Treatment of Hypertension in Patients with Chronic Kidney Disease

Login or register to view PDF.
Abstract

Chronic kidney disease (CKD) is increasingly prevalent in the US. At least 85% of patients with stage 3 CKD or greater have hypertension. Goal blood pressure is 130/80mmHg or less in most patients with CKD. A lower goal blood pressure of 125/75mmHg should be aimed for in CKD patients with significant proteinuria. Most CKD patients will require at least three or four medications, if not more, to achieve these goal blood pressures. Dosages, particularly of diuretics, may need to be adjusted in subjects with CKD. Side effects of medication such as an acute increase in creatinine or hyperkalemia are more frequent in this population and clinical vigilance with appropriate biochemical monitoring is necessary. Patients with CKD are more likely to die from cardiovascular disease before requiring dialysis or a kidney transplant. As a result, CKD patients should be considered at high risk for cardiovascular disease. Aggressive management of blood pressure in this population is vital.

Disclosure
Debbie L Cohen, MD, provides speaker services for Boehringer Ingelheim. Raymond R Townsend, MD, has provided consultancy services for GSK, NiCox, and Roche and has received research grants from the National Institutes of Health and Novartis.
Correspondence
Debbie L Cohen, MD, University of Pennsylvania, Renal, Electrolyte, and Hypertension Division, 1 Founders Building, 3400 Spruce Street, Philadelphia, PA 19104. E: debbie.cohen@uphs.upenn.edu
Received date
23 July 2009
Accepted date
16 September 2009

Chronic kidney disease (CKD) is increasingly prevalent, with an estimated 26 million adults with CKD in the US.1 Hypertension is the most common comorbidity in chronic kidney disease. At least 85% of patients with stage 3 CKD or greater have hypertension, making parenchymal kidney disease the most common ‘secondary’ form of hypertension. Treatment of hypertension can often be challenging, as these patients often have severe hypertension requiring the use of multiple medications to achieve target blood pressure (BP) goals. Target BP goals are lower in patients with CKD than in the general population. Ideally, BP should be less than 130/80mmHg and less than 125/75mmHg if patients also have significant proteinuria (>1g/ 24 hours).

How Common Is Hypertension in Chronic Kidney Disease?

Hypertension is very common in CKD, with more than 80% of CKD patients having coexistent hypertension. Patients with more severe CKD are more likely to have more severe hypertension2 that is more difficult to control, requiring a greater number of medications. Conversely, patients with more severe hypertension are more likely to develop CKD.3 The type of renal disease also influences the likelihood of hypertension, and classically tubulo-interstitial diseases have had less prevalence of elevated BP than glomerular diseases.4

Pathogenesis of Hypertension in Chronic Kidney Disease

A large and growing number of factors influence BP regulation in CKD, as shown in Table 1. Most of the increase in BP results from a subset of three primary systems, which include:

  • salt retention;
  • renin–angiotensin–aldosterone axis activation; and
  • sympathetic nervous system activation.

These factors are all potentially treatable and the physician should take these into account when selecting medications in CKD patients.

Salt Retention

Volume expansion is common in hypertension of CKD. As renal function declines, so does the ability to excrete sodium. If heart failure occurs, this adds to the challenge of maintaining euvolemia. Sodium increases BP, so as kidney function declines it does so to an even greater extent than simple volume expansion would predict at the lowest levels of kidney function.This finding suggests that the effect of salt intake on BP as further kidney function loss occurs is likely to be enhanced by the CKD milieu. Moreover, salt administration is well known to abet the pro-hypertensive effects of angiotensin-II6 and norepinephrine.7

Renin–Angiotensin System

The renin–angiotensin system plays a major role in hypertension, especially in patients with CKD. This is clear in the antihypertensive response to both ACE inhibitors and angiotensin II receptor blockers (ARBs) in patients with CKD. Aside from the hemodynamic consequences of renin activation, the excess angiotensin II produced probably contributes to progressive renal function loss and other target organ damage. It does this through its stimulation of aldosterone release, potentiation of the effects of various growth factors, and, in particular, its stimulating effects on the fibrogenic cytokine transforming growth factor-β.8,9

Sympathetic Nervous System

Some of the increase in renin system activation may be the result of sympathetic input into the juxtaglomerular apparatus through the β1-adrenoreceptor.6 The sympathetic nervous system also appears to be overactive in CKD.10 Like the renin–angiotension system, the kidney is both the source and the recipient of neurogenic activity.11 There is a fairly extensive network of sensory nerve fibers in the kidney, and many laboratory data indicate that sympathetic activation plays a role in hypertension in CKD through direct vascular constriction and the aforementioned interactions with renin and salt. The recent investigation showing that renal sympathetic nerve ablation improves BP control in drug-resistant hypertension is further evidence of the importance of this system.12 Several other systems are also active to a pathological degree in some patients with CKD. Among those amenable to potential drug treatment are endothelin13 and aldosterone.14 Finally, there is also some progress on the genetic front. Several rare phenotypes in which the kidney and hypertension are linked have been described, shedding light on important intra-renal BP regulation pathways.15 Aside from diagnostic value, there has been little benefit achieved to date from genetic studies with respect to guiding intervention.

Evaluation of Patients with Chronic Kidney Disease and Hypertension

The evaluation of patients with hypertension and CKD requires some additional consideration aside from the standard hypertensive work-up suggested by JNC-716, in the following ways:

  • In addition to the usual history taken for determining primary versus secondary hypertension, target organ damage, and presence of other cardiovascular risk factors,17 historical details should focus on prior drug therapies and why they were stopped. This should include issues such as drugs that caused hyperkalemia, worsening renal function, edema or dyspnea, and intolerable side effects.
  • In addition to the usual exam findings that may suggest secondary forms of hypertension, document the presence of rales, presence of an S3, pedal edema, and carotid bruits. These findings are more common in CKD patients, and their subsequent clinical course may guide diuretic management.
  • In addition to the standard lab tests (detailing target organ damage, suggesting secondary hypertension, or reflecting drug side effects), consider checking for presence/degree of anemia. Proteinuria should also be quantified since erythropoietin (which can raise BP) may be necessary and because the antihypertensive benefits and treatment goals of intervention are more impressive and lower, respectively, when proteinuria is present.18
Management of Hypertension in Chronic Kidney Disease

The importance of hypertension in CKD rests on two management principles. The first principle is that hypertension confers a substantial risk for heart disease, stroke, peripheral arterial disease, and further kidney failure.16 This risk is amplified when proteinuria is present.19 The second is that although hypertension ranks technically as the second most common cause of end-stage renal disease behind diabetes, it is clear that the majority of patients with diabetes and CKD also have hypertension. Therefore, the decision to lower BP in CKD is undertaken to preserve target organ function on several fronts. Moreover, it is more likely that a CKD patient will die from heart disease than reach end-stage renal disease.20 Therefore, it is critical to adequately manage hypertension in CKD, as this plays a central role in reducing the likelihood of developing cardiovascular disease.

Behavioral Recommendations

Restricting dietary sodium intake to 2,400mg/day is strongly recommended. So is weight loss (in the overweight) or weight management (in those not overweight), frequent physical activity, and restricting alcohol intake to a maximum of two drinks/day for men and one drink/day for women.16 The Dietary Approaches to Stop Hypertension (DASH) diet is often recommended.21 The greatest experience in this intervention category is with salt-intake reduction. In addition to helping to control BP, reducing salt intake can enable better reduce proteinuria when present22 and perhaps slow progressive kidney function loss.23

Drug Treatment Recommendations

The first step is defining the goal BP. This is recommended by both JNC-7 and the National Kidney Foundation as 130/80mmHg or less in patients with CKD.16 A lower BP of 125/75mmHg is recommended in patients with CKD and more than 1g of proteinuria/24 hours.24

Pathogenetic Mechanisms

Figure 1 represents a useful scaffold for sorting pharmacological therapy into a physiological approach to BP reduction. There are four basic inter- related mechanisms by which BP is regulated. Most drug therapies principally address one of these influences. Hypertension associated with CKD frequently requires treatment that addresses several of these mechanisms. It is not uncommon for patients to require three or four drugs, and sometimes even more, to achieve BP goals. BP goals are based on risk stratification from JNC-6:25

  • <140/90mmHg in lower-risk patients;
  • <130/80mmHg in higher-risk patients (those with diabetes, co-existing non-renal target organ damage, such as heart failure, stroke, etc.); and
  • <125/75mmHg in those with significant proteinuria.

At times patients show discordant BPs (for example 128/86mmHg in a higher-risk situation) that fall in between goals, in that systolic BP may be controlled but diastolic BP is not at goal. In clinical practice the authors preferentially aim for systolic control in most cases. Figure 2 presents a step-wise approach to using BP medications (sequentially adapted from multiple sources).24,26

Salt Retention

Diuretic treatment is a frequent mainstay of managing the salt volume excess component of hypertension in CKD. As glomerular filtration rate falls below about 30cc/minute, the effectiveness of hydrochlorothiazide as an antihypertensive wanes and the use of indapamide, metolazone, or a loop diuretic is recommended. When using generic loop diuretics, such as furosemide or bumetanide, the dosing is usually on a twice-a-day basis. This is because the duration of action is relatively short at six to eight hours and rebound sodium reabsorption may occur at the end of this time, mitigating the antihypertensive effect.

Dose adjustment for CKD may be required, particularly with loop diuretics in advanced CKD. Diuretics typically abet the antihypertensive effect of other classes of drug, such as angiotensin-converting enzyme (ACE) inhibitors, angiotensin-receptor blockers (ARBs), and drugs that block sympathetic nervous system components. If not used first, they are usually given as second-line agents.

Renin–Angiotensin System

The use of ACE inhibition in the hypertension of CKD is well accepted. A meta-analysis of the treatment of hypertension in 1,860 non-diabetic CKD patients demonstrated less renal disease progression when an ACE inhibitor was used, particularly in patients with more than 1,000mg of urinary protein loss/day.27 The putative benefits of ACE inhibition beyond BP may relate to their ability to selectively reduce efferent glomerular arteriolar tone more than afferent tone. This reduces glomerular hypertension and protein losses28 and is due to ACE inhibitor stimulation of bradykinin activity.29

There is less published information on the use of ARB therapy in non-diabetic CKD. This is a newer class of drug compared with ACE inhibitors; losartan was introduced in 1995, whereas captopril was launched in 1981. ARBs are clearly established as efficacious and beneficial in type 2 diabetic CKD.30,31 In general it is thought that ARBs provide similar BP reduction and kidney function preservation to ACE inhibitors,32 perhaps with less hyperkalemia.33

The brief review by Imai of nine ACE inhibitor versus ARB studies illustrates the lack of both sizable head-to-head comparisons in CKD and generally the equivalence in short-term end-points, such as proteinuria reduction and BP control.34

The recent ONTARGET study35 did not show a major BP reduction benefit with the combination of ACE inhibitors plus ARBs in patients with CKD and hypertension. There was also a small but significant decline in glomerular filtration rate and an increased incidence of hyperkalemia with combination therapy. Combination therapy, however, does reduce proteinuria in patients with CKD, but data are lacking to show whether this slows the progression of CKD in this population.

Sympathetic Nervous System

Several studies attest to the efficacy of beta-blocker therapy in hypertensive patients with CKD.36,37 The association of an increased occurrence of heart failure when β1-adrenoreceptor blocking agents are used has limited enthusiasm for the use of this class as first- or even second-line agents except in men with prostate hypertrophy symptoms. In the African American Study of Kidney Disease (AASK) trial of CKD progression there were no significant differences between the three drug groups (ramipril, metoprolol, and amlodipine) in glomerular filtration rate slope change over time.37 Given the significant role of the sympathetic nervous system in CKD progression and interim heart disease occurrence or progression, a strong case is made for their use in managing hypertension in CKD patients.38 A number of recent studies, however (Losartan Intervention For Endpoint [LIFE]39 reduction in hypertension study and the Anglo Scandinavian Cardiac Outcomes Trial [ASCOT]40), have shown adverse cardiovascular outcomes and greater incidence of stroke in high-risk hypertensive populations where atenolol was the beta-blocker of choice. These patients did not specifically have CKD and the data on atenolol may not be applicable in this population.

General Comments

Calcium-channel blocking drugs are commonly used in managing hypertension in CKD.41 Although much has been written about those that are dihydropyridine-based (nifedipine, amlodipine, etc.) compared with non-dihydropyridine (verapamil or diltiazem), there are no head-to-head trials to support substantial differences in hard outcomes such as CKD progression.42 Choosing an agent within this class is often guided by issues such as heart rate or concurrent beta-blocker use.42 They are efficacious in reducing BP in CKD and work well in a regimen using a diuretic and renin–angiotensin system inhibiting drug.43

When None of the Above Is Working

Several alternatives remain in the antihypertensive pharmacopeia. Minoxidil is sometimes used when a regimen of diuretic/renin– angiotensin blocker/calcium-channel blocker is already in place.44 Minoxidil appears to be better tolerated in the dialysis population. When monoxidil is used in patients with earlier stages of CKD, it should be prescribed together with a loop diuretic, such as furosemide, which should be taken twice daily. Beta-blocking drugs may need dose adjustments to offset the tachycardia associated with minoxidil therapy. Topical clonidine can be a very effective addition in a patient with CKD who continues to have uncontrolled BP.45 Clonidine should be used cautiously if the patient is on a beta-blocker because of added bradycardia. Topical administration is attractive as it enhances adherence in patients on multiple drugs. Combining an ACE inhibitor and an ARB is occasionally desirable when residual proteinuria remains,46,47 but this combination has not yet been shown to improve outcomes. Preliminary data with mineralocorticoid-receptor antagonists suggest they are beneficial, particularly in proteinuric chronic renal diseases.48 Their role in long-term outcomes in hypertensive treatment in patients with CKD is still largely undetermined. Their use in drug-resistant hypertension in non-CKD patients, however, is encouraging.49 Endothelin levels are elevated in many CKD cases and patients with uncontrolled BP may respond to an endothelin antagonist,13,50 although this class of drugs is technically only US Food and Drug Administration (FDA)-approved for pulmonary hypertension.

Dosing Caveats in Chronic Kidney Disease

Diuretic dosing has already been covered. Several beta-blockers (such as atenolol and nadolol) are predominantly excreted by the kidney, and their doses or dosing intervals can be altered to capitalize on this effect.51

What To Do When the Undesired Occurs
Hypotension

Drug dose reduction is usually needed if hypotension occurs. Typically hypotension will be more likely in volume depletion, positional change in older people, and in those with greater than expected renin system activation. Withholding and then reducing the dose of diuretic, alpha-adrenergic blocker or ACE/ARB (respectively) drug is a logical approach.

Creatinine Increase

Anticipate a small (<25%) creatinine increase with the use of an ACE inhibitor or an ARB as this may reflect reduction in glomerular hyperfiltration and is actually a welcome finding.52 Significant increases in creatinine (doubling, for example) are usually a sign of volume depletion, even in the absence of orthostasis, and it is empirically worth reducing the diuretic dose in this instance. In some cases a significant rise in creatinine on an ACE inhibitor or an ARB may signal underlying renal artery stenosis. Also look for intercurrent drug use (such as non-steroidal anti-inflammatory drugs [NSAIDs]), hypotension, and intercurrent illness, such as vomiting/diarrhea, which can challenge volume and increase creatinine levels.

Hyperkalemia

Look for dietary (high-potassium foods) or drug indiscretion (the use of potassium-sparing diuretics, salt substitutes, and NSAIDs).

Expectations, Drug Horizons, and Newer Technologies

The proportion of patients achieving antihypertensive goals in CKD remains a concern.2 The AASK trial shows that the frequency of patients achieving <130/80mmHg, at least among motivated research patients with non-diabetic CKD, can be in the range of 50–60%.53 Two lessons from the experience in treating hypertension in the non-diabetic CKD population should be kept in mind:

  • It is not just renal function that is at risk. The odds of dying from cardiovascular disease are many-fold that of reaching end-stage renal disease.20
  • Despite good treatment, CKD tends to progress as reported by the AASK cohort follow-up study.54 This needs to be interpreted in light of the point above; it also indicates a continued opportunity to ‘dig deeper’ into the pathophysiology of CKD progression, since addressing hypertension is only part of the solution.

Several drug classes are on the horizon or have recently been introduced that may prove valuable in treating hypertension in CKD. Direct renin inhibitors provide an alternative approach to renin system abrogation. Preliminary evidence indicates that they lower BP and reduce proteinuria in CKD,55 but their role in overall management of this population is still unclear. Third-generation beta-blockade with nebivolol holds the promise of both addressing sympathetic activation and stimulating nitric oxide production.56 Endothelin antagonists, as mentioned before, are another area of potential value in CKD. Darusentan appears to be in late-phase development and may be useful in managing those subjects with CKD who are drug-refractory. A CKD outcome trial evaluating the endothelin antagonist avosentan in diabetic nephropathy was terminated in December 2006, however, because of edema in those receiving the endothelin antagonist. Thus, more work is necessary in this class to determine its place in the clinical management of hypertension in CKD.

A Few Additional Points About Hypertension Management
  • Most medication changes will have their effect in about two weeks. This is thus a convenient scheduling time to re-check BP and any attendant lab tests. Sometimes it is prudent to check potassium earlier (clinical judgment situation).
  • Dietary sodium restriction does help; sometimes it is necessary to collect a 24-hour urine to document the excess salt intake.
  • Attention to cigarette use, if present, can make a substantial difference in CKD progression.57
  • Attention to alcohol consumption is important. Men should be encouraged to limit alcohol consumption to no more than drinks (1oz or 30ml of ethanol) per day, and women and lighter-weight persons to no more than one drink per day.
  • In truly resistant hypertension (a three- or four-drug regimen in good doses, with at least one diuretic), consider renovascular disease, aldosterone excess (aldosterone:renin ratio) and obstructive sleep apnea, as each of these can be addressed specifically.
  • Usual follow-up intervals in stable patients are three to six months.
  • If home BP monitoring is carried out, have the patient demonstrate at least once in your office how they actually do it. You are able to check their BP; the question is whether they can do it correctly.42
  • When altering a diuretic dose, either down or up, the authors strongly recommend having daily weights taken for at least one week afterwards, with instructions for the patient to call if they lose or gain two pounds a day two days in a row.

Finally, some new ways to evaluate the vascular basis of target organ protection are afforded by using antihypertensive medications in CKD that involve the employment of technologies evaluating arterial stiffness and vascular compliance. These have proved useful in end-stage renal disease populations.58 Prospective cohort studies determining the utility of such technologies in CKD are in progress, but it is pre-emptive to incorporate these techniques into routine clinical care now.59

References
  1. Coresh J SE, et al., JAMA, 2007;298:2038–47.
    Crossref | PubMed
  2. Coresh J, et al., Arch Intern Med, 2001;161:1207–16.
    Crossref
  3. Hsu CY, et al., Arch Intern Med, 2005;165:923–8.
    Crossref | PubMed
  4. Brown MA, et al., J Hypertens, 1992;10:701–12.
    PubMed
  5. Koomans HA, et al., Hypertension, 1982;4:190–97.
    Crossref | PubMed
  6. Schalekamp MA, et al., Clin Sci Mol Med, 1973;45:417–28.
    Crossref | PubMed
  7. Ito Y, et al., Clin Exp Hypertens A, 1989;11(Suppl. 1):363–70.
    Crossref | PubMed
  8. Struthers AD, et al., Cardiovasc Res, 2004;61:663–70.
    Crossref | PubMed
  9. Mehta PK, et al., Am J Physiol Cell Physiol, 2007;292:C82–97.
    Crossref | PubMed
  10. Klein IH Ligtenberg G, et al., J Am Soc Nephrol, 2003;14:3239–44.
    Crossref | PubMed
  11. Campese VM, Kidney Int, 2006;69:967–73.
    Crossref | PubMed
  12. Krum H, et al., Lancet, 2009;373:1275–81.
    Crossref | PubMed
  13. Dhaun N Goddard J, et al., J Am Soc Nephrol, 2006;17:943–55.
    Crossref | PubMed
  14. Del Vecchio L, et al., Nat Clin Pract Nephrol, 2007;3:42–9.
    Crossref | PubMed
  15. Lifton RP, Harvey Lect, 2004–2005;100:71–101.
    PubMed
  16. Chobanian AV Bakris GL, et al., Hypertension, 2003;42:1206–52.
    Crossref | PubMed
  17. Mohler ER, III, et al., Advanced Therapy in Hypertension and Vascular Disease, 1st ed, Hamilton, Ontario: BC Decker, 2006.
  18. De Zeeuw D, et al., Kidney Int Suppl, 2005;(98):S25–S29.
    Crossref | PubMed
  19. Mulrow CD, et al., Ann Intern Med, 2003;139:296–8.
    Crossref | PubMed
  20. Foley RN, et al., J Am Soc Nephrol, 2005;16:489–95.
    Crossref | PubMed
  21. Appel LJ, et al., N Engl J Med, 1997;336:1117–24.
    Crossref | PubMed
  22. Jones-Burton C, et al., Am J Nephrol, 2006;26:268–75.
    Crossref | PubMed
  23. Weir MR, et al., Am J Kidney Dis, 2005;45:176–88.
    Crossref | PubMed
  24. Bakris GL, et al., Am J Kidney Dis, 2000;36:646–61.
    Crossref | PubMed
  25. Arch Intern Med, 1997;157:2413–46.
    Crossref | PubMed
  26. Schieppati A, et al., Kidney Int, 2003;64:1947–55.
    Crossref | PubMed
  27. Jafar TH, et al., Ann Intern Med, 2003;139:244–52.
    Crossref | PubMed
  28. Anderson S, et al., Am J Hypertens, 1988;1:380S–383S.
    PubMed
  29. Schanstra JP, et al., Int J Mol Med, 1999;3:185–91.
    Crossref | PubMed
  30. Brenner BM, et al., N Engl J Med, 2001;345:861–9.
    Crossref | PubMed
  31. Parving HH, et al., N Engl J Med, 2001;345:870–78.
    Crossref | PubMed
  32. Parvanova A Chiurchiu C, et al., Expert Opin Pharmacother, 2005;6:1931–42.
    Crossref | PubMed
  33. Bakris GL, et al., Kidney Int, 2000;58:2084–92.
    Crossref | PubMed
  34. Imai E, Intern Med, 2006;45:179–81.
    Crossref | PubMed
  35. YS, Teo KK, et al., N Engl J Med, 2008;358:1547–59.
    Crossref | PubMed
  36. Kohno M, et al., Drugs, 1988;36(Suppl. 6):129–35.
    Crossref | PubMed
  37. Wright JT Jr, et al., JAMA, 2002;288:2421–31.
    Crossref | PubMed
  38. Bakris GL,et al., Kidney Int, 2006;70:1905–13.
    Crossref | PubMed
  39. Lindholm LH, et al., Lancet, 2002;359(9311).
  40. Dahlöf B, et al., Lancet, 2005;366(9489):895–906.
    Crossref | PubMed
  41. Segura J, et al., J Am Soc Nephrol, 2005;16(Suppl. 1): S64–6.
    Crossref | PubMed
  42. Townsend RR, J Clin Hypertens (Greenwich), 2000;2:288–9.
    PubMed
  43. Tobe S, et al., Curr Hypertens Rep, 2002;4:191–4.
    Crossref | PubMed
  44. Toto RD, et al., et al., Kidney Int, 1995;48:851–9.
    Crossref | PubMed
  45. Lowenthal DT, et al., Clin Nephrol, 1993;39:37–43.
    PubMed
  46. Nakao N, et al., Lancet, 2003;361:117–24.
    Crossref | PubMed
  47. Fernandez-Juárez G, et al., J Am Soc Nephrol, 2006;17 (Suppl. 3):S250–54.
    Crossref | PubMed
  48. Epstein M, Am J Med, 2006;119:912–19.
    Crossref | PubMed
  49. Calhaun DA, J Clin Hypertens (Greenwich), 2007;9:19–24.
    PubMed
  50. Richter CM, Rheumatology (Oxford), 2006;45(Suppl. 3): iii36–8.
    Crossref | PubMed
  51. Borchard U, Clin Physiol Biochem, 1990;8(Suppl. 2):28–34.
    PubMed
  52. Bakris GL, et al., Arch Intern Med, 2000;160:685–93.
    PubMed
  53. Wright JT Jr, et al.,Arch Intern Med, 2002;162:1636–43.
    Crossref | PubMed
  54. Appel L, et al., J Am Soc Nephrol, 2006;17:54–5.
  55. Staessen JA, et al., Lancet, 2006;368:1449–56.
    Crossref | PubMed
  56. Kalinowski L, et al., Circulation, 2003;107:2747–52.
    Crossref | PubMed
  57. Regalado M, et al., Am J Kidney Dis, 2000;35:687–94.
    Crossref | PubMed
  58. Guerin AP, et al., Circulation, 2001;103:987–92.
    Crossref | PubMed
  59. Townsend RR, Curr Opin Nephrol Hypertens, 2008;17:93–8.
    Crossref | PubMed