Author + information
- Received February 27, 1996
- Revision received December 17, 1996
- Accepted January 9, 1997
- Published online April 1, 1997.
- Philip J Chowienczyk, MRCPA,*,
- Gerald F Watts, MD, MRCPB,
- Anthony S Wierzbicki, DPhil, BM, BChA,
- John R Cockcroft, MRCPA,
- Sally E Brett, BNA and
- James M Ritter, DPhil, FRCPA
- ↵*Dr. Philip J. Chowienczyk, Department of Clinical Pharmacology, St. Thomas’ Hospital, Lambeth Palace Road, London SE1 7EH, England, United Kingdom.
Objectives. We sought to determine whether hypertriglyceridemia in patients with lipoprotein lipase (LPL) dysfunction is associated with endothelial dysfunction in resistance vessels of the forearm vasculature.
Background. Vasodilator responses to acetylcholine, acting through stimulation of nitric oxide (NO) release from the endothelium, are impaired in hypercholesterolemia and normalized by l-arginine, suggesting dysfunction of the l-arginine/NO pathway. Similar abnormalities have been reported in conditions associated with hypertriglyceridemia, such as non–insulin-dependent diabetes. The relation between endothelial function and plasma triglyceride concentrations has, however, not previously been studied in vivo.
Methods. We examined forearm blood flow responses to brachial artery infusions of acetylcholine (alone and with l-arginine) and nitroprusside (an NO donor) in 17 patients with severe hypertriglyceridemia (mean [±SD] plasma triglyceride concentration 1,914 ± 1,288 mg/dl) but normal low density lipoprotein cholesterol (89 ± 31 mg/dl) and in 34 normolipidemic control subjects. Severe LPL dysfunction was demonstrated in 10 of 17 patients.
Results. Acetylcholine (7.5 and 15 μg/min) produced similar forearm blood flow responses in hypertriglyceridemic patients (mean [±SEM] 7.7 ± 0.9 and 10.5 ± 1.2 ml/min per 100 ml) and in control subjects (7.5 ± 0.6 and 11.0 ± 0.8 ml/min per 100 ml, p = 0.78 by analysis of variance). Responses to acetylcholine co-infused with l-arginine (10 mg/min) and nitroprusside (3 and 10 μg/min) were also similar in hypertriglyceridemic patients and control subjects (p = 0.93 and p = 0.27 for acetylcholine with l-arginine and nitroprusside, respectively). The ratio response to acetylcholine/response to nitroprusside differed between hypertriglyceridemic patients and control subjects by only 1%. The study had >90% power (alpha = 0.05) to detect a difference >30% in this ratio.
Conclusions. Severe hypertriglyceridemia associated with LPL dysfunction is not associated with the degree of endothelial dysfunction seen in moderate hypercholesterolemia when responses to acetylcholine are impaired by >40%.
(J Am Coll Cardiol 1997;29:964–8)
© 1997 by the American College of Cardiology
Endothelial dysfunction, characterized by impaired responses to the endothelium-dependent vasodilator acetylcholine but preserved responses to nitrovasodilators, is a feature of early atherosclerosis ([1, 2]). In hypercholesterolemia such endothelial dysfunction precedes clinical evidence of atherosclerosis occurring both in epicardial coronary vessels and in coronary and forearm resistance vessels ([3, 4]). Acetylcholine stimulates the synthesis of nitric oxide (NO) from l-arginine within the endothelium, whereas nitrovasodilators act directly on vascular smooth muscle through the same effector mechanism. l-Arginine corrects deficient endothelium-dependent vasodilation in hypercholesterolemic humans ([5, 6]). In addition to vasodilation, NO inhibits platelet adhesion and aggregation ([7, 8]), vascular smooth muscle cell proliferation (), monocyte adhesion to endothelial cells () and oxidation of low density lipoprotein (LDL) (). Such actions suggest that NO may inhibit atherosclerosis. This is consistent with studies involving modulation of the l-arginine/NO pathway in hypercholesterolemic rabbits: inhibition of NO synthase promotes atherosclerosis ([12, 13]), whereas l-arginine inhibits development of intimal lesions (). The antiatherogenic actions of NO may thus underlie the association between endothelial dysfunction and the development of vascular disease. In vitro studies show that LDL cholesterol, particularly oxidized LDL has powerful inhibitory effects on vasorelaxation induced by acetylcholine (). Lipoproteins other than LDL have also been shown to inhibit endothelium-dependent relaxation (); the relative importance of these is less clear, however. Impaired vasodilator responses to acetylcholine have been observed in conditions associated with hypertriglyceridemia, such as non–insulin-dependent diabetes (), and by some ([18, 19]) but not all () investigators in essential hypertension. The relation between endothelial function and hypertriglyceridemia has not previously been studied in vivo. In the present study we examined whether endothelial function was impaired in patients with LPL dysfunction with the Fredrickson type V phenotype. Such patients have severe hypertriglyceridemia but normal or low levels of LDL cholesterol. Furthermore, in contrast to mixed hyperlipidemia (), the production of small, dense LDL with the increased propensity to oxidation () may not be increased ([23, 24]). Patients with LPL dysfunction provide a useful model to assess the effects of hypertriglyceridemia without confounding effects of elevated or altered LDL cholesterol.
Seventeen patients with severe hypertriglyceridemia (Fredrickson type V) were recruited from St. Thomas’ Hospital Lipid Clinic. Liproprotein lipase activity after an intravenous infusion of heparin, 50 U/kg, was assessed in 10 of 17 patients and was reduced or absent. All patients were on a low fat (<25 g/day) diet. Two patients had recently developed diabetes. For each patient two gender-matched normolipidemic control subjects (total cholesterol <230 mg/dl, triglycerides <350 mg/dl) were recruited from the staff of St. Thomas’ Hospital and from the Metropolitan Traffic Police and were studied contemporaneously. Control subjects were nonsmokers and had no family history of premature coronary artery disease.
1.2 Lipid measurements.
Lipid profiles were obtained after a 12-h fast using standard enzymatic methods. High density lipoprotein (HDL) cholesterol was measured after precipitation of apoprotein B–containing lipoproteins using dextran sulphate/magnesium chloride. In control subjects LDL cholesterol was calculated using the Friedewald equation. In hypertriglyceridemic patients preparative ultracentrifugation was used to separate lipoprotein fractions and cholesterol and triglyceride concentrations measured in each fraction. Low density lipoprotein cholesterol was separated at 1.006 to 1.063 g/ml. Patient characteristics and lipid profiles are summarized in Table 1.
1.3 Experimental protocol.
The study was approved by the West Lambeth Health Authority Ethics Committee, and all subjects gave written informed consent. Studies were performed in a quiet clinical laboratory (temperature controlled between 24°C and 26°C during each study). Forearm blood flow was measured in both arms using venous occlusion plethysmography with electrically calibrated strain gauges ([25, 26]). Collecting cuff pressure was 40 mm Hg and wrist cuff occlusion pressure was 180 mm Hg. A 27-gauge needle was inserted into the left brachial artery under sterile conditions using <1 ml of 1% lidocaine hydrochloride to provide local anesthesia. Drugs were dissolved in saline (0.9% NaCl) and saline or drug solution infused at a rate of 1.0 ml/min by constant rate infusion pumps. Basal blood flow was recorded after a 15-min infusion of saline. Two cumulative doses of sodium nitroprusside (3 and 10 μg/min, each dose for 6 min) were then infused, followed by saline until blood flow returned to baseline (between 6 and 12 min). After repeat basal flow measurements, a similar cumulative infusion of two doses of acetylcholine chloride (7.5 and 15 μg/min, each dose for 6 min) was given. After further basal flow measurements obtained after 12 min of saline, l-arginine (10 mg/min) was infused alone for 6 min and then co-infused with acetylcholine (7.5 and 15 μg/min, as done earlier). We previously demonstrated severe impairment of responses to acetylcholine in hypercholesterolemic men at these doses ([4, 27]). At higher doses the proportion of the response mediated through the l-arginine/NO pathway appears to diminish with acetylcholine acting through an NO-independent mechanism (). Forearm blood flow was measured during the last 3 min of each infusion period. Flows were recorded for 10 s every 15 s, and the mean of the last five measurements in each recording period was used for analysis. Blood flow was measured in ml/min per 100 ml of forearm volume ().
Sodium nitroprusside, acetylcholine chloride and l-arginine were obtained from Roche (Basel, Switzerland), Coopervision (Southampton, United Kingdom) and Torbay Hospital Pharmacy (Torbay, United Kingdom).
1.5 Statistical methods.
Data (Table 1) are presented as mean value ± SD. Forearm blood flow data (Table 2) are presented as mean value ± SE. Repeated measures analysis of variance (ANOVA) was used to compare forearm blood flow responses to acetylcholine and sodium nitroprusside in hypertriglyceridemic patients and control subjects. Influences of baseline blood flow and forearm length were taken into account by incorporating these measurements as covariates (). Age and gender were also incorporated as covariates. As in previous studies (), we observed a nearly linear relation between the standard deviation of blood flow responses and the mean blood flow, suggesting the data to be log-normally distributed. We therefore log-transformed blood flows before using these as repeated measures in ANOVA. The ratio response to acetylcholine (mean for both doses)/response to nitroprusside (mean for both doses) was calculated, and 95% confidence intervals were calculated for the difference in this ratio between hypertriglyceridemic patients and control subjects. Differences were considered significant at p < 0.05.
Blood flow in the noninfused (control) arm did not change significantly in response to drug infusions in the contralateral arm, confirming that at the doses used acetylcholine and nitroprusside did not cause systemic effects when infused into the brachial artery. Blood flow responses in the infused arm of hypertriglyceridemic patients and control subjects are shown in Table 2, and mean responses to both doses of each drug are shown in Fig. 1. Nitroprusside and acetylcholine produced similar responses in both hypertriglyceridemic patients and control subjects (p = 0.27 and p = 0.78 for the difference between responses in hypertriglyceridemic patients and control subjects for nitroprusside and acetylcholine, respectively). In the control subjects there was no significant correlation between responses to acetylcholine and LDL cholesterol (p = 0.89). The ratio response to acetylcholine/response to nitroprusside differed between hypertriglyceridemic patients and control subjects by only 1% (p = 0.89), with 95% confidence limits −19 to 25%. l-Arginine infused alone or co-infused with acetylcholine did not increase blood flow responses significantly. Responses to acetylcholine during co-infusion of l-arginine were similar in hypertriglyceridemic patients and control subjects (p = 0.93). A subgroup analysis was performed in the men only, and findings of similar blood flow responses to nitroprusside and acetylcholine were unchanged. The difference in the acetylcholine/nitroprusside ratio between hypertriglyceridemic men (n = 10) and control men (n = 20) was −9% (p = 0.45) with 95% confidence intervals −27 to 15%. Exclusion of the hypertriglyceridemic patients in whom LPL activity was not assessed also did not alter the findings of similar blood flow responses to nitroprusside and acetylcholine in hypertriglyceridemic patients and control subjects.
3.1 Hypertriglyceridemia and atherosclerosis.
Coronary artery disease is independently associated with elevated LDL cholesterol and with depressed HDL cholesterol (). Hypertriglyceridemia is less strongly associated with coronary atherosclerosis (), and much of this association disappears when multivariate analysis is used to correct for LDL and HDL cholesterol (). Such analyses are associated with many problems, however, and it remains uncertain whether elevated triglycerides are a causal factor in the development of atherosclerosis (). The association between plasma triglyceride concentration and coronary artery disease may be secondary to other disorders such as glucose intolerance, obesity or hypertension (). Endothelial dysfunction occurs before the development of structurally evident atherosclerosis and may be an initiating event in the atherosclerotic process (). We sought to establish whether endothelial function was impaired in patients with severe hypertriglyceridemia but normal or low LDL cholesterol.
3.2 Major findings in hypertriglyceridemic patients with LPL dysfunction.
Patients with LPL dysfunction in whom fasting triglyceride values were more than 10-fold those in control subjects but in whom LDL cholesterol was normal or low did not show impaired vasodilator responses to acetylcholine or nitroprusside. The power of our study to detect a blunting of the response to acetylcholine relative to the response to nitroprusside of >30% was >90%. The 95% confidence intervals for the difference in this ratio between hypertriglyceridemic patients and control subjects were −19 to 25%. Therefore, we cannot exclude a modest effect of hypertriglyceridemia on endothelial function, as reflected by these confidence intervals. However, responses to acetylcholine in patients with type II hypercholesterolemia and modest (twofold greater than control) elevations in LDL cholesterol are reduced by >40% ([4, 36]). Our results suggest, therefore, that any effect of hypertriglyceridemia on endothelial function in patients with LPL dysfunction is minor compared with that of LDL cholesterol in type II hypercholesterolemic patients.
3.3 Possible confounding factors and limitations of this study.
Two of our patients had non–insulin-dependent diabetes. Vasodilator responses to acetylcholine are impaired in non–insulin-dependent diabetes (), so this could have biased our study toward the opposite conclusion. Premenopausal women are protected from the adverse effects of hypercholesterolemia on endothelial function (). It is possible, therefore, that the inclusion of women in our study could have biased the study toward a negative result. This is unlikely, though, because only three of seven women were premenopausal and exclusion of women from the analysis demonstrated similar blood flow responses to acetylcholine and nitroprusside in hypertriglyceridemic men and control men. Although impaired acetylcholine responsiveness has been regarded as indicative of generalized dysfunction of the endothelium, this may not be the case in all conditions. The present study does not exclude other types of endothelial dysfunction relating, for example, to basal NO synthesis or to other receptor-operated pathways ().
3.4 Implications for patients with other phenotypes.
In many conditions hypertriglyceridemia may act as a surrogate marker for factors such as insulin resistance and/or abnormalities of LDL particle size, which may influence endothelial function (). Thus, in many conditions hypertriglyceridemia may be associated with endothelial dysfunction without being the causal factor. The present study suggests that hypertriglyceridemia itself is unlikely to be a causal factor in endothelial dysfunction in dyslipidemic states. We must be cautious, however, in extrapolating the results of the present study to patients with other phenotypes because we cannot be certain of the relative influence of other lipids. The low LDL cholesterol in patients with LPL dysfunction could mask an adverse effect of hypertriglyceridemia. However, we did not observe any relation between acetylcholine responsiveness within the normal range of plasma concentrations of LDL cholesterol in the control subjects, making this unlikely. Our patients with LPL dysfunction were inevitably hypercholesterolemic because both triglycerides and cholesterol are carried in chylomicrons and VLDL fractions. This type V lipid phenotype is typical of LPL-deficient patients on a low fat/high carbohydrate diet ([39, 40]). Our study also suggests that chylomicron and VLDL cholesterol influence acetylcholine responsiveness less than LDL cholesterol does. Elevated HDL cholesterol appears to ameliorate impaired acetylcholine responsiveness in the coronary vasculature of patients with established coronary atherosclerosis (). In our patients with LPL dysfunction HDL cholesterol was lower than that in control subjects. Our findings therefore suggest that low HDL cholesterol is not necessarily associated with impaired acetylcholine responsiveness in the absence of elevated LDL cholesterol.
We have shown that severe hypertriglyceridemia in patients with LPL dysfunction is not associated with significant dysfunction of the l-arginine/NO pathway in resistance vessels of the forearm vasculature during stimulation with acetylcholine. These results suggest that hypertriglyceridemia is unlikely to be a causal factor in the development of endothelial dysfunction in dyslipidemic states.
☆ This work was supported by the British Heart Foundation, London, England, United Kingdom.
- analysis of variance
- high density lipoprotein
- low density lipoprotein
- lipoprotein lipase
- nitric oxide
- very low density lipoprotein
- Received February 27, 1996.
- Revision received December 17, 1996.
- Accepted January 9, 1997.
- The American College of Cardiology
- Zeiher AM,
- Drexler H,
- Wollschläger H,
- Just H
- Radomski MW,
- Palmer RMJ,
- Moncada S
- Kubes P,
- Suzuki M,
- Granger DN
- Naruse K,
- Shimizu K,
- Muramatsu M,
- et al.
- Cayatte AJ,
- Palacino JJ,
- Horten K,
- Cohen RA
- Takahashi M,
- Yui Y,
- Yasumoto H,
- et al.
- Linder L,
- Kiowski W,
- Bühler FR,
- Lüscher TF
- Dejager S,
- Brucker E,
- Chapman MJ
- Demant T,
- Gaw A,
- Watts GF,
- et al.
- Casino PR,
- Kilcoyne CM,
- Quyyumi AA,
- Hoeg JM,
- Panza JA
- Zeiher AM,
- Schächinger V,
- Hohnloser SH,
- Saurbier B,
- Just H