2.1 Literature search strategy
A literature search was carried out via computer and hand searches. We used vitamin D receptor or VDR or rs2228570 or FokI and tuberculosis and pulmonary search published articles in Chinese and English database. We use these databases: SinoMed data-base, CNKI, Wanfang, VIP, PubMed, Cochrane Library, Elsevier ScienceDirect, Embase, Web of Science and SpringLink.
2.2 Literature inclusion and exclusion criteria
The literature inclusion criteria: (1) type of study: case-control studies; (2) content of study: FokⅠ gene polymorphisms and TB susceptibility; (3) research object: cases of inclusion were diagnosed TB, and exclusion autoimmune diseases, diseases of the endocrine system, long time use of adrenal cortical hormones or immunosuppressive drugs, and HIV infection[2, 3]; (4) control of inclusion were healthy control group: X-ray chest showed no abnormalities, PPD test less than 5mm3; (5) the studies with enough data to calculate odds ratio(OR) and 95% confidence interval(CI)[4, 5]; (6) there were no restrictions on race and gender, and aged ≥ 16years.
Literature exclusion criteria: (1) the studies with incomplete data or required data cannot be calculated OR and 95%CI; (2) there were reporting the clinical consequences; (3) repeated research; (4) controls were healthy subjects with no history of contact with TB patients.
2.3 Data extraction
All data were extraction independently by 2 investigators. The data were extraction included: the first author’s name, publication year, the area of research, case group and control group sample.
2.4 Literature quality evaluation
We evaluated the literature with 2 investigators using Newcastle-Ottawa Scale(NOS). The scale includes a total of 8 items including the selection, comparability and exposure. The highest score is 9 points, and the score more than 7 points as high-quality literature[6].
2.5 Statistical analysis
In this study, we considered the f allele is a gene that increases risk. Therefore, we use the five models: allele contrast(f vs. F), dominant model(ff + fF vs. FF), recessive model(ff vs. fF + FF), co-dominant model(ff vs. FF, fF vs. FF) to calculate OR and 95%CI. The het-erogeneity between studies was assessed with a based Q test and I2 statistics. If there is no significant heterogeneity (I2 < 50%, Q test P > 0.05), select the fixed effect model; otherwise, choose random Effect model. Sensitivity analysis to verify the stability and reliability of meta-analysis results[4, 5, 7]. The Begg rank correlation method was used to statistically assess publication bias. All the statistical tests were conducted with R software(version 4.0.5), and the statistical significance was defined as p < 0.05. All p values were two-sided.