Effect of intravenous glucagon on urinary catecholamine excretion in normal man.

We have evaluated intravenous glucagon as pharmacologic stimulus to adrenal catecholamine secretion in normal human subjects. Urinary epinephrine and norepinephrine were measured at two hourly intervals before and after the intravenous injection of saline, 4 mg glucagon, or 0.1 mu/kg crystalline insulin. Urinary epinephrine was increased 2.3-fold over baseline after both saline and glucagon. By contrast, insulin hypoglycemia produced a 24-fold rise in urinary epinephrine. No rise in urinary norepinephrine was detected in any test. Under the conditions of this study, we conclude that epinephrine excreted after glucagon injection is due to the stress of the test itself. In normal man, glucagon either does not stimulate adrenal catecholamine secretion, or the effect is too small to measure. By contrast, in pheochromocytoma, glucagon may be specific for catecholamine secretion, based on data from the literature. In normal subjects, insulin hypoglycemia remains the only proved method for assessing adrenal catecholamine reserve.     

Quantitation of apolipoprotein A-I of human plasma high density lipoprotein.

High density lipoproteins (HDL) may be controlled via their major apolipoprotein, A-I. To study this apolipoprotein, a simple, precise, and accurate immunodiffusion assay for A-I was developed and applied in a sample of Bell Telephone Company employees. A-I showed a slight increase with age in men (r=0.11, n=263) and women (r=0.15, n=257). A-I correlated closely with HDL cholesterol (r=0.72). It was weakly related to total triglyceride in women (r=0.24) but was inversely related in men (r=-0.17). Women on estrogen had the highest A-I levels (149 mg/dl +/- 26, x +/- S.D., n=29, p is less than 0.05), followed by women on combination oral contraceptives (141 +/- 26, n=80) whereas women on no medication had lower levels (129 +/- 25, n=99, p is less than 0.01) but men had the lowest levels (120 +/- 20, p is less than 0.01) In a separate group of 14 women given estrogen for 2 wks (1 mug/kg/day), A-I increased by 24%. Thus A-I is increased by exogenous and, most likely, endogenous estrogen, Among hyperlipidemic referral subjects, those with hypercholesterolemia (n=43) and hypertriglyceridemic women (n=33) had normal A-I levels. Among hypertriglyceridemic men both A-I and HDL cholesterol values were decreased (115 +/- 20, p is less than 0.01 and 37 +/- 3, p is less than 0.01, respectively, n=68) but were significantly lower among a group of myocardial infarction survivors (107 +/- 16, p is less than 0.01, and 27 +/- 6, p is less than 0.01, respectively, n=24). High density lipoprotein levels and the content of cholesterol in HDL associated with A-I appear to be decreased in coronary heart disease.     

Lp(a) lipoprotein: relationship to sinking pre-beta lipoprotein hyperlipoproteinemia, and apolipoprotein B.

To assess the relationship between the Lp(a) and the "sinking pre-beta" (d smaller than 1.006) lipoprotein, the concentration of Lp(a) was quantified by radial immunodiffusion and the presence or absence of sinking pre-beta was assessed by agarose electrophoresis in overnight fasting plasma samples from 485 adults, comprised of 320 with normal lipid levels, 48 with type IIa, 40 with type IIb, and 77 with type IV lipoprotein phenotypes. The median Lp(a) level was 7.6 mg/100 ml, 89% (433 of 485) having detectable Lp(a) levels. Twenty-two per cent (107 of 485) had detectable pre-beta lipoprotein in the d greater than 1.006 plasma fraction (sinking pre-beta). Of the sinking pre-beta positive plasma samples, 96% (102 and 107) exceeded the median Lp(a) level, and sinking pre-beta was detected in all 44 samples with an Lp(a) concentration exceeding 40 mg/100 ml. The relationship of Lp(a) and sinking pre-beta to lipoprotein phenotype was assessed. Compared to the normolipidemic group, the type IIa group had higher Lp(a) percentile values (p smaller than 0.02), whereas the IIb and type IV groups had significantly lower Lp(a) values than the normolipidemic group. Ninety-two per cent (296 of 320) of the normolipidemic subjects had detectable levels of Lp(a) and 22% (70 of 320) had detectable sinking pre-beta lipoprotein. Ninety-four per cent (45 of 48) of the type IIa plasmas had detectable Lp(a) levels and 27% (13 of 48) had sinking pre-beta lipoproteins. Contrasted with the IIa group, only 80% (32 of 40) of the IIb plasmas had detectable Lp(a) levels and 18% (7 of 40) had sinking pre-beta lipoprotein. In the type IV plasmas 78% (60 of 77) had detectable Lp(a) and 22% (17 of 77) had sinking pre-beta lipoprotein. Lp(a) or log Lp(a) levels were not correlated with apolipoprotein B levels (n = 485, r = 0.002 or 0.037, respectively). Furthermore, Lp(a) levels remained essentially constant in three subjects whose aprptein B levels were altered in response to pharmacological and/or dietary manipulation. A fourth subject had a 50% increase in Lp(a) but this change did not correlate with apoprotein B changes. Thus, these findings suggest that Lp(a) is metabolically independnet of low density lipoprotein even though it shares the same structural protein, apoprotein B.     

Delayed clearance of chylomicron remnants following vitamin-A-containing oral fat loads in broad-beta disease (type III hyperlipoproteinemia).

Chylomicron "remnants" are formed by the selective removal of triglyceride catalyzed by lipoprotein lipase. To investigate a possible defect in the clearance of these remnants in the pathophysiology of broad-beta disease (type III hyperlipoproteinemia), subjects with this disorder and comparison subjects with endogenous hypertriglyceridemia (and type IV lipoprotein patterns) ingested an oral fat load (corn oil: cocoa butter, 1:1, 50 g/sq M) containing retinyl ester, 100 mg, with or without 15 muCi 15-(14) C-retinol (43.7 mCi/mg). The content of triglyceride and vitamin A was sequentially determined in chylomicrons (Sf more than 400) and very low density lipoproteins (VLDS, Sf20-400) over the ensuing 24-72 hr. Vitamin A was chosen as a marker for exogenous sterol assimilation since, like cholesterol, it is absorbed in the small intestine and cosecreted in esterified form with triglyceride in the chylomicron core; however, unlike cholesterol, once having been removed by the liver, it cannot be recycled inot VLDL, but subsequently circulates only as a complex with the high density retinol binding protein. Thus measurements of the vitamin A/triglyceride ratio in Sf greater than 20 lipoproteins reflected the relative efficiency of vitamin A versus triglyceride removal within these lipoproteins. These studies confirmed the intital concentration of exogenous vitamin A in chylomicrons but invariably disclosed an increasing proportion of the remaining Sf greater than 20 vitamin A in VLDL 24 hr after its ingestion. The vitamin A/triglyceride ratio also invariably increased between 6 and 24 hr in the Sf20-30 subfraction, reflecting the formation of vitamin A-rich "remnants" as intermediate species in the catabolism of chylomicrons and VLDL. Among those with mild to moderate endogenous hypertriglyceridemia the Sf greater than 400 vitamin A/triglyceride ratio declined between 6 and 24 hr, reflecting the efficient passage of the vitamin A through this fraction and/or continued secretion of Sf greater than 400 particles rich in triglyceride. Among those with severe endogenous hypertriglyceridemia, both the peak and decline in the Sf greater than 400 vitamin A/triglyceride ratio were delayed. However, among those with broad-beta disease, an increasing vitamin A/triglyceride ratio between 6 and 24 hr was frequent within all VLDL subfractions and invariable among lipoproteins of Sf greater than 400 regardless of the degree of antecedent hypertriglyceridemia. Although additional experiments disclosed a similar delay in both vitamin A and triglyceride assimilation when basal triglyceride levels were high in these subjects, marked reduction of triglyceride levels did not correct the rise in the Sf greater than 400 vitamin A/triglyceride ratio between 6 and 24 hr. Experiments employing preparative electrophoresis confirmed the identity of VLDL containing a high vitamin A/triglyceride ratio with the beta-VLDL which accumulate in broad-beta disease...     

Broad-beta disease versus endogenous hypertriglyceridemia: levels and lipid composition of chylomicrons and very low density lipoproteins during fat-free feeding and alimentary lipemia.

To assess the roles of endogenous and exogenous lipid in the production of the abnormal lipoprotein patterns characteristic of broad-beta desease (with a type IVIII lipoprotein pattern) and endogenous hypertriglyceridemia (with a type IV pattern), oral fat loads (50 g/M-2) were administered to six subjects with broad-beta disease and to eight with endogenous hypertriglyceridemia following at leat 72 hr of 0% fat, 85% carbohydrate isocaloric formula feeding. Total plasma and Sf greater than 400, 100-400, 60-100, 30-60, and 20-30 lipoprotein cholesterol, triglyceride, and phospholipid levels were measured at 0 hr, 6 hr (at or before the peak of alimentary lipemia), and 24 hr following the fat load. Following fat-free feeding the levels and composition of the endogenous Sf greater than 400 lipoproteins were similar in both disorders; whereas total Sf20-400, and most notably, Sf 30-60 and 20-30 levels were increased and enriched in cholesterol in the subjects with broad-beta disease.     

Thyrotropin levels in hyperlipidemic survivors of myocardial infarction.

In an earlier investigations, the prevalence of hyperlipidemia in a group of patients who survived myocardial infarction was determined, and family studies were performed to allow genetic classification of patients with hyperlipidemia. Radioimmunoassay for thyrotropin (TSH) has now been performed on plasmas from most of these hyperlipidemic survivors. Elevated TSH values were found in five of the 18 hyperlipidemic women over age 60, and in seven of the remaining 104 hyperlipidemic subjects. Among the various genetically defined types of hyperlipidemia, the highest prevalence of TSH elevations was seen in women with sporadic (nonfamilial) hypertriglyceridemia; four of the ten had an abnormal TSH level, and two of the remainder were receiving thyroid medication. Hypercholesterolemia was not strongly correlated with TSH abnormalities. These data support the hypothesis that clinically inapparent thyroid damage may be associated with coronary artery disease.     

Postheparin plasma triglyceride lipases in chronic hemodialysis: evidence for a role for hepatic lipase in lipoprotein metabolism.

Elevated plasma triglyceride levels frequently occur in patients with chronic renal failure receiving longterm hemodialysis. Postheparin plasma lipolytic activity, an indirect measure of triglyceride removal, is low in hemodialysis patients, but this activity measures both hepatic triglyceride lipase (HTGL) and lipoprotein lipase (LPL). To determine if HTGL and/or LPL are low in hemodialysis patients and related to lipoprotein lipid levels, both activities were measured by a selective antibody-inhibition technique in postheparin plasma from 20 hemodialysis patients with a wide range of plasma triglyceride levels ($\text{104–676}\text{mg}\text{100}\text{ml}$), and the relationships between the enzyme activities and lipoprotein lipid levels were examined. To more accurately compare subjects, the heparin doses were adjusted for the differences in plasma volumes between the hemodialysis patients and the nonuremic control subjects. Hemodialysis patients with elevated plasma triglyceride levels (↑TG) had HTGL levels (148 ± 67 nmole/min/ml, n=10) which were similar to the dialysis patients with normal triglyceride levels (nlTG) (134 ± 64 nmole/min/ml, n=10) and both groups were significantly lower (p<0.05, p<0.02, respectively) than the levels of the control subjects (208 ± 61 nmole/min/ml, n=11). The HTGL levels of the hemodialysis patients with ↑TG correlated inversely with plasma total cholesterol (r_s=−0.833, p<0.01) and the d>1.006 fraction cholesterol (low + high density lipoproteins, r_s=−0.863,p<0.01), but not triglyceride. The activity of HTGL of the entire group of hemodialysis patients correlated with the plasma total cholesterol (r_s=−0.615, p<0.01), d>1.006 fraction cholesterol (r_s=−0.731, p<0.01) and low density lipoprotein cholesterol (r_s=−0.659, p<0.01). The LPL levels of the hemodialysis patients with the ↑TG (52 ± 24 nmole/min/ml) were lower than those with nlTG (70 ± 25 nmole/min/ml) and the levels of both hemodialysis groups were significantly lower (p<0.01, p<0.02, respectively) than the LPL levels in the control subjects (110 ± 43 nmole/min/ml). The ratio of LPL to total postheparin plasma lipolytic activity was lower in the hemodialysis patients with ↑TG (0.32 ± 0.15), than in the hemodialysis patients with nlTG (0.47 ± 0.18, p<0.06) or the control subjects (0.45 ± 0.09, p<0.05). Unlike HTGL, the levels of LPL did not correlate with lipid levels in the hemodialysis patients. Thus, both postheparin plasma HTGL and LPL are low in hemodialysis patients. The relationship between HTGL and low density lipoprotein cholesterol levels suggests a possible role for HTGL in low density lipoprotein catabolism.     

Lipid metabolism in pregnancy. VIII. Effects of dietary fat versus carbohydrate on lipoprotein and hepatic lipids and tissue triglyceride lipases

We have asked, is hypertriglyceridemia in the fed state in pregnancy due to intolerance to exogenous fat, accumulation of endogenous triglycerides, or accumulation of remnants of d < 1.006 lipoprotein metabolism? To answer these questions, we fed rat-free diets high in starch or sucrose, or diets containing fat or fat plus cholesterol to pregnant and nonpregnant rats for 12 days until gestational day 21 (term = 22 days). Blood was obtained 0, 4, or 8 hr after removal of food from the cages. Lipid concentrations were determined in chylomicrons and very low, low, and high density lipoproteins. Hypertriglyceridemia in pregnancy exists on both starch and sucrose containing fat-free diets and is exaggerated 4 and 8 hr after food is removed from the cage. The triglyceride rise occurs in d < 1.006 lipoproteins. With fat feeding, chylomicron triglyceride concentrations are not significantly elevated in pregnant rats, 0 or 8 hr postabsorptively despite greater food intake in pregnancy. In contrast, very low density lipoprotein (VLDL) triglyceride concentrations are elevated at all times following fat feeding in pregnant compared to nonpregnant animals. A significant contribution of lipoprotein remnants to the triglyceride rise in d < 1.006 lipoproteins seems unlikely since an isolated increase in VLDL cholesterol is not observed. No statistically significant accumulation of hepatic triglycerides occurs on any diet in pregnancy. Diet induced shifts in adipose tissue and muscle lipoprotein lipase activity are exaggerated in pregnancy while hepase in unaffected. Fetal weight is similar on all diets except sucrose where weight is reduced. Conclusions: Hypertriglyceridemia in fed pregnant rats is due to an increase in endogenous triglycerides. Remnant lipid accumulation does not appear to contribute to the endogenous hypertriglyceridemia. There is no intolerance to exogenous (dietary) fat. The results are compatible with an unimpaired delivery of exogenous fat to fat oxidizing tissues thereby maximizing glucose availability for fetal growth.     

High-density lipoproteins in myocardial infarction survivors.

Subjects with existing coronary heart disease and those with many of the conditions associated with increased risk of coronary disease have reduced levels of high-density lipoprotein (HDL) cholesterol. Since HDL cholesterol is only one index of HDL composition, a reduction of HDL cholesterol could reflect a change in HDL composition and/or a decrease in all HDL constituents. Therefore the present studies assessed the major apolipoproteins of HDL, A-I and A-II, in addition to HDL cholesterol in 90 male myocardial infarction (MI) survivors and their lipid-matched male controls. The MI survivors had significantly lower (p < 0.01) A-I (112 ± 2 mg/dl, mean ± SEM), A-II (29 ± 1 mg/dl), and HDL cholesterol (39 ± 1 mg/dl) than the lipid-matched control group (A-I, 121 ± 2; A-II, 33 ± 1; HDL cholesterol, 43 ± 1) and than a population-based male control group (n = 172; A-I, 121 ± 2; A-II, 33 ± 1; HDL cholesterol, 45 ± 1). The HDL cholesterol/A-I ratio in the MI survivors was slightly lower than the ratio in the lipidmatched control group but significantly lower (P < 0.02) than that in the population-based control group. The HDL cholesterol of both control groups was significantly negatively related to log triglyceride (r = ?0.43). Similarly the HDL cholesterol of the MI survivors was inversely correlated with log triglyceride (r = ?0.51), but the slope of this relationship was significantly steeper in the MI survivors. These results are consistent with a relative decrease of HDL in MI survivors over and above that attibutable to their increased triglyceride levels.     

Effects of pregnancy, postpartum lactation, and oral contraceptive use on the lipoprotein cholesterol/triglyceride ratio

Lipoprotein cholesterol/triglyceride ratio changes have been observed previously with sex hormone use. To determine if the lipoprotein cholesterol/triglyceride ratio is similarly changed by pregnancy and postpartum lactation, we examined pregnant subjects at 36 weeks gestation and the same women at 6 weeks postpartum and compared them to age-matched, nonpregnant women using or not using oral contraceptives. The cholesterol/triglyceride ratios were examined as means and medians and as curvilinear functions of increasing triglyceride concentration. Median ratios did not predict all ratio changes identified graphically. At very-low-density lipoprotein (VLDL) triglyceride concentrations below 40 mg/dL, the VLDL ratio is less than control in oral contraceptive users and further reduced in pregnant women. Above triglyceride concentrations of 40-60 mg/dL, the curves in the three groups are indistinguishable. No effect of lactation is observed. The low-density lipoprotein (LDL) cholesterol/triglyceride ratio is comparably lower in pregnant subjects and oral contraceptive users at all concentrations of lipoprotein triglyceride and again there is no effect of lactation. In high-density lipoprotein (HDL), there is no effect of either pregnancy or oral contraceptive use on the cholesterol/triglyceride ratio, while it is significantly higher with lactation. Postpartum decreases in the VLDL and LDL cholesterol/triglyceride ratio are seen at all lipoprotein concentrations independent of lactation. We conclude that triglyceride enriches VLDL at low concentrations and LDL at all concentrations in pregnancy and with oral contraceptive use, suggesting a common, hormonal mechanism. HDL is enriched with cholesterol during postpartum lactation, consistent with decreased transfer of cholesterol to other lipoproteins.(ABSTRACT TRUNCATED AT 250 WORDS)     

Adipose tissue lipoprotein lipase activity in type III hyperlipoproteinemia.

An impairment in the catabolism of chylomicron and very low density lipoprotein remnants appears to cause the lipid abnormalities in type III hyperlipoproteinemia. A reduction in the activity of lipoprotein lipase (LPL) has been suggested as the catabolic defect. Results in this study indicate that the activity of adipose tissue LPL measured in the fasted and fed states are in the normal range in type III hyperlipoproteinemia (fasted: type III = 2.7 +/- 1.8 mU/10(6) cells, N = 8; normals = 3.4 +/- 2.5, N = 23, p, not significant; fed: type III = 3.6 +/- 2.1, N = 7; normals = 4.8 +/- 1.8, N = 12, p, not significant). This suggests that perhaps another mechanism, such as the interaction between LPL and its lipid substrate, is abnormal, or that the activity of LPL derived from another tissue source is deficient.     

Genetic transmission of isoapolipoprotein E phenotypes in a large kindred: Relationship to dysbetalipoproteinemia and hyperlipidemia

The largest reported kindred of a proband with type III hyperlipoproteinemia was investigated by assessment of lipid and lipoprotein levels and very low density lipoprotein (VLDL) isoapolipoprotein E distributions in all accessible family members (56% of the 124 living blood relatives and 59% of the 37 spouses). The results confirm in this kindred a trimodal distribution of apoE₃/E₂ ratios, and segregation analysis of 16 informative matings classified according to $\text{E}{}_3\text{E}{}_2$ ratio demonstrated classical Mendelian inheritance of the autosomal codominant type: the $\text{E}{}_3\text{E}{}_2$ ratio is determined by two alleles, apoE₃^d and apoE₃ⁿ, which produce three phenotypes apoE₃-D, apoE₃-ND, and apoE₃-N, corresponding to the low, intermediate, and high modes, respectively. Vertical transmission of the apoE₃-D phenotype occurred in two branches of the second generation. In both instances this represented pseudodominance; i.e., products of heterozygous (apoE₃-ND) × homozygous (apoE₃-D) matings. Hyperlipidemia (defined as a low density lipoprotein cholesterol and/or plasma triglyceride level exceeding the respective age-, sex-, and sex-steroid-specific 95th percentiles derived from Lipid Research Clinics population studies) was present in 15 blood relatives in multiple lipoprotein patterns, consistent with the presence of familial combined hyperlipidemia in this kindred. Eight of nine members with the apoE₃-D phenotype had either type III hyperlipoproteinemia or, in the absence of hyperlipidemia, β-VLDL and at least marginally cholesterol-rich VLDL (VLDL-cholesterol/plasma triglyceride >0.25) (defined as dysbetalipoproteinemia). The ninth such member, the only child with this phenotype, was normal. β-VLDL and marginally cholesterol-rich VLDL was seen in but one of six hyperlipidemic family members of phenotype apoE₃-ND, in none of seven hyperlipidemic blood relatives of phenotype apoE₃-N, in no normolipidemic family members of phenotype apoE₃-ND or apoE₃-N, and in no spouses (three of whom were hyperlipidemic and nine of phenotype apoE₃-ND). Thus, among adult members of the O'D kindred the apo₃-D phenotype was nearly specifically associated with dysbetalipoproteinemia or, when hyperlipidemia was present, type III hyperlipoproteinemia.     

Contrasting effects of unmodified and time-release forms of niacin on lipoproteins in hyperlipidemic subjects: clues to mechanism of action of niacin

To minimize the cutaneous flushing symptoms associated with niacin use, a time-release capsule form of niacin has been formulated. Thus study compares the effects of time-release niacin with those of unmodified niacin on lipoprotein lipids, including HDL2 and HDL3, apoproteins A-I and A-II, clinical chemistries, symptomatic side effects, and adherence to the medication regimen. Seventy-one primarily hypercholesterolemic subjects were randomized to either unmodified niacin or time-release niacin ad took medication for a six-month period. The two groups were closely matched on anthropomorphic and lipid variables. Adherence to the therapeutic regimen at a dose of 1.5 g/d in the first month of treatment was similar in the two groups. Thereafter, at a dose of 3.0 g/d, adherence was in excess of 90% among subjects taking unmodified niacin but only 64% among those taking time-release niacin, chiefly because of aggravated gastrointestinal symptoms; cutaneous flushing side effects, however, were slightly less common with time-release niacin. At these levels of adherence, LDL cholesterol (C) was reduced 21% by unmodified niacin and 13% by the time release form. Plasma total triglyceride was reduced more with unmodified niacin (27%) than with time-release niacin (8% maximum), and HDL-C and HDL2-C were increased significantly with unmodified niacin (26% and 36%) and were not significantly changed by time-release niacin. Increased to a similar degree on both regimens were HDL3-C (approximately 35%) and apoA-I (approximately 12%). ApoA-II was not affected by either drug regimen.(ABSTRACT TRUNCATED AT 250 WORDS)     

Lipid and lipoprotein triglyceride and cholesterol interrelationships: Effects of sex,hormone use,and hyperlipidemia

The interrelationships of lipid and lipoprotein cholesterol and triglyceride concentrations in normolipidemic and hyperlipidemic employees of the Pacific Northwest Bell Telephone Company were examined bivariately using correlation analysis and multivariately by factor analysis. Application of the latter resulted in the identification of three distinct lipoprotein lipid clusters, which succinctly describes their metabolic relationships. Among normolipidemic subjects, the interrelationships were found to be similar in male and female subjects, but hormone use by women considerably altered interrelationships that involved high-density lipoprotein cholesterol (HDL-C) and triglyceride. Among hyperlipidemic subjects, we found that elevation in cholesterol level alone rarely altered relationships, but elevation in triglyceride level either alone or in conjunction with an elevation in cholesterol concentration was associated with substantial changes in relationships involving the low-density lipoprotein (LDL) fraction. In many instances, positive relationships between LDL cholesterol (LDL-C) and other lipoprotein lipids became inverse in the presence of triglyceride elevation. We conclude that hormone use by women and hypertriglyceridemia with or without an elevation in cholesterol level clearly alter lipoprotein relationships, whereas pure hypercholesterolemia does not. These alterations provide a basis for investigating pathophysiologic mechanisms in hypertriglyceridemia.     

Thyroid function and lipids in patients with chronic renal disease treated by hemodialysis: with comments on the "free thyroxine index".

It is known that hypertriglyceridemia and abnormal thyroid function tests are common in patients with chronic renal disease treated by hemodialysis, but a possible association between these abnormalities has not been investigated. We measured serum thyroid hormone and lipid levels in one hundred patients receiving chronic hemodialysis. There were ten patients with elevations in serum thyrotropin (TSH). In the remaining patients with normal TSH, those receiving no drugs known to affect thyroid function had low mean values for thyroxine (T₄) and triiodothyronine (T₃), while the mean T₃ resin uptake (fraction of T₃ bound to resin, T₃U) was normal, resulting in low mean values for free T₄ index (FT₄I = T₄ × T₃U) and free T₃ index (FT₃I = T₃ × T₃U). The depression in T₃ was greater than the depression in T₄, so that $\text{T}{}_3\text{T}{}_4$ ratios were also low. Androgen-treated male patients had the lowest T₃ and T₄ levels; their T₃Us were high, presumably reflecting androgen-induced depression of T₄-binding globulin, but their mean FT₄I and FT₃I were still lower than those of the other patients. Evidence is presented that the conventional T₃U underestimates the proportion of free hormone when binding protein levels are low, and that a “new T₃U,” the ratio of resin-bound T₃ to serum protein-bound T₃, more accurately reflects the proportion of free hormone. Computing a “new FT₄I,” as new T₃U × T₄, yielded values in the androgen-treated group similar to those of the other patients. Patients receiving propranolol had lower $\text{T}{}_3\text{T}{}_4$ ratios than the other patients, presumably due to propranolol's inhibitory effect on peripheral conversion of T₄ to T₃. No relationship was found between indexes of thyroid function and lipid levels. Subjects with high TSH levels had levels of triglyceride and cholesterol similar to the remaining patients. Among those with normal TSH levels, low levels of T₃ or T₄ did not correlate with abnormal lipid levels. It is concluded that thyroid hormone abnormalities do not explain the hypertriglyceridemia of patients on chronic hemodialysis. The frequency of low T₄ and T₃ values in such patients, in the absence of clinical evidence of hypothyroidism and, in most instances, in the absence of TSH hypersecretion, remains unexplained.     

Mechanism of the repetitive ventricular response in man

The electrophysiologic characteristics of the repetitive ventricular response that followed an electrically induced single premature ventricular complex were evaluated to determine its mechanism during atrial pacing or sinus rhythm in 30 patients. Seven patients had preexisting bundle branch block. His bundle or right bundle branch deflections did not precede the repetitive complex in 29 of the 30 patients, which implies that the proximal His-Purkinje system was not involved in the reentry circuit. In 24 of 30 patients the QRS axis of the repetitive complex was divergent 45 ° or more from the stimulated complex. In 22 of 30 patients the repetitive complex had a right bundle branch block configuration. In 14 of 18 patients with two or more repetitive complexes, the QRS pattern changed from beat to beat, which implies that either the reentry pathway or conduction was changing. Thus, the repetitive ventricular response, which can be associated with clinically important ventricular arrhythmias, probably represents intraventricular rather than proximal His-Purkinje system reentry.     

The variability of arterial pressure

Echocardiograms from 562 patients were examined for evidence of the pattern of tricuspid valve prolapse. Criteria for the diagnosis can be established similar to those applicable to mitral prolapse. In 500 consecutive patients without mitral valve prolapse, there were no cases of isolated tricuspid valve prolapse. Eleven of 53 (21 per cent) patients with mitral valve prolapse also had tricuspid valve prolapse. Four of six (67 per cent) patients with Marfan's syndrome and mitral valve prolapse also had tricuspid valve prolapse. The occurrence of this echocardiographic pattern as an isolated finding as well as associated with mitral valve prolapse was significantly less than previous angiographic reports. Patients with both these findings tended to be older than those with mitral valve prolapse alone, but clinically differed in no other way. Use of standardized technique can minimize errors in diagnosis.     

Reversible abnormalities in postheparin lipolytic activity during the late phase of release in diabetes mellitus (postheparin lipolytic activity in diabetes).

To test whether abnormalities in multiphasic release of lipoprotein lipase are associated with hypertriglyceridemia in diabetes mellitus, postheparin lipolytic activity (PHLA) was measured during a high-dose, constant heparin infusion in 20 diabetic subjects with hypertriglyceridemia, 25 nondiabetic hypertriglyceridemic subjects and 7 normal subjects. The standard low heparin dose PHLA and the PHLa during the early phase of the heparin infusion were the same in all groups. In constrast, the PHLA during the late phase of the heparin infusion was lower in the 12 untreated diabetic subjects than in the 25 nondiabetic hypertriglyceridemic and the 7 normal subjects (p less than 0.001). An abnormality in late phase PHLA in the untreated diabetic subjects was more apparent when it was compared to the level of PHLA attained during the early phase of the heparin infusion (Equilibrium PHLA/60 min PHLA). The relative PHLA in the late phase of the infusion was lower in the untreated diabetic subjects (0.671 +/- 0.147) than in the nondiabetic hypertriglyceridemic subjects (0.847 +/- 0.019, p less than 0.001), or in the chronically treated diabetic subjects (0.823 +/- 0.108, p less than 0.05). Among the untreated diabetic subjects, increasing fasting glucose levels were associated with both decreasing absolute PHLA levels at the late phase of the infusion (r = 0.61, p less than 0.02) and greater decreases in relative PHLA during the infusion (r = -0.80, p less than 0.001). Treatment of the diabetes with long-term oral sulfonylurea or insulin therapy corrected the abnormality in the late phase PHLA with an associated decrease in plasma triglyceride levels (p less than 0.001). In five subjects with a deficient PHLA response to a standard, low dose of heparin, the PHLA response was low throughout the heparin infusion. With treatment, the PHLA response to the low heparin dose corrected rapidly toward normal in those two diabetic subjects with PHLa deficiency, and the early PHLA response during the heparin infusion increased. However, the late phase abnormality in all untreated diabetic subjects did not correct to normal until after several months of antihyperglycemic therapy. In the untreated diabetic subjects the degree of elevation of the plasma triglyceride level appeared to result from the interaction of the abnormality in PHLA with the presence or absence of an inherited familial lipid disorder.