The risk factors for children with primary nephrotic syndrome: a systematic review and meta-analysis.

BACKGROUND: To evaluate the risk factors of primary nephrotic syndrome (PNS) with urinary tract infection (UTI) in children.
METHODS: Multiple databases including PubMed, Excerpta Medica Database (EMBASE), Web of Science (WOS), the Cochrane Library, and China National Knowledge Infrastructure (CNKI) were used to search for relevant studies, which were full-text articles involved in the evaluation of differences between PNS with UTI and without UTI. All included articles were assessed for quality and the data analyses were conducted with Review Manager (5.4). Forest plots, sensitivity analysis, and bias analysis were also performed on the included articles.
RESULTS: Eight studies were included in this meta-analysis, with a total of 3,274 patients. Meta-analysis showed that a low level of serum albumin [mean difference (MD): -0.32 g/dL; 95% confidence interval (CI): (-0.55, -0.08); P=0.008], a low level of serum total protein [MD: -0.16 g/dL; 95% CI: (-0.20, -0.12); P<0.00001], a high level of urinary protein [MD: 5.09 mg/d; 95% CI: (3.13, 7.05); P<0.00001], a lower level of serum urea nitrogen [MD: -0.10 mg/dL; 95% CI: (-0.18, -0.02); P=0.01], and a higher level of serum cholesterol [MD: 2.26 mg/dL; 95% CI: (0.74, 3.78)] had a higher risk of PNS with UTI. There was no obvious publication bias among included studies. DISCUSSION: Our research demonstrated that a low level of serum albumin, a low level of serum total protein, a high level of urinary protein, a low level of serum urea nitrogen, and a high level of serum cholesterol were the risk factors of PNS with UTI in children. 2021 Translational Pediatrics. All rights reserved.

Introduction

Primary nephrotic syndrome (PNS) is the most common glomerular disease and chronic kidney disease in childhood (1-3). Due to the increased permeability of the glomerular filtration membrane, a large amount of protein is lost from urine, resulting in significant proteinuria, hypoproteinemia, edema, and hyperlipidemia (4,5). PNS refers to those caused by glomerular disease, the exact cause and pathogenesis are not yet fully understood. It is unanimously recognized that it is an immune-mediated inflammatory disease. Compared with adults, children are more prone to PNS, with 2–7 cases per 100,000 children every year (6). In children, it is mainly minimal degenerative nephropathy, mesangial proliferative nephritis and focal segmental glomerulosclerosis, but in adult, it is mainly mesangial proliferative nephritis and membranous proliferative nephritis (3).

Urinary tract infection (UTI) is the second most common complication of PNS (7). Among all the infections, UTI is of special concern as most of the UTIs in PNS are asymptomatic, which easily leads to missed diagnosis and misdiagnosis of asymptomatic bacteriuria, and affects the efficacy of PNS treatment (8,9). The risk of UTI before the age of 14 is about 1–3% in boys and 3–10% in girls (10). The incidence rate of UTI varies from 4% to 0.4% in school age and preschool children. In children with PNS, the incidence of UTI is about 30% (11).

Studies have shown that UTI has a higher incidence in children with recurrent PNS (12-14). UTI may be the cause of decreased hormone sensitivity and the recurrence of renal disease (15). In severe cases, UTI may lead to persistent renal damage, scarring, and even end-stage renal disease (16). The possibility of UTI should be considered in PNS patients who have been hospitalized repeatedly or for a long time. Therefore, early and accurate diagnosis and treatment of UTI is important in these patients (17,18).

Although it is important to explore the risk factors of PNS with UTI in children for treatment and prevention, there is no systematic evaluation to estimate the risk factors worldwide. Therefore, we conducted a search for relevant studies about the risk factors for PNS with UTI, and performed the first systematic review and meta-analysis to study the risk factors of PNS with UTI in children.

We present the following article in accordance with the PRISMA reporting checklist (available at https://dx.doi.org/10.21037/tp-21-468).

Methods

Literature search strategy

Seven electronic databases, including PubMed, Excerpta Medica Database (EMBASE), Web of Science (WOS), the Cochrane Library, and China National Knowledge Infrastructure (CNKI), were systematically searched from their inception until July 2021. We used the following keywords: (I) primary nephrotic syndrome; (II) urinary tract infection; (III) children. After that, the terms were combined using the Boolean operator “AND” in order to obtain articles that included two or more of the words used in the search. We performed a comprehensive search to identify all relevant studies regardless of language or publication status (published, unpublished, in press, and ongoing). Additional reports were identified by reviewing the reference lists of eligible studies and review articles.

Study selection

The studies had to fulfill the following criteria to be included in the analysis:

Patients younger than 18 years old;

Patients with PNS;

Containing indicators evaluating the difference between PNS with UTI and without UTI;

Only articles with the full text available were selected.

Studies were excluded for the following reasons:

Research did not meet the inclusion criteria;

The risk factors of interest were not reported or impossible to use;

Review, abstract, or duplicate publication.

Data extraction and quality assessment

Data were extracted in duplicate by two investigators independently and were input into a dedicated database. The following details were extracted whenever available: first author’s name, patient age and gender, country of origin, year of publication, language, sample size, years of onset, and risk factors of interest. The quality of studies included in the review was evaluated by two independent reviewers through subjective judgment, with differences resolved by consensus or through a third reviewer if required. The Cochrane risk of bias tool 16 was used to reveal the risk of bias, and Review Manager was used to generate the risk of bias graph.

Statistical analysis

Review Manager (version 5.4, The Cochrane Collaboration, 2020) was adopted to estimate the difference of risk factors among selected reports. Continuous variables were evaluated by the mean difference (MD). Dichotomous variables were evaluated by the relative risk (RR). The chi-square test was used to test the heterogeneity among the included studies, and the heterogeneity was quantified by combining with I2. I2 values of 25%, 50%, and 75% were regarded as low, moderate, and high heterogeneity, respectively. A random effects model was used if heterogeneity was observed, while a fixed effects model was applied in the absence of inter-study heterogeneity. We also planned to perform sensitivity analyses based on the quality and weight of the trials by excluding each individual trial in turn. Publication bias assessments using a funnel plot and Egger’s test were conducted to assess publication bias.

Results

Search process

A total of 1,177 articles were identified by the screening electronic search strategy. After removal of duplicates and screening through the titles and abstracts, 1,020 articles were excluded. Of the remaining 157 full-text articles, 149 were also removed as they were ineligible article designs or had no relevant data. Finally, 8 studies met our inclusion criteria and were selected for the present meta-analysis (19-26). shows the details of our literature search and selection process.

Figure 1

Flowchart of study selection.

Characteristics of included studies

The basic features of the included studies are shown in . Of the 8 studies, 6 studies were published in English, 2 were in Chinese, 3 studies were from Bangladesh, 2 were from China, 2 were from India, and the other 1 was from Iran. These studies contained a total of 3,274 PNS patients (696 with UTI and 2,578 without UTI). The prevalence of UTI ranged from 15.0% to 50.8%. The risk factors of interest included serum albumin, serum total protein, urinary protein, serum urea nitrogen, serum cholesterol, serum triglycerides, and serum creatinine.

Table 1

Basic characteristics of the included studies in the systematic review

Study	Country	Language	Prevalence of UTI (%)	No. of patients			Gender (M/F)			Age		Years of onset	Risk factors*
				Case	Control		Case	Control		Case	Control
Gulati, 1996	India	English	15.2	37	206		27/10	156/50		<16.0	<16.0	–	1,3,5
Zhao, 1999	China	Chinese	31.4	32	70		–	–		1–14	1–14	–	1,6
Afroz, 2013	Bangladesh	English	40.9	45	56		25/20	32/24		1–12	1–12	–	1,5
Chen, 2014	China	Chinese	50.8	62	60		20/42	22/38		2–14	2–14	March 2008 to March 2012	1,3,7
Barua, 2016	Bangladesh	English	30.8	16	36		8/8	22/14		2–6	2–6	January 2009 to December 2009	1,5
Salarzaei, 2017	Iran	English	37.9	47	77		25/22	46/31		5.48±3.36	6.57±3.18	2015	1,2,3,4,5,6,7
Kundu, 2018	Bangladesh	English	40.3	25	37		15/10	24/13		6.12±3.25	7.26±3.39	June 2015 to March 2016	1,2,3,4,5,6,7
Narain, 2018	India	English	15.0	432	2,036		333/99	1,507/529		7.8±2.1	8.2±2.5	January 2000 to November 2016	1,2,3,4,5

*, 1: serum albumin; 2: serum total protein; 3: urinary protein; 4: serum urea nitrogen; 5: serum cholesterol; 6: serum triglycerides; 7: serum creatinine. UTI, urinary tract infection.

Results of quality assessment

Direct comparisons and the risk of bias assessment were performed using Review Manager (version 5.4). Among the included studies, a high risk of attrition bias was found in 3 studies, 2 studies showed a high risk of reporting bias, and 1 showed a high risk of performance bias (). presents a summary of the risk of bias for each included study.

Figure 2

Proportion of studies with low (green), high (red), or unclear (yellow) risk of bias.

Figure 3

Risk of bias summary of the included studies.

Results of heterogeneity tests

Serum albumin and serum total protein

In 8 studies involving 3,274 patients, a lower level of serum albumin had a higher risk of PNS with UTI [MD: −0.32 g/dL; 95% confidence interval (CI): (−0.55, −0.08); P=0.008; ], with significant heterogeneity (I2=92%; P<0.00001). The I2 of the sensitivity analysis for serum albumin decreased to 80% by removing the study by Narain 2018 (21), indicating that the heterogeneity was mainly due to Narain 2018. The result was not significantly altered after sensitivity analysis [MD: −0.25 g/dL; 95% CI: (−0.44, −0.05); P=0.01].

Figure 4

Forest plot: PNS with UTI versus PNS without UTI for serum albumin and serum total protein. PNS, primary nephrotic syndrome; UTI, urinary tract infection.

In 2,654 patients across 3 studies, a lower level of serum total protein had a higher risk of PNS with UTI [MD: −0.16 g/dL; 95% CI: (−0.20, −0.12); P<0.00001; ], and the pooled studies were homogeneous (P=0.25; I2=29%).

Urinary protein and serum urea nitrogen

Five studies involving 3,019 patients reported on urinary protein, and the pooled analysis showed that a higher level of urinary protein significantly increased the risk of PNS with UTI [MD: 5.09 mg/d; 95% CI: (3.13, 7.05); P<0.00001; ], with significant heterogeneity (I2=97%; P<0.00001). The I2 of the sensitivity analysis for urinary protein did not change significantly, suggesting that the meta-analysis result was robust.

Figure 5

Forest plot: PNS with UTI versus PNS without UTI for urinary protein and serum urea nitrogen. PNS, primary nephrotic syndrome; UTI, urinary tract infection.

Three studies reported on serum urea nitrogen, and the meta-analysis showed that a lower level of serum urea nitrogen had a higher risk of PNS with UTI [MD: −0.10 mg/dL; 95% CI: (−0.18, −0.02); P=0.01; ], and the pooled studies were homogeneous (P=0.57; I2=0%).

Serum cholesterol and serum triglycerides

Six studies reported on serum cholesterol, and the meta-analysis showed that the higher the level of serum cholesterol, the higher the risk of PNS with UTI [MD: 2.26 mg/dL; 95% CI: (0.74, 3.78); P=0.004; ]. The pooled studies were heterogeneous (P=0.0004; I2=78%). The I2 of the sensitivity analysis for serum cholesterol decreased to 52% by removing the study by Salarzaei 2017 (20), indicating that the heterogeneity was mainly due to Salarzaei 2017. However, the result was not significantly changed after sensitivity analysis [MD: 2.72 g/dL; 95% CI: (−1.79, 3.65); P<0.00001].

Figure 6

Forest plot: PNS with UTI versus PNS without UTI for serum cholesterol and serum triglycerides. PNS, primary nephrotic syndrome; UTI, urinary tract infection.

In 288 patients across 3 studies, the pooled analysis showed that the level of serum triglycerides had no association with the risk of PNS with UTI [MD: −56.92 mg/dL; 95% CI: (−143.44, 29.61); P=0.20; ]. The pooled studies were heterogeneous (P=0.02; I2=74%), and the result was robust after sensitivity analysis.

Serum creatinine

Four studies involving 2,776 patients reported on serum creatinine, and the pooled analysis showed that the level of serum creatinine had no association with the risk of PNS with UTI [MD: 0.02 mg/dL; 95% CI: (−0.17, 0.20); P=0.87; ], with significant heterogeneity (I2=92%; P<0.00001). The I2 of the sensitivity analysis for serum albumin decreased to 50% by removing the study by Chen 2014 (24), indicating that the heterogeneity was mainly due to Chen 2014. The result was not significantly changed after sensitivity analysis [MD: −0.08 mg/dL; 95% CI: (−0.16, 0.01); P=0.09].

Figure 7

Forest plot: PNS with UTI versus PNS without UTI for serum creatinine. PNS, primary nephrotic syndrome; UTI, urinary tract infection.

Publication bias

A funnel plot was generated to qualitatively evaluate the publication bias for serum albumin. The shape of the funnel plot showed some evidence of asymmetry (), however, Egger’s test was nonsignificant (P=0.58), which indicated that there was no obvious publication bias.

Figure 8

Funnel plot for potential publication bias.

Discussion

Clinical studies suggest that the reasons for PNS combined with UTI are as follows: (I) renal function is damaged and urinary protein level is increased; (II) residual proteinuria in the urethra is a good medium for pathogen growth; (III) hypoproteinemia leads to a decrease in leukocyte production; (IV) the most effective drugs for PNS are glucocorticoids and immunosuppressants (27,28). At present, the incidence of PNS combined with UTI is increasing year by year, and has become the primary problem in the prevention and control of infectious diseases all over the world (29,30).

Due to tissue edema, loss of immunoglobulins, and the use of immunosuppressants, the defense system of children with PNS is damaged, leading to low immunity and infection, which may even be life-threatening (31). In this paper, the risk factors of PNS with UTI were studied by meta-analysis.

This study showed that the levels of serum albumin, serum total protein, urinary protein, serum urea nitrogen, and serum cholesterol can affect the development of PNS with UTI. Specifically, a low level of serum albumin [MD: −0.32 g/dL; 95% CI: (−0.55, −0.08); P=0.008], a low level of serum total protein [MD: −0.16 g/dL; 95% CI: (−0.20, −0.12); P<0.00001], a high level of urinary protein [MD: 5.09 mg/d; 95% CI: (3.13, 7.05); P<0.00001], a lower level of serum urea nitrogen [MD: −0.10 mg/dL; 95% CI: (−0.18, −0.02); P=0.01], and a higher level of serum cholesterol [MD: 2.26 mg/dL; 95% CI: (0.74, 3.78); P=0.004] were the risk factors for PNS with UTI in children.

The loss of plasma protein from urine and the protein filtered out from the glomerulus are metabolized in renal tubules, leading to hypoproteinemia, low resistance, malnutrition, and urethral mucosal edema in children, making them more prone to UTI (32). Some studies have pointed out that hypercholesterolemia can inhibit the synthesis of cell membrane proteins, block the expression of cell membrane molecules and adhesion molecules, and can affect the mutual recognition and contact between cells. It can further affect the recognition, proliferation, activation, and differentiation of immune cells, damage humoral and cellular immune functions, and weaken the antigen presentation, phagocytosis, and bactericidal effect of macrophages, ultimately leading to the occurrence of UTI (33,34). Urinary residual proteinuria is a good culture medium for pathogens, so the higher the level of proteinuria, the higher the risk of UTI (35,36).

There were some limitations in this meta-analysis. Firstly, some studies did not explain the diagnostic criteria of UTI and PNS, the description of the determination method of research factors was not detailed, some data was incomplete, the inclusion criteria were limited, and the final sample size was limited. Secondly, the heterogeneity tests showed that several outcome variables were highly heterogeneous, which may be related to the differences in measurement methods across different countries. In addition, the research factors were not comprehensive, such as only the analysis of serum albumin, urinary protein and other factors that be of interest to most researchers.

In conclusion, a low level of serum albumin, a low level of serum total protein, a high level of urinary protein, a low level of serum urea nitrogen, and a high level of serum cholesterol were the risk factors of PNS with UTI in children.

The article’s supplementary files as