DIETARY FATS and CARDIOVASCULAR DISEASE
Abstract: Direct research on the cardiovascular (CVD) health risk of saturated fat confirms results from its effect on serum cholesterol. But both methods show that the effect of saturated fat on risk of CVD is positive but far smaller than that assumed in the past. A small risk benefit probably derives from added mono-unsaturated fat in accord with its effect on serum cholesterol. But polyunsaturated fats reduce CVD risks much more and trans fats increase risk much more than would be estimated from their effects on cholesterol. Life Ahead includes formulas shown here that reflect the health risk values of each of these fats individually and in various combinations.
Background: Saturated fat long has been deemed the key villain causing heart disease. The large measured variation of serum cholesterol with changes in dietary saturated fat and the strong relationship between cholesterol and heart disease implies this. The now classic research relating saturated fat to heart disease was done by Ancel Keys in his Seven Countries Study (Harvard University Press, 1980). Keys showed that a 5-fold change in coronary disease for men was associated with differences of 5% to 20% in dietary saturated fat in different countries.
Today's research does confirm that dietary fat can be harmful, but to a much lesser extent than was found by Keys in the different countries. There are a number of reasons why this could be true. Those having lowest dietary saturated fat were in developing countries where men were more physical activity and ate diets differing in other nutrients in addition to differences in amounts of fat.
Today’s Most Probable Health Risk from Dietary Saturated Fat: The present Life Ahead program develops the effect of Saturated Fat in diet based on its effect on serum cholesterol. A dozen studies now have investigated the direct effects of dietary fat on heart disease. A problem in direct research on diet is that any change in diet usually is so inter-related with many other factors that identification of a correct effect of any one factor becomes very difficult. For example, factors that changed coincidently with a doubling of saturated fat in the large Study #1 in Table F following included a change in alcohol consumption from 15 to 8 units per day, a change in intake of fiber from 26 to 16 gm per day, a change in unsaturated fat from 17 to 24% of calories, a doubling of trans-fats, and a change in Vitamin E from 127 to 72 units per day. Statistical methods used today cannot adjust accurately for all of these inter-related factor complications
With a caution because of these problems, a likely average result from the more useful recent studies #1-4 in the Table F following was a risk ratio for heart disease of about 1.16 – with a possible variation from at perhaps 1.05 to 1.25 - for a change in saturated fat from about 8 to 15% of calories. A problem here is “What change in dietary cholesterol accompanied these changes in saturated fat?” The Life Ahead formulas relating changes in cholesterol from fat to risk of heart disease forecasts for this same difference a change in heart disease risk of 1.10 at constant dietary cholesterol. But a change in risk of 1.25 is computed when recognizing the usual change of about 13 mg in dietary cholesterol that usually accompanies each change of 1% in dietary saturated fat. Either including or not including this probable association of dietary cholesterol, these available direct population study results do confirm within their margin of error the effect derived from the serum cholesterol based Life Ahead formula. This is a much lower risk ratio than that proposed by Keys.
It is surprising that this effect of saturated fat on heart disease that now is confirmed similarly by each of these two quite different methods is so small. A reduction in dietary fat from that of the highest 20% of our population to that of the lowest 20% changes risk of disease just 15%. And except for the single and now discounted Key’s inter-country estimate, none of the available research studies of saturated fat shows a much larger effect than this. In comparison, use of various vitamins reduces risk of cardiovascular disease by twice larger amounts, and good exercise and Cardiofitness can reduce risk up to several times. Diet saturated fat can be harmful to some extent and of course can increase body weight that produces another added risk over time. And dietary saturated fat increases risk of cancer very significantly. But other diet factors discussed following and elsewhere herein also may reduce risk of cancer, and now appear to be more potentially important to the risk of cardiovascular disease and overall health.
Monounsaturated Fat and Heart Disease: Most research relating amounts of monounsaturated fats to cholesterol shows that this fat reduces cholesterol, but by only a modest amount. A problem is that the transfatty acids that are included in diet partly as monounsaturated fats are strongly hyper-cholesteremic and increase risk of disease. Thus it is necessary to segregate these dietary components for useful analysis. This is accomplished at least approximately via the Life Ahead formula that relates serum cholesterol to dietary factors.
Few research results relating dietary monounsaturated fats to cardiovascular disease were found. The single estimate of Hu in the Table F following shows a risk ratio of 1.18, but another estimate in the same study after adjusting for trans-fats shows a risk ratio of 0.81. Each estimate is subject to a substantial margin of error. The Life Ahead estimate for change in risk using the formula for a change in monounsaturated fat from 11 to 19% of diet calories is: A reduction in serum cholesterol of about 2.5 mg/dl and risk ratio of 0.98. This small computed effect for a difference of from the lowest 20% and highest 20% of a population suggests that the effect of monounsaturated fats on heart disease probably is very small and will not be measurable via direct population study research. In the absence of contrary evidence it is assumed tentatively that the effect of monounsaturated fats on disease can be estimated from its effect on serum cholesterol. But it remains possible that monounsaturated fat could be somewhat more beneficial than this. Another problem is that specific monounsaturated fats as olive oil can have quite differing effects on both cholesterol and risks than average monounsaturated fat.
Polyunsaturated & Omega-3 Fats and Cardiovascular Disease Death: Today's research shows that these fats are enormously important to the risk of cardiovascular disease and death. Thus this subject thus is discussed separately and in depth in a separate article of this title.
Transfatty Acids and Heart Disease: As for polyunsaturated fat, transfatty acids affect risk of disease far differently than would be estimated from their effect on serum cholesterol. The eight different measurements in Table F all show substantial risks for what are very small differences in transfatty acids. An average risk factor of about 1.4 or 40% more heart disease is associated with only 1.5% of these added fats. Even using a transfatty acid factor for increasing serum cholesterol as large as that for saturated fat, the increase in serum cholesterol from 1.5% added trans-fats could account for an added risk of disease of just 2.5% vs. the actually measured increase of 40%. Accordingly, an additional factor for diet trans-fats is included in Life Ahead to reflect a risk ratio in addition to the above amount that is included in the cholesterol factor of:
Risk of heart disease = 1.24 ^ (Trans-fats as % of diet calories – 1.2)
Although measurements of risk usually have been made only for coronary disease, it seems likely that these risks will apply similarly to stroke and other forms of heart disease. The model thus applies these risks to all cardiovascular disease. Key foods that contribute trans-fats are margarine and especially hard margarine, cookies, bread and baked goods, and cooked beef and other meats.
A problem now is that the transfatty acid values of foods found were far from adequate. Highest values were for breads and cereals, and some values on these found from different sources were quite inconsistent. Also researchers have cautioned food providers to reduce content of transfatty acids on key foods. Thus some high values measured in the past may no longer be valid. The Life Ahead diet library thus includes quite conservative values for trans-fats on some foods that produce the above 1.2% of calorie value for an average diet. These values thus may underestimate somewhat both the level of and the potential negative value to health of trans-fats.
The Life Ahead Technology on Dietary Fats can Contribute: This new technology can provide some quite different than previous insights on the health values of different foods. The information from individual health studies such as those in Table F is of basic importance, and a needed foundation for more detailed knowledge. But the individual study results have high margins of error and contribute only rough general ideas about the values of actual real foods. For example, these studies indicate that saturated fats are “not good”; polyunsaturated fats are “good”, etc. This provides general advice as “eat more of this, less of this.” But this does not give us a useful perspective of just how bad, or how good these things really are in terms of future Well-Days of life. The question still is, “Is it worth giving up a pleasure of eating for some possible but only qualitative effect on health?”
The Life Ahead technology can help in two different ways. First, a quantification of benefits of any food in terms of its contribution to future Well-Days of life helps answer this question about pleasure vs. benefit. Second the technology can value overall the benefits of combinations of up to twenty nutrients in each food and combine these benefits with those from accompanying supplements. For example a food high in fat can be beneficial to health if it also contributes enough fiber and unsaturated fat. Or one low in fat can be harmful to health if has low fiber, high diet cholesterol, or is high in transfatty acids. Fats can be important but these still comprise only a portion of the net health benefits of foods. Each food includes other nutrients including vitamins, minerals, antioxidants that contribute.
Most population health research studies are limited to deriving only one or a few very approximate statistical ratios of to the effect of one nutrient on health. By integrating the results of hundreds of studies via a biochemical oriented framework Life Ahead derives a probable quantified effect of up to twenty dietary nutrients. In the absence of a computer quantified analysis it is virtually impossible to estimate an overall probable result from this number of contributing factors. These derived values as shown herein appear consistent with results of direct research on the values of individual food groups. These present estimates via Life Ahead should be refined by more research and analysis and by inclusion of even additional nutrient factors. But the present estimates appear to provide useful guidance on health habits that has not previously been available.
|
TABLE F EFFECT of DIETARY FATS on CORONARY HEART DISEASE |
|
|||||||||||||
|
No |
Study |
Sex |
Scope |
Risk Ratio |
Error Margin |
Basis |
Diff, % in Calories |
Cause |
Notes |
|
||||
|
SATURATED FATS |
||||||||||||||
|
1 |
Ascherio, A, BMJ 1996, 313:84 |
M |
Hlth Professionals Study, N=734 |
1.22 0.96 |
0.96-1.56 0.73-1.27 |
5ths |
7.2-14.8% of calories |
All CHD |
Adj multivariables Also adj Fiber diff |
|||||
|
2
|
Hu, FB N Engl J Med, 1997, 337:1491
|
W |
Nurses St N= 658 |
1.38 1.16 1.17 |
1.13-1.68 0.93-1.14 0.97-1.41 |
5ths |
10.7-18.8 for +5% of calories |
All CHD
|
Age adjustonly Multivariable adjust Multivar plus transfat |
|||||
|
3 |
Tzonou A Epidemiology 1993. 4:492 |
MW |
Italy, N=329 case control |
1.19 |
0.96-1.48 |
5ths |
|
CHD |
Some adjustments |
|||||
|
4 |
Kromhout D, 1995, Prev Med 24:308 |
M |
Seven country study |
1.25 |
P <0.001 |
|
|
Death CHD |
25 yrs followup 4 major sat fats |
|||||
|
|
||||||||||||||
|
MONO UNSATURATED FATS |
||||||||||||||
|
5 |
See (2) above
|
|
|
1.18 0.81 |
0.95-1.46 0.65-1.00 |
5ths
|
11.0-19.1 for +5% of calories |
All CHD |
Multivariable adjusted Multivar plus Trans fat |
|||||
|
|
||||||||||||||
|
POLY UNSATURATED FATS |
||||||||||||||
|
6 |
Same as (1) above
|
|
|
1.04 0.80
|
0.82-1.33 0.63-1.03
|
5ths
|
3.5-7.1% 0.4-0.7%
|
All CHD
|
Linoleic Acid Linolenic Acid (key poly-fats) |
|||||
|
7 |
Same as (2) above |
|
|
0.83 0.62 |
0.67-1.02 0.46-0.85 |
5ths |
2.9-6.4% for + 5% of Calories |
All CHD |
Multivariable adjust Multivar + transfat |
|||||
|
|
||||||||||||||
|
TRANSFATTY ACIDS |
||||||||||||||
|
8 |
Same as (1) above |
|
|
1.40 1.21 |
1.10-1.79 0.93-1,58 |
5ths 5ths |
0.69-2.0 same |
All CHD |
Multivariable adj + adj for Fiber |
|||||
|
9 |
Same as (2) above |
|
|
1.27 |
1.03-1.56 |
5ths |
1.3-2.9 |
All CHD |
Multivariable adj |
|||||
|
10 |
Willet, WC Lancet 1993,341:581 |
W |
Nurses St N = 431 |
1.45 |
1.08-1.94 |
5ths |
1.1-2.6 |
All CHD |
Standard risk factors |
|||||
|
11 |
Tavani, A Eur J Clin Nutr 1997, 51:20 |
W |
Italy, N = 429 |
1.5 |
1.0-2.20 |
|
|
|
Med-High intake of Margerine |
|||||
|
12 |
Same as (3) above |
|
|
1.87 |
.92-4.28 |
5ths |
|
|
Cooking w. margerine |
|||||
|
13 |
Pietinen P Am J, 1997, Epidemiol 145:876 |
M |
Finland smokers, N=1399 |
1.39 |
1.09-1.78 |
5ths |
Avg 3.0 |
All CHD |
Multivariable adj |
|||||
|
14 |
Same as (4) above |
|
|
1.28 |
P < 0.001 |
|
|
CHD Deaths |
elaidic acid |
|||||