Eczema as a protective factor for brain cancer: a meta-analysis

Background Brain cancer is one of the most aggressive cancer types owing to poor treatment effects. Epidemiological studies have demonstrated that allergies may increase the disease risk. Therefore, this study evaluated the association between eczema and the risk of various brain cancers. Methods We systematically searched the PubMed and Embase databases from their inception until June 23, 2022. Two reviewers independently reviewed and screened the articles, extracted data, assessed the study quality, and pooled the results. Stata software was used to generate pooled odds ratios and 95% confidence intervals (CIs). Results We included 20 studies comprising 5,117,222 patients that investigated the relationship between eczema and brain cancer. Eczema was significantly inversely associated with the risk of brain cancer (odds ratio [OR], 0.82; 95% CI, 0.77–0.87), glioma (OR, 0.53; 95% CI, 0.14–2.02), meningioma (OR, 0.74; 95% CI, 0.66–0.84), and acoustic neuroma (OR, 0.60; 95% CI, 0.41–0.88). Interesting, The strong correlation between eczema and the reduced risk of brain cancer was observed in people over 16 years old (OR, 0.79; 95% CI, 0.71–0.88), but not in those under 16 years old (OR, 0.94; 95% CI, 0.79–1.11). In addition, subgroup analyses found that eczema significantly decreased the glioma risk in Europeans (OR, 0.73; 95% CI, 0.65–0.82) but not Australians (OR, 0.53; 95% CI, 0.14–2.02) or Americans (OR, 1.01; 95% CI, 0.69–1.46). Conclusion Eczema may be considered as a potential protective factor of brain cancer in population aged over 16 years. However, this relationship requires verification using large-scale clinical data. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-022-10471-0.


Introduction
Brain cancer is one of the most aggressive tumors owing to poor treatment effects [1,2]. More than 90% of brain tumors occur in the brain parenchyma, with the remainder occurring in the meninges, spinal cord, and cranial nerves [3,4]. Brain cancers are the primary cause of cancer-related deaths in children and adults worldwide. In 2016, 330,000 central nervous system cancer cases were reported globally, of which 227,000 patients died [5]. The reasons for high rates of brain tumor recurrence and mortality remain unclear. However, several have investigated the potential risk factors for brain tumors. For example, epidemiological studies have shown that air pollution, diet, allergies, medications, genetic predispositions, and some demographic characteristics (e.g., age, sex, and race) may increase the disease risk [6]. However, these factors require confirmation. Therefore, the need to identify new reliable risk factors for early brain tumor interventions is urgent.
Eczema is a common chronic cutaneous inflammatory disease affecting 10-30% of children and 2-10% of adults worldwide [7]. Some reports indicate that eczema affects the cancer risk, but the results are controversial [8,9]. Therefore, this study performed a meta-analysis to assess the relationship between eczema and brain cancer to clarify the risk factors.

Search strategy
This meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) [10] and the Meta-analysis of Observational Studies in Epidemiology (MOOSE) [11] guidelines. We searched the PubMed and Embase databases for studies focusing on the relationship between eczema and brain cancers from inception until June 23, 2022. Supplementary Material 1 describes the specific search parameters. Two reviewers (LJ and ZY) independently performed a preliminary screening of relevant literature based on the titles and abstracts. Then, the full text of these papers was reviewed to identify those suitable for analysis. Differences in opinions were resolved through a discussion with the corresponding author. In addition, articles in the bibliographies of the papers mentioned above that met the screening criteria were also assessed for analysis potential.

Selection criteria
The inclusion criteria were: a) human studies; b) observational studies including cohort and case-control study designs; c) published in English; and d) studies focusing on the relationship between eczema and the risk of brain tumors.
The exclusion criteria were: a) reviews, animal studies, abstracts, or conference proceedings; b) unpublished data or case reports; c) duplicate literature and duplicate data; and d) studies lacking specific data on exposure (eczema) and endpoints (brain tumors).

Data extraction and quality assessment
The following relevant information was independently extracted by two reviewers (XSL and XYD) from the included studies: first author's name, country of origin, publication year, study design, sample size, age, percentage of females, tumor type, number of patients, years of follow-up, and odds ratios (ORs) with 95% confidence intervals (CIs). For this meta-analysis, we defined eczema as the exposure factor and the occurrence of various brain tumors as the outcome. Data not directly obtainable from the text were extracted from the associated information (e.g., graphs). The corresponding authors were contacted to retrieve missing data.
Two authors (TYR and YWM) used the Newcastle-Ottawa Scale (NOS) [12] to evaluate the quality of the included studies based on three aspects: selection, comparability, and outcome assessments. The scores were summed, and studies with NOS scores of ≥5 were considered high quality [13].

Statistical analyses
Stata software (version 12.0; Stata Corp. LLC, College Station, TX, USA) was used for the meta-analysis. The heterogeneity of the included papers was estimated using a homogeneity test (Q test) and the I 2 value. A P-value of > 0.1 and I 2 value of < 50% indicated acceptable heterogeneity, and then a fixed effects model was used to calculate pooled ORs and 95% CIs [14]; otherwise, the random effects model was used [15]. Subgroup analyses were performed to determine differences in the associations between eczema and brain tumors by age (≥16 years and <16 years) and study location (Australians, Americans, Europe and others). Sensitivity analyses were also performed by excluding one study at a time and then rerunning the analyses to calculate the effects for the remaining studies to determine if the pooled results were markedly affected by a single study. Finally, publication bias was evaluated using Begg's funnel plot and Egger's weighted regression test.

Eczema and brain cancer risk
Eczema was significantly associated with a decreased brain cancer risk in the analysis with all 20 studies [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35] (OR, 0.82; 95% CI, 0.77-0.87; Fig. 2A). Insignificant heterogeneity greatly enhanced the reliability of the results (P = 0.004, I 2 = 48.1%; Fig. 2A). The sensitivity analysis indicated that excluding each study in turn all studies were distributed between the upper and lower 95% CIs with the estimation line, and excluding any one study did not alter the overall combined results (Fig. 2B). Publication bias, assessed by Begg's rank correlation and Egger's linear regression, did not identify publication bias among the studies (Begg's: P > |z| = 0.362; Egger's: P = 0.183, 95% CI: − 1.900-0.384; Fig. 2C). Next, to test this hypothesis that the brain cancer risk might be influenced in different aged eczema patients, subgroup analyses based on age were performed; we The association between eczema and brain cancer risk. A Forest plot of the estimated effects of eczema on brain cancer risk. B Sensitivity analysis conducted by recalculating the pooled results of the primary analysis following the exclusion of one study per iteration. C Begg's and Egger's tests indicate the lack of publication bias among such studies. D Subgroup analysis of the estimated effects of eczema on brain cancer risk. CI, Confidence interval; ES, Effect size identified a strong association between eczema and brain cancer risk in those aged over 16 years (OR, 0.79; 95% CI, 0.71-0.88) but no association for those under 16 years old (OR, 0.94; 95% CI, 0.79-1.11; Fig. 2D).

Discussion
This meta-analysis aimed to advance our understanding of the association between eczema and brain cancer by analyzing 20 related papers. The strong correlation between eczema and the reduced risk of brain cancer was observed in people over the age of 16, but not in the age of 16, which indicated that eczema may be considered as a potential protective factor Fig. 4 The association between eczema and meningioma risk. A Forest plot of the estimated effects of eczema on meningioma risk. B Sensitivity analysis. CI, Confidence interval when assessing the risk of brain cancer in population aged over 16 years. At the same time, eczema was significantly associated with a decreased risk of glioma, meningioma, and acoustic neuroma. However, the relationship between eczema and the risk of glioma differed regionally. Specifically, eczema significantly decreased the glioma risk in the European subgroup without heterogeneity but not in the Australian and American subgroups.
Many previous studies have investigated associations between allergic diseases and the risk of various cancers [29,36,37] but with inconsistent results. For instance, Hwang et al. reported complex and sitespecific relationships between allergic diseases (e.g., eczema, allergic rhinitis, and asthma) and cancer risk [8]. Unexpectedly, the incidence of brain cancer was higher in patients with eczema [36]; however, the sample size was too small to be reliable. Conversely, another study reported that children with eczema did not have a lower risk of brain tumors (OR, 0.52; 95% CI, 0.17-1.57) [29], which supported our subgroup analysis results. Additionally, a recent study demonstrated that eczema, an allergic disease that increases immune surveillance, may be related to the low risk of many cancers, supporting our results [37]. Unlike our meta-analysis, these reviews preferentially pooled crude estimates and included broader criteria, focusing on the effects of all allergic diseases on the incidence of several cancers; nonetheless, their results were consistent with ours.
Numerous studies have reported inverse associations between eczema and several cancer sites [31,38,39]. However, how eczema affects the incidence of tumors remains debatable. Many researchers agree that the severity of eczema varies widely, and only the most severe cases are identifiable [34]. However, in our study, eczema was associated with a reduced risk of brain cancer. Some theories have been proposed to support this conclusion.
Specifically, the effects of allergic diseases on the risk of glioma are close to consensus. One hypothesis is that since eczema is an allergic disease, it may indicate a high state of immune surveillance [40][41][42]. Consequently, the overactive immune system suppresses abnormal cell growth or proliferation. However, the mechanisms by which the enhanced immune surveillance hypothesis helps suppress tumor growth remain unclear. An additional explanation for this protective effect is that allergies are a defense mechanism against infestation by large parasites and environmental toxins [43]. As a result, individuals with stronger allergic reactions excrete and eliminate environmental toxins or carcinogens more efficiently throughout their lives than those without strong reactions. Interestingly, the association between eczema and glioma risk varied in the European, Australian, and American regions. However, the mechanisms underlying these differences have not been elucidated. Studies have shown that the glioma incidence rate varies globally, with the highest rates observed in America, Australia, and Northern Europe [44]. Therefore, the risk of glioma is closely related to the region, descent, and race or ethnicity [45].
This study had some limitations. First, some of our results had heterogeneity, which might affect their reliability. However, heterogeneity did not exist between eczema and meningiomas and acoustic neuromas; the relationship between eczema and glioma was an exception. Therefore, heterogeneity minimally affected our conclusions. Second, our study population spanned a wide age range, but data limitations prevented us from investigating the effects of age on the results; we could only evaluate associations. Finally, we did not find publication bias in our analyses because researchers only reported positive results. Therefore, we inevitably could not assess the effects of unpublished studies with negative results. Nonetheless, our study included many patients and individually eliminated studies to evaluate sensitivity, making our results more comprehensive and reliable.