Main Menu Health Library Exercise and Cardiofitness
EXERCISE and DEATH from ALL CAUSES
Abstract: Research shows with high significance that the exercise that produces Cardiofitness can reduce substantially the risk of death from all causes. Actual differences in levels of Cardiofitness - both estimated or actually measured - explain best the differing risk ratios in the Global data set of research on exercise and physical activity. The risks of death from all causes related to Cardiofitness compare well with the summed risks developed independently for cardiovascular diseases and cancer. Directly developed risks of death from all causes suggest that Cardiofitness may be a more important population risk factor than other major risk factors such as cholesterol, smoking, or blood pressure.
Reduced Risk of Early Death is the Ultimate Benefit of any Health Action: A question about research on health risks of an individual disease as heart disease often is this: “Does a reduced risk from this disease really extend to a lowered risk of death?” There always is the possibility that reducing risk of one disease really increases risk of one or more others, and does not necessarily reduce the risk that really counts: The overall change in risk of death.
Life Ahead produces risks of disease and death by individual cause. Thus it does not use an overall risk of death in its computations. But a question becomes: "Do the risks of death from individual diseases add up to and confirm the actually measured total risk of death?" Life Ahead identifies risks mostly for cardiovascular diseases and cancer. Thus if the risks identified for these individual diseases add up to and explain the actually measured risk of death from all causes, this provides a further confirmation that the risks by both cause and disease probably are correct.
The Research: The research found relating exercise, physical activity and Cardiofitness to death from all causes is summarized in Table D following. Risks are shown for 39 different study comparisons from 16 studies. These risks are compared with estimated differences in Activity Calories, Cardiofitness at study baselines, and estimated average Cardiofitness during the study durations. Note that 10 of the 39 measured risk ratios are below 0.40, 6 are below 0.35, and 3 are below 0.30. It is shown separately that the risk of death can be improved more via improvements in Cardiofitness than from improvements in other well recognized risk factors.
Smoking long has been claimed to be largest preventable cause of premature death. But only about 30% of adults still smoke. At least 80% of adults could and should improve their Cardiofitness. A scenario can be developed showing that a real program to improve population Cardiofitness might prolong many more lives than would a complete cessation of smoking. Yet today physiologists and health researchers have failed to develop an accepted useful measure of Cardiofitness such as the CFR that can be used jointly by themselves and the population to understand this most important of major risk factors.
Risks vs. Cardiofitness: The best correlation between risks of all cause death and measures of exercise was obtained on the Global data set using the average Cardiofitness over the term of the studies. The analysis of the 39 different study comparisons from 16 studies produces this result:
1) Risk Ratio = exp (0.0372 * Average Diff in CFR). Correlation coefficient 0.956, t value
of coefficient = 19.7; and 5-95% limits of coefficient are 0.344 to 0.408.
These are exceptional levels of significance. A near identical result was obtained from studies A1 thru A6 that were based on actually measured differences in Cardiofitness:
2) Risk ratio = exp (0.0358 * actual diff in CFR). Correlation coefficient 0.955, t value of
coefficient = 12.0; and 5-95% limits of coefficient of 0.030-0.042
The most probable effect of actual fitness in CFR on all-cause death assuming 16% regression of fitness differences per year for the 7-8 year studies . (See cvr-paper-cvd-risk.htm)
(3) log risk ratio -0.0458 (-0.0374 to –0.0541) * Differnce in-CFR
t of coeff = 11.0 s = 0.310 r = 0.96 reduction in risk for an increase of 10 in CFR: 37%
A relationship of high significance is obtained relating CFR at study start to risk, but with a lesser coefficient of 0.0238. This is due to the higher values of fitness measured at most study baselines than those that probably actually prevailed during the term of the studies.
Risk ratios for death from all causes from (1) have for differences in average CFR are: for 5 CFR, 0.84; for 10 CFR, 0.70, for 15 CFR, 0.58; for 20 CFR, 0.49; for 30 CFR, 0.34. Values from (3) are somewhat larger
Physical Activity Calories and Risk of Death: As was found for research results on cardiovascular diseases and cancer, the relationship between Calories of Activity and risk was much poorer than that for Cardiofitness and risk. As compared with the very high correlation coefficient r of 0.935 with Cardiofitness, the correlation coefficient of death from all causes with Calories was only 0.819, and the standard error of estimate was nearly twice higher than that obtained for Cardiofitness.
A Reconciliation: The effect of Cardiofitness on risk of death from all causes appears consistent with the results for deaths from cardiovascular diseases and cancer. A test using computations of risk for an average man at age 60 produced an estimate of 3.52% per CFR from formula (1) above vs. 3.63% per CFR from the totals of formulas for the individual diseases times their respective amounts. This provides confidence that the values for the individual diseases are probably correct, and that Cardiofitness produces its benefits primarily from reducing risks of Cardiovascular diseases and Cancer.
What do These Risk Ratios Mean? A risk of all-cause death translates quite directly into probable length of life. Because a fit body makes living better as well as life longer, good Cardiofitness may be the best elixir of life now known.
|
|
Table D |
||||||||||||||
|
|
EXERCISE, CARDIOFITNESS and ALL CAUSE DEATH |
||||||||||||||
|
|
No |
Study
|
M Or W |
Scope |
Risk Ratio |
5%-95% Error |
Estimated Difference Activity Calories/ Week |
Difference in CFR |
Notes |
||||||
|
Study Start
|
Average During Study |
||||||||||||||
|
|
|
||||||||||||||
|
|
GROUPS SEPARATED by MEASURED CARDIOFITNESS |
||||||||||||||
|
|
A1 |
Ekelund, LG, N Engl J Med 1988, 319:1379 |
M |
91 CVD deaths on 3100 men |
0.55 0.34 |
0.38-0.83 0.20-0.59 |
2000 n/a |
45 45 |
24 24 |
For Healthy Men Those w. prior CVD 8.5 yrs, with sub-max Balke test |
|||||
|
|
A2 |
Blair, SN, 1989, JAMA 262:2395 |
M
W |
240 deaths on 10,200 Men
43 deaths on 3120 Women |
1.00 0.73 0.68 0.29 1.00 0.52 0.31 0.21 |
Base 0.40-1.32 0.38-1.23 0.17-0.48 Base 0.19-0.92 0.29-0.67 0.10-0.45 |
0 200 500 1400 0 200 500 1060 |
0 14 26 54 0 19 32 70 |
0 8 14 29 0 10 17 37 |
Basis VO2-Max test Most M&W probably used aerobic exercise. Followed average 8 years. |
|||||
|
|
A3 |
Morris, JN, Britsh Heart Journal 1990 63;325-334 |
M |
heart attack, ages 45 to 64 |
1.00 0.38
|
base high signif |
|
|
|
No aerobic exercise vigorous aerobic same for all age groups, 60-64 at start |
|||||
|
|
A4 |
Arraiz, GA J Clin Epidemiol, 1992, 45:419 |
M & W |
113 deaths of 31,600 in Canada |
0.40 |
0.23-0.80 |
1000 |
37 |
22 |
Est from Step test, 7 year follow-up |
|||||
|
|
A5 |
Blair, SN, 1995, JAMA 1995, 273:1093 |
M |
157 deaths on 1512 men with 2 exams |
0.33 |
0.23-0.47 |
1400 |
54 |
40 |
Men that maintained VO2 max fitness in 2 exams, 10 years follow-up |
|||||
|
|
A6 |
Sawada, S, Nippon Koshu Eisei Zasshi 1999, 46:113 |
M |
247 deaths on 10,000 Japanese |
1.00 0.66 0.46 |
Base 0.47-0.94 0.27-0.78 |
0 1200 2500 |
0 25 50 |
0 12 25 |
14 years follow-up Bicycle fitness test, Est VO2 Max |
|||||
|
|
A7 |
Wei, M, Ann Intern Med 2000:132:605 |
M |
180 deaths of 1263 men
|
0.36
|
0.26-0.50
|
1400
|
46
|
25
|
VO2 max test, 12 Years |
|||||
| A8 |
Myers J N Engl J Medicine 2002; 346:793 |
M |
2534 deaths of 6213 men referred for exercise testing, treadmill, 6.2 yrs |
102 0.84 |
Avg Base 0.79-0.89 |
n/a |
0 19 |
0 11 |
Risk= 0.84 for each met in intensity. Those with CVD initially had risk ratio of 0.91 per met. |
||||||
|
|
A9 |
Mora S JAMA 2003, 290:1600 |
F |
427 all cause deaths, 147 coronary deaths from 2994 age 47 |
1.00 0.72 0.58 0.46 0.34 |
p<0.001 for slope of risk vs. VO2 Max |
0 250 500 760 1060 |
0 15 40 57 80 |
0 7
|
90% of VO2 Max on Bruce Protocol Calories est as Blair groups |
|||||
| A10 |
Gulati, M, Circulation 2003,108:1554 |
F |
180 deaths of 5541 in Chicago area, 8 yrs |
1.00 0.53 0.32 |
Base p=0.002 p<0.0001 |
0 375 910 |
0 26 57 |
0 14 30 |
Symptom limited Bruce test, target rates not used. Calories as Blain groups |
||||||
|
|
A11 |
Laukkanen, European Heart Journal 2004,25:1428 |
M |
425 deaths of 2682 age 42-60 for 13 years |
1.0 0.65 0.43 0.22 |
Base
0.12-0.38 |
0 17 33 55 |
0 8 16 27 |
Slightly larger effects of those with disease at start, similar effects for cardiovascular death. |
||||||
|
|
GROUPS SEPARATED by ESTIMATED ACTIVITY and CARDIOFITNESS – ALL CAUSE DEATH |
||||||||||||||
|
|
B1 |
Leon, AS, JAMA 1987, 258:2388 |
M |
781 CHD deaths on 12,000 from MRFIT study |
0.83 |
0.67-1.01 |
3950 start, 1300 avg |
25 |
11 |
7 years follow-up based on estimated activity calories |
|||||
|
|
B2 |
Mensink GB, Epidemiology 1996, 7:391 |
M & W |
7,700 men, 7,700 women in Germany |
0.36 0.28 |
0.16-0.79 0.07-1.17 |
2500 2500 |
25 25 |
25 25 |
For men For women 2+ hrs/week on sports, exercise |
|||||
|
|
B3 |
Haapanen N, Am J Epidemiol 1996, 143:870 |
M |
168 deaths, on 1072 Finland |
0.36 |
0.19-0.68 |
2500 |
25 |
17 |
Both work and leisure, 11 years |
|||||
|
|
B4 |
Kushi, LH JAMA 1997, 277:1287 |
W |
2260 deaths on 40,400 Iowa women |
1.00 0.76 0.70 0.62 |
Base P=<0.001 for trend |
0 600 1000 1600 |
0 6 12 18 |
0 5 10 16 |
3 years follow-up |
|||||
|
|
B5 |
Kujala UM, JAMA 1998, 279:440 |
M & W |
1253 deaths, 7900 M and 7900 W Finland twins |
1.00 0.71 0.57 |
Base 0.62-0.81 0.45-0.74
|
0 800 1600 |
0 10 20 |
0 6 12 |
17 Year study, avg of M and W |
|||||
|
|
B6 |
Hakim, AA, N Eng J Med, 1998, 338:94 |
M |
209 deaths on 707 non-smoking men |
1.00 0.68 0.59 |
Base n/a |
0 450 1200 |
0 10 18
|
0 6 12 |
<1 hr/week 1.5 hr/wk 4 hr/wk est 12 year study |
|||||
|
|
|||||||||||||||