Breast Cancer Res Treat (2014) 145:513–524 DOI 10.1007/s10549-014-2969-8
EPIDEMIOLOGY
Folate intake and the risk of breast cancer: a systematic review and meta-analysis Martin Tio • Juliana Andrici • Guy D. Eslick
Received: 10 April 2014 / Accepted: 12 April 2014 / Published online: 29 April 2014 Ó Springer Science+Business Media New York 2014
Abstract There is conflicting epidemiological evidence on the role of folate and breast cancer risk. We conducted a systematic review and quantitative meta-analysis of folate intake and folate blood levels and the risk of breast cancer. Four electronic databases (Medline, PubMed, Embase, and Current Contents Connect) were searched to April 11, 2014, with no language restrictions for observational studies that measured folate intake or blood levels and the risk of breast cancer. The meta-analysis of dietary folate intake comprising 36 studies with 34,602 cases, and a total sample size of 608,265 showed a decreased risk of breast cancer, with an odds ratio (OR) of 0.84 [95 % confidence interval (CI) 0.77–0.91]. When stratified by menopausal status and by study design, none of the meta-analyses of prospective studies showed any statistically significant decrease in the risk of breast cancer. The meta-analysis of total folate showed no statistically significant association with breast cancer OR of 0.98 (95 % CI 0.91–1.07). There was no significant association between either dietary or total folate intake and breast cancer when stratified by hormonal receptor status. The meta-analysis of blood folate levels found no significant association with the risk of breast cancer, with an OR of 0.86 (95 % CI 0.60–1.25). Breast cancer does not appear to be associated with folate intake, and this did not vary by menopausal status or hormonal receptor status. Folate blood levels also do not appear to be associated with breast cancer risk.
M. Tio J. Andrici G. D. Eslick (&) The Whiteley-Martin Research Centre, The Discipline of Surgery, Sydney Medical School, The University of Sydney, Nepean Hospital, Level 5, South Block, Penrith, NSW, Australia e-mail: [email protected]
Keywords Breast cancer Folate Folic acid Meta-analysis Systematic review
Introduction Breast cancer is both the most common form of cancer in females, and the leading cause of cancer mortality in females [1]. With the possible exception of alcohol [2], nutritional intake has not been well established as a risk factor for breast cancer [3]. Folate has been implicated as one possible nutritional risk factor for cancer in general, as it plays a major role in the one-carbon metabolism of DNA. Evidence supporting folate with carcinogenesis ranges from in vitro evidence indicating that low folate levels contribute to DNA damage and epigenetic modification of DNA methylation [4, 5], all the way up to a meta-analysis of randomized clinical trials suggesting an increased risk of cancer with individuals given supplemental doses of folate [6]. The association between folate and breast cancer specifically though, is much more equivocal. There is conflicting evidence both at the level of basic science investigations, and at the level of human observational studies [7–10]. Two meta-analyses of observational studies in humans have indicated that folate is not significantly associated with the risk of breast cancer [11, 12]. Subsequent to the publication of these meta-analyses, however, has been the further publication of a number of large-scale observational studies on folate and breast cancer. Additionally, these meta-analyses did not include any evaluation of folate and the risk of different hormonal receptor subtypes of breast cancer. As such, we performed a systematic review and metaanalysis of observational studies on folate intake and the
123
514
risk of breast cancer, including analyses of hormone receptor subtypes of breast cancer.
Methods Search strategy We followed the meta-analysis of observational studies in epidemiology (MOOSE) guidelines in performing our systematic review [13]. Relevant articles were identified by two reviewers (M.T. and J.A.) by systematically searching through MEDLINE (from 1950), PubMed (from 1946), EMBASE (from 1949) and Current Contents Connect (from 1998) through to March 2nd, 2014. The search used the terms ‘‘folate’’ or ‘‘folic acid’’ or ‘‘vitamin B9’’ and ‘‘breast cancer’’ or ‘‘breast neoplasms.’’ The search terms used were searched as text word and as exploded medical subject headings where possible. The reference lists of relevant articles were also searched for appropriate studies. No language restrictions were used in either the search or study selection. A search for unpublished literature was not performed. Disagreement on article inclusion between the two reviewers was resolved via a third reviewer (G.E.). Study selection In order to be included, eligible studies needed to (1) have a study design of either a cohort or case control; (2) report the risk of breast cancer in association with total folate intake, dietary folate intake, or folate blood levels. Dietary folate was defined as all folate intake via foods, and the total folate intake was defined as all folate intake via foods and supplements; (3) report the risk point estimate as an odds ratio (OR), hazard ratio, relative risk, or incidence rate ratio that compared a higher level of folate intake to a lower level of folate intake. When multiple levels of folate intake were presented, the ratio comparing the highest intake versus the lowest intake was chosen; (4) report the 95 % confidence interval (CI) for the point estimate; and (5) use an internal comparison when calculating the risk estimate. Studies were excluded if they did not meet the inclusion criteria, if they were duplicates of other studies eligible for inclusion, or if they were meta-analyses of studies. In the case of duplicate studies, the most recently published study was chosen for inclusion.
Breast Cancer Res Treat (2014) 145:513–524
direction, population type, country, continent, economic development, response rate, case control matching, mean age, number of adjusted variables, years of follow up, lowest folate level, highest folate level, difference between highest and lowest folate levels, and the risk estimates and CIs on breast cancer. Authors were not contacted if data were missing. Adjusted ratios were extracted in preference to non-adjusted ratios. When more than one adjusted ratio was reported, the ratio with the highest number of adjusted variables was chosen. Statistical analysis Pooled estimates of the OR and 95 % CI for the risk of breast cancer in association with folate were calculated using the random-effects model of DerSimonian and Laird [14]. Heterogeneity was assessed using the I2 statistic, which determines the proportion of variability across studies that is due to heterogeneity as opposed to sampling error [15]. Sensitivity analyses were performed when statistically significantly heterogeneity was detected. Subgroup analyses stratified by either study design or population group were also performed. Meta-regression was performed when heterogeneity remained after sensitivity and subgroup analysis. Our pre-specified meta-regression analyses compared the log odd ratios versus highest absolute folate level, lowest absolute folate level, and difference between highest and lowest absolute folate level. Publication bias was assessed with Egger’s regression model [16]. If publication bias was detected, then the additional publication bias methods consisting of the failsafe number method and the trim-and-fill method were employed to quantify the effect of the bias. The fail-safe number method calculates the number of unpublished studies needed to convert the observed result to statistical non-significance at the alpha level of significance p \ 0.05 level. Publication bias is considered to be an issue if the fail-safe number is less than 5n ? 10, where n is the number of studies included in the meta-analysis [17]. The trim-and-fill method simulates unpublished studies in the meta-analysis to calculate a new pooled OR, which is then compared to the original pooled OR. If the new pooled OR is similar to the original pooled OR, then this indicates that publication bias has little effect on the meta-analysis results. Results were regarded as statistically significant if p \ 0.05. All analyses were done with Comprehensive Meta-analysis (version 2.0), Englewood, NJ, USA (2005).
Data extraction One reviewer (M.T.) performed the data extraction via a standardized data extraction form. Information was extracted on the publication year, study design, number of cases, number of controls, total sample size, temporal
123
Results Our search identified 3,725 citations, from which 49 articles were extracted for inclusion in our systematic review
Breast Cancer Res Treat (2014) 145:513–524
515
Fig. 1 Study selection flowchart
(Fig. 1). Table 1 shows selected characteristics of the identified studies [9, 10, 18–64]. The majority of the included studies originated from either Europe [10, 18, 25, 26, 31, 33, 34, 36, 41, 44, 48, 53] or North America [9, 21, 22, 24, 27, 29, 30, 37, 40, 42, 43, 45–47, 51, 54, 56–58, 60, 62, 64] with the remainder from Asia [23, 28, 35, 39, 49, 50, 52, 55, 59, 63] Central and South America [19, 32, 38], and Oceania [20, 61]. Two studies that met our inclusion criteria could not be examined with meta-analysis due to a lack of other studies to pool them with. Kotsopolous et al. [31] found no significant association with plasma folate levels and the risk of breast cancer in BRCA1 carriers, and Sellers et al. [47] found a significantly decreased risk of breast cancer with high dietary folate intake in individuals with a family history of breast cancer, but no significant association in individuals without a family history of breast cancer. Dietary folate intake Figure 2 shows the meta-analysis of 36 studies (14 prospective studies and 22 retrospective studies) comprising 34,602 cases with a total sample size of 608,265 on dietary
folate intake and risk of breast cancer. High dietary folate intake was associated with a statistically significant decreased risk of breast cancer, with an OR of 0.84 (95 % CI 0.77–0.91). There was statistically significant heterogeneity (I2 = 71.2 %; p \ 0.01), and a sensitivity analysis did not find any single study that contributed significantly to the heterogeneity. There was statistically significant publication bias (Eggers test, p \ 0.01), and the adjusted results using the trim-and-fill method gave a non-statistically significant OR of 0.93 (95 % CI 0.85–1.02). The failsafe N was 364. When stratified by menopausal status, the meta-analysis of postmenopausal breast cancer risk included 20 studies (9 prospective studies and 11 retrospective studies) comprising 15,484 cases with a total sample size of 360,634. There was a statistically significant association between high dietary folate intake and risk of postmenopausal breast cancer, with an OR of 0.84 (95 % CI 0.75–0.94). There was significant heterogeneity (I2 = 62.6 %; p \ 0.01), and sensitivity analysis did not find any single study that contributed significantly to the heterogeneity. There was statistically significant publication bias (Eggers test, p \ 0.01), and the adjusted OR via the trim-and-fill method
123
123
Cohort
Gong et al. [62]
Larsson et al. [34]
Case control
Gao et al. [28]
Cohort
Case control
Freudenheim et al. [27]
Lajous et al. [33]
Case control
Ericson et al. [26]
Case control
Nested case control
Ericson et al. [25]
Lajous et al. [32]
Nested case control
Ericson et al. [10]
Case control
Cohort
Duffy et al. [24]
Kotsopolous et al. [31]
Cohort
Chou et al. [23]
Cohort
Case control
Cho et al. [22]
Kabat et al. [30]
Cohort
Chen et al. [21]
Case control
Case control
Beilby et al. [20]
Case control
Case control
Bassett et al. [61]
Islam et al. [63]
Cohort
Aune et al. [19]
Graham et al. [29]
Case control
Case control
Adzersen et al. [18]
Study type
Source
Sweden
France
Mexico
Poland
Canada
Japan
USA
USA
China
USA
Sweden
Sweden
Sweden
USA
Taiwan
USA
USA
Australia
Australia
Uruguay
Germany
Country
Table 1 Studies included in the systematic review
2,952
1,812
475
48
2,491
1,754
430
1,493
669
297
203
313
392
1,783
146
221
1,426
141
936
461
310
Cases
61,433
62,739
1,794
144
49,654
5,262
917
3,016
1,351
608
404
939
11,619
86,747
431
90,663
2,908
250
20,756
2,493
663
Total size
Population
Population
Hospital cases; population controls
Hospital
Population
Hospital
Population
Mixed cases; population controls
Hospital and population cases; population controls
Hospital cases; population controls
Population
Population
Population
Hospital
Hospital
Population
Population
Hospital cases; population controls
Population
Hospital
Hospital
Population derivation
Adults born 1914–1948
Adults born 1925–1950
Adults aged 20–75
Adults
Adults aged 40–59
Adults aged 20–79
White adults
Adults aged 20–75
Adults
White adults
Adults born 1926–1945
Adults born 1926–1945
Adults born 1926–1945
Postmenopausal adults
Adults aged 20–80
Registered nurses
Adults
Adults aged 30–84
Adults aged 27–80
Dietary folate total folate
Dietary folate
Dietary folate
Plasma folate
Dietary folate
Dietary folate
Dietary folate
Dietary folate
Dietary folate
Dietary folate
Plasma folate
Plasma folate
Dietary folate total folate
Total folate
Total folate plasma folate
Dietary folate total folate
Dietary folate total folate
Serum folate
Dietary folate
Dietary folate Dietary folate
Adults aged 25–75
Folate measurement
Adults aged \90
Population
PR subtype risk
ER subtype risk
Overall risk
Postmenopausal risk
Postmenopausal risk
Premenopausal risk
Overall risk
BRCA1 carrier risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
Postmenopausal risk
ER subtype risk
Overall risk
Overall risk
Premenopausal risk
Postmenopausal ER subtype risk
Postmenopausal risk
Postmenopausal risk
Postmenopausal risk
Overall risk
Premenopausal ER subtype risk
Premenopausal risk
Overall risk
Overall risk
PR subtype risk
ER subtype risk
Overall risk
Overall risk
Overall risk
Reported breast cancer risks
516 Breast Cancer Res Treat (2014) 145:513–524
Nested case control
Cohort
Cohort
Cohort
Cohort
Case control
Case control
Rohan et al. [43]
Roswall et al. [44]
Sellers et al. [45]
Sellers et al. [46]
Sellers et al. [47]
Sharp et al. [48]
Shrubsole et al. [49]
Italy
Cohort
Maruti et al. [40]
Case control
Case control
Ma et al. [39]
Case control
Case control
Ma et al. [38]
Negri et al. [41]
Nested case control
Lin et al. [37]
Potischman et al. [42]
USA
Case control
Levi et al. [36]
China
UK
USA
USA
USA
Denmark
Canada
USA
Japan
Brazil
USA
Switzerland
South Korea
Case control
Lee et al. [35]
Country
Study type
Source
Table 1 continued
1,321
62
1,823
1,416
1,586
1,072
1,336
568
2,569
743
398
458
848
289
323
Cases
2,703
128
33,552
34,393
35,973
26,224
6,718
2,019
5,157
35,023
796
916
1,696
731
646
Total size
Population
Hospital
Population
Population
Population
Population
Population
Population
Hospital
Population
Hospital
Hospital
Population
Hospital
Hospital
Population derivation
Adults aged 25–64
Adults
Adults 55–69
Adults 55–69
Adults 55–69
Adults aged 50–64
Adults aged 40–59
Adults aged 20–44
Adults
Adults aged 50–76
Adults aged 20–74
Adults aged 20–74
Dietary folate
Dietary folate
Dietary folate total folate
Dietary folate total folate
Dietary folate total folate
Dietary folate total folate
Dietary folate
Dietary folate total folate
Dietary folate
Dietary folate total folate
Dietary folate
Dietary folate
Dietary folate total folate plasma folate
Dietary folate
Adults aged \75
Adults aged 45 or older
Dietary folate
Folate measurement
Adults
Population
Postmenopausal risk
Premenopausal risk
Overall risk
Overall risk
Negative family History risk
Positive family History risk
PR subtype risk
ER subtype risk
Postmenopausal risk
PR subtype risk
ER subtype risk
Postmenopausal risk
Postmenopausal risk
Premenopausal risk
Premenopausal risk
Overall risk
ER subtype risk
Postmenopausal risk
PR subtype risk
ER subtype risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
PR subtype risk
ER subtype risk
Postmenopausal risk
Premenopausal risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
Reported breast cancer risks
Breast Cancer Res Treat (2014) 145:513–524 517
123
123
Case control
Case control
Zhu et al. [60]
Case Control
Yang et al. [64]
Zhang et al. [59]
Case control
Yang et al. [55]
Cohort
Nested case control
Wu et al. [54]
Zhang et al. [58]
Nested case control
Wu et al. [54]
Cohort
Case control
Thorand et al. [53]
Nested case control
Case control
Suzuki et al. [52]
Zhang et al. [56]
Cohort
Zhang et al. [57]
USA
Cohort
Stevens et al. [9]
Stolzenberg-Solomen et al. [51]
USA
China
USA
USA
USA
USA
South Korea
USA
USA
Germany
Japan
USA
China
Cohort
Shrubsole et al. [50]
Country
Study type
Source
Table 1 continued
304
438
2,812
712
3,483
2,325
362
110
133
43
456
691
3,898
718
Cases
609
876
88,744
1,424
88,818
4,850
724
220
266
149
1,368
25,400
70,656
73,237
Total size
Population
Hospital
Population
Population
Population
Hospital cases; population controls
Hospital
Population
Population
Population
Hospital
Population
Population
Population
Population derivation
African-American adults
Adults aged 25–70
Adults aged 30–55
Adults aged 30–55
Adults aged 30–55
Adults aged 25–79
Adults aged 30–65
Adults with serum sampled in 1989
Adults with serum sampled in 1974
Adults aged 38–80
Adults aged 20–79
Adults aged 55–74
Adults aged 50–74
Adults aged 40–70
Population
Dietary folate
Dietary Folate
Dietary folate total folate
Plasma folate
Total folate
Total folate
Dietary folate
Serum folate
Serum folate
Dietary folate
Dietary folate total folate
Dietary folate total folate
Dietary folate total folate
Dietary folate
Folate measurement
Overall risk
PR subtype risk
ER subtype risk
Overall risk
PR subtype risk
ER subtype risk
Overall risk
Overall risk
ER subtype risk
Overall risk
Overall risk
Overall risk
Overall risk
Postmenopausal
Postmenopausal Risk
Premenopausal risk
Overall risk
Postmenopausal risk
Postmenopausal risk
PR subtype risk
ER subtype risk
Postmenopausal risk
Premenopausal risk
Overall risk
Reported breast cancer risks
518 Breast Cancer Res Treat (2014) 145:513–524
Breast Cancer Res Treat (2014) 145:513–524 Fig. 2 Dietary folate intake and risk of breast cancer
519
Study name
Statistics for each study
Odds ratio and 95% CI
Odds ratio
Lower limit
Upper limit
p-Value
0.47 1.05 0.99 0.85 1.08 0.56 0.76 0.70 0.91 0.70 0.79 1.02 0.64 0.78 1.01 0.50 0.45 0.84 1.26 1.01 0.91 0.73 0.89 1.72 1.22 0.83 0.49 0.62 0.79 1.12 Stolzenberg-Solomon et al (2006) 0.97 Suzuki et al (2008) 0.65 Thorand et al (1998) 1.14 Yang et al (2010) 1.22 Zhang et al (2011) 0.32 Zhu et al (2003) 0.58 0.84
0.26 0.54 0.83 0.64 0.86 0.34 0.43 0.53 0.67 0.48 0.68 0.89 0.45 0.67 0.90 0.29 0.27 0.52 0.89 0.62 0.68 0.60 0.68 0.97 0.95 0.62 0.20 0.46 0.59 1.01 0.77 0.46 0.73 0.65 0.21 0.25 0.77
0.86 2.03 1.19 1.13 1.35 0.92 1.36 0.92 1.23 1.02 0.92 1.16 0.91 0.90 1.13 0.87 0.74 1.36 1.78 1.64 1.22 0.88 1.17 3.05 1.57 1.11 1.20 0.83 1.06 1.24 1.23 0.91 1.79 2.28 0.49 1.36 0.91
0.01 0.88 0.91 0.27 0.50 0.02 0.35 0.01 0.54 0.06 0.00 0.77 0.01 0.00 0.86 0.01 0.00 0.48 0.19 0.97 0.53 0.00 0.40 0.06 0.12 0.20 0.12 0.00 0.11 0.03 0.80 0.01 0.57 0.53 0.00 0.21 0.00
Adzersen et al (2003) Aune et al (2010) Bassett et al (2013) Chen et al (2005) Cho et al (2007) Ericson et al (2007) Freudenheim et al (1996) Gao et al (2009) Gong et al (2014) Graham et al (1991) Islam et al (2013) Kabat et al (2008) Lajous et al (2006) a Lajous et al (2006) b Larsson et al (2008) Lee et al (2011) Levi et al (2001) Lin et al (2008) Ma et al (2009) a Ma et al (2009) b Maruti et al (2009) Negri et al (2000) Potischman et al (1999) Rohen et al (2000) Roswall et al (2010) Sellers et al (2001) Sharp et al (2002) Shrubsole et al (2001) Shrubsole et al (2011) Stevens et al (2010)
0.1
0.2
0.5
Protective
showed no significant association, with an OR of 0.94 (95 % CI 0.83–1.06). The fail-safe N was 95. Further stratification via prospective study design found no statistically significant association with postmenopausal breast cancer risk, with an OR of 0.93 (95 %CI 0.81–1.07). There was statistically significant heterogeneity (I2 = 68.4 %; p \ 0.01), but this became non-significant on sensitivity analysis with the removal of Stevens et al. [9] (I2 = 48.0 %; p = 0.06). The removal of Stevens et al. [9] did not significantly change the results, with an OR of 0.90 (95 % CI 0.79–1.02). There was no publication bias (Eggers test, p = 0.24). The stratification via retrospective study design found a statistically significant decreased risk of postmenopausal breast cancer risk, with an OR of 0.76 (95 % CI 0.66–0.86). There was no significant heterogeneity (I2 = 20.9 %; p = 0.24). There was no significant publication bias (Eggers test, p = 0.14). The meta-analysis of dietary folate intake and premenopausal breast cancer risk included 13 studies (3 prospective studies and 10 retrospective studies) with 6,376 cases and a total sample size of 181,199. There was a statistically significant association between dietary folate intake and the risk of premenopausal breast cancer, with an OR of 0.81 (95 % CI 0.66–1.00). There was significant heterogeneity (I2 = 69.0 %; p \ 0.01), which was not eliminated after sensitivity analysis. There was no significant publication bias (Eggers test, p = 0.99).
1
2
5
10
Harmful
Stratifying by prospective study design showed no significant association for premenopausal breast cancer risk, with an OR 1.02 (95 % CI 0.62–1.68). There was significant heterogeneity (I2 = 73.8 %; p = 0.02), but sensitivity analysis and meta-regression were not appropriate given that there were only 3 studies. There was no publication bias (Eggers test, p = 0.93). Stratifying by retrospective study design showed a significant association, with an OR of 0.75 (95 % CI 0.61–0.93). There was significant heterogeneity (I2 = 58.8 %; p = 0.01), which became non-significant on the removal of Ma et al. [38] (I2 = 0.0 %; p = 0.52),. This did not significantly change the results, with an OR of 0.70 (95 % CI 0.61–0.79). There was no publication bias (Eggers test, p = 0.67). Total folate intake Figure 3 shows the meta-analysis of total folate intake and breast cancer risk, which included 15 studies (11 prospective and 4 retrospective) comprising 21,001 cases with a total sample size of 521,474. There was no statistically significant association between total folate intake and breast cancer risk, with an OR of 0.98 (95 % CI 0.91–1.07). There was significant heterogeneity (I2 = 52.7 %; p = 0.01) which was not removed on sensitivity analysis. There was no publication bias (Eggers test, p = 0.44). When stratified by menopausal status, the meta-analysis of postmenopausal included 7 prospective studies comprising
123
520 Fig. 3 Total folate intake and risk of breast cancer
Breast Cancer Res Treat (2014) 145:513–524 Study name
Statistics for each study
Odds ratio Chen et al (2005) 0.95 Cho et al (2007) 1.09 Chou et al (2007) 0.51 Duffy et al (2009) 0.97 Ericson et al (2007) 0.56 Larsson et al (2008) 1.06 Lin et al (2008) 1.24 Maruti et al (2009) 0.78 Potischman et al (1999) 1.11 Roswall et al (2010) 1.23 Sellers et al (2001) 0.84 Stevens et al (2010) 1.03 Stolzenberg-Solomon et al (2006) 1.27 Yang et al (2013) 0.90 Zhang et al (1999) 0.93 0.98
Lower limit 0.74 0.88 0.30 0.84 0.35 0.87 0.88 0.61 0.81 0.97 0.63 0.93 1.00 0.64 0.83 0.91
Upper limit 1.22 1.35 0.87 1.12 0.90 1.29 1.75 0.99 1.52 1.56 1.11 1.15 1.62 1.26 1.04 1.07
Odds ratio and 95% CI
p-Value 0.69 0.42 0.01 0.68 0.02 0.55 0.22 0.04 0.52 0.09 0.23 0.59 0.05 0.54 0.19 0.71 0.1 0.2 0.5 Protective
10,165 cases with a total sample size of 293,425. Total folate intake had no statistically significant association with postmenopausal breast cancer, with an OR of 0.97 (95 % CI 0.84–1.12). There was significant heterogeneity (I2 = 67.5 %; p = 0.01) which was not removed on sensitivity analysis. Meta-regression was significant for an association with high total folate intake being more protective as the absolute level of low total folate intake decreased (p \ 0.01), but was not associated with the absolute level of high total folate intake (p = 0.18), or the difference between the two levels (p = 0.48). The meta-analysis of total folate intake and premenopausal breast cancer risk included 2 studies (1 prospective study and 1 retrospective study) comprising 1,600 cases with a total sample size of 92,682. Total folate intake had no statistically significant association with premenopausal breast cancer, with an OR of 1.1 (95 % CI 0.92–1.31). There was no significant heterogeneity (I2 = 0.0 %; p = 0.93), and publication bias could not be assessed due to an insufficient number of studies. Blood folate levels Figure 4 shows the meta-analysis of blood folate levels and breast cancer risk, consisting of 7 studies (5 prospective studies and 2 retrospective studies) comprising 2,403 cases with a total sample size of 5,226. There was no significant association between blood folate levels and the risk of breast cancer, with an OR of 0.86 (95 % CI 0.60–1.25). There was significant heterogeneity (I2 = 70.3 %; p \ 0.01), which became insignificant with the removal of Beilby et al. [20] (I2 = 48.7 %; p = 0.08). The result remained non-significant with Beilby et al. [20] removed, with an OR of 1.01 (95 % CI 0.76–1.33). Hormone receptor subtypes Table 2 shows the results for dietary folate intake, total folate intake, and blood folate levels for breast cancer risk
123
1
2 5 Harmful
10
by hormonal receptor subtype. Sensitivity analysis of the heterogeneous results showed that Zhang et al. [59] contributed significantly to the heterogeneity of the dietary folate and estrogen receptor positive, progesterone receptor positive, and estrogen receptor positive/progesterone receptor positive breast cancer risks. Maruti et al. [40] contributed significantly to the heterogeneity of the total folate and estrogen receptor negative breast cancer risk. Removal of these studies resulted in homogeneity, with no substantial change to the overall risk (data not shown).
Discussion Our systematic review and meta-analysis show that dietary folate intake, total folate intake, and folate blood levels are all unlikely to have any significant association with the risk of breast cancer. Although our analysis of high dietary folate intake and breast cancer was significantly associated with a decreased risk of breast cancer, this result was highly heterogeneous, and become non-significant in the majority of subsequent subgroup analyses. Given that the only significantly positive analyses were those stratified by retrospective study design, it is likely that the results were influenced by methodological biases, such as selection bias and recall bias, rather than representing an accurate association. None of the analyses of total folate and breast cancer risk found any statistically significant associations, although meta-regression of the heterogeneous postmenopausal breast cancer risk analysis did find a significant trend between the absolute level of lowest folate intake level, and an increasing risk of breast cancer. This is suggestive that perhaps those at the lowest levels of total folate intake may indeed have a higher risk of postmenopausal breast cancer, but that our analyses could not detect this due to the majority of studies having folate intakes that were too high. It is not clear, however, as to why this trend was seen in the total folate intake analysis, but not the
Breast Cancer Res Treat (2014) 145:513–524 Fig. 4 Blood folate level and risk of breast cancer
521 Study name
Statistics for each study
Odds ratio Beilby et al (2004) 0.23 Chou et al (2007) 0.64 Ericson et al (2009) 1.20 Lin et al (2008) 1.42 Wu et al (1999) a 0.93 Wu et al (1999) b 1.27 Zhang et al (2003) 0.73 0.86
Lower limit 0.09 0.35 0.84 1.00 0.43 0.53 0.50 0.60
Upper limit 0.56 1.17 1.71 2.02 2.03 3.04 1.07 1.25
Odds ratio and 95% CI p-Value 0.00 0.15 0.31 0.05 0.86 0.59 0.10 0.44 0.01 0.1 1 10 100 Protective Harmful
Meta Analysis
Table 2 Folate and risk of breast cancer by hormonal receptor status Hormone receptor subtype
Dietary folate Included study references
Results
Included study references
Results
Included study references
Results
ER?
[34], [39], [44], [46], [58], [62]
OR 0.91 (95 % CI 0.77–1.08); I2 = 67.8 %, p \ 0.01; Eggers p = 0.08
[40], [44], [46], [58], [64]
OR 1.00 (95 % CI 0.97–1.04); I2 = 0.0 %, p = 0.69; Eggers p = 0.10
[26], [37]
OR 1.59 (95 % CI 1.19–2.12); I2 = 0.0 %, p = 0.45
ER-
[22], [34], [39], [44], [46], [50], [58], [59], [62]
OR 0.96 (95 % CI 0.83–1.11); I2 = 8.95 %, p = 0.36; Eggers p = 0.45
[22], [40], [44], [46], [58], [64]
OR 0.93 (95 % CI 0.82–1.05); I2 = 60.5 %, p = 0.02; Eggers p = 0.06
[26], [37]
OR 1.02 (95 % CI 0.52-2.00); I2 = 0.0 %, p = 0.48
PR?
[44], [46], [58], [59]
OR 0.81 (95 % CI 0.53–1.24); I2 = 86.2 %, p \ 0.01; Eggers p = 0.42
[44], [46], [58]
OR 1.01 (95 % CI 0.97–1.04); I2 = 0.0 %, p = 0.69; Eggers p = 0.32
–
–
PR-
[34], [39], [44], [46], [58], [59]
OR 1.01 (95 % CI 0.90–1.13); I2 = 0.0 %, p = 0.80; Eggers p = 0.38
[44], [46], [58]
OR 1.00 (95 % CI 0.94–1.05); I2 = 0.0 %, p = 0.57; Eggers p = 0.24
–
–
ER?/ PR?
[34], [39], [44], [50], [59], [61]
OR 0.92 (95 % CI 0.69–1.22); I2 = 78.8 %, p \ 0.01; Eggers p = 0.46
–
–
–
–
ER?/ PR-
[34], [39], [44], [59]
OR 0.89 (95 % CI 0.72–1.10); I2 = 0.0 %, p = 0.68; Eggers p = 0.97
[34], [44]
OR 0.83 (95 % CI 0.68–1.02); I2 = 0.0 %, p = 0.71
–
–
ER-/ PR?
[44], [59]
OR 0.49 (95 % CI 0.17–1.42); I2 = 55.9 %, p = 0.13
–
–
–
–
ER-/ PR-
[34], [39], [44], [50], [59], [61]
OR 0.97 (95 % CI 0.81–1.17); I2 = 0.0 %, p = 0.82; Eggers p = 0.21
–
–
–
–
Total folate
dietary folate intake analysis, particularly since a number of studies reporting both total and dietary folate intake were included in both analyses. Given the number of subgroup analyses undertaken in our systematic review, one possibility is that this represents a false positive result secondary to multiple testing. With the prevalence of dietary supplement use approaching 50 % in developed countries [65, 66], further
Blood folate
studies incorporating total folate use rather than dietary folate alone are needed. This is particularly evident given that the majority of studies on folate intake and breast cancer risk have been performed in developed countries. We similarly found no association between blood folate levels and the risk of overall breast cancer. An elevated risk for estrogen receptor positive breast cancer with high folate levels was seen on subgroup analysis, but with only 2 studies
123
522
Breast Cancer Res Treat (2014) 145:513–524
available to pool, robust conclusions cannot be made. Further research on whether folate blood levels influence the risk of breast cancer is needed. A lack of significant association between dietary and total folate intake and breast cancer risk remained in all of the hormonal receptor subtype analyses, although the inclusion of Zhang et al. [59] in the dietary folate analyses caused significant heterogeneity, as it reported a substantially decreased risk of breast cancer in comparison to the other included studies. Zhang et al. [59] is one of only 4 studies [28, 49, 50] included in our systematic review that originates from China, and it is notable that the other three Chinese studies reported either statistically significant decreased risks of breast cancer [28, 48], or a non-statistically significant decreased risk with a comparatively lower point estimate [50]. These studies also had lower folate intakes compared to the majority of the other included studies, and are suggestive that perhaps very low intakes of folate may increase breast cancer risk. Further research in developing countries and populations that include lower average folate intakes is needed to further clarify this. Our systematic review had a number of strengths. This is the first systematic review to report on folate intake and breast cancer by hormone receptor status. We followed the MOOSE guidelines in reporting our review. Multiple databases were searched, and were not limited by language. The subgroup analyses were all pre-specified. There were also a number of limitations. As the majority of included studies were primarily from developed countries, firm conclusions about populations in developing countries could not be made. Few studies on blood folate levels were available to analyze. We were also unable to compare risks before and after folate fortification in US-based studies.
Conclusions Folate intake and folate blood levels do not appear to be associated with breast cancer risk at the level of intake encountered in developed countries. This lack of association remained after stratification by menopausal status and hormone receptor status. An association with very low levels of folate intake cannot be ruled out, but given the scarcity of such studies; further research examining populations with very low folate intakes will be needed.
Conflicts of interest interest.
The authors have declared no conflicts of
References 1. Jemal A, Bray F, Center MM et al (2011) Global cancer statistics. CA Cancer J Clin 61(2):69–90
123
2. Seitz HK, Pelucchi C, Bagnardi V et al (2012) Epidemiology and pathophysiology of alcohol and breast cancer: update 2012. Alcohol Alcohol 47(3):204–212 3. Trichopoulos D, Adami HO, Ekbom A et al (2008) Early life events and conditions and breast cancer risk: from epidemiology to etiology. Int J Cancer 122(3):481–485 4. Blount BC, Mack MM, Wehr CM et al (1997) Folate deficiency causes uracil misincorporation into human DNA and chromosomal breakage; implications for cancer and neuronal damage. Proc Natl Acad Sci USA 94(7):3290–3295 5. Stempak JM, Sohn KY, Chiang EP et al (2005) Cell and stage of transformation-specific effects of folate deficiency on methionine cycle intermediates and DNA methylation in an in vitro model. Carcinogenesis 26(5):981–990 6. Wien TN, Pike E, Wisloff T et al (2012) Cancer risk with folic acid supplements: a systematic review and meta-analysis. BMJ Open 12(1):e000653 7. Ly A, Lee H, Chen J et al (2011) Effect of maternal and postweaning folic acid supplementation on mammary tumor risk in the offspring. Cancer Res 71(3):988–997 8. Kotsopoulos J, Medline A, Renlund R et al (2005) Effects of dietary folate on the development and progression of mammary tumors in rats. Carcinogenesis 26(9):1603–1612 9. Stevens VL, McCullough ML, Sun J et al (2010) Folate and other one-carbon metabolism-related nutrients and risk of postmenopausal breast cancer in the Cancer Prevention Study II Nutrition Cohort. Am J Clin Nutr 91(6):1708–1715 10. Ericson U, Sonestedt E, Gullberg B et al (2007) High folate intake is associated with lower breast cancer incidence in postmenopausal women in the Malmo Diet and Cancer cohort. Am J Clin Nutr 86(2):434–443 11. Larsson SC, Giovannucci E, Wolk A (2007) Folate and risk of breast cancer: a meta-analysis. J Natl Cancer Inst 99(1):64–76 12. Lewis SJ, Harbord RM, Harris R et al (2006) Meta-analyses of observational and genetic association studies of folate intakes or levels and breast cancer risk. J Natl Cancer Inst 98(22):1607–1622 13. Stroup DF, Berlin JA, Morton SC et al (2000) Metaanalysis of Observational Studies in Epidemiology (MOOSE) Group. Metaanalysis of observational studies in epidemiology: a proposal for reporting. JAMA 283(15):2008–2012 14. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7(3):177–188 15. Higgins JP, Thompson SG, Deeks JJ et al (2003) Measuring inconsistency in meta-analyses. BMJ 327(7414):557–560 16. Egger M, Davey Smith G, Schneider M et al (1997) Bias in metaanalysis detected by a simple, graphical test. BMJ 315(7109): 629–634 17. Orwin R (1983) A fail safe N for effect size in meta-analysis. J Educ Stat 8:157–159 18. Adzersen KH, Jess P, Freivogel KW et al (2003) Raw and cooked vegetables, fruits, selected micronutrients, and breast cancer risk: a case control study in Germany. Nutr Cancer 46(2):131–137 19. Aune D, Deneo-Pellegrini H, Ronco AL et al (2011) Dietary folate intake and the risk of 11 types of cancer: a case control study in Uruguay. Ann Oncol 22(2):444–451 20. Beilby J, Ingram D, Hahnel R et al (2004) Reduced breast cancer risk with increasing serum folate in a case-control study of the C677T genotype of the methylenetetrahydrofolate reductase gene. Eur J Cancer 40(8):1250–1254 21. Chen J, Gammon MD, Chan W et al (2005) One-carbon metabolism, MTHFR polymorphisms, and risk of breast cancer. Cancer Res 65(4):1606–1614 22. Cho E, Holmes M, Hankinson SE et al (2007) Nutrients involved in one-carbon metabolism and risk of breast cancer among premenopausal women. Cancer Epidemiol Biomarkers Prev 16(12):2787–2790
Breast Cancer Res Treat (2014) 145:513–524 23. Chou YC, Lee MS, Wu MH et al (2007) Plasma homocysteine as a metabolic risk factor for breast cancer: findings from a case-control study in Taiwan. Breast Cancer Res Treat 101(2):199–205 24. Duffy CM, Assaf A, Cyr M et al (2009) Alcohol and folate intake and breast cancer risk in the WHI Observational Study. Breast Cancer Res Treat 116(3):551–562 25. Ericson UC, Ivarsson MI, Sonestedt E et al (2009) Increased breast cancer risk at high plasma folate concentrations among women with the MTHFR 677T allele. Am J Clin Nutr 90(5): 1380–1389 26. Ericson U, Borqquist S, Ivarsson MI et al (2010) Plasma folate concentrations are positively associated with risk of estrogen receptor beta negative breast cancer in a Swedish nested case control study. J Nutr 140(9):1661–1668 27. Freudenheim JL, Marshall JR, Vena JE et al (1996) Premenopausal breast cancer risk and intake of vegetables, fruits, and related nutrients. J Natl Cancer Inst 88(6):340–348 28. Gao CM, Tang JH, Cao HX et al (2009) MTHFR polymorphisms, dietary folate intake and breast cancer risk in Chinese women. J Hum Genet 54(7):414–418 29. Graham S, Hellmann R, Marshall J et al (1991) Nutritional epidemiology of postmenopausal breast cancer in western New York. Am J Epidemiol 134(6):552–566 30. Kabat GC, Miller AB, Jain M et al (2008) Dietary intake of selected B vitamins in relation to risk of major cancers in women. Br J Cancer 99(5):816–821 31. Kotsopoulos J, Sukiennicki G, Muszynska M et al (2012) Plasma micronutrients, trace elements, and breast cancer in BRCA1 mutation carriers: an exploratory study. Cancer Causes Control 23(7):1065–1074 32. Lajous M, Lazcano-Ponce E, Hernandez-Avila M et al (2006) Folate, vitamin B6, and vitamin B12 intake and the risk of breast cancer. Cancer Epidemiol Biomarkers Prev 15(3):443–448 33. Lajous M, Romieu I, Sabia S et al (2006) Folate, vitamin B12 and postmenopausal breast cancer in a prospective study of French women. Cancer Causes Control 17(9):1209–1213 34. Larsson SC, Berqkvist L, Wolk A (2008) Folate intake and risk of breast cancer by estrogen and progesterone receptor status in a Swedish cohort. Cancer Epidemiol Biomarkers Prev 17(12): 3444–3449 35. Lee SA, Lee KM, Lee SJ et al (2010) Antioxidant vitamins intake, ataxia telangiectasia mutated (ATM) genetic polymorphisms, and breast cancer risk. Nutr Cancer 62(8):1087–1094 36. Levi F, Pasche C, Lucchini F et al (2001) Dietary intake of selected micronutrients and breast-cancer risk. Int J Cancer 91(2):260–263 37. Lin J, Lee IM, Cook NR et al (2008) Plasma folate, vitamin B-6, vitamin B-12, and risk of breast cancer in women. Am J Clin Nutr 87(3):734–743 38. Ma E, Iwasaki M, Junko I et al (2009) Dietary intake of folate, vitamin B6, and vitamin B12, genetic polymorphism of related enzymes, and risk of breast cancer: a case-control study in Brazilian women. BMC Cancer 9:122 39. Ma E, Iwasaki M, Kobayashi M et al (2009) Dietary intake of folate, vitamin B2, vitamin B6, vitamin B12, genetic polymorphism of related enzymes, and risk of breast cancer: a casecontrol study in Japan. Nutr Cancer 61(4):447–456 40. Maruti SS, Ulrich CM, White E (2009) Folate and one-carbon metabolism nutrients from supplements and diet in relation to breast cancer risk. Am J Clin Nutr 89(2):624–633 41. Negri E, La Vecchia C, Franceschi S (2000) Re: dietary folate consumption and breast cancer risk. J Natl Cancer Inst 92(15): 1270–1271 42. Potischman N, Swanson CA, Coates RJ et al (1999) Intake of food groups and associated micronutrients in relation to risk of early-stage breast cancer. Int J Cancer 82(3):315–321
523 43. Rohan TE, Jain MG, Howe GR et al (2000) Dietary folate consumption and breast cancer risk. J Natl Cancer Inst 92(3):266–269 44. Roswall N, Olsen A, Christensen J et al (2010) Micronutrient intake and breast cancer characteristics among postmenopausal women. Eur J Cancer Prev 19(5):360–365 45. Sellers TA, Kushi LH, Cerhan JR et al (2001) Dietary folate intake, alcohol, and risk of breast cancer in a prospective study of postmenopausal women. Epidemiology 12(4):420–428 46. Sellers TA, Vierkant RA, Cerhan JR et al (2002) Interaction of dietary folate intake, alcohol, and risk of hormone receptordefined breast cancer in a prospective study of postmenopausal women. Cancer Epidemiol Biomarkers Prev 11:1104–1107 47. Sellers TA, Gabrick DM, Vierkant RA et al (2004) Does folate intake decrease risk of postmenopausal breast cancer among women with a family history? Cancer Causes Control 15(2): 113–120 48. Sharp L, Little J, Schofield AC et al (2002) Folate and breast cancer: the role of polymorphisms in methylenetetrahydrofolate reductase (MTHFR). Cancer Lett 181(1):65–71 49. Shrubsole MJ, Jin F, Dai Q et al (2001) Dietary folate intake and breast cancer risk: results from the Shanghai Breast Cancer Study. Cancer Res 61(19):7136–7141 50. Shrubsole MJ, Shu XO, Li HL et al (2011) Dietary B vitamin and methionine intakes and breast cancer risk among Chinese women. Am J Epidemiol 173(10):1171–1182 51. Stolzenberg-Solomon RZ, Chang SC, Leitzmann MF et al (2006) Folate intake, alcohol use, and postmenopausal breast cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Am J Clin Nutr 83(4):895–904 52. Suzuki T, Matsuo K, Hirose K et al (2008) One-carbon metabolism-related gene polymorphisms and risk of breast cancer. Carcinogenesis 29(2):356–362 53. Thorand B, Kohlmeier L, Simonsen N et al (1998) Intake of fruits, vegetables, folic acid, and related nutrients and risk of breast cancer in postmenopausal women. Public Health Nutr 1(3):147–156 54. Wu K, Helzlsouer KJ, Comstock GW et al (1999) A prospective study on folate, B12, and pyridoxal 50 -phosphate (B6) and breast cancer. Cancer Epidemiol Biomarkers Prev 8(3):209–217 55. Yang YJ, Hwang SH, Kim HJ et al (2010) Dietary intake of nitrate relative to antioxidant vitamin in relation to breast cancer risk: a case-control study. Nutr Cancer 62(5):555–566 56. Zhang S, Hunter DJ, Hankinson SE et al (1999) A prospective study of folate intake and the risk of breast cancer. JAMA 281(17):1632–1637 57. Zhang SM, Willett WC, Selhub J et al (2003) Plasma folate, vitamin B6, vitamin B12, homocysteine, and risk of breast cancer. J Natl Cancer Inst 95(5):373–380 58. Zhang SM, Hankinson SE, Hunter DJ et al (2005) Folate intake and risk of breast cancer characterized by hormone receptor status. Cancer Epidemiol Biomarkers Prev 14(8):2004–2008 59. Zhang CX, Ho SC, Chen YM et al (2011) Dietary folate, vitamin B6, vitamin B12, and methionine intake and the risk of breast cancer by oestrogen and progesterone receptor status. Br J Nutr 106(6):936–943 60. Zhu K, Davidson NE, Hunter S et al (2003) Methyl-group dietary intake and risk of breast cancer among African-American women: a case-control study by methylation status of the estrogen receptor alpha genes. Cancer Causes Control 14(9):827–836 61. Bassett JK, Baglietto L, Hodge AMm et al (2013) Dietary intake of B vitamins and methionine and breast cancer risk. Cancer Causes Control 24(8):1555–1563 62. Gong Z, Ambrosone CB, McCann SE et al (2014) Associations of dietary folate, Vitamins B6 and methionine intake with risk of breast cancer among African American and European American women. Int J Cancer 134(6):1422–1435
123
524 63. Islam T, Ito H, Sueta A et al (2013) Alcohol and dietary folate intake and the risk of breast cancer: a case-control study in Japan. Eur J Cancer Prev 22(4):358–366 64. Yang D, Baumgartner RN, Slattery ML et al (2013) Dietary intake of folate, B-vitamins and methionine and breast cancer risk among Hispanic and non-Hispanic white women. PLoS One 8(2):e54495
123
Breast Cancer Res Treat (2014) 145:513–524 65. Bailey RL, Gahche JJ, Lentino CV, Dwyer JT, Engel JS, Thomas PR et al (2011) Dietary supplement use in the United States, 2003–2006. J Nutr 141(2):261–266 66. Skeie G, Braaten T, Hjartaker A, Lentjes M, Amiano P, Jakszyn P et al (2009) Use of dietary supplements in the European Prospective Investigation into Cancer and Nutrition calibration study. Eur J Clin Nutr 63(S4):S226–S238
EPIDEMIOLOGY
Folate intake and the risk of breast cancer: a systematic review and meta-analysis Martin Tio • Juliana Andrici • Guy D. Eslick
Received: 10 April 2014 / Accepted: 12 April 2014 / Published online: 29 April 2014 Ó Springer Science+Business Media New York 2014
Abstract There is conflicting epidemiological evidence on the role of folate and breast cancer risk. We conducted a systematic review and quantitative meta-analysis of folate intake and folate blood levels and the risk of breast cancer. Four electronic databases (Medline, PubMed, Embase, and Current Contents Connect) were searched to April 11, 2014, with no language restrictions for observational studies that measured folate intake or blood levels and the risk of breast cancer. The meta-analysis of dietary folate intake comprising 36 studies with 34,602 cases, and a total sample size of 608,265 showed a decreased risk of breast cancer, with an odds ratio (OR) of 0.84 [95 % confidence interval (CI) 0.77–0.91]. When stratified by menopausal status and by study design, none of the meta-analyses of prospective studies showed any statistically significant decrease in the risk of breast cancer. The meta-analysis of total folate showed no statistically significant association with breast cancer OR of 0.98 (95 % CI 0.91–1.07). There was no significant association between either dietary or total folate intake and breast cancer when stratified by hormonal receptor status. The meta-analysis of blood folate levels found no significant association with the risk of breast cancer, with an OR of 0.86 (95 % CI 0.60–1.25). Breast cancer does not appear to be associated with folate intake, and this did not vary by menopausal status or hormonal receptor status. Folate blood levels also do not appear to be associated with breast cancer risk.
M. Tio J. Andrici G. D. Eslick (&) The Whiteley-Martin Research Centre, The Discipline of Surgery, Sydney Medical School, The University of Sydney, Nepean Hospital, Level 5, South Block, Penrith, NSW, Australia e-mail: [email protected]
Keywords Breast cancer Folate Folic acid Meta-analysis Systematic review
Introduction Breast cancer is both the most common form of cancer in females, and the leading cause of cancer mortality in females [1]. With the possible exception of alcohol [2], nutritional intake has not been well established as a risk factor for breast cancer [3]. Folate has been implicated as one possible nutritional risk factor for cancer in general, as it plays a major role in the one-carbon metabolism of DNA. Evidence supporting folate with carcinogenesis ranges from in vitro evidence indicating that low folate levels contribute to DNA damage and epigenetic modification of DNA methylation [4, 5], all the way up to a meta-analysis of randomized clinical trials suggesting an increased risk of cancer with individuals given supplemental doses of folate [6]. The association between folate and breast cancer specifically though, is much more equivocal. There is conflicting evidence both at the level of basic science investigations, and at the level of human observational studies [7–10]. Two meta-analyses of observational studies in humans have indicated that folate is not significantly associated with the risk of breast cancer [11, 12]. Subsequent to the publication of these meta-analyses, however, has been the further publication of a number of large-scale observational studies on folate and breast cancer. Additionally, these meta-analyses did not include any evaluation of folate and the risk of different hormonal receptor subtypes of breast cancer. As such, we performed a systematic review and metaanalysis of observational studies on folate intake and the
123
514
risk of breast cancer, including analyses of hormone receptor subtypes of breast cancer.
Methods Search strategy We followed the meta-analysis of observational studies in epidemiology (MOOSE) guidelines in performing our systematic review [13]. Relevant articles were identified by two reviewers (M.T. and J.A.) by systematically searching through MEDLINE (from 1950), PubMed (from 1946), EMBASE (from 1949) and Current Contents Connect (from 1998) through to March 2nd, 2014. The search used the terms ‘‘folate’’ or ‘‘folic acid’’ or ‘‘vitamin B9’’ and ‘‘breast cancer’’ or ‘‘breast neoplasms.’’ The search terms used were searched as text word and as exploded medical subject headings where possible. The reference lists of relevant articles were also searched for appropriate studies. No language restrictions were used in either the search or study selection. A search for unpublished literature was not performed. Disagreement on article inclusion between the two reviewers was resolved via a third reviewer (G.E.). Study selection In order to be included, eligible studies needed to (1) have a study design of either a cohort or case control; (2) report the risk of breast cancer in association with total folate intake, dietary folate intake, or folate blood levels. Dietary folate was defined as all folate intake via foods, and the total folate intake was defined as all folate intake via foods and supplements; (3) report the risk point estimate as an odds ratio (OR), hazard ratio, relative risk, or incidence rate ratio that compared a higher level of folate intake to a lower level of folate intake. When multiple levels of folate intake were presented, the ratio comparing the highest intake versus the lowest intake was chosen; (4) report the 95 % confidence interval (CI) for the point estimate; and (5) use an internal comparison when calculating the risk estimate. Studies were excluded if they did not meet the inclusion criteria, if they were duplicates of other studies eligible for inclusion, or if they were meta-analyses of studies. In the case of duplicate studies, the most recently published study was chosen for inclusion.
Breast Cancer Res Treat (2014) 145:513–524
direction, population type, country, continent, economic development, response rate, case control matching, mean age, number of adjusted variables, years of follow up, lowest folate level, highest folate level, difference between highest and lowest folate levels, and the risk estimates and CIs on breast cancer. Authors were not contacted if data were missing. Adjusted ratios were extracted in preference to non-adjusted ratios. When more than one adjusted ratio was reported, the ratio with the highest number of adjusted variables was chosen. Statistical analysis Pooled estimates of the OR and 95 % CI for the risk of breast cancer in association with folate were calculated using the random-effects model of DerSimonian and Laird [14]. Heterogeneity was assessed using the I2 statistic, which determines the proportion of variability across studies that is due to heterogeneity as opposed to sampling error [15]. Sensitivity analyses were performed when statistically significantly heterogeneity was detected. Subgroup analyses stratified by either study design or population group were also performed. Meta-regression was performed when heterogeneity remained after sensitivity and subgroup analysis. Our pre-specified meta-regression analyses compared the log odd ratios versus highest absolute folate level, lowest absolute folate level, and difference between highest and lowest absolute folate level. Publication bias was assessed with Egger’s regression model [16]. If publication bias was detected, then the additional publication bias methods consisting of the failsafe number method and the trim-and-fill method were employed to quantify the effect of the bias. The fail-safe number method calculates the number of unpublished studies needed to convert the observed result to statistical non-significance at the alpha level of significance p \ 0.05 level. Publication bias is considered to be an issue if the fail-safe number is less than 5n ? 10, where n is the number of studies included in the meta-analysis [17]. The trim-and-fill method simulates unpublished studies in the meta-analysis to calculate a new pooled OR, which is then compared to the original pooled OR. If the new pooled OR is similar to the original pooled OR, then this indicates that publication bias has little effect on the meta-analysis results. Results were regarded as statistically significant if p \ 0.05. All analyses were done with Comprehensive Meta-analysis (version 2.0), Englewood, NJ, USA (2005).
Data extraction One reviewer (M.T.) performed the data extraction via a standardized data extraction form. Information was extracted on the publication year, study design, number of cases, number of controls, total sample size, temporal
123
Results Our search identified 3,725 citations, from which 49 articles were extracted for inclusion in our systematic review
Breast Cancer Res Treat (2014) 145:513–524
515
Fig. 1 Study selection flowchart
(Fig. 1). Table 1 shows selected characteristics of the identified studies [9, 10, 18–64]. The majority of the included studies originated from either Europe [10, 18, 25, 26, 31, 33, 34, 36, 41, 44, 48, 53] or North America [9, 21, 22, 24, 27, 29, 30, 37, 40, 42, 43, 45–47, 51, 54, 56–58, 60, 62, 64] with the remainder from Asia [23, 28, 35, 39, 49, 50, 52, 55, 59, 63] Central and South America [19, 32, 38], and Oceania [20, 61]. Two studies that met our inclusion criteria could not be examined with meta-analysis due to a lack of other studies to pool them with. Kotsopolous et al. [31] found no significant association with plasma folate levels and the risk of breast cancer in BRCA1 carriers, and Sellers et al. [47] found a significantly decreased risk of breast cancer with high dietary folate intake in individuals with a family history of breast cancer, but no significant association in individuals without a family history of breast cancer. Dietary folate intake Figure 2 shows the meta-analysis of 36 studies (14 prospective studies and 22 retrospective studies) comprising 34,602 cases with a total sample size of 608,265 on dietary
folate intake and risk of breast cancer. High dietary folate intake was associated with a statistically significant decreased risk of breast cancer, with an OR of 0.84 (95 % CI 0.77–0.91). There was statistically significant heterogeneity (I2 = 71.2 %; p \ 0.01), and a sensitivity analysis did not find any single study that contributed significantly to the heterogeneity. There was statistically significant publication bias (Eggers test, p \ 0.01), and the adjusted results using the trim-and-fill method gave a non-statistically significant OR of 0.93 (95 % CI 0.85–1.02). The failsafe N was 364. When stratified by menopausal status, the meta-analysis of postmenopausal breast cancer risk included 20 studies (9 prospective studies and 11 retrospective studies) comprising 15,484 cases with a total sample size of 360,634. There was a statistically significant association between high dietary folate intake and risk of postmenopausal breast cancer, with an OR of 0.84 (95 % CI 0.75–0.94). There was significant heterogeneity (I2 = 62.6 %; p \ 0.01), and sensitivity analysis did not find any single study that contributed significantly to the heterogeneity. There was statistically significant publication bias (Eggers test, p \ 0.01), and the adjusted OR via the trim-and-fill method
123
123
Cohort
Gong et al. [62]
Larsson et al. [34]
Case control
Gao et al. [28]
Cohort
Case control
Freudenheim et al. [27]
Lajous et al. [33]
Case control
Ericson et al. [26]
Case control
Nested case control
Ericson et al. [25]
Lajous et al. [32]
Nested case control
Ericson et al. [10]
Case control
Cohort
Duffy et al. [24]
Kotsopolous et al. [31]
Cohort
Chou et al. [23]
Cohort
Case control
Cho et al. [22]
Kabat et al. [30]
Cohort
Chen et al. [21]
Case control
Case control
Beilby et al. [20]
Case control
Case control
Bassett et al. [61]
Islam et al. [63]
Cohort
Aune et al. [19]
Graham et al. [29]
Case control
Case control
Adzersen et al. [18]
Study type
Source
Sweden
France
Mexico
Poland
Canada
Japan
USA
USA
China
USA
Sweden
Sweden
Sweden
USA
Taiwan
USA
USA
Australia
Australia
Uruguay
Germany
Country
Table 1 Studies included in the systematic review
2,952
1,812
475
48
2,491
1,754
430
1,493
669
297
203
313
392
1,783
146
221
1,426
141
936
461
310
Cases
61,433
62,739
1,794
144
49,654
5,262
917
3,016
1,351
608
404
939
11,619
86,747
431
90,663
2,908
250
20,756
2,493
663
Total size
Population
Population
Hospital cases; population controls
Hospital
Population
Hospital
Population
Mixed cases; population controls
Hospital and population cases; population controls
Hospital cases; population controls
Population
Population
Population
Hospital
Hospital
Population
Population
Hospital cases; population controls
Population
Hospital
Hospital
Population derivation
Adults born 1914–1948
Adults born 1925–1950
Adults aged 20–75
Adults
Adults aged 40–59
Adults aged 20–79
White adults
Adults aged 20–75
Adults
White adults
Adults born 1926–1945
Adults born 1926–1945
Adults born 1926–1945
Postmenopausal adults
Adults aged 20–80
Registered nurses
Adults
Adults aged 30–84
Adults aged 27–80
Dietary folate total folate
Dietary folate
Dietary folate
Plasma folate
Dietary folate
Dietary folate
Dietary folate
Dietary folate
Dietary folate
Dietary folate
Plasma folate
Plasma folate
Dietary folate total folate
Total folate
Total folate plasma folate
Dietary folate total folate
Dietary folate total folate
Serum folate
Dietary folate
Dietary folate Dietary folate
Adults aged 25–75
Folate measurement
Adults aged \90
Population
PR subtype risk
ER subtype risk
Overall risk
Postmenopausal risk
Postmenopausal risk
Premenopausal risk
Overall risk
BRCA1 carrier risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
Postmenopausal risk
ER subtype risk
Overall risk
Overall risk
Premenopausal risk
Postmenopausal ER subtype risk
Postmenopausal risk
Postmenopausal risk
Postmenopausal risk
Overall risk
Premenopausal ER subtype risk
Premenopausal risk
Overall risk
Overall risk
PR subtype risk
ER subtype risk
Overall risk
Overall risk
Overall risk
Reported breast cancer risks
516 Breast Cancer Res Treat (2014) 145:513–524
Nested case control
Cohort
Cohort
Cohort
Cohort
Case control
Case control
Rohan et al. [43]
Roswall et al. [44]
Sellers et al. [45]
Sellers et al. [46]
Sellers et al. [47]
Sharp et al. [48]
Shrubsole et al. [49]
Italy
Cohort
Maruti et al. [40]
Case control
Case control
Ma et al. [39]
Case control
Case control
Ma et al. [38]
Negri et al. [41]
Nested case control
Lin et al. [37]
Potischman et al. [42]
USA
Case control
Levi et al. [36]
China
UK
USA
USA
USA
Denmark
Canada
USA
Japan
Brazil
USA
Switzerland
South Korea
Case control
Lee et al. [35]
Country
Study type
Source
Table 1 continued
1,321
62
1,823
1,416
1,586
1,072
1,336
568
2,569
743
398
458
848
289
323
Cases
2,703
128
33,552
34,393
35,973
26,224
6,718
2,019
5,157
35,023
796
916
1,696
731
646
Total size
Population
Hospital
Population
Population
Population
Population
Population
Population
Hospital
Population
Hospital
Hospital
Population
Hospital
Hospital
Population derivation
Adults aged 25–64
Adults
Adults 55–69
Adults 55–69
Adults 55–69
Adults aged 50–64
Adults aged 40–59
Adults aged 20–44
Adults
Adults aged 50–76
Adults aged 20–74
Adults aged 20–74
Dietary folate
Dietary folate
Dietary folate total folate
Dietary folate total folate
Dietary folate total folate
Dietary folate total folate
Dietary folate
Dietary folate total folate
Dietary folate
Dietary folate total folate
Dietary folate
Dietary folate
Dietary folate total folate plasma folate
Dietary folate
Adults aged \75
Adults aged 45 or older
Dietary folate
Folate measurement
Adults
Population
Postmenopausal risk
Premenopausal risk
Overall risk
Overall risk
Negative family History risk
Positive family History risk
PR subtype risk
ER subtype risk
Postmenopausal risk
PR subtype risk
ER subtype risk
Postmenopausal risk
Postmenopausal risk
Premenopausal risk
Premenopausal risk
Overall risk
ER subtype risk
Postmenopausal risk
PR subtype risk
ER subtype risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
PR subtype risk
ER subtype risk
Postmenopausal risk
Premenopausal risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
Postmenopausal risk
Premenopausal risk
Overall risk
Reported breast cancer risks
Breast Cancer Res Treat (2014) 145:513–524 517
123
123
Case control
Case control
Zhu et al. [60]
Case Control
Yang et al. [64]
Zhang et al. [59]
Case control
Yang et al. [55]
Cohort
Nested case control
Wu et al. [54]
Zhang et al. [58]
Nested case control
Wu et al. [54]
Cohort
Case control
Thorand et al. [53]
Nested case control
Case control
Suzuki et al. [52]
Zhang et al. [56]
Cohort
Zhang et al. [57]
USA
Cohort
Stevens et al. [9]
Stolzenberg-Solomen et al. [51]
USA
China
USA
USA
USA
USA
South Korea
USA
USA
Germany
Japan
USA
China
Cohort
Shrubsole et al. [50]
Country
Study type
Source
Table 1 continued
304
438
2,812
712
3,483
2,325
362
110
133
43
456
691
3,898
718
Cases
609
876
88,744
1,424
88,818
4,850
724
220
266
149
1,368
25,400
70,656
73,237
Total size
Population
Hospital
Population
Population
Population
Hospital cases; population controls
Hospital
Population
Population
Population
Hospital
Population
Population
Population
Population derivation
African-American adults
Adults aged 25–70
Adults aged 30–55
Adults aged 30–55
Adults aged 30–55
Adults aged 25–79
Adults aged 30–65
Adults with serum sampled in 1989
Adults with serum sampled in 1974
Adults aged 38–80
Adults aged 20–79
Adults aged 55–74
Adults aged 50–74
Adults aged 40–70
Population
Dietary folate
Dietary Folate
Dietary folate total folate
Plasma folate
Total folate
Total folate
Dietary folate
Serum folate
Serum folate
Dietary folate
Dietary folate total folate
Dietary folate total folate
Dietary folate total folate
Dietary folate
Folate measurement
Overall risk
PR subtype risk
ER subtype risk
Overall risk
PR subtype risk
ER subtype risk
Overall risk
Overall risk
ER subtype risk
Overall risk
Overall risk
Overall risk
Overall risk
Postmenopausal
Postmenopausal Risk
Premenopausal risk
Overall risk
Postmenopausal risk
Postmenopausal risk
PR subtype risk
ER subtype risk
Postmenopausal risk
Premenopausal risk
Overall risk
Reported breast cancer risks
518 Breast Cancer Res Treat (2014) 145:513–524
Breast Cancer Res Treat (2014) 145:513–524 Fig. 2 Dietary folate intake and risk of breast cancer
519
Study name
Statistics for each study
Odds ratio and 95% CI
Odds ratio
Lower limit
Upper limit
p-Value
0.47 1.05 0.99 0.85 1.08 0.56 0.76 0.70 0.91 0.70 0.79 1.02 0.64 0.78 1.01 0.50 0.45 0.84 1.26 1.01 0.91 0.73 0.89 1.72 1.22 0.83 0.49 0.62 0.79 1.12 Stolzenberg-Solomon et al (2006) 0.97 Suzuki et al (2008) 0.65 Thorand et al (1998) 1.14 Yang et al (2010) 1.22 Zhang et al (2011) 0.32 Zhu et al (2003) 0.58 0.84
0.26 0.54 0.83 0.64 0.86 0.34 0.43 0.53 0.67 0.48 0.68 0.89 0.45 0.67 0.90 0.29 0.27 0.52 0.89 0.62 0.68 0.60 0.68 0.97 0.95 0.62 0.20 0.46 0.59 1.01 0.77 0.46 0.73 0.65 0.21 0.25 0.77
0.86 2.03 1.19 1.13 1.35 0.92 1.36 0.92 1.23 1.02 0.92 1.16 0.91 0.90 1.13 0.87 0.74 1.36 1.78 1.64 1.22 0.88 1.17 3.05 1.57 1.11 1.20 0.83 1.06 1.24 1.23 0.91 1.79 2.28 0.49 1.36 0.91
0.01 0.88 0.91 0.27 0.50 0.02 0.35 0.01 0.54 0.06 0.00 0.77 0.01 0.00 0.86 0.01 0.00 0.48 0.19 0.97 0.53 0.00 0.40 0.06 0.12 0.20 0.12 0.00 0.11 0.03 0.80 0.01 0.57 0.53 0.00 0.21 0.00
Adzersen et al (2003) Aune et al (2010) Bassett et al (2013) Chen et al (2005) Cho et al (2007) Ericson et al (2007) Freudenheim et al (1996) Gao et al (2009) Gong et al (2014) Graham et al (1991) Islam et al (2013) Kabat et al (2008) Lajous et al (2006) a Lajous et al (2006) b Larsson et al (2008) Lee et al (2011) Levi et al (2001) Lin et al (2008) Ma et al (2009) a Ma et al (2009) b Maruti et al (2009) Negri et al (2000) Potischman et al (1999) Rohen et al (2000) Roswall et al (2010) Sellers et al (2001) Sharp et al (2002) Shrubsole et al (2001) Shrubsole et al (2011) Stevens et al (2010)
0.1
0.2
0.5
Protective
showed no significant association, with an OR of 0.94 (95 % CI 0.83–1.06). The fail-safe N was 95. Further stratification via prospective study design found no statistically significant association with postmenopausal breast cancer risk, with an OR of 0.93 (95 %CI 0.81–1.07). There was statistically significant heterogeneity (I2 = 68.4 %; p \ 0.01), but this became non-significant on sensitivity analysis with the removal of Stevens et al. [9] (I2 = 48.0 %; p = 0.06). The removal of Stevens et al. [9] did not significantly change the results, with an OR of 0.90 (95 % CI 0.79–1.02). There was no publication bias (Eggers test, p = 0.24). The stratification via retrospective study design found a statistically significant decreased risk of postmenopausal breast cancer risk, with an OR of 0.76 (95 % CI 0.66–0.86). There was no significant heterogeneity (I2 = 20.9 %; p = 0.24). There was no significant publication bias (Eggers test, p = 0.14). The meta-analysis of dietary folate intake and premenopausal breast cancer risk included 13 studies (3 prospective studies and 10 retrospective studies) with 6,376 cases and a total sample size of 181,199. There was a statistically significant association between dietary folate intake and the risk of premenopausal breast cancer, with an OR of 0.81 (95 % CI 0.66–1.00). There was significant heterogeneity (I2 = 69.0 %; p \ 0.01), which was not eliminated after sensitivity analysis. There was no significant publication bias (Eggers test, p = 0.99).
1
2
5
10
Harmful
Stratifying by prospective study design showed no significant association for premenopausal breast cancer risk, with an OR 1.02 (95 % CI 0.62–1.68). There was significant heterogeneity (I2 = 73.8 %; p = 0.02), but sensitivity analysis and meta-regression were not appropriate given that there were only 3 studies. There was no publication bias (Eggers test, p = 0.93). Stratifying by retrospective study design showed a significant association, with an OR of 0.75 (95 % CI 0.61–0.93). There was significant heterogeneity (I2 = 58.8 %; p = 0.01), which became non-significant on the removal of Ma et al. [38] (I2 = 0.0 %; p = 0.52),. This did not significantly change the results, with an OR of 0.70 (95 % CI 0.61–0.79). There was no publication bias (Eggers test, p = 0.67). Total folate intake Figure 3 shows the meta-analysis of total folate intake and breast cancer risk, which included 15 studies (11 prospective and 4 retrospective) comprising 21,001 cases with a total sample size of 521,474. There was no statistically significant association between total folate intake and breast cancer risk, with an OR of 0.98 (95 % CI 0.91–1.07). There was significant heterogeneity (I2 = 52.7 %; p = 0.01) which was not removed on sensitivity analysis. There was no publication bias (Eggers test, p = 0.44). When stratified by menopausal status, the meta-analysis of postmenopausal included 7 prospective studies comprising
123
520 Fig. 3 Total folate intake and risk of breast cancer
Breast Cancer Res Treat (2014) 145:513–524 Study name
Statistics for each study
Odds ratio Chen et al (2005) 0.95 Cho et al (2007) 1.09 Chou et al (2007) 0.51 Duffy et al (2009) 0.97 Ericson et al (2007) 0.56 Larsson et al (2008) 1.06 Lin et al (2008) 1.24 Maruti et al (2009) 0.78 Potischman et al (1999) 1.11 Roswall et al (2010) 1.23 Sellers et al (2001) 0.84 Stevens et al (2010) 1.03 Stolzenberg-Solomon et al (2006) 1.27 Yang et al (2013) 0.90 Zhang et al (1999) 0.93 0.98
Lower limit 0.74 0.88 0.30 0.84 0.35 0.87 0.88 0.61 0.81 0.97 0.63 0.93 1.00 0.64 0.83 0.91
Upper limit 1.22 1.35 0.87 1.12 0.90 1.29 1.75 0.99 1.52 1.56 1.11 1.15 1.62 1.26 1.04 1.07
Odds ratio and 95% CI
p-Value 0.69 0.42 0.01 0.68 0.02 0.55 0.22 0.04 0.52 0.09 0.23 0.59 0.05 0.54 0.19 0.71 0.1 0.2 0.5 Protective
10,165 cases with a total sample size of 293,425. Total folate intake had no statistically significant association with postmenopausal breast cancer, with an OR of 0.97 (95 % CI 0.84–1.12). There was significant heterogeneity (I2 = 67.5 %; p = 0.01) which was not removed on sensitivity analysis. Meta-regression was significant for an association with high total folate intake being more protective as the absolute level of low total folate intake decreased (p \ 0.01), but was not associated with the absolute level of high total folate intake (p = 0.18), or the difference between the two levels (p = 0.48). The meta-analysis of total folate intake and premenopausal breast cancer risk included 2 studies (1 prospective study and 1 retrospective study) comprising 1,600 cases with a total sample size of 92,682. Total folate intake had no statistically significant association with premenopausal breast cancer, with an OR of 1.1 (95 % CI 0.92–1.31). There was no significant heterogeneity (I2 = 0.0 %; p = 0.93), and publication bias could not be assessed due to an insufficient number of studies. Blood folate levels Figure 4 shows the meta-analysis of blood folate levels and breast cancer risk, consisting of 7 studies (5 prospective studies and 2 retrospective studies) comprising 2,403 cases with a total sample size of 5,226. There was no significant association between blood folate levels and the risk of breast cancer, with an OR of 0.86 (95 % CI 0.60–1.25). There was significant heterogeneity (I2 = 70.3 %; p \ 0.01), which became insignificant with the removal of Beilby et al. [20] (I2 = 48.7 %; p = 0.08). The result remained non-significant with Beilby et al. [20] removed, with an OR of 1.01 (95 % CI 0.76–1.33). Hormone receptor subtypes Table 2 shows the results for dietary folate intake, total folate intake, and blood folate levels for breast cancer risk
123
1
2 5 Harmful
10
by hormonal receptor subtype. Sensitivity analysis of the heterogeneous results showed that Zhang et al. [59] contributed significantly to the heterogeneity of the dietary folate and estrogen receptor positive, progesterone receptor positive, and estrogen receptor positive/progesterone receptor positive breast cancer risks. Maruti et al. [40] contributed significantly to the heterogeneity of the total folate and estrogen receptor negative breast cancer risk. Removal of these studies resulted in homogeneity, with no substantial change to the overall risk (data not shown).
Discussion Our systematic review and meta-analysis show that dietary folate intake, total folate intake, and folate blood levels are all unlikely to have any significant association with the risk of breast cancer. Although our analysis of high dietary folate intake and breast cancer was significantly associated with a decreased risk of breast cancer, this result was highly heterogeneous, and become non-significant in the majority of subsequent subgroup analyses. Given that the only significantly positive analyses were those stratified by retrospective study design, it is likely that the results were influenced by methodological biases, such as selection bias and recall bias, rather than representing an accurate association. None of the analyses of total folate and breast cancer risk found any statistically significant associations, although meta-regression of the heterogeneous postmenopausal breast cancer risk analysis did find a significant trend between the absolute level of lowest folate intake level, and an increasing risk of breast cancer. This is suggestive that perhaps those at the lowest levels of total folate intake may indeed have a higher risk of postmenopausal breast cancer, but that our analyses could not detect this due to the majority of studies having folate intakes that were too high. It is not clear, however, as to why this trend was seen in the total folate intake analysis, but not the
Breast Cancer Res Treat (2014) 145:513–524 Fig. 4 Blood folate level and risk of breast cancer
521 Study name
Statistics for each study
Odds ratio Beilby et al (2004) 0.23 Chou et al (2007) 0.64 Ericson et al (2009) 1.20 Lin et al (2008) 1.42 Wu et al (1999) a 0.93 Wu et al (1999) b 1.27 Zhang et al (2003) 0.73 0.86
Lower limit 0.09 0.35 0.84 1.00 0.43 0.53 0.50 0.60
Upper limit 0.56 1.17 1.71 2.02 2.03 3.04 1.07 1.25
Odds ratio and 95% CI p-Value 0.00 0.15 0.31 0.05 0.86 0.59 0.10 0.44 0.01 0.1 1 10 100 Protective Harmful
Meta Analysis
Table 2 Folate and risk of breast cancer by hormonal receptor status Hormone receptor subtype
Dietary folate Included study references
Results
Included study references
Results
Included study references
Results
ER?
[34], [39], [44], [46], [58], [62]
OR 0.91 (95 % CI 0.77–1.08); I2 = 67.8 %, p \ 0.01; Eggers p = 0.08
[40], [44], [46], [58], [64]
OR 1.00 (95 % CI 0.97–1.04); I2 = 0.0 %, p = 0.69; Eggers p = 0.10
[26], [37]
OR 1.59 (95 % CI 1.19–2.12); I2 = 0.0 %, p = 0.45
ER-
[22], [34], [39], [44], [46], [50], [58], [59], [62]
OR 0.96 (95 % CI 0.83–1.11); I2 = 8.95 %, p = 0.36; Eggers p = 0.45
[22], [40], [44], [46], [58], [64]
OR 0.93 (95 % CI 0.82–1.05); I2 = 60.5 %, p = 0.02; Eggers p = 0.06
[26], [37]
OR 1.02 (95 % CI 0.52-2.00); I2 = 0.0 %, p = 0.48
PR?
[44], [46], [58], [59]
OR 0.81 (95 % CI 0.53–1.24); I2 = 86.2 %, p \ 0.01; Eggers p = 0.42
[44], [46], [58]
OR 1.01 (95 % CI 0.97–1.04); I2 = 0.0 %, p = 0.69; Eggers p = 0.32
–
–
PR-
[34], [39], [44], [46], [58], [59]
OR 1.01 (95 % CI 0.90–1.13); I2 = 0.0 %, p = 0.80; Eggers p = 0.38
[44], [46], [58]
OR 1.00 (95 % CI 0.94–1.05); I2 = 0.0 %, p = 0.57; Eggers p = 0.24
–
–
ER?/ PR?
[34], [39], [44], [50], [59], [61]
OR 0.92 (95 % CI 0.69–1.22); I2 = 78.8 %, p \ 0.01; Eggers p = 0.46
–
–
–
–
ER?/ PR-
[34], [39], [44], [59]
OR 0.89 (95 % CI 0.72–1.10); I2 = 0.0 %, p = 0.68; Eggers p = 0.97
[34], [44]
OR 0.83 (95 % CI 0.68–1.02); I2 = 0.0 %, p = 0.71
–
–
ER-/ PR?
[44], [59]
OR 0.49 (95 % CI 0.17–1.42); I2 = 55.9 %, p = 0.13
–
–
–
–
ER-/ PR-
[34], [39], [44], [50], [59], [61]
OR 0.97 (95 % CI 0.81–1.17); I2 = 0.0 %, p = 0.82; Eggers p = 0.21
–
–
–
–
Total folate
dietary folate intake analysis, particularly since a number of studies reporting both total and dietary folate intake were included in both analyses. Given the number of subgroup analyses undertaken in our systematic review, one possibility is that this represents a false positive result secondary to multiple testing. With the prevalence of dietary supplement use approaching 50 % in developed countries [65, 66], further
Blood folate
studies incorporating total folate use rather than dietary folate alone are needed. This is particularly evident given that the majority of studies on folate intake and breast cancer risk have been performed in developed countries. We similarly found no association between blood folate levels and the risk of overall breast cancer. An elevated risk for estrogen receptor positive breast cancer with high folate levels was seen on subgroup analysis, but with only 2 studies
123
522
Breast Cancer Res Treat (2014) 145:513–524
available to pool, robust conclusions cannot be made. Further research on whether folate blood levels influence the risk of breast cancer is needed. A lack of significant association between dietary and total folate intake and breast cancer risk remained in all of the hormonal receptor subtype analyses, although the inclusion of Zhang et al. [59] in the dietary folate analyses caused significant heterogeneity, as it reported a substantially decreased risk of breast cancer in comparison to the other included studies. Zhang et al. [59] is one of only 4 studies [28, 49, 50] included in our systematic review that originates from China, and it is notable that the other three Chinese studies reported either statistically significant decreased risks of breast cancer [28, 48], or a non-statistically significant decreased risk with a comparatively lower point estimate [50]. These studies also had lower folate intakes compared to the majority of the other included studies, and are suggestive that perhaps very low intakes of folate may increase breast cancer risk. Further research in developing countries and populations that include lower average folate intakes is needed to further clarify this. Our systematic review had a number of strengths. This is the first systematic review to report on folate intake and breast cancer by hormone receptor status. We followed the MOOSE guidelines in reporting our review. Multiple databases were searched, and were not limited by language. The subgroup analyses were all pre-specified. There were also a number of limitations. As the majority of included studies were primarily from developed countries, firm conclusions about populations in developing countries could not be made. Few studies on blood folate levels were available to analyze. We were also unable to compare risks before and after folate fortification in US-based studies.
Conclusions Folate intake and folate blood levels do not appear to be associated with breast cancer risk at the level of intake encountered in developed countries. This lack of association remained after stratification by menopausal status and hormone receptor status. An association with very low levels of folate intake cannot be ruled out, but given the scarcity of such studies; further research examining populations with very low folate intakes will be needed.
Conflicts of interest interest.
The authors have declared no conflicts of
References 1. Jemal A, Bray F, Center MM et al (2011) Global cancer statistics. CA Cancer J Clin 61(2):69–90
123
2. Seitz HK, Pelucchi C, Bagnardi V et al (2012) Epidemiology and pathophysiology of alcohol and breast cancer: update 2012. Alcohol Alcohol 47(3):204–212 3. Trichopoulos D, Adami HO, Ekbom A et al (2008) Early life events and conditions and breast cancer risk: from epidemiology to etiology. Int J Cancer 122(3):481–485 4. Blount BC, Mack MM, Wehr CM et al (1997) Folate deficiency causes uracil misincorporation into human DNA and chromosomal breakage; implications for cancer and neuronal damage. Proc Natl Acad Sci USA 94(7):3290–3295 5. Stempak JM, Sohn KY, Chiang EP et al (2005) Cell and stage of transformation-specific effects of folate deficiency on methionine cycle intermediates and DNA methylation in an in vitro model. Carcinogenesis 26(5):981–990 6. Wien TN, Pike E, Wisloff T et al (2012) Cancer risk with folic acid supplements: a systematic review and meta-analysis. BMJ Open 12(1):e000653 7. Ly A, Lee H, Chen J et al (2011) Effect of maternal and postweaning folic acid supplementation on mammary tumor risk in the offspring. Cancer Res 71(3):988–997 8. Kotsopoulos J, Medline A, Renlund R et al (2005) Effects of dietary folate on the development and progression of mammary tumors in rats. Carcinogenesis 26(9):1603–1612 9. Stevens VL, McCullough ML, Sun J et al (2010) Folate and other one-carbon metabolism-related nutrients and risk of postmenopausal breast cancer in the Cancer Prevention Study II Nutrition Cohort. Am J Clin Nutr 91(6):1708–1715 10. Ericson U, Sonestedt E, Gullberg B et al (2007) High folate intake is associated with lower breast cancer incidence in postmenopausal women in the Malmo Diet and Cancer cohort. Am J Clin Nutr 86(2):434–443 11. Larsson SC, Giovannucci E, Wolk A (2007) Folate and risk of breast cancer: a meta-analysis. J Natl Cancer Inst 99(1):64–76 12. Lewis SJ, Harbord RM, Harris R et al (2006) Meta-analyses of observational and genetic association studies of folate intakes or levels and breast cancer risk. J Natl Cancer Inst 98(22):1607–1622 13. Stroup DF, Berlin JA, Morton SC et al (2000) Metaanalysis of Observational Studies in Epidemiology (MOOSE) Group. Metaanalysis of observational studies in epidemiology: a proposal for reporting. JAMA 283(15):2008–2012 14. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7(3):177–188 15. Higgins JP, Thompson SG, Deeks JJ et al (2003) Measuring inconsistency in meta-analyses. BMJ 327(7414):557–560 16. Egger M, Davey Smith G, Schneider M et al (1997) Bias in metaanalysis detected by a simple, graphical test. BMJ 315(7109): 629–634 17. Orwin R (1983) A fail safe N for effect size in meta-analysis. J Educ Stat 8:157–159 18. Adzersen KH, Jess P, Freivogel KW et al (2003) Raw and cooked vegetables, fruits, selected micronutrients, and breast cancer risk: a case control study in Germany. Nutr Cancer 46(2):131–137 19. Aune D, Deneo-Pellegrini H, Ronco AL et al (2011) Dietary folate intake and the risk of 11 types of cancer: a case control study in Uruguay. Ann Oncol 22(2):444–451 20. Beilby J, Ingram D, Hahnel R et al (2004) Reduced breast cancer risk with increasing serum folate in a case-control study of the C677T genotype of the methylenetetrahydrofolate reductase gene. Eur J Cancer 40(8):1250–1254 21. Chen J, Gammon MD, Chan W et al (2005) One-carbon metabolism, MTHFR polymorphisms, and risk of breast cancer. Cancer Res 65(4):1606–1614 22. Cho E, Holmes M, Hankinson SE et al (2007) Nutrients involved in one-carbon metabolism and risk of breast cancer among premenopausal women. Cancer Epidemiol Biomarkers Prev 16(12):2787–2790
Breast Cancer Res Treat (2014) 145:513–524 23. Chou YC, Lee MS, Wu MH et al (2007) Plasma homocysteine as a metabolic risk factor for breast cancer: findings from a case-control study in Taiwan. Breast Cancer Res Treat 101(2):199–205 24. Duffy CM, Assaf A, Cyr M et al (2009) Alcohol and folate intake and breast cancer risk in the WHI Observational Study. Breast Cancer Res Treat 116(3):551–562 25. Ericson UC, Ivarsson MI, Sonestedt E et al (2009) Increased breast cancer risk at high plasma folate concentrations among women with the MTHFR 677T allele. Am J Clin Nutr 90(5): 1380–1389 26. Ericson U, Borqquist S, Ivarsson MI et al (2010) Plasma folate concentrations are positively associated with risk of estrogen receptor beta negative breast cancer in a Swedish nested case control study. J Nutr 140(9):1661–1668 27. Freudenheim JL, Marshall JR, Vena JE et al (1996) Premenopausal breast cancer risk and intake of vegetables, fruits, and related nutrients. J Natl Cancer Inst 88(6):340–348 28. Gao CM, Tang JH, Cao HX et al (2009) MTHFR polymorphisms, dietary folate intake and breast cancer risk in Chinese women. J Hum Genet 54(7):414–418 29. Graham S, Hellmann R, Marshall J et al (1991) Nutritional epidemiology of postmenopausal breast cancer in western New York. Am J Epidemiol 134(6):552–566 30. Kabat GC, Miller AB, Jain M et al (2008) Dietary intake of selected B vitamins in relation to risk of major cancers in women. Br J Cancer 99(5):816–821 31. Kotsopoulos J, Sukiennicki G, Muszynska M et al (2012) Plasma micronutrients, trace elements, and breast cancer in BRCA1 mutation carriers: an exploratory study. Cancer Causes Control 23(7):1065–1074 32. Lajous M, Lazcano-Ponce E, Hernandez-Avila M et al (2006) Folate, vitamin B6, and vitamin B12 intake and the risk of breast cancer. Cancer Epidemiol Biomarkers Prev 15(3):443–448 33. Lajous M, Romieu I, Sabia S et al (2006) Folate, vitamin B12 and postmenopausal breast cancer in a prospective study of French women. Cancer Causes Control 17(9):1209–1213 34. Larsson SC, Berqkvist L, Wolk A (2008) Folate intake and risk of breast cancer by estrogen and progesterone receptor status in a Swedish cohort. Cancer Epidemiol Biomarkers Prev 17(12): 3444–3449 35. Lee SA, Lee KM, Lee SJ et al (2010) Antioxidant vitamins intake, ataxia telangiectasia mutated (ATM) genetic polymorphisms, and breast cancer risk. Nutr Cancer 62(8):1087–1094 36. Levi F, Pasche C, Lucchini F et al (2001) Dietary intake of selected micronutrients and breast-cancer risk. Int J Cancer 91(2):260–263 37. Lin J, Lee IM, Cook NR et al (2008) Plasma folate, vitamin B-6, vitamin B-12, and risk of breast cancer in women. Am J Clin Nutr 87(3):734–743 38. Ma E, Iwasaki M, Junko I et al (2009) Dietary intake of folate, vitamin B6, and vitamin B12, genetic polymorphism of related enzymes, and risk of breast cancer: a case-control study in Brazilian women. BMC Cancer 9:122 39. Ma E, Iwasaki M, Kobayashi M et al (2009) Dietary intake of folate, vitamin B2, vitamin B6, vitamin B12, genetic polymorphism of related enzymes, and risk of breast cancer: a casecontrol study in Japan. Nutr Cancer 61(4):447–456 40. Maruti SS, Ulrich CM, White E (2009) Folate and one-carbon metabolism nutrients from supplements and diet in relation to breast cancer risk. Am J Clin Nutr 89(2):624–633 41. Negri E, La Vecchia C, Franceschi S (2000) Re: dietary folate consumption and breast cancer risk. J Natl Cancer Inst 92(15): 1270–1271 42. Potischman N, Swanson CA, Coates RJ et al (1999) Intake of food groups and associated micronutrients in relation to risk of early-stage breast cancer. Int J Cancer 82(3):315–321
523 43. Rohan TE, Jain MG, Howe GR et al (2000) Dietary folate consumption and breast cancer risk. J Natl Cancer Inst 92(3):266–269 44. Roswall N, Olsen A, Christensen J et al (2010) Micronutrient intake and breast cancer characteristics among postmenopausal women. Eur J Cancer Prev 19(5):360–365 45. Sellers TA, Kushi LH, Cerhan JR et al (2001) Dietary folate intake, alcohol, and risk of breast cancer in a prospective study of postmenopausal women. Epidemiology 12(4):420–428 46. Sellers TA, Vierkant RA, Cerhan JR et al (2002) Interaction of dietary folate intake, alcohol, and risk of hormone receptordefined breast cancer in a prospective study of postmenopausal women. Cancer Epidemiol Biomarkers Prev 11:1104–1107 47. Sellers TA, Gabrick DM, Vierkant RA et al (2004) Does folate intake decrease risk of postmenopausal breast cancer among women with a family history? Cancer Causes Control 15(2): 113–120 48. Sharp L, Little J, Schofield AC et al (2002) Folate and breast cancer: the role of polymorphisms in methylenetetrahydrofolate reductase (MTHFR). Cancer Lett 181(1):65–71 49. Shrubsole MJ, Jin F, Dai Q et al (2001) Dietary folate intake and breast cancer risk: results from the Shanghai Breast Cancer Study. Cancer Res 61(19):7136–7141 50. Shrubsole MJ, Shu XO, Li HL et al (2011) Dietary B vitamin and methionine intakes and breast cancer risk among Chinese women. Am J Epidemiol 173(10):1171–1182 51. Stolzenberg-Solomon RZ, Chang SC, Leitzmann MF et al (2006) Folate intake, alcohol use, and postmenopausal breast cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Am J Clin Nutr 83(4):895–904 52. Suzuki T, Matsuo K, Hirose K et al (2008) One-carbon metabolism-related gene polymorphisms and risk of breast cancer. Carcinogenesis 29(2):356–362 53. Thorand B, Kohlmeier L, Simonsen N et al (1998) Intake of fruits, vegetables, folic acid, and related nutrients and risk of breast cancer in postmenopausal women. Public Health Nutr 1(3):147–156 54. Wu K, Helzlsouer KJ, Comstock GW et al (1999) A prospective study on folate, B12, and pyridoxal 50 -phosphate (B6) and breast cancer. Cancer Epidemiol Biomarkers Prev 8(3):209–217 55. Yang YJ, Hwang SH, Kim HJ et al (2010) Dietary intake of nitrate relative to antioxidant vitamin in relation to breast cancer risk: a case-control study. Nutr Cancer 62(5):555–566 56. Zhang S, Hunter DJ, Hankinson SE et al (1999) A prospective study of folate intake and the risk of breast cancer. JAMA 281(17):1632–1637 57. Zhang SM, Willett WC, Selhub J et al (2003) Plasma folate, vitamin B6, vitamin B12, homocysteine, and risk of breast cancer. J Natl Cancer Inst 95(5):373–380 58. Zhang SM, Hankinson SE, Hunter DJ et al (2005) Folate intake and risk of breast cancer characterized by hormone receptor status. Cancer Epidemiol Biomarkers Prev 14(8):2004–2008 59. Zhang CX, Ho SC, Chen YM et al (2011) Dietary folate, vitamin B6, vitamin B12, and methionine intake and the risk of breast cancer by oestrogen and progesterone receptor status. Br J Nutr 106(6):936–943 60. Zhu K, Davidson NE, Hunter S et al (2003) Methyl-group dietary intake and risk of breast cancer among African-American women: a case-control study by methylation status of the estrogen receptor alpha genes. Cancer Causes Control 14(9):827–836 61. Bassett JK, Baglietto L, Hodge AMm et al (2013) Dietary intake of B vitamins and methionine and breast cancer risk. Cancer Causes Control 24(8):1555–1563 62. Gong Z, Ambrosone CB, McCann SE et al (2014) Associations of dietary folate, Vitamins B6 and methionine intake with risk of breast cancer among African American and European American women. Int J Cancer 134(6):1422–1435
123
524 63. Islam T, Ito H, Sueta A et al (2013) Alcohol and dietary folate intake and the risk of breast cancer: a case-control study in Japan. Eur J Cancer Prev 22(4):358–366 64. Yang D, Baumgartner RN, Slattery ML et al (2013) Dietary intake of folate, B-vitamins and methionine and breast cancer risk among Hispanic and non-Hispanic white women. PLoS One 8(2):e54495
123
Breast Cancer Res Treat (2014) 145:513–524 65. Bailey RL, Gahche JJ, Lentino CV, Dwyer JT, Engel JS, Thomas PR et al (2011) Dietary supplement use in the United States, 2003–2006. J Nutr 141(2):261–266 66. Skeie G, Braaten T, Hjartaker A, Lentjes M, Amiano P, Jakszyn P et al (2009) Use of dietary supplements in the European Prospective Investigation into Cancer and Nutrition calibration study. Eur J Clin Nutr 63(S4):S226–S238
