Table 12 Studies examining the relationship between physical activity and cardiovascular disease.

Paffenbarger and Hale 1975 [47]

To evaluate the role of PA in reducing coronary mortality among longshoreman

• n = 6,351

22 years of follow up, or until reached the age of 75 yr

RR (95% CI) Sudden death

VPA is associated with reduced risk of coronary mortality, particularly sudden cardiac death.

USA

• Sex: Men

   • G1 = 1.00 (referent)

• Age: 35-74 yr

   • G2 = 3.5

• Characteristics: Longshoreman

PA assessment: Energy and oxygen cost requirements of longshoring jobs

   • G3 = 2.8

Prospective cohort

Delayed death

   • G1 = 1.00 (referent)

D & B score = 12

   • G2 = 1.4

Activity level

   • G3 = 1.5

G1 = Heavy (5.2-7.5 kcal/min)

Unspecified death

G2 = Moderate (2.4-5.0 kcal/min)

   • G1 = 1.00 (referent)

G3 = Light (1.5-2.0 kcal/min)

   • G2 = 1.1

   • G3 = 1.6

Outcome measure: Death from CHD

Manson et al 2002 [56]

To compare the roles of walking and vigorous exercise in the prevention of CV events in a large, ethnically diverse cohort of postmenopausal women.

• n = 73,743

Enrolment from 1994-98 Clinic visit for baseline screening,

   • Number of New Cases: 345

Both walking and VPA are associated with substantial reductions in the incidence of CHD events.

USA

• Sex: Women

   • Total Number of CVD events: 1551

• Age: 50-79 yr

• Characteristics: Healthy, Post Menopausal

Age adjusted RR (95% CI) Total exercise (MET-hr/wk)

PA assessment: Questionnaire for: Total exercise (MET- hr/wk)

Prospective cohort

G1 = 0-2.4

   • G1 = 1.00 (referent)

• Women's Health Initiative Observational Study

G2 = 2.5-7.2

   • G2 = 0.73 (0.53-0.99)

G3 = 7.3-13.4

   • G3 = 0.69 (0.51-0.95)

D & B score = 12

G4 = 13.5-23.3

   • G4 = 0.68 (0.50-0.93)

G5 = ≥ 23.4

   • G5 = 0.47 (0.33-0.67)

p = <0.001

Walking (MET-hr/wk)

Walking (MET-hr/wk)

   • G1 = 1.00 (referent)

G1 = None

   • G2 = 0.71 (0.53-0.96)

G2 = 0.1-2.5

   • G3 = 0.60 (0.44-0.83)

G3 = 2.6-5.0

   • G4 = 0.54 (0.39-0.76)

G4 = 5.1-10.0

   • G5 = 0.61 (0.44-0.84)

G5 > 10

p = 0.004

Time for VPA (min)

Vigorous exercise

G1 = None

   • G1 = 1.00 (referent)

G2 = 1-60

   • G2 = 1.12 (0.79-1.60)

G3 = 61-100

   • G3 = 0.56 (0.32-0.98)

G4 = 101-150

   • G4 = 0.73 (0.43-1.25)

G5 = >150

   • G5 = 0.58 (0.34-0.99)

p = 0.008

Outcome Measure: Incidence of CVD and CHD

Wisloff et al 2006 [58]

To study the association between the amount and intensity of exercise and CVD mortality.

• n = 56,072 (27,143 men; 28,929 women)

Length of follow-up: 16 ± 4 yr

   • Number of Cases: 1,603 male, 993 female

Men and women who exercise to a moderate degree and spend less than the recommended energy (< 1000 kcal/wk) are at lower risk of dying from heart disease than those who never exercise.

Norway

• Sex: Men and women

PA assessment: Questionnaire for LTPA, 4 groups

Multivariate RR (95% CI)

Prospective cohort

• Age: ≥ 20 yr

Men

• Characteristics: Free form CVD

Men

   • Q1 = 1.00 (referent)

Q1 = None

   • Q2 = 0.66 (0.50-0.87)

D & B score = 12

• HUNT study

Q2 = 1/wk >30 min high

   • Q3 = 0.83 (0.65-1.06)

Q3 = 2-3/wk > 30 min high

   • Q4 = 0.77 (0.59-1.01)

Q4 = ≥ 4/wk > 30 min high

Women

   • Q1 = 1.00 (referent)

Women

   • Q2 = 0.63 (0.31-1.29)

Q1 = None

   • Q3 = 0.66 (0.32-1.34)

Q2 = 1/wk ≤ 30 min low

   • Q4 = 0.86 (0.45-1.62)

Q3 = 1/wk ≤ 30 min high

Q4 = 2-3/wk ≤ 30 min low

Outcome Measure: Ischaemic heart disease mortality

Cox proportional HR

Lee et al 2001 [59]

To examine the relationship between PA (specifically walking) and CHD among women, including those at high risk for CHD.

• n = 39,372

Recruitment of Participants: Sept 1992-May 1995

   • Number of Cases: 244

Even light to moderate activity is associated with lower CHD rates in women.

USA and Puerto Rico

• Sex: Women

• Age: ≥ 45 yr

Multivariate RR (95% CI) Time spent walking

• Characteristics: Healthy

PA assessment: Questionnaires Divided into 4 or 5 groups:

   • G1 = 1.00 (referent)

• Women's Health Study

   • G2 = 0.86 (0.57-1.29)

As little as 1 hour of walking per week predicted lower risk.

Prospective cohort

   • G3 = 0.49 (0.28-0.86)

   • G4 = 0.48 (0.29-0.78)

p = <0.001

D & B score = 12

Time spent walking

G1 = No regular walking

Walking pace

G2 = 1-59 min/wk

   • G1 = 1.00 (referent)

G3 = 1.0-1.5 h/wk

   • G2 = 0.56 (0.32-0.97)

G4 = ≥ 2.0 h/wk

   • G3 = 0.71 (0.47-1.05)

Walking pace (km/h)

   • G4 = 0.52 (0.30-0.90)

G1 = No regular walking

p = 0.02

G2 = 3.2

G3 = 3.2-4.7

G4 = ≥ 4.8

EE (kcal/wk)

   • Q1 = 1.00 (referent)

EE (kcal/wk)

   • Q2 = 0.79 (0.56-1.12)

G1 = 200

   • Q3 = 0.55 (0.37-0.82)

G2 = 200-599

   • Q4 = 0.75 (0.50-1.12)

G3 = 600-1499 and

p = 0.03

G4 = 1500 or more

Energy expended VPA (kcal/wk)

Energy expenditure for VPA (kcal/wk)

   • G1 = 1.00 (referent)

G1 = No vigorous, <200 kcal/wk

   • G2 = 0.65 (0.46-0.91)

G2 = No vigorous, ≥ 200 kcal/wk

   • G3 = 1.18 (0.79-1.78)

   • G4 = 0.96 (0.60-1.55)

   • G5 = 0.63 (0.38-1.04)

G3 = Vigorous, 1-199 kcal/wk

G4 = Vigorous, 200-499 kcal/wk

G5 = Vigorous, ≥ 500 kcal/wk

Paffenbarger et al 1993 [67]

To analyze changes in the lifestyle of Harvard Alumni and the associations of these changes to mortality.

• n = 10,269

Baseline measure in 1962 or 1967 with a follow up in 1977

Alumni who increased their PA index to 2000 kcal or more per week had a 17% lower risk of death from CHD then those who were sedentary (p = 0.507)

Moderately vigorous sports activity was associated with lower rates of death from CHD among middle aged and older men

• Sex: Men

• Age: 45-84 yr

USA

• Characteristics: Health, Harvard College Alumni

Prospective cohort

PA assessment: Mailed questionnaires included questions on type, duration, intensity, frequency of PA.

Men who took up moderate took up moderately vigorous activity had a 41% lower risk than those who continued not to engage in such activity (p = 0.044)

D & B score = 13

Outcome Measure: CHD deaths between 1977 and 1985

Cox proportional hazards model

Poisson regression methods

The Mantel extension of the Mantel-Haenszel test

Haapanen et al 1997 [77]

To examine the association between duration and intensity of LTPA and the risk of CHD.

• n = 2,840 (1,500 men; 1,340 women)

Length of Follow-up: 10 yrs

   • Incident Rates (per 1000 person-years) for CHD = 108 for men and 75 for women.

Total EE had an inverse and independent association with risk of CHD in middle aged Finnish men but not among women.

Finland

• Sex: Men and women

PA assessment: Questionnaire for LTPA EE (kcal/wk)

Multivariate RR (95% CI) LTPA and CHD mortality

Prospective cohort

• Age: 35-63 yr

Men

• Characteristics: Healthy

Men

   • G1 = 1.98

G1 = 0-1100

   • G2 = 1.33

D & B score = 13

G2 = 1101-1900

   • G3 = 1.00 (referent)

G3 = >1900

Women

Women

   • G1 = 1.25

G1 = 0-900

   • G2 = 0.73

G2 = 901-1500

   • G3 = 1.00 (referent)

G3 = >1500

Outcome Measure: CHD mortality

Cox proportional HR

Barengo et al 2004 [164]

To investigate whether moderate or high LTPA are associated with a reduced CVD and all-cause mortality, independent of CVD risk factors and other forms of PA in men and women.

• n = 31,677 (15,853 men; 16,824 women)

20 year follow-up

   • Number of Cases (Men): 1,661

Moderate and high levels of LTPA and OPA are associated with reduced CVD mortality.

PA assessment: Questionnaire for LTPA and OPA, 3 groups

   • Number of Cases (Women): 778

Finland

• Sex: Men and women

HR (95% CI) LTPA, men

Prospective cohort

• Age: 30-59

G1 = Low activity

   • G1 = 1.00 (referent)

• Characteristics: Participant from eastern and south-western Finland

G2 = Moderate activity

   • G2 = 0.91 (0.82-1.00)

D & B score = 14

G3 = High activity

   • G3 = 0.83 (0.69-0.99)

LTPA, women (referent)

   • G1 = 1.00

   • G2 = 0.83 (0.71-0.96)

   • G3 = 0.89 (0.68-1.18)

OPA, men

   • G1 = 1.00 (referent)

   • G2 = 0.75 (0.64-0.87)

   • G3 = 0.77 (0.69-0.87)

OPA, women

   • G1 = 1.00 (referent)

   • G2 = 0.73 (0.60-0.88)

   • G3 = 0.77 (0.65-0.91)

Bijnen et al 1998 [166]

To describe the association between the PA pattern of elderly men and CHD mortality.

• n = 802

Length of Follow-up: 10

   • Number of Cases: 90

PA did not show a protective effect on death from CHD.

• Sex: Men

• Age: 64-84 yr

PA assessment: Questionnaire, divided into 3 groups

RR (95% CI)

Netherlands

• Characteristics: Free from Serious Illness

   • G1 = 1.00 (referent)

   • G2 = 0.63 (0.38-1.05)

Prospective cohort

G1 = Lowest

   • G3 = 0.85 (0.51-1.44)

• Ethnicity: Dutch

G2 = Middle

• Zutphen Elderly Study

G3 = Highest

D & B score = 13

Outcome Measure: CHD Mortality

Cox Proportional HR

Davey-Smith et al 2000 [174]

To examine the association between two measures of physical activity (LTPA and usual walking pace) with cause specific mortality (CHD).

• n = 6,702

Length of Follow-up: 25 yrs

   • Number of Cases: 955

Inverse associations of both LTPA and walking pace with mortality from CHD were seen.

• Sex: Men

• Age: 40-64 yr

RR (95% CI) by walking pace

England

• Whitehall Study

PA assessment: Questionnaire during examination for walking pace and LTPA

   • G1 = 1.45 (0.9-2.2)

   • G2 = 1.30 (1.1-1.6)

Prospective cohort

   • G3 = 1.00 (referent)

p < 0.01

D & B score = 11

Walking pace

Multivariate RR (95% CI) by LTPA level

G1 = Slower

   • G1 = 1.24 (1.0-1.5)

G2 = Same

   • G2 = 0.94 (0.8-1.2)

G3 = Faster

   • G3 = 1.00

p < 0.05

LTPA

G1 = Inactive

G2 = Moderate

G3 = Active

Outcome Measure: CHD Mortality

Cox Proportional HR

Eaton et al 1995 [175]

To determine whether self reported PA predicts a decreased risk of CHD.

• n = 8,463 (LTPA), 8,418 (OPA)

Length of Follow-up: 21 yrs

   • Number of Cases: 709

Baseline levels of self reported LTPA predicted a decreased rate of CHD.

Age adjusted RR (95% CI) by LTPA level

USA

• Sex: Men

PA assessment: Interview

   • G1 = 1.00 (referent)

• Age: 40 yr

   • G2 = 0.79 (0.63-0.99)

Prospective cohort

• Characteristics: Healthy, free of CHD

LTPA

   • G3 = 0.73 (0.59-0.89)

G1 = Sedentary

   • G4 = 0.71 (0.52-0.98)

G2 = Light

D & B score = 11

Ethnicity: Israeli

G3 = Light Daily

Age adjusted RR (95% CI) by OPA level

G4 = Heavy

   • G1 = 1.00 (referent)

   • G2 = 0.99 (0.75-1.18)

OPA

   • G3 = 0.94 (0.78-1.12)

G1 = Sitting

   • G4 = 0.87 (0.67-1.10)

G3 = Walking

G4 = Physical Labour

Outcome Measure: CHD Death

Cox Proportional HR

Hillsdon et al 2004 [183]

To examine whether a short, easily administered measure of PA is associated with the risk of death from all causes and specific causes.

• n = 10,522 (4,929 men; 5,593 women)

Length of Follow-up: > 10 yrs

   • Number of Cases: 155

Self reported VPA is associated with the risk of future mortality.

Multivariate RR (95% CI) by PA level

UK

• Sex: Men and women

PA assessment: Questionnaire, 3 groups:

   • G1 = 1.00 (referent)

   • G2 = 0.46 (0.19-1.12)

Prospective cohort

• Age: 35-64 yr

G1 = Never / <1 time/month

   • G3 = 0.96 (0.53-1.75)

• Characteristics: no history of chest pain

G2 = <2 times/wk

D & B score = 11

G3 = ≥ 2 times/wk

Outcome Measure: IHD mortality

Cox proportional HR

Leon et al 1997 [199]

To study the relationship of PA to CHD in a well defined population at above average risk for CHD over a 16 yr observation period.

• n = 12,138

Follow up for 16 years

Age Adjusted RR (95% CI)

A relatively small amount (10-36 min/d) of daily moderate intensity LTPA can significantly reduce premature mortality from CHD in middle aged men at high risk for CHD.

USA

• Sex: Men

   • G1 = 1.00 (referent)

• Age: 35-57 yr

PA assessment: Questionnaire at baseline (Minnesota LTPA questionnaire), divided/grouped into deciles of LTPA (min/d)

   • G2 = 0.71 (0.56-0.91)

• Characteristics: Free of CHD but in the upper 10-15% of a CHD probability risk score

   • G3 = 0.75 (0.59-0.96)

   • G4 = 0.69 (0.54-0.96)

Prospective cohort

Multivariate adjusted RR (95% CI)

   • G1 = 1.00 (referent)

D & B score = 11

G1 = D1: (0-9 min/d)

   • G2 = 0.75 (0.54-0.96)

Multiple risk factor intervention trial

G2 = D2-4: (10-36 min/d)

   • G3 = 0.81 (0.64-1.04)

G3 = D5-7: (37-75 min/d)

   • G4 = 0.75 (0.59-0.96)

G4 = D8-10: (76-359 min/d)

Outcome Measure: CHD Mortality

Rosengren et al 1997 [211]

To examine the long term effect of OPA and LTPA on the risk of death from CHD.

• n = 7,142

Length of Follow-up: 20 yrs

Number of Cases: 684

There appears to be a protective effect of LTPA on CHD-related death.

• Sex: Men

• Age: 47-55 yr

Multivariate RR (95% CI) for LTPA

Sweden

• Characteristics: Swedish men

PA assessment: Questionnaire for LTPA, 3 groups

   • G1 = 1.00 (referent)

   • G2 = 0.84 (0.71-1.00)

Prospective cohort

   • G3 = 0.84 (0.73-0.96)

G1 = Sedentary

G2 = Moderately active

D & B score = 13

G3 = Regular exercise

Outcome Measure: CHD death

Proportional HR

Schnohr et al 2006 [214]

To describe the associations between different levels of LTPA and subsequent causes of death.

• n = 4,894 (2,136 men; 2,758 women)

Participants included in the study were only those whose PA levels did not change over 5 years

   • Number of Cases: 292

There was an inverse and significant dose- response association between LTPA and CHD-related mortality.

Adjusted RR (95% CI) Whole group

Denmark

• Sex: Men and women

   • G1 = 1.00 (referent)

   • G2 = 0.71 (0.51-0.99)

Prospective cohort

Age: 20--79 yr

PA assessment:

   • G3 = 0.56 (0.38-0.82)

• Characteristics: Healthy

Questionnaire LTPA

D & B score = 12

• Copenhagen City Heart Study

G1 = <4 METS

Men

G2 = 4-6 METS

   • G1 = referent

G3 = >6 METS

   • G2 = survived 4.9 yrs longer

   • G3 = survived 6.8 yrs longer

Cox proportional HR

Women

   • G1 = referent

   • G2 = survived 5.5 yrs longer

   • G3 = survived 6.4 yrs longer

Weller et al 1998 [220]

To examine the relationship between PA and mortality.

• n = 6,620

Length of Follow-up: 7 yrs

   • Number of Cases: 109

LTPA is inversely associated with risk of fatal MI.

• Sex: Women

• Age: ≥ 30 yr

PA assessment:

OR (95% CI) by LTPA

Canada

• Characteristics: Canadian Women

Questionnaire, 4 groups for LTPA (kcal/kg/day) and non-LTPA (kcal/kg/day)

   • Q1 = 1.00 (referent)

   • Q2 = 0.61 (0.07-1.19)

Prospective cohort

   • Q3 = 0.84 (0.52-1.37)

   • Q4 = 0.63 (0.36-1.09)

D & B score = 9

LTPA (kcal/kg/day)

OR (95% CI) by non-LTPA

Q1 = ≥ 0

   • Q1 = 1.00 (referent)

Q2 = ≥ 0.1

   • Q2 = 0.71 (0.44-1.16)

Q3 = ≥ 0.5

   • Q3 = 0.57 (0.33-0.97)

Q4 = ≥ 1.6

   • Q4 = 0.49 (0.26-0.92)

Non-LTPA (kcal/kg/day)

Q1 = ≥ 0

Q2 = ≥ 2.8

Q3 = ≥ 5.9

Q4 = ≥ 9.9

Outcome Measure: Fatal MI

Logistic regression analysis

Yu et al 2003 [221]

To examine the optimal intensity of LTPA to decrease the risk of CHD mortality in middle aged British men.

• n = 1,975

10 year follow-up

   • Number of Cases: 82

Strong significant inverse relationship between heavy LTPA and CHD mortality.

• Sex: Men

PA assessment: Questionnaire (Minnesota LTPA questionnaire), 3 groups

Multivariate adjusted HR (95% CI)

• Age: 49-64 yr

   • G1 = 1.00 (referent)

UK

• Characteristics: Healthy, no previous history of CHD

   • G2 = 0.74 (0.44-1.25)

   • G3 = 0.55 (0.31-0.98)

Prospective cohort

p = 0.039

Relationship was not significant for low- moderate intensity LTPA and OPA.

• Caerphilly collaborative heart study

Total activity level (kcal/day)

D & B score = 11

G1 = 0.0 - 161.6

G2 = 161.8 - 395.3

G3 = 395.5 - 2747.2

Cox proportional HR

Altieri et al 2004 [222]

To assess the possible protective role of PA on CHD.

• n = 985 (507 men; 478 women)

PA assessment: Questionnaire for OPA, divided into quartiles

Number of Cases: 507

LTPA from 15-19 yrs as well as OPA from 30 - 39 yrs both have a significant inverse relationship with risk of non fatal acute MI.

OR (95% CI) for CHD and OPA

Italy

• Sex: Men and women

Q1 = lowest

   • Q1 = 1.00 (referent)

Q2

   • Q2 = 0.63 (0.39-1.03)

Case Control

• Age: < 79 yr

Q3

   • Q3 = 0.56 (0.35-0.90)

• Characteristics: Case: Patients admitted to Hospital with non-fatal Acute MI. Controls: Patients admitted to hospital for acute condition unrelated to known or potential risk factors for acute MI

Q4 = highest

   • Q4 = 0.57 (0.34-0.95)

D & B score = 11

p = 0.045

Outcome Measure: Non Fatal acute MI

Unconditional logistic regression

Batty et al 2003 [223]

To examine the relationship between physical activity and three mortality endpoints in healthy persons.

• n = 6,474

Length of Follow-up: 25 yr

   • Number of Cases: 837

A suggestion that the symptomatic nature of ischemia appeared to modify the affects of

• Sex: Men

   • Number of Dropouts: 158

• Age: 40-64 yr

PA assessment: Questionnaire for LTPA, divided into 3 groups:

UK

• Characteristics: British civil servants who underwent a resting ECG

HR (95% CI) for CHD and LTPA

   • G1 = 1.14 (0.9-1.4)

PA on total and CHD mortality.

Prospective cohort

G1 = Inactive

   • G2 = 0.94 (0.8-1.1)

G2 = Moderate

   • G3 = 1.00 (referent)

G3 = Active

D & B score = 13

Outcome Measure: CHD mortality

Cox Proportional HR

Chen and Millar [224]

To examine the potential protective effect of LTPA on the incidence of heart disease and depression.

• n = 15,670

Length of Follow-up: 2 yrs

• 100 cases

Regular and at least MPA can be beneficial to heart health.

• Sex: Men and women

• Age: ≥ 20 yr

PA assessment: EE from self administered questionnaire, 4 groups (kcal/kg/day)

Adjusted OR (95% CI)

Canada

• Characteristics: Healthy and free from heart disease

   • G1 = 5.0 (1.84-13.59)

   • G2 = 3.7 (1.26-10.67)

Prospective cohort

   • G3 = 1.00 (referent)

G1 = Sedentary

   • G4 = 1.3 (0.41-3.89)

G2 = Light (<1.5)

D & B score = 11

National Population Health Survey

G3 = Moderate (1.5-2.9)

G4 = Active (≥ 3)

Outcome Measure: CHD incidence

Multiple logistic regression

Conroy et al 2005 [225]

To examine the relationship between 1) PA during young adulthood and middle age, and 2) PA during each time period and CHD during middle age and older women.

• n = 37,169

Length of Follow-up: 9 yrs

   • Number of Cases: 477

PA during middle age predicts lower risk of CHD

• Sex: Women

• Age: ≥ 45 yr

Multivariate RR (95% CI) Baseline PA and incidence of CHD

US

• Characteristics: Healthy women health professionals

PA assessment: Questionnaire for EE (kcal/wk) and months/yr

• Women's Health Study

   • G1 = 1.00 (referent)

Cohort study

   • G2 = 0.62 (0.48-0.80)

   • G3 = 0.61 (0.48-0.79)

D & B score = 11

Baseline PA (kcal/wk)

   • G4 = 0.61 (0.46-0.81)

G1 = <200

p = <0.001

G2 = 200-599

G3 = 600-1499

Past PA and incidence of CHD

G4 = ≥ 1500

   • G1 = 1.00 (referent)

   • G2 = 0.76 (0.57-1.02)

Past PA

   • G3 = 0.95 (0.72-1.24)

Months per year

   • G4 = 1.04 (0.78-1.39)

G1 = 0

   • G5 = 0.81 (0.58-1.14)

G2 = 1-3

G3 = 4-6

G4 = 7-9

G5 = 10-12

Outcome Measure: Incidence of CHD

Cox proportional hazard regression

Dorn et al 1999 [226]

To examine the long-term relationships between total PA and mortality from all causes and CHD in the general population.

• n = 1,461 (698 men; 763 women)

Length of Follow-up: 29 years

   • Number of Cases: 109 men, 81 women

PA favorably influences mortality risks in non- obese men and younger women.

USA

• Sex: Men and women

PA assessment: Questionnaire

Multivariate RR (95% CI) for PAI in non- obese men

Prospective cohort

• Age: 15-96 yr

   • 0.40 (0.19-0.88) for 1 kcal/kg/h

• Characteristics:

Outcome Measure: CHD

Multivariate RR (95% CI) for PAI in obese men

Healthy, free from CHD, diabetes, and Stroke.

Mortality

   • 1.86 (0.86-4.03) for 1 kcal/kg/h

D & B score = 11

Cox Proportional Hazard

Ratio

Multivariate RR (95% CI) for PAI in women < 60 yrs

• Ethnicity: White.

   • 0.42 (0.11-1.52) for 1 kcal/kg/h

Multivariate RR (95% CI) for PAI in women > 60 yrs

   • 1.78 (0.77-4.09) for 1 kcal/kg/h

Folsom et al 1997 [227]

To examine the association of PA at baseline with CHD incidence.

• n = 13,999 (6,166 men; 7833 women)

Length of Follow-up: 4-7 yrs

   • Number of Cases: 223 men, 97 women,

No significant relationships.

Multivariate RR (95% CI) LTPA, men

USA

• Sex: Men and women

PA assessment: Questionnaire during home interview, divided into quartiles of LTPA and sports activity

   • Q1 = 1.00 (referent)

Prospective cohort

• Age: 45-64 yr

   • Q2 = 1.08 (0.75-1.55)

• Characteristics: no CHD at baseline

   • Q3 = 0.83 (0.51-1.36)

   • Q4 = 0.89 (0.59-1.35)

D & B score = 9

Q1 = Low

• Ethnicity: Black and non Black

Q2

LTPA, women

Q3

   • Q1 = 1.00 (referent)

• Atherosclerosis Risk in Communities Study

Q4 = High

   • Q2 = 0.74 (0.42-1.31)

   • Q3 = 1.07 (0.55-2.09)

Outcome Measure: CHD incidence Poisson Regression

   • Q4 = 0.64 (0.34-1.24)

Multivariate RR (95% CI) Sports, men

   • Q1 = 1.00 (referent)

   • Q2 = 1.15 (0.79-1.68)

   • Q3 = 1.03 (0.68-1.54)

   • Q4 = 0.83 (0.56-1.23)

Sports, women

   • Q1 = 1.00 (referent)

   • Q2 = 0.99 (0.58-1.67)

   • Q3 = 0.64 (0.32-1.27)

   • Q4 = 0.72 (0.37-1.38)

Fransson et al 2004 [228]

To estimate the influence of LTPA and OPA on acute MI.

• n = 4069 (2,742 men; 1,327 women)

PA assessment: Questionnaire for LTPA, 5 groups

   • Number of Cases: 1,204 men, 550 women

Exercise seems to reduce the risk of MI.

Sweden

• Sex: Men and Women

G1 = Seldom

OR (95% CI)

G2 = Sometimes

Case Control

• Age: 45-70 yr

G3 = 1×/wk

LTPA, men

• Characteristics: Cases: Diagnosed with acute MI

G4 = 2-3×/wk

   • G1 = 1.00 (referent)

D & B score = 12

G5 = >3×/wk

   • G2 = 0.76 (0.61-0.95)

   • G3 = 0.67 (0.51-0.88)

   • G4 = 0.63 (0.49-0.83)

• Stockholm Heart Epidemiology

Questionnaire for total physical activity, 3 groups

   • G5 = 0.53 (0.38-0.73)

G1 = Passive

G2 = Somewhat active

LTPA, women

G3 = Active

   • G1 = 1.00 (referent)

Questionnaire for sitting at work, 3 groups

   • G2 = 0.69 (0.49-0.98)

   • G3 = 0.38 (0.25-0.58)

G1 = Less than half the time

   • G4 = 0.62 (0.38-1.01)

G2 = About half the time

   • G5 = 0.31 (0.15-0.66)

G3 = More than half the time

Total physical activity, men

   • G1 = 1.00 (referent)

   • G2 = 0.66 (0.47-0.94)

Outcome Measure: Acute MI

   • G3 = 0.46 (0.31-0.69)

Total physical activity, women

Conditional and unconditional logistics regression

   • G1 = 1.00 (referent)

   • G2 = 0.34 (0.22-0.53)

   • G3 = 0.16 (0.07-0.37)

Sitting at work, men

   • G1 = 1.00 (referent)

   • G2 = 0.91 (0.73-1.15)

   • G3 = 0.90 (0.72-1.12)

Sitting at work, women

   • G1 = 1.00 (referent)

   • G2 = 0.77 (0.51-1.17)

   • G3 = 0.47 (0.31-0.69)

Fransson et al 2006 [229]

To evaluate whether LTPA compensates for the increased risk of acute MI associated with overweight and obesity.

• n = 4069 (2,742 men; 1,327 women)

PA Assessment: Questionnaire for LTPA, 3 groups

Number of Cases: 1204 men, 550 women

Regular LTPA seems to provide protection against MI and non- fatal MI.

Multivariate OR (95% CI) for acute MI

Sweden

• Sex: Men and women

G1 = Very little /occasional walks

LTPA, men

   • G1 = 1.00 (referent)

Case Control

• Age: 45-70 yr

G2 = Occasional / once per week

   • G2 = 0.70 (0.58-0.84)

• Characteristics: Cases: had acute MI

   • G3 = 0.57 (0.46-0.71)

D & B score = 12

G3 = Twice per week or more

LTPA, women

   • G1 = 1.00 (referent)

Outcome measure: Acute MI

   • G2 = 0.52 (0.40-0.68)

   • G3 = 0.44 (0.30-0.65)

Multivariate OR (95% CI) for non-fatal MI

Conditional and unconditional logistics regression

LTPA, men

   • G1 = 1.00 (referent)

   • G2 = 0.79 (0.65-0.96)

   • G3 = 0.63 (0.50-0.79)

LTPA, women

   • G1 = 1.00 (referent)

   • G2 = 0.64 (0.48-0.86)

   • G3 = 0.58 (0.39-0.87)

Haapanen-Niemi 2000 [230]

To investigate the independent associations and the possible interaction of BMI LTPA and perceived physical performance and functional capacity with the risk of mortality.

• n = 2,212 (1,090 men; 1,122 women)

Length of Follow-up: 16 yrs

   • Number of Cases: 208 all cause deaths, 54% of those CVD. 73% of CVD deaths due to CHD

Increase perceived PF is associated with a reduced risk of CHD mortality in men.

Finland

• Sex: Men and women

PA assessment: Postal Survey

Multivariate RR (95% CI)

Prospective cohort

• Age: 35-63 yr

Total LTPA energy expenditure (kcal/wk)

Total LTPA EE index and CHD mortality, men

• Characteristics:

Healthy

   • G1 = 1.00 (referent)

G1 = High

   • G2 = 0.88 (0.44-1.76)

D & B score = 13

• Ethnicity:

Finnish

G2 = Moderate

   • G3 = 1.70 (0.90-3.21)

G3 = Low

p = 0.056

Perceived physical fitness compared to age-mates

Multivariate RR (95% CI) Perceived physical fitness, men

G1 = Better

   • G1 = 1.00 (referent)

G2 = Similar

   • G2 = 2.82 (1.06-7.46)

G3 = Worse

   • G3 = 4.64 (1.56-13.84)

Outcome Measure: CHD mortality

p = 0.011

Total LTPA EE index and CHD mortality, women

Cox proportional HR

   • G1 = 1.00 (referent)

   • G2 = 0.43 (0.16-1.16)

   • G3 = 1.17 (0.51-2.68)

p = 0.046

Multivariate RR (95% CI) Perceived physical fitness, women

   • G1 = 1.00 (referent)

   • G2 = 0.82 (0.32-2.16)

   • G3 = 1.89 (0.57-6.27)

p = 0.154

Kannel et al 1986 [231]

To examine the role of low levels of OPA and LTPA in the development of CV morbidity and mortality over the short and long term.

• n = 1,166

Length of Follow-up: 24 yrs

   • Number of Cases: 220 mortality, 371 morbidity

Rate of CHD Mortality and Morbidity decreases with increased level of PA but no association was found with physical demand of work

• Sex: Men

• Age: 45-65 yr

USA

• Characteristics:

PA assessment: Questionnaire during examination

Cumulative 24 year age adjusted rate per 1000 people

Prospective cohort

24 hr PA index for LTPA CHD mortality

PA index:

   • G1 = 255

D & B score = 11

G1 = <29

   • G2 = 184

G2 = 30-34

   • G3 = 152

G3 = >34

p < 0.01

Physical demand of work

24 hr PA index for LTPA CHD incidence

G1 = Sedentary

   • G1 = 414

G2 = Light

   • G2 = 353

G3 = Medium

   • G3 = 311

G4 = Heavy

Outcome Measure: CHD mortality and Morbidity

Physical demand of work and CHD mortality

   • G1 = 216

Cox proportional HR

   • G2 = 209

   • G3 = 169

   • G4 = 170

Physical demand of work and CHD incidence:

   • G1 = 355

   • G2 = 405

   • G3 = 307

   • G4 = 325

Kaprio et al 2000 [232]

To examine the contribution of genetic and other familial factors to the relationship between LTPA and CHD.

• n = 8,205

Length of Follow-up: 18 yrs

   • Number of Cases: 723

LTPA compared to being sedentary helps prevent CHD in men.

• Sex: Men

• Age: 25-69 yr

Multivariate RR (95% CI)

Finland

• Characteristics: Same sex twin pairs, free of CVD

PA assessment: Questionnaire for LTPA, 3 groups:

   • G1 = 1.00 (referent)

   • G2 = 0.84 (0.70-1.01)

Prospective cohort

   • G3 = 0.68 (0.50-0.92)

G1 = Sedentary

p = 0.010

G2 = Occasional

D & B score = 12

Exercisers

G3 = Conditioning

Exercisers

Outcome Measure: Hospitalization or death from CHD

Poisson regression

Lakka et al 1994 [233]

To investigate the independent associations of LTPA and maximal oxygen uptake with the risk of acute MI.

• n = 1,166

Baseline examination: 1984-1989

Conditioning LTPA and VO₂ max had an inverse, graded and independent association with the risk

• Sex: Men

• Age: 42-61 yr

Adjusted RH (95% CI) by conditioning PA level

Finland

• Characteristics: Healthy with normal ECG

PA assessment: Questionnaire for conditioning PA (h/wk), 3 groups (h/wk)

G1 = <0.7

   • G1 = 1.00 (referent)

Prospective cohort

G2 = 0.7

   • G2 = 1.11 (0.58-2.12)

• Kuopio Ischaemic Heart Disease Risk Factor Study

G3 = >2.2

   • G3 = 0.31(0.12-0.85)

D & B score = 13

Adjusted RG (95% CI) by VO₂ max

   • G1 = 1.00

PF assessment: VO₂ max (ml/kg/min)

   • G2 = 0.76 (0.38-1.50)

   • G3 = 0.26 (0.10-0.68)

G1 = <28.0

G2 = 28.0-33.6

G3 = >33.6

Outcome event: acute MI

Cox proportional HR

Laukkanen at al 2004 [234]

To determine whether VO_2peak predicts CVD morbidity and mortality in a sample of men as related to conventional risk factors, medications or underlying chronic disease.

• 1,294 healthy; 1,057 unhealthy

PF Assessment: VO₂ peak (ml/kg/min) measured by exercise test with an electrically braked cycle ergometer, divided into quartiles

   • Number of Cases: 204 CV deaths, 323 non-fatal coronary events

Dose-response relationship between directly measured PF and CVD death among healthy men at baseline.

Finland

• Sex: Men

Healthy men with low VO₂ peak (lowest quartile) had an increased risk

• Age: 42-60 yr

Prospective cohort

• Characteristics: Healthy and not healthy participants

Q1 = <27.6

Adjusted RR (95% CI) by PF quartile Fatal MI

Q2 = 27.6-32.2

Unfit men with unfavorable risk profiles are the risk group that would benefit the most from preventative measures.

D & B score = 11

Q3 = 32.3-37.1

   • 3.29 (0.86-12.90)

• Kuopio Ischaemic Heart Disease Risk Factor Study

Q4 = >37.2

Non-Fatal MI

Outcome Measure: Incidence of fatal and non fatal CVD during 13 year follow-up

   • 2.16 (1.12-4.18)

Cox proportional HR

Lee at al 2000 [235]

To investigate whether different durations of exercise episode are associated with different risk of CHD.

• n = 7,307

Baseline survey in 1988

   • Number of Cases: 482

Longer durations of PA bouts are not associated with decreased CHD risk compared with shorter bouts, once total EE is taken into account.

• Sex: Men

USA

• Age: Mean 66.1 ± 7.5

PA assessment: Survey for EE (kJ/wk), divided into 5 groups and episodes of PA (min), divided into 6 groups

Multivariate adjusted RR (95% CI) by EE

   • G1 = 1.00 (referent)

• Characteristics: Healthy

   • G2 = 0.80 (0.57-1.12)

   • G3 = 0.80 (0.55-1.16)

Prospective cohort

• Harvard Alumni Study

   • G4 = 0.74 (0.47-1.17)

   • G5 = 0.62 (0.41-0.94)

D & B score = 12

Energy expenditure (kJ/wk)

As long as the total EE is similar, more frequent shorter bouts or longer less frequent bouts have an equivalent reduction in CHD risk.

G1 = <4,200

Multivariate adjusted RR (95% CI) by duration of PA episode

G2 = 4,200-8,399

G3 = 8,400-12,599

G4 = 12,600-16,799

   • G1 = 1.00 (referent)

G5 = ≥ 16,800

   • G2 = 1.15 (0.70-1.87)

   • G3 = 1.01 (0.68-1.51)

   • G4 = 1.11 (0.67-1.84)

Duration of PA episode (min)

   • G5 = 1.18 (0.77-1.80)

G1 = None

   • G6 = 1.25 (0.83-1.87)

G2 = 1-15

G3 = 16-30

G4 = 31-45

G5 = 46-60

G6 = >60

Outcome Measure: Fatal and Non Fatal CHD

Proportional hazards regression

Lee et al 2003 [236]

To investigate whether moderate- intensity exercise is associated with reduced CHD.

• n = 7,337

PA assessment: Survey rating usual level of exertion when exercising, divided into tertiles

   • Number of Cases: 551

Inverse association between relative intensity of PA and the risk of CHD.

USA

• Sex: Male

Multivariate adjustment RR (95% CI)

• Age: Mean 66.1 yr

   • T1 = 1.00 (referent)

   • T2 = 0.87 (0.70-1.09)

• Characteristics: Healthy

   • T3 = 0.92 (0.75-1.14)

Prospective cohort

Energy expenditure (kcal/wk)

Harvard Alumni Study

T1 = <1000

D & B score = 13

T2 = 1000-2499

T3 = ≥ 2500

Cox proportional HR

Lemaitre et al 1999 [237]

To investigate whether regular participation in moderate intensity activity confers overall protection from sudden primary cardiac arrest.

• n = 355 cases, 503 controls

PA assessment: Interview (with spouses) for LTPA, 7 groups

   • 355 cases

Participation in moderate intensity LTPA was associated with a decreased risk of primary cardiac arrest.

• Sex: Men and women

RR (95% CI)

USA

G1 = No activity

   • G1 = 1.00 (referent)

• Age: 25-74 yr

G2 = Gardening only≤ 60 min/wk

   • G2 = 0.52 (0.21-1.28)

Case control

• Characteristics: Previously healthy prior to primary cardiac arrest. Control Subjects: Individually matched to case patients on age (within 7 years) and sex at a ratio of about 2:1 were randomly selected from community by random-digit dialing

G3 = Gardening only > 60 min/wk

   • G3 = 0.34 (0.13 0.89)

G4 = Walking ≤ 60 min/wk

   • G4 = 0.45 (0.17-1.19)

D & B score = 11

G5 = Walking > 60 min/wk

   • G5 = 0.27 (0.11-0.67)

G6 = Moderate intensity

   • G6 = 0.31 (0.13-0.74)

LTPA (not walking or gardening)

G7 = 0.34 (0.16-0.75)

G7 = High intensity LTPA

Logistic regression analysis

Lemaitre et al 1995 [238]

To examine whether LTPA decreases the risk of MI in postmenopausal women.

• n = 1,193

PA assessment: Phone interview for LTPA, divided into quartiles of EE (mean kcal/wk)

   • Number of Cases: 268

Risk of MI among postmenopausal women is decreased by 50% with modest LT energy expenditures, equivalent to 30-45 min of walking for exercise three times per week

• Sex: Women

• Age: Mean 67 yr

Multivariate RR (95% CI)

USA

   • Q1 = 1.00 (referent)

• Characteristics: Postmenopaus al Cases: Diagnosed with non-fatal MI Controls: free from MI

Q1 = 71

   • Q2 = 0.52 (0.34-0.80)

Case control

Q2 = 472

   • Q3 = 0.40 (0.26-0.63)

Q3 = 1183

   • Q4 = 0.40 (0.25-0.63)

D & B score = 11

Q4 = 3576

p = <0.001

Outcome Measure: Diagnosed with non-fatal MI

Logistic regression analysis

Li et al 2006 [239]

To examine independent and joint associations of PA and adiposity with CHD incidence.

• n = 88,393

Length of Follow-up: 20 yrs

   • Number of Cases: 2,358

Physical inactivity independently contributes to the development of CHD in women.

• Sex: Women

   • Number of Dropouts: <2% lost to follow-contributes to the development of CHD in women.

USA

• Age: 34-59 yr

up

• Characteristics: Nurses

PA assessment: Questionnaire for LTPA (hr/wk), 3 groups

Prospective cohort

Multivariate HR (95% CI)

• Nurses' Health Study

   • G1 = 1.00 (referent)

G1 = ≥3.5

   • G2 = 1.34 (1.18-1.51)

D & B score = 12

G2 = 1-3.49

   • G3 = 1.43 (1.26-1.63)

G3 = <1

Outcome Measure: CHD incidence

Cox proportional HR

Lemaitre et al 1995 [240]

To evaluate the effect of PA on MI occurrence.

• n = 1,107 (726 controls, 381 cases)

PA assessment: Questionnaire, 3-5 groups depending on variable

OR (95% CI),

PA level was inversely associated with occurrence of MI in both sexes, although the association presented a significant linear trend only for women; in men it suggested a u-shaped relation.

Total PA, men

Portugal

• Sex: Men and women

   • G1 = 1.00 (referent)

Total PA (MET hr/day), men

   • G2 = 0.54 (0.33-0.88)

Case control

• Age: ≥ 40 yr

   • G3 = 0.34 (0.20-0.59)

• Characteristics: Case: Admitted to Hospital and diagnosed with first episode of MI Control: Healthy, no history of CHD

G1 = 28.3-32.1

   • G4 = 0.59 (0.36-0.98)

D & B score = 12

G2 = 32.2-33.3

   • G5 = 0.90 (0.56-1.45)

G3 = 33.4-36.5

Trend p = 0.827

G4 = 36.6-40.3

Total PA, women

G5 = 40.4-83.1

   • Q1 = 1.00 (referent)

Total PA (MET hr/day), women

   • Q2 = 0.39 (0.21-0.73)

Q1 = 28.9-32.7

   • Q3 = 0.33 (0.17-0.64)

Q2 = 32.8-34.1

   • Q4 = 0.22 (0.11-0.47)

Q3 = 34.2-37.8

p = <0.001

Q4 = 37.8-70.6

Sport participation, men

Sport participation (MET hr/day), men

   • G1 = 1.00 (referent)

G1 = 0.0

   • G2 = 0.36 (0.19-0.69),

G2 = 0.1-1.0

   • G3 = 0.72 (0.41-1.26),

G3 = 1.1-2.0

   • G4 = 0.42 (0.23-0.76),

G4 = 2.1-3.6

   • G5 = 0.31 (0.16-0.62)

G5 = 3.7-15.4

p = <0.001

Lovasi et al 2007 [241]

To investigate the shape of the relationship between LTPA and MI risk.

• n = 4,094

PA assessment: Telephone interview (Minnesota LTPA Questionnaire)

   • Number of Cases: 697

Time engaged in LTPA, even non strenuous LTPA was associated with a lower risk of MI, and the shape of this relationship was non- linear

• Sex: Men and women

Adjusted OR (95% CI)

USA

• Age: 64 ± 9 yr

LTPA and non fatal CHD

• Characteristics: Group Health Cooperative Members

   • G1 = 1.00 (referent)

Case control

LTPA

   • G2 = 0.88 (0.66-1.17)

G1 = None

   • G3 = 0.62 (0.46-0.83)

D & B score = 11

G2 = <2

   • G4 = 0.61 (0.45-0.82)

G3 = 2-5

   • G5 = 0.59 (0.44-0.80)

G4 = 5-9

G5 = >9 h/wk

Adjusted RR (95% CI) Strenuous LTPA and non Fatal CHD

Strenuous LTPA

   • G1 = 1.00 (referent)

G1 = None

   • G2 = 0.76 (0.59-0.99)

G2 = non strenuous LTPA

   • G3 = 0.53 (0.40-0.70)

G3 = Any Strenuous

LTPA

Outcome measure: non fatal CHD

Logistic regression

Manson et al 1999 [242]

To assess the comparative roles of walking and vigorous exercise in the prevention of coronary events in women.

• n = 72,488

PA assessment:

   • Number of Cases: 645 coronary events

Both walking and VPA are associated with a substantial reductions in incidence of CHD. Risk reductions for each were similar hen total PAy was similar. Walking 3 or more hours per week could reduce the risk of CHD by 30-40%.

• Sex: Women

Questionnaire with detailed information on PA.

• Age: 40-65 yr

Multivariate RR (95% CI) by total PA score

USA

• Characteristics: Healthy, no Previous history of CHD

   • G1 = 1.00 (referent)

   • G2 = 0.88 (0.71-1.10)

Prospective cohort

Total PA score

   • G3 = 0.81(0.64-1.02)

G1 = 1-2.0

   • G4 = 0.74 (0.58-0.95)

Nurses' Health Study

G2 = 2.1-4.6

   • G5 = 0.66 (0.51-0.86)

D & B score = 12

G3 = 4.7-10.4

p = 0.002

G4 = 10.5-21.7

G5 = >21.7

Multivariate RR (95% CI) by walking activity

   • G1 = 1.00 (referent)

Walking, in those who did not participate in VPA: (MET hr/wk)

   • G2 = 0.78 (0.57-1.06)

G1 = 0.5

   • G3 = 0.88 (0.65-1.21)

G2 = 0.6-2.0

   • G4 = 0.70 (0.51-0.95)

G3 = 2.1-3.8

   • G5 = 0.65 (0.47-0.91)

G4 = 3.9-9.9

p = 0.02

G5 = ≥ 10

Multivariate RR (95% CI) by walking pace

   • 1.00 (referent)

Walking pace (mph)

   • 0.75 (0.59-0.96)

G1 = <2.0

   • 0.64 (0.47-0.88)

G2 = 2.0-2.9

G3 = ≥ 3.0

Mora et al 2007 [243]

To investigate whether differences in several CV risk factors mediate the effect of PA on reduced risk of CVD.

• n = 27,055

10.9 ± 1.6 yr of follow up

   • Number of Cases: 640

There remained a borderline significant inverse association between PA and risk of CHD after adjustment for all sets of risk factors.

• Sex: Women

• Age: ≥ 45 yr

PA assessment: Questionnaires at study entry for categories of EE from PA (kcal/wk), 4 groups

HR (95% CI), basic model

USA

• Characteristics: Healthy

   • G1 = 1.00 (referent)

   • G2 = 0.84 (0.67-1.06)

Prospective cohort

• Women's health study

   • G3 = 0.76 (0.61-0.96)

   • G4 = 0.62 (0.48-0.82)

G1 = <200

p = 0.001

D & B score = 13

G2 = 200-599

While all sets of risk factors should some mediation on the effect of PA on CHD none made the relationship insignificant

G3 = 600-1499

Multivariate adjusted HR (95% CI)

G4 = ≥ 1500

   • G1 = 1.00 (referent)

   • G2= 0.71 (0.58-0.87)

Outcome measure:

   • G3 = 0.64 (0.52-0.78)

Incidence of CVD and

   • G4 = 0.48 (0.38-0.62)

p = <0.001

Cox proportional HR

O'Connor et al 1995 [244]

To examine the association between intensity of exercise and CHD risk.

• n = 680 (532 men and 148 women)

PA assessment: Home interview for PA, divided into quartiles

   • Number of Cases: 340

Significant inverse association between PA level and the risk of non fatal MI in men, which persisted after adjustment for other risk factors.

Adjusted OR (95% CI) by PA level, men

   • Q1 = 1.00 (referent)

USA

• Sex: Men and women

Q1 = Lowest

   • Q2 = 0.60 (0.32-1.13)

Q2

   • Q3 = 0.41 (0.21-0.78)

Case control

• Age: < 76 yr

Q3

   • Q4 = 0.41 (0.22-0.77)

• Characteristics: Cases: Diagnosed MI (non-fatal), no previous history of CHD. Controls: no history of CHD.

Q4 = Highest

p = 0.003

D & B score = 12

Outcome Measure: non-fatal MI

Adjusted OR (95% CI) by PA level, women

   • Q1 = 1.00 (referent)

Moderate- vigorous sports men Cut-points kcal/wk

   • Q2 = 1.07 (0.27-4.17)

Q1 = Lowest

   • Q3 = 2.02 (0.56-7.38)

Q2

   • Q4 = 1.29 (0.31-5.35)

Q3

p = 0.51

Q4 = Highest

Adjusted OR (95% CI) by moderate-vigorous sports, men

   • Q1 = 1.00 (referent)

Moderate- vigorous sports Women

   • Q2 = 1.12 (0.60-2.10)

Cut-points kcal/wk

   • Q3 = 0.61 (0.30-1.24)

Q1 = Lowest

   • Q4 = 0.43 (0.20-0.92)

Q2

p = 0.02

Q3

Q4 = Highest

Adjusted OR (95% CI) by moderate-vigorous sports, women

Logistic regression analysis

   • Q1 = 1.00 (referent)

   • Q2 = 1.31 (0.37-4.66)

   • Q3 = 1.90 (0.44-8.28)

   • Q4 = 0.35 (0.07-1.84)

p = 0.62

Rastogi et al 2004 [245]

To examine the relation between PA and CHD risk in India.

• n = 1,050

PA assessment: Questionnaire

Number of Cases: 350

Observed a strong and dose dependent inverse association between LTPA and non fatal CHD.

• Sex: Men and women

Multivariate OR (95% CI) by LTPA

USA

• Age: 21-74 yr

LTPA (MET min/d)

   • G1 = 1.00 (referent)

• Characteristics: Cases: Diagnosed with MI (non fatal) Controls: non- cardiac patients

G1 = 0

   • G2 = 0.96 (0.59-1.55)

Case control

G2 = 0-145

   • G3 = 0.44 (0.27-0.71)

D & B score = 12

G3 = ≥145

p = 0.001

Sedentary time (min/d)

Multivariate OR (95% CI) by sedentary time

G1 = <70

   • G1 = 1.00 (referent)

G2 = 70-130

   • G2 = 1.15 (0.68-1.95)

G3 = 130-215

   • G3 = 1.04 (0.61-1.76)

G4 = ≥215

   • G4 = 1.88 (1.09-3.21)

p = 0.02

Outcome Measure: Non-fatal MI

Conditional logistic regression

Rodriguez et al 1994 [246]

To examine the relationship between PA and 23 yr incidence of CHD morbidity and mortality.

• n = 7,074

23 year follow-up

   • Number of Cases: 789

PA was associated with a significant reduction in the risk of CHD morbidity and mortality.

• Sex: Men

• Age: 45-64 yr

PA assessment: Questionnaire for PA index, divided into tertiles

Age adjusted RR (95% CI), CHD incidence

USA

• Characteristics: Japanese- American living in Oahu, Hawaii in 1965, < 65 years to reduce effect of retirement on PA levels

   • T1 = 1.00 (referent)

   • T2 = 1.01 (.86-1.19)

Prospective cohort

T1 = Low

   • T3 = 0.83 (0.86-1.19)

These data support the hypothesis that PA is associated with a favorable profile of CVD risk factors.

T2 = Moderate

T3 = High

Multivariate adjusted RR (95% CI), CHD incidence

D & B score = 11

Cox proportional regression model

   • T1 = 1.00 (referent)

   • T2 = 1.07 (0.90-1.26)

This study did not show a dose- response relationship since the medium tertile of PA showed increased rates of CHD compared to the inactive group.

• The Honolulu Heart Program

   • T3 = 0.95 (0.80-1.14)

Age adjusted RR (95% CI), CHD mortality

   • T1 = 1.00 (referent)

   • T2 = 1.12 (0.88-1.44)

   • T3 = 0.74 (0.56-0.97)

Multivariate adjusted RR (95% CI)

   • T1 = 1.00 (referent)

   • T2 = 1.19 (0.93-1.53)

   • T3 = 0.85 (0.65-1.13)

Rothenbacher et al 2003 [247]

To estimate the risk for CHD associated with LTPA.

• n = 791 (312 cases; 479 controls)

PA assessment: Interview

Number of Cases: 312

LTPA showed a clear inverse association with risk of CHD.

LTPA (h/wk)

Multivariate OR (95% CI), LTPA

Germany

• Sex: Men and Women

G1 = 0

Winter

G2 = <1

   • G1 = 1.00 (referent)

Case control

Age: 40-68 yr

G3 = 1-2

   • G2 = 0.48 (0.27-0.84)

Characteristics: Cases: stable CHD diagnosed within 2 years, no recent MI, Controls: no history of CHD.

G4 = >2

   • G3 = 0.54 (0.369-0.82)

D & B score = 12

   • G4 = 0.27 (0.19-0.47)

Workday activity by

bike/foot, (min/workday)

Summer

G1 = <15

   • G1 = 1.00 (referent)

G2 = 15-30

   • G2 = 0.85 (0.47-1.53)

G3 = 30-60

   • G3 = 0.60 (0.38-0.95)

G4 = >60

   • G4 = 0.39 (0.26-0.59)

Outcome Measure: non fatal CHD

Multivariate OR (95% CI), workday activity by bike/foot

Unconditional logistic regression, linear regression model

   • G1 = 1.00 (referent)

   • G2 = 0.53 (0.30-0.93)

   • G3 = 0.36 (0.21-0.62)

   • G4 = 0.58 (0.36-0.94)

Seccareccia and Menotti 1992 [248]

To examine the relationship between OPA and the risk of CHD death.

• n = 1,621

25 year of follow-up

• 189 cases

Increase in OPA is inversely related to risk of CHD death.

• Sex: Men

• Age: 40-59 yr

PA assessment: Questionnaire for OPA (kcal/d), 3 groups

Age Standardized CHD and deaths rates:

• Characteristics: Healthy

   • G1 = 18.9 ± 3.1

Italy

   • G2 = 13.1 ± 1.7

G1 = Sedentary, < 2400

   • G3 = 11.0 ± 0.9

Prospective cohort

G2 = Moderate, 2400-3199

D & B score = 11

G3 = Heavy ≥ 3200

Indicators of PF including HR, vital capacity, FEV in 3/4 of sec, and corrected arm circumference (minus contribution of fat).

End Point: Fatal CHD

Sesso et al 2000 [249]

To examine the association of the quantity and intensity of PA with CHD risk and the impact of other coronary risk factors.

• n = 12,516

PA assessment: Questionnaire

Number of Cases: 2,135

L-Shaped association between PA and the risk of CHD, with a reduction in CHD risk of approximately 20% for total PA levels >4200 kJ/wk

• Sex: Men

• Age: 39-88 yr

Multivariate HR (95% CI)

USA

• Characteristics: Healthy

PA Index (kJ/wk)

   • G1 = 1.00 (referent)

G1 = <2100

   • G2 = 0.90 (0.79-1.03)

Prospective cohort

• Harvard Alumni

G2 = 2100-4199

   • G3 = 0.81 (0.71-0.92)

Study

G3 = 4200-8399

   • G4 = 0.80 (0.69-0.93)

G4 = 8400-12599

   • G5 = 0.81 (0.71-0.94)

D & B score = 12

G5 = >12600

p = 0.003

Suggests that vigorous activities are associated with a reduced risk of CHD, whereas moderate or light PA has no clear association with risk of CHD.

Cox proportional HR

Sundquist et al 2005 [250]

To examine the long term effect of LTPA on incident cases of CHD.

• n = 5,196 (2,645 men, 2,551 women)

PA assessment: Questionnaire Levels of PA

Age and sex adjusted RR (95% CI)

Positive long term effect of LTPA on CHD risk among men and women.

   • Q1 = 1.00 (referent)

Sweden

• Sex: Men and women

Q1 = None

   • Q2 = 0.72 (0.51-1.00)

Q2 = Occasionally

   • Q3 = 0.64 (0.46-0.89)

Prospective cohort

Age: 35-74 yr

Q3 = 1-2 times per week

   • Q4 = 0.46 (0.29-0.74)

• Characteristics: Those not hospitalized for CHD in the last 2 years and those who rate their general health as poor were excluded

Q4 = Vigorous ≥2 times per week

Multivariate adjusted RR (95% CI)

D & B score = 11

Outcome Measure: Fatal or non fatal CHD

   • Q1 = 1.00 (referent)

   • Q2 = 0.76 (0.55-1.07)

   • Q3 = 0.74 (0.53-1.04)

   • Q4 = 0.59 (0.37-0.95)

Cox regression model

Talbot et al 2002 [251]

To examine the contributions of LTPA and aerobic fitness to the risk of coronary events in healthy younger and older adults.

• n = 689

Surveys began in 1960 and were completed on every visit

   • Number of Cases: 63

In younger men PF predicts a reduced risk of CHD but not LTPA.

• Sex: Men

• Age:

After adjusting for coronary risk factors there was:

USA

51.6 ± 16.8 yr

• Characteristics: Community dwelling

PA assessment: Survey for LTPA (97 activities) at every visit.

RR: 0.53 (p < 0.001) and

In older men, high intensity LTPA and PF appear to be of similar importance in reducing CHD risk.

Prospective cohort

RR: 0.61 (p = 0.024) in older men.

D & B score = 12

• Baltimore Longitudinal Study of Aging

PF assessment: Treadmill VO₂ max test on alternate visits

Total LTPA was unrelated to coronary risk in either age group.

With 3 levels of LTPA intensity substituted for total LTPA:

Unpaired t-tests and chi square tests. Cox Proportional hazards Analysis

RR = 0.39 for tertile 3 vs. tertile 1

Tanasescu et al 2002 [252]

To assess the amount, type and intensity of PA in relation to risk of CHD in men.

• n = 44,452

PA assessment: Questionnaire

   • Number of Cases: 1,700

Total PA, running, weight training, and walking were associated with a reduced risk for CVD.

• Sex: Men

• Age: 40-75 yr

Age adjusted HR (95% CI) by total PA

USA

• Characteristics: Health professionals, no history of CHD and in good health

Total PA (MET hr/wk)

   • Q1 = 1.00 (referent)

Q1 = 0-6.32

   • Q2 = 0.85 (0.74 0.98)

Prospective cohort

Q2 = 6.33-14.49

   • Q3 = 0.78 (0.67-0.92)

Q3 = 14.50-25.08

   • Q4 = 0.72 (0.62-0.83)

The average exercise intensity was associated with a reduced risk (independent of total PA).

Q4 = 25.09-41.98

   • Q5 = 0.58 (0.49-0.68)

D & B score = 11

Q5 = > 41.99

p = .001

• Health Professionals follow-up study

Exercise intensity (METs)

Age adjusted HR (95% CI) by exercise intensity

G1 = Low-1-4

   • G1 = .00 (referent)

G2 = Mod.-4-6

   • G2 = 0.94 (0.83-1.04)

G3 = High 6-12

Walking pace independent of total volume of PA (mph)

   • G3 = 0.83 (0.72-0.97)

p = 0.02

Q1 = <2

Age adjusted HR (95% CI) by walking pace

Q2 = 2-3

   • Q1 = 1.00 (referent)

Q3 = 3-4

   • Q2 = 0.72 (0.54-0.94)

Q4 = > 4

   • Q3 = 0.61 (0.45-0.81)

   • Q4 = 0.51 (0.31-0.84)

Outcome Measure: Nonfatal MI or Fatal CHD occurring during follow-up

p <0.001

Cox proportional HR

Vatten et al 2006 [253]

To investigate whether obesity- related CV mortality could be modified by PA.

• n = 54,284 (27,769 men; 26,515 women)

Length of Follow-up: 16 years

   • Number of Cases: 2,462

Increased PA reduces the risk of death in women, but not in men.

Multivariate HR (95% CI), men

Norway

• Sex: Men and women

PA assessment:

   • Q1 = 1.00 (referent)

Questionnaire

   • Q2 = 1.01 (0.89-1.16)

Prospective cohort

Age: ≥ 20 yr

Divided into 4 groups

   • Q3 = 0.98 (0.84-1.14)

• Characteristics: Free from CVD at baseline

Q1 = High

   • Q4 = 1.18 (1.00-1.38)

Q2 = Medium

p = 0.11

D & B score = 12

Q3 = Low

• HUNT study

Q4 = Never

Multivariate HR (95% CI), women

Outcome Measure: Ischemic heart disease mortality

   • Q1 = 1.00 (referent)

   • Q2 = 1.23 (1.01-1.51)

   • Q3 = 1.54 (1.24-1.91)

   • Q4 = 1.52 (1.23-1.88)

Cox proportional HR

p <0.001

Wagner et al 2002 [254]

To investigate if the association between PA patterns and incidence of coronary events could explain the gradient in CHD observed between 2 countries.

• n = 9,758

Length of Follow-up: 5 yrs

Number of Cases: 167 hard CHD, 154 angina events

Beneficial effect of LTPA EE on hard CHD incidence in middle aged men.

• Sex: Men and women

PA assessment: Questionnaire for LTPA, 3 groups:

Number of Dropouts: < 2%

Ireland/France

• Age: 50-59 yr

• Characteristics: Healthy at Baseline

HR (95% CI), hard events

Prospective cohort

G1 = Lowest

   • G1 = 1.00 (referent)

G2 = Middle

   • G2 = 0.73 (0.51-1.05)

G3 = Highest

   • G3 = 0.66 (0.46-0.96)

D & B score = 12

Outcome Measure: CHD hard events and Angina

p = 0.04

HR (95% CI), angina

   • G1 = 1.00 (referent)

Cox proportional HR

   • G2 = 0.83 (0.55-1.25)

   • G3 = 1.28 (0.88-1.86)

p = 0.10