Table 3.
Study characteristics and results by reported outcomes (order by outcomes, study design and author).
| Mortality and BC Recurrence | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Study | Country of Study |
Study Design |
Sample Size/Number in Analysis |
Dietary Assessment and/or MD Adherence Assessment |
Duration/ Follow-Up |
Exposure | Comparator | Main Result (HR/RR, 95% CI) |
Variables Used for Adjustment |
| Di Maso et al., 2020 [37] | Italy | Cohort | 1453/1453 (<5% missing data on BMI and education) | FFQ (2 years before BC diagnosis) and 9-component MD score [26] |
Truncated 15 years after diagnosis (cohort follow-up median: 12.6 years; maximum: 16.8 years) |
MDS: 6−9 | MDS: 0−3 | All-cause mortality: HR-adjusted: 0.72 (0.32–0.92) Breast-cancer mortality: HR-adjusted: 0.83 (0.62–1.11) Non-breast cancer mortality: HR-adjusted: 0.58 (0.36–0.93) |
Age (at diagnosis), total energy intake, years of education, menopausal status, TNM stage, ER/PR status, area of residence and calendar period at diagnosis, (BC and non-BC mortality further adjusted for competing risk according to Fine–Gray model) |
| Ergas et al., 2021 [36] | USA | Cohort | 4505/3660 | FFQ and aMED-diet score (adapted from 9-component MD score) [26] |
Recruit 2005–2013, end of follow-up in December 2018, mean: 9.08 years (SD 2.77) |
aMDS: 6–9 | aMEDS: 0–2 | Breast-cancer recurrence: HR model 2: 1.08 (0.79–1.47) All-cause mortality: HR model 1: 0.56 (0.43–0.71) HR model 2: 0.79 (0.61–1.03) HR model 3: 0.87(0.66–1.14) Breast-cancer mortality: HR model 2: 0.79 (0.54–1.16) Non-breast cancer mortality: HR model 2: 0.73 (0.5–1.05) |
Model 1 (minimally adjusted): age (at diagnosis), total energy intake. Model 2 (medium-adjusted): model 1 + race and ethnicity, education, menopausal status, cancer stage, ER and PR status, physical activity, smoking, HER2 status. Model 3 (maximally adjusted): model 2 + BMI, surgery type, chemotherapy, radiation, HT |
| Karavasiloglou et al., 2019 [19] | Switzerland and USA | Cohort | 110/110 | 24-h dietary recall and 9-component MD score [26] |
Recruit 1988–1994, end of follow-up on 31 December 2011, mean 14.2 years (SEM 0.8) |
MDS: 5–9 | MDS: 0–4 | All-cause mortality: HR model 1: 0.47 (0.29–0.76) HR model 2: 0.78 (0.47–1.32) |
Model 1 (minimally adjusted): age (questionnaire completion), race/ethnicity. Model 2 (maximally adjusted): model 1 + race and ethnicity, total energy intake, BMI, moderate-to-vigorous physical activity, smoking, marital status, socioeconomic status, history of menopausal-hormone-therapy used, period since cancer diagnosis, prevalent chronic diseases |
| Kim et al., 2011 [38] | USA | Cohort | 6367/2729 | FFQ and aMED-diet score (adapted from 9-components MD score, aMEDs) [26] |
Recruitment period 1978–1998, end of follow-up in June 2004 |
aMEDS Quintile 5 |
aMEDS Quintile 1 |
All-cause mortality: RR model 1: 0.74 (0.55–0.99) RR model 2: 0.87 (0.64–1.17) Breast-cancer mortality: RR model 1: 1.11 (0.74–1.66) RR model 2: 1.15 (0.74–1.77) Non-breast cancer mortality: RR model 1:0.58 (0.38–0.88) RR model 2: 0.8 (0.5–1.26) |
Model 1 (minimally adjusted): age, time since diagnosis. Model 2 (maximally adjusted): model 1 + race and ethnicity, energy, BMI, physical activity, smoking, menopausal status, cancer stage, physical activity, smoking, treatment (chemotherapy, radiation, TAM), oral-contraceptive use, postmenopausal-hormone-therapy use, multivitamin usage at first birth and parity, alcohol intake, weight change |
| QoL | |||||||||
| Study |
Country of
Study |
Study
Design |
Sample Size/Number in Analysis |
Dietary
Assessment and/or MD Adherence Assessment |
Duration/
Follow-up |
Exposure | Comparator |
Main result
(Mean, SD) |
Variables Used for Adjustment |
| Long Parma et al., 2022 [46]; (Zuniga et al., 2019 [47]; Ramirez et al., 2017 [48]) |
USA | RCT | I: 76, C: 77/ I: 60, C: 65 |
14-item PREDIMED questionnaire [50] |
6 months/ 12 months |
PREDIMED 6-month mean score (SD): 8.7 (0.3) Individualised anti-inflammatory dietary prescriptions and behaviour change, 6-month monthly workshops, 12-month monthly navigation, motivational interviewing and tailored newsletters |
PREDIMED 6-month mean score (SD): 7.6 (0.3) Minimal nutritional information and two telephone calls prior to assessment appointments. |
FACT-G a: p = 0.41 6-month: I 87.96 (12.48), C 84.47 (15.81) 12-month: I 85.21 (13.38), C 84.57 (16.42) FACT-G subscales: Social Well-Being p = 0.77 6-month: I 20.96 (5.37), C 20.54 (5.94) 12-month: I 20.77 (5.15), C 20.46 (6.17) Emotional Well-Being p = 0.76 6-month: I 20.91 (2.75), C 19.97 (3.67) 12-month: I 20.22 (3.23), C 19.76 (3.98) Functional Well-Being p = 0.98 6-month: I 21.76 (4.29), C 20.71 (5.24) 12-month: I 20.60 (4.68), C 20.78 (5.42) Physical Well-Being p = 0.62 6-month: I 24.13 (3.91), C 23.25 (4.42) 12-month: I 23.60 (4.09), C 23.57 (4.13) BCS b: p = 0.82 6-month: I 25.01 (5.38), C 24.15 (5.86) 12-month: I 24.77 (5.34), C 24.31 (6.37) CES-D c: p = 0.51 6-month: I 2.45 (2.18), C 2.65 (2.39) 12-month: I 2.85 (2.74), C 2.88 (2.70) Perceived Stress Scale d: p = 0.01 Baseline: I 21.77 (7.63), C 19.75 (7.60) 6-month: I 20.64 (7.61), C 20.32 (8.31) 12-month: I 21.59 (7.44), C 20.01 (8.23) (p = 0.019 for main effect in I: reduction between baseline and 6-month) |
None |
| Alvarez-Bustos et al., 2021 [39]; (Ruiz-Casado et al., 2020 [40]) |
Spain | Cross- sectional |
180/180 | 14-item PREDIMED questionnaire [50] |
NA | MDS > 7 | MDS ≤ 7 | Cancer related fatigue: No strong evidence for an association between adherence to the MD and cancer-related fatigue (numerical results were not reported) |
None |
| Barchitta et al., 2020 [41] | Italy | Cross- sectional |
68/68 | 14-item PREDIMED questionnaire [50] |
NA | PREDIMED ≥10 positive items |
PREDIMED ≤5 positive items |
No strong evidence for an association between MD adherence and overall QoL or QoL subscales (EORTC QLQ-C30 c) (numerical results were not reported) |
None |
| Porciello et al., 2020 [42]; (Porciello et al., 2019 [43]) |
Italy | Cross- sectional |
309/309 | 14-item PREDIMED questionnaire [50] |
NA | PREDIMED > 7 | PREDIMED ≤ 7 | EORTC QLQ-C30 e subscales: Physical functioning: MDH 83.3 (14.5), MDL 78.9 (17.8), p = 0.02 β-model 1: 0.199, p = 0.001 β-model 2: 0.207, p = 0.001 β-model 3: 0.169, p = 0.006 Pain: MDH 23.1 (21.7), MDL 28.5 (24.3), p = 0.04 β-model 1: −0.175, p = 0.002 β-model 2: −0.174, p = 0.005 β-model 3: −0.131, p = 0.027 Dyspnoea: β-model 1: −0.115, p = 0.045 Insomnia: β -model 1: −0.114, p = 0.048 β -model 2: −0.131, p = 0.029 EQ-5D-3L Scale f: MDH 0.87 (0.11), MDL 0.84 (0.12), p = 0.05 β-model 1: 0.167, p = 0.004 β-model 2: 0.190, p = 0.003 (Results in other subscales and EORTC QLQ-B23 e are presented in Supplementary Materials Table S7) |
Model 1: age, cancer stag. Model 2: age, cancer stage, BMI, type of surgery, comorbidities, combined therapy. Model 3: age, cancer stage, smoking status, step count, education, civil status (married or single) |
| Health-Related Parameters | |||||||||
| Skouroliakou et al., 2017 [49] | Greece | RCT | I: 35, C: 35/ I: 26, C: 24 |
FFQ and 0–9 score (revised to include fish intake) [25] |
6 months | MDS at 6 months mean (SD): 7.65 (0.68) Personalized dietary intervention based on MD and physical-activity recommendations from ACS |
MDS at 6 months mean (SD): 4.44 (1.04) Updated American Cancer Society Guidelines |
BMI (kg/m2): p = 0.97 I 27.55 (4.69), C 27.73 (5.7) Body weight (kg): p = 0.89 I 72.69 (13.83), C 72.53 (15.61) Waist circumference (cm): p = 0.48 I 94.36 (11.37), C 96.97 (13.06) Blood glucose (mg/dL): p < 0.002 (ANCOVA p = 0.01) I 91.03 (9.96), C 105.95 (21.04) TC (mg/dL): p = 0.62 I 203.83 (44.56), C 209.15 (36.36) LDL-C (mg/dL): p = 0.56 I 123.18 (46.73), C 130.78 (34.39) HDL-C (mg/dL): p = 0.08 I 66.52 (17.56), C 57.36 (13.83) TAG (mg/dL): p = 0.86 I 89 (61.13), C 86.79 (43.74) |
Blood-glucose levels adjusted for BMI and estimated weekly MET-mins in ANCOVA analysis |
| Lorenzo et al., 2020 [45] | Spain | Cross- sectional |
90/67 | FFQ and 12 questions from the 14-item PREDIMED questionnaire [50] |
NA | PREDIMED (12 questions) >7 |
PREDIMED (12 questions) ≤7 |
BMI (kg/m2): p ≥ 0.05 MDH 27.8 (3.2), MDL 28.3 (5.7) Body weight (kg): p ≥ 0.05 MDH 68.9 (8.9), MDL 72.3 (14.1) Waist circumference (cm): p ≥ 0.05 MDH 87.8 (9.1), MDL 91.7 (15.3) Hip circumference (cm): p ≥ 0.05 MDH 106.3 (11.7), MDL 104.7 (11.7) Waist to hip ratio: p ≥ 0.05 MDH 0.82 (0.14), MDL 0.87 (0.18) Prevalence of obesity: p ≥ 0.05 MDH 68.9%, MDL 80% |
Age and BMI |
| Negrati et al., 2021 [44] | Italy | Cross- sectional |
139/80 | Diet score (range 0–55) [51] | NA | Diet score (range 0–55). Quartile 4: mean 38 |
Diet score (range 0–55). Quartile 1: mean 28.5 |
BMI (kg/m2): r = −0.110, p ≥ 0.05 MDH 29.3 (6.30), MDL 30.8 (6.20) Blood glucose (mg/dL): r = −0.216, p ≥ 0.05 MDH 85.3 (14.72), MDL 91.2 (17.32) Insulin: r = −0.20, p ≥ 0.05 MDH 8.7 (11.282), MDL 12.8 (4.69) HOMA-IR: r = −0.176, p ≥ 0.05 MDH 1.92 (3.05), MDL 3.06 (1.25) TC (mg/dL): r = −0.024, p ≥ 0.05 MDH 239.1 (31.08), MDL 230 (94.94), LDL-C (mg/dL): r = −0.192, p ≥ 0.05 MDH 132.4 (34.74), MDL 148.8 (33.09) HDL-C (mg/dL): r = −0.02, p ≥ 0.05 MDH 60.3 (10.58), MDL 59.1 (13.97) TAG (mg/dL): r = 0.11, p ≥ 0.05 MDH 143 (62.39), MDL 135.5 (68.44) |
None |
| Porciello et al., 2020 [42]; (Porciello et al., 2019 [43]) |
Italy | Cross- sectional |
309/309 | 14-item PREDIMED questionnaire [50] |
NA | PREDIMED > 7 | PREDIMED ≤ 7 | BMI (kg/m2): MDH 27.21 (6.13), MDL 28.24 (5.97) |
None |
aMEDS: Alternative Mediterranean Diet Score; BC: breast cancer; BMI: body-mass index; C: control group; CNT: cannot tell; ER: estrogen receptor; FFQ: Food Frequency Questionnaire; HDL-C: high-density-lipoprotein cholesterol; HER-2: human epidermal growth factor receptor 2; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HR: hazard ratio; HT: hormonal therapy; I: intervention group; LDL-C: low-density-lipoprotein cholesterol; MD: Mediterranean diet; MDH: high-MD-adherence group; MDL: low-MD-adherence group; MDS: Mediterranean Diet Score; NA: not applicable; NR: not reported; PR: progesterone receptor; QoL: quality of life; RCT: randomised controlled trial; RR: relative risk; SD: standard deviation; SEM: standard error of the mean; TAG: triacylglycerol; TAM: tamoxifen; TC: total cholesterol; WC: waist circumference; 95% CI: 95% confidence interval. a: FACT-G (Functional Assessment of Cancer Therapy–General): a 27-item questionnaire (range 0–108) designed to measure four domains of QoL in cancer patients, physical (range 0–28), social (range 0–28), emotional (range 0–24) and functional well-being (range 0–28). The higher the score, the better the QoL [52]. b: BCS: breast cancer subscale (range 0–40) of FACT-B (FACT-G + BCS). The higher the score, the better the QoL [53]. c: CES-D (Centre for Epidemiologic Studies Depression Scale): the higher the score, the greater frequency and number of depression symptoms (range 0–60) [54]. d: Perceived Stress Scale: the higher the score, the higher the stress (range 0–56: 0–18 low stress, 19–37 moderate stress, 38–56 high stress) [55]. e: EORTC QLQ-C30 (European Organization for Research and Treatment of Cancer Quality-of-Life Questionnaire Core 30) and EORTC QLQ-BC23 (breast-cancer module): include functional scales (a high score for a functional scale represents a high/healthy level of functioning), symptom scales and single items (a high score for a symptom scale/item represents a high level of symptomatology/problems) and a global health status/QoL scale (a high score represents a high QoL), range 0–100 for all of the scales individual items) [56]. f: EQ-5D-3L (European Quality of Life 5 Dimensions 3 Level): comprises the following five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. The digits for the five dimensions can be combined into a five-digit number and converted to a single summary index, with higher scores indicating higher health utility (0: a health state equivalent to death; negative: worse than death; 1: perfect health) [57].