Table 1 Main characteristics of the included studies

 Ammitzbøll 2016 [34], DCHS, Denmark, prospective cohort study, USA

n = 959

144 cases

Median (IQR), y = 10 (7)

Cancer registry;

Central Population Register.

Questionnaire

MET-h/week:

0

0–41

> 41–63

> 63–97

> 97–379

1.5 >(0.83–2.72)

1

0.75 (0.45–1.24)

0.83 (0.49–1.40)

0.75 (0.42–1.33)

BMI, baseline alcohol, smoking status, education, comorbidity, nodal status, operation type, chemotherapy, recreational and household PA.

Serious

Bao 2015 [35], SBCSS, prospective cohort study, China

n = 518

128 cases

Median (range), y = 9.1 (0.6–11.8)

Shanghai Cancer Registry; Shanghai Vital Statistics Registry.

Questionnaire

MET-h/week:

0

< 7.6

≥ 7.6

1

0.79 (0.5–1.27)

0.61 (0.41–0.91)

Age at diagnosis, BMI at baseline, education, marital status, menopausal status, Charlson comorbidity index, chemotherapy, radiotherapy, tumor-node metastasis stage, soy protein intake, tea consumption at baseline.

Moderate

 Bertram 2011 [28], WHEL, follow-up study of RCT, USA

n = 2361

195 cases

Median (range), y = 7.1 (1.0–10.8)

Medical records and death certificates, telephone interviews, confirmation obtained for > 95% of participants.

Questionnaire

MET-h/week:

0–2.5

2.5–7.5

7.5–14.9

14.9–24.7

24.7–107

1

1.01 (0.66–1.55)

0.85 (0.53- 1.35)

0.75 (0.46–1.23)

0.47 (0.26- 0.84)

Age and BMI at randomization, race, fruit and vegetable consumption, menopausal status, tumor type, tumor grade, tumor stage, anti-estrogen use, clinical site, time from diagnosis to randomization, hot flashes, and study group.

Moderate

Bradshaw 2014 [36], LIBCSP, prospective cohort follow-up of case-control study, USA

n = 1423

420 cases

Median (range), y = 12.7 (0.23–13.42)

Physician confirmed diagnosis;

National Death Index.

Interviewer-administered questionnaire

0

0.1–9

> 9

1

0.43 (0.20–0.83)

0.32 (0.23–0.47)

Age, pre-diagnosis BMI, chemotherapy treatment, tumor size, missing PA data.

Serious

Chen 2011 [37], SCR, prospective cohort study, China

n = 4826

436 cases

Median, y = 4.3

Population-based Shanghai Cancer Registry;

Annual linkage with the Shanghai Vital Statistics database.

Interview

MET-h/week:

0

< 8.3

> 8.3

1

0.81 (0.63–1.05)

0.65 (0.51- 0.84)

Date of birth, BMI at baseline, waist-to-hip ratio at baseline, menopausal status, income, education, quality of life, cruciferous vegetable intake, soy protein intake, tea consumption, chemotherapy, radiotherapy, tamoxifen use, tumor-node metastasis status, estrogen progesterone receptor status.

Moderate

 de Glas 2014 [30], TEAM-L, prospective cohort study, Netherlands

n = 435

58 cases

Followed until 2012 (from 2004 to 5)

Histologically/cytologically confirmed diagnosis.

Questionnaire

MET-h/week:

0–21

21.1–40

40.1–65.5

65.6–258

1

0.43 (0.19- 0.94)

0.60 (0.29- 1.24)

0.57 (0.26- 1.40)

Age at 1 year after diagnosis, number of comorbidities, tumor stage, node stage, BMI, and chemotherapy.

Moderate

Holick 2008 [38], CWLS, prospective cohort follow-up of case-control study, USA

n = 4482

412 cases

Mean, y = 5.5 ± 1.1

Self-reported and state cancer registries;

National Death Index.

Questionnaire

MET-h/week:

< 2.8

2.8–7.9

8.0–20.9

≥ 21.0

1

0.58 (0.45–0.76)

0.53 (0.40- 0.69)

0.44 (0.32- 0.60)

Age at diagnosis, stage of disease at diagnosis, state of residence, interval between diagnosis and PA assessment, BMI, post-diagnosis menopausal status, post-diagnosis hormone therapy use, total energy intake year before enrollment in the CWLS, education level at diagnosis, family history of breast cancer at diagnosis, and initial treatment modality (radiation, chemotherapy, tamoxifen).

Serious

 Holmes 2005 [39], NHS, prospective cohort study, USA

n = 2987

463 cases

Median, y = 8

Self-reported and confirmed from medical records and pathology reports, family, postal authorities;

National Death Index.

Questionnaire

MET-h/week:

< 3

3–8.9

9–14.9

15–23.9

≥24

1

0.71 (0.56- 0.89)

0.59 (0.41- 0.84)

0.56 (0.41- 0.77)

0.65 (0.48- 0.88)

Age, interval between diagnosis and PA assessment, smoking status, BMI, menopausal status and hormone therapy use, age at first birth and parity, oral contraceptive use, energy intake, energy-adjusted protein intake, disease stage, radiation treatment, chemotherapy, and tamoxifen treatment.

Serious

 Irwin 2008 [40], HEAL, prospective cohort study, USA

n post-diagnosis = 688

53 cases

Median (range), y = 6 (5–8)

Surveillance, Epidemiology, and End Results registries.

Questionnaire

MET-h/week:

0

> 0–8.9

≥ 9

1

0.36 (0.17–0.73)

0.33 (0.15–0.73)

Age, race, disease stage, initial treatment, tamoxifen use, BMI, and fruit/vegetable servings per day.

Serious

 Irwin 2011 [32], WHI, prospective cohort study, USA

n = 2910

186 cases

Mean (SD), y = 3.3 (1.8)

Physician-confirmed diagnosis, clinical center follow-up of participants and surrogates; National Death Index.

Questionnaire

MET-h/week:

0

> 0–3.0

3.1–8.9

≥ 9

1

0.42 (0.21–0.82)

0.72 (0.48- 1.07)

0.54 (0.38-0.79)

Age, stage, estrogen receptor, progesterone receptor, grade, human epidermal growth factor receptor 2, ethnicity, study arm, previous therapy use, time from diagnosis to PA assessment, BMI, diabetes, alcohol consumption, smoking, total calories, percentage calories from fat, servings of fruit and vegetables.

Moderate

 Maliniak 2018 [41], CPS-II NC, prospective cohort study, USA

n = 3689

185 cases

Median (IQR) = 7.5 (5.8)

Self-reported diagnosis of breast cancer; National Death Index. Both verified through medical records or state cancer registries.

Questionnaire

MET-h/week:

<  3.5

3.5 - <  8.75

8.75 - < 17.5

> 17.5

<  3.5

3.5 - <  8.75

8.75 - < 17.5

> 17.5

Age 45–64 years:

1.36 (0.89–2.07)

1

0.67 (0.44–1.02)

0.56 (0.37–0.83)

Age 65–92 years:

1.34 (1.14–1.58)

1

0.81 (0.67–0.99)

0.74 (0.61–0.90)

Age at diagnosis, race, calendar year of diagnosis, post-diagnosis BMI, Surveillance, Epidemiology, and End Results summary stage at diagnosis, post-diagnosis number of co-morbidities, post-diagnosis use of hormone replacement therapy, post-diagnosis alcohol intake, and post-diagnosis other cancer diagnosis, smoking status.

Serious

 Sternfeld 2009 [42], LACE, prospective cohort study, USA

n = 1970

187 cases

Mean (SD), mo = 87 (18)

Kaiser Permanente Northern California Cancer Registry, Utah Cancer Registry; death certificates.

Questionnaire

MET-h/week:

< 29

29- < 44

44- < 62

≥ 62

1

0.89 (0.59–1.33)

0.82 (0.54–1.25)

0.76 (0.48–1.19)

Age, BMI, number of positive nodes, stage, weight at 18 years, type of treatment, type of surgery, education level, smoking status.

Serious

 Glenn 2015 [31], SCCS, prospective cohort study, USA

n = 15,645/ m-f

2370 cases

Median (range), y = 6.2 (0.01–9.8)

Self-reported diagnosis; Social Security Administration vital status service for epidemiologic researchers and the National Death Index.

Questionnaire

MET-h/day:

<  6.9

6.9–14.1

14.2–24.8

>  24.9

1

0.77 (0.69–0.86)

0.66 (0.58–0.74)

0.64 (0.57–0.73)

Age, sex, race, BMI, income, education, comorbidities (hypertension, high cholesterol, myocardial infarction, stroke), smoking, insulin use, time since diagnosis, sedentary time.

Serious

 Gregg 2003 [43], NHIS, prospective cohort study, USA

n = 2896/ m-f

671 cases

y = 8

Self-reported diagnosis; National Death Index.

Interview

h/week:

0

> 0–1.9

≥ 2

1

0.95 (0.77-1.17).

0.71 (0.59-0.87)

Age, BMI, sex, race, self-rated health, smoking, weight loss approaches, hospitalizations, hypertension, use of antihypertensive medications, physician visits, limitations caused by cancer and CVD, functional limitations.

Serious

 Hu 2004 [33], six independent population surveys, prospective cohort study, Finland

n = 3316/ m-f

1410 cases

Mean, y = 18.5

Self-reported, hospital discharge diagnosis, or drug-treated cases in the Drug Registry; Statistics Finland.

Questionnaire

Low

Moderate

High

1

0.95 (0.81- 1.12)

0.96 (0.80- 1.15)

Age, sex and study year, BMI, systolic blood pressure, cholesterol, smoking, occupational PA, commuting PA; individuals with comorbidities, severe disease or disability at baseline, and who died the first two years of follow-up were excluded.

Serious

 Sluik 2012 [44], EPIC, prospective cohort study, 10 European Countries

n = 5859/ m-f

755 case

Median, y = 9.4

Self-reported diabetes confirmed from a physician, or use of medication, or self-reported confirmation during follow-up, or diabetes registries, or HbA1c level > 6% at baseline; Linkages with local, regional, or central cancer registries, boards of health, or death indices.

Questionnaire

MET-h/week:

< 45

45–74

75–113

> 113

1

0.85 (0.70–1.04)

0.80 (0.64–0.99)

0.73 (0.57–0.93)

Age, sex, study center, diabetes medication, disease duration, myocardial infarction, stroke, cancer, alcohol consumption, smoking behavior, education, energy intake, scores for dietary patterns.

Serious

 Sone 2013 [45], JDCS, prospective cohort study, Japan

n = 1702/ m-f

69 cases

Median, y = 8.05

HbA1c levels ≥ 6.5% (51 mmol/mol) referring to the Japan Diabetes Society; annual reports form.

Questionnaire

MET-h/week:

≤ 3.7

3.8–15.3

≥ 15.4

1

0.88 (0.47–1.64)

0.47 (0.22–0.99)

Age, sex, BMI, diabetes duration, smoking, energy/ethanol intake, dietary fiber, saturated fatty acid, type of occupation, HbA1c, systolic blood pressure, LDL-cholesterol, HDL-cholesterol, triacylglycerol, treatment (insulin, oral hypoglycaemic agents, antihypertensive agents or lipid-lowering agents).

Serious

 Tanasescu 2003 [46], HPFS, prospective cohort study, USA

n = 2803/ m

355 cases

y = 14

Self-reported physician’s diagnosis confirmed from diagnostic criteria from the National Diabetes Data Group; next of kin, work associates, postal authorities, and National Death Index.

Questionnaire

MET-h/week:

0–5.1

5.2–12.0

12.1–21.7

21.8–37.1

≥37.2

1

0.88 (0.64–1.21)

0.64 (0.45–0.91)

0.64 (0.45–0.90)

0.65 (0.45–0.93)

BMI, alcohol intake, smoking status, family history of myocardial infarction, use of vitamin E supplements, diabetes duration, diabetes medication, dietary intake of trans fat, saturated fat, fiber, and folate, history of angina and coronary artery bypass graft (CABG), hypertension at baseline, high serum cholesterol at baseline.

Moderate

 Cheng 2018 [14], HSE and SES, prospective cohort study, England and Scotland

n = 2398/ m-f

571 cases

Mean (SD), y = 8.5 (3.9)

Confirmed by spirometry according to GOLD criteria; National Health Service mortality data.

Questionnaire

MET-h/week:

0

< 3.75

3.75- < 7.5

≥ 7.5

1

0.86 (0.67–1.10)

0.75 (0.56–1)

0.56 (0.45–0.69)

Age, gender, BMI, COPD severity, history of CVD, cancer and diabetes, self-reported longstanding illness, smoking status, alcohol consumption, education.

Serious

 Garcia-Aymerich 2006 [47], CCHS, prospective cohort study, Denmark

n = 2386/ m-f

1425 cases

Mean (SD) = 12 (5.9)

Confirmed by a fixed FEV1/FVC ratio test; Danish National Board of Health.

Questionnaire

Very low

Low

Moderate

High

1

1.04 (0.84–1.27)

0.73 (0.61–0.86)

0.72 (0.59–0.86)

Age, gender, BMI, income, smoking status, glucose, systolic blood pressure FEV₁, IHD, myocardial infarction, stroke, asthma, sputum, asthma and smoking interaction.

Serious

 Gerber 2011 [48], ISFAMI, prospective cohort study, Israel

n = 1521/ m-f

427 cases

Median (IQR), y = 13.2 (12.0–13.5)

Index hospitalization; Israeli Population Registry, death certificates, hospital charts, family physicians and members.

Questionnaire

None

Irregularly

Regularly

1

0.71 (0.54–0.95)

0.56 (0.42–0.74)

Age, gender, hypertension, diabetes, dyslipidemia, smoking, obesity, chronic IHD, comorbidity index, Killip class, self-rated health, thrombolytic therapy, CABG, PTCA within 45 days, education, income, pre-MI employment, living with a steady partner, recurrent MI (and unstable angina pectoris, heart failure, CABG and PTCA) and cardiac rehabilitation during follow-up.

Moderate

 Janssen 2006 [49], CHS, prospective cohort study, USA

n = 1045/ m-f

489 cases

y = 9

Self-reported IHD confirmed from medications, medical records, clinical examinations; reviews of obituaries, medical records, death certificates, and the U.S. Health Care Financing Administration health care utilization database.

Interview

EE (kcal/week)

< 500

500–999

1000–1999

2000–2999

≥ 3000

1

0.87 (0.68–1.26)

0.77 (0.59–0.99)

0.54 (0.36–0.81)

0.63 (0.44–0.91)

Age, sex, race, smoking, alcohol, socioeconomic status, adiposity, prevalent disease (diabetes, lung disease, cancer, hypertension, stroke, congestive heart failure), and type of IHD (angina, MI, coronary revascularization).

Moderate

 Lahtinen 2017 [13], ARTEMIS, follow-up study of RCT, Finland

n = 1746/ m-f

147 cases

Median (IQR), mo = 54 (41–69)

Angiographically confirmed IHD with coronary stenosis > 50% of ≥ 1 coronary arteries; national death registries, mailing, telephone calls to family, electronic patient records.

Questionnaire

Inactive

Irregularly active

Active

Highly active

4.6 (2.1–10.0)

2.4 (1.1–5.3)

2.0 (0.9–4.3)

1

Age, gender, BMI at baseline, diabetes mellitus, history of MI, left ventricular ejection fraction, Canadian Cardiovascular Society grading for angina pectoris, cardiovascular event at 2-years follow-up, smoking status and alcohol consumption at 2 years.

Moderate

 Moholdt 2017 [50], HUNT, prospective cohort study, Norway

n = 6493/ m-f

3818 cases

Median (IQR), y = 12.5 (14.1)

Self-reported IHD; National Cause of Death Registry in Norway.

Questionnaire

Inactive

Low

Recommended

High

1

0.85 (0.79–0.92)

0.81 (0.72–0.90)

0.82 (0.70–0.95)

Age, examination year, smoking status, diabetes mellitus, alcohol consumption, hypertension, health status, PA for BMI, BMI for physical activity (stratified by sex).

Moderate

 Mons 2014 [51], KAROLA, prospective cohort study, Germany

n = 1038/ m-f

Mean (SD), y = 8.1 (3.1)

Physician-confirmed diagnosis; death certificates.

Questionnaire

Frequency/strenuous PA:

Rarely/never

1-4x/month

2-4x/week

5-6x/week

Daily

3.81 (2.17–6.70)

1.74 (0.59–3.21)

1

1.69 (0.81–3.50)

1.77 (0.90–3.47)

Age, gender, BMI, education, study site, employment status, cotinine-validated smoking status, self-reported poor health, history of MI, diabetes mellitus, hypertension, left ventricular function, number of affected vessels.

Serious

 Stewart 2017 [29], STABILITY, follow-up study of RCT, Global (39 countries)

n = 15,467/ m-f

1081 cases

Median (IQR), y = 3.79 (0.31)

IHD confirmed from coronary angiography.

Questionnaire

Median (IQR), MET-h/w:

14.0 ± 12.0

40.0 ± 14.0

90.0 ± 52.0

1

0.75 (0.65–0.87)

0.70 (0.60–0.82)

Age, gender, BMI, randomized treatment, systolic blood pressure, hypertension, geographic region for final reporting, prior myocardial infarction (MI), prior coronary revascularization percutaneous coronary intervention or coronary artery bypass graft, IHD, diabetes mellitus, smoking status, polyvascular disease, significant renal dysfunction, hemoglobin, white blood cell count, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, estimated glomerular filtration rate according to the Chronic Kidney Disease-Epidemiology research group calculator, congestive heart failure.

Serious

 Tian 2017 [52], CKB, prospective cohort study, China

n (IHD) = 13,936 / m-f

1442 cases

y = 7

Physician-diagnosed hypertension or CVD (stroke, transient ischemic attack, ischemic heart disease); China CDC’s Disease Surveillance Points system, local residential and health insurance records, active confirmation from street committee or village administrators.

Interviewer-administered questionnaire

MET-h/d:

< 8.91

8.91–16.53

≥ 16.53

1 (0.93 - 1.08)

0.80 (0.73–0.88)

0.59 (0.50–0.70)

Smoking status, alcohol consumption, fruit consumption, dairy products, meat and preserved vegetables, education, income, survey season, diabetes status, family history of CVD, CVD medication, poor health status (Analysis stratified by age-at-risk, gender, region, baseline CVD status)

Serious

 Wannamethee 2000 [53], BRHS, prospective cohort study, Great Britain

n = 772/ m

131 cases

Mean, y = 5

Self-reported physician-diagnosed IHD; National Health Service registers in Southport and Edinburgh.

Interviewer-administered questionnaire

Inactive/occasional

Light

Moderate

Moderate to vigorous

1

0.42 (0.25–0.71)

0.47 (0.24–0.92)

0.63 (0.39–1.03)

Age, smoking, social class, self-rated health status, diabetes mellitus, history of MI and stroke.

Moderate