Citrus fruit intake and lung cancer risk: A meta-analysis of observational studies

To explore the hypothesis that Citrus intake may reduce the risk of lung cancer.

Meta-analyses of Dichotomy and dose-response relationship.

We searched online literature databases including PubMed, Embase, and Cochrane Library to screen relevant articles available up to 27 July 2020. Search terms included (i) Citrus, Fruit, Diet, Dietary; (ii) cancer, neoplasm, tumor (iii)lung; (iv)case-control, cohort, prospective.

The selection of studies and the meta-analysis were carried out by following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The following inclusion criteria were chosen: (i) epidemiological studies with case-control or cohort design; (ii) human participants; (iii) studies investigated the relationship between Citrus fruit intake and lung cancer risk; (iv) if data were duplicated in more than two studies, we brought the most recent or all-sided study into this analysis. We collected all full-text articles that met the inclusion criteria. We applied the following exclusion criteria to the full-text articles, including possible articles listed by manual search: (i) there was no represented odds ratio (OR) or relative risk (RR) estimate and its corresponding 95 % confidence interval (95 % CI) (or data to calculate them) for the highest versus lowest levels of Citrus fruit consumption (ii) reviews, systematic reviews and meta-analyses; (iii) there was no data of Citrus fruit intake at the individual level.

Two reviewers independently performed the extraction of data from eligible studies.

Adjusted odds ratios (ORs) and 95 % CIs were combined and weighted by the method of "Dersimonian and Laird" to produce pooled ORs using a random-effects model. Moreover, we utilized the method reported by "Longnecker and Greenland" to evaluate linear trends and 95 % CIs by the ORs' natural logs and corresponding CIs from categories of Citrus intake. Finally, we evaluated the risk of publication bias and selection bias by inspecting for asymmetry in the pre-specified funnel plots of the study OR against the standard error of the OR's logarithm and by "Egger's test".

We included twenty-one studies in the final review. Pooled analyses suggested that those with the highest Citrus fruit intake compared to the lowest intake had a 9% reduction in lung cancer risk [OR 0.91 (95 % CI 0.84-0.98)]. We found a nonlinear association between Citrus intake and lung cancer risk in the dose-response analysis (p = 0.0054) and that the risk reached the minimum (OR = 0.91) around 60 g/d. However, no obvious dose-response association was observed with intakes above 80 g/d.

We found that Citrus fruit intake was negatively associated with the risk of lung cancer. Besides, there was a nonlinear dose-response relationship between Citrus intake and lung cancer risk within a certain range.