SUBCLINICAL HYPOTHYROIDISM (sht) is a disorder

Transcription

1 X/03/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 88(8): Printed in U.S.A. Copyright 2003 by The Endocrine Society doi: /jc Impaired Endothelium-Dependent Vasodilatation in Subclinical Hypothyroidism: Beneficial Effect of Levothyroxine Therapy STEFANO TADDEI, NADIA CARACCIO, AGOSTINO VIRDIS, ANGELA DARDANO, DANIELE VERSARI, LORENZO GHIADONI, ANTONIO SALVETTI, ELE FERRANNINI, AND FABIO MONZANI Department of Internal Medicine, University of Pisa, Pisa, Italy Subclinical hypothyroidism (sht) is associated with enhanced cardiovascular risk. To test the hypothesis that patients with sht are characterized by endothelial dysfunction and impaired nitric oxide (NO) availability, in 14 patients [serum cholesterol, mg/dl ( mm)] and 28 euthyroid subjects, subdivided into groups A and B [serum cholesterol, mg/dl ( mm) and mg/dl ( mm), respectively], we studied the forearm blood flow (strain-gauge plethysmography) response to intrabrachial acetylcholine, an endothelium-dependent vasodilator, at baseline and during infusion of N G -monomethyl-l-arginine (L-NMMA), a NO synthase inhibitor. Response to sodium nitroprusside and minimal forearm vascular resistances were also evaluated. In sht patients, vasodilation to acetylcholine was reduced, compared with group B ( % vs. 503 SUBCLINICAL HYPOTHYROIDISM (sht) is a disorder characterized by elevated serum TSH levels despite normal free thyroid hormone (FT 3 and FT 4 ) values (1). The prevalence of this condition is higher in women than men, and it increases with age, reaching a peak of 21% in women and 16% in men over 74 yr of age (2). Although sht is often asymptomatic, nearly 30% of patients have symptoms that are suggestive of thyroid hormone deficiency (3 5). Therefore, important questions have been raised regarding its clinical relevance and appropriate management (6). Previous studies have suggested an association between sht and coronary heart disease (7, 8). Whether this association is related to sht-induced changes in serum lipid levels is still debated (9 11). In a recent population-based survey, sht emerged as a significant risk factor for aortic atherosclerosis and myocardial infarction in elderly women, independently of serum cholesterol levels (12). Therefore, the association between sht and atherosclerosis may not be entirely explained by dyslipidemia, and further mechanisms must be taken into consideration (13 15). The endothelium plays a major role in the maintenance of vascular function and integrity through the production of Abbreviations: ApoB, Apolipoprotein B; BMI, Body mass index; ESR, erythrocyte sedimentation rate; FBF, forearm blood flow; FT, free thyroid hormone; HDL, high-density lipoprotein; Lp (a), Lipoprotein (a); L-NMMA, N G -monomethyl-l-arginine; l-t 4, levothyroxine ; LDL, lowdensity lipoprotein; MFVRs, minimal forearm vascular resistances; NO, nitric oxide; sht, subclinical hypothyroidism; Tg-Ab, antithyroglobulin; TPO-Ab, antithyroid peroxidase. 19%, P ) and group A (663 65%, P 0.02 vs. group B and P vs. sht). L-NMMA blunted the vasodilation to acetylcholine in groups A and B ( % and % maximal forearm blood flow reduction, respectively, P < vs. acetylcholine), whereas it was ineffective in sht patients ( %). Response to sodium nitroprusside and minimal vascular resistances were similar. In sht (n 9) patients, 6 months of euthyroidism by levothyroxine replacement increased acetylcholine-vasodilation and restored L-NMMA inhibition. Patients with sht are characterized by endothelial dysfunction resulting from a reduction in NO availability, an alteration partially independent of dyslipidemia and reversed by levothyroxine supplementation. (J Clin Endocrinol Metab 88: , 2003) vasodilator and vasoconstrictor substances (16, 17). The most important vasodilator substance produced by the endothelium is nitric oxide (NO) (18, 19), derived from the conversion of l-arginine into citrulline via the activity of the enzyme NO-synthase (20). In the presence of major cardiovascular risk factors, including hypercholesterolemia, the endothelium can generate oxidative stress, which, in turn, causes NO breakdown (16, 21). It is now well established that endothelial dysfunction, a condition characterized by decreased NO availability, acts as a promoter of atherosclerosis and is associated with an increased risk of cardiovascular events (22 24). The aim of this study was to assess vascular reactivity in sht patients and its relation to the serum lipid profile. The effect of restored euthyroidism was also evaluated. Patients Patients and Methods The study population included 14 sht patients with Hashimoto s thyroiditis, who were positive for both antithyroid peroxidase (TPO-Ab) and antithyroglobulin (Tg-Ab) antibodies (Table 1). To be enrolled, patients had to have documented sht (TSH 3.6 IU/ml) for at least 6 months before the study. As controls, 28 euthyroid subjects, matched to the patient group for sex, age, and body mass index (BMI), were recruited among staff and relatives of patients. According to the lipid profile, they were divided into group A [n 14; serum cholesterol less than 200 mg/dl (5.2 mm)] and group B (n 14; lipid profile matched to sht patients). Women were premenopausal, with regular menses, and none were pregnant. Obese (BMI 30 kg/m 2 ) subjects, smokers, and those with hypertension, diabetes mellitus, or other systemic diseases were excluded from the study. No individual was taking any drug. 3731

2 3732 J Clin Endocrinol Metab, August 2003, 88(8): Taddei et al. Endothelial Dysfunction in Hypothyroidism TABLE 1. Clinical characteristics in sht patients and controls Group A (n 14) Group B (n 14) sht (n 14) Age (yr) Sex (F/M) 11/3 11/3 12/2 SBP (mm Hg) DBP (mm Hg) BMI (kg/m 2 ) ESR (mm) a Fibrinogen (mg/dl) Total cholesterol (mg/dl) ( ) b ( ) b ( ) HDL cholesterol (mg/dl) 58 7 (46 75) (42 86) 59 9 (36 79) LDL cholesterol (mg/dl) (65 117) c (89 215) c (82 195) Triglycerides (mg/dl) (54 147) (64 164) (62 201) ApoB (g/liter) c c Lp(a) (g/liter) TSH ( IU/ml) ( ) ( ) d ( ) Free T 4 (pg/ml) a Free T 3 (pg/ml) a TPO-Ab (U/ml) Negative Negative Tg-Ab (U/ml) Negative Negative Data are reported as mean SEM and ranges. a P 0.05 vs. Groups A and B. b P vs. Groups A and B. c P 0.01 vs. Group A. d P vs. Group A. SBP, Systolic blood pressure; DBP, diastolic blood pressure. Conversion factors: total, HDL, and LDL cholesterol: ; triglycerides: ; free T 3 : 1.536; free T 4 : All study subjects gave their signed informed consent to the study, which was approved by the Institutional Ethics Committee. Protocol Vascular reactivity was assessed by the perfused forearm technique, as already described (21). Briefly, the brachial artery was cannulated for drug infusion at systemically ineffective rates, and intraarterial blood pressure and heart rate were monitored. Forearm blood flow (FBF) was measured in both forearms by strain-gauge venous occlusion plethysmography (25). Circulation to the hand was excluded 1 min before FBF measurement by inflating a pediatric cuff around the wrist at suprasystolic blood pressure. Forearm volume was measured by the water displacement method. NO contribution to endothelium-dependent vasodilatation was estimated by performing a dose-response curve to acetylcholine (at infusion rates of 0.15, 0.45, 1.5, and 4.5 g min 1 dl 1 ; 5 min at each dose), in the absence and presence of N G -monomethyl-l-arginine (L-NMMA; 100 g min 1 dl 1 ), to block NO synthase (26). To avoid L-NMMAinduced basal flow modification, the NO-clamp (which allows assessment of endothelial agonists in the presence of NO-synthase blockade without changes in basal flow) was applied (27). Thus, after 10 min of L-NMMA infusion, sodium nitroprusside was coinfused (0.3 g min 1 dl 1 for 5 min) to neutralize the L-NMMA-induced vasoconstriction and restore baseline FBF. Endothelium-independent vasodilatation was assessed with a doseresponse curve to sodium nitroprusside (1, 2, and 4 g min 1 dl 1 ; 5 min at each dose) (28). These rates were selected to induce vasodilatation comparable with that obtained with acetylcholine. A 30-min washout was allowed between dose-response curves. Finally, minimal forearm vascular resistances (MFVRs), an index of structural vascular alteration (29), were evaluated by calculating the ratio between mean intraarterial blood pressure and maximal forearm vasodilation induced by 13 min of forearm ischemia (obtained by inflating the plethysmographic cuff at 300 mm Hg) plus 1 min of hand exercise (30). Patients were then given levothyroxine (l-t 4 ) replacement therapy. Treatment always started with 25 g, the dose being then gradually increased. To confirm patient compliance and to adjust the l-t 4 dose, serum TSH was measured every 3 months. The l-t 4 dose required for restoring euthyroidism averaged 67.5 g/d. Nine sht patients were reevaluated after 6-months of stable euthyroidism [achieved by an average of 11 months of treatment (range, 6 15)], whereas five patients dropped out because of poor compliance. Analytical measurements Serum FT 3 and FT 4 levels were measured by specific RIA (Techno- Genetics Recordati, Milan, Italy). TSH was determined with an ultrasensitive immunoradiometric assay method (Cis Diagnostici, Tronzano Vercellese, Italy). Tg-Ab was measured by specific immunoradiometric assay (Biocode, Sclessin, Belgium); TPO-Ab was measured by specific RIA (Sorin Biomedica, Saluggia, Italy). Total serum cholesterol, triglycerides, and high-density lipoprotein (HDL)-cholesterol were assayed by enzymatic methods (Roche, Diagnostic, Mannheim, Germany). Lowdensity lipoprotein (LDL)-cholesterol level was calculated by Friedewald s equation. Apolipoprotein B (ApoB) and lipoprotein (a) [Lp(a)] were assayed by nephelometry (Behring Diagnostics, Marburg, Germany). Normal ranges in our laboratory are as follows: FT pm; FT pm, TSH IU/ml, Tg-Ab less than 50 U/ml, and TPO-Ab less than 10 U/ml; ApoB g/liter; and Lp(a) less than 0.3 g/liter. Statistical analysis Data were analyzed in terms of FBF increase or decrease, taken as evidence of local vasodilation and vasoconstriction, respectively. Student s unpaired or paired t test, 2 test, and ANOVA for repeated measures were used as appropriate. Because of the highly skewed distribution of Lp(a) values, the nonparametric Mann-Whitney U test was applied to assess differences between groups. Linear regression analysis was carried out by standard techniques. Unless otherwise stated, results were expressed as mean sem. Results Humoral parameters At baseline, serum TSH levels were significantly higher in sht patients than in controls (P ), whereas serum FT 3 and FT 4 levels, though still within the normal range, were lower (P and P 0.05, respectively). With regard to the serum lipid profile, sht patients had serum total and

3 Taddei et al. Endothelial Dysfunction in Hypothyroidism J Clin Endocrinol Metab, August 2003, 88(8): LDL-cholesterol and ApoB levels that were higher than group A subjects (P , P 0.001, and P 0.01, respectively) but were comparable with those of group B subjects (Table 1). Mean Lp(a) values proved similar in the three subgroups, although four sht patients and one group B subject had elevated Lp(a) values. A positive relationship between serum TSH and LDL-cholesterol levels (r 0.66, P 0.05) was observed. Finally, sht patients showed a significantly higher erythrocyte sedimentation rate (ESR) than both control groups (P 0.01), whereas fibrinogen was similar in the three study groups (Table 1). Vascular parameters Throughout all experimental steps, blood pressure and contralateral FBF remained stable (data not shown). MFVRs were similar in sht patients and control subjects ( , , and mm Hg. min. ml -1. dl -1 in sht patients, group A, and group B, respectively). Response to acetylcholine and sodium nitroprusside The FBF increase induced by acetylcholine was significantly smaller in sht patients (from to a maximum of ml min 1 dl 1, % increase above baseline), compared with group B (from to a maximum of ml min 1 dl 1 ; % increase, P vs. sht). Moreover, both sht patients and group B subjects had a lower response to acetylcholine, in comparison with group A (from to a maximum of ml min 1 dl 1 ; % increase, P 0.02 vs. group B and P vs. sht). In contrast, vasodilatation to sodium nitroprusside was not significantly different between sht patients (396 38% increase) and group A (472 33%) or group B subjects (424 41%) (Fig. 1). In the sht group, significant relationships were found between maximal response to acetylcholine and both serum TSH (r 0.55, P 0.05) (Fig. 2) and LDL-cholesterol levels (r 0.68, P 0.01). Effect of L-NMMA on the response to acetylcholine L-NMMA significantly blunted the maximum vasodilating effect of acetylcholine in group A (saline, ml min 1 dl 1 ; L-NMMA, ; % maximal FBF reduction, P vs. acetylcholine alone) and, to a lesser extent, in group B subjects (saline, ml min 1 dl 1 ; L-NMMA, ; % maximal FBF reduction; P 0.02 vs. group A). In contrast, in sht patients L-NMMA coinfusion only slightly attenuated the response to acetylcholine (saline, ml min 1 dl 1 ; L-NMMA, ; % maximal FBF reduction; P vs. acetylcholine alone) (Fig. 3). Subgroup analysis. In the sht group, 11 of 14 patients were characterized by TSH levels lower than 10 IU/ml. In these individuals, the vasodilating effect to acetylcholine (from to a maximum of ml min 1 dl 1, FIG. 2. Scatterplots describe the relationship between TSH plasma values and the maximal degree of vasodilation (Max VD) to acetylcholine (Ach) in sht patients. Vasodilation to acetylcholine is expressed as percentage increase above basal. FIG. 1. Line graphs show FBF increase above basal induced by acetylcholine and sodium nitroprusside in sht patients (triangles), group B (filled circles), and group A (empty circles) control subjects. Data are shown as means SE and expressed as percentage increase above basal. *, Significant difference between groups (P 0.05).

4 3734 J Clin Endocrinol Metab, August 2003, 88(8): Taddei et al. Endothelial Dysfunction in Hypothyroidism FIG. 3. Acetylcholine-induced increase in FBF under control conditions (empty circles) and in the presence of L-NMMA (filled circles) in group A and group B controls and sht patients. Data are shown as means SE and expressed as percentage increase above basal. *, Significant difference between infusion with and without L-NMMA (P 0.05). 30.2% increase above baseline) and the blunting effect of L-NMMA ( ml min 1 dl 1, % maximal FBF reduction; P vs. acetylcholine alone) still proved to be significantly reduced, compared with groups A (P ) and B (P ). Moreover, in the sht group, 9 out of 14 patients were characterized by LDL levels lower than 160 mg/dl (4 mm). In these individuals, the vasodilating effect to acetylcholine (from to a maximum of ml min 1 dl 1 ; % increase above baseline) and the blunting effect of L-NMMA ( ml min 1 dl 1 ; % maximal FBF reduction; P vs. acetylcholine alone) still proved to be significantly reduced, compared with groups A (P 0.002) and B (P 0.001). Finally, even excluding patients with elevated Lp(a), the results do not substantially change (data not shown). Effect of L-T 4 replacement therapy Serum TSH levels decreased significantly (P vs. baseline) and returned within the normal range; FT 3 and FT 4 values remained within the normal range throughout the entire treatment period, although FT 4 rose significantly (P 0.01 vs. baseline) (Table 2). Though remaining not significantly different from group B, sht patients showed a slight decrease in both total and LDL-cholesterol (P 0.04 and P 0.05, respectively), whereas ApoB did not change. Moreover, the Lp(a) value was unchanged both as a mean and in patients with elevated baseline values. A significant relationship between baseline serum TSH level and LDL-cholesterol reduction was detected (r 0.69, P 0.05). Interestingly, the inflammatory status was not modified, because ESR and autoantibody titer did not change (Table 2). TABLE 2. Clinical features in sht patients at baseline and after L-T 4 therapy Baseline (n 9) post L-T 4 (n 9) SBP (mm Hg) DBP (mm Hg) BMI (kg/m 2 ) ESR (mm) Fibrinogen (g/liter) Total cholesterol (mg/dl) a HDL cholesterol (mg/dl) LDL cholesterol (mg/dl) b Triglycerides (mg/dl) ApoB (g/liter) Lp(a) (g/liter) TSH ( IU/ml) c Free T 4 (pg/ml) d Free T 3 (pg/ml) TPO-Ab (U/ml) Tg-Ab (U/ml) a P 0.04 vs. baseline. b P 0.05 vs. baseline. c P vs. baseline. d P 0.01 vs. baseline. Data are expressed as mean SEM. Conversion factors: total, HDL, and LDL cholesterol: ; triglycerides: ; free T 3 : 1.536; free T 4 : Stable euthyroidism induced a significant improvement in vasodilation to acetylcholine (FBF, from to ml min 1 dl 1 at baseline vs. from to ml min 1 dl 1 at follow-up, P 0.05; ANOVA for repeated measures) (Fig. 4). Moreover, the inhibitory effect of L-NMMA on the response to acetylcholine was restored (saline, ml min 1 dl 1 ; L-NMMA, ; P vs. saline), so that the degree of inhibition exerted by the

5 Taddei et al. Endothelial Dysfunction in Hypothyroidism J Clin Endocrinol Metab, August 2003, 88(8): FIG. 4. Acetylcholine-induced increase in FBF under control conditions (empty circles) and in the presence of L-NMMA (filled circles) in sht patients at baseline and after L-T 4 replacement therapy. Data are shown as means SE and expressed as percentage increase above basal. *, Significant difference between infusion with and without L-NMMA (P 0.05). NO-synthase inhibitor was now similar to that obtained in group B subjects ( vs % reduction of maximal FBF response, respectively). The maximal response to acetylcholine was positively associated with both baseline serum TSH levels (r 0.82, P 0.01) and treatment-induced changes in TSH (r 0.80, P 0.01). A significant correlation between maximal response to acetylcholine and change in LDL-cholesterol (r 0.84, P 0.01) was also observed. Subgroup analysis. In six of nine treated patients characterized by pretreatment TSH levels lower than 10 IU/ml, after l-t 4 administration, the vasodilating effect to acetylcholine increased slightly and nonsignificantly (baseline: from to a maximum of ml min 1 dl 1 ; % increase above baseline; after l-t 4 : from to a maximum of ml min 1 dl 1 ; % increase above baseline). Moreover, the blunting effect of L-NMMA ( ml min 1 dl 1 ; % increase above baseline) was found to be potentiated, although nonstatistically significant, compared with baseline ( ml min 1 dl 1, % increase above baseline; % vs % reduction of maximal FBF response, respectively; P 0.09). In six of nine treated patients characterized by pretreatment LDL levels lower than 160 mg/dl (4 nm), after l-t 4 administration, the vasodilating effect to acetylcholine increased slightly and nonsignificantly (baseline: from to a maximum of ml min 1 dl 1, % increase above baseline; after l-t 4 : from to a maximum of ml min 1 dl 1, % increase above baseline). However, the blunting effect of L-NMMA ( ml min 1 dl 1 ; % maximal FBF reduction) was found to be significantly potentiated, compared with baseline ( ml min 1 dl 1 ; % maximal FBF reduction; P 0.04). Discussion The present study was designed to evaluate whether patients with sht are characterized by reduced endotheliumdependent vasodilation and impaired NO availability. Because endothelial dysfunction has been reported in association with aging, postmenopause, hypertension, and smoking status (31), all our patients were young, nonobese, nonsmokers with normal blood pressure and rigorously matched to a control group of euthyroid subjects. Additionally, because sht patients typically show moderately increased serum total and LDL-cholesterol levels (9), which per se could alter endothelial function (32), the vascular parameters of sht patients were specifically compared with those of control group B with a matching lipoprotein profile. It must be observed, however, that four sht patients and only one control subject had elevated Lp(a) values. The present results indicate that the vasodilating effect of acetylcholine was significantly reduced in sht patients, compared with healthy subjects, an effect which was evident even in those patients with TSH lower than 10 IU/ml. Because the response to sodium nitroprusside (which specifically acts on smooth muscle cells) (28) and calculated MFVRs (an index of vascular structural alterations) (29) were similar in the two study groups, the present results indicate that sht is characterized by the presence of endothelial dysfunction. This alteration cannot be accounted for exclusively by the alteration in lipid profile, because the response to acetylcholine in group B control subjects was found to be impaired, compared with normocholesterolemic controls (group A), but significantly better than that of sht patients. Moreover, in sht patients with normal Lp(a) values or LDL cholesterol lower than 160 mg/dl (4 mm), vasodilation to acetylcholine proved to be impaired, compared with control groups. Thus, it is conceivable that sht per se causes an impairment of endothelium-dependent vasodilation. On the other hand, though a significant relationship between the maximal response to acetylcholine and serum LDL-cholesterol level was observed in both patients and controls, in sht patients a significant correlation between the maximal response to acetylcholine and baseline serum TSH value was also obtained.

6 3736 J Clin Endocrinol Metab, August 2003, 88(8): Taddei et al. Endothelial Dysfunction in Hypothyroidism In the euthyroid subjects, the vasodilator response to acetylcholine was blunted by L-NMMA, confirming earlier observations (21, 26). This finding indicates that the relaxing activity of this agonist must be predominantly mediated by activation of the l-arginine-no pathway. In contrast, in sht patients, both the entire population and the subgroup with TSH lower than 10 IU/ml, the vasodilation to acetylcholine was almost fully resistant to L-NMMA, indicating the presence of an alteration in the l-arginine-no pathway, leading to impaired NO availability. Again, this alteration is probably, in part, related to sht per se because, in group B subjects, the degree of inhibition exerted by L-NMMA on vasodilation to acetylcholine was reduced, compared with group A subjects, but greater than that observed in sht patients. Moreover, in sht patients with normal Lp(a) values or LDL cholesterol lower than 160 mg/dl (4 mm), the blunting effect of L-NMMA on the vasodilation to acetylcholine proved to be reduced. A causal relationship between the presence of sht and endothelial dysfunction is also supported by the effect of hormone replacement. Thus, after 6 months of stable euthyroidism, a significant improvement in acetylcholine-induced vasodilatation (along with an unchanged response to sodium nitroprusside) was observed. Moreover, the most relevant issue is probably the finding that the slight increased response to acetylcholine was associated with a restoration of the inhibitory activity of L-NMMA, indicating that l-t 4 treatment improves endothelium-dependent vasodilation by restoring NO availability. A tendency to improvement after l-t 4 treatment was also observed in those patients with TSH lower than 10 IU/ml. However, the low numerosity of this subgroup (n 6) does not allow a definite conclusion to be reached. It is important to note that, after treatment, the degree of inhibition exerted by L-NMMA on acetylcholinedependent vasodilation was similar to that observed in group B subjects. This beneficial effect of l-t 4 administration was associated with a reduction of serum LDL-cholesterol levels, which were shown to be not significantly different from the corresponding values of group B. On the other hand, it is of interest that l-t 4 treatment increased the L- NMMA-induced inhibition on vasodilation to acetylcholine, even in patients with LDL cholesterol plasma level lower than 160 mg/dl (4 mm). Finally, significant relationships between the maximal response to acetylcholine and both serum TSH and LDL-cholesterol reduction were found. However, it is important to consider that the effect of l-t 4 treatment would have been more appropriately addressed by a placebo-controlled design. In the present investigation, this issue was assessed as a secondary end-point, when several patients were already on l-t 4 administration We therefore decided to perform an open study. Although this represents a study limitation, the good reproducibility of the forearm technique and lack of time-dependent effects on endothelial responses confer substantial plausibility on the results obtained with l-t 4 treatment. Overall, these findings suggest the presence of an early, reversible impairment of endothelium-dependent vasodilatation in sht patients (possibly because of reduced NO availability), which is, only in part, explained by the mild dyslipidemia. This finding, together with the evidence of a relationship between maximal vasodilation to acetylcholine and serum TSH levels, supports the hypothesis that sht per se may conspire with dyslipidemia to promote endothelial dysfunction. It can be argued that, by definition, sht patients have thyroid hormone levels within the normal range. However, given the narrow limit of individual thyroid hormone level variations, about half of laboratory reference ranges, a test result within the laboratory reference limits is not necessarily normal for that subject (33). These results extend the previous evidence demonstrating the presence of impaired flow-mediated endotheliumdependent vasodilation in the conduit arteries of patients with sht (34) and provide new information on NO availability alteration and the effect of l-t 4 replacement therapy. In addition to a direct effect of thyroid hormones, other mechanisms could lead to endothelial dysfunction in sht patients. In this regard, all the patients had thyroiditis; and therefore, a possible concomitant effect of inflammation and autoimmunity should be not overlooked, because inflammatory states are associated with endothelial dysfunction (35). Although in the present study, the presence of increased ESR values in sht patients seems to confirm this hypothesis, the lack of measurements of more specific inflammation parameters, including interleukins and C-reactive protein, may represent a major methodological limitation restricting the substantiation of this likely interpretation. Furthermore, four sht patients showed increased Lp(a) levels, in contrast to only one subject in group B. Because elevated Lp(a) values are associated with endothelial dysfunction in human coronary circulation (36), this alteration could, in part, participate in the impaired endothelium-dependent vasodilation. It is therefore conceivable that these factors (direct effect of thyroid hormone, inflammatory status, and lipid profile) can jointly contribute to impairing endothelium-dependent vasodilation in sht patients. The finding that after 6 months of stable euthyroidism, inflammatory status did not change (as supported by the lack of ESR and autoantibody titer modifications) and lipid profile improved only slightly, with no change in Lp(a) values, is in keeping with only partial restoration of endothelium-dependent vasodilation. In conclusion, endothelial dysfunction may represent early development of cardiovascular damage in sht patients. Therefore, early l-t 4 replacement therapy may be advised not only to prevent progression to frank hypothyroidism but also to slow down atherogenesis. Acknowledgments Received January 8, Accepted April 13, Address all correspondence and requests for reprints to: Stefano Taddei, M.D., Department of Internal Medicine, University of Pisa, Via Roma, 67, Pisa, Italy. s.taddei@med.unipi.it. References 1. Surks MI, Ocampo E 1996 Subclinical thyroid disease. Am J Med 100: Canaris GJ, Manowitz NR, Mayor G, Ridgway EC 2000 The Colorado thyroid disease prevalence study. Arch Intern Med 160: Cooper DS, Halpern R, Wood LC, Levin AA, Ridgway EC 1984 l-thyroxine therapy in subclinical hypothyroidism. A double-blind, placebo-controlled trial. Ann Intern Med 101: Monzani F, Del Guerra P, Caraccio N, Pruneti CA, Pucci E, Luisi M, Banchieri

7 Taddei et al. Endothelial Dysfunction in Hypothyroidism J Clin Endocrinol Metab, August 2003, 88(8): L 1993 Subclinical hypothyroidism: neurobehavioural features and beneficial effect of l-thyroxine treatment. Clin Investig 71: Monzani F, Caraccio N, Del Guerra P, Casolaro A, Ferrannini E 1999 Neuromuscular symptoms and dysfunction in subclinical hypothyroid patients: beneficial effect of l-t 4 replacement therapy. Clin Endocrinol (Oxf) 51: Chu JW, Crapo LM 2001 Clinical perspective. The treatment of subclinical hypothyroidism is seldom necessary. J Clin Endocrinol Metab 86: Bastenie PA, Vanhaelst L, Bonnyns M, Neve P, Staquet M 1971 Preclinical hypothyroidism: a risk factor for coronary heart-disease. Lancet 1: Bastenie PA, Vanhaelst L, Glodstein J, Smets P 1977 Asymptomatic autoimmune thyroiditis and coronary heart-disease. Cross-sectional and prospective studies. Lancet 2: Caraccio N, Ferrannini E, Monzani F 2002 Lipoprotein profile in subclinical hypothyroidism: effect of levothyroxine replacement therapy. A randomized, placebo-controlled study. J Clin Endocrinol Metab 87: Meier C, Staub JJ, Roth CB, Guglielmetti M, Kunz M, Miserez AR, Drewe J, Huber P, Herzog R, Muller B 2001 TSH-controlled l-thyroxine therapy reduces cholesterol levels and clinical symptoms in subclinical hypothyroidism: a double blind, placebo-controlled trial (Basel Thyroid Study). J Clin Endocrinol Metab 86: Tzotzas T, Krassas GE, Konstantinidis T, Bougoulia M 2000 Changes in lipoprotein(a) levels in overt and subclinical hypothyroidism before and during treatment. Thyroid 10: Hak AE, Pols HAP, Visser TJ, Drexhage HA, Hofman A, Witteman JC 2000 Subclinical hypothyroidism is an independent risk factor for atherosclerosis and myocardial infarction in elderly women. The Rotterdam Study. Ann Intern Med 132: Mathews JD, Whittingham S, Mackay IR 1974 Autoimmune mechanism in human vascular disease. Lancet 2: Masaki H, Nishikawa M, Urakami M, Yoshimura M, Toyoda N, Mori Y, Yoshikawa N, Inada M ,3,5 -Triiodothyronine inhibits collagen-induced human platelet aggregation. J Clin Endocrinol Metab 75: Chadarevian R, Bruckert E, Ankri A, Beucler I, Giral P, Turpin G 1998 Relationship between thyroid hormones and d-dimer levels. Thromb Haemost 79: Luscher TF, Vanhoutte PM 1990 The endothelium: modulator of cardiovascular function. Boca Raton, FL: CRC Press; Furchgott RF, Zawadzky JV 1980 The obligatory role of endothelial cells in the relaxation of artherial smooth muscle by acetylcholine. Nature 288: Palmer RMJ, Ferrige AG, Moncada S 1987 Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 327: Palmer RMJ, Ashton DS, Moncada S 1988 Vascular endothelial cells synthesize nitric oxide from l-arginine. Nature 333: Bredt DS, Hwang PM, Glatt CE, Lowenstein C, Reed RR, Snyder SH 1991 Cloned and expressed nitric oxide synthase structurally resembles cytochrome P-450 reductase. Nature 351: Taddei S, Virdis A, Ghiadoni L, Magagna A, Salvetti A 1998 Vitamin C improves endothelium-dependent vasodilation by restoring nitric oxide activity in essential hypertension. Circulation 97: Suwaidi JA, Hamasaki S, Higano ST, Nishimura RA, Holmes Jr DR, Lerman A 2000 Long-term follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation 101: Schacetylcholineinger V, Britten MB, Zeiher AM 2000 Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation 101: Heitzer T, Schlinzig T, Krohn K Meinertz T, Munzel T 2001 Endothelial dysfunction, oxidative stress, and risk of cardiovascular events in patients with coronary artery disease. Circulation 104: Whitney RJ 1953 The measurement of volume changes in human limbs. J Physiol (Lond) 121: Vallance P, Collier J, Moncada S 1989 Effect of endothelium-derived nitric oxide on peripheral arteriolar tone in man. Lancet 2: Stroes ESG, Lüscher TF, de Groot FG, Koomans HA, Rabelink TJ 1997 Cyclosporin A increases nitric oxide activity in vivo. Hypertension 29: Schultz KD, Schultz K, Schultz G 1977 Sodium nitroprusside and other smooth muscle relaxants increase cyclic GMP levels in rat ductus deferens. Nature 265: Folkow B 1995 The structural factor in hypertension with special emphasis on the altered geometric design of the systemic arteries. In: Laragh JH, Brenner BM, eds. Hypertension: pathophysiology, diagnosis and treatment. New York: Raven Press; Pedrinelli R, Taddei S, Spessot M, Salvetti A 1987 Maximal postischemic forearm vasodilation in human hypertension: a reassessment of the method. J Hypertens Suppl 5:S431 S Anderson TJ 1999 Assessment and treatment of endothelial dysfunction in humans. J Am Coll Cardiol 34: Creager MA, Cooke JP, Mendelsohn ME, Gallagher SJ, Coleman SM, Loscalzo J, Dzau VJ 1990 Impaired vasodilation of forearm resistance vessel in hypercholesterolemic humans. J Clin Invest 86: Andersen S, Pedersen KM, Bruun NH, Laurberg P 2002 Narrow individual variations in serum T(4) and T(3) in normal subjects: a clue to the understanding of subclinical thyroid disease. J Clin Endocrinol Metab 87: Lekakis J, Papamichael C, Alevizaki M, Piperingos G, Marafelia P, Mantzos J, Stamatelopoulos S, Koutras DA 1997 Flow-mediated, endothelium-dependent vasodilation is impaired in subjects with hypothyroidism, borderline hypothyroidism, and high-normal serum thyrotropin (TSH) values. Thyroid 7: Fichtlscherer S, Rosenberger G, Walter DH, Breuer S, Dimmeler S, Zeiher AM 2000 Elevated C-reactive protein levels and impaired endothelial vasoreactivity in patients with coronary artery disease. Circulation 102: Kugiyama K, Doi H, Motoyama T, Motoyama T, Soejima H, Misumi K, Kawano H, Nakagawa O, Yoshimura M, Ogawa H, Matsumura T, Sugiyama S, Nakano T, Nakajima K, Yasue H 1998 Association of remnant lipoprotein levels with impairment of endothelium-dependent vasomotor function in human coronary arteries. Circulation 97: