Interpreting and investigating proteinuriaBMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e2339 (Published 05 April 2012) Cite this as: BMJ 2012;344:e2339
- Oliver T Browne, medical student,
- Sunil Bhandari, consultant nephrologist
- 1Department of Renal Medicine, Hull and East Yorkshire Hospitals NHS Trust and Hull York Medical School, Kingston upon Hull HU3 2JZ, UK
- Correspondence to: S Bhandari
- Accepted 29 December 2011
A 46 year old woman who had been treated for an uncomplicated E coli urinary tract infection three weeks previously returns to check her urine, as advised by her GP, since her initial urine dipstick test showed traces of protein, blood, leucocytes, and nitrites. She is a lawyer with no important medical history or previous urinary infections. She has recently taken a non-steroidal anti-inflammatory drug (NSAID) for menstrual pain. She has had no recent urinary symptoms or episodes of visible haematuria. There is no family history of renal disease, hypertension, or diabetes, and she is not pregnant.
On physical examination, her blood pressure is 144/82 mm Hg, and there is no oedema. However, the urine dipstick test shows ++ result for protein but no blood (see box 1 for glossary of terms).
Box1: Glossary of terms
The presence of dipstick positive proteinuria in two or more consecutive urine samples over a one to two week period. A previous urinary tract infection may theoretically affect the test, from the presence of leucocytes or alkalinisation of the urine leading to a false positive result for proteinuria. If the evidence is weak, the test should be repeated about two weeks later to check for persistent proteinuria. Repeat tests are more reliable for detecting renal pathology as they reduce the likelihood of transient proteinuria from exercise, stress, or fever.
This is rare in people over 30 years old. Usually it is manifest only when a person stands erect for long periods and disappears on lying down. The diagnosis is made by collecting split urine specimens for comparison. The daytime specimen typically has an increased concentration of protein, whereas a night time specimen has a normal concentration.
Proteinuria and albuminuria
These are an abnormal presence of protein and albumin in the urine. The normal laboratory ranges are
Protein <150 mg/day/1.73 m2
Albumin <100 mg/day/1.73 m2.
The composition of urinary protein can be considered to be about 40% albumin and low molecular weight immunoglobulins (including IgA and light chains), 40% secreted proteins (such as Tamm-Horsfall protein, synthesised by the kidney tubule), and the remainder being other immunoglobulins. Dipstick results of + or greater indicate at least 30 mg/dL (equivalent to about 600 mg/day for a 2 litre urine volume) of proteinuria or albuminuria.
Protein:creatinine ratio and albumin:creatinine ratio
Increased ratios of protein or albumin to creatinine in the urine indicate raised protein or albumin as a result of glomerular pathology or hyperfiltration. They are calculated by dividing the urine protein or albumin concentration by the urine creatinine concentration (which is assumed to be excreted uniformly over 24 hours) to take into account differences in urine volume. Normal ranges are
Albumin:creatinine ratio <3.5 mg/mmol for men, <2.5 mg/mmol for women
Protein:creatinine ratio <15 mg/mmol.
An albumin:creatinine ratio of 30 mg/mmol is equivalent to a protein:creatinine ratio of 50 mg/mmol, and a 24 hour protein of 500 mg/day, while an albumin:creatinine ratio of 70 mg/mmol is equivalent to a protein:creatinine ratio of 100 mg/mmol and a 24 hour protein of 1000 mg/day.
What is the next investigation?
Who should be tested?
The prevalence of proteinuria in the general population is about 2% and is higher in elderly people and those with comorbidities.1 2 Guidelines from the National Institute for Health and Clinical Excellence (NICE), based on systematic reviews of observational studies, recommend screening of groups considered at high risk of chronic kidney disease. These include patients with hypertension, diabetes, vascular disease (cardiovascular, peripheral, or cerebrovascular), a family history of renal disease, or multisystem disease with potential for renal involvement (such as systemic lupus erythematosus) and those taking nephrotoxic drugs.3 Screening is by estimation of glomerular filtration rate from serum creatinine and calculating urine albumin:creatinine ratio.
In this case, although the patient is asymptomatic, further evaluation is warranted, primarily because of the presence of proteinuria on a background of NSAIDs, to avoid missing significant renal pathology.
What to look for on physical assessment
Clinical history and examination often reveal few abnormalities unless features of multisystem disease are present or the degree of proteinuria is sufficient to cause physical signs such as frothy urine or peripheral oedema. In the current case further clinical assessment is likely to be negative, but a review of the patient’s history for symptoms such as joint pains, weight loss, or fever may be relevant in a case of persistent proteinuria (see box 1) to rule out systemic disorders such as lupus; indeed, a rash may suggest a vasculitis.
Confirmation of proteinuria
Ideally evidence of persistent proteinuria should be confirmed with an early morning urine sample to exclude postural proteinuria, but this is rarely practical. Urine dipstick testing for protein can detect a urinary concentration of albumin of 100-200 mg/L but is relatively insensitive to other plasma proteins such as immunoglobulin light chains, which might be present in a patient with myeloma-associated renal disease. Dipstick urine analysis is a poor method of quantifying proteinuria. Indeed, over-estimation or under-estimation may be compounded if the result is read manually rather than by an automated urine dipstick analyser. False negative tests are often seen in dilute urine (specific gravity <1.005) and when protein other than albumin is present in the urine. False positives can be seen in concentrated urine, basic urine (pH >8), and in the presence of haematuria.
A midstream urine sample should be sent for culture to exclude urinary tract infection, but the presence of leucocytes and nitrites on urine analysis (81% sensitive and 59% specific, negative predictive value 93%) with symptoms and signs should direct initial therapy because of the potential for contamination of a midstream urine sample, resulting in variable sensitivity (60-100%) and specificity (49-100%) for infection (see BMJ article on urinary tract infections).4 5
Quantification of proteinuria (albumin:creatinine ratio or protein:creatinine ratio)
A persistent positive urine dipstick test for proteinuria should prompt sending a sample for laboratory quantification by the ratio of albumin or protein to creatinine (see box 1). A systematic review has shown that testing of a random urine sample by protein:creatinine ratio had a sensitivity >90% to rule out pathology but low specificity (67%) to rule in a diagnosis.6
NICE recommends use of albumin:creatinine ratio because it has greater sensitivity (96.8%) than protein:creatinine ratio for low levels of proteinuria, whereas the Scottish Intercollegiate Guidelines Network (SIGN) recommends protein:creatinine ratio.3 7 8 Historically, measuring proteinuria by means of a 24 hour urine collection was the gold standard, but this is impractical, subject to inaccurate sample collection, expensive, and inconvenient. Studies such as RENAAL have shown that both protein:creatinine ratio and albumin:creatinine ratio correlate well with 24 hour urine measurements (r=0.93 and 0.8 respectively).9 10
In the current case the patient’s results for albumin:creatinine ratio and protein:creatinine ratio testing are 170 mg/mmol and 240 mg/mmol respectively, confirming significant proteinuria.
Creatinine and estimated glomerular filtration rate
Serum creatinine and estimated glomerular filtration rate should be measured to assess renal function as suggested by NICE.3 Other tests could be considered as detailed in box 2, but referral to a nephrologist for specific investigations to further characterise the pathology is warranted irrespective of estimated glomerular filtration rate (box 3).
Box 2: Summary of investigations for proteinuria
Basic investigations useful before referral to a nephrologist
Full blood count*
Biochemical profile or estimated glomerular filtration rate*
Protein:creatinine ratio or albumin:creatinine ratio*
Ultrasound scan of kidneys, ureters, and bladder*
Other potential specialist tests performed by nephrologist
Complement (C3, C4)
Serum or urine electrophoresis
Antistreptolysin O titre
Antineutophil cytoplasmic antibody
Antibodies to double stranded DNA
Antibody to glomerular basement membrane
*Useful to carry out in primary care
Box 3 Summary of individual criteria for referral to nephrologist*
Multisystem or collagen disease (such as systemic lupus erythematosus)
Resistant hypertension (>4 antihypertensive drugs including a diuretic)
Family history of renal disease (such as polycystic kidney disease)
Estimated glomerular filtration rate <60 μmol/L/1.73 m2
Albumin:creatinine ratio >70 mg/mmol or protein:creatinine ratio >100 mg/mmol
Albumin:creatinine ratio >30 mg/mmol or protein:creatinine ratio >50 mg/mmol if associated non-visible haematuria
Nephrotic range proteinuria (protein:creatinine ratio >300 mg/mmol) and hypoalbuminaemia (<30 g/L) refer urgently
*Based on NICE guidelines3
Guidelines based on expert opinion suggest a renal ultrasound scan should be considered to confirm the presence of two kidneys (a single kidney occurs in 1 in 1000 people and may lead to hyperfiltration and proteinuria), measure kidney size, and identify structural abnormalities such as polycystic kidneys.
Relation between proteinuria and cardiovascular morbidity or mortality
Large epidemiological studies involving the general population, cohorts at high risk for chronic kidney disease (for example, with diabetes), and cohorts with known chronic kidney disease indicate that proteinuria is an independent risk factor for adverse cardiovascular events and faster progression of kidney disease.11 12 13 14 The PREVEND (prospective study of a cohort of 40 845 in the general population) and ADVANCE (observational analysis of 10 640 patients with type 2 diabetes) studies found that both the degree of albuminuria and estimated glomerular filtration rate correlated with increased renal risk (up to 120-fold).15 The rate of cardiovascular events more than doubled for every 10-fold increase in baseline albumin:creatinine ratio (relative risk 2.48 (95% confidence interval 1.74 to 3.52)) and for every halving of baseline estimated glomerular filtration rate (relative risk 2.2 (1.09 to 4.33)).16 In addition, these two measurements had additive effects on the number of events.
This patient has clinically significant proteinuria (protein:creatinine ratio >200 mg/mmol), which invariably indicates glomerular pathology rather than a tubular problem. The wide differential diagnosis could include minimal change glomerulonephritis, focal segmental glomerulosclerosis, membranous glomerulonephritis, and diabetic nephropathy. The temporal association with NSAIDs and the known association between NSAID use and minimal change glomerulonephritis suggest that this is the most likely diagnosis, although >85% of cases are usually idiopathic. Modest proteinuria (protein:creatinine ratio 100-200 mg/mmol) might indicate an alternative of tubular pathology such as drug induced interstitial nephritis from possible antibiotic use or acute tubular necrosis. In this case serum creatinine may be abnormal. Isolated proteinuria of <100 mg/mmol requires monitoring only.
The presence or absence of concomitant haematuria is diagnostically helpful since haematuria is uncommon in patients with minimal change glomerulonephritis, focal segmental glomerulosclerosis, or diabetic nephropathy. The absence of non-visible haematuria is less in favour of other possibilities such as IgA nephropathy, polycystic kidney disease, vasculitis, collagen or multisystem disease, and post-infectious glomerulonephritis. Therefore cessation of NSAIDs may lead to resolution, but referral should be made to a nephrologist (box 2).
Explain to the patient that, because there is significant protein in the urine, further specialist evaluation is required. Also mention that the patient’s use of an NSAID might be responsible.
Withdrawal of the patient’s NSAID led to complete remission within two weeks, as seen in most cases (over 90% recover within 2-6 weeks after stopping NSAIDs). If this did not occur a trial of corticosteroids might be considered and renal biopsy performed to exclude other pathologies.
Recommend future use of NSAIDs should be avoided. Biannual monitoring of urinary protein is an important means of assessing response to therapy. Therefore, advise subsequent monitoring with dipstick testing (potentially by the patient) and suggest that the patient seek medical advice if oedema should occur as relapses tend to be abrupt and often clinically apparent. Identification of relapses allows early intervention and potential resolution in minimal change glomerulonephritis.
Proteinuria may indicate important glomerular or tubular pathology related to side effects from drugs such as NSAIDs
In patients with persistently positive dipstick tests, proteinuria is best quantified by measuring albumin:creatinine ratio or protein:creatinine ratio
Refer the patient if albumin:creatinine ratio >70 mg/mmol (or >30 mg/mmol if concomitant haematuria); protein:creatinine ratio >100 mg/mmol (or >50 mg/mmol if concomitant haematuria), or other renal pathology is present
Urgent referral for heavy proteinuria (protein:creatinine ratio >300 mg/mmol) and a low serum albumin (<30 g/L)
Proteinuria and impaired renal function (low estimated glomerular filtration rate) independently and additively correlate with risk of progression of renal disease and with cardiovascular events
Cite this as: BMJ 2012;344:e2339
This series of occasional articles provides an update on the best use of key diagnostic tests in the initial investigation of common or important clinical presentations. The series advisers are Steve Atkin, professor, head of department of academic endocrinology, diabetes, and metabolism, Hull York Medical School; and Eric Kilpatrick, honorary professor, department of clinical biochemistry, Hull Royal Infirmary, Hull York Medical School. To suggest a topic for this series, please email us at firstname.lastname@example.org.
Contributors: All the authors have contributed according to ICMJE guidelines. OTB contributed to planning, drafting, and revising the article and to approval of the final version. SB conceived of the article; contributed to planning, drafting, and revising the article; and approved the final version. SB is guarantor for the overall content. Professors Eric Kilpatrick, Ian Watt, and Unna Macleod reviewed the article and gave advice on its content.
Competing interests: All authors have completed the Unified Competing Interest form at www.icmje.org/coi_discosure.pdf (available on request from the corresponding author) and declare that SB has received honoraria from Pfizer, Amgen, and Pharmacosmos for lectures (all created solely by SB and not influenced by the sponsor); no financial relationships with any organisations that might have an interest in the submitted work in the previous three years, no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent not required (patient anonymised, dead, or hypothetical).