Breast cancer risk factors in Iran: a systematic review & meta-analysis.

Objectives Breast cancer is known as one of the deadliest forms of cancer, and it is increasing globally. There are a variety of proven and controversial risk factors for this malignancy. Herein, we aimed to undertake a systematic review and meta-analysis focus on the epidemiology of breast cancer risk factors in Iran. Methods We performed a systematic search via PubMed, Scopus, Web of Science, and Persian databases for identifying studies published on breast cancer risk factors up to March 2019. Meta-analyses were done for risk factors reported in more than one study. We calculated odds ratios (ORs) with corresponding 95% confidence intervals (CIs) using a fixed/random-effects models. Results Thirty-nine studies entered into the meta-analysis. Pooling of ORs showed a significant harmful effect for risk factors including family history (OR: 1.80, 95%CI 1.47-2.12), hormonal replacement therapy (HRT) (OR: 5.48, 95%CI 0.84-1.74), passive smokers (OR: 1.68, 95%CI 1.34-2.03), full-term pregnancy at age 30 (OR: 3.41, 95%CI 1.19-5.63), abortion (OR: 1.84, 95%CI 1.35-2.33), sweets consumption (OR: 1.71, 95%CI 1.32-2.11) and genotype Arg/Arg (crude OR: 1.59, 95%CI 1.07-2.10), whereas a significant protective effect for late menarche (OR: 0.58, 95%CI 0.32-0.83), nulliparity (OR: 0.68, 95%CI 0.39-0.96), 13-24 months of breastfeeding (OR: 0.68, 95%CI 0.46-0.90), daily exercise (OR: 0.59, 95%CI 0.44-0.73) and vegetable consumption (crude OR: 0.28, 95%CI 0.10-0.46). Conclusions This study suggests that factors such as family history, HRT, passive smokers, late full-term pregnancy, abortion, sweets consumption and genotype Arg/Arg might increase risk of breast cancer development, whereas late menarche, nulliparity, 13-24 months breastfeeding, daily exercise and vegetable consumption had an inverse association with breast cancer development.


Search Strategy
We followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline for study design, search protocol, screening and reporting. A systematic search was

Criteria study selection
Two group members (A.SH and K.HD) selected the papers independently and discussed to solve the disagreements. Studies met the following criteria included into meta-analysis: 1) studies were case-control or cohort; 2) studies included the risk factors of breast cancer patients in Iran; and 3) studies were considered the female breast cancer. Studies were excluded if they were: 1) conference abstracts, comments, letters, animal studies, reviews, case reports, ecological studies, cross-sectional studies, and in vitro studies; 2) duplicate publications; 3) insufficient for calculating of desired parameters, and 4) male breast cancer studies.

Data extraction & quality assessment
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 7 April 2020 doi:10.20944/preprints202004.0074.v1 Two researchers (Z.SH and K.HD) have independently evaluated the quality of studies and extracted data from included papers. The supervisor (R.AN) resolved any disagreements in this part. Data extraction checklist included the name of the first author, publication year, a region of study, number of patients, mean age, quantitative information of risk factors, clinicopathological features, and available correlations.
The Newcastle-Ottawa Scale (NOS) checklist was used to value the selected papers in relation to various aspects of the methodology and study process.

Data analysis
Statistical analysis was performed using STATA v.11 software. To assess the heterogeneities, we used the I-square (I 2 ) test. According to the studies heterogeneity, we pooled results using a fixedeffects or random-effects model as appropriate for heterogeneity more or less than 50%, respectively.

Study selection process
Our initial search through databases resulted in 562 papers. After excluding duplicated papers, remained papers screened using title and abstract. Finally, after eligibility assessment of 84 full texts, 41 studies entered into qualitative synthesis and finally, 39 studies entered into the metaanalysis. PRISMA flow diagram for the study selection process presented in Figure 1.

Benign Breast Conditions
Data from five studies were combined for the meta-analysis of benign breast history as a risk factor to develop breast cancer. No significant difference was observed in this regard (crude OR: 1.24, 95%CI 0.82-1.67) (Additional file).

X-ray exposure
This factor was studied in four papers. No significant differences were observed regarding the history of x-rays exposure between cases and controls  (Fig. 7).

Genotype status
Two studies were investigated p53 codon 72 polymorphisms as a breast cancer risk factor.
Although genotype Arg/Pro (crude OR: 0.69, 95%CI 0.37-1.00) was not related to the odds of breast cancer development, a significant higher chance found for genotype Arg/Arg (crude OR: 1.59, 95%CI 1.07-2.10) in this regard (Additional file).

Genital surgery
Genital surgery was considered in two studies. No significant differences were found regarding the history of genital surgery for breast cancer development (crude OR: 1.37, 95%CI 0.87-1.86) (Additional file).   Family history of breast cancer was one of the associated risk factors for breast cancer development in our study. In one of the first meta-analysis on "Family history and the risk of breast cancer",

Residency status
Pharoah et al. [64] pooled estimate of relative risk (RR) indicated that the probability of breast cancer occurrence is higher in those individuals with a family history of this malignancy (RR: 1.9, 95%CI, 1.7-2.0). They also found that this probability is higher in the first-degree relatives, especially mother and sister (RR: 3.6, 95%CI 2.5-5.0). There are many other studies that reported the association of family history with the risk of breast cancer [65][66][67].
High levels of estrogen can increase the chance of breast cancer development through genotoxic stress induction and breast tissue mutations [68,69]. Therefore, receiving external estrogen through HRT may increase the risk of breast cancer development. In this regard, HRT users showed the highest chance of developing breast cancer in our meta-analysis (OR: 5.48, 95%CI 0.84-1.74), in consistent with several studies [70][71][72]. In contrast, study of Bae et al. reported no significant association in this regard among Korean Women [73].
The first systematic review on "Adverse life-events and risk of breast cancer" [74] indicated no significant relationships in this regard (OR: 0.8, 95%CI 0.96-1.06) and two other systematic reviews reported the same conclusion [75,76]. Our results were contrary to these studies, but was in the same line with the study of Lin et [79][80][81].
In regard to the age at first full-term pregnancy, our results were in the same line with previous studies regarding the higher risk of breast cancer at age 30 or older ages [82,83]. For nulliparity condition, our study showed an inverse association with breast cancer development, which was in contrast with several previous reports [84,85].
Numerous investigators have studied the association of induced abortion and risk of breast cancer throughout the world. One of the oldest studies titled "Induced abortion as a cancer risk factor" discussed the induced abortion as a breast cancer risk factor [86]. Similarly, some meta-analysis also reported the same conclusions [87,88]. In this regard, our meta-analysis found that induced abortion was significantly associated with risk of breast cancer in Iranian women. Besides, the  [89,90].
The meta-analysis findings showed the significant association between sweet foods consumption and risk of breast cancer. Although we did not find a specific systematic review in this regard, several epidemiological studies in different regions reported the association of sweet foods consumption and risk of breast cancer [91][92][93][94]. For example, Tavani et al. performed a comprehensive case-control study in Italy and found a direct relationship between sweet foods consumption and risk of breast cancer development [94]. In fact, excessive sweets intake with a high glycemic index may cause insulin resistance as well as insulin-related growth factors as promoters of breast carcinogenesis. Moreover, ovarian steroid secretion including estrogens and androgens might be stimulated by insulin. Altogether, these processes end up in increased risk of breast cancer [95,96].
Previous studies have investigated the relationships of P53 codon 72 polymorphisms and risk of breast cancer development in different regions [97]. In this investigation, we found that the direction with two other meta-analyses carried out by Li et al. [101] and Collaborative Group on Hormonal Factors in Breast Cancer [102].
We found that longer breastfeeding duration (13 to 24 months) plays a protective rule against breast cancer development, in consistent with numerous meta-analysis in various populations [103][104][105][106] Perspective" in 2007, recommended physical activity as a protective factor against cancers, especially postmenopausal breast cancer [108]. It was also recommended by "American Cancer Society Guidelines on nutrition and physical activity for cancer prevention" [109]. Our findings also support this hypothesis as a preventive factor for breast cancer development. In fact, physical activities effect on the risk of cancers development through mechanisms such as metabolic, reproductive effects, hormonal, and immunity enhancement, etc. [110].  [111][112][113]. In contrast, several systematic reviews and meta-analysis showed no significant relationships in this regard [114,115] and some others were controversial regarding the types of vegetables and its combination intake with fruits as well as breast cancer types [116,117].
According to the retrospective nature of the included studies, it is recommended to design some longitudinal cohort investigations in order to examine the accurate role of these risk factors in breast cancer development.

Conclusion:
Based on this systematic review and meta-analysis, factors including a family history of breast cancer, HRT, ER and PR positive status, the experience of stress condition, passive smokers, abortion, sweets consumption and genotype Arg/Arg, play a significant role in the development of breast cancer. In contrast, late menarche, nulliparity, long breastfeeding duration, regular physical activity and consumption of vegetables showed a significant inverse association with breast cancer occurrence.

Ethics approval
It is only remarkable that the protocol of this systematic review and meta-analysis registered in International Prospective Register of Systematic Reviews (PROSPERO) due to code CRD42019127382.