Saudi Journal of Kidney Diseases and Transplantation

: 2014  |  Volume : 25  |  Issue : 4  |  Page : 830--836

Correlation of fractional excretion of magnesium with steroid responsiveness in children with nephrotic syndrome

Jubaida Rumana1, Mohammed Hanif2, Golam Muinuddin3, Mohammed Maruf-ul-Quader4,  
1 Department of Pediatrics and Neonatology, Bangladesh Institute of Health Sciences and Hospital, Darus Salam, Mirpur, Bangladesh
2 Department of Pediatric Nephrology, Dhaka Shishu (Children) Hospital, Sher-a-Bangla Nagar, Bangladesh
3 Department of Pediatric Nephrology, Bangabandhu Sheikh Mujib Medical University, Shahbag, Bangladesh
4 Upazilla Health Complex Banshkhali, Chittagong, Bangladesh

Correspondence Address:
Jubaida Rumana
Department of Pediatrics, Bangladesh Institute of Health Sciences and Hospital, Darus Salam, Mirpur, Dhaka


Steroid-resistant nephrotic syndrome (SRNS) patients are candidates for other alter­native drug regimes, and the non-responsiveness to steroid is more common among glomerulo-nephritides other than minimal change disease. Without performing biopsy and proper renal histology, progression of the disease cannot be assessed. Fractional excretion of magnesium (FE Mg) has been found to correlate directly with various renal histologies. The aim of this study is to evaluate the relationship of FE Mg in children with the histological pattern in SRNS. In this prospective observational study, 40 children of nephrotic syndrome, both with the first episode as well as relapse, aged 1-12 years were included in the study. Of them, 20 were steroid-responsive cases and 20 were steroid-resistant cases. FE Mg was determined in all the patients and renal histology was performed in the steroid-resistant cases. A correlation was found between FE Mg and renal histology. Data were analyzed in SPSS program version-16. Comparison of two groups was performed by the Fisher exact test and unpaired t test. P-value less than 0.05 were considered to be significant. The results of histo-pathology showed that the mean difference in FE Mg was significant (P <0.001), as FE Mg was 7.0 ± 2.3% in mesangiocapillary glomerulonephritis, 6.9 ± 1.3% in focal segmental glomerulosclerosis, 4.7 ± 0.6% in immunoglobulin M nephropathy, 4.5 ± 1.2% in focal segmental proliferative glomerulo-nephritis, 4.4 ± 1.6% in minimal change disease, 4.2 ± 0.4% in diffuse mesangial proliferative glome-rulonephritis and 3.8 ± 1.3% in mesangial proliferative glomerulonephritis. There was a statistically significant difference between FE Mg in steroid-resistant nephrotic syndrome (4.9 ± 1.9) and steroid-responsive syndrome (1.2 ± 0.3). FE Mg is a simple, minimally invasive screening marker for SRNS, and is an early predictor of clinical outcome. It can be considered as an initial investigation where biopsy cannot be performed or indications are not clear.

How to cite this article:
Rumana J, Hanif M, Muinuddin G, Maruf-ul-Quader M. Correlation of fractional excretion of magnesium with steroid responsiveness in children with nephrotic syndrome.Saudi J Kidney Dis Transpl 2014;25:830-836

How to cite this URL:
Rumana J, Hanif M, Muinuddin G, Maruf-ul-Quader M. Correlation of fractional excretion of magnesium with steroid responsiveness in children with nephrotic syndrome. Saudi J Kidney Dis Transpl [serial online] 2014 [cited 2023 Feb 4 ];25:830-836
Available from:

Full Text


About 90% of cases of nephrotic syndrome (NS) in children are idiopathic and steroid-sensitive, with a favorable prognosis, [1] while the remaining 10% of cases do not respond to steroids within four weeks of steroid therapy and are defined as steroid-resistant nephrotic syndrome (SRNS). [2] NS is the most common presentation of glomerulonephritis in chil-dren. [3] Histologically, minimal change disease is common in 76.4%, membranoproliferative glomerulonephritis in 7.5%, focal segmental glomerulosclerosis in 6.9%, mesangial proli-ferative glomerulonephritis in 2.3% and mem­branous nephropathy in 1.5% of NS, and the remaining 5.4% are of other histological pat-terns. [4] More than 95% of children with mini­mal change NS respond to corticosteroid the­rapy. The percentage of steroid responsiveness in mesangial proliferative minimal change di­sease is 50% and that with focal segmental glomerulosclerosis is 20%. [5]

Glomerular diseases cause tubulointerstitial damage through multiple pathways, such as impaired glomerular selectivity of toxic subs­tances, altered glomerular hemodynamics, im-munologic mechanisms, inflammatory media­tors and leukocyte migration. These cause des­truction of the glomeruli, attack tubules and induce tubulointerstitial injury. [6] Thirty percent to 40% of steroid-resistant cases progress to end-stage renal disease within five years of disease process. Mesangial proliferative in 50%, membranoproliferative glomeruloneph-ritis in 20-30%, focal segmental glomerulo-sclerosis in 21% of cases and membranous nephropathy in 7.3% progress to end-stage renal disease within five years of disease. [7],[8] Minimal change lesion or mild mesangial proliferative glomerulonephritis shows benign clinico-pathology and is not associated with tubulointerestitial damage. Whereas, focal seg-mental glomerulosclerosis, a moderate to dif­fuse mesangial proliferative glomeruloneph-ritis and membrano proliferative glomerulo-nephritis, is associated with tubulointerstitial lesions and progresses to chronic renal insuf­ficiency and end-stage renal disease sooner or later. Hence, the early recognition of its cha­racteristic nature would help select a proper therapeutic regimen. For the diagnostic pur­pose, determination of the presence or absence of tubulointerstitial disease in any renal patient would possibly address this specific issue.

Tubular function test by mean of fractional excretion of magnesium (FE Mg) has recently been reported to correlate directly with the magnitude of tubulointerstitial fibrosis (TIF). It has been shown that FE Mg can reflect tubular function for both the ability of tubules for re-absorption of the filtered magnesium and for retaining the intracellular magnesium, which is the second most abundant intra-cellular cation after potassium. [9] In contrast, any disturbance to the structure and function of the tubular epithelium would affect tubular magnesium wastage and its reabsorption capa­city, thereby increasing the value of FE Mg above the normal range. [10] FE Mg of <2.2% indicates intact tubular function, whereas ele­vated FE Mg reflects tubular damage. The higher the value, the greater is the magnitude of TIF. [11] Therefore, FE Mg is a marker for TIF, which is a reflection of chronic kidney disease in an early stage where other markers of chronic kidney disease are normal.

In Bangladesh, evaluation of FE Mg has been performed in cases of glomerulonephritis and diabetic nephropathy, but its association with the NS was first assessed in the present study. The aim of the study was to evaluate FE Mg in children with NS and to determine the asso­ciation of FE Mg with the histological pattern in steroid-resistant NS. FE Mg is an early screening of clinical severity and predictor of clinical outcome associated with SRNS. An early detection of tubular damage through increased FE Mg will help identify the disease process and initiate the appropriate treatment so that the deterioration of the pathological process could be prevented by modifying the treatment policy.

 Materials and Methods

In this prospective, observational study, the patients of NS aged between 1 and 12 years were included. This patients had generalized edema, proteinuria (40 mg/m [2] /h), hypoalbu-minemia (serum albumin of <2.5 g/dL) and hypercholesterolemia (serum cholesterol of >200 mg/dL). This study was performed in the Pediatric Nephrology Department of Dhaka Shishu (Children) Hospital and Bangabandhu Sheikh Mujib Medical University during June 2009‒July 2010. During this period, we had come across 20 patients of SRNS who were not responding to steroids within four weeks of steroid therapy. Among them, 16 patients were initially steroid resistant and four patients had late steroid resistance. The first 20 hospitalized patients who were responding to steroids within four weeks were included in the steroid-sensitive nephrotic syndrome (SSNS) group. Of these patients, 14 patients had the first episode of NS and six patients had infrequent relapses of NS.

The patients having NS due to systemic diseases such as lupus nephritis, Henoch Scholen purpura and poly arteritis nodosa were excluded. Also, patients who were suf­fering from diseases wherein magnesium is lost, such as gastro-intestinal disorders and burn, were excluded. The patients were also screened for absence of hepatitis B surface antigen, tuberculosis, heart failure and lupus nephritis. None of the patients had received nephrotoxic drugs such as loop diuretics, aminoglycoside and radiocontrast dye before the collection of their samples. Patients having concomitant urinary tract infections were trea­ted with an appropriate antibiotic before the collection of the urinary samples.

From each patient, 2 mL of blood and 10 mL of spot urine was collected. After collection, the samples were preserved appropriately till the time of testing. The levels of magnesium in blood and urine were then measured using the magnesium flex (R) reagent cartridge (Cat no. DF57) using a biochemistry auto analyzer machine (model: The Dimension® RxL Max® integrated chemistry system; Siemens Healthcare Diagnostics Inc., Newyork, DE, USA) fol­lowing the atomic absorption spectrophoto­meter method. The serum and urine creatinine levels were also measured by standard methods in our laboratory. The same calibrators and reagents were used for every patient. FE Mg was then calculated using the following formula:


The person who had performed the tests was blinded about the type of NS and also the patients did not have any idea about the test results. Renal biopsy was performed in all the 20 cases of SRNS. Histological and direct immunofluorescent examinations were per­formed by the histo-pathologist who had no information about the tubular functional value (FE Mg) of each individual.

The SPSS software (version 16.0) was used for analyzing data (SPSS, Chicago, IL, USA). The comparison of the sample mean of two quantitative variables was determined by the non-parametric method using the Fisher's exact test and unpaired t-test. A P-value of < 0.05 was considered significant. Each patient was on follow-up for one year and none of the SSNS patients developed SRNS.

Informed consent was obtained from the pa­rents of each patient. The institutional ethics committee approved the protocol of the study.


[Figure 1] shows the distribution pattern of NS among the patients. [Table 1] presents the basic characteristics of the study subjects. Their mean age, gender, weight, height, serum albu­min level, serum cholesterol level, spot protein creatinine ratio and serum creatinine level did not show any significant association between the two groups. Hypertension was present in 30% of the cases in the SRNS group, but no patient had hypertension in the SSNS group. Presence of hypertension was statistically sig­nificant between the two groups (P-value = 0.010).{Figure 1}{Table 1}

[Table 2] shows that all the patients had a high level of FE Mg in SRNS (mean ± SD, 4.9 ± 1.9). However, all the patients had normal FE Mg (1.2 ± 0.3) in SSNS. The mean difference in FE Mg was significant (P <0.001) between the two groups.{Table 2}

Histopathology of the steroid-resistant pa­tients was as follows: Five patients had mini­mal change disease; four patients had focal segmental glomerulosclerosis, three patients had focal segmental proliferative glomerulonephritis, two patients had mesangial proliferative glomerulonephritis, two patients had immunoglo-bulin M nephropathy, two patients had diffuse mesangial proliferative glomerulonephritis and two patients had mesangio capillary glome-rulonephritis [Figure 2].{Figure 2}

The mean ± SD of FE Mg was 7.0 ± 2.3% in mesangiocapillary glomerulonephritis, 6.9 ± 1.3% in focal segmental glomerulosclerosis, 4.7 ± 0.6% in immunoglobulin M nephropathy, 4.5 ± 1.2% in focal segmental proliferative glomerulonephritis, 4.4 ± 1.6% in minimal change disease, 4.2 ± 0.4% in diffuse mesangio-proliferative glomerulonephritis and 3.8 ± 1.3% in mesangial proliferative glomerulonephritis [Table 3]. {Table 3}

The mean difference of FE Mg between the histopathological groups was significant (P <0.001). Five SRNS cases showed minimal change disease, and their FE Mg mean was 4.4 ± 1.6, i.e. more than normal (normal FE Mg: <2.2%). Other SRNS cases of different histo-pathologies also showed increased FE Mg.


The results of the study show that FE Mg was higher in the SRNS patients than in the steroid-responsive patients. FE Mg varied according to the different histopathologies of the steroid-resistant cases. FE Mg was more than normal in minimal change disease and was abnor­mally elevated and significant in diffuse me-sangial proliferative glomerulonephritis, focal segmental glomerulosclerosis, immunoglobulin M nephropathy, mesangial proliferative glo-merulonephritis, mesangiocapillary glomerulo-nephritis and focal segmental glomuloneph-ritis. Futrakul and Deekajorndech showed ab­normally elevated FE Mg in focal segmental glomerulosclerosis, which is similar to the fin­dings of the present study. [12],[13] Results of a study by Futrakul et al also showed increased FE Mg in diffuse mesangial proliferative glo-merulonephritis cases, which is also similar to the findings of the present study. Mesangial proliferative glomerulonephritis and focal seg-mental glomulonephritis were not included in other studies (Futrakul et al 1999; Deeka-jorndech 2007), which were important causes of SRNS included in our study. Of the 20 patients of SRNS, five (25%) showed minimal change in histopathology in our study. In a retrospective study by Malades among 171 patients, 62 cases (34%) showed a mini­mal change in histopathology. In another study by ISKDC among 354 patients of NS, 55 were SRNS; of them, 45.5% had minimal change disease. [8] High FE Mg in minimal change di­sease may be explained by the fact that some focal segmental glomerulosclerosis patients behave like minimal change disease initially, and subsequent fully evolve into focal seg-mental glomerulosclerosis. Also, the site of biopsy taken could be another explanation. In focal segmental glomerulosclerosis, less than 70% of the glomerulus is affected focally, and if the biopsy was obtained from the remaining 30% area, it can show a picture of minimal change disease.

In an intrarenal hemodynamic study, it was shown that the tubular transport defect corre­lates inversely with the magnitude of peri-tubular capillary blood flow. [14] Dysfunctioning glomerular endothelium was demonstrated in chronic kidney disease. [13] Therefore, all these act in concert for damaging and dysfunc-tioning of the tubule. TIF is the final common pathway to end-stage renal disease. Under­standing the mechanism of TIF is, therefore, essential in establishing the novel therapeutic strategies for the prevention or arrest of prog­ressive kidney diseases. [13] FE Mg appears to be the most sensitive index to discriminate the early lesion of TIF. [8] Futrakul et al inves­tigated patients with NS for the most sensitive markers of tubular dysfunction. [13] They de­monstrated FE Mg as a sensitive index to de­tect early abnormality of the tubular structure and function. In the clinical settings where kidney biopsy cannot be performed or the indication for kidney biopsy is not obvious, FE Mg is an alternative approach that may provide useful information relevant to the pre­sence or absence of tubulointerstitial disease or fibrosis. [15]

NS is the most common cause of chronic kidney disease in children, and is more com­mon in children aged over five years. [16] FE Mg is the most sensitive marker in detecting the low level of tubular injury in humans. [17] FE Mg can detect the tubulointerstitial damage in SRNS in the earliest possible time. The results of the present non-invasive FE Mg investi­gation shows that it will help the early detec­tion of a potential progressive kidney disease and will guide us to the early initiation of appropriate treatment when there is still an adequate renal functional reserve. The FE Mg test does not cost more than US $3 (Tk 250), and the test can be performed in any laboratory where determination of magnesium level and creatinine level is possible. Therefore, an early appropriate treatment in non-responding NS will reduce the morbidity and mortality of the disease and will be convenient for both phy­sicians and patients in our country set-up.

The main limitation of the present study is that it was performed in only 40 NS patients; of them, 20 patients were steroid resistant. The sample size obtained was based on the time factor of the research as the allocated time was only one year. During this period, we had come across only 20 SRNS patients. Secondly, only two centers were included in this study. A large-scale study with a longer follow-up pe­riod is necessary for further evaluation of these findings.


Based on the findings of the study, in the clinical situations where kidney biopsy cannot be performed, FE Mg is an alternate approach, which appears as a valid, statistically signi­ficant test for SRNS and can help one to implement early appropriate treatment.


The authors are grateful to the ACME Laboratories for their financial support, which helped them to do the work smoothly. The authors owe their heartiest thanks to M. Shamsul Islam Khan, Advisor for Health Informatics and Library Sciences and Prof. M. A. Hafez, departmental head of bio-statistics at the Bangladesh Institute of Health Sciences and Hospital, for their generous help and in­formative suggestions. They are also grateful to the study patients and their parents for their full cooperation.


1Eddy AA , Symons JM. Nephrotic syndrome in childhood. Lancet 2003;362:629-39.
2Hoyer PF, Vester U, Becker U. Steroid resistant Nephrotic Syndrome. Comprehensive Pediatric nephrology. 1 st ed. Mosby, Elsevier, Philadelphia, USA: 2008. p. 257-67.
3Madani A, Fahimi D, Esfehani ST, et al. Glo-merular diseases in Iranian children clinico-pathological correlation. Pediatr Nephrol 2003; 18:925-8.
4ISKDC Primary Nephrotic Syndrome in chil­dren. Identification of patients with minimal change nephrotic syndrome from initial res­ponse to predinisolon. J Pediatr 1981;98:561-4.
5Pais P, Avner DE. Idiopathic Nephrotic syn­drome. Nelson text book of Pediatrics. 19 th ed. Philadelphia: Elsevier; 2011. p. 1804-5.
6Nath KA. The tubulointerestitium in progress-sive renal disease. Kidney Int 1998;54:992-4.
7Bagga A, Srivastava RN. Nephotic syndrome. Pediatric nephrology, 5 th ed. New Delhi: Jaypee Brothers Medical Publishers; 2011. p. 219-24.
8Niaudut P, Boyer O. Idiopathic nephrotic syn­drome in children: Clinical aspects. Pediatric nephrology, 6 th ed. Heidelberg: Springer; 2009. p. 667-702.
9Futrakul P, Yenrudi S, Futrakul N, et al. Tubular function and tubulointerstitial disease. Am J Kidney Dis 1999;33:886-91.
10Deekajorndech T. Fractional excretion of mag­nesium in systemic lupus erythematososus. J Med Assoc Thai 2005;88:743-5.
11Futrakul N, Futrakul P. Prevention of end-stage renal disease: An innovative strategy. Bangkok: Chulalongkorn University Printing House; 2008. p. 2-31.
12Futrakul N, Panichakul T, Sirisinha S, Futrakul P, Siriviriyakul P. Glomerular endothelial dys­function in chronic kidney disease. Ren Fail 2004;26:259-64.
13Iwano M, Neilson EG. Mechanism of tubulo-interstitial fibrosis. Curr Opin Nephrol Hypertens 2004;13:279-84.
14Gheissari A, Andalib A, Labibzadeh N, Modarresi M, Azhir A, Merrikhi A. Fractional excretion of magnesium (FEMg), a marker for tubular dysfunction in children with clinically recovered ischemic acute tubular necrosis. Saudi J Kidney Dis Transplant 2011;22:476-81.
15Deekajorndech T. A biomarker for detecting early tubulointerstial disease and ischaemia in glomerulonephropathy. Ren Fail 2007;29:1013-7.
16Bhimma R, Adhikari M, Asharam K, Connolly C. The spectrum of kidney disease (stage 2-5) in KwaZulu-Natal, South Africa. Pediatr Nephrol 2008;23:1841-6.
17Emeigh Hart SG. Assesment of renal injury in vivo. J Pharmacol Toxicol Methods 2005;52: 30-45.