National Heart, Lung, and Blood Institute

Working Group



September 17, 2007



Workshop Report (in-page links)
Background and Significance
Summary Recommendation

Background and Significance

There are over 70 million hypertensive individuals in the United States. Over 60% of them or more than 45 million are on drug treatment. Of these, only about one-fourth are on treatment regimens that include a thiazide-type diuretic. According to the data extracted from the National Disease and Therapeutic Index (conducted by IMS Health) by Stafford et al., the percent of hypertensive patients treated with diuretics fell from 60% to 19% between 1980 and 2002. Publication of the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) results in JAMA in December 2002 reversed this trend. The percent of all hypertension visits with patients prescribed thiazide-type diuretics rose to 29% in 2003, and then stabilized at ~26%. Similar trends were reported by Ma et al. from two US National Ambulatory Surveys conducted by the National Center for Health Statistics. The percentage of antihypertensive drug visits receiving thiazide diuretics doubled from 23% (19%-28%) in 1993 to 46% (42%-51%) in 2004. In the Department of Veterans Affairs, which participated in several of the studies supporting the use of thiazide-type diuretics for treatment of hypertension, in 2005 less than 50% of hypertensive patients on drug treatment had a thiazide-type diuretic included in their hypertension treatment regimens. As of Fiscal Year 2007 a new thiazide diuretic performance measure was initiated in the VA, so these rates are expected to improve to some degree. However, this approach may not be feasible in medical settings where the majority of Americans receive medical care.

ALLHAT was a randomized, double-blind, active-controlled antihypertensive treatment trial in 42,418 patients assigned to a thiazide-type diuretic, an ACE-inhibitor, a calcium channel-blocker, or an a-blocker-based treatment. After an average follow-up of 4.9 years (3.2 for the a-blocker/diuretic comparison which was terminated early), it showed that the diuretic was at least as good as the comparator drugs in lowering blood pressure and preventing cardiovascular and renal outcomes. The diuretic was superior for preventing heart failure (versus all), combined cardiovascular events (versus a-blocker and ACE-inhibitor) and stroke (versus ACE-inhibitor [Blacks only] and a-blocker).

Thiazide-type diuretics are recommended by the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC7) guidelines as preferred initial treatment of hypertension in most patients (with and without diabetes), either alone or in combination with other antihypertensive drugs. This recommendation is based on over four decades of clinical trials, including active-controlled trials where diuretics were tested against other drugs for their efficacy in preventing hard clinical outcomes such as myocardial infarction, death, stroke, heart failure and renal failure.

Clinical trials have also frequently shown potentially undesirable metabolic biochemical effects during diuretic treatment compared to other drugs, including an increase in serum glucose levels (dysglycemia). Diuretic-induced increases in serum glucose levels are likely small and appear to attenuate over time. (“Diuretic-induced” indicates the part of the diuretic-associated increase in serum glucose levels that is above the increase related to aging, weight gain, sedentary life style, and other risk factors.) Nevertheless, opinion leaders in the medical community have raised concerns about the potential for long-term adverse cardiovascular and renal effects of the observed dysglycemia. They argue that the average length of follow-up in clinical trials, 4-5 years, is not long enough to recognize the potential adverse effects of the known biochemical changes. In addition, they express a concern that patients who develop thiazide-associated diabetes will now require monitoring and treatment for diabetes that they would not have experienced without the thiazide. The ongoing debate regarding the implications of diuretic-induced biochemical changes hampers adoption of the hypertension treatment guidelines with respect to thiazide-type diuretics. Avoidance of diuretics leaves millions of patients on diuretic-free regimens that impart a higher risk of new-onset heart failure and, especially in black patients, also a higher risk of stroke, while providing no known advantages other than small biochemical changes that have no documented long-term clinical effect. On the other hand, it is possible that the clinical advantages of the thiazide-type diuretics could be enhanced by eliminating or diminishing their biochemical effects.

On September 17, 2007, the NHLBI convened an expert working group to discuss potential research approaches that could lead to amelioration of the metabolic effects of diuretics and/or removal of a scientific barrier to adoption of evidence-based guidelines by furthering understanding of the mechanistic underpinnings and consequences of these effects.

Back to top


The specific purpose of the Working Group was to discuss a potential new initiative for short-term clinical trials to:

a) Test a hypothesis that maintenance of potassium balance prevents an increase in serum glucose levels;
b) Evaluate approaches to preventing the increase in serum glucose levels, such as potassium supplementation, combining diuretics with ACE-inhibitors, ARBs, or potassium-sparing diuretics; and
c) Explore the mechanistic underpinnings of diuretic-induced dysglycemia and its prevention.

The charge to the group consisted of the following:

a) Provide specific recommendations on the design and conduct of short-term clinical trials, including advice regarding:
- feasibility
- single multi-arm trial versus several independent trials conducted with or without central coordination
- primary and secondary endpoints
- study population
- optimal duration of follow-up
- ranking of potential interventions
- biologic mechanisms to be assessed with consideration given to feasibility, participant burden and cost
- other issues that may result from the group’s discussion;
b) Recommendations regarding potential use of existing patient and population cohorts and databases; and
c) Recommendations regarding potential animal and in-vitro studies.

Back to top


a) Specific recommendations regarding short-term clinical trials:

  • Short-term clinical trials are feasible and have a high likelihood of informing both science and clinical practice.
  • A single multi-arm, parallel design randomized trial, with a thiazide-type diuretic alone as a control arm would be ideal.
  • Primary endpoint: serum glucose level following 3 months of follow-up was considered most practical; recommended minimum detectable effect-size: 5 mg/dL difference between the control and each of the intervention arms.
  • Outcomes:
    • Fasting serum glucose, serum potassium, serum Mg – at baseline and at 2, 4, 6, 8, and 12 weeks
    • 2-hour glucose tolerance test, serum insulin (fasting, 30 min, 2 hour), serum renin, serum triglycerides and free fatty acids at baseline, 4 weeks, and 12 weeks
    • Hemoglobin A1c at baseline and 12 weeks
    • Urine potassium and magnesium, timed specimens
    • Incident diabetes
    • Blood pressure and body weight (BMI) at baseline and at 2, 4, 6, 8, and 12 weeks
  • Study population: hypertensive individuals 50 years or older, and with a fasting blood sugar (FBS) 90 – 120 mg/dL (50% with FBS≥100mg), 20% over age 65, 50% black; appropriate representation of Hispanics, women, and overweight individuals
  • Intervention:
    • Wash-out: 4 weeks (diet, and a CCB if needed for BP control)
    • Treatment arms: 1) thiazide-type diuretic alone (control arm); 2) thiazide-type diuretic plus a potassium-sparing diuretic (amiloride 10-20 mg); 3) thiazide-type diuretic plus an ARB or an ACE-inhibitor; 4) CCB (amlodipine 5-10 mg) alone - as a metabolically neutral control
    • Dose of the thiazide-type diuretic should be an equivalent of 12.5 – 25 mg of chlorthalidone
    • Life-style advice for all participants
    • CCB as needed for BP control in all arms except for the CCB-alone arm; reserpine or clonidine to be used in the CCB-alone arm
    • Post-intervention wash-out to test for persistence of the effects should be considered
    • Safety concerns, including those related to impaired renal functions can be addressed by appropriate exclusion criteria and via in-trial monitoring

b) Use of existing population cohorts:

  • In analyses of observational studies, consider separating thiazide-type diuretics from other drugs; include potassium supplements in the analyses as well as BMI and potassium levels.
  • Several observational studies analyzed the relationship between the use of diuretics and the occurrence of diabetes. Their conclusions were mixed, including no relationship between diuretics and incident diabetes in the Atherosclerosis Research in Communities (ARIC) study, a well-conducted large observational study. One important difference across the studies was the type of control group employed; another was the definition of diabetes. ARIC collected information on self-reported diabetes and also diagnosed diabetes using biochemical parameters. As a result, the investigators were in a position to analyze data comparing self-reported diabetes with diabetes diagnosed using various biochemical parameters. These types of analyses could inform both the interpretation of published reports and future epidemiological research.
  • Consider designing an experimental study of a challenge/response maneuver (thiazide-type diuretic challenge and serum glucose response) nested within a population-based or an outpatient clinical cohort. Individuals who respond to the challenge with elevated glucose levels (responders) and those who do not (non-responders) would be then followed for several years for drug treatment history (especially for hypertension and diabetes) and clinical outcomes.
  • In the future, use ACCORD data to estimate CVD benefit related to changes in HbA1c.
  • Evaluate extended follow-up of clinical trial cohorts such as ALLHAT to determine long-term clinical impact of thiazide-type diuretics.

c) Recommendations regarding animal and in-vitro studies:

  • Results of this multi-arm clinical trial will guide animal experimentation, which will help identify and quantify mechanisms that underlie the relationships that are established in the clinical trial.
  • The overarching questions for mechanistic research are whether thiazides independently cause or exacerbate hyperglycemia and whether the degree of hypokalemia caused by diuretic therapy can cause impaired insulin secretion. In addition, other mechanisms involving the sympathetic nervous and the renin-angiotensin systems, the related decrease in blood pressure and the effects of potassium on vascular smooth muscle membrane potentials may all contribute to or drive the hypothesized benefits of potassium supplementation. From the clinical application perspective, it raises the need for better understanding of issues related to co-administration of drugs affecting these complex and interrelated processes.
  • Research is needed to guide selection of animal models given inter-species differences. To-date animal studies focused on slow potassium depletion with rapid repletion. However, studies are needed to understand the effects of chronic (repletion and depletion) changes in potassium levels on glucose metabolism, especially on insulin release. As in humans, studies should be conducted in two settings--normal versus obese (or other stressed-system/metabolic syndrome equivalent) models.
  • There is a need to understand changes in fasting versus postprandial insulin, effects of potassium supplementation on potassium transport, and effects of insulin and other factors on skeletal muscle uptake of potassium versus glucose.
  • There is a need to better understand the complex relationships between chronic low potassium levels and the renin-angiotensin system, volume and the sympathetic nervous system, and the interrelation between flow to the skeletal muscle and the metabolic rate. It is not known if decreases in blood pressure may cause hyperglycemia, and the role of epinephrine needs to be elucidated.

Back to top

Summary Recommendation

The group enthusiastically endorsed development of a clinical trial initiative as outlined above as an important public health priority which will address the following elements of the NHLBI Strategic Plan:

  • Goal 2; Challenge 2.1: To promote translation of clinical research findings back to the laboratory
  • Goal 2; Challenge 2.4: To enhance the evidence available to guide the practice of medicine and improve public health
  • Goal 3; Challenge 3.3: To promote the development and implementation of evidence-based guidelines

With respect to mechanistic and observational studies, the group recommended that these be pursued as investigator-initiated projects, future initiatives, or as a part of ongoing projects.

Back to top

May 2008
Skip footer links and go to content
Twitter iconTwitterExternal link Disclaimer         Facebook iconFacebookimage of external link icon         YouTube iconYouTubeimage of external link icon         Google+ iconGoogle+image of external link icon