Table 1 Characteristics of included studies and quality score

Petrick et al. [13], 2015, USA

Population-based case–control study

Max: 7.5

Median^a:

0.8 for EAC,

0.9 for ESCC

274 EAC

191 ESCC

Primary invasive cases:

localized, regional, distant, unknown

Total flavonoids, six classes of flavonoids (anthocyanidins, flavan-3-ols, flavanones, flavones, flavanols, and isoflavones), lignans

A 104-item modified FFQ

Each were divided into 4 categories (mg/day):

Total flavonoids:

0–62.35,

62.36–103.39,

103.40–253.24, ≥253.25;

Anthocyanidins:

0–6.23,

6.24–10.11,

10.12–16.23,

≥16.24;

Flavan-3-ols:

0–10.90,

10.91–26.67,

26.68–210.51, ≥210.52;

Flavanones:

0–8.63,

8.64–32.94,

34.95–49.00,

≥49.01;

Flavones:

0–1.20,

1.21–1.81,

1.82–2.64,

≥2.65;

Flavonols:

0–8.16,

8.17–12.30,

12.31–19.34,

≥19.35;

Isoflavones:

0–0.31,

0.32–0.46,

0.47–0.62,

≥0.63;

Lignans:

0–0.044,

0.045–0.060,

0.061–0.079,

≥0.080

All-cause mortality

For EAC:

Total flavonoids:

0.98 (0.68, 1.41);

Anthocyanidins:

0.87 (0.60, 1.26);

Flavan-3-ols:

0.93 (0.65, 1.33);

Flavanones:

1.15 (0.79, 1.68);

Flavones:

0.83 (0.58, 1.19);

Flavonols:

0.94 (0.65, 1.37);

Isoflavones:

0.75 (0.49, 1.13);

Lignans:

0.78 (0.54, 1.14)

For ESCC:

Total flavonoids:

0.91 (0.58, 1.44);

Anthocyanidins:

2.272 (0.66, 1.56);

Flavan-3-ols:

1.09 (0.69, 1.74);

Flavanones:

1.24 (0.76, 2.03);

Flavones:

2.272 (0.64, 1.54);

Flavonols:

0.93 (0.61, 1.40);

Isoflavones:

0.97 (0.60, 1.58);

Lignans:

0.61 (0.39, 0.96)

Cancer stage and dietary energy intake

Miles et al. [12], 2016, USA

Population-based case–control study

Median: 12.1

108 EC

74 EAC

Well differentiated, Poorly differentiated, Undetermined

Sugary beverages including soft drinks and fruit juices (classified into SB1 and SB2^b)

NCI-block FFQ

Median for soft drinks and fruit juices(g/day):

soft drinks:4.0;

fruit juices:0.71

SB1 intake quartile point(g/day):

(Q1)0.71,

(Q2)11.81,

(Q3) 40.00;

SB2 intake quartile points(g/day)

(Q1)3.04,

(Q2)20.76,

(Q3)45.29

All-cause mortality

For EC:

soft drinks:

2.29 (1.32,3.93);

fruit juices:

2.39 (1.34,4.30);

SB1:

2.58 (1.45,4.60);

SB2:

1.94 (1.06,3.53)

For EAC:

soft drinks:

1.84 (0.92,3.68);

fruit juices:

1.60 (0.79,3.25);

SB1:

1.51 (0.72,3.16);

SB2:

1.44 (0.57,3.62)

Age, sex, ethnicity, education, smoking, alcohol drinking, caloric intake, pathology type, and tumor differentiation grade

Shi et al. [30], 2018, China

A new established patient cohort

Median: 4.08

185 ESCC

AJCC stage: Only included I and II stage

Preserved vegetables

A modified FFQ

2 categories (time/week):

< 1 and ≥ 1

All-cause mortality

1.58 (1.01,2.47)

Age and sex

Lu et al. [24], 2011, China

A new established patient cohort

Median: 3.5, (Min-Max: 0.03–4.66)

120 ESCC patients underwent esophagectomy

T stage: T1~T4

N stage: N0, N1

M stage: M0, M1

clinical stage: 1/2, 3/4

Folate

NIH-modified FFQ

3 categories(ug/day):

< 30.0,

30.0–95.4,

≥95.5

Cancer-specific mortality

0.39 (0.20,0.78)

Age, sex, TNM stage

Jing et al. [25], 2012, China

A new established patient cohort

Median: 3.25,

(Min-Max: 0.25–5)

167 ESCC

T stage: T1~T4

N stage: N0, N1

M stage: M0, M1

Folate

A 65-item self-administered structured questionnaire

3 categories(ug/day):

< 230,

230–300,

> 300

Cancer-specific mortality:

0.45 (0.18,0.87)

Age, sex, smoking, drinking, tumor sites, TNM stage, chemo-therapy and radio-therapy

Trivers et al. [23], 2005, USA

population- based, case-control study

Max: 7.5

Median ^a:

0.8 for EAC and 0.9 for ESCC

293 EAC

220 ESCC

Incident invasive cases: localized, regional, distant, unknown

Alcohol

Baseline interviews

2 categories:

non-drinkers

and ever drinkers ^c

All-cause mortality ^d:

EAC:

1.08 (0.81,1.44)

ESCC:

1.77 (0.93,3.35)

None

Park et al. [28], 2006, South Korea

A cohort of male participated in a national health examination program

Mean:2.05 (Max:6.8)

272 EC

Not collected

Alcohol

A self-administered questionnaire

3 categories(g of alcohol/week):

0,

0–124.1,

≥124.2

All-cause mortality

1.44 (0.81,2.55)

Age, BMI, fasting serum glucose level, cholesterol level, physical activity, food preference, blood pressure, and other comorbidities

Samadi et al. [33], 2007, Iran

Patients initially diagnosed in Aras Clinic

Max: 5

122 EC

Differentiation:well, moderate/ poor, nondifferentiated

Alcohol

Questionnaire completed at the time of diagnosis

2 categories: no and yes

All-cause mortality:

7.51(0.82,69.10)

Age, sex, residence, treatment, smoking, opium use, differentiation, education

Sundelof et al. [14], 2008, Sweden

Nationwide case-control study

From 1994.12.1–1997.12.31 to 2004.12.31

177 EAC

159 ESCC

TNM stage: I, II, III, IV

Alcohol (including beer, wine and liquor)

Computer-aided face-to-face interview with separate questions

4 categories(g of pure alcohol/week): never, 1–15, 16–70, > 70

All-cause mortality

EAC:1.0 (0.5,1.7)

ESCC:0.6 (0.3,1.4)

Age, sex, educational level, symptomatic reflux, BMI, smoking, physical activity, tumor stage and for esophagectomy

Shitara et al. [26], 2010, Japan

Patients in Aichi Cancer Center Hospital

Median:5.6 (Min-Max: 2.1–7.9)

363 ESCC

UICC stage: I, II, III, IV

Alcohol

HERPACC questionnaire

2 categories^e (g of ethanol/week): <230 and ≥ 230

All-cause mortality

0.85(0.61,1.18)

Age, sex, smoking, ECOG PS, tumor length, UICC stage, treatment

Lu et al. [24], 2011, China

A new established patient cohort

Median: 3.5, (Min-Max: 0.03–4.66).

120 ESCC patients underwent esophagectomy

T stage: T1~T4

N stage: N0, N1

M stage: M0, M1

clinical stage: 1/2, 3/4

Alcohol

Risk factor questionnaire

2 categories:

Never drinkers and ever drinkers

Cancer-specific mortality:

1.02(0.61–1.72)

None

Jing et al. [25], 2012, China

Patients in General Hospital of Chengdu Military Area

Median:

3.25

(Min-Max: 0.25–5)

167 ESCC

T stage: T1~T4

N stage: N0, N1

M stage: M0, M1

Alcohol

A 65-item self-administered structured questionnaire

4 categories^f (g of ethanol/week): never, former, moderate, heavy

Cancer-specific mortality: 1.42(0.83,1.84)

None

Thrift et al. [29], 2012, Australia

Population-based case-control study

Median: 6.4

(Min-Max: 4.8–8.9)

301 ESCC

AJCC stage: I, II, III, IV

Alcohol (including light beer, regular beer, white wine, red wine, port/sherry and spirits/liqueur)

A health and life style questionnaire

4 categories(g of ethanol/week) < 10, 10–60, 70–200, ≥210

All-cause mortality ESCC:

2.08 (1.18,3.69);

Age, sex, AJCC stage, treatment intent, number of comorbidities and smoking

Thrift et al. [31], 2012, Australia

Population-based case-control study

Median: 6.4

(Min-Max: 4.8–8.9)

362 EAC

AJCC stage: I, II, III, IV

Alcohol (including light beer, regular beer, white wine, red wine, port/sherry and spirits/liqueur)

A health and life style questionnaire

4 categories(g of ethanol/week)

< 10, 10–60, 70–200, ≥210

All-cause mortality

EAC: 1.02(0.64,1.64)

Age, sex, AJCC stage, treatment intent, number of comorbidities and smoking

Wu et al. [32], 2013, China

Hospital-based case-control study

Max: 5

718 ESCC

AJCC stage: I + II, III + IV

Alcohol

A standardized questionnaire

2 categories^g:

users and non-users

All-cause mortality: 1.30(1.01,1.67);

Age, sex, education levels and AJCC stages

Huang et al. [27], 2014, China

Prospectively created esophageal carcinoma database

Median:5.3

2151EC

1851 ESCC,

AJCC stage: 0 + I, II, III

Alcohol (including wine, spirit and beer)

Baseline interviews

4 categories^h (g of alcohol/week):

0, 0–90.09, 91–272.09, ≥273

All-cause mortality ^d:

EC: 1.46(1.19,1.79)

ESCC: 1.37(1.11,1.70)

Age, sex, weight loss, stage, radicality of surgery, adjuvant treatment, smoking

Ma et al. [15], 2016, China

Esophageal cancer database of the Department of Thoracic Surgery at Sun Yat-sen University Cancer Center

Median: 6.5 (Min-Max:1–20)

643 ESCC with negative lymphatic metastasis having undergone esophagectomy

Post-operation staging of AJCC stage: IA, IB, IIA

Alcohol

Medical records

2 categoriesⁱ:

non-drinkers and drinkers

All-cause mortality ^d:

1.58(1.21,2.07)

Age, sex, smoking, family history, tumor location, surgery technique, post-operation staging, tumor grade