Prevalence of Gestational Diabetes Mellitus in the Middle East and North Africa, 2000–2019: A Systematic Review, Meta-Analysis, and Meta-Regression

Women in the Middle East and North Africa (MENA) region are burdened with several risk factors related to gestational diabetes mellitus (GDM) including overweight and high parity. We systematically reviewed the literature and quantified the weighted prevalence of GDM in MENA at the regional, subregional, and national levels. Studies published from 2000 to 2019 reporting the prevalence of GDM in the MENA region were retrieved and were assessed for their eligibility. Overall and subgroup pooled prevalence of GDM was quantified by random-effects meta-analysis. Sources of heterogeneity were investigated by meta-regression. The risk of bias (RoB) was assessed by the National Heart, Lung, and Blood Institute’s tool. One hundred and two research articles with 279,202 tested pregnant women for GDM from 16 MENA countries were included. Most of the research reports sourced from Iran (36.3%) and Saudi Arabia (21.6%), with an overall low RoB. In the 16 countries, the pooled prevalence of GDM was 13.0% (95% confidence interval [CI], 11.5–14.6%, I2, 99.3%). Nationally, GDM was highest in Qatar (20.7%, 95% CI, 15.2–26.7% I2, 99.0%), whereas subregionally, GDM was highest in Gulf Cooperation Council (GCC) countries (14.7%, 95% CI, 13.0–16.5%, I2, 99.0%). The prevalence of GDM was high in pregnant women aged ≥30 years (21.9%, 95% CI, 18.5–25.5%, I2, 97.1%), in their third trimester (20.0%, 95% CI, 13.1–27.9%, I2, 98.8%), and who were obese (17.2%, 95% CI, 12.8–22.0%, I2, 93.8%). The prevalence of GDM was 10.6% (95% CI, 8.1–13.4%, I2, 98.9%) in studies conducted before 2009, whereas it was 14.0% (95% CI, 12.1–16.0%, I2, 99.3%) in studies conducted in or after 2010. Pregnant women in the MENA region are burdened with a substantial prevalence of GDM, particularly in GCC and North African countries. Findings have implications for maternal health in the MENA region and call for advocacy to unify GDM diagnostic criteria. Systematic Review Registration PROSPERO CRD42018100629


INTRODUCTION
Gestational diabetes mellitus (GDM) (1) is usually diagnosed during the second and third trimesters of pregnancy (2). Risk factors of GDM include excessive body weight, low level of physical activity, consanguineous marriage, previous history of GDM, glycated hemoglobin >5.7%, and history of cardiovascular disease (3). As the toll of overweight and obese reproductive-age females soars, the risk of developing hyperglycemia in pregnancy increases (4).
GDM has a global public health burden (5) with both shortand long-term consequences on health. The short-term ramifications of GDM include adverse perinatal outcomes for the affected women (e.g., preeclampsia, polyhydramnios, and increased cesarean section ["C-section"] risk) and their neonates (e.g., macrosomia and shoulder dystocia) (1,6), whereas the long-term complications of GDM incorporate the risk of type 2 diabetes mellitus (T2DM) for the mother and the risk of childhood obesity, impaired glucose tolerance, and/or metabolic syndrome for their neonates (6). Since increased blood glucose levels are associated with certain perinatal complications, gestational blood glucose control is vital (7).
Understanding population-specific healthcare needs at specific points of time is essential, and prevalence estimates are ideal for such purposes (8). Unfortunately, the global GDM prevalence estimates (<1%-28%) show a wide variation due to ethnicity, ethnic variation among various populations, and inconsistent use of screening and diagnostic criteria (4,9). To precisely estimate the burden of GDM of a particular geographic area, it is essential to determine the region-specific prevalence estimate. There is scant literature on the prevalence of GDM in the Middle East and North Africa (MENA) region, although two of the main risk factors [physical inactivity and above-normal body mass index (BMI)] are identified as being highly prevalent in this region (10). Moreover, three of the world's top ten most prevalent countries for diabetes mellitus belong to this region: Saudi Arabia (24%), Kuwait (23%), and Qatar (23%) (11). For the entire Eastern Mediterranean region, the existing prevalence estimate of GDM is 14.5%, although this includes only cases diagnosed according to the World Health Organization (WHO) 1999 criteria (4). One previous survey showed that physicians and hospitals in this region use different criteria to diagnose GDM (12).
A systematic review and meta-analysis of prevalence studies is considered to be an ideal method to understand the burden of GDM at regional and national levels. In this systematic review, meta-analysis, and meta-regression, we estimated the weighted pooled prevalence of GDM in the MENA region, at the regional, subregional, and national levels, based on literature published between January 2000 and December 2019.

This review follows the Preferred Reporting Items for Systematic
Review and Meta-Analysis (PRISMA) 2009 guidelines (13).
The PRISMA checklist is provided elsewhere (Supplementary File 1). Following our published protocol, we report here "systematic review 2" (14). We implemented minor amendments whenever needed, including an updated database search.

Data Source and Searches
To identify eligible studies reporting the prevalence of GDM in the MENA countries, we conducted a comprehensive search of five electronic databases (MEDLINE, EMBASE, Web of Science, SCOPUS, and Cochrane library) from January 1, 2000, to December 31, 2019, using variant Medical Subject Headings and free-text terms. Restricting the literature search to 2000 was to estimate changes in the GDM prevalence over the past two decades (before and after 2010), at national, sub-regional, and regional levels, whenever enough data is available for the meta-analysis. The literature search strategy was developed in consultation with an expert librarian at the National Medical Library at the United Arab Emirates University. The full search strategy available in the published protocol (14). Retrieved references were imported to the Covidence software (Covidence, Melbourne, Australia) (15). Deduplication of similar references was performed automatically by the Covidence software.

Study Selection
To identify and select studies for inclusion, we followed the PECO(T) framework: participants, exposure, comparator, outcome(s), and type of study (16). However, we considered only participants and outcomes because the focus of this review was on studies reporting the prevalence of GDM. Study eligibility criteria are presented in Table 1.

Identifying Eligible Studies
Titles and abstracts were screened by RHA, NMA, and MSP to detect eligible research reports on the prevalence of GDM. For studies that appeared eligible, the full text was reviewed (RHA, NMA, and MSP). Screening of all titles and abstracts and full text articles was performed independently by two reviewers. Disagreements among reviewers were resolved by discourse. We also searched the reference lists of eligible studies for studies that might have been missed. Figure 1 shows the PRISMA flowchart of study selection.
In this review, the term "research report" is used to refer to a full published research document. The term "study" is used to refer to a single study on a specific population group. One big observational study (one research report) provides GDM data stratified into four age groups (four studies). Hence, one research report could contribute several studies on GDM prevalence.

Data Extraction and Quality Assessment
Relevant data from eligible studies were extracted into a predesigned Excel sheet using a predefined list of numerical and string variables. The outcome of interest was the weighted prevalence of GDM in pregnant women in the MENA countries, according to various characteristics including, but not limited to, age, BMI, trimester, and time period. We extracted author names, publication year, country, city, and study setting.
In addition, data on the implemented methodology (design, data collection period, sampling strategy, and GDM diagnosis and ascertainment methodology) and characteristics of the studied pregnant women (age, pregnancy trimester, sample size, number of women with GDM and GDM prevalence) were extracted whenever available.
In addition to the overall prevalence of GDM, some research reports also reported the prevalence of GDM stratified according to different characteristics, such as age, parity, comorbidity, pregnancy trimester, and BMI. In such reports, data extraction was performed for the stratified GDM prevalence, following the rule that the study had to have at least ten tested subjects per strata; otherwise, information on the entire tested sample was extracted. A predefined sequential order was established when extracting stratified GDM prevalence estimates as follows: GDM stratified first according to comorbidities followed by parity, age, and BMI. This prioritization was used to identify the strata with more information on the tested pregnant women. When there was no stratification for the prevalence of GDM, we extracted the overall GDM prevalence measured.
For each research report reporting the stratified prevalence of GDM according to more than one category (i.e., age and BMI), one category per research report was considered and included based on the aforementioned prioritization scheme, to avoid double counting. In studies in which GDM was ascertained using different guidelines, the most sensitive and reliable ascertainment assay was considered (i.e., prioritizing fasting blood glucose over self-reported) or was based on the most recent and updated criteria (i.e., prioritizing WHO 2010 over 2006 criteria).
The risk of bias (RoB) assessment was performed at the level of the research report rather than the study. The quality of each research report was evaluated according to criteria of the National Heart, Lung, and Blood Institute (18). Six of 14 items from the quality assessment tool for prevalence studies were used (18). The six quality-related items assessed the research question/ objectives, studied population, sample size justification, and outcome measures and assessment. Eight items were not used because they are applicable only to follow-up cohort studies. For additional quality assessment, we also assessed the robustness of the implemented methodology using three additional quality-ofevidence criteria: sampling methodology, GDM ascertainment methodology, and precision of the estimate. Studies were considered to have "high" precision if at least 100 women were tested for GDM. We computed the overall proportion of research reports with potentially low RoB across each of these nine quality criteria and also computed the proportion (out of nine) of quality items with a potentially low RoB for each of the included research reports.
Data abstraction and quality assessment were performed independently by two reviewers (NA and MP) and crosschecked for disagreements. Any discrepancies in the extraction phase or in the quality assessment between the reviewers were discussed and resolved with a consultation of a senior reviewer (RA-R).

Data Synthesis and Analysis
To estimate the weighted pooled prevalence of GDM and the corresponding 95% confidence interval (CI), we performed meta-analyses of the extracted data. The Freeman-Tukey double arcsine transformation method was applied to stabilize the variances of the prevalence measures (19). The inverse variance method was used to weight the estimated pooled prevalence measures (20). Dersimonian-Laird random-effects model was used to estimate the overall pooled GDM  (22). Along with the pooled estimates, ranges and median were also reported to describe the dispersion of the GDM prevalence measures reported in the literature. The prediction interval, which estimates the 95% interval in which the true effect size in a new prevalence study will lie, was also quantified and reported (22). For the subgroup meta-analysis, country-level pooled estimates were generated overall and based on time period. In addition, to estimate the change in GDM both at the country level and overall, the data collection period was stratified into two time periods: 2000-2009 and 2010-2019. For studies in which the data collection period overlapped, the collection period was defined as "overlap" so as not to miss any important data when estimating country-level, subregional, and regional prevalence. The median (~2 years) was used in studies with an unclear data collection period. In these studies, the median was subtracted from the year of publication to estimate the year of data collection.
The weighted pooled prevalence, regardless of country, was also estimated according to the age of the pregnant women, trimester, BMI, study period, GDM ascertainment guidelines, and sample size (<100 or ≥100). The provision of pooled estimates regardless of the ascertainment guidelines was justified by the fact that the women were defined and treated as GDM patients following each specific ascertainment guideline.
Accumulated evidence has shown that GDM is associated with an increased risk of C-section (23,24) and maternal mortality (4). Independent of the research report and the characteristics of the tested pregnant women for GDM, we estimated the pooled GDM prevalence according to the Csection rate and maternal mortality ratio (MMR). Information on the C-section rate (25,26) and MMR were retrieved from various resources (27). Depending on data availability, information on C-section rate and MMR was extracted in the same or the closest year to the estimated GDM prevalence. For every GDM study, the rate of C-section was then categorized as <15%, 15-29%, >30%, or unclear, whereas the MMR was categorized as either ≤100/100,000 live births, >100/100,000 live births, or unclear.
To provide prevalence estimates at a subregional level, we regrouped MENA countries into four subregions, namely, North Africa, Gulf Cooperation Council (GCC) countries, Levant, and Iran/Iraq region. We estimated the overall pooled prevalence in these subregions and according to patient age, trimester, BMI, study period, GDM ascertainment guidelines, rate of C-section, and MMR.
Random-effects univariate and multivariable meta-regression models were implemented to identify sources of between-study heterogeneity and to quantify their contribution to variability in the prevalence of GDM. In univariate meta-regression models, analysis was performed by country, age, pregnancy trimester, BMI, and sample size. All variables with a p-value <0.1 in the univariate models were included in the multivariable model. In the final multivariable model, a p-value ≤0.05 was considered statistically significant, which contributed to the heterogeneity in prevalence estimates.

Publication Bias
A funnel plot was generated to explore the small-study effect on the pooled GDM prevalence estimates. The funnel plot was created by plotting each GDM prevalence measure against its standard error. The asymmetry of the funnel plot was tested using Egger's test (28).
The study is registered with PROSPERO, number CRD42018100629.

Database Search and Scope of the Review
Of the 13,139 citations retrieved from the 5 databases, 102 research reports were deemed eligible and included in this review ( Figure 1). The research reports were from 16 countries in the MENA region: Algeria (one), Bahrain (two), Egypt (four), Iraq (three), Iran (37), Jordan (four), Lebanon (two), Libya (one), Morocco (one), Oman (five), Qatar (six), Saudi Arabia (22), Sudan (two), Tunisia (one), United Arab Emirates (UAE) (eight), and Yemen (one). The prevalence data for both decades (time periods) were available from six countries (Bahrain, Iran, Oman, Qatar, Saudi Arabia, and the UAE); for the other countries, data were available for the time period 2010-2019 ( Table 2). Self-reported GDM status was documented in five research reports (31,73,83,90,119). The predominantly used GDM diagnostic criteria in the MENA region were from the American Diabetes Association and the International Association of Diabetes and Pregnancy Study Group (ADA/IADPSG; 48.5% of studies).
For the six countries reporting data on both decades, the overall GDM prevalence was estimated separately for each decade. There was a rise in the prevalence of GDM by 4% to 8% in Iran, Oman, and Saudi Arabia and a decrease of 2%

Subregional Specific Pooled GDM Prevalence
In Sudan, one of the North African countries with a C-section rate of 15-29%, a lower GDM prevalence (weighted prevalence of 7.9%) was observed compared with countries with a C-section rate of <15% (weighted prevalence of 23.0%). In North African countries with an MMR of >100/100,000 live births, the prevalence of GDM was 32.0% higher than in countries with an MMR of ≤100/100,000 live births (Supplementary File 2).
In GCC countries, the prevalence of GDM rose from 11.9% to 15.9% over the two successive decades. Overweight (12.5%) and obese (18.5%) pregnant women and pregnant women with a C-section rate of 15-29% (15.5%) were burdened with high GDM prevalence (Supplementary File 3). In these countries, pregnant women aged ≥30 years were burdened with higher GDM prevalence than the other subregions. As compared with the first decade, the weighted GDM prevalence in the subsequent decade increased by almost 4% in Iraq.
Tables 2-4 in the appendix provide additional weighted GDM prevalence estimates in each subregion according to different measured characteristics (Supplementary Files 2-5). Q: Cochran's Q statistic is a measure assessing the existence of heterogeneity in estimates of GDM prevalence.
2 I 2 is a measure assessing the percentage of between-study variation due to differences in GDM prevalence estimates across studies rather than chance. 3 Prediction intervals estimate the 95% confidence interval in which the true GDM prevalence estimate in a new study is expected to fall. 4 Heterogeneity between subgroups using random-effects model (fixed-effect model). 5 Overall pooled estimates in the 16 countries regardless of the tested population, sample size, and data collection period, using the most updated criteria when GDM is ascertained using different criteria in the same population.

Predictors of Heterogeneity in GDM
In the univariate meta-regression models, country, age, pregnancy trimester, BMI, and sample size were associated with variability in the prevalence of GDM at p<0.1. In the multivariate meta-regression model, only pregnancy trimester was retained, with no significant association with the prevalence of GDM at p<0.05. Compared with Saudi Arabia, the adjusted GDM prevalence was 135% (adjusted odds ratio [aOR], Q: Cochran's Q statistic is a measure assessing the existence of heterogeneity in estimates of GDM prevalence. 2 I 2 is a measure assessing the percentage of between-study variation due to differences in GDM prevalence estimates across studies rather than chance. 3 Prediction intervals estimate the 95% confidence interval in which the true GDM prevalence estimate in a new study is expected to fall. 4 Heterogeneity between subgroups using random-effects model (fixed-effect model). 5 Regardless of the year of the guidelines for the most updated criteria when GDM was ascertained, based on different criteria in the same population. 6 Overall pooled estimates in the 16 countries regardless of the tested population, sample size, and data collection period, using the most updated criteria when GDM was ascertained using different criteria in the same population.

Publication Bias in GDM Prevalence
Both the visual (funnel plot asymmetry) and statistical assessment (Egger's test, p<0.001) of publication bias suggested the role of a small-study effect (Supplementary File 7).

Quality Assessment of the GDM Research Reports
Supplementary Figure 2 presents the findings of the research report-specific quality assessment for relevant GDM prevalence studies. In all 102 research reports, the research question(s) and/ or objective(s) were clearly stated, and the study population group was clearly specified and defined. Half of the research reports (49.5%) did not provide information on the sample size calculation or justification. Most (79.2%) of the research reports used biological assays or extracted data from medical records to ascertain GDM, whereas the GDM status was self-reported in only five reports. In more than half (58.4%) of the 102 research reports, the tested sample size was at least 100 pregnant women. Overall, the research reports were judged to be of potentially low RoB, with an average of seven of the nine measured assessment items. Four

Main Findings
A total of 102 eligible research reports comprising 198 GDM prevalence studies were reported in 16 countries in the MENA region between 2000 and 2019. Most of these reports (58.41%) were from Iran and Saudi Arabia. The pooled prevalence of GDM in the 16 MENA countries was appreciably high (13.0%, 95% CI, 11.5-14.6%, I 2 , 99.3%), particularly in the GCC and North African countries. The prevalence of GDM increased with maternal age, gestational age, and BMI. It was also high in countries with a C-section rate of 15-29% and an MMR of >100/ 100,000 live births. The pooled GDM prevalence (13.0%) was alarmingly higher than that of European countries (2-6%) (133) but was similar to the sub-Saharan Africa region (14.0%). In contrast to the pooled prevalence estimates of Asia (11.5%) (134), the prevalence estimated in the present meta-analysis was slightly higher. The Asian meta-analysis included prevalence estimates from Saudi Arabia, Iran, and Qatar, and when compared with our estimates, they were 3.5% and 7.4% lower for Iran and Saudi Arabia, respectively, and 7.4% higher for Qatar (134). Such variations might be due to the differences in the literature search dates and languages, eligible sample size, GDM ascertainment criteria, and differences in the type of observational studies used for the prevalence estimation.
Our overall weighted GDM prevalence estimate depicted substantial heterogeneity (I 2 , 99.3%). This could be attributable to the less restrictive inclusion criteria in this review. In addition, the prevalence estimates of GDM can significantly differ with the variation in the GDM diagnostic criteria (135,136). We noted clinical inconsistency in GDM diagnostic criteria used in the prevalence studies we reviewed ( Table 4). This corresponds to the common use of existing nonuniform GDM diagnostic criteria in different countries (12,134). Given the importance of the prevalence of GDM in meaningful intervention development, its estimation can be affected by the inclusion of studies that use different GDM diagnosing criteria (137,138). The prevalence of GDM estimated based on the IADPSG criteria is usually high due to the low threshold for fasting blood glucose level relevant to other criteria. In our study, more than 25% of the studies used IADPSG criteria. To obtain homogenous and comparable prevalence estimates and to avoid confusion in practices of screening, diagnosis, and follow-up of GDM, health authorities should consider implementing uniform GDM diagnostic criteria nationally and across the MENA region.
The GDM prevalence estimates in our analysis suggested an increasing trend, parallel to the increase in BMI, correlating with the known fact that overweight and obesity are risk factors of GDM (139,140). Although this does not prove a causal link between these parameters, it inevitably might significantly reflect the impact of the high burden of overweight and obesity in several countries in the MENA region, such as Egypt and the six GCC countries (141). This highlights the importance of investigating dietitians' role in ensuring the appropriate caloric intake of GDM patients based on their BMI as per the recommendations of the ADA (142) and promoting exercise, especially among those with increased BMI (143).
GDM can have devastating maternal and birth consequences. Mothers with GDM are at higher risk of developing T2DM, dying, and undergoing C-section (23,24,144). Children born to mothers with untreated GDM face an increased risk of neonatal death and long-term disability (145,146). Notably, diabetes in pregnancy is a neglected cause of maternal mortality globally, affecting one of every sixth pregnancy in the world, and some of the known GDM morbidities that may cause maternal death are postpartum hemorrhage, obstructed labor, and preeclampsia (147). In our analysis, although the prevalence of GDM was higher (16.35%) in countries with high MMR (>100/100,000 live births), it was also substantial in countries with lower MMR (≤100/100,000 live births). Although this does not prove temporality, it highlights the importance of researching complications of GDM (if any) leading to maternal deaths, to help healthcare providers in the MENA region establish protocols to prevent these anticipated adversities. GCC countries with the highest GDM prevalence, as presented in this study, are also burdened with high T2DM (148). There is no doubt that controlling GDM would have multiple benefits in avoiding unfavorable health consequences for both mothers and their babies.

Strengths, Implications, and Limitations
The strengths of our review included its comprehensive characterization of the burden of GDM among pregnant women in several MENA countries. The review provides several weighted estimates in different population groups of the pregnant women at national, subregional, and regional levels that could be used, in addition to future work, to guide the planning, implementation, and evaluation of programs to prevent and control GDM. The overall and national-based pooled prevalence estimates might help policy makers of the respective MENA countries to contrast and quantify the local burden of GDM and introduce better policy initiatives regarding the flow of resources and funds for GDM care and management. Moreover, the finding of higher GDM prevalence corresponding to higher BMI categories might help in developing BMI-specific dietary and exercise guidelines. Furthermore, health authorities and organizations in the region are encouraged to review and consider standardizing the GDM diagnostic criteria at least at the national levels to improve the measurability and comparability of GDM rates and burden across the country and over time. Since we found a wide range of GDM diagnostic criteria used in the MENA region, health organizations across this region might consider moving toward the use of uniform GDM diagnostic criteria to produce better comparable statistical estimates in the future. For instance, in the UAE, different hospitals within the country use different GDM screening and ascertainment criteria (12). Having different GDM diagnostic criteria will preclude understanding the exact burden of the GDM.
Limitations included that our review did not provide any prevalence estimate for about 29% of the MENA region countries, as no prevalence data were available. This might have compromised the comprehensiveness of our prevalence estimates at the regional level. Since we believe that this study is the first to determine the prevalence of GDM in the MENA region, a comparison with previous similar estimates was not possible. This study offers scarce help regarding the prevalence of GDM with its associated comorbidities, such as gestational hypertension, preterm birth, and traumatic vaginal delivery (149), and separate review articles are warranted. The prevalence of GDM can also vary depending on several sociodemographic and maternal characteristics as well as within [urban or rural setting (150,151)] and between countries and regions; however, our study does not provide such distinction on the prevalence data. In some of the reviewed studies, detailed information on the methodology and GDM measurement procedures was missing, and this limits the category-based generalizability of the measured pooled GDM prevalence. For instance, the 3.35-times increase in the prevalence of GDM in studies reported before 2009 compared with studies reported after 2009 should be cautiously interpreted, as there was an overlap in the time period in 14 studies that tested 29,983 women. The various thresholds for fasting blood glucose level to diagnose GDM, applied on the several criteria considered from the studies, might suggest a bias in the estimated GDM prevalence. Unless estimated by rigorous comparable survey and testing methodology in individual population-based studies, the burden of GDM at the country, subregional, or regional level should not be interpreted as the burden of the measured outcomes at the population level. Moroever, this review did not explore the associations between various maternal and neaonatal characterstics and GDM. Therefore, future systematic reviews and meta-analyses studies focusing on the burden of GDM according to different maternal and neonatal characteristics as well as on the strength of association between various maternal characteristics and GDM are warranted.

CONCLUSIONS
Pregnant women in the MENA region are burdened with a relatively high GDM prevalence. Particularly, in the GCC and North African countries, the observed high burden of GDM may be mainly driven by the high prevalence of several risk factors for DM including overweight and obesity, parity, and late maternal age. To avoid maternal and newborn consequences, vigilant risk factor prevention programs and screening and management programs are necessary in the context of GDM. Moreover, unifying the GDM screening and diagnostic criteria, at least at the country level, is warranted to understand the precise burden of GDM. In countries that lack GDM burden data, high-quality research and surveillance programs are also warranted.

DATA AVAILABILITY STATEMENT
The data sets used and/or analyzed in the current study and the supplementary information files are available from the corresponding author on reasonable request.  The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Supplementary File 2 | Overall weighted prevalence of GDM in pregnant women in North African countries (Morocco, Algeria, Tunisia, Libya, Egypt) by pregnancy trimester, body mass index, study period, ascertainment methodology, rate of cesarean section deliveries, and maternal mortality.
Supplementary File 3 | Overall weighted prevalence of GDM in pregnant women in GCC countries and Yemen (the UAE, Qatar, Saudi Arabia, Kuwait, Yemen, Oman, Bahrain) by pregnancy trimester, body mass index, study period, ascertainment methodology, rate of cesarean section deliveries, and maternal mortality.
Supplementary File 4 | Overall weighted prevalence of GDM in pregnant women in Iran/Iraq by pregnancy trimester, body mass index, study period, ascertainment methodology, rate of cesarean section deliveries, and maternal mortality.
Supplementary File 5 | Overall weighted prevalence of GDM in pregnant women in the Levant region (Jordan, Syria, Palestine, Lebanon) by pregnancy trimester, body mass index, study period, ascertainment methodology, rate of cesarean section deliveries, and maternal mortality.
Supplementary File 6 | Univariate and multivariable meta-regression analyses to identify sources of heterogeneity in studies reporting on the prevalence of GDM in pregnant women by different measured characteristics.