Association of Mediterranean diet with survival after breast cancer diagnosis in women from nine European countries: results from the EPIC cohort study

Abstract

Background: The Mediterranean diet has been associated with lower risk of breast cancer (BC) but evidence from prospective studies on the role of Mediterranean diet on BC survival remains sparse and conflicting. We aimed to investigate whether adherence to Mediterranean diet prior to diagnosis is associated with overall and BC-specific mortality.

Methods: A total of 13,270 incident breast cancer cases were identified from an initial sample of 318,686 women in 9 countries from the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Adherence to Mediterranean diet was estimated through the adapted relative Mediterranean diet (arMED), a 16-point score that includes 8 key components of the Mediterranean diet and excludes alcohol. The degree of adherence to arMED was classified as low (score 0-5), medium (score 6-8), and high (score 9-16). Multivariable Cox proportional hazards models were used to analyze the association between the arMED score and overall mortality, and Fine-Gray competing risks models were applied for BC-specific mortality.

Results: After a mean follow-up of 8.6 years from diagnosis, 2340 women died, including 1475 from breast cancer. Among all BC survivors, low compared to medium adherence to arMED score was associated with a 13% higher risk of all-cause mortality (HR 1.13, 95%CI 1.01-1.26). High compared to medium adherence to arMED showed a non-statistically significant association (HR 0.94; 95% CI 0.84-1.05). With no statistically significant departures from linearity, on a continuous scale, a 3-unit increase in the arMED score was associated with an 8% reduced risk of overall mortality (HR_3-unit 0.92, 95% CI: 0.87-0.97). This result sustained when restricted to postmenopausal women and was stronger among metastatic BC cases (HR_3-unit 0.81, 95% CI: 0.72-0.91).

Conclusions: Consuming a Mediterranean diet before BC diagnosis may improve long-term prognosis, particularly after menopause and in cases of metastatic breast cancer. Well-designed dietary interventions are needed to confirm these findings and define specific dietary recommendations.

Keywords: Breast cancer; Cancer survivors; Dietary patterns; Mediterranean diet; Prospective studies.

Fig. 1

Associations between arMED score and overall survival among subgroups of breast cancer survivors. Abbreviations: BC, breast cancer; BMI, body mass index; HR, hazard ratio; CI, confidence interval; ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; DRRD, Diabetes Risk Reduction Diet; ISD, Inflammatory Score of the Diet; N, number of breast cancer cases. ¹Each model has excluded of their stratification/adjustment the specific variable that represents the subgroup in each case. Models stratified by country, stage (metastatic, non-metastatic, unknown) and menopausal status at diagnosis and adjusted for age at diagnosis, attained level of education, physical activity, body mass index (modeled as restricted cubic spline with five internal knots placed at equally spaced percentiles), alcohol consumption reported at recruitment, smoking habit and intensity as cigarettes per day at recruitment, ever use of hormones for menopause at diagnosis, grade of differentiation, and tumor receptor status: ER, PR, HER2. ²Only BC survivors with known status of the receptor (positive or negative) have been considered. ³Categories of physical activity. Inactive: includes moderately inactive and inactive. Active: includes moderately active and active

Fig. 2

Direct adjusted cumulative incidence function curves for overall mortality by categories of adherence to arMED score. HR hazard ratio, CI confidence interval, arMED adapted relative Mediterranean diet. Categories of arMED: low adherence, 0–5; medium adherence, 6–8; high adherence, 9–16. HRs and CI 95% from multivariable Cox model stratified by country, stage (metastatic, non-metastatic, unknown) and menopausal status at diagnosis and adjusted for age at diagnosis, attained level of education, physical activity, body mass index (modeled as restricted cubic spline), alcohol consumption reported at recruitment, smoking habit and intensity at recruitment, ever use of hormones for menopause at diagnosis, grade of differentiation, and tumor receptor status (ER, PR, HER2)