0.003). There was no change in the number of ventricular ectopic beats. CONCLUSIONS: Short-term i.v. administration of terbutaline produced hypokalemia partly due to an increase in the number of sodium-potassium pumps. Furthermore, terbutaline induced changes in ECG with a highly significant lengthening of the QTc interval but with an unchanged number of ventricular ectopic beats in healthy subjects.

Magnesium, sodium, and potassium content and [3H]ouabain binding capacity of skeletal muscle in relatives of patients with type 2 diabetes: effect of dexamethasone.

Theoretically, disturbancies in sodium (Na) and potassium (K) homeostasis and a magnesium (Mg) deficit could be possible factors in the development of obesity, type 2 diabetes, and hypertension. Therefore, we measured electrolyte content and [3H]ouabain binding capacity of skeletal muscle in 20 relatives of type 2 diabetic patients and in 20 controls before and after glucose infusion and before and after treatment with dexamethasone, which decreases insulin sensitivity. Muscle electrolyte content and [3H]ouabain binding capacity did not differ between groups. Infusion of glucose increased muscle Na content 25%, decreased muscle potassium content 9%, and muscle Mg content 5%. Muscle potassium/Mg ratio decreased only in relatives. Treatment with dexamethasone increased muscle Na content 15% and decreased muscle Mg content 7%, whereas muscle potassium/Na ratio decreased 17%. Dexamethasone increased muscle [3H]ouabain binding capacity by 42% in both groups. Basal and 1-hour intramuscular glucose content correlated inversely with basal muscle potassium/Na ratio in relatives only. In conclusion, persons who were predisposed to the development of type 2 diabetes exhibited an increased interdependency between glucose, Na, and potassium handling in skeletal muscle. Muscle Na content and [3H]ouabain binding capacity increased during treatment with dexamethasone, and muscle potassium/Na ratio decreased. Intravenous (IV) glucose injection decreases muscle Mg content, as does a decrease in insulin sensitivity, without any differences between relatives and controls.

Magnesium reduces insulin-stimulated glucose uptake and serum lipid concentrations in type 1 diabetes.

A magnesium (Mg) deficit has been described in patients with type 1 diabetes, and it has been related to the development of cardiovascular disease. We tested the hypothesis that type 1 diabetic patients have deficits in dietary Mg intake and that proper long-term (24 weeks) oral Mg supplementation would reduce cardiovascular risk factors. Therefore, the Mg status, dietary Mg intake, and the effect of Mg supplementation were evaluated in 10 type 1 diabetic patients and 5 control subjects. Muscle Mg content was decreased by 7% in the type 1 diabetic patients, and it increased by 5% after 24 weeks of oral MgO supplementation. Acute and chronic Mg supplementation decreased serum total cholesterol, serum low-density lipoprotein (LDL)-cholesterol, and apolipoprotein B. Insulin-stimulated glucose uptake decreased by 35% after 24 weeks of oral MgO supplementation. Eight of 10 patients with type 1 diabetes had a daily intake of Mg below 90% of the recommended daily allowance. In conclusion, a Mg deficit was found in type 1 diabetic patients. The deficit might be due partly to a relatively Mg-deficient diet. Mg repletion was associated with a decrease in atherogenic lipid fractions and a reduced insulin-stimulated glucose uptake.

Effects of detraining on endurance capacity and metabolic changes during prolonged exhaustive exercise.

The effects of 4 wk of detraining on maximal O2 uptake (VO2max) and on endurance capacity defined as the maximal time to exhaustion at 75% of VO2max were studied in nine well-trained endurance athletes. Detraining consisted of one short 35-min high-intensity bout per week as opposed to the normal 6-10 h/wk. Detraining had no effect on VO2max (4.57 +/- 0.10 vs. 4.54 +/- 0.08 l/min), but endurance capacity decreased by 21% from 79 +/- 4 to 62 +/- 4 min (P < 0.001). Endurance exercise respiratory exchange ratio was higher in the detrained than in the trained state (0.91 +/- 0.01 vs. 0.89 +/- 0.01; P < 0.01). Muscle [K+] values were unchanged during exercise and were similar in the trained and detrained states. Muscle [Mg2+] values were similar at rest and at minute 40 (30.3 +/- 0.9 vs. 30.8 +/- 0.6 mmol/kg dry wt) but increased significantly at exhaustion to 33.8 +/- 1.0 mmol/kg dry wt in the trained state and to 33.9 +/- 0.9 mmol/kg dry wt in the detrained state. The elevated muscle [Mg2+] at exhaustion could contribute to fatigue in prolonged exercise through an inhibition of Ca2+ release from sarcoplasmic reticulum. It is concluded that the endurance capacity can vary considerably during detraining without changes in VO2max. Altered substrate utilization or changes in electrolyte regulation may account for the reduced endurance capacity.

Skeletal muscle magnesium content in identical twins, discordant for type 2 diabetes.

BACKGROUND: Low magnesium (Mg) status has been implied as a factor in the development of type 2 diabetes mellitus (DM) and its complications. We therefore studied Mg-status in identical twins, discordant for type 2 DM and in matched controls. Through correlation analysis, possible associations between Mg-status and glucose uptake were evaluated. METHODS: Plasma Mg concentration was measured in 12 monozygote twin pairs, discordant for type 2 DM and in 12 matched controls. Muscle Mg content was measured in 10 persons from each group. An oral glucose tolerance test and a euglycaemic, hyperinsulinaemic clamp were utilized. RESULTS: Neither muscle Mg content nor plasma Mg concentration differed among groups. Plasma Mg concentration decreased during the euglycaemic, hyperinsulinaemic clamp. In the control group, muscle Mg content correlated positively with insulin stimulated glucose disposal rate (r=0.77, p<0.01) and negatively with two hour plasma glucose concentration during an oral glucose tolerance test (OGTT) (r=- 0.64, p<0.05). In the control group, the two hour plasma glucose concentration during an oral glucose tolerance test correlated with the decrease in plasma Mg concentration (r=- 0.80, p<0.002) and with the change in muscle Mg content (r=0.90, p<0.0005) induced by the clamp. None of these associations were found in the two twin groups. CONCLUSIONS: Normal plasma Mg concentration and muscle Mg content were found in persons with type 2 DM and in persons, who were heavily predisposed to the development of type 2 DM, indicating a normal whole-body Mg content. However, the missing associations between measures of glucose disposal and changes in both plasma Mg concentration and muscle Mg content in the two twin groups indicates, that physiological mechanisms, which partly regulates insulin sensitivity and Mg status in healthy individuals are either exhausted or fully utilized in both type 2 DM and in genetically identical twins without DM.

Improved running economy following intensified training correlates with reduced ventilatory demands.

PURPOSE: To compare the effects of three types of intensive run training on running economy (RE) during exhaustive running and to establish possible relationships with changes in ventilatory function and/or muscle fiber type distribution. METHODS: Thirty-six male recreational runners were divided into three groups and assigned to either exhaustive distance training (DT), long-interval training (LIT), or short-interval training (SIT) three times 20-30 minxwk(-1) for 6 wk. VO(2 max) and RE were measured during treadmill running before and after training. Muscle fiber type distribution of the vastus lateralis muscle was established from biopsy material. RESULTS: VO(2max) (Lxmin(-1) increased by 5.9% (P < 0.0001), 6.0% (P < 0.0001), and 3.6% (P < 0.01) in DT, LIT, and SIT, respectively, and running speed at VO(2max) by 9% (P < 0.0001), 10% (P < 0.0001), and 4% (P < 0.05), respectively. Time-to-exhaustion at 87% of pretraining VO(2max) (mean 3.83) mxs(-1) increased by 94% in DT (P < 0.0001), 67% in LIT (P < 0.0001). Running economy improved by 3.1% in DT (P < 0.05), 3.0% in LIT (P < 0.01), and 0.9% SIT (NS): pulmonary ventilation (VE) was on average 11 Lxmin(-1) lower following training (P < 0.0001). The individual decrements in VE correlated with improvements in RE (r = 0.77; P < 0.0001) and may account for 25-70% of the decrease in aerobic demand. Muscle fiber composition, and respiratory exchange ratio, stride length, and stride frequency during running were unaltered with training. CONCLUSIONS: Recreational runners can improve RE and aerobic run performance by exchanging parts of their conventional aerobic distance training with intensive distance or long-interval running, whereas short-interval running is less efficient. The improvement in RE may relate to reduced ventilatory demands. Muscle fiber type distribution was unaltered with training and showed no associations with RE.

The basis for common reference intervals for serum potassium.

In order to investigate the relevance of the currently used lower reference limit for S-Potassium in Danish hospital laboratories, analytical bias in the measurement of S-Potassium was compared with the lower reference limit in each of 52 Danish hospital laboratories. The acceptable bias range was estimated according to Gowans et al on the basis of the result of two different reference sample groups. The estimated acceptable 0.95 bias range was 0.24 mmol/L, so the observed bias range of 0.23 mmol/L was within this limit. As all preanalytical errors tend to increase the measured S-Potassium, all acceptable bias should be in the direction of decreasing the measured value. It can be concluded that analytical performance allows for more uniform (even common) reference interval(s) in all Danish and perhaps Nordic hospital laboratories, provided that preanalytical errors can be controlled.

New data on the mechanisms of hypermagnesuria in type I diabetes mellitus.

It has been known for long that renal Mg excretion is increased in patients with type I diabetes mellitus, and that these patients have a Mg deficit. It can be hypothesized, that this deficit might be related to the development of late complications in the diabetic. In recent years it has been shown that the increased renal Mg excretion in patients with type I diabetes is due primarily to an elevated plasma glucose concentration. An increase in plasma glucose concentration from 5 to 12 mmol/l more than doubles renal Mg excretion, if everything else is kept constant. Hyperinsulinism may also contribute to the increased renal Mg excretion. However, since improved metabolic control in patients with type I diabetes reduces the renal Mg loss despite an increase in insulin dosage, hyperinsulinism is probably of minor importance in the aetiology of hypermagnesuria in patients with type I diabetes mellitus, compared with the effect of hyperglyeaemia.

[Heart rate disorders in potassium and magnesium deficiency]. / Hjerterytmeforstyrrelser ved kalium- og magnesiummangel.

Potassium and magnesium deficiencies are common in patients with heart disease. These are often coexistent and pathophysiologically related. Potassium deficiency cannot be treated without correction of concomitant magnesium deficiency. Correlations between serum levels and body stores are very poor for both ions. Therefore diagnosis and treatment of these conditions based on serum levels alone are erroneous. There is some evidence that it is primarily the intracellular depletion of these ions which is arrhythmogenic. Magnesium infusion has been proved effective in treatment of torsade de pointes ventricular tachycardia and arrhythmias induced by digoxin-intoxication, and is recommended in these conditions. Whether it is effective in other forms of arrhythmia is not yet elucidated.

[The effect of beta blockers combined with thiazides on serum electrolytes]. / Virkningen af beta-blokkere kombineret med tiazider på serumelektrolytter.

The case reports of a total of 9,670 patients admitted to a medical department were reviewed manually. From this material, all patients receiving treatment with beta-blockers alone, bendroflumethiazide alone or combinations of these were selected. In addition, a control group was selected. The patients should have received treatment for at least seven days and they should not be "ill". Patients receiving treatment with beta-blockers had higher S-K+ (4.2 mmol/l) and lower standard bicarbonate levels (23 mmol/l) than matched controls. Patients receiving treatment with bendroflumethiazide had low S-K+ (3.6 mmol/l) and high standard bicarbonate (26 mmol/l). Patients receiving treatment with both bendroflumethiazide and beta-blocker had low S-K+ (3.4 mmol/l) and the same standard bicarbonate as patients who received bendroflumethiazide alone. A more limited group chosen according to stricter criteria confirmed the above mentioned results. The investigation demonstrated that treatment with both beta-blocker and bendroflumethiazide caused deterioration of the hypokalaemia induced by bendroflumethiazide despite the higher S-K+ during beta-blockade. This may be due to renal excretion of potassium during beta-blockade.

[Magnesium deficiency and development of late diabetic complications]. / Magnesiummangel og udviklingen af diabetiske senkomplikationer.

Diabetics have a higher prevalence of recurrent angina, myocardial infarction and hypertension compared with healthy individuals. This is partly due to the metabolic control, and partly to the development of nephropathy, but inherited factors also seem to contribute to their development. Another factor, to some extent inherited, could be magnesium-deficiency. Up to 30% of all diabetics suffer from magnesiumdepletion, which in other disease entities has been related to angina pectoris, myocardial infarction, and changes in blood lipids. Furthermore, magnesium-deficiency might be related to the development of hypertension. In diabetes, magnesium-depletion has been linked with the development of retinopathy, and an increase in abortions and malformations in diabetic pregnancies. No intervention trials with magnesium, to elucidate the question of magnesium depletion and diabetic late complications have hitherto been made. No methods for assessing magnesium-status in diabetes are available as yet, to select patients for this type of trial.

[Bartter's syndrome. A condition with chronic hypokalemia]. / Bartters syndrom. En tilstand med kronisk hypokalioemi.

Bartter's syndrome (BS) is a disease with severe hypokalaemia due to renal potassium wasting. The potassium loss is due to lesions at different sites within the renale tubule. Additional features include metabolic alkalosis, excess renal production of prostaglandins, hyperreninaemia, hyperaldosteronism and impaired pressor responses to exogenous angiotensin II. These secondary features are the result of renal potassium wasting. Symptoms are due to potassium deficiency, but many adult patients feel well despite marked hypokalaemia. The hypocalciuric variant of BS is called Gitelman's syndrome. These patients have a more benign course. The diagnosis of BS is one of exclusion, mainly of surreptitious vomiting, diuretic or laxative abuse. The primary treatment is potassium supplementation often in combination with potassium-sparing diuretics, prostaglandin inhibitors or ACE-inhibitors. With coexisting magnesium deficiency, magnesium supplementation might be effective.

[Percutaneous transluminal coronary angioplasty in acute myocardial infarction]. / Perkutan transluminar koronar angioplastik ved akut myokardieinfarkt.

The attitude towards the treatment of acute myocardial infarction is currently changing. As an example of one of the new methods, a successful case of acute percutaneous transluminal coronary angioplasty is presented. Which of the new therapeutic regimens provides the best prognosis is unknown, but considering the resources and the hitherto published studies, intravenous thrombolysis seems in general to be the most advantageous, while the others ought only to be used on special indications.

[Injection site for quick-acting insulin. Significance for glycemic control in basal bolus insulin regimen]. / Injektionssted for hurtigtvirkende insulin. Betydning for glykoemisk kontrol ved basal-bolus insulinregimet.

The impact on glycaemic control of soluble insulin injected either intramuscularly into the thigh (IMT), subcutaneously into the abdominal wall (SCA) or subcutaneously into the thigh (SCT) was evaluated in 49 Type 1 diabetic outpatients following a randomised three-month intervention study. Insulin doses were adjusted based on patients' self-monitored blood glucose values and reported hypoglycaemic episodes. More patients in the SCA and IMT groups than in the SCT group had serum fructosamine values within normal limits following intervention. Blood glucose at 03.00 was lower in the SCT group than in the SCA and IMT groups, due to a higher number of low nocturnal blood glucose values (less than 4 mmol/l) in the SCT group. In conclusion, s.c. injection of soluble insulin into the abdominal wall or intramuscularly into the thigh is preferable compared to s.c. injection into the thigh in the basal bolus insulin delivery regimen. Soluble insulin injection s.c. into the thigh during daytime is a risk factor for nocturnal hypoglycaemia.

Antioxidant status and lipid peroxidation after short-term maximal exercise in trained and untrained humans.

The purpose of this study was to measure resting muscle and blood antioxidant status in untrained (n = 8) and jump-trained (n = 8) humans and to evaluate free radical-mediated muscle damage after a strenuous jump test consisting of six bouts of 30-s continuous jumping separated by 2 min of rest. Resting muscle antioxidant activities [superoxide dismutase (SOD), glutathione peroxidase (GPX), glutathione reductase (GR), and manganese SOD] were significantly higher in jump-trained compared with untrained subjects. Blood antioxidant enzyme activities and muscle catalase, however, were not different between the two groups. Creatine kinase activities increased significantly (P < 0.0001) after the jump test in untrained individuals, but remained unchanged in the jump trained. Plasma and muscle malonaldehyde (MDA) after the jump test were not significantly different from rest. These data suggest that jump training is associated with elevated activities of SOD and the coupled enzymes GPX and GR in muscle tissue, but other antioxidants remain unchanged. High-intensity jump exercise induces muscle enzyme leakage in untrained humans, but muscle lipid peroxidation, measured as changes in MDA, was not different in the two groups despite the varied muscle antioxidant enzyme levels.

Insulin increases renal magnesium excretion: a possible cause of magnesium depletion in hyperinsulinaemic states.

The effects of insulin upon renal magnesium excretion were examined. Urinary magnesium excretion rates were measured in seven healthy volunteers (three men, four women) before and during a euglycaemic, hyperinsulinaemic clamp. Insulin was infused at 120 pmol m-2 min-1 and at 240 pmol m-2 min-1. Compared to baseline, the renal magnesium excretion increased 30% during the infusion of insulin at a rate of 120 pmol m-2 min-1. During infusion of insulin, 240 pmol m-2 min-1, renal magnesium excretion increased 50% compared to baseline. There were no changes in either glomerular filtration rates, plasma magnesium, urinary volume or general changes in the renal handling of divalent ions as judged by an unchanged urinary excretion rate of calcium (0% during infusion of insulin, 120 pmol m-2 min-1, and 8% increase during infusion of 240 pmol m-2 min-1 (NS). During the 120 pmol m-2 min-1 insulin infusion rate, plasma insulin rose from 46.1 pmol I-1 to 158.8 pmol I-1 and during the 240 pmol m-2 min-1 insulin infusion rate, mean plasma insulin concentration was 361.4 pmol I-1. Thus, physiological concentrations of insulin induce a specific increase in the renal excretion of magnesium. This might partly explain the magnesium depletion observed in various hyperinsulinaemic states, diabetes mellitus, atherosclerosis, hypertension, and obesity.

Effect of glucose/insulin infusion and magnesium supplementation on serum and muscle sodium and potassium and muscle [3H]ouabain binding capacity in Type 1 diabetes mellitus.

Glucose has recently been found to decrease muscle potassium content. The aim of this study was to evaluate the effect of the infusion of glucose and insulin and the effect of magnesium supplementation on serum and muscle sodium and potassium and muscle [3H]ouabain binding capacity in patients with Type 1 diabetes mellitus and in controls. Muscle potassium and sodium content, muscle [3H]ouabain binding capacity and serum potassium and sodium concentrations were determined in 10 patients with Type 1 diabetes mellitus and in 5 controls before and after an euglycaemic, hyperinsulinaemic clamp, and after an intravenous magnesium load test. Nine of the patients with Type 1 diabetes mellitus were restudied after 24 weeks of oral magnesium oxide supplementation. Basic serum and muscle sodium and potassium and muscle [3H]ouabain binding capacity did not differ between groups. The infusion of glucose and insulin reduced muscle potassium content, whereas muscle sodium content was unchanged. There were no differences between groups. Oral magnesium oxide supplementation increased muscle potassium content by 6%. Muscle [3H]ouabain binding capacity was unchanged. In patients with Type 1 diabetes mellitus, the intravenous infusion of magnesium increased serum potassium concentration before but not after oral magnesium oxide supplementation. In controls, the infusion of magnesium did not affect serum potassium concentration. It was found that intravenous infusion of glucose and insulin decreases muscle potassium content, probably by shifting potassium from the muscle cells to the splanchnic organs. Oral magnesium oxide supplementation increases muscle potassium content in patients with Type 1 diabetes mellitus. The increase in serum potassium concentration owing to the intravenous infusion of magnesium could be used in the evaluation of magnesium status in patients with Type 1 diabetes mellitus. This, however, requires further investigation.

Reference intervals based on hospitalized 'healthy' patients and medical students in relation to analytical bias for serum potassium.

The reliability of reference intervals for measurements of serum (S)-potassium in Danish hospital laboratories was investigated (i) by estimation of reference interval based on two different, healthy subpopulations and (ii) by comparison of reference intervals for S-potassium with analytical bias in each of 52 Danish laboratories. (i) The reference values from 227 hospitalized 'healthy' patients were obtained during the period 1979 to 1987 from the first-drawn serum specimen from the hospitalized patients, who were later discharged from the hospital without a diagnosis. The estimated 0.95 reference interval was 3.34 to 4.52 mmol l-1. The other reference sample group consisted of 314 medical students from whom blood was collected in the period from 1983 to 1987. Here the estimated reference interval was from 3.44 to 4.53 mmol l-1. The concentration values from both reference sample groups were corrected for analytical bias (+0.05 mmol l-1). (ii) The 52 Danish laboratories revealed a considerable variability in reference intervals which, regarding the lower reference limit, ranged from 3.2 to 3.7 mmol l-1 in strong contrast to the analytical bias (ranging from -0.08 to +0.15 mmol l-1) in 50 laboratories (two outliers). There was no relationship between lower reference limit and analytical bias in the individual laboratories. It is concluded that analytical performance allows for more uniform (even common) reference intervals throughout the Danish and perhaps Nordic hospital laboratories.