† These authors contributed equally.
Background: Many studies have assessed the role of circulating
neutrophil gelatinase-associated lipocalin (NGAL) on the risk of gestational
diabetes mellitus (GDM), but the results remain uncertain. Thus, this study aimed
to assess the association between NGAL and GDM risk by performing a
meta-analysis. Methods: We carried out a systematic search of electronic
databases (PubMed, Embase, Wanfang and Chinese National Knowledge Infrastructure
databases) to retrieve all related studies. The estimates of standardized mean
difference (SMD) and its 95% confidence interval (CI) were calculated in a
random-effects model. Between-study heterogeneity was assessed using I
Gestational diabetes mellitus (GDM), which is the most common metabolic disorder
during pregnancy, is defined as glucose intolerance that is first recognized in
pregnancy and affects around 5% of all pregnancies [1, 2, 3]. Although there is no
consensus on the national guidance of GDM, pregnant women with indicative results
of DM on the basic of diagnostic criteria of World Health Organization (WHO)
(fasting glucose
Insulin resistance, which has been known as one of the key pathophysiological features observed in type 2 diabetes as well as GDM [8], was found to present prior to conception and persisted across pregnancy [9], and is suggested to play a vital role in the mechanism of GDM. The insulin sensitivity is defined by the reduced sensitivity by approximately 50–60% in late pregnancy as opposed to pre-pregnancy among pregnant woman with/without normal glucose tolerance [10]. Neutrophil gelatinase-associated lipocalin (NGAL), also named lipocalin-2, first identified as a matrix protein of specific granules of human neutrophils [11], was found to exert an effect on the regulation of insulin sensitivity and was associated with insulin resistance [12]. To further investigate the potential role of NGAL in the GDM development, a broad range of studies [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30] was conducted to evaluated the differences of circulating NGAL levels between GDM cases and healthy pregnant women. But large variations in the results were observed across studies, a little difference was obtained from studies by Sweeting AN et al. [28] and Wang YY et al. [29] compared to the relatively significant differences obtained from most studies [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]. There is a need for more convincing pooled estimate, and further explorations for the source of underlying between-study heterogeneity that influences the concentrations of circulating NGAL.
Up-to-date, the most clinically and cost-effective methods of GDM screening remain controversial, and the ability to diagnose GDM in the first trimester compares to the second or third trimester of pregnancy keep inconsistent [31]. Thus, this meta-analysis was conducted to verify the circulating NGAL levels in those with GDM compared to healthy pregnant women, and potential influencing factors, which might be useful in providing more evidence to detect GDM in early pregnancy.
Two investigators (Zhu Chen and Shuyu Wang) conducted 2-step literature searches. Firstly, the webs of knowledge database were systematically searched, including PubMed, Embase, Chinese National Knowledge Infrastructure (CNKI) [32] and Wanfang databases [33], to collect articles reported on the levels of circulating NGAL among patients with GDM. The search was extended up to October 2019 without language restriction. The following key words were used for several combinations searching: (lipocalin or NGAL or LCN-2) and (gestational diabetes[mesh] or gestational diabetes). Secondly, the reference lists of related articles or citations of possible previous reviews or meta-analyses were also investigated to perform a manual search of other eligible studies. Articles identified by these two steps of searching were then screened for the selection bias based on the abstracts. The eligibility of articles identified with the selection process were then assessed based on the full-text review.
Studies were included in this meta-analysis on the basis of the following inclusion criteria: (1) original observational studies (case-control and cohort studies); (2) the diagnosis of GDM followed well accepted guidance or diagnostic criteria; (3) the circulating NGAL levels among GDM patients and healthy pregnant women were reported; (4) at least 20 cases were included; and (5) full text and complete data were available.
The exclusion criteria were: (1) abstracts, reviews, case reports, letters, comments or conference; (2) studies that did not provide sufficient data; (3) NGAL levels were not measured using blood sample; and (4) republished studies. For studies with overlapping study populations, only the largest study with the most recent datasets would be included in the final analysis.
Two reviewers applied the inclusion and exclusion criteria independently. The following data were extracted from each eligible study: first author’s name, publication year, country (region) and ethnicity of the population, GDM diagnostic criteria, number of GDM patients and controls, matching methods, age, gestation weeks of recruitment, mean circulating NGAL level with standard deviation (SD) and their ranges of the participants, sample sources. The principal investigators would be contacted for further information when these data were missing, and studies would be excluded if the principal investigator could not or deny to provide further data. Any discrepancies between two reviewers were resolved by discussion among all co-authors.
The qualities of observational studies (case-control and cohort studies) were
evaluated using the Newcastle–Ottawa scale (NOS) [34]. Using this scale, the
quality of each study is judged on three broad perspectives (eight items),
including participant selection (3 or 4 stars), comparability of study groups (1
or 2 stars) and assessment of outcome or exposure (2 or 3 stars). One star would
be awarded as one point and the highest quality of studies could be up to nine
star (point), with a higher score (
To assess the variations in circulating levels of NGAL between GDM cases and
control groups, the pooled standardized mean difference (SMD) with 95%
confidence interval (CI) was calculated to yield an overall effect size by using
a random-effects model. Between-study heterogeneity was assessed using Cochran’s
Q statistic and the Higgins’ I
As it showed in Fig. 1, a total of 164 articles were retrieved after initial searching and 31 duplicates were excluded by endnote. After reviewing the title and abstract, 101 articles were excluded and the full texts of 32 articles were identified, and 15 articles were further excluded because they did not meet the selection criteria. In total, 17 articles (19 results) met the criteria and were finally included in our analysis [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30].
Flow chart of study selection in the meta-analysis.
General characteristics of the included studies are presented in Table 1 (Ref. [13, 15, 16, 17, 18, 19, 20, 22, 23, 24, 25, 26, 27, 28, 30, 37, 38]). All the data from 17 included articles, published from 2009 to 2020, covered 1080 GDM patients and 1736 normal pregnant women. Almost all included studies (15/17) were carried out in China, one study in Italy [13], and one conducted in multi-ethnicities populations (Caucasians, East Asians, and South Asians) [28]. The sample size ranged from 26 to 248 among GDM cases, and 21 to 732 among healthy pregnant women. There were three studies with circulating NGAL levels measured during first trimester, three studies during second trimester, eight during third trimester, and two studies at each trimester. 11 studies with matched BMI between cases and controls, whereas BMI did not match for the rest 6 studies. Almost all included studies measured circulating NGAL levels using ELISA methods and serum sample, aside from one study conducted by Sweeting AN et al. [28] which used DELFIA method and one study by Lou Y et al. [22] which used the plasma sample. The diagnose of GDM was based on various criteria: American College of Obstetricians and Gynecologists (ACOG) (n = 2), American Diabetes Association (ADA) (n = 6), The Australasian Diabetes in Pregnancy Society (ADIPS) (n = 1), Carpenter and Couston (C&C) (n = 1), The International Association of Diabetes and Pregnancy Study Groups (IADPSG) (n = 2), and WHO (n = 5). The mean/median levels of NGAL ranged from 4.8 to 105.9 ng/mL in patients with GDM, and from 3.66 to 87.2 ng/mL in healthy controls. The quality score assessed by the NOS of included studies ranged from 5 to 8 points.
First author | Year | Study location | Case group | Control group | Measurement trimester | BMI matched | GDM diagnostic | Quality score | ||||||
Sample size | Age, years | Gestation weeks | NGAL, ng/mL | Sample size | Age, years | Gestation weeks | NGAL, ng/mL | criterion | ||||||
Yin X [30] | 2020 | China | 49 | 32.47 |
39.14 |
4.80 |
39 | 31.77 |
39.00 |
4.80 |
3 | 0 | IADPSG | 7 |
Sweeting | 2019 | Australia | 248 | 33 (30–36) | 11–13 | Caucasians: 105.9 (73.9–141.7) | 732 | 32 (29–35) | 11–13 | Caucasians: 87.2 (65.9–111.6) | 1 | 1 | ADIPS | 9 |
AN [28] | East Asians: 70.9 (57.9–95.8) | East Asians: 73.7 (55.1–94.9) | ||||||||||||
South Asians: 76.0 (51.9–91.5) | South Asians: 64.8 (47.2–109.0) | |||||||||||||
He XJ [26] | 2018 | China | 37 | 31.6 |
38.9 |
49.811 |
34 | 29.3 |
38.8 |
14.438 |
2 | 0 | ADA | 6 |
Kang YS [27] | 2018 | China | 107 | 28.8 |
9.3 |
13.72 |
110 | 29.3 |
9.4 |
8.045 |
1 | 0 | WHO | 6 |
Lu SL [25] | 2017 | China | 42 | 28.2 |
38.5 |
1st: 21.8 |
42 | 28.4 |
38.9 |
1st: 10.2 |
1–3 | 0 | ADA | 6 |
2nd: 49.6 |
2nd: 14.6 |
|||||||||||||
3rd: 50.1 |
3rd: 14.7 |
|||||||||||||
Ma QP [24] | 2015 | China | 97 | 29.9 |
38.5 |
1st: 25.32 |
100 | 28.95 |
39.06 |
1st: 10.32 |
1–3 | 1 | IADPSG | 7 |
2nd: 56.84 |
2nd: 17.64 |
|||||||||||||
3rd: 49.87 |
3rd: 14.32 |
|||||||||||||
Liu YH [23] | 2015 | China | 30 | 32.5 |
35.2 |
32.61 |
30 | 29.3 |
36.5 |
14.37 |
3 | 0 | ACOG | 5 |
Hu J [20] | 2014 | China | 55 | 33.02 |
34.98 |
42.56 |
55 | 28.12 |
34.92 |
17.63 |
3 | 0 | ACOG | 6 |
Guo J [37] | 2014 | China | 28 | 28.5 |
37.3 |
57.5 |
21 | 27.9 |
37.5 |
21.7 |
3 | 1 | WHO | 6 |
Fu XM [19] | 2014 | China | 30 | 29.4 |
35.7 |
32.5 |
30 | 28.5 |
36.3 |
14.5 |
3 | 1 | ADA | 6 |
Lou Y [22] | 2014 | China | 96 | 27.88 |
NA | 49.47 |
164 | 28.38 |
271.4 |
15.95 |
3 | 1 | ADA | 7 |
Wang F [18] | 2013 | China | 26 | NA | 20–32 | 52.94 |
66 | NA | 20–32 | 15.44 |
2 | 0 | ADA | 5 |
Ren GH [17] | 2012 | China | 35 | 29.4 |
38.2 |
61.9 |
32 | 28.7 |
39.1 |
24.0 |
3 | 1 | WHO | 5 |
Duan DM [16] | 2012 | China | 77 | 29.5 |
35.7 |
43.99 |
77 | 28.0 |
36.1 |
17.80 |
3 | 1 | ADA | 6 |
Jiang J [15] | 2011 | China | 42 | 29.07 |
14–18 | 45.83 |
42 | 28.78 |
14–18 | 19.72 |
2 | 1 | WHO | 7 |
Chen Q [38] | 2011 | China | 40 | 31.7 |
24–28 | 42.2 (32.4–55.5) | 80 | 31.28 |
24–28 | 28.7 (22.8–35.4) | 2 | 1 | WHO | 6 |
D’Anna R [13] | 2009 | Italy | 41 | 27.2 |
9–12 | 51.3 (39.8–66.1) | 82 | 28.4 |
9–12 | 17.8 (15.5–20.9) | 1 | 1 | C&C | 7 |
Abbreviations: GDM, gestational diabetes mellitus; BMI, body mass index; ADIPS, The Australasian Diabetes in Pregnancy Society; IADPSG, The International Association of Diabetes and Pregnancy Study Groups; ADA, American Diabetes Association; C&C, Carpenter and Couston; ACOG, American College of Obstetricians and Gynecologists; WHO, World Health Organization; NGAL, neutrophil gelatinase-associated lipocalin; NA, not available.
|
The overall estimates were calculated in a random-effect model due to the
tremendous heterogeneity between studies (I
Forest plots of association between circulating neutrophil gelatinase-associated lipocalin and gestational diabetes mellitus by trimesters.
Subgroups | Studies | GDM | Control | SMD | 95% CI | p for Z |
I |
p for I | |
Total 1 |
19 | 1078 | 1722 | 3.16 | 2.28–4.04 | 98.4 | |||
Total 2 |
19 | 1078 | 1722 | 3.43 | 2.51–4.36 | 98.6 | |||
Total 3 |
19 | 1078 | 1722 | 3.36 | 2.45–4.28 | 98.5 | |||
Gestational weeks | |||||||||
First trimester | 6 | 491 | 1010 | 2.34 | 0.79–3.89 | 0.003 | 99.1 | ||
Second trimester | 3 | 108 | 188 | 3.58 | 0.75–6.40 | 0.013 | 98.1 | ||
Third trimester | 10 | 479 | 524 | 3.54 | 2.55–4.53 | 96.0 | |||
Countries | |||||||||
Asians | 17 | 929 | 1110 | 3.34 | 2.34–4.35 | 98.5 | |||
Caucasians | 2 | 149 | 612 | 1.68 | –0.65–3.99 | 0.157 | 98.3 | ||
Maternal age, years | |||||||||
12 | 651 | 796 | 4.23 | 3.50–4.96 | 93.3 | ||||
7 | 427 | 926 | 1.30 | 0.55–2.04 | 0.001 | 96.5 | |||
BMI matched | |||||||||
Yes | 13 | 781 | 1385 | 2.63 | 1.64–3.62 | 98.5 | |||
No | 6 | 297 | 337 | 4.29 | 3.48–5.10 | 87.2 | |||
Abbreviations: GDM, gestational diabetes mellitus; NGAL, neutrophil gelatinase-associated lipocalin; BMI, body Mass Index; SMD, standardized mean difference. |
In the stratified analyses, the significant findings were observed among all
strata except for studies conducted among Caucasians (SMD: 1.68; 95% CI: –0.68,
3.99; p = 0.157). According to the overlaps of the 95%
CIs among strata, we observed that higher difference among those with lower
maternal age (
According to the result of influence analysis, no significant effect of any individual study has been found to been exerted on the association between circulating NGAL levels between GDM patients and controls, with the SMD ranged from 1.57 (1.46, 1.68) to 2.14 (2.02, 2.26) (Fig. 3).
Funnel plot of association between circulating neutrophil gelatinase-associated lipocalin and gestational diabetes mellitus.
The possibility of publication bias was detected by the Egger’s test (p
Influence analysis for association between circulating neutrophil gelatinase-associated lipocalin and gestational diabetes mellitus.
In this meta-analysis, we firstly pool the current evidence on the role of
circulating NGAL in detection of GDM. Result from comparation between 1080 GDM
cases and 1736 controls, our finding shows that the circulating NGAL levels in
GDM patients are higher than that of healthy pregnant women, although significant
heterogeneity of results was detected. Subgroup analyses further indicated that
these differences were more significant in women with younger maternal age (
Over the last decades, the need and interest in the identification of molecules
for the detection of disease onset and progression has wildly emerged. Since
first identified in 1994, NGAL has become increasingly relevant as a biomarker in
several diseases (e.g., acute kidney injury, Alzheimer’s disease, multiple
sclerosis, cardiovascular diseases, depression, etc.) [12, 39]. By pooling the
current evidence, our results were in chord with findings from previous studies
mainly conducted in Chinese population [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30], showing that circulating NGAL
levels were higher in GDM cases compared to healthy controls. For instance, in a
cohort study involving 41 women with a singleton pregnancy, who developed GDM in
the past 12 months, and healthy group of 82 normal pregnancies, the levels of
circulating NGAL were significantly higher in those with GDM than that of control
group [51.3 ng/mL (39.8–66.1) vs. 17.8 ng/mL (15.5–20.9); p
Although the broad functions of NGAL have been described, including innate immune response (e.g., chelator of bacterial siderophores and anti-plasmodial regulator) [40], iron homeostasis (e.g., inductor of cell proliferation/cell differentiation and inductor of cell death ) [41], modulation of the inflammatory response (e.g., anti-inflammatory modulator and pro-inflammatory modulator) [11], the exact mechanism by which NGAL has potential role in the occurrence and development of GDM were still unclear. Agent-mediated decreases in NGAL concentrations significantly correlated with increases in insulin sensitivity [12]. Moreover, NGAL was found to promote insulin resistance in cultured adipocytes and hepatocytes [42]. According to the current evidence, NGAL has not only been proposed to be an iron delivery protein [41], but also been probably involved in the regulation of insulin sensitivity through the influence on iron homeostasis [39, 42]. During an oral glucose tolerance test, the GDM Serum ferritin levels might be independent predictors of 2-h glucose [43], and iron intake in healthy women has been found to be positively associated with higher risk of type 2 diabetes [44]. Consistent with this surmise that iron is required for the effect of NGAL on insulin action, the iron-free NGAL might be ineffective in causing insulin resistance in cultured hepatocytes [42]. Consequently, NGAL may play a role in the development of GDM through insulin resistance; still, whether the increase circulating NGAL levels contributes to GDM progression is not yet known.
Subgroup analyses indicated that the variation of circulating NGAL levels
between GDM cases and healthy pregnant women increased when studies were
performed in women with maternal age
At present, GDM presents a particular public health challenge given its rapidly increasing prevalence in the context of the global obesity epidemic [51]. Screening and diagnosing GDM in early pregnancy are beneficial to reduce the burden of disease caused by GDM. Our study suggested that NGAL may be helpful for early identification of GDM, but this potential role needed more evidence to prove, further prospective studies are needed.
This meta-analysis had several limitations. Firstly, between-study heterogeneity
was significantly assessed for overall estimate (I
Our findings imply that the expression levels of NGAL in patients with GDM were higher than in healthy pregnant women, indicating that NGAL might be a useful detecting index for the judgment of the occurrence of GDM. This suggests that NGAL should be valued at the GDM screening in order to improve the maternal and fetus health. However, more evidence from larger prospective studies are needed to demonstrate the exact role of circulating NGAL during the early pregnancy.
Study concept and design—LX and HH. Data extraction—ZC, and SYW. Data analysis—JCH and SYW. Manuscript drafting—ZC and HH. HH, JCH and LX contributed to the article revise. All authors gave final approval of the version to be published.
Not applicable.
We would like to express my gratitude to all those who helped me during the writing of this manuscript. Thanks to all the peer reviewers for their opinions and suggestions.
This research was supported by Ningbo Clinical Research Center for Digestive System Tumors (GrantNo.2019A21003).
The authors declare no conflict of interest.