-
The systematic review has been registered on PROSPERO under ID number CRD42023388047, and this article is the result of a research project approved by the ethics committee of the Iran University of Medical Sciences (ID number IR. IUMS. REC.1401.827).
-
To diagnose anemia, your healthcare provider is likely to ask you about your medical and family histories, perform physical exams, and order blood tests. Tests might include: 1) Complete blood count (CBC). CBC is used to count the number of blood cells in a blood sample. For anemia, the test measures the number of red blood cells in the blood, called hematocrit, and the hemoglobin level in the blood. Typical adult hemoglobin values are 14–18 g/dL for men and 12–16 g/dL for women. Typical adult hematocrit values vary among different medical practices. However, they are generally between 40% and 52% for men and 35% and 47% for women. 2) A test to show the size and shape of red blood cells. It considers the size, shape, and color of the red blood cells.
-
This meta-analysis included studies conducted from 2010 to 2021. The study population included all studies that analyzed samples from patients with diabetes mellitus and anemia. Authentic English and Farsi electronic resources were used to extract information. Reliable English and Persian databases such as PubMed, Scopus, Web of Sciences, Google Scholar, Magiran, and SID were searched. The following search strategy was used to search national and international data banks: (“Anemia”) and (“Diabetes Mellitus”) using the OR operator to connect synonyms, and the AND operator was used to combine the results obtained from the OR operator. The search strategy for each database was consistent with the database’s search strategy (Supplementary Table S1, available in www.besjournal.com).
Search strategy Keywords Anemia*[Title/Abstract] OR “Anemia”[Title/Abstract] OR iron deficiency,Anemia[MeSH] OR “iron deficiency”[Title/Abstract] OR iron deficiency* [Title/Abstract] OR Anamia [Title/Abstract] OR Anaemia [Title/Abstract] OR Anamia* [Title/Abstract] OR Anaemia*[ Title/Abstract] OR anemi*[tw] Anemia Diabetes Mellitus[MeSH Terms] OR Diabetes Mellitus,Type 2[MeSH] OR Diabetes Mellitus,Type 1[MeSH] OR Diabetes Mellitus,Type 2 OR [Title/Abstract] OR Diabetic[MeSH Terms] OR Diabet*[tw] OR Diabetes Mellitus,Type 2[Title/Abstract] Diabetes Mellitus Prevalence[MeSH] OR Epidemiology[MeSH] OR Survey[MeSH] Prevalence 1 AND 2 AND 3 AND Search Table S1. Search keywords
After completing the search and removing duplicates, the articles were imported into EndNote. All articles were first evaluated based on their titles, then based on their abstracts, and finally, the full texts of the relevant studies were evaluated and entered into the quality review stage. The study was conducted in a transparent and unbiased manner, and all articles were carefully evaluated and presented without bias. We made every effort to obtain the full texts of the relevant articles and carefully evaluated them. All articles used in this study were appropriately referenced to ensure the credibility of the scientific materials.
-
This review included observational and primary studies that examined the prevalence of anemia in patients with diabetes. Studies that followed the World Health Organization’s (WHO) definition of anemia to determine its prevalence were also included. Primary studies written in Farsi or English were eligible for inclusion. Additionally, the reference lists of related primary and review articles were scanned to identify relevant studies. Studies that did not report anemia prevalence according to the World Health Organization cutoff for anemia were excluded. In this study, anemia was defined using the WHO definition of hemoglobin levels < 11 g/dL.
-
In the next step, researchers used the Newcastle-Ottawa Scale (NOS) tool, a standard checklist to evaluate the quality of the observational studies, including cross-sectional, case-control, and cohort studies. The criteria used to measure bias included sample size calculations, information bias, and selection bias. Information from studies that met the necessary quality criteria was extracted and entered into a prepared electronic form. The extracted information generally included the name of the article’s main author, year of publication, study location, quality assessment score, type of study, sample size, and main findings of the article.
To prepare the data and information for the meta-analysis, all steps (search, selection, quality assessment, extraction, and homogenization of results) were carried out independently by two researchers. In case of any disagreement between the two researchers, a third researcher was involved in resolving the issue through discussions.
To ensure the correct selection of articles related to the prevalence of anemia in patients with diabetes mellitus worldwide in accordance with the inclusion criteria, two researchers (MFG and NA) were independently responsible for the article selection process. Initially, all articles related to the prevalence of anemia in patients with diabetes worldwide were collected, and articles that contained relevant keywords in their titles or abstracts were included in the initial list, whereas unrelated articles were removed. Finally, only articles that investigated the prevalence of anemia in individuals with diabetes mellitus worldwide were included in this study. After selecting the studies, the required variables, such as the study type, sample size, number of patients with diabetes with anemia, demographic characteristics of the study participants, and the time and place of the study were extracted.
Implementation limitations of the plan and their reduction methods were: 1) Abstracts and summaries of articles presented at international conferences and meetings were excluded; 2) Interventional studies or those focused solely on clinical status were excluded; 3) Only primary research studies were included; review articles and secondary studies were excluded. 4) If the required information was not available in the selected article, the study was excluded.
-
To assess possible biases in the studies, the NOS checklist[29] was used to evaluate quality. Two researchers (MFG and NA) independently conducted the evaluations. Any disagreements between the researchers were resolved through discussion or with the guidance of respected professors and research teams. If the information necessary to assess possible bias in the study was not available, the author of the article was contacted. Finally, the studies were scored according to the table provided in Supplementary Figure S1, available in www.besjournal.com.
-
The WHO defines anemia as a blood hemoglobin level of < 7.7 mmol/L (13 g/dL) in men and 7.4 mmol/L (12 g/dL) in women. Typically, the evaluation of the cause of anemia includes CBC, peripheral smear, reticulocyte count, and serum iron indices.
-
All meta-analyses were conducted using random models. Heterogeneity between studies was reported in the meta-analysis using the chi-square, I2, and Tau2 tests. Subgroup analyses were also performed based on publication year, continent, age group, medications, disease control status, and disease duration. Publication bias was calculated using Egger’s and Begg’s tests. The meta-analysis was performed using STATA 17 software at a significance level of 05.
-
After conducting an initial search of the relevant databases, 27,283 articles were retrieved. Of these, 606 were selected for further screening based on their titles. The abstracts of these articles were reviewed, and 88 that did not contain the desired information were excluded. The remaining 234 articles underwent a full-text assessment for eligibility. Of these, 146 were subjected to qualitative evaluation using the NOS checklist. After this critical evaluation, 51 articles, including 26,485 patients with diabetes, were included. A flowchart of the study selection is shown in Figure 1.
-
Baseline characteristics of the included studies we identified 51 articles related to the study objectives (Figure 1), of which 31 (59%) were conducted in Asia, 17 (35%) in Africa, two (4%) in America, and one (2%) in Europe. Ethiopia, India, and Pakistan had the highest number of studies among the countries analyzed (Supplementary Table S2, available in www.besjournal.com).
NO. First author Publication years Country Average age Sample size
N−DiabeticType DM N. treatment−type N. Duration of DM (year) N. Poor Glycemic
control (HBA1C > 7)Prevalence anemia (%) Quality score 1 Abat et al.[1] 2013 Africa 40/9 (16/8) 193
191DM1
DM2Metformin (160)
Insulin (224)< 5 (206)
6–10 (124)
> 11 (54)4/2
78/17 2 adejumo et al.[2] 2012 Africa 48/8 (9/09) 144 DM2 80 15/3 6 3 Al−ghazaly et al.[3] 2019 Asia 53 (12) 324 DM2 Metformin (123) < 5 (228)
6–10 (51)
> 11 (45)23/8 5 4 AL−Salman et al.[4] 2015 Asia 52/3 (14/1) 227 DM2 55/5 8 5 Argoon et al.[5] 2014 Africa 43/08 (9/3) 100 DM2 34 5 6 Arkew et al.[6] 2021 Africa 50 (16) 134 DM2 Metformin (24) < 7 (24)
> 7 (24)78 17/9 7 7 Barbieri et al.[7] 2015 Africa 60/9 (8/9) 146 DM2 34/2 7 8 Awofisoy et al.[8] 2021 American 56/3 (11/5) 150 DM2 Metformin (133)
Insulin (36)45/5 7 9 Bekel et al.[9] 2019 Africa 62/3 (0/98) 410 DM2 < 5 (195)
> 5 (177)231 34/8 7 10 Bin−Bin He et al.[10] 2015 Asia > 60 1 997 DM2 22 7 11 Chung et al.[11] 2017 Asia 57/8 (13/5) 2 230 DM2 Insulin (382) 31 7 12 Conway et al.[12] 2018 American 51/3 (9/1) 5 210 DM2 13 Shaheen et al.[13] 2021 Africa 48/6 (15/9) 100 DM2 14 Engidaw et al.[14] 2020 Africa 57 (11/5) 297 DM2 15 Rathod et al.[15] 2018 Asia 24−72 100 DM2 16 Gunvanti. Rathod et al.[16] 2016 Asia 58 (14) 200 DM2 17 Hodel et al.[17] 2020 Africa 54 (19−91) 64 DM2 18 Hosseini et al.[18] 2014 Asia 53/9 (1) 305 DM2 Insulin (29) 19 Mirghani et al.[19] 2018 Asia 59/6 (9/9) 170 DM2 20 Kebede et al.[20] 2021 Africa > 60 372 DM2 < 5 (65)
6−10 (193)
> 10 (114)21 Kim et al.[21] 2021 Asia 60/09 (0/31) 2 903 DM2 22 Manglunia et al.[22] 2018 Asia 55/5 (13/7) 120 DM2 23 Jin Ook Chung et al.[11] 2018 Asia 56/8 (9/5) 1 300 DM2 24 Pehlivanoglu et al.[23] 2020 Asia 54/6 (11/8) 229 DM2 25 Danish Qureshi et al.[24] 2020 Asia 56/9 (7/2) 117 DM2 56/4 26 Panda et al.[25] 2018 Asia 51/3 (8/8) 54 DM2 63 27 Shabeeb et al.[26] 2021 Asia 30−79 150 DM2 65/3 28 Arshad et al.[27] 2021 Asia 54 227 DM2 < 5 (86)
6−10 (82)
> 10 (83)80 29 Taderegew et al.[28] 2020 Africa 53/7 (10/4) 249 DM2 < 5 (102)
6−10 (84)
> 10 (63)30 Tujuba et al.[29] 2021 Africa 40 (IQR:20) 325 DM2 < 5 (91)
6−10 (111)
> 10 (123)31 Yorke et al.[30] 2021 Africa 54/8 (10/8) 195 DM2 < 5 (57)
6−10 (60)
> 10 (75)32 Umeshchandara G et al.[31] 2021 Asia 54/1 (12/1) 230 DM2 < 5 (88)
6−10 (52)
> 10 (25)33 Chio et al.[32] 2015 Asia 63 (11) 1 142 DM2 34 Rathore et al.[33] 2018 Asia < 50
> 60200 DM2 Metformin (126) 35 Wang et al.[34] 2020 Asia 59/8 (12/9) 901 DM2 36 Shams et al.[35] 2015 Asia 51 (12/4) 130 DM2 Insulin (32) 98 37 Ahmed et al.[36] 2017 Asia 56/5 (10/5) 640 DM2 Metformin (98)
Insulin (32)38 Trevest et al.[37] 2014 European 83/6 (5/2) 115 DM2 Insulin (57) 39 sarosh et al.[38] 2022 Asia 49/6 (13) 200 DM2 40 Mulavu et al.[39] 2020 Asia 54/4 (14) 101 DM2 Metformin (7)
Insulin (63)41 Aynalem et al.[40] 2022 Africa 54 (12) 357 DM2 42 Idris et al.[41] 2018 Asia 60/5 (9/5) 808 DM2 < 5 (231)
6−10 (310)
> 10 (266)43 Wali et al.[42] 2022 Asia 48/5 (7/5) 215 DM2 44 Arani et al.[43] 2022 Asia 57/7 (8/6) 415 DM2 45 Feteh et al.[44] 2016 Africa 56/5 (10/6) 636 DM2 Metformin (512)
Insulin (68)46 Fiseha et al.[45] 2019 Africa 45 (14/6) 123
289DM1
DM2< 5 (253)
6−10 (107)
> 10 (52)47 Grossman et al.[46] 2013 Asia 63 (9/6) 445 DM2 Metformin (335)
Insulin (52)48 Hailu et al.[47] 2020 Africa 18−80 54
204DM1
DM249 Wang et al.[34] 2020 Asia 60/6 (11/4) 367 DM2 50 sharif et al.[48] 2014 Asia 54/4 (9/5) 200 DM2 143 51 Kaushik et al.[49] 2018 Asia 51−60 100 DM2 Table S2. Characteristics of the included studies in Meta-analysis
Table 1 presents the meta-analysis results of the pooled prevalence of anemia in patients with diabetes and its subgroups by year, continent, age, and sex. The pooled prevalence of anemia across all samples was 35.45 (95% CI: 30.30–40.76) (Figure 2), with an effect size of 35.0 (95% CI: 29.85–40.32) excluding the study by Trevest et al.[24], with approximately similar estimates of 36.18 (95% CI: 29.77–42.83) for men and 36.52 (95% CI: 29.31–44.04) for women. Owing to the limited number of studies conducted in Europe and the US, it was not possible to estimate their effect size accurately. However, in Asia and Africa (Figures 3 and 4), the prevalence of anemia in patients with diabetes was significantly higher in Asia (40.02, 95% CI: 32.72–47.54) than in Africa (28.46; 95% CI: 21.90–35.50) (P for heterogeneity = 0.029). This difference was more pronounced in Asian men and women than in their African counterparts, with Asian women showing a larger effect size than Asian men (44.54, 95% CI: 39.43–49.44 vs. 29.85, 95% CI: 23.17–37.35). The trend of reporting the prevalence of anemia in patients with diabetes has increased over time in recent years. However, many factors may have influenced these changes and their distribution.
Group Number of studies Pooled ES (95% CI) Heterogeneity Publication bias (P-value) I2 (%) P-value Begg’s Egger’s Total Total 51 35.45 (30.30–40.76) 98.68 0.001 without (Trevest) 50 35.0 (29.85–40.32) 98.69 0.001 Subgroup analysis Year of publication 0.01 0.001 2012 1 15.28 (9.83–22.21) − − 2013 2 15.70 (12.96–18.64) − − 2014 4 46.48 (29.02–64.39) 95.57 0.001 2015 5 38.97 (26.85–51.83) 97.81 0.001 2016 2 35.30 (32.09–38.58) − − 2017 2 26.19 (24.60–27.82) − − 2018 9 46.43 (30.34–62.90) 99.27 0.001 2019 3 29.80 (23.53–36.48) − − 2020 9 23.97 (16.36–32.52) 95.22 0.001 2021 10 38.28 (21.41–56.73) 99.15 0.001 2022 4 40.70 (10.21–75.93) 99.37 0.001 Continent Africa 18 28.46 (21.90–35.50) 96.15 0.001 Asia 30 40.02 (32.72–47.54) 98.90 0.001 America 2 12.47 (11.60–13.38) − − Europe 1 59.13 (49.57–68.21) − − Age category, years < 56 28 39.43 (31.92–47.19) 97.13 0.001 ≥ 56 23 30.82 (24.22–37.83) 99.07 0.001 Male Total 34 36.18 (29.77–42.83) 96.37 0.001 0.005 0.001 Year of publication 2014 4 43.32 (27.49–59.87) 89.78 0.001 2015 3 37.58 (14.87–63.54) − − 2018 5 52.42 (34.44–70.08) 94.09 0.001 2019 3 33.11 (22.96–44.10) − − 2020 4 26.57 (11.41–45.21) 93.95 0.001 2021 10 39.13 (21.95–57.79) 98.06 0.001 Continent Africa 14 29.83 (21.9–38.40) 93.27 0.001 Asia 18 39.43 (29.85–49.44) 97.45 0.001 Female Total 32 36.52 (29.31−44.04) 97.17 0.001 0.02 0.003 Year of publication 2015 3 48.17 (19.89−77.09) − − 2018 5 56.33 (33.44−77.93) 95.32 0.001 2019 3 24.15 (19.01−29.68) − − 2020 4 34.56 (24.73−45.09) 79.54 0.001 2021 10 37.58 (20.05−56.95) 98.61 0.001 Continent Africa 14 26.21 (18.25−35.01) 94.23 < 0.001 Asia 17 44.54 (33.28−56.08) 98.08 0.002 Note. ES, effect size. Table 1. The overall and subgroup prevalence of anemia in diabetes mellitus
Table 2 shows the indicators related to patients with diabetes. The prevalence of anemia was higher in insulin recipients than in metformin recipients (46.97% vs. 43.59%). Patients with diabetes and poor diabetes control had a higher prevalence of anemia (31.96%, 95% CI: 20.82%–44.24%). The pooled prevalence of anemia was also higher in patients with longer disease duration. The effect size was estimated to be 46.93% in patients with > 10 years of illness, 36.71% in those with 5–10 years of illness, and 21.62% in those with < 5 years of illness.
Variables Number of studies Pooled ES (95% CI) Heterogeneity Publication bias (P-value) I2 (%) Tau2 Begg’s Egger’s Metformin 4 43.59 (28.75–59.03) 89.50 0.08 0.497 0.183 Insulin 6 46.97 (25.54–68.98) 96.59 0.28 0.091 0.033 Poor control 10 31.96 (20.82–44.24) 96.64 0.16 0.421 0.179 Good control 10 23.84 (17.03–31.36) 84.20 0.06 0.421 0.919 Duration < 5 (years) 13 21.62 (13.76–30.65) 94.56 0.13 0.100 0.239 Duration 5–10 (years) 13 36.71 (25.16–49.07) 95.94 0.19 0.393 0.321 Duration > 10 (years) 10 46.93 (30.35–63.85) 95.98 0.28 0.835 0.585 Note. ES, effect size. Table 2. Summary of the analysis of the indicators related to patients with diabetes
In Table 1, although the values of some subgroups are reported despite the limited number of studies, only subgroups with at least three pooled studies reported the prevalence of anemia in other strata to determine the size of their effect.
Publication bias and sensitivity analysis were performed using Begg's and Egger's tests. Publication bias was not observed for indicators related to diabetes (metformin, insulin, poor and good control, and duration; P > 0.05) but was observed in the sex subgroup (P < 0.05). The small number of studies in some subgroups and the context of the present review affected the publication bias. In any case, the prevalence is positive, and the lack of studies with negative prevalence can cause errors in estimating prevalence. Furthermore, the single group used in this review affected the values. Sensitivity analysis indicated that the results were not influenced by a single study and did not alter the direction of estimation (Figure 5, Table 1). Figure 6 shows the prevalence of the pooled effect size according to year.
HTML
Diagnosis
Searching Strategy
Inclusion/Exclusion Criteria
Data Extraction
Quality Assessment
The WHO Diagnostic Criteria for Anemia
Heterogeneity Assessment and Statistical Analysis
Search and Selection
Study Characteristics
bes23105+Supplementary Materials.pdf |