THE INTERHEART ANALYSIS
“INTEGRATED CARE REDUCES RISK FACTORS FOR CARDIOVASCUOLAR DISEASE”
JEFFREY L. PROBSTFIELD, MD, FACP, FACC, FAHA, FESC, FSCT
Director, Clinical Trials Service Unit Professor of Medicine (Cardiology)
University of Washington School of Medicine
Research grants- NHLBI, NCI, Sanofi- Aventis, Abbott and Lilly; Consultantships: Genentech, Amylin, Sanofi-Aventis, no stocks,
options or BOD positions
INTERHEART: Background
• Over 80% of the CVD burden occurs in low (L) and middle income countries (MIC), but there are few data on risk factors for CVD from these countries
• Current thinking suggests that “only 50%” of the risk of CHD is accounted by known risk factors
Rosenman RH, et al. Med Clin N Amer 1974; 58: 269-279.
“The classic risk factors enhancing the incidence of CHD are now well known. However, only a minority of individuals with such attributes in any prospective study have been victimized with the passage of time (REF to his own work). A considerable rate of CHD also was observed in individuals who did not exhibit many of these risk factors (REF to his own work). More recently, Keys et al (1972, CIRC) showed that the classic risk factors account for only about half of the cases of CHD.”
Magnus and Beaglehole, 2001
“The only-50% claim goes back at least to 1975, and we could find no reference or published source that plausibly supports it with empirical data. The claim consists of simple assertions with no supporting data or rationale, or assertions made with inappropriate citations. In both cases, the claim has been secondarily quoted, perpetuating the myth.”
WHERE DID THE 50% MYTH COME FROM?
SATURDAY EVENING POST COVER. MARCH 6, 1948
GOSSIP
Magnus P, Beaglehole R. The real contribution of the major risk factors to the coronary epidemic: time to end the “only 50%” myth. Arch Intern Med 2001; 161: 2657-60.
Reviews evidence that a more realistic estimate of the extent to
which traditional risk factors account for CHD is between 75 % and 87 % of CHD deaths. – Cites Stamler J, et al (JAMA 1999; 282:2012-2018). – About 10% of people in several cohorts studied had favorable levels of
4 major CHD risk factors (serum cholesterol <200 mg/dL, blood pressure < 120/80 mm Hg, no current cigarette smoking, and no history of diabetes).
– CHD death rates were 80 - 90% lower when these 4 major risk factors were all favorable.
– Similar data were reported in a French cohort (Thomas, et al. Hypertension 2001) and a Swedish cohort (Rosengren, et al. Eur Heart J, 2001).
Proportion of Framingham ppts with higher than favorable major risk factor levels --- Fatal and non-fatal CHD over
20+ Yrs (Average of 3 measurements for each RF at baseline)
Risk Factor 35-39 Men
35-39 Women
40-59 Men
40-59 Women
Number of people w/ event
59 35 433 296
Chol 200+ 78.0% 68.6% 89.8% 86.5% DBP 80+ or SBP 120+
86.4% 40.0% 81.8% 78.7%
Smoker 72.9% 54.3% 70.4% 44.3% Diabetes 1.7% 5.7% 3.2% 4.4% > one of above 98.3% 85.7% 99.3% 98.7%
Canto editorial JAMA. 2003; 290:947-949.
The 3 articles in this issue of JAMA provide evidence that convincingly challenges the frequent claim that "only 50%" of CHD is attributable to the conventional risk factors of smoking, diabetes, hypertension, and hyperlipidemia and clearly point out that additional research is needed to establish the role of other novel CHD risk markers.
Perhaps more important, these studies
emphasize that to reduce the burden of cardiovascular disease, physicians should have even greater vigilance in identifying
conventional CHD risk factors and must redouble efforts to control them effectively.
INTERHEART: History (1991 to 2004)
1. Single center case-control study (300 + 300) in India
1991-1994
2. Multicenter study in India (1200 + 2400) 1994-2003
3. Idea and protocol development of INTERHEART
1997-1999
4. Vanguard phase 1999-2000 5. Full scale recruitment 2000-2003 6. Data analysis/publications 2004
INTERHEART: Aims
1. To evaluate the association (odds ratio) of risk factors for MI globally, and in each region; and among major ethnic groups in the world.
2. To quantify the impact of each risk factor alone and their combination on the population’s risk (population attributable risk, PAR) overall and in each region, ethnic group, in males and females and in young and old.
Methods Cases: First MI. Controls: Matched to cases by age (+/-5 yr and sex) at each site Data collected from 262 sites in 52 countries: Questionnaire: demographics, lifestyle, health hx, psychosocial, medications Physical measures: height, weight, waist & hip circum, blood pressure, heart rate Blood sample: 20 ml Statistical OR and PAR both presented with 99% confidence intervals. methods: All analyses adjusted for age, sex and region.
Coordinated by the Population Health Research Institute, McMaster University, Canada
Argentina Australia Bahrain Bangladesh Benin Botswana Brazil Cameroon Canada Chile China/Hong Kong Colombia
Croatia Czech Rep Egypt Germany Greece Guatemala Hungary India Iran Israel Italy Japan Kenya
Kuwait Malaysia Mexico Mozambique Nepal New Zealand Netherlands Nigeria Pakistan Philippines Poland Portugal Qatar
Russia Seychelles Singapore S Africa Spain Sri Lanka Sultanate of Oman Sweden Thailand UAE UK USA Zimbabwe
52 Countries Representing Every Inhabited Continent
Recruitment by Region
MI Cases Controls Actual Actual
W Eur 664 761 C/E Eur 1,758 1,927 Middle E/Egypt 1,651 1,789 Africa 579 788 S Asia 1,742 2,204 China/HongKong 3,056 3,056 SE Asia/Japan 968 1,199 ANZ 592 681 S Amer/Mexico 1,246 1,888 N Amer 297 340
W Eur5% CE Eur
14%
MEC13%
Africa5%
S Asia15%
S Am12%
N Am2%
ANZ5% SE Asia
8%
China/HongKong21%
Arab10%
Latin Am11%
Oth Asian6%
Col Afr2%
Other1%
Euro26%
Bl Afr2%S Asian
18%
Chinese24%
Distribution by region Distribution by ethnicity
15,152 MI cases and 14,820 controls
Mean age of male and female cases, overall and by region
40455055606570
Overal
l
W Eur
C/E Eur
MidE/Egyp
t
Africa
SAsia
China/H
ongK
SE Asia/Ja
pan
ANZ
S Amer/
Mex
N Amer
Age
, yea
rsMen Women
Risk of AMI associated with Risk Factors in the Overall Population
Risk factor % Cont % Cases PAR 1 (99% CI) PAR 2 (99% CI) ApoB/ApoA-1(5 v 1) 20.0 33.5 54.1 (49.6, 58.6) 49.2 (43.8, 54.5) Curr smoking 26.8 45.2 36.4(33.9,39.0) 35.7,(32.5,39.1) Diabetes 7.5 18.5 12.3 (11.2, 13.5) 9.9 (8.5, 11.5) Hypertension 21.9 39.0 23.4 (21.7, 25.1) 17.9 (15.7, 20.4) Abd Obesity (3 v 1) 33.3 46.3 33.7 (30.2, 37.4) 20.1 (15.3, 26.0) Psychosocial - - 28.8 (22.6, 35.8) 32.5 (25.1, 40.8) Veg & fruits daily 42.4 35.8 12.9 (10.0, 16.6) 13.7 (9.9, 18.6) Exercise 19.3 14.3 25.5 (20.1, 31.8) 12.2 (5.5, 25.1) Alcohol 24.5 24.0 13.9 (9.3, 20.2) 6.7 (2.0, 20.2) Combined - - 90.4 (88.1, 92.4) 90.4 (88.1, 92.4)
INTERHEART:Risk of AMI with Multiple Risk Factors
Smk DM HTN APoB/A 1+2+3 all4 +O +PS All RFs
2.9 2.4 1.9 3.3 13.0 42.3 68.5 182.9 333.7
1
2
4
8
16
32
64
128
256
512
OR
(99%
CI)
INTERHEART: Decreased Risk of AMI with Avoidance of Smoking; Daily Fruits/Veg, Reg
Phys Activity & Alcohol
0.35 0.70 0.86 0.91 0.24 0.21 0.19
0.125
0.25
0.5
1.0
No smk Frt/Veg Exer Alc Nosmk+Fvg +Exer +Alc
OR
(99%
CI)
Population Attributable Risk by Region and Overall
LIFESTYLE FACTORS Region Smoke % Fr/vg % Exer % Alc % All LS W. Europe 28.9 12.9 38.8 18.9 67.8 E/C Europe 30.2 10.2 11.3 12.9 49.6 Middle East 44.8 8.1 4.0 -4.4 45.5 Africa 38.0 3.8 11.1 27.3 63.2 S. Asia 37.5 18.4 24.3 -5.3 55.2 China 35.8 17.8 21.1 5.3 62.4 S.E. Asia 36.2 11.2 31.4 27.9 69.9 Australia/NZ 44.7 10.7 23.8 18.5 65.8 S. America 38.5 6.7 27.2 -3.1 56.9 N. America 26.3 19.8 25.3 25.3 59.8 Overall 1 36.2 12.9 25.5 13.9 62.8 Overall 2 35.7 13.7 12.2 6.7 54.6
Population Attributable Risk by Region and Overall
NON-LIFESTYLE RISK FACTORS Region HTN % Diab % Abd Obes % All PS% Lipids % All 9 RF W. Europe 22.0 14.9 63.6 38.9 44.6 94.0 E/C Europe 24.5 9.1 28.0 4.9 35.0 72.5 Middle East 9.7 15.5 26.7 41.6 70.5 95.0 Africa 29.9 17.1 58.3 40.0 74.1 97.4 S. Asia 19.4 12.1 37.0 15.9 58.7 89.4 China 22.1 10.0 5.5 35.6 43.8 89.9 S.E. Asia 38.4 21.0 58.0 26.7 67.7 93.7 Austral/NZ 22.8 7.2 61.6 28.9 43.4 89.5 S. America 32.8 12.8 45.4 35.6 47.6 89.4 N. America 18.9 7.9 59.6 51.4 50.5 98.7 Overall 1 23.4 12.4 33.7 28.8 53.8 90.4 Overall 2 17.9 9.9 20.1 32.5 49.2 90.4
Risk Factor Impact by Age Odds Ratio PAR
Young Old Young Old Smoking 3.33 2.44* 40.7 33.1 Fruit/Veg 0.69 0.72 16.9 11.9 Exercise 0.95 0.79 7.5 13.4 Alcohol 1.00 0.85 -4.1 11.1 Hypertension 2.24 1.72* 19.2 17.0 Diabetes 2.96 2.05* 12.4 8.6 Abd Obesity 1.79 1.50 24.8 18.1 All Psych 2.87 2.43 43.5 25.2 ApoB/ApoA-1 4.35 2.50* 58.9 43.6 All 9 RF 216.47 81.99* 93.8 87.9
P for interactions: *p<0.001
INTERHEART: Association of Risk Factors with AMI in Men & Women (1)
Risk Factor Gender Cont. % Curr Smok F 9.3
M 33.0
Diabetes F 7.9 M 7.4
Hypertension F 28.3 M 19.7
Abd Obesity F 33.3 M 33.3
0.25 0.5 1 2 4 8
OR (99% CI)
INTERHEART: Association of Risk Factors with AMI in Men & Women (2)
Risk Factor Gender Cont. %
PS Index F - M -
Fruits/Veg F 50.3 M 39.6
Exercise F 16.5 M 20.3
Alcohol F 11.2 M 29.1
ApoB/ApoA-1 F 14.1 M 21.9
0.25 0.5 1 2 4 8
OR (99% CI)
Risk factors by sex
INTERHEART: ApoB/ApoA-1 ratio (top quintile vs lowest quintile) and MI
0.5 1 2 4 8 16
OR (99% CI)
Region N Cont. % Overall 21408 20.0 W Eur 1047 13.8 CE Eur 2618 20.3 MEC 3291 29.9 Afr 1037 18.0 S Asia 2820 27.7 China/HK 5400 7.3 SE Asia 1858 22.7 ANZ 487 13.8 S Am 2644 27.1 N Am 206 12.4
INTERHEART: Apolipoprotein B/A-1 and MI
Deciles: 1 2 3 4 5 6 7 8 9 10 Cont 1210 1206 1208 1207 1210 1209 1207 1208 1208 1209 Cases 435 496 610 720 790 893 1063 1196 1366 1757 Median 0.43 0.53 0.60 0.66 0.72 0.78 0.85 0.93 1.04 1.28
1
2
4
8
OR
(99%
CI)
0.25
0.50
0.75
1
2
4
Dec1 Dec2 Dec3 Dec4 Dec5 Dec6 Dec7 Dec8 Dec9 Dec10
OR (9
5% C
I)
OR OF MI FOR INCREASING DECILES OF LIPIDS, LIPOPROTEINS AND APOLIPOPROTEINS
Decile Medians overal l (cases+controls)
Decile 1 2 3 4 5 6 7 8 9 10
Apo-A1 Lipoprotein 0.80 0.95 1.03 1.10 1.16 1.23 1.30 1.37 1.49 1.69
Apo-B Lipoprotein 0.54 0.66 0.74 0.80 0.88 0.93 0.99 1.07 1.17 1.36
LDL Cholesterol 1.69 2.18 2.49 2.75 2.99 3.23 3.50 3.80 4.20 4.98
HDL Cholesterol 0.59 0.73 0.82 0.90 0.98 1.06 1.15 1.28 1.45 1.78
Total Cholesterol 3.32 3.95 4.34 4.65 4.94 5.23 5.53 5.89 6.34 7.22
apoA-1
apoB
LDL-C
TC
HDL-C
0.75
1
2
4
8
OR
(95%
CI)
Dec1 Dec2 Dec3 Dec4 Dec5 Dec6 Dec7 Dec8 Dec9 Dec10
Decile Medians overal l (cases+controls)
Decile 1 2 3 4 5 6 7 8 9 10
ApoB/ApoA1 Ratio 0.43 0.53 0.60 0.66 0.72 0.78 0.85 0.93 1.04 1.28
LDL/HDL Ratio 1.37 1.87 2.22 2.52 2.82 3.13 3.49 3.92 4.55 5.93
TC/HDL Ratio 2.74 3.37 3.82 4.23 4.64 5.08 5.58 6.21 7.15 9.20
OR OF MI FOR INCREASING DECILES OF VARIOUS LIPID OR LIPOPROTEIN RATIOS
apoB/apoA1
LDL-C/HDL-C
TC/HDL-C
INTERHEART: Smoking and MI
1
2
4
8
16
Cont 7489 727 1031 446 1058 96 230 168 56 Cases 4223 469 1021 623 1832 254 538 459 218 OR 1 1.38 2.10 2.99 3.83 5.80 5.26 6.34 9.16
Never 1-5 6-10 11-15 16-20 21-25 26-30 31-40 41+
OR
(99%
CI)
Low Levels of Smoking and MI
0.75
1
2
4
8
Never 1-2 3-4 5-6 7-8 9-10 11-12 13-14 15-16 17-18 19-20 21+
OR
(95%
CI)
Adjusted for age sex and region
INTERHEART: Current or Former Smoking & MI
0.5 1 2 4 8
OR (99% CI)
Region N Cont. % Overall 26527 47.9 W Eur 1403 55.0 CE Eur 3624 54.2 MEC 3301 45.4 Afr 1339 53.8 S Asia 3706 41.0 China/HK 6062 42.7 SE Asia 2131 57.1 ANZ 1267 54.2 S Am 3068 48.9 N Am 626 64.6
Tobacco and Risk of AMI
• Tobacco use one of the most important avoidable causes of CVD world wide
• 1.5 billion smokers, 82% in developing countries • 100 million died from tobacco related diseases in 20th
century • 1 billion expected to die in 21st century if current trends
continue • 50% deaths in middle aged adults, lose 22 years of life • Most deaths occurred in men, but increasingly in
women, with increasing tobacco use
Tobacco and Risk of AMI
Fig 3
0.75
1
1.5
2
3
4
8O
R (9
5% C
I)
Never filter nonfilter beedies pipes chew chew+smoke
Risk of AMI Associated with Type of Tobacco Used
---------- Former Smokers ( Years since cessation )----------------------
Figure 5:■ Overall
● 1-9 Cigs/day
10-19 cigs/day
20+ Cigs/day
Current >1-3 >3-5 >5-10 >10-15 >15-20 >200.50
0.75
1
2
4
OR
(95%
CI)
0.75
1
1.5
2
3
OR
(95%
CI)
In all subjects, adjusted for smoking status
In never-smokers
Never 1-7hrs/wk 8-14hrs/wk 15-21hrs/wk 22+hrs/wk
ETS Exposure (hours/week)
Risk of AMI Associated with Second Hand Smoke
INTERHEART: Self-reported Hypertension and MI
0.5 1 2 4 8 OR (99% CI)
N Cont.% 26916 22.3 1425 16.4 3636 32.7 3404 20.2 1355 21.6 3881 13.8 6075 21.1 2141 15.3 1269 22.0 3100 27.7 630 28.6
Region
W Eur Overall
CE Eur MEC Afr
S Asia China/HK SE Asia
ANZ S Am N Am
INTERHEART: Self-reported Diabetes and MI overall
Region N Cont.% Overall 26903 7.6 W Eur 1422 4.2 CE Eur 3636 6.8 MEC 3401 11.6 Afr 1355 8.0
S Asia 3882 10.6 China/HK 6075 2.9 SE Asia 2140 9.2
ANZ 1269 4.8 S Am 3093 9.0 N Am 630 9.7
0.5 1 2 4 8 OR (99% CI)
CDC. www.cdc.gov.
Parallel epidemics of diabetes and obesity
Diabetes
Obesity (BMI ≥30 kg/m2)
<4% 4%–4.9% 5%–5.9% >6%
10%–14% 15%–19% 20%– 24% >25%
2004 1994
DO YOU SEE THE TSUNAMI HEADED
TOWARD US?
OBESITY, ATTENDENT DIABETES AND CHD/CVD
Numbers of People with Diabetes 2000-2030
Wild S et al: Diabetes Care 27:1047-1053; 2004
33.0 66.8
102%
Numbers are millions
33.3 50.0 44%
15.2 35.9
136% 7.0 18.2
160%
46.9 119.5 155%
35.8 71.0 99%
World 2000 = 171 million 2030 = 366 million
Increase 114%
INTERHEART: Analysis of Markers of Obesity and MI
1. Prevalence of obesity in cases & controls overall and in various regions using:
BMI as a marker of overall obesity WHR as a marker of abdominal obesity
2. To assess the association of BMI alone, WHR alone and the two together to MI.
3. To assess regional variations in OR and PAR for obesity
Waist
Waist?
ASSOCIATION OF BMI AND WHR TO MI RISK
Yusuf S, et al.2005
CASE HISTORY
WGT BP T C LDL-C HDL-C TG HbA1c RX Diet
9-3-99 208 160/100 243 170 44 143 6.0
6-7-00 193 126/78 198 129 43 128 6.2 60 x 6 LF+RC
11-20-00 186 118/80 189 121 55 67 5.4 60 x 6 LF+RC
3-14-02 182 124/78 60 x 6 LF+RC
1-05-05 179 124/78 186 121 50 74 60 x 6 LF+RC-C
DIABETES PREVENTION PROGRAM:
NEJM 2002;346:393-403
Lifestyle Modification
Modification Approximate SBP reduction (range)
Weight reduction 5–20 mmHg/10 kg weight loss
Adopt DASH eating plan
8–14 mmHg
Dietary sodium reduction
2–8 mmHg
Physical activity 4–9 mmHg
Moderation of alcohol consumption
2–4 mmHg
INTERHEART: Composite Psychosocial Index and MI
0.5 1 2 4 8
OR (99% CI)
Region
N
OR
Overall
24767
3.11
W.Eur
1375
4.76
C.E.Eur
3473
1.31 M East
2892
3.21
Africa
1259
4.34 S.Asia
3300
2.47
China/HK
5894
7.72 SE Asia
1921
2.76
ANZ
1255
2.81 S.Am.
2783
2.80
N.Am.
615
4.11
Psychosocial stressors
• Stress at work • Stress at home
• Financial stress • Stressful life events
• Depression • Locus of control
OTHER RISK FACTORS?
What about them?
OTHER RISK FACTOR CANDIDATES
• Family History • Elevated prothrombotic factors- e.g. fibrinogen, PAI-I* • Markers of inflammation--HS- e.g. CRP* • Previous Infections—e.g. Chlamydia* • Elevated homocysteine* • Elevated Lp(a)* • Low Socioeconomic Status*
• *Calculations in INTERHEART not yet determined
INTERHEART
Impact of adding “family history”-less than 1%
increment of PAR either unadjusted or adjusted.
INTERHEART: Summary
1. Nine simple risk factors are strongly associated with AMI worldwide.
2. These risk factors are even more important in the young, and their effects are consistent in men and women, across all ethnic groups and all regions.
3. Abnormal Apo-B/ApoA-1 ratio and smoking are the most important risk factors and account for >2/3 of the PAR. All 9 risk factors account for >90% of the PAR globally and in most regions.
IMPLICATIONS: Implementing preventive strategies based
on our current knowledge would avert the majority of premature CHD worldwide.
INTERHEART: Implications
1. Current knowledge provides a basis for a global strategy for prevention of CHD and suggests the potential for preventing the large majority of premature MI. These strategies can utilize similar principles in various regions of the world, but with consideration for the prevalence of risk factors and local economic and cultural factors.
2. Future research is likely to be particularly fruitful if focussed on why known risk factors develop and how they can be substantially modified.
INTERHEART: Bottom Line
KEEP OUR EYES ON THE TARGETS!! --the targets here are 9 identified risk factors,
all of which have identifiable modifications that can reduce risk for CHD/CVD.
“Although substantial gains can be achieved
through control or elimination of established risk factors for cardiovascular disease, it is also important to consider that in data from the United Kingdom Heart Disease Prevention Project and other cohorts, approximately half of all patients suffering a CHD event have no established risk factors.”
Hennekens CH. Circulation 1998; 97: 1095-1102
“Although much has been learned about the
causes of coronary heart disease, the gaps in knowledge are noteworthy; for example, fully half of all patients with this condition do not have any of the established risk factors (hypertension, hypercholesterolemia, cigarette smoking, diabetes mellitus, marked obesity, and physical inactivity.”
Braunwald E. Shattuck Lecture. NEJM 1997; 337: 1360-9.
Many others have made similar claims Lefkowitz RJ, Willerson JT. Prospects for cardiovascular research. JAMA
2001; 285: 581-7 Ridker PM. Evaluating novel cardiovascular risk factors: can we better
predict heart attacks? Ann Intern Med 1999;130: 933-7. Rosenman RH, Friedman M. Neurogenic factors in pathogenesis of
coronary heart disease. Med Clin North Am 1974; 58: 269-79. Gorelick PB. Stroke prevention therapy beyond antithrombotics: unifying
mechanisms in ischemic stroke pathogenesis and implications for therapy: an invited review. Stroke. 2002; 33: 862-75.
Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med 2002; 347: 1557-65.
Mosca L. C-reactive protein--to screen or not to screen? N Engl J Med 2002; 347: 1615-7.
Proportion of CHA ppts with clinically elevated major risk factor levels --- Fatal CHD over 20+ Yrs
Risk Factor 18-39 y.o. Men
18-39 y.o. Women
40-59 y.o. Men
40-59 y.o. Women
Number of people w/ event
240 25 1068 465
Chol 240+ 27.1 % 16.0 % 30.3 % 38.9 % DBP 90+ or SBP 140+
62.9 % 32.0 % 73.4 % 71.4 %
Smoker 72.1 % 80.0 % 52.1 % 51.8 % Diabetes 3.3 % 0 % 7.1 % 5.6 % > one clinically elevated
95.0 % 92.0 % 92.7 % 93.8 %
Proportion of MRFIT ppts with above optimal major risk factor levels --- CHD Death as Outcome over 17 Years (Single Measurements at baseline for each RF)
Risk Factor 35-39 Men 40-57 Men Number of people w/ event 1,442 17,416
Chol 200+ 71.7% 76.5% DBP 80+ or SBP 120+ 83.6% 88.7%
Smoker 66.1% 48.9%
Diabetes 3.0% 5.6% At least one of above 98.1% 98.7%
DEFINITION
POPULATION ATTRIBUTABLE RISK --- i.e., the proportion of all cases attributable to the
relevant risk factor if causality were proven.