Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Oct 1.
Published in final edited form as: Cancer Epidemiol Biomarkers Prev. 2022 Apr 1;31(4):839–850. doi: 10.1158/1055-9965.EPI-21-1023

Plasma Metabolomics and Breast Cancer Risk Over 20 Years of Follow-up Among Postmenopausal Women in the Nurses’ Health Study

Kristen D Brantley 1, Oana A Zeleznik 2, Bernard Rosner 2,3, Rulla M Tamimi 4, Julian Avila-Pacheco 5, Clary B Clish 5, A Heather Eliassen 1,2,6
PMCID: PMC8983458  NIHMSID: NIHMS1773833  PMID: 35064065

Abstract

Background

Metabolite profiles provide insight into biologic mechanisms contributing to breast cancer development. We explored the association between pre-diagnostic plasma metabolites (N=307) and invasive breast cancer among postmenopausal women in a nested case-control study within the Nurses’ Health Study (N=1,531 matched pairs).

Methods

Plasma metabolites were profiled via liquid chromatography tandem mass spectrometry using samples taken ≥10 years (distant, N=939 cases), and <10 years (proximate, N=592 cases) before diagnosis. Multivariable conditional logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) comparing the 90th to 10th percentile of individual metabolite level, using the number of effective tests (NEF) to account for testing multiple correlated hypotheses. Associations of metabolite groups with breast cancer were evaluated using metabolite set enrichment analysis (MSEA) and weighted gene co-expression network analysis (WGCNA), with adjustment for the false discovery rate.

Results

No individual metabolites were significantly associated with breast cancer risk. MSEA showed negative enrichment of cholesteryl esters at the distant timepoint (normalized enrichment score (NES)=−2.26, padj=0.02). Positive enrichment of triacylglycerols (TAGs) with <3 double bonds was observed at both timepoints. TAGs with ≥3 double bonds were inversely associated with breast cancer at the proximate timepoint (NES=−2.91, padj=0.03).

Conclusion

Cholesteryl esters measured earlier in disease etiology were inversely associated with breast cancer. TAGs with many double bonds measured closer to diagnosis were inversely associated with breast cancer risk.

Impact

The discovered associations between metabolite subclasses and breast cancer risk can expand our understanding of biochemical processes involved in cancer etiology.

INTRODUCTION

Metabolite profiles reflect the integrated impact of the genome and exogenous exposures on the metabolic state and may provide insight into biologic mechanisms contributing to disease development. Breast cancer is the most common cancer among women worldwide.1 While key sex hormone related metabolic pathways are well-established in breast cancer etiology, knowledge on metabolic pathways in aggregate may reveal additional targets for prevention.

A handful of recent studies have explored metabolite associations with breast cancer incidence,2–9 though only a few have taken an agnostic approach to explore the metabolomics of breast cancer,3,8,9 instead focusing on weight-associated or nutritional metabolites. Among the studies that have explored metabolites overall with respect to breast cancer risk, one had a very small sample size (N=84 cases),8 and all used different metabolomic platforms for measurement. All studies thus far have only captured metabolite profiles at a single point in time. Previous studies suggest inverse associations between carnitines3,9 and phosphatidylcholines9 and breast cancer risk, and positive associations between amino acids and breast cancer risk,5 though importance of individual metabolites varied by study.

Here we used an agnostic approach in the Nurses’ Health Study to investigate associations between metabolite levels, measured prior to diagnosis, and future breast cancer risk. We also examined how these measures changed over time, using measures from two different blood collections, approximately 10 years apart.

MATERIAL & METHODS

Cohort

We conducted a nested case-control study within the Nurses’ Health Study (NHS), a prospective cohort of 121,700 female nurses started in 1976. Biennial follow up questionnaires collect risk factor information as well as new disease diagnoses. Blood samples were collected in 1989–1990 from 32,826 cohort members, aged 43–69 years at blood collection. A subset of these women (N=18,743) provided a second blood sample between 2000–2002. The study protocol was approved by the institutional review boards of the Brigham and Women’s Hospital and Harvard T.H. Chan School of Public Health, and those of participating registries as required.

Breast cancer cases were identified by self-report and confirmed by medical record review. Deaths were captured by next of kin, postal service, or review of the National Death Index. Cases were all women diagnosed with invasive or in situ breast cancer between 2000–2010 who provided a blood sample (N=939 for distant 1989–90 blood collection; N=592 for proximate 2000–02 blood collection) and had no prior reported cancer (other than non-melanoma skin). All those with proximate blood samples also had distant blood measures. Controls were matched to cases on factors at each blood draw, including age (+/− 1 year), month (±1 month), time of day (±2 h), fasting status (≥10 h since a meal vs.<10 h or unknown), and combined menopausal status and postmenopausal hormone use (premenopausal/postmenopausal, not on hormones//postmenopausal, on hormones, unknown).

Metabolite Profiling

Plasma metabolites were profiled at the Broad Institute of MIT and Harvard (Cambridge, MA). Two liquid chromatography tandem mass spectrometry (LC/MS-MS) platforms were used for identification of metabolites, designed to measure polar metabolites and lipids, and free fatty acids, described elsewhere.10–13 Specifics on measurement procedures are described in a previous publication.14 Briefly, matched case-control pairs were distributed randomly within batch, pooled reference samples were included every 20 samples, and 64 quality controls were randomly distributed. Measures were standardized using the ratio of the value of the sample to the value of the nearest pooled reference multiplied by the median of all reference values for the metabolite. For metabolites measured with multiple metabolomics platforms, the assay laboratory provided a list of the preferred measurement platform. For metabolites measured multiple times with the same platform, the metabolite with the lowest CV was used for analysis. Metabolites that had poor stability due to delay in processing were excluded (N=51).13 Following this initial data cleaning, a total of 307 known metabolites were successfully measured and included in the study. Metabolites were annotated by superclass, class, and subclass distinctions.

Covariates

Identified risk factors for breast cancer were included as covariates in the analyses: BMI at age 18 (kg/m2), weight change since age 18 (kg), age at first birth and parity (nulliparous, 1–2 kids <25 y, 1–2 kids 25+ y, 3+ kids <25y, 3+ kids 25+ y), age at menarche (years), breastfeeding history (yes/no), history of benign breast disease (yes/no), family history of breast cancer (yes/no), physical activity (MET-hours/week), and alcohol intake (g/day).

Statistical Analysis

Metabolites with <10% missing were imputed with ½ the minimum value (N=39 at distant blood collection, N=0 at proximate blood collection). Metabolites with ≥10% missing (N=15 at distant blood collection, N=16 at proximate blood collection) were not imputed. We used probit transformation for all metabolites. We used multivariable conditional logistic regression (CLR) to calculate odds ratios (OR) and 95% confidence intervals (CI) for individual metabolites with breast cancer at both distant and proximate blood collections. Unconditional logistic regression (UCLR) with adjustment for matching factors was used for estrogen receptor positive (ER+) and negative (ER-) breast cancers due to limited ER− cases. ORs represent a 2.5 standard deviation (SD) increase in metabolites, equivalent to the comparison for 90th-10th percentile of metabolite value under the assumption of a normal distribution.

We accounted for testing for multiple correlated hypotheses by calculating the number of effective tests by performing a principal components (PC) analysis of all metabolites among controls and calculating the number of PCs that explained 99.5% of the total variance.15 For this method, padj=punadjusted/number of effective tests (padj distant=0.0003, padj proximate=0.0002).

In a separate analysis, we explored the association of presence vs. absence of metabolites with ≥10% missingness with breast cancer risk.

Correlations between metabolite measurements at distant and proximate timepoints were assessed using unadjusted Spearman correlations, and adjusted for fasting, age at blood draw, and weight change since age 18. We used unconditional logistic regression models including metabolite measures at both timepoints in the same regression along with an interaction term; the p-value for the interaction term was used to determine potential interest in the difference measures.

The difference of metabolite levels was analyzed via unconditional logistic regression, with ORs representing comparison of the 90th – 10th percentile metabolite level from distant to proximate blood, adjusted for distant blood. For average and difference analyses, fasting status and menopausal status were assessed as a combination of the two timepoints. The remaining covariates were from the proximate blood collection.

Metabolites were grouped based on structural similarities by subclasses; triacylglycerols (TAGs) were further divided as TAGs with ≥3 vs. TAGs with <3 double bonds. Metabolite set enrichment analysis (MSEA) combines the effect estimates from logistic regressions performed on individual metabolites by defined groups, to determine a summary Enrichment Score (ES) and Normalized Enrichment Score (NES) adjusted for group size.16 The ES represents the degree to which the metabolite set is overrepresented compared to other sets; where a positive ES represents a significant positive enrichment in breast cancer, while a significant negative score indicates a group that is negatively enriched in breast cancer. P-values were adjusted using the False Discovery Rate (FDR) to account for multiple comparisons.17

Weighted gene (metabolite) co-expression network analysis (WGCNA) was used to identify metabolite modules associated with breast cancer risk. This analysis process is described in detail elsewhere.18 Briefly, a co-expression network is constructed using the absolute values of the correlation coefficients between metabolites to identify interconnected “nodes” based on a threshold value of similarity. Hierarchical clustering based on scale-free topology identifies densely interconnected metabolites from the network, and modules are grouped by using a Dynamic Tree Cut method.19 Within each analysis, all metabolites were assigned a module score, derived in control subjects at each timepoint separately, based on their loading on the first principal component of each module. Module scores were then included in UCLR models for breast cancer risk. The resulting OR represents the association of a particular module with breast cancer risk. The loading status of individual metabolites into each module was examined to determine the influence of individual metabolites on the resultant module association.18

Datasets for analysis were created in SAS version 9 (SAS Institute Inc., Cary, NC, USA). All analyses were conducted using R programming language, version 4.0.3.

RESULTS

A total of 939 cases and 939 matched controls were included for distant blood collection analysis, and 592 cases and 592 controls were included for the proximate blood collection. At first blood collection, mean age was 55 years (SD=6.9); 25% of women were premenopausal (Table 1). At the second blood draw, 98% of women were postmenopausal. Family history of breast cancer, particularly at second collection, was higher among cases (23%) compared to controls (15%). As expected, weight gain since age 18 was higher at the second blood draw; at both timepoints, cases tended to have approximately 2kg more weight gain compared to controls.

Table 1.

Descriptive characteristics of participants in NHS who provided blood samples at distant and proximate dates.^

Distant Blood Proximate Blood
Characteristic Case (N=939) Control (N=939) Case (N=592) Control (N=592)
Age at blood draw (mean (SD)) 55.5 (6.9) 55.6 (6.9) 66.4 (6.9) 66.5 (6.8)
Fasting at blood draw (N (%)) 626 (67%) 683 (73%) 515 (87%) 547 (92%)
Menopausal status & PMH use at blood draw (N (%))
 Premenopausal 239 (26%) 240 (26%) 3 (1%) 5 (1%)
 Postmenopausal, no PMH use 288 (31%) 289 (31%) 188 (32%) 186 (31%)
 Postmenopausal, PMH use 293 (31%) 292 (31%) 393 (66%) 395 (67%)
 Unknown 0 0 8 (1%) 6 (1%)
Age at menarche (mean (SD)) 12.5 (1.4) 12.6 (1.4) 12.5 (1.4) 12.6 (1.4)
Nulliparous (N (%)) 90 (10%) 75 (8%) 51 (9%) 35 (6%)
Parity (mean (SD))† 3.1 (1.4) 3.2 (1.6) 3.1 (1.3) 3.2 (1.6)
Age at first birth (mean (SD))† 25 (3.1) 25 (3.1) 24.9 (3.1) 24.7 (3.0)
Breastfeeding history (N (%))† 604 (64%) 583 (62%) 399 (67%) 381 (64%)
History of benign breast disease (N (%)) 492 (52%) 430 (46%) 383 (65%) 346 (56%)
Family history of breast cancer (N (%)) 136 (15%) 101 (11%) 135 (23%) 87 (15%)
Weight change from age 18 to blood draw in kg (mean (SD)) 12.3 (10.9) 10.6 (11.2) 15.1 (12.8) 13.5 (12.8)
BMI at blood draw kg/m2 (mean (SD)) 25.7 (4.3) 25.2 (4.7) 26.7 (5.0) 26.4 (5.1)
Average alcohol consumption at blood draw in g/day (mean (SD)) 7.0 (9.9) 5.9 (8.2) 6.7 (9.2) 5.8 (7.7)
Activity level at blood draw in MET-hours/week (mean (SD)) 15.4 (18.8) 15.9 (17.6) 25.7 (42.0) 23.4 (31.7)
Open in a new tab
^

Distant blood draw was >10 years before diagnosis date for cases. Proximate blood draw was <=10 years before diagnosis date for cases.

†

Among parous women

No individual metabolites at either distant or proximate timepoints were significantly associated with breast cancer risk after adjusting for the number of effective tests (NEF distant=193, proximate=186, padj distant=0.0003, proximate=0.0002). Despite the lack of significance at this level, several metabolites and metabolite classes stood out as nominally significant (Table 2, Tables S1a & S1b). The amino acid phenylalanine was positively associated with breast cancer risk at both distant (OR=1.41, 95% CI=1.08–1.85; nominal p-value=0.01), and proximate timepoints (OR=1.76, 95% CI=1.25–2.48; nominal p-value=0.001). Similar positive associations at both timepoints were observed for the amino acid proline. We observed strong positive associations for TAGs with <3 double bonds at the distant timepoint, (e.g.: C51:0 TAG OR=1.30, 95% CI=1.01–1.68; nominal p-value=0.04). At the proximate timepoint, several TAGs with high numbers of double bonds were inversely associated with breast cancer risk (e.g.: C54:9 TAG OR=0.64, 95% CI=0.47–0.87; nominal p-value=0.005).

Table 2.

Odds ratios for breast cancer risk comparing 90th to 10th percentiles of selected metabolites^ measured at distant or proximate blood.

Unadjusted Multivariate Adjusted†
Metabolite Name HMDB ID Class Sub Class OR (95% CI) p value OR (95% CI) p value
Distant Blood
phenylalanine HMDB0000159 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.50 (1.17–1.94) 0.002 1.41 (1.08–1.85) 0.012
Proline HMDB0000162 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.37 (1.07–1.75) 0.012 1.33 (1.03–1.72) 0.032
homoarginine HMDB0000670* Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.41 (1.11–1.80) 0.005 1.3 (1.01–1.68) 0.039
Lysine HMDB0000182 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.38 (1.08–1.77) 0.011 1.31 (1.01–1.69) 0.040
C5:1 carnitine HMDB0002366 Fatty Acyls Fatty acid esters 0.80 (0.64–1.01) 0.064 0.73 (0.57–0.93) 0.010
C5-DC carnitine HMDB0013130 Fatty Acyls Fatty acid esters 0.72 (0.57–0.92) 0.007 0.73 (0.57–0.93) 0.012
C51:0 TAG HMDB0031106* Glycerolipids Triacylglycerols 1.46 (1.15–1.86) 0.002 1.30 (1.01–1.68) 0.044
C22:5 LPC HMDB0010403* Glycerophospholipids Glycerophosphocholines 0.78 (0.61–0.99) 0.041 0.78 (0.60–1.00) 0.047
C22:0 LPE HMDB0011520 Glycerophospholipids Glycerophosphoethanolamines 0.69 (0.54–0.89) 0.004 0.75 (0.58–0.98) 0.035
C38:6 PE plasmalogen HMDB0011387* Glycerophospholipids Glycerophosphoethanolamines 0.82 (0.65–1.03) 0.088 0.78 (0.61–0.99) 0.039
Thyroxine HMDB0000248 NA NA 1.50 (1.16–1.95) 0.002 1.56 (1.19–2.05) 0.001
acetyl-galactosamine HMDB0000212 Organooxygen compounds Carbohydrates and carbohydrate conjugates 1.42 (1.1–1.84) 0.008 1.35 (1.02–1.77) 0.035
2-methylguanosine HMDB0005862 Purine nucleosides NA 1.38 (1.07–1.77) 0.014 1.32 (1.01–1.72) 0.039
Guanosine HMDB0000133 Purine nucleosides NA 0.77 (0.61–0.97) 0.027 0.78 (0.61–0.99) 0.041
C22:5 CE HMDB0010375* Steroids and steroid derivatives Cholesterol esters 0.61 (0.48–0.77) <0.001 0.67 (0.52–0.86) 0.002
C18:3 CE HMDB0010370* Steroids and steroid derivatives Cholesterol esters 0.70 (0.55–0.88) 0.003 0.69 (0.54–0.89) 0.004
C20:5 CE HMDB0006731 Steroids and steroid derivatives Cholesterol esters 0.75 (0.60–0.95) 0.016 0.74 (0.58–0.95) 0.017
Proximate Blood
2-aminohippuric acid NA** Benzene and substituted derivatives Benzoic acids and derivatives 1.39 (1.01–1.93) 0.046 1.45 (1.02–2.06) 0.038
N1,N12-diacetylspermine HMDB0002172 Carboximidic acids and derivatives Carboximidic acids 1.38 (1.02–1.85) 0.034 1.41 (1.03–1.94) 0.032
phenylalanine HMDB0000159 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.77 (1.29–2.42) <0.001 1.76 (1.25–2.48) 0.001
Proline HMDB0000162 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.52 (1.12–2.07) 0.007 1.59 (1.13–2.22) 0.007
Isoleucine HMDB0000172 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.55 (1.15–2.08) 0.004 1.56 (1.12–2.17) 0.009
Leucine HMDB0000687 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.50 (1.12–2.02) 0.007 1.48 (1.06–2.06) 0.02
N-alpha-acetylarginine HMDB0004620* Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.39 (1.03–1.89) 0.033 1.45 (1.06–2.00) 0.022
Serine HMDB0000187 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.35 (0.99–1.85) 0.058 1.46 (1.05–2.02) 0.023
N-acetylornithine HMDB0003357 Carboxylic acids and derivatives Amino acids, peptides, and analogues 0.78 (0.58–1.03) 0.081 0.71 (0.53–0.96) 0.026
Betaine HMDB0000043 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.27 (0.91–1.79) 0.164 1.47 (1.03–2.12) 0.035
C5-DC carnitine HMDB0013130 Fatty Acyls Fatty acid esters 0.67 (0.5–0.91) 0.010 0.71 (0.52–0.97) 0.030
myristoleic acid HMDB0002000 Fatty Acyls Fatty acids and conjugates 1.5 (1.07–2.1) 0.018 1.58 (1.11–2.24) 0.012
C58:7 TAG HMDB0005471* Glycerolipids Triacylglycerols 0.60 (0.44–0.82) 0.001 0.59 (0.42–0.82) 0.002
C56:9 TAG HMDB0005448* Glycerolipids Triacylglycerols 0.68 (0.51–0.91) 0.010 0.64 (0.46–0.87) 0.004
C56:10 TAG HMDB0010513* Glycerolipids Triacylglycerols 0.69 (0.52–0.93) 0.013 0.63 (0.46–0.86) 0.004
C54:9 TAG HMDB0010498* Glycerolipids Triacylglycerols 0.70 (0.52–0.94) 0.017 0.64 (0.47–0.87) 0.005
C54:8 TAG HMDB0010518* Glycerolipids Triacylglycerols 0.70 (0.52–0.94) 0.017 0.65 (0.47–0.88) 0.006
C58:11 TAG HMDB0010531* Glycerolipids Triacylglycerols 0.70 (0.52–0.94) 0.017 0.64 (0.47–0.88) 0.006
C56:8 TAG HMDB0005392* Glycerolipids Triacylglycerols 0.69 (0.51–0.93) 0.015 0.66 (0.48–0.90) 0.008
C58:9 TAG HMDB0005463* Glycerolipids Triacylglycerols 0.67 (0.50–0.91) 0.010 0.66 (0.47–0.90) 0.010
C58:10 TAG HMDB0005476* Glycerolipids Triacylglycerols 0.69 (0.51–0.93) 0.014 0.66 (0.48–0.90) 0.010
C56:7 TAG HMDB0005462* Glycerolipids Triacylglycerols 0.74 (0.55–0.99) 0.044 0.68 (0.50–0.94) 0.017
C58:6 TAG HMDB0005458* Glycerolipids Triacylglycerols 0.68 (0.49–0.92) 0.013 0.68 (0.49–0.94) 0.018
C52:7 TAG HMDB0010517* Glycerolipids Triacylglycerols 0.76 (0.57–1.02) 0.065 0.69 (0.50–0.94) 0.019
C54:7 TAG HMDB0005447* Glycerolipids Triacylglycerols 0.73 (0.54–0.97) 0.031 0.70 (0.51–0.95) 0.021
C60:12 TAG HMDB0005478* Glycerolipids Triacylglycerols 0.76 (0.56–1.02) 0.071 0.71 (0.51–0.98) 0.035
C52:6 TAG HMDB0005436* Glycerolipids Triacylglycerols 0.79 (0.59–1.06) 0.114 0.73 (0.53–0.99) 0.046
C18:3 LPC HMDB0010387* Glycerophospholipids Glycerophosphocholines 1.40 (1.04–1.90) 0.026 1.40 (1.02–1.93) 0.035
C16:1 LPC HMDB0010383* Glycerophospholipids Glycerophosphocholines 1.39 (1.04–1.87) 0.028 1.39 (1.02–1.89) 0.038
C16:0 LPC HMDB0010382 Glycerophospholipids Glycerophosphocholines 1.40 (1.04–1.89) 0.026 1.38 (1.01–1.89) 0.042
C18:1 LPC HMDB0002815* Glycerophospholipids Glycerophosphocholines 1.32 (0.97–1.79) 0.072 1.39 (1.01–1.93) 0.046
C38:6 PE HMDB0009102* Glycerophospholipids Glycerophosphoethanolamines 0.76 (0.55–1.05) 0.091 0.69 (0.49–0.97) 0.035
Tryptophan HMDB0000929 Indoles and derivatives Indolyl carboxylic acids and derivatives 1.39 (1.04–1.87) 0.028 1.40 (1.03–1.9) 0.030
C16:0 Ceramide (d18:1) HMDB0004949 Sphingolipids Ceramides 1.62 (1.18–2.22) 0.003 1.72 (1.23–2.40) 0.002
C24:1 Ceramide (d18:1) HMDB0004953* Sphingolipids Ceramides 1.46 (1.08–1.98) 0.014 1.42 (1.04–1.94) 0.028
C22:0 Ceramide (d18:1) HMDB0004952 Sphingolipids Ceramides 1.43 (1.06–1.94) 0.020 1.39 (1.01–1.92) 0.044
Open in a new tab
^

Selected metabolites are those with nominal p value <0.05 in fully adjusted models among metabolites with <10% missingness. Missing values were imputed with 1/2 the minimum value. Results sorted by class, subclass & p-value for fully adjusted model. Significant p-value with NEF adjustment: distant blood p-value=0.0003, proximate blood p-value=0.0002.

†

Multivariate conditional logistic regression model adjusted for: BMI at age 18, weight change since age 18, age at menarche, combined age at first birth and parity, breastfeeding history, history of benign breast disease, family history of breast cancer, alcohol use (g/day), activity level (MET-hrs/week). P values are nominal p-values before correction for multiple testing.

*

Representative HMDBID

**

No HMDBID

The majority of metabolites with ≥10% missingness were drug related, and none of these metabolites were associated with breast cancer risk in presence v. absence assessment (Tables S2a & S2b).

While most associations were consistent between ER+ and ER− breast cancer, some metabolites were associated in opposite directions for ER+ vs. ER− breast cancers (Table 3; Tables S3a & S3b (ER+), Tables S4a & S4b (ER−)), though most were not significantly heterogeneous. For example, at the proximate timepoint TAGs with <3 double bonds were strongly positively associated with ER+ breast cancers, but inversely associated with ER− breast cancers (e.g.: C52:0 TAG ER+ OR=1.49, 95% CI=1.04–2.15; nominal p-value=0.03; ER− OR=0.86, 95% CI=0.42–0.74; nominal p-value=0.668, nominal p-value=0.09, p-het=0.25).

Table 3.

Odds ratios for breast cancer risk comparing 90th to 10th percentiles of selected metabolite levels^ measured at distant or proximate blood, by ER status of case.

ER+ (N=585) ER− (N=91)
Metabolite Name HMDB ID Class Subclass OR (95% CI)† p value OR (95% CI)† p value
Distant Blood
hippurate HMDB0000714 Benzene and substituted derivatives Benzoic acids and derivatives 0.67 (0.50–0.90) 0.007 1.02 (0.57–1.82) 0.952
N-alpha-acetylarginine HMDB0004620* Carboxylic acids and derivatives Amino acids, peptides, and analogues 0.70 (0.53–0.94) 0.016 0.69 (0.39–1.23) 0.214
citrulline HMDB0000904 Carboxylic acids and derivatives Amino acids, peptides, and analogues 0.74 (0.55–0.99) 0.045 0.82 (0.46–1.46) 0.496
C5:1 carnitine HMDB0002366 Fatty Acyls Fatty acid esters 0.59 (0.44–0.79) <0.001 1.12 (0.61–2.06) 0.715
C3 carnitine HMDB0000824 Fatty Acyls Fatty acid esters 0.72 (0.54–0.97) 0.029 1.16 (0.64–2.10) 0.621
C4 carnitine HMDB0002013 Fatty Acyls Fatty acid esters 0.69 (0.52–0.92) 0.012 0.91 (0.50–1.64) 0.748
C5-DC carnitine HMDB0013130 Fatty Acyls Fatty acid esters 0.70 (0.53–0.93) 0.015 0.94 (0.52–1.68) 0.823
C34:2 DAG HMDB0007103* Fatty Acyls Lineolic acids and derivatives 1.50 (1.12–2.02) 0.007 1.49 (0.81–2.78) 0.203
C36:3 DAG HMDB0007219* Fatty Acyls Lineolic acids and derivatives 1.35 (1.01–1.80) 0.040 1.29 (0.70–2.39) 0.419
C32:0 DAG HMDB0007098* Glycerolipids Diacylglycerols 1.44 (1.06–1.94) 0.018 1.49 (0.80–2.77) 0.209
C34:1 DAG HMDB0007102* Glycerolipids Diacylglycerols 1.38 (1.03–1.87) 0.034 1.41 (0.75–2.66) 0.284
C52:4 TAG HMDB0005363* Glycerolipids Triacylglycerols 1.43 (1.07–1.90) 0.015 1.22 (0.67–2.25) 0.518
C50:2 TAG HMDB0005377* Glycerolipids Triacylglycerols 1.44 (1.06–1.96) 0.021 1.26 (0.67–2.38) 0.476
C50:1 TAG HMDB0005360* Glycerolipids Triacylglycerols 1.42 (1.04–1.93) 0.027 1.24 (0.66–2.33) 0.508
C52:2 TAG HMDB0005369* Glycerolipids Triacylglycerols 1.38 (1.02–1.87) 0.035 1.25 (0.67–2.36) 0.479
C50:3 TAG HMDB0005433* Glycerolipids Triacylglycerols 1.38 (1.02–1.88) 0.035 1.45 (0.78–2.73) 0.245
C51:1 TAG HMDB0042104* Glycerolipids Triacylglycerols 1.38 (1.02–1.86) 0.036 1.46 (0.79–2.68) 0.227
C43:2 TAG HMDB0043169* Glycerolipids Triacylglycerols 1.37 (1.01–1.85) 0.041 1.46 (0.80–2.67) 0.220
C55:2 TAG HMDB0042226* Glycerolipids Triacylglycerols 1.35 (1.00–1.82) 0.047 1.18 (0.64–2.19) 0.591
C22:5 LPC HMDB0010403* Glycerophospholipids Glycerophosphocholines 0.58 (0.43–0.77) <0.001 0.71 (0.39–1.27) 0.248
C18:2 LPC HMDB0010386* Glycerophospholipids Glycerophosphocholines 0.64 (0.47–0.87) 0.005 0.78 (0.41–1.49) 0.453
C20:5 LPC HMDB0010397 Glycerophospholipids Glycerophosphocholines 0.65 (0.48–0.88) 0.006 0.82 (0.43–1.54) 0.535
C18:1 LPC HMDB0002815* Glycerophospholipids Glycerophosphocholines 0.68 (0.51–0.91) 0.011 0.96 (0.52–1.74) 0.889
C18:0 LPC HMDB0010384 Glycerophospholipids Glycerophosphocholines 0.69 (0.51–0.93) 0.014 1.09 (0.58–2.03) 0.789
C36:5 PC plasmalogen-B HMDB0011220* Glycerophospholipids Glycerophosphocholines 0.75 (0.57–0.99) 0.043 0.84 (0.47–1.47) 0.534
C22:0 LPE HMDB0011520 Glycerophospholipids Glycerophosphoethanolamines 0.65 (0.48–0.88) 0.005 0.96 (0.51–1.80) 0.900
C38:6 PE plasmalogen HMDB0011387* Glycerophospholipids Glycerophosphoethanolamines 0.72 (0.54–0.95) 0.022 0.72 (0.41–1.27) 0.254
C36:5 PE plasmalogen HMDB0011410* Glycerophospholipids Glycerophosphoethanolamines 0.73 (0.55–0.97) 0.028 0.84 (0.48–1.47) 0.542
serotonin HMDB0000259 Indoles and derivatives Tryptamines and derivatives 1.37 (1.04–1.81) 0.025 1.03 (0.59–1.80) 0.925
C20:4 LPC HMDB0010395 NA NA 0.66 (0.50–0.88) 0.004 0.94 (0.52–1.67) 0.823
thyroxine HMDB0000248 NA NA 1.50 (1.11–2.04) 0.009 2.16 (1.15–4.13) 0.018
C20:1 LPE HMDB0011512* NA NA 0.69 (0.52–0.92) 0.011 0.71 (0.39–1.30) 0.270
trigonelline HMDB0000875 NA NA 0.71 (0.53–0.95) 0.021 0.66 (0.37–1.18) 0.161
carnitine HMDB0000062 Organonitrogen compounds Quaternary ammonium salts 0.73 (0.54–0.99) 0.041 1.12 (0.60–2.07) 0.729
acetyl-galactosamine HMDB0000212 Organooxygen compounds Carbohydrates and carbohydrate conjugates 1.26 (0.95–1.69) 0.110 1.85 (1.04–3.33) 0.037
2-methylguanosine HMDB0005862 Purine nucleosides NA 1.34 (1.00–1.79) 0.054 2.02 (1.13–3.64) 0.019
C22:5 CE HMDB0010375* Steroids and steroid derivatives Cholesterol esters 0.52 (0.39–0.70) <0.001 0.65 (0.35–1.18) 0.160
C20:5 CE HMDB0006731 Steroids and steroid derivatives Cholesterol esters 0.61 (0.46–0.82) 0.001 0.60 (0.33–1.10) 0.103
C18:3 CE HMDB0010370* Steroids and steroid derivatives Cholesterol esters 0.65 (0.49–0.86) 0.003 0.55 (0.30–1.01) 0.054
C20:4 CE HMDB0006726 Steroids and steroid derivatives Cholesterol esters 0.67 (0.50–0.89) 0.006 0.77 (0.42–1.38) 0.374
C18:0 CE HMDB0010368 Steroids and steroid derivatives Cholesterol esters 0.68 (0.51–0.91) 0.010 0.99 (0.53–1.83) 0.965
C20:3 CE HMDB0006736* Steroids and steroid derivatives Cholesterol esters 0.73 (0.55–0.96) 0.026 0.69 (0.38–1.25) 0.225
C18:1 CE HMDB0000918* Steroids and steroid derivatives Cholesterol esters 0.72 (0.54–0.97) 0.030 0.77 (0.41–1.42) 0.404
Proximate Blood
hippurate HMDB0000714 Benzene and substituted derivatives Benzoic acids and derivatives 0.64 (0.45–0.91) 0.014 0.52 (0.26–1.03) 0.062
N1,N12-diacetylspermine HMDB0002172 Carboximidic acids and derivatives Carboximidic acids 1.46 (1.02–2.08) 0.038 2.33 (1.14–4.81) 0.020
proline HMDB0000162 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.52 (1.04–2.21) 0.029 0.89 (0.41–1.91) 0.760
phenylalanine HMDB0000159 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.48 (1.03–2.14) 0.035 1.67 (0.80–3.50) 0.171
isoleucine HMDB0000172 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.47 (1.00–2.16) 0.049 1.31 (0.62–2.79) 0.487
C5-DC carnitine HMDB0013130 Fatty Acyls Fatty acid esters 0.60 (0.42–0.86) 0.004 1.46 (0.72–2.97) 0.289
C52:0 TAG HMDB0005365* Glycerolipids Triacylglycerols 1.49 (1.04–2.15) 0.030 0.86 (0.42–1.74) 0.668
C54:9 TAG HMDB0010498* Glycerolipids Triacylglycerols 0.68 (0.48–0.98) 0.037 0.60 (0.29–1.23) 0.160
C58:11 TAG HMDB0010531* Glycerolipids Triacylglycerols 0.69 (0.48–0.98) 0.038 0.61 (0.29–1.27) 0.189
C58:9 TAG HMDB0005463* Glycerolipids Triacylglycerols 0.69 (0.48–0.99) 0.043 0.70 (0.34–1.46) 0.345
C58:7 TAG HMDB0005471* Glycerolipids Triacylglycerols 0.68 (0.47–0.99) 0.044 0.55 (0.27–1.10) 0.093
C56:10 TAG HMDB0010513* Glycerolipids Triacylglycerols 0.69 (0.49–0.99) 0.045 0.55 (0.26–1.15) 0.113
C58:10 TAG HMDB0005476* Glycerolipids Triacylglycerols 0.70 (0.49–0.99) 0.046 0.67 (0.32–1.37) 0.272
C52:1 TAG HMDB0005367* Glycerolipids Triacylglycerols 1.45 (1.01–2.09) 0.047 0.83 (0.41–1.70) 0.614
tryptophan HMDB0000929 Indoles and derivatives Indolyl carboxylic acids and derivatives 1.56 (1.09–2.23) 0.015 0.98 (0.49–1.97) 0.956
guanosine HMDB0000133 Purine nucleosides NA 1.45 (1.02–2.06) 0.039 0.71 (0.35–1.42) 0.332
C22:0 Ceramide (d18:1) HMDB0004952 Sphingolipids Ceramides 1.50 (1.05–2.16) 0.027 1.28 (0.63–2.59) 0.488
C24:1 Ceramide (d18:1) HMDB0004953* Sphingolipids Ceramides 1.48 (1.03–2.13) 0.035 1.26 (0.61–2.6) 0.533
C16:0 Ceramide (d18:1) HMDB0004949 Sphingolipids Ceramides 1.48 (1.02–2.15) 0.037 1.89 (0.95–3.82) 0.071
C22:5 CE HMDB0010375* Steroids and steroid derivatives Cholesteryl esters 0.69 (0.49–0.98) 0.040 1.08 (0.56–2.10) 0.814
hydroxyproline HMDB0000725 Carboxylic acids and derivatives Amino acids, peptides, and analogues 0.77 (0.54–1.10) 0.157 2.03 (1.01–4.15) 0.048
C5:1 carnitine HMDB0002366 Fatty Acyls Fatty acid esters 0.80 (0.55–1.14) 0.218 2.44 (1.19–5.07) 0.015
C45:0 TAG HMDB0042093* Glycerolipids Triacylglycerols 1.16 (0.80–1.67) 0.465 0.67 (0.33–1.34) 0.039
C22:0 LPE HMDB0011520 Glycerophospholipids Glycerophosphoethanolamines 0.96 (0.66–1.39) 0.818 2.39 (1.16–5.01) 0.018
deoxyguanosine HMDB0000085 NA NA 1.18 (0.83–1.68) 0.346 0.47 (0.23–0.98) 0.043
methyl N-methylanthra-nilate HMDB0034169 NA NA 1.10 (0.78–1.55) 0.580 0.50 (0.25–0.98) 0.045
kynurenic acid HMDB0000715 Quinolines and derivatives Quinoline carboxylic acids 0.97 (0.67–1.41) 0.888 2.14 (1.03–4.5) 0.041
Open in a new tab
^

Selected metabolites are those with <10% missingness and a nominal p<0.05 for either ER+ or ER− breast cancers (those identified as significant in ER− breast cancers are in bold). Missing values were imputed with 1/2 the minimum value. Results sorted by class, subclass, and p-value for fully adjusted model. Metabolites in bold represent those chosen as top hits for ER− breast cancer.

†

Multivariate unconditional logistic regression model adjusted for: BMI at age 18, weight change since age 18, age at menarche, combined age at first birth and parity, breastfeeding history, history of benign breast disease, family history of breast cancer, alcohol use (g/day), activity level (met hrs/week). P values are nominal p-values before correction for multiple testing.

*

Representative HMDBID

Due to the differences in TAG associations by number of double bonds, we further explored TAGs by number of carbon atoms and number of double bonds (Figure 1). We observed a strong inverse association for TAGs with increasing carbon atoms and double bonds at the proximate timepoint. This inverse association was not notable for the distant timepoint, though we observed a trend of more positive associations with lower number of carbon atoms and double bonds at the distant timepoint.

Figure 1.

Figure 1.

Open in a new tab

Odds ratios for breast cancer risk comparing 90th to 10th percentile of triacylglycerols, by number of Carbon atoms and double bonds at (a) distant and (b) proximate blood. Models are for CLR, fully adjusted for: BMI at age 18, weight change since age 18, age at menarche, combined age at first birth and parity, breastfeeding history, history of benign breast disease, family history of breast cancer, alcohol use (g/day), activity level (MET-hrs/week). Y axis is number of double bonds, X axis is number of Carbon atoms. Protective associations are shown in blue, harmful associations are shown in red.

MSEA results mirrored individual metabolite analyses and revealed several subclasses of metabolites significantly associated with breast cancer risk after FDR correction (Figure 2 and Tables S5a & S5b). TAGs with <3 double bonds at the distant timepoint were strongly positively associated with risk of overall (padj=0.02 ER+, and ER− breast cancers. This trend remained at the proximate blood draw for ER+ breast cancers, but no association was observed for ER− breast cancers. TAGs with ≥3 double bonds were strongly inversely associated with breast cancer risk at the proximate blood draw for overall (padj=0.03), ER+, and ER− breast cancers; however, at the distant blood draw this group was significantly positively associated with ER+ breast cancer.

Figure 2.

Figure 2.

Open in a new tab

Gene set enrichment analysis by subclass of metabolites for overall, ER+, and ER− breast cancer, distant (≥10y before dx) & proximate blood (<10y before dx). Overall breast cancer results use conditional logistic regression; ER status specific models use unconditional logistic regression models adjusted for matched factors. Models are fully adjusted for the following: BMI at age 18, weight change from age 18 to blood draw, age at menarche, combined age at first birth and parity, breastfeeding history, history of benign breast disease, family history of breast cancer, alcohol use (g/day), activity level (MET-hrs/week). Stars denote p-values adjusted by FDR: * (padj <0.2); ** (padj<0.05). Darker blue is a more negative enrichment score; darker red is a more positive enrichment score.

Cholesteryl esters were strongly inversely associated with risk at the distant timepoint (padj overall BC=0.02), and less strongly, though still inverse, at the proximate timepoint. Glycerophospholipids, glycerophosphoethanolamines, and glycerophosphocholines were inversely associated with risk at the distant timepoint for ER+ breast cancers, though associations were weaker and not significant at the proximate timepoint. Similarly, diacylglycerols (DAGs) were strongly positively associated with risk at the distant timepoint, but less strongly associated at the proximate timepoint. Further, the group amino acids, peptides, and analogues was positively associated with overall (padj=0.02), ER+ and ER− breast cancer at the proximate blood, though the result was stronger for ER− than ER+ breast cancers. This group was not significantly associated with breast cancer risk at the distant timepoint.

WGCNA defined 12 metabolite modules at distant collection, and 11 at proximate (Figure S1 & S2, Table S6). Module 1, the grey module, represents those metabolites that remained after correlation analyses determined other metabolite groupings. While modules were not defined by one particular subclass, most had a majority of one subclass, or a split between two subclass distinctions.

At the distant timepoint, no modules were significantly associated with overall breast cancer risk (Table S7a). One module, defined by several glycerophospholipids and TAGs with high numbers of double bonds, was suggestively inversely associated with ER+ breast cancer (M7 OR=0.66, 95% CI=0.49–0.89; nominal p-value=0.01, FDR adjusted p-value=0.08). TAGs with high numbers of double bonds were negatively weighted in this module, while glycerophospholipids were mainly positively weighted (Figure S3a). This finding, with higher glycerophospholipids and lower TAGs with ≥3 double bonds, corresponds with MSEA results (Figure 2). While glycerophospholipids were not significantly associated with ER+ breast cancer in MSEA, our results from the WGCNA highlight the importance of a few key glycerophospholipids including C20:4 LPC (OR comparing 90th to 10th percentile=0.66, 95% CI=0.50–0.88, p=0.004) and C18:2 LPC (OR =0.64, 95% CI=0.47–0.87, p=0.005) (Table S3a). At the proximate timepoint, no modules were associated with breast cancer risk (Table S7b). Despite this, associative patterns that arose in module groupings aligned with MSEA results (Figure S3b).

Metabolites with the most significant difference measures between blood draws included TAGs with ≥3 double bonds (Table 4). An increase in TAGs with ≥3 double bonds from distant to proximate measures was associated with a reduced breast cancer risk (e.g.: for C56:10 TAG OR 90th-10th percentile=0.62, 95% CI=0.43–0.88; nominal p-value=0.007).

Table 4.

Odds ratios for breast cancer risk comparing 90th–10th percentiles in proximate-distant metabolite measures for metabolites with nominal p-value <0.05.

Metabolite HMDB ID Class Subclass OR (95% CI) ^ p value Spearman correlation†
N-alpha-acetylarginine HMDB0004620 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.86 (1.25–2.78) 0.002 0.655
2-aminooctanoic acid HMDB0000991 Carboxylic acids and derivatives Amino acids, peptides, and analogues 0.67 (0.47–0.93) 0.019 0.458
aminoisobutyric acid HMDB0001906 Carboxylic acids and derivatives Amino acids, peptides, and analogues 0.65 (0.45–0.93) 0.020 0.502
isoleucine HMDB0000172 Carboxylic acids and derivatives Amino acids, peptides, and analogues 1.43 (1.01–2.03) 0.044 0.422
myristoleic acid HMDB0002000 Fatty Acyls Fatty acids and conjugates 1.47 (1.03–2.11) 0.033 0.513
C58:9 TAG HMDB0005463 Glycerolipids Triacylglycerols 0.60 (0.42–0.85) 0.004 0.475
C56:9 TAG HMDB0005448 Glycerolipids Triacylglycerols 0.60 (0.42–0.86) 0.005 0.489
C58:11 TAG HMDB0010531 Glycerolipids Triacylglycerols 0.60 (0.41–0.86) 0.005 0.520
C56:10 TAG HMDB0010513 Glycerolipids Triacylglycerols 0.62 (0.43–0.88) 0.007 0.493
C56:7 TAG HMDB0005462 Glycerolipids Triacylglycerols 0.62 (0.43–0.88) 0.008 0.500
C58:7 TAG HMDB0005471 Glycerolipids Triacylglycerols 0.64 (0.45–0.90) 0.010 0.423
C56:8 TAG HMDB0005392 Glycerolipids Triacylglycerols 0.64 (0.45–0.90) 0.010 0.426
C54:8 TAG HMDB0010518 Glycerolipids Triacylglycerols 0.65 (0.46–0.91) 0.013 0.438
C54:9 TAG HMDB0010498 Glycerolipids Triacylglycerols 0.65 (0.46–0.92) 0.015 0.469
C58:10 TAG HMDB0005476 Glycerolipids Triacylglycerols 0.64 (0.45–0.92) 0.015 0.488
C60:12 TAG HMDB0005478 Glycerolipids Triacylglycerols 0.64 (0.45–0.92) 0.017 0.504
C58:8 TAG HMDB0005413 Glycerolipids Triacylglycerols 0.67 (0.48–0.95) 0.023 0.459
C52:7 TAG HMDB0010517 Glycerolipids Triacylglycerols 0.68 (0.48–0.97) 0.032 0.473
C58:6 TAG HMDB0005458 Glycerolipids Triacylglycerols 0.70 (0.49–0.98) 0.037 0.434
C54:7 TAG HMDB0005447 Glycerolipids Triacylglycerols 0.71 (0.51–0.99) 0.045 0.388
C18:1 LPC HMDB0002815 Glycerophospholipids Glycerophosphocholines 1.47 (1.06–2.06) 0.022 0.390
C18:0 LPC HMDB0010384 Glycerophospholipids Glycerophosphocholines 1.45 (1.05–2.01) 0.026 0.351
C40:9 PC HMDB0008731 Glycerophospholipids Glycerophosphocholines 0.67 (0.46–0.98) 0.039 0.552
C38:6 PC HMDB0007991 Glycerophospholipids Glycerophosphocholines 0.67 (0.46–0.98) 0.040 0.547
C18:3 LPC HMDB0010387 Glycerophospholipids Glycerophosphocholines 1.38 (1.00–1.90) 0.047 0.301
C38:6 PE HMDB0009102 Glycerophospholipids Glycerophosphoethanolamines 0.64 (0.44–0.92) 0.017 0.549
C22:0 LPE HMDB0011520 Glycerophospholipids Glycerophosphoethanolamines 1.44 (1.03–2.03) 0.035 0.438
C20:1 LPE HMDB0011512 Glycerophospholipids Glycerophosphoethanolamines 1.40 (1.00–1.94) 0.048 0.375
C22:6 LPE HMDB0011526 Glycerophospholipids Glycerophosphoethanolamines 0.70 (0.49–1.00) 0.048 0.434
tryptophan HMDB0000929 Indoles and derivatives Indolyl carboxylic acids and derivatives 1.42 (1.02–1.98) 0.039 0.395
ribothymidine HMDB0000884 Pyrimidine nucleosides Pyrimidine nucleosides 1.49 (1.03–2.16) 0.033 0.552
C16:0 Ceramide (d18:1) HMDB0004949 Sphingolipids Ceramides 1.57 (1.12–2.20) 0.009 0.417
C22:0 Ceramide (d18:1) HMDB0004952 Sphingolipids Ceramides 1.48 (1.04–2.12) 0.030 0.509
C24:1 Ceramide (d18:1) HMDB0004953 Sphingolipids Ceramides 1.44 (1.01–2.04) 0.043 0.463
C18:3 CE HMDB0010370 Steroids and steroid derivatives Cholesteryl esters 1.53 (1.07–2.19) 0.021 0.495
C18:0 CE HMDB0010368 Steroids and steroid derivatives Cholesteryl esters 1.44 (1.04–2.02) 0.030 0.403
Open in a new tab
^

All models adjusted for distant blood measure. Estimate is for difference proximate-distant blood measure. Results are sorted by class, subclass, and p-value for Model 2. ORs are for unconditional logistic regressions adjusted for BMI at age 18, weight change from 18 to blood draw, age at menarche, combined age at first birth and parity, breastfeeding history, history of benign breast disease, family history of breast cancer, alcohol use (g/day), activity level (met hrs/week). P values are nominal p-values before correction for multiple testing.

†

Correlations between proximate and distant time point, adjusted for fasting status and age at blood draw

*

Representative HMDBID

The majority of metabolites were moderately correlated between timepoints (Spearman correlation 0.40–0.50) (Table S8 & Figure S4). Analysis of metabolite associations with breast cancer risk taking the associations of averaged metabolites from both timepoints with breast cancer risk generally were similar to individual timepoint analyses. However, some associations were weakened due to opposing associations, while others were strengthened by consistent associations (Table S9). MSEA analysis of average values highlighted the strong inverse association seen for cholesteryl esters (Figure S5), and the strong positive association seen for TAGS with <3 double bonds. TAGS with ≥3 double bonds showed strong inverse associations with overall and ER− BC on average, but null associations with ER+ breast cancer on average, due to opposing directions of association at distant and proximate bloods.

DISCUSSION

In this nested case-control study examining the association between 307 plasma metabolites and breast cancer risk, we identified several metabolite groups, defined based on similar biochemical structure, that were associated with risk. Individual metabolites did not reach statistical significance with correction for multiple comparisons; however, common patterns appeared for structurally similar metabolites. By subclass, cholesteryl esters were inversely associated with breast cancer risk, whereas amino acids and derivatives were associated with increased risk. The association between TAGs and breast cancer risk was dependent on the number of double bonds; TAGs with ≥3 double bonds were inversely associated, while TAGs with <3 double bonds were positively associated with risk. The unique ability to assess metabolite measures at two different timepoints also highlighted the potential for metabolites to influence different stages of breast cancer development, as several associations differed by time between blood draw and diagnosis.

Our results add novel knowledge and provide support for several findings from other similar agnostic metabolomic approaches. For example, a nested case-control study in EPIC (with 1,624 cases), examined 127 metabolites in pre-diagnostic blood samples.9 While individual metabolite results were not consistent with our study, associations by metabolite classes were similar. For example, C2 carnitine was inversely associated with breast cancer risk in EPIC but not in our study, though we observed an inverse association between high levels of carnitines and risk in general, with C5-DC carnitine appearing inversely related to breast cancer with a nominal p-value <0.05 at both timepoints. This finding is also reflected in a recent nested case-control study in Cancer Prevention Study 2 (CPS 2, n=782 cases), with 1,275 metabolites,3 where acyl fatty acid derivatives of carnitine were inversely associated with risk. Carnitine deficiency is associated with increased insulin sensitivity,20 suggesting that the inverse association between carnitines and breast cancer may be due to insulin-dependent signaling pathways. In fact, carnitine supplementation has been shown to improve glucose homeostasis.20 Among BMI-associated metabolites in the Prostate, Lung, Colorectal, and Ovarian cancer screening (PLCO) cohort, acylcarnitines 3-methylglutarylcarnitine and 2-methylbutyrylcarnitine were associated with increased breast cancer risk,2 contrasting with the finding of the agnostic analysis within CPS 2.3 While we generally observed inverse associations with breast cancer risk for carnitines and derivatives, a few carnitines were suggestively positively associated with risk, including C14 carnitine. Higher levels of acylcarnitines have been associated with increased meat consumption, and higher blood concentrations may be indicative of changes in mitochondrial function and β-oxidation;21,22 thus, the positive association with breast cancer seen here may represent breakdown products of animal sources of protein. In addition, accumulation of long chain (C14-C20) acylcarnitines has been associated with decreased insulin sensitivity.23,24

Phosphatidylcholines were inversely associated with risk in EPIC; we found nominally significant inverse association between several glycerophosphocholines (derivatives of phosphatidylcholines) and risk, especially for ER+ breast cancer. Dietary choline intake from glycerophosphocholines was inversely associated with breast cancer risk in a Chinese case-control study.25 In an earlier study within the EPIC-Heidelberg cohort, higher levels of lysophosphatidylcholines (lysoPCs) were associated with decreased breast cancer risk,6 which was suggestive, though not statistically significant, in our study. In contrast, a positive association between glycerophosphocholines and breast cancer risk was observed in CPS 2,3 which was not observed in our study or in EPIC.9 The associations may be dependent on side chains of interest; because each study measured a slightly different set of metabolites, direct comparisons are not possible, though this group may be important in breast cancer development.

Although the Korean Cancer Prevention Study II was of much smaller scale (N=84 cases), amino acid metabolism, fatty acid metabolism, and linoleic acid metabolism differed between cases and controls, similar to some of our findings.8 In pathway analysis, increased breast cancer risk was observed for metabolites involved in phenylalanine, tyrosine, and tryptophan biosynthesis, suggesting that amino acid metabolism may be an important driver in breast cancer development. Lower circulating levels of amino acid were apparent in cases at diagnosis compared to controls,26,27 suggesting that the tumor-specific metabolic reprogramming focuses on amino acids. Perhaps a high level of amino acids many years prior to diagnosis provides a hospitable environment for tumor cells that will later use these amino acids to drive their formation. In our analysis phenylalanine was one of the strongest hits at both the distant and proximate timepoints. The importance of this metabolite may be further highlighted by the need for cancer cells to uptake phenylalanine to survive; in fact, a recent study used nanoparticles coated with phenylalanine to target and cause cancer cells to self-destruct.28 Proline also appeared nominally significant in our analyses; this amino acid plays a key role in metabolic reprogramming important for cancer cell survival.29 Further supporting our amino acid findings, plasma levels of amino acids including valine, lysine, arginine, glutamine were associated with increased breast cancer risk in SU.VI.MAX cohort,5 which used untargeted NMR metabolomic profiles (N=206 cases). In contrast to our current findings and those reported separately,14 in the Women’s Health Study, branched-chain amino acids (BCAAs) valine, leucine, and isoleucine were not associated with breast cancer risk.

Uniquely, in our study we observed a strong inverse association between cholesteryl esters and breast cancer risk. Cholesteryl esters form the components of cholesterol, high density lipoprotein (HDL) and low-density lipoprotein (LDL), levels of which have been associated with breast cancer risk,30 though epidemiologic evidence for these associations remain inconsistent.31 In addition, laboratory studies and in vivo studies suggest cholesterol metabolism as a driver for breast cancer tumor growth.32 The role of cholesteryl esters in lipid metabolism and transport is of interest, as lipid metabolic reprogramming occurs in cancer cells.33 Our finding of an inverse association of cholesteryl esters with breast cancer risk was more notable at the distant timepoint. While the biologic basis for this difference over time is unclear, women with more prominent cholesteryl ester profiles earlier in life may be more likely to benefit from their protective effect, making cholesterol metabolism across the life course an important avenue for further research.

Here we found TAGs, a metabolite subclass not measured in EPIC9 and CPS23, were significantly associated with breast cancer risk, but in opposite directions depending on the size and number of carbon atom double bonds. Recent studies of diabetes suggest that lipid composition is important in the association between lipids and diabetes, and may reflect insulin activity.34 TAGs with low carbon atoms and low double bonds are associated with insulin resistance and consequently with diabetes, while TAGs with high carbon atoms and high double bonds are higher in those with normal insulin function.34 Insulin signaling is a marker for metabolic health, which is a predictor of breast cancer risk.35,36 Insulin signaling is responsible for activating the mitogen activated protein kinase (MAPK) and PI3K/Akt pathways, which promote cancer cell proliferation and invasion.37 As noted above, the role of carnitines in the insulin-signaling pathway may also underlie their associations with breast cancer risk. We also found several ceramides (C16:0, C24:1, C22:0), also potential markers for insulin resistance,38 to be associated with an increased risk of breast cancer. In addition, several TAGS with many double bonds and carbon atoms are associated with higher fish intake,39 indicating a potential protective mechanism of dietary fish intake. More generally, polyunsaturated, omega-3, and omega-6 fatty acids are associated with a higher alternative healthy eating index (AHEI) score,40 further demonstrating potential of dietary intake to influence metabolite levels and future breast cancer risk. In contrast to our findings, lower plasma levels of unsaturated lipids were associated with a higher breast cancer risk in the SU.VI.MAX cohort.5 Further research is needed to fully understand our findings, including why these relationships changed over time.

There are several differences between our study and previous studies. The platforms used for metabolomic profiling differed, which may account for the inconsistencies between studies,3 and actual metabolites measured, constituting various stages of breakdown pathways, differed between studies. Moreover, the timing of blood collections and median time from blood draw to diagnosis differed across studies. This may have contributed to observed differences between studies. While lag-time between blood draw and diagnosis was explored in the EPIC-Heidelberg cohort,6 median follow-up time was <10 years from blood draw.

There are several strengths of our study. The large sample size allowed adequate power for analyses. We had the ability to control for covariates at the time of metabolite measure, as data were collected for all pertinent covariates every two years with follow-up questionnaires, and on blood draw specific questionnaires. Our study assessed how metabolite associations with breast cancer change over time, with samples taken covering a period 0–20 years prior to cancer diagnosis.

While the metabolomic platforms used for profiling were a strength of our study, these also represent a limitation given the inability to directly compare results to those of others. Assessment by ER status was limited by ER− cases. We were unable to investigate premenopausal breast cancer, as most women in NHS were already postmenopausal by the second blood draw.

In conclusion, we found several metabolite subclasses that may be of further interest to explore in breast cancer etiology, including cholesteryl esters, amino acids, and TAGs. Our findings clarified some previous findings, supporting the idea that carnitine metabolism and glycerophosphocholines may be involved in reduction of breast cancer risk. Notably, several metabolite-breast cancer associations we observed may be explained, at least in part, by their role in insulin-signaling pathways. Future studies are needed to determine the intricacies of the biologic mechanisms contributing to breast cancer risk.

Supplementary Material

1
2

ACKNOWLEDGEMENTS

This study was funded by the National Cancer Institute. Eliassen AH received UM1 CA186107, P01 CA87969, and T32 CA009001 grants from NCI; Hankinson SE received R01 CA49449 from NCI. We would like to thank the participants and staff of the Nurses’ Health Study for their valuable contributions as well as the following state cancer registries for their help: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, WY. The authors assume full responsibility for analyses and interpretation of these data.

Footnotes

The authors declare no potential conflicts of interest.

REFERENCES

  • 1.Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global Cancer Statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin, in press. [DOI] [PubMed] [Google Scholar]
  • 2.Moore SC, Playdon MC, Sampson JN, Hoover RN, Trabert B, Matthews CE, et al. A Metabolomics Analysis of Body Mass Index and Postmenopausal Breast Cancer Risk. J Natl Cancer Inst. 2018;110(6):588–597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Moore SC, Mazzilli KM, Sampson JN, Matthews CE, Carter BD, Playdon MC, et al. A Metabolomics Analysis of Postmenopausal Breast Cancer Risk in the Cancer Prevention Study II. Metabolites. 2021;11(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Tobias DK, Hazra A, Lawler PR, Chandler PD, Chasman DI, Buring JE, et al. Circulating branched-chain amino acids and long-term risk of obesity-related cancers in women. Sci Rep. 2020;10(1):16534. doi: 10.1038/s41598-020-73499-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lécuyer L, Victor Bala A, Deschasaux M, Bouchemal N, Triba MN, Vasson MP, et al. NMR metabolomic signatures reveal predictive plasma metabolites associated with long-term risk of developing breast cancer. Int J Epidemiol. 2018;47(2):484–494. doi: 10.1093/ije/dyx271 [DOI] [PubMed] [Google Scholar]
  • 6.Kühn T, Floegel A, Sookthai D, Johnson T, Rolle-Kampczyk U, Otto W, et al. Higher plasma levels of lysophosphatidylcholine 18:0 are related to a lower risk of common cancers in a prospective metabolomics study. BMC Med. 2016;14:13. doi: 10.1186/s12916-016-0552-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Playdon MC, Ziegler RG, Sampson JN, Stolzenberg-Solomon R, Thompson HJ, Irwin ML, et al. Nutritional metabolomics and breast cancer risk in a prospective study. Am J Clin Nutr. 2017;106(2):637–649. doi: 10.3945/ajcn.116.150912 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Yoo HJ, Kim M, Kim M, Kang M, Jung KJ, Hwang S, et al. Analysis of metabolites and metabolic pathways in breast cancer in a Korean prospective cohort: the Korean Cancer Prevention Study-II. Metabolomics. 2018;14(6):85. doi: 10.1007/s11306-018-1382-4 [DOI] [PubMed] [Google Scholar]
  • 9.His M, Viallon V, Dossus L, Gicquiau A, Achaintre D, Scalbert A, et al. Prospective analysis of circulating metabolites and breast cancer in EPIC. BMC Medicine. 2019;17(1):178. doi: 10.1186/s12916-019-1408-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Mascanfroni ID, Takenaka MC, Yeste A, Patel B, Wu Y, Kenison JE, et al. Metabolic control of type 1 regulatory T cell differentiation by AHR and HIF1-α. Nat Med. 2015;21(6):638–646. doi: 10.1038/nm.3868 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.O’Sullivan JF, Morningstar JE, Yang Q, Zheng B, Gao Y, Jeanfavre S, et al. Dimethylguanidino valeric acid is a marker of liver fat and predicts diabetes. J Clin Invest. 2017;127(12):4394–4402. doi: 10.1172/JCI95995 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Paynter NP, Balasubramanian R, Giulianini, Wang DD, Tinker LF, Gopal S, et al. , Metabolic predictors of incident coronary heart disease in women. Circulation, 2018. 137(8): p. 841–853. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Townsend MK, Clish CB, Kraft P, Wu C, Souza AL, Deik AA, et al. Reproducibility of metabolomic profiles among men and women in 2 large cohort studies. Clin Chem. 2013;59(11):1657–1667. doi: 10.1373/clinchem.2012.199133 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Zeleznik OA, Eliassen AH, Kraft P, Poole EM, Rosner BA, Jeanfavre S, et al. A Prospective Analysis of Circulating Plasma Metabolites Associated with Ovarian Cancer Risk. Cancer Res. 2020;80(6):1357–1367. doi: 10.1158/0008-5472.CAN-19-2567 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Gallois A, Mefford J, Ko A, Vaysse A, Julienne H, Ala-Korpela M, et al. A comprehensive study of metabolite genetics reveals strong pleiotropy and heterogeneity across time and context. Nature Communications. 2019;10(1):4788. doi: 10.1038/s41467-019-12703-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences. 2005;102(43):15545–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Benjamini Y, Krieger AM, Yekutieli D. Adaptive linear step-up procedures that control the false discovery rate. Biometrika. 2006;93(3):491–507 [Google Scholar]
  • 18.Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Langfelder P, Zhang B, Horvath S. Defining clusters from a hierarchical cluster tree: the Dynamic Tree Cut package for R. Bioinformatics. 2008;24(5):719–720. doi: 10.1093/bioinformatics/btm563 [DOI] [PubMed] [Google Scholar]
  • 20.Ringseis R, Keller J, Eder K. Role of carnitine in the regulation of glucose homeostasis and insulin sensitivity: evidence from in vivo and in vitro studies with carnitine supplementation and carnitine deficiency. Eur J Nutr. 2012. Feb;51(1):1–18. doi: 10.1007/s00394-011-0284-2. Epub 2011 Dec 2. [DOI] [PubMed] [Google Scholar]
  • 21.Rebholz CM, Zheng Z, Grams ME, Appel LJ, Sarnak MJ, Inker LA, et al. Serum metabolites associated with dietary protein intake: results from the Modification of Diet in Renal Disease (MDRD) randomized clinical trial. Am J Clin Nutr. 2019;109(3):517–525. doi: 10.1093/ajcn/nqy202 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bouchard-Mercier A, Rudkowska I, Lemieux S, Couture P, Vohl M-C. The metabolic signature associated with the Western dietary pattern: a cross-sectional study. Nutr J. 2013;12:158. doi: 10.1186/1475-2891-12-158 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Vilks K, Videja M, Makrecka-Kuka M, Katkevics M, Sevostjanovs E, Grandane A, et al. Long-Chain Acylcarnitines Decrease the Phosphorylation of the Insulin Receptor at Tyr1151 Through a PTP1B-Dependent Mechanism. Int J Mol Sci. 2021;22(12):6470. doi: 10.3390/ijms22126470 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.McCann MR, George De la Rosa MV, Rosania GR, Stringer KA. L-Carnitine and Acylcarnitines: Mitochondrial Biomarkers for Precision Medicine. Metabolites. 2021;11(1):51. doi: 10.3390/metabo11010051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Zhang C-X, Pan M-X, Li B, Wang L, Mo X-F, Chen Y-M, et al. Choline and betaine intake is inversely associated with breast cancer risk: a two-stage case-control study in China. Cancer Sci. 2013;104(2):250–258. doi: 10.1111/cas.12064 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Yuan B, Schafferer S, Tang Q, Scheffler M, Nees J, Jӧrg Heil, et al. A plasma metabolite panel as biomarkers for early primary breast cancer detection. Int J Cancer. 2019;144(11):2833–2842. doi: 10.1002/ijc.31996 [DOI] [PubMed] [Google Scholar]
  • 27.Jové M, Collado R, Quiles JL, Ramírez-Tortosa M-C, Sol J, Ruiz-Sanjuan M, et al. A plasma metabolomic signature discloses human breast cancer. Oncotarget. 2017;8(12):19522–19533. doi: 10.18632/oncotarget.14521 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Wu Zhuoran, Lim Hong Kit, Tan Shao Jie, Gautam Archana, Hou Han Wei, Ng Kee Woei, et al. Potent-By-Design: Amino Acids Mimicking Porous Nanotherapeutics with Intrinsic Anticancer Targeting Properties. Small, 2020; 16 (34) [DOI] [PubMed] [Google Scholar]
  • 29.Burke L, Guterman I, Palacios Gallego R, Britton RG, Burschowsky D, Tufarelli C, et al. The Janus-like role of proline metabolism in cancer. Cell Death Discov. 2020;6:104. doi: 10.1038/s41420-020-00341-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Pelton K, Coticchia CM, Curatolo AS, Schaffner CP, Zurakowski D, Solomon KR, et al. Hypercholesterolemia induces angiogenesis and accelerates growth of breast tumors in vivo. Am J Pathol. 2014;184(7):2099–2110. doi: 10.1016/j.ajpath.2014.03.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Cedó L, Reddy ST, Mato E, Blanco-Vaca F, Escolà-Gil JC. HDL and LDL: Potential New Players in Breast Cancer Development. J Clin Med. 2019;8(6). doi: 10.3390/jcm8060853 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Danilo C, Frank PG. Cholesterol and breast cancer development. Curr Opin Pharmacol. 2012;12(6):677–682. doi: 10.1016/j.coph.2012.07.009 [DOI] [PubMed] [Google Scholar]
  • 33.Beloribi-Djefaflia S, Vasseur S, Guillaumond F. Lipid metabolic reprogramming in cancer cells. Oncogenesis. 2016;5:e189. doi: 10.1038/oncsis.2015.49 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Rhee EP, Cheng S, Larson MG, Walford GA, Lewis GD, McCabe E, et al. Lipid profiling identifies a triacylglycerol signature of insulin resistance and improves diabetes prediction in humans. J Clin Invest. 2011;121(4):1402–1411. doi: 10.1172/JCI44442 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Iyengar NM, Arthur R, Manson JE, Chlebowski RT, Kroenke CH, Peterson L, et al. Association of Body Fat and Risk of Breast Cancer in Postmenopausal Women with Normal Body Mass Index: A Secondary Analysis of a Randomized Clinical Trial and Observational Study. JAMA Oncol. 2019;5(2):155–163. doi: 10.1001/jamaoncol.2018.5327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Park Y-MM, White AJ, Nichols HB, O’Brien KM, Weinberg CR, Sandler DP. The association between metabolic health, obesity phenotype and the risk of breast cancer. Int J Cancer. 2017;140(12):2657–2666. doi: 10.1002/ijc.30684 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Yee LD, Mortimer JE, Natarajan R, Dietze EC, Seewaldt VL. Metabolic Health, Insulin, and Breast Cancer: Why Oncologists Should Care About Insulin. Front Endocrinol (Lausanne). 2020;11:58. doi: 10.3389/fendo.2020.00058 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Sokolowska E, Blachnio-Zabielska A. The Role of Ceramides in Insulin Resistance. Front Endocrinol (Lausanne). 2019;10:577. doi: 10.3389/fendo.2019.00577 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Mazzilli KM, McClain KM, Lipworth L, Playdon MC, Sampson JN, Clish CB, et al. Identification of 102 Correlations between Serum Metabolites and Habitual Diet in a Metabolomics Study of the Prostate, Lung, Colorectal, and Ovarian Cancer Trial. J Nutr. 2020;150(4):694–703. doi: 10.1093/jn/nxz300 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Akbaraly T, Würtz P, Singh-Manoux A, Shipley MJ, Haapakoski R, Lehto M, et al. Association of circulating metabolites with healthy diet and risk of cardiovascular disease: analysis of two cohort studies. Sci Rep. 2018;8(1):8620. doi: 10.1038/s41598-018-26441-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1773833-supplement-1.xlsx (266.3KB, xlsx)
2
NIHMS1773833-supplement-2.pptx (7MB, pptx)

RESOURCES