Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Obesity and Risk for Brain/CNS Tumors, Gliomas and Meningiomas: A Meta-Analysis

  • Theodoros N. Sergentanis,

    Affiliation Department of Hygiene, Epidemiology and Medical Statistics, Medical School, National University of Athens, Athens, Greece

  • Georgios Tsivgoulis,

    Affiliation Second Department of Neurology, “Attikon” University Hospital, Medical School, National University of Athens, Athens, Greece

  • Christina Perlepe,

    Affiliation Department of Hygiene, Epidemiology and Medical Statistics, Medical School, National University of Athens, Athens, Greece

  • Ioannis Ntanasis-Stathopoulos,

    Affiliation Department of Hygiene, Epidemiology and Medical Statistics, Medical School, National University of Athens, Athens, Greece

  • Ioannis-Georgios Tzanninis,

    Affiliation Department of Hygiene, Epidemiology and Medical Statistics, Medical School, National University of Athens, Athens, Greece

  • Ioannis N. Sergentanis,

    Affiliation Hôpital de Psychiatrie, Hôpitaux Universitaires de Genève, Geneva, Switzerland

  • Theodora Psaltopoulou

    tpsaltop@med.uoa.gr

    Affiliation Department of Hygiene, Epidemiology and Medical Statistics, Medical School, National University of Athens, Athens, Greece

Obesity and Risk for Brain/CNS Tumors, Gliomas and Meningiomas: A Meta-Analysis

  • Theodoros N. Sergentanis, 
  • Georgios Tsivgoulis, 
  • Christina Perlepe, 
  • Ioannis Ntanasis-Stathopoulos, 
  • Ioannis-Georgios Tzanninis, 
  • Ioannis N. Sergentanis, 
  • Theodora Psaltopoulou
PLOS
x

Abstract

Objective

This meta-analysis aims to examine the association between being overweight/obese and risk of meningiomas and gliomas as well as overall brain/central nervous system (CNS) tumors.

Study Design

Potentially eligible publications were sought in PubMed up to June 30, 2014. Random-effects meta-analysis and dose-response meta-regression analysis was conducted. Cochran Q statistic, I-squared and tau-squared were used for the assessment of between-study heterogeneity. The analysis was performed using Stata/SE version 13 statistical software.

Results

A total of 22 studies were eligible, namely 14 cohort studies (10,219 incident brain/CNS tumor cases, 1,319 meningioma and 2,418 glioma cases in a total cohort size of 10,143,803 subjects) and eight case-control studies (1,009 brain/CNS cases, 1,977 meningioma cases, 1,265 glioma cases and 8,316 controls). In females, overweight status/obesity was associated with increased risk for overall brain/CNS tumors (pooled RR = 1.12, 95%CI: 1.03–1.21, 10 study arms), meningiomas (pooled RR = 1.27, 95%CI: 1.13–1.43, 16 study arms) and gliomas (pooled RR = 1.17, 95%CI: 1.03–1.32, six arms). Obese (BMI>30 kg/m2) females seemed particularly aggravated in terms of brain/CNS tumor (pooled RR = 1.19, 95%CI: 1.05–1.36, six study arms) and meningioma risk (pooled RR = 1.48, 95%CI: 1.28–1.71, seven arms). In males, overweight/obesity status correlated with increased meningioma risk (pooled RR = 1.58, 95%CI: 1.22–2.04, nine study arms), whereas the respective association with overall brain/CNS tumor or glioma risk was not statistically significant. Dose-response meta-regression analysis further validated the findings.

Conclusion

Our findings highlight obesity as a risk factor for overall brain/CNS tumors, meningiomas and gliomas among females, as well as for meningiomas among males.

Introduction

Glioma and meningioma are the two most common primary central nervous system (CNS) tumors, representing 70% and 20% of brain tumors, respectively [1, 2]. Gliomas originate from glial cells, are as a rule histologically malignant and are more frequent among males [1]. On the other hand, meningiomas originate from the arachnoidal cells of the leptomeninges, are typically histologically benign and are two-fold more frequent among females [3]. The risk factors for brain tumors are poorly understood [4], but may include genetic conditions and ionizing radiation [5]; occupational exposures seem also to be meaningful, as glioma has been linked with occupational exposure to arsenic, mercury and petroleum products, whereas meningioma has been associated with lead exposure [6]. The Million Women Study highlighted attained height as a risk factor for the incidence of all central nervous system tumors with an excess risk of about 20% per 10 cm increase in height [7]. Currently, there is a vivid debate regarding the effects of mobile phone use [8]. On the other hand, a meta-analysis has highlighted atopy as a potential factor inversely associated with glioma but not meningioma [9], whereas another meta-analysis did not find any significant association between smoking and glioma risk [10]. Head injury may entail at most only a small increase in the overall risk of brain tumors[11]. Hormonal and reproductive risk factors appear to be associated with meningioma risk, as positive associations with hormone replacement therapy use [12], uterine fibroids [13] and endometriosis [13] have been reported, culminating at a possibly positive association with breast cancer [14].

Obesity is a well established risk factor for several cancer types [15]; it has been postulated that obesity may account for approximately 20% of all cancer cases [16]. The spectrum of obesity-related cancer may span colon [17], endometrial [18], postmenopausal breast [16], renal [15, 16], esophageal [15], thyroid [15, 16], prostate [19] cancer and hematological malignancies [20], whereas a recent meta-analysis performed by our team highlighted the association between obesity and melanoma among males [21]. Nevertheless, the association between obesity and CNS tumors remains rather obscure, given that CNS tumors are rather uncommon and individual studies have yielded mutually conflicting results. Recently, a meta-analysis focused especially on meningioma, synthesizing data from six studies [22]; the authors concluded that being obese, but not overweight, was associated with increased risk for meningioma, especially among females.

To our knowledge, however, no effort has been undertaken till now, to quantitatively synthesize all published cohort and case-control studies, in order to evaluate the potential association between glioma, or overall brain/CNS tumors and obesity.

In view of the former considerations, our aim was to comprehensively examine the association between obesity and risk for brain/CNS tumors, glioma as well as meningioma in adults, synthesizing all available evidence. Separate analyses were conducted in males and females, in order to evaluate potential sex-specific discrepancies.

Materials and Methods

Search algorithm and eligibility of studies

This meta-analysis was performed in line with the PRISMA guidelines. Eligible studies were identified in PubMed; no restriction regarding publication language was adopted and the end-of-search date was June 30, 2014. The details about the search algorithm are provided in S1 File.

Cohort and case-control studies that examined the association between BMI and risk of overall brain/CNS tumors, glioma or meninigioma were deemed eligible. A systematic search in the reference lists of eligible articles and relevant reviews for potentially eligible articles was performed (“snowball” procedure). In case of overlapping study populations the larger study was ultimately included. Two reviewers (TNS, TP) independently performed the selection of studies; in case of disagreement, the final decision was reached by team consensus.

Data extraction

The data extraction from eligible studies included: first author’s name, publication year, journal where the study was published, type of study (cohort or case-control), study period including the follow-up, geographical region, age of subjects, adjustment factors in multivariable analyses. Specifically for case-control studies, the number, definition and eligibility criteria of cases and controls, as well as the matching factors were extracted. Regarding cohort studies, the cohort size and number of incident cases were abstracted. Two reviewers (IN-S, IGT) independently extracted the data and, in case of disagreement, final decision was reached by team consensus.

Regarding effect estimates, the maximally adjusted Odds Ratios (ORs) for case-control studies as well as the maximally adjusted Relative Risks (RRs) for cohort studies were abstracted, together with their Confidence Intervals (CIs) for each BMI category. Separate abstraction was undertaken for males and females. When the aforementioned information was not available, the data of 2x2 tables in the articles were used for the calculation of crude effect estimates and 95% CIs.

Letters were sent to the authors of studies which did not separately report relative risks pertaining to overweight and obese subjects, as well as to articles reporting exclusively on overall brain/CNS cancer, requesting the subgroup analyses on gliomas and meningiomas. The corresponding authors were contacted twice (a reminder e-mail was sent one week after the first e-mail).

Statistical analyses: meta-analysis

Being overweight was defined as BMI 25–30 kg/m2 and obesity was defined as BMI >30 kg/m2. In this analysis, the term “study arms” refers to the various BMI categories presented in the included studies. The overweight and/or obese “study arms” were pooled, as appropriate. Apart from the overall pooling of overweight/obese subjects (vs. normal weight), subgroup analyses were performed for overweight and obese subjects. Separate models were constructed regarding overall brain/CNS tumors, meningiomas and gliomas, stratified by gender (females; males; study arms not distinguishing between the two sexes). The pooled effect estimate was calculated on the basis of random effects models (DerSimonian-Laird, with the estimate of heterogeneity being taken from the Mantel-Haenszel model). Cochran Q statistic and I2 were used for the assessment of between-study heterogeneity [23]; nevertheless, in view of the limitations of I2 as an index [24, 25] the values of τ2 (tau^2) were also provided.

Statistical analyses: dose-response meta-regression analysis

Dose-response meta-regression analysis serves the investigation of the relationship between different BMI levels and the outcomes. A specific BMI value was allocated to each study arm according to the algorithm described by our previous meta-analysis [21]. Specifically, in case of category a to b the arithmetic mean was adopted. Regarding the upper, open ended categories (≥a), two alternative approaches were undertaken [21], namely: (i) the lower bound was multiplied by 1.2 (approach according to Berlin et al. [26]) and (ii) the formula an+ (an—an-1) was implemented (approach according to Il’yasova et al. [27]), where an stands for the open-ended category threshold (>an) and an-1 pertains to the immediately lower category (an-1 to an). During pooling of study arms, the values according to the approach by Berlin et al. [26] were preferred, because this algorithm allocates values of upper bounds to a greater number of study arms versus that by Il’yasova et al. [27]. As described in our previous meta-analysis [21], the latter algorithm cannot allocate values when the n-1 category is open, as is the case in studies adopting a binary categorization.

At the meta-regression analysis, the log of the study effect estimates was set as the dependent variable in a general linear model. BMI was analyzed in increments of 5 kg/m2 [21]; we present the exponentiated slope coefficient from the linear regression. Both alternative analyses (approaches by Berlin et al. [26] and Il’yasova et al. [27]) were presented. All analyses were performed using Stata/SE version 13 (Stata Corp, College Station, TX, USA).

Risk of bias

The Newcastle-Ottawa Quality scale [28] was used for the evaluation of quality regarding the included studies. With respect to longitudinal cohort studies, the cut-off value was a priori set at 5 years regarding the desirable length of follow-up, whereas the cut-off value for completeness of follow-up was set at 85%. The studies were rated by two independent reviewers (IN-S and IGT); in case of disagreements, final decision was reached by team consensus.

Concerning the assessment of publication bias, the overall (overweight and obese pooled together) pooling analyses in males and females were chosen, in order to maximize the power of the relevant tests [23]. The Egger’s [29] formal statistical test was adopted; for the interpretation of Egger’s test, statistical significance was defined as p<0.1.

Results

Selection and description of eligible studies

Fig 1 presents the flow chart describing the successive steps during the selection of eligible studies. A total of 1,282 abstracts were identified and screened; all details regarding the selection of eligible studies are presented in S1 File.

thumbnail
Fig 1. Flow chart describing the successive steps during the selection of eligible studies.

http://dx.doi.org/10.1371/journal.pone.0136974.g001

Taken as a whole, 14 cohort studies [7, 3042] including 10,219 incident brain/CNS tumor cases [7, 30, 31, 3740, 42], 1,319 incident meningioma cases [7, 31, 3335, 41], 2,418 incident glioma cases [7, 31, 32, 35, 36, 41], in a total cohort size of 10,143,803 subjects (Table 1) and eight case-control studies [4350] including 1,009 brain/CNS cases, 1,977 meningioma cases, 1,265 glioma cases and 8,316 controls (Table 2) were included in the present meta-analysis. The dataset of the study is available in “S1 Dataset” and the PRISMA Checklist in “S1 Checklist”.

Meta-analysis: brain/CNS tumors

The middle panels of Table 3 present the results of the meta-analyses regarding the association between being overweight/obese and risk of overall brain/CNS tumors. Overweight status/obesity was associated with increased risk for brain/CNS tumors among females (pooled RR = 1.12, 95%CI: 1.03–1.21, 10 study arms, I2 = 12.8%, Fig 2), with the association being evident in the subset of cohort studies (pooled RR = 1.13, 95%CI: 1.03–1.24, eight study arms, I2 = 24.4%). The positive association was mainly due to the obese female subjects, as increased brain/CNS tumor risk was observed among them (pooled RR = 1.19, 95%CI: 1.05–1.36, six study arms, I2 = 17.8%, Fig A in S1 File).

thumbnail
Fig 2. Forest plot describing the association between overweight status/obesity and brain/CNS tumor risk among females.

Apart from the overall analysis, the subanalyses on cohort (upper panels) and case-control (lower panels) studies are presented.

http://dx.doi.org/10.1371/journal.pone.0136974.g002

thumbnail
Table 3. Results of the meta-analyses examining the association between obesity and risk brain/CNS tumors, meningiomas and gliomas.

Bold cells denote statistically significant associations.

http://dx.doi.org/10.1371/journal.pone.0136974.t003

On the other hand, no association between overweight status/obesity and brain/CNS tumor risk was identified in males (pooled RR = 1.01, 95%CI: 0.94–1.08, 14 study arms, I2 = 12.1%, Fig B in S1 File). Subgroup analyses by overweight or obesity status replicated the null association (Fig C in S1 File).

In the instances where studies did not distinguish their findings by gender (referred to as “study arms which did not distinguish sexes”), the pooled analysis did not yield any statistically significant association regarding brain/CNS tumor risk (pooled RR = 1.09, 95%CI: 0.97–1.23, seven study arms, I2 = 43.9%, Figs D and E in S1 File). Most probably, this was due to the admixture of the two genders, with their differing profiles.

A post hoc sensitivity analysis including studies reporting exclusively on brain tumor risk (therefore excluding the study arms with collective reporting on “brain/CNS” as a whole) was hampered by the smaller number of eligible study arms; no significant associations were noted therein (Fig F in S1 File).

Meta-analysis: meningiomas

The middle panels of Table 3 present the results of the meta-analyses regarding the association between overweight status/obesity and risk of meningioma. Overweight status/obesity was associated with increased risk for meningioma among females (pooled RR = 1.27, 95%CI: 1.13–1.43, 16 study arms, I2 = 35.7%, Fig 3), with the association replicated both in the subset of cohort (pooled RR = 1.39, 95%CI: 1.17–1.64, 10 study arms, I2 = 39.9%) as well as case-control studies. Once again, the positive association could be mainly attributed to the obese female subjects, as increased meningioma risk was noted among them (pooled RR = 1.48, 95%CI: 1.28–1.71, seven study arms, I2 = 0.0%, Fig G in S1 File); once again the correlation was reproducible upon the subgroups of cohort as well as case-control studies. On the other hand, the overweight females presented only with a marginal trend towards increased risk for meningioma (pooled RR = 1.11, 95%CI: 0.99–1.25, p = 0.085, nine study arms, I2 = 8.3%, Fig G in S1 File).

thumbnail
Fig 3. Forest plot describing the association between overweight status /obesity and meningioma risk among females,.

Apart from the overall analysis, the subanalyses on cohort (upper panels) and case-control (lower panels) studies are presented.

http://dx.doi.org/10.1371/journal.pone.0136974.g003

The results among males were compatible with those among females. Overweight status/obesity was associated with increased risk for meningioma among males (pooled RR = 1.58, 95%CI: 1.22–2.04, nine study arms, I2 = 27.2%, Fig 4); however, the positive association appeared to be confined to the case-control studies. The positive association was mainly attributable to the subgroup of obese males (pooled RR = 1.78, 95%CI: 1.22–2.61, four study arms, I2 = 29.6%, Fig H in S1 File).

thumbnail
Fig 4. Forest plot describing the association between overweight status /obesity and meningioma risk among males.

Apart from the overall analysis, the subanalyses on cohort (upper panels) and case-control (lower panels) studies are presented.

http://dx.doi.org/10.1371/journal.pone.0136974.g004

As expected, the positive association between overweight status/obesity and meningioma risk was also noted during the pooling of the six study arms that did not distinguish between the two sexes (pooled RR = 1.31, 95%CI: 1.12–1.53, I2 = 0.0%, Fig I in S1 File).

Meta-analysis: gliomas

The lower panels of Table 3 present the results of the meta-analyses regarding the association between overweight status/obesity and glioma risk. Being overweight or obese correlated with increased risk for glioma among females (pooled RR = 1.17, 95%CI: 1.03–1.32, six study arms, I2 = 0.0%, Fig 5), with the association spanning both the subsets of cohort (pooled RR = 1.14, 95%CI: 1.00–1.29, four study arms, I2 = 0.0%) as well as case-control studies. Statistical significance was achieved regarding overweight females (pooled RR = 1.19, 95%CI: 1.02–1.38, three study arms, I2 = 0.0%, Fig J-i in S1 File), whereas the association among obese females did not reach significance (pooled RR = 1.13, 95%CI: 0.92–1.38, I2 = 0.0%, Fig J-ii in S1 File); of note though only three study arms reported on the latter subgroup, denoting the scarcity of data.

thumbnail
Fig 5. Forest plot describing the association between overweight status/obesity and glioma risk among females.

Apart from the overall analysis, the subanalyses on cohort (upper panels) and case-control (lower panels) studies are presented.

http://dx.doi.org/10.1371/journal.pone.0136974.g005

Among males, there was no association between being overweight/obese and glioma risk (pooled RR = 0.96, 95%CI: 0.76–1.23, six study arms, I2 = 22.4%, Fig K in S1 File); the null pattern was reproduced at the subgroup analyses on cohort studies, case-control studies, overweight and obese subjects (Fig L in S1 File). The pooling of the 14 study arms that did not distinguish between the two sexes pointed to a null association (pooled RR = 1.03, 95%CI: 0.94–1.14, I2 = 0.0%, Fig M in S1 File); the null pattern was replicated at the subgroup analyses on cohort studies, case-control studies, overweight and obese subjects (Fig N in S1 File).

Dose-response meta-regression analysis

The results of the dose-response meta-regression analysis are presented in Table A in S1 File. The positive association between overweight status/obesity and risk for meningioma among females was reflected upon the findings of the meta-regression analysis (exponentiated coefficient = 1.15, 95%CI: 1.03–1.28, BMI in increments of 5 kg/m2, p = 0.018 following the approach by Berlin et al. [26], Fig O in S1 File), with similar results according to the approach by Il’yasova et al. [27].

On the other hand, the remaining sets of meta-regression analyses did not yield any significant associations; of note, however, it should be noted that the majority of these meta-regression analyses should be deemed explorative, as 10 study arms represent a minimum requirement for satisfactory power according to the Cochrane Handbook [23].

Evaluation of quality of studies and risk of bias

The evaluation of quality of studies is presented in Table B in S1 File for cohort studies and in Table C in S1 File for case-control studies. The quality of cohort studies was mainly compromised by the fact that the ascertainment of BMI was based on questionnaires rather than actual measurements. Regarding the case-control studies, their quality was often compromised by the lack of comparability of non-response rates among cases and controls, as well as the definition of controls. Regarding the overall brain/CNS analysis (Fig P in S1 File), publication bias was not significant at the overall (overweight and obese, pooled together) analysis on females (p = 0.429), males (p = 0.238), or study arms not distinguishing between the two sexes (p = 0.971). Similarly, lack of publication bias was noted regarding meningioma (p = 0.119; p = 0.761; p = 0.696, respectively, Fig Q in S1 File), whereas concerning glioma publication bias was noted only in the analysis pertaining to study arms not distinguishing between the two sexes (p = 0.421; p = 0.120; p = 0.030, respectively, Fig R in S1 File).

Discussion

This meta-analysis highlights obesity as a risk factor for overall brain/CNS, meningiomas and gliomas in females. More specifically, being overweight or obese conferred a substantial increase in risk for all three conditions (12%, 27% and 17%, respectively). Moreover, overweight status or obesity also emerged as a risk factor for meningioma in males, conferring an excess risk of 58%.

From a methodological point of view, dose-response meta-regression analysis was often hampered by the small number of eligible study arms; nevertheless, it was capable of further validating the association between being overweight/obese and increased risk of meningioma among females. Indeed, the dose-response pattern may reinforce the causal nature of the association, according to the Bradford-Hill criteria [51]. Apart from the meta-regression analysis, the dose-response pattern could also be indirectly observed, given that obese males and females seemed sizably more burdened than overweight ones in terms of meningioma risk. Further commenting on methodological issues, this meta-analysis clearly underlines the need for separate reporting by gender in the forthcoming studies, as the study arms that did not distinguish between the two sexes seemed to merely reflect the effects of subjects’ admixture.

Obesity may act through a variety of cancer-promoting pathways, such as chronic insulin resistance, hyperinsulinemia and increased activity of insulin-like growth factor (IGF)-I [52]. Importantly, more than half of meningiomas express the IGF-1 receptor, with IGF-I inducing the growth of meningioma cells in culture; this may provide a direct mechanism of potential tumor-promoting action in both genders [53]. Similarly, IGF-I receptors have been identified in glioma cells, possessing mitogenic [54] and anti-apoptotic [55] actions; this physiological background, shared by meningiomas and gliomas, may provide evidence for the biological plausibility of the associations observed by this meta-analysis.

The underlying mechanisms regarding the sex-specific pattern regarding gliomas and overall brain/CNS tumors remain unclear, but may encompass sex hormone-related effects. The well established, positive correlation between body fat mass and the levels of estrogen [56], highlighting the role of adipose tissue as a highly active endocrine organ, seems of special importance in this context. Being overweight/obese was consistently associated with meningioma risk in females, among both subgroups (cohort and case-control studies), whereas the association pertaining to males seemed mainly due to case-control studies. Sex steroid hormone receptors are indeed present in the majority of meningiomas, namely 80 to 90% of meningiomas express the progesterone receptor and 40% express the estrogen receptor [57]; this fact is consistent with the higher incidence of meningiomas among women compared to men [4, 58], as well as the progression of meningiomas during pregnancy [59].

Limitations of our meta-analysis essentially reflect the shortcomings of individual studies. Unadjusted effect estimates have been occasionally calculated, suffering essentially from confounding, whereas gender-specific effect estimates were not always provided, as discussed earlier. Furthermore, the number of eligible study arms regarding males was substantially smaller than that regarding females, potentially limiting the power of the analyses. Moreover, heterogeneity is very often present but may remain undetected, especially for small meta-analyses [60], as was occasionally the case in the present effort. Regarding the external generalizability of the findings documented herein, it seems worth mentioning that only two studies on Chinese/Asian populations were eligible [37, 38] and they reported on brain/CNS tumors, a fact which underlines the need for relevant studies especially regarding histotypes (meningioma/glioma), so as to shed light into race-specific effects, if any. In cohort studies, the ascertainment of BMI was based on questionnaires rather than actual measurements, a fact which may have introduced reporting bias; case-control studies often suffered by the differential non-response rates among cases and controls and are inherently prone to reverse causation.

It would also be tempting to envisage future studies assessing more elaborate indices of obesity, such as the waist to hip ratio as a marker of central, obesity, or measurements of subcutaneous fat, and whole body fat proportion in an attempt to provide further insight for the underlying pathophysiological links. Future studies may also broaden the perspective of this meta-analysis, examining the associations of obesity with schwannomas and other CNS tumor types, regarding which there is currently limited data. Individual patient data (IPD) pooled analyses seem also particularly desirable, as they might effectively investigate mediating and moderating effects regarding both patient and study characteristics, whereas they might effectively account for missing data as well as minimize or more adequately explain heterogeneity.

Despite the limitations, this meta-analysis bears certain strengths that can also be acknowledged. The lack of publication bias might be one of them; nevertheless, tests for publication bias are rather underpowered, especially in the context of less than 10 studies [61]. Other assets include the broad search algorithm, the meticulous “snowball” procedure for the maximization of eligible articles and synthesized information, as well as the rigorous contact [62] with the authors who subsequently provided us with supplementary data. Indeed, for instance regarding meningioma, our meta-analysis collectively synthesized 16 study arms, whereas the most recent meta-analysis on the field focused only on six studies [22], underlying the comprehensiveness of our approach, which may well have supplied us with substantial statistical power to detect the associations, contrary to the previous work. Another strength of this meta-analysis pertains to the fact that the synthesized studies are relatively recent, published between 2001 and 2014. From a public health perspective, the association between obesity and risk of the CNS tumors seems extremely meaningful, as the modifiable nature of obesity [63] may provide opportunities for prevention regarding these rare, yet considered unpreventable, tumor types.

In conclusion, this meta-analysis is the first to highlight obesity as a risk factor for overall brain/CNS tumors, meningiomas and gliomas among females, as well as for meningiomas among males. Further accumulation of data seems mandatory regarding more elaborate obesity indices. These results open novel perspectives in the prevention of CNS tumors and warrant further investigation regarding the mechanisms underlying the pathophysiological links and potential sex-specific effects.

Supporting Information

S1 Checklist. PRISMA Checklist.

doi:10.1371/journal.pone.0136974.s001

(DOC)

S1 Dataset. Dataset underlying the study findings.

doi:10.1371/journal.pone.0136974.s002

(XLSX)

S1 File. Supporting Information Methods, Results, Tables and Figures.

doi:10.1371/journal.pone.0136974.s003

(DOCX)

Acknowledgments

The authors would like to thank the authors of the individual studies for their replies as well as additional data, which are detailed in the S1 File.

Author Contributions

Conceived and designed the experiments: TNS TP. Performed the experiments: TNS GT CP IN-S I-GT INS TP. Analyzed the data: TNS GT CP IN-S I-GT INS TP. Contributed reagents/materials/analysis tools: TNS GT CP IN-S I-GT INS TP. Wrote the paper: TNS GT CP IN-S I-GT INS TP.

References

  1. 1. Ohgaki H and Kleihues P (2005) Epidemiology and etiology of gliomas. Acta Neuropathol 109: 93–108. pmid:15685439
  2. 2. Longstreth WT Jr., Dennis LK, McGuire VM, Drangsholt MT and Koepsell TD (1993) Epidemiology of intracranial meningioma. Cancer 72: 639–648. pmid:8334619
  3. 3. Sanson M and Cornu P (2000) Biology of meningiomas. Acta Neurochir (Wien) 142: 493–505.
  4. 4. Adami HO, Hunter D and Trichopoulos D (2008) Textbook of Cancer Epidemiology. NY: Oxford University Press.
  5. 5. Umansky F, Shoshan Y, Rosenthal G, Fraifeld S and Spektor S (2008) Radiation-induced meningioma. Neurosurg Focus 24: E7.
  6. 6. Navas-Acien A, Pollan M, Gustavsson P and Plato N (2002) Occupation, exposure to chemicals and risk of gliomas and meningiomas in Sweden. Am J Ind Med 42: 214–227. pmid:12210690
  7. 7. Benson VS, Pirie K, Green J, Casabonne D and Beral V (2008) Lifestyle factors and primary glioma and meningioma tumours in the Million Women Study cohort. Br J Cancer 99: 185–190. doi: 10.1038/sj.bjc.6604445. pmid:18560401
  8. 8. Little MP, Rajaraman P, Curtis RE, Devesa SS, Inskip PD, Check DP, et al. (2012) Mobile phone use and glioma risk: comparison of epidemiological study results with incidence trends in the United States. BMJ 344: e1147. doi: 10.1136/bmj.e1147. pmid:22403263
  9. 9. Linos E, Raine T, Alonso A and Michaud D (2007) Atopy and risk of brain tumors: a meta-analysis. J Natl Cancer Inst 99: 1544–1550. pmid:17925535
  10. 10. Mandelzweig L, Novikov I and Sadetzki S (2009) Smoking and risk of glioma: a meta-analysis. Cancer Causes Control 20: 1927–1938. doi: 10.1007/s10552-009-9386-z. pmid:19568697
  11. 11. Inskip PD, Mellemkjaer L, Gridley G and Olsen JH (1998) Incidence of intracranial tumors following hospitalization for head injuries (Denmark). Cancer Causes Control 9: 109–116. pmid:9486470
  12. 12. Cowppli-Bony A, Bouvier G, Rue M, Loiseau H, Vital A, Lebailly P, et al. (2011) Brain tumors and hormonal factors: review of the epidemiological literature. Cancer Causes Control 22: 697–714. doi: 10.1007/s10552-011-9742-7. pmid:21359526
  13. 13. Claus EB, Calvocoressi L, Bondy ML, Schildkraut JM, Wiemels JL and Wrensch M (2011) Family and personal medical history and risk of meningioma. J Neurosurg 115: 1072–1077. doi: 10.3171/2011.6.JNS11129. pmid:21780859
  14. 14. Custer BS, Koepsell TD and Mueller BA (2002) The association between breast carcinoma and meningioma in women. Cancer 94: 1626–1635. pmid:11920521
  15. 15. Renehan AG, Tyson M, Egger M, Heller RF and Zwahlen M (2008) Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet 371: 569–578. doi: 10.1016/S0140-6736(08)60269-X. pmid:18280327
  16. 16. Wolin KY, Carson K and Colditz GA (2010) Obesity and cancer. Oncologist 15: 556–565. doi: 10.1634/theoncologist.2009-0285. pmid:20507889
  17. 17. Ning Y, Wang L and Giovannucci EL (2010) A quantitative analysis of body mass index and colorectal cancer: findings from 56 observational studies. Obes Rev 11: 19–30. doi: 10.1111/j.1467-789X.2009.00613.x. pmid:19538439
  18. 18. Crosbie EJ, Zwahlen M, Kitchener HC, Egger M and Renehan AG (2010) Body mass index, hormone replacement therapy, and endometrial cancer risk: a meta-analysis. Cancer Epidemiol Biomarkers Prev 19: 3119–3130. doi: 10.1158/1055-9965.EPI-10-0832. pmid:21030602
  19. 19. Discacciati A, Orsini N and Wolk A (2012) Body mass index and incidence of localized and advanced prostate cancer—a dose-response meta-analysis of prospective studies. Ann Oncol.
  20. 20. Lichtman MA (2010) Obesity and the risk for a hematological malignancy: leukemia, lymphoma, or myeloma. Oncologist 15: 1083–1101. doi: 10.1634/theoncologist.2010-0206. pmid:20930095
  21. 21. Sergentanis TN, Antoniadis AG, Gogas HJ, Antonopoulos CN, Adami HO, Ekbom A, et al. (2013) Obesity and risk of malignant melanoma: a meta-analysis of cohort and case-control studies. Eur J Cancer 49: 642–657. doi: 10.1016/j.ejca.2012.08.028. pmid:23200191
  22. 22. Shao C, Bai LP, Qi ZY, Hui GZ and Wang Z (2014) Overweight, obesity and meningioma risk: a meta-analysis. PLoS One 9: e90167. doi: 10.1371/journal.pone.0090167. pmid:24587258
  23. 23. Higgins JPT and Green S (2011 Available from www.cochrane-handbook.org.) Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration.
  24. 24. Hardy RJ and Thompson SG (1998) Detecting and describing heterogeneity in meta-analysis. Stat Med 17: 841–856. pmid:9595615
  25. 25. Mittlbock M and Heinzl H (2006) A simulation study comparing properties of heterogeneity measures in meta-analyses. Stat Med 25: 4321–4333. pmid:16991104
  26. 26. Berlin JA, Longnecker MP and Greenland S (1993) Meta-analysis of epidemiologic dose-response data. Epidemiology 4: 218–228. pmid:8512986
  27. 27. Il'yasova D, Hertz-Picciotto I, Peters U, Berlin JA and Poole C (2005) Choice of exposure scores for categorical regression in meta-analysis: a case study of a common problem. Cancer Causes Control 16: 383–388. pmid:15953980
  28. 28. Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, et al. (2011) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed March 01, 2015
  29. 29. Egger M, Davey Smith G, Schneider M and Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629–634. pmid:9310563
  30. 30. Calle EE, Rodriguez C, Walker-Thurmond K and Thun MJ (2003) Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med 348: 1625–1638. pmid:12711737
  31. 31. Edlinger M, Strohmaier S, Jonsson H, Bjorge T, Manjer J, Borena WT, et al. (2012) Blood pressure and other metabolic syndrome factors and risk of brain tumour in the large population-based Me-Can cohort study. J Hypertens 30: 290–296. doi: 10.1097/HJH.0b013e32834e9176. pmid:22179083
  32. 32. Holick CN, Giovannucci EL, Rosner B, Stampfer MJ and Michaud DS (2007) Prospective study of intake of fruit, vegetables, and carotenoids and the risk of adult glioma. Am J Clin Nutr 85: 877–886. pmid:17344512
  33. 33. Jhawar BS, Fuchs CS, Colditz GA and Stampfer MJ (2003) Sex steroid hormone exposures and risk for meningioma. J Neurosurg 99: 848–853. pmid:14609164
  34. 34. Johnson DR, Olson JE, Vierkant RA, Hammack JE, Wang AH, Folsom AR, et al. (2011) Risk factors for meningioma in postmenopausal women: results from the Iowa Women's Health Study. Neuro Oncol 13: 1011–1019. doi: 10.1093/neuonc/nor081. pmid:21750006
  35. 35. Michaud DS, Bove G, Gallo V, Schlehofer B, Tjonneland A, Olsen A, et al. (2011) Anthropometric measures, physical activity, and risk of glioma and meningioma in a large prospective cohort study. Cancer Prev Res (Phila) 4: 1385–1392.
  36. 36. Moore SC, Rajaraman P, Dubrow R, Darefsky AS, Koebnick C, Hollenbeck A, et al. (2009) Height, body mass index, and physical activity in relation to glioma risk. Cancer Res 69: 8349–8355. doi: 10.1158/0008-5472.CAN-09-1669. pmid:19808953
  37. 37. Oh SW, Yoon YS and Shin SA (2005) Effects of excess weight on cancer incidences depending on cancer sites and histologic findings among men: Korea National Health Insurance Corporation Study. J Clin Oncol 23: 4742–4754. pmid:16034050
  38. 38. Parr CL, Batty GD, Lam TH, Barzi F, Fang X, Ho SC, et al. (2010) Body-mass index and cancer mortality in the Asia-Pacific Cohort Studies Collaboration: pooled analyses of 424,519 participants. Lancet Oncol 11: 741–752. doi: 10.1016/S1470-2045(10)70141-8. pmid:20594911
  39. 39. Samanic C, Gridley G, Chow WH, Lubin J, Hoover RN and Fraumeni JF Jr. (2004) Obesity and cancer risk among white and black United States veterans. Cancer Causes Control 15: 35–43. pmid:14970733
  40. 40. Samanic C, Chow WH, Gridley G, Jarvholm B and Fraumeni JF Jr. (2006) Relation of body mass index to cancer risk in 362,552 Swedish men. Cancer Causes Control 17: 901–909. pmid:16841257
  41. 41. Wiedmann M, Brunborg C, Lindemann K, Johannesen TB, Vatten L, Helseth E, et al. (2013) Body mass index and the risk of meningioma, glioma and schwannoma in a large prospective cohort study (The HUNT Study). Br J Cancer 109: 289–294. doi: 10.1038/bjc.2013.304. pmid:23778522
  42. 42. Wolk A, Gridley G, Svensson M, Nyren O, McLaughlin JK, Fraumeni JF, et al. (2001) A prospective study of obesity and cancer risk (Sweden). Cancer Causes Control 12: 13–21. pmid:11227921
  43. 43. Aghi MK, Eskandar EN, Carter BS, Curry WT Jr. and Barker FG 2nd (2007) Increased prevalence of obesity and obesity-related postoperative complications in male patients with meningiomas. Neurosurgery 61: 754–760; discussion 760–751. pmid:17986936
  44. 44. Claus EB, Calvocoressi L, Bondy ML, Wrensch M, Wiemels JL and Schildkraut JM (2013) Exogenous hormone use, reproductive factors, and risk of intracranial meningioma in females. J Neurosurg 118: 649–656. doi: 10.3171/2012.9.JNS12811. pmid:23101448
  45. 45. Custer B, Longstreth WT Jr., Phillips LE, Koepsell TD and Van Belle G (2006) Hormonal exposures and the risk of intracranial meningioma in women: a population-based case-control study. BMC Cancer 6: 152. pmid:16759391
  46. 46. Lee E, Grutsch J, Persky V, Glick R, Mendes J and Davis F (2006) Association of meningioma with reproductive factors. Int J Cancer 119: 1152–1157. pmid:16570277
  47. 47. Little RB, Madden MH, Thompson RC, Olson JJ, Larocca RV, Pan E, et al. (2013) Anthropometric factors in relation to risk of glioma. Cancer Causes Control 24: 1025–1031. doi: 10.1007/s10552-013-0178-0. pmid:23456313
  48. 48. Cabaniols C, Giorgi R, Chinot O, Ferahta N, Spinelli V, Alla P, et al. (2011) Links between private habits, psychological stress and brain cancer: a case-control pilot study in France. J Neurooncol 103: 307–316. doi: 10.1007/s11060-010-0388-1. pmid:20835749
  49. 49. Pan SY, Johnson KC, Ugnat AM, Wen SW and Mao Y (2004) Association of obesity and cancer risk in Canada. Am J Epidemiol 159: 259–268. pmid:14742286
  50. 50. Schildkraut JM, Calvocoressi L, Wang F, Wrensch M, Bondy ML, Wiemels JL, et al. (2014) Endogenous and exogenous hormone exposure and the risk of meningioma in men. J Neurosurg 120: 820–826. doi: 10.3171/2013.12.JNS131170. pmid:24484233
  51. 51. Hill AB (1965) The Environment and Disease: Association or Causation? Proc R Soc Med 58: 295–300. pmid:14283879
  52. 52. Renehan AG, Frystyk J and Flyvbjerg A (2006) Obesity and cancer risk: the role of the insulin-IGF axis. Trends Endocrinol Metab 17: 328–336. pmid:16956771
  53. 53. Kurihara M, Tokunaga Y, Tsutsumi K, Kawaguchi T, Shigematsu K, Niwa M, et al. (1989) Characterization of insulin-like growth factor I and epidermal growth factor receptors in meningioma. J Neurosurg 71: 538–544. pmid:2552046
  54. 54. Friend KE, Khandwala HM, Flyvbjerg A, Hill H, Li J and McCutcheon IE (2001) Growth hormone and insulin-like growth factor-I: effects on the growth of glioma cell lines. Growth Horm IGF Res 11: 84–91. pmid:11472074
  55. 55. Yin D, Tamaki N, Parent AD and Zhang JH (2005) Insulin-like growth factor-I decreased etoposide-induced apoptosis in glioma cells by increasing bcl-2 expression and decreasing CPP32 activity. Neurol Res 27: 27–35. pmid:15829155
  56. 56. Hankinson SE, Willett WC, Manson JE, Hunter DJ, Colditz GA, Stampfer MJ, et al. (1995) Alcohol, height, and adiposity in relation to estrogen and prolactin levels in postmenopausal women. J Natl Cancer Inst 87: 1297–1302. pmid:7658481
  57. 57. Omulecka A, Papierz W, Nawrocka-Kunecka A and Lewy-Trenda I (2006) Immunohistochemical expression of progesterone and estrogen receptors in meningiomas. Folia Neuropathol 44: 111–115. pmid:16823693
  58. 58. Wiemels J, Wrensch M and Claus EB (2010) Epidemiology and etiology of meningioma. J Neurooncol 99: 307–314. doi: 10.1007/s11060-010-0386-3. pmid:20821343
  59. 59. Ismail K, Coakham HB and Walters FJ (1998) Intracranial meningioma with progesterone positive receptors presenting in late pregnancy. Eur J Anaesthesiol 15: 106–109. pmid:9522150
  60. 60. Kontopantelis E, Springate DA and Reeves D (2013) A re-analysis of the Cochrane Library data: the dangers of unobserved heterogeneity in meta-analyses. PLoS One 8: e69930. doi: 10.1371/journal.pone.0069930. pmid:23922860
  61. 61. Kicinski M, Springate DA and Kontopantelis E (2015) Publication bias in meta-analyses from the Cochrane Database of Systematic Reviews. Stat Med.
  62. 62. Mullan RJ, Flynn DN, Carlberg B, Tleyjeh IM, Kamath CC, LaBella ML, et al. (2009) Systematic reviewers commonly contact study authors but do so with limited rigor. J Clin Epidemiol 62: 138–142. doi: 10.1016/j.jclinepi.2008.08.002. pmid:19013767
  63. 63. World Cancer Research Fund / American Institute for Cancer Research (2007) Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. Washington, DC: AICR.