Blood Lead Level and Renal Impairment among Adults: A Meta-Analysis

Background: The adult population in lead-related occupations or environmentally exposed to lead may be at risk for renal impairment and lead nephropathy. This meta-analysis aims to determine the impact of blood lead level (BLL) on renal function among middle-aged participants. Methods: Cross-sectional, longitudinal, or cohort studies that reported BLL and renal function tests among adult participants were retrieved from PubMed, Scopus, and ISI Web of Science. Relevant studies were included and assessed for quality using the Newcastle–Ottawa Scale (NOS). The pooled mean BLL of participants with a high BLL (≥30 µg/dL), moderate BLL (20–30 µg/dL), and low BLL (<20 µg/dL) was estimated using the random effects model. The pooled mean differences in BLL, blood urea nitrogen (BUN), creatinine, uric acid, and creatinine clearance between the exposed and non-exposed participants were estimated using the random effects model. Meta-regression was performed to demonstrate the association between the effect size (ES) of the pooled mean BLL and renal function. Heterogeneity among the included studies was assessed using the Cochrane Q and I2 statistics. Cochrane Q with a p value less than 0.05 and I2 more than 50% demonstrated substantial heterogeneity among the studies included. Publication bias was assessed using the funnel plot between the effect size and standard error of the effect size. Results: Out of 1657 articles, 43 were included in the meta-analysis. The meta-analysis demonstrated that the pooled mean BLL in the participants with a high BLL, moderate BLL, and low BLL was 42.41 µg/dL (95% confidence interval (CI): 42.14–42.67, I2: 99.1%), 22.18 µg/dL (95% CI: 21.68–22.68, I2: 60.4%), and 2.9 µg/dL (95% CI: 2.9–2.9, I2: 100%), respectively. The mean BLL of the exposed participants was higher than that of the non-exposed participants (weighted mean difference (WMD): 25.5, p < 0.0001, 95% CI: 18.59–32.45, I2: 99.8%, 17 studies). The mean BUN (WMD: 1.66, p < 0.0001, 95% CI: 0.76–2.55, I2: 76%, 10 studies) and mean creatinine (WMD: 0.05, p = 0.007, 95% CI: 0.01–0.08, I2: 76.8%, 15 studies) in the exposed participants were higher than those in the non-exposed participants. The mean creatinine clearance in the exposed participants was lower than that in the non-exposed participants (standard mean difference (SMD): −0.544, p = 0.03, 95% CI: −1.035–(−0.054), I2: 96.2%). The meta-regression demonstrated a significant positive effect of BLL on BUN (p = 0.022, coefficient: 0.75, constant: −3.7, 10 studies). Conclusions: BLL was observed to be associated with abnormal renal function test parameters, including high BUN, high creatinine, and low creatinine clearance. Moreover, BUN seemed to be the most valuable prognostic marker for lead-induced renal impairment. Therefore, regular checks for renal function among lead-exposed workers should be a priority and publicly promoted.


Background
Lead is a heavy metal and toxicant to the human body [1]. The most common sources of lead in lead-related occupations come from batteries, radiator manufacturing, lead refineries, paints, and ceramics [2]. In addition, lead is distributed in the environment as contaminated dust, in drinking water, and in soil where humans can be exposed through (<20 µg/dL). Cutoff values of <20 µg/dL, 20-30 µg/dL, and >30 µg/dL were used to indicate lead exposure among participants, as previously described by Lim et al. [25].

Statistical Analysis
The mean BLL reported in the included studies was used to analyze the pooled mean BLL among the exposed participants. However, as the non-exposed participants were a low-risk group and showed a low BLL, the mean BLL of non-exposed or controls was not estimated in this study. The median with rank or interquartile rank of BLL and renal parameters reported in the included studies was transformed into the mean and standard deviation (SD), as reported elsewhere [26]. The unit of BLL and renal parameters, including BUN, creatinine, and uric acid, used for analyses were µg/dL and mg/dL, respectively; hence, any studies that reported a different unit then had data converted to µg/dL of BLL and mg/dL of renal parameters using the calculator available online [27]. The pooled mean BLL and 95% confidence interval of the included studies were estimated using the random-effects model. The pooled mean differences in BLL, BUN, creatinine, and uric acid between the exposed and non-exposed participants were estimated using the random-effects model and presented as weighted mean differences (WMDs) with 95% CIs. WMD is the difference in means between the mean value in exposed and non-exposed participants. As the mean creatinine clearance was reported in the included studies in different units, the standard mean difference (SMD) was used to estimate the difference in mean creatinine clearance between exposed and non-exposed participants. Meta-regression was performed to demonstrate the association between the effect size (ES) or WMD of BLL and renal function test parameters, including BUN, creatinine, BUN/creatinine ratio, creatinine clearance, and uric acid. Heterogeneity among the included studies was assessed using the Cochrane Q and I 2 statistics. Cochrane Q with a p value less than 0.05 and I 2 more than 50% demonstrated substantial heterogeneity among the included studies [28]. Of the heterogeneity that existed, the random-effects model was used for estimating the pooled variables, and if heterogeneity did not exist, the fixed-effects model was used for estimating the pooled variables. Subgroup analysis of BLL was performed to demonstrate any differences among the groups of exposed participants. Publication bias among the included studies was assessed by visualizing the funnel plot asymmetry. If the funnel plot demonstrated an asymmetrical distribution, Egger's test was used to confirm whether the asymmetrical distribution of the funnel plot was caused by the small-study effects. All analyses were performed using Stata Version 14.2 (StataCorp, College Station, TX, USA).

Search Results
Overall, 1657 articles were retrieved from the searches of three databases. After removing 676 duplicate articles, 981 articles were screened for potentially relevant articles through title and abstract screening. As a result, 754 articles were excluded due to their having no relevance to the present study. The full texts of the 227 articles that remained were examined according to the eligibility criteria, and 184 articles were excluded ( Figure 1). Finally, 43 articles [3,4,7,11,13,[15][16][17]19,20,25, met the study criteria and were included in the study.

Quality of the Included Studies
The quality of the included studies is shown in Table S2. Seventeen studies were highquality studies, as BLL was reported in both the exposed and non-exposed participants. However, the rest of the included studies were low-quality studies, as they did not enroll a control group. Low-quality studies were included in the present study to analyze the pooled mean BLL.
Low mean BLL (<20 µg/dL) Sources of contamination: polluted areas, heavy metal pollution, battery manufacturing and lead recycling plants, auto repair, smelting factory

Pooled Mean Difference in BLL between Exposed and Control Participants
The pooled mean difference in BLL between the exposed and non-exposed participants was estimated using the mean BLL from 17 studies [7,[15][16][17]20,30,[32][33][34]39,41,42,47,49,51,56,59] (Figure 3). Overall, the mean BLL of the exposed group was higher than that of the non-exposed participants (weighted mean difference: 25.5, p < 0.0001, 95% CI: 18.59-32.45, I 2 : 99.8%). Subgroup analysis demonstrated that the difference in BLL between exposed and non-exposed participants was larger for those with high mean BLL (weighted mean difference: 32.28, p < 0.0001, 95% CI: 28.91-35.65, I 2 : 96.4%), whereas the difference between exposed and non-exposed participants was smallest for those with a low mean BLL (weighted mean difference: 4.73, p < 0.0001, 95% CI: 2.69-6.76, I 2 : 93.6%). The mean difference in BLL between exposed and non-exposed participants. WMD: Weighted Mean Difference (µg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (µg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between the two groups. The mean difference in BLL between exposed and non-exposed participants. WMD: Weighted Mean Difference (µg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (µg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between the two groups.

BLL and Gender
The pooled mean difference in BLL between the exposed men and women was estimated using five studies [11,19,[52][53][54]. The results demonstrated that the mean BLL in the exposed males was higher than that in the female participants (weighted mean difference: 2.45, p < 0.0001, 95% CI: 1.11-3.80, I 2 : 95.8%) (Figure 4). Three studies [19,53,54] demonstrated a higher mean BLL in male participants than in female participants.

BLL and Gender
The pooled mean difference in BLL between the exposed men and women was estimated using five studies [11,19,[52][53][54]. The results demonstrated that the mean BLL in the exposed males was higher than that in the female participants (weighted mean difference: 2.45, p < 0.0001, 95% CI: 1.11-3.80, I 2 : 95.8%) (Figure 4). Three studies [19,53,54] demonstrated a higher mean BLL in male participants than in female participants. The mean difference in BLL between male and female participants. The mean BLL in the exposed males was higher than that in the female participants (weighted mean difference: 2.45, p < 0.0001, 95% CI: 1.11-3.80, I 2 : 95.8%) (white diamond symbol). WMD: Weighted Mean Difference (µg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (µg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between the two groups.

Renal Function Tests and BLL
Meta-regression analyses were performed to determine the association between the effect size (weighted mean difference, WMD) of renal function test parameters (dependent variable) and mean BLL (independent variable). The meta-regression of BUN (weighted mean difference) and mean BLL was performed using the data from 10 studies [7,15,17,20,32,39,41,47,49,59] because these studies reported the mean BLL and mean BUN. The results demonstrated a significant positive effect of BLL on BUN (weighted mean difference) (p = 0.022, coefficient: 0.75, constant: −3.7) (Figure 7). The meta-regression of creatinine (weighted mean difference) and mean BLL was performed using the data from 15 studies [7,[15][16][17]20,30,32,33,35,39,41,47,49,51,59]. The results demonstrated a nonsignificant effect of mean BLL on creatinine level (weighted mean difference) (p = 0.989) (Figure 8). The meta-regression of mean BLL and the BUN/creatinine ratio (weighted mean difference) was performed using the data from 10 studies [7,15,17,20,32,39,41,47,49,59]. The results demonstrated a non-significant effect of mean BLL on the BUN/creatinine ratio (weighted mean difference) (p = 0.889, coefficient: 0.12, constant: 0.034) (Figure 9). No significant effect of mean BLL on creatinine clearance or uric acid was found ( Supplementary  Figures S3 and S4). The mean difference in blood urea nitrogen (BUN) levels between exposed and nonexposed participants. WMD: Weighted Mean Difference (mg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (mg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between the two groups. Figure 6. The mean difference in creatine levels between exposed and non-exposed participants. WMD: Weighted Mean Difference (mg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (mg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between two groups. Figure 5. The mean difference in blood urea nitrogen (BUN) levels between exposed and non-exposed participants. WMD: Weighted Mean Difference (mg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (mg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between the two groups. The mean difference in blood urea nitrogen (BUN) levels between exposed and nonexposed participants. WMD: Weighted Mean Difference (mg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (mg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between the two groups. Figure 6. The mean difference in creatine levels between exposed and non-exposed participants. WMD: Weighted Mean Difference (mg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (mg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between two groups. Figure 6. The mean difference in creatine levels between exposed and non-exposed participants. WMD: Weighted Mean Difference (mg/dL), % Weighted: the impact proportion of each study to the pooled effect, CI: Confidence Interval (mg/dL), Black diamond symbol: point estimate for each study, White diamond symbol: pooled WMD in each subgroup or all groups, Solid line in the middle of the graph at 0: no difference in WMD between the two groups, Dashed line: pooled WMD between two groups.

Publication Bias
The funnel plot between the effect size (weighted mean difference) and standard error of the effect size demonstrated the likelihood of asymmetry ( Figure 10). Therefore, Egger's test was performed to confirm the funnel plot asymmetry. The results showed that no small-study effects among the included studies were found (p < 0.728), indicating no publication bias across the included studies.

Publication Bias
The funnel plot between the effect size (weighted mean difference) and s ror of the effect size demonstrated the likelihood of asymmetry ( Figure 10) Egger's test was performed to confirm the funnel plot asymmetry. The resu that no small-study effects among the included studies were found (p < 0.728) no publication bias across the included studies.

Discussion
The present meta-analysis demonstrated that the mean BLL among partic

Publication Bias
The funnel plot between the effect size (weighted mean difference) an ror of the effect size demonstrated the likelihood of asymmetry (Figure Egger's test was performed to confirm the funnel plot asymmetry. The r that no small-study effects among the included studies were found (p < 0.7 no publication bias across the included studies.

Discussion
The present meta-analysis demonstrated that the mean BLL among pa high BLL was 42.41 µg/dL, moderate BLL was 22.18 µg/dL, and low BLL w The mean BLL was significantly higher in lead-exposed participants than i participants for all 18 included studies. This finding was similar to a met formed in Iran, which demonstrated high mean BLL in Iranian lead-exposed Moreover, the high difference in BLL seemed to be observed clearly amon with high BLL compared to those with moderate or low BLL. Various stud

Discussion
The present meta-analysis demonstrated that the mean BLL among participants with high BLL was 42.41 µg/dL, moderate BLL was 22.18 µg/dL, and low BLL was 2.90 µg/dL. The mean BLL was significantly higher in lead-exposed participants than in non-exposed participants for all 18 included studies. This finding was similar to a meta-analysis performed in Iran, which demonstrated high mean BLL in Iranian leadexposed workers [21]. Moreover, the high difference in BLL seemed to be observed clearly among participants with high BLL compared to those with moderate or low BLL. Various studies supported this difference in the mean BLL among the two groups of participants. In Brazil, de Pinto Almeida et al. demonstrated that the mean BLL was 64.1 ± 16.3 µg/dL and 25.5 ± 4.4 µg/dL in primary lead smelting workers and in non-exposed participants, respectively [16]. In Germany, it was reported that the mean BLL was 40.6 (20.2-70.6) µg/dL in workers who were exposed to lead dust in an accumulator plant, whereas the mean BLL was 6.8 (4.8-10.6) µg/dL in the control group [12]. In South Africa, Ehrlich et al. reported that the mean BLL in battery factory workers was 53.5 ± 12.7 µg/dL [13]. In India, Patil et al. showed that the mean BLL in battery manufacturing workers, silver jewelry workers, spray painters, and controls was 53.63 ± 16.98, 48.56±7.39, 22.32 ± 8.87, and 12.52 ± 4.08 µg/dL, respectively [49]. In Nigeria, Alasia et al. showed that the mean BLL was 50.37 ± 24.58 µg/dL in lead-exposed workers and 41.40 ± 26.85 µg/dL in non-exposed participants [61]. Onuegbu et al. performed a study on automobile mechanics, battery repair workers, and petrol station attendants and demonstrated that the mean BLL was 69.7 ± 13.2 µg/dL in lead-exposed group and 18.5 ± 3.6 µg/dL in non-exposed participants [17]. Recently, a study in India also showed that the mean BLL was 30.5 ± 12.2 µg/dL in spray paint workers and 5.46 ± 2.58 µg/dL in the control group [7]. In South Korea, Jung et al. performed a study among workers who worked in secondary lead smelter, plastic stabilizer, and radiator manufacturing industries and showed that the mean BLL in highly exposed, moderately exposed, lowly exposed, and non-exposed participants was 74.6 ± 7.8, 46.5 ± 5.9, 24.3 ± 2.7, and 7.9 ± 1.4 µg/dL, respectively [39]. In Taiwan, a study by Hsiao et al. among lead battery factory workers showed that the mean BLL was 15.8 µg/dL and 11.6 µg/dL in males and females, respectively [62].
The pooled mean difference in BLL between the exposed males and females showed that the mean BLL in the exposed males was higher than that in the female participants. A significant difference in gender was clearly observed in a study by Staessen et al. [54] and Wang et al. [19]. However, a study by Lai et al. [11] demonstrated that BLL in exposed males was lower than that in female participants. The heterogeneity of the results between the studies might have been because males were more likely to be exposed to lead than females. Another possible explanation is that estrogen is higher in females than males; therefore, estrogen may increase lead distribution to the bone and slow the release of lead from the bone in women as well [63,64].
The present meta-analysis demonstrated the difference in the mean BUN, serum creatinine, and mean creatinine clearance in lead-exposed participants compared to non-exposed participants. In addition, the present meta-analysis showed that the mean BUN was significantly higher in lead-exposed participants than in non-exposed participants, especially in participants with high and moderate mean BLL. These results demonstrated that an increase in BLL could induce renal impairment among exposed participants. The difference in BUN was clearly observed in five included studies [15,17,39,49,59]. Nevertheless, some included studies demonstrated no difference in BUN between the two groups of participants [7,20,32] and caused heterogeneity among the studies included in the meta-analysis. A high mean BUN was also reported in lead battery workers and spray painters in India [7] and Taiwan [19], in a secondary lead refinery worker in South Korea [39] and Japan [65], and in lead workers in Nigeria [17] and India [49]. Moreover, Wang et al. demonstrated that every increment of 10 µg/dL BLL produced an increase of 0.62 mg/dL in BUN levels [19]. The increase in BUN might be caused by the reduction in renal plasma flow and the decrease in the glomerular filtration rate (GFR), leading to high accumulations of urea nitrogen in the plasma [66]. The meta-regression analysis between BLL and BUN demonstrated that the mean BLL was an independent factor affecting BUN levels. This result suggested that BUN is a sensitive marker of lead-induced renal impairment. In addition to the lead that affected the BUN levels, there were other factors, such as age, work duration, gender, and smoking habit [19].
The present meta-analysis showed that the mean creatinine was significantly higher in lead-exposed participants than in non-exposed participants, especially in participants with a high mean BLL. The higher mean creatinine among the exposed participants with a high mean BLL was clearly demonstrated in four studies [7,15,16,59]. High mean levels of creatinine were observed in various studies, such as the study by de Pinto Almeida et al., which studied Brazilian lead workers [16]; the studies of Onuegbu et al. [17] and Alasia et al. [61], which examined Nigerian lead workers; and the study of Kshirsagar Mandakini et al., which studied spray painters in India [7]. Nevertheless, five studies [17,32,33,39,49] demonstrated no difference in the mean creatinine between the two groups of participants. A study by Roels et al. [20] showed a lower mean creatinine in exposed participants than in non-exposed participants. Despite the high mean level of creatinine in the leadexposed workers that was observed, the meta-regression showed no relationship between the mean BLL and creatinine level. Some previous studies reported similar findings to ours [32,39,49]. This might be because kidneys have millions of nephrons and have reserve capacity; therefore, the clinical manifestations of renal impairment would not be demonstrated until the nephrons were destroyed by more than 50% [49]. This indicated that serum creatinine was insufficiently sensitive for the early detection of renal impairment induced by lead. The non-association of BLL and creatinine might be due to factors related to creatinine balance, such as gender, age, weight, work duration, smoking habits, and alcohol consumption, which also affect serum creatinine [19].
Creatinine clearance has been widely used to determine GFR. It is commonly used in routine laboratory work for evaluating renal function. This study demonstrated that the mean creatinine clearance was significantly lower in lead-exposed workers than in non-exposed participants. This finding was observed in the studies by Alasia et al. [61], Gennart et al. [33], Weaver et al. [59], Chen et al. [30], and Reilly et al. [51]. Nevertheless, the meta-regression analyses did not show the relationship between the mean BLL and creatinine clearance. This finding is consistent with a study on lead-exposed workers in Japan [48], which indicated that a BLL less than 70 µg/dL might not affect the function of the glomeruli [54]. Furthermore, various confounding factors, such as ethnicity, age, gender, work duration, muscle mass, and protein intake, might influence creatinine clearance [19,44]. These confounding factors might, in part, affect the analysis of BLL and creatinine clearance.
Uric acid is the product of purine metabolism; moreover, it is derived from the degradation of a cell or nucleic acid within a cell, and elimination of uric acid occurs in the proximal tubule and distal tubule [67]. A previous study indicated that chronic lead exposure may interfere with the secretion of uric acid in the distal tubule, leading to hyperuricemia [68]. However, certain mechanisms of hyperuricemia induced by lead are still unclear. The present meta-analysis demonstrated no difference in the uric acid level between lead-exposed and non-exposed participants. Nevertheless, the exposed participants with a high BLL seemed to have a higher uric acid level, as demonstrated in four included studies [7,15,16,32]. In addition, a study by Kshirsagar et al. [41] demonstrated that exposed participants with a moderate BLL had a higher uric acid level than the control participants. Three studies demonstrated no difference in the serum uric acid level between the two groups of participants [20,39,47]. In addition, some previous studies contradicted our study [16,41,61]. These studies reported an increase in the uric acid level of lead-exposed workers who had a BLL greater than 60 µg/dL. Although the meta-analysis demonstrated a difference in the uric acid level between the two groups of participants, the meta-regression showed no association between BLL and mean uric acid. Therefore, the change in uric acid was insufficient as a sensitive marker to detect early renal impairment induced by lead exposure.
In addition to renal impairment induced by lead, lead exposure also increased the severity of underlying diseases, especially in susceptible populations with hypertension, diabetes mellitus, and chronic kidney disease [8]. Moreover, cadmium, mercury, and other heavy metals contaminating the environment and workplace may result in combined adverse effects on the human body. Therefore, protection from heavy metal exposure is crucial; for example, factory owners should provide occupational health educational programs to prevent workers from being poisoned by lead. In addition, lead exposure prevention should be implemented before, while and after the work is finished, for example, wearing personal protection devices, such as gloves, masks, and aprons, before starting to work, hand washing prior to eating, not smoking or eating in the workplace, and cleaning the body and mandatorily changing clothes before leaving the workplace to reduce the distribution of lead into the environment [65]. Although the removal and return of lead-exposed workers at 60 and 40 µg/dL, respectively, is used by the United States Occupational Safety and Health Administration (US OSHA) [69] and presumably by other countries, the results of the present study suggested that workers who have an excessive BLL of 30 µg/dL should be removed from their job and return to work when their BLL drops below 20 µg/dL.

Limitations
The present study had limitations. First, there were a limited number of included studies based on the eligibility criteria, which limited the study to adult or middle-aged participants. Second, the relationship between BLL and work duration was not assessed due to data unavailability among the included studies. Third, there are several factors that affect the progression of lead nephropathy in addition to lead, including individual susceptibility, race, and the pattern of lead exposure [39]. These might be the reasons for the heterogeneity among the included studies, where renal impairment was found to be related to lead exposure.

Conclusions
BLL was associated with abnormal renal function test parameters, including high BUN, high creatinine, and low creatinine clearance. Moreover, BUN seemed to be the most valuable prognostic marker for lead-induced renal impairment. Therefore, regular checks for renal function among lead-exposed workers should be important and publicly advocated for.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/ijerph18084174/s1. Supplementary Figure S1. The mean difference in uric acid levels between exposed and control participants, Supplementary Figure S2. The mean difference in creatinine clearance between exposed and control participants, Supplementary Figure S3. The meta-regression analysis of WMD and creatinine clearance, Supplementary Figure S4. The meta-regression analysis of WMD and uric acid, Supplementary Tables, Table S1. Search terms, Table S2. Quality of the included studies, PRISMA Checklist S1.
Author Contributions: S.K. and M.K. designed the study. M.K. performed the data collection and statistical analysis, in addition to drafting the methods and results sections of the manuscript. S.K. drafted the introduction and discussion sections of the manuscript. All authors have read and agreed to the published version of the manuscript.