Liraglutide Promotes Natriuresisbut Does Not Increase CirculatingLevels of Atrial Natriuretic Peptidein Hypertensive Subjects WithType 2 DiabetesDiabetes Care 2015;38:132–139 | DOI: 10.2337/dc14-1958

OBJECTIVE

GLP-1 receptor (GLP-1R) agonists induce natriuresis and reduce blood pressure(BP) through incompletely understood mechanisms. We examined the effects ofacute and 21-day administration of liraglutide on plasma atrial natriuretic peptide(ANP), urinary sodium excretion, office and 24-h BP, and heart rate (HR).

RESEARCH DESIGN AND METHODS

Liraglutide or placebowas administered for 3 weeks to hypertensive subjects withtype 2 diabetes in a double-blinded, randomized, placebo-controlled crossoverclinical trial in the ambulatory setting. End points included within-subject changefrom baseline in plasma ANP, Nt-proBNP, office BP, and HR at baseline and over 4h following a single dose of liraglutide (0.6 mg) and after 21 days of liraglutide(titrated to 1.8 mg) versus placebo administration. Simultaneous 24-h ambulatoryBP and HR monitoring and 24-h urine collections were measured at baseline andfollowing 21 days of treatment.

RESULTS

Plasma ANP levels did not change significantly after acute (+16.72 pg/mL, P = 0.24,95% CI [212.1, +45.5] at 2 h) or chronic (217.42 pg/mL, 95% CI [236.0, +1.21] at 2h) liraglutide administration. Liraglutide significantly increased 24-h and nighttimeurinary sodium excretion; however, 24-h systolic BP was not significantly differ-ent. Small but significant increases in 24-h and nighttime diastolic BP and HR wereobserved with liraglutide. Body weight, HbA1c, and cholesterol were lower, andoffice-measured HR was transiently increased (for up to 4 h) with liraglutideadministration.

CONCLUSIONS

Sustained liraglutide administration for 3 weeks increases urinary sodium excre-tion independent of changes in ANP or BP in overweight and obese hypertensivepatients with type 2 diabetes.

Patients with type 2 diabetes experience a greater than twofold excess risk for thedevelopment of cardiovascular disease (1,2), with coexistent hypertension furtherincreasing the risk of cardiovascular complications (3). Even modest reductions inblood pressure (BP) reduce the risk of stroke and myocardial infarction (4).

1Lunenfeld-Tanenbaum Research Institute,Mount Sinai Hospital, Toronto, Ontario, Canada2Leadership Sinai Centre for Diabetes, Mount Si-nai Hospital, Toronto, Ontario, Canada3Department of Medicine, Mount Sinai Hospital,University of Toronto, Toronto, Ontario, Canada

Corresponding author: Daniel J. Drucker,drucker@lunenfeld.ca.

Received 14 August 2014 and accepted 9 Octo-ber 2014.

Clinical trial reg. no. NCT01755572, clinicaltrials.gov.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc14-1958/-/DC1.

A slide set summarizing this article is availableonline.

© 2015 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered.

Julie A. Lovshin,1 Annette Barnie,2

Ariana DeAlmeida,3 Alexander Logan,1

Bernard Zinman,1,2 and Daniel J. Drucker1

132 Diabetes Care Volume 38, January 2015

CARDIOVASC

ULA

RANDMETABOLICRISK

Accordingly, there is great interest intherapies that might not only improveglucose but also enhance BP control inhypertensive diabetic subjects.Among various classes of antidia-

betic agents, both sodium-glucosecotransporter 2 (SGLT2) inhibitors andGLP-1 receptor (GLP-1R) agonists areassociated with further reduction of BPin hypertensive subjects. Although theantihypertensive actions of SGLT2 inhib-itors are thought to be largely mediatedthrough enhanced urinary sodium excre-tion and osmotic diuresis (5), the mech-anisms through which GLP-1R agonistsreduce BP are less well understood andmay include weight loss, direct or in-direct vasorelaxation, or stimulation ofnatriuresis (6).Both SGLT2 inhibitors and GLP-1R ag-

onists exert their glycemic effectsthrough mechanisms that are glucosedependent, resulting in low rates of hy-poglycemia. Similarly, both drug classesreduce BP in hypertensive subjectsthrough mechanisms that are attenu-ated in normotensive subjects, therebydiminishing the likelihood of treatment-associated hypotension. Interestingly,natriuresis is induced by both SGLT2 in-hibitors and GLP-1R agonists throughdifferent mechanisms; notably, thepathways and mechanisms linking GLP-1R signaling to renal sodium excretionare controversial and less well under-stood (7). Although short-term infusion(3 h) of GLP-17-36amide increases natri-uresis in healthy volunteers (8,9) andin insulin-resistant obese males (8),whether these actions are sustainedwith chronic GLP-1R activation is un-clear. Furthermore, the majority ofstudies examining how GLP-1R agonistsincrease sodium excretion are often3–72 h in duration (7–11). Hence, thereis limited information on how sustainedadministration of GLP-1R agonists regu-lates mechanisms leading to BP reduc-tion in hypertensive subjects with type 2diabetes.We recently examined pathways link-

ing GLP-1R signaling to control of BP innondiabetic hypertensive mice. Thesestudies demonstrated that structurally di-verse GLP-1R agonists, including nativeGLP-1, liraglutide, and exenatide, in-creased plasma levels of atrial natriureticpeptide (ANP), enhanced natriuresis,and reduced BP in mice with angiotensinII–induced hypertension (12). In this

context, we examinedwhether liraglutideincreased ANP and urinary sodium excre-tion in a prospective, double-blinded, ran-domized, placebo-controlled, crossovertrial in 20 hypertensive patients withtype 2 diabetes.

RESEARCH DESIGN AND METHODS

Study DesignThis study was a single-center, prospec-tive, double-blinded, randomized (1:1),placebo-controlled, crossover studythat took place at Mount Sinai Hospital,Toronto, Ontario, Canada. Subjectswere randomized to liraglutide or to pla-cebo for 21 days during two separatetreatment periods (treatment period 1and treatment period 2) separated by a21-day treatment washout period (Sup-plementary Fig. 1A and B). Ten patientsstarted in treatment sequence A (firstrandomized to receive liraglutide), and10 patients were randomized to startin treatment sequence B (first ran-domized to receive placebo). The studycoordinators enrolled patients andpharmacy-assigned participants to in-terventions according to the random-ization sequence that was computergenerated. Treatment allocation wasblinded to patients and study personneluntil the database was locked for analy-sis. Recruitment commenced January2013 and ended October 2013, andfollow-up ended January 2014.

During treatment period 1, subjectsself-administered study drug (liraglutideor placebo) daily, at an initial dose of 0.6mg for the 1st week (day 1, clinic visit 1)and then 1.2 mg for the 2nd week fol-lowed by 1.8 mg until day 21 (clinic visit2) (Supplementary Fig. 1A). To preserveblinding, placebo was volume matched,with a two-step sham titration, and de-livered in a pen device identical to thatused for liraglutide. After completion ofthe first 21-day treatment period, sub-jects underwent a 21-day washoutperiod and started treatment period2 (crossover to opposite treatmentsequence) on day 42 (clinic visit 3).All subjects continued their routinemedications throughout the study;however, if a patient was treated witha sulfonylurea prior to study entry, thedose of the sulfonylurea was reducedby 50% at the start and for the durationof each treatment phase. Antihyperten-sive medications were temporarily dis-continued for 1–2 days prior to each

clinical visit. One patient could not tol-erate temporary discontinuation of theirantihypertensive medications and com-pleted test days on therapy, and twopatients neglected to withhold theirantihypertensive medications at clinicvisit 1. To maintain subject consistencybetween visits, they were instructed tocomplete the remaining clinic visits andtests on therapy. All of these patientswere included in the study populationfor the analysis.

Study PopulationStudy entry criteria included subjectswith type 2 diabetes with an HbA1c

$6.5 and #10.1%, with systolic hyper-tension (systolic BP [SBP] $130 and#180 mmHg) and aged .18 years. Ex-clusion criteria reflected conditions con-traindicated with the recommended useof liraglutide (http://www.pbm.va.gov/clinicalguidance/drugmonographs/Liraglutidemonograph.pdf). Recruit-ment occurred directly from familyphysician or specialty clinics or via self-referral in response to advertising assummarized in Supplementary Fig. 2.

Experimental ProtocolThe study protocol was approved by theMount Sinai Hospital Research EthicsBoard (Toronto, Ontario, Canada) andperformed in accordance with the Dec-laration of Helsinki and the Tri-CouncilPolicy Statement: Ethical Conduct forResearch Involving Humans. This studywas registered at www.clinicaltrials.gov(NCT01755572).

Baseline TestsAfter study eligibility criteria were con-firmed (Supplementary Fig. 2), subjectsentered the baseline test period, whichconsisted of a 24-h ambulatory bloodpressure monitoring (ABPM) and a24-h urine collection (1–7 days prior today 1) as described below.

Clinic Visits

Subjects attended four clinic visits (days1, 21, 42, and 63) at the start and end ofeach treatment period (SupplementaryFig. 1A) and were instructed to withholdantihypertensive medications (24–48 h)and oral antidiabetic agents (12 h) be-fore clinic visits. Patients fasted prior toeach assessment (12 h) and were in-structed to avoid caffeine, smoking,and exercise on the day before andmorning of all clinic visits. Office BPand heart rate (HR) were measured

care.diabetesjournals.org Lovshin and Associates 133

serially (average of five readings over 6min, first reading discarded) by a cali-brated, automated oscillometric sphyg-momanometer (BpTRU; BpTRU MedicalDevices, Coquitlam, British Columbia,Canada). These measurements wereperformed prior to study drug adminis-tration (time = 0 h) and at the end of theclinic visit (time = 4 h).

Blood Collection

A venous cannula was inserted in thesubject’s forearm and blood was sampledat25min prior to study drug administra-tion and then hourly for 4 h thereafter(Supplementary Fig. 1B). Blood was col-lected at room temperature or, for ANPand angiotensin II collection, into pre-chilled (248C) 4-mL K2EDTA tubes towhich 50 mL of aprotinin (A6279; Sigma-Aldrich [for ANP]) or 100 mL bestatin(002-A22/B-BST ALPCO; Fisher Scientific[for angiotensin II]) had been added. Sam-ples were either immediately centrifuged(1,800g) for 10 min at 48C, aliquoted, andstored (2808C) or sent to the clinical lab-oratory for biochemical analysis.

Biochemical Assays

Plasma ANP concentrations were mea-sured in duplicate using an EIA kit (pro-tocol III, catalog no. S-1131; PeninsulaLaboratories, Inc., Bachem International).Angiotensin II concentrations were de-termined using an RIA kit (BuhlmannLaboratories, ALPCO Diagnostics, Salem,NH). All other analytes were measuredin the clinical laboratories of MountSinai Hospital.

Ambulatory BP (24 h)

Ambulatory BP (24 h) was assessedusing a Spacelabs system (model90207–30) at baseline and 1–2 daysprior to the end of each treatment pe-riod (days 21 and 63). Subjects recordedtheir sleep and awake times over 24 h.The ambulatory BP readings (CardiologyInformation Management System, ver-sion 9.0.2.4475; Sentinel, SpacelabsHealthcare) were interpreted by a ne-phrologist blinded to treatment alloca-tion. One individual had incomplete24-h ABPM recordings; hence, in addi-tion to the 2 subjects excluded due tosignificant protocol deviations (Supple-mentary Fig. 2), the number of individ-uals included in the BP analysis was 17.

Urine Collections (24 h)

Urine collections were performed simul-taneously with the 24-h ABPM 1–2 daysprior to clinic visits. Patients did not

undergo sodium restriction; however,they were instructed to maintain a con-sistent dietary intake throughout thestudy.

Statistical MethodsIndependent statisticians performedthe statistical analyses. The primary out-come was the within-subject hourly andbaseline change in plasma ANP betweenliraglutide or placebo assessed after thefirst injection or after 3 weeks of dailyadministration. Secondary outcomesincluded within-subject differences in24-h ABPM and 24-h urine sodium ex-cretion after 3 weeks of treatment be-tween liraglutide and placebo. For theprimary outcome, a sample size of 16participants would provide 80% powerto detect a 20% difference in plasmaANP at a significance level (a) of 0.05.This calculation was based upon the es-timated treatment effect observed inpreclinical studies for plasma ANP andliraglutide (12). Supplementary Table 1provides descriptive statistics for pa-tient baseline characteristics where themean and standard deviation are re-ported for continuous variables andthe median with interquartile rangeare presented for skewed variables.For categorical variables, the countsand percentages are provided.

The treatment effect between the twointerventions, liraglutide and placebo,was estimated from a linear random-effects model with a random effect forsubject and fixed effects for period, se-quence, and treatment. For the linearregression models, residual diagnosticswere performed to check for the validityof the models. To determine the treat-ment significance for nonnormal dis-tributed data, nonparametric Wilcoxonsigned rank test was used.

For secondary end points (24-h ABPMand urine sodium), the within-subjectdifferences were calculated between lira-glutide and placebo following 21 days oftreatment as compared with baselinemeasurements. All statistical analyseswere performed using SAS, version9.1 (SAS Institute, Cary, NC), with two-sided probability values ,0.05 consid-ered to be statistically significant.

RESULTS

SubjectsEighteen participants were included inthe analysis (2 subjects were excluded

due to significant protocol deviations),with the exception of the 24-h ABPMfor which 17 individuals were included(1 subject excluded due to incomplete24-h ABPM tests). The study populationprimarily consisted of overweight andobese (BMI = 29.50 kg/m2) males, medi-an age 62 years, with a history of type 2diabetes of 5.8 years and a mean base-line HbA1c of 7% (Supplementary Table1). Themajority of subjects were treatedwith metformin (95%) and/or a sulfonyl-urea (35%). At screening, the medianSBP and diastolic BP (DBP) was 144and 85 mmHg, respectively; 58% of sub-jects were nondippers (failure to reducenocturnal SBP by 10% relative to day)and 18% were risers (increase in noctur-nal SBP relative to day). Surprisingly,44% of the study population was notbeing treated for hypertension. The me-dian ACR was +1.38 mg/mmol, and sev-eral individuals had microalbuminuria(.30 mg/24 h).

Cardiac Natriuretic HormonesPlasma ANP levels were unchanged af-ter the first single 0.6-mg injection ofliraglutide (+16.72 pg/mL, P = 0.24,95% CI [–12.1, +45.5] at 2 h) (Fig. 1Aand Supplementary Fig. 3A). Similarly,after 21 days of liraglutide (1.8 mg) ther-apy (Fig. 1B and Supplementary Fig. 3B),no statistically significant changeswere observed in plasma ANP levelswith liraglutide compared with placebo(217.42 pg/mL, P = 0.07, 95% CI [–36.0,+1.21] at 2 h). No significant changes inserum Nt-proBNP concentrations wereobserved at day 1 or following 21 days(Supplementary Fig. 3C and D).

Urine SodiumA significant increase in 24-h urinary so-dium excretion (median change +14.18mmol/L liraglutide vs. placebo) andnighttime urinary sodium excretion(median change +4.24 mmol/L night-time, liraglutide vs. placebo) was ob-served following 21 days of liraglutide(Fig. 2A and B). No significant changesin the albumin-to-creatinine ratio wereobserved with liraglutide (Supplemen-tary Fig. 4A and B).

BP and HRNo differences were observed in 24-h,daytime, or nighttime (Table 1) or officeSBP (Supplementary Table 3). Althoughliraglutide did not produce significantreductions in SBP, a robust drop in SBPfor liraglutide was observed between

134 Liraglutide Increases Natriuresis Diabetes Care Volume 38, January 2015

2100 h and 0200 h (Fig. 3). Small, yetstatistically significant, increases in 24-h(least squares mean difference [LSMD]+3.78 6 1.34 mmHg, P = 0.01) and noc-turnal DBP (LSMD +3.77 6 1.74 mmHg,P = 0.05) were observed (Table 1) withliraglutide treatment as compared withplacebo. The within-subject differencefor 24-h (LSMD +5.21 6 2.42, P = 0.05)and nighttime HR (LSMD +7.34 6 2.38,P = 0.008) was statistically increasedwith liraglutide treatment comparedwith placebo. A significant rise in baseline(time = 0 h) HR measured in the office onclinic days was observed after 21 days ofliraglutide (LSMD, +9.25 6 3.5 bpm, P =0.02) (Supplementary Table 3); how-ever, by 4 h after liraglutide injection,the within-subject increases in office-measured HR were no longer statisticallysignificant (LSMD +3.696 3.02 bpm, P =0.24) (Supplementary Table 3).

Metabolic End PointsLiraglutide significantly reduced levelsof HbA1c (mean 7.04–6.61%, LSMD20.7%, P = 0.005), fasting plasma

glucose (+8.39 to 5.68 mmol/L, LSMD23.4 mmol/L, P = 0.0004), and choles-terol (mean +4.16 to 3.64 mmol/L,LSMD20.63 mmol/L, P = 0.002 for totalcholesterol; +2.00 to 1.78 mmol/L,LSMD 20.37 mmol/L, P = 0.04 for LDLcholesterol) (Supplementary Table 2).No significant changes in serum creati-nine or plasma angiotensin II concen-trations were detected. There was astatistically significant decrease in esti-mated glomerular filtration rate (Supple-mentary Table 2). Liraglutide reducedbody weight (LSMD 21.35 6 0.46 kg,P = 0.009) and BMI (LSMD 20.43 6 0.18kg/m2, P = 0.03), without changes in waistcircumference (Supplementary Table 3).

Adverse EventsAll 20 patients completed the researchstudy, and no serious adverse events oc-curred (Supplementary Table 4). Themost common complaints were a lossof appetite (30%) or were gastrointesti-nal in nature, including nausea (20%),dyspepsia (15%), and flatulence (15%).One individual was unable to tolerate

dose titration due to dyspepsia andcompleted the study at the lower dose(1.2 mg) following retitration.

CONCLUSIONS

A gut-cardiac GLP-1R–ANP axis was re-cently identified in nondiabetic rodentswith angiotensin II–induced hyperten-sion (12); however, the current findingsdo not suggest a role for a GLP-1R–ANPaxis in transducing the renovascular ef-fects of liraglutide in hypertensive sub-jects with type 2 diabetes. Our datademonstrating no acute increase inplasma ANP levels after liraglutide ad-ministration are consistent with resultsfroma studyof nativeGLP-1 (1.25pmol/kg)in 12 healthy, young, normal-weight,male human subjects. Although a 2-h in-fusion of GLP-1 increased urinary so-dium excretion (11), there were nochanges in levels of circulating pro-ANP over 2 h.

Our current findings extend existingknowledge of the relationships en-compassing GLP-1 action and changesin cardiorenal parameters in several

Figure 1—Treatment effect of 1- or 21-day liraglutide treatment on cardiac natriuretic peptides compared with crossover treatment with placebo.The difference in mean change from baseline in plasma ANP concentrations was measured hourly following 1-day (A) and 21-day (B) treatment withliraglutide and following crossover treatment with placebo. The data are presented as the hourly mean difference from baseline with 95% CI.Treatment doses: 1 day (0.6 mg, single dose) and 21 day (1.8 mg, final dose).

care.diabetesjournals.org Lovshin and Associates 135

ways. First, we did not restrict measure-ment of natriuretic peptides to a singletime point; rather, we assessed serialchanges in plasma levels of ANP andNt-proBNP over time, in acute studies,and after 21 days of liraglutide adminis-tration. Second, we simultaneously as-sessed whether sustained liraglutidetherapy was associated with changes inurinary sodium excretion. Third, ourstudies were carried out in hypertensiveoverweight and obese subjects withtype 2 diabetes, characteristics rep-resentative of patients treated withGLP-1R agonists in a clinical setting.

Furthermore, we were able to assesschanges in BP in the same group ofsubjects, enabling an assessment ofwhether changes in ANP or urinary so-dium were associated with potentialreduction of BP. Our data clearly dem-onstrate that liraglutide increased uri-nary sodium excretion independent ofchanges in circulating ANP or reductionof SBP. These findings, taken togetherwith the lack of significant SBP reductionover 3 weeks in the same patient popu-lation, suggest that enhanced urinary so-dium excretion alone may be insufficientto explain the BP-lowering effects

described with liraglutide and otherGLP-1R agonists (7,13).

Intriguingly, an observational non-randomized study of 31 obese (BMI31.7 kg/m2), prehypertensive (mean BP138.2/85.9 mmHg) subjects with type 2diabetes reported significant increasesin plasma levels of ANP and Nt-proBNPafter 12 weeks of daily liraglutide ad-ministration (14). Furthermore, themagnitude of changes in ANP andNt-proBNP correlated with the extentof weight loss after 12 weeks. In a sec-ondary analysis, we observed robustchanges in plasma ANP concentrationsin four patients after acute administra-tion of liraglutide. Taken together, thesefindings are consistent with the knownheterogeneity in control of ANP secre-tion and the pathophysiology of hyper-tension in human subjects (15) andleave open the possibility that a smallsubset of hypertensive subjects mayexhibit increased ANP secretion afteradministration of GLP-1R agonists. Fur-thermore previous studies have dem-onstrated that levels of circulatingnatriuretic peptides may be increasedafter weight loss secondary to lifestylechanges (16,17) or bariatric surgery (18).Hence, although the available datasuggest that acute GLP-1R activationdoes not stimulate secretion of ANP orNt-proBNP in humans, chronic therapywith GLP-1R agonists may indirectlyincrease circulating levels of ANP andNt-proBNP in obese subjects throughincompletely understood mechanismsrelated to weight loss.

Our study has several limitations. Atstudy entry, the majority of patientswere receiving concomitant therapywith single or multiple antihypertensiveagents, including 55% receiving adju-vant diuretic therapy. Hence, one ormore of these antihypertensive agentsmay have impacted the effect of liraglu-tide to modulate ANP secretion, urinarysodium excretion, or BP. Second, we didnot measure dietary changes in sodiumintake. As patients lost body weight fol-lowing 21-day liraglutide treatment, so-dium intake may have increased ordecreased with liraglutide treatment.As several study end points, such asplasma ANP levels and urinary sodiumexcretion, are sensitive to changes in di-etary sodium intake, this represents amajor potential limitation in study de-sign. Furthermore, our study population

Figure 2—The effect of 21-day treatment with liraglutide on 24-h, daytime, and nighttimeurinary sodium excretion compared with crossover treatment with placebo. A: Median (95%CI) urine sodium for 24 h, daytime, and nighttime for the treatment groups (liraglutide andplacebo). B: The within-subject median change in urine sodium excretion for liraglutide minusplacebo is presented in the box plot and whiskers graph. For the box plots and whiskers graph,the horizontal line indicates the median change, the box represents interquartile range of thechange, and outliers are presented as single points. Nonparametric tests were used for com-parison (Wilcoxon rank sum). *, the normal approximation for two-sided P value,0.05; **, thenormal approximation for two-sided P value ,0.005 for liraglutide compared with placebo.

136 Liraglutide Increases Natriuresis Diabetes Care Volume 38, January 2015

was predominantly overweight or obesemiddle-aged males with established di-abetes and hypertension, and it remainsunclear whether other cohorts of dia-betic subjects may have responded dif-ferently to liraglutide.Three principal features of our study

include the small sample size, briefstudy period, and the achievement ofmaximally approved liraglutide dosing(1.8 mg) for only the last week of thestudy. These limitations may partiallyexplain why we were unable to detectsignificant changes in plasma ANP levelsor SBP in this patient population. Never-theless, our BP findings are in agreementwith the results of a trial examiningthe BP-lowering effects of twice-daily ex-enatide, which reported trends towardlowering of SBP, daytime DBP, andnighttime BP; however, none of thedifferences reported were statisticallysignificant (19). In contrast, Ferdinandet al. (20) randomized 755 patients todulaglutide or placebo and reported asignificant reduction in SBP after 16weeks in 251 subjects randomized toreceive 1.5 mg of dulaglutide adminis-tered once weekly. Furthermore, severallarge outcome studies have demon-strated significant reductions in SBP fol-lowing 26-week or longer treatment with

liraglutide against placebo and ac-tive comparator controls (21–23). Hence,the small sample size of the current ex-ploratory study or the short exposureto maximal treatment with liraglu-tide (1.8 mg for only 7 days) may havelimited detection of small changes inSBP associated with liraglutide.

We demonstrate that overweightor obese subjects with type 2 diabetesand hypertension exhibit significant in-creases in urinary sodium excretion inresponse to sustained liraglutide ad-ministration, independent of concomi-tant changes in BP or circulating levelsof natriuretic peptides. Hence, the anti-hypertensive mechanism(s) engagedby GLP-1R signaling in this patient pop-ulation are independent of a cardio-renal GLP-1–ANP axis described inrodents (12). Our findings provide newinsight into temporal pharmacody-namic changes arising in hypertensivediabetic subjects treated with liraglu-tide and refine our understanding ofthe relationships between GLP-1R–dependent increases in urinary sodiumand potential reductions in BP. In con-trast, the chronotropic actions of lira-glutide were easily detected in ourstudy; however, the increment in HRwas transient and was no longer

statistically significant 4 h after the lastdose of liraglutide. Collectively, our dataemphasize that the mechanisms linkingactivation of GLP-1R signaling to controlof BP in hypertensive diabetic humans re-main incompletely understood and re-quire further investigation.

Funding. This investigator-initiated clinical trialwas supported with a grant-in-aid from NovoNordisk, which also supplied the liraglutide andplacebo. J.A.L. is a postdoctoral research fellowin the Division of Endocrinology & Metabolismat the University of Toronto and is supported byan Eliot Phillipson Clinician Scientist TrainingFellowship Award. B.Z. holds the Sam and JudyPencer Family Chair in Diabetes Research atMount Sinai Hospital and University of Toronto.D.J.D. is supported in part by the Canada Re-search Chairs Program and the Banting & BestDiabetes Centre–Novo Nordisk Chair in IncretinBiology.Duality of Interest. J.A.L. has received speaker’shonoraria from Novo Nordisk. B.Z. has receivedresearch support and/or consulting honorariafrom AstraZeneca, Boehringer Ingelheim, Eli Lillyand Company, Johnson & Johnson, Merck, NovoNordisk, and Takeda. D.J.D. has served as an ad-visor or consultant within the past 12 months toArisaph Pharmaceuticals, Inc., Intarcia Therapeu-tics, Merck Research Laboratories, MedImmune,NovoNordisk, NPSPharmaceuticals, Inc., Receptos,Inc., Sanofi, and Transition Therapeutics Inc. D.J.D.receives support for preclinical studies throughgrants to Mount Sinai Hospital from Merck, GSK,

Table 1—Within-subject treatment difference for liraglutide compared with crossover treatment with placebo on 24-h ABPM

Variable Baseline median (ICR)

Treatment group liraglutide(1.8 mg) 21 days, (n = 17)LSMD from baseline (6SE)

Treatment group placebo21 days, (n = 17) LSMDfrom baseline (6SE)

Treatment difference liraglutide(1.8 mg) vs. placeboLSMD (6SE), P value

SBP (mmHg)24 h 130.45 (136.19, 127.10) +1.62 6 2.04 20.72 6 2.04 +2.33 6 1.67, P = 0.18Daytime 133.07 (141.70) +1.68 6 2.14 20.58 6 2.14 +2.25 6 1.78, P = 0.23Nighttime 122.73 (129.69, 118.58) +0.87 6 2.64 20.83 6 2.64 +1.70 6 2.77, P = 0.54

DBP (mmHg)24 h 79.74 (83.05, 68.79) +3.39 6 1.12 20.39 6 1.12 +3.78 6 1.34, P = 0.01Daytime 79.67 (85.44, 70.48) +3.66 6 1.24 +0.19 6 1.24 +3.48 6 1.48, P = 0.03Nighttime 71 (75.75, 63.98) +2.32 6 1.46 21.44 6 1.46 +3.77 6 1.74, P = 0.05

HR (bpm)24 h 67.59 (86.17, 63) +7.39 6 2.08 +2.18 6 2.08 +5.21 6 2.42, P = 0.05Daytime 73.32 (89.68, 64.93) +6.51 6 2.28 +1.94 6 2.27 +4.57 6 2.69, P = 0.11Nighttime 63.25 (76.02, 58.53) +9.73 6 1.77 +2.39 6 1.77 +7.34 6 2.38, P = 0.008

MAP (mmHg)24 h 97.03 (100.32, 91.26) +1.71 6 1.83 20.60 6 1.83 +2.31 6 2.18, P = 0.30Daytime 101.43 (103.09, 93.74) +2.65 6 1.48 20.06 6 1.48 +2.71 6 1.53, P = 0.10Nighttime 89.57 (95.07, 84.56) 20.10 6 1.53 +1.21 6 1.54 21.32 6 2.18, P = 0.55

PP (mmHg)24 h 54.62 (61.76, 46.88) 21.99 6 1.36 20.76 6 1.36 21.22 6 1.33, P = 0.37Daytime 54.75 (63.07, 47.42) 20.98 6 0.94 21.77 6 0.94 +0.78 6 1.33, P = 0.56Nighttime 52.64 (60.12, 46.13) 21.45 6 1.63 +0.62 6 1.63 22.07 6 1.50, P = 1.19

Baseline data are presented as median (ICR), and the baseline-subtracted within-subject difference for 21-day liraglutide treatment compared with21-day treatment with placebo are presented as LSMD6 SE. Boldface indicates statistical significance. ICR, interquartile range; LSMD, least squaresmean difference; MAP, mean arterial pressure; PP, pulse pressure.

care.diabetesjournals.org Lovshin and Associates 137

Novo Nordisk, and Sanofi. Neither D.J.D. nor hisfamily members hold stock directly or indirectlyin any of these companies. No other potentialconflicts of interest relevant to this articlewere reported.Author Contributions. J.A.L. codesigned thestudy, authored the study protocol, authoredthe Research Ethics Board (REB) submission,conducted and carried out the study, inter-preted the results, and directed the statisticalanalysis and was the primary author of themanuscript. A.B. was the primary researchnurse and study coordinator for this study,assisted with the REB submission, and reviewedand commented on the manuscript. A.D. as-sisted with data analysis, produced figures forthe manuscript, and reviewed and commentedon the manuscript. A.L. assisted in study designand interpreted the 24-h ABPM reports. B.Z.codesigned the study, assisted in interpretingthe results, and reviewed and commented on themanuscript. D.J.D. codesigned the study, re-viewed and commented on the study protocoland REB submission, interpreted the results,critically appraised and revised the manuscript,and served as the principal investigator for thisstudy. D.J.D. is the guarantor of this work and, assuch, had full access to all the data in the studyand takes responsibility for the integrity of thedata and the accuracy of the data analysis.

References1. Seshasai SR, Kaptoge S, Thompson A, et al.;Emerging Risk Factors Collaboration. Diabetesmellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011;364:829–8412. Sarwar N, Gao P, Seshasai SR, et al.; EmergingRisk Factors Collaboration. Diabetes mellitus,fasting blood glucose concentration, and risk ofvascular disease: a collaborative meta-analysis of102 prospective studies [published correctionappears in Lancet 2010;376:958]. Lancet 2010;375:2215–22223. Lewington S, Clarke R, Qizilbash N, Peto R,Collins R; Prospective Studies Collaboration.Age-specific relevance of usual blood pressureto vascular mortality: a meta-analysis of individ-ual data for one million adults in 61 prospectivestudies. Lancet 2002;360:1903–19134. UK Prospective Diabetes Study Group. Tightblood pressure control and risk of macrovascularand microvascular complications in type 2 dia-betes: UKPDS 38. BMJ 1998;317:703–7135. Oliva RV, Bakris GL. Blood pressure effects ofsodium-glucose co-transport 2 (SGLT2) inhibi-tors. J Am Soc Hypertens 2014;8:330–3396. Ussher JR, Baggio LL, Campbell JE, et al.Inactivation of the cardiomyocyte glucagon-like peptide-1 receptor (GLP-1R) unmaskscardiomyocyte-independent GLP-1R-mediatedcardioprotection. Mol Metab 2014;3:507–5177. Ussher JR, Drucker DJ. Cardiovascular actionsof incretin-based therapies. Circ Res 2014;114:1788–18038. Gutzwiller JP, Tschopp S, Bock A, et al.Glucagon-like peptide 1 induces natriuresisin healthy subjects and in insulin-resistantobese men. J Clin Endocrinol Metab 2004;89:3055–30619. Skov J, Dejgaard A, Frøkiær J, et al. Glucagon-likepeptide-1 (GLP-1): effect on kidney hemodynamics

Figure 3—Effect(s) of 21-day liraglutide treatment on mean 24-h ambulatory BP and HR com-pared with crossover treatment with placebo. Twenty four–hour ABPM was used to measurehourly SBP (A), DBP (B), and HR (C) during a 24-h period at baseline and following 21-daytreatment with liraglutide and with placebo. Data are presented as hourly means (6SD). Solidblack line with square markers represents liraglutide (1.8 mg), solid gray line with open circularmarkers represents placebo, and dashed line with closed circular markers represents baseline.Least-squares mean difference in 24-h ABPM is presented in Table 1.

138 Liraglutide Increases Natriuresis Diabetes Care Volume 38, January 2015

and renin-angiotensin-aldosterone system inhealthy men. J Clin Endocrinol Metab 2013;98:E664–E67110. Gutzwiller JP,Hruz P,HuberAR, et al. Glucagon-like peptide-1 is involved in sodium and water ho-meostasis in humans. Digestion 2006;73:142–15011. Skov J, Holst JJ, Gotze JP, Frokiaer J,Christiansen JS. Glucagon-like peptide-1: effecton pro-atrial natriuretic peptide in healthymales. Endocr Connect 2014;3:11–1612. Kim M, Platt MJ, Shibasaki T, et al. GLP-1receptor activation and Epac2 link atrial natri-uretic peptide secretion to control of bloodpressure. Nat Med 2013;19:567–57513. Fonseca VA, Devries JH, Henry RR, DonsmarkM, Thomsen HF, Plutzky J. Reductions in systolicblood pressure with liraglutide in patients withtype 2 diabetes: insights from a patient-levelpooled analysis of six randomized clinical trials. JDiabetes Complications 2014;28:399–40514. Li CJ, Yu Q, Yu P, et al. Changes in liraglutide-induced body composition are related to mod-ifications in plasma cardiac natriuretic peptideslevels in obese type 2 diabetic patients. Cardio-vasc Diabetol 2014;13:36

15. Padmanabhan S, Newton-Cheh C, DominiczakAF. Genetic basis of blood pressure and hyperten-sion. Trends Genet 2012;28:397–40816. Bertoni AG, Wagenknecht LE, Kitzman DW,Marcovina SM, Rushing JT, Espeland MA; BrainNatriuretic Peptide Subgroup of the Look AHEADResearch Group. Impact of the Look AHEAD in-tervention on NT-pro brain natriuretic peptide inoverweight and obese adults with diabetes. Obe-sity (Silver Spring) 2012;20:1511–151817. Chainani-Wu N, Weidner G, Purnell DM,et al. Relation of B-type natriuretic peptide lev-els to body mass index after comprehensivelifestyle changes. Am J Cardiol 2010;105:1570–157618. Chen-Tournoux A, Khan AM, Baggish AL,et al. Effect of weight loss after weight losssurgery on plasma N-terminal pro-B-type natri-uretic peptide levels. Am J Cardiol 2010;106:1450–145519. Gill A, Hoogwerf BJ, Burger J, et al. Effect ofexenatide on heart rate and blood pressure insubjects with type 2 diabetes mellitus: a double-blind, placebo-controlled, randomized pilotstudy. Cardiovasc Diabetol 2010;9:6

20. Ferdinand KC, WhiteWB, Calhoun DA, et al.Effects of the once-weekly glucagon-like peptide-1 receptor agonist dulaglutide on ambulatoryblood pressure and heart rate in patients withtype 2 diabetes mellitus. Hypertension 2014;64:731–73721. Zinman B, Gerich J, Buse JB, et al. Efficacyand safety of the human glucagon-like peptide-1 analog liraglutide in combination with metfor-min and thiazolidinedione in patients with type 2diabetes (LEAD-4 Met+TZD). Diabetes Care 2009;32:1224–123022. Vilsbøll T, Zdravkovic M, Le-Thi T, et al. Lira-glutide, a long-acting human glucagon-likepeptide-1 analog, given as monotherapy sig-nificantly improves glycemic control and low-ers body weight without risk of hypoglycemiain patients with type 2 diabetes. Diabetes Care2007;30:1608–161023. Russell-Jones D, Vaag A, Schmitz O, et al.Liraglutide vs insulin glargine and placebo in com-bination with metformin and sulfonylurea ther-apy in type 2 diabetes mellitus (LEAD-5 met+SU):a randomised controlled trial. Diabetologia 2009;52:2046–2055

care.diabetesjournals.org Lovshin and Associates 139