The review protocol follows the recommendations of the Joanna Briggs Institute Manual of Evidence Synthesis . In addition, to ensure transparency and non-duplicity of the publication, as well as to minimize the presence of biases during its execution, this protocol was publicly registered in the PROSPERO database (CRD42020188474). The present study protocol is being reported in accordance with the reporting guidance provided in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) statement .
What is the prevalence of vitamin B complex deficiency in women of reproductive age in Brazil?
Studies will be selected according to the following criteria:
The participants are women of reproductive age (10 to 49 years of age), including pregnant and lactating mothers.
Outcome(s)/clinical situation of interest
The outcome of interest in this review is the deficiency of any vitamins from the B complex, diagnosed by laboratory/biochemical analysis following the recommended methods. No cut-off values were taken as an inclusion criterion. Studies in which the deficiency was identified by clinical signals and symptoms will not be eligible.
This systematic review will include cross-sectional studies or cohorts. In case of a cohort with multiple investigations of our interesting outcome, in global analysis, we will use the rate of the most recent analysis in which it was investigated, but for temporal analysis, all rates investigated in distinct moments of the follow-up will be included. Only fully published studies in the form of a scientific article or academic thesis, dissertation, or monograph will be eligible. No published studies or studies without full-text access, besides conference proceedings, editorials, case reports, case-control, letters, and trials will be not eligible.
Studies that provided the prevalence rates of the outcome will be eligible. Original studies that did not present the prevalence rates but described the proportion of eligible women with deficiency of any vitamin from B complex will be eligible if the prevalence may be calculated.
Only studies carried out in Brazil will be eligible. Studies in any setting (primary care, hospital, outpatient, clinic patients) representing the general population or subgroups that reflect populational conditions will be eligible. It will be ineligible: studies carried out only with patients with diseases or any pathological condition (except obesity associated with their comorbidities); situations that affect the digestion, absorption, metabolization rates, and daily requirements of the vitamins such as post-bariatric surgery, athletes, and vegetarians; and studies on which the vitamins deficiencies are assessed in a woman with genetic polymorphisms/variants that affect the body status of the vitamin B complex. However, studies that determine the prevalence of the polymorphism related to and the deficiency rates of vitamin B, not conditioning one to another, will be eligible.
Information sources and search strategy
The source of literature will be a systematic and structured search of the following electronic databases, planned to take place in 2020, June, and updated in 2022, June: MEDLINE (PubMed), Embase, Web of Science (main collection), Scopus, SciELO Citation Index (by Web of Science), Literatura Latino Americana e do Caribe em Ciências da Saúde (by Biblioteca Virtual em Saúde - BIREME), and a Brazilian database of academic thesis and dissertations (Biblioteca Digital Brasileira de Teses e Dissertações - BDTD). In addition, we will be performing a consultation with Brazilian researchers in the areas of health, medicine, nutrition, and public health allocated in public universities of the country, by e-mail, requesting potentially eligible studies after the explanation of the objectives and inclusion criteria. Hand-searching of reference lists of included studies, relevant reviews, or other relevant documents will also be performed to search for potentially eligible studies. The literature searches in electronic databases were designed using search terms underpinning the inclusion criteria and ensured by MeSH Terms and Emtree terms. The search strategies were created by combining search terms with Boolean operators and using search facilitators (quotation marks, truncation symbol, proximity operators) and were adapted by each database. Before the official electronic search, the search strategies were piloted to ensure sufficient specificity and sensitivity. The search strategies that will be used in the official literature search are presented in Supplementary file 1. No limits on publication date and languages will be defined.
The results from the literature search will be transferred to Ryyan QCRI, the Systematic Reviews web app  for the management of the references. In the app, the repeated references will be identified by one reviewer and only one of its multiple records will remain for the next step. In the phase of study selection, two reviewers will be analyzing the titles, abstracts, and full texts for eligibility. They will be blinded, and at the conclusion of the process, the judgment of each will be confronted. In case of discrepant judgment about the study eligibility, this study will be analyzed by the reviewers, and a consensus will be reached by discussion. Initially, the titles will be read, and if it appears eligible, the abstracts and full texts will be read to confirm eligibility.
Data extraction of the eligible studies will be performed by two reviewers, independently, registering the data in a customized Microsoft Office Excel® sheet, that will be pilot-tested prior to initiating the data extraction process. The extracted data by the two reviewers will be confronted and any disagreement resolved by new consultation of the study. In this process, the original authors may be consulted for any additional information not (fully) presented in the publication by e-mail or by Research Gate®. The information that will be extracted from the studies is the study identification, year of publication, type of publication, local (city and state), and period in which the study was conducted, study design, population group (pregnant, lactating woman, or women not pregnant or lactating), and their characteristics (age range or mean/median, setting of recruitment, the proportion of obese or malnourished participants, and the proportion of chronic diseases; the use of medicines and vitamin supplements; and other clinical and demographic characteristics; for pregnant, trimester of the analysis), sampling process, the number of participants recruited and analyzed, sample size calculation, vitamin B complex deficiencies investigated and their diagnostic method, the prevalence data, and their 95% CI or the number of cases for prevalence calculation.
Initially, we will perform a narrative description of the summary of the results. For the selection process, the results will be presented in accordance with the PRISMA flow diagram . The information extracted about each study (characteristics) and the prevalence rates (and 95% IC) of vitamin B complex deficiencies will be presented in a summary table. The results will be presented separately for pregnant, lactating women, and women not pregnant or no lactating and for each vitamin B assessed.
Meta-analysis will be done using a random model and using STATA 16.0 software and the Metaprop and Metapreg packages. Random effects model will be applied to estimate the pooled prevalence since we will expect heterogeneity among the studies a priori. The proportion (prevalence) of cases will be transformed to logit before combining the data in the metanalytic approach. Statistics for heterogeneity (test Q de Cochran and I2) and sensitivity analysis will be also done. For the meta-analysis, we entered the software the proportion of cases and the total numbers o evaluated subjects.
For each subpopulation (pregnant, lactating women, and women not pregnant or no lactating) and each vitamin B, if there are two or more studies that fulfilled each subgroup criteria, we intend to stratify the analysis by the following:
Time/year (temporal analysis), showing the prevalence rates in a timeline graph, and specific for folate, we will stratify in before 2004 and after 2004, because this year, it became mandatory to enrich wheat and corn flours with folic acid in the country
Age group: 15 to 19 years; 20 to 35 years; >35 years
State and Geographic Region of Brazil: 24 states + federal district; 5 macro-regions of the country (north, south, southeast; midwest; and northeast)
For pregnant: trimester of gestation (1st, 2nd, or 3th)
Lactation period: 1st semester, 2nd semester, and 2nd year
Other that may reduce the possible heterogeneity among the studies and/or modify the prevalence rates. Individual meta-analysis will be done by each subgroup above described, including only studies that attend these subgroup criteria. Meta-regression may be applied to investigate these factors/variables associated with the modification of the prevalence rates.
Risk of bias
The risk of bias will be assessed following the JBI Critical Appraisal Checklist for Studies Reporting Prevalence Data (available in https://jbi.global/sites/default/files/2020-08/Checklist_for_Prevalence_Studies.pdf). This tool encompasses nine items about the sampling process, sample size, description of the study subjects, identification/diagnosis of the clinical condition, statistical analysis, and response rate. Adequation for each domain/item will be classified as adequate (YES), inappropriate (NO), absence of clear information to judgment (UNCLEAR), or not applicable (N / A). This procedure will be done by two reviewers, independently, and the judgments will be confronted. Eventual disagreements in the risk of the bias assessment process will be solved by discussion until consensus. The results of this step will be presented in a table, describing the judgment for each domain assessed and their reason(s).