Cleveland Clinic Journal of Medicine

Great effort has been spent on identifying easily measured biomarkers to predict the progression of coronary disease and chronic kidney disease (CKD). Interestingly, these two disease processes seem to share some biomarkers and perhaps some pathogenic mechanisms. An ultimate hope is that some of these markers will be found to also contribute directly to organ dysfunction and be amenable to therapy. Blood pressure and (in many people’s minds) low-density lipoprotein cholesterol fulfill this hope. The jury remains out on C-reactive protein and serum urate. There are others.

In this issue of the Journal, Stephen et al review the data indicating that albuminuria helps predict the progression of CKD, coronary disease, ventricular remodeling, and, in some studies, all-cause mortality. Proteinuria has generally been assumed to be a marker of renal injury, but, the authors point out, albumin can under some circumstances initiate inflammatory mechanisms and stimulate mediators of fibrosis.

Although not mentioned by Stephen et al, albumin (like hemoglobin) is susceptible to nonenzymatic glycosylation in patients with diabetes. There is a hint in the literature that glycosylated albumin may be preferentially excreted. Its effects on various tissues are incompletely studied, but it strikes me that perhaps this molecule plays a unique pathogenic role in diabetic renal and vascular disease, even more than native albumin. Further evaluation of this specific marker may lead to even stronger associations (although in a select population of patients with poorly controlled diabetes).

The focus on urine as a fluid with diagnostic and predictive characteristics is certainly not new. Both Hippocrates and Galen recognized the value of inspecting urine. Uroscopy (now urinalysis) may be the oldest surviving laboratory test. Recently, my friend Joe Nally, a coauthor with Stephen et al, shared with me a paper detailing the romantic yet checkered history of urinalysis.¹

Gilles de Corbeil in the 12th century wrote a poem on judging urine, intending it as an aid for remembering the supposed 20 different diagnostic colors of urine and describing in detail the use of the urine flask, a bladder-shaped container for studying the partitioning of the urine colors and substance as representative of the diseased parts of the body. A urine flask was even illustrated in a version of Chaucer’s Canterbury Tales as a recognized accoutrement of the stylish physician (Figure 1). The “art” of uroscopy grew so successful over the centuries as a component of rampant medical charlatanry (casting no aspersions, of course, on current nephrologists) that the Royal College of Physicians in 1601 felt pressed to attack the “pisse-mongers” by stating, “It is ridiculous and foolish to divine the…course of disease…from the inspection of urine.”¹ This dictate was apparently ignored then, but seemingly is too frequently followed by clinicians today, contributing to the oft-delayed diagnosis of glomerulonephritis and other renal diseases.

In 1637, Thomas Brian published The Pisse-Prophet or Certaine Pisse Pot Lectures, in which he railed against the witchcraft of uroscopy, which he said should only be performed by university-trained physicians. Jump forward to 1827, when Richard Bright elegantly described acute glomerulonephritis, although not the microscopic findings that would be illustrated in accurate detail by Golding Bird in his 1844 treatise, Urinary Deposits. Sitting on the bookshelf behind my desk is a copy of Richard W. Lippman’s Urine and Urinary Sediment: A Practical Manual and Atlas (1957). I have no urine flask—rheumatologists know their limitations.

As we enter 2014, all of us at the Journal offer you our sincere wishes for a personally healthy and universally peaceful new year.

Great effort has been spent on identifying easily measured biomarkers to predict the progression of coronary disease and chronic kidney disease (CKD). Interestingly, these two disease processes seem to share some biomarkers and perhaps some pathogenic mechanisms. An ultimate hope is that some of these markers will be found to also contribute directly to organ dysfunction and be amenable to therapy. Blood pressure and (in many people’s minds) low-density lipoprotein cholesterol fulfill this hope. The jury remains out on C-reactive protein and serum urate. There are others.

In this issue of the Journal, Stephen et al review the data indicating that albuminuria helps predict the progression of CKD, coronary disease, ventricular remodeling, and, in some studies, all-cause mortality. Proteinuria has generally been assumed to be a marker of renal injury, but, the authors point out, albumin can under some circumstances initiate inflammatory mechanisms and stimulate mediators of fibrosis.

Although not mentioned by Stephen et al, albumin (like hemoglobin) is susceptible to nonenzymatic glycosylation in patients with diabetes. There is a hint in the literature that glycosylated albumin may be preferentially excreted. Its effects on various tissues are incompletely studied, but it strikes me that perhaps this molecule plays a unique pathogenic role in diabetic renal and vascular disease, even more than native albumin. Further evaluation of this specific marker may lead to even stronger associations (although in a select population of patients with poorly controlled diabetes).

The focus on urine as a fluid with diagnostic and predictive characteristics is certainly not new. Both Hippocrates and Galen recognized the value of inspecting urine. Uroscopy (now urinalysis) may be the oldest surviving laboratory test. Recently, my friend Joe Nally, a coauthor with Stephen et al, shared with me a paper detailing the romantic yet checkered history of urinalysis.¹

Gilles de Corbeil in the 12th century wrote a poem on judging urine, intending it as an aid for remembering the supposed 20 different diagnostic colors of urine and describing in detail the use of the urine flask, a bladder-shaped container for studying the partitioning of the urine colors and substance as representative of the diseased parts of the body. A urine flask was even illustrated in a version of Chaucer’s Canterbury Tales as a recognized accoutrement of the stylish physician (Figure 1). The “art” of uroscopy grew so successful over the centuries as a component of rampant medical charlatanry (casting no aspersions, of course, on current nephrologists) that the Royal College of Physicians in 1601 felt pressed to attack the “pisse-mongers” by stating, “It is ridiculous and foolish to divine the…course of disease…from the inspection of urine.”¹ This dictate was apparently ignored then, but seemingly is too frequently followed by clinicians today, contributing to the oft-delayed diagnosis of glomerulonephritis and other renal diseases.

In 1637, Thomas Brian published The Pisse-Prophet or Certaine Pisse Pot Lectures, in which he railed against the witchcraft of uroscopy, which he said should only be performed by university-trained physicians. Jump forward to 1827, when Richard Bright elegantly described acute glomerulonephritis, although not the microscopic findings that would be illustrated in accurate detail by Golding Bird in his 1844 treatise, Urinary Deposits. Sitting on the bookshelf behind my desk is a copy of Richard W. Lippman’s Urine and Urinary Sediment: A Practical Manual and Atlas (1957). I have no urine flask—rheumatologists know their limitations.

As we enter 2014, all of us at the Journal offer you our sincere wishes for a personally healthy and universally peaceful new year.

Great effort has been spent on identifying easily measured biomarkers to predict the progression of coronary disease and chronic kidney disease (CKD). Interestingly, these two disease processes seem to share some biomarkers and perhaps some pathogenic mechanisms. An ultimate hope is that some of these markers will be found to also contribute directly to organ dysfunction and be amenable to therapy. Blood pressure and (in many people’s minds) low-density lipoprotein cholesterol fulfill this hope. The jury remains out on C-reactive protein and serum urate. There are others.

In this issue of the Journal, Stephen et al review the data indicating that albuminuria helps predict the progression of CKD, coronary disease, ventricular remodeling, and, in some studies, all-cause mortality. Proteinuria has generally been assumed to be a marker of renal injury, but, the authors point out, albumin can under some circumstances initiate inflammatory mechanisms and stimulate mediators of fibrosis.

Although not mentioned by Stephen et al, albumin (like hemoglobin) is susceptible to nonenzymatic glycosylation in patients with diabetes. There is a hint in the literature that glycosylated albumin may be preferentially excreted. Its effects on various tissues are incompletely studied, but it strikes me that perhaps this molecule plays a unique pathogenic role in diabetic renal and vascular disease, even more than native albumin. Further evaluation of this specific marker may lead to even stronger associations (although in a select population of patients with poorly controlled diabetes).

The focus on urine as a fluid with diagnostic and predictive characteristics is certainly not new. Both Hippocrates and Galen recognized the value of inspecting urine. Uroscopy (now urinalysis) may be the oldest surviving laboratory test. Recently, my friend Joe Nally, a coauthor with Stephen et al, shared with me a paper detailing the romantic yet checkered history of urinalysis.¹

Gilles de Corbeil in the 12th century wrote a poem on judging urine, intending it as an aid for remembering the supposed 20 different diagnostic colors of urine and describing in detail the use of the urine flask, a bladder-shaped container for studying the partitioning of the urine colors and substance as representative of the diseased parts of the body. A urine flask was even illustrated in a version of Chaucer’s Canterbury Tales as a recognized accoutrement of the stylish physician (Figure 1). The “art” of uroscopy grew so successful over the centuries as a component of rampant medical charlatanry (casting no aspersions, of course, on current nephrologists) that the Royal College of Physicians in 1601 felt pressed to attack the “pisse-mongers” by stating, “It is ridiculous and foolish to divine the…course of disease…from the inspection of urine.”¹ This dictate was apparently ignored then, but seemingly is too frequently followed by clinicians today, contributing to the oft-delayed diagnosis of glomerulonephritis and other renal diseases.

In 1637, Thomas Brian published The Pisse-Prophet or Certaine Pisse Pot Lectures, in which he railed against the witchcraft of uroscopy, which he said should only be performed by university-trained physicians. Jump forward to 1827, when Richard Bright elegantly described acute glomerulonephritis, although not the microscopic findings that would be illustrated in accurate detail by Golding Bird in his 1844 treatise, Urinary Deposits. Sitting on the bookshelf behind my desk is a copy of Richard W. Lippman’s Urine and Urinary Sediment: A Practical Manual and Atlas (1957). I have no urine flask—rheumatologists know their limitations.

As we enter 2014, all of us at the Journal offer you our sincere wishes for a personally healthy and universally peaceful new year.

“One can obtain considerable information concerning the general health by examining the urine.” 
—Hippocrates (460?–355? BCE)

Chronic kidney disease is a notable public health concern because it is an important risk factor for end-stage renal disease, cardiovascular disease, and death. Its prevalence¹ exceeds 10% and is considerably higher in high-risk groups, such as those with diabetes or hypertension, which are growing in the United States.

While high levels of total protein in the urine have always been recognized as pathologic, a growing body of evidence links excretion of the protein albumin to adverse cardiovascular outcomes, and most international guidelines now recommend measuring albumin specifically. Albuminuria is a predictor of declining renal function and is independently associated with adverse cardiovascular outcomes. Thus, clinicians need to detect it early, manage it effectively, and reduce concurrent risk factors for cardiovascular disease.

Therefore, this review will focus on albuminuria. However, because the traditional standard for urinary protein measurement was total protein, and because a few guidelines still recommend measuring total protein rather than albumin, we will also briefly discuss total urinary protein.

MOST URINARY PROTEIN IS ALBUMIN

Most of the protein in the urine is albumin filtered from the plasma. Less than half of the rest is derived from the distal renal tubules (uromodulin or Tamm-Horsfall mucoprotein), ² and urine also contains a small and varying proportion of immunoglobulins, low-molecular-weight proteins, and light chains.

Normal healthy people lose less than 30 mg of albumin in the urine per day. In greater amounts, albumin is the major urinary protein in most kidney diseases. Other proteins in urine can be specific markers of less-common illnesses such as plasma cell dyscrasia, glomerulopathy, and renal tubular disease.

MEASURING PROTEINURIA AND ALBUMINURIA

Albumin is not a homogeneous molecule in urine. It undergoes changes to its molecular configuration in the presence of certain ions, peptides, hormones, and drugs, and as a result of proteolytic fragmentation both in the plasma and in renal tubules.³ Consequently, measuring urinary albumin involves a trade-off between convenience and accuracy.

A 24-hour timed urine sample has long been the gold standard for measuring albuminuria, but the collection is cumbersome and time-consuming, and the test is prone to laboratory error.

Dipstick measurements are more convenient and are better at detecting albumin than other proteins in urine, but they have low sensitivity and high interobserver variation.^3–5

The albumin-to-creatinine ratio (ACR). As the quantity of protein in the urine changes with time of day, exertion, stress level, and posture, spot-checking of urine samples is not as good as timed collection. However, a simultaneous measurement of creatinine in a spot urine sample adjusts for protein concentration, which can vary with a person’s hydration status. The ACR so obtained is consistent with the 24-hour timed collection (the gold standard) and is the recommended method of assessing albuminuria.³ An early morning urine sample is favored, as it avoids orthostatic variations and varies less in the same individual.

In a study in the general population comparing the ACR in a random sample and in an early morning sample, only 44% of those who had an ACR of 30 mg/g or higher in the random sample had one this high in the early morning sample.⁶ However, getting an early morning sample is not always feasible in clinical practice. If you are going to measure albuminuria, the Kidney Disease Outcomes and Quality Initiative⁷ suggests checking the ACR in a random sample and then, if the test is positive, following up and confirming it within 3 months with an early morning sample.

Also, since creatinine excretion differs with race, diet, and muscle mass, if the 24-hour creatinine excretion is not close to 1 g, the ACR will give an erroneous estimate of the 24-hour excretion rate.³

Table 1 compares the various methods of measuring protein in the urine.^3,5,8,9 Of note, methods of measuring albumin and total protein vary considerably in their precision and accuracy, making it impossible to reliably translate values from one to the other.⁵

National and international guidelines (Table 2)^7,10–13 agree that albuminuria should be tested in diabetic patients, as it is a surrogate marker for early diabetic nephropathy.^3,13 Most guidelines also recommend measuring albuminuria by a urine ACR test as the preferred measure, even in people without diabetes.

Also, no single cutoff is universally accepted for distinguishing pathologic albuminuria from physiologic albuminuria, nor is there a universally accepted unit of measure.¹⁴ Because approximately 1 g of creatinine is lost in the urine per day, the ACR has the convenient property of numerically matching the albumin excretory rate expressed in milligrams per 24 hours. The other commonly used unit is milligrams of albumin per millimole of creatinine; 30 mg/g is roughly equal to 3 mg/mmol.

The term microalbuminuria was traditionally used to refer to albumin excretion of 30 to 299 mg per 24 hours, and macroalbuminuria to 300 mg or more per 24 hours. However, as there is no pathophysiologic basis to these thresholds (see outcomes data below), the current Kidney Disease Improving Global Outcomes (KDIGO) guidelines do not recommend using these terms.^13,15

RENAL COMPLICATIONS OF ALBUMINURIA

A failure of the glomerular filtration barrier or of proximal tubular reabsorption accounts for most cases of pathologic albuminuria.¹⁶ Processes affecting the glomerular filtration of albumin include endothelial cell dysfunction and abnormalities with the glomerular basement membrane, podocytes, or the slit diaphragms among the podocytic processes.¹⁷

Ultrafiltrated albumin has been directly implicated in tubulointerstitial damage and glomerulosclerosis through diverse pathways. In the proximal tubule, albumin up-regulates interleukin 8 (a chemoattractant for lymphocytes and neutrophils), induces synthesis of endothelin 1 (which stimulates renal cell proliferation, extracellular matrix production, and monocyte attraction), and causes apoptosis of tubular cells.¹⁸ In response to albumin, proximal tubular cells also stimulate interstitial fibroblasts via paracrine release of transforming growth factor beta, either directly or by activating complement or macrophages.^18,19

Studies linking albuminuria to kidney disease

Albuminuria has traditionally been associated with diabetes mellitus as a predictor of overt diabetic nephropathy,^20,21 although in type 1 diabetes, established albuminuria can spontaneously regress.²²

Albuminuria is also a strong predictor of progression in chronic kidney disease.²³ In fact, in the last decade, albuminuria has become an independent criterion in the definition of chronic kidney disease; excretion of more than 30 mg of albumin per day, sustained for at least 3 months, qualifies as chronic kidney disease, with independent prognostic implications (Table  3).¹³

Astor et al,²⁴ in a meta-analysis of 13 studies with more than 21,000 patients with chronic kidney disease, found that the risk of end-stage renal disease was three times higher in those with albuminuria.

Gansevoort et al,²³ in a meta-analysis of nine studies with nearly 850,000 participants from the general population, found that the risk of end-stage renal disease increased continuously as albumin excretion increased. They also found that as albuminuria increased, so did the risk of progression of chronic kidney disease and the incidence of acute kidney injury.

Hemmelgarn et al,²⁵ in a large pooled cohort study with more than 1.5 million participants from the general population, showed that increasing albuminuria was associated with a decline in the estimated glomerular filtration rate (GFR) and with progression to end-stage renal disease across all strata of baseline renal function. For example, in persons with an estimated GFR of 60 mL/min/1.73 m2

• 1 per 1,000 person-years for those with no proteinuria
• 2.8 per 1,000 person-years for those with mild proteinuria (trace or 1+ by dipstick or ACR 30–300 mg/g)
• 13.4 per 1,000 person-years for those with heavy proteinuria (2+ or ACR > 300 mg/g).

Rates of progression to end-stage renal disease were:

• 0.03 per 1,000 person-years with no proteinuria
• 0.05 per 1,000 person-years with mild proteinuria
• 1 per 1,000 person-years with heavy proteinuria.²⁵

CARDIOVASCULAR CONSEQUENCES OF ALBUMINURIA

The exact pathophysiologic link between albuminuria and cardiovascular disease is unknown, but several mechanisms have been proposed.

One is that generalized endothelial dysfunction causes both albuminuria and cardiovascular disease.²⁶ Endothelium-derived nitric oxide has vasodilator, antiplatelet, antiproliferative, antiadhesive, permeability-decreasing, and anti-inflammatory properties. Impaired endothelial synthesis of nitric oxide has been independently associated with both microalbuminuria and diabetes.²⁷

Low levels of heparan sulfate (which has antithrombogenic effects and decreases vessel permeability) in the glycocalyx lining vessel walls may also account for albuminuria and for the other cardiovascular effects.^28–30 These changes may be the consequence of chronic low-grade inflammation, which precedes the occurrence and progression of both albuminuria and atherothrombotic disease. The resulting abnormalities in the endothelial glycocalyx could lead to increased glomerular permeability to albumin and may also be implicated in the pathogenesis of atherosclerosis.²⁶

In an atherosclerotic aorta and coronary arteries, the endothelial dysfunction may cause increased leakage of cholesterol and glycated end-products into the myocardium, resulting in increasing wall stiffness and left ventricular mass. A similar atherosclerotic process may account for coronary artery microthrombi, resulting in subendocardial ischemia that could lead to systolic and diastolic heart dysfunction.³¹

Studies linking albuminuria to heart disease

There is convincing evidence that albuminuria is associated with cardiovascular disease. An ACR between 30 and 300 mg/g was independently associated with myocardial infarction and ischemia.³² People with albuminuria have more than twice the risk of severe coronary artery disease, and albuminuria is also associated with increased intimal thickening in the carotid arteries.^33,34 An ACR in the same range has been associated with increased incidence and progression of coronary artery calcification.³⁵ Higher brachial-ankle pulse-wave velocity has also been demonstrated with albuminuria in a dose-dependent fashion.^36,37

An ACR of 30 to 300 mg/g has been linked to left ventricular hypertrophy independently of other risk factors,³⁸ and functionally with diastolic dysfunction and abnormal midwall shortening.³⁹ In a study of a subgroup of patients with diabetes from a population-based cohort of Native American patients (the Strong Heart Study),³⁹ the prevalence of diastolic dysfunction was:

• 16% in those with no albuminuria
• 26% in those with an ACR of 30 to 300 mg/g
• 31% in those with an ACR greater than 300 mg/g.

The association persisted even after controlling for age, sex, hypertension, and other covariates.

Those pathologic associations have been directly linked to clinical outcomes. For patients with heart failure (New York Heart Association class II–IV), a study found that albuminuria (an ACR > 30 mg/g) conferred a 41% higher risk of admission for heart failure, and an ACR greater than 300 mg/g was associated with an 88% higher risk.⁴⁰

In an analysis of a prospective cohort from the general population with albuminuria and a low prevalence of renal dysfunction (the Prevention of Renal and Vascular Endstage Disease study),⁴¹ albuminuria was associated with a modest increase in the incidence of the composite end point of myocardial infarction, stroke, ischemic heart disease, revascularization procedures, and all-cause mortality per doubling of the urine albumin excretion (hazard ratio 1.08, range 1.04 –1.12).

The relationship to cardiovascular outcomes extends below traditional lower-limit thresholds of albuminuria (corresponding to an ACR > 30 mg/g). A subgroup of patients from the Framingham Offspring Study without prevalent cardiovascular disease, hypertension, diabetes, or kidney disease, and thus with a low to intermediate probability of cardiovascular events, were found to have thresholds for albuminuria as low as 5.3 mg/g in men and 10.8 mg/g in women to discriminate between incident coronary artery disease, heart failure, cerebrovascular disease, other peripheral vascular disease, or death.⁴²

In a meta-analysis including more than 1 million patients in the general population, increasing albuminuria was associated with an increase in deaths from all causes in a continuous manner, with no threshold effect.⁴³ In patients with an ACR of 30 mg/g, the hazard ratio for death was 1.63, increasing to 2.22 for those with more than 300 mg/g compared with those with no albuminuria. A similar increase in the risk of myocardial infarction, heart failure, stroke, or sudden cardiac death was noted with higher ACR.⁴³

Important prospective cohort studies and meta-analyses related to albuminuria and kidney and cardiovascular disease and death are summarized in the eTable.^23,39–50

THE CASE FOR TREATING ALBUMINURIA

Reduced progression of renal disease

Treating patients who have proteinuric chronic kidney disease with an angiotensin-converting enzyme (ACE) inhibitor or an angiotensin receptor blocker (ARB) can reduce the risk of progression of renal failure. However, it is unclear whether this benefit is the result of treating concomitant risk factors independent of the reduction in albuminuria, and there is no consistent treatment effect in improving renal outcomes across studies.

Fink et al,⁵¹ in a meta-analysis, found that chronic kidney disease patients with diabetes, hypertension, and macroalbuminuria had a 40% lower risk of progression to end-stage renal disease if they received an ACE inhibitor (relative risk [RR] 0.60, 95% confidence interval [CI] 0.43–0.83). In the same meta-analysis, ARBs also reduced the risk of progression to end-stage renal disease (RR 0.77, 95% CI 0.66–0.90).

Jafar et al,⁵² in an analysis of pooled patient-level data including only nondiabetic patients on ACE inhibitor therapy (n = 1,860), found that the risk of progression of renal failure, defined as a doubling of serum creatinine or end-stage renal disease, was reduced (RR 0.70, 95% CI 0.55–0.88). Patients with higher levels of albuminuria showed the most benefit, but the effect was not conclusive for albuminuria below 500 mg/day at baseline.

Maione et al,⁵³ in a meta-analysis that included patients with albuminuria who were treated with ACE inhibitors vs placebo (n = 8,231), found a similar reduction in risk of:

• Progression to end-stage renal disease (RR 0.67, 95% CI 0.54–0.84)
• Doubling of serum creatinine (RR 0.62, 95% CI 0.46–0.84)
• Progression of albuminuria (RR 0.49, 95% CI 0.36–0.65)
• Normalization of pathologic albuminuria (as defined by the trialists in the individual studies) (RR 2.99, 95% CI 1.82–4.91).

Similar results were obtained for patients with albuminuria who were treated with ARBs.⁵³

ONTARGET.⁵⁴ In contrast, in the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial, the combination of an ACE inhibitor and an ARB showed no benefit in reducing the progression of renal failure, and in those patients with chronic kidney disease there was a higher risk of a doubling of serum creatinine or of the development of end-stage renal disease and hyperkalemia.

Also, in a pooled analysis of the ONTARGET and Telmisartan Randomized Assessment Study in ACE Intolerant Subjects With Cardiovascular Disease (TRANSCEND) trials, a 50% reduction in baseline albuminuria was associated with reduced progression of renal failure in those with a baseline ACR less than 10 mg/g.⁵⁵

Improved cardiovascular outcomes

There is also evidence of better cardiovascular outcomes with treatment of albuminuria. Again, it is uncertain whether this is a result of treating risk factors other than albuminuria with ACE inhibitors or ARBs, and there is no parallel benefit demonstrated across all studies.

LIFE.^47,48 In the Losartan Intervention for Endpoint Reduction in Hypertension trial, survival analyses suggested a decrease in risk of cardiovascular adverse events as the degree of proteinuria improved with ARB therapy.

Maione et al,⁵³ in a meta-analysis including 8,231 patients with albuminuria and at least one other risk factor, found a significant reduction in the rate of nonfatal cardiovascular outcomes (angina, myocardial infarction, revascularization, stroke, transient ischemic attack, or heart failure) with ACE inhibitors vs placebo (RR 0.88, CI 0.82–0.94) and also in 3,888 patients treated with ARBs vs placebo (RR 0.77, CI 0.61–0.98). However, the meta-analysis did not show that ACE inhibitor or ARB therapy reduced rate of cardiovascular or all-cause mortality.

Fink et al,⁵¹ in their meta-analysis of 18 trials of ACE inhibitors and four trials of ARBs, also found no evidence that ACE inhibitor or ARB therapy reduced cardiovascular mortality rates.³⁸

The ONTARGET trial evaluated the combination of an ACE inhibitor and ARB therapy in patients with diabetes or preexisting peripheral vascular disease. Reductions in the rate of cardiovascular disease or death were not observed, and in those with chronic kidney disease, there was a higher risk of doubling of serum creatinine or development of end-stage renal disease and adverse events of hyperkalemia.⁵⁶ And although an increase in baseline albuminuria was associated with worse cardiovascular outcomes, its reduction in the ONTARGET and TRANSCEND trials did not demonstrate better outcomes when the baseline ACR was greater than 10 mg/g.⁵⁵

WHO SHOULD BE TESTED?

The benefit of adding albuminuria to conventional cardiovascular risk stratification such as Framingham risk scoring is not conclusive. However, today’s clinician may view albuminuria as a biomarker for renal and cardiovascular disease, as albuminuria might be a surrogate marker for endothelial dysfunction in the glomerular capillaries or other vital vascular beds.

High-risk populations and chronic kidney disease patients

Nearly all the current guidelines recommend annual screening for albuminuria in patients with diabetes and hypertension (Table 2).^7,10–13 Other high-risk populations include people with cardiovascular disease, a family history of end-stage renal disease, and metabolic syndrome. Additionally, chronic kidney disease patients whose estimated GFR defines them as being in stage 3 or higher (ie, GFR < 60 mL/min/1.73m²), regardless of other comorbidities, should be tested for albuminuria, as it is an important risk predictor.

Most experts prefer that albuminuria be measured by urine ACR in a first morning voided sample, though this is not the only option.

Screening the general population

Given that albuminuria has been shown to be such an important prognosticator for patients at high risk and those with chronic kidney disease, the question arises whether screening for albuminuria in the asymptomatic low-risk general population would foster earlier detection and therefore enable earlier intervention and result in improved outcomes. However, a systematic review done for the United States Preventive Services Task Force and for an American College of Physicians clinical practice guideline did not find robust evidence to support this.⁵¹

OUR RECOMMENDATIONS

Who should be tested?

• Patients with chronic kidney disease stage 3, 4, or 5 (GFR < 60 mL/min/1.73m²) who are not on dialysis
• Patients who are considered at higher risk of adverse outcomes, such as those with diabetes, hypertension, a family history of end-stage renal disease, or cardiovascular disease. Testing is useful for recognizing increased renal and cardiovascular risk and may lead clinicians to prescribe or titrate a renin-angiotensin system antagonist, a statin, or both, or to modify other cardiovascular risk factors.
• Not recommended: routine screening in the general population who are asymptomatic or are considered at low risk.

Which test should be used?

Based on current evidence and most guidelines, we recommend the urine ACR test as the screening test for people with diabetes and others deemed to be at high risk.

What should be done about albuminuria?

• Controlling blood pressure is important, and though there is debate about the target blood pressure, an individualized plan should be developed with the patient based on age, comorbidities, and goals of care.
• An ACE inhibitor or ARB, if not contraindicated, is recommended for patients with diabetes whose ACR is greater than 30 mg/g and for patients with chronic kidney disease and an ACR greater than 300 mg/g.
• Current evidence does not support the combined use of an ACE inhibitor and an ARB, as proof of benefit is lacking and the risk of adverse events is higher.
• Refer patients with high or unexplained albuminuria to a nephrologist or clinic specializing in chronic kidney disease.

“One can obtain considerable information concerning the general health by examining the urine.” 
—Hippocrates (460?–355? BCE)

Chronic kidney disease is a notable public health concern because it is an important risk factor for end-stage renal disease, cardiovascular disease, and death. Its prevalence¹ exceeds 10% and is considerably higher in high-risk groups, such as those with diabetes or hypertension, which are growing in the United States.

While high levels of total protein in the urine have always been recognized as pathologic, a growing body of evidence links excretion of the protein albumin to adverse cardiovascular outcomes, and most international guidelines now recommend measuring albumin specifically. Albuminuria is a predictor of declining renal function and is independently associated with adverse cardiovascular outcomes. Thus, clinicians need to detect it early, manage it effectively, and reduce concurrent risk factors for cardiovascular disease.

Therefore, this review will focus on albuminuria. However, because the traditional standard for urinary protein measurement was total protein, and because a few guidelines still recommend measuring total protein rather than albumin, we will also briefly discuss total urinary protein.

MOST URINARY PROTEIN IS ALBUMIN

Most of the protein in the urine is albumin filtered from the plasma. Less than half of the rest is derived from the distal renal tubules (uromodulin or Tamm-Horsfall mucoprotein), ² and urine also contains a small and varying proportion of immunoglobulins, low-molecular-weight proteins, and light chains.

Normal healthy people lose less than 30 mg of albumin in the urine per day. In greater amounts, albumin is the major urinary protein in most kidney diseases. Other proteins in urine can be specific markers of less-common illnesses such as plasma cell dyscrasia, glomerulopathy, and renal tubular disease.

MEASURING PROTEINURIA AND ALBUMINURIA

Albumin is not a homogeneous molecule in urine. It undergoes changes to its molecular configuration in the presence of certain ions, peptides, hormones, and drugs, and as a result of proteolytic fragmentation both in the plasma and in renal tubules.³ Consequently, measuring urinary albumin involves a trade-off between convenience and accuracy.

A 24-hour timed urine sample has long been the gold standard for measuring albuminuria, but the collection is cumbersome and time-consuming, and the test is prone to laboratory error.

Dipstick measurements are more convenient and are better at detecting albumin than other proteins in urine, but they have low sensitivity and high interobserver variation.^3–5

The albumin-to-creatinine ratio (ACR). As the quantity of protein in the urine changes with time of day, exertion, stress level, and posture, spot-checking of urine samples is not as good as timed collection. However, a simultaneous measurement of creatinine in a spot urine sample adjusts for protein concentration, which can vary with a person’s hydration status. The ACR so obtained is consistent with the 24-hour timed collection (the gold standard) and is the recommended method of assessing albuminuria.³ An early morning urine sample is favored, as it avoids orthostatic variations and varies less in the same individual.

In a study in the general population comparing the ACR in a random sample and in an early morning sample, only 44% of those who had an ACR of 30 mg/g or higher in the random sample had one this high in the early morning sample.⁶ However, getting an early morning sample is not always feasible in clinical practice. If you are going to measure albuminuria, the Kidney Disease Outcomes and Quality Initiative⁷ suggests checking the ACR in a random sample and then, if the test is positive, following up and confirming it within 3 months with an early morning sample.

Also, since creatinine excretion differs with race, diet, and muscle mass, if the 24-hour creatinine excretion is not close to 1 g, the ACR will give an erroneous estimate of the 24-hour excretion rate.³

Table 1 compares the various methods of measuring protein in the urine.^3,5,8,9 Of note, methods of measuring albumin and total protein vary considerably in their precision and accuracy, making it impossible to reliably translate values from one to the other.⁵

National and international guidelines (Table 2)^7,10–13 agree that albuminuria should be tested in diabetic patients, as it is a surrogate marker for early diabetic nephropathy.^3,13 Most guidelines also recommend measuring albuminuria by a urine ACR test as the preferred measure, even in people without diabetes.

Also, no single cutoff is universally accepted for distinguishing pathologic albuminuria from physiologic albuminuria, nor is there a universally accepted unit of measure.¹⁴ Because approximately 1 g of creatinine is lost in the urine per day, the ACR has the convenient property of numerically matching the albumin excretory rate expressed in milligrams per 24 hours. The other commonly used unit is milligrams of albumin per millimole of creatinine; 30 mg/g is roughly equal to 3 mg/mmol.

The term microalbuminuria was traditionally used to refer to albumin excretion of 30 to 299 mg per 24 hours, and macroalbuminuria to 300 mg or more per 24 hours. However, as there is no pathophysiologic basis to these thresholds (see outcomes data below), the current Kidney Disease Improving Global Outcomes (KDIGO) guidelines do not recommend using these terms.^13,15

RENAL COMPLICATIONS OF ALBUMINURIA

A failure of the glomerular filtration barrier or of proximal tubular reabsorption accounts for most cases of pathologic albuminuria.¹⁶ Processes affecting the glomerular filtration of albumin include endothelial cell dysfunction and abnormalities with the glomerular basement membrane, podocytes, or the slit diaphragms among the podocytic processes.¹⁷

Ultrafiltrated albumin has been directly implicated in tubulointerstitial damage and glomerulosclerosis through diverse pathways. In the proximal tubule, albumin up-regulates interleukin 8 (a chemoattractant for lymphocytes and neutrophils), induces synthesis of endothelin 1 (which stimulates renal cell proliferation, extracellular matrix production, and monocyte attraction), and causes apoptosis of tubular cells.¹⁸ In response to albumin, proximal tubular cells also stimulate interstitial fibroblasts via paracrine release of transforming growth factor beta, either directly or by activating complement or macrophages.^18,19

Studies linking albuminuria to kidney disease

Albuminuria has traditionally been associated with diabetes mellitus as a predictor of overt diabetic nephropathy,^20,21 although in type 1 diabetes, established albuminuria can spontaneously regress.²²

Albuminuria is also a strong predictor of progression in chronic kidney disease.²³ In fact, in the last decade, albuminuria has become an independent criterion in the definition of chronic kidney disease; excretion of more than 30 mg of albumin per day, sustained for at least 3 months, qualifies as chronic kidney disease, with independent prognostic implications (Table  3).¹³

Astor et al,²⁴ in a meta-analysis of 13 studies with more than 21,000 patients with chronic kidney disease, found that the risk of end-stage renal disease was three times higher in those with albuminuria.

Gansevoort et al,²³ in a meta-analysis of nine studies with nearly 850,000 participants from the general population, found that the risk of end-stage renal disease increased continuously as albumin excretion increased. They also found that as albuminuria increased, so did the risk of progression of chronic kidney disease and the incidence of acute kidney injury.

Hemmelgarn et al,²⁵ in a large pooled cohort study with more than 1.5 million participants from the general population, showed that increasing albuminuria was associated with a decline in the estimated glomerular filtration rate (GFR) and with progression to end-stage renal disease across all strata of baseline renal function. For example, in persons with an estimated GFR of 60 mL/min/1.73 m2

• 1 per 1,000 person-years for those with no proteinuria
• 2.8 per 1,000 person-years for those with mild proteinuria (trace or 1+ by dipstick or ACR 30–300 mg/g)
• 13.4 per 1,000 person-years for those with heavy proteinuria (2+ or ACR > 300 mg/g).

Rates of progression to end-stage renal disease were:

• 0.03 per 1,000 person-years with no proteinuria
• 0.05 per 1,000 person-years with mild proteinuria
• 1 per 1,000 person-years with heavy proteinuria.²⁵

CARDIOVASCULAR CONSEQUENCES OF ALBUMINURIA

The exact pathophysiologic link between albuminuria and cardiovascular disease is unknown, but several mechanisms have been proposed.

One is that generalized endothelial dysfunction causes both albuminuria and cardiovascular disease.²⁶ Endothelium-derived nitric oxide has vasodilator, antiplatelet, antiproliferative, antiadhesive, permeability-decreasing, and anti-inflammatory properties. Impaired endothelial synthesis of nitric oxide has been independently associated with both microalbuminuria and diabetes.²⁷

Low levels of heparan sulfate (which has antithrombogenic effects and decreases vessel permeability) in the glycocalyx lining vessel walls may also account for albuminuria and for the other cardiovascular effects.^28–30 These changes may be the consequence of chronic low-grade inflammation, which precedes the occurrence and progression of both albuminuria and atherothrombotic disease. The resulting abnormalities in the endothelial glycocalyx could lead to increased glomerular permeability to albumin and may also be implicated in the pathogenesis of atherosclerosis.²⁶

In an atherosclerotic aorta and coronary arteries, the endothelial dysfunction may cause increased leakage of cholesterol and glycated end-products into the myocardium, resulting in increasing wall stiffness and left ventricular mass. A similar atherosclerotic process may account for coronary artery microthrombi, resulting in subendocardial ischemia that could lead to systolic and diastolic heart dysfunction.³¹

Studies linking albuminuria to heart disease

There is convincing evidence that albuminuria is associated with cardiovascular disease. An ACR between 30 and 300 mg/g was independently associated with myocardial infarction and ischemia.³² People with albuminuria have more than twice the risk of severe coronary artery disease, and albuminuria is also associated with increased intimal thickening in the carotid arteries.^33,34 An ACR in the same range has been associated with increased incidence and progression of coronary artery calcification.³⁵ Higher brachial-ankle pulse-wave velocity has also been demonstrated with albuminuria in a dose-dependent fashion.^36,37

An ACR of 30 to 300 mg/g has been linked to left ventricular hypertrophy independently of other risk factors,³⁸ and functionally with diastolic dysfunction and abnormal midwall shortening.³⁹ In a study of a subgroup of patients with diabetes from a population-based cohort of Native American patients (the Strong Heart Study),³⁹ the prevalence of diastolic dysfunction was:

• 16% in those with no albuminuria
• 26% in those with an ACR of 30 to 300 mg/g
• 31% in those with an ACR greater than 300 mg/g.

The association persisted even after controlling for age, sex, hypertension, and other covariates.

Those pathologic associations have been directly linked to clinical outcomes. For patients with heart failure (New York Heart Association class II–IV), a study found that albuminuria (an ACR > 30 mg/g) conferred a 41% higher risk of admission for heart failure, and an ACR greater than 300 mg/g was associated with an 88% higher risk.⁴⁰

In an analysis of a prospective cohort from the general population with albuminuria and a low prevalence of renal dysfunction (the Prevention of Renal and Vascular Endstage Disease study),⁴¹ albuminuria was associated with a modest increase in the incidence of the composite end point of myocardial infarction, stroke, ischemic heart disease, revascularization procedures, and all-cause mortality per doubling of the urine albumin excretion (hazard ratio 1.08, range 1.04 –1.12).

The relationship to cardiovascular outcomes extends below traditional lower-limit thresholds of albuminuria (corresponding to an ACR > 30 mg/g). A subgroup of patients from the Framingham Offspring Study without prevalent cardiovascular disease, hypertension, diabetes, or kidney disease, and thus with a low to intermediate probability of cardiovascular events, were found to have thresholds for albuminuria as low as 5.3 mg/g in men and 10.8 mg/g in women to discriminate between incident coronary artery disease, heart failure, cerebrovascular disease, other peripheral vascular disease, or death.⁴²

In a meta-analysis including more than 1 million patients in the general population, increasing albuminuria was associated with an increase in deaths from all causes in a continuous manner, with no threshold effect.⁴³ In patients with an ACR of 30 mg/g, the hazard ratio for death was 1.63, increasing to 2.22 for those with more than 300 mg/g compared with those with no albuminuria. A similar increase in the risk of myocardial infarction, heart failure, stroke, or sudden cardiac death was noted with higher ACR.⁴³

Important prospective cohort studies and meta-analyses related to albuminuria and kidney and cardiovascular disease and death are summarized in the eTable.^23,39–50

THE CASE FOR TREATING ALBUMINURIA

Reduced progression of renal disease

Treating patients who have proteinuric chronic kidney disease with an angiotensin-converting enzyme (ACE) inhibitor or an angiotensin receptor blocker (ARB) can reduce the risk of progression of renal failure. However, it is unclear whether this benefit is the result of treating concomitant risk factors independent of the reduction in albuminuria, and there is no consistent treatment effect in improving renal outcomes across studies.

Fink et al,⁵¹ in a meta-analysis, found that chronic kidney disease patients with diabetes, hypertension, and macroalbuminuria had a 40% lower risk of progression to end-stage renal disease if they received an ACE inhibitor (relative risk [RR] 0.60, 95% confidence interval [CI] 0.43–0.83). In the same meta-analysis, ARBs also reduced the risk of progression to end-stage renal disease (RR 0.77, 95% CI 0.66–0.90).

Jafar et al,⁵² in an analysis of pooled patient-level data including only nondiabetic patients on ACE inhibitor therapy (n = 1,860), found that the risk of progression of renal failure, defined as a doubling of serum creatinine or end-stage renal disease, was reduced (RR 0.70, 95% CI 0.55–0.88). Patients with higher levels of albuminuria showed the most benefit, but the effect was not conclusive for albuminuria below 500 mg/day at baseline.

Maione et al,⁵³ in a meta-analysis that included patients with albuminuria who were treated with ACE inhibitors vs placebo (n = 8,231), found a similar reduction in risk of:

• Progression to end-stage renal disease (RR 0.67, 95% CI 0.54–0.84)
• Doubling of serum creatinine (RR 0.62, 95% CI 0.46–0.84)
• Progression of albuminuria (RR 0.49, 95% CI 0.36–0.65)
• Normalization of pathologic albuminuria (as defined by the trialists in the individual studies) (RR 2.99, 95% CI 1.82–4.91).

Similar results were obtained for patients with albuminuria who were treated with ARBs.⁵³

ONTARGET.⁵⁴ In contrast, in the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial, the combination of an ACE inhibitor and an ARB showed no benefit in reducing the progression of renal failure, and in those patients with chronic kidney disease there was a higher risk of a doubling of serum creatinine or of the development of end-stage renal disease and hyperkalemia.

Also, in a pooled analysis of the ONTARGET and Telmisartan Randomized Assessment Study in ACE Intolerant Subjects With Cardiovascular Disease (TRANSCEND) trials, a 50% reduction in baseline albuminuria was associated with reduced progression of renal failure in those with a baseline ACR less than 10 mg/g.⁵⁵

Improved cardiovascular outcomes

There is also evidence of better cardiovascular outcomes with treatment of albuminuria. Again, it is uncertain whether this is a result of treating risk factors other than albuminuria with ACE inhibitors or ARBs, and there is no parallel benefit demonstrated across all studies.

LIFE.^47,48 In the Losartan Intervention for Endpoint Reduction in Hypertension trial, survival analyses suggested a decrease in risk of cardiovascular adverse events as the degree of proteinuria improved with ARB therapy.

Maione et al,⁵³ in a meta-analysis including 8,231 patients with albuminuria and at least one other risk factor, found a significant reduction in the rate of nonfatal cardiovascular outcomes (angina, myocardial infarction, revascularization, stroke, transient ischemic attack, or heart failure) with ACE inhibitors vs placebo (RR 0.88, CI 0.82–0.94) and also in 3,888 patients treated with ARBs vs placebo (RR 0.77, CI 0.61–0.98). However, the meta-analysis did not show that ACE inhibitor or ARB therapy reduced rate of cardiovascular or all-cause mortality.

Fink et al,⁵¹ in their meta-analysis of 18 trials of ACE inhibitors and four trials of ARBs, also found no evidence that ACE inhibitor or ARB therapy reduced cardiovascular mortality rates.³⁸

The ONTARGET trial evaluated the combination of an ACE inhibitor and ARB therapy in patients with diabetes or preexisting peripheral vascular disease. Reductions in the rate of cardiovascular disease or death were not observed, and in those with chronic kidney disease, there was a higher risk of doubling of serum creatinine or development of end-stage renal disease and adverse events of hyperkalemia.⁵⁶ And although an increase in baseline albuminuria was associated with worse cardiovascular outcomes, its reduction in the ONTARGET and TRANSCEND trials did not demonstrate better outcomes when the baseline ACR was greater than 10 mg/g.⁵⁵

WHO SHOULD BE TESTED?

The benefit of adding albuminuria to conventional cardiovascular risk stratification such as Framingham risk scoring is not conclusive. However, today’s clinician may view albuminuria as a biomarker for renal and cardiovascular disease, as albuminuria might be a surrogate marker for endothelial dysfunction in the glomerular capillaries or other vital vascular beds.

High-risk populations and chronic kidney disease patients

Nearly all the current guidelines recommend annual screening for albuminuria in patients with diabetes and hypertension (Table 2).^7,10–13 Other high-risk populations include people with cardiovascular disease, a family history of end-stage renal disease, and metabolic syndrome. Additionally, chronic kidney disease patients whose estimated GFR defines them as being in stage 3 or higher (ie, GFR < 60 mL/min/1.73m²), regardless of other comorbidities, should be tested for albuminuria, as it is an important risk predictor.

Most experts prefer that albuminuria be measured by urine ACR in a first morning voided sample, though this is not the only option.

Screening the general population

Given that albuminuria has been shown to be such an important prognosticator for patients at high risk and those with chronic kidney disease, the question arises whether screening for albuminuria in the asymptomatic low-risk general population would foster earlier detection and therefore enable earlier intervention and result in improved outcomes. However, a systematic review done for the United States Preventive Services Task Force and for an American College of Physicians clinical practice guideline did not find robust evidence to support this.⁵¹

OUR RECOMMENDATIONS

Who should be tested?

• Patients with chronic kidney disease stage 3, 4, or 5 (GFR < 60 mL/min/1.73m²) who are not on dialysis
• Patients who are considered at higher risk of adverse outcomes, such as those with diabetes, hypertension, a family history of end-stage renal disease, or cardiovascular disease. Testing is useful for recognizing increased renal and cardiovascular risk and may lead clinicians to prescribe or titrate a renin-angiotensin system antagonist, a statin, or both, or to modify other cardiovascular risk factors.
• Not recommended: routine screening in the general population who are asymptomatic or are considered at low risk.

Which test should be used?

Based on current evidence and most guidelines, we recommend the urine ACR test as the screening test for people with diabetes and others deemed to be at high risk.

What should be done about albuminuria?

• Controlling blood pressure is important, and though there is debate about the target blood pressure, an individualized plan should be developed with the patient based on age, comorbidities, and goals of care.
• An ACE inhibitor or ARB, if not contraindicated, is recommended for patients with diabetes whose ACR is greater than 30 mg/g and for patients with chronic kidney disease and an ACR greater than 300 mg/g.
• Current evidence does not support the combined use of an ACE inhibitor and an ARB, as proof of benefit is lacking and the risk of adverse events is higher.
• Refer patients with high or unexplained albuminuria to a nephrologist or clinic specializing in chronic kidney disease.

“One can obtain considerable information concerning the general health by examining the urine.” 
—Hippocrates (460?–355? BCE)

Chronic kidney disease is a notable public health concern because it is an important risk factor for end-stage renal disease, cardiovascular disease, and death. Its prevalence¹ exceeds 10% and is considerably higher in high-risk groups, such as those with diabetes or hypertension, which are growing in the United States.

While high levels of total protein in the urine have always been recognized as pathologic, a growing body of evidence links excretion of the protein albumin to adverse cardiovascular outcomes, and most international guidelines now recommend measuring albumin specifically. Albuminuria is a predictor of declining renal function and is independently associated with adverse cardiovascular outcomes. Thus, clinicians need to detect it early, manage it effectively, and reduce concurrent risk factors for cardiovascular disease.

Therefore, this review will focus on albuminuria. However, because the traditional standard for urinary protein measurement was total protein, and because a few guidelines still recommend measuring total protein rather than albumin, we will also briefly discuss total urinary protein.

MOST URINARY PROTEIN IS ALBUMIN

Most of the protein in the urine is albumin filtered from the plasma. Less than half of the rest is derived from the distal renal tubules (uromodulin or Tamm-Horsfall mucoprotein), ² and urine also contains a small and varying proportion of immunoglobulins, low-molecular-weight proteins, and light chains.

Normal healthy people lose less than 30 mg of albumin in the urine per day. In greater amounts, albumin is the major urinary protein in most kidney diseases. Other proteins in urine can be specific markers of less-common illnesses such as plasma cell dyscrasia, glomerulopathy, and renal tubular disease.

MEASURING PROTEINURIA AND ALBUMINURIA

Albumin is not a homogeneous molecule in urine. It undergoes changes to its molecular configuration in the presence of certain ions, peptides, hormones, and drugs, and as a result of proteolytic fragmentation both in the plasma and in renal tubules.³ Consequently, measuring urinary albumin involves a trade-off between convenience and accuracy.

A 24-hour timed urine sample has long been the gold standard for measuring albuminuria, but the collection is cumbersome and time-consuming, and the test is prone to laboratory error.

Dipstick measurements are more convenient and are better at detecting albumin than other proteins in urine, but they have low sensitivity and high interobserver variation.^3–5

The albumin-to-creatinine ratio (ACR). As the quantity of protein in the urine changes with time of day, exertion, stress level, and posture, spot-checking of urine samples is not as good as timed collection. However, a simultaneous measurement of creatinine in a spot urine sample adjusts for protein concentration, which can vary with a person’s hydration status. The ACR so obtained is consistent with the 24-hour timed collection (the gold standard) and is the recommended method of assessing albuminuria.³ An early morning urine sample is favored, as it avoids orthostatic variations and varies less in the same individual.

In a study in the general population comparing the ACR in a random sample and in an early morning sample, only 44% of those who had an ACR of 30 mg/g or higher in the random sample had one this high in the early morning sample.⁶ However, getting an early morning sample is not always feasible in clinical practice. If you are going to measure albuminuria, the Kidney Disease Outcomes and Quality Initiative⁷ suggests checking the ACR in a random sample and then, if the test is positive, following up and confirming it within 3 months with an early morning sample.

Also, since creatinine excretion differs with race, diet, and muscle mass, if the 24-hour creatinine excretion is not close to 1 g, the ACR will give an erroneous estimate of the 24-hour excretion rate.³

Table 1 compares the various methods of measuring protein in the urine.^3,5,8,9 Of note, methods of measuring albumin and total protein vary considerably in their precision and accuracy, making it impossible to reliably translate values from one to the other.⁵

National and international guidelines (Table 2)^7,10–13 agree that albuminuria should be tested in diabetic patients, as it is a surrogate marker for early diabetic nephropathy.^3,13 Most guidelines also recommend measuring albuminuria by a urine ACR test as the preferred measure, even in people without diabetes.

Also, no single cutoff is universally accepted for distinguishing pathologic albuminuria from physiologic albuminuria, nor is there a universally accepted unit of measure.¹⁴ Because approximately 1 g of creatinine is lost in the urine per day, the ACR has the convenient property of numerically matching the albumin excretory rate expressed in milligrams per 24 hours. The other commonly used unit is milligrams of albumin per millimole of creatinine; 30 mg/g is roughly equal to 3 mg/mmol.

The term microalbuminuria was traditionally used to refer to albumin excretion of 30 to 299 mg per 24 hours, and macroalbuminuria to 300 mg or more per 24 hours. However, as there is no pathophysiologic basis to these thresholds (see outcomes data below), the current Kidney Disease Improving Global Outcomes (KDIGO) guidelines do not recommend using these terms.^13,15

RENAL COMPLICATIONS OF ALBUMINURIA

A failure of the glomerular filtration barrier or of proximal tubular reabsorption accounts for most cases of pathologic albuminuria.¹⁶ Processes affecting the glomerular filtration of albumin include endothelial cell dysfunction and abnormalities with the glomerular basement membrane, podocytes, or the slit diaphragms among the podocytic processes.¹⁷

Ultrafiltrated albumin has been directly implicated in tubulointerstitial damage and glomerulosclerosis through diverse pathways. In the proximal tubule, albumin up-regulates interleukin 8 (a chemoattractant for lymphocytes and neutrophils), induces synthesis of endothelin 1 (which stimulates renal cell proliferation, extracellular matrix production, and monocyte attraction), and causes apoptosis of tubular cells.¹⁸ In response to albumin, proximal tubular cells also stimulate interstitial fibroblasts via paracrine release of transforming growth factor beta, either directly or by activating complement or macrophages.^18,19

Studies linking albuminuria to kidney disease

Albuminuria has traditionally been associated with diabetes mellitus as a predictor of overt diabetic nephropathy,^20,21 although in type 1 diabetes, established albuminuria can spontaneously regress.²²

Albuminuria is also a strong predictor of progression in chronic kidney disease.²³ In fact, in the last decade, albuminuria has become an independent criterion in the definition of chronic kidney disease; excretion of more than 30 mg of albumin per day, sustained for at least 3 months, qualifies as chronic kidney disease, with independent prognostic implications (Table  3).¹³

Astor et al,²⁴ in a meta-analysis of 13 studies with more than 21,000 patients with chronic kidney disease, found that the risk of end-stage renal disease was three times higher in those with albuminuria.

Gansevoort et al,²³ in a meta-analysis of nine studies with nearly 850,000 participants from the general population, found that the risk of end-stage renal disease increased continuously as albumin excretion increased. They also found that as albuminuria increased, so did the risk of progression of chronic kidney disease and the incidence of acute kidney injury.

Hemmelgarn et al,²⁵ in a large pooled cohort study with more than 1.5 million participants from the general population, showed that increasing albuminuria was associated with a decline in the estimated glomerular filtration rate (GFR) and with progression to end-stage renal disease across all strata of baseline renal function. For example, in persons with an estimated GFR of 60 mL/min/1.73 m2

• 1 per 1,000 person-years for those with no proteinuria
• 2.8 per 1,000 person-years for those with mild proteinuria (trace or 1+ by dipstick or ACR 30–300 mg/g)
• 13.4 per 1,000 person-years for those with heavy proteinuria (2+ or ACR > 300 mg/g).

Rates of progression to end-stage renal disease were:

• 0.03 per 1,000 person-years with no proteinuria
• 0.05 per 1,000 person-years with mild proteinuria
• 1 per 1,000 person-years with heavy proteinuria.²⁵

CARDIOVASCULAR CONSEQUENCES OF ALBUMINURIA

The exact pathophysiologic link between albuminuria and cardiovascular disease is unknown, but several mechanisms have been proposed.

One is that generalized endothelial dysfunction causes both albuminuria and cardiovascular disease.²⁶ Endothelium-derived nitric oxide has vasodilator, antiplatelet, antiproliferative, antiadhesive, permeability-decreasing, and anti-inflammatory properties. Impaired endothelial synthesis of nitric oxide has been independently associated with both microalbuminuria and diabetes.²⁷

Low levels of heparan sulfate (which has antithrombogenic effects and decreases vessel permeability) in the glycocalyx lining vessel walls may also account for albuminuria and for the other cardiovascular effects.^28–30 These changes may be the consequence of chronic low-grade inflammation, which precedes the occurrence and progression of both albuminuria and atherothrombotic disease. The resulting abnormalities in the endothelial glycocalyx could lead to increased glomerular permeability to albumin and may also be implicated in the pathogenesis of atherosclerosis.²⁶

In an atherosclerotic aorta and coronary arteries, the endothelial dysfunction may cause increased leakage of cholesterol and glycated end-products into the myocardium, resulting in increasing wall stiffness and left ventricular mass. A similar atherosclerotic process may account for coronary artery microthrombi, resulting in subendocardial ischemia that could lead to systolic and diastolic heart dysfunction.³¹

Studies linking albuminuria to heart disease

There is convincing evidence that albuminuria is associated with cardiovascular disease. An ACR between 30 and 300 mg/g was independently associated with myocardial infarction and ischemia.³² People with albuminuria have more than twice the risk of severe coronary artery disease, and albuminuria is also associated with increased intimal thickening in the carotid arteries.^33,34 An ACR in the same range has been associated with increased incidence and progression of coronary artery calcification.³⁵ Higher brachial-ankle pulse-wave velocity has also been demonstrated with albuminuria in a dose-dependent fashion.^36,37

An ACR of 30 to 300 mg/g has been linked to left ventricular hypertrophy independently of other risk factors,³⁸ and functionally with diastolic dysfunction and abnormal midwall shortening.³⁹ In a study of a subgroup of patients with diabetes from a population-based cohort of Native American patients (the Strong Heart Study),³⁹ the prevalence of diastolic dysfunction was:

• 16% in those with no albuminuria
• 26% in those with an ACR of 30 to 300 mg/g
• 31% in those with an ACR greater than 300 mg/g.

The association persisted even after controlling for age, sex, hypertension, and other covariates.

Those pathologic associations have been directly linked to clinical outcomes. For patients with heart failure (New York Heart Association class II–IV), a study found that albuminuria (an ACR > 30 mg/g) conferred a 41% higher risk of admission for heart failure, and an ACR greater than 300 mg/g was associated with an 88% higher risk.⁴⁰

In an analysis of a prospective cohort from the general population with albuminuria and a low prevalence of renal dysfunction (the Prevention of Renal and Vascular Endstage Disease study),⁴¹ albuminuria was associated with a modest increase in the incidence of the composite end point of myocardial infarction, stroke, ischemic heart disease, revascularization procedures, and all-cause mortality per doubling of the urine albumin excretion (hazard ratio 1.08, range 1.04 –1.12).

The relationship to cardiovascular outcomes extends below traditional lower-limit thresholds of albuminuria (corresponding to an ACR > 30 mg/g). A subgroup of patients from the Framingham Offspring Study without prevalent cardiovascular disease, hypertension, diabetes, or kidney disease, and thus with a low to intermediate probability of cardiovascular events, were found to have thresholds for albuminuria as low as 5.3 mg/g in men and 10.8 mg/g in women to discriminate between incident coronary artery disease, heart failure, cerebrovascular disease, other peripheral vascular disease, or death.⁴²

In a meta-analysis including more than 1 million patients in the general population, increasing albuminuria was associated with an increase in deaths from all causes in a continuous manner, with no threshold effect.⁴³ In patients with an ACR of 30 mg/g, the hazard ratio for death was 1.63, increasing to 2.22 for those with more than 300 mg/g compared with those with no albuminuria. A similar increase in the risk of myocardial infarction, heart failure, stroke, or sudden cardiac death was noted with higher ACR.⁴³

Important prospective cohort studies and meta-analyses related to albuminuria and kidney and cardiovascular disease and death are summarized in the eTable.^23,39–50

THE CASE FOR TREATING ALBUMINURIA

Reduced progression of renal disease

Treating patients who have proteinuric chronic kidney disease with an angiotensin-converting enzyme (ACE) inhibitor or an angiotensin receptor blocker (ARB) can reduce the risk of progression of renal failure. However, it is unclear whether this benefit is the result of treating concomitant risk factors independent of the reduction in albuminuria, and there is no consistent treatment effect in improving renal outcomes across studies.

Fink et al,⁵¹ in a meta-analysis, found that chronic kidney disease patients with diabetes, hypertension, and macroalbuminuria had a 40% lower risk of progression to end-stage renal disease if they received an ACE inhibitor (relative risk [RR] 0.60, 95% confidence interval [CI] 0.43–0.83). In the same meta-analysis, ARBs also reduced the risk of progression to end-stage renal disease (RR 0.77, 95% CI 0.66–0.90).

Jafar et al,⁵² in an analysis of pooled patient-level data including only nondiabetic patients on ACE inhibitor therapy (n = 1,860), found that the risk of progression of renal failure, defined as a doubling of serum creatinine or end-stage renal disease, was reduced (RR 0.70, 95% CI 0.55–0.88). Patients with higher levels of albuminuria showed the most benefit, but the effect was not conclusive for albuminuria below 500 mg/day at baseline.

Maione et al,⁵³ in a meta-analysis that included patients with albuminuria who were treated with ACE inhibitors vs placebo (n = 8,231), found a similar reduction in risk of:

• Progression to end-stage renal disease (RR 0.67, 95% CI 0.54–0.84)
• Doubling of serum creatinine (RR 0.62, 95% CI 0.46–0.84)
• Progression of albuminuria (RR 0.49, 95% CI 0.36–0.65)
• Normalization of pathologic albuminuria (as defined by the trialists in the individual studies) (RR 2.99, 95% CI 1.82–4.91).

Similar results were obtained for patients with albuminuria who were treated with ARBs.⁵³

ONTARGET.⁵⁴ In contrast, in the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial, the combination of an ACE inhibitor and an ARB showed no benefit in reducing the progression of renal failure, and in those patients with chronic kidney disease there was a higher risk of a doubling of serum creatinine or of the development of end-stage renal disease and hyperkalemia.

Also, in a pooled analysis of the ONTARGET and Telmisartan Randomized Assessment Study in ACE Intolerant Subjects With Cardiovascular Disease (TRANSCEND) trials, a 50% reduction in baseline albuminuria was associated with reduced progression of renal failure in those with a baseline ACR less than 10 mg/g.⁵⁵

Improved cardiovascular outcomes

There is also evidence of better cardiovascular outcomes with treatment of albuminuria. Again, it is uncertain whether this is a result of treating risk factors other than albuminuria with ACE inhibitors or ARBs, and there is no parallel benefit demonstrated across all studies.

LIFE.^47,48 In the Losartan Intervention for Endpoint Reduction in Hypertension trial, survival analyses suggested a decrease in risk of cardiovascular adverse events as the degree of proteinuria improved with ARB therapy.

Maione et al,⁵³ in a meta-analysis including 8,231 patients with albuminuria and at least one other risk factor, found a significant reduction in the rate of nonfatal cardiovascular outcomes (angina, myocardial infarction, revascularization, stroke, transient ischemic attack, or heart failure) with ACE inhibitors vs placebo (RR 0.88, CI 0.82–0.94) and also in 3,888 patients treated with ARBs vs placebo (RR 0.77, CI 0.61–0.98). However, the meta-analysis did not show that ACE inhibitor or ARB therapy reduced rate of cardiovascular or all-cause mortality.

Fink et al,⁵¹ in their meta-analysis of 18 trials of ACE inhibitors and four trials of ARBs, also found no evidence that ACE inhibitor or ARB therapy reduced cardiovascular mortality rates.³⁸

The ONTARGET trial evaluated the combination of an ACE inhibitor and ARB therapy in patients with diabetes or preexisting peripheral vascular disease. Reductions in the rate of cardiovascular disease or death were not observed, and in those with chronic kidney disease, there was a higher risk of doubling of serum creatinine or development of end-stage renal disease and adverse events of hyperkalemia.⁵⁶ And although an increase in baseline albuminuria was associated with worse cardiovascular outcomes, its reduction in the ONTARGET and TRANSCEND trials did not demonstrate better outcomes when the baseline ACR was greater than 10 mg/g.⁵⁵

WHO SHOULD BE TESTED?

The benefit of adding albuminuria to conventional cardiovascular risk stratification such as Framingham risk scoring is not conclusive. However, today’s clinician may view albuminuria as a biomarker for renal and cardiovascular disease, as albuminuria might be a surrogate marker for endothelial dysfunction in the glomerular capillaries or other vital vascular beds.

High-risk populations and chronic kidney disease patients

Nearly all the current guidelines recommend annual screening for albuminuria in patients with diabetes and hypertension (Table 2).^7,10–13 Other high-risk populations include people with cardiovascular disease, a family history of end-stage renal disease, and metabolic syndrome. Additionally, chronic kidney disease patients whose estimated GFR defines them as being in stage 3 or higher (ie, GFR < 60 mL/min/1.73m²), regardless of other comorbidities, should be tested for albuminuria, as it is an important risk predictor.

Most experts prefer that albuminuria be measured by urine ACR in a first morning voided sample, though this is not the only option.

Screening the general population

Given that albuminuria has been shown to be such an important prognosticator for patients at high risk and those with chronic kidney disease, the question arises whether screening for albuminuria in the asymptomatic low-risk general population would foster earlier detection and therefore enable earlier intervention and result in improved outcomes. However, a systematic review done for the United States Preventive Services Task Force and for an American College of Physicians clinical practice guideline did not find robust evidence to support this.⁵¹

OUR RECOMMENDATIONS

Who should be tested?

• Patients with chronic kidney disease stage 3, 4, or 5 (GFR < 60 mL/min/1.73m²) who are not on dialysis
• Patients who are considered at higher risk of adverse outcomes, such as those with diabetes, hypertension, a family history of end-stage renal disease, or cardiovascular disease. Testing is useful for recognizing increased renal and cardiovascular risk and may lead clinicians to prescribe or titrate a renin-angiotensin system antagonist, a statin, or both, or to modify other cardiovascular risk factors.
• Not recommended: routine screening in the general population who are asymptomatic or are considered at low risk.

Which test should be used?

Based on current evidence and most guidelines, we recommend the urine ACR test as the screening test for people with diabetes and others deemed to be at high risk.

What should be done about albuminuria?

• Controlling blood pressure is important, and though there is debate about the target blood pressure, an individualized plan should be developed with the patient based on age, comorbidities, and goals of care.
• An ACE inhibitor or ARB, if not contraindicated, is recommended for patients with diabetes whose ACR is greater than 30 mg/g and for patients with chronic kidney disease and an ACR greater than 300 mg/g.
• Current evidence does not support the combined use of an ACE inhibitor and an ARB, as proof of benefit is lacking and the risk of adverse events is higher.
• Refer patients with high or unexplained albuminuria to a nephrologist or clinic specializing in chronic kidney disease.

An 85-year-old man with hypertension, hyperlipidemia, and coronary artery disease presented to our clinic with diffuse muscle pain. The pain had been present for about 3 months, but it had become noticeably worse over the past few weeks.

He was not aware of any trauma. He described the muscle pain as dull and particularly severe in his lower extremities (his thighs and calves). The pain did not limit his daily activities, nor did physical exertion or the time of day have any effect on the level of the pain.

His medications at that time included metoprolol, aspirin, hydrochlorothiazide, simvastatin, and a daily multivitamin.

He was not in acute distress. On neurologic and musculoskeletal examinations, all deep-tendon reflexes were intact, with no tenderness to palpation of the upper and lower extremities. No abnormalities were noted on the joint examination. He had full range of motion, with 5/5 muscle strength in the upper and lower extremities bilaterally and normal muscle tone. He was able to walk with ease. Results of initial laboratory testing, including creatine kinase and erythrocyte sedimentation rate, were normal.

1. What should be the next best step in the evaluation of this patient’s muscle pain?

• Order tests for cyclic citrullinated peptide (CCP) antibody and rheumatoid factor
• Advise him to refrain from physical activity until his symptoms resolve
• Take a more detailed history, including a review of medications and supplements
• Recommend a trial of a nonsteroidal anti-inflammatory drug (NSAID)
• Send him for radiographic imaging

Since his muscle pain has persisted for several months without improvement, a more detailed history should be taken, including a review of current medications and supplements.

Testing CCP antibody and rheumatoid factor would be useful if rheumatoid arthritis were suspected, but in the absence of demonstrable arthritis on examination, these tests would have low specificity even if the results were positive.

An NSAID may temporarily alleviate his pain, but it will not help establish a diagnosis. And in elderly patients, NSAIDs are not without complications and so should be prescribed only in appropriate situations.

Imaging would be appropriate at this point only if there was clinical suspicion of a specific disease. However, our patient has no focal deficits, and the suspicion of fracture or malignancy is low.

The medical history should include asking about current drug regimens, recent medication changes, and the use of herbal supplements, since polypharmacy is common in elderly patients with multiple comorbidities.

On further questioning, our patient said that his dose of simvastatin had been increased from 40 mg daily to 80 mg daily about 1 month before his symptoms appeared. He was taking a daily multivitamin but was not using herbal supplements or other over-the-counter products. He did not recall any constitutional symptoms before the onset of his current symptoms, and he had never had similar muscle pain in the past.

2. Based on the additional information from the history, what is the most likely cause of his muscle pain?

• Limited myositis secondary to recent viral infection
• Rhabdomyolysis
• Hypothyroidism
• Drug-drug interaction
• Statin-induced myalgia

Our patient’s history provided nothing to suggest viral myositis. Hypothyroidism should always be considered in patients with myalgia, but this is not likely in our patient, as he does not display other characteristics, such as diminished reflexes, hypotonia, cold intolerance, and mood instability. Even though calcium channel blockers have been known to cause myalgia in patients on statins, a drug-drug reaction is not likely, as he had not started taking a calcium channel blocker before his symptoms began. This patient did not show signs or symptoms of rhabdomyolysis, a type of myopathy in which necrosis of the muscle tissue occurs, generally causing profound weakness and pain.¹

Therefore, statin-induced myopathy is the most likely cause of his diffuse muscle pain, particularly since his simvastatin had been increased 1 month before the onset of symptoms.

3. What should be the next step in his management?

• Decrease the dose of simvastatin to the last known dose he was able to tolerate
• Continue simvastatin at the same dose and then monitor
• Switch to another statin
• Add coenzyme Q10
• Stop simvastatin

Decreasing the statin dosage to the last well-tolerated dose would not be appropriate in a patient with myopathy, as the symptoms would probably not improve.^2–4 Also, one should not switch to a different statin while a patient is experiencing symptoms. Rather, the statin should be stopped for at least 6 weeks or until the symptoms have fully resolved.¹

Adding coenzyme Q10 is another option, especially in a patient with previously diagnosed coronary artery disease,⁵ when continued statin therapy is thought necessary to reduce the likelihood of repeat coronary events.

We discontinued his simvastatin. Followup 3 weeks later in the outpatient clinic showed that his symptoms were slowly improving. The symptoms had resolved completely 4 months later.

4. How should we manage our patient’s hyperlipidemia once his symptoms have resolved?

• Restart simvastatin at the 80-mg dose
• Restart simvastatin at the 40-mg dose
• Start a hydrophilic statin at full dose
• Use a drug from another class of lipid-lowering drugs
• Wait another 3 months before prescribing any lipid-lowering drug

His treatment for hyperlipidemia should be continued, considering his comorbidities. However, restarting the same statin, even at a lower dose, will likely cause his symptoms to recur. Thus, a different statin should be tried once his muscle pain has resolved.

Other classes of lipid-lowering drugs are usually less efficacious than statins, particularly when trying to control low-density lipoprotein (LDL) cholesterol, so a drug from another class should not be used until other statin options have been attempted.^2,6,7

Simvastatin is lipophilic. Trying a statin with hydrophilic properties (eg, pravastatin, rosuvastatin, fluvastatin) has been shown to convey similar cardioprotective effects with a lower propensity for myalgia, as lipophilic statins have a higher propensity to penetrate muscle tissue than do hydrophilic statins.^3,4,8

Once his symptoms resolved, our patient was started on a hydrophilic statin, fluvastatin 20 mg daily. Unfortunately, his pain recurred 3 weeks later. The statin was stopped, and his symptoms again resolved.

5. Since our patient was unable to tolerate a second statin, what should be the next step in his management?

• Restart simvastatin 
• Use a drug from another class to control the hyperlipidemia
• Wait at least 6 months after symptoms resolve before trying any lipid-lowering drug
• Initiate therapy with coenzyme Q10 and fish oil
• Wait for symptoms to resolve, then restart a hydrophilic statin at a lower dose and lower frequency

Restarting simvastatin will likely cause a recurrence of the myalgia. Other lipid-lowering drugs such as nicotinic acid, bile acid resins, and fibrates are not as efficacious as statins. Coenzyme Q10 and fish oil can reduce lipid levels, but they are not as efficacious as statins.

In view of our patient’s lipid profile—LDL cholesterol elevated at 167 mg/dL, high-density lipoprotein cholesterol 31 mg/dL, triglycerides 47 mg/dL—it is important to treat his hyperlipidemia. Therefore, another attempt at statin therapy should be made once his symptoms have resolved.

Studies have shown that restarting a statin at a low dose and low frequency is effective in patients who have experienced intolerance to a statin.^3,4 Our patient was treated with low-dose pravastatin (20 mg), resulting in a moderate improvement in his LDL cholesterol to 123 mg/dL.

STATIN-INDUCED MYOPATHY: ADDRESSING THE DILEMMA

Treating hyperlipidemia is important to prevent vascular events in patients with or without coronary artery disease. Statins are the most effective agents available for controlling hypercholesterolemia, specifically LDL levels, as well as for preventing myocardial infarction.

Unfortunately, significant side effects have been reported, and myopathy is the most prevalent. Statin-induced myopathy includes a combination of muscle tenderness, myalgia, and weakness.^2–11 In randomized controlled trials, the risk of myopathy was estimated to be between 1.5% and 5%.⁶ In unselected clinic patients on high-dose statins, the rate of muscle complaints may be as high as 20%.¹²

The cause of statin-induced myopathy is not known, although studies have linked it to genetic defects.⁷ Risk factors have been identified and include personal and family history of myalgia, Asian ethnicity, hypothyroidism, and type 1 diabetes. The incidence of statin-induced myalgia is two to three times higher in patients on corticosteroid therapy. Other risk factors include female sex, liver disease, and renal dysfunction.^7,8

A less common etiology is anti-HMG coenzyme A reductase antibodies. Studies have shown that these antibody levels correlate well with the amount of myositis as measured by creatine kinase levels. However, there is no consensus yet on screening for these antibodies.¹³

Statin therapy poses a dilemma, as there is a thin line between the benefits and the risks of side effects, especially statin-induced myopathy.^3,4 Current recommendations include discontinuing the statin until symptoms fully resolve. Creatine kinase levels may be useful in assessing for potential muscle breakdown, especially in patients with reduced renal function, as this predisposes them to statin-induced myopathy, yet normal values do not preclude the diagnosis of statin-induced myopathy.^3,4,7,8

Once symptoms resolve and laboratory test results normalize, a trial of a different statin is recommended. If patients become symptomatic, a trial of a low-dose hydrophilic statin at a once- or twice-weekly interval has been recommended. Several studies have assessed the efficacy of a low-dose statin with decreased frequency of administration and have consistently shown significant improvement in lipid levels.^3,4 For instance, once-weekly rosuvastatin at a dose between 5 mg and 20 mg resulted in a 29% reduction in LDL cholesterol levels, and 80% of patients did not experience a recurrence of myalgia.³ Furthermore, a study of patients treated with 5 mg to 10 mg of rosuvastatin twice a week resulted in a 26% decrease in LDL cholesterol levels.⁴ This study also showed that when an additional non-statin lipid-lowering drug was prescribed (eg, ezetimibe, bile acid resin, nicotinic acid), more than half of the patients reached their goal lipid level.⁴

The addition of coenzyme Q10 and fish oil has also been suggested. Although, the evidence to support this is inconclusive, the potential benefit outweighs the risk, since the side effects are minimal.¹ However, no study yet has evaluated the risks vs the benefits in patients with elevated creatine kinase.

Statin-induced myopathy is a commonly encountered adverse effect. Currently, there are no guidelines on restarting statin therapy after statin-induced myopathy; however, data suggest that statin therapy should be restarted once symptoms resolve, and that variations in dose and frequency may be necessary.^1–8,14

An 85-year-old man with hypertension, hyperlipidemia, and coronary artery disease presented to our clinic with diffuse muscle pain. The pain had been present for about 3 months, but it had become noticeably worse over the past few weeks.

He was not aware of any trauma. He described the muscle pain as dull and particularly severe in his lower extremities (his thighs and calves). The pain did not limit his daily activities, nor did physical exertion or the time of day have any effect on the level of the pain.

His medications at that time included metoprolol, aspirin, hydrochlorothiazide, simvastatin, and a daily multivitamin.

He was not in acute distress. On neurologic and musculoskeletal examinations, all deep-tendon reflexes were intact, with no tenderness to palpation of the upper and lower extremities. No abnormalities were noted on the joint examination. He had full range of motion, with 5/5 muscle strength in the upper and lower extremities bilaterally and normal muscle tone. He was able to walk with ease. Results of initial laboratory testing, including creatine kinase and erythrocyte sedimentation rate, were normal.

1. What should be the next best step in the evaluation of this patient’s muscle pain?

• Order tests for cyclic citrullinated peptide (CCP) antibody and rheumatoid factor
• Advise him to refrain from physical activity until his symptoms resolve
• Take a more detailed history, including a review of medications and supplements
• Recommend a trial of a nonsteroidal anti-inflammatory drug (NSAID)
• Send him for radiographic imaging

Since his muscle pain has persisted for several months without improvement, a more detailed history should be taken, including a review of current medications and supplements.

Testing CCP antibody and rheumatoid factor would be useful if rheumatoid arthritis were suspected, but in the absence of demonstrable arthritis on examination, these tests would have low specificity even if the results were positive.

An NSAID may temporarily alleviate his pain, but it will not help establish a diagnosis. And in elderly patients, NSAIDs are not without complications and so should be prescribed only in appropriate situations.

Imaging would be appropriate at this point only if there was clinical suspicion of a specific disease. However, our patient has no focal deficits, and the suspicion of fracture or malignancy is low.

The medical history should include asking about current drug regimens, recent medication changes, and the use of herbal supplements, since polypharmacy is common in elderly patients with multiple comorbidities.

On further questioning, our patient said that his dose of simvastatin had been increased from 40 mg daily to 80 mg daily about 1 month before his symptoms appeared. He was taking a daily multivitamin but was not using herbal supplements or other over-the-counter products. He did not recall any constitutional symptoms before the onset of his current symptoms, and he had never had similar muscle pain in the past.

2. Based on the additional information from the history, what is the most likely cause of his muscle pain?

• Limited myositis secondary to recent viral infection
• Rhabdomyolysis
• Hypothyroidism
• Drug-drug interaction
• Statin-induced myalgia

Our patient’s history provided nothing to suggest viral myositis. Hypothyroidism should always be considered in patients with myalgia, but this is not likely in our patient, as he does not display other characteristics, such as diminished reflexes, hypotonia, cold intolerance, and mood instability. Even though calcium channel blockers have been known to cause myalgia in patients on statins, a drug-drug reaction is not likely, as he had not started taking a calcium channel blocker before his symptoms began. This patient did not show signs or symptoms of rhabdomyolysis, a type of myopathy in which necrosis of the muscle tissue occurs, generally causing profound weakness and pain.¹

Therefore, statin-induced myopathy is the most likely cause of his diffuse muscle pain, particularly since his simvastatin had been increased 1 month before the onset of symptoms.

3. What should be the next step in his management?

• Decrease the dose of simvastatin to the last known dose he was able to tolerate
• Continue simvastatin at the same dose and then monitor
• Switch to another statin
• Add coenzyme Q10
• Stop simvastatin

Decreasing the statin dosage to the last well-tolerated dose would not be appropriate in a patient with myopathy, as the symptoms would probably not improve.^2–4 Also, one should not switch to a different statin while a patient is experiencing symptoms. Rather, the statin should be stopped for at least 6 weeks or until the symptoms have fully resolved.¹

Adding coenzyme Q10 is another option, especially in a patient with previously diagnosed coronary artery disease,⁵ when continued statin therapy is thought necessary to reduce the likelihood of repeat coronary events.

We discontinued his simvastatin. Followup 3 weeks later in the outpatient clinic showed that his symptoms were slowly improving. The symptoms had resolved completely 4 months later.

4. How should we manage our patient’s hyperlipidemia once his symptoms have resolved?

• Restart simvastatin at the 80-mg dose
• Restart simvastatin at the 40-mg dose
• Start a hydrophilic statin at full dose
• Use a drug from another class of lipid-lowering drugs
• Wait another 3 months before prescribing any lipid-lowering drug

His treatment for hyperlipidemia should be continued, considering his comorbidities. However, restarting the same statin, even at a lower dose, will likely cause his symptoms to recur. Thus, a different statin should be tried once his muscle pain has resolved.

Other classes of lipid-lowering drugs are usually less efficacious than statins, particularly when trying to control low-density lipoprotein (LDL) cholesterol, so a drug from another class should not be used until other statin options have been attempted.^2,6,7

Simvastatin is lipophilic. Trying a statin with hydrophilic properties (eg, pravastatin, rosuvastatin, fluvastatin) has been shown to convey similar cardioprotective effects with a lower propensity for myalgia, as lipophilic statins have a higher propensity to penetrate muscle tissue than do hydrophilic statins.^3,4,8

Once his symptoms resolved, our patient was started on a hydrophilic statin, fluvastatin 20 mg daily. Unfortunately, his pain recurred 3 weeks later. The statin was stopped, and his symptoms again resolved.

5. Since our patient was unable to tolerate a second statin, what should be the next step in his management?

• Restart simvastatin 
• Use a drug from another class to control the hyperlipidemia
• Wait at least 6 months after symptoms resolve before trying any lipid-lowering drug
• Initiate therapy with coenzyme Q10 and fish oil
• Wait for symptoms to resolve, then restart a hydrophilic statin at a lower dose and lower frequency

Restarting simvastatin will likely cause a recurrence of the myalgia. Other lipid-lowering drugs such as nicotinic acid, bile acid resins, and fibrates are not as efficacious as statins. Coenzyme Q10 and fish oil can reduce lipid levels, but they are not as efficacious as statins.

In view of our patient’s lipid profile—LDL cholesterol elevated at 167 mg/dL, high-density lipoprotein cholesterol 31 mg/dL, triglycerides 47 mg/dL—it is important to treat his hyperlipidemia. Therefore, another attempt at statin therapy should be made once his symptoms have resolved.

Studies have shown that restarting a statin at a low dose and low frequency is effective in patients who have experienced intolerance to a statin.^3,4 Our patient was treated with low-dose pravastatin (20 mg), resulting in a moderate improvement in his LDL cholesterol to 123 mg/dL.

STATIN-INDUCED MYOPATHY: ADDRESSING THE DILEMMA

Treating hyperlipidemia is important to prevent vascular events in patients with or without coronary artery disease. Statins are the most effective agents available for controlling hypercholesterolemia, specifically LDL levels, as well as for preventing myocardial infarction.

Unfortunately, significant side effects have been reported, and myopathy is the most prevalent. Statin-induced myopathy includes a combination of muscle tenderness, myalgia, and weakness.^2–11 In randomized controlled trials, the risk of myopathy was estimated to be between 1.5% and 5%.⁶ In unselected clinic patients on high-dose statins, the rate of muscle complaints may be as high as 20%.¹²

The cause of statin-induced myopathy is not known, although studies have linked it to genetic defects.⁷ Risk factors have been identified and include personal and family history of myalgia, Asian ethnicity, hypothyroidism, and type 1 diabetes. The incidence of statin-induced myalgia is two to three times higher in patients on corticosteroid therapy. Other risk factors include female sex, liver disease, and renal dysfunction.^7,8

A less common etiology is anti-HMG coenzyme A reductase antibodies. Studies have shown that these antibody levels correlate well with the amount of myositis as measured by creatine kinase levels. However, there is no consensus yet on screening for these antibodies.¹³

Statin therapy poses a dilemma, as there is a thin line between the benefits and the risks of side effects, especially statin-induced myopathy.^3,4 Current recommendations include discontinuing the statin until symptoms fully resolve. Creatine kinase levels may be useful in assessing for potential muscle breakdown, especially in patients with reduced renal function, as this predisposes them to statin-induced myopathy, yet normal values do not preclude the diagnosis of statin-induced myopathy.^3,4,7,8

Once symptoms resolve and laboratory test results normalize, a trial of a different statin is recommended. If patients become symptomatic, a trial of a low-dose hydrophilic statin at a once- or twice-weekly interval has been recommended. Several studies have assessed the efficacy of a low-dose statin with decreased frequency of administration and have consistently shown significant improvement in lipid levels.^3,4 For instance, once-weekly rosuvastatin at a dose between 5 mg and 20 mg resulted in a 29% reduction in LDL cholesterol levels, and 80% of patients did not experience a recurrence of myalgia.³ Furthermore, a study of patients treated with 5 mg to 10 mg of rosuvastatin twice a week resulted in a 26% decrease in LDL cholesterol levels.⁴ This study also showed that when an additional non-statin lipid-lowering drug was prescribed (eg, ezetimibe, bile acid resin, nicotinic acid), more than half of the patients reached their goal lipid level.⁴

The addition of coenzyme Q10 and fish oil has also been suggested. Although, the evidence to support this is inconclusive, the potential benefit outweighs the risk, since the side effects are minimal.¹ However, no study yet has evaluated the risks vs the benefits in patients with elevated creatine kinase.

Statin-induced myopathy is a commonly encountered adverse effect. Currently, there are no guidelines on restarting statin therapy after statin-induced myopathy; however, data suggest that statin therapy should be restarted once symptoms resolve, and that variations in dose and frequency may be necessary.^1–8,14

An 85-year-old man with hypertension, hyperlipidemia, and coronary artery disease presented to our clinic with diffuse muscle pain. The pain had been present for about 3 months, but it had become noticeably worse over the past few weeks.

He was not aware of any trauma. He described the muscle pain as dull and particularly severe in his lower extremities (his thighs and calves). The pain did not limit his daily activities, nor did physical exertion or the time of day have any effect on the level of the pain.

His medications at that time included metoprolol, aspirin, hydrochlorothiazide, simvastatin, and a daily multivitamin.

He was not in acute distress. On neurologic and musculoskeletal examinations, all deep-tendon reflexes were intact, with no tenderness to palpation of the upper and lower extremities. No abnormalities were noted on the joint examination. He had full range of motion, with 5/5 muscle strength in the upper and lower extremities bilaterally and normal muscle tone. He was able to walk with ease. Results of initial laboratory testing, including creatine kinase and erythrocyte sedimentation rate, were normal.

1. What should be the next best step in the evaluation of this patient’s muscle pain?

• Order tests for cyclic citrullinated peptide (CCP) antibody and rheumatoid factor
• Advise him to refrain from physical activity until his symptoms resolve
• Take a more detailed history, including a review of medications and supplements
• Recommend a trial of a nonsteroidal anti-inflammatory drug (NSAID)
• Send him for radiographic imaging

Since his muscle pain has persisted for several months without improvement, a more detailed history should be taken, including a review of current medications and supplements.

Testing CCP antibody and rheumatoid factor would be useful if rheumatoid arthritis were suspected, but in the absence of demonstrable arthritis on examination, these tests would have low specificity even if the results were positive.

An NSAID may temporarily alleviate his pain, but it will not help establish a diagnosis. And in elderly patients, NSAIDs are not without complications and so should be prescribed only in appropriate situations.

Imaging would be appropriate at this point only if there was clinical suspicion of a specific disease. However, our patient has no focal deficits, and the suspicion of fracture or malignancy is low.

The medical history should include asking about current drug regimens, recent medication changes, and the use of herbal supplements, since polypharmacy is common in elderly patients with multiple comorbidities.

On further questioning, our patient said that his dose of simvastatin had been increased from 40 mg daily to 80 mg daily about 1 month before his symptoms appeared. He was taking a daily multivitamin but was not using herbal supplements or other over-the-counter products. He did not recall any constitutional symptoms before the onset of his current symptoms, and he had never had similar muscle pain in the past.

2. Based on the additional information from the history, what is the most likely cause of his muscle pain?

• Limited myositis secondary to recent viral infection
• Rhabdomyolysis
• Hypothyroidism
• Drug-drug interaction
• Statin-induced myalgia

Our patient’s history provided nothing to suggest viral myositis. Hypothyroidism should always be considered in patients with myalgia, but this is not likely in our patient, as he does not display other characteristics, such as diminished reflexes, hypotonia, cold intolerance, and mood instability. Even though calcium channel blockers have been known to cause myalgia in patients on statins, a drug-drug reaction is not likely, as he had not started taking a calcium channel blocker before his symptoms began. This patient did not show signs or symptoms of rhabdomyolysis, a type of myopathy in which necrosis of the muscle tissue occurs, generally causing profound weakness and pain.¹

Therefore, statin-induced myopathy is the most likely cause of his diffuse muscle pain, particularly since his simvastatin had been increased 1 month before the onset of symptoms.

3. What should be the next step in his management?

• Decrease the dose of simvastatin to the last known dose he was able to tolerate
• Continue simvastatin at the same dose and then monitor
• Switch to another statin
• Add coenzyme Q10
• Stop simvastatin

Decreasing the statin dosage to the last well-tolerated dose would not be appropriate in a patient with myopathy, as the symptoms would probably not improve.^2–4 Also, one should not switch to a different statin while a patient is experiencing symptoms. Rather, the statin should be stopped for at least 6 weeks or until the symptoms have fully resolved.¹

Adding coenzyme Q10 is another option, especially in a patient with previously diagnosed coronary artery disease,⁵ when continued statin therapy is thought necessary to reduce the likelihood of repeat coronary events.

We discontinued his simvastatin. Followup 3 weeks later in the outpatient clinic showed that his symptoms were slowly improving. The symptoms had resolved completely 4 months later.

4. How should we manage our patient’s hyperlipidemia once his symptoms have resolved?