Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The metabolic cost of lowering blood pressure with hydrochlorothiazide

The metabolic cost of lowering blood pressure with hydrochlorothiazide Background: The landmark Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial (ALLHAT) placed a new spotlight on thiazide diuretics as the first-line therapy for hypertension. This is concerning as thiazide- diuretics may contribute to comorbidities associated with the current epidemic of obesity. Previous randomized clinical trials have linked thiazide diuretic treatment to insulin resistance, metabolic syndrome, and increased incidence of type 2 diabetes. Methods: This proof of concept, longitudinal, randomized, double–blind study evaluated the effects of the angiotensin II receptor blocker Valsartan and the specific thiazide diuretic Hydrochlorothiazide (HCTZ) on hepatic triglyceride level (primary outcome), as well as triglyceride levels within other organs including the heart, skeletal muscle, and pancreas. Additionally, we evaluated whether myocardial function, insulin sensitivity, and insulin secretion were affected by these treatments. Results: Hepatic TG levels increased by 57% post HCTZ treatment: ΔhTG = 4.12% and remained unchanged HCTZ post Valsartan treatment: ΔhTG = 0.06%. The elevation of hepatic TG levels after HCTZ treatment was additionally accompanied by a reduction in insulin sensitivity: ΔSI = −1.14. Treatment with Valsartan resulted in improved HCTZ insulin sensitivity: ΔSI = 1.24. Treatment-induced changes in hepatic TG levels and insulin sensitivity were statistically significant between groups (p = 0.0098 and p = 0.0345 respectively). Disposition index, DI, remained hTG SI unchanged after HCTZ treatment: ΔDI = −141 but it was increased by a factor of 2 after treatment with HCTZ Valsartan: ΔDI =1018). However, the change between groups was not statistically significant. Both therapies did not modify abdominal visceral and subcutaneous fat mass as well as myocardial structure and function. Additionally, myocardial, pancreatic, and skeletal muscle triglyceride deposits remained unchanged in both therapeutic arms. Conclusions: Our findings are two-fold and relate to hepatic steatosis and insulin sensitivity. HCTZ treatment worsened hepatic steatosis measured as hepatic triglyceride content and reduced insulin sensitivity. Valsartan treatment did not affect hepatic triglyceride levels and improved insulin sensitivity. The results of this study reinforce the message that in patients at risk for type 2 diabetes it is particularly important to choose an antihypertensive regimen that lowers blood pressure without exacerbating patient’s metabolic profile. Keywords: Type 2 diabetes, Valsartan, Hydrochlorothiazide, Proton magnetic resonance spectroscopy, Insulin sensitivity, Insulin secretion * Correspondence: [email protected] Equal contributors Cedars -Sinai Medical Center, The Heart Institute, Los Angeles, California, USA Full list of author information is available at the end of the article © 2013 Price et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 2 of 7 http://www.dmsjournal.com/content/5/1/35 The incidence of obesity and obesity-related complica- insulin sensitivity, and insulin secretion were affected by tions such as hypertension and type 2 diabetes are rising these treatments. steadily despite the increased public and scientific awareness of this multifactorial problem. Although spe- Study subjects cific efforts to turn the obesity tide concentrate on the Eighty-two individuals were screened for eligibility to development of new treatment strategies, it is important participate in the study. Qualifying individuals were to revisit old therapies and review their side effect pro- young adults (age range 18–55 years)with 3 of the fol- files as some treatments may silently augment the meta- lowing 5 conditions: fasting glucose > 100 mg/dl; waist bolic syndrome. circumference: men > 102 cm, women >88 cm; HDL: The landmark Antihypertensive and Lipid-Lowering men < 40 mg/dl, women <50 mg/dl; TG > 150 mg/dl; BP treatment to prevent Heart Attack Trial (ALLHAT) placed > 130/85 mm Hg. Individuals with a previous diagnosis a new spotlight on thiazide diuretics as the first-line ther- of type 2 diabetes, stage 2 hypertension (BP > 160/ apy for hypertension [1].This is concerning as thiazide- 110 mm Hg), or those exposed to thiazolidinediones, diuretics may contribute to comorbidities associated with statins, diuretics, ARB, ACEI, or any investigational the current epidemic of obesity. Previous randomized clin- agents within 6 months prior to the study did not qual- ical trials have linked treatment with thiazide diuretic to ify. Claustrophobia and presence of metallic implants in insulin resistance, metabolic syndrome, and increased in- the body were magnetic resonance imaging (MRI) exclu- cidence of type 2 diabetes [2,3]. sion criteria. Additionally, women of child bearing age On the contrary, evidence accumulates that therapies who were not using reliable contraception or those which interfere with the adverse metabolic effects of breast feeding did not qualify. angiotensin II, such as angiotensin II receptor blocking Twenty-six individuals who qualified and agreed to (ARB) or/and angiotensin converting enzyme (ACE I) participate were randomized in blocks of 4 to either therapies, cause no metabolic harm as confirmed by the therapy (12 to HCTZ and 14 to Valsartan). Eight indi- DREAM [4] and NAVIGATOR [5-7] studies. The favor- viduals (5 in HCTZ and 3 in Valsartan) did not complete able metabolic action of ARB and ACE-I agents could the protocol for various personal reasons. Results from 4 originate from improvement of insulin sensitivity [8] or individuals who completed therapy (1 in HCTZ and 3 in could be facilitated through the recruitment and differ- Valsartan) were excluded from analysis due to either sig- entiation of adipocytes [9]. Both mechanisms could lead nificant lifestyle changes (N = 1) that resulted in a large to reduction in ectopic deposition of triglyceride in or- amount of weight loss or because individuals were not gans such as liver, heart, pancreas and skeletal muscle, a available for the end-of-study evaluation (N = 3). Four- hypothesis that has not yet been tested. teen individuals (6 in HCTZ and 8 in Valsartan) com- We present the results of a randomized study compar- pleted all study procedures and were considered in final ing the metabolic effects of treatment with hydrochloro- analysis (Figure 1). thiazide (HCTZ) and Valsartan in individuals at high risk for development of type 2 diabetes. We specifically Experimental protocol evaluated the effect of these treatments on intra-hepatic Qualified participants were randomized to once-daily triglyceride content as well as insulin sensitivity, beta- 320 mg Valsartan or 25 mg HCTZ therapy for 8 months, cell function, and ectopic triglyceride deposition in the with both agents started at half dose and increased to heart, pancreas, and skeletal muscle. full dose after the first month. Notably, one study sub- ject did not tolerate full dose Valsartan due to relative Methods hypotension and was; therefore, continued on half dose This proof of concept, longitudinal, randomized, double– (Valsartan 160 mg) for the entire study duration. All blind study evaluated two antihypertensive treatments in study measurements and procedures were performed at individuals at high risk for diabetes. The study was regis- baseline, 1–7 days prior to randomization and repeated tered as clinical trial # NCT00745953. The research proto- at the end of the 8-month treatment period. col was approved by Institutional Review Board at UT Southwestern Medical Center. All participants gave in- Procedures formed written consent prior to experiments. Oral glucose tolerance test (OGTT) Our objective was to compare the effects of the angioten- A standard 75 g oral glucose tolerance test was adminis- sin II receptor blocker Valsartan and the thiazide diuretic tered only at baseline to evaluate each study subject’s Hydrochlorothiazide (HCTZ) on hepatic triglyceride level glycemic status according to American Diabetes Associ- (primary outcome), as well as triglyceride levels within other ation criteria [10], and to screen for the presence of un- organs including the heart, skeletal muscle, and pancreas. diagnosed diabetes. The test was performed at 8:30 AM Additionally, we evaluated whether myocardial function, after an overnight fast (10–12 hrs) and within 10 days of Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 3 of 7 http://www.dmsjournal.com/content/5/1/35 Figure 1 Study consort diagram. all other baseline measurements. Blood was sampled at pancreatic, myocardial and skeletal muscle TG content baseline: time ‘0’ and following the standard glucose using a1.5 Tesla Gyroscan Achieva whole body clinical drink at 15, 30, 60, 120, and 180 minutes. system (Philips Medical Systems, Cleveland, USA) equipped with software for localized spectroscopy as de- Frequently sampled intravenous glucose tolerance test scribed before [16,17].In short, high-resolution morpho- (FSIVGTT) logical images were collected to serve as a “roadmap” for The protocol was initiated at 8:30 AM after an overnight selection of a testing volume of 27 cc within the upper fast (10–12 hours). Two intravenous (antecubical vein) right hepatic lobe, 2 cc within pancreatic tail, 6 cc in polyethylene catheters were inserted, one for infusions of myocardial septum, and 1 cc within the skeletal muscle. glucose and regular human insulin and another for blood All spectra were collected using PRESS sequence (Point sampling. A bolus of 50% dextrose solution (0.3 g/kg body RESolved Spectroscopy) for spatial localization and the weight) was injected at time 0 and a bolus of regular hu- signal acquisition with the following data acquisition pa- man insulin (0.03 U/kg body weight) was injected at rameters: T =27 ms, T = 3 s. All data were collected e r 20 min. Blood samples were collected for determination without water suppression. Sixteen acquisitions were av- of plasma glucose and insulin levels at: -15, -10, -5, -1, 2, eraged for liver, pancreas, and skeletal muscle, and 32 3, 4, 5, 6, 8, 10, 14, 19, 22, 25, 30, 40, 50, 70, 100, 140, and for heart. Areas of resonances from protons in water 180 minutes. Data were analyzed using the Millennium molecules and in methylenes of fatty acid chains were Minimal Model (MINMOD) [11]. We report the Acute evaluated with line-fit procedure using a commercial Insulin Response to glucose (AIR ) – a measure of glucose software (NUTS-ACORNNMR, Freemont, CA) [14-17]. stimulated insulin secretion, Insulin Sensitivity (SI), and Disposition Index (DI) – a measure of insulin secretion adjusted for the prevailing insulin sensitivity which pre- Cardiac imaging dicts progression to type 2 diabetes [11]. Dynamic cine images were used to quantify left ventricu- lar (LV) volume [16,18,19]. Image analysis was performed Proton magnetic resonance spectroscopy (MRS) by an observer blinded to the subject’s clinical history and To study the role of steatosis in the clinical setting, we treatment, using a commercially available workstation and others have developed non-invasive, in vivo tech- (MASS, Philips Medical Systems). Endocardial and epicar- nique that permits the precise and reproducible quantifi- dial LV borders were traced manually at end diastole and cation of intracellular triglyceride in various human end systole from short-axis slices, and the papillary mus- organs, including skeletal muscle [12-15], liver [16,17], cles were excluded from the LV cavity volume. LV mass myocardium [16,18-21], and pancreas [22]. This method was computed as the product of end-diastolic LV volume offers a technological advantage as it distinguishes the and myocardial density (1.05 g/mL). The fraction of blood large compartments of triglyceride in adipose tissue cells pumped out of the left ventricle with each heart beat, the from the triglyceride droplets that are stored within the ejection fraction (EF), was calculated as the difference be- cytosol of parenchymal cells. This method is now widely tween left ventricular end diastolic volume and left ven- accepted and has become extremely useful in obesity tricular end systolic volume divided by left ventricular end and diabetes clinical studies as these evaluations are fast, diastolic volume. EF was used as an index of global LV safe, and reliable. In this study, we evaluated hepatic, function. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 4 of 7 http://www.dmsjournal.com/content/5/1/35 Abdominal MRI Statistics The amount of subcutaneous and visceral abdominal fat Responses to therapies, measured as a difference be- was determined from a single abdominal axial image at tween baseline and end of the study, were compared be- the level between vertebral bodies L2 and L3 [23]. The tween the groups. The tests for normality (Shapiro-Wilk, image analysis was performed by a single observer who chi-square and Kolmogorov-Smirnov) showed that the was blinded to the volunteer’s treatment, using commer- hepatic triglyceride content (hTG) response to HCTZ cially available software (MASS, Philips Medical Sys- and disposition index - DI response to Valsartan were tems) that maps the subcutaneous and intra-abdominal not normally distributed with 95% of confidence based adipose tissue compartments. on the results of at least one of the listed tests. All other responses, including SI response to both treatments were normally distributed with 95% confidence. There- Laboratory measurements fore we used two sample t-test for comparison of the All blood was processed immediately and was analyzed central tendency for SI between the two groups. The F- within 7 days. Lipid profile, liver function tests, glucose, test was used to compare the variability of the responses and insulin were analyzed in a commercial laboratory, between the two groups. We used non-parametric Quest Diagnostics, Irving, TX. HbA1c was analyzed by Wilcoxon-Mann–Whitney test to compare hTG and DI HPLC at UT Southwestern Medical Center. responses. Data were analyzed with Statgraphics Centur- ion XVI software. Statistical significance was set at p < 0.05. Clinical measurements Blood pressure was measured with a Space Labs con- Results tinuous home monitor for at least a 24 hr period. The The characteristics of study participants and the main average of all results obtained during this monitoring study results are listed in Table 1 and Table 2 respect- period is reported. Waist circumference was measured ively. Demographic, clinical, and biochemical character- at the level of the umbilicus in neutral respiratory pos- istics of both groups were similar at baseline and did not ition, using the same standard tape for all measurements change following either treatment. However, hepatic TG during the entire study. Hip circumference was mea- levels, a measure of hepatic steatosis, increased by 57% sured at the widest part of the hips. after HCTZ (baseline average hTG = 7.18% +/− 3.30%, Table 1 Characteristics of study participants (mean ± standard error) HCTZ Valsartan Variable Baseline End Baseline End N 69 Male,% 71 22 Age, years 35 ±11 37 ±5 Hip circumference, cm 110 ±6 115 ±9 116 ±10 116 ±10 Waist circumference, cm 101 ±12 104 ±16 107 ±9 107 ±11 BMI, kg/m 30.7 ± 2.4 31.6 ± 2.7 34.7 ± 3.4 34.9 ± 3.9 SBP, mmHg 120 ±11 119 ±9 114 ±8 108 ±9 DBP, mmHg 72 ±8 71 ±5 71 ±7 67 ±5 HR, beat/min 74 ±9 77 ±8 80 ±7 81 ±7 Glucose, mg/dL 95 ±11 98 ±11 100 ±10 94 ±11 Insulin, /mL 7 ±6 8 ±4 7 ±5 9 ±7 Cholesterol, mg/dL 209 ±17 217 ±34 194 ±40 192 ±30 Triglycerides, mg/dL 182 ± 118 186 ±52 153 ± 106 151 ±83 HDL, mg/dL 40 ±8 41 ±6 47 ±14 49 ±20 LDL, mg/dL 142 ±17 139 ±32 119 ±31 113 ±27 HbA1c,% 5.4 ± 0.1 55 ± 0.3 5.4 ± 0.5 55 ± 0.3 ALT, u/L 31 ±8 31 ±12 27 ±25 22 ±15 AST, u/L 29 ±8 26 ±6 21 ±11 19 ±6 Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 5 of 7 http://www.dmsjournal.com/content/5/1/35 Table 2 Hepatic triglyceride content (hTG), insulin secretion (SI) glucose stimulated insulin response (AIR ) and disposition index (DI) following the eight months treatment with either hydrochlorothiazide (HCTZ) or Valsartan (mean ± standard error) HCTZ Valsartan Variable Baseline End Baseline End hTG, f/w, % 7.18 ± 3.30 11.53 ± 4.56 8.29 ± 3.90 8.27 ± 3.38 -5 SI, (x10 /min pmol/l) 3.72 ± 0.98 2.57 ± 0.02 2.27 ± 0.68 3.51 ± 0.8 AIR , pmol/l 445 ± 291 466 ± 263 701 ± 608 708 ± 405 -5 DI, (x10 /min) 1332 ± 399 1191 ± 312 1016 ± 157 2034 ± 566 Of note: DI, SI*AIRg and describes the progression to type 2 diabetes when lowered or improvement of metabolic status when raised [11]. hTG range 0.59% - 21.97%; post HCTZ average hTG = induced changes in hepatic TG levels and insulin sensi- 11.30% +/− 4.56%, hTG range 3.13% - 32.37%; and tivity were statistically significant between groups (p hTG ΔhTG = 4.12%). Hepatic TG levels were unchanged = 0.0098and p = 0.0345 respectively). The individual re- HCTZ SI after Valsartan therapy (baseline average hTG = 8.21% sults as well as the changes in hTG and SI are shown in +/− 3.90%, hTG range 1.62% - 34.92%; post Valsartan Figures 2 and 3. average hTG = 8.27% +/− 3.38%, hTG range 2.83% - DI remained unchanged after HCTZ treatment (base- 31.49%, and ΔhTG = 0.06%). The increased inhepatic line average DI = 1332 +/− 399, DI range 471–3255; post TG levels in the HCTZ group were accompanied by a HCTZ average DI = 1191 +/− 312, DI range 372–2339; reduction in insulin sensitivity (baseline average SI = and ΔDI = −141) but DI increased by a factor of 2 HCTZ 3.71 +/− 0.98, SI range 1.17 - 7.78; post HCTZ average after treatment with Valsartan (baseline average DI = SI = 2.57 +/− 0.20, SI range 1.66-3.04; and ΔSI = 1016 +/− 157, DI range 473–1819; post Valsartan aver- HCTZ −1.14). Treatment with Valsartan resulted in improved age DI 2034 +/− 556, DI range 737 – 5739; and ΔDI insulin sensitivity (baseline average SI = 2.27 +/− 0.68, SI =1018) – primarily due to improvement in SI. However, range 0.46- 6.67; post Valsartan average SI 3.51 +/− 0.80, there change between treatment groups was not signifi- SI range 1.28-7.42; and ΔSI = 1.24). Treatment- cant. Abdominal (visceral) and subcutaneous fat mass as Figure 2 Impact of hydrochlorothiazide (HCTZ) and Valsartan treatments on hepatic triglyceride levels (hTG) and insulin sensitivity (SI). Results of hTG and SI are color coded relative to patient. Black points represent the averages. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 6 of 7 http://www.dmsjournal.com/content/5/1/35 accepted as first-line agent for treating hypertension - should be re-evaluated. HCTZ offer affordable and effi- cient blood pressure lowering but come with the added cost of worsening metabolic profiles for diabetes. This can occur even at doses of 25 mg and is sustained over a lon- ger treatment period (8 months). By the same means, the ARB agents improve insulin sensitivity and did not cause fatty liver. Primum non nocere - our treatment choices should not only improve the primary condition for which they are prescribed, but we must ensure that our patients suf- fer no harm. The ultimate goal of any antihypertensive therapy is to prevent cardiovascular events. The use of an antihypertensive agent that worsens hepatic steatosis and insulin resistance, both of which promote cardiovas- cular disease, negates the ultimate cardiovascular- preventive goal of the treatment. Hypertension clusters with metabolic syndrome, diabetes, and hepatic steatosis, and requires life-long pharmacologic treatment. When all aspects are balanced i.e. the blood pressure lowering effect and the worsening metabolic profile the use of HCTZ as a first line therapeutic choice should be questioned. Figure 3 (a) Changes in hepatic triglyceride content (hTG) after The underlying mechanism of the deleterious meta- treatment with Valsartan and HCTZ ( p = 0.0098). (b) Changes in bolic action of HCTZ is still debated although our re- insulin sensitivity (SI) after treatment with HCTZ and Valsartan (p = −0.0345). sults implicate concomitant worsening of fatty liver and decreased insulin sensitivity. Chronic exposure to angio- tensin II may render fat cells less efficient in their cap- well as myocardial structure and function remained un- acity to adequately store excess triglyceride, resulting in changed in both therapeutic arms. Additionally, we did tissue overflow with ectopic triglyceride and ultimately not detect significant changes in myocardial, pancreatic, hepatic steatosis. Interestingly, blocking the renin- and skeletal muscle triglyceride deposits with these angiotensin system by Valsartan did not result in fatty interventions. liver. The favorable metabolic action of Valsartan is probably complex and could originate from improve- Discussion ment of insulin sensitivity [8] or it could be facilitated We have demonstrated a differential metabolic effect of through the recruitment and differentiation of adipo- two frequently prescribed antihypertensive agents in in- cytes [25]. In light of the above-mentioned observations dividuals at risk for type 2 diabetes. The findings of our it seems natural to hypothesize that pairing ARB with study relate to hepatic steatosis and insulin sensitivity. HCTZ could block angiotensin II and mitigate the ad- HCTZ treatment worsened hepatic steatosis measured verse metabolic effects induced by HCTZ. Regrettably, as hepatic TG content and reduced insulin sensitivity. an attempt to block these unfavorable metabolic effects Valsartan treatment did not affect hepatic TG levels and of HCTZ by combining it with losartan was not success- improved insulin sensitivity. The results of this study ful [26]. We acknowledge that this study is small and these re- reinforce the message that in patients at risk for type 2 diabetes it is particularly important to choose an antihy- sults should be replicated, yet the effect size was consid- pertensive regimen that lowers blood pressure without erable and very comparable to that found in a similar independently conducted study [24]. We also note that exacerbating patient’s metabolic profile. Our results parallel the earlier reported findings from small sample size did not allow us to further explore the the MEDICA study [24]. MEDICA investigators reported mechanisms that contribute to our observations. similar worsening in hepatic triglyceride content and insu- lin sensitivity following 3 months of treatment with 50 mg of HCTZ and observed no change of hepatic triglyceride Conclusions levels following treatment with another ARB class medica- We have documented that HCTZ therapy leads to the tion Candesartan. In light of such tight agreement of development of hepatic steatosis and compromised insu- two independent studies, therapy with HCTZ - widely lin sensitivity in subjects at high risk for type 2 diabetes. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 7 of 7 http://www.dmsjournal.com/content/5/1/35 Conversely, treatment with Valsartan did not cause ec- 7. Karnes JH, Cooper-DeHoff RM: Antihypertensive medications: benefits of blood pressure lowering and hazards of metabolic effects. Expert Rev topic fat redistribution, and in fact, leads to improved in- Cardiovasc Ther 2009, 7:689–702. sulin sensitivity. Further studies are needed to determine 8. Kurtz TW, Pravenec M: Antidiabetic mechanisms of angiotensin- the exact mechanism by which HCTZ exerts its deleteri- converting enzyme inhibitors and angiotensin II receptor antagonists: beyond the renin-angiotensin system. J Hypertens 2004, 22:2253–2261. ous effects and whether these changes in metabolic 9. Sharma AM, Janke J, Gorzelniak K, Engeli S, Luft FC: Angiotensin Blockade parameters will translate into an increase in major long- Prevents Type 2 Diabetes by Formation of Fat Cells. Hypertension 2002, term adverse cardiovascular events. While the clinical 40:609–611. 10. Nathan DM, Davidson MB, DeFronzo RA, Heine RJ, Henry RR, Pratley R, relevance of thiazide-induced metabolic derangements Zinman B: Impaired fasting glucose and impaired glucose tolerance: remains uncertain at this time, hypertensive individuals implications for care. Diabetes Care 2007, 30:753–759. at risk for diabetes and those with known hepatic 11. Bergman RN: Lilly Lecture 1989. Toward Physiological Understanding of Glucose Tolerance. Minimal-Model Approach. Diabetes 1989, steatosis should opt for antihypertensive agents that are 38:1512–1527. metabolically benign – i.e. ARB or ACE inhibitors – 12. Boesch C, Slotboom J, Hoppeler H, Kreis R: In vivo determination of intra- until this issue is clarified. myocellular lipids in human muscle by means of localized 1H-MR -spectroscopy. Magn Reson Med 1997, 37:484–493. 13. Schick F, Eismann B, Jung WI, Bongers H, Bunse M, Lutz O: Comparison of Competing interests localized proton NMR signals of skeletal muscle and fat tissue in vivo: This study was funded as the investigator initiated research award to LSS two lipid compartments in muscle tissue. Magn Reson Med 1993, and RGB by Novartis. 29:158–167. 14. Szczepaniak LS, Babcock EE, Schick F, Dobbins RL, Garg A, Burns DK, Authors’ contributions McGarry JD, Stein DT: Measurement of intracellular triglyceride stores by ALP and IL designed and carried OGTT and FSIVGTT experiments. H spectroscopy: validation in vivo. Am J Physiol 1999, 276:E977–E989. Participated in results interpretation and writing the manuscript. EWS 15. Szczepaniak LS, Dobbins RL, Stein DT, McGarry JD: Bulk magnetic processed MR imaging and spectroscopy data, participated in data susceptibility effects on the assessment of intra- and extramyocellular interpretation, and assisted with statistical analysis.JW assisted in GCRC lipids in vivo. Magn Reson Med 2002, 47:607–610. experiments, created and maintained data base for the study, processed 16. Szczepaniak LS, Dobbins RL, Metzger GJ, Sartoni-D’Ambrosia G, Arbique D, FSIVGTT data using MinMode software, contributed to discussion on results Vongpatanasin W, Unger R, Victor RG: Myocardial triglycerides and systolic interpretation.RGV participated in study design and results interpretation.LSS function in humans: in vivo evaluation by localized proton spectroscopy designed the study, overlooked all study procedures, executed MR and cardiac imaging. Magn Reson Med 2003, 49:417–423. Spectroscopy and wrote the paper. All authors read and approved the final 17. Szczepaniak LS, Nurenberg P, Leonard D, Browning JD, Reingold JS, Grundy manuscript. S, Hobbs HH, Dobbins RL: Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the Funding sources general population. Am J Physiol Endocrinol Metab 2005, 288:E462–E468. This study was supported in part by Novartis Investigator Initiated Award, 18. McGavock JM, Victor RG, Unger RH, Szczepaniak LS: Adiposity of the heart, NIH UL1RR024982, NIH K23 RR024470, R01 NIDDK081524. revisited. Ann Intern Med 2006, 144:517–524. 19. McGavock JM, Lingvay I, Zib I, Tillery T, Salas N, Unger R, Levine BD, Raskin Author details P, Victor RG, Szczepaniak LS: Cardiac steatosis in diabetes mellitus: a 1H- Department of Internal Medicine, UT Southwestern Medical Center, Dallas, magnetic resonance spectroscopy study. Circulation 2007, Texas, USA. Cedars -Sinai Medical Center, The Heart Institute, Los Angeles, 116:1170–1175. California, USA. 20. Reingold JS, McGavock JM, Kaka S, Tillery T, Victor RG, Szczepaniak LS: Determination of triglyceride in the human myocardium by magnetic Received: 9 July 2013 Accepted: 5 July 2013 resonance spectroscopy: reproducibility and sensitivity of the method. Published: 9 July 2013 Am J Physiol Endocrinol Metab 2005, 289:E935–E939. 21. Lingvay I, Raskin P, Szczepaniak LS: The fatty hearts of patients with References diabetes. Nat Rev Cardiol 2009, 6:268–269. 1. Barzilay JI, Davis BR, Cutler JA, Pressel SL, Whelton PK, Basile J, Margolis KL, 22. Lingvay I, Esser V, Legendre JL, Price AL, Wertz KM, Adams-Huet B, Zhang S, Ong ST, Sadler LS, Summerson J: Fasting glucose levels and incident Unger RH, Szczepaniak LS: Noninvasive quantification of pancreatic fat in diabetes mellitus in older nondiabetic adults randomized to receive 3 humans. J Clin Endocrinol Metab 2009, 94:4070–4076. different classes of antihypertensive treatment: a report from the 23. Abate N, Garg A, Peshock RM, Stray-Gundersen J, Grundy SM: Relationships Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack of generalized and regional adiposity to insulin sensitivity in men. Trial (ALLHAT). Arch Intern Med 2006, 166:2191–2201. J Clin Invest 1995, 96:88–98. 2. Stump CS, Hamilton MT, Sowers JR: Effect of antihypertensive agents on 24. Eriksson JW, Jansson PA, Carlberg B, Hagg A, Kurland L, Svensson MK, the development of type 2 diabetes mellitus. Mayo Clin Proc 2006, Ahlstrom H, Strom C, Lonn L, Ojbrandt K, Johansson L, Lind L: 81:796–806. Hydrochlorothiazide, but not candesartan, aggravates insulin resistance 3. Elliott WJ, Meyer PM: Incident diabetes in clinical trials of and causes visceral and hepatic fat accumulation: the Mechanisms for antihypertensive drugs: a network meta-analysis. Lancet 2007, the Diabetes Preventing Effect of Candesartan (MEDICA) Study. 369:201–207. Hypertension 2008, 52:1030–1037. 4. Gerstein HC, Yusuf S, Bosch J, Pogue J, Sheridan P, Dinccag N, Hanefeld M, 25. Sharma AM: Does it matter how blood pressure is lowered in patients Hoogwerf B, Laakso M, Mohan V, Shaw J, Zinman B, Holman RR: Effect of with metabolic risk factors? J Am Soc Hypertens 2008(4):S23–S29. rosiglitazone on the frequency of diabetes in patients with impaired 26. Bakris G, Molitch M, Hewkin A, Kipnes M, Sarafidis P, Fakouhi K, Bacher P, glucose tolerance or impaired fasting glucose: a randomised controlled Sowers J: Differences in glucose tolerance between fixed-dose trial. Lancet 2006, 368:1096–1105. antihypertensive drug combinations in people with metabolic 5. Basile JN: Antihypertensive therapy, new-onset diabetes, and syndrome. Diab Care 2006, 29:2592–2597. cardiovascular disease. Int J Clin Pract 2009, 63:656–666. 6. Califf RM, Boolell M, Haffner SM, Bethel MA, McMurray J, Duggal A, Holman doi:10.1186/1758-5996-5-35 RR: Prevention of diabetes and cardiovascular disease in patients with Cite this article as: Price et al.: The metabolic cost of lowering blood impaired glucose tolerance: rationale and design of the Nateglinide and pressure with hydrochlorothiazide. Diabetology & Metabolic Syndrome Valsartan in Impaired Glucose Tolerance Outcomes Research 2013 5:35. (NAVIGATOR) Trial. Am Heart J 2008, 156:623–632. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diabetology & Metabolic Syndrome Springer Journals

The metabolic cost of lowering blood pressure with hydrochlorothiazide

Download PDF
7 pages

Loading next page...
 
/lp/springer-journals/the-metabolic-cost-of-lowering-blood-pressure-with-hydrochlorothiazide-GsY1IVK9be

References (34)

Publisher
Springer Journals
Copyright
Copyright © 2013 by Price et al.; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Diabetes; Metabolic Diseases; Endocrinology
eISSN
1758-5996
DOI
10.1186/1758-5996-5-35
pmid
23837919
Publisher site
See Article on Publisher Site

Abstract

Background: The landmark Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial (ALLHAT) placed a new spotlight on thiazide diuretics as the first-line therapy for hypertension. This is concerning as thiazide- diuretics may contribute to comorbidities associated with the current epidemic of obesity. Previous randomized clinical trials have linked thiazide diuretic treatment to insulin resistance, metabolic syndrome, and increased incidence of type 2 diabetes. Methods: This proof of concept, longitudinal, randomized, double–blind study evaluated the effects of the angiotensin II receptor blocker Valsartan and the specific thiazide diuretic Hydrochlorothiazide (HCTZ) on hepatic triglyceride level (primary outcome), as well as triglyceride levels within other organs including the heart, skeletal muscle, and pancreas. Additionally, we evaluated whether myocardial function, insulin sensitivity, and insulin secretion were affected by these treatments. Results: Hepatic TG levels increased by 57% post HCTZ treatment: ΔhTG = 4.12% and remained unchanged HCTZ post Valsartan treatment: ΔhTG = 0.06%. The elevation of hepatic TG levels after HCTZ treatment was additionally accompanied by a reduction in insulin sensitivity: ΔSI = −1.14. Treatment with Valsartan resulted in improved HCTZ insulin sensitivity: ΔSI = 1.24. Treatment-induced changes in hepatic TG levels and insulin sensitivity were statistically significant between groups (p = 0.0098 and p = 0.0345 respectively). Disposition index, DI, remained hTG SI unchanged after HCTZ treatment: ΔDI = −141 but it was increased by a factor of 2 after treatment with HCTZ Valsartan: ΔDI =1018). However, the change between groups was not statistically significant. Both therapies did not modify abdominal visceral and subcutaneous fat mass as well as myocardial structure and function. Additionally, myocardial, pancreatic, and skeletal muscle triglyceride deposits remained unchanged in both therapeutic arms. Conclusions: Our findings are two-fold and relate to hepatic steatosis and insulin sensitivity. HCTZ treatment worsened hepatic steatosis measured as hepatic triglyceride content and reduced insulin sensitivity. Valsartan treatment did not affect hepatic triglyceride levels and improved insulin sensitivity. The results of this study reinforce the message that in patients at risk for type 2 diabetes it is particularly important to choose an antihypertensive regimen that lowers blood pressure without exacerbating patient’s metabolic profile. Keywords: Type 2 diabetes, Valsartan, Hydrochlorothiazide, Proton magnetic resonance spectroscopy, Insulin sensitivity, Insulin secretion * Correspondence: [email protected] Equal contributors Cedars -Sinai Medical Center, The Heart Institute, Los Angeles, California, USA Full list of author information is available at the end of the article © 2013 Price et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 2 of 7 http://www.dmsjournal.com/content/5/1/35 The incidence of obesity and obesity-related complica- insulin sensitivity, and insulin secretion were affected by tions such as hypertension and type 2 diabetes are rising these treatments. steadily despite the increased public and scientific awareness of this multifactorial problem. Although spe- Study subjects cific efforts to turn the obesity tide concentrate on the Eighty-two individuals were screened for eligibility to development of new treatment strategies, it is important participate in the study. Qualifying individuals were to revisit old therapies and review their side effect pro- young adults (age range 18–55 years)with 3 of the fol- files as some treatments may silently augment the meta- lowing 5 conditions: fasting glucose > 100 mg/dl; waist bolic syndrome. circumference: men > 102 cm, women >88 cm; HDL: The landmark Antihypertensive and Lipid-Lowering men < 40 mg/dl, women <50 mg/dl; TG > 150 mg/dl; BP treatment to prevent Heart Attack Trial (ALLHAT) placed > 130/85 mm Hg. Individuals with a previous diagnosis a new spotlight on thiazide diuretics as the first-line ther- of type 2 diabetes, stage 2 hypertension (BP > 160/ apy for hypertension [1].This is concerning as thiazide- 110 mm Hg), or those exposed to thiazolidinediones, diuretics may contribute to comorbidities associated with statins, diuretics, ARB, ACEI, or any investigational the current epidemic of obesity. Previous randomized clin- agents within 6 months prior to the study did not qual- ical trials have linked treatment with thiazide diuretic to ify. Claustrophobia and presence of metallic implants in insulin resistance, metabolic syndrome, and increased in- the body were magnetic resonance imaging (MRI) exclu- cidence of type 2 diabetes [2,3]. sion criteria. Additionally, women of child bearing age On the contrary, evidence accumulates that therapies who were not using reliable contraception or those which interfere with the adverse metabolic effects of breast feeding did not qualify. angiotensin II, such as angiotensin II receptor blocking Twenty-six individuals who qualified and agreed to (ARB) or/and angiotensin converting enzyme (ACE I) participate were randomized in blocks of 4 to either therapies, cause no metabolic harm as confirmed by the therapy (12 to HCTZ and 14 to Valsartan). Eight indi- DREAM [4] and NAVIGATOR [5-7] studies. The favor- viduals (5 in HCTZ and 3 in Valsartan) did not complete able metabolic action of ARB and ACE-I agents could the protocol for various personal reasons. Results from 4 originate from improvement of insulin sensitivity [8] or individuals who completed therapy (1 in HCTZ and 3 in could be facilitated through the recruitment and differ- Valsartan) were excluded from analysis due to either sig- entiation of adipocytes [9]. Both mechanisms could lead nificant lifestyle changes (N = 1) that resulted in a large to reduction in ectopic deposition of triglyceride in or- amount of weight loss or because individuals were not gans such as liver, heart, pancreas and skeletal muscle, a available for the end-of-study evaluation (N = 3). Four- hypothesis that has not yet been tested. teen individuals (6 in HCTZ and 8 in Valsartan) com- We present the results of a randomized study compar- pleted all study procedures and were considered in final ing the metabolic effects of treatment with hydrochloro- analysis (Figure 1). thiazide (HCTZ) and Valsartan in individuals at high risk for development of type 2 diabetes. We specifically Experimental protocol evaluated the effect of these treatments on intra-hepatic Qualified participants were randomized to once-daily triglyceride content as well as insulin sensitivity, beta- 320 mg Valsartan or 25 mg HCTZ therapy for 8 months, cell function, and ectopic triglyceride deposition in the with both agents started at half dose and increased to heart, pancreas, and skeletal muscle. full dose after the first month. Notably, one study sub- ject did not tolerate full dose Valsartan due to relative Methods hypotension and was; therefore, continued on half dose This proof of concept, longitudinal, randomized, double– (Valsartan 160 mg) for the entire study duration. All blind study evaluated two antihypertensive treatments in study measurements and procedures were performed at individuals at high risk for diabetes. The study was regis- baseline, 1–7 days prior to randomization and repeated tered as clinical trial # NCT00745953. The research proto- at the end of the 8-month treatment period. col was approved by Institutional Review Board at UT Southwestern Medical Center. All participants gave in- Procedures formed written consent prior to experiments. Oral glucose tolerance test (OGTT) Our objective was to compare the effects of the angioten- A standard 75 g oral glucose tolerance test was adminis- sin II receptor blocker Valsartan and the thiazide diuretic tered only at baseline to evaluate each study subject’s Hydrochlorothiazide (HCTZ) on hepatic triglyceride level glycemic status according to American Diabetes Associ- (primary outcome), as well as triglyceride levels within other ation criteria [10], and to screen for the presence of un- organs including the heart, skeletal muscle, and pancreas. diagnosed diabetes. The test was performed at 8:30 AM Additionally, we evaluated whether myocardial function, after an overnight fast (10–12 hrs) and within 10 days of Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 3 of 7 http://www.dmsjournal.com/content/5/1/35 Figure 1 Study consort diagram. all other baseline measurements. Blood was sampled at pancreatic, myocardial and skeletal muscle TG content baseline: time ‘0’ and following the standard glucose using a1.5 Tesla Gyroscan Achieva whole body clinical drink at 15, 30, 60, 120, and 180 minutes. system (Philips Medical Systems, Cleveland, USA) equipped with software for localized spectroscopy as de- Frequently sampled intravenous glucose tolerance test scribed before [16,17].In short, high-resolution morpho- (FSIVGTT) logical images were collected to serve as a “roadmap” for The protocol was initiated at 8:30 AM after an overnight selection of a testing volume of 27 cc within the upper fast (10–12 hours). Two intravenous (antecubical vein) right hepatic lobe, 2 cc within pancreatic tail, 6 cc in polyethylene catheters were inserted, one for infusions of myocardial septum, and 1 cc within the skeletal muscle. glucose and regular human insulin and another for blood All spectra were collected using PRESS sequence (Point sampling. A bolus of 50% dextrose solution (0.3 g/kg body RESolved Spectroscopy) for spatial localization and the weight) was injected at time 0 and a bolus of regular hu- signal acquisition with the following data acquisition pa- man insulin (0.03 U/kg body weight) was injected at rameters: T =27 ms, T = 3 s. All data were collected e r 20 min. Blood samples were collected for determination without water suppression. Sixteen acquisitions were av- of plasma glucose and insulin levels at: -15, -10, -5, -1, 2, eraged for liver, pancreas, and skeletal muscle, and 32 3, 4, 5, 6, 8, 10, 14, 19, 22, 25, 30, 40, 50, 70, 100, 140, and for heart. Areas of resonances from protons in water 180 minutes. Data were analyzed using the Millennium molecules and in methylenes of fatty acid chains were Minimal Model (MINMOD) [11]. We report the Acute evaluated with line-fit procedure using a commercial Insulin Response to glucose (AIR ) – a measure of glucose software (NUTS-ACORNNMR, Freemont, CA) [14-17]. stimulated insulin secretion, Insulin Sensitivity (SI), and Disposition Index (DI) – a measure of insulin secretion adjusted for the prevailing insulin sensitivity which pre- Cardiac imaging dicts progression to type 2 diabetes [11]. Dynamic cine images were used to quantify left ventricu- lar (LV) volume [16,18,19]. Image analysis was performed Proton magnetic resonance spectroscopy (MRS) by an observer blinded to the subject’s clinical history and To study the role of steatosis in the clinical setting, we treatment, using a commercially available workstation and others have developed non-invasive, in vivo tech- (MASS, Philips Medical Systems). Endocardial and epicar- nique that permits the precise and reproducible quantifi- dial LV borders were traced manually at end diastole and cation of intracellular triglyceride in various human end systole from short-axis slices, and the papillary mus- organs, including skeletal muscle [12-15], liver [16,17], cles were excluded from the LV cavity volume. LV mass myocardium [16,18-21], and pancreas [22]. This method was computed as the product of end-diastolic LV volume offers a technological advantage as it distinguishes the and myocardial density (1.05 g/mL). The fraction of blood large compartments of triglyceride in adipose tissue cells pumped out of the left ventricle with each heart beat, the from the triglyceride droplets that are stored within the ejection fraction (EF), was calculated as the difference be- cytosol of parenchymal cells. This method is now widely tween left ventricular end diastolic volume and left ven- accepted and has become extremely useful in obesity tricular end systolic volume divided by left ventricular end and diabetes clinical studies as these evaluations are fast, diastolic volume. EF was used as an index of global LV safe, and reliable. In this study, we evaluated hepatic, function. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 4 of 7 http://www.dmsjournal.com/content/5/1/35 Abdominal MRI Statistics The amount of subcutaneous and visceral abdominal fat Responses to therapies, measured as a difference be- was determined from a single abdominal axial image at tween baseline and end of the study, were compared be- the level between vertebral bodies L2 and L3 [23]. The tween the groups. The tests for normality (Shapiro-Wilk, image analysis was performed by a single observer who chi-square and Kolmogorov-Smirnov) showed that the was blinded to the volunteer’s treatment, using commer- hepatic triglyceride content (hTG) response to HCTZ cially available software (MASS, Philips Medical Sys- and disposition index - DI response to Valsartan were tems) that maps the subcutaneous and intra-abdominal not normally distributed with 95% of confidence based adipose tissue compartments. on the results of at least one of the listed tests. All other responses, including SI response to both treatments were normally distributed with 95% confidence. There- Laboratory measurements fore we used two sample t-test for comparison of the All blood was processed immediately and was analyzed central tendency for SI between the two groups. The F- within 7 days. Lipid profile, liver function tests, glucose, test was used to compare the variability of the responses and insulin were analyzed in a commercial laboratory, between the two groups. We used non-parametric Quest Diagnostics, Irving, TX. HbA1c was analyzed by Wilcoxon-Mann–Whitney test to compare hTG and DI HPLC at UT Southwestern Medical Center. responses. Data were analyzed with Statgraphics Centur- ion XVI software. Statistical significance was set at p < 0.05. Clinical measurements Blood pressure was measured with a Space Labs con- Results tinuous home monitor for at least a 24 hr period. The The characteristics of study participants and the main average of all results obtained during this monitoring study results are listed in Table 1 and Table 2 respect- period is reported. Waist circumference was measured ively. Demographic, clinical, and biochemical character- at the level of the umbilicus in neutral respiratory pos- istics of both groups were similar at baseline and did not ition, using the same standard tape for all measurements change following either treatment. However, hepatic TG during the entire study. Hip circumference was mea- levels, a measure of hepatic steatosis, increased by 57% sured at the widest part of the hips. after HCTZ (baseline average hTG = 7.18% +/− 3.30%, Table 1 Characteristics of study participants (mean ± standard error) HCTZ Valsartan Variable Baseline End Baseline End N 69 Male,% 71 22 Age, years 35 ±11 37 ±5 Hip circumference, cm 110 ±6 115 ±9 116 ±10 116 ±10 Waist circumference, cm 101 ±12 104 ±16 107 ±9 107 ±11 BMI, kg/m 30.7 ± 2.4 31.6 ± 2.7 34.7 ± 3.4 34.9 ± 3.9 SBP, mmHg 120 ±11 119 ±9 114 ±8 108 ±9 DBP, mmHg 72 ±8 71 ±5 71 ±7 67 ±5 HR, beat/min 74 ±9 77 ±8 80 ±7 81 ±7 Glucose, mg/dL 95 ±11 98 ±11 100 ±10 94 ±11 Insulin, /mL 7 ±6 8 ±4 7 ±5 9 ±7 Cholesterol, mg/dL 209 ±17 217 ±34 194 ±40 192 ±30 Triglycerides, mg/dL 182 ± 118 186 ±52 153 ± 106 151 ±83 HDL, mg/dL 40 ±8 41 ±6 47 ±14 49 ±20 LDL, mg/dL 142 ±17 139 ±32 119 ±31 113 ±27 HbA1c,% 5.4 ± 0.1 55 ± 0.3 5.4 ± 0.5 55 ± 0.3 ALT, u/L 31 ±8 31 ±12 27 ±25 22 ±15 AST, u/L 29 ±8 26 ±6 21 ±11 19 ±6 Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 5 of 7 http://www.dmsjournal.com/content/5/1/35 Table 2 Hepatic triglyceride content (hTG), insulin secretion (SI) glucose stimulated insulin response (AIR ) and disposition index (DI) following the eight months treatment with either hydrochlorothiazide (HCTZ) or Valsartan (mean ± standard error) HCTZ Valsartan Variable Baseline End Baseline End hTG, f/w, % 7.18 ± 3.30 11.53 ± 4.56 8.29 ± 3.90 8.27 ± 3.38 -5 SI, (x10 /min pmol/l) 3.72 ± 0.98 2.57 ± 0.02 2.27 ± 0.68 3.51 ± 0.8 AIR , pmol/l 445 ± 291 466 ± 263 701 ± 608 708 ± 405 -5 DI, (x10 /min) 1332 ± 399 1191 ± 312 1016 ± 157 2034 ± 566 Of note: DI, SI*AIRg and describes the progression to type 2 diabetes when lowered or improvement of metabolic status when raised [11]. hTG range 0.59% - 21.97%; post HCTZ average hTG = induced changes in hepatic TG levels and insulin sensi- 11.30% +/− 4.56%, hTG range 3.13% - 32.37%; and tivity were statistically significant between groups (p hTG ΔhTG = 4.12%). Hepatic TG levels were unchanged = 0.0098and p = 0.0345 respectively). The individual re- HCTZ SI after Valsartan therapy (baseline average hTG = 8.21% sults as well as the changes in hTG and SI are shown in +/− 3.90%, hTG range 1.62% - 34.92%; post Valsartan Figures 2 and 3. average hTG = 8.27% +/− 3.38%, hTG range 2.83% - DI remained unchanged after HCTZ treatment (base- 31.49%, and ΔhTG = 0.06%). The increased inhepatic line average DI = 1332 +/− 399, DI range 471–3255; post TG levels in the HCTZ group were accompanied by a HCTZ average DI = 1191 +/− 312, DI range 372–2339; reduction in insulin sensitivity (baseline average SI = and ΔDI = −141) but DI increased by a factor of 2 HCTZ 3.71 +/− 0.98, SI range 1.17 - 7.78; post HCTZ average after treatment with Valsartan (baseline average DI = SI = 2.57 +/− 0.20, SI range 1.66-3.04; and ΔSI = 1016 +/− 157, DI range 473–1819; post Valsartan aver- HCTZ −1.14). Treatment with Valsartan resulted in improved age DI 2034 +/− 556, DI range 737 – 5739; and ΔDI insulin sensitivity (baseline average SI = 2.27 +/− 0.68, SI =1018) – primarily due to improvement in SI. However, range 0.46- 6.67; post Valsartan average SI 3.51 +/− 0.80, there change between treatment groups was not signifi- SI range 1.28-7.42; and ΔSI = 1.24). Treatment- cant. Abdominal (visceral) and subcutaneous fat mass as Figure 2 Impact of hydrochlorothiazide (HCTZ) and Valsartan treatments on hepatic triglyceride levels (hTG) and insulin sensitivity (SI). Results of hTG and SI are color coded relative to patient. Black points represent the averages. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 6 of 7 http://www.dmsjournal.com/content/5/1/35 accepted as first-line agent for treating hypertension - should be re-evaluated. HCTZ offer affordable and effi- cient blood pressure lowering but come with the added cost of worsening metabolic profiles for diabetes. This can occur even at doses of 25 mg and is sustained over a lon- ger treatment period (8 months). By the same means, the ARB agents improve insulin sensitivity and did not cause fatty liver. Primum non nocere - our treatment choices should not only improve the primary condition for which they are prescribed, but we must ensure that our patients suf- fer no harm. The ultimate goal of any antihypertensive therapy is to prevent cardiovascular events. The use of an antihypertensive agent that worsens hepatic steatosis and insulin resistance, both of which promote cardiovas- cular disease, negates the ultimate cardiovascular- preventive goal of the treatment. Hypertension clusters with metabolic syndrome, diabetes, and hepatic steatosis, and requires life-long pharmacologic treatment. When all aspects are balanced i.e. the blood pressure lowering effect and the worsening metabolic profile the use of HCTZ as a first line therapeutic choice should be questioned. Figure 3 (a) Changes in hepatic triglyceride content (hTG) after The underlying mechanism of the deleterious meta- treatment with Valsartan and HCTZ ( p = 0.0098). (b) Changes in bolic action of HCTZ is still debated although our re- insulin sensitivity (SI) after treatment with HCTZ and Valsartan (p = −0.0345). sults implicate concomitant worsening of fatty liver and decreased insulin sensitivity. Chronic exposure to angio- tensin II may render fat cells less efficient in their cap- well as myocardial structure and function remained un- acity to adequately store excess triglyceride, resulting in changed in both therapeutic arms. Additionally, we did tissue overflow with ectopic triglyceride and ultimately not detect significant changes in myocardial, pancreatic, hepatic steatosis. Interestingly, blocking the renin- and skeletal muscle triglyceride deposits with these angiotensin system by Valsartan did not result in fatty interventions. liver. The favorable metabolic action of Valsartan is probably complex and could originate from improve- Discussion ment of insulin sensitivity [8] or it could be facilitated We have demonstrated a differential metabolic effect of through the recruitment and differentiation of adipo- two frequently prescribed antihypertensive agents in in- cytes [25]. In light of the above-mentioned observations dividuals at risk for type 2 diabetes. The findings of our it seems natural to hypothesize that pairing ARB with study relate to hepatic steatosis and insulin sensitivity. HCTZ could block angiotensin II and mitigate the ad- HCTZ treatment worsened hepatic steatosis measured verse metabolic effects induced by HCTZ. Regrettably, as hepatic TG content and reduced insulin sensitivity. an attempt to block these unfavorable metabolic effects Valsartan treatment did not affect hepatic TG levels and of HCTZ by combining it with losartan was not success- improved insulin sensitivity. The results of this study ful [26]. We acknowledge that this study is small and these re- reinforce the message that in patients at risk for type 2 diabetes it is particularly important to choose an antihy- sults should be replicated, yet the effect size was consid- pertensive regimen that lowers blood pressure without erable and very comparable to that found in a similar independently conducted study [24]. We also note that exacerbating patient’s metabolic profile. Our results parallel the earlier reported findings from small sample size did not allow us to further explore the the MEDICA study [24]. MEDICA investigators reported mechanisms that contribute to our observations. similar worsening in hepatic triglyceride content and insu- lin sensitivity following 3 months of treatment with 50 mg of HCTZ and observed no change of hepatic triglyceride Conclusions levels following treatment with another ARB class medica- We have documented that HCTZ therapy leads to the tion Candesartan. In light of such tight agreement of development of hepatic steatosis and compromised insu- two independent studies, therapy with HCTZ - widely lin sensitivity in subjects at high risk for type 2 diabetes. Price et al. Diabetology & Metabolic Syndrome 2013, 5:35 Page 7 of 7 http://www.dmsjournal.com/content/5/1/35 Conversely, treatment with Valsartan did not cause ec- 7. Karnes JH, Cooper-DeHoff RM: Antihypertensive medications: benefits of blood pressure lowering and hazards of metabolic effects. Expert Rev topic fat redistribution, and in fact, leads to improved in- Cardiovasc Ther 2009, 7:689–702. sulin sensitivity. Further studies are needed to determine 8. Kurtz TW, Pravenec M: Antidiabetic mechanisms of angiotensin- the exact mechanism by which HCTZ exerts its deleteri- converting enzyme inhibitors and angiotensin II receptor antagonists: beyond the renin-angiotensin system. J Hypertens 2004, 22:2253–2261. ous effects and whether these changes in metabolic 9. Sharma AM, Janke J, Gorzelniak K, Engeli S, Luft FC: Angiotensin Blockade parameters will translate into an increase in major long- Prevents Type 2 Diabetes by Formation of Fat Cells. Hypertension 2002, term adverse cardiovascular events. While the clinical 40:609–611. 10. Nathan DM, Davidson MB, DeFronzo RA, Heine RJ, Henry RR, Pratley R, relevance of thiazide-induced metabolic derangements Zinman B: Impaired fasting glucose and impaired glucose tolerance: remains uncertain at this time, hypertensive individuals implications for care. Diabetes Care 2007, 30:753–759. at risk for diabetes and those with known hepatic 11. Bergman RN: Lilly Lecture 1989. Toward Physiological Understanding of Glucose Tolerance. Minimal-Model Approach. Diabetes 1989, steatosis should opt for antihypertensive agents that are 38:1512–1527. metabolically benign – i.e. ARB or ACE inhibitors – 12. Boesch C, Slotboom J, Hoppeler H, Kreis R: In vivo determination of intra- until this issue is clarified. myocellular lipids in human muscle by means of localized 1H-MR -spectroscopy. Magn Reson Med 1997, 37:484–493. 13. Schick F, Eismann B, Jung WI, Bongers H, Bunse M, Lutz O: Comparison of Competing interests localized proton NMR signals of skeletal muscle and fat tissue in vivo: This study was funded as the investigator initiated research award to LSS two lipid compartments in muscle tissue. Magn Reson Med 1993, and RGB by Novartis. 29:158–167. 14. Szczepaniak LS, Babcock EE, Schick F, Dobbins RL, Garg A, Burns DK, Authors’ contributions McGarry JD, Stein DT: Measurement of intracellular triglyceride stores by ALP and IL designed and carried OGTT and FSIVGTT experiments. H spectroscopy: validation in vivo. Am J Physiol 1999, 276:E977–E989. Participated in results interpretation and writing the manuscript. EWS 15. Szczepaniak LS, Dobbins RL, Stein DT, McGarry JD: Bulk magnetic processed MR imaging and spectroscopy data, participated in data susceptibility effects on the assessment of intra- and extramyocellular interpretation, and assisted with statistical analysis.JW assisted in GCRC lipids in vivo. Magn Reson Med 2002, 47:607–610. experiments, created and maintained data base for the study, processed 16. Szczepaniak LS, Dobbins RL, Metzger GJ, Sartoni-D’Ambrosia G, Arbique D, FSIVGTT data using MinMode software, contributed to discussion on results Vongpatanasin W, Unger R, Victor RG: Myocardial triglycerides and systolic interpretation.RGV participated in study design and results interpretation.LSS function in humans: in vivo evaluation by localized proton spectroscopy designed the study, overlooked all study procedures, executed MR and cardiac imaging. Magn Reson Med 2003, 49:417–423. Spectroscopy and wrote the paper. All authors read and approved the final 17. Szczepaniak LS, Nurenberg P, Leonard D, Browning JD, Reingold JS, Grundy manuscript. S, Hobbs HH, Dobbins RL: Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the Funding sources general population. Am J Physiol Endocrinol Metab 2005, 288:E462–E468. This study was supported in part by Novartis Investigator Initiated Award, 18. McGavock JM, Victor RG, Unger RH, Szczepaniak LS: Adiposity of the heart, NIH UL1RR024982, NIH K23 RR024470, R01 NIDDK081524. revisited. Ann Intern Med 2006, 144:517–524. 19. McGavock JM, Lingvay I, Zib I, Tillery T, Salas N, Unger R, Levine BD, Raskin Author details P, Victor RG, Szczepaniak LS: Cardiac steatosis in diabetes mellitus: a 1H- Department of Internal Medicine, UT Southwestern Medical Center, Dallas, magnetic resonance spectroscopy study. Circulation 2007, Texas, USA. Cedars -Sinai Medical Center, The Heart Institute, Los Angeles, 116:1170–1175. California, USA. 20. Reingold JS, McGavock JM, Kaka S, Tillery T, Victor RG, Szczepaniak LS: Determination of triglyceride in the human myocardium by magnetic Received: 9 July 2013 Accepted: 5 July 2013 resonance spectroscopy: reproducibility and sensitivity of the method. Published: 9 July 2013 Am J Physiol Endocrinol Metab 2005, 289:E935–E939. 21. Lingvay I, Raskin P, Szczepaniak LS: The fatty hearts of patients with References diabetes. Nat Rev Cardiol 2009, 6:268–269. 1. Barzilay JI, Davis BR, Cutler JA, Pressel SL, Whelton PK, Basile J, Margolis KL, 22. Lingvay I, Esser V, Legendre JL, Price AL, Wertz KM, Adams-Huet B, Zhang S, Ong ST, Sadler LS, Summerson J: Fasting glucose levels and incident Unger RH, Szczepaniak LS: Noninvasive quantification of pancreatic fat in diabetes mellitus in older nondiabetic adults randomized to receive 3 humans. J Clin Endocrinol Metab 2009, 94:4070–4076. different classes of antihypertensive treatment: a report from the 23. Abate N, Garg A, Peshock RM, Stray-Gundersen J, Grundy SM: Relationships Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack of generalized and regional adiposity to insulin sensitivity in men. Trial (ALLHAT). Arch Intern Med 2006, 166:2191–2201. J Clin Invest 1995, 96:88–98. 2. Stump CS, Hamilton MT, Sowers JR: Effect of antihypertensive agents on 24. Eriksson JW, Jansson PA, Carlberg B, Hagg A, Kurland L, Svensson MK, the development of type 2 diabetes mellitus. Mayo Clin Proc 2006, Ahlstrom H, Strom C, Lonn L, Ojbrandt K, Johansson L, Lind L: 81:796–806. Hydrochlorothiazide, but not candesartan, aggravates insulin resistance 3. Elliott WJ, Meyer PM: Incident diabetes in clinical trials of and causes visceral and hepatic fat accumulation: the Mechanisms for antihypertensive drugs: a network meta-analysis. Lancet 2007, the Diabetes Preventing Effect of Candesartan (MEDICA) Study. 369:201–207. Hypertension 2008, 52:1030–1037. 4. Gerstein HC, Yusuf S, Bosch J, Pogue J, Sheridan P, Dinccag N, Hanefeld M, 25. Sharma AM: Does it matter how blood pressure is lowered in patients Hoogwerf B, Laakso M, Mohan V, Shaw J, Zinman B, Holman RR: Effect of with metabolic risk factors? J Am Soc Hypertens 2008(4):S23–S29. rosiglitazone on the frequency of diabetes in patients with impaired 26. Bakris G, Molitch M, Hewkin A, Kipnes M, Sarafidis P, Fakouhi K, Bacher P, glucose tolerance or impaired fasting glucose: a randomised controlled Sowers J: Differences in glucose tolerance between fixed-dose trial. Lancet 2006, 368:1096–1105. antihypertensive drug combinations in people with metabolic 5. Basile JN: Antihypertensive therapy, new-onset diabetes, and syndrome. Diab Care 2006, 29:2592–2597. cardiovascular disease. Int J Clin Pract 2009, 63:656–666. 6. Califf RM, Boolell M, Haffner SM, Bethel MA, McMurray J, Duggal A, Holman doi:10.1186/1758-5996-5-35 RR: Prevention of diabetes and cardiovascular disease in patients with Cite this article as: Price et al.: The metabolic cost of lowering blood impaired glucose tolerance: rationale and design of the Nateglinide and pressure with hydrochlorothiazide. Diabetology & Metabolic Syndrome Valsartan in Impaired Glucose Tolerance Outcomes Research 2013 5:35. (NAVIGATOR) Trial. Am Heart J 2008, 156:623–632.

Journal

Diabetology & Metabolic Syndrome – Springer Journals

Published: Jul 9, 2013

There are no references for this article.